Theory Reference

ANSYS, Inc. Release 2024 R1

Southpointe January 2024
2600 Ansys Drive
Canonsburg, PA 15317 ANSYS, Inc. and
ansysinfo@[Link] ANSYS Europe,
Ltd. are UL
[Link] registered ISO
(T) 724-746-3304 9001: 2015
(F) 724-514-9494 companies.
Copyright and Trademark Information

© 2024 ANSYS, Inc. Unauthorized use, distribution or duplication is prohibited.

ANSYS, Ansys Workbench, AUTODYN, CFX, FLUENT and any and all ANSYS, Inc. brand, product, service and feature
names, logos and slogans are registered trademarks or trademarks of ANSYS, Inc. or its subsidiaries located in the
United States or other countries. ICEM CFD is a trademark used by ANSYS, Inc. under license. CFX is a trademark
of Sony Corporation in Japan. All other brand, product, service and feature names or trademarks are the property
of their respective owners. FLEXlm and FLEXnet are trademarks of Flexera Software LLC.

Disclaimer Notice

THIS ANSYS SOFTWARE PRODUCT AND PROGRAM DOCUMENTATION INCLUDE TRADE SECRETS AND ARE CONFID-
ENTIAL AND PROPRIETARY PRODUCTS OF ANSYS, INC., ITS SUBSIDIARIES, OR LICENSORS. The software products
and documentation are furnished by ANSYS, Inc., its subsidiaries, or affiliates under a software license agreement
that contains provisions concerning non-disclosure, copying, length and nature of use, compliance with exporting
laws, warranties, disclaimers, limitations of liability, and remedies, and other provisions. The software products
and documentation may be used, disclosed, transferred, or copied only in accordance with the terms and conditions
of that software license agreement.

ANSYS, Inc. and ANSYS Europe, Ltd. are UL registered ISO 9001: 2015 companies.

U.S. Government Rights

For U.S. Government users, except as specifically granted by the ANSYS, Inc. software license agreement, the use,
duplication, or disclosure by the United States Government is subject to restrictions stated in the ANSYS, Inc.
software license agreement and FAR 12.212 (for non-DOD licenses).

Third-Party Software

See the legal information in the product help files for the complete Legal Notice for ANSYS proprietary software
and third-party software. If you are unable to access the Legal Notice, contact ANSYS, Inc.

Published in the U.S.A.

Table of Contents
1. Introduction ............................................................................................................................................ 1
1.1. Overview of the Theory Reference ..................................................................................................... 1
1.2. Understanding Theory Reference Notation ........................................................................................ 1
1.3. Applicable Products .......................................................................................................................... 3
2. Structures ................................................................................................................................................ 5
2.1. Structural Fundamentals ................................................................................................................... 5
2.1.1. Stress-Strain Relationships ........................................................................................................ 5
2.1.2. Orthotropic Material Transformation for Axisymmetric Models .................................................. 9
2.1.3. Temperature-Dependent Coefficient of Thermal Expansion ..................................................... 10
2.2. Derivation of Structural Matrices ..................................................................................................... 12
2.3. Structural Strain and Stress Evaluations ........................................................................................... 15
2.3.1. Integration Point Strains and Stresses ..................................................................................... 15
2.3.2. Surface Stresses ..................................................................................................................... 16
2.3.3. Shell Element Output ............................................................................................................. 16
2.4. Combined Stresses and Strains ........................................................................................................ 18
2.4.1. Combined Strains ................................................................................................................... 18
2.4.2. Combined Stresses ................................................................................................................. 19
2.4.3. Failure Criteria ........................................................................................................................ 20
[Link]. Maximum Strain Failure Criteria ..................................................................................... 20
[Link]. Maximum Stress Failure Criteria ..................................................................................... 21
[Link]. Tsai-Wu Failure Criteria ................................................................................................... 22
[Link]. Physical Failure Criteria .................................................................................................. 23
[Link].1. Hashin Fiber Failure Criterion ................................................................................ 23
[Link].2. Hashin Matrix Failure Criterion .............................................................................. 23
[Link].3. Puck Fiber Failure Criterion ([398]) ......................................................................... 23
[Link].4. Puck Matrix Failure Criterion ([398]) ....................................................................... 24
[Link].5. LaRc03 Fiber Failure Criterion ([399]) ...................................................................... 24
[Link].6. LaRc03 Matrix Failure Criterion ([399]) .................................................................... 25
[Link].7. LaRc04 Fiber Failure Criterion ([400]) ...................................................................... 25
[Link].8. LaRc04 Matrix Failure Criterion ([400]) .................................................................... 25
2.5. Perfectly Matched Layers (PML) in Elastic Media ............................................................................... 26
3. Structures with Geometric Nonlinearities ............................................................................................ 29
3.1. Understanding Geometric Nonlinearities ......................................................................................... 29
3.2. Large Strain .................................................................................................................................... 30
3.2.1. Theory .................................................................................................................................. 30
3.2.2. Implementation ..................................................................................................................... 32
3.2.3. Definition of Thermal Strains .................................................................................................. 33
3.2.4. Element Formulation .............................................................................................................. 34
3.2.5. Applicable Input .................................................................................................................... 35
3.2.6. Applicable Output .................................................................................................................. 35
3.3. Large Rotation ................................................................................................................................ 35
3.3.1. Theory ................................................................................................................................... 35
3.3.2. Implementation ..................................................................................................................... 36
3.3.3. Element Transformation ......................................................................................................... 36
3.3.4. Deformational Displacements ................................................................................................ 37
3.3.5. Updating Rotations ................................................................................................................ 39
3.3.6. Applicable Input .................................................................................................................... 39
3.3.7. Applicable Output .................................................................................................................. 39
3.3.8. Consistent Tangent Stiffness Matrix and Finite Rotation ........................................................... 39

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 iii
Theory Reference

3.4. Stress Stiffening .............................................................................................................................. 40

3.4.1. Overview and Usage .............................................................................................................. 40
3.4.2. Theory ................................................................................................................................... 41
3.4.3. Implementation ..................................................................................................................... 43
3.4.4. Pressure Load Stiffness ........................................................................................................... 45
3.4.5. Applicable Input .................................................................................................................... 46
3.4.6. Applicable Output .................................................................................................................. 46
3.5. Spin Softening ................................................................................................................................ 46
3.5.1. Spring-Mass System ............................................................................................................... 46
3.5.2. General Equation ................................................................................................................... 47
3.6. General Element Formulations ........................................................................................................ 49
3.6.1. Fundamental Equations ......................................................................................................... 50
3.6.2. Classical Pure Displacement Formulation ................................................................................ 52
3.6.3. Mixed u-P Formulations .......................................................................................................... 53
3.6.4. u-P Formulation I .................................................................................................................... 55
3.6.5. u-P Formulation II ................................................................................................................... 57
3.6.6. u-P Formulation III .................................................................................................................. 58
3.6.7. Volumetric Constraint Equations in u-P Formulations .............................................................. 58
3.6.8. Inverse Formulations .............................................................................................................. 59
[Link]. Displacement-Based Inverse Formulation ....................................................................... 59
3.7. Constraints and Lagrange Multiplier Method ................................................................................... 60
3.8. Steady-State Rolling ........................................................................................................................ 61
3.8.1. General ALE Formulation ........................................................................................................ 61
3.8.2. ALE Formulation for Steady-State Rolling ................................................................................ 62
3.8.3. Kinematics ............................................................................................................................. 63
4. Structures with Material Nonlinearities ................................................................................................ 65
4.1. Understanding Material Nonlinearities ............................................................................................ 65
4.2. Rate-Independent Plasticity ............................................................................................................ 67
4.2.1. Theory ................................................................................................................................... 67
4.2.2. Yield Criterion ........................................................................................................................ 67
4.2.3. Flow Rule ............................................................................................................................... 69
4.2.4. Hardening Rule ...................................................................................................................... 70
4.2.5. Plastic Strain Increment .......................................................................................................... 71
4.2.6. Implementation ..................................................................................................................... 73
4.2.7. Elastoplastic Stress-Strain Matrix ............................................................................................. 74
4.2.8. Bilinear Isotropic Hardening ................................................................................................... 74
4.2.9. Nonlinear Isotropic Hardening ................................................................................................ 75
4.2.10. Bilinear Kinematic Hardening ............................................................................................... 76
[Link]. Applying an Initial Backstress ....................................................................................... 78
4.2.11. Multilinear Kinematic Hardening .......................................................................................... 78
4.2.12. Nonlinear Kinematic Hardening ............................................................................................ 80
[Link]. Applying an Initial Backstress ....................................................................................... 81
4.2.13. Extended Drucker-Prager Model ........................................................................................... 81
4.2.14. Extended Drucker-Prager Cap Model .................................................................................... 83
[Link]. Shear Failure Envelope Function .................................................................................. 83
[Link]. Compaction Cap Function ........................................................................................... 84
[Link]. Expansion Cap Function .............................................................................................. 85
[Link]. Lode Angle Function .................................................................................................... 86
[Link]. Hardening Functions ................................................................................................... 86
4.2.15. Gurson's Model .................................................................................................................... 89
[Link]. Gurson Plasticity with Isotropic/Chaboche Kinematic Hardening .................................. 92

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
iv Published: 2023-10-23
 Theory Reference

4.2.16. Cast Iron Material Model ....................................................................................................... 93

4.3. Rate-Dependent Plasticity (Including Creep and Viscoplasticity) ....................................................... 96
4.3.1. Creep ..................................................................................................................................... 96
4.3.2. Rate-Dependent Plasticity ..................................................................................................... 96
[Link]. Perzyna Option ............................................................................................................. 97
[Link]. Peirce Option ................................................................................................................ 97
[Link]. Exponential Visco-Hardening (EVH) Option .................................................................... 98
[Link]. Anand Viscoplasticity Option ......................................................................................... 98
4.3.3. Extended Drucker-Prager (EDP) Creep Model ........................................................................ 100
[Link]. EDP Inelastic Strain Rate Decomposition ...................................................................... 101
[Link]. EDP Yielding and Hardening Conditions ....................................................................... 101
[Link]. EDP Creep Measurements ............................................................................................ 102
[Link]. EDP Equivalent Creep Stress ......................................................................................... 103
[Link]. EDP Elastic Creep and Stress Projection ........................................................................ 104
4.3.4. Cap Creep Model .................................................................................................................. 105
[Link]. Assumptions and Restrictions ...................................................................................... 105
[Link]. Functions and Potentials .............................................................................................. 106
[Link]. Cap Zones ................................................................................................................... 106
[Link]. Equivalent Creep Stress ................................................................................................ 107
[Link].1. Equivalent Creep Stress in the Compaction Zone ................................................. 108
[Link]. Elastic Creep and Stress Projection ............................................................................... 108
[Link]. Commands Used for Cap Creep ................................................................................... 109
4.4. Gasket Material ............................................................................................................................. 109
4.4.1. Stress and Deformation ........................................................................................................ 109
4.4.2. Material Definition ............................................................................................................... 110
4.4.3. Thermal Deformation ........................................................................................................... 110
4.5. Nonlinear (Multilinear) Elasticity .................................................................................................... 111
4.5.1. Guidelines for Use ................................................................................................................ 111
4.6. Hyperelasticity .............................................................................................................................. 112
4.6.1. Finite-Strain Elasticity ........................................................................................................... 113
4.6.2. Deviatoric-Volumetric Multiplicative Split ............................................................................. 115
4.6.3. Isotropic Hyperelasticity ....................................................................................................... 115
[Link]. Arruda-Boyce Model ..................................................................................................... 115
[Link]. Blatz-Ko Model ............................................................................................................ 116
[Link]. Extended Tube Model ................................................................................................... 116
[Link]. Gent Model ................................................................................................................. 116
[Link]. Hencky Model ............................................................................................................. 117
[Link]. Mooney-Rivlin ............................................................................................................. 117
[Link]. Neo-Hookean .............................................................................................................. 118
[Link]. Ogden Compressible Foam Model ................................................................................. 118
[Link]. Ogden Potential ........................................................................................................... 119
[Link]. Polynomial Form ........................................................................................................ 119
[Link]. Yeoh Model ................................................................................................................ 120
4.6.4. Anisotropic Hyperelasticity ................................................................................................... 120
4.6.5. USER Subroutine .................................................................................................................. 122
4.6.6. Output Quantities ................................................................................................................ 122
4.6.7. Hyperelasticity Material Curve-Fitting ................................................................................... 122
[Link]. Uniaxial Tension (Equivalently, Equibiaxial Compression) .............................................. 125
[Link]. Equibiaxial Tension (Equivalently, Uniaxial Compression) .............................................. 126
[Link]. Pure Shear ................................................................................................................... 126
[Link]. Volumetric Deformation .............................................................................................. 127

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 v
Theory Reference

[Link]. Least Squares Fit Analysis ............................................................................................. 128

4.6.8. Experimental Response Functions ........................................................................................ 128
4.6.9. Material Stability Check ........................................................................................................ 130
4.7. Bergstrom-Boyce .......................................................................................................................... 130
4.8. Mullins Effect ............................................................................................................................... 132
4.8.1. The Pseudo-Elastic Model ..................................................................................................... 132
4.9. Viscoelasticity ............................................................................................................................... 134
4.9.1. Small Strain Viscoelasticity .................................................................................................... 134
4.9.2. Constitutive Equations ......................................................................................................... 134
4.9.3. Numerical Integration .......................................................................................................... 136
4.9.4. Thermorheological Simplicity ............................................................................................... 137
4.9.5. Large-Deformation Viscoelasticity ......................................................................................... 138
4.9.6. Visco-Hypoelasticity ............................................................................................................. 138
4.9.7. Large-Strain Visco-Hyperelasticity ......................................................................................... 139
4.9.8. Large-Strain Visco-Anisotropic Hyperelasticity ....................................................................... 140
4.9.9. Shift Functions ..................................................................................................................... 141
[Link]. Williams-Landel-Ferry Shift Function ............................................................................ 141
[Link]. Tool-Narayanaswamy Shift Function ............................................................................. 141
[Link]. Tool-Narayanaswamy Shift Function with Fictive Temperature ...................................... 141
[Link]. User-Defined Shift Function ......................................................................................... 142
4.10. Swelling ...................................................................................................................................... 143
4.11. Cohesive Zone Material (CZM) Model .......................................................................................... 143
4.11.1. Interface Elements .............................................................................................................. 143
[Link]. Material Model – Exponential Behavior ....................................................................... 144
[Link]. Material Model – Bilinear Behavior ............................................................................. 146
[Link]. Material Model – Exponential Behavior for Preventing Surface Penetration ................. 149
[Link]. Material Model – Linear Behavior for Preventing Surface Penetration .......................... 150
[Link]. Viscous Regularization ............................................................................................... 150
[Link]. Material Model – Rigid Exponential Behavior .............................................................. 150
[Link]. Material Model – Frictional Behavior ........................................................................... 152
4.11.2. Contact Elements ............................................................................................................... 153
[Link]. Material Models for Debonding .................................................................................. 153
[Link].1. Bilinear Behavior - Contact (TBOPT = CBDD and CBDE) ...................................... 153
[Link].2. Bilinear Behavior - Interface (TBOPT = BILI) ........................................................ 158
[Link].3. Exponential Behavior ........................................................................................ 159
4.12. Fluid Material Models .................................................................................................................. 159
4.12.1. Liquid ................................................................................................................................ 160
4.12.2. Gas .................................................................................................................................... 161
4.12.3. Pressure-Volume Data ........................................................................................................ 162
5. Electromagnetics ................................................................................................................................ 163
5.1. Electromagnetic Field Fundamentals ............................................................................................. 163
5.1.1. Magnetic Scalar Potential ..................................................................................................... 166
5.1.2. Solution Strategies ............................................................................................................... 166
[Link]. RSP Strategy ................................................................................................................ 168
[Link]. DSP Strategy ................................................................................................................ 168
[Link]. GSP Strategy ................................................................................................................ 169
5.1.3. Magnetic Vector Potential ..................................................................................................... 170
5.1.4. Edge-Based Magnetic Vector Potential ................................................................................. 171
5.1.5. Harmonic Analysis Using Complex Formalism ....................................................................... 171
5.1.6. Electric Scalar Potential ......................................................................................................... 173
[Link]. Quasistatic Electric Analysis ......................................................................................... 175

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
vi Published: 2023-10-23
 Theory Reference

[Link]. Electrostatic Analysis ................................................................................................... 175

5.2. Derivation of Electromagnetic Matrices ......................................................................................... 176
5.2.1. Magnetic Scalar Potential ..................................................................................................... 176
[Link]. Degrees of freedom ..................................................................................................... 176
[Link]. Coefficient Matrix ........................................................................................................ 176
[Link]. Applied Loads ............................................................................................................. 176
5.2.2. Magnetic Vector Potential ..................................................................................................... 177
[Link]. Degrees of Freedom .................................................................................................... 177
[Link]. Coefficient Matrices ..................................................................................................... 177
[Link]. Applied Loads ............................................................................................................. 178
5.2.3. Edge-Based Magnetic Vector Potential .................................................................................. 179
5.2.4. Electric Scalar Potential ......................................................................................................... 181
[Link]. Quasistatic Electric Analysis ......................................................................................... 182
[Link]. Electrostatic Analysis ................................................................................................... 182
5.3. Electromagnetic Field Evaluations ................................................................................................. 183
5.3.1. Magnetic Scalar Potential Results .......................................................................................... 183
5.3.2. Magnetic Vector Potential Results ......................................................................................... 184
5.3.3. Edge-Based Magnetic Vector Potential .................................................................................. 185
5.3.4. Magnetic Forces ................................................................................................................... 186
[Link]. Lorentz forces .............................................................................................................. 186
[Link]. Maxwell Forces ............................................................................................................ 187
[Link].1. Surface Integral Method ...................................................................................... 187
[Link].2. Volumetric Integral Method ................................................................................ 188
[Link]. Virtual Work Forces ...................................................................................................... 188
[Link].1. Element Shape Method ....................................................................................... 188
[Link].2. Nodal Perturbation Method ................................................................................ 189
5.3.5. Joule Heat in a Magnetic Analysis ......................................................................................... 190
5.3.6. Electric Scalar Potential Results ............................................................................................. 190
[Link]. Quasistatic Electric Analysis ......................................................................................... 191
[Link]. Electrostatic Analysis ................................................................................................... 192
5.3.7. Electrostatic Forces ............................................................................................................... 192
5.4. Stranded Coil Analyses .................................................................................................................. 193
5.4.1. Governing Equations ............................................................................................................ 193
5.4.2. A-VOLT-EMF Formulation ...................................................................................................... 194
5.5. Inductance, Flux and Energy Computation ..................................................................................... 196
5.5.1. Differential Inductance Definition ......................................................................................... 196
5.5.2. Review of Inductance Computation Methods ........................................................................ 197
5.5.3. Inductance Computation Method Used ................................................................................ 198
5.5.4. Transformer and Motion Induced Voltages ............................................................................ 198
5.5.5. Absolute Flux Computation .................................................................................................. 199
5.5.6. Inductance Computations .................................................................................................... 199
5.5.7. Absolute Energy Computation .............................................................................................. 200
5.6. Electromagnetic Particle Tracing .................................................................................................... 201
5.7. Capacitance Computation ............................................................................................................. 202
5.8. Conductance Computation ........................................................................................................... 204
5.9. Hall Effect ..................................................................................................................................... 205
6. Heat Flow ............................................................................................................................................. 207
6.1. Heat Flow Fundamentals ............................................................................................................... 207
6.1.1. Conduction, Convection, and Mass Transport (Advection) ..................................................... 207
6.1.2. Radiation ............................................................................................................................. 210
[Link]. View Factors ................................................................................................................ 211

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 vii
Theory Reference

[Link]. View Factor Matrix Properties ....................................................................................... 213

[Link]. Radiation Usage .......................................................................................................... 215
6.1.3. Mass Transport (Advection) .................................................................................................. 215
[Link]. Diffusive Flux and Total Flux Neumann Boundary Conditions ........................................ 216
6.2. Derivation of Heat Flow Matrices ................................................................................................... 217
6.3. Heat Flow Evaluations ................................................................................................................... 219
6.3.1. Integration Point Output ...................................................................................................... 219
6.3.2. Surface Output ..................................................................................................................... 220
6.4. Radiosity Solution Method ............................................................................................................ 220
6.4.1. Radiosity Equations Simplified for Models with Symmetry ..................................................... 221
6.4.2. View Factor Calculation (3D): Hemicube Method ................................................................... 222
6.4.3. View Factor Calculation (2D) ................................................................................................. 224
[Link]. Non-Hidden Method ................................................................................................... 224
[Link]. Hidden Method ........................................................................................................... 224
[Link]. View Factors of Axisymmetric Bodies ............................................................................ 225
[Link]. Space Node ................................................................................................................. 226
6.5. Radiation Matrix Method ............................................................................................................... 226
6.5.1. View Factor Calculation (2D): Radiation Matrix Method .......................................................... 227
7. Thin Fluid Film Flow ............................................................................................................................ 229
7.1. Squeeze Film ................................................................................................................................ 229
7.1.1. Flow Between Flat Surfaces .................................................................................................. 229
7.1.2. Flow in Channels .................................................................................................................. 232
7.2. Slide Film ...................................................................................................................................... 234
7.3. Hydrodynamic Bearing ................................................................................................................. 236
7.3.1. Finite Length Formulation for 2D Elements ........................................................................... 237
7.3.2. Finite Element Formulation for 3D Element ........................................................................... 239
8. Acoustics ............................................................................................................................................. 241
8.1. Acoustic Fundamentals ................................................................................................................. 241
8.1.1. Governing Equations ............................................................................................................ 241
8.1.2. Finite Element Formulation of the Wave Equation ................................................................. 243
8.1.3. Governing Equations with Mean Flow Effect ......................................................................... 244
8.1.4. Finite Element Formulation of the Convective Wave Equation ................................................ 246
8.1.5. Governing Equations for Axially Symmetric Elements ............................................................ 247
8.2. Derivation of Acoustic Matrices ..................................................................................................... 248
8.3. Propagation, Radiation, and Scattering of Acoustic Pressure Waves ................................................. 249
8.3.1. Acoustic Boundary Conditions .............................................................................................. 249
[Link]. Dirichlet and Neumann Boundary Conditions ............................................................... 249
[Link]. Robin Boundary Condition ........................................................................................... 249
[Link]. Floquet Periodic Boundary Condition ........................................................................... 250
8.3.2. Absorbing Boundary Condition (ABC) ................................................................................... 251
8.3.3. Artificially Matched Layers .................................................................................................... 254
[Link]. Perfectly Matched Layers (PML) .................................................................................... 254
[Link]. Perfectly Matched Layer (PML) for Axially Symmetric Elements ..................................... 257
[Link]. Irregular Perfectly Matched Layers (IPML) ..................................................................... 257
8.3.4. Acoustic Excitation Sources .................................................................................................. 258
[Link]. Specified Pressure or Normal Velocity on the Domain Boundary ................................... 259
[Link]. Mass Source in the Wave Equation ............................................................................... 259
[Link]. Analytic Wave Sources ................................................................................................. 260
[Link]. Analytic Port Modes in a Duct ...................................................................................... 263
[Link]. Power Source in the Diffusion Equation ....................................................................... 264
8.3.5. Sophisticated Acoustic Media ............................................................................................... 265

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
viii Published: 2023-10-23
 Theory Reference

[Link]. Non-uniform Acoustic Media ....................................................................................... 265

[Link]. Equivalent Fluid of Perforated Materials ....................................................................... 266
[Link].1. Johnson-Champoux-Allard Model ....................................................................... 266
[Link].2. Delany-Bazley and Miki Models ........................................................................... 268
[Link].3. Complex Media Properties .................................................................................. 269
[Link]. Impedance Sheet Approximation ................................................................................. 269
[Link]. Viscous-Thermal Media ................................................................................................ 270
[Link].1. Boundary Layer Impedance (BLI) Model ............................................................... 270
[Link].2. Low Reduced Frequency (LRF) Model .................................................................. 271
8.4. Acoustic Fluid-Structural Interaction (FSI) ...................................................................................... 272
8.4.1. Coupled Acoustic Fluid-Structural System with an Unsymmetric Matrix Equation ................... 273
8.4.2. Coupled Acoustic Fluid-Structural System with Symmetric Matrix Equation for Full Harmonic
Analysis ........................................................................................................................................ 275
8.5. Pure Scattered Pressure Formulation ............................................................................................. 275
8.6. Acoustic Output Quantities ........................................................................................................... 277
8.6.1. Sound Pressure .................................................................................................................... 277
8.6.2. Far-field Parameters .............................................................................................................. 279
8.6.3. Sound Power ........................................................................................................................ 280
8.6.4. Acoustic Surface Quantities .................................................................................................. 281
8.6.5. Acoustic Volumetric Quantities ............................................................................................. 282
8.6.6. Band Sound Pressure Level ................................................................................................... 282
8.6.7. Equivalent Radiated Power .................................................................................................. 283
8.7. Transfer Admittance Matrix ........................................................................................................... 283
8.7.1. Transfer Admittance Matrix Connected to Acoustic Domains ................................................. 284
8.7.2. Transfer Admittance Matrix Connected to the Structural and Acoustic Domain ...................... 285
8.8. Random Acoustics ........................................................................................................................ 287
8.8.1. Acoustic Diffuse Sound Field ................................................................................................ 287
8.8.2. Diffuse Sound Field Power Spectral Density .......................................................................... 287
8.8.3. Diffuse Sound Field Physical Sampling .................................................................................. 289
8.9. Room Acoustics ............................................................................................................................ 289
8.9.1. Diffuse Model for Room Acoustics ........................................................................................ 290
8.9.2. Mixed Boundary Conditions ................................................................................................. 291
8.9.3. Coupled Rooms with a Partition Wall ..................................................................................... 291
8.10. The Full Linear Navier-Stokes (FLNS) Model .................................................................................. 291
8.10.1. The FLNS Model for Viscous-Thermal Acoustics ................................................................... 292
8.10.2. The Finite Element FLNS Model ........................................................................................... 293
8.10.3. Boundary Conditions of the FLNS Model ............................................................................. 293
8.10.4. Coupling Conditions on the FSI Interface for the FLNS Model .............................................. 294
8.10.5. Boundary Layers in the FLNS Model .................................................................................... 294
8.11. Poroelastic Acoustics ................................................................................................................... 295
8.11.1. Mixed Displacement and Pressure Formulation for Poroelastic Acoustic Material ................. 295
8.11.2. The Finite Element Model ................................................................................................... 297
8.11.3. Boundary Conditions of Poroelastic Acoustics ..................................................................... 297
8.11.4. Coupling Conditions of Poroelastic Acoustics ...................................................................... 298
8.12. Nonlinear Acoustics .................................................................................................................... 300
8.12.1. Westervelt Equation ........................................................................................................... 300
8.12.2. Kuznetsov Equation ............................................................................................................ 300
8.12.3. The Finite Element Model ................................................................................................. 301
8.12.4. Fluid-Structure Interaction in Nonlinear Acoustics .............................................................. 302
9. Diffusion .............................................................................................................................................. 305
9.1. Diffusion Fundamentals ................................................................................................................ 305

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 ix
Theory Reference

9.2. Normalized Concentration Approach ............................................................................................ 306

9.3. Derivation of Diffusion Matrices .................................................................................................... 307
9.4. Diffusion Analysis Results .............................................................................................................. 307
10. Coupling ............................................................................................................................................ 309
10.1. Coupled Effects ........................................................................................................................... 309
10.1.1. Elements ............................................................................................................................ 309
[Link]. Advantages ............................................................................................................... 310
[Link]. Disadvantages ........................................................................................................... 310
10.1.2. Coupling Methods .............................................................................................................. 311
[Link]. Thermal-Structural Analysis ........................................................................................ 313
[Link]. Electromagnetic Analysis ........................................................................................... 314
[Link]. Stranded Coil Analysis ................................................................................................ 314
[Link]. Structural-Magnetic Analysis ...................................................................................... 315
[Link]. Structural-Electromagnetic Analysis ........................................................................... 316
[Link]. Structural-Stranded Coil Analysis ................................................................................ 316
[Link]. Electro-Thermo-Structural Analysis ............................................................................ 317
[Link]. Electro-Magneto-Thermo-Structural Analysis ............................................................. 317
[Link]. Thermal-Electromagnetic Analysis .............................................................................. 318
[Link]. Piezoelectric Analysis ............................................................................................... 319
[Link]. Electrostatic-Structural Analysis ................................................................................ 320
[Link]. Thermo-Piezoelectric Analysis .................................................................................. 320
[Link]. Piezoresistive Analysis .............................................................................................. 321
[Link]. Thermo-Pressure Analysis ......................................................................................... 321
[Link]. Acoustic-Structural Analysis ..................................................................................... 322
[Link]. Thermo-Electric Analysis .......................................................................................... 322
[Link]. Thermal-Magnetic Analysis ...................................................................................... 323
[Link]. Circuit-Magnetic Analysis ......................................................................................... 323
[Link]. Structural-Diffusion Analysis .................................................................................... 323
[Link]. Thermal-Diffusion Analysis ....................................................................................... 324
[Link]. Structural-Thermal-Diffusion Analysis ....................................................................... 324
10.2. Thermoelasticity ......................................................................................................................... 329
10.3. Thermoplasticity ......................................................................................................................... 332
10.4. Thermoviscoelasticity .................................................................................................................. 333
10.5. Piezoelectrics .............................................................................................................................. 334
10.5.1. Constitutive Equations of Piezoelectricity ............................................................................ 334
10.5.2. Losses in a Piezoelectric Analysis ......................................................................................... 336
10.5.3. Derivation of Piezoelectric Matrices .................................................................................... 339
10.5.4. Piezoelectric Results ........................................................................................................... 340
10.5.5. Perfectly Matched Layers (PML) in Piezoelectric Medium ..................................................... 343
10.6. Electroelasticity ........................................................................................................................... 345
10.7. Piezoresistivity ............................................................................................................................ 347
10.8. Thermoelectrics .......................................................................................................................... 348
10.9. Review of Coupled Electromechanical Methods ........................................................................... 351
10.10. Porous Media Flow .................................................................................................................... 351
10.11. Structural-Diffusion Coupling .................................................................................................... 354
10.12. Thermal-Diffusion Coupling ....................................................................................................... 357
10.13. Electric-Diffusion Coupling ........................................................................................................ 360
10.14. Magnetoelasticity ..................................................................................................................... 361
11. Shape Functions ................................................................................................................................ 365
11.1. Understanding Shape Function Labels ......................................................................................... 365
11.2. 2D Lines ...................................................................................................................................... 366

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
x Published: 2023-10-23
 Theory Reference

11.2.1. 2D Lines Without RDOF ...................................................................................................... 367

11.2.2. 2D Lines With RDOF ............................................................................................................ 367
11.3. 3D Lines ...................................................................................................................................... 367
11.3.1. 3D 2-Node Lines (Not Combining Translations and Rotations) .............................................. 367
11.3.2. 3D 2-Node Lines (Combining Translations and Rotations) .................................................... 368
11.3.3. 3D 3-Node Lines ................................................................................................................. 368
11.3.4. 3D 4-Node Lines ................................................................................................................. 368
11.4. Axisymmetric Shells .................................................................................................................... 369
11.4.1. Axisymmetric Shell Without ESF .......................................................................................... 369
11.5. Axisymmetric Harmonic Shells and General Axisymmetric Surfaces .............................................. 369
11.5.1. Axisymmetric Harmonic Shells ............................................................................................ 369
[Link]. Axisymmetric Harmonic Shells Without ESF ................................................................ 370
[Link]. Axisymmetric Harmonic Shells With ESF ..................................................................... 370
11.5.2. General Axisymmetric Surfaces ........................................................................................... 370
11.6. 3D Shells ..................................................................................................................................... 372
11.6.1. 3D 3-Node Triangular Shells Without RDOF (Constant Strain Triangle Elements [CST]) .......... 373
11.6.2. 3D 6-Node Triangular Shells Without RDOF (Linear Strain Triangle Elements [LST]) ............... 373
11.6.3. 3D 3-Node Triangular Shells With RDOF but Without SD ...................................................... 373
11.6.4. 3D 4-Node Quadrilateral Shells Without RDOF and Without ESF (Q4) ................................... 374
11.6.5. 3D 4-Node Quadrilateral Shells Without RDOF but With ESF (Quadrilateral Modified 6-Node
Elements [QM6]) ........................................................................................................................... 374
11.6.6. 3D 8-Node Quadrilateral Shells without RDOF ..................................................................... 374
11.6.7. 3D 4-Node Quadrilateral Shells With RDOF but Without SD and Without ESF ........................ 375
11.6.8. 3D 4-Node Quadrilateral Shells With RDOF but Without SD and With ESF ............................. 375
11.7. 2D and Axisymmetric Solids ........................................................................................................ 376
11.7.1. 2D and Axisymmetric 3-Node Triangular Solids (CST) ........................................................... 376
11.7.2. 2D and Axisymmetric 6-Node Triangular Solids (LST) ........................................................... 377
11.7.3. 2D and Axisymmetric 4-node Quadrilateral Solid Without ESF (Q4) ...................................... 377
11.7.4. 2D and Axisymmetric 4-node Quadrilateral Solids With ESF (QM6) ....................................... 378
11.7.5. 2D and Axisymmetric 8-Node Quadrilateral Solids (Q8) ....................................................... 378
11.7.6. 2D and Axisymmetric 4-Node Quadrilateral Infinite Solids ................................................... 378
[Link]. Shape Functions ........................................................................................................ 379
[Link]. Mapping Functions .................................................................................................... 379
11.7.7. 2D and Axisymmetric 8-Node Quadrilateral Infinite Solids ................................................... 379
[Link]. Shape Functions ........................................................................................................ 379
[Link]. Mapping Functions .................................................................................................... 380
11.8. Axisymmetric Harmonic Solids .................................................................................................... 380
11.8.1. Axisymmetric Harmonic 3-Node Triangular Solids ............................................................... 381
11.8.2. Axisymmetric Harmonic 6-Node Triangular Solids ............................................................... 381
11.8.3. Axisymmetric Harmonic 4-Node Quadrilateral Solids Without ESF ....................................... 381
11.8.4. Axisymmetric Harmonic 4-Node Quadrilateral Solids With ESF ............................................. 381
11.8.5. Axisymmetric Harmonic 8-Node Quadrilateral Solids ........................................................... 382
11.9. 3D Solids .................................................................................................................................... 382
11.9.1. 4-Node Tetrahedra .............................................................................................................. 382
11.9.2. 10-Node Tetrahedra ............................................................................................................ 383
11.9.3. 5-Node Pyramids ................................................................................................................ 383
11.9.4. 13-Node Pyramids .............................................................................................................. 384
11.9.5. 6-Node Wedges Without ESF ............................................................................................... 385
11.9.6. 6-Node Wedges With ESF .................................................................................................... 385
11.9.7. 15-Node Wedges ................................................................................................................ 386
11.9.8. 8-Node Bricks Without ESF .................................................................................................. 387

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xi
Theory Reference

11.9.9. 8-Node Bricks With ESF ....................................................................................................... 388

11.9.10. 20-Node Bricks ................................................................................................................. 388
11.9.11. 8-Node Infinite Bricks ........................................................................................................ 389
[Link]. Shape Functions ...................................................................................................... 389
[Link]. Mapping Functions .................................................................................................. 390
11.9.12. 20-Node Infinite Bricks ...................................................................................................... 390
[Link]. Shape Functions ...................................................................................................... 391
[Link]. Mapping Functions .................................................................................................. 392
11.9.13. General Axisymmetric Solids ............................................................................................. 392
[Link]. General Axisymmetric Solid With 4 Base Nodes ......................................................... 394
[Link]. General Axisymmetric Solid With 3 Base Nodes ......................................................... 394
[Link]. General Axisymmetric Solid With 8 Base Nodes ......................................................... 394
[Link]. General Axisymmetric Solid With 6 Base Nodes ......................................................... 395
11.10. Electromagnetic Tangential Vector Elements .............................................................................. 395
11.10.1. Tetrahedral Elements ........................................................................................................ 396
11.10.2. Hexahedral Elements ........................................................................................................ 396
12. Element Tools .................................................................................................................................... 399
12.1. Element Shape Testing ................................................................................................................ 399
12.1.1. Overview ........................................................................................................................... 399
12.1.2. 3D Solid Element Faces and Cross-Sections ......................................................................... 399
12.1.3. Aspect Ratio ....................................................................................................................... 402
12.1.4. Aspect Ratio Calculation for Triangles .................................................................................. 402
12.1.5. Aspect Ratio Calculation for Quadrilaterals .......................................................................... 403
12.1.6. Parallel Deviation ............................................................................................................... 404
12.1.7. Parallel Deviation Calculation .............................................................................................. 405
12.1.8. Maximum Corner Angle ...................................................................................................... 406
12.1.9. Maximum Corner Angle Calculation .................................................................................... 406
12.1.10. Jacobian Ratio .................................................................................................................. 407
[Link]. Jacobian Ratio Calculation ....................................................................................... 407
12.1.11. Warping Factor ................................................................................................................. 410
[Link]. Warping Factor Calculation for Quadrilateral Shell Elements ...................................... 410
[Link]. Warping Factor Calculation for 3D Solid Elements ..................................................... 412
12.2. Integration Point Locations ......................................................................................................... 414
12.2.1. Lines (1, 2, or 3 Points) ......................................................................................................... 414
12.2.2. Quadrilaterals (2 x 2 or 3 x 3 Points) ..................................................................................... 414
12.2.3. Bricks and Pyramids (2 x 2 x 2 Points) .................................................................................. 415
12.2.4. Triangles (1, 3, or 6 Points) ................................................................................................... 415
12.2.5. Tetrahedra (1, 4, 5, or 11 Points) ........................................................................................... 416
12.2.6. Triangles and Tetrahedra (2 x 2 or 2 x 2 x 2 Points) ................................................................ 418
12.2.7. Wedges (3 x 2 or 3 x 3 Points) .............................................................................................. 418
12.2.8. Wedges (2 x 2 x 2 Points) ..................................................................................................... 418
12.2.9. Bricks (14 Points) ................................................................................................................ 419
12.2.10. Nonlinear Bending (5 Points) ............................................................................................. 420
12.2.11. General Axisymmetric Elements ........................................................................................ 420
12.3. Temperature-Dependent Material Properties ............................................................................... 420
12.4. Positive Definite Matrices ............................................................................................................ 421
12.4.1. Matrices Representing the Complete Structure ................................................................... 421
12.4.2. Element Matrices ................................................................................................................ 422
12.5. Lumped Matrices ........................................................................................................................ 422
12.5.1. Diagonalization Procedure ................................................................................................. 422
12.5.2. Special Handling of Rotational Degrees of Freedom ............................................................ 423

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xii Published: 2023-10-23
 Theory Reference

12.5.3. Limitations of Lumped Mass Matrices ................................................................................. 423

12.6. Reuse of Matrices ........................................................................................................................ 424
12.6.1. Element Matrices ................................................................................................................ 424
12.6.2. Structure Matrices .............................................................................................................. 424
12.6.3. Override Option ................................................................................................................. 425
12.7. Hydrostatic Loads ....................................................................................................................... 425
12.7.1. Internal and External Pressures ........................................................................................... 425
12.7.2. Buoyancy ........................................................................................................................... 426
12.7.3. Effective Tension ................................................................................................................ 426
12.7.4. Instability Checking ............................................................................................................ 427
12.7.5. Effect of Water Pressure on Some Elements ......................................................................... 427
12.8. Hydrodynamic Loads .................................................................................................................. 428
12.8.1. Regular Waves on Line Elements ......................................................................................... 429
[Link]. Small Amplitude Linear (Airy) Wave (Kw = 0, 1) ............................................................ 429
[Link]. Stoke’s Fifth Order Wave (Kw = 2) ................................................................................ 431
[Link]. Dean’s Stream Function Wave (Kw = 3) ........................................................................ 431
12.8.2. Irregular Waves on Line Elements (Kw = 5 through 7) ........................................................... 431
[Link]. Random Wave ........................................................................................................... 432
[Link]. Shell New Wave ......................................................................................................... 434
[Link]. Constrained New Wave .............................................................................................. 434
12.8.3. Diffracted Wave on Line and Surface Elements (Kw = 8) ........................................................ 436
[Link]. Diffracted Waves on Line Elements ............................................................................. 436
[Link]. Diffracted Waves on Surface Elements ........................................................................ 437
12.8.4. Presence of Both Waves and Current ................................................................................... 438
12.8.5. MacCamy-Fuchs Corrections ............................................................................................... 439
12.8.6. Morison's Equation ............................................................................................................. 440
12.8.7. Dynamic Pressure Head ...................................................................................................... 442
12.9. Nodal and Centroidal Data Evaluation ......................................................................................... 442
13. Element Library ................................................................................................................................. 445
13.1. Reserved for Future Use .............................................................................................................. 445
13.2. Not Documented ........................................................................................................................ 445
13.3. Reserved for Future Use .............................................................................................................. 445
13.4. Reserved for Future Use .............................................................................................................. 445
13.5. SOLID5 - 3D Coupled-Field Solid .................................................................................................. 445
13.5.1. Other Applicable Sections .................................................................................................. 446
13.6. Reserved for Future Use .............................................................................................................. 446
13.7. Reserved for Future Use .............................................................................................................. 447
13.8. Reserved for Future Use .............................................................................................................. 447
13.9. Reserved for Future Use .............................................................................................................. 447
13.10. Reserved for Future Use ............................................................................................................. 447
13.11. LINK11 - Linear Actuator ............................................................................................................ 447
13.11.1. Assumptions and Restrictions ........................................................................................... 447
13.11.2. Element Matrices and Load Vector .................................................................................... 447
13.11.3. Force, Stroke, and Length .................................................................................................. 449
13.12. Reserved for Future Use ............................................................................................................. 449
13.13. PLANE13 - 2D Coupled-Field Solid ............................................................................................. 450
13.13.1. Other Applicable Sections ................................................................................................ 451
13.14. COMBIN14 - Spring-Damper ...................................................................................................... 451
13.14.1. Types of Input ................................................................................................................... 451
13.14.2. Stiffness Pass .................................................................................................................... 452
13.14.3. Output Quantities ............................................................................................................ 453

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xiii
Theory Reference

13.15. Reserved for Future Use ............................................................................................................. 454

13.16. Reserved for Future Use ............................................................................................................. 454
13.17. Reserved for Future Use ............................................................................................................. 454
13.18. Reserved for Future Use ............................................................................................................. 454
13.19. Reserved for Future Use ............................................................................................................. 454
13.20. Reserved for Future Use ............................................................................................................. 454
13.21. MASS21 - Structural Mass .......................................................................................................... 455
13.22. Reserved for Future Use ............................................................................................................. 456
13.23. Reserved for Future Use ............................................................................................................. 456
13.24. Reserved for Future Use ............................................................................................................. 456
13.25. PLANE25 - Axisymmetric-Harmonic 4-Node Structural Solid ....................................................... 456
13.25.1. Other Applicable Sections ................................................................................................ 457
13.25.2. Assumptions and Restrictions ........................................................................................... 457
13.25.3. Use of Temperature .......................................................................................................... 457
13.26. Not Documented ...................................................................................................................... 458
13.27. MATRIX27 - Stiffness, Damping, or Mass Matrix ........................................................................... 458
13.27.1. Assumptions and Restrictions ........................................................................................... 458
13.28. Reserved for Future Use ............................................................................................................. 458
13.29. FLUID29 - 2D Acoustic Fluid ....................................................................................................... 458
13.29.1. Other Applicable Sections ................................................................................................ 459
13.30. FLUID30 - 3D Acoustic Fluid ....................................................................................................... 459
13.30.1. Other Applicable Sections ................................................................................................ 459
13.31. LINK31 - Radiation Link .............................................................................................................. 459
13.31.1. Standard Radiation (KEYOPT(3) = 0) .................................................................................. 460
13.31.2. Empirical Radiation (KEYOPT(3) = 1) .................................................................................. 460
13.31.3. Solution ........................................................................................................................... 461
13.32. Reserved for Future Use ............................................................................................................. 461
13.33. LINK33 - 3D Conduction Bar ...................................................................................................... 461
13.33.1. Other Applicable Sections ................................................................................................ 461
13.33.2. Matrices and Load Vectors for 2-node element .................................................................. 462
13.33.3. Matrices and Load Vectors for 3-node element .................................................................. 462
13.33.4. Output ............................................................................................................................. 462
13.34. LINK34 - Convection Link ........................................................................................................... 463
13.34.1. Conductivity Matrix .......................................................................................................... 463
13.34.2. Output ............................................................................................................................. 464
13.35. PLANE35 - 2D 6-Node Triangular Thermal Solid .......................................................................... 465
13.35.1. Other Applicable Sections ................................................................................................ 465
13.36. SOURC36 - Current Source ......................................................................................................... 465
13.36.1. Description ...................................................................................................................... 466
13.37. COMBIN37 - Control .................................................................................................................. 466
13.37.1. Element Characteristics .................................................................................................... 466
13.37.2. Element Matrices .............................................................................................................. 467
13.37.3. Adjustment of Real Constants ........................................................................................... 468
13.37.4. Evaluation of Control Parameter ........................................................................................ 468
13.38. FLUID38 - Dynamic Fluid Coupling ............................................................................................. 469
13.38.1. Description ...................................................................................................................... 469
13.38.2. Assumptions and Restrictions ........................................................................................... 470
13.38.3. Mass Matrix Formulation ................................................................................................... 470
13.38.4. Damping Matrix Formulation ............................................................................................ 471
13.39. COMBIN39 - Nonlinear Spring .................................................................................................... 472
13.39.1. Input ................................................................................................................................ 472

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xiv Published: 2023-10-23
 Theory Reference

13.39.2. Element Stiffness Matrix and Load Vector .......................................................................... 473

13.39.3. Choices for Element Behavior ............................................................................................ 473
13.40. COMBIN40 - Combination .......................................................................................................... 476
13.40.1. Characteristics of the Element ........................................................................................... 476
13.40.2. Element Matrices for Structural Applications ..................................................................... 476
13.40.3. Determination of F1 and F2 for Structural Applications ...................................................... 477
13.40.4. Thermal Analysis ............................................................................................................... 478
13.41. Reserved for Future Use ............................................................................................................. 478
13.42. Reserved for Future Use ............................................................................................................. 478
13.43. Reserved for Future Use ............................................................................................................. 478
13.44. Reserved for Future Use ............................................................................................................. 478
13.45. Reserved for Future Use ............................................................................................................. 479
13.46. Reserved for Future Use ............................................................................................................. 479
13.47. INFIN47 - 3D Infinite Boundary .................................................................................................. 479
13.47.1. Introduction ..................................................................................................................... 479
13.47.2. Theory .............................................................................................................................. 480
13.47.3. Reduced Scalar Potential .................................................................................................. 483
13.47.4. Difference Scalar Potential ................................................................................................ 483
13.47.5. Generalized Scalar Potential .............................................................................................. 484
13.48. Not Documented ...................................................................................................................... 484
13.49. Not Documented ...................................................................................................................... 484
13.50. MATRIX50 - Superelement (or Substructure) .............................................................................. 485
13.50.1. Other Applicable Sections ................................................................................................ 485
13.51. Not Documented ...................................................................................................................... 485
13.52. Reserved for Future Use ............................................................................................................. 485
13.53. Reserved for Future Use ............................................................................................................. 485
13.54. Reserved for Future Use ............................................................................................................. 485
13.55. PLANE55 - 2D Thermal Solid ...................................................................................................... 486
13.55.1. Other Applicable Sections ................................................................................................ 486
13.55.2. Mass Transport Option ...................................................................................................... 486
13.56. Not Documented ...................................................................................................................... 487
13.57. Reserved for Future Use ............................................................................................................. 487
13.58. Not Documented ...................................................................................................................... 487
13.59. Reserved for Future Use ............................................................................................................. 487
13.60. Reserved for Future Use ............................................................................................................. 488
13.61. SHELL61 - Axisymmetric-Harmonic Structural Shell .................................................................... 488
13.61.1. Other Applicable Sections ................................................................................................ 488
13.61.2. Assumptions and Restrictions ........................................................................................... 488
13.61.3. Stress, Force, and Moment Calculations ............................................................................. 488
13.62. Reserved for Future Use ............................................................................................................. 491
13.63. Reserved for Future Use ............................................................................................................. 491
13.64. Not Documented ...................................................................................................................... 491
13.65. Reserved for Future Use ............................................................................................................. 491
13.66. Reserved for Future Use ............................................................................................................. 491
13.67. Reserved for Future Use ............................................................................................................. 492
13.68. LINK68 - Coupled Thermal-Electric Line ...................................................................................... 492
13.68.1. Other Applicable Sections ................................................................................................ 492
13.69. Reserved for Future Use ............................................................................................................. 492
13.70. SOLID70 - 3D Thermal Solid ....................................................................................................... 492
13.70.1. Other Applicable Sections ................................................................................................ 493
13.70.2. Fluid Flow in a Porous Medium ......................................................................................... 493

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xv
Theory Reference

13.71. MASS71 - Thermal Mass ............................................................................................................. 495

13.71.1. Specific Heat Matrix .......................................................................................................... 495
13.71.2. Heat Generation Load Vector ............................................................................................ 495
13.72. Reserved for Future Use ............................................................................................................. 496
13.73. Reserved for Future Use ............................................................................................................. 496
13.74. Not Documented ...................................................................................................................... 496
13.75. PLANE75 - Axisymmetric-Harmonic 4-Node Thermal Solid ......................................................... 496
13.75.1. Other Applicable Sections ................................................................................................ 496
13.76. Reserved for Future Use ............................................................................................................. 497
13.77. PLANE77 - 2D 8-Node Thermal Solid .......................................................................................... 497
13.77.1. Other Applicable Sections ................................................................................................ 497
13.77.2. Assumptions and Restrictions ........................................................................................... 497
13.78. PLANE78 - Axisymmetric-Harmonic 8-Node Thermal Solid ......................................................... 498
13.78.1. Other Applicable Sections ................................................................................................ 498
13.78.2. Assumptions and Restrictions ........................................................................................... 498
13.79. Reserved for Future Use ............................................................................................................. 498
13.80. Reserved for Future Use ............................................................................................................. 498
13.81. Reserved for Future Use ............................................................................................................. 498
13.82. Reserved for Future Use ............................................................................................................. 499
13.83. PLANE83 - Axisymmetric-Harmonic 8-Node Structural Solid ....................................................... 499
13.83.1. Other Applicable Sections ................................................................................................ 499
13.83.2. Assumptions and Restrictions ........................................................................................... 499
13.84. Not Documented ...................................................................................................................... 500
13.85. Reserved for Future Use ............................................................................................................. 500
13.86. Not Documented ...................................................................................................................... 500
13.87. SOLID87 - 3D 10-Node Tetrahedral Thermal Solid ....................................................................... 500
13.87.1. Other Applicable Sections ................................................................................................ 500
13.88. Reserved for Future Use ............................................................................................................. 501
13.89. Reserved for Future Use ............................................................................................................. 501
13.90. SOLID90 - 3D 20-Node Thermal Solid ......................................................................................... 501
13.90.1. Other Applicable Sections ................................................................................................ 501
13.91. Reserved for Future Use ............................................................................................................. 501
13.92. Reserved for Future Use ............................................................................................................. 501
13.93. Reserved for Future Use ............................................................................................................. 502
13.94. CIRCU94 - Piezoelectric Circuit ................................................................................................... 502
13.94.1. Electric Circuit Elements ................................................................................................... 502
13.94.2. Piezoelectric Circuit Element Matrices and Load Vectors .................................................... 502
13.95. Reserved for Future Use ............................................................................................................. 505
13.96. SOLID96 - 3D Magnetic Scalar Solid ........................................................................................... 505
13.96.1. Other Applicable Sections ................................................................................................ 506
13.97. Reserved for Future Use ............................................................................................................. 506
13.98. SOLID98 - Tetrahedral Coupled-Field Solid ................................................................................. 506
13.98.1. Other Applicable Sections ................................................................................................ 507
13.99. Reserved for Future Use ............................................................................................................. 507
13.100. Reserved for Future Use ........................................................................................................... 507
13.101. Reserved for Future Use ........................................................................................................... 507
13.102. Reserved for Future Use ........................................................................................................... 507
13.103. Reserved for Future Use ........................................................................................................... 507
13.104. Reserved for Future Use ........................................................................................................... 507
13.105. Reserved for Future Use ........................................................................................................... 507
13.106. Reserved for Future Use ........................................................................................................... 507

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xvi Published: 2023-10-23
 Theory Reference

13.107. Reserved for Future Use ........................................................................................................... 507

13.108. Reserved for Future Use ........................................................................................................... 508
13.109. Reserved for Future Use ........................................................................................................... 508
13.110. INFIN110 - 2D Infinite Solid ...................................................................................................... 508
13.110.1. Mapping Functions ......................................................................................................... 508
13.110.2. Matrices ......................................................................................................................... 510
13.111. INFIN111 - 3D Infinite Solid ...................................................................................................... 512
13.111.1. Other Applicable Sections ............................................................................................... 513
13.112. Reserved for Future Use ........................................................................................................... 513
13.113. Reserved for Future Use ........................................................................................................... 513
13.114. Reserved for Future Use ........................................................................................................... 513
13.115. Reserved for Future Use ........................................................................................................... 513
13.116. FLUID116 - Coupled Thermal-Fluid Pipe ................................................................................... 513
13.116.1. Assumptions and Restrictions ......................................................................................... 514
13.116.2. Combined Equations ...................................................................................................... 514
13.116.3. Thermal Matrix Definitions .............................................................................................. 514
13.116.4. Fluid Equations ............................................................................................................... 517
13.117. Reserved for Future Use ........................................................................................................... 519
13.118. Reserved for Future Use ........................................................................................................... 519
13.119. Reserved for Future Use ........................................................................................................... 519
13.120. Reserved for Future Use ........................................................................................................... 519
13.121. PLANE121 - 2D 8-Node Electrostatic Solid ................................................................................ 520
13.121.1. Other Applicable Sections ............................................................................................... 520
13.121.2. Assumptions and Restrictions ......................................................................................... 520
13.122. SOLID122 - 3D 20-Node Electrostatic Solid ............................................................................... 520
13.122.1. Other Applicable Sections ............................................................................................... 521
13.123. SOLID123 - 3D 10-Node Tetrahedral Electrostatic Solid ............................................................. 521
13.123.1. Other Applicable Sections ............................................................................................... 521
13.124. CIRCU124 - Electric Circuit ....................................................................................................... 521
13.124.1. Electric Circuit Elements .................................................................................................. 522
13.124.2. Electric Circuit Element Matrices ..................................................................................... 522
13.125. CIRCU125 - Diode .................................................................................................................... 524
13.125.1. Diode Elements .............................................................................................................. 524
13.125.2. Norton Equivalents ......................................................................................................... 525
13.125.3. Element Matrix and Load Vector ...................................................................................... 525
13.126. TRANS126 - Electromechanical Transducer ............................................................................... 526
13.127. Reserved for Future Use ........................................................................................................... 529
13.128. Reserved for Future Use ........................................................................................................... 529
13.129. FLUID129 - 2D Infinite Acoustic ................................................................................................ 529
13.129.1. Other Applicable Sections ............................................................................................... 529
13.130. FLUID130 - 3D Infinite Acoustic ................................................................................................ 530
13.130.1. Mathematical Formulation and F.E. Discretization ............................................................ 530
13.130.2. Finite Element Discretization ........................................................................................... 532
13.131. SHELL131 - 4-Node Layered Thermal Shell ................................................................................ 534
13.131.1. Other Applicable Sections ............................................................................................... 534
13.132. SHELL132 - 8-Node Layered Thermal Shell ................................................................................ 534
13.132.1. Other Applicable Sections ............................................................................................... 535
13.133. Reserved for Future Use ........................................................................................................... 535
13.134. Reserved for Future Use ........................................................................................................... 535
13.135. Reserved for Future Use ........................................................................................................... 535
13.136. FLUID136 - 3D Squeeze Film Fluid Element .............................................................................. 535

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xvii
Theory Reference

13.136.1. Other Applicable Sections ............................................................................................... 536

13.136.2. Assumptions and Restrictions ......................................................................................... 536
13.137. Reserved for Future Use ........................................................................................................... 536
13.138. FLUID138 - 3D Viscous Fluid Link Element ................................................................................ 536
13.138.1. Other Applicable Sections ............................................................................................... 536
13.139. FLUID139 - 3D Slide Film Fluid Element .................................................................................... 537
13.139.1. Other Applicable Sections ............................................................................................... 537
13.140. Reserved for Future Use ........................................................................................................... 537
13.141. Reserved for Future Use ........................................................................................................... 537
13.142. Reserved for Future Use ........................................................................................................... 537
13.143. Not Documented .................................................................................................................... 537
13.144. Reserved for Future Use ........................................................................................................... 538
13.145. Reserved for Future Use ........................................................................................................... 538
13.146. Reserved for Future Use ........................................................................................................... 538
13.147. Reserved for Future Use ........................................................................................................... 538
13.148. Reserved for Future Use ........................................................................................................... 538
13.149. Reserved for Future Use ........................................................................................................... 538
13.150. Reserved for Future Use ........................................................................................................... 538
13.151. SURF151 - 2D Thermal Surface Effect ....................................................................................... 538
13.152. SURF152 - 3D Thermal Surface Effect ....................................................................................... 539
13.152.1. Matrices and Load Vectors .............................................................................................. 539
13.152.2. Adiabatic Wall Temperature as Bulk Temperature ............................................................. 541
13.152.3. Film Coefficient Adjustment ............................................................................................ 542
13.152.4. Radiation Form Factor Calculation ................................................................................... 542
13.153. SURF153 - 2D Structural Surface Effect ..................................................................................... 543
13.154. SURF154 - 3D Structural Surface Effect ..................................................................................... 545
13.155. SURF155 - 3D Thermal Surface Line Load ................................................................................. 548
13.156. SURF156 - 3D Structural Surface Line Load Effect ..................................................................... 549
13.157. SHELL157 - Thermal-Electric Shell ............................................................................................ 549
13.157.1. Other Applicable Sections ............................................................................................... 550
13.158. Not Documented .................................................................................................................... 550
13.159. SURF159 - General Axisymmetric Surface with 2 or 3 Nodes ..................................................... 550
13.159.1. Other Applicable Sections ............................................................................................... 551
13.159.2. Assumptions and Restrictions ......................................................................................... 551
13.160. Reserved for Future Use ........................................................................................................... 551
13.161. Reserved for Future Use ........................................................................................................... 551
13.162. Reserved for Future Use ........................................................................................................... 551
13.163. Reserved for Future Use ........................................................................................................... 551
13.164. Reserved for Future Use ........................................................................................................... 551
13.165. Reserved for Future Use ........................................................................................................... 552
13.166. Reserved for Future Use ........................................................................................................... 552
13.167. Reserved for Future Use ........................................................................................................... 552
13.168. Reserved for Future Use ........................................................................................................... 552
13.169.TARGE169 - 2D Target Segment ................................................................................................ 552
13.169.1. Other Applicable Sections ............................................................................................... 552
13.169.2. Segment Types ............................................................................................................... 552
13.170.TARGE170 - 3D Target Segment ................................................................................................ 553
13.170.1. Introduction ................................................................................................................... 553
13.170.2. Segment Types ............................................................................................................... 554
13.170.3. Reaction Forces .............................................................................................................. 554
13.171. Reserved for Future Use ........................................................................................................... 554

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xviii Published: 2023-10-23
 Theory Reference

13.172. CONTA172 - 2D 3-Node Surface-to-Surface Contact ................................................................. 555

13.172.1. Other Applicable Sections ............................................................................................... 555
13.173. Reserved for Future Use ........................................................................................................... 555
13.174. CONTA174 - 3D 8-Node Surface-to-Surface Contact ................................................................. 555
13.174.1. Introduction ................................................................................................................... 556
13.174.2. Contact Kinematics ......................................................................................................... 556
13.174.3. Frictional Model .............................................................................................................. 559
13.174.4. Contact Algorithm .......................................................................................................... 563
13.174.5. Viscous Damping ............................................................................................................ 566
13.174.6. Energy and Momentum Conserving Contact ................................................................... 567
13.174.7. Debonding ..................................................................................................................... 569
13.174.8. Contact Surface Wear ...................................................................................................... 571
13.174.9. Thermal/Structural Contact ............................................................................................. 573
13.174.10. Electric Contact ............................................................................................................ 574
13.174.11. Magnetic Contact ......................................................................................................... 575
13.174.12. Pore Fluid Contact ........................................................................................................ 575
13.174.13. Diffusive Contact .......................................................................................................... 576
13.175. CONTA175 - 2D/3D Node-to-Surface Contact ........................................................................... 577
13.175.1. Other Applicable Sections ............................................................................................... 577
13.175.2. Contact Models .............................................................................................................. 577
13.175.3. Contact Forces ................................................................................................................ 578
13.176. Reserved for Future Use ........................................................................................................... 578
13.177. CONTA177 - 3D Line-to-Surface Contact ................................................................................... 579
13.177.1. Other Applicable Sections ............................................................................................... 579
13.177.2. Contact Kinematics ......................................................................................................... 579
13.177.3. Contact Model ................................................................................................................ 581
13.177.4. Contact Forces ................................................................................................................ 581
13.178. CONTA178 - 3D Node-to-Node Contact .................................................................................... 582
13.178.1. Introduction ................................................................................................................... 582
13.178.2. Contact Algorithms ........................................................................................................ 582
13.178.3. Element Damper ............................................................................................................ 584
13.178.4. Rigid Coulomb Friction ................................................................................................... 584
13.179. PRETS179 - Pretension ............................................................................................................. 585
13.179.1. Introduction ................................................................................................................... 585
13.179.2. Assumptions and Restrictions ......................................................................................... 585
13.180. LINK180 - 3D Spar (or Truss) ..................................................................................................... 585
13.180.1. Assumptions and Restrictions ......................................................................................... 586
13.181. SHELL181 - 4-Node Shell .......................................................................................................... 586
13.181.1. Other Applicable Sections ............................................................................................... 587
13.181.2. Assumptions and Restrictions ......................................................................................... 587
13.181.3. Assumed Displacement Shape Functions ........................................................................ 588
13.181.4. Membrane Option .......................................................................................................... 588
13.181.5. Warping ......................................................................................................................... 588
13.181.6. Shear Correction ............................................................................................................. 588
13.182. PLANE182 - 2D 4-Node Structural Solid .................................................................................... 588
13.182.1. Other Applicable Sections ............................................................................................... 589
13.182.2. Theory ............................................................................................................................ 589
13.183. PLANE183 - 2D 8-Node Structural Solid .................................................................................... 589
13.183.1. Other Applicable Sections ............................................................................................... 590
13.183.2. Assumptions and Restrictions ......................................................................................... 590
13.184. MPC184 - Multipoint Constraint ............................................................................................... 590

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xix
Theory Reference

13.184.1. Slider Element ................................................................................................................ 591

13.184.2. Joint Elements ................................................................................................................ 592
13.185. SOLID185 - 3D 8-Node Structural Solid ..................................................................................... 592
13.185.1. SOLID185 - 3D 8-Node Structural Solid ............................................................................ 592
13.185.2. SOLID185 - 3D 8-Node Layered Solid ............................................................................... 593
13.185.3. Other Applicable Sections ............................................................................................... 594
13.185.4. Theory ............................................................................................................................ 594
13.185.5. Shear Correction ............................................................................................................. 594
13.186. SOLID186 - 3D 20-Node homogeneous/Layered Structural Solid .............................................. 594
13.186.1. SOLID186 - 3D 20-Node Homogeneous Structural Solid ................................................... 594
13.186.2. SOLID186 - 3D 20-Node Layered Structural Solid ............................................................. 595
13.186.3. Other Applicable Sections ............................................................................................... 596
13.186.4. Shear Correction ............................................................................................................. 596
13.187. SOLID187 - 3D 10-Node Tetrahedral Structural Solid ................................................................. 597
13.187.1. Other Applicable Sections ............................................................................................... 597
13.188. BEAM188 - 3D 2-Node Beam .................................................................................................... 597
13.188.1. Assumptions and Restrictions ......................................................................................... 599
13.188.2. Ocean Effects .................................................................................................................. 600
13.188.3. Stress Evaluation ............................................................................................................. 601
13.189. BEAM189 - 3D 3-Node Beam .................................................................................................... 601
13.190. SOLSH190 - 3D 8-Node Layered Solid Shell .............................................................................. 602
13.190.1. Other Applicable Sections ............................................................................................... 603
13.190.2. Theory ............................................................................................................................ 603
13.190.3. Shear Correction ............................................................................................................. 603
13.191. Reserved for Future Use ........................................................................................................... 603
13.192. INTER192 - 2D 4-Node Gasket .................................................................................................. 603
13.192.1. Other Applicable Sections ............................................................................................... 604
13.193. INTER193 - 2D 6-Node Gasket .................................................................................................. 604
13.193.1. Other Applicable Sections ............................................................................................... 604
13.194. INTER194 - 3D 16-Node Gasket ................................................................................................ 604
13.194.1. Element Technology ....................................................................................................... 605
13.195. INTER195 - 3D 8-Node Gasket .................................................................................................. 606
13.195.1. Other Applicable Sections ............................................................................................... 606
13.196. Reserved for Future Use ........................................................................................................... 606
13.197. Reserved for Future Use ........................................................................................................... 606
13.198. Reserved for Future Use ........................................................................................................... 606
13.199. Reserved for Future Use ........................................................................................................... 606
13.200. Reserved for Future Use ........................................................................................................... 607
13.201. Reserved for Future Use ........................................................................................................... 607
13.202. INTER202 - 2D 4-Node Cohesive ............................................................................................... 607
13.202.1. Other Applicable Sections ............................................................................................... 607
13.203. INTER203 - 2D 6-Node Cohesive ............................................................................................... 607
13.203.1. Other Applicable Sections ............................................................................................... 608
13.204. INTER204 - 3D 16-Node Cohesive ............................................................................................. 608
13.204.1. Element Technology ....................................................................................................... 608
13.205. INTER205 - 3D 8-Node Cohesive ............................................................................................... 609
13.205.1. Other Applicable Sections ............................................................................................... 609
13.206. Reserved for Future Use ........................................................................................................... 609
13.207. Reserved for Future Use ........................................................................................................... 609
13.208. SHELL208 - 2-Node Axisymmetric Shell .................................................................................... 610
13.208.1. Other Applicable Sections ............................................................................................... 610

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xx Published: 2023-10-23
 Theory Reference

13.208.2. Assumptions and Restrictions ......................................................................................... 610

13.208.3. Shear Correction ............................................................................................................. 611
13.209. SHELL209 - 3-Node Axisymmetric Shell .................................................................................... 611
13.209.1. Other Applicable Sections ............................................................................................... 611
13.209.2. Assumptions and Restrictions ......................................................................................... 611
13.209.3. Shear Correction ............................................................................................................. 612
13.210. Reserved for Future Use ........................................................................................................... 612
13.211. Reserved for Future Use ........................................................................................................... 612
13.212. CPT212 - 2D 4-Node Coupled Pore-Pressure-Thermal Mechanical Solid .................................... 612
13.212.1. Other Applicable Sections ............................................................................................... 613
13.213. CPT213 - 2D 8-Node Coupled Pore-Pressure-Thermal Mechanical Solid .................................... 613
13.213.1. Other Applicable Sections ............................................................................................... 613
13.213.2. Assumptions and Restrictions ......................................................................................... 614
13.214. COMBI214 - 2D Spring-Damper Bearing ................................................................................... 614
13.214.1. Matrices ......................................................................................................................... 614
[Link]. Matrices for User-Input Characteristics (KEYOPT(1) = 0) ........................................... 614
[Link].1. General Case ................................................................................................. 614
[Link].2. The Case of Rotating Reference Frame Dynamics ............................................ 615
[Link]. Equations for Reynolds Equation Integration (KEYOPT(1) > 0) .................................. 616
13.214.2. Output Quantities ........................................................................................................... 616
[Link]. Output for User-Input Characteristics (KEYOPT(1) = 0) ............................................. 616
[Link]. Output for Reynolds Equation Integration (KEYOPT(1) > 0) ...................................... 617
13.215. CPT215 - 3D 8-Node Coupled Pore-Pressure-Thermal Mechanical Solid ..................................... 617
13.215.1. Other Applicable Sections ............................................................................................... 618
13.216. CPT216 - 3D 20-Node Coupled Pore-Pressure-Thermal Mechanical Solid ................................... 618
13.216.1. Other Applicable Sections ............................................................................................... 619
13.217. CPT217 - 3D 10-Node Coupled Pore-Pressure-Thermal Mechanical Solid ................................... 619
13.217.1. Other Applicable Sections ............................................................................................... 620
13.218. FLUID218 - 3D Hydrodynamic Bearing Element ........................................................................ 620
13.218.1. Other Applicable Sections ............................................................................................... 620
13.218.2. Assumptions and Restrictions ......................................................................................... 620
13.219. Reserved for Future Use ........................................................................................................... 621
13.220. Reserved for Future Use ........................................................................................................... 621
13.221. Reserved for Future Use ........................................................................................................... 621
13.222. PLANE222 - 2D 4-Node Coupled-Field Solid ............................................................................. 621
13.222.1. Other Applicable Sections ............................................................................................... 622
13.222.2. Theory ............................................................................................................................ 622
13.223. PLANE223 - 2D 8-Node Coupled-Field Solid ............................................................................. 623
13.223.1. Other Applicable Sections ............................................................................................... 624
13.224. Reserved for Future Use ........................................................................................................... 624
13.225. SOLID225 - 3D 8-Node Coupled-Field Solid .............................................................................. 625
13.225.1. Other Applicable Sections ............................................................................................... 626
13.225.2. Theory ............................................................................................................................ 626
13.226. SOLID226 - 3D 20-Node Coupled-Field Solid ............................................................................ 627
13.226.1. Other Applicable Sections ............................................................................................... 629
13.227. SOLID227 - 3D 10-Node Coupled-Field Solid ............................................................................ 629
13.227.1. Other Applicable Sections ............................................................................................... 630
13.228. LINK228 - 3D Coupled-field Link ............................................................................................... 631
13.228.1. Other Applicable Sections ............................................................................................... 631
13.228.2. Assumptions and Restrictions ......................................................................................... 632
13.229. Reserved for Future Use ........................................................................................................... 632

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxi
Theory Reference

13.230. PLANE230 - 2D 8-Node Electric Solid ........................................................................................ 632

13.230.1. Other Applicable Sections ............................................................................................... 632
13.230.2. Assumptions and Restrictions ......................................................................................... 632
13.231. SOLID231 - 3D 20-Node Electric Solid ....................................................................................... 633
13.231.1. Other Applicable Sections ............................................................................................... 633
13.232. SOLID232 - 3D 10-Node Tetrahedral Electric Solid ..................................................................... 633
13.232.1. Other Applicable Sections ............................................................................................... 633
13.233. PLANE233 - 2D 8-Node Electromagnetic Solid .......................................................................... 634
13.233.1. Other Applicable Sections ............................................................................................... 634
13.233.2. Assumptions and Restrictions ......................................................................................... 634
13.234. Reserved for Future Use ........................................................................................................... 634
13.235. Reserved for Future Use ........................................................................................................... 634
13.236. SOLID236 - 3D 20-Node Electromagnetic Solid ......................................................................... 635
13.236.1. Other Applicable Sections ............................................................................................... 635
13.237. SOLID237 - 3D 10-Node Electromagnetic Solid ......................................................................... 635
13.237.1. Other Applicable Sections ............................................................................................... 636
13.238. PLANE238 - 2D 8-Node Diffusion Solid ..................................................................................... 636
13.238.1. Other Applicable Sections ............................................................................................... 637
13.238.2. Assumptions and Restrictions ......................................................................................... 637
13.239. SOLID239 - 3D 20-Node Diffusion Solid .................................................................................... 637
13.239.1. Other Applicable Sections ............................................................................................... 637
13.240. SOLID240 - 3D 10-Node Tetrahedral Diffusion Solid .................................................................. 638
13.240.1. Other Applicable Sections ............................................................................................... 638
13.241. HSFLD241 - 2D Hydrostatic Fluid .............................................................................................. 638
13.242. HSFLD242 - 3D Hydrostatic Fluid .............................................................................................. 639
13.242.1. Introduction ................................................................................................................... 640
13.242.2. Element Matrices and Load Vectors ................................................................................. 640
[Link]. Incompressible Fluid .............................................................................................. 642
[Link]. Liquid .................................................................................................................... 643
[Link]. Gas ........................................................................................................................ 644
[Link]. Pressure-Volume Data ............................................................................................ 644
13.243. Reserved for Future Use ........................................................................................................... 645
13.244. Reserved for Future Use ........................................................................................................... 645
13.245. Reserved for Future Use ........................................................................................................... 645
13.246. Reserved for Future Use ........................................................................................................... 645
13.247. Reserved for Future Use ........................................................................................................... 645
13.248. Reserved for Future Use ........................................................................................................... 645
13.249. Reserved for Future Use ........................................................................................................... 645
13.250. COMBI250 - 3D Bushing Element ............................................................................................. 645
13.250.1. Matrices ......................................................................................................................... 645
[Link]. Stiffness Matrix ...................................................................................................... 646
[Link]. Viscous Damping Matrix ......................................................................................... 646
[Link]. Structural Damping Matrix ..................................................................................... 646
13.251. SURF251 - 2D Radiosity Surface ............................................................................................... 647
13.252. SURF252 - 3D Thermal Radiosity Surface .................................................................................. 647
13.253. Reserved for Future Use ........................................................................................................... 647
13.254. Reserved for Future Use ........................................................................................................... 648
13.255. Reserved for Future Use ........................................................................................................... 648
13.256. Reserved for Future Use ........................................................................................................... 648
13.257. INFIN257 - Structural Infinite Solid ........................................................................................... 648
13.257.1. Structural Infinite Element for Static Analysis ................................................................... 649

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxii Published: 2023-10-23
 Theory Reference

13.257.2. Structural Infinite Element for Dynamic Analysis .............................................................. 652

13.258. Reserved for Future Use ........................................................................................................... 654
13.259. Reserved for Future Use ........................................................................................................... 654
13.260. Reserved for Future Use ........................................................................................................... 654
13.261. Reserved for Future Use ........................................................................................................... 654
13.262. Reserved for Future Use ........................................................................................................... 655
13.263. REINF263 - 2D Smeared Reinforcing ......................................................................................... 655
13.263.1. Other Applicable Sections ............................................................................................... 656
13.264. REINF264 - 3D Discrete Reinforcing .......................................................................................... 656
13.264.1. Structural ....................................................................................................................... 657
13.264.2. Thermal .......................................................................................................................... 657
13.264.3. Other Applicable Sections ............................................................................................... 657
13.265. REINF265 - 3D Smeared Reinforcing ......................................................................................... 658
13.265.1. Structural ....................................................................................................................... 658
13.265.2. Thermal .......................................................................................................................... 659
13.265.3. Other Applicable Sections ............................................................................................... 659
13.265.4. Stiffness and Mass Matrices of a Reinforcing Layer ........................................................... 659
13.266. Reserved for Future Use ........................................................................................................... 660
13.267. Reserved for Future Use ........................................................................................................... 660
13.268. Reserved for Future Use ........................................................................................................... 660
13.269. Reserved for Future Use ........................................................................................................... 660
13.270. Reserved for Future Use ........................................................................................................... 660
13.271. Reserved for Future Use ........................................................................................................... 660
13.272. SOLID272 - General Axisymmetric Solid with 4 Base Nodes ....................................................... 661
13.272.1. Other Applicable Sections ............................................................................................... 662
13.272.2. Assumptions and Restrictions ......................................................................................... 662
13.273. SOLID273 - General Axisymmetric Solid with 8 Base Nodes ....................................................... 662
13.273.1. Other Applicable Sections ............................................................................................... 663
13.273.2. Assumptions and Restrictions ......................................................................................... 663
13.274. Reserved for Future Use ........................................................................................................... 663
13.275. Reserved for Future Use ........................................................................................................... 663
13.276. Reserved for Future Use ........................................................................................................... 663
13.277. Reserved for Future Use ........................................................................................................... 663
13.278. SOLID278 - 3D 8-Node Homogeneous/Layered Thermal Solid .................................................. 664
13.278.1. SOLID278 - 3D 8-Node Homogeneous Thermal Solid ....................................................... 664
13.278.2. SOLID278 - 3D 8-Node Layered Thermal Solid .................................................................. 664
13.278.3. Other Applicable Sections ............................................................................................... 665
13.279. SOLID279 - 3D 20-Node Homogeneous/Layered Thermal Solid ................................................ 665
13.279.1. SOLID279 - 3D 20-Node Homogeneous Thermal Solid ..................................................... 665
13.279.2. SOLID279 - 3D 20-Node Layered Thermal Solid ................................................................ 666
13.280. CABLE280 - 3D 3-Node Cable ................................................................................................... 666
13.281. SHELL281 - 8-Node Shell .......................................................................................................... 667
13.281.1. Other Applicable Sections ............................................................................................... 669
13.281.2. Assumptions and Restrictions ......................................................................................... 669
13.281.3. Membrane Option .......................................................................................................... 669
13.281.4. Shear Correction ............................................................................................................. 669
13.282. Reserved for Future Use ........................................................................................................... 669
13.283. Reserved for Future Use ........................................................................................................... 669
13.284. Reserved for Future Use ........................................................................................................... 669
13.285. SOLID285 - 3D 4-Node Tetrahedral Structural Solid ................................................................... 670
13.285.1. Other Applicable Sections ............................................................................................... 670

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxiii
Theory Reference

13.285.2. Theory ............................................................................................................................ 670

13.286. Reserved for Future Use ........................................................................................................... 670
13.287. Reserved for Future Use ........................................................................................................... 670
13.288. PIPE288 - 3D 2-Node Pipe ........................................................................................................ 671
13.288.1. Assumptions and Restrictions ......................................................................................... 672
13.288.2. Ocean Effects .................................................................................................................. 673
[Link]. Location of the Element ......................................................................................... 673
[Link]. Mass Matrix ........................................................................................................... 674
[Link]. Load Vector ............................................................................................................ 674
[Link]. Hydrostatic Effects ................................................................................................. 675
[Link]. Hydrodynamic Effects ............................................................................................ 675
13.288.3. Stress Evaluation ............................................................................................................. 675
13.289. PIPE289 - 3D 3-Node Pipe ........................................................................................................ 675
13.290. ELBOW290 - 3D 3-Node Elbow ................................................................................................. 676
13.290.1. Other Applicable Sections ............................................................................................... 677
13.290.2. Assumptions and Restrictions ......................................................................................... 677
13.290.3. Shear Correction ............................................................................................................. 677
13.291. SOLID291 - 3D 10-Node Tetrahedral Thermal Solid ................................................................... 678
13.291.1. Other Applicable Sections ............................................................................................... 678
13.292. PLANE292 - 2D 4-Node Thermal Element ................................................................................. 678
13.292.1. Other Applicable Sections ............................................................................................... 679
13.293. PLANE293 - 2D 8-Node Thermal Solid ...................................................................................... 679
13.293.1. Other Applicable Sections ............................................................................................... 679
13.293.2. Assumptions and Restrictions ......................................................................................... 679
14. Analysis Tools .................................................................................................................................... 681
14.1. Acceleration Effect ...................................................................................................................... 681
14.1.1. Acceleration Due to One Rotation ....................................................................................... 682
14.1.2. Acceleration Due to Two Rotations ...................................................................................... 682
[Link]. Rotation 1 and Rotation 3 .......................................................................................... 683
[Link]. Rotation 1 and Rotation 2 .......................................................................................... 683
[Link]. Rotation 2 and Rotation 3 .......................................................................................... 684
[Link]. Additional Term for Point Mass Element ...................................................................... 685
14.2. Inertia Relief ................................................................................................................................ 685
14.2.1. Mass-Related Information Calculation ................................................................................. 687
14.2.2. Supported Analysis Types for Inertia Relief .......................................................................... 689
[Link]. Linear Static Analysis .................................................................................................. 689
[Link]. Linear Static Analysis Without Supports ...................................................................... 690
[Link]. Nonlinear Static Analysis ............................................................................................ 691
[Link]. Linear Perturbation Static Analysis with Inertia Relief .................................................. 692
[Link]. Linear Perturbation Eigenvalue Buckling Analysis ....................................................... 692
14.3. Damping Matrices ....................................................................................................................... 694
[Link] (FULL) Analysis ..................................................................................................... 694
14.3.2. Damped Modal Analysis ..................................................................................................... 696
14.3.3. Harmonic (FULL) Analysis ................................................................................................... 696
14.3.4. Mode-Superposition Analysis ............................................................................................. 698
[Link]. Mode-Superposition Analyses Following a QR Damp Modal Analysis .......................... 700
14.4. Rotating Structures ..................................................................................................................... 700
14.4.1. Rotating Reference Frame Equations ................................................................................... 701
[Link]. Coriolis Matrix, Spin-Softening Matrix, and Centrifugal Load ........................................ 702
[Link]. Coriolis Force ............................................................................................................. 703
14.4.2. Stationary Reference Frame Equations ................................................................................ 703

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxiv Published: 2023-10-23
 Theory Reference

[Link]. Gyroscopic Matrix ...................................................................................................... 704

[Link].1. Kinetic Energy for the Gyroscopic Matrix Calculation of Lumped Mass and Beam
Element ............................................................................................................................ 704
[Link].2. General Expression of the Kinetic Energy for the Gyroscopic Matrix Calculation ... 705
[Link]. Rotating Damping Matrix ........................................................................................... 705
[Link]. Gyroscopic Force ....................................................................................................... 707
[Link]. Mass Unbalance Transient Forces ............................................................................... 707
14.5. Automatic Time-Stepping ........................................................................................................... 708
14.5.1. Time-Step Prediction .......................................................................................................... 709
14.5.2. Time-Step Bisection ............................................................................................................ 710
14.5.3. The Response Eigenvalue for 1st Order Transients ............................................................... 710
[Link]. Additional Considerations for Thermal Transient Analysis ............................................ 711
[Link] Response Frequency for Structural Dynamics ................................................................ 711
14.5.5. Creep Time Increment ........................................................................................................ 712
14.5.6. Plasticity Time Increment .................................................................................................... 712
14.5.7. Midstep Residual for Structural Dynamic Analysis ................................................................ 712
14.6. Solving for Unknowns and Reactions ........................................................................................... 713
14.6.1. Reaction Forces .................................................................................................................. 714
14.6.2. Disequilibrium .................................................................................................................... 716
14.7. Equation Solvers ......................................................................................................................... 717
14.7.1. Direct Solvers ..................................................................................................................... 717
14.7.2. Sparse Direct Solver ............................................................................................................ 718
14.7.3. Iterative Solver ................................................................................................................... 719
14.8. Mode-Superposition Method ...................................................................................................... 721
14.8.1. General Equations .............................................................................................................. 721
14.8.2. Equations for QR Damped Eigensolver Based Analysis ......................................................... 724
14.8.3. Equations for Unsymmetric Eigensolver Based Analysis ....................................................... 725
14.8.4. Modal Damping ................................................................................................................. 725
14.8.5. Residual Vector Method ...................................................................................................... 725
14.8.6. Residual Response Method ................................................................................................. 726
14.9. Extraction of Modal Damping Parameter for Squeeze Film Problems ............................................ 727
14.10. Newton-Raphson Procedure ...................................................................................................... 730
14.10.1. Overview .......................................................................................................................... 730
14.10.2. Convergence .................................................................................................................... 734
14.10.3. Predictor .......................................................................................................................... 736
14.10.4. Adaptive Descent ............................................................................................................. 737
14.10.5. Line Search ....................................................................................................................... 738
14.10.6. Arc-Length Method .......................................................................................................... 739
14.11. Constraint Equations ................................................................................................................. 742
14.11.1. Derivation of Matrix and Load Vector Operations ............................................................... 742
14.11.2. Constraints: Automatic Selection of Dependent Degrees of Freedom ................................ 744
14.12. Eigenvalue and Eigenvector Extraction ...................................................................................... 745
14.12.1. Supernode Method .......................................................................................................... 747
14.12.2. Block Lanczos ................................................................................................................... 748
14.12.3. PCG Lanczos ..................................................................................................................... 748
14.12.4. Unsymmetric Method ....................................................................................................... 748
14.12.5. Subspace Method ............................................................................................................. 750
14.12.6. Damped Method .............................................................................................................. 751
14.12.7. QR Damped Method ......................................................................................................... 752
[Link]. QR Damped Method with Constant Structural Damping ........................................... 753
[Link]. QR Damped Method Output .................................................................................... 754

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxv
Theory Reference

14.12.8. Shifting ............................................................................................................................ 754

14.12.9. Repeated Eigenvalues ....................................................................................................... 755
14.12.10. Complex Eigensolutions ................................................................................................. 756
14.13. Analysis of Cyclically Symmetric Structures ................................................................................ 758
14.13.1. Modal Analysis ................................................................................................................. 758
14.13.2. Complete Mode Shape Derivation .................................................................................... 759
14.13.3. Mode-Superposition Harmonic Analysis ............................................................................ 760
[Link].Transform to Modal Coordinates ............................................................................... 761
[Link]. Cyclic Coordinates ................................................................................................... 762
[Link]. Properties of Paired Eigenmodes .............................................................................. 762
[Link]. >Forcing .................................................................................................................. 763
[Link]. Damping ................................................................................................................. 766
14.13.4. Aerodynamic Coupling ..................................................................................................... 767
14.13.5. Expansion to Output Quantities ........................................................................................ 768
14.13.6. Mistuning ......................................................................................................................... 768
[Link]. Equation of Motion for a Mistuned System ............................................................... 768
[Link]. Transform to Tuned-System Modal Coordinates ........................................................ 768
[Link]. Craig-Bampton Reduction for Mistuning .................................................................. 769
[Link]. Component Mode Projection ................................................................................... 769
[Link]. Mistuning Equation of Motion with Cyclic Quantities ................................................ 770
[Link].1. Evaluation of Participation Factors ................................................................... 770
[Link]. Mistuning Parameters .............................................................................................. 771
14.13.7. Cyclic Symmetry Transformations ...................................................................................... 771
14.14. Mass Related Information .......................................................................................................... 773
14.14.1. Precise Calculation of Mass Related Information ................................................................ 773
14.14.2. Lumped Calculation of Mass Related Information .............................................................. 775
[Link]. Accuracy of the Lumped Calculation ........................................................................ 777
[Link]. Effect of KSUM, LSUM, ASUM, and VSUM Commands ................................................. 777
14.15. Energies .................................................................................................................................... 778
14.15.1. Potential Energy (or Stiffness Energy) ................................................................................ 778
[Link]. Potential Energy (or Stiffness Energy) for Complex Results ........................................ 779
[Link]. Potential Energy Limitations ..................................................................................... 780
14.15.2. Kinetic Energy .................................................................................................................. 781
[Link]. Kinetic Energy for Complex Results ........................................................................... 781
[Link]. Kinetic Energy Limitations ........................................................................................ 782
14.15.3. Damping Energy .............................................................................................................. 782
[Link]. Damping Energy for Complex Results ....................................................................... 783
[Link]. Damping Energy Limitations .................................................................................... 783
14.15.4. Work Done by External Loads ............................................................................................ 783
[Link]. Work Done by External Loads for Complex Results .................................................... 784
[Link]. Work Limitations ...................................................................................................... 784
14.15.5. Artificial Energy ................................................................................................................ 785
14.15.6. Element Support for Energies ........................................................................................... 785
14.16. Reduced-Order Modeling for State-Space Matrices Export .......................................................... 786
14.17. Enforced Motion in Structural Analysis ....................................................................................... 788
14.17.1. Full Method for Transient and Harmonic Analyses .............................................................. 789
14.17.2. Enforced Motion Method for Transient and Harmonic Analyses .......................................... 789
[Link]. Structure Subjected to Differential Support Motion .................................................. 790
[Link]. Structure Subjected to Global Support Acceleration ................................................. 791
14.17.3. Large Mass Method .......................................................................................................... 792
15. Analysis Procedures .......................................................................................................................... 793

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxvi Published: 2023-10-23
 Theory Reference

15.1. Static Analysis ............................................................................................................................. 793

15.1.1. Assumptions and Restrictions ............................................................................................. 794
15.1.2. Description of Structural Systems ....................................................................................... 794
15.1.3. Description of Thermal, Magnetic and Other First Order Systems ......................................... 795
15.2. Transient Analysis ........................................................................................................................ 796
15.2.1. Assumptions and Restrictions ............................................................................................. 796
15.2.2. Description of Structural and Other Second Order Systems ................................................. 796
[Link]. Time Integration Scheme for Linear Systems ............................................................... 797
[Link]. Time Integration Scheme for Nonlinear Systems ......................................................... 801
[Link]. Solution ..................................................................................................................... 803
[Link].1. Full Solution Method ......................................................................................... 804
[Link].2. Mode-Superposition Method ............................................................................ 807
15.2.3. Description of Thermal, Magnetic and Other First Order Systems ......................................... 807
15.3. Modal Analysis ............................................................................................................................ 810
15.3.1. Assumptions and Restrictions ............................................................................................. 811
15.3.2. Description of Analysis for Symmetric Undamped Systems .................................................. 811
15.3.3. Participation Factors ........................................................................................................... 812
15.3.4. Effective Mass and Cumulative Mass Fraction ...................................................................... 812
15.3.5. Modal Mass and Kinetic Energy ........................................................................................... 812
15.4. Harmonic Analysis ....................................................................................................................... 813
15.4.1. Harmonic Analysis Assumptions and Restrictions ................................................................ 813
15.4.2. Description of Harmonic Analysis ........................................................................................ 814
15.4.3. Harmonic Analysis Complex Displacement Output .............................................................. 815
15.4.4. Nodal and Reaction Load Computation in a Harmonic Analysis ........................................... 815
15.4.5. Harmonic Analysis Solution ................................................................................................ 817
[Link]. Full Solution Method ................................................................................................. 817
[Link]. Frequency-Sweep Method ......................................................................................... 817
[Link].1. Viscous or Hysteretic Damping .......................................................................... 817
[Link]. Mode-Superposition Method ..................................................................................... 819
[Link].1. Expansion Pass .................................................................................................. 821
15.4.6. Automatic Frequency Spacing in a Harmonic Analysis ......................................................... 821
15.4.7. Logarithm Frequency Spacing in a Harmonic Analysis ......................................................... 822
15.4.8. Harmonic Analysis with Rotating Forces on Rotating Structures ........................................... 823
[Link]. General Asynchronous Rotating Force ........................................................................ 823
[Link]. Specific Synchronous Forces: Mass Unbalance ............................................................ 824
15.4.9. Harmonic Ocean Wave Procedure (HOWP) .......................................................................... 825
15.5. Buckling Analysis ........................................................................................................................ 826
15.5.1. Assumptions and Restrictions ............................................................................................. 826
15.5.2. Description of Analysis ....................................................................................................... 827
15.6. Substructuring Analysis ............................................................................................................... 828
15.6.1. Assumptions and Restrictions (Within Superelement) ......................................................... 828
15.6.2. Description of Analysis ....................................................................................................... 828
15.6.3. Statics ................................................................................................................................ 828
15.6.4. Transients ........................................................................................................................... 831
15.6.5. Component Mode Synthesis (CMS) ..................................................................................... 831
15.7. Spectrum Analysis ....................................................................................................................... 838
15.7.1. Assumptions and Restrictions ............................................................................................. 838
15.7.2. Description of Analysis ....................................................................................................... 839
15.7.3. Single-Point Response Spectrum ........................................................................................ 839
15.7.4. Damping ............................................................................................................................ 839
15.7.5. Participation Factors and Mode Coefficients ........................................................................ 840

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxvii
Theory Reference

15.7.6. Combination of Modes ....................................................................................................... 843

[Link]. Complete Quadratic Combination Method ................................................................. 844
[Link]. Grouping Method ...................................................................................................... 844
[Link]. Double Sum Method ................................................................................................. 845
[Link]. Square Root of the Sum of the Squares (SRSS) Method ............................................... 845
[Link]. Naval Research Laboratory Sum (NRL-SUM) Method ................................................... 845
[Link]. Closely Spaced Modes (CSM) Method ......................................................................... 846
[Link]. Rosenblueth Method ................................................................................................. 847
15.7.7. Effective Mass and Cumulative Mass Fraction ...................................................................... 847
15.7.8. Dynamic Design Analysis Method ....................................................................................... 848
15.7.9. Random Vibration Method ................................................................................................. 849
15.7.10. Description of Method ...................................................................................................... 850
15.7.11. Response Power Spectral Densities and Mean Square Response ........................................ 851
[Link]. Variance of Element Results and Reaction Forces ...................................................... 853
[Link]. Equivalent Stress Mean Square Response ................................................................. 854
15.7.12. Cross Spectral Terms for Partially Correlated Input PSDs ..................................................... 854
15.7.13. Spatial Correlation ............................................................................................................ 855
15.7.14. Wave Propagation ............................................................................................................ 856
15.7.15. Multi-Point Response Spectrum Method ........................................................................... 857
[Link]. Absolute Sum Combination Method ........................................................................ 857
15.7.16. Missing-Mass Response .................................................................................................... 858
15.7.17. Rigid Responses ............................................................................................................... 860
15.8. Linear Perturbation Analysis ........................................................................................................ 861
15.8.1. Assumptions and Restrictions ............................................................................................. 861
15.8.2. Description of Analysis ....................................................................................................... 862
15.8.3. Static Analysis Based on Linear Perturbation ....................................................................... 863
15.8.4. Modal Analysis Based on Linear Perturbation ...................................................................... 863
15.8.5. Eigenvalue Buckling Analysis Based on Linear Perturbation ................................................. 864
15.8.6. Harmonic Analysis Based on Linear Perturbation ................................................................. 867
15.8.7. Substructure or CMS Generation Based on Linear Perturbation ............................................ 868
15.8.8. Application of Perturbation Loads ....................................................................................... 868
15.8.9. Downstream Analysis Using the Solution of a Linear Perturbation Analysis .......................... 870
15.9. Semi-Implicit Analysis ................................................................................................................. 870
15.9.1. Assumptions and Restrictions ............................................................................................. 871
15.9.2. Description of Analysis ....................................................................................................... 871
15.9.3. Stability and Selective Mass Scaling .................................................................................... 872
15.9.4. Artificial Bulk Viscosity ........................................................................................................ 873
16. Preprocessing and Postprocessing Tools .......................................................................................... 875
16.1. Integration and Differentiation Procedures .................................................................................. 875
16.1.1. Single Integration Procedure .............................................................................................. 875
16.1.2. Double Integration Procedure ............................................................................................ 876
16.1.3. Differentiation Procedure ................................................................................................... 876
16.1.4. Double Differentiation Procedure ....................................................................................... 877
16.2. Fourier Coefficient Evaluation ...................................................................................................... 877
16.3. Statistical Procedures .................................................................................................................. 878
16.3.1. Mean, Covariance, Correlation Coefficient ............................................................................ 879
16.3.2. Random Samples of a Uniform Distribution ......................................................................... 880
16.3.3. Random Samples of a Gaussian Distribution ....................................................................... 880
16.3.4. Random Samples of a Triangular Distribution ...................................................................... 881
16.3.5. Random Samples of a Beta Distribution .............................................................................. 882
16.3.6. Random Samples of a Gamma Distribution ......................................................................... 883

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxviii Published: 2023-10-23
 Theory Reference

17. Postprocessing .................................................................................................................................. 885

17.1. POST1 - Derived Nodal Data Processing ....................................................................................... 885
17.1.1. Derived Nodal Data Computation ....................................................................................... 885
17.2. POST1 - Vector Operations ........................................................................................................... 886
17.3. POST1 - Path Operations ............................................................................................................. 887
17.3.1. Defining the Path ............................................................................................................... 887
17.3.2. Defining Orientation Vectors of the Path ............................................................................. 887
17.3.3. Mapping Nodal and Element Data onto the Path ................................................................. 889
17.3.4. Operating on Path Data ...................................................................................................... 889
17.4. POST1 - Stress Linearization ......................................................................................................... 890
17.4.1. Cartesian Case .................................................................................................................... 891
17.4.2. Axisymmetric Case (General) .............................................................................................. 893
17.4.3. Axisymmetric Case ............................................................................................................. 897
17.5. POST1 - Electromagnetic Macros ................................................................................................. 898
17.5.1. Flux Passing Thru a Closed Contour ..................................................................................... 899
17.5.2. Magnetomotive Forces ....................................................................................................... 899
17.5.3. Power Loss ......................................................................................................................... 900
17.5.4. Energy Supplied ................................................................................................................. 900
17.5.5. Terminal Inductance ........................................................................................................... 901
17.5.6. Flux Linkage ....................................................................................................................... 901
17.5.7. Terminal Voltage ................................................................................................................. 901
17.5.8. Energy in a Magnetic Field .................................................................................................. 902
17.5.9. Relative Error in Electrostatic or Electromagnetic Field Analysis ............................................ 902
[Link]. Electrostatics ............................................................................................................. 903
[Link].1. Electric Field ...................................................................................................... 903
[Link].2. Electric Flux Density .......................................................................................... 903
[Link]. Electromagnetics ....................................................................................................... 903
[Link].1. Magnetic Field Intensity .................................................................................... 903
[Link].2. Magnetic Flux Density ....................................................................................... 904
17.5.10. Electromotive Force .......................................................................................................... 904
17.5.11. Computation of Equivalent Transmission-line Parameters .................................................. 905
17.6. POST1 - Error Approximation Technique ...................................................................................... 906
17.6.1. Error Approximation Technique for Displacement-Based Problems ...................................... 906
17.6.2. Error Approximation Technique for Temperature-Based Problems ........................................ 908
17.6.3. Error Approximation Technique for Magnetics-Based Problems ........................................... 910
17.7. POST1 - Harmonic Solid and Shell Element Postprocessing .......................................................... 912
17.7.1. Thermal Solid Elements (PLANE75, PLANE78) ...................................................................... 912
17.7.2. Structural Solid Elements (PLANE25, PLANE83) .................................................................... 913
17.7.3. Structural Shell Element (SHELL61) ..................................................................................... 913
17.8. POST26 - Data Operations ........................................................................................................... 914
17.9. POST26 - Response Spectrum Generator (RESP) ........................................................................... 915
17.9.1. Time Step Size .................................................................................................................... 917
17.10. POST1 and POST26 - Interpretation of Equivalent Strains ............................................................ 917
17.10.1. Physical Interpretation of Equivalent Strain ....................................................................... 918
17.10.2. Elastic Strain ..................................................................................................................... 918
17.10.3. Plastic Strain ..................................................................................................................... 918
17.10.4. Creep Strain ..................................................................................................................... 919
17.10.5. Total Strain ....................................................................................................................... 919
17.11. POST26 - Response Power Spectral Density ................................................................................ 919
17.12. POST26 - Computation of Covariance ........................................................................................ 920
17.13. POST1 and POST26 – Complex Results Postprocessing ............................................................... 920

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxix
Theory Reference

17.14. POST1 - Modal Assurance Criterion (MAC) .................................................................................. 922

17.15. POST1 – Frequency response function correlation ..................................................................... 923
Bibliography ............................................................................................................................................. 925

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxx Published: 2023-10-23
List of Figures
2.1. Stress Vector Definition ........................................................................................................................... 6
2.2. Material Coordinate Systems ................................................................................................................... 9
2.3. Effects of Consistent Pressure Loading ................................................................................................... 15
3.1. Position Vectors and Motion of a Deforming Body ................................................................................. 30
3.2. Polar Decomposition of a Shearing Deformation ................................................................................... 31
3.3. Element Transformation Definitions ...................................................................................................... 37
3.4. Definition of Deformational Rotations ................................................................................................... 38
3.5. General Motion of a Fiber ...................................................................................................................... 41
3.6. Motion of a Fiber with Rigid Body Motion Removed ............................................................................... 42
3.7. Spinning Spring-Mass System ............................................................................................................... 47
3.8. Effects of Spin Softening and Stress Stiffening ....................................................................................... 49
3.9. Domains and Maps in General ALE Formulation ..................................................................................... 62
3.10. Steady-State Rolling Using ALE Formulation ........................................................................................ 62
4.1. Stress-Strain Behavior of Each of the Plasticity Options .......................................................................... 69
4.2. Various Yield Surfaces ........................................................................................................................... 69
4.3. Types of Hardening Rules ...................................................................................................................... 70
4.4. Uniaxial Behavior .................................................................................................................................. 75
4.5. Uniaxial Behavior for Multilinear Kinematic Hardening ........................................................................... 79
4.6. Shear Failure Envelope Functions .......................................................................................................... 84
4.7. Compaction Cap Function ..................................................................................................................... 85
4.8. Expansion Cap Function ........................................................................................................................ 85
4.9. Yielding Surface in π-Plane .................................................................................................................... 86
4.10. Cap Model .......................................................................................................................................... 87
4.11. Growth, Nucleation, and Coalescence of Voids in Microscopic Scale ...................................................... 90
4.12. Idealized Response of Gray Cast Iron in Tension and Compression ........................................................ 93
4.13. Cross-Section of Yield Surface ............................................................................................................. 94
4.14. Yield Surface in the Meridional Plane ................................................................................................... 94
4.15. Flow Potential for Cast Iron .................................................................................................................. 95
4.16. Material Point in Yielding Condition Elastically Predicted .................................................................... 101
4.17. Uniaxial Compression Test ................................................................................................................. 102
4.18. Creep Isosurface ............................................................................................................................... 104
4.19. Stress Projection ............................................................................................................................... 105
4.20. Cap Zoning ....................................................................................................................................... 106
4.21. Uniaxial Compression Test and Creep Isosurface for Shear and Expansion Zones ................................. 107
4.22. Pressure vs. Deflection Behavior of a Gasket Material .......................................................................... 109
4.23. Stress-Strain Behavior for Nonlinear Elasticity ..................................................................................... 111
4.24. Illustration of Deformation Modes ..................................................................................................... 123
4.25. Equivalent Deformation Modes ......................................................................................................... 124
4.26. Pure Shear from Direct Components .................................................................................................. 127
4.27. Bergstrom-Boyce Material Model Representation ............................................................................... 130
4.28. Schematic of Interface Elements ........................................................................................................ 144
4.29. Mode I Dominated Bilinear CZM Law ................................................................................................. 146
4.30. Mode II Dominated Bilinear CZM Law ................................................................................................ 147
4.31. Traction-Separation Relationship Along the Interface ......................................................................... 151
4.32. Normal Contact Stress and Contact Gap Curve for Bilinear Cohesive Zone Material ............................. 154
5.1. Electromagnetic Field Regions ............................................................................................................ 166
5.2. Energy and Co-energy for Non-Permanent Magnets ............................................................................ 200
5.3. Energy and Co-energy for Permanent Magnets .................................................................................... 200
5.4. Lumped Capacitor Model of Two Conductors and Ground ................................................................... 203

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxxi
Theory Reference

5.5. Lumped Conductor Model of Two Conductors and Ground .................................................................. 205
6.1. Specified Heat Flows ........................................................................................................................... 208
6.2. Specified Convection Surfaces ............................................................................................................. 209
6.3. Cooling Passages in a Solid .................................................................................................................. 209
6.4. View Factor Calculation Terms ............................................................................................................. 211
6.5. View Factor Facets for a General Model (No Symmetry) ........................................................................ 213
6.6. View Factor Facets for a Model with Symmetry .................................................................................... 214
6.7. Mass Transport Problem in a Flow Field ................................................................................................ 216
6.8. The Hemicube .................................................................................................................................... 223
6.9. Derivation of Delta-View Factors for Hemicube Method ...................................................................... 223
6.10. Receiving Surface Projection ............................................................................................................. 225
6.11. Axisymmetric Geometry .................................................................................................................... 225
6.12. End View of Showing n = 8 Segments ................................................................................................ 226
7.1. Flow Theory, Cut-off, and Maximum Frequency Interrelation ................................................................. 235
7.2. Hydrodynamic Bearing ....................................................................................................................... 238
8.1. Absorbing Boundary ........................................................................................................................... 252
8.2. Equivalent Source Principle ................................................................................................................. 280
8.3. Spherical Coordinates ......................................................................................................................... 280
8.4. Two-Port Transfer Admittance Matrix ................................................................................................... 283
8.5. Plane Wave in Spherical Coordinates ................................................................................................... 287
11.1. 2D Line Element ................................................................................................................................ 366
11.2. 3–D Line Element .............................................................................................................................. 367
11.3. Axisymmetric Harmonic Shell Element .............................................................................................. 370
11.4. General Axisymmetric Surface Elements (when Nnp = 3) ..................................................................... 371
11.5. 3D Shell Elements ............................................................................................................................. 373
11.6. 2D and Axisymmetric Solid Element .................................................................................................. 376
11.7. 4-Node Quadrilateral Infinite Solid Element ....................................................................................... 378
11.8. 8-Node Quadrilateral Infinite Solid Element ....................................................................................... 379
11.9. Axisymmetric Harmonic Solid Elements ............................................................................................. 380
11.10. 3D Solid Elements ........................................................................................................................... 382
11.11. 10-Node Tetrahedra Element ........................................................................................................... 383
11.12. 8-Node Brick Element ...................................................................................................................... 384
11.13. 13-Node Pyramid Element ............................................................................................................... 384
11.14. 6-Node Wedge Element ................................................................................................................... 385
11.15. 15-Node Wedge Element ................................................................................................................. 386
11.16. 8-Node Brick Element ...................................................................................................................... 387
11.17. 20-Node Brick Element .................................................................................................................... 388
11.18. 3D 8-Node Brick Element ................................................................................................................. 389
11.19. 20-Node Solid Brick Infinite Element ................................................................................................ 390
11.20. General Axisymmetric Solid Elements (when Nnp = 3) ....................................................................... 393
11.21. 1st-Order Tetrahedral Element ......................................................................................................... 396
11.22. 1st-Order Brick Element ................................................................................................................... 397
12.1. Brick Element .................................................................................................................................... 400
12.2. Pyramid Element ............................................................................................................................... 401
12.3. Wedge Element ................................................................................................................................. 401
12.4. Tetrahedron Element ......................................................................................................................... 402
12.5. Aspect Ratios for Triangles ................................................................................................................. 403
12.6. Quadrilateral Aspect Ratio Calculation ............................................................................................... 403
12.7. Aspect Ratios for Quadrilaterals ......................................................................................................... 404
12.8. Parallel Deviation Unit Vectors ........................................................................................................... 405
12.9. Parallel Deviations for Quadrilaterals ................................................................................................. 405

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxxii Published: 2023-10-23
 Theory Reference

12.10. Maximum Corner Angles for Triangles .............................................................................................. 406

12.11. Maximum Corner Angles for Quadrilaterals ...................................................................................... 406
12.12. Jacobian Ratios for Triangles ............................................................................................................ 409
12.13. Jacobian Ratios for Quadrilaterals .................................................................................................... 409
12.14. Jacobian Ratios for Quadrilaterals .................................................................................................... 409
12.15. Shell Average Normal Calculation .................................................................................................... 410
12.16. Shell Element Projected onto a Plane ............................................................................................... 411
12.17. Quadrilateral Shell Having Warping Factor ....................................................................................... 411
12.18. Warping Factor for Bricks ................................................................................................................. 412
12.19. Integration Point Locations for Quadrilaterals .................................................................................. 415
12.20. Integration Point Locations for Bricks and Pyramids ......................................................................... 415
12.21. Integration Point Locations for Triangles .......................................................................................... 416
12.22. Integration Point Locations for Tetrahedra ....................................................................................... 417
12.23. Integration Point Locations for Triangles and Tetrahedra .................................................................. 418
12.24. 6 and 9 Integration Point Locations for Wedges ................................................................................ 418
12.25. 8 Integration Point Locations for Wedges ......................................................................................... 419
12.26. Integration Point Locations for 14 Point Rule .................................................................................... 419
12.27. Nonlinear Bending Integration Point Locations ................................................................................ 420
12.28. Velocity Profiles for Wave-Current Interactions ................................................................................. 438
13.1. Element Behavior .............................................................................................................................. 467
13.2. Input Force-Deflection Curve ............................................................................................................. 472
13.3. Stiffness Computation ....................................................................................................................... 473
13.4. Input Force-Deflection Curve Reflected Through Origin ..................................................................... 474
13.5. Force-Deflection Curve with KEYOPT(2) = 1 ........................................................................................ 474
13.6. Nonconservative Unloading (KEYOPT(1) = 1) ...................................................................................... 474
13.7. No Origin Shift on Reversed Loading (KEYOPT(1) = 1) ......................................................................... 475
13.8. Origin Shift on Reversed Loading (KEYOPT(1) = 1) .............................................................................. 475
13.9. Crush Option (KEYOPT(2) = 2) ............................................................................................................ 475
13.10. Force-Deflection Relationship .......................................................................................................... 476
13.11. A Semi-infinite Boundary Element Zone and the Corresponding Boundary Element IJK .................... 480
13.12. Infinite Element IJML and the Local Coordinate System .................................................................... 481
13.13. Stress Locations .............................................................................................................................. 489
13.14. Element Orientations ...................................................................................................................... 491
13.15. Global to Local Mapping of a 1D Infinite Element ............................................................................. 509
13.16. Mapping of 2D Solid Infinite Element ............................................................................................... 509
13.17. I-V (Current-Voltage) Characteristics of CIRCU125 ............................................................................. 525
13.18. Norton Current Definition ............................................................................................................... 525
13.19. Electromechanical Transducer ......................................................................................................... 526
13.20. Absorbing Boundary ....................................................................................................................... 531
13.21. Form Factor Calculation ................................................................................................................... 543
13.22. 2D Segment Types .......................................................................................................................... 553
13.23. 3D Segment Types .......................................................................................................................... 554
13.24. Contact Detection Point Location at Gauss Point .............................................................................. 556
13.25. Surface-Projection-Based Contact .................................................................................................... 557
13.26. Penetration Distance ....................................................................................................................... 558
13.27. Smoothing Convex Corner .............................................................................................................. 558
13.28. Friction Model ................................................................................................................................. 559
13.29. Beam Sliding Inside a Hollow Beam ................................................................................................. 580
13.30. Parallel Beams in Contact ................................................................................................................ 580
13.31. Crossing Beams in Contact .............................................................................................................. 580
13.32. Force-Deflection Relations for Rigid Coulomb Option ....................................................................... 585

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxxiii
Theory Reference

13.33. 184.2 Slider Constraint Geometry .................................................................................................... 591

13.34. Section Model ................................................................................................................................. 600
13.35. 1D Infinite Element Mapping ........................................................................................................... 649
13.36. 1D Infinite Element Mapping ........................................................................................................... 651
13.37. Section Cell Model .......................................................................................................................... 673
14.1. Rotational Coordinate System (Rotations 1 and 3) .............................................................................. 683
14.2. Rotational Coordinate System (Rotations 1 and 2) .............................................................................. 684
14.3. Rotational Coordinate System (Rotations 2 and 3) .............................................................................. 685
14.4. Reference Frames .............................................................................................................................. 701
14.5. Mass Unbalance ................................................................................................................................ 707
14.6. Single Degree of Freedom Oscillator .................................................................................................. 723
14.7. Damping and Amplitude Ratio vs. Frequency ..................................................................................... 727
14.8. Fluid Pressure From Modal Excitation Distribution ............................................................................. 728
14.9. Newton-Raphson Solution - One Iteration ......................................................................................... 732
14.10. Newton-Raphson Solution - Next Iteration ....................................................................................... 732
14.11. Incremental Newton-Raphson Procedure ......................................................................................... 733
14.12. Initial-Stiffness Newton-Raphson ..................................................................................................... 733
14.13. Arc-Length Approach with Full Newton-Raphson Method ................................................................ 740
14.14. Typical Cyclically Symmetric Structure ............................................................................................. 758
14.15. Base Sector Definition ..................................................................................................................... 759
14.16. Full Model with the Cyclic-Symmetric Sector Highlighted ................................................................. 761
14.17. Forcing Sign and Numbering Convention ........................................................................................ 764
14.18. Zigzag Diagram for an Even Number of Sectors ................................................................................ 765
15.1. Applied and Reaction Load Vectors .................................................................................................... 795
15.2. Effect of Number of Substeps (NSUBST) and Ramping (KBC) on Initial Velocity for TIMINT,OFF ............. 805
15.3. Frequency Spacing ............................................................................................................................ 822
15.4. Mass Unbalance at Node I ................................................................................................................. 824
15.5. Force History ..................................................................................................................................... 826
15.6. Types of Buckling Problems ............................................................................................................... 827
15.7. Sphere of Influence Relating Spatially Correlated PSD Excitation ........................................................ 856
15.8. Linear Perturbation Started from Loadstep = 2, Substep = 3 ............................................................... 869
16.1. Integration Procedure ....................................................................................................................... 876
16.2. Uniform Density ................................................................................................................................ 880
16.3. Cumulative Probability Function ........................................................................................................ 880
16.4. Gaussian Density .............................................................................................................................. 881
16.5. Triangular Density ............................................................................................................................. 882
16.6. Beta Density ..................................................................................................................................... 883
16.7. Gamma Density ................................................................................................................................ 884
17.1. Typical Path Segment ........................................................................................................................ 887
17.2. Position and Unit Vectors of a Path .................................................................................................... 888
17.3. Mapping Data ................................................................................................................................... 889
17.4. Coordinates of Cross Section ............................................................................................................. 891
17.5. Typical Stress Distribution ................................................................................................................. 891
17.6. Axisymmetric Cross-Section .............................................................................................................. 893
17.7. Geometry Used for Axisymmetric Evaluations .................................................................................... 893
17.8. Centerline Sections ........................................................................................................................... 898
17.9. Non-Perpendicular Intersections ....................................................................................................... 898
17.10. Equivalent Two-Wire Transmission Line ............................................................................................ 905
17.11. Single Mass Oscillators .................................................................................................................... 916

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxxiv Published: 2023-10-23
List of Tables
1.1. General Terms ......................................................................................................................................... 1
1.2. Superscripts and Subscripts ................................................................................................................... 2
3.1. Interpolation Functions of Hydrostatic Pressure of Current-Technology Elements ................................... 54
3.2. Interpolation Functions of Hydrostatic Pressure for SOLID285 ................................................................ 55
4.1. Notation ............................................................................................................................................... 67
4.2. Summary of Plasticity Options ............................................................................................................... 70
4.3. Material Parameter Units for the Anand Option .................................................................................... 100
8.1. Coefficients of the Approximation Functions for Delany-Bazley and Miki Models .................................. 268
10.1. Elements Used for Coupled Effects .................................................................................................... 309
10.2. Coupling Methods ............................................................................................................................ 311
10.3. Nomenclature of Coefficient Matrices ................................................................................................ 325
10.4. Losses in Piezoelectric Analysis .......................................................................................................... 338
10.5. Electroelastic Analysis Results ............................................................................................................ 346
11.1. Shape Function Labels ...................................................................................................................... 365
12.1. Aspect Ratio Limits ............................................................................................................................ 404
12.2. Parallel Deviation Limits .................................................................................................................... 405
12.3. Maximum Corner Angle Limits .......................................................................................................... 406
12.4. Jacobian Ratio Limits ......................................................................................................................... 409
12.5. Warping Factor Limits ........................................................................................................................ 412
12.6. Gauss Numerical Integration Constants ............................................................................................. 414
12.7. Numerical Integration for Triangles .................................................................................................... 415
12.8. Numerical Integration for Tetrahedra ................................................................................................. 416
12.9. Numerical Integration for 20-Node Brick ............................................................................................ 419
12.10. Thru-Thickness Numerical Integration .............................................................................................. 420
12.11. Wave Theory Table ........................................................................................................................... 428
12.12. Assumed Data Variation of Stresses .................................................................................................. 443
14.1. Procedures Used for Eigenvalue and Eigenvector Extraction ............................................................... 746
14.2. Aliased Engine Order (Excited Harmonic Index) .................................................................................. 765
14.3. Element Types Supporting Additional Energy Calculations ................................................................. 785
14.4. Exceptions for Element Energies ........................................................................................................ 786
15.1. Nomenclature ................................................................................................................................... 795
15.2. Nomenclature ................................................................................................................................... 808
15.3. Types of Spectrum Loading ............................................................................................................... 839
17.1. POST26 Operations ........................................................................................................................... 914

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 xxxv
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
xxxvi Published: 2023-10-23
Chapter 1: Introduction
The Mechanical APDL Theory Reference (p. 1) presents theoretical descriptions of many elements,
commands, and procedures in Mechanical APDL and related products (p. 3). It is available to any
customer who wants to better understand how the program uses input data to calculate output, and
how to interpret various element and command results.

The following topics are available:

1.1. Overview of the Theory Reference
1.2. Understanding Theory Reference Notation
1.3. Applicable Products

1.1. Overview of the Theory Reference

By understanding the theory underlying the Mechanical APDL product capabilities, you can use those
capabilities more intelligently and with greater confidence while also being aware of their limitations.
Studying the entire reference is unnecessary; you need only refer to specific sections as needed.

This reference does not present all theory relating to finite element analysis. If you require the theory
behind the basic finite element method, many third-party sources for the topic are available. If you
need theory information beyond what is presented here, consult the indicated bibliographic reference,
run a simple test problem to try the feature of interest, or contact your Ansys Support Distributor for
more information.

1.2. Understanding Theory Reference Notation

The notation defined below is a partial list of the notation used throughout the manual. There are also
some tables of definitions given in following sections:

• Coupling (p. 309)

• Rate-Independent Plasticity (p. 67)

Due to the wide variety of topics covered in this manual, some exceptions will exist.

Table 1.1: General Terms

Term Meaning
[B] strain-displacement matrix
[C] damping matrix
[Ct] specific heat matrix
[D] elasticity matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 1
Introduction

Term Meaning
E Young's modulus
{F} force vector
[I] identity matrix
{I} current vector, associated with electrical potential degrees of
freedom
{J} current vector, associated with magnetic potential degrees of
freedom
[K] stiffness matrix
t
[K ] conductivity matrix
[M] mass matrix
[O] null matrix
P, {P} pressure (vector)
{Q} heat flow vector
[S] stress stiffness matrix
{T} temperature vector
t time, thickness
[TR] local to global conversion matrix
u, v, w, {u} displacement, displacement vector
{V} electric potential vector
δU virtual internal work
δV virtual external work
{W} fluid flow vector
x, y, z element coordinate
X, Y, Z nodal coordinates (usually global Cartesian)
α coefficient of thermal expansion
ε strain
ν Poisson's ratio
σ stress

Below is a partial list of superscripts and subscripts used on [K], [M], [C], [S], {u}, {T}, and/or {F}. See also
Coupling (p. 309). The absence of a subscript on the above terms implies the total matrix in final form,
ready for solution.

Table 1.2: Superscripts and Subscripts

Term Meaning
ac nodal effects caused by an acceleration field
c convection surface
cr creep
e based on element in global coordinates

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
2 Published: 2023-10-23
 Applicable Products

Term Meaning
el elastic
g internal heat generation
i equilibrium iteration number
based on element in element coordinates
I independent
n substep number (time step)
nd effects applied directly to node
pl plasticity
pr pressure
D dependent
sw swelling
t, th thermal
^ (flex over term) reduced matrices and vectors
. (dot over term) time derivative

1.3. Applicable Products

This document applies to the following family of products:

Ansys Mechanical Pro (structural, full thermal, and vibration analysis)

Ansys Mechanical Premium (nonlinear structural, full thermal, vibration, linear dynamics analysis)
Ansys Mechanical Enterprise (advanced nonlinear structural, full thermal, vibration, linear dy-
namics and nonlinear transient dynamics, fracture, acoustics, full coupled-field, hydrodynamics)
Ansys Mechanical Enterprise PrepPost
Ansys Mechanical Enterprise Solver (batch mode)

Note:

While connection capabilities and High Performance Computing are included as part of
the release distribution, they are separately licensed and separately installed products.
Contact your Support Representative if you want to install and run any of these separately
licensed products at your site.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 3
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
4 Published: 2023-10-23
Chapter 2: Structures
The following topics are available for structures:
2.1. Structural Fundamentals
2.2. Derivation of Structural Matrices
2.3. Structural Strain and Stress Evaluations
2.4. Combined Stresses and Strains
2.5. Perfectly Matched Layers (PML) in Elastic Media

2.1. Structural Fundamentals

The following topics concerning structural fundamentals are available:
2.1.1. Stress-Strain Relationships
2.1.2. Orthotropic Material Transformation for Axisymmetric Models
[Link]-Dependent Coefficient of Thermal Expansion

2.1.1. Stress-Strain Relationships

This section discusses material relationships for linear materials. Nonlinear materials are discussed in
Structures with Material Nonlinearities (p. 65). The stress is related to the strains by:
(2.1)

where:

{σ} = stress vector = (output as S)

[D] = elasticity or elastic stiffness matrix or stress-strain matrix (defined in Equation 2.14 (p. 8)
through Equation 2.19 (p. 8)) or inverse defined in Equation 2.4 (p. 7) or, for a few aniso-
tropic elements, defined by full matrix definition (input with TB,ANEL.)
{εel} = {ε} - {εth} = elastic strain vector (output as EPEL)

{ε} = total strain vector =

{εth} = thermal strain vector (defined in Equation 2.3 (p. 6)) (output as EPTH)

Note:

{εel} (output as EPEL) are the strains that cause stresses.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 5
Structures

The shear strains (εxy, εyz, and εxz) are the engineering shear strains, which are twice the
tensor shear strains. The ε notation is commonly used for tensor shear strains, but is used
here as engineering shear strains for simplicity of output.

A related quantity used in POST1 labeled "component total strain" (output as EPTO) is de-
scribed in Structures with Material Nonlinearities (p. 65).

The stress vector is shown in the figure below. The sign convention for direct stresses and strains
used in Mechanical APDL is that tension is positive and compression is negative. For shears, positive
is when the two applicable positive axes rotate toward each other.

Figure 2.1: Stress Vector Definition

Equation 2.1 (p. 5) may also be inverted to:

(2.2)

For the 3D case, the thermal strain vector is:

(2.3)

where:

= x component of secant coefficient of thermal expansion (see Temperature-Dependent

Coefficient of Thermal Expansion (p. 10))
ΔT = T - Tref
T = current temperature at the point in question
Tref = reference (strain-free) temperature (input on TREF command or as REFT on MP com-
mand)

The flexibility or compliance matrix, [D]-1 is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
6 Published: 2023-10-23
 Structural Fundamentals

(2.4)

where typical terms are:

Ex = x component of Young's modulus (input as EX on MP command)

νxy = major Poisson's ratio (input as PRXY on MP command)
νyx = minor Poisson's ratio (input as NUXY on MP command)
Gxy = xy component of shear modulus (input as GXY on MP command)

The difference between νxy and νyx is described below.

Also, the [D]-1 matrix is presumed to be symmetric, so that:

(2.5)

(2.6)

(2.7)

Because of the above three relationships, νxy, νyz, νxz, νyx, νzy, and νzx are not independent quantities
and therefore the user should input either νxy, νyz, and νxz (input as PRXY, PRYZ, and PRXZ), or νyx,
νzy, and νzx (input as NUXY, NUYZ, and NUXZ). The use of Poisson's ratios for orthotropic materials
sometimes causes confusion, so that care should be taken in their use. Assuming that Ex is larger than
Ey, νxy (PRXY) is larger than νyx (NUXY). Hence, νxy is commonly referred to as the "major Poisson's
ratio", because it is larger than νyx, which is commonly referred to as the "minor" Poisson's ratio. For
orthotropic materials, the user needs to inquire of the source of the material property data as to
which type of input is appropriate. In practice, orthotropic material data are most often supplied in
the major (PR-notation) form. For isotropic materials (Ex = Ey = Ez and νxy = νyz = νxz), so it makes no
difference which type of input is used.

Expanding Equation 2.2 (p. 6) with Equation 2.3 (p. 6) through Equation 2.7 (p. 7) and writing out
the six equations explicitly,
(2.8)

(2.9)

(2.10)

(2.11)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 7
Structures

(2.12)

(2.13)

where typical terms are:

εx = x component of direct strain

σx = x component of direct stress
εxy = xy component of shear strain
σxy = xy component of shear stress

Alternatively, Equation 2.1 (p. 5) may be expanded by first inverting Equation 2.4 (p. 7) and then
combining that result with Equation 2.3 (p. 6) and Equation 2.5 (p. 7) through Equation 2.7 (p. 7)
to give six explicit equations:

(2.14)

(2.15)

(2.16)

(2.17)
(2.18)
(2.19)

where:
(2.20)

If the shear moduli Gxy, Gyz, and Gxz are not input for isotropic materials, they are computed as:

(2.21)

For orthotropic materials, the user needs to inquire of the source of the material property data as to
the correct values of the shear moduli, as there are no defaults provided by the program.

The [D] matrix must be positive definite. The program checks each material property as used by each
active element type to ensure that [D] is indeed positive definite. Positive definite matrices are defined
in Positive Definite Matrices (p. 421). In the case of temperature dependent material properties, the
evaluation is done at the uniform temperature (input as BFUNIF,TEMP) for the first load step. The
material is always positive definite if the material is isotropic or if νxy, νyz, and νxz are all zero. When
using the major Poisson's ratios (PRXY, PRYZ, PRXZ), h as defined in Equation 2.20 (p. 8) must be
positive for the material to be positive definite.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
8 Published: 2023-10-23
 Structural Fundamentals

2.1.2. Orthotropic Material Transformation for Axisymmetric Models

The transformation of material property data from the R-θ-Z cylindrical system to the x-y-z system
used for the input requires special care. The conversion of the Young's moduli is fairly direct, whereas
the correct method of conversion of the Poisson's ratios is not obvious. Consider first how the Young's
moduli transform from the global cylindrical system to the global Cartesian as used by the axisym-
metric elements for a disc:

Figure 2.2: Material Coordinate Systems

Thus, ER → Ex, Eθ → Ez, EZ → Ey. Starting with the global Cartesian system, the input for x-y-z coordinates
gives the following stress-strain matrix for the non-shear terms (from Equation 2.4 (p. 7)):

(2.22)

Rearranging so that the R-θ-Z axes match the x-y-z axes (that is, x → R, y → Z, z → θ):

(2.23)

If one coordinate system uses the major Poisson's ratios, and the other uses the minor Poisson's ratio,
an additional adjustment will need to be made.

Comparing Equation 2.22 (p. 9) and Equation 2.23 (p. 9) gives:

(2.24)
(2.25)
(2.26)
(2.27)
(2.28)
(2.29)

This assumes that: νxy, νyz, νxz and νRZ, νRθ, νZθ are all major Poisson's ratios (that is, Ex EY Ez and
ER EZ Eθ).

If this is not the case (for example, Eθ > EZ):

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 9
Structures

(2.30)

2.1.3. Temperature-Dependent Coefficient of Thermal Expansion

Considering a typical component, the thermal strain from Equation 2.3 (p. 6) is:
(2.31)

where:

α se(T) = temperature-dependent secant coefficient of thermal expansion (SCTE)

α se(T) is input in one of three ways:

1. Input α se(T) directly (input as ALPX, ALPY, or ALPZ on MP command)

2. Computed using Equation 2.34 (p. 10) from α in(T), the instantaneous coefficients of thermal
expansion (input as CTEX, CTEY, or CTEZ on MP command)

3. Computed using Equation 2.32 (p. 10) from εith(T), the input thermal strains (input as THSX,
THSY, or THSZ on MP command)

α se(T) is computed from εith(T) by rearranging Equation 2.31 (p. 10):

(2.32)

Equation 2.32 (p. 10) assumes that when T = Tref, εith = 0. If this is not the case, the εith data is shifted
se
automatically by a constant so that it is true. α at Tref is calculated based on the slopes from the
adjacent user-defined data points. Hence, if the slopes of εith above and below Tref are not identical,
se
a step change in α at Tref will be computed.

εth(T) (thermal strain) is related to α in(T) by:

(2.33)

Combining this with Equation 2.32 (p. 10),

(2.34)

No adjustment is needed for α in(T) as α se(T) is defined to be α in(T) when T = Tref.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
10 Published: 2023-10-23
 Structural Fundamentals

As seen above, α se(T) is dependent on what was used for Tref. If α se(T) was defined using Tref as one
value but then the thermal strain was zero at another value, an adjustment needs to be made (using
the MPAMOD command). Consider:

(2.35)

(2.36)

Equation 2.35 (p. 11) and Equation 2.36 (p. 11) represent the thermal strain at a temperature T for
two different starting points, To and Tref. Now let To be the temperature about which the data has
been generated (definition temperature), and Tref be the temperature at which all strains are zero
(reference temperature). Thus, is the supplied data, and is what is needed as program input.

The right-hand side of Equation 2.35 (p. 11) may be expanded as:

(2.37)

also,

(2.38)

or

(2.39)

Combining Equation 2.35 (p. 11) through Equation 2.38 (p. 11),

(2.40)

Thus, Equation 2.40 (p. 11) must be accounted for when making an adjustment for the definition
temperature being different from the strain-free temperature. This adjustment may be made (using
the MPAMOD command).

Note that:

Equation 2.40 (p. 11) is nonlinear. Segments that were straight before may be no longer straight.
Hence, extra temperatures may need to be specified initially (using the MPTEMP command).
If Tref = To, Equation 2.40 (p. 11) is trivial.
If T = Tref, Equation 2.40 (p. 11) is undefined.

se
The values of T as used here are the temperatures used to define α (input via the MPTEMP com-
se
mand). Thus, when using the α adjustment procedure, avoid defining a T value to be the same as

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 11
Structures

se
T = Tref (to a tolerance of one degree). If a T value is the same as Tref, the value of α remains un-
changed.

2.2. Derivation of Structural Matrices

The principle of virtual work states that a virtual (very small) change of the internal strain energy must
be offset by an identical change in external work due to the applied loads, or:
(2.41)

where:

U = strain energy (internal work) = U1 + U2

V = external work = V1 + V2 + V3
δ = virtual operator

The virtual strain energy is:

(2.42)

where:

{ε} = strain vector

{σ} = stress vector
vol = volume of element

Continuing the derivation assuming linear materials and geometry, Equation 2.41 (p. 12) and Equa-
tion 2.42 (p. 12) are combined to give:
(2.43)

The strains may be related to the nodal displacements by:

(2.44)

where:

[B] = strain-displacement matrix, based on the element shape functions

{u} = nodal displacement vector

It will be assumed that all effects are in the global Cartesian system. Combining Equation 2.44 (p. 12)
with Equation 2.43 (p. 12), and noting that {u} does not vary over the volume:

(2.45)

Another form of virtual strain energy is when a surface moves against a distributed resistance, as in a
foundation stiffness. This may be written as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
12 Published: 2023-10-23
 Derivation of Structural Matrices

(2.46)

where:

{wn} = motion normal to the surface

{σ} = stress (or pressure) carried by the surface
areaf = area of the distributed resistance

Both {wn} and {σ} will usually have only one nonzero component. The point-wise normal displacement
is related to the nodal displacements by:
(2.47)

where:

[Nn] = matrix of shape functions for normal motions at the surface

The stress, {σ}, is

(2.48)

where:

k = the foundation stiffness in units of force per length per unit area

Combining Equation 2.46 (p. 13) through Equation 2.48 (p. 13), and assuming that k is constant over
the area,
(2.49)

Next, the external virtual work will be considered. The inertial effects will be studied first:

(2.50)

where:

{w} = vector of displacements of a general point

{Fa} = acceleration (D'Alembert) force vector

According to Newton's second law:

(2.51)

where:

ρ = density (input as DENS on MP command)

t = time

The displacements within the element are related to the nodal displacements by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 13
Structures

(2.52)

where [N] = matrix of shape functions. Combining Equation 2.50 (p. 13), Equation 2.51 (p. 13), and
Equation 2.52 (p. 14), and assuming that ρ is constant over the volume,
(2.53)

The pressure force vector formulation starts with:

(2.54)

where:

{P} = the applied pressure vector (normally contains only one nonzero component)
area p = area over which pressure acts

Combining equations Equation 2.52 (p. 14) and Equation 2.54 (p. 14),
(2.55)

Unless otherwise noted, pressures are applied to the outside surface of each element and are normal
to curved surfaces, if applicable.

Nodal forces applied to the element can be accounted for by:

(2.56)

where:

Finally, Equation 2.41 (p. 12), Equation 2.45 (p. 12), Equation 2.49 (p. 13), Equation 2.53 (p. 14), Equa-
tion 2.55 (p. 14), and Equation 2.56 (p. 14) may be combined to give:

(2.57)

Noting that the {δu}T vector is a set of arbitrary virtual displacements common in all of the above terms,
the condition required to satisfy Equation 2.57 (p. 14) reduces to:
(2.58)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
14 Published: 2023-10-23
 Structural Strain and Stress Evaluations

Equation 2.58 (p. 14) represents the equilibrium equation on a one element basis.

The above matrices and load vectors were developed as "consistent". Other formulations are possible.
For example, if only diagonal terms for the mass matrix are requested (LUMPM,ON), the matrix is called
"lumped" (see Lumped Matrices (p. 422)). For most lumped mass matrices, the rotational degrees of
freedom (DOFs) are removed. If the rotational DOFs are requested to be removed (KEYOPT commands
with certain elements), the matrix or load vector is called "reduced". Thus, use of the reduced pressure
load vector does not generate moments as part of the pressure load vector. Use of the consistent
pressure load vector can cause erroneous internal moments in a structure. An example of this would
be a thin circular cylinder under internal pressure modelled with irregular shaped shell elements. As
suggested by Figure 2.3: Effects of Consistent Pressure Loading (p. 15), the consistent pressure loading
generates an erroneous moment for two adjacent elements of dissimilar size.

Figure 2.3: Effects of Consistent Pressure Loading

2.3. Structural Strain and Stress Evaluations

2.3.1. Integration Point Strains and Stresses

The element integration point strains and stresses are computed by combining equations Equa-
tion 2.1 (p. 5) and Equation 2.44 (p. 12) to get:
(2.59)
(2.60)

where:

{εel} = strains that cause stresses (output as EPEL)

[B] = strain-displacement matrix evaluated at integration point

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 15
Structures

{u} = nodal displacement vector

{εth} = thermal strain vector
{σ} = stress vector (output as S)
[D] = elasticity matrix

Nodal and centroidal stresses are computed from the integration point stresses as described in Nodal
and Centroidal Data Evaluation (p. 442).

2.3.2. Surface Stresses

Surface stress output may be requested on "free" faces of 2D and 3D elements. "Free" means not
connected to other elements as well as not having any imposed displacements or nodal forces normal
to the surface. The following steps are executed at each surface Gauss point to evaluate the surface
stresses. The integration points used are the same as for an applied pressure to that surface.

1. Compute the in-plane strains of the surface at an integration point using:

(2.61)

Hence, , and are known. The prime (') represents the surface coordinate system, with z
being normal to the surface.

2. A each point, set:

(2.62)
(2.63)
(2.64)

where P is the applied pressure. Equation 2.63 (p. 16) and Equation 2.64 (p. 16) are valid, as the
surface for which stresses are computed is presumed to be a free surface.

3. At each point, use the six material property equations represented by:
(2.65)

to compute the remaining strain and stress components ( , , , , and ).

4. Repeat and average the results across all integration points.

2.3.3. Shell Element Output

For elastic shell elements, the forces and moments per unit length (using shell nomenclature) are
computed as:

(2.66)

(2.67)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
16 Published: 2023-10-23
 Structural Strain and Stress Evaluations

(2.68)

(2.69)

(2.70)

(2.71)

(2.72)

(2.73)

where:

Tx, Ty, Txy = in-plane forces per unit length (output as TX, TY, and TXY)
Mx, My, Mxy = bending moments per unit length (output as MX, MY, and MXY)
Nx, Ny = transverse shear forces per unit length (output as NX and NY)
t = thickness at midpoint of element, computed normal to center plane
σx, etc. = direct stress (output as SX, etc.)
σxy, etc. = shear stress (output as SXY, etc.)

For shell elements with linearly elastic material, Equation 2.66 (p. 16) to Equation 2.73 (p. 17) reduce
to:
(2.74)

(2.75)

(2.76)

(2.77)

(2.78)

(2.79)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 17
Structures

(2.80)

(2.81)

For shell elements with nonlinear materials, Equation 2.66 (p. 16) to Equation 2.73 (p. 17) are numer-
ically integrated.

It should be noted that the shell nomenclature and the nodal moment conventions are in apparent
conflict with each other. For example, a cantilever beam located along the x axis and consisting of
shell elements in the x-y plane that deforms in the z direction under a pure bending load with coupled
nodes at the free end, has the following relationship:
(2.82)

where:

b = width of beam
FMY = nodal moment applied to the free end (input as VALUE on F command with Lab = MY
(not MX))

The shape functions of the shell element result in constant transverse strains and stresses through
the thickness. Some shell elements adjust these values so that they will peak at the mid-surface with
3/2 of the constant value and be zero at both surfaces, as noted in the element discussions in Element
Library (p. 445).

The through-thickness stress (σz) is set equal to the negative of the applied pressure at the surfaces
of the shell elements, and linearly interpolated in between.

2.4. Combined Stresses and Strains

When a model has only one functional direction of strains and stress, comparison with an allowable
value is straightforward. But when there is more than one component, the components are normally
combined into one number to allow a comparison with an allowable. This section discusses different
ways of doing that combination, representing different materials and/or technologies.

2.4.1. Combined Strains

The principal strains are calculated from the strain components by the cubic equation:

(2.83)

where is the principal strain (three values).

The three principal strains are labeled , , and (output as 1, 2, and 3 with strain items such as
EPEL). The principal strains are ordered so that is the most positive and is the most negative.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
18 Published: 2023-10-23
 Combined Stresses and Strains

The strain intensity (output as INT with strain items such as EPEL) is the largest of the absolute
values of , , or . That is:
(2.84)

The von Mises or equivalent strain (output as EQV with strain items such as EPEL) is calculated as:

(2.85)

where is the effective Poisson's ratio (defined via AVPRIN). If is not defined, the default values
are as follows:

Calculation Default
Elastic Equivalent Strain Defined elastic property
Thermal Equivalent Strain 0.0
Plastic Equivalent Strain 0.5
Creep Equivalent Strain 0.5
Hyperelastic Materials 0.5
User Materials 0.5
Beam, Link, Pipe, Discrete Element 0.0
types
Cyclic Symmetry 0.0
Load Case Operations 0.0

2.4.2. Combined Stresses

The principal stresses ( ) are calculated from the stress components by the cubic equation:

(2.86)

where is the principal stress (three values).

The three principal stresses are labeled , , and (output quantities S1, S2, and S3). The principal
stresses are ordered so that is the most positive (tensile) and is the most negative (compressive).

The stress intensity (output as SINT) is the largest of the absolute values of , , or .
That is:
(2.87)

The von Mises or equivalent stress (output as SEQV) is calculated as:

(2.88)

or
(2.89)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 19
Structures

When (input as PRXY or NUXY via MP), the equivalent stress is related to the equivalent strain
through
(2.90)

where is the Young's modulus (input as EX via MP).

2.4.3. Failure Criteria

Use failure criteria to assess the possibility of failure of a material. Doing so allows the consideration
of orthotropic materials, which might be much weaker in one direction than another. Failure criteria
are available in POST1 for all plane, shell, and solid structural elements (using the TB or FC family of
commands).

Possible failure of a material can be evaluated by up to 20 different criteria, of which 11 are predefined.
They are evaluated at the top and bottom (or middle) of each layer at each of the in-plane integration
points.

A failure criterion value ≥ 1 indicates failure.

The following topics related to the predefined failure criteria are available:
[Link]. Maximum Strain Failure Criteria
[Link]. Maximum Stress Failure Criteria
[Link].Tsai-Wu Failure Criteria
[Link]. Physical Failure Criteria

[Link]. Maximum Strain Failure Criteria

(2.91)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
20 Published: 2023-10-23
 Combined Stresses and Strains

[Link]. Maximum Stress Failure Criteria

(2.92)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 21
Structures

[Link]. Tsai-Wu Failure Criteria

If the criterion used is the "strength index" (TWSI – see PRESOL,FAIL):
(2.93)

and if the criterion used is the inverse of the "strength ratio" (TWSR – see PRESOL,FAIL):
(2.94)

where:

ξ3 = value of Tsai-Wu failure criterion

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
22 Published: 2023-10-23
 Combined Stresses and Strains

Cxy, Cyz, Cxz = x-y, y-z, x-z, respectively, coupling coefficient for Tsai-Wu theory

For consistency, the program uses the inverse of "strength ratio" as the TWSR failure criterion.
Similar to other failure criteria, a TWSR criterion indicates failure if its value is ≥ 1.

The Tsai-Wu failure criteria used here are 3D versions of the failure criterion reported in Tsai and
Hahn([192] (p. 935)) for the "strength index" and in Tsai([94] (p. 930)) for the "strength ratio". Apparent
differences are:

1. The program input uses negative values for compression limits, whereas Tsai uses positive values
for all limits.

2. The program uses Cxy instead of the used by Tsai and Hahn with Cxy being twice the value
of .

[Link]. Physical Failure Criteria

The physical failure criteria are specially formulated to account for different damage mechanisms
(fiber and matrix failure) in fiber-reinforced composite materials.

Predefined physical failure criteria include the following:

[Link].1. Hashin Fiber Failure Criterion

(2.95)

[Link].2. Hashin Matrix Failure Criterion

(2.96)

[Link].3. Puck Fiber Failure Criterion ([398] (p. 947))

(2.97)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 23
Structures

[Link].4. Puck Matrix Failure Criterion ([398] (p. 947))

(2.98)

where:

= stresses on an action plane parallel to the fibers

= four inclination parameters

[Link].5. LaRc03 Fiber Failure Criterion ([399] (p. 947))

(2.99)

where:

= fracture toughness ratio between mode I and mode II

= 2D stresses on the fiber misalignment (kinking) plane
= longitudinal friction coefficient

= in-plane shear strength

= tensile transverse strength

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
24 Published: 2023-10-23
 Combined Stresses and Strains

[Link].6. LaRc03 Matrix Failure Criterion ([399] (p. 947))

(2.100)

where:

= fracture toughness ratio between mode I and mode II

= effective transverse and longitudinal stresses on an action plane parallel to fibers
= effective stresses on an action plane parallel to the misaligned fibers

= transverse shear strength

= initial fracture angle under pure transverse compression

[Link].7. LaRc04 Fiber Failure Criterion ([400] (p. 947))

(2.101)

= 3D stresses on the fiber misalignment (kinking) plane

= derived from orthotropic material constants

[Link].8. LaRc04 Matrix Failure Criterion ([400] (p. 947))

(2.102)

where:

= transverse friction coefficient

= effective normal, transverse, and longitudinal stresses on an action plane
parallel to the fibers
= effective stresses on the action plane parallel to the misaligned fibers

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 25
Structures

2.5. Perfectly Matched Layers (PML) in Elastic Media

Perfectly matched layers are artificial anisotropic materials that absorb all incoming elastic waves without
any reflections, except for the gazing wave that parallels the PML interface in the propagation direction.
PMLs are constructed to propagate elastic waves in homogeneous isotropic elastic media in harmonic
response analyses (or in an infinite static fields in static analyses).

Consider two Cartesian coordinate systems: a global system with respect to an orthogonal basis
, and a local system with respect to another orthogonal basis . The displacements
of such an elastic PML medium in the local coordinate system are governed by the following equations:

(2.103)

(2.104)

(2.105)

where:

= components of the stress tensor in the local Cartesian coordinate system

= components of the infinitesimal strain tensor in the local Cartesian coordinate system

= components of the material stiffness tensor in the local Cartesian coordinate system
= nonzero complex-valued coordinate stretching functions in the direction of the local Cartesian
coordinate system
= density of the elastic medium

On multiplying Equation 2.103 (p. 26) with the and noting that is a function of only,
Equation 2.103 (p. 26) through Equation 2.105 (p. 26) are rewritten in matrix notation:
(2.106)
(2.107)

(2.108)

where:

The tensors and vectors in the local coordinate system are transformed into tensors and vectors in the
global coordinate system using the rotation-of-basis transformation matrix with the component
, such that . The governing equations of the displacements of the
elastic PML medium in the global coordinate system are written as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
26 Published: 2023-10-23
 Perfectly Matched Layers (PML) in Elastic Media

(2.109)
(2.110)
(2.111)

By using the Galerkin method and setting the displacements to zero (input as UX, UY and UZ on the D
command) on the backed boundary, the "weak" formulation of an elastic wave in the elastic PML medium
is expressed as:
(2.112)

where:
(2.113)

The PML material is defined by PLANE182, PLANE183, SOLID185, SOLID186, and SOLID187 elements
with KEYOPT(15) = 1. Because the PML material is constructed in Cartesian coordinates, the edges of
the 2D or 3D PML region must be aligned with the axes of the global Cartesian coordinate system. More
than one 1D PML region may exist in a finite element model. The PML element coordinate system
(PSYS) uniquely identifies each PML region. A parabolic attenuation distribution minimizes numerical
reflections in the PML.

Ansys, Inc. recommends using at least three PML element layers to obtain satisfactory accuracy. Some
buffer elements between the PML region and objects should be utilized. Because a PML acts as an in-
finite open domain, any boundary conditions and material properties must be carried over into the
PML region. The displacements must be set to zero (input as UX, UY and UZ on the D command) on
the exterior surface of the PML, excluding Neumann boundaries (symmetric planes in the model). Any
excitation sources such as body force are prohibited in the PML region.

For more information, see Perfectly Matched Layers (PML) in the Acoustic Analysis Guide.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 27
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
28 Published: 2023-10-23
Chapter 3: Structures with Geometric Nonlinearities
This chapter discusses the various geometrically nonlinear options available, including large strain, large
deflection, stress stiffening, pressure load stiffness, and spin softening. Only elements with displacement
degrees of freedom are applicable. Not included in this section are multi-status elements (such as
COMBIN40, discussed in Element Library (p. 445)) and the eigenvalue buckling capability (discussed in
Buckling Analysis (p. 826)).

The following topics are available:

3.1. Understanding Geometric Nonlinearities
3.2. Large Strain
3.3. Large Rotation
3.4. Stress Stiffening
3.5. Spin Softening
3.6. General Element Formulations
3.7. Constraints and Lagrange Multiplier Method
3.8. Steady-State Rolling

3.1. Understanding Geometric Nonlinearities

Geometric nonlinearities refer to the nonlinearities in the structure or component due to the changing
geometry as it deflects. That is, the stiffness [K] is a function of the displacements {u}. The stiffness
changes because the shape changes and/or the material rotates. The program can account for four
types of geometric nonlinearities:

1. Large strain (p. 30) assumes that the strains are no longer infinitesimal (they are finite). Shape
changes (e.g. area, thickness, etc.) are also accounted for. Deflections and rotations may be arbitrarily
large.

2. Large rotation (p. 35) assumes that the rotations are large but the mechanical strains (those that
cause stresses) are evaluated using linearized expressions. The structure is assumed not to change
shape except for rigid body motions. The elements of this class refer to the original configuration.

3. Stress stiffening (p. 40) assumes that both strains and rotations are small. A 1st order approximation
to the rotations is used to capture some nonlinear rotation effects.

4. Spin softening (p. 46) also assumes that both strains and rotations are small. This option accounts
for the coupling between the transverse vibrational motion and the centrifugal force due to an an-
gular velocity.

All elements support the spin softening capability, while only some of the elements support the other
options. For more information, see the Element Reference.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 29
Structures with Geometric Nonlinearities

3.2. Large Strain

When the strains in a material exceed more than a few percent, the changing geometry due to this
deformation can no longer be neglected. Analyses which include this effect are called large strain, or
finite-strain, analyses. A large strain analysis is performed in a static (ANTYPE,STATIC) or transient (AN-
TYPE,TRANS) analysis while flagging large deformations (NLGEOM,ON) when the appropriate element
type(s) is used.

The remainder of this section addresses the large strain formulation for elastic-plastic elements. These
elements use a hypoelastic formulation so that they are restricted to small elastic strains (but allow for
arbitrarily large plastic strains). Hyperelasticity (p. 112) addresses the large strain formulation for hypere-
lastic elements, which allow arbitrarily large elastic strains.

3.2.1. Theory
The theory of large strain computations can be addressed by defining a few basic physical quantities
(motion and deformation) and the corresponding mathematical relationship. The applied loads acting
on a body make it move from one position to another. This motion can be defined by studying a
position vector in the "deformed" and "undeformed" configuration. Say the position vectors in the
"deformed" and "undeformed" state are represented by {x} and {X} respectively, then the motion
(displacement) vector {u} is computed by (see Figure 3.1: Position Vectors and Motion of a Deforming
Body (p. 30)):
(3.1)

Figure 3.1: Position Vectors and Motion of a Deforming Body

The deformation gradient is defined as:

(3.2)

which can be written in terms of the displacement of the point via Equation 3.1 (p. 30) as:
(3.3)

where:

[I] = identity matrix

The information contained in the deformation gradient [F] includes the volume change, the rotation
and the shape change of the deforming body. The volume change at a point is
(3.4)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
30 Published: 2023-10-23
 Large Strain

where:

Vo = original volume
V = current volume
det [ ] = determinant of the matrix

The deformation gradient can be separated into a rotation and a shape change using the right polar
decomposition theorem:
(3.5)

where:

[R] = rotation matrix ([R]T[R] = [I])

[U] = right stretch (shape change) matrix

Once the stretch matrix is known, a logarithmic or Hencky strain measure is defined as:
(3.6)

([ε] is in tensor (matrix) form here, as opposed to the usual vector form {ε}). Since [U] is a 2nd order
tensor (matrix), Equation 3.6 (p. 31) is determined through the spectral decomposition of [U]:

(3.7)

where:

λi = eigenvalues of [U] (principal stretches)

{ei} = eigenvectors of [U] (principal directions)

The polar decomposition theorem (Equation 3.5 (p. 31)) extracts a rotation [R] that represents the
average rotation of the material at a point. Material lines initially orthogonal will not, in general, be
orthogonal after deformation (because of shearing), see Figure 3.2: Polar Decomposition of a Shearing
Deformation (p. 31). The polar decomposition of this deformation, however, will indicate that they
will remain orthogonal (lines x-y' in Figure 3.2: Polar Decomposition of a Shearing Deformation (p. 31)).
For this reason, non-isotropic behavior (e.g. orthotropic elasticity or kinematic hardening plasticity)
should be used with care with large strains, especially if large shearing deformation occurs.

Figure 3.2: Polar Decomposition of a Shearing Deformation

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 31
Structures with Geometric Nonlinearities

3.2.2. Implementation
Computationally, the evaluation of Equation 3.6 (p. 31) is performed by one of two methods using
the incremental approximation (since, in an elastic-plastic analysis, we are using an incremental
solution procedure):
(3.8)

with
(3.9)

where [ΔUn] is the increment of the stretch matrix computed from the incremental deformation
gradient:
(3.10)

where [ΔFn] is:

(3.11)

[Fn] is the deformation gradient at the current time step and [Fn-1] is at the previous time step.

(Hughes([157] (p. 933))) uses the approximate 2nd order accurate calculation for evaluating Equa-
tion 3.9 (p. 32):
(3.12)

where [R1/2] is the rotation matrix computed from the polar decomposition of the deformation
gradient evaluated at the midpoint configuration:
(3.13)

where [F1/2] is (using Equation 3.3 (p. 30)):

(3.14)

and the midpoint displacement is:

(3.15)

{un} is the current displacement and {un-1} is the displacement at the previous time step. [Δεn] is the
"rotation-neutralized" strain increment over the time step. The strain increment is also computed
from the midpoint configuration:
(3.16)

{Δun} is the displacement increment over the time step and [B1/2] is the strain-displacement relationship
evaluated at the midpoint geometry:
(3.17)

This method is an excellent approximation to the logarithmic strain if the strain steps are less than
~10%. This method is used by the standard 2D and 3D solid and shell elements.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
32 Published: 2023-10-23
 Large Strain

The computed strain increment [Δεn] (or equivalently {Δεn}) can then be added to the previous strain
{εn-1} to obtain the current total Hencky strain:
(3.18)

This strain can then be used in the stress updating procedures, see Rate-Independent Plasticity (p. 67)
and Rate-Dependent Plasticity (Including Creep and Viscoplasticity) (p. 96) for discussions of the rate-
independent and rate-dependent procedures respectively.

3.2.3. Definition of Thermal Strains

According to Callen([244] (p. 938)), the coefficient of thermal expansion is defined as the fractional
increase in the length per unit increase in the temperature. Mathematically,
(3.19)

where:

α = coefficient of thermal expansion

= current length
T = temperature

Rearranging Equation 3.19 (p. 33) gives:

(3.20)

On the other hand, the logarithmic strain is defined as:

(3.21)

where:

= logarithmic strain
o = initial length

Differential of Equation 3.21 (p. 33) yields:

(3.22)

Comparison of Equation 3.20 (p. 33) and Equation 3.22 (p. 33) gives:
(3.23)

Integration of Equation 3.23 (p. 33) yields:

(3.24)

where:

= initial (reference) strain at temperature To

To = reference temperature

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 33
Structures with Geometric Nonlinearities

In the absence of initial strain ( ), then Equation 3.24 (p. 33) reduces to:
(3.25)

The thermal strain corresponds to the logarithmic strain. As an example problem, consider a line
element of a material with a constant coefficient of thermal expansion α. If the length of the line is
o at temperature To, then the length after the temperature increases to T is:
(3.26)

Now if one interpreted the thermal strain as the engineering (or nominal) strain, then the final length
would be different.
(3.27)

where:

εe = engineering strain

The final length is then:

(3.28)

However, the difference should be very small as long as:

(3.29)

because
(3.30)

3.2.4. Element Formulation

The element matrices and load vectors are derived using an updated Lagrangian formulation. This
produces equations of the form:
(3.31)

where the tangent matrix has the form:

(3.32)

[Ki] is the usual stiffness matrix:

(3.33)

[Bi] is the strain-displacement matrix in terms of the current geometry {Xn} and [Di] is the current
stress-strain matrix.

[Si] is the stress stiffness (or geometric stiffness) contribution, written symbolically as:

(3.34)

where [Gi] is a matrix of shape function derivatives and [τi] is a matrix of the current Cauchy (true)
stresses {σi} in the global Cartesian system. The Newton-Raphson restoring force is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
34 Published: 2023-10-23
 Large Rotation

(3.35)

Some of the plane stress and shell elements account for the thickness changes due to the out-of-
plane strain εz using an approach similar to that of Hughes and Carnoy ([158] (p. 933)). Shells, however,
do not update their reference plane (as might be required in a large strain out-of-plane bending de-
formation); the thickness change is assumed to be constant through the thickness. General element
formulations using finite deformation are developed in General Element Formulations (p. 49) and
apply to current-technology elements only.

3.2.5. Applicable Input

NLGEOM,ON activates large strain computations in those elements which support it. NLGEOM,ON
also activates the stress-stiffening contribution to the tangent matrix.

3.2.6. Applicable Output

For elements which have large strain capability, stresses (output as S) are true (Cauchy) stresses in
the rotated element coordinate system (the element coordinate system follows the material as it ro-
tates). Strains (output as EPEL, EPPL, etc.) are the logarithmic or Hencky strains, also in the rotated
element coordinate system.

An exception is for the hyperelastic elements. For these elements, stress and strain components
maintain their original orientations and some of these elements use other strain measures.

3.3. Large Rotation

If the rotations are large but the mechanical strains (those that cause stresses) are small, then a large-
rotation procedure can be used.

A large-rotation analysis is performed in a static (ANTYPE,STATIC) or transient (ANTYPE,TRANS) analysis

while flagging large deformations (NLGEOM,ON) when the appropriate element type is used.

All large-strain elements also support this capability, as both options account for the large rotations
and for small strains, the logarithmic strain measure and the engineering strain measure coincide.

3.3.1. Theory
Large Strain (p. 30) presented the theory for general motion of a material point. Large-rotation theory
follows a similar development, except that the logarithmic strain measure (Equation 3.6 (p. 31)) is
replaced by the Biot, or small (engineering) strain measure:
(3.36)

where:

[U] = stretch matrix

[I] = 3 x 3 identity matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 35
Structures with Geometric Nonlinearities

3.3.2. Implementation
A corotational (or convected coordinate) approach is used in solving large-rotation/small-strain
problems (Rankin and Brogan([67] (p. 928))). "Corotational" may be thought of as "rotated with". The
nonlinearities are contained in the strain-displacement relationship which for this algorithm takes on
the special form:
(3.37)

where:

[Bv] = usual small strain-displacement relationship in the original (virgin) element coordinate
system
[Tn] = orthogonal transformation relating the original element coordinates to the convected
(or rotated) element coordinates

The convected element coordinate frame differs from the original element coordinate frame by the
amount of rigid body rotation. Hence [Tn] is computed by separating the rigid body rotation from
the total deformation {un} using the polar decomposition theorem, Equation 3.5 (p. 31). From Equa-
tion 3.37 (p. 36), the element tangent stiffness matrix has the form:
(3.38)

and the element restoring force is:

(3.39)

where the elastic strain is computed from:

(3.40)

is the element deformation which causes straining as described in a subsequent subsection.

The large-rotation process can be summarized as a three step process for each element:

1. Determine the updated transformation matrix [Tn] for the element.

2. Extract the deformational displacement from the total element displacement {un} for com-
puting the stresses as well as the restoring force .

3. After the rotational increments in {Δu} are computed, update the node rotations appropriately.
All three steps require the concept of a rotational pseudovector in order to be efficiently imple-
mented (Rankin and Brogan([67] (p. 928)), Argyris([68] (p. 928))).

3.3.3. Element Transformation

The updated transformation matrix [Tn] relates the current element coordinate system to the global
Cartesian coordinate system as shown in Figure 3.3: Element Transformation Definitions (p. 37).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
36 Published: 2023-10-23
 Large Rotation

Figure 3.3: Element Transformation Definitions

[Tn] can be computed directly or the rotation of the element coordinate system [Rn] can be computed
and related to [Tn] by
(3.41)

where [Tv] is the original transformation matrix. The determination of [Tn] is unique to the type of
element involved, whether it is a solid element, shell element, beam element, or spar element.

Solid Elements. The rotation matrix [Rn] for these elements is extracted from the displacement
field using the deformation gradient coupled with the polar decomposition theorem (see Mal-
vern([88] (p. 929))).
Shell Elements. The updated normal direction (element z direction) is computed directly from the
updated coordinates. The computation of the element normal is given in Element Library (p. 445)
for each particular shell element. The extraction procedure outlined for solid elements is used
coupled with the information on the normal direction to compute the rotation matrix [Rn].
Beam Elements. The nodal rotation increments from {Δu} are averaged to determine the average
rotation of the element. The updated average element rotation and then the rotation matrix [Rn]
is computed using Rankin and Brogan([67] (p. 928)). In special cases where the average rotation
of the element computed in the above way differs significantly from the average rotation of the
element computed from nodal translations, the quality of the results will be degraded.
Link Elements. The updated transformation [Tn] is computed directly from the updated coordinates.
Generalized Mass Element (MASS21). The nodal rotation increment from {Δu} is used to update
the element rotation which then yields the rotation matrix [Rn].

3.3.4. Deformational Displacements

The displacement field can be decomposed into a rigid body translation, a rigid body rotation, and
a component which causes strains:
(3.42)

where:

{ur} = rigid body motion

{ud} = deformational displacements which cause strains

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 37
Structures with Geometric Nonlinearities

{ud} contains both translational as well as rotational degrees of freedom.

The translational component of the deformational displacement can be extracted from the displacement
field by
(3.43)

where:

= translational component of the deformational displacement

[Rn] = current element rotation matrix
{xv} = original element coordinates in the global coordinate system
{u} = element displacement vector in global coordinates

{ud} is in the global coordinate system.

For elements with rotational degrees of freedom, the rotational components of the deformational
displacement must be computed. The rotational components are extracted by essentially "subtracting"
the nodal rotations {u} from the element rotation given by {ur}. In terms of the pseudovectors this
operation is performed as follows for each node:

1. Compute a transformation matrix from the nodal pseudovector {θn} yielding [Tn].

2. Compute the relative rotation [Td] between [Rn] and [Tn]:

(3.44)

This relative rotation contains the rotational deformations of that node as shown in Figure 3.4: Defin-
ition of Deformational Rotations (p. 38).

3. Extract the nodal rotational deformations {ud} from [Td].

Because of the definition of the pseudovector, the deformational rotations extracted in step 3 are
limited to less than 30°, since 2sin(θ /2) no longer approximates θ itself above 30°. This limitation only
applies to the rotational distortion (that is, bending) within a single element.

Figure 3.4: Definition of Deformational Rotations

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
38 Published: 2023-10-23
 Large Rotation

3.3.5. Updating Rotations

Once the transformation [T] and deformational displacements {ud} are determined, the element
matrices Equation 3.38 (p. 36) and restoring force Equation 3.39 (p. 36) can be determined. The
solution of the system of equations yields a displacement increment {Δu}. The nodal rotations at the
element level are updated with the rotational components of {Δu}. The global rotations (in the output
and on the results file) are not updated with the pseudovector approach, but are simply added to
the previous rotation in {un-1}.

3.3.6. Applicable Input

The large-rotation computations in those elements which support it are activated by the large-deform-
ation key (NLGEOM,ON). Stress stiffening is also be included and contributes to the tangent stiffness
matrix (which may be required for structures weak in bending resistance).

3.3.7. Applicable Output

Stresses (output as S) are engineering stresses in the rotated element coordinate system (the element
coordinate system follows the material as it rotates). Strains (output as EPEL, EPPL, etc.) are engineering
strains, also in the rotated element coordinate system. This applies to element types that do not have
large-strain capability. For element types that have large-strain capability, see Large Strain (p. 30).

3.3.8. Consistent Tangent Stiffness Matrix and Finite Rotation

It has been found in many situations that the use of consistent tangent stiffness in a nonlinear ana-
lysis can speed up the rate of convergence greatly. It normally results in a quadratic rate of conver-
gence. A consistent tangent stiffness matrix is derived from the discretized finite element equilibrium
equations without the introduction of various approximations. The terminology of finite rotation in
the context of geometrical nonlinearity implies that rotations can be arbitrarily large and can be up-
dated accurately. A consistent tangent stiffness accounting for finite rotations derived by Nour-Omid
and Rankin([177] (p. 934)) for beam/shell elements is used. The technology of consistent tangent
matrix and finite rotation makes the buckling and postbuckling analysis a relatively easy task. The
theory of finite rotation representation and update has been described in Large Rotation (p. 35) using
a pseudovector representation. The following will outline the derivations of a consistent tangent
stiffness matrix used for the corotational approach.

The nonlinear static finite element equations solved can be characterized by at the element level by:

(3.45)

where:

N = number of total elements

= element internal force vector in the element coordinate system, generally see
Equation 3.46 (p. 40)
[Tn]T = transform matrix transferring the local internal force vector into the global coordinate
system

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 39
Structures with Geometric Nonlinearities

= applied load vector at the element level in the global coordinate system
(3.46)

Hereafter, we shall focus on the derivation of the consistent tangent matrix at the element level
without introducing an approximation. The consistent tangent matrix is obtained by differentiating
Equation 3.45 (p. 39) with respect to displacement variables {ue}:

(3.47)

It can be seen that Part I is the main tangent matrix Equation 3.38 (p. 36) and Part II is the stress
stiffening matrix (Equation 3.34 (p. 34), Equation 3.61 (p. 44) or Equation 3.64 (p. 44)). Part III is an-
other part of the stress stiffening matrix (see Nour-Omid and Rankin([177] (p. 934))) traditionally neg-
lected in the past. However, many numerical experiments have shown that Part III of is essential
to the faster rate of convergence. In some cases, Part III of is unsymmetric; when this occurs, a
procedure of symmetrizing is invoked.

As Part III of the consistent tangent matrix utilizes the internal force vector to form the matrix,
it is required that the internal vector not be so large as to dominate the main tangent matrix
(Part I). This can normally be guaranteed if the realistic material and geometry are used, that is, the
element is not used as a rigid link and the actual thicknesses are input.

It is also noted that the consistent tangent matrix Equation 3.47 (p. 40) is very suitable for use with
the arc-length solution method.

3.4. Stress Stiffening

3.4.1. Overview and Usage

Stress stiffening (also called geometric stiffening, incremental stiffening, initial stress stiffening, or
differential stiffening by other authors) is the stiffening (or weakening) of a structure due to its stress
state. This stiffening effect normally needs to be considered for thin structures with bending stiffness
very small compared to axial stiffness, such as cables, thin beams, and shells and couples the in-plane
and transverse displacements. This effect also augments the regular nonlinear stiffness matrix produced
by large-strain or large-deflection effects (NLGEOM,ON). The effect of stress stiffening is accounted
for by generating and then using an additional stiffness matrix, hereinafter called the "stress stiffness
matrix". The stress stiffness matrix is added to the regular stiffness matrix in order to give the total
stiffness. Stress stiffening may be used for static (ANTYPE,STATIC) or transient (ANTYPE,TRANS) ana-
lyses. Working with the stress stiffness matrix is the pressure load stiffness, discussed in Pressure Load
Stiffness (p. 45).

The stress stiffness matrix is computed based on the stress state of the previous equilibrium iteration.
Thus, to generate a valid stress-stiffened problem, at least two iterations are normally required, with

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
40 Published: 2023-10-23
 Stress Stiffening

the first iteration being used to determine the stress state that will be used to generate the stress
stiffness matrix of the second iteration. If this additional stiffness affects the stresses, more iterations
need to be done to obtain a converged solution.

In some linear analyses, the static (or initial) stress state may be large enough that the additional
stiffness effects must be included for accuracy. Modal (ANTYPE,MODAL) and substructure (ANTYPE,SUB-
STR) analyses are linear analyses for which the prestressing effects can be requested to be included
(PSTRES,ON command). Note that in these cases the stress stiffness matrix is constant, so that the
stresses computed in the analysis are assumed small compared to the prestress stress.

If membrane stresses should become compressive rather than tensile, then terms in the stress stiffness
matrix may "cancel" the positive terms in the regular stiffness matrix and therefore yield a nonpositive-
definite total stiffness matrix, which indicates the onset of buckling. If this happens, it is indicated
with the message: "Large negative pivot value ___, at node ___ may be because buckling load has been
exceeded". It must be noted that a stress stiffened model with insufficient boundary conditions to
prevent rigid body motion may yield the same message.

The linear buckling load can be calculated directly by adding an unknown multiplier of the stress
stiffness matrix to the regular stiffness matrix and performing an eigenvalue buckling problem (AN-
TYPE,BUCKLE) to calculate the value of the unknown multiplier. This is discussed in more detail in
Buckling Analysis (p. 826).

3.4.2. Theory
The strain-displacement equations for the general motion of a differential length fiber are derived
below. Two different results have been obtained and these are both discussed below. Consider the
motion of a differential fiber, originally at dS, and then at ds after deformation.

Figure 3.5: General Motion of a Fiber

One end moves {u}, and the other end moves {u + du}, as shown in Figure 3.5: General Motion of a
Fiber (p. 41). The motion of one end with the rigid body translation removed is {u + du} - {u} = {du}.
{du} may be expanded as

(3.48)

where u is the displacement parallel to the original orientation of the fiber. This is shown in Fig-
ure 3.6: Motion of a Fiber with Rigid Body Motion Removed (p. 42). Note that X, Y, and Z represent
global Cartesian axes, and x, y, and z represent axes based on the original orientation of the fiber. By
the Pythagorean theorem,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 41
Structures with Geometric Nonlinearities

(3.49)

The stretch, Λ, is given by dividing ds by the original length dS:

(3.50)

Figure 3.6: Motion of a Fiber with Rigid Body Motion Removed

As dS is along the local x axis,

(3.51)

Next, Λ is expanded and converted to partial notation:

(3.52)

The binominal theorem states that:

(3.53)

when A2 < 1. One should be aware that using a limited number of terms of this series may restrict
its applicability to small rotations and small strains. If the first two terms of the series in Equa-
tion 3.53 (p. 42) are used to expand Equation 3.52 (p. 42),

(3.54)

The resultant strain (same as extension since strains are assumed to be small) is then

(3.55)

If, more accurately, the first three terms of Equation 3.53 (p. 42) are used and displacement derivatives
of the third order and above are dropped, Equation 3.53 (p. 42) reduces to:

(3.56)

The resultant strain is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
42 Published: 2023-10-23
 Stress Stiffening

(3.57)

For most 2D and 3D elements, Equation 3.55 (p. 42) is more convenient to use as no account of the
loaded direction has to be considered. The error associated with this is small as the strains were as-
sumed to be small. For 1D structures, and some 2D elements, Equation 3.57 (p. 43) is used for its
greater accuracy and causes no difficulty in its implementation.

3.4.3. Implementation
The stress-stiffness matrices are derived based on Equation 3.34 (p. 34), but using the nonlinear strain-
displacement relationships given in Equation 3.55 (p. 42) or Equation 3.57 (p. 43) (Cook([5] (p. 925))).

For a spar, the stress-stiffness matrix is given as:

(3.58)

The stress stiffness matrix for a 2D beam is given in Equation 3.59 (p. 43), which is the same as reported
by Przemieniecki([29] (p. 926)). All beam and straight pipe elements use the same type of matrix.
Forces used by straight pipe elements are based on not only the effect of axial stress with pipe wall,
but also internal and external pressures on the "end-caps" of each element. This force is sometimes
referred to as effective tension.

(3.59)

where:

F = force in member
L = length of member

The stress stiffness matrix for 2D and 3D solid elements is generated by the use of numerical integra-
tion. A 3D solid element is used here as an example:

(3.60)

where the matrices shown in Equation 3.60 (p. 43) have been reordered so that first all x-direction
degrees of freedom are given, then y, and then z. [So] is an 8 by 8 matrix given by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 43
Structures with Geometric Nonlinearities

(3.61)

The matrices used by this equation are:

(3.62)

where σx, σxy etc. are stress based on the displacements of the previous iteration, and,

(3.63)

where Ni represents the ith shape function. This is the stress stiffness matrix for small strain analyses.
For large-strain elements in a large-strain analysis (NLGEOM,ON), the stress stiffening contribution is
computed using the actual strain-displacement relationship (Equation 3.6 (p. 31)).

One further case requires some explanation: axisymmetric structures with nonaxisymmetric deforma-
tions. As any stiffening effects may only be axisymmetric, only axisymmetric cases are used for the
prestress case. Axisymmetric cases are defined as (input as MODE on MODE command) = 0. Then,
any subsequent load steps with any value of (including 0 itself ) uses that same stress state, until
another, more recent, = 0 case is available. Also, torsional stresses are not incorporated into any
stress stiffening effects.

Specializing this to SHELL61 (Axisymmetric-Harmonic Structural Shell), only two stresses are used for
prestressing: σs, σθ, the meridional and hoop stresses, respectively. The element stress stiffness matrix
is:
(3.64)

(3.65)

where [As] is defined below and [N] is defined by the element shape functions. [As] is an operator
matrix and its terms are:

(3.66)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
44 Published: 2023-10-23
 Stress Stiffening

The three columns of the [As] matrix refer to u, v, and w motions, respectively. As suggested by the
definition for [Sm], the first two rows of [As] relate to σs and the second two rows relate to σθ. The
first row of [As] is for motion normal to the shell varying in the s direction and the second row is for
hoop motions varying in the s direction. Similarly, the third row is for normal motions varying in the
hoop direction. Thus Equation 3.57 (p. 43), rather than Equation 3.55 (p. 42), is the type of nonlinear
strain-displacement expression that has been used to develop Equation 3.66 (p. 44).

3.4.4. Pressure Load Stiffness

Quite often concentrated forces are treated numerically by equivalent pressure over a known area.
This is especially common in the context of a linear static analysis. However, it is possible that different
buckling loads may be predicted from seemingly equivalent pressure and force loads in an eigenvalue
buckling analysis. The difference can be attributed to pressure being considered as a follower load.
The force on the surface depends on the prescribed pressure magnitude and also on the surface
orientation. Concentrated loads are not considered as follower loads. The follower effects is a preload
stiffness and plays a significant role in nonlinear and eigenvalue buckling analysis. The follower effects
manifest in the form of a load stiffness matrix in addition to the normal stress stiffening effects. As
with any numerical analysis, you should use the type of loading that best models the in-service
component.

The effect of change of direction and/or area of an applied pressure is responsible for the pressure
load stiffness matrix ([Spr]) (see section 6.5.2 of Bonet and Wood([237] (p. 938))). It is used either for a
large-deflection analysis (NLGEOM,ON), for an eigenvalue buckling analysis, or for a modal, linear
transient, or harmonic analysis that has prestressing flagged (PSTRES,ON command).

The need of [Spr] is most dramatically seen when modeling the collapse of a ring due to external
pressure using eigenvalue buckling. The expected answer is:
(3.67)

where:

Pcr = critical buckling load

E = Young's modulus
I = moment of inertia
R = radius of the ring
C = 3.0

This value of C = 3.0 is achieved when using [Spr], but when it is missing, C = 4.0, a 33% error.

For eigenvalue buckling analyses, all elements use pressure load stiffness automatically.

[Spr] is derived as an unsymmetric matrix. Symmetricizing is done, unless the command NROPT,UNSYM
is used. Processing unsymmetric matrices takes more running time and storage, but may be more
convergent.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 45
Structures with Geometric Nonlinearities

3.4.5. Applicable Input

In a nonlinear analysis (ANTYPE,STATIC or ANTYPE,TRANS), the stress stiffness contribution is activated
and then added to the stiffness matrix. When not using large deformations (NLGEOM,OFF), the rotations
are presumed to be small and the additional stiffness induced by the stress state is included. When
using large deformations (NLGEOM,ON), the stress stiffness augments the tangent matrix, affecting
the rate of convergence but not the final converged solution.

The stress stiffness contribution in the prestressed analysis is activated by the prestress flag
(PSTRES,ON) and directs the preceding analysis to save the stress state.

3.4.6. Applicable Output

In a small deflection/small strain analysis (NLGEOM,OFF), the 2D and 3D elements compute their
strains using Equation 3.55 (p. 42). The strains (output as EPEL, EPPL, etc.) therefore include the
higher-order terms (e.g. in the strain computation. Also, nodal and reaction loads (output
quantities F and M) will reflect the stress stiffness contribution, so that moment and force equilibrium
include the higher order (small rotation) effects.

3.5. Spin Softening

The vibration of a spinning body will cause relative circumferential motions, which will change the dir-
ection of the centrifugal load which, in turn, will tend to destabilize the structure. As a small deflection
analysis cannot directly account for changes in geometry, the effect can be accounted for by an adjust-
ment of the stiffness matrix, called spin softening.

The spin softening contribution is included if any of the following criteria are met:

• Large deformations are active (NLGEOM,ON).

• Prestress effects are active (PSTRES,ON) and the analysis is not full harmonic.

• The analysis type is linear perturbation and the spin softening matrix is not set to zero (PERTURB).

• Coriolis effect is activated in a rotating reference frame (CORIOLIS,ON).

For the first three criteria, spin softening is an additional contribution to the tangent matrix (Equa-
tion 3.32 (p. 34)). For the last criteria, it is part of the equations of a rotating structure when expressed
in a rotating reference frame (Equation 14.12 (p. 687)).

In the following sections, equations are first developed for a spring-mass system, and then a general
system equation is formed:
3.5.1. Spring-Mass System
3.5.2. General Equation

3.5.1. Spring-Mass System

Consider a simple spring-mass system, with the spring oriented radially with respect to the axis of
rotation, as shown in Figure 3.7: Spinning Spring-Mass System (p. 47). In the rotating reference frame,
the equilibrium of the spring and centrifugal forces on the mass using small deflection logic requires:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
46 Published: 2023-10-23
 Spin Softening

(3.68)

where:

u = radial displacement of the mass from the rest position along the rotating reference frame
direction X'
r = radial rest position of the mass along the rotating reference frame direction X'
= angular velocity of rotation

Figure 3.7: Spinning Spring-Mass System

Y'

X'

However, to account for large-deflection effects, Equation 3.68 (p. 47) must be expanded to:
(3.69)

Rearranging terms,
(3.70)

Defining:
(3.71)

and
(3.72)

Equation 3.70 (p. 47) becomes simply,

(3.73)

is the stiffness needed in a small deflection solution to account for large-deflection effects. is
the same as that derived from small deflection logic. This decrease in the effective stiffness matrix is
called spin (or centrifugal) softening. See also Carnegie([105] (p. 930)) for additional development.

3.5.2. General Equation

Based on the expression of the centrifugal acceleration (Equation 14.53 (p. 702)), the spin softening
element matrix is expressed as:
(3.74)

where:

= spin softening element matrix

= shape function matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 47
Structures with Geometric Nonlinearities

= element density
= rotational matrix associated with the angular velocity vector (input with OMEGA
or CMOMEGA) – see Equation 14.55 (p. 703)
The tangent matrix can then be written as:
(3.75)

with

(3.76)

where:

ωx, ωy, ωz = x, y, and z components of the angular velocity

If there are more than one non-zero component of angular velocity of rotation, the stiffness matrix
may become unsymmetric. For example, for a diagonal mass matrix with a different mass in each
direction, the matrix becomes nonsymmetric with the expression in Equation 3.75 (p. 48) expanded
as:
(3.77)
(3.78)
(3.79)
(3.80)
(3.81)
(3.82)
(3.83)
(3.84)
(3.85)

where:

Kxx, Kyy, Kzz = x, y, and z components of stiffness matrix as computed by the element
Kxy, Kyx, Kxz, Kzx, Kyz, Kzy = off-diagonal components of stiffness matrix as computed by
the element

Mxx, Myy, Mzz = x, y, and z components of mass matrix

From Equation 3.77 (p. 48) thru Equation 3.85 (p. 48), it may be seen that there are spin softening
effects only in the plane of rotation, not normal to the plane of rotation. Using the example of a
modal analysis, Equation 3.71 (p. 47) can be combined with Equation 15.52 (p. 811) to give:
(3.86)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
48 Published: 2023-10-23
 General Element Formulations

or
(3.87)

where:

ω = the natural circular frequencies of the rotating body.

If stress stiffening is added to Equation 3.87 (p. 49), the resulting equation is:
(3.88)

Stress stiffening is normally applied whenever spin softening is activated, even though they are inde-
pendent theoretically. The modal analysis of a thin fan blade is shown in Figure 3.8: Effects of Spin
Softening and Stress Stiffening (p. 49).

Figure 3.8: Effects of Spin Softening and Stress Stiffening

On Fan Blade Natural Frequencies

3.6. General Element Formulations

Element formulations developed in this section are applicable for general finite-strain deformation.
Naturally, they are applicable to small deformations, small deformation with large rotations, and stress
stiffening as particular cases. The formulations are based on the principle of virtual work. Minimal as-
sumptions are used in arriving at the slope of nonlinear force-displacement relationship (the element
tangent stiffness). Hence, they are also called consistent formulations. These formulations have been
implemented in PLANE182, PLANE183, SOLID185, and SOLID186. SOLID187, SOLID272, SOLID273, SOL-
ID285, SOLSH190, LINK180, SHELL181, BEAM188, BEAM189, SHELL208, SHELL209, REINF263, REINF264,
REINF265, SHELL281, SOLID285, PIPE288, PIPE289, and ELBOW290 are further specializations of the
general theory.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 49
Structures with Geometric Nonlinearities

In this section, the convention of index notation will be used. For example, repeated subscripts imply
summation on the possible range of the subscript, usually the space dimension, so that σii = σ11 + σ22
+ σ33, where 1, 2, and 3 refer to the three coordinate axes x1, x2, and x3, otherwise called x, y, and z.

The following general formulation topics are available:

3.6.1. Fundamental Equations
3.6.2. Classical Pure Displacement Formulation
3.6.3. Mixed u-P Formulations
3.6.4. u-P Formulation I
3.6.5. u-P Formulation II
3.6.6. u-P Formulation III
3.6.7. Volumetric Constraint Equations in u-P Formulations
3.6.8. Inverse Formulations

3.6.1. Fundamental Equations

General finite-strain deformation has the following characteristics:

• Geometry changes during deformation. The deformed domain at a particular time is generally dif-
ferent from the undeformed domain and the domain at any other time.

• Strain is no longer infinitesimal so that a large-strain definition has to be employed.

• Cauchy stress cannot be updated simply by adding its increment. It has to be updated by a partic-
ular algorithm in order to take into account the finite deformation.

• Incremental analysis is necessary to simulate the nonlinear behaviors.

The updated Lagrangian method is applied to simulate geometric nonlinearities (accessed with
NLGEOM,ON). Assuming all variables, such as coordinates xi, displacements ui, strains εij, stresses σij,
velocities vi, volume V and other material variables have been solved for and are known at time t;
one solves for a set of linearized simultaneous equations having displacements (and hydrostatic
pressures in the mixed u-P formulation) as primary unknowns to obtain the solution at time t + Δt.
These simultaneous equations are derived from the element formulations which are based on the
principle of virtual work:

(3.89)

where:

σij = Cauchy stress component

ui = displacement
xi = current coordinate

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
50 Published: 2023-10-23
 General Element Formulations

V = volume of deformed body

S = surface of deformed body on which tractions are prescribed

The internal virtual work can be indicated by:

(3.90)

where:

W = internal virtual work

Element formulations are obtained by differentiating this virtual work expression (Bonet and
Wood([237] (p. 938)) and Gadala and Wang([292] (p. 941))). In derivation, only linear differential terms
are kept and all higher order terms are ignored so that finally a linear set of equations can be obtained.

In element formulation, material constitutive law is used to create the relation between stress increment
and strain increment. The constitutive law only reflects the stress increment due to straining. However,
the Cauchy stress is affected by the rigid body rotation and is not objective (not frame invariant). An
objective stress is needed, therefore, to be able to be applied in constitutive law. One of these is
Jaumann rate of Cauchy stress expressed by McMeeking and Rice([293] (p. 941))
(3.91)

where:

= Jaumann rate of Cauchy stress

= time rate of Cauchy stress

Therefore, the Cauchy stress rate is:

(3.92)

Using the constitutive law, the stress change due to straining can be expressed as:
(3.93)

where:

cijkl = material constitutive tensor

vi = velocity

The Cauchy stress rate can be written as:

(3.94)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 51
Structures with Geometric Nonlinearities

3.6.2. Classical Pure Displacement Formulation

Pure displacement formulation only takes displacements or velocities as primary unknown variables.
All other quantities such as strains, stresses and state variables in history-dependent material models
are derived from the displacements. It is the most widely used formulation and is able to handle most
nonlinear deformation problems.

The differentiation of δW:

(3.95)

From Equation 3.94 (p. 51), the stress differentiation can be derived as:
(3.96)

where:

The differentiation of dV is:

(3.97)

where:

ev = eii

Substitution of Equation 3.96 (p. 52) and Equation 3.97 (p. 52) into Equation 3.95 (p. 52) yields:

(3.98)

The third term is unsymmetric and is usually insignificant in most of deformation cases. Hence, it is
ignored. The final pure displacement formulation is:

(3.99)

The above equation is a set of linear equations in terms of Dui or displacement change, which can
be solved by standard linear solvers. This formulation is exactly the same as the one published by
McMeeking and Rice([293] (p. 941)). The stiffness has two terms: the first one is material stiffness due
to straining; the second one is stiffness due to geometric nonlinearity (stress stiffness).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
52 Published: 2023-10-23
 General Element Formulations

Since no other assumption is made on the size or type of deformation, the formulation can be applied
to any deformation problems (small deformation, finite deformation, small deformation with large
rotation, stress stiffening, etc.) so it is called a general element formulation.

To achieve higher efficiency, the second term or stress stiffness is included only if requested for analyses
with geometric nonlinearities (NLGEOM,ON, or PSTRES,ON), buckling analysis (ANTYPE,BUCKLE), or
linear perturbation analysis (PERTURB).

3.6.3. Mixed u-P Formulations

The above pure displacement formulation is computationally efficient. However, the accuracy of any
displacement formulation is dependent on Poisson's ratio or the bulk modulus. In such formulations,
volumetric strain is determined from derivatives of displacements, which are not as accurately predicted
as the displacements themselves. Under nearly incompressible conditions (Poisson's ratio is close to
0.5 or bulk modulus approaches infinity), any small error in the predicted volumetric strain will appear
as a large error in the hydrostatic pressure and subsequently in the stresses. This error will, in turn,
also affect the displacement prediction since external loads are balanced by the stresses, and may
result in displacements very much smaller than they should be for a given mesh--this is called locking-
- or, in some cases, in no convergence at all.

Another disadvantage of pure displacement formulation is that it is not to be able to handle fully
incompressible deformation, such as fully incompressible hyperelastic materials.

To overcome these difficulties, mixed u-P formulations were developed. In these u-P formulations of
the current-technology elements, the hydrostatic pressure or volume change rate is interpolated
on the element level and solved on the global level independently in the same way as displacements.
The final stiffness matrix has the format of:

(3.100)

where:

Δu = displacement increment
= hydrostatic pressure increment

Since hydrostatic pressure is obtained on a global level instead of being calculated from volumetric
strain, the solution accuracy is independent of Poisson's ratio and bulk modulus. Hence, it is more
robust for nearly incompressible material. For fully incompressible material, mixed u-P formulation
has to be employed in order to get solutions.

The pressure degrees of freedom are brought to global level by using internal or external nodes. The
internal nodes are different from the regular (external) nodes in the following aspects:

• Each internal node is associated with only one element.

• The location of internal nodes is not important. They are used only to bring the pressure degrees
of freedom into the global equations.

• Internal nodes are created automatically and are not accessible by users.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 53
Structures with Geometric Nonlinearities

The interpolation function of pressure is determined according to the order of elements. To remedy
the locking problem, they are one order less than the interpolation function of strains or stresses. For
most current-technology elements, the hydrostatic pressure degrees of freedom are introduced by
the internal nodes. The number of pressure degrees of freedom, number of internal nodes, and inter-
polation functions are shown in Table 3.1: Interpolation Functions of Hydrostatic Pressure of Current-
Technology Elements (p. 54).

Table 3.1: Interpolation Functions of Hydrostatic Pressure of Current-Technology Elements

Internal
Element KEYOPT(6) Functions
nodes
PLANE182

selective
reduced
1 1 1
integration and
uniform
reduced
integration
PLANE182

Enhanced strain 1 2 3
formulation
PLANE183
SOLID185

selective
reduced
1 1 1
integration and
uniform
reduced
integration
SOLID185

Enhanced strain
formulation
SOLID186 1 2 4
Uniform
reduced
integration and
full integration
SOLID187 1 1 1
SOLID187 2 2 4
SOLSH190 1 2 4
SOLID272 KEYOPT(2) on r-z plane and
1 KEYOPT(2)
/3 Fourier interpolation in

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
54 Published: 2023-10-23
 General Element Formulations

Internal
Element KEYOPT(6) Functions
nodes
the circumferential (θ)
direction
SOLID273 on r-z
KEYOPT(2) plane and Fourier
1 KEYOPT(2) interpolation in the
x3
circumferential (θ)
direction

In Table 3.1: Interpolation Functions of Hydrostatic Pressure of Current-Technology Elements (p. 54),
, , , , and are the pressure degrees of freedom at internal node i. s, t, and r are the natural
coordinates.

For SOLID285, the hydrostatic pressure degrees of freedom are introduced by extra degrees of freedom
(HDSP) at each node. The total number of pressures and interpolation function of hydrostatic pressure
are shown in Table 3.2: Interpolation Functions of Hydrostatic Pressure for SOLID285 (p. 55).

Table 3.2: Interpolation Functions of Hydrostatic Pressure for SOLID285

Element Functions
285 4

, , , and are the pressure degrees of freedom at each element node i. s, t, and r are the
natural coordinates.

3.6.4. u-P Formulation I

This formulation is for nearly incompressible materials other than hyperelastic materials. For these
materials, the volumetric constraint equations or volumetric compatibility can be defined as (see
Bathe([2] (p. 925)) for details):
(3.101)

where:

K = bulk modulus

P can also be defined as:

(3.102)

In mixed formulation, stress is updated and reported by:

(3.103)

where:

δij = Kronecker delta

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 55
Structures with Geometric Nonlinearities

σij = Cauchy stress from constitutive law

so that the internal virtual work Equation 3.90 (p. 51) can be expressed as:

(3.104)

Introduce the constraint Equation 3.101 (p. 55) by Lagrangian multiplier , the augmented internal
virtual work is:

(3.105)

Substitute Equation 3.103 (p. 55) into above; it is obtained:

(3.106)

where:

ev = δij eij = eii

Take differentiation of Equation 3.105 (p. 56), ignore all higher terms of Dui and D than linear term,
the final formulation can be expressed as:

(3.107)

This is a linear set of equations of Dui and D (displacement and hydrostatic pressure changes). In
the final mixed u-P formulation, the third term is the stress stiffness and is included only if requested
(NLGEOM,ON or PSTRES,ON). The rest of the terms are based on the material stiffness. The first term
is from material constitutive law directly or from straining; the second term is because of the stress
modification (Equation 3.103 (p. 55)); the fourth and fifth terms are the extra rows and columns in
stiffness matrix due to the introduction of the extra degree of freedom: pressure, that is, KuP, KPu and
KPP as in Equation 3.100 (p. 53).

The stress stiffness in the above formulation is the same as the one in pure displacement formulation.
All other terms exist even for small deformation and are the same as the one derived by
Bathe([2] (p. 925)) for small deformation problems.

It is worthwhile to indicate that in the mixed formulation of the higher order elements (PLANE183 ,
SOLID186 and SOLID187 with KEYOPT(6) = 1), elastic strain only relates to the stress in the element
on an averaged basis, rather than pointwise. The reason is that the stress is updated by Equa-
tion 3.103 (p. 55) and pressure is interpolated independently in an element with a function which
is one order lower than the function for volumetric strain. For lower order elements (PLANE182,
SOLID185), this problem is eliminated since either B-bar technology or uniform reduced integration
is used; volumetric strain is constant within an element, which is consistent with the constant pressure
interpolation functions (see Table 3.1: Interpolation Functions of Hydrostatic Pressure of Current-

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
56 Published: 2023-10-23
 General Element Formulations

Technology Elements (p. 54)). In addition, this problem will not arise in element SOLID187 with linear
interpolation function of (KEYOPT(6) = 2). This is because the order of interpolation function of
is the same as the one for volumetric strain. In other words, the number of degrees of freedom in
one element is large enough to make consistent with the volumetric strain at each integration
point. Therefore, when mixed formulation of element SOLID187 is used with nearly incompressible
material, the linear interpolation function of or KEYOPT(6) = 2 is recommended.

3.6.5. u-P Formulation II

A special formulation is necessary for fully incompressible hyperelastic material because the volume
constraint equation is different and hydrostatic pressure can not be obtained from material constitutive
law. Instead, it must be calculated separately. For these types of materials, the stress is updated by:
(3.108)

where:

= deviatoric component of Cauchy stress tensor

The deviatoric component of deformation tensor defined by the eij term of Equation 3.89 (p. 50) can
be expressed as:
(3.109)

The internal virtual work (Equation 3.90 (p. 51)) can be shown using and :

(3.110)

The volume constraint is the incompressible condition. For a fully incompressible hyperelastic mater-
ial, it can be as defined by Sussman and Bathe([125] (p. 931)), Bonet and Wood([237] (p. 938)), Cris-
field([294] (p. 941)):
(3.111)

where:

= determinant of deformation gradient tensor

Xi = original coordinate
Vo = original volume

As in the mixed u-P formulation I (u-P Formulation I (p. 55)), the constraint Equation 3.111 (p. 57)
was introduced to the internal virtual work by the Lagrangian multiplier . Then, differentiating the
augmented internal virtual work, the final formulation is obtained.

This formulation is similar to the formulation for nearly incompressible materials, i.e. Equa-
tion 3.107 (p. 56). The only major difference is that [KPP] = [0] in this formulation. This is because
material in this formulation is fully incompressible.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 57
Structures with Geometric Nonlinearities

3.6.6. u-P Formulation III

When material behavior is almost incompressible, the pure displacement formulation may be applicable.
The bulk modulus of material, however, is usually very large and thus often results in an ill-conditioned
matrix. To avoid this problem, a special mixed u-P formulation is therefore introduced. The almost
incompressible material usually has small volume changes at all material integration points. A new
variable is introduced to quantify this small volume change, and the constraint equation
(3.112)

is enforced by introduction of the modified potential:

(3.113)

where:

W = hyperelastic strain energy potential

Q = energy augmentation due to volume constraint condition

3.6.7. Volumetric Constraint Equations in u-P Formulations

The final set of linear equations of mixed formulations (see Equation 3.100 (p. 53)) can be grouped
into two:
(3.114)
(3.115)

Equation 3.114 (p. 58) are the equilibrium equations and Equation 3.115 (p. 58) are the volumetric
constraint equations. The total number of active equilibrium equations on a global level (indicated
by Nd) is the total number of displacement degrees of freedom without any prescribed displacement
boundary condition. The total number of volumetric constraint equations (indicated by Np) is the
total number of pressure degrees of freedom in all mixed u-P elements. The optimal ratio of Nd/Np
is 2 for 2D elements and 3 for 3D elements. When Nd/Np is too small, the system may have too many
constraint equations which may result in a severe locking problem. On the other hand, when Nd/Np
is too large, the system may have too few constraint equations which may result in too much deform-
ation and loss of accuracy.

When Nd/Np < 1, the system has more volumetric constraint equations than equilibrium equations,
thus the system is over-constrained. In this case, if the u-P formulation I is used, the system equations
will be very ill-conditioned so that it is hard to keep accuracy of solution and may cause divergence.
If the u-P formulation II is used, the system equation will be singular because [KPP] = [0] in this formu-
lation so that the system is not solvable. Therefore, over-constrained models should be avoided as
described in the Element Reference.

Volumetric constraint is incorporated into the final equations as extra conditions. A check is made at
the element level for elements with internal nodes for pressure degrees of freedom and at degrees
of freedom (HDSP) at global level for SOLID285 to see if the constraint equations are satisfied. The
number of elements in which constraint equations have not been satisfied is reported for current-
technology elements if the check is done at element level.

For u-P formulation I, the volumetric constraint is met if:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
58 Published: 2023-10-23
 General Element Formulations

(3.116)

and for u-P formulation II, the volumetric constraint is met if:

(3.117)

and for u-P formulation III, the volumetric constraint is met if:

(3.118)

where:

tolV = tolerance for volumetric compatibility (input as COMP on the CNVTOL command)

3.6.8. Inverse Formulations

When element formulations (or simply formulations) are mentioned in this reference, they refer to
traditional forward solving, where input consists of an undeformed geometry/configuration, and the
solution is the deformed geometry/configuration and the stresses and strains on it.

The term inverse formulation refers to formulations used for inverse solving, where input consists of
a deformed geometry/configuration under loading, and the solution is the undeformed geometry/con-
figuration and the stresses and strains on the deformed input geometry/configuration. Such formula-
tions are therefore called inverse element formulations (or simply inverse formulations).

In this section, the term forward formulation appears (instead of simply formulation) to clearly differ-
entiate between forward and inverse formulations.

For more information, see Nonlinear Static Analysis with Inverse Solving in the Structural Analysis
Guide and Inverse-Solving Analysis of a Rotor Fan Blade with Disk in the Technology Showcase: Example
Problems.

[Link]. Displacement-Based Inverse Formulation

Displacement-based inverse formulation starts with the principal of virtual work (Equation 3.89 (p. 50)
and Equation 3.90 (p. 51)) used in forward formulation.

The diffentiation of :

(3.119)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 59
Structures with Geometric Nonlinearities

is similar to Equation 3.96 (p. 52) but without the third term, as volume is the volume of the input
configuration and is fixed during solution, as are the virtual strains (Govindjeet and Mihal-
ic[[433] (p. 949)] and [[434] (p. 949)], and Lu and Li[[435] (p. 949)]).

The stress differentiation is the same as Equation 3.96 (p. 52), but the displacements for calculating
and are the inverse displacements (rather than as in forward formula-
tion). Here, and are coordinates in the input configuration and the current configuration, re-
spectively.

The final displacement-based formulation is the same as Equation 3.99 (p. 52), but with:

Therefore, the stiffness matrices are unsymmetric in general, requiring the unsymmetric linear
solver.

3.7. Constraints and Lagrange Multiplier Method

Constraints are generally implemented using the Lagrange Multiplier Method (See Belytsch-
ko([345] (p. 944))). This formulation has been implemented in MPC184 as described in the Element Refer-
ence. In this method, the internal energy term given by Equation 3.90 (p. 51) is augmented by a set of
constraints, imposed by the use of Lagrange multipliers and integrated over the volume leading to an
augmented form of the virtual work equation:
(3.120)

where:

W' = augmented potential

and
(3.121)

is the set of constraints to be imposed.

The variation of the augmented potential is zero provided (and, hence ) and, simultan-
eously:
(3.122)

The equation for augmented potential (Equation 3.120 (p. 60)) is a system of ntot equations, where:
(3.123)

where:

ndof = number of degrees of freedom in the model

nc = number of Lagrange multipliers

The solution vector consists of the displacement degrees of freedom and the Lagrange multipliers.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
60 Published: 2023-10-23
 Steady-State Rolling

The stiffness matrix is of the form:

(3.124)

where:

= increments in displacements and Lagrange multiplier, respectively.

3.8. Steady-State Rolling

Stress analysis on rolling bodies is common in many practical fields. A few examples include vehicle
wheels, ball bearings, and rollers in printers. A stress analysis using transient analysis is typically time-
consuming. It can be inefficient when fine meshes are necessary for contact analysis. In certain applica-
tions where the detailed transient dynamic behavior is not the primary concern, one can significantly
simplify the stress analysis with the arbitrary Lagrangian Eulerian (ALE) formulation.

For a rolling body, the mesh for ALE analysis deforms but does not rotate in space; rather, the material
flows through the mesh, where a steady-state solution can be obtained for each deformed mesh. An
advantage of the non-rotating mesh is that it allows the mesh in the critical region (contact area) to be
selectively refined, enhancing simulation efficiency and accuracy.

The following topics related to the ALE formulation and steady-state rolling are available:
3.8.1. General ALE Formulation
3.8.2. ALE Formulation for Steady-State Rolling
3.8.3. Kinematics

For further information, including supported element types, see SSTATE.

3.8.1. General ALE Formulation

In a finite element analysis based on a Lagrangian mesh, quantities are expressed in both material
and spatial domains. Both domains are related via mapping. In the ALE formulation, an additional
ALE domain is introduced:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 61
Structures with Geometric Nonlinearities

Figure 3.9: Domains and Maps in General ALE Formulation

In the ALE formulation, the mesh motion is prescribed or calculated. The inverse map of Ψ is then
used to calculate the material motion. The mesh displacement is obtained as:

3.8.2. ALE Formulation for Steady-State Rolling

The steady-state rolling analysis using ALE formulation is greatly simplified from the general ALE
formulation. The following figure shows a body rotating at a constant angular velocity of ω that deforms
from the initial shape (solid line) to its steady-state deformed shape (dotted line). y

Figure 3.10: Steady-State Rolling Using ALE Formulation

For steady-state ALE analysis, all time derivatives vanish in the ALE domain. Specifically, the mesh
velocity is shown as:

The material velocity is equal to the convective velocity , and can be obtained from the spatial
gradient and the referential particle velocity , as shown:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
62 Published: 2023-10-23
 Steady-State Rolling

where the referential particle velocity is simply the velocity due to the prescribed material rotation.

Based on the kinematics, the calculation can then be performed in the same way as well established
methods for a Lagrangian mesh.

3.8.3. Kinematics
The balance of mass, expressed within the ALE domain, is given by:

where:

= referential mass density

Under steady state, the relative velocity as well as the time derivative vanish. Therefore:

Of special significance is the second term, where . The conclusion therefore is that steady-
state rolling, within the ALE framework, is possible only if the body is homogeneous at least with re-
spect to the axis of rotation, so holds. In this case, the balance of mass is reduced to:

From the balance of linear momentum, the first Piola-Kirchoff, , is placed in the ALE domain. As-
suming that represents the vector field of body forces, then

is integrated over the volume of the body and multiplied by the test function to obtain the
weak formulation of the balance of momentum.

Considering the Dirichlet and Neumann boundary conditions, the weak formulation of the equation
of motion can be expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 63
Structures with Geometric Nonlinearities

The first and the second term of the left side of the equation describe the virtual work of the inertial
forces and the internal forces, respectively. On the right side, the virtual work of the volume forces
and the traction are considered.

Special attention is given to the inertia term, which inserting the time derivative leads to:

To avoid the higher order gradients that may be obtained inserting again the spatial gradient from
the material velocity, Nackenhorst [[424] (p. 949)] proposes that this term be expressed as:

The first term of the right side of the equation describes the steady-state rolling that, under steady-
state conditions, is composed only of first order gradients. The second term is related to the weighted
impulse flux going with the convective velocity over the boundary as it vanishes at natural boundaries,
common in the continuous formulation of rotating bodies.

For the finite element discretization, the weak formulation of the principle of virtual work which in-
cludes the inertia effects is expressed as:

where:

= tangential stiffness matrix

= inertia matrix

= external forces
= inertial forces
= internal forces

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
64 Published: 2023-10-23
Chapter 4: Structures with Material Nonlinearities
This chapter discusses the structural material nonlinearities of plasticity, creep, nonlinear elasticity, hy-
perelasticity, viscoelasticity, concrete and swelling. Not included in this section are the slider, frictional,
or other nonlinear elements (discussed in Element Library (p. 445)) that can represent other nonlinear
material behavior.

The following topics are available:

4.1. Understanding Material Nonlinearities
4.2. Rate-Independent Plasticity
4.3. Rate-Dependent Plasticity (Including Creep and Viscoplasticity)
4.4. Gasket Material
4.5. Nonlinear (Multilinear) Elasticity
4.6. Hyperelasticity
4.7. Bergstrom-Boyce
4.8. Mullins Effect
4.9. Viscoelasticity
4.10. Swelling
4.11. Cohesive Zone Material (CZM) Model
4.12. Fluid Material Models

4.1. Understanding Material Nonlinearities

Material nonlinearities occur because of the nonlinear relationship between stress and strain; that is,
the stress is a nonlinear function of the strain. The relationship is also path-dependent (except for the
case of nonlinear elasticity and hyperelasticity), so that the stress depends on the strain history as well
as the strain itself.

The program can account for many material nonlinearities, as follows:

1. Rate-independent plasticity (p. 67) is characterized by the irreversible instantaneous straining that
occurs in a material.

2. Rate-dependent plasticity (p. 96) allows the plastic-strains to develop over a time interval. This is
also termed viscoplasticity (p. 98).

3. Creep (p. 96) is also an irreversible straining that occurs in a material and is rate-dependent so that
the strains develop over time. The time frame for creep is usually much larger than that for rate-
dependent plasticity.

4. Gasket material (p. 109) may be modeled using special relationships.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 65
Structures with Material Nonlinearities

5. Nonlinear elasticity (p. 111) allows a nonlinear stress-strain relationship to be specified. All straining
is reversible.

6. Hyperelasticity (p. 112) is defined by a strain-energy density potential that characterizes elastomeric
and foam-type materials. All straining is reversible.

7. Viscoelasticity (p. 134) is a rate-dependent material characterization that includes a viscous contribution
to the elastic straining.

8. Swelling (p. 143) allows materials to enlarge in the presence of neutron flux.

Listed in this table are the number of stress and strain components involved. One component uses X
(for example, SX, EPELX, etc.), four components use X, Y, Z, XY, and six components use X, Y, Z, XY, YZ,
XZ.

Strain Definitions

For the case of nonlinear materials, the definition of elastic strain given with Equation 2.1 (p. 5) has
the form of:
(4.1)

where:

= elastic strain vector (output as EPEL)

= total strain vector
= thermal strain vector (output as EPTH)
= plastic strain vector (output as EPPL)
= creep strain vector (output as EPCR)
= swelling strain vector (output as EPSW)

In this case, is the strain measured by a strain gauge. Equation 4.1 (p. 66) is only intended to show
the relationships between the terms. See subsequent sections for more detail).

In POST1, total strain is reported as:

(4.2)

where:

= component total strain (output as EPTO)

Comparing the last two equations,

(4.3)

The difference between these two "total" strains stems from the different usages: can be used to
compare strain gauge results and can be used to plot nonlinear stress-strain curves.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
66 Published: 2023-10-23
 Rate-Independent Plasticity

4.2. Rate-Independent Plasticity

Rate-independent plasticity is characterized by the irreversible straining that occurs in a material once
a certain level of stress is reached. The plastic strains are assumed to develop instantaneously (that is,
independent of time). Several options are available to characterize different types of material behaviors:

• Material Behavior Option

• Bilinear Isotropic Hardening
• Multilinear Isotropic Hardening
• Nonlinear Isotropic Hardening
• Classical Bilinear Kinematic Hardening
• Multilinear Kinematic Hardening
• Nonlinear Kinematic Hardening
• Anisotropic
• Extended Drucker-Prager
• Extended Drucker-Prager Cap
• Cast Iron
• User-specified behavior (described in User Routines and Non-Standard Uses)

Except for user-specified behavior (TB,USER), each material-behavior option is described in greater detail
later in this chapter. Figure 4.1: Stress-Strain Behavior of Each of the Plasticity Options (p. 69) represents
the stress-strain behavior of each of the options.

4.2.1. Theory
Plasticity theory provides a mathematical relationship that characterizes the elastoplastic response
of materials. There are three ingredients in the rate-independent plasticity theory: the yield cri-
terion (p. 67), flow rule (p. 69) and the hardening rule (p. 70). Table 4.1: Notation (p. 67) summarizes
the notation used in the remainder of this chapter.

4.2.2. Yield Criterion

The yield criterion determines the stress level at which yielding is initiated. For multi-component
stresses, this is represented as a function of the individual components, , which can be inter-
preted as an equivalent stress :
(4.4)

where:

= stress vector

Table 4.1: Notation

Mechanical APDL
Variable Definition
Output Label
elastic strains EPEL
plastic strains EPPL

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 67
Structures with Material Nonlinearities

Mechanical APDL
Variable Definition
Output Label
trial strain
equivalent plastic strain EPEQ

stresses S
equivalent stress
material yield parameter
mean or hydrostatic stress HPRES
equivalent stress parameter SEPL

plastic multiplier
yield surface translation
plastic work
translation multiplier
stress-strain matrix
tangent modulus
yield criterion
stress ratio SRAT
plastic potential
deviatoric stress

When the equivalent stress is equal to a material yield parameter σy,

(4.5)

the material will develop plastic strains. If is less than , the material is elastic and the stresses
will develop according to the elastic stress-strain relations. Note that the equivalent stress can never
exceed the material yield since in this case plastic strains would develop instantaneously, thereby
reducing the stress to the material yield. Equation 4.5 (p. 68) can be plotted in stress space as shown
in Figure 4.2: Various Yield Surfaces (p. 69) for some of the plasticity options. The surfaces in Fig-
ure 4.2: Various Yield Surfaces (p. 69) are known as the yield surfaces and any stress state inside the
surface is elastic, that is, they do not cause plastic strains.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
68 Published: 2023-10-23
 Rate-Independent Plasticity

Figure 4.1: Stress-Strain Behavior of Each of the Plasticity Options

Figure 4.2: Various Yield Surfaces

4.2.3. Flow Rule

The flow rule determines the direction of plastic straining and is given as:
(4.6)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 69
Structures with Material Nonlinearities

= plastic multiplier (which determines the amount of plastic straining)

= function of stress termed the plastic potential (which determines the direction of plastic
straining)

If is the yield function (as is normally assumed), the flow rule is termed associative and the plastic
strains occur in a direction normal to the yield surface.

4.2.4. Hardening Rule

The hardening rule describes the changing of the yield surface with progressive yielding, so that the
conditions (stress states) for subsequent yielding can be established. Two hardening rules are available:
work (or isotropic) hardening and kinematic hardening. In work hardening, the yield surface remains
centered about its initial center line and expands in size as the plastic strains develop. For materials
with isotropic plastic behavior this is termed isotropic hardening and is shown in Figure 4.3: Types of
Hardening Rules (p. 70) (a). Kinematic hardening assumes that the yield surface remains constant in
size and the surface translates in stress space with progressive yielding, as shown in Figure 4.3: Types
of Hardening Rules (p. 70) (b).

The yield criterion, flow rule and hardening rule for each option are summarized in Table 4.2: Summary
of Plasticity Options (p. 70) and are discussed in detail later in this chapter.

Figure 4.3: Types of Hardening Rules

Table 4.2: Summary of Plasticity Options

Name TB Lab Yield Flow Rule Hardening Material

Criterion Rule Response
Bilinear Isotropic PLASTIC von associative work bilinear
Hardening (TBOPT Mises/Hill hardening
= BISO)
Nonlinear NLISO von associative work nonlinear
Isotropic Mises/Hill hardening
Hardening
Classical Bilinear PLASTIC von associative kinematic bilinear
Kinematic (TBOPT Mises/Hill (Prandtl-Reuss hardening
Hardening = BKIN) equations)
Nonlinear CHAB von associative kinematic nonlinear
Kinematic Mises/Hill hardening
Hardening

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
70 Published: 2023-10-23
 Rate-Independent Plasticity

Name TB Lab Yield Flow Rule Hardening Material

Criterion Rule Response
Extended EDP von Mises associative work multilinear
Drucker-Prager with or hardening
dependence non-associative
on
hydrostatic
stress
Cast Iron CAST von Mises non-associativework multilinear
with hardening
dependence
on
hydrostatic
stress
Gurson GURS von Mises associative work multilinear
with hardening
dependence
pressure
and porosity

4.2.5. Plastic Strain Increment

If the equivalent stress calculated using elastic properties exceeds the material yield, then plastic
straining must occur. Plastic strains reduce the stress state so that it satisfies the yield criterion,
Equation 4.5 (p. 68). Based on the theory presented in the previous section, the plastic strain increment
is readily calculated.

The hardening rule states that the yield criterion changes with work hardening and/or with kinematic
hardening. Incorporating these dependencies into Equation 4.5 (p. 68), and recasting it into the fol-
lowing form:
(4.7)

where:

= plastic work
= translation of yield surface

and are termed internal or state variables. Specifically, the plastic work is the sum of the plastic
work done over the history of loading:
(4.8)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 71
Structures with Material Nonlinearities

and translation (or shift) of the yield surface is also history dependent and is given as:
(4.9)

where:

= material parameter
= backstress (location of the center of the yield surface)

Equation 4.7 (p. 71) can be differentiated so that the consistency condition is:
(4.10)

Noting from Equation 4.8 (p. 71) that

(4.11)

and from Equation 4.9 (p. 72) that

(4.12)

Equation 4.10 (p. 72) becomes

(4.13)

The stress increment can be calculated via the elastic stress-strain relations
(4.14)

where:

= stress-strain matrix

with
(4.15)

since the total strain increment can be divided into an elastic and plastic part. Substituting Equa-
tion 4.6 (p. 69) into Equation 4.13 (p. 72) and Equation 4.15 (p. 72) and combining Equation 4.13 (p. 72),
Equation 4.14 (p. 72), and Equation 4.15 (p. 72) yields

(4.16)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
72 Published: 2023-10-23
 Rate-Independent Plasticity

The size of the plastic strain increment is therefore related to the total increment in strain, the current
stress state, and the specific forms of the yield and potential surfaces. The plastic strain increment is
then calculated using Equation 4.6 (p. 69):
(4.17)

4.2.6. Implementation
An Euler backward scheme is used to enforce the consistency condition Equation 4.10 (p. 72). This
ensures that the updated stress, strains and internal variables are on the yield surface. The algorithm
proceeds as follows:

1. The material parameter Equation 4.5 (p. 68) is determined for this time step (for example, the
yield stress at the current temperature).

2. The stresses are calculated based on the trial strain , which is the total strain minus the plastic
strain from the previous time point (thermal and other effects are ignored):
(4.18)

where the superscripts are described with Understanding Theory Reference Notation (p. 1) and
subscripts refer to the time point. Where all terms refer to the current time point, the subscript
is dropped. The trial stress is then
(4.19)

3. The equivalent stress is evaluated at this stress level by Equation 4.4 (p. 67). If is less than
, the material is elastic and no plastic strain increment is calculated.

4. If the stress exceeds the material yield, the plastic multiplier is determined by a local Newton-
Raphson iteration procedure (Simo and Taylor([156] (p. 933))).

5. is calculated via Equation 4.17 (p. 73).

6. The current plastic strain is updated

(4.20)

where:

= current plastic strains (output as EPPL)

and the elastic strain calculated

(4.21)

where:

= elastic strains (output as EPEL)

The stress vector is:

(4.22)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 73
Structures with Material Nonlinearities

where:

= stresses (output as S)

7. The increments in the plastic work and the center of the yield surface are calculated via
Equation 4.11 (p. 72) and Equation 4.12 (p. 72) and the current values updated
(4.23)

and
(4.24)

where the subscript n-1 refers to the values at the previous time point.

8. For output purposes, an equivalent plastic strain (output as EPEQ), equivalent plastic strain
increment (output with the label "MAX PLASTIC STRAIN STEP"), equivalent stress parameter
(output as SEPL) and stress ratio (output as SRAT) are calculated. The stress ratio is given
as
(4.25)

where is evaluated using the trial stress . is therefore greater than or equal to one when
yielding is occurring and less than one when the stress state is elastic. The equivalent plastic strain
increment is given as:
(4.26)

The equivalent plastic strain and equivalent stress parameters are developed for each option in
the next sections.

Note that the Euler backward integration scheme in step 4 is the radial return algorithm
(Krieg([47] (p. 927))) for the von Mises yield criterion.

4.2.7. Elastoplastic Stress-Strain Matrix

The tangent or elastoplastic stress-strain matrix is derived from the local Newton-Raphson iteration
scheme used in step 4 above (Simo and Taylor([156] (p. 933))). It is therefore the consistent (or al-
gorithmic) tangent. If the flow rule is nonassociative ( ), then the tangent is unsymmetric.

4.2.8. Bilinear Isotropic Hardening

These options use the von Mises yield criterion with the associated flow rule and isotropic (work)
hardening (accessed via TB,PLASTIC,,,,BISO).

The equivalent stress Equation 4.4 (p. 67) is:

(4.27)

where {s} is the deviatoric stress Equation 4.37 (p. 76). When is equal to the current yield stress
the material is assumed to yield. The yield criterion is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
74 Published: 2023-10-23
 Rate-Independent Plasticity

(4.28)

For work hardening, is a function of the amount of plastic work done. For the case of isotropic
plasticity assumed here, can be determined directly from the equivalent plastic strain of
Equation 4.43 (p. 77) (output as EPEQ) and the uniaxial stress-strain curve as depicted in Figure 4.4: Uni-
axial Behavior (p. 75). is output as the equivalent stress parameter (output as SEPL).

Figure 4.4: Uniaxial Behavior

For Determination

4.2.9. Nonlinear Isotropic Hardening

Both the Voce([253] (p. 939)) hardening law, and the nonlinear power hardening law can be used to
model nonlinear isotropic hardening. The Voce hardening law for nonlinear isotropic hardening be-
havior (accessed with TB,NLISO,,,,VOCE) is specified by the following equation:
(4.29)

where:

= elastic limit
, , b = material parameters characterizing the isotropic hardening behavior of materials

= equivalent plastic strain

The constitutive equations are based on linear isotropic elasticity, the von Mises yield function and
the associated flow rule. The yield function is:
(4.30)

The plastic strain increment is:

(4.31)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 75
Structures with Material Nonlinearities

= plastic multiplier

The equivalent plastic strain increment is then:

(4.32)

The accumulated equivalent plastic strain is:

(4.33)

The power hardening law for nonlinear isotropic hardening behavior (accessed with TB,NLISO,,,,POWER)
which is used primarily for ductile plasticity and damage is developed in the Gurson's Model (p. 89):
(4.34)

where:

= current yield strength

= initial yield strength
= shear modulus

is the microscopic equivalent plastic strain and is defined by:

(4.35)

where:

= macroscopic plastic strain tensor

= rate change of variables
= Cauchy stress tensor
= inner product operator of two second order tensors
= porosity

4.2.10. Bilinear Kinematic Hardening

This option uses the von Mises yield criterion with the associated flow rule and kinematic hardening
(TB,PLASTIC,,,,BKIN).

The equivalent stress Equation 4.4 (p. 67) is therefore:

(4.36)

where = deviatoric stress vector:

(4.37)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
76 Published: 2023-10-23
 Rate-Independent Plasticity

= mean or hydrostatic stress

= yield surface translation vector Equation 4.9 (p. 72)

Because Equation 4.36 (p. 76) is dependent on the deviatoric stress, yielding is independent of the
hydrostatic stress state. When is equal to the uniaxial yield stress, , the material is assumed to
yield. The yield criterion Equation 4.7 (p. 71) is therefore:
(4.38)

The associated flow rule yields:

(4.39)

so that the increment in plastic strain is normal to the yield surface. The associated flow rule with the
von Mises yield criterion is known as the Prandtl-Reuss flow equation.

The yield surface translation is defined as:

(4.40)

where:

= shear modulus

= Young's modulus (input as EX on MP command)

= Poisson's ratio (input as PRXY or NUXY on MP command)

The shift strain is calculated analogously to Equation 4.24 (p. 74):

(4.41)

where:

(4.42)

where:

= Young's modulus (input as EX on MP command)

= tangent modulus from the bilinear uniaxial stress-strain curve

The yield surface translation is initially zero and changes with subsequent plastic straining.

The equivalent plastic strain is dependent on the loading history and is defined to be:
(4.43)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 77
Structures with Material Nonlinearities

= equivalent plastic strain for this time point (output as EPEQ)

= equivalent plastic strain from the previous time point

The equivalent stress parameter is equal to the yield stress.

[Link]. Applying an Initial Backstress

For bilinear kinematic hardening, shifting the yield center is possible by applying an initial backstress
in the yield criteria. The yield function becomes:
(4.44)

and the associated flow rule is:

(4.45)

where is the initial backstress (defined via INISTATE). The equivalent stress becomes:

(4.46)

The effective backstress is still constituted according to Rice's hardening rule, which behaves
as a linear relationship with the plastic strain. In the case of a non-zero initial backstress for this
model, the total backstress does not maintain a straightforward linear relationship
with the plastic strain. As a result, the yield center is naturally shifted due to the definition of .

4.2.11. Multilinear Kinematic Hardening

This option (TB,PLASTIC,,,,KINH) uses the Besseling(54 (p. 927)) model, also called the sublayer or
overlay model (Zienkiewicz(55 (p. 927))), to characterize the material behavior. The material behavior
is assumed to be composed of various portions (or subvolumes), all subjected to the same total strain,
but each subvolume having a different yield strength. (For a plane stress analysis, the material can
be thought to be made up of a number of different layers, each with a different thickness and yield
stress.) Each subvolume has a simple stress-strain response but when combined the model can rep-
resent complex behavior. This allows a multilinear stress-strain curve that exhibits the Bauschinger
(kinematic hardening) effect (Figure : Stress-Strain Behavior of Each of the Plasticity Options (p. 69)
(b)).

The following steps are performed in the plasticity calculations:

1. The portion of total volume for each subvolume and its corresponding yield strength are determ-
ined.

2. The increment in plastic strain is determined for each subvolume assuming each subvolume is
subjected to the same total strain.

3. The individual increments in plastic strain are summed using the weighting factors determined
in step 1 to compute the overall or apparent increment in plastic strain.

4. The plastic strain is updated and the elastic strain is calculated.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
78 Published: 2023-10-23
 Rate-Independent Plasticity

The portion of total volume (the weighting factor) and yield stress for each subvolume is determined
by matching the material response to the uniaxial stress-strain curve. A perfectly plastic von Mises
material is assumed and this yields for the weighting factor for subvolume

(4.47)

where:

= weighting factor (portion of total volume) for subvolume and is evaluated sequentially
from 1 to the number of subvolumes
= slope of the kth segment of the stress-strain curve (see Figure 4.5: Uniaxial Behavior
for Multilinear Kinematic Hardening (p. 79))

= sum of the weighting factors for the previously evaluated subvolumes

Figure 4.5: Uniaxial Behavior for Multilinear Kinematic Hardening

The yield stress for each subvolume is given by

(4.48)

where ( , ) is the breakpoint in the stress-strain curve. The number of subvolumes corresponds
to the number of breakpoints specified.

The increment in plastic strain for each subvolume is calculated using a von Mises yield criterion
with the associated flow rule. The section on specialization for bilinear kinematic hardening is followed
but since each subvolume is elastic-perfectly plastic, and therefore is zero.

The plastic strain increment for the entire volume is the sum of the subvolume increments:

(4.49)

where:

= number of subvolumes

The current plastic strain and elastic strain can then be calculated for the entire volume via Equa-
tion 4.20 (p. 73) and Equation 4.21 (p. 73).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 79
Structures with Material Nonlinearities

The equivalent plastic strain (output as EPEQ) is defined by Equation 4.43 (p. 77) and equivalent
stress parameter (output as SEPL) is calculated by evaluating the input stress-strain curve at
(after adjusting the curve for the elastic strain component). The stress ratio (output as SRAT,
Equation 4.25 (p. 74)) is defined using the and values of the first subvolume.

4.2.12. Nonlinear Kinematic Hardening

The material model considered is a rate-independent version of the nonlinear kinematic hardening
model proposed by Chaboche([245] (p. 938), [246] (p. 938)) (accessed via TB,CHABOCHE). The constitutive
equations are based on linear isotropic elasticity, a von Mises yield function and the associated flow
rule. Like the bilinear hardening option, the model can be used to simulate the monotonic hardening
and the Bauschinger effect. The model is also applicable to simulate the ratcheting effect of materials.
In addition, the model allows the superposition of several kinematic models as well as isotropic
hardening models. It is thus able to model the complicated cyclic plastic behavior of materials, such
as cyclic hardening or softening and ratcheting or shakedown.

The model uses the von Mises yield criterion with the associated flow rule, the yield function is:
(4.50)

where:

= isotropic hardening variable

According to the normality rule, the flow rule is written:

(4.51)

where:

= plastic multiplier

The backstress is superposition of several kinematic models as:

(4.52)

where:

= number of kinematic models to be superposed.

The evolution of the backstress (kinematic hardening rule) for each component is defined as:
(4.53)

where:

= material constants for kinematic hardening

The associated flow rule yields:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
80 Published: 2023-10-23
 Rate-Independent Plasticity

(4.54)

The plastic strain increment, Equation 4.51 (p. 80) is rewritten as:
(4.55)

The equivalent plastic strain increment is then:

(4.56)

The accumulated equivalent plastic strain is:

(4.57)

The isotropic hardening variable can be defined by:

(4.58)

where:

= elastic limit
, , = material constants characterizing the material isotropic hardening behavior.

The material hardening behavior in Equation 4.50 (p. 80) can also be defined via bilinear or multi-
linear isotropic hardening options, discussed earlier in Bilinear Isotropic Hardening (p. 74).

The return mapping approach with consistent elastoplastic tangent moduli that was proposed by
Simo and Hughes([252] (p. 939)) is used for numerical integration of the constitutive equation described
above.

[Link]. Applying an Initial Backstress

The Chaboche model supports an initial backstress definition (INISTATE,SET,DTYP,BSTR). The yield
function in Equation 4.50 (p. 80) becomes:

(4.59)

Adhering to the Chaboche constitutive framework, only the associated flow rule changes:

(4.60)

4.2.13. Extended Drucker-Prager Model

This option is an extension of the linear Drucker-Prager yield criterion (input with TB,EDP). Both yield
surface and the flow potential, (input with TBOPT on TB,EDP command) can be taken as linear, hy-

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 81
Structures with Material Nonlinearities

perbolic and power law independently, and thus results in either an associated or nonassociated flow
rule. The yield surface can be changed with progressive yielding of the isotropic hardening plasticity
material options, see hardening rule Figure 4.1: Stress-Strain Behavior of Each of the Plasticity Op-
tions (p. 69) (c) Bilinear Isotropic and (d) Multilinear Isotropic.

The yield function with linear form (input with TBOPT = LYFUN) is:
(4.61)

where:

= material parameter referred to pressure sensitive parameter (input as C1 on TBDATA

command using TB,EDP)

where is the deviatoric Cauchy stress tensor

= yield stress of material (input as C2 on TBDATA command or input as TB,BISO,

TB,NLISO, or TB,PLAST)

The yield function with hyperbolic form (input with TBOPT = HYFUN) is:
(4.62)

where:

= material parameter characterizing the shape of yield surface (input as C2 on TBDATA

command using TB,EDP)

The yield function with power law form (input with TBOPT = PYFUN) is:
(4.63)

where:

= material parameter characterizing the shape of yield surface (input as C2 on TBDATA

command using TB,EDP):

Similarly, the flow potential for linear form (input with TBOPT = LFPOT) is:
(4.64)

The flow potential for hyperbolic form (input with TBOPT = HFPOT) is:
(4.65)

The flow potential for power law form (input with TBOPT = PFPOT) is:
(4.66)

The plastic strain is defined as:

(4.67)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
82 Published: 2023-10-23
 Rate-Independent Plasticity

= plastic multiplier

When the flow potential is the same as the yield function, the plastic flow rule is associated, which
in turn results in a symmetric stiffness matrix. When the flow potential is different from the yield
function, the plastic flow rule is nonassociated, and this results in an unsymmetric material stiffness
matrix. By default, the unsymmetric stiffness matrix (accessed by NROPT,UNSYM) will be symmetricized.

4.2.14. Extended Drucker-Prager Cap Model

The cap model focuses on geomaterial plasticity resulting from compaction at low mean stresses
followed by significant dilation before shear failure ([383] (p. 946)). A three-invariant cap plasticity
model with three smooth yielding surfaces including a compaction cap, an expansion cap, and a
shear envelope proposed by Pelessone ([384] (p. 946)) is described here.

Geomaterials typically have much higher triaxial strength in compression than in tension. The cap
model accounts for this by incorporating the third-invariant of stress tensor (J3) into the yielding
functions.

Introduced first are functions to be used in the cap model, including shear failure envelope function,
compaction cap function, expansion cap function, the Lode angle function, and hardening functions.
Then, a unified yielding function for the cap model that is able to describe all the behaviors of shear,
compaction, and expansion yielding surfaces is derived using the shear failure envelope and cap
functions.

The following topics are covered:

[Link]. Shear Failure Envelope Function
[Link]. Compaction Cap Function
[Link]. Expansion Cap Function
[Link]. Lode Angle Function
[Link]. Hardening Functions

[Link]. Shear Failure Envelope Function

A typical geomaterial shear envelope function is based on the exponential format given below:
(4.68)

where:

= first invariant of Cauchy stress tensor

subscript s = shear envelope function
superscript y = yielding related material constants
= current cohesion-related material constant (input using TB,EDP with TBOPT = CYFUN)
= material constants (input using TB,EDP with TBOPT = CYFUN)

Equation 4.68 (p. 83) reduces to the Drucker-Prager yielding function if parameter is set to zero.
All material constants in Equation 4.68 (p. 83) are defined based on and , which are different
from those in the previous sections. The effect of hydrostatic pressure on material yielding may be

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 83
Structures with Material Nonlinearities

exaggerated at high pressure range by only using the linear term (Drucker-Prager) in Equa-
tion 4.68 (p. 83). Such an exaggeration is reduced by using both the exponential term and linear
term in the shear function. Figure 4.6: Shear Failure Envelope Functions (p. 84) shows the configur-
ation of the shear function. In Figure 4.6: Shear Failure Envelope Functions (p. 84) the dots are the
testing data points, the finer dashed line is the fitting curve based on the Drucker-Prager linear
yielding function, the solid curved line is the fitting curve based on Equation 4.68 (p. 83), and the
coarser dashed line is the limited state of Equation 4.68 (p. 83) at very high pressures. In the figure
is the current modified cohesion obtained by setting I1 in Equation 4.68 (p. 83) to zero.

For positive values of , the shear envelope function of the EDP Cap model is evaluated at ,
giving the constant value .

Figure 4.6: Shear Failure Envelope Functions

[Link]. Compaction Cap Function

The compaction cap function is formulated using the shear envelope function defined in Equa-
tion 4.68 (p. 83).

(4.69)

where:

= Heaviside (or unit step) function

subscript C = compaction cap-related function or constant
= ratio of elliptical x-axis to y-axis ( to )
= key flag indicating the current transition point at which the compaction cap surface
and shear portion intersect.

In Equation 4.69 (p. 84), is an elliptical function combined with the Heaviside function. is
plotted in Figure 4.7: Compaction Cap Function (p. 85).

This function implies:

1. When , the first invariant of stress, is greater than , the compaction cap takes no effect
on yielding. The yielding may happen in either shear or expansion cap portion.

2. When is less than , the yielding may only happen in the compaction cap portion,
which is shaped by both the shear function and the elliptical function.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
84 Published: 2023-10-23
 Rate-Independent Plasticity

Figure 4.7: Compaction Cap Function

[Link]. Expansion Cap Function

Similarly, is an elliptical function combined with the Heaviside function designed for the expansion
cap. is shown in Figure 4.8: Expansion Cap Function (p. 85).

(4.70)

where:

subscript t = expansion cap-related function or constant

This function implies that:

1. When is negative, the yielding may happen in either shear or compaction cap portion,
while the tension cap has no effect on yielding.

2. When is positive, the yielding may only happen in the tension cap portion. The tension
cap is shaped by both the shear function and by another elliptical function.

Equation 4.70 (p. 85) assumes that is only a function of and not a function of as is set
to zero in function .

Figure 4.8: Expansion Cap Function

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 85
Structures with Material Nonlinearities

[Link]. Lode Angle Function

Unlike metals, the yielding and failure behaviors of geomaterials are affected by their relatively
weak (compared to compression) tensile strength. The ability of a geomaterial to resist yielding is
lessened by nonuniform stress states in the principal directions. The effect of reduced yielding ca-
pacity for such geomaterials is described by the Lode angle and the ratio of triaxial extension
strength to compression strength. The Lode angle can be written in a function of stress invariants
and :

(4.71)

where:

and = second and third invariants of the deviatoric tensor of the Cauchy stress tensor.

The Lode angle function is defined by:

(4.72)

where:

= ratio of triaxial extension strength to compression strength

The three-invariant plasticity model is formulated by multiplying in the yielding function by the
Lode angle function described by Equation 4.72 (p. 86). The profile of the yielding surface in a
three-invariant plasticity model is presented in Figure 4.9: Yielding Surface in π-Plane (p. 86).

Figure 4.9: Yielding Surface in π-Plane

[Link]. Hardening Functions

The cap hardening law is defined by describing the evolution of the parameter , the intersection
point of the compaction cap and the axis. The evolution of is related only to the plastic volume
strain . A typical cap hardening law has the exponential form proposed in Fossum and Fre-
drich([93] (p. 930)):
(4.73)

where:

= initial value of at which the cap takes effect in the plasticity model.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
86 Published: 2023-10-23
 Rate-Independent Plasticity

= maximum possible plastic volumetric strain for geomaterials.

Parameters and have units of 1 / (Force / Length) and 1 / (Force / Length)2, respectively. All
constants in Equation 4.73 (p. 86) are non-negative.

Besides cap hardening, another hardening law defined for the evolution of the cohesion parameter
used in the shear portion described in Equation 4.68 (p. 83) is considered. The evolution of the
modified cohesion is assumed to be purely shear-related and is the function of the effective
deviatoric plastic strain :
(4.74)

The effective deviatoric plastic strain is defined by its rate change as follows:

(4.75)

where:

= plastic strain tensor

= rate change of variables
= second order identity tensor

The unified and compacted yielding function for the cap model with three smooth surfaces is for-
mulated using above functions as follows:

(4.76)

where:

= function of both and

Again, the parameter is the intersection point of the compaction cap and the axis. The para-
meter is the state variable and can be implicitly described using and given below:
(4.77)

The yielding model described in Equation 4.76 (p. 87) is used and is drawn in the and plane
in Figure 4.10: Cap Model (p. 87).

Figure 4.10: Cap Model

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 87
Structures with Material Nonlinearities

The cap model also allows non-associated models for all compaction cap, shear envelope, and ex-
pansion cap portions. The nonassociated models are defined through using the yielding functions
in Equation 4.76 (p. 87) as its flow potential functions, while providing different values for some
material constants. It is written below:

(4.78)

where:

(4.79)

where:

superscript f = flow-related material constant

This document refers to the work of Schwer ([385] (p. 947)) and Foster ([386] (p. 947)) for the numer-
ical formulations used in the Pelessone ([384] (p. 946)) smooth cap plasticity model. Ansys, Inc. de-
veloped a new material integrator that is able to achieve a faster convergence rate for the transition
zone between the cap and shear portions.

The flow functions in Equation 4.78 (p. 88) and Equation 4.79 (p. 88) are obtained by replacing ,
, , and in Equation 4.76 (p. 87) and Equation 4.77 (p. 87) with , , , and . The
nonassociated cap model is input by using TB,EDP with TBOPT = CFPOT.

Shear hardening can be taken into account on by providing (via TB,PLASTIC or TB,NLISO). The
initial value of must be consistent to . This input regulates the relationship between the
modified cohesion and the effective deviatoric plastic strain.

As the smooth models have more numerical advantages, it is often necessary to transfer nonsmooth
caps such as the Sandler model ([383] (p. 946)) to a smooth model. To facilitate the model transform-
ation from the nonsmooth cap model to the Pelessone smooth cap model ([384] (p. 946)), two simple
and robust methods are used ([387] (p. 947)); rather than solving a group of nonlinear equations,
the Ansys, Inc. implementation solves only one scalar nonlinear equation.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
88 Published: 2023-10-23
 Rate-Independent Plasticity

Calibrating CAP Constants

Calibrating the CAP constants , , , , , and the hardening input for differs significantly
from the other EDP options. The CAP parameters are all defined in relation to and , while the
other EDP coefficients are defined according to and .

4.2.15. Gurson's Model

The Gurson model is used to represent plasticity and damage in ductile porous metals. The model
theory is based on Gurson([363] (p. 945)) and Tvergaard and Needleman([364] (p. 945)). When plasticity
and damage occur, ductile metal goes through a process of void growth, nucleation, and coalescence.
Gurson’s method models the process by incorporating these microscopic material behaviors into
macroscopic plasticity behaviors based on changes in the void volume fraction (porosity) and pressure.
A porosity index increase corresponds to an increase in material damage, which implies a diminished
material load-carrying capacity.

The microscopic porous metal representation in Figure 4.11: Growth, Nucleation, and Coalescence of
Voids in Microscopic Scale (p. 90)(a), shows how the existing voids dilate (a phenomenon, called void
growth) when the solid matrix is in a hydrostatic-tension state. The solid matrix portion is assumed
to be incompressible when it yields, therefore any material volume growth (solid matrix plus voids)
is due solely to the void volume expansion.

The second phenomenon is void nucleation which means that new voids are created during plastic
deformation. Figure 4.11: Growth, Nucleation, and Coalescence of Voids in Microscopic Scale (p. 90)(b),
shows the nucleation of voids resulting from the debonding of the inclusion-matrix or particle-matrix
interface, or from the fracture of the inclusions or particles themselves.

The third phenomenon is the coalescence of existing voids. In this process, shown in Fig-
ure 4.11: Growth, Nucleation, and Coalescence of Voids in Microscopic Scale (p. 90)(c), the isolated
voids establish connections. Although coalescence may not discernibly affect the void volume, the
load carrying capacity of this material begins to decay more rapidly at this stage.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 89
Structures with Material Nonlinearities

Figure 4.11: Growth, Nucleation, and Coalescence of Voids in Microscopic Scale

The evolution equation of porosity is given by

(4.80)

where:

= porosity
= rate change of variables

The evolution of the microscopic equivalent plastic work is:

(4.81)

where:

= microscopic equivalent plastic strain

= Cauchy stress
= inner product operator of two second order tensors
= macroscopic plastic strain
= current yielding strength

The evolution of porosity related to void growth and nucleation can be stated in terms of the micro-
scopic equivalent plastic strain, as follows:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
90 Published: 2023-10-23
 Rate-Independent Plasticity

(4.82)

where:

= second order identity tensor

The void nucleation is controlled by either the plastic strain or stress, and is assumed to follow a
normal distribution of statistics. In the case of strain-controlled nucleation, the distribution is described
in terms of the mean strain and its corresponding deviation. In the case of stress-controlled nucleation,
the distribution is described in terms of the mean stress and its corresponding deviation. The porosity
rate change due to nucleation is then given as follows:

(4.83)

where:

= volume fraction of the segregated inclusions or particles

= mean strain
= strain deviation
= mean stress
= stress deviation (scalar with stress units)

= pressure

It should be noted that "stress controlled nucleation" means that the void nucleation is determined
by the maximum normal stress on the interfaces between inclusions and the matrix. This maximum
normal stress is measured by . Thus, more precisely, the "stress" in the mean stress refers to
. This relationship better accounts for the effect of triaxial loading conditions on nucleation.

Given Equation 4.80 (p. 90) through Equation 4.83 (p. 91), the material yielding rule of the Gurson
model is defined as follows:

(4.84)

where:

, , and = Tvergaard-Needleman constants

= yield strength of material

= equivalent stress

, the Tvergaard-Needleman function is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 91
Structures with Material Nonlinearities

(4.85)

where:

= critical porosity

= failure porosity

The Tvergaard-Needleman function is used to model the loss of material load carrying capacity, which
is associated with void coalescence. When the current porosity reaches a critical value , the ma-
terial load carrying capacity decreases more rapidly due to the coalescence. When the porosity
reaches a higher value , the material load carrying capacity is lost completely. The associative
plasticity model for the Gurson model has been implemented.

[Link]. Gurson Plasticity with Isotropic/Chaboche Kinematic Hardening

Gurson's model (p. 89) accounts for hydrostatic pressure and material isotropic hardening effects
on porous metals. The Chaboche model is essentially the von Mises plasticity model incorporating
Chaboche-type kinematic hardening.

The Gurson-Chaboche model, also used for modeling porous metals, is an extension of the Gurson
model, combining both material isotropic and kinematic hardening. The model is based on the
work of M hlich and Brocks ([427] (p. 949)).

The yielding function of the Gurson-Tvergaard and Needleman model is modified as follows:

(4.86)

where:

is the total effective backstress, and the total backstress is the summation of several sub-
backstresses:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
92 Published: 2023-10-23
 Rate-Independent Plasticity

where the evolution of each sub-backstress is defined by the Chaboche kinematic law:
(4.87)

The evolution of the equivalent plastic strain is defined through:

This model first requires the input parameters for Gurson plasticity with isotropic hardening, and
then additional input parameters for Chaboche kinematic hardening. For more information, see
Hardening in the Material Reference.

4.2.16. Cast Iron Material Model

The cast iron plasticity model is designed to model gray cast iron. The microstructure of gray cast
iron can be looked at as a two-phase material, graphite flakes inserted into a steel matrix
(Hjelm([334] (p. 944))). This microstructure leads to a substantial difference in behavior in tension and
compression. In tension, the material is more brittle with low strength and cracks form due to the
graphite flakes. In compression, no cracks form, the graphite flakes behave as incompressible media
that transmit stress and the steel matrix only governs the overall behavior.

The model assumes isotropic elastic behavior, and the elastic behavior is assumed to be the same in
tension and compression. The plastic yielding and hardening in tension may be different from that
in compression. (See Figure 4.12: Idealized Response of Gray Cast Iron in Tension and Compres-
sion (p. 93).)

Figure 4.12: Idealized Response of Gray Cast Iron in Tension and Compression

Compression

Tension

Yield Criteria

A composite yield surface is used to describe the different behavior in tension and compression. The
tension behavior is pressure-dependent and the Rankine maximum stress criterion is used. The com-
pression behavior is pressure-independent and the von Mises yield criterion is used. In principal stress
space, the yield surface is a cylinder with a tension cutoff (cap). Figure 4.13: Cross-Section of Yield
Surface (p. 94) shows a cross section of the yield surface on principal deviatoric-stress space and
Figure 4.14: Yield Surface in the Meridional Plane (p. 94) shows the composite yield surface in the
meridional plane.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 93
Structures with Material Nonlinearities

Figure 4.13: Cross-Section of Yield Surface

(Viewed along the hydrostatic pressure axis)

Figure 4.14: Yield Surface in the Meridional Plane

The yield surface for tension and compression "regimes" are described by Equation 4.88 (p. 94) and
Equation 4.89 (p. 94) (Chen and Han([332] (p. 943))).

The yield function for the tension cap is:

(4.88)

and the yield function for the compression regime is:

(4.89)

where:

= hydrostatic pressure

= von Mises equivalent stress

= deviatoric stress tensor

= Lode angle

= second invariant of deviatoric stress tensor

= third invariant of deviatoric stress tensor

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
94 Published: 2023-10-23
 Rate-Independent Plasticity

= tension yield stress

= compression yield stress

Flow Rule

The plastic strain increments are defined as:

(4.90)

where is the so-called plastic flow potential, which consists of the von Mises cylinder in compression
and modified to account for the plastic Poisson's ratio in tension, and takes the form:
(4.91)

(4.92)

where:

= plastic Poisson's ratio (input via TB,CAST)

= flow potential in compression
= flow potential in tension

Experimental evidence suggests that is always less than 0.5. For = 0.5, Equation 4.92 (p. 95)
reduces to the von Mises cylinder. The flow potentials in the meridional plane are shown in this figure:

Figure 4.15: Flow Potential for Cast Iron

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 95
Structures with Material Nonlinearities

The region to the left of the uniaxial compression line is the compressive flow potential, while the
region to the right is the tensile flow potential. As the flow potential is different from the yield function
(nonassociated flow rule), the resulting material Jacobian is unsymmetric.

Hardening

The yield stress in uniaxial tension, , depends on the equivalent uniaxial plastic strain in tension,
, and the temperature . Also the yield stress in uniaxial compression, , depends on the equivalent
uniaxial plastic strain in compression, , and the temperature .

To calculate the change in the equivalent plastic strain in tension, the plastic work expression in the
uniaxial tension case is equated to general plastic work expression as:
(4.93)

where:

= plastic strain vector increment

The change in the equivalent plastic strain in compression is defined as:

(4.94)

where:

4.3. Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

Rate-dependent plasticity describes the flow rule of materials, which depends on time. The deformation
of materials is now assumed to develop as a function of the strain rate (or time). An important class of
applications of this theory is high temperature creep. Several options are provided in Mechanical APDL
to characterize the different types of rate-dependent material behaviors. The creep option is used for
describing material creep over a relative long period or at low strain. The rate-dependent plasticity
option adopts a unified creep approach to describe material behavior that is strain rate dependent.
Anand's viscoplasticity option is another rate-dependent plasticity model for simulations such as metal
forming. Other than other these built-in options, a rate-dependent plasticity model may be incorporated
as user material option through the user-programmable feature.

4.3.1. Creep
For information about the creep options available, see Rate-Dependent Plasticity (Viscoplasticity) in
the Material Reference.

Also see Extended Drucker-Prager (EDP) Creep Model (p. 100) and Cap Creep Model (p. 105).

4.3.2. Rate-Dependent Plasticity

The rate-dependent plasticity model includes four options: Perzyna([296] (p. 941)), Peirce et
al.([297] (p. 941)), EVH([245] (p. 938)), and Anand([160] (p. 934)). The options are defined via the TB,RATE

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
96 Published: 2023-10-23
 Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

command’s TBOPT field (where TBOPT = PERZYNA, PEIRCE, EVH, and ANAND, respectively), and are
available with most current-technology elements.

The material hardening behavior is assumed to be isotropic. The integration of the material constitutive
equations are based a return mapping procedure (Simo and Hughes([252] (p. 939))) to enforce both
stress and material tangential stiffness matrix are consistent at the end of time step. A typical applic-
ation of this material model is the simulation of material deformation at high strain rate, such as impact.

The following topics describe each rate-dependent plasticity option:

[Link]. Perzyna Option
[Link]. Peirce Option
[Link]. Exponential Visco-Hardening (EVH) Option
[Link]. Anand Viscoplasticity Option

[Link]. Perzyna Option

The Perzyna option has the following form:
(4.95)

where:

= equivalent plastic strain rate

m = strain rate hardening parameter (input as C1 via TBDATA command)
γ = material viscosity parameter (input as C2 via TBDATA command)
σ = equivalent stress
σo = static yield stress of material (defined via TB,BISO, or TB,NLISO commands)

Note:

σo is a function of some hardening parameters in general.

As γ tends to , or m tends to zero or tends to zero, the solution converges to the static (rate-
independent) solution. However, for this material option when m is very small (< 0.1), the solution
shows difficulties in convergence (Peric and Owen([298] (p. 941))).

[Link]. Peirce Option

The Peirce option has the following form:

(4.96)

Similar to the Perzyna model, the solution converges to the static (rate-independent) solution, as
γ tends to , or m tends to zero, or tends to zero. For small value of m, this option shows much
better convergency than PERZYNA option (Peric and Owen([298] (p. 941))).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 97
Structures with Material Nonlinearities

[Link]. Exponential Visco-Hardening (EVH) Option

For information about this option, see Exponential Visco-Hardening (EVH) Option in the Material
Reference.

[Link]. Anand Viscoplasticity Option

Metal under elevated temperature, such as the hot-metal-working problems, the material physical
behaviors become very sensitive to strain rate, temperature, history of strain rate and temperature,
and strain hardening and softening. The systematical effect of all these complex factors can be
taken into account and modeled via Anand viscoplasticity([160] (p. 934), [148] (p. 933)). The Anand
option is categorized into the group of the unified plasticity models where the inelastic deformation
refers to all irreversible deformation that can not be simply or specifically decomposed into the
plastic deformation derived from the rate-independent plasticity theories and the part resulted
from the creep effect. Compare to the traditional creep approach, the Anand option introduces a
single scalar internal variable "s", called the deformation resistance, which is used to represent the
isotropic resistance to inelastic flow of the material.

Although the Anand option was originally developed for the metal forming application ([160] (p. 934),
[148] (p. 933)), it is however applicable for general applications involving strain and temperature
effect, including but not limited to such as solder join analysis, high temperature creep etc.

The inelastic strain rate is described by the flow equation as follows:

(4.97)

where:

= inelastic strain rate tensor

= rate of accumulated equivalent plastic strain

S, the deviator of the Cauchy stress tensor, is:

(4.98)

and q, equivalent stress, is:

(4.99)

where:

p = one-third of the trace of the Cauchy stress tensor

σ = Cauchy stress tensor
I = second order identity tensor
":" = inner product of two second-order tensors

The rate of accumulated equivalent plastic strain, , is defined as follows:

(4.100)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
98 Published: 2023-10-23
 Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

The equivalent plastic strain rate is associated with equivalent stress, q, and deformation resistance,
s, by:
(4.101)

A = constant with the same unit as the strain rate

Q = activation energy with unit of energy/volume
R = universal gas constant with unit of energy/volume/temperature
θ = absolute temperature
ξ = dimensionless scalar constant
s = internal state variable
m = dimensionless constant

Equation 4.101 (p. 99) implies that the inelastic strain occurs at any level of stress (more precisely,
deviation stress). This theory is different from other plastic theories with yielding functions where
the plastic strain develops only at a certain stress level above yielding stress.

The evolution of the deformation resistance is dependent of the rate of the equivalent plastic strain
and the current deformation resistance. It is:
(4.102)

where:

a = dimensionless constant
h0 = constant with stress unit
s* = deformation resistance saturation with stress unit

The sign, , is determined by:

(4.103)

The deformation resistance saturation s* is controlled by the equivalent plastic strain rate as follows:

(4.104)

where:

= constant with stress unit

n = dimensionless constant

Because of the , Equation 4.102 (p. 99) is able to account for both strain hardening and strain
softening. The strain softening refers to the reduction on the deformation resistance. The strain
softening process occurs when the strain rate decreases or the temperature increases. Such changes
cause a great reduction on the saturation s* so that the current value of the deformation resistance
s may exceed the saturation.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 99
Structures with Material Nonlinearities

The material constants and their units specified for the Anand option are listed in Table 4.3: Mater-
ial Parameter Units for the Anand Option (p. 100). All constants must be positive, except constant
"a", which must be 1.0 or greater. The inelastic strain rate in Anand's definition of material is tem-
perature and stress dependent as well as dependent on the rate of loading. Determination of the
material parameters is performed by curve-fitting a series of the stress-strain data at various tem-
peratures and strain rates as in Anand([160] (p. 934)) or Brown et al.([148] (p. 933)).

Table 4.3: Material Parameter Units for the Anand Option

TBDATA
Parameter Meaning Units
Constant
1 so Initial value of deformation stress, e.g. psi, MPa
resistance
2 Q/R Q = activation energy energy / volume, e.g. kJ /
mole
R = universal gas constant energy / (volume
temperature), e.g. kJ / (mole
-K
3 A pre-exponential factor 1 / time e.g. 1 / second
4 ξ multiplier of stress dimensionless
5 m strain rate sensitivity of stress dimensionless
6 ho hardening/softening constant stress e.g. psi, MPa
7 coefficient for deformation stress e.g. psi, MPa
resistance saturation value
8 n strain rate sensitivity of saturation dimensionless
(deformation resistance) value
9 a strain rate sensitivity of hardening dimensionless
or softening

where:

kJ = kilojoules
K = kelvin

If h0 is set to zero, the deformation resistance goes away and the Anand option reduces to the
traditional creep model.

4.3.3. Extended Drucker-Prager (EDP) Creep Model

Long term loadings such as gravity and other dead loadings greatly contribute inelastic responses of
geomaterials. In such cases the inelastic deformation is resulted not only from material yielding but
also from material creep. The part of plastic deformation is rate-independent and the creep part is
time or rate-dependent.

In the cases of loading at a low level (not large enough to cause the material to yield), the inelastic
deformation may still occur because of the creep effect. To account for the creep effect, the material
model extended Drucker-Prager (EDP model) combines rate-independent EDP with implicit creep
functions. The combination has been done in such a way that yield functions and flow rules defined

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
100 Published: 2023-10-23
 Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

for rate-independent plasticity are fully exploited for creep deformation, advantageous for complex
models as the required data input is minimal.

[Link]. EDP Inelastic Strain Rate Decomposition

We first assume that the material point yields so that both plastic deformation and creep deformation
occur. Figure 4.16: Material Point in Yielding Condition Elastically Predicted (p. 101) illustrates such
a stress state. We next decompose the inelastic strain rate as follows:
(4.105)

where:

The plastic strain rate is further defined as follows:

(4.106)

where:

= plastic multiplier
Q = flow function that has been previously defined in Equation 4.64 (p. 82), Equa-
tion 4.65 (p. 82), and Equation 4.66 (p. 82) in Extended Drucker-Prager Model (p. 81)

Here we also apply these plastic flow functions to the creep strain rate as follows:
(4.107)

where:

= creep multiplier

Figure 4.16: Material Point in Yielding Condition Elastically Predicted

[Link]. EDP Yielding and Hardening Conditions

As material yields, the real stress should always be on the yielding surface, implying that:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 101
Structures with Material Nonlinearities

(4.108)

where:

F = yielding function defined in Equation 4.64 (p. 82), Equation 4.65 (p. 82), and Equa-
tion 4.66 (p. 82) in Extended Drucker-Prager Model (p. 81)
σ Y = yielding stress

It is assumed that material hardening is related to plasticity only and not to creep. The yield stress,
, evolves as a function of the hardening variable, :

where the hardening variable is the plastic work and is calculated from the summation of:

[Link]. EDP Creep Measurements

The creep behaviors could be measured through a few simple tests such as the uniaxial compression,
uniaxial tension, and shear tests. We here assume that the creep is measured through the uniaxial
compression test described in Figure 4.17: Uniaxial Compression Test (p. 102).

Figure 4.17: Uniaxial Compression Test

The measurements in the test are the vertical stress and vertical creep strain at temperature
T. The creep test is targeted to be able to describe material creep behaviors in a general implicit
rate format as follows:
(4.109)

We define the equivalent creep strain and the equivalent creep stress through the equal creep work
as follows:
(4.110)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
102 Published: 2023-10-23
 Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

where:

and = equivalent creep strain and equivalent creep stress to be defined.

For this particular uniaxial compression test, the stress and creep strain are:

(4.111)

Inserting (Equation 4.111 (p. 103)) into (Equation 4.110 (p. 102)) , we conclude that for this special
test case the equivalent creep strain and the equivalent creep stress just recover the corresponding
test measurements. Therefore, we are able to simply replace the two test measurements in (Equa-
tion 4.109 (p. 102)) with two variables of the equivalent creep strain and the equivalent creep stress
as follows:
(4.112)

Once the equivalent creep stress for any arbitrary stress state is obtained, we can insert it into
(Equation 4.112 (p. 103)) to compute the material creep rate at this stress state. We next focus on
the derivation of the equivalent creep stress for any arbitrary stress state.

[Link]. EDP Equivalent Creep Stress

We first introduce the creep isosurface concept. Figure 4.18: Creep Isosurface (p. 104) shows any two
material points A and B at yielding but they are on the same yielding surface. We say that the creep
behaviors of point A and point B can be measured by the same equivalent creep stress if any and
the yielding surface is called the creep isosurface. We now set point B to a specific point, the inter-
section between the yielding curve and the straight line indicating the uniaxial compression test.
From previous creep measurement discussion, we know that point B has for the coordinate
p and for the coordinate q. Point B is now also on the yielding surface, which immediately implies:
(4.113)

It is interpreted from (Equation 4.113 (p. 103)) that the yielding stress σ Y is the function of the
equivalent creep stress . Therefore, we have:
(4.114)

We now insert (Equation 4.114 (p. 103)) into the yielding condition (Equation 4.108 (p. 102)) again:
(4.115)

We then solve (Equation 4.115 (p. 103)) for the equivalent creep stress for material point A on
the isosurface but with any arbitrary coordinates (p,q). (Equation 4.115 (p. 103)) is, in general, a
nonlinear equation and the iteration procedure must be followed for searching its root. In the local
material iterations, for a material stress point not on the yielding surface but out of the yielding
surface like the one shown in Figure 4.16: Material Point in Yielding Condition Elastically Pre-
dicted (p. 101), (Equation 4.115 (p. 103)) is also valid and the equivalent creep stress solved is always
positive.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 103
Structures with Material Nonlinearities

Figure 4.18: Creep Isosurface

[Link]. EDP Elastic Creep and Stress Projection

When the loading is at a low level or the unloading occurs, the material doesn’t yield and is at an
elastic state from the point view of plasticity. However, the inelastic deformation may still exist fully
due to material creep. In this situation, the equivalent creep stress obtained from (Equa-
tion 4.115 (p. 103)) may be negative in some area. If this is the case, (Equation 4.115 (p. 103)) is not
valid any more. To solve this difficulty, the stress-projection method shown in Figure 4.19: Stress
Projection (p. 105) is used, where the real stress σ is multiplied by an unknown scalar β so that the
projected stress σ* = βσ is on the yielding surface. The parameter β can be obtained by solving the
following equation:
(4.116)

Again, Equation 4.116 (p. 104) is a nonlinear equation except in the linear Drucker-Prager model.
Because the projected stress σ* is on the yielding surface, the equivalent creep stress denoted as
and calculated by inserting σ* into (Equation 4.115 (p. 103)), as follows:
(4.117)

is always positive. The real equivalent creep stress is obtained by simply rescaling as follows:
(4.118)

For creep flow in this situation, (Equation 4.107 (p. 101)) can be simply modified as follows:
(4.119)

For stress in a particular continuous domain indicated by the shaded area in Figure 4.19: Stress
Projection (p. 105), the stresses are not able to be projected on the yielding surface (that is, Equa-
tion 4.116 (p. 104)) has no positive value of solution for β). For stresses in this area, no creep is as-
sumed. This assumption makes some sense partially because this area is pressure-dominated and
the EDP models are shear-dominated.

Having Equation 4.105 (p. 101),Equation 4.106 (p. 101), Equation 4.107 (p. 101), or Equation 4.119 (p. 104),
Equation 4.108 (p. 102), Equation 4.110 (p. 102), and Equation 4.112 (p. 103), the EDP creep model is
a mathematically well-posed problem.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
104 Published: 2023-10-23
 Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

Figure 4.19: Stress Projection

4.3.4. Cap Creep Model

The cap creep model is an extension of the cap (rate-independent plasticity) model (p. 83). The ex-
tension is based on creep theory similar to that of the Extended Drucker-Prager (EDP) creep mod-
el (p. 100).

Several concepts related to material creep theory (such as the equivalent creep stress, equivalent
creep strain, and isosurfaces) are used in the cap creep model and are described in the EDP creep
theory.

Unlike EDP which requires only one creep test measurement, a cap creep model requires two inde-
pendent creep test measurements (as described in Commands Used for Cap Creep (p. 109)) to account
for both shear-dominated creep and compaction-dominated creep behaviors.

The following topics describing the cap creep model are available:
[Link]. Assumptions and Restrictions
[Link]. Functions and Potentials
[Link]. Cap Zones
[Link]. Equivalent Creep Stress
[Link]. Elastic Creep and Stress Projection
[Link]. Commands Used for Cap Creep

For related information, see Extended Drucker-Prager Cap Model (p. 83) and Extended Drucker-Prager
(EDP) Creep Model (p. 100) in this reference.

[Link]. Assumptions and Restrictions

The following assumptions and restrictions apply to the cap creep model:

• Creep in both the shear and creep-expansion zones is measured via uniaxial testing.

• The compaction creep zone is measured via hydrostatic testing.

• Shear creep and compaction creep are independent of each other.

• Shear creep exists in all zones insofar as possible.

• The Lode angle effect (p. 86) is ignored.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 105
Structures with Material Nonlinearities

• Material hardening is dependent on plastic strain only.

[Link]. Functions and Potentials

The following yielding equation applies of the cap portion of the cap creep model:

(4.120)

This equation represents the flow function of the cap creep model:

(4.121)

The equations are similar to those used for yielding (p. 88) and flow function (p. 88) (respectively)
in the cap (rate-independent plasticity) model (p. 86).

The tension cap portion for the cap creep model is nearly identical to the cap (rate-independent
plasticity) (p. 83) yielding function except that the J3 (Lode angle) effect is ignored. The notations
and conventions remain the same.

Equation 4.120 (p. 106) is also used to calculate the equivalent creep stresses, and Equa-
tion 4.121 (p. 106) is used to calculate the creep strain.

[Link]. Cap Zones

For the sake of convenience, the cap domain shown in the following figure is divided into six zones
(labeled 1, 2, 3, 10, 20, and 30):

Figure 4.20: Cap Zoning

Following is a brief description of each zone:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
106 Published: 2023-10-23
 Rate-Dependent Plasticity (Including Creep and Viscoplasticity)

• Zone 1 shows the compaction zone where the material yields and creeps. Zone 10 represents
a compaction-dominated zone; its inelastic behavior is still subject to a certain degree of creep
due to compaction stress.

• Zone 2 shows the shear zone where the material yields and creeps due to the shear effect. Zone
20 represents an elastic shear creep zone.

• Zone 3 shows the expansion zone with both yield and creep. Zone 30 represents an elastic creep
zone.

Creep in zones 2, 20, 3, and 30 is measured via uniaxial compression testing. Those zones do not
experience compaction creep contributed from the compaction cap.

Zones 1 and 10 may also experience shear creep because a certain amount of shear creep exists
in all zones. It is assumed that stress points below line CD are not subject to shear creep.

[Link]. Equivalent Creep Stress

The shear and expansion zones share the same uniaxial compression (p. 105) for creep measurement.
The derivation for the equivalent creep stress for the two zones is similar to that used for extended
Drucker-Prager creep stress (p. 103).

In coordinate system shown in the following figure, the coordinates of the stress points

from the uniaxial compression test are , where is the measured stress or the
equivalent creep stress.

Figure 4.21: Uniaxial Compression Test and Creep Isosurface for Shear and Expansion Zones

The point also on the shear yielding surface should satisfy the corresponding yielding function, as
follows:
(4.122)

or

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 107
Structures with Material Nonlinearities

(4.123)

Equation 4.123 (p. 108) implies , which is used in the yielding condition

resulting in:

(4.124)

Equation 4.124 (p. 108) is then solved for the equivalent creep stress for stresses at either the
shear zone or the expansion zone. As shown in the EDP creep model (p. 100), the measured creep
strain component in the axial direction is the equivalent creep strain; therefore, the creep rate
function for both shear and expansion zones is:

where is the equivalent creep strain for the shear or expansion zone and T is the temperature.

[Link].1. Equivalent Creep Stress in the Compaction Zone

The equivalent creep stress for compaction cap is assumed to be independent of J2 and only re-
lated to the first invariant of stress:

The creep test measurement for the compaction cap is regulated via hydrostatic compression
testing. The measurement and record are

• where p is the hydrostatic pressure, and

• where is the volume creep strain.

Thus, the creep rate function for the compaction cap is:
(4.125)

[Link]. Elastic Creep and Stress Projection

In zones 30, 20, 10, and 1 (where the material does not yield at the shear envelope criterion),
loading is light; it also possible that unloading occurs. Inelastic deformation may still exist, however,
due exclusively to material shear creep. In such a case, the equivalent creep stress obtained from
the previous derivation may be negative. To avoid a negative equivalent creep stress, the stress-
projection method proposed in EDP Elastic Creep and Stress Projection (p. 104) is used.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
108 Published: 2023-10-23
 Gasket Material

[Link]. Commands Used for Cap Creep

To model cap creep, use the input from the cap (rate-independent plasticity) model (p. 83). The
cap creep model ignores any tension-to-compression-strength ratio values and sets them to 1.

Creep data is input via the TB,CREEP command. Because the cap creep model requires two inde-
pendent creep function inputs, an additional material table data command, TBEO, allows shear and
compaction creep data to be defined separately.

4.4. Gasket Material

Gasket joints are essential components in most of structural assemblies. Gaskets as sealing components
between structural components are usually very thin and made of many materials, such as steel, rubber
and composites. From a mechanics point of view, gaskets act to transfer the force between mating
components. The gasket material is usually under compression. The material under compression exhibits
high nonlinearity. The gasket material also shows quite complicated unloading behavior. The primary
deformation of a gasket is usually confined to 1 direction, that is through-thickness. The stiffness con-
tribution from membrane (in-plane) and transverse shear are much smaller, and are neglected.

The table option GASKET allows gasket joints to be simulated with the interface elements, in which the
through-thickness deformation is decoupled from the in-plane deformation, see INTER192 - 2D 4-Node
Gasket (p. 603), INTER193 - 2D 6-Node Gasket (p. 604), INTER194 - 3D 16-Node Gasket (p. 604), and INTER195
- 3D 8-Node Gasket (p. 606) for detailed description of interface elements. The user can directly input
the experimentally measured complex pressure-closure curve (compression curve) and several unloading
pressure-closure curves for characterizing the through thickness deformation of gasket material.

Figure 4.22: Pressure vs. Deflection Behavior of a Gasket Material (p. 109) shows the experimental pressure
vs. closure (relative displacement of top and bottom gasket surfaces) data for a graphite composite
gasket material. The sample was unloaded and reloaded 5 times along the loading path and then un-
loaded at the end of the test to determine the unloading stiffness of the material.

Figure 4.22: Pressure vs. Deflection Behavior of a Gasket Material

4.4.1. Stress and Deformation

The gasket pressure and deformation are based on the local element coordinate systems. The gasket
pressure is actually the stress normal to the gasket element midsurface in the gasket layer. Gasket

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 109
Structures with Material Nonlinearities

deformation is characterized by the closure of top and bottom surfaces of gasket elements, and is
defined as:
(4.126)

Where, uTOP and uBOTTOM are the displacement of top and bottom surfaces of interface elements in
the local element coordinate system based on the mid-plane of element.

4.4.2. Material Definition

The input of material data of a gasket material is specified by the command (TB,GASKET). The input
of material data considers of 2 main parts: general parameters and pressure closure behaviors. The
general parameters defines initial gasket gap, the stable stiffness for numerical stabilization, and the
stress cap for gasket in tension. The pressure closure behavior includes gasket compression (loading)
and tension data (unloading).

The GASKET option has followings sub-options:

Sub-option Description
PARA Define gasket material general parameters
COMP Define gasket compression data
LUNL Define gasket linear unloading data
NUNL Define gasket nonlinear unloading data

A gasket material can have several options at the same time. When no unloading curves are defined,
the material behavior follows the compression curve while it is unloaded.

4.4.3. Thermal Deformation

The thermal deformation is taken into account by using an additive decomposition in the total de-
formation, d, as:
(4.127)

where:

d = relative total deformation between top and bottom surfaces of the interface element
di = relative deformation between top and bottom surfaces causing by the applying stress,
this can be also defined as mechanical deformation
dth = relative thermal deformation between top and bottom surfaces due to free thermal
expansion
do = initial gap of the element and is defined by sub-option PARA

The thermal deformation causing by free thermal expansion is defined as:

(4.128)

where:

α = coefficient of thermal expansion (input as ALPX on MP command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
110 Published: 2023-10-23
 Nonlinear (Multilinear) Elasticity

ΔT = temperature change in the current load step

h = thickness of layer at the integration point where thermal deformation is of interest

4.5. Nonlinear (Multilinear) Elasticity

The following multilinear elasticity topic is available:
4.5.1. Guidelines for Use

4.5.1. Guidelines for Use

Mechanical APDL provides a capability to model nonlinear (multilinear) elastic materials (input using
TB,MELAS). Unlike plasticity, no energy is lost. (The process is conservative.)

Figure 4.23: Stress-Strain Behavior for Nonlinear Elasticity (p. 111) represents the stress-strain behavior
of this option. Note that the material unloads along the same curve, so that no permanent inelastic
strains are induced.

Figure 4.23: Stress-Strain Behavior for Nonlinear Elasticity

The total strain components {εn} are used to compute an equivalent total strain measure:

(4.129)

is used with the input stress-strain curve to get an equivalent value of stress σe .

The elastic (linear) component of strain can then be calculated:

(4.130)

and the "plastic" or nonlinear portion is therefore:

(4.131)

In order to avoid an unsymmetric matrix, only the symmetric portion of the tangent stress-strain
matrix is used:
(4.132)

which is the secant stress-strain matrix.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 111
Structures with Material Nonlinearities

4.6. Hyperelasticity
Hyperelasticity refers to a constitutive response derivable from an elastic free-energy potential. It is
typically used for materials experiencing large elastic deformation. Applications for elastomers such as
vulcanized rubber and synthetic polymers, along with some biological materials, often fall into this
category.

The microstructure of polymer solids consists of chain-like molecules. The flexibility of the molecules
allows for an irregular molecular arrangement, leading to complex behavior. Polymers are usually iso-
tropic at small deformation and anisotropic at larger deformation, as the molecule chains realign to the
loading direction. Under an essentially monotonic loading condition, however, many polymer materials
can be approximated as isotropic.

Some classes of hyperelastic materials cannot be modeled as isotropic, such as:

• Fiber-reinforced polymer composites. Typical fiber patterns include unidirectional and bidirectional,
and the fibers can have a stiffness that is 50-1000 times that of the polymer matrix, resulting in
a strongly anisotropic material behavior.

• Biomaterials (such as muscles and arteries). These anisotropic materials can experience large
deformation in which the anisotropic behavior is due to their fibrous structure.

The typical volumetric behavior of hyperelastic materials can be grouped into two classes:

• Materials typically having small volumetric changes during deformation, such as polymers. These
are incompressible or nearly-incompressible materials.

• Materials having large volumetric changes during deformation, such as foams. These are compress-
ible materials.

The hyperelastic material constitutive models are derived from strain-energy potentials that are functions
of the deformation invariants. (An exception is the response function (p. 128) model, as it obtains the
constitutive response functions directly from experimental data.)

Hyperelastic models are defined via material data tables (TB,HYPER or TB,AHYPER) and include:

• Incompressible or nearly-incompressible isotropic models: Neo-Hookean (p. 118), Mooney-Rivlin (p. 117),
Polynomial Form (p. 119), Ogden Potential (p. 119), Arruda-Boyce (p. 115), Gent (p. 116), Yeoh (p. 120),
and Extended Tube (p. 116).

• Compressible isotropic models: Blatz-Ko (p. 116) and Ogden Compressible Foam (p. 118) options are
applicable to compressible foam or foam-type materials.

• Incompressible or nearly-incompressible isotropic response function (p. 128) hyperelastic model: This
model (TB,EXPE) uses experimental data.

• Invariant-based anisotropic strain-energy potential: This model (TB,AHYPER) is a potential-based function

with parameters for defining the volumetric part, the isochoric part, and the material directions.

Element support for material models is given in Material Model Support for Elements in the Material
Reference.

The following hyperelasticity theory topics are available:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
112 Published: 2023-10-23
 Hyperelasticity

4.6.1. Finite-Strain Elasticity

4.6.2. Deviatoric-Volumetric Multiplicative Split
4.6.3. Isotropic Hyperelasticity
4.6.4. Anisotropic Hyperelasticity
4.6.5. USER Subroutine
4.6.6. Output Quantities
4.6.7. Hyperelasticity Material Curve-Fitting
4.6.8. Experimental Response Functions
4.6.9. Material Stability Check

4.6.1. Finite-Strain Elasticity

A material is said to be hyperelastic if there exists an elastic potential function W (or strain-energy
density function) which is a scalar function of one of the strain or deformation tensors, whose deriv-
ative with respect to a strain component determines the corresponding stress component. It can be
expressed by:
(4.133)

where:

Sij = components of the second Piola-Kirchhoff stress tensor

W = strain-energy function per unit undeformed volume
Eij = components of the Lagrangian strain tensor
Cij = components of the right Cauchy-Green deformation tensor

The Lagrangian strain can be expressed as follows:

(4.134)

where:

δij = Kronecker delta (δij = 1, i = j; δij = 0, i ≠ j)

The nominal strain in the current configuration is defined by:

(4.135)

The deformation tensor Cij is composed of the products of the deformation gradients Fij:
(4.136)

where:

Fij = components of the deformation gradient tensor

Xi = undeformed position of a point in direction i
xi = Xi + ui = deformed position of a point in direction i

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 113
Structures with Material Nonlinearities

ui = displacement of a point in direction i

and is the right stretch tensor.

The Kirchhoff stress is defined:

(4.137)

and the Cauchy stress is obtained by:

(4.138)

The eigenvalues (principal stretch ratios) of Cij are , , and , and exist only if:
(4.139)

which can be re-expressed as:

(4.140)

where:

I1, I2, and I3 = invariants of Cij,

(4.141)

and
(4.142)

J is also the ratio of the deformed elastic volume over the reference (undeformed) volume of materials
(Ogden([295] (p. 941)) and Crisfield([294] (p. 941))).

When there is thermal volume strain, the volume ratio J is replaced by the elastic volume ratio Jel
which is defined as the total volume ratio J over thermal volume ratio Jth, as:
(4.143)

and the thermal volume ratio Jth is:

(4.144)

where:

α = coefficient of the thermal expansion

ΔT = temperature difference about the reference temperature

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
114 Published: 2023-10-23
 Hyperelasticity

4.6.2. Deviatoric-Volumetric Multiplicative Split

Under the assumption that material response is isotropic, it is convenient to express the strain-energy
function in terms of strain invariants or principal stretches (Simo and Hughes([252] (p. 939))).
(4.145)

or
(4.146)

Define the volume-preserving part of the deformation gradient, , as:

(4.147)

and thus
(4.148)

The modified principal stretch ratios and invariants are then:

(4.149)
(4.150)

The strain-energy potential can then be defined as:

(4.151)

4.6.3. Isotropic Hyperelasticity

Following are several forms of strain-energy potential (W) provided (as options TBOPT in TB,HYPER)
for the simulation of incompressible or nearly incompressible hyperelastic materials.

[Link]. Arruda-Boyce Model

The form of the strain-energy potential for Arruda-Boyce model is:

(4.152)

where:

μ = initial shear modulus of material (input on TBDATA commands with TB,HYPER)

λL = limiting network stretch (input on TBDATA commands with TB,HYPER)
d = material incompressibility parameter (input on TBDATA commands with TB,HYPER)

The initial bulk modulus is:

(4.153)

As the parameter λL goes to infinity, the model is converted to Neo-Hookean form.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 115
Structures with Material Nonlinearities

[Link]. Blatz-Ko Model

The form of strain-energy potential for the Blatz-Ko model is:
(4.154)

where:

μ = initial shear modulus of material (input on TBDATA commands with TB,HYPER)

The initial bulk modulus is defined as:

(4.155)

[Link]. Extended Tube Model

The extended-tube model is a physics-based polymer model which introduces the physical consid-
eration on the molecular scale into the formulation of the strain-energy potential. The model con-
siders the topological constraints as well as the limited chain extensibility of network chains in the
filled rubbers.

The elastic strain-energy potential consists of the elastic energy from the crosslinked network and
the constraint network as well as the volumetric strain energy:

(4.156)

where the initial shear modulus is given by G = Gc + Ge, and:

Ge = constraint contribution to modulus

Gc = crosslinked contribution to modulus
δ = extensibility parameter
β = empirical parameter (0 < β ≤1)
d1 = material incompressibility parameter

The model is equivalent to a two-term Ogden model with the following parameters:

[Link]. Gent Model

The form of the strain-energy potential for the Gent model is:

(4.157)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
116 Published: 2023-10-23
 Hyperelasticity

μ = initial shear modulus of material (input on TBDATA commands with TB,HYPER)

Jm = limiting value of (input on TBDATA commands with TB,HYPER)
d = material incompressibility parameter (input on TBDATA commands with TB,HYPER)

The initial bulk modulus is:

(4.158)

As the parameter Jm goes to infinity, the model is converted to Neo-Hookean form.

[Link]. Hencky Model

For information about the Hencky model, see Hencky Hyperelasticity (TB,HYPER,,,,HENCKY) in the
Material Reference.

[Link]. Mooney-Rivlin
This option includes two-, three-, five-, and nine-term Mooney-Rivlin models. The form of the strain-
energy potential for a two-parameter Mooney-Rivlin model is:
(4.159)

where:

c10, c01, d = material constants (input on TBDATA commands with TB,HYPER)

The form of the strain-energy potential for a three-parameter Mooney-Rivlin model is

(4.160)

where:

c10, c01, c11, d = material constants (input on TBDATA commands with TB,HYPER)

The form of the strain-energy potential for five-parameter Mooney-Rivlin model is:

(4.161)

where:

c10, c01, c20, c11, c02, d = material constants (input on TBDATA commands with TB,HYPER)

The form of the strain-energy potential for nine-parameter Mooney-Rivlin model is:

(4.162)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 117
Structures with Material Nonlinearities

where:

c10, c01, c20, c11, c02, c30, c21, c12, c03, d = material constants (input on TBDATA commands
with TB,HYPER)

The initial shear modulus is given by:

(4.163)

The initial bulk modulus is:

(4.164)

[Link]. Neo-Hookean
The form Neo-Hookean strain-energy potential is:
(4.165)

where:

μ = initial shear modulus of materials (input on TBDATA commands with TB,HYPER)

d = material incompressibility parameter (input on TBDATA commands with TB,HYPER)

The initial bulk modulus is related to the material incompressibility parameter by:
(4.166)

where:

K = initial bulk modulus

[Link]. Ogden Compressible Foam Model

The strain-energy potential of the Ogden compressible foam model is based on the principal
stretches of left-Cauchy strain tensor, which has the form:

(4.167)

where:

N = material constant (input as NPTS on TB,HYPER)

μi, α i, βi = material constants (input on TBDATA commands with TB,HYPER)

The initial shear modulus, μ, is given as:

(4.168)

The initial bulk modulus K is defined by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
118 Published: 2023-10-23
 Hyperelasticity

(4.169)

For N = 1, α 1 = -2, μ1= –μ, and β = 0.5, the Ogden option is equivalent to the Blatz-Ko option.

[Link]. Ogden Potential

The Ogden form of strain-energy potential is based on the principal stretches of left-Cauchy strain
tensor, which has the form:

(4.170)

where:

N = material constant (input as NPTS on TB,HYPER)

μi, α i, dk = material constants (input on TBDATA commands with TB,HYPER)

Similar to the Polynomial form, there is no limitation on N. A higher N can provide better fit the
exact solution, however, it may, on the other hand, cause numerical difficulty in fitting the material
constants and also it requests to have enough data to cover the entire range of interest of the de-
formation. Therefore a value of N > 3 is not usually recommended.

The initial shear modulus, μ, is given as:

(4.171)

The initial bulk modulus is:

(4.172)

For N = 1 and α 1 = 2, the Ogden potential is equivalent to the Neo-Hookean potential. For N = 2,
α 1 = 2 and α 2 = -2, the Ogden potential can be converted to the 2 parameter Mooney-Rivlin
model.

[Link]. Polynomial Form

The polynomial form of strain-energy potential is

(4.173)

where:

N = material constant (input as NPTS on TB,HYPER)

cij, dk = material constants (input on TBDATA commands with TB,HYPER)

In general, there is no limitation on N. A higher N may provide better fit the exact solution, however,
it may, on the other hand, cause numerical difficulty in fitting the material constants and requires

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 119
Structures with Material Nonlinearities

enough data to cover the entire range of interest of deformation. Therefore a very higher N value
is not usually recommended.

The Neo-Hookean model can be obtained by setting N = 1 and c01 = 0. Also for N = 1, the two
parameters Mooney-Rivlin model is obtained, for N = 2, the five parameters Mooney-Rivlin model
is obtained and for N = 3, the nine parameters Mooney-Rivlin model is obtained.

The initial shear modulus is defined:

(4.174)

The initial bulk modulus is:

(4.175)

[Link]. Yeoh Model

The Yeoh model is also called the reduced polynomial form. The strain-energy potential is:

(4.176)

where:

N = material constant (input as NPTS on TB,HYPER)

Ci0 = material constants (input on TBDATA commands with TB,HYPER)
dk = material constants (input on TBDATA commands with TB,HYPER)

The Neo-Hookean model can be obtained by setting N = 1.

The initial shear modulus is defined:

(4.177)

The initial bulk modulus is:

(4.178)

4.6.4. Anisotropic Hyperelasticity

The anisotropic constitutive strain-energy density function W is:
(4.179)

where:

Wv = volumetric part of the strain energy

Wd = deviatoric part of strain energy (often called isochoric part of the strain energy)

It is assumed that the material is nearly incompressible or purely incompressible. The volumetric part
Wv is absolutely independent of the isochoric part Wd.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
120 Published: 2023-10-23
 Hyperelasticity

The volumetric part, Wv, is assumed to be only function of J as:

(4.180)

The isochoric part Wd is a function of the invariants of the isochoric part

of the right Cauchy Green tensor and the two constitutive material directions A, B in the undeformed
configuration. The material directions yield so-called structural tensors of the microstructure
of the material. Thus, the strain-energy density yields:

(4.181)

where:

Two strain energy potentials, as forms of polynomial or exponential functions, are available for char-
acterizing the isochoric part of the strain-energy potential.

The third invariant is ignored due to the incompressible assumption. The parameter ς is defined
as:
(4.182)

In Equation 4.181 (p. 121) the irreducible basis of invariants:

(4.183)

The exponential-function-based strain energy potential function is given as:

(4.184)

For information about the splitting of the volumetric and deviatoric parts of the strain-density function,
see Finite-Strain Elasticity (p. 113). For further information about this material model, see Anisotropic
Hyperelasticity (TB,AHYPER) in the Material Reference.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 121
Structures with Material Nonlinearities

4.6.5. USER Subroutine

The option of user subroutine enables you to define your own strain-energy potential. Use of subroutine
userhyper is necessary to provide the derivatives of the strain-energy potential with respect to the
strain invariants. See the Programmer's Reference for more information about writing a user hypere-
lasticity subroutine.

4.6.6. Output Quantities

Stresses (output quantities S) are true (Cauchy) stresses in the global coordinate system. They are
calculated from the second Piola-Kirchhoff stresses using:

(4.185)

where:

ρ, ρo = mass densities in the current and initial configurations

Strains (output as EPEL) are the Hencky (logarithmic) strains (see Equation 3.6 (p. 31)). They are in
the global coordinate system. Thermal strain (output as EPTH) is reported as:

(4.186)

4.6.7. Hyperelasticity Material Curve-Fitting

The hyperelastic constants in the strain-energy density function of a material determine its mechan-
ical response. Therefore, in order to obtain successful results during a hyperelastic analysis, it is neces-
sary to accurately assess the material constants of the materials being examined. Material constants
are generally derived for a material using experimental stress-strain data. It is recommended that this
test data be taken from several modes of deformation over a wide range of strain values. In fact, it
has been observed that to achieve stability, the material constants should be fit using test data in at
least as many deformation states as will be experienced in the analysis.

For hyperelastic materials, simple deformation tests (consisting of six deformation modes) can be
used to accurately characterize the material constants. (See Material Curve-Fitting in the Material
Reference.) All available laboratory test data will be used to determine the hyperelastic material con-
stants. The six different deformation modes are graphically illustrated in Figure 4.24: Illustration of
Deformation Modes (p. 123). Combinations of data from multiple tests will enhance the characterization
of the hyperelastic behavior of a material.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
122 Published: 2023-10-23
 Hyperelasticity

Figure 4.24: Illustration of Deformation Modes

Although the algorithm accepts up to six different deformation states, it can be shown that apparently
different loading conditions have identical deformations, and are thus equivalent. Superposition of
tensile or compressive hydrostatic stresses on a loaded incompressible body results in different
stresses, but does not alter deformation of a material. As depicted in Figure 4.25: Equivalent Deform-
ation Modes (p. 124), we find that upon the addition of hydrostatic stresses, the following modes of
deformation are identical:

1. Uniaxial Tension and Equibiaxial Compression.

2. Uniaxial Compression and Equibiaxial Tension.

3. Planar Tension and Planar Compression.

With several equivalent modes of testing, we are left with only three independent deformation states
for which one can obtain experimental data.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 123
Structures with Material Nonlinearities

Figure 4.25: Equivalent Deformation Modes

The following sections outline the development of hyperelastic stress relationships for each independ-
ent testing mode. In the analyses, the coordinate system is chosen to coincide with the principal
directions of deformation. Thus, the right Cauchy-Green strain tensor can be written in matrix form
by:

(4.187)

where:

λi = 1 + εi principal stretch ratio in the ith direction

εi = principal value of the engineering strain tensor in the ith direction

The principal invariants of Cij are:

(4.188)
(4.189)
(4.190)

For each mode of deformation, fully-incompressible material behavior is also assumed so that third
principal invariant, I3, is identically one:
(4.191)

Finally, the hyperelastic Piola-Kirchhoff stress tensor, Equation 4.133 (p. 113) can be algebraically ma-
nipulated to determine components of the Cauchy (true) stress tensor. In terms of the left Cauchy-
Green strain tensor, the Cauchy stress components for a volumetrically constrained material can be
shown to be:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
124 Published: 2023-10-23
 Hyperelasticity

(4.192)

where:

p = pressure

[Link]. Uniaxial Tension (Equivalently, Equibiaxial Compression)

As shown in Figure 4.24: Illustration of Deformation Modes (p. 123), a hyperelastic specimen is loaded
along one of its axis during a uniaxial tension test. For this deformation state, the principal stretch
ratios in the directions orthogonal to the 'pulling' axis will be identical. Therefore, during uniaxial
tension, the principal stretches, λi, are given by:
(4.193)
(4.194)

Due to incompressibility Equation 4.191 (p. 124):

(4.195)

and with Equation 4.194 (p. 125),

(4.196)

For uniaxial tension, the first and second strain invariants then become:
(4.197)

and
(4.198)

Substituting the uniaxial tension principal stretch ratio values into the Equation 4.192 (p. 125), we
obtain the following stresses in the 1 and 2 directions:
(4.199)

and
(4.200)

Subtracting Equation 4.200 (p. 125) from Equation 4.199 (p. 125), we obtain the principal true stress
for uniaxial tension:
(4.201)

The corresponding engineering stress is:

(4.202)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 125
Structures with Material Nonlinearities

[Link]. Equibiaxial Tension (Equivalently, Uniaxial Compression)

During an equibiaxial tension test, a hyperelastic specimen is equally loaded along two of its axes,
as shown in Figure 4.24: Illustration of Deformation Modes (p. 123). For this case, the principal stretch
ratios in the directions being loaded are identical. Hence, for equibiaxial tension, the principal
stretches, λi, are given by:
(4.203)
(4.204)

Utilizing incompressibility Equation 4.191 (p. 124), we find:

(4.205)

For equibiaxial tension, the first and second strain invariants then become:
(4.206)

and
(4.207)

Substituting the principal stretch ratio values for equibiaxial tension into the Cauchy stress Equa-
tion 4.192 (p. 125), we obtain the stresses in the 1 and 3 directions:
(4.208)

and
(4.209)

Subtracting Equation 4.209 (p. 126) from Equation 4.208 (p. 126), we obtain the principal true stress
for equibiaxial tension:
(4.210)

The corresponding engineering stress is:

(4.211)

[Link]. Pure Shear

(Uniaxial Tension and Uniaxial Compression in Orthogonal Directions)

Pure shear deformation experiments on hyperelastic materials are generally performed by loading
thin, short and wide rectangular specimens, as shown in Figure 4.26: Pure Shear from Direct Com-
ponents (p. 127). For pure shear, plane strain is generally assumed so that there is no deformation
in the 'wide' direction of the specimen: λ2 = 1.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
126 Published: 2023-10-23
 Hyperelasticity

Figure 4.26: Pure Shear from Direct Components

Due to incompressibility Equation 4.191 (p. 124), it is found that:

(4.212)

For pure shear, the first and second strain invariants are:
(4.213)

and
(4.214)

Substituting the principal stretch ratio values for pure shear into the Cauchy stress Equa-
tion 4.192 (p. 125), we obtain the following stresses in the 1 and 3 directions:
(4.215)

and
(4.216)

Subtracting Equation 4.216 (p. 127) from Equation 4.215 (p. 127), we obtain the principal pure shear
true stress equation:
(4.217)

The corresponding engineering stress is:

(4.218)

[Link]. Volumetric Deformation

The volumetric deformation is described as:
(4.219)

As nearly incompressible is assumed, we have:

(4.220)

The pressure, P, is directly related to the volume ratio J through:

(4.221)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 127
Structures with Material Nonlinearities

[Link]. Least Squares Fit Analysis

By performing a least squares fit analysis the Mooney-Rivlin constants can be determined from ex-
perimental stress-strain data and Equation 4.200 (p. 125), Equation 4.210 (p. 126), and Equa-
tion 4.217 (p. 127). Briefly, the least squares fit minimizes the sum of squared error between experi-
mental and Cauchy predicted stress values. The sum of the squared error is defined by:

(4.222)

where:

E = least squares residual error

Ti(Cj) = engineering stress values (function of hyperelastic material constants

n = number of experimental data points

Equation 4.222 (p. 128) is minimized by setting the variation of the squared error to zero: δ E2 = 0.
This yields a set of simultaneous equations which can be used to solve for the hyperelastic constants:

(4.223)

It should be noted that for the pure shear case, the hyperelastic constants cannot be uniquely de-
termined from Equation 4.217 (p. 127). In this case, the shear data must by supplemented by either
or both of the other two types of test data to determine the constants.

4.6.8. Experimental Response Functions

From Equation 4.192 (p. 125), the deviatoric stress is determined solely by the deformation and the
derivatives of the elastic potential function , , and . These derivatives are
called the response functions and are determined analytically for the hyperelastic potentials in Iso-
tropic Hyperelasticity (p. 115) and Anisotropic Hyperelasticity (p. 120). These response functions can
also be determined directly from experimental data, bypassing the need to fit the potential function's
parameters to the experimental data.

The constitutive Equation 4.201 (p. 125), Equation 4.210 (p. 126), and Equation 4.217 (p. 127) give an
explicit relationship between the stress, deformation, and response functions. The experimental data
consists of the measured deformation and stress so that the only unknowns in the constitutive
equations are the response functions.

The volumetric response function ( ) is determined either analytically (from the polynomial
volumetric potential function given as the second term on the right side of Equation 4.173 (p. 119))
or as experimental data of volume ratio versus pressure. Given the experimental data, the volume
ratio for a general deformation is used to determine the pressure and hence the volumetric response
function.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
128 Published: 2023-10-23
 Hyperelasticity

The response functions for the first and second deformation invariant are determined from the exper-
imental data from uniaxial tension, equibiaxial tension, pure shear or combined uniaxial tension and
compression experiments. Additionally, for incompressible materials, uniaxial compression experiments
are equivalent to equibiaxial tension and can be used in place of equibiaxial data to determine the
response functions.

Combined uniaxial tension plus compression data cannot be combined with other data sets (except
pressure-volume), and gives only a material behavior that depends on the first invariant response
function. Combinations of the other experimental data sets can be used to determine the first and
second invariant response functions. Any combination of uniaxial tension, equibiaxial tension or pure
shear data can be used.

For a single set of experimental data, the respective constitutive equation has two unknown response
functions. Setting the second invariant response function ( ) to zero, the constitutive equation
can be solved for the first invariant response function ( ).

Given two sets of experimental data for independent deformations (uniaxial-equibiaxial, uniaxial-pure
shear, or equibiaxial-pure shear), the resulting two equations can be solved for the two unknown re-
sponse functions. With experimental data for uniaxial tension, equibiaxial tension, and pure shear
deformations, the resulting three constitutive equations can be solved for a best fit of the two response
functions.

The response functions can therefore be determined from experimental data over the range of exper-
imental deformation; however, the constitutive response is needed for any general deformation. The
constitutive response is obtained by determining an experimental deformation that is nearest to the
general deformation, where the definition of nearest is that the general deformation and the experi-
mental deformation have the same first invariant I1.

For example, given any arbitrary deformation state, Equation 4.188 (p. 124) gives the equation for the
first invariant. Equating this equation to the uniaxial deformation first invariant of Equation 4.197 (p. 125)
gives
(4.224)

where:

λ1,λ2, and λ3 = stretches for the general deformation

λ = unknown uniaxial stretch that yields an equivalent value of the first invariant I1.

Solving the resulting equation for the tensile value of λ gives the uniaxial deformation that is nearest
to the general deformation. Then, λ is used to determine the response function from the uniaxial
experimental data, which corresponds to the required response function for the general deformation.

In general, Equation 4.224 (p. 129) has two valid solutions, one in tension and one in compression. For
use with the uniaxial tension experimental data, the tensile solution is used. For use with the combined
uniaxial tension and compression data, the tension or compression λ value is chosen so that the dif-
ference between the uniaxial deformation gradient's eigenvalues are closest to the eigenvalues of
the actual deformation gradient.

The nearest equibiaxial and pure shear deformations are determined in a similar manner by equating
the first invariant for the general deformation with the first invariant for the equibiaxial or pure shear

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 129
Structures with Material Nonlinearities

deformation. Then the experimental stretch is determined by choosing the tensile value from the
resulting valid solutions for λ.

4.6.9. Material Stability Check

Stability checks are provided for the Mooney-Rivlin hyperelastic materials. A nonlinear material is
stable if the secondary work required for an arbitrary change in the deformation is always positive.
Mathematically, this is equivalent to:
(4.225)

where:

dσ = change in the Cauchy stress tensor corresponding to a change in the logarithmic strain

Since the change in stress is related to the change in strain through the material stiffness tensor,
checking for stability of a material can be more conveniently accomplished by checking for the pos-
itive definiteness of the material stiffness.

The material stability checks are done at the end of preprocessing but before an analysis actually
begins. At that time, the program checks for the loss of stability for six typical stress paths (uniaxial
tension and compression, equibiaxial tension and compression, and planar tension and compression).
The range of the stretch ratio over which the stability is checked is chosen from 0.1 to 10. If the ma-
terial is stable over the range then no message will appear. Otherwise, a warning message appears
that lists the Mooney-Rivlin constants and the critical values of the nominal strains where the mater-
ial first becomes unstable.

4.7. Bergstrom-Boyce
The Bergstrom-Boyce material model (TB,BB) is a phenomenological-based, highly nonlinear material
model used to model typical elastomers and biological materials. The model allows for a nonlinear
stress-strain relationship, creep, and rate-dependence.

The Bergstrom-Boyce model is based on a spring (A) in parallel with a spring and damper (B) in series,
as shown in Figure 4.27: Bergstrom-Boyce Material Model Representation (p. 130). The material model
is associated with time-dependent stress-strain relationships without complete stress relaxation. All
components (springs and damper) are highly nonlinear.

Figure 4.27: Bergstrom-Boyce Material Model Representation

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
130 Published: 2023-10-23
 Bergstrom-Boyce

The stress state in A can be found in the tensor form of the deformation gradient tensor (F = dxi / dXj)
and material parameters, as follows:

(4.226)

where:

= stress state in A
= initial shear modulus of A

= limiting chain stretch of A

= bulk modulus
= det[F]

= inverse Langevin function (where the Langevin function is given by Equa-

tion 4.227 (p. 131))

For numerical efficiency, the Padé ([370] (p. 946)) approximation shown below is used for the inverse
Langevin function ( ). This approximation will differ from the polynomial approximation used for the
Arruda-Boyce model.
(4.227)

The stress in the viscoelastic component of the material (B) is a function of the deformation and the
rate of deformation. Of the total deformation in B, a portion takes place in the elastic component while
the rest of the deformation takes place in the viscous component. Because the stress in the elastic
portion is equal to the stress in the plastic portion, the total stress can be written merely as a function
of the elastic deformation, as shown in Equation 4.228 (p. 131):

(4.228)

All variables in this equation are analogous to the variables in Equation 4.226 (p. 131). The viscous de-
formation can be found from the total deformation and the elastic deformation:

Correct solutions for and will satisfy:

(4.229)

where:

= direction of the stress tensor

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 131
Structures with Material Nonlinearities

= effective creep rate

= Frobenius norm of the stress tensor

= deviatoric Kirchoff stress

= inelastic network stretch

, ε, C, and m = material input parameters

As is a function of the deformation ( ), and τ is based on the stress tensor, Equation 4.229 (p. 131)
is expanded to:

(4.230)

After Equation 4.230 (p. 132) is satisfied, the corresponding stress tensor from component B is added to
the stress tensor from component A to find the total stress, as shown in Equation 4.231 (p. 132):
(4.231)

Energy dissipation occurs in the viscous component of material B. The total dissipation is given by
summing the incremental dissipation for each time step, where the incremental dissipation is:
(4.232)

Results for the dissipation are stored in ITEM = SEND and component VDAM.

For more information about the constitutive model, see references [368] (p. 946) and [369] (p. 946). For
more information about the implementation, see reference [370] (p. 946).

4.8. Mullins Effect

The Mullins effect (TB,CDM) is a phenomenon typically observed in compliant filled polymers. It is
characterized by a decrease in material stiffness during loading and is readily observed during cyclic
loading as the material response along the unloading path differs noticeably from the response that
along the loading path. Although the details about the mechanisms responsible for the Mullins effect
have not yet been settled, they might include debonding of the polymer from the filler particles,
cavitation, separation of particle clusters, and rearrangement of the polymer chains and particles.

In the body of literature that exists concerning this phenomenon, a number of methods have been
proposed as constitutive models for the Mullins effect. The model is a maximum load modification to
the nearly- and fully-incompressible hyperelastic constitutive models already available. In this model,
the virgin material is modeled using one of the available hyperelastic potentials, and the Mullins effect
modifications to the constitutive response are proportional to the maximum load in the material history.

4.8.1. The Pseudo-Elastic Model

The pseudo-elastic model of the Mullins effect [375] (p. 946) is a modification of the standard thermo-
dynamic formulation for hyperelastic materials and is given by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
132 Published: 2023-10-23
 Mullins Effect

(4.233)

where

= virgin material deviatoric strain-energy potential

= evolving scalar damage variable
= damage function

The arbitrary limits 0 < 1 are imposed with = 1 defined as the state of the material without
any changes due to the Mullins effect. Then, along with equilibrium, the damage function is defined
by:

(4.234)

which implicitly defines the damage parameter . Using Equation 4.234 (p. 133), the deviatoric part
of the second Piola-Kirchhoff stress tensor is then:

(4.235)

The modified Ogden-Roxburgh damage function [376] (p. 946) has the following functional form of
the damage variable:
(4.236)

where , , and are material parameters and Wm is the maximum virgin potential over the time
interval :
(4.237)

The tangent stiffness tensor for a constitutive model defined by Equation 4.233 (p. 133) is expressed
as follows:

(4.238)

The derivative for in Equation 4.236 (p. 133) is:

(4.239)

For user-defined pseudo-elastic Mullins effects, define the damage function and the derivative
via the user-programmable subroutine userMullins.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 133
Structures with Material Nonlinearities

4.9. Viscoelasticity
A material is said to be viscoelastic if the material has an elastic (recoverable) part as well as a viscous
(nonrecoverable) part. Upon application of a load, the elastic deformation is instantaneous while the
viscous part occurs over time.

The viscoelastic model usually depicts the deformation behavior of glass or glass-like materials and may
simulate cooling and heating sequences of such material. These materials at high temperatures turn
into viscous fluids and at low temperatures behave as solids. Further, the material is restricted to be
thermorheologically simple (TRS), which assumes the material response to a load at a high temperature
over a short duration is identical to that at a lower temperature but over a longer duration. The mater-
ial model is available with elements LINK180, SHELL181, PLANE182, PLANE183, SOLID185, SOLID186,
SOLID187, BEAM188, BEAM189, SOLSH190, SHELL208, SHELL209, REINF264, REINF265, SOLID272, SOLID273,
SHELL281, SOLID285, PIPE288, PIPE289, and ELBOW290 for small-deformation and large-deformation
viscoelasticity.

The following topics related to viscoelasticity are available:

4.9.1. Small Strain Viscoelasticity
4.9.2. Constitutive Equations
4.9.3. Numerical Integration
[Link] Simplicity
4.9.5. Large-Deformation Viscoelasticity
4.9.6. Visco-Hypoelasticity
4.9.7. Large-Strain Visco-Hyperelasticity
4.9.8. Large-Strain Visco-Anisotropic Hyperelasticity
4.9.9. Shift Functions

4.9.1. Small Strain Viscoelasticity

In this section, the constitutive equations and the numerical integration scheme for small strain vis-
coelasticity are discussed. Large-strain viscoelasticity is presented in Large-Deformation Viscoelasti-
city (p. 138).

4.9.2. Constitutive Equations

A material is viscoelastic if its stress response consists of an elastic part and viscous part. Upon applic-
ation of a load, the elastic response is instantaneous while the viscous part occurs over time. Generally,
the stress function of a viscoelastic material is given in an integral form. Within the context of small
strain theory, the constitutive equation for an isotropic viscoelastic material can be written as:

(4.240)

where:

σ = Cauchy stress
e = deviatoric part of the strain

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
134 Published: 2023-10-23
 Viscoelasticity

Δ = volumetric part of the strain

G(t) = shear relaxation kernel function
K(t) = bulk relaxation kernel function
t = current time
τ = past time
I = unit tensor

For the elements LINK180, SHELL181, PLANE182, PLANE183, SOLID185, SOLID186, SOLID187, BEAM188,
SOLSH190, SHELL208, SHELL209, REINF264, REINF265, SOLID272, SOLID273, SHELL281, SOLID285,
PIPE288, PIPE289, and ELBOW290, the kernel functions are represented in terms of Prony series, which
assumes that:

(4.241)

(4.242)

where:

, Gi = shear elastic moduli

, Ki = bulk elastic moduli
, = relaxation times for each Prony component

Introducing the relative moduli:

(4.243)
(4.244)

where:

The kernel functions can be equivalently expressed as:

(4.245)

The integral function Equation 4.240 (p. 134) can recover the elastic behavior at the limits of very slow
and very fast load. Here, G0 and K0 are, respectively, the shear and bulk moduli at the fast load limit
(i.e. the instantaneous moduli), and and are the moduli at the slow limit. The elasticity para-
meters input correspond to those of the fast load limit. Moreover by admitting Equation 4.241 (p. 135),
the deviatoric and volumetric parts of the stress are assumed to follow different relaxation behavior.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 135
Structures with Material Nonlinearities

The number of Prony terms for shear nG and for volumetric behavior nK need not be the same, nor
do the relaxation times and .

The Prony representation has a prevailing physical meaning in that it corresponds to the solution of
the classical differential model (the parallel Maxwell model) of viscoelasticity. This physical rooting is
the key to understand the extension of the above constitutive equations to large-deformation cases
as well as the appearance of the time-scaling law (for example, pseudo time) at the presence of time-
dependent viscous parameters.

4.9.3. Numerical Integration

To perform finite element analysis, the integral Equation 4.240 (p. 134) need to be integrated. The in-
tegration scheme proposed by Taylor([113] (p. 931)) and subsequently modified by Simo([327] (p. 943))
is adapted. We will delineate the integration procedure for the deviatoric stress. The pressure response
can be handled in an analogous way. To integrate the deviatoric part of Equation 4.240 (p. 134), first,
break the stress response into components and write:

(4.246)

where:

s = deviatoric stress

In addition,

(4.247)

One should note that

(4.248)

where:

Δt = tn+1 - tn.

The first term of Equation 4.248 (p. 136) is readily recognized as: .

Using the middle point rule for time integration for the second term, a recursive formula can be ob-
tained as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
136 Published: 2023-10-23
 Viscoelasticity

(4.249)

where:

Δe = en+1 - en.

4.9.4. Thermorheological Simplicity

Materials viscous property depends strongly on temperature. For example, glass-like materials turn
into viscous fluids at high temperatures while behave like solids at low temperatures. In reality, the
temperature effects can be complicated. The so called thermorheological simplicity is an assumption
based on the observations for many glass-like materials, of which the relaxation curve at high temper-
ature is identical to that at a low temperature if the time is properly scaled (Scherer([326] (p. 943))).
In essence, it stipulates that the relaxation times (of all Prony components) obey the scaling law

(4.250)

Here, A(T, Tr) is called the shift function. Under this assumption (and in conjunction with the differential
model), the deviatoric stress function can be shown to take the form

(4.251)

likewise for the pressure part. Here, notably, the Prony representation still holds with the time t, τ in
the integrand being replaced by:

here ξ is called pseudo (or reduced) time. In Equation 4.251 (p. 137), is the decay time at a given
temperature.

The assumption of thermorheological simplicity allows for not only the prediction of the relaxation
time over temperature, but also the simulation of mechanical response under prescribed temperature
histories. In the latter situation, A is an implicit function of time t through T = T(t). In either case, the
stress equation can be integrated in a manner similar to Equation 4.246 (p. 136). Indeed,

(4.252)

Using the middle point rule for time integration on Equation 4.252 (p. 137) yields

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 137
Structures with Material Nonlinearities

(4.253)

where:

Two widely used shift functions, namely the Williams-Landel-Ferry shift function and the Tool-Naray-
anaswamy shift function, are available. The form of the functions are given in Shift Functions (p. 141).

4.9.5. Large-Deformation Viscoelasticity

Two types of large-deformation viscoelasticity models are implemented: large-deformation small
strain, and large-deformation large strain. The first is associated with hypo-type constitutive equations,
and the latter is based on hyperelasticity, which includes isotropic and anisotropic hyperelasticity.

4.9.6. Visco-Hypoelasticity
For visco-hypoelasticity model, the constitutive equations are formulated in terms of the rotated
stress RTσR, here R is the rotation arising from the polar decomposition of the deformation gradient
F. Let RTσR = Σ + pI where Σ is the deviatoric part and p is the pressure. It is evident that Σ = RTSR.
The stress response function is given by:

(4.254)

(4.255)

where:

d = deviatoric part of the rate of deformation tensor D.

This stress function is consistent with the generalized differential model in which the stress rate is
replaced by Green-Naghdi rate.

To integrate the stress function, one perform the same integration scheme in Equation 4.246 (p. 136)
to the rotated stress Equation 4.254 (p. 138) to yield:

(4.256)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
138 Published: 2023-10-23
 Viscoelasticity

= rotation tensor arising from the polar decomposition of the middle point deformation
gradient

In the actual implementations, the rate of deformation tensor is replaced by the strain increment and
we have
(4.257)

where:

symm[.] = symmetric part of the tensor.

From Σ = RTsR and using Equation 4.256 (p. 138) and Equation 4.257 (p. 139), it follows that the deviat-
oric Cauchy stress is given by

(4.258)

where:

The pressure response can be integrated in a similar manner and the details are omitted.

4.9.7. Large-Strain Visco-Hyperelasticity

Large-strain visco-hyperelasticity is based on the formulation proposed by (Simo([327] (p. 943))),
amended here to take into account the viscous volumetric response and the thermorheological sim-
plicity. Simo's formulation is an extension of the small strain theory. Again, the viscoelastic behavior
is specified separately by the underlying elasticity and relaxation behavior.

This decomposition of the energy function is consistent with hyperelasticity described in Hyperelasti-
city (p. 112).
(4.259)

where:

J = det (F)

As is well known, the constitutive equations for hyperelastic material with strain-energy function Φ
is given by:
(4.260)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 139
Structures with Material Nonlinearities

S2d = second Piola-Kirchhoff stress tensor

The true stress can be obtained as:

(4.261)

Using Equation 4.259 (p. 139) in Equation 4.261 (p. 140) results

(4.262)

It has been shown elsewhere that is deviatoric, therefore Equation 4.262 (p. 140) already as-
sumes the form of deviatoric/pressure decomposition.

Following Simo([327] (p. 943)) and Holzapfel([328] (p. 943)), the viscoelastic constitutive equations, in
terms of the second Piola-Kirchhoff stress, is given by

(4.263)

Denote

(4.264)

(4.265)

and applying the recursive formula to Equation 4.264 (p. 140) and Equation 4.265 (p. 140) yields,

(4.266)

(4.267)

The above are the updating formulas used in the implementation. Cauchy stress can be obtained
using Equation 4.261 (p. 140).

4.9.8. Large-Strain Visco-Anisotropic Hyperelasticity

Large-strain visco-anisotropic hyperelasticity assumes that viscoelastic behavior is isotropic, while the
underlying elasticity is anisotropic hyperelasticity. The algorithm is similar to that of viscoelastic hy-
perelasticity (p. 139). The strain energy potential is now represented as follows:
(4.268)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
140 Published: 2023-10-23
 Viscoelasticity

A and B are the two unit constitutive material directions, |A| = 1 and |B| = 1. Further information
about the anisotropic hyperelastic potential is available in Anisotropic Hyperelasticity (p. 120).

4.9.9. Shift Functions

Mechanical APDL offers the following forms of the shift function:
[Link]. Williams-Landel-Ferry Shift Function
[Link].Tool-Narayanaswamy Shift Function
[Link].Tool-Narayanaswamy Shift Function with Fictive Temperature
[Link]. User-Defined Shift Function

The shift function is activated via the TB,SHIFT command. For detailed information, see Viscoelasticity
in the Material Reference.

[Link]. Williams-Landel-Ferry Shift Function

The Williams-Landel-Ferry shift function (Williams [277] (p. 940)) is defined by

(4.269)

where:

T = temperature
C1, C2, C3 = material parameters

[Link]. Tool-Narayanaswamy Shift Function

The Tool-Narayanaswamy shift function (Narayanaswamy [111] (p. 931)) is defined by
(4.270)

where:

Tr = material parameter

= material parameter

[Link]. Tool-Narayanaswamy Shift Function with Fictive Temperature

This extension of the Tool-Narayanaswamy shift function includes a fictive temperature. The shift
function is defined by
(4.271)

where:

TF = fictive temperature

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 141
Structures with Material Nonlinearities

= material parameter

The fictive temperature is given by

where:

nf = number of partial fictive temperatures

Cfi = fictive temperature relaxation coefficients
Tfi = partial fictive temperatures

An integrator for the partial fictive temperatures (Markovsky [109] (p. 930)) is given by

where:

= time increment
= temperature relaxation times
The superscript 0 represents values from the previous time step.

The fictive temperature model also modifies the volumetric thermal strain model and gives the in-
cremental thermal strain as

where the glass and liquid coefficients of thermal expansion are given by

The total thermal strain is given by the sum over time of the incremental thermal strains

[Link]. User-Defined Shift Function

Other shift functions can be accommodated via the user-provided subroutine UsrShift, described
in the Programmer's Reference. The inputs for this subroutine are the user-defined parameters, the
current value of time and temperature, their increments, and the current value of user state variables

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
142 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

(if any). The outputs from the subroutine are Δξ, Δξ1/2 as well as the current value of user state
variables.

4.10. Swelling
Mechanical APDL can account for irradiation-induced swelling (accessed via TB,SWELL). Swelling is
defined as a material enlarging volumetrically in the presence of neutron flux. The amount of swelling
can also be a function of temperature. The material is assumed to be isotropic and the basic solution
technique used is the initial stress method.

For a solid element, the swelling strain vector is simply:

(4.272)

It is seen that the swelling strains are handled in a manner totally analogous to temperature strains in
an isotropic medium and that shearing strains are not used.

For more information, see Swelling in the Material Reference.

4.11. Cohesive Zone Material (CZM) Model

Fracture or delamination along an interface between phases plays a major role in limiting the toughness
and the ductility of the multi-phase materials, such as matrix-matrix composites and laminated composite
structure. This has motivated considerable research on the failure of the interfaces. Interface delamination
can be modeled by traditional fracture mechanics methods such as the nodal release technique. Altern-
atively, you can use techniques that directly introduce fracture mechanism by adopting softening rela-
tionships between tractions and the separations, which in turn introduce a critical fracture energy that
is also the energy required to break apart the interface surfaces. This technique is called the cohesive
zone material (CZM) model. The interface surfaces of the materials can be represented by a special set
of interface elements or contact elements, and a CZM model can be used to characterize the constitutive
behavior of the interface.

The CZM model consists of a constitutive relation between the traction T acting on the interface and
the corresponding interfacial separation δ (displacement jump across the interface). The definitions of
traction and separation depend on the element and the material model.

The following related topics are available:

4.11.1. Interface Elements
4.11.2. Contact Elements

4.11.1. Interface Elements

For interface elements, the interfacial separation is defined as the displacement jump, δ (that is, the
difference of the displacements of the adjacent interface surfaces):
(4.273)

The definition of the separation is based on local element coordinate system, Figure 4.28: Schematic
of Interface Elements (p. 144). The normal of the interface is denoted as local direction n, and the
local tangent direction is denoted as t. Thus:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 143
Structures with Material Nonlinearities

(4.274)
(4.275)

Figure 4.28: Schematic of Interface Elements

The following topics about using interface elements are available:

[Link]. Material Model – Exponential Behavior
[Link]. Material Model – Bilinear Behavior
[Link]. Material Model – Exponential Behavior for Preventing Surface Penetration
[Link]. Material Model – Linear Behavior for Preventing Surface Penetration
[Link]. Viscous Regularization
[Link]. Material Model – Rigid Exponential Behavior
[Link]. Material Model – Frictional Behavior

[Link]. Material Model – Exponential Behavior

An exponential form of the CZM model (input via TB,CZM with TBOPT = EXPO), originally proposed
by Xu and Needleman([360] (p. 945)), uses a surface potential:
(4.276)

where:

ϕ(δ) = surface potential

e = 2.7182818
σmax = maximum normal traction at the interface (input on TBDATA command as C1 using
TB,CZM)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
144 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

The traction is defined as:

(4.277)

or
(4.278)

and

(4.279)

From equations Equation 4.278 (p. 145) and Equation 4.279 (p. 145), we obtain the normal traction
of the interface
(4.280)

and the shear traction

(4.281)

The normal work of separation is:

(4.282)

and shear work of separation is assumed to be the same as the normal work of separation, ϕn, and
is defined as:
(4.283)

For the 3D stress state, the shear or tangential separations and the tractions have two components,
δt1 and δt2 in the element's tangential plane, and we have:
(4.284)

The traction is then defined as:

(4.285)

and

(4.286)

(In POST1 and POST26 the traction, T, is output as SS and the separation, δ, is output as SD.)

The tangential direction t1 is defined along ij edge of element and the direction t2 is defined along
direction perpendicular to the plane formed by n and t1. Directions t1, t2, and n follow the right-
hand side rule.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 145
Structures with Material Nonlinearities

[Link]. Material Model – Bilinear Behavior

The bilinear CZM model (input via TB,CZM with TBOPT = BILI) can be used with interface elements.
The model is based on the model proposed by Alfano and Crisfield [362] (p. 945).

Mode I Dominated Bilinear CZM Model

The Mode I dominated bilinear CZM model assumes that the separation of the material interfaces
is dominated by the displacement jump normal to the interface, as shown in the following figure:

Figure 4.29: Mode I Dominated Bilinear CZM Law

The relation between normal cohesive traction Tn and normal displacement jump δn can be expressed
as:
(4.287)

Where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
146 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

For Mode I dominated cohesive law, the tangential cohesive traction and tangential displacement
jump behavior is assumed to follow the normal cohesive traction and normal displacement jump
behavior and is expressed as:
(4.288)

Were:

Mode II Dominated Bilinear CZM Model

The Mode II dominated bilinear CZM model assumes that the separation of the material interfaces
is dominated by the displacement jump that is tangent to the interface, as shown in the following
figure:

Figure 4.30: Mode II Dominated Bilinear CZM Law

The relation between tangential cohesive traction Tt and tangential displacement jump δt can be
expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 147
Structures with Material Nonlinearities

(4.289)

Where:

For Mode II dominated cohesive law, the normal cohesive traction and normal displacement jump
behavior is assumed to follow the tangential cohesive traction and tangential displacement jump
behavior and is expressed as:
(4.290)

Where:

Mixed-Mode Bilinear Cohesive Zone Material Model

For bilinear cohesive law under the mixed-mode fracture, the separation of material interfaces de-
pends on both the normal and tangential components of displacement jumps. To take into account
the difference in their contributions to the separation of material interfaces, a non-dimensional ef-
fective displacement jump λ for mixed-mode fracture is defined as

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
148 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

(4.291)

where the non-dimensional parameter β (input via the TBDATA command as C6 using TB,CZM)
assigns different weights to the tangential and normal displacement jumps.

The normal and tangential components of the cohesive tractions are expressed as:

(4.292)

The damage parameter Dm associated with mixed mode bilinear cohesive law is defined as:

Where:

Fracture Mode Identification of a CZM Model

Determining the fracture mode of a CZM model is based on the input data, as follows:

Case Input on the TBDATA command as follows:

Mode I Dominated C1, C2, C3, C4, C5 (where C3 = -τmax)
Mode II/III Dominated C1, C2, C3, C4, C5 (where C1 = -σmax)
Mixed-Mode C1, C2, C3, C4, C5, C6 (where C1 = σmax and C3 = τmax)

[Link]. Material Model – Exponential Behavior for Preventing Surface Penet-

ration
An exponential form of the CZM model for preventing surface penetration (input via TB,CZM with
TBOPT = CEXP) is similar to the standard exponential behavior of the CZM model when upper and
lower surfaces contact each other. There is no interface traction when surfaces are separated.

(4.293)

where:

= traction in the normal direction

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 149
Structures with Material Nonlinearities

= traction in the tangential direction

is a CZM material model constant that helps to define tractions and stiffness at crack-surface
closure. To calibrate its value, refer to in the standard EXPO CZM model.

Definitions for the remaining terms here are identical to those for the standard exponential
form of the CZM model in Material Model – Exponential Behavior (p. 144).

[Link]. Material Model – Linear Behavior for Preventing Surface Penetration

A linear form of the CZM model for preventing surface penetration (input via TB,CZM with TBOPT
= CLIN) describes the interface traction in the normal direction when upper and lower surfaces
contact each other. There is no interface traction when surfaces are separated.

(4.294)

where:

= traction in the normal direction

= traction in the tangential direction
= constant for penalty method
= normal separation

[Link]. Viscous Regularization

To avoid convergence difficulties, a small fictitious viscosity is introduced in cohesive zone material
(CZM) models ([413] (p. 948)). The traction definitions are modified by adding viscous damping, as
follows:

(4.295)

where:

Tn = normal traction before viscous regularization

Tt = tangential traction before viscous regularization

ζ = damping coefficient (input on the TBDATA command as C1 (after defining the CZM model with
the viscous regularization option (TB,CZM,,,,VREG))

[Link]. Material Model – Rigid Exponential Behavior

Unlike the standard exponential behavior of the CZM model, a rigid exponential form of the CZM
model (TB,CZM with TBOPT = REXP) bypasses positive stiffness region and proceeds directly to
the damage stage.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
150 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

Figure 4.31: Traction-Separation Relationship Along the Interface

When the separation is positive without unloading, the potential is:

(4.296)

where:

= surface potential
= 2.7182818
= maximum traction at surface
= separation across the interface, including both normal ( ) and tangential ( ) components:

= separation where the maximum traction is attained:

As shown in Figure 4.31: Traction-Separation Relationship Along the Interface (p. 151), a linear rule
is applied for the unloading-reloading path. For interface closures, a penalty algorithm is adopted
to prevent penetration. By default, the penalty slope is a constant, set according to the reference
stiffness.

Alternatively, a penalty slope can be applied by specifying exponential material behavior to prevent
surface penetration (TB,CZM with TBOPT = CEXP):

Example 4.1: Applying a Penalty Slope via TB,CZM

! Apply penalty slope via TB,CZM with TBOPT = CEXP.
! The material ID (MATID) for the following TB commands must be
! identical (in this case 1).

TB,CZM,1,1,,REXP ! Activate REXP material model

TBDATA,1, max_r, cn_r, ct_r ! Define material constants for REXP

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 151
Structures with Material Nonlinearities

TB,CZM,1,1,,CEXP ! Activate CEXP material model

TBDATA,1, max_c, cn_c, ct_c ! Define material constants for CEXP

[Link]. Material Model – Frictional Behavior

An isotropic Coulomb’s friction law (TB,CZM with TBOPT = FRIC) is provided to simulate tangential
stresses of the CZM model when upper and lower surfaces contact each other. There is no friction
traction when surfaces are separated. A penalty algorithm is applied.

This method requires tangential stiffness, and the amount of penetration between the two surfaces
depends on the stiffness. Higher stiffness values decrease the amount of penetration but can lead
to ill-conditioning of the global stiffness matrix and to convergence difficulties. Ideally, the stiffness
is high enough that contact penetration is acceptably small, but low enough so that the problem
is well-behaved in terms of convergence or matrix ill-conditioning.

The interface traction vector is:

where:

= normal contact pressure

= frictional stress in direction 1
= frictional stress in direction 2

The frictional stress for isotropic friction is obtained via Coulomb's law:

(4.297)

where:

= tangential penalty stiffness

= slip increment in direction over the current substep; in case of sticking, the elastic slip
increment represents the reversible tangential motion within the current substep
= equivalent slip increment over the current substep
= coefficient of friction
= frictional stress in direction = 1, 2 at the end of the previous substep
= limit frictional stress, such that when the equivalent shear stress is < , no motion
occurs between two surfaces
=

is typically defined as but can be changed (TB,CZM with TBOPT = FRIC) so that,
regardless of the magnitude of the contact pressure, sliding occurs if the magnitude of the equivalent
frictional stress reaches the defined value.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
152 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

4.11.2. Contact Elements

Delamination with contact elements is referred to as debonding. The interfacial separation is defined
in terms of contact gap or penetration and tangential slip distance. The computation of contact and
tangential slip is based on the type of contact element and the location of contact detection point.
The cohesive zone model can only be used for bonded contact (KEYOPT(12) = 2, 3, 4, 5, or 6) with
the augmented Lagrangian method (KEYOPT(2) = 0) or the pure penalty method (KEYOPT(2) = 1). See
CONTA174 - 3D 8-Node Surface-to-Surface Contact (p. 555) for details.

[Link]. Material Models for Debonding

Cohesive zone modeling with contact elements includes two different traction separation law forms:
a bilinear traction separation law and an exponential traction separation law. Different variations
of the bilinear material models are available, with each variation being activated by a different
TBOPT value on the TB,CZM command, as described in the following sections:
[Link].1. Bilinear Behavior - Contact (TBOPT = CBDD and CBDE)
[Link].2. Bilinear Behavior - Interface (TBOPT = BILI)
[Link].3. Exponential Behavior

[Link].1. Bilinear Behavior - Contact (TBOPT = CBDD and CBDE)

The bilinear cohesive zone material model (input using TB,CZM with TBOPT = CBDD or CBDE)
is based on the model proposed by Alfano and Crisfield [362] (p. 945).

Mode I Debonding

Mode I debonding defines a mode of separation of the interface surfaces where the separation
normal to the interface dominates the slip tangent to the interface. The normal contact stress
(tension) and contact gap behavior is plotted in Figure 4.32: Normal Contact Stress and Contact
Gap Curve for Bilinear Cohesive Zone Material (p. 154). It shows linear elastic loading (OA) followed
by linear softening (AC). The maximum normal contact stress is achieved at point A. Debonding
begins at point A and is completed at point C when the normal contact stress reaches zero value;
any further separation occurs without any normal contact stress. The area under the curve OAC
is the energy released due to debonding and is called the critical fracture energy. The slope of
the line OA determines the contact gap at the maximum normal contact stress and, hence,
characterizes how the normal contact stress decreases with the contact gap, that is, whether the
fracture is brittle or ductile. After debonding has been initiated it is assumed to be cumulative
and any unloading and subsequent reloading occurs in a linear elastic manner along line OB at
a more gradual slope.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 153
Structures with Material Nonlinearities

Figure 4.32: Normal Contact Stress and Contact Gap Curve for Bilinear Cohesive Zone
Material

The equation for curve OAC can be written as:

(4.298)

where:

P = normal contact stress (tension)

Kn = normal contact stiffness
un = contact gap
= contact gap at the maximum normal contact stress (tension)
= contact gap at the completion of debonding (input on TBDATA command as C2
using TB,CZM)
dn = debonding parameter

The debonding parameter for Mode I Debonding is defined as:

(4.299)

with dn = 0 for Δn 1 and 0 < dn 1 for Δn > 1.

where:

The normal critical fracture energy is calculated as:

(4.300)

where:

σmax = maximum normal contact stress (input on TBDATA command as C1 using TB,CZM).

For mode I debonding the tangential contact stress and tangential slip behavior follows the
normal contact stress and contact gap behavior and is written as:
(4.301)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
154 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

where:

τt = tangential contact stress

Kt = tangential contact stiffness
ut = tangential slip distance

Mode II Debonding

Mode II debonding defines a mode of separation of the interface surfaces where tangential slip
dominates the separation normal to the interface. The equation for the tangential contact stress
and tangential slip distance behavior is written as:
(4.302)

where:

= tangential slip distance at the maximum tangential contact stress

= tangential slip distance at the completion of debonding (input on TBDATA command
as C4 using TB,CZM)
dt = debonding parameter

The debonding parameter for Mode II Debonding is defined as:

(4.303)

with dt = 0 for Δt 1 and 0 < dt 1 for Δt > 1.

where:

For the 3D stress state an "isotropic" behavior is assumed and the debonding parameter is calcu-
lated using an equivalent tangential slip distance:
(4.304)

where:

u1 and u2 = slip distances in the two principal directions in the tangent plane

The components of the tangential contact stress are defined as:

(4.305)

and
(4.306)

The tangential critical fracture energy is calculated as:

(4.307)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 155
Structures with Material Nonlinearities

where:

τmax = maximum tangential contact stress (input on TBDATA command as C3 using

TB,CZM).

The normal contact stress and contact gap behavior follows the tangential contact stress and
tangential slip behavior and is written as:
(4.308)

Mixed-Mode Debonding

In mixed-mode debonding the interface separation depends on both normal and tangential
components. The equations for the normal and the tangential contact stresses are written as:
(4.309)

and
(4.310)

The debonding parameter is defined as:

(4.311)

with dm = 0 for Δm 1 and 0 < dm 1 for Δm > 1, and Δm and χ are defined below.

where:

and

The constraint on χ that the ratio of the contact gap distances be the same as the ratio of tan-
gential slip distances is enforced automatically by appropriately scaling the contact stiffness value,
Kt, as follows:

(4.312)

Kn in the above equation is the contact stiffness value. If the contact opening stiffness is specified
by real constant FKOP of contact element, then contact opening stiffness is used in this equation.
If contact opening stiffness is not specified, the contact closing stiffness is used.

For mixed-mode debonding, both normal and tangential contact stresses contribute to the total
fracture energy, and debonding is completed before the critical fracture energy values are reached
for the components. Therefore, a combined energy criterion is used to define the completion of
debonding:

(4.313)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
156 Published: 2023-10-23
 Cohesive Zone Material (CZM) Model

and

are, respectively, the normal and tangential fracture energies, and α is the power law exponent
for mixed-mode debonding (default =2). Verification of satisfaction of energy criterion can be
done during postprocessing of results.

Identifying Debonding Modes

The debonding modes are based on input data:

1. Mode I for normal data (input on TBDATA command as C1, C2, and C5).

2. Mode II for tangential data (input on TBDATA command as C3, C4, and C5).

3. Mixed-mode for normal and tangential data (input on TBDATA command as C1, C2, C3,
C4, C5, C6, and C7).

Artificial Damping

Debonding is accompanied by convergence difficulties in the Newton-Raphson solution. Artificial

damping is used in the numerical solution to overcome these problems. The debonding parameter
with viscous damping, dv is calculated using the debonding parameter d (dn or dt or dm) and the
viscous debonding parameter of the previous substep as:
(4.314)

where:

Δt = time increment
dVold = viscous debonding parameter at previous substep
η = damping coefficient (input on TBDATA command as C5 using TB,CZM).

This viscous regularized debonding parameter, dv, is used in the calculation of contact tractions
when damping is activated. This can result in traction and maximum separation values higher
than specified by the CZM law (TB,CZM). However, to ensure that the total debonding occurs
close to the specified maximum separation (or critical fracture energy), the surfaces are deemed
fully debonded and the interface behavior is changed to standard contact when the non-viscous
regularized debonding parameter reaches 1 (d = 1). It is recommended you set the damping
coefficient, η, (which has units of time) to a value much smaller than the minimum time step size
so that the maximum traction and maximum separation (or critical fracture energy) values are
not exceeded in the debonding calculations.

The energy dissipated due to damping (accessed via the AENE label on the ETABLE command)
should be compared to element potential energy. The energies can be output in the Job-
[Link] file (via the OUTPR command). You can also access the energies as follows:

• In POST1 via PRENERGY, PRESOL, PLESOL, and ETABLE commands

• In POST26 via ENERSOL and ESOL commands

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 157
Structures with Material Nonlinearities

The damping energy should be much smaller than the potential energy.

Tangential Slip Under Normal Compression

An option is provided to control tangential slip under compressive normal contact stress for
mixed-mode debonding. By default, no tangential slip is allowed for this case, but it can be activ-
ated by setting the flag β (input on TBDATA command as C6 using TB,CZM) to 1. Settings on β
are:

β = 0 (default) no tangential slip under compressive normal contact stress for mixed-
mode debonding
β = 1 tangential slip under compressive normal contact stress for mixed-mode debonding

Debonding under the Presence of Friction

When friction is defined between contact surfaces undergoing debonding, tangential stress is
calculated as the maximum between the tangential stress as governed by the debonding model
and the tangential stress as governed by the friction law.

Post Separation Behavior

After debonding is completed the surface interaction is governed by standard contact constraints
for normal and tangential directions. Frictional contact is used if friction is specified for contact
elements.

Results Output for POST1 and POST26

All applicable output quantities for contact elements are also available for debonding: normal
contact stress P (output as PRES), tangential contact stress τt (output as SFRIC) or its components
τ1 and τ2 (output as TAUR and TAUS), contact gap un (output as GAP), tangential slip ut (output
as SLIDE) or its components u1 and u2 (output as TASR and TASS), etc.

The following debonding-specific output quantities are also available (output as NMISC data):
debonding time history (output as DTSTART), debonding parameter dn, dt, or dm (output as
DPARAM), fracture energies Gn and Gt (output as DENERI and DENERII).

[Link].2. Bilinear Behavior - Interface (TBOPT = BILI)

The bilinear interface law described in Material Model – Bilinear Behavior (p. 146) (input via TB,CZM
with TBOPT = BILI) can be used with contact elements. This law differs from the bilinear model
described in Bilinear Behavior - Contact (TBOPT = CBDD and CBDE) (p. 153) in a few subtle ways.

• In this model, the response under the compressive traction is also governed by the traction
separation law. This differs from bilinear behavior (contact) where, under compressive
traction, the response is governed by standard contact constraints in the normal direction,
thus preventing penetration.

• The separation at which maximum traction is achieved is an input quantity in this law. In
the bilinear behavior (contact) model, this quantity is calculated based on contact stiffness
and maximum traction.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
158 Published: 2023-10-23
 Fluid Material Models

• The mixed mode behavior is also different for this law, with user control over the contri-
bution of normal and tangential separations to the equivalent slip (Equation 4.291 (p. 149)).

• The viscous regularization is also different, with this model requiring use of TB,VREG (see
Viscous Regularization (p. 150)).

Bilinear behavior (contact) is recommended over bilinear behavior (interface) for use with contact
elements since it is can prevent nonphysical penetration under compressive tractions.

Post Separation Behavior

After debonding is completed, the surface interaction is governed by standard contact constraints
for normal and tangential directions. Frictional contact is used if friction is specified for contact
elements.

Results Output for POST1 and POST26

All applicable output quantities for contact elements are also available for debonding: normal
contact stress P (output as PRES), tangential contact stress τt (output as SFRIC) or its components
τ1 and τ2 (output as TAUR and TAUS), contact gap un (output as GAP), tangential slip ut (output
as SLIDE) or its components u1 and u2 (output as TASR and TASS), etc.

The following debonding-specific output quantities are also available (output as NMISC data):
debonding time history (output as DTSTART), damage parameter dn, dt, or dm (output as DPARAM),
total debonding energy Gtotal (output as DENER).

[Link].3. Exponential Behavior

The exponential interface law described in Material Model – Exponential Behavior (p. 144) (input
via TB,CZM with TBOPT = EXPO) can be used with contact elements. The traction-separation re-
lationship is determined by the surface potential (Equation 4.276 (p. 144)) under all circumstances.
Viscous regularization as described in Viscous Regularization (p. 150) can be used to improve
convergence behavior.

Results Output for POST1 and POST26

All applicable output quantities for contact elements are also available for debonding: normal
contact stress P (output as PRES), tangential contact stress τt (output as SFRIC) or its components
τ1 and τ2 (output as TAUR and TAUS), contact gap un (output as GAP), tangential slip ut (output
as SLIDE) or its components u1 and u2 (output as TASR and TASS), etc.

The following debonding-specific output quantity is also available (output as NMISC data): total
debonding energy Gtotal (output as DENER).

4.12. Fluid Material Models

Fluid material models define constitutive relations for fluids used in structural analyses. These models
are available for hydrostatic fluid elements, HSFLD241 and HSFLD242.

For more details on the fluid material models presented here, see Fluids in the Material Reference.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 159
Structures with Material Nonlinearities

4.12.1. Liquid
Liquid material can be used to model a compressible linear (Newtonian) isotropic elastic non-viscous
fluid. Using the stress-strain relationship for linear elastic material as given in Structural Fundament-
als (p. 5), the constitutive equation for liquid can be written as:
(4.315)

with

(4.316)

(4.317)

where:

K = bulk modulus
εvolf = volumetric strain for fluid
ΔVf = change in fluid volume
V0f = initial fluid volume
P = current fluid pressure
P0 = initial fluid pressure (use the IC command)

The thermal properties of liquid material are defined by coefficient of linear thermal expansion, which
determines the change of fluid volume due to change of temperature:
(4.318)

where:

α = coefficient of linear thermal expansion

T = current fluid temperature
Tref = reference temperature (use TREF or MP command)

Initial fluid density (needed when prescribing fluid mass flow rate) can be specified using the
TBcommand with Lab = FLUID and TBOPT = LIQUID.

Change in fluid volume also affects fluid density. If fluid mass is kept fixed, current density for liquid
material can defined as:

(4.319)

where:

ρ0f = initial fluid density

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
160 Published: 2023-10-23
 Fluid Material Models

This material model is defined through theTB command with Lab = FLUID and TBOPT = LIQUID; in
addition, use the TBDATA command to define the bulk modulus, K, coefficient of linear thermal ex-
pansion, α, and initial fluid density, ρ0f, as material constants C1, C2, and C3, respectively.

4.12.2. Gas
The Ideal Gas Law is written as:
(4.320)

where:

ρ = density
P = pressure
R = universal gas constant
T = temperature

The Ideal Gas Law is used to define the constitutive equation for gas material:
(4.321)

where:

m = mass of fluid
R = universal gas constant
Tt = Toff + T = total fluid temperature
Toff = temperature offset from absolute zero to zero (use TOFFST command)
Pt = Pref + P = total fluid pressure
Pref = reference fluid pressure (specified as a real constant PREF)

Current density for gas material is defined in terms of initial density as:
(4.322)

where:

T0t = Toff + Tref = initial total temperature

Tref = reference temperature (use TREF or MP command)
P0t = Pref + P0 = initial total fluid pressure

To define a gas material, use theTB command with Lab = FLUID, TBOPT = GAS, and initial fluid
density, ρ0f, as material constant C1 on the TBDATA command. To completely define the initial state
of the gas material, also specify a reference pressure, Pref, as a real constant (use the R command) for
the hydrostatic fluid element, a temperature offset from absolute zero to zero temperature, Toff, (use
the TOFFST command), and a reference temperature, Tref (use the TREF or MP command).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 161
Structures with Material Nonlinearities

4.12.3. Pressure-Volume Data

For compressible fluids that do not follow the constitutive equation for liquid or Ideal Gas Law, a
pressure-volume curve may be used to define a pressure-volume relationship.

Current fluid density for this material model can be defined as:

(4.323)

where the current fluid volume is given by the pressure-volume curves based on the current fluid
pressure and temperature.

A pressure-volume curve can be defined through the TB command with Lab = FLUID and TBOPT =
PVDATA, along with the TBTEMP command to define temperature, and the TBPT command to define
pressure-volume data points for each temperature. The pressure-volume data points must be defined
in terms of total pressure and total volume of the fluid in the containing vessel.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
162 Published: 2023-10-23
Chapter 5: Electromagnetics
The following topics concerning electromagnetic are available:
5.1. Electromagnetic Field Fundamentals
5.2. Derivation of Electromagnetic Matrices
5.3. Electromagnetic Field Evaluations
5.4. Stranded Coil Analyses
5.5. Inductance, Flux and Energy Computation
5.6. Electromagnetic Particle Tracing
5.7. Capacitance Computation
5.8. Conductance Computation
5.9. Hall Effect

5.1. Electromagnetic Field Fundamentals

Electromagnetic fields are governed by the following Maxwell's equations (Smythe([151] (p. 933))):
(5.1)

(5.2)
(5.3)
(5.4)

where:

x = curl operator
= divergence operator
{H} = magnetic field intensity vector
{J} = total current density vector
{Js} = applied source current density vector
{Je} = induced eddy current density vector
{Jvs} = velocity current density vector
{D} = electric flux density vector (Maxwell referred to this as the displacement vector, but to
avoid misunderstanding with mechanical displacement, the name electric flux density is used
here.)
t = time
{E} = electric field intensity vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 163
Electromagnetics

{B} = magnetic flux density vector

ρ = electric charge density

The continuity equation follows from taking the divergence of both sides of Equation 5.1 (p. 163).
(5.5)

The continuity equation must be satisfied for the proper setting of Maxwell's equations. Users should
prescribe Js taking this into account.

The above field equations are supplemented by the constitutive relation that describes the behavior
of electromagnetic materials. For problems considering saturable material without permanent magnets,
the constitutive relation for the magnetic fields is:
(5.6)

where:

μ = magnetic permeability matrix, in general a function of {H}

The magnetic permeability matrix [μ] may be input either as a function of temperature or field. Specific-
ally, if [μ] is only a function of temperature,

(5.7)

where:

μo = permeability of free space (input on EMUNIT command)

μrx = relative permeability in the x-direction (input as MURX on MP command)

If [μ] is only a function of field,

(5.8)

where:

μh = permeability derived from the input B versus H curve (input with TB,BH).

Mixed usage is also permitted, e.g.:

(5.9)

When permanent magnets are considered, the constitutive relation becomes:

(5.10)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
164 Published: 2023-10-23
 Electromagnetic Field Fundamentals

where:

{Mo} = remanent intrinsic magnetization vector

Rewriting the general constitutive equation in terms of reluctivity it becomes:

(5.11)

where:

[ν] = reluctivity matrix = [μ]-1

The constitutive relations for the related electric fields are:

(5.12)
(5.13)

where:

σxx = conductivity in the x-direction (input as inverse of RSVX on MP command)

εxx = permittivity in the x-direction (input as PERX on MP command)

The solution of magnetic field problems is commonly obtained using potential functions. Two kinds of
potential functions, the magnetic vector potential and the magnetic scalar potential, are used depending
on the problem to be solved. Factors affecting the choice of potential include: field dynamics, field di-
mensionality, source current configuration, domain size and discretization.

The applicable regions are shown below. These will be referred to with each solution procedure discussed
below.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 165
Electromagnetics

Figure 5.1: Electromagnetic Field Regions

where:

Ω0 = free space region

Ω1 = nonconducting permeable region
Ω2 = conducting region
μ = permeability of iron
μo = permeability of air
Mo = permanent magnets
S1 = boundary of W1
σ = conductivity
Ω = Ω1 + Ω2 + Ω0

5.1.1. Magnetic Scalar Potential

The scalar potential method as implemented in SOLID5, SOLID96, and SOLID98 for 3D magnetostatic
fields is discussed in this section. Magnetostatics means that time varying effects are ignored. This
reduces Maxwell's equations for magnetic fields to:
(5.14)
(5.15)

5.1.2. Solution Strategies

In the domain Ω0 and Ω1 of a magnetostatic field problem (Ω2 is not considered for magnetostatics)
a solution is sought which satisfies the relevant Maxwell's Equation 5.14 (p. 166) and Equa-
tion 5.15 (p. 166) and the constitutive relation Equation 5.10 (p. 164) in the following form (Gy-
imesi([142] (p. 932)) and Gyimesi([150] (p. 933))):
(5.16)
(5.17)

where:

{Hg} = preliminary or "guess" magnetic field

φg = generalized potential

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
166 Published: 2023-10-23
 Electromagnetic Field Fundamentals

The development of {Hg} varies depending on the problem and the formulation. Basically, {Hg} must
satisfy Ampere's law (Equation 5.14 (p. 166)) so that the remaining part of the field can be derived as
the gradient of the generalized scalar potential φg. This ensures that φg is singly valued. Additionally,
the absolute value of {Hg} must be greater than that of Δφg. In other words, {Hg} should be a good
approximation of the total field. This avoids difficulties with cancellation errors (Gyimesi([150] (p. 933))).

This framework allows for a variety of scalar potential formulation to be used. The appropriate formu-
lation depends on the characteristics of the problem to be solved. The process of obtaining a final
solution may involve several steps (controlled by the MAGOPT solution option).

As mentioned above, the selection of {Hg} is essential to the development of any of the following
scalar potential strategies. The development of {Hg} always involves the Biot-Savart field {Hs} which
satisfies Ampere's law and is a function of source current {Js}. {Hs} is obtained by evaluating the integral:

(5.18)

where:

{Js} = current source density vector at d(volc)

{r} = position vector from current source to node point
volc = volume of current source

The above volume integral can be reduced to the following surface integral (Gyimesi et al.([175] (p. 934)))

(5.19)

where:

surfc = surface of the current source

Evaluation of this integral is automatically performed upon initial solution execution or explicitly
(controlled by the BIOT command). The values of {Js} are obtained either directly as input by:

SOURC36 - Current Source

or indirectly calculated by electric field calculation using:

SOLID5 - 3D Coupled-Field Solid

LINK68 - Coupled Thermal-Electric Line
SOLID98 - Tetrahedral Coupled-Field Solid
PLANE230 - 2D Electric Solid
SOLID231 or SOLID232 - 3D Electric Solids

Depending upon the current configuration, the integral given in Equation 5.19 (p. 167) is evaluated
in a closed form and/or a numerical fashion (Smythe([151] (p. 933))).

Three different solution strategies emerge from the general framework discussed above:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 167
Electromagnetics

Reduced Scalar Potential (RSP) Strategy

Difference Scalar Potential (DSP) Strategy
General Scalar Potential (GSP) Strategy

[Link]. RSP Strategy

Applicability

If there are no current sources ({Js} = 0), the RSP strategy is applicable. The RSP strategy is also ap-
plicable, in general, if there are current sources but there is no iron ([μ] = [μo]) within the problem
domain. This formulation is developed by Zienkiewicz([76] (p. 929)).

Procedure

The RSP strategy uses a one-step procedure (MAGOPT,0). Equation 5.16 (p. 166) and Equa-
tion 5.17 (p. 166) are solved making the following substitution:
(5.20)

Saturation is considered if the magnetic material is nonlinear. Permanent magnets are also con-
sidered.

[Link]. DSP Strategy

Applicability

The DSP strategy is applicable when current sources and singly connected iron regions exist within
the problem domain ({Js} ≠ {0}) and ([μ] ≠ [μo]). A singly connected iron region does not enclose
a current. In other words, a contour integral of {H} through the iron must approach zero as u → .
(5.21)

This formulation is developed by Mayergoyz([120] (p. 931)).

Procedure

The DSP strategy uses a two-step solution procedure. The first step (MAGOPT,2) makes the following
substitution into Equation 5.16 (p. 166) and Equation 5.17 (p. 166):
(5.22)

subject to:
(5.23)

This boundary condition is satisfied by using a very large value of permeability in the iron (internally
set by the program). Saturation and permanent magnets are not considered. This step produces a
near zero field in the iron region which is subsequently taken to be zero according to:
(5.24)

and in the air region:

(5.25)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
168 Published: 2023-10-23
 Electromagnetic Field Fundamentals

The second step (MAGOPT,3) uses the fields calculated on the first step as the preliminary field for
Equation 5.16 (p. 166) and Equation 5.17 (p. 166):
(5.26)
(5.27)

Here saturation and permanent magnets are considered. This step produces the following fields:
(5.28)

and
(5.29)

which are the final results to the applicable problems.

[Link]. GSP Strategy

Applicability

The GSP strategy is applicable when current sources ({Js ≠ {0}) exist within the problem domain in
conjunction with a multiply connected iron ([μ] ≠ [μo]) region. A multiply connected iron region
encloses some current source. This means that a contour integral of {H} through the iron region is
not zero:
(5.30)

where:

= refers to the dot product

This formulation is developed by Gyimesi([142] (p. 932), [150] (p. 933), [203] (p. 936)).

Procedure

The GSP strategy uses a three-step solution procedure. The first step (MAGOPT,1) performs a
solution only in the iron with the following substitution into Equation 5.16 (p. 166) and Equa-
tion 5.17 (p. 166):
(5.31)

subject to:
(5.32)

Here S1 is the surface of the iron air interface. Saturation can optimally be considered for an improved
approximation of the generalized field but permanent magnets are not. The resulting field is:
(5.33)

The second step (MAGOPT,2) performs a solution only in the air with the following substitution
into Equation 5.16 (p. 166) and Equation 5.17 (p. 166):
(5.34)

subject to:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 169
Electromagnetics

(5.35)

This boundary condition is satisfied by automatically constraining the potential solution φg at the
surface of the iron to be what it was on the first step (MAGOPT,1). This step produces the following
field:
(5.36)

Saturation or permanent magnets are of no consequence since this step obtains a solution only in
air.

The third step (MAGOPT,3) uses the fields calculated on the first two steps as the preliminary field
for Equation 5.16 (p. 166) and Equation 5.17 (p. 166):
(5.37)
(5.38)

Here saturation and permanent magnets are considered. The final step allows for the total field to
be computed throughout the domain as:
(5.39)

5.1.3. Magnetic Vector Potential

The vector potential method, implemented in PLANE13 and PLANE233 for 2D electromagnetic fields,
is discussed in this section. Considering static and dynamic fields and neglecting displacement currents
(quasi-stationary limit), the following subset of Maxwell's equations apply:
(5.40)
(5.41)
(5.42)

The usual constitutive equations for magnetic and electric fields apply as described by Equa-
tion 5.11 (p. 165) and Equation 5.12 (p. 165). Although some restrictions on anisotropy and nonlinearity
do occur in the formulations mentioned below.

In the entire domain, Ω, of an electromagnetic field problem a solution is sought which satisfies the
relevant Maxwell's Equation 5.40 (p. 170) thru Equation 5.41 (p. 170). See Figure 5.1: Electromagnetic
Field Regions (p. 166) for a representation of the problem domain Ω.

A solution can be obtained by introducing potentials which allow the magnetic field {B} and the
electric field {E} to be expressed as (Biro([121] (p. 931))):
(5.43)
(5.44)

where:

{A} = magnetic vector potential

V = electric scalar potential

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
170 Published: 2023-10-23
 Electromagnetic Field Fundamentals

These specifications ensure the satisfaction of two of Maxwell's equations, Equation 5.41 (p. 170) and
Equation 5.42 (p. 170). What remains to be solved is Ampere's law, Equation 5.40 (p. 170) in conjunction
with the constitutive relations, Equation 5.11 (p. 165), and the divergence free property of current
density. The resulting differential equations are:

(5.45)

(5.46)

(5.47)

These equations are subject to the appropriate boundary conditions.

This system of simplified Maxwell's equations with the introduction of potential functions is used for
the solutions of 2D static and dynamic fields. Silvester([73] (p. 928)) presents a 2D static formulation.
Chari([70] (p. 928)), Brauer([71] (p. 928)) and Tandon([72] (p. 928)) discuss the 2D eddy current problem
and Weiss([95] (p. 930)) and Garg([96] (p. 930)) discuss 2D eddy current problems which allow for skin
effects (eddy currents present in the source conductor).

5.1.4. Edge-Based Magnetic Vector Potential

An edge-based method, using edge-based elements with a discontinuous normal component of
magnetic vector potential, is implemented in the 3D electromagnetic SOLID236 and SOLID237 elements.

The differential equations governing SOLID236 and SOLID237 elements are the following:

(5.48)

(5.49)

(5.50)

These equations are subject to the appropriate magnetic and electrical boundary conditions.

The uniqueness of edge-based magnetic vector potential is ensured by the tree gauging procedure
(GAUGE command) that sets the edge-flux degrees of freedom corresponding to the spanning tree
of the finite element mesh to zero.

5.1.5. Harmonic Analysis Using Complex Formalism

In a general dynamic problem, any field quantity, q(r,t) depends on the space, r, and time, t, variables.
In a harmonic analysis, the time dependence can be described by periodic functions:
(5.51)

or
(5.52)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 171
Electromagnetics

r = location vector in space

t = time
ω = angular frequency of time change.
a(r) = amplitude (peak)
φ(r) = phase angle
c(r) = measurable field at ωt = 0 degrees
s(r) = measurable field at ωt = -90 degrees

In an electromagnetic analysis, q(r,t) can be the flux density, {B}, the magnetic field, {H}, the electric
field, {E}, the current density, J, the vector potential, {A}, or the scalar potential, V. Note, however, that
q(r,t) can not be the Joule heat, Qj, the magnetic energy, W, or the force, Fjb, because they include a
time-constant term.

The quantities in Equation 5.51 (p. 171) and Equation 5.52 (p. 171) are related by
(5.53)
(5.54)
(5.55)
(5.56)

In Equation 5.51 (p. 171)) a(r), φ(r), c(r) and s(r) depend on space coordinates but not on time. This
separation of space and time is taken advantage of to minimize the computational cost. The originally
4 (3 space + 1 time) dimensional real problem can be reduced to a 3 (space) dimensional complex
problem. This can be achieved by the complex formalism.

The measurable quantity, q(r,t), is described as the real part of a complex function:
(5.57)

Q(r) is defined as:

(5.58)

where:

j = imaginary unit
Re { } = denotes real part of a complex quantity
Qr(r) and Qi(r) = real and imaginary parts of Q(r). Note that Q depends only on the space
coordinates.

The complex exponential in Equation 5.57 (p. 172) can be expressed by sine and cosine as
(5.59)

Substituting Equation 5.59 (p. 172) into Equation 5.57 (p. 172) provides Equation 5.58 (p. 172)
(5.60)

Comparing Equation 5.51 (p. 171) with Equation 5.60 (p. 172) reveals:
(5.61)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
172 Published: 2023-10-23
 Electromagnetic Field Fundamentals

(5.62)

In words, the complex real, Qr(r), and imaginary, Qi(r), parts are the same as the measurable cosine,
c(r), and sine, s(r), amplitudes.

A harmonic analysis provides two solution sets: the real and imaginary components of a complex
solution. According to Equation 5.51 (p. 171), and Equation 5.61 (p. 172) the magnitude of the real and
imaginary sets describe the measurable field at t = 0 and at ωt = -90 degrees, respectively. Comparing
Equation 5.52 (p. 171) and Equation 5.61 (p. 172) provides:
(5.63)

(5.64)

Equation 5.63 (p. 173) expresses the amplitude (peak) and phase angle of the measurable harmonic
field quantities by the complex real and imaginary parts.

The time average of harmonic fields such as A, E, B, H, J, or V is zero at point r. This is not the case
for P, W, or F because they are quadratic functions of B, H, or J. To derive the time dependence of a
quadratic function - for the sake of simplicity - we deal only with a Lorentz force, F, which is product
of J and B. (This is a cross product but, for simplicity, the components are not shown. The space de-
pendence is also omitted.)

(5.65)

where:

Fjb = Lorentz Force density (output as FMAG on PRESOL command)

The time average of cos2 and sin2 terms is 1/2 whereas that of the sin cos term is zero. Therefore, the
time average force density is:
(5.66)

Thus, the force can be obtained as the sum of "real" and "imaginary" forces. In a similar manner the
time averaged Joule power density, Qj, and magnetic energy density, W, can be obtained as:
(5.67)
(5.68)

5.1.6. Electric Scalar Potential

Neglecting the time-derivative of magnetic flux density (the quasistatic approximation), the
system of Maxwell's equations (Equation 5.1 (p. 163) through Equation 5.4 (p. 163)) reduces to:
(5.69)
(5.70)
(5.71)
(5.72)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 173
Electromagnetics

As follows from Equation 5.70 (p. 173), the electric field {E} is irrotational, and can be derived from:
(5.73)

where:

V = electric scalar potential

In the time-varying electromagnetic field governed by Equation 5.69 (p. 173) through Equa-
tion 5.72 (p. 173), the electric and magnetic fields are uncoupled. If only the electric solution is of in-
terest, replacing Equation 5.69 (p. 173) by the continuity Equation 5.5 (p. 164) and eliminating Equa-
tion 5.71 (p. 173) produces the system of differential equations governing the quasistatic electric field.

Repeating Equation 5.12 (p. 165) and Equation 5.13 (p. 165) without velocity effects, the constitutive
equations for the electric fields become:
(5.74)
(5.75)

where:

ρxx = resistivity in the x-direction (input as RSVX on MP command)

εxx = permittivity in the x-direction (input as PERX on MP command)

The conditions for {E}, {J}, and {D} on an electric material interface are:
(5.76)

(5.77)

(5.78)

where:

Et1, Et2 = tangential components of {E} on both sides of the interface

Jn1, Jn2 = normal components of {J} on both sides of the interface
Dn1, Dn2 = normal components of {D} on both sides of the interface
ρs = surface charge density

Two cases of the electric scalar potential approximation are considered below.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
174 Published: 2023-10-23
 Electromagnetic Field Fundamentals

[Link]. Quasistatic Electric Analysis

In this analysis, the relevant governing equations are Equation 5.73 (p. 174) and this continuity
equation:
(5.79)

Substituting the constitutive Equation 5.74 (p. 174) and Equation 5.75 (p. 174) into Equation 5.79 (p. 175),
and taking into account Equation 5.73 (p. 174), one obtains the differential equation for electric
scalar potential:
(5.80)

Equation 5.80 (p. 175) is used to approximate a time-varying electric field in elements PLANE230,
SOLID231, and SOLID232. It takes into account both the conductive and dielectric effects in electric
materials. Neglecting time-variation of the electric potential, Equation 5.80 (p. 175) reduces to the
governing equation for steady-state electric conduction:
(5.81)

In the case of a time-harmonic electric field analysis, the complex formalism allows Equa-
tion 5.80 (p. 175) to be re-written as:
(5.82)

where:

ω = angular frequency

Equation 5.82 (p. 175) is the governing equation for a time-harmonic electric analysis using elements
PLANE121, SOLID122, and SOLID123.

In a time-harmonic analysis, the loss tangent tan δ can be used instead of or in addition to the
electrical conductivity [σ] to characterize losses in dielectric materials. In this case, the conductivity
matrix [σ] is replaced by the effective conductivity [σeff] defined as:
(5.83)

where:

tan δ = loss tangent (input as LSST on MP command)

[Link]. Electrostatic Analysis

Electric scalar potential equation for electrostatic analysis is derived from governing Equa-
tion 5.72 (p. 173) and Equation 5.73 (p. 174), and constitutive Equation 5.75 (p. 174):
(5.84)

Equation 5.84 (p. 175), subject to appropriate boundary conditions, is solved in an electrostatic field
analysis of dielectrics using elements PLANE121, SOLID122, and SOLID123.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 175
Electromagnetics

5.2. Derivation of Electromagnetic Matrices

The finite element matrix equations can be derived by variational principles. These equations exist for
linear and nonlinear material behavior as well as static and transient response. Based on the presence
of linear or nonlinear materials (as well as other factors), the program chooses the appropriate Newton-
Raphson method. The user may select another method with the (NROPT command (see Newton-
Raphson Procedure (p. 730))). When transient affects are to be considered a first order time integration
scheme must be involved (TIMINT command (see Transient Analysis (p. 796))).

5.2.1. Magnetic Scalar Potential

The scalar potential formulations are restricted to static field analysis with partial orthotropic nonlinear
permeability. The degrees of freedom (DOFs), element matrices, and load vectors are presented here
in the following form (Zienkiewicz([76] (p. 929)), Chari([74] (p. 928)), and Gyimesi([142] (p. 932))):

[Link]. Degrees of freedom

{φe} = magnetic scalar potentials at the nodes of the element (input/output as MAG)

[Link]. Coefficient Matrix

(5.85)

(5.86)

(5.87)

[Link]. Applied Loads

(5.88)

where:

{N} = element shape functions (φ = {N}T{φe})

vol = volume of the element

{Hg} = preliminary or "guess" magnetic field (see Electromagnetic Field Fundamentals (p. 163))
{Hc} = coercive force vector (input as MGXX, MGYY, MGZZ on MP command))
[μ] = permeability matrix (derived from input material property MURX, MURY, and MURZ
(MP command) and/or material curve B versus H (accessed with TB,BH))(see Equa-
tion 5.7 (p. 164), Equation 5.8 (p. 164), and Equation 5.9 (p. 164))

= derivative of permeability with respect to magnitude of the magnetic field intensity

(derived from the input material property curve B versus H (accessed with TB,BH))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
176 Published: 2023-10-23
 Derivation of Electromagnetic Matrices

The material property curve is input in the form of B values versus H values and is then converted
to a spline fit curve of μ versus H from which the permeability terms μh and are evaluated.

The coercive force vector is related to the remanent intrinsic magnetization vector as:
(5.89)

where:

μo = permeability of free space (input as MUZRO on EMUNIT command)

The Newton-Raphson solution technique (Option on the NROPT command) is necessary for nonlinear
analyses. Adaptive descent is also recommended (Adaptky on the NROPT command). When adaptive
descent is used Equation 5.85 (p. 176) becomes:
(5.90)

where:

ξ = descent parameter (see Newton-Raphson Procedure (p. 730))

5.2.2. Magnetic Vector Potential

The vector potential formulation is applicable to both static and dynamic fields with partial orthotropic
nonlinear permeability. The basic equation to be solved is of the form:
(5.91)

The terms of this equation are defined below (Biro([121] (p. 931))).

[Link]. Degrees of Freedom

(5.92)

where:

= magnetic vector potentials (input/output as AZ)

= time integrated electric scalar potential (input/output as VOLT)

The VOLT degree of freedom is a time integrated electric potential to allow for symmetric matrices.

[Link]. Coefficient Matrices

(5.93)

(5.94)

(5.95)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 177
Electromagnetics

(5.96)

(5.97)

(5.98)

(5.99)

(5.100)

(5.101)

[Link]. Applied Loads

(5.102)

(5.103)

(5.104)

(5.105)

(5.106)

where:

[NA] = matrix of element shape functions for {A}

[N] = vector of element shape functions for {V} (V = {N}T{ve})

{Js} = source current density vector (input as JS on BFE command)
{Jt} = total current density vector (input as JS on BFE command)
vol = volume of the element
{Hc} = coercive force vector (input as MGXX, MGYY, MGZZ on MP command)
νo = reluctivity of free space (derived from value using MUZRO on EMUNIT command)
[ν] = partially orthotropic reluctivity matrix (inverse of [μ], derived from input material
property curve B versus H (input using TB,BH command))

= derivative of reluctivity with respect to the magnitude of magnetic flux squared

(derived from input material property curve B versus H (input using TB,BH command))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
178 Published: 2023-10-23
 Derivation of Electromagnetic Matrices

[σ] = orthotropic conductivity (input as RSVX, RSVY, RSVZ on MP command (inverse)) (see
Equation 5.12 (p. 165)).
{v} = velocity vector

The coercive force vector is related to the remanent intrinsic magnetization vector as:
(5.107)

The material property curve is input in the form of B values versus H values and is then converted
to a spline fit curve of ν versus |B|2 from which the isotropic reluctivity terms νh and are
evaluated.

The above element matrices and load vectors are presented for the most general case of a vector
potential analysis. Many simplifications can be made depending on the conditions of the specific
problem. Restricting the formulation to 2D, there is only one component of the vector potential
(AZ).

Combining some of the above equations, the variational equilibrium equations may be written as:
(5.108)
(5.109)

Here T denotes transposition.

Static analyses require only the magnetic vector potential degrees of freedom (KEYOPT controlled)
and the K coefficient matrices. If the material behavior is nonlinear then the Newton-Raphson
solution procedure is required (Option on the NROPT command (see Newton-Raphson Proced-
ure (p. 730))).

For 2D dynamic analyses, a current density load of either source ({Js}) or total {Jt} current density is
valid. Jt input represents the impressed current expressed in terms of a uniformly applied current
density. This loading is only valid in a skin-effect analysis with proper coupling of the VOLT degrees
of freedom. The electric scalar potential must be constrained properly in order to satisfy the funda-
mentals of electromagnetic field theory. This can be achieved by direct specification of the potential
value (using the D command) as well as with coupling and constraining (using the CP and CE
commands).

The general transient analysis (ANTYPE,TRANS accepts nonlinear material behavior and permanent
magnets (MGXX, MGYY, MGZZ). Harmonic transient analyses (ANTYPE,HARMIC (see Harmonic
Analysis (p. 813))) is a linear analyses with sinusoidal loads; therefore, it is restricted to linear mater-
ial behavior without permanent magnets.

5.2.3. Edge-Based Magnetic Vector Potential

The following section describes the derivation of the electromagnetic finite element equations used
by SOLID236 and SOLID237 elements.

In an edge-based electromagnetic analysis, the magnetic vector potential {A} is approximated using
the edge-based shape functions:
(5.110)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 179
Electromagnetics

where:

[W] = matrix of element vector (edge-based) shape functions.

{Ae} = edge-flux at the element mid-side nodes (input/output as AZ). Edge-flux is defined as the line

integral = of the magnetic vector potential along the element edge L.

The electric scalar potential V is approximated using scalar (node-based) element shape functions:
(5.111)

where:

{N} = vector of element scalar (node-based) shape functions,

{Ve} = electric scalar potential at the element nodes (input/output as VOLT).

Applying the variational principle to the governing electromagnetic equations (see Equation 5.48 (p. 171)
- Equation 5.50 (p. 171)), we obtain the system of finite element equations:

(5.112)

where:

= element magnetic reluctivity matrix,

= element linear magnetic

reluctivity matrix,

= element nonlinear magnetic reluctivity

matrix,

= element electric conductivity matrix,

= element magneto-electric coupling matrix,

= element electromagnetic coupling matrix,

= element eddy current damping matrix,

= element displacement current damping matrix,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
180 Published: 2023-10-23
 Derivation of Electromagnetic Matrices

= element magneto-dielectric coupling matrix,

= element displacement current mass matrix,

= element source current density vector,

= element remnant magnetization load vector,

vol = element volume,

[ν] = reluctivity matrix (inverse of the magnetic permeability matrix input as MURX, MURY, MURZ on
MP command or derived from the B-H curve input on TB command),

= derivative of reluctivity with respect to the magnitude of magnetic flux squared (derived from
the B-H curve (input via TB,BH command))

[σ] = electrical conductivity matrix (inverse of the electrical resistivity matrix input as RSVX, RSVY,
RSVZ on MP command),

[ε]= dielectric permittivity (input as PERX, PERY, PERZ on MP command) (applicable to a harmonic
electromagnetic analysis (KEYOPT(1) = 1) only),

{v} = velocity vector (input as VELO on BF command) (applicable to electromagnetic analysis option
(KEYOPT(1) = 1) only),

{Js} = source current density vector (input as JS on BFE command) (applicable to the stranded con-
ductor analysis option (KEYOPT(1) = 0) only),

{Hc} = coercive force vector (input as MGXX, MGYY, MGZZ on MP command),

{Ie} = nodal current vector (input/output as AMPS).

Equation 5.112 (p. 180) describes the strong coupling between the magnetic edge-flux and the electric
potential degrees of freedom is nonsymmetric. It can be made symmetric by either using the weak
coupling option (KEYOPT(2) = 1) in static or transient analyses or using the time-integrated electric
potential (KEYOPT(2) = 2) in transient or harmonic analyses. In the latter case, the VOLT degree of

freedom has the meaning of the time-integrated electric scalar potential , and Equa-
tion 5.112 (p. 180) becomes:

(5.113)

5.2.4. Electric Scalar Potential

The electric scalar potential V is approximated over the element as follows:
(5.114)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 181
Electromagnetics

where:

{N} = element shape functions

{Ve} = nodal electric scalar potential (input/output as VOLT)

[Link]. Quasistatic Electric Analysis

The application of the variational principle and finite element discretization to the differential
Equation 5.80 (p. 175) produces the matrix equation of the form:
(5.115)

where:

vol = element volume

[σeff] = "effective" conductivity matrix (defined by Equation 5.83 (p. 175))
{Ie} = nodal current vector (input/output as AMPS)

Equation 5.115 (p. 182) is used in the finite element formulation of PLANE230, SOLID231, and SOL-
ID232. These elements model both static (steady-state electric conduction) and dynamic (time-
transient and time-harmonic) electric fields. In the former case, matrix [Cv] is ignored.

A time-harmonic electric analysis can also be performed using elements PLANE121, SOLID122, and
SOLID123. In this case, the variational principle and finite element discretization are applied to the
differential Equation 5.82 (p. 175) to produce:
(5.116)

where:

[Link]. Electrostatic Analysis

The matrix equation for an electrostatic analysis using elements PLANE121, SOLID122, and SOLID123
is derived from Equation 5.84 (p. 175):
(5.117)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
182 Published: 2023-10-23
 Electromagnetic Field Evaluations

{ρ} = charge density vector (input as CHRGD on BF command)

{ρs} = surface charge density vector (input as CHRGS on SF command)

5.3. Electromagnetic Field Evaluations

The basic magnetic analysis results include magnetic field intensity, magnetic flux density, magnetic
forces and current densities. These types of evaluations are somewhat different for magnetic scalar and
vector formulations. The basic electric analysis results include electric field intensity, electric current
densities, electric flux density, Joule heat and stored electric energy.

5.3.1. Magnetic Scalar Potential Results

The first derived result is the magnetic field intensity which is divided into two parts (see Electromag-
netic Field Fundamentals (p. 163)); a generalized field {Hg} and the gradient of the generalized potential
- ϕg. This gradient (referred to here as {Hϕ) is evaluated at the integration points using the element
shape function as:
(5.118)

where:

{N} = shape functions

{ωg} = nodal generalized potential vector

The magnetic field intensity is then:

(5.119)

where:

{H} = magnetic field intensity (output as H)

Then the magnetic flux density is computed from the field intensity:
(5.120)

where:

{B} = magnetic flux density (output as B)

[μ] = permeability matrix (defined in Equation 5.7 (p. 164), Equation 5.8 (p. 164), and Equa-
tion 5.9 (p. 164))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 183
Electromagnetics

Nodal values of field intensity and flux density are computed from the integration points values as
described in Nodal and Centroidal Data Evaluation (p. 442).

Magnetic forces are also available and are discussed below.

5.3.2. Magnetic Vector Potential Results

The magnetic flux density is the first derived result. It is defined as the curl of the magnetic vector
potential. This evaluation is performed at the integration points using the element shape functions:
(5.121)

where:

{B} = magnetic flux density (output as B)

x = curl operator
[NA] = shape functions
{Ae} = nodal magnetic vector potential

Then the magnetic field intensity is computed from the flux density:
(5.122)

where:

{H} = magnetic field intensity (output as H)

[ν] = reluctivity matrix

Nodal values of field intensity and flux density are computed from the integration point value as de-
scribed in Nodal and Centroidal Data Evaluation (p. 442).

Magnetic forces are also available and are discussed below.

For a vector potential transient analysis current densities are also calculated.
(5.123)

where:

{Jt} = total current density

(5.124)

where:

{Je} = current density component due to {A}

[σ] = conductivity matrix
n = number of integration points
[NA] = element shape functions for {A} evaluated at the integration points

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
184 Published: 2023-10-23
 Electromagnetic Field Evaluations

{Ae} = time derivative of magnetic vector potential

and

(5.125)

where:

{Js} = current density component due to V

= divergence operator
{Ve} = electric scalar potential
{N} = element shape functions for V evaluated at the integration points

and
(5.126)

where:

{Jv} = velocity current density vector

{v} = applied velocity vector
{B} = magnetic flux density (see Equation 5.121 (p. 184))

5.3.3. Edge-Based Magnetic Vector Potential

The following section describes the results derived from an edge-based electromagnetic analysis using
SOLID236 and SOLID237 elements.

The electromagnetic fields and fluxes are evaluated at the integration points as follows:
(5.127)
(5.128)

(5.129)

(5.130)

(5.131)

where:

{B} = magnetic flux density (output as B at the element nodes),

{H} = magnetic field intensity (output as H at the element nodes),

{E} = electric field intensity (output as EF at the element nodes),

{Jc} = conduction current density (output as JC at the element nodes and as JT at the element centroid),

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 185
Electromagnetics

{Js} = total (conduction + displacement) current density (output as JS at the element centroid; same
as JT in a static or transient analysis),

{Ae}= edge-flux at the element mid-side nodes (input/output as AZ),

{Ve} = electric scalar potential at the element nodes (input/output as VOLT),

[W] = matrix of element vector (edge-based) shape functions,

{N} = vector of element scalar (node-based) shape functions,

[ν] = reluctivity matrix (inverse of the magnetic permeability matrix (input as MURX, MURY, MURZ on
MP command or derived from the B-H curve input on TB command),

[σ] = electrical conductivity matrix (inverse of the electrical resistivity matrix input as RSVX, RSVY,
RSVZ on MP command),

[ε] = dielectric permittivity (input as PERX, PERY, PERZ on MP command) (applicable to a harmonic
electromagnetic analysis (KEYOPT(1) = 1) only).

{ν} = velocity vector (input as VELO on BF command) (applicable to electromagnetic analyses (KEY-
OPT(1) = 1) only)

Nodal values of the above quantities are computed from the integration point values as described
in Nodal and Centroidal Data Evaluation (p. 442).

5.3.4. Magnetic Forces

Magnetic forces are computed by elements using the vector potential method (PLANE13, PLANE223,
and PLANE233), the edge-based magnetic vector potential method (SOLID226, SOLID227, SOLID236
and SOLID237), and the scalar potential method (SOLID5, SOLID96, and SOLID98). Three different
techniques are used to calculate magnetic forces at the element level.

[Link]. Lorentz forces

Magnetic forces in current carrying conductors (element output quantity FJB) are numerically integ-
rated using:
(5.132)

where:

{N} = vector of shape functions

For a 2D analysis, the corresponding electromagnetic torque about +Z is given by:

(5.133)

where:

{Z} = unit vector along +Z axis

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
186 Published: 2023-10-23
 Electromagnetic Field Evaluations

{r} = position vector in the global Cartesian coordinate system

In a time-harmonic analysis, the time-averaged Lorentz force and torque are computed by:
(5.134)

and
(5.135)

respectively.

where:

{J}* = complex conjugate of {J}

[Link]. Maxwell Forces

The Maxwell stress tensor is used to determine forces on ferromagnetic regions. Depending on
whether the magnetic forces are derived from the Maxwell stress tensor using surface or volumetric
integration, one distinguishes between the surface and the volumetric integral methods.

[Link].1. Surface Integral Method

This method is used by PLANE13, SOLID5, SOLID96, and SOLID98 elements.

The force calculation is performed on surfaces of air material elements which have a nonzero
face loading specified (MXWF on SF commands) (Moon([78] (p. 929))). For the 2D application, this
method uses extrapolated field values and results in the following numerically integrated surface
integral:

(5.136)

where:

{Fmx} = Maxwell force (output as FMX)

μo = permeability of free space (input on EMUNIT command)

T12 = Bx By
T21 = Bx By

3D applications are an extension of the 2D case.

For a 2D analysis, the corresponding electromagnetic torque about +Z axis is given by:
(5.137)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 187
Electromagnetics

where:

= unit surface normal in the global Cartesian coordinate system

In a time-harmonic analysis, the time-averaged Maxwell stress tensor force and torque are com-
puted by:

(5.138)

and

(5.139)

respectively.

where:

{B}* = complex conjugate of {B}

Re{ } = denotes real part of a complex quantity

[Link].2. Volumetric Integral Method

This method is used by PLANE223, SOLID226, SOLID227, PLANE233, SOLID236, and SOLID237
elements with KEYOPT(8) = 0.

The Maxwell forces are calculated by the following volumetric integral:

(5.140)

where:

= element magnetic Maxwell forces (output as FMAG at all the element nodes with KEY-
OPT(7) = 0 or at the element corner nodes only with KEYOPT(7) = 1),

[B] = strain-displacement matrix

{Tmx} = Maxwell stress vector = {T11 T22 T33 T12 T23 T13}T

The EMFT macro can be used with this method to sum up Maxwell forces and torques.

[Link]. Virtual Work Forces

Electromagnetic nodal forces (including electrostatic forces) are calculated using the virtual work
principle. The two formulations currently used for force calculations are the element shape method
(magnetic forces) and nodal perturbations method (electromagnetic forces).

[Link].1. Element Shape Method

Magnetic forces calculated using the virtual work method (element output quantity FVW) are
obtained as the derivative of the energy versus the displacement (MVDI on BF commands) of

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
188 Published: 2023-10-23
 Electromagnetic Field Evaluations

the movable part. This calculation is valid for a layer of air elements surrounding a movable part
(Coulomb([77] (p. 929))). To determine the total force acting on the body, the forces in the air
layer surrounding it can be summed. The basic equation for force of an air material element in
the s direction is:

(5.141)

where:

Fs = force in element in the s direction

s = virtual displacement of the nodal coordinates taken alternately to be in the X, Y, Z

global directions
vol = volume of the element

For a 2D analysis, the corresponding electromagnetic torque about +Z axis is given by:
(5.142)

In a time-harmonic analysis, the time-averaged virtual work force and torque are computed by:

(5.143)

and

(5.144)

respectively.

[Link].2. Nodal Perturbation Method

This method is used by PLANE121, SOLID122 and SOLID123 elements.

Electromagnetic (both electric and magnetic) forces are calculated as the derivatives of the total
element coenergy (sum of electrostatic and magnetic coenergies) with respect to the element
nodal coordinates (Gyimesi et al.([343] (p. 944))):

(5.145)

where:

Fxi = x-component (y- or z-) of electromagnetic force calculated in node i

xi = nodal coordinate (x-, y-, or z-coordinate of node i)
vol = volume of the element

Nodal electromagnetic forces are calculated for each node in each element. In an assembled
model the nodal forces are added up from all adjacent to the node elements. The nodal perturb-

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 189
Electromagnetics

ation method provides consistent and accurate electric and magnetic forces (using the EMFT
command macro).

5.3.5. Joule Heat in a Magnetic Analysis

Joule heat is computed by elements using the vector potential method (PLANE13, SOLID236, and
SOLID237) if the element has a nonzero resistivity (material property RSVX) and a nonzero current
density (either applied Js or resultant Jt). It is available as the output power loss (output as JHEAT) or
as the coupled field heat generation load (LDREAD,HGEN).

Joule heat per element is computed as:

1. Static or Transient Magnetic Analysis

(5.146)

where:

Qj = Joule heat per unit volume

n = number of integration points
[ρ] = resistivity matrix (input as RSVX, RSVY, RSVZ on MP command)
{Jti} = total current density in the element at integration point i

2. Harmonic Magnetic Analysis

(5.147)

where:

Re = real component
{Jti} = complex total current density in the element at integration point i
{Jti}* = complex conjugate of {Jti}

5.3.6. Electric Scalar Potential Results

The first derived result in this analysis is the electric field. By definition (Equation 5.73 (p. 174)), it is
calculated as the negative gradient of the electric scalar potential. This evaluation is performed at
the integration points using the element shape functions:
(5.148)

Nodal values of electric field (output as EF) are computed from the integration points values as de-
scribed in Nodal and Centroidal Data Evaluation (p. 442). The derivation of other output quantities
depends on the analysis types described below.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
190 Published: 2023-10-23
 Electromagnetic Field Evaluations

[Link]. Quasistatic Electric Analysis

The conduction current and electric flux densities are computed from the electric field (see Equa-
tion 5.74 (p. 174) and Equation 5.75 (p. 174)):
(5.149)
(5.150)

where:

Both the conduction current {J} and electric flux {D} densities are evaluated at the integration point
locations; however, whether these values are then moved to nodal or centroidal locations depends
on the element type used to do a quasistatic electric analysis:

• In a current-based electric analysis using elements PLANE230, SOLID231, and SOLID232, the
conduction current density is stored at both the nodal (output as JC) and centoidal (output as
JT) locations. The electric flux density vector components are stored at the element centroidal
location and output as nonsummable miscellaneous items;

• In a charge-based analysis using elements PLANE121, SOLID122, and SOLID123 (harmonic analysis),
the conduction current density is stored at the element centroidal location (output as JT), while
the electric flux density is moved to the nodal locations (output as D).

The total electric current {Jtot} density is calculated as a sum of conduction {J} and displacement
current densities:
(5.151)

The total electric current density is stored at the element centroidal location (output as JS). It can
be used as a source current density in a subsequent magnetic analysis (LDREAD,JS).

The Joule heat is computed from the centroidal values of electric field and conduction current
density. In a steady-state or transient electric analysis, the Joule heat is calculated as:
(5.152)

where:

Q = Joule heat generation rate per unit volume (output as JHEAT)

In a harmonic electric analysis, the Joule heat generation value per unit volume is time-averaged
over a one period and calculated as:
(5.153)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 191
Electromagnetics

Re = real component
{E}* = complex conjugate of {E}

The value of Joule heat can be used as heat generation load in a subsequent thermal analysis
(LDREAD,HGEN).

In a transient electric analysis, the element stored electric energy is calculated as:

(5.154)

where:

W = stored electric energy (output as SENE)

In a harmonic electric analysis, the time-averaged electric energy is calculated as:

(5.155)

[Link]. Electrostatic Analysis

The derived results in an electrostatic analysis are:

Electric field (see Equation 5.148 (p. 190)) at nodal locations (output as EF);
Electric flux density (see Equation 5.150 (p. 191)) at nodal locations (output as D);
Element stored electric energy (see Equation 5.154 (p. 192)) output as SENE

Electrostatic forces are also available and are discussed below.

5.3.7. Electrostatic Forces

Electrostatic forces are determined using the nodal perturbation method (recommended) described
in Nodal Perturbation Method (p. 189) or the Maxwell stress tensor described here. This force calculation
is performed on surfaces of elements which have a nonzero face loading specified (MXWF on SF
commands). For the 2D application, this method uses extrapolated field values and results in the
following numerically integrated surface integral:

(5.156)

where:

εo = free space permittivity (input as PERX on MP command)

T12 = Ex Ey
T21 = Ey Ex

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
192 Published: 2023-10-23
 Stranded Coil Analyses

n1 = component of unit normal in x-direction

n2 = component of unit normal in y-direction
s = surface area of the element face

3D applications are an extension of the 2D case.

5.4. Stranded Coil Analyses

In the magnetic vector potential formulations discussed in previous sections, the source current density
{Js} is either known and input as BFE,,JS (stranded conductor analysis) or determined from a fully-coupled
electromagnetic solution as a sum of eddy Je (Equation 5.124 (p. 184)) and DC conduction Js (Equa-
tion 5.125 (p. 185)) current densities (massive conductor analysis).

A stranded coil refers to a coil consisting of many fine turns of conducting wires. Because in the low-
frequency approximation the cross-section of the wires is small compared to the skin depth, the eddy
currents in the wire can be neglected and the magnitude of the current density within the wires can
be considered constant. The coil can be energized by an applied voltage or current, or by a controlling
electric circuit. The stranded coil analysis can be viewed as a special case of electromagnetic field ana-
lysis where the current flow direction in wires is known and determined by the winding, while the
magnitude of the current in the coil or voltage drop across the coil can either be imposed or determined
from the solution.

The following sections describe the magnetic and electric equations that govern the stranded coil and
the different finite element formulations used by the electromagnetic elements. The formulations apply
to static, transient and harmonic analysis types.

5.4.1. Governing Equations

In a stranded coil analysis, the magnetic field equation
(5.157)

is coupled to the electric circuit equation

(5.158)

by the following expressions for the electric current density {J} and the magnetic flux Φ in the coil:
(5.159)

(5.160)

where:

{A} = magnetic vector potential,

ν = magnetic reluctivity,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 193
Electromagnetics

ΔV = voltage drop across the coil,

I = total electric current,

R = total DC resistance of the coil winding,

Sc = coil cross-sectional area,

Nc = number of coil turns,

Vc = coil volume,

{t} = winding direction vector (direction of {J}).

The coupled set of electromagnetic equations governing the stranded coil is obtained by eliminating
the current density {J} and magnetic flux Φ from Equation 5.157 (p. 193) and Equation 5.158 (p. 193)
using Equation 5.159 (p. 193) and Equation 5.160 (p. 193), respectively. If the voltage drop across the
coil ΔV is considered to be an independent electric variable in the stranded coil analysis, the A-VOLT-
EMF formulation results.

5.4.2. A-VOLT-EMF Formulation

This formulation is available with the stranded coil analysis option (KEYOPT(1) = 2) of the current-
technology electromagnetic elements PLANE233, SOLID236 and SOLID237.

In addition to the magnetic vector potential ({A}) and the voltage drop across the coil (ΔV) unknowns,
the coil electromotive force (E) is introduced as a degree of freedom to prevent the distribution of
eddy currents in the wire:

(5.161)

The corresponding finite element matrix equation can be written as:

(5.162)

where:

[KAA] = element magnetic reluctivity matrix defined in Equation 5.112 (p. 180),

Sc = coil cross-sectional area (input as R2 (2D) or R1 (3D) on R command),

Nc = number of coil turns (input as R3 (2D) or R2 (3D) on R command),

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
194 Published: 2023-10-23
 Stranded Coil Analyses

Vc = coil volume (calculated (2D) or input as R3 (3D) on R command),

{t} = current direction vector (input as R5 (2D) or R4-R6 (3D) on R command),

R = total DC resistance of the coil (input as R6 (2D) or R7 (3D) on R command),

s = symmetry factor of the coil (input as R7 (2D) or R8 (3D) on R command),

{N} = vector of element scalar (node-based) shape functions,

[W] = {N} (2D) or matrix of vector (edge-based) shape functions (3D),

vol = element volume,

{Ie} = nodal current vector (input/output as AMPS),

{Ae} = nodal magnetic vector potential - Z-component of magnetic vector potential at element nodes
(2D) or edge-flux at element midside nodes (3D) (input/output as AZ),

{ΔVe} = nodal voltage drop across the coil (input/output as VOLT),

{Ee} = nodal electromotive force in the coil (input/output as EMF).

Note that the VOLT and EMF degrees of freedom should be coupled for each coil using the CP,,VOLT
and CP,,EMF commands.

Equation 5.162 (p. 194) that strongly couples magnetic and electric degrees of freedom in a stranded
coil is nonsymmetric. It can be made symmetric by either using the weak coupling option (KEYOPT(2)
= 1) in static or transient analyses or, provided the symmetry factor s = 1, by using the time-integrated
voltage drop and emf formulation (KEYOPT(2) = 2) in transient or harmonic analyses. In the latter
case, the VOLT and EMF degrees of freedom have the meaning of the time-integrated voltage drop

across the coil and time-integrated electromotive force , and Equation 5.162 (p. 194)
becomes:

(5.163)

With the A-VOLT-EMF formulation where VOLT and EMF are 'true' (not time-integrated) voltage drop
and emf respectively, the stranded coil can be node-coupled coupled to the current-based circuit
elements (CIRCU124) as they share the same VOLT degree of freedom.

The stranded coil analysis results include:

• Nodal magnetic flux density calculated as in Equation 5.121 (p. 184) (output as B)

• Nodal magnetic field intensity calculated as in Equation 5.122 (p. 184) (output as H)

• Element electric current density calculated as in Equation 5.159 (p. 193) (output as JT or JS)

• Nodal magnetic forces calculated as in Equation 5.140 (p. 188) (output as FMAG)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 195
Electromagnetics

• Element Joule heat generation rate calculated as in Equation 5.146 (p. 190) (output as JHEAT)

Note that the calculated current density JT (or JS) and the Joule heat generation rate JHEAT are ef-
fective in the sense that they are calculated based on the coil cross-sectional area (SC) and coil volume
(VC), respectively, and those real constants include the wire and the non-conducting material filling
the space between the winding.

5.5. Inductance, Flux and Energy Computation

Inductance plays an important role in the characterization of magnetic devices, electrical machines,
sensors and actuators. The concept of a non-variant (time-independent), linear inductance of wire-like
coils is discussed in every electrical engineering book. However, its extension to variant, nonlinear, dis-
tributed coil cases is far from obvious.

Time-variance is essential when the geometry of the device is changing: actuators and electrical machines,
for example. In this case, the inductance depends on a stroke (in a 1D motion case) which, in turn, de-
pends on time.

Many magnetic devices apply iron for the conductance of magnetic flux. Most iron has a nonlinear B-
H curve. Because of this nonlinear feature, two kinds of inductance must be considered: differential and
secant. The secant inductance is the ratio of the total flux over current. The differential inductance is
the ratio of flux change over a current excitation change.

The flux of a single wire coil can be defined as the surface integral of the flux density. However, when
the size of the wire is not negligible, it is not clear which contour spans the surface. The field within
the coil must be taken into account. Even larger difficulties occur when the current is not constant: for
example solid rotor or squirrel-caged induction machines.

The voltage induced in a variant coil can be decomposed into two major components: transformer
voltage and motion-induced voltage.

The transformer voltage is induced in coils by the rate change of exciting currents. It is present even if
the geometry of the system is constant, meaning the coils don't move or expand. To obtain the trans-
former voltage, the knowledge of flux change (that is, that of differential flux) is necessary when the
exciting currents are perturbed.

The motion-induced voltage (sometimes called back-EMF) is related to the geometry change of the
system. It is present even if the currents are kept constant. To obtain the motion-induced voltage, the
knowledge of absolute flux in the coils as a function of stroke is necessary.

Obtaining the proper differential and absolute flux values requires consistent calculations of magnetic
absolute and incremental energies and co-energies. For current-technology elements, the linear perturb-
ation procedure can be used to calculate the differential inductance and the absolute flux using the
incremental (IENE) and the co-energy (COEN) element records, respectively.

5.5.1. Differential Inductance Definition

Consider a magnetic excitation system consisting of n coils each fed by a current, Ii. The flux linkage
ψi of the coils is defined as the surface integral of the flux density over the area multiplied by the
number of turns, Ni, of the of the pertinent coil. The relationship between the flux linkage and currents
can be described by the secant inductance matrix, [Ls]:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
196 Published: 2023-10-23
 Inductance, Flux and Energy Computation

(5.164)

where:

{ψ} = vector of coil flux linkages

t = time
{I} = vector of coil currents.
{ψo} = vector of flux linkages for zero coil currents (effect of permanent magnets)

Main diagonal element terms of [Ls] are called self inductance, whereas off diagonal terms are the
mutual inductance coefficients. [Ls] is symmetric which can be proved by the principle of energy
conservation.

In general, the inductance coefficients depend on time, t, and on the currents. The time dependent
case is called time variant which is characteristic when the coils move. The inductance computation
used by the program is restricted to time invariant cases. Note that time variant problems may be
reduced to a series of invariant analyses with fixed coil positions. The inductance coefficient depends
on the currents when nonlinear magnetic material is present in the domain.

The voltage vector, {U}, of the coils can be expressed as:

(5.165)

In the time invariant nonlinear case

(5.166)

The expression in the bracket is called the differential inductance matrix, [Ld]. The circuit behavior of
a coil system is governed by [Ld]: the induced voltage is directly proportional to the differential in-
ductance matrix and the time derivative of the coil currents. In general, [Ld] depends on the currents,
therefore it should be evaluated for each operating point.

5.5.2. Review of Inductance Computation Methods

After a magnetic field analysis, the secant inductance matrix coefficients, Lsij, of a coupled coil system
could be calculated at postprocessing by computing flux linkage as the surface integral of the flux
density, {B}. The differential inductance coefficients could be obtained by perturbing the operating
currents with some current increments and calculating numerical derivatives. However, this method
is cumbersome and neither accurate nor efficient. A much more convenient and efficient method is
offered by the energy perturbation method developed by Demerdash and Arkadan([227] (p. 937)),
Demerdash and Nehl([228] (p. 937)) and Nehl et al.([229] (p. 937)). The energy perturbation method is
based on the following formula:

(5.167)

where W is the magnetic energy, Ii and Ij are the currents of coils i and j. The first step of this procedure
is to obtain an operating point solution for nominal current loads by a nonlinear analysis. In the
second step, linear analyses are carried out with properly perturbed current loads and a tangent re-

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 197
Electromagnetics

luctivity tensor, νt, evaluated at the operating point. For a self coefficient, two, and for a mutual
coefficient, four, incremental analyses are required. In the third step the magnetic energies are obtained
from the incremental solutions and the coefficients are calculated according to Equation 5.167 (p. 197).

5.5.3. Inductance Computation Method Used

The inductance computation method is based on Gyimesi and Ostergaard([231] (p. 937)) who used
Smythe's procedure([151] (p. 933)).

The incremental energy Wij is defined by

(5.168)

where {ΔH} and {ΔB} denote the increase of magnetic field and flux density due to current increments,
ΔIi and ΔIj. The coefficients can be obtained from
(5.169)

This allows an efficient method that has the following advantages:

1. For any coefficient, self or mutual, only one incremental analysis is required.

2. There is no need to evaluate the absolute magnetic energy. Instead, an "incremental energy" is
calculated according to a simple expression.

3. The calculation of incremental analysis is more efficient because the factorized stiffness matrix
can be applied (no inversion is needed). Only incremental load vectors should be evaluated.

For elements PLANE233, SOLID236 and SOLID237, the linear perturbation procedure can be used to
derive the differential inductance from the element incremental energy (output as IENE). The incre-
mental energy is calculated using Equation 5.168 (p. 198), where ΔH and ΔB are the linear perturbation
analysis magnetic field and flux densities corresponding to the perturbation current loads ΔIi and
ΔIj.

5.5.4. Transformer and Motion Induced Voltages

The absolute flux linkages of a time-variant multi-coil system can be written in general:
(5.170)

where:

{X} = vector of strokes

The induced voltages in the coils are the time derivative of the flux linkages, according to Equa-
tion 5.165 (p. 197). After differentiation:
(5.171)

(5.172)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
198 Published: 2023-10-23
 Inductance, Flux and Energy Computation

{V} = vector of stroke velocities

The first term is called transformer voltage (it is related to the change of the exciting current). The
proportional term between the transformer voltage and current rate is the differential inductance
matrix according to Equation 5.166 (p. 197).

The second term is the motion-induced voltage or back-EMF (it is related to the change of strokes).
The time derivative of the stroke is the velocity, hence the motion-induced voltage is proportional
to the velocity.

5.5.5. Absolute Flux Computation

Whereas the differential inductance can be obtained from the differential flux due to current perturb-
ation as described in Differential Inductance Definition (p. 196), Review of Inductance Computation
Methods (p. 197), and Inductance Computation Method Used (p. 198). The computation of the motion
induced voltage requires the knowledge of absolute flux. In order to apply Equation 5.172 (p. 198),
the absolute flux should be mapped out as a function of strokes for a given current excitation ad the
derivative provides the matrix link between back-EMF and velocity.

The absolute flux is related to the system co-energy by:

(5.173)

According to Equation 5.173 (p. 199), the absolute flux can be obtained with an energy perturbation
method by changing the excitation current for a given stroke position and taking the derivative of
the system co-energy.

The increment of co-energy can be obtained by:

(5.174)

where:

= change of co-energy due to change of current Ii

ΔHi = change of magnetic field due to change of current Ii

The incremental co-energy in Equation 5.174 (p. 199) is output as COEN in a linear perturbation ana-
lysis using elements PLANE233, SOLID236 and SOLID237.

5.5.6. Inductance Computations

The differential inductance matrix and the absolute flux linkages of coils can be calculated using linear
perturbation analysis for the current-technology electromagnetic elements.

The differential inductance computation is based on the energy perturbation procedure using Equa-
tion 5.168 (p. 198) and Equation 5.169 (p. 198).

The absolute flux computation is based on the co-energy perturbation procedure using Equa-
tion 5.173 (p. 199) and Equation 5.174 (p. 199).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 199
Electromagnetics

The output can be applied to compute the voltages induced in the coils using Equation 5.172 (p. 198).

5.5.7. Absolute Energy Computation

The absolute magnetic energy density is defined by:

(5.175)

and the absolute magnetic co-energy density is defined by:

(5.176)

See Figure 5.2: Energy and Co-energy for Non-Permanent Magnets (p. 200) and Figure 5.3: Energy and
Co-energy for Permanent Magnets (p. 200) for the graphical representation of these energy definitions.

For current-technology electromagnetic elements, the absolute magnetic energy ( ) and

co-energy ( ) are output as MENE and COEN element records, respectively.

Figure 5.2: Energy and Co-energy for Non-Permanent Magnets

Figure 5.3: Energy and Co-energy for Permanent Magnets

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
200 Published: 2023-10-23
 Electromagnetic Particle Tracing

Equation 5.168 (p. 198) and Equation 5.174 (p. 199) provide the incremental magnetic energy and in-
cremental magnetic co-energy definitions used for inductance and absolute flux computations.

In addition to the magnetic energy and co-energy, elements PLANE233, SOLID236 and SOLID237
calculate the apparent energy defined as:

(5.177)

5.6. Electromagnetic Particle Tracing

Once the electromagnetic field is computed, particle trajectories can be evaluated by solving the
equations of motion:
(5.178)

where:

m = mass of particle
q = charge of particle
{E} = electric field vector
{B} = magnetic field vector
{F} = Lorentz force vector
{a} = acceleration vector
{v} = velocity vector

The tracing follows from element to element: the exit point of an old element becomes the entry point
of a new element. Given the entry location and velocity for an element, the exit location and velocity
can be obtained by integrating the equations of motion.

The particle tracing algorithm is based on Gyimesi et al.([230] (p. 937)) with these assumptions:

1. No relativistic effects (the velocity is much smaller than speed of light).

2. Pure electric tracing ({B} = {0}), pure magnetic tracing ({E} = {0}), or combined {E-B} tracing.

3. Electrostatic and/or magnetostatic analysis

4. Constant {E} and/or {B} within an element.

5. Quadrangle, triangle, hexahedron, tetrahedron, wedge or pyramid element shapes bounded by

planar surfaces.

These simplifications significantly reduce the computation time of the tracing algorithm because the
trajectory can be given in an analytic form:

1. parabola in the case of electric tracing

2. helix in the case of magnetic tracing.

3. generalized helix in the case of coupled E-B tracing.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 201
Electromagnetics

The exit point from an element is the point where the particle trajectory meets the plane of bounding
surface of the element. It can be easily computed when the trajectory is a parabola. However, to compute
the exit point when the trajectory is a helix, a transcendental equation must be solved. A Newton-
Raphson algorithm is used to obtain the solution. The starting point is carefully selected to ensure
convergence to the correct solution. This can be quite involved, as about 70 sub-cases are considered
by the algorithm. This tool allows particle tracing within an element with accuracy limited only by ma-
chine precision. This does not mean that the tracing is exact, since the element field solution may be
inexact. However, with mesh refinement, this error can be controlled.

Once a trajectory is computed, any available physical items can be printed or plotted along the path
(see PLTRAC, PATH, PAPUT). For example, elapsed time, traveled distance, particle velocity components,
temperature, field components, potential values, acoustic pressure, mechanical strain, etc. Animation
is also available.

The plotted particle traces consist of two branches: the first is a trajectory for a given starting point at
a given velocity (forward ballistic); the second is a trajectory for a particle to hit a given target location
at a given velocity (backward ballistics).

5.7. Capacitance Computation

Capacitance computation is one of the primary goals of an electrostatic analysis. For the definition of
ground (partial) and lumped capacitance matrices see Vago and Gyimesi([240] (p. 938)). The knowledge
of capacitance is essential in the design of electrostatic devices, Micro Electro Mechanical Systems
(MEMS), transmission lines, printed circuit boards (PCB), electromagnetic interference and compatibility
(EMI/EMC) etc. The computed capacitance can be the input of a subsequent MEMS analysis using the
electrostructural transducer element TRANS126; for the theory, see TRANS126 - Electromechanical
Transducer (p. 526).

The capacitance matrix of an electrostatic system can be computed (by the CMATRIX command macro).
The capacitance calculation is based on the energy principle. For details, see Gyimesi and Oster-
gaard([250] (p. 939)) and for its successful application, see Hieke([251] (p. 939)). The energy principle
constitutes the basis for inductance matrix computation, as shown in Inductance, Flux and Energy
Computation (p. 196).

The electrostatic energy of a linear three electrode (the third is ground) system is:
(5.179)

where:

W = electrostatic energy
V1 = potential of first electrode with respect to ground
V2 = potential of second electrode with respect to ground

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
202 Published: 2023-10-23
 Capacitance Computation

By applying appropriate voltages on electrodes, the coefficients of the ground capacitance matrix can
be calculated from the stored static energy.

The charges on the conductors are:

(5.180)

(5.181)

where:

Q1 = charge of first electrode

Q2 = charge of second electrode

The charge can be expressed by potential differences, too:

(5.182)
(5.183)

where:

The lumped capacitances can be obtained by lumped capacitors as shown in Figure 5.4: Lumped Capa-
citor Model of Two Conductors and Ground (p. 203). Lumped capacitances are suitable for use in circuit
simulators.

Figure 5.4: Lumped Capacitor Model of Two Conductors and Ground

In some cases, one of the electrodes may be located very far from the other electrodes. This can be
modeled as an open electrode problem with one electrode at infinity. The open boundary region can
be modeled by infinite elements or simply closing the FEM region far enough away by an artificial Di-
richlet boundary condition. In this case, the ground key parameter (GRNDKEY on the CMATRIX command
macro) should be activated. This key assumes that there is a ground electrode at infinity.

The previous case should be distinguished from an open boundary problem without an electrode at
infinity. In this case the ground electrode is one of the modeled electrodes. The FEM model size can
be minimized in this case, too, by infinite elements. When performing the capacitance calculation,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 203
Electromagnetics

however, the ground key (GRNDKEY on the CMATRIX command macro) should not be activated since
there is no electrode at infinity.

5.8. Conductance Computation

Conductance computation is one of the primary goals of an electrostatic analysis. For the definition of
ground (partial) and lumped conductance matrices see Vago and Gyimesi([240] (p. 938)). The knowledge
of conductance is essential in the design of electrostatic devices, Micro Electro Mechanical Systems
(MEMS), transmission lines, printed circuit boards (PCB), electromagnetic interference and compatibility
(EMI/EMC) etc. The computed capacitance can be the input of a subsequent MEMS analysis using the
electrostructural transducer element TRANS126; for theory see TRANS126 - Electromechanical Trans-
ducer (p. 526).

The conductance matrix of an electrostatic system can be computed (by the GMATRIX command macro).
The conductance calculation is based on the energy principle. For details see Gyimesi and Oster-
gaard([250] (p. 939)), and for its successful application, see Hieke([251] (p. 939)). The energy principle
constitutes the basis for inductance matrix computation, as shown in Inductance, Flux and Energy
Computation (p. 196).

The electrostatic energy of a linear three conductor (the third is ground) system is:
(5.184)

where:

W = electrostatic energy
V1 = potential of first conductor with respect to ground
V2 = potential of second conductor with respect to ground

By applying appropriate voltages on conductors, the coefficients of the ground conductance matrix
can be calculated from the stored static energy.

The currents in the conductors are:

(5.185)
(5.186)

where:

I1 = current in first conductor

I2 = current in second conductor

The currents can be expressed by potential differences, too:

(5.187)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
204 Published: 2023-10-23
 Hall Effect

(5.188)

where:

The lumped conductances can be obtained by lumped conductors as shown in Figure 5.5: Lumped
Conductor Model of Two Conductors and Ground (p. 205). Lumped conductances are suitable for use
in circuit simulators.

Figure 5.5: Lumped Conductor Model of Two Conductors and Ground

5.9. Hall Effect

The Hall effect is available using the electromagnetic analysis option (KEYOPT(1) = 1) of the 3D electro-
magnetic elements SOLID236 and SOLID237.

The Hall effect analysis is nonlinear and requires at least two iterations to achieve a converged solution.
The Newton-Raphson algorithm will be turned on automatically when the Hall constant is specified. An
electromagnetic analysis with the Hall effect can be steady-state or transient.

The electric constitutive relation is generalized as follows to include the Hall effect:
(5.189)

where:

{J} = electric current density vector

= electric conductivity matrix without a magnetic field

ρ xx = electrical resistivity in the X-direction (input as RSVX on MP command)

{E} = electric field intensity vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 205
Electromagnetics

{E H } = –R H [{J} x {B}] = Hall field intensity vector

R H = Hall coefficient (input as RH on MP command)

{B}={B x, B y ,B z }T = magnetic flux density vector

Combining ohmic and Hall conductivity terms, Equation 5.189 (p. 205) can be rewritten using an effective
anisotropic and nonsymmetric conductivity:
(5.190)

where:

(5.191)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
206 Published: 2023-10-23
Chapter 6: Heat Flow
The following heat flow topics are available:
6.1. Heat Flow Fundamentals
6.2. Derivation of Heat Flow Matrices
6.3. Heat Flow Evaluations
6.4. Radiosity Solution Method
6.5. Radiation Matrix Method

6.1. Heat Flow Fundamentals

The following topics concerning heat flow fundamentals are available:
6.1.1. Conduction, Convection, and Mass Transport (Advection)
6.1.2. Radiation
6.1.3. Mass Transport (Advection)

6.1.1. Conduction, Convection, and Mass Transport (Advection)

The first law of thermodynamics states that thermal energy is conserved. Specializing this to a differ-
ential control volume:
(6.1)

where:

ρ = density (input as DENS on MP command)

c = specific heat (input as C on MP command)
T = temperature (=T(x,y,z,t))
t = time

= vector operator

= velocity vector for mass transport of heat (input , , using the R command

for PLANE55 and SOLID70 and the BF command for SOLID278 or PLANE292. For details, see
Mass Transport (Advection) (p. 215).)
{q} = heat flux vector (output as TFX, TFY, and TFZ)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 207
Heat Flow

= heat generation rate per unit volume (input on BF or BFE commands)

It should be realized that the terms {L}T and {L}T{q} may also be interpreted as T and {q}, re-
spectively, where represents the grad operator and represents the divergence operator.

Next, Fourier's law is used to relate the heat flux vector to the thermal gradients:
(6.2)

where:

Kxx, Kyy, Kzz = conductivity in the element x, y, and z directions, respectively (input as KXX,
KYY, KZZ on MP command)

Combining Equation 6.1 (p. 207) and Equation 6.2 (p. 208),
(6.3)

Expanding Equation 6.3 (p. 208) to its more familiar form:

(6.4)

It will be assumed that all effects are in the global Cartesian system.

Three types of boundary conditions are considered. It is presumed that these cover the entire element.

1. Specified temperatures acting over surface S1:

(6.5)

where T* is the specified temperature (input on D command).

2. Specified heat flows acting over surface S2:

(6.6)

where:

{n} = unit outward normal vector

q* = specified heat flow (input on SF or SFE commands)

Figure 6.1: Specified Heat Flows

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
208 Published: 2023-10-23
 Heat Flow Fundamentals

3. Specified convection surfaces acting over surface S3 (Newton's law of cooling):

(6.7)

where:

hf = film coefficient (input on SF or SFE commands) Evaluated at (TB + TS)/2 unless oth-
erwise specified for the element
TB = bulk temperature of the adjacent fluid (input on SF or SFE commands)
TS = temperature at the surface of the model

Figure 6.2: Specified Convection Surfaces

For a fluid flowing past a solid surface, the bulk temperature TB is equal to the free stream tem-
perature, TFS.

For the case of bleed holes in a solid, film effectiveness (η) accounts for the coolant bleeding
through the cooling holes to the external surface of the solid. The bulk temperature TB is then a
combination of the free stream temperature TFS and the temperature of the coolant exiting the
bleed hole, TEX:
(6.8)

where 0 < η < 1.

Typically, TEX is obtained from a FLUID116 element using the extra node option for SURF151 or
SURF152.

Figure 6.3: Cooling Passages in a Solid

Note that positive specified heat flow is into the boundary (that is, in the direction opposite of {n}),
which accounts for the negative signs in Equation 6.6 (p. 208) and Equation 6.7 (p. 209).

Combining Equation 6.2 (p. 208) with Equation 6.6 (p. 208) and Equation 6.7 (p. 209)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 209
Heat Flow

(6.9)
(6.10)

Premultiplying Equation 6.3 (p. 208) by a virtual change in temperature, integrating over the volume
of the element, and combining with Equation 6.9 (p. 210) and Equation 6.10 (p. 210) with some manip-
ulation yields:

(6.11)

where:

vol = volume of the element

δT = an allowable virtual temperature (=δT(x,y,z,t))

6.1.2. Radiation
Radiant energy exchange between neighboring surfaces of a region or between a region and its
surroundings can produce large effects in the overall heat transfer problem. Though the radiation
effects generally enter the heat transfer problem only through the boundary conditions, the coupling
is especially strong due to nonlinear dependence of radiation on surface temperature.

Extending the Stefan-Boltzmann Law for a system of N enclosures, the energy balance for each surface
in the enclosure for a gray diffuse body is given by Siegal and Howell([89] (p. 929)(Equation 8-19)),
which relates the energy losses to the surface temperatures:

(6.12)

where:

N = number of radiating surfaces

δji = Kronecker delta
εi = effective emissivity (input on EMIS or MP command) of surface i
Fji = radiation view factors (see below)
Ai = area of surface i
Qi = energy loss of surface i
σ = Stefan-Boltzmann constant (input on STEF or R command)
Ti = absolute temperature of surface i

For a system of two surfaces radiating to each other, Equation 6.12 (p. 210) can be simplified to give
the heat transfer rate between surfaces i and j as (see Chapman([353] (p. 945))):

(6.13)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
210 Published: 2023-10-23
 Heat Flow Fundamentals

where:

Ti, Tj = absolute temperature at surface i and j, respectively

If Aj is much greater than Ai, Equation 6.13 (p. 210) reduces to:
(6.14)

where:

[Link]. View Factors

The view factor, Fij, is defined as the fraction of total radiant energy that leaves surface i which arrives
directly on surface j, as shown in Figure 6.4: View Factor Calculation Terms (p. 211). It can be expressed
by the following equation:

Figure 6.4: View Factor Calculation Terms

(6.15)

where:

Ai,Aj = area of surface i and surface j

r = distance between differential surfaces i and j
θi = angle between Ni and the radius line to surface d(Aj)
θj = angle between Nj and the radius line to surface d(Ai)
Ni,Nj = surface normal of d(Ai) and d(Aj)

To ensure a good energy balance, it is important to satisfy both row sum and reciprocity relationships
for the view factor matrix (VFSM command).

For a perfect enclosure, the row sum must satisfy the following requirement:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 211
Heat Flow

where Fij are the view factor matrix values.

For a leaky enclosure, the row sum must satisfy the following requirement:

For a perfect enclosure, the following residual must be less than the specified convergence value
(input as CONV on the VFSM command).

where ' indicates the new view factor values.

For a leaky enclosure, the following residual must be less than the specified convergence value
(input as CONV on the VFSM command).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
212 Published: 2023-10-23
 Heat Flow Fundamentals

where Fij is the original view factor values.

To ensure a good energy balance, the following reciprocity relationship must also be met:

where Ai is the area of the ith facet.

[Link]. View Factor Matrix Properties

View Factor Matrix: General
For a perfect enclosure with N facets, the view factor matrix is:

(6.16)

where:

Fij represents the fraction of radiation energy leaving facet i and hitting facet j.

Figure 6.5: View Factor Facets for a General Model (No Symmetry)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 213
Heat Flow

Consequently, the following requirements must be satisfied:

summation: for all i = 1,2, … N

reciprocity: for all i and j, where Ai is the area of facet i.

View Factor Matrix for a Model with Symmetry

For a model with symmetry defined by issuing the RSYMM command, we can decompose the view
factor matrix as:

(6.17)

where the subscripts, and , denote independent and dependent. Independent facets have un-
derlying solid elements underneath while dependent facets are created from independent facets
using RSYMM settings, as illustrated in Figure 6.6: View Factor Facets for a Model with Sym-
metry (p. 214).

Figure 6.6: View Factor Facets for a Model with Symmetry

The reciprocity condition is written as:

(6.18)

where: .

Here, and represent a block diagonal matrix of the independent and dependent facet areas,
respectively.

Based on symmetry:
(6.19)

Substituting Equation 6.19 (p. 214) and Equation 6.17 (p. 214) into Equation 6.18 (p. 214) and simplifying
gives:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
214 Published: 2023-10-23
 Heat Flow Fundamentals

(6.20)
(When numerical methods are used, Equation 6.20 (p. 215) is approximate.)

This means matrix is symmetric, .

Also note that block lumping gives

(6.21)

where .

When condensed view factor is turned on (VFCO is issued with LEVEL=1 or 2), the view factor

matrix produced is .

When condensed view factor is turned off (VFCO command is not used (default) or is issued with

LEVEL = 0), it is .

[Link]. Radiation Usage

Four methods for analysis of radiation problems are included:

1. Radiation link element LINK31(LINK31 - Radiation Link (p. 459)). For simple problems involving
radiation between two points or several pairs of points. The effective radiating surface area, the
form factor and emissivity can be specified as real constants for each radiating point.

2. Surface effect elements - SURF151 in 2D and SURF152 in 3D for radiating between a surface
and a point (SURF151 - 2D Thermal Surface Effect (p. 538) and SURF152 - 3D Thermal Surface
Effect (p. 539) ). The form factor between a surface and the point can be specified as a real
constant or can be calculated from the basic element orientation and the extra node location.

3. Radiation matrix method (Radiation Matrix Method (p. 226)). For more generalized radiation
problems involving two or more surfaces. The method involves generating a matrix of view
factors between radiating surfaces and using the matrix as a superelement in the thermal ana-
lysis.

4. Radiosity solver method (Radiosity Solution Method (p. 220)). For generalized problems in 3D
involving two or more surfaces. The method involves calculating the view factor for the flagged
radiating surfaces using the hemicube method and then solving the radiosity matrix coupled
with the conduction problem.

6.1.3. Mass Transport (Advection)

A mass transport option is available for the current-technology elements, SOLID278 and PLANE292,
and the legacy elements, PLANE55 and SOLID70, to simulate the mass transport of heat by a flow
with a prescribed velocity field in your analysis. The option to include mass transport is controlled by
KEYOPT (8) for PLANE55, SOLID70, and PLANE292 and KEYOPT(11) for SOLID278. This section discusses
the governing equations and restrictions for mass transport simulation. For procedural details and

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 215
Heat Flow

information on setting up an analysis to simulate mass transport and example problems, see Mass
Transport (Advection) in the Thermal Analysis Guide.

[Link]. Diffusive Flux and Total Flux Neumann Boundary Conditions

Figure 6.7: Mass Transport Problem in a Flow Field (p. 216) shows the directed flow field (orange
arrows) over a domain with a surface and the unit vector normal to its surface. The surface
can be split into the inflow side where and the outflow side where , and

Figure 6.7: Mass Transport Problem in a Flow Field

The equation in Figure 6.7: Mass Transport Problem in a Flow Field (p. 216) is the first law of thermo-
dynamics in differential form (Equation 6.1 (p. 207) rewritten), and the second term describes mass
transport. Typically, the Dirichlet boundary condition is applied on the inflow side and the Neumann
boundary condition is applied on the outflow side as follows:

Dirichlet boundary condition:

Neumann boundary condition:

These are the only boundary condition options available for legacy elements PLANE55 and SOLID70.

For current-technology elements PLANE292 and SOLID278, you can choose between two types of
Neumann boundary conditions described below.

1. The diffusive flux (Dflux) formulation is, again, the same as that for PLANE55 and SOLID70:

Neumann boundary condition:

You specify the Dflux Neumann boundary condition by setting KEYOPT(8) =1 for
PLANE292 and KEYOPT(11) =1 for SOLID278.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
216 Published: 2023-10-23
 Derivation of Heat Flow Matrices

2. The total flux (Tflux) formulation includes the advective term in the Neumann boundary
condition as follows:

Neumann boundary condition:

You specify this Tflux Neumann boundary condition by setting KEYOPT(8) = 2 for
PLANE292 and KEYOPT(11) = 2 for SOLID278.

There are pros and cons associated with either of these boundary conditions. The main advantage
of the Tflux formulation is that it will satisfy an energy balance with the PRRSOL command, but
setting up its Neumann boundary condition is slightly more complex since it includes both a diffusive
and an advective flux. Vice versa, if you are not interested in the energy balance, you could use
the Dflux boundary condition which is easy to set up.

Imposing the Neumann Boundary Condition

Specifying a known heat flux

To specify a heat flux at the outlet as the Neumann boundary condition, issue the SF or SFE com-
mand with Lab = HFLUX and specify its value. Specifying a known heat flux Neumann boundary
condition is the same for both the Dflux and the Tflux formulation.

Specifying zero diffusion flux for high Peclet number

Since the diffusion flux becomes negligible at high flow rate, it is typical for high Pe to specify a
soft boundary condition of zero diffusion flux on the outflow side, .

Dflux formulation

This is easily accomplished with the Dfulx formulation of the Neumann boundary condition.
To set on the boundary, no commands are required since the default value of
Lab = HFLUX (on SF or SFE) is zero.

Tflux formulation

It is less straightforward to specify zero diffusion flux on the outflow side under the Tflux
formulation. Setting , the Tflux Neumann boundary condition becomes
, which you can set as a convection boundary condition. To do so, issue
the SF or SFE command with Lab = CONV, and specify the bulk temperature = 0 and the
known value of the film coefficient to .

6.2. Derivation of Heat Flow Matrices

As stated before, the variable T was allowed to vary in both space and time. This dependency is separated
as:
(6.22)

where:

T = T(x,y,z,t) = temperature
{N} = {N(x,y,z)} = element shape functions

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 217
Heat Flow

{Te} = {Te(t)} = nodal temperature vector of element

Thus, the time derivatives of Equation 6.22 (p. 217) may be written as:
(6.23)

δT has the same form as T:

(6.24)

The combination {L}T is written as

(6.25)

where:

[B] = {L}{N}T

Now, the variational statement of Equation 6.11 (p. 210) can be combined with Equation 6.22 (p. 217)
thru Equation 6.25 (p. 218) to yield:

(6.26)

Terms are defined in Heat Flow Fundamentals (p. 207). ρ is assumed to remain constant over the volume
of the element. On the other hand, c and may vary over the element. Finally, {Te}, , and {δTe} are
nodal quantities and do not vary over the element, so that they also may be removed from the integral.
Now, since all quantities are seen to be premultiplied by the arbitrary vector {δTe}, this term may be
dropped from the resulting equation. Thus, Equation 6.26 (p. 218) may be reduced to:

(6.27)

Equation 6.27 (p. 218) may be rewritten as:

(6.28)

where:

= element specific heat (thermal damping) matrix

= element mass transport conductivity matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
218 Published: 2023-10-23
 Heat Flow Evaluations

= element diffusion conductivity matrix

= element convection surface conductivity matrix

= element mass flux vector

= element convection surface heat flow vector

= element heat generation load

Comments on and modifications of the above definitions:

1. is not symmetric.

2. is calculated as defined above, for SOLID90 only. All other elements use a diagonal matrix,

with the diagonal terms defined by the vector .

3. is frequently diagonalized, as described in Lumped Matrices (p. 422).

4. If exists and has been diagonalized and also the analysis is a transient (Key = ON on the TIMINT
command), has its terms adjusted so that they are proportioned to the main diagonal terms
of . , the heat generation rate vector for Joule heating is treated similarly, if present. This
adjustment ensures that elements subjected to uniform heating will have a uniform temperature
rise. However, this adjustment also changes nonuniform input of heat generation to an average
value over the element.

5. For phase change problems, is evaluated from the enthalpy curve (Tamma and
Namnuru([43] (p. 927))) if enthalpy is input (input as ENTH on MP command). This option should be
used for phase change problems.

6.3. Heat Flow Evaluations

6.3.1. Integration Point Output

The element thermal gradients at the integration points are:
(6.29)

where:

{a} = thermal gradient vector (output as TG)

{L} = vector operator
T = temperature

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 219
Heat Flow

Using shape functions, Equation 6.29 (p. 219) may be written as:
(6.30)

where:

[B] = shape function derivative matrix evaluated at the integration points

{Te} = nodal temperature vector of element

Then, the heat flux vector at the integration points may be computed from the thermal gradients:
(6.31)

where:

{q} = heat flux vector (output as TF)

[D] = conductivity matrix (see Equation 6.2 (p. 208))

Nodal gradient and flux vectors may be computed from the integration point values as described in
Nodal and Centroidal Data Evaluation (p. 442).

6.3.2. Surface Output

The convection surface output is:
(6.32)

where:

qc = heat flow per unit area due to convection

hf = film coefficient (input on SF or SFE commands)
TS = temperature at surface of model
TB = bulk temperature of the adjacent fluid (input on SF or SFE commands)

6.4. Radiosity Solution Method

In the radiosity solution method for the analysis of gray diffuse radiation between N surfaces, Equa-
tion 6.12 (p. 210) is solved in conjunction with the basic conduction problem.

For the purpose of computation it is convenient to rearrange Equation 6.12 (p. 210) into the following
series of equations

(6.33)

and

(6.34)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
220 Published: 2023-10-23
 Radiosity Solution Method

Equation 6.33 (p. 220) and Equation 6.34 (p. 220) are expressed in terms of the outgoing radiative fluxes
(radiosity) for each surface, , and the net flux from each surface qi. For known surface temperatures,
Ti, in the enclosure, Equation 6.34 (p. 220) forms a set of linear algebraic equations for the unknown,
outgoing radiative flux (radiosity) at each surface. Equation 6.34 (p. 220) can be rewritten as
(6.35)

where:

[A] is a full matrix due to the surface to surface coupling represented by the view factors and is a
function of temperature due to the possible dependence of surface emissivities on temperature. Equa-
tion 6.35 (p. 221) is solved using a Newton-Raphson procedure for the radiosity flux {qo}.

When the qo values are available, Equation 6.34 (p. 220) then allows the net flux at each surface to be
evaluated. The net flux calculated during each iteration cycle is under-relaxed, before being updated
using
(6.36)

where:

φ = radiosity flux relaxation factor

k = iteration number

The net surface fluxes provide boundary conditions to the finite element model for the conduction
process. The radiosity Equation 6.35 (p. 221) is solved coupled with the conduction Equation 6.12 (p. 210)
using a segregated solution procedure until convergence of the radiosity flux and temperature for each
time step or load step.

The surface temperatures used in the above computation must be uniform over each surface in order
to satisfy conditions of the radiation model. In the finite element model, each surface in the radiation
problem corresponds to a face or edge of a finite element. The uniform surface temperatures needed
for use in Equation 6.35 (p. 221) are obtained by averaging the nodal point temperatures on the appro-
priate element face.

For open enclosure problems using the radiosity method, an ambient temperature needs to be specified
using a space temperature (SPCTEMP command) or a space node (SPCNOD command), to account for
energy balance between the radiating surfaces and the ambient.

6.4.1. Radiosity Equations Simplified for Models with Symmetry

When condensed view factor is turned off (VFCO command is not used (default) or issued with LEVEL
= 0 ), the radiosity equations are Equation 6.35 (p. 221) expressed in matrix form:
(6.37)
where

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 221
Heat Flow

(assuming emissivity, , is constant for simplicity)

is the identity matrix.

For a model with symmetry, , where and are the loads on independent and dependent
facets, respectively. Since loading is symmetric, , and

(6.38)

Here, matrix is decomposed into independent and dependent blocks. See View Factor Matrix for a
Model with Symmetry (p. 214) for more details.

Solving Equation 6.38 (p. 222) for independent and dependent fluxes, and , it is typically found
that they are close in value.

When condensed view factor is turned on (VFCO command is issued with LEVEL = 1 or 2), the con-
straint can be imposed by adding dependent row to independent row and dependent column
to independent column, and recalling that (Equation 6.19 (p. 214)) and (Equa-
tion 6.20 (p. 215)), Equation 6.38 (p. 222) becomes:
(6.39)
(6.40)
Thus, the problem is reduced to solving only for .

Note:

Similar results are true for models with multiple planes of symmetry.

6.4.2. View Factor Calculation (3D): Hemicube Method

For solving radiation problems in 3D, the radiosity solution method calculates the view factors using
the hemicube method as compared to the traditional double area integration method for 3D geometry.
Details related to using the hemicube method for view factor calculation are given in Glass
([272] (p. 940)) and Cohen and Greenberg ([276] (p. 940)). For 2D and axisymmetric models, see View
Factor Calculation (2D): Radiation Matrix Method (p. 227) and View Factors of Axisymmetric Bod-
ies (p. 225).

The hemicube method is based upon Nusselt's hemisphere analogy. Nusselt's analogy shows that
any surface, which covers the same area on the hemisphere, has the same view factor. From this it
is evident that any intermediate surface geometry can be used without changing the value of the
view factors. In the hemicube method, instead of projecting onto a sphere, an imaginary cube is
constructed around the center of the receiving patch. A patch in a finite element model corresponds
to an element face of a radiating surface in an enclosure. The environment is transformed to set the
center of the patch at the origin with the normal to the patch coinciding with the positive Z axis. In
this orientation, the imaginary cube is the upper half of the surface of a cube, the lower half being
below the 'horizon' of the patch. One full face is facing in the Z direction and four half faces are facing
in the +X, -X, +Y, and -Y directions. These faces are divided into square 'pixels' at a given resolution,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
222 Published: 2023-10-23
 Radiosity Solution Method

and the environment is then projected onto the five planar surfaces. Figure 6.8: The Hemicube (p. 223)
shows the hemicube discretized over a receiving patch from the environment.

Figure 6.8: The Hemicube

Figure 6.9: Derivation of Delta-View Factors for Hemicube Method

The contribution of each pixel on the cube's surface to the form-factor value varies and is dependent
on the pixel location and orientation as shown in Figure 6.9: Derivation of Delta-View Factors for
Hemicube Method (p. 223). A specific delta form-factor value for each pixel on the cube is found from
modified form of Equation 6.15 (p. 211) for the differential area to differential area form-factor. If two
patches project on the same pixel on the cube, a depth determination is made as to which patch is
seen in that particular direction by comparing distances to each patch and selecting the nearer one.
After determining which patch (j) is visible at each pixel on the hemicube, a summation of the delta
form-factors for each pixel occupied by patch (j) determines the form-factor from patch (i) at the
center of the cube to patch (j). This summation is performed for each patch (j) and a complete row
of N form-factors is found.

At this point the hemicube is positioned around the center of another patch and the process is repeated
for each patch in the environment. The result is a complete set of form-factors for complex environ-
ments containing occluded surfaces. The overall view factor for each surface on the hemicube is
given by:

(6.41)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 223
Heat Flow

N = number of pixels
ΔF = delta-view factor for each pixel

The hemicube resolution (input on the HEMIOPT command) determines the accuracy of the view
factor calculation and the speed at which they are calculated using the hemicube method. Default
is set to 10. Higher values increase accuracy of the view factor calculation.

6.4.3. View Factor Calculation (2D)

Two methods in the AUX 12 processor are available for calculating 2D view factors: the Non-Hidden
and Hidden methods.

[Link]. Non-Hidden Method

The non-hidden procedure calculates a view factor for every surface to every other surface whether
the view is blocked by an element or not. In this procedure, the following equation is used and the
integration is performed adaptively.

For a finite element discretized model, Equation 6.15 (p. 211) for the view factor Fij between two
surfaces i and j can be written as:

(6.42)

where:

m = number of integration points on surface i

n = number of integration points on surface j

When the dimensionless distance between two viewing surfaces D, defined in Equation 6.43 (p. 224),
is less than 0.1, the accuracy of computed view factors is known to be poor (Siegal and How-
ell([89] (p. 929))).

(6.43)

where:

dmin = minimum distance between the viewing surfaces A1 and A2

Amax = max (A1, A2)

So, the order of surface integration is adaptively increased from order one to higher orders as the
value of D falls below 8. The area integration is changed to contour integration when D becomes
less than 0.5 to maintain the accuracy. The contour integration order is adaptively increased as D
approaches zero.

[Link]. Hidden Method

The hidden procedure is a simplified method which uses Equation 6.15 (p. 211) and assumes that
all the variables are constant, so that the equation becomes:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
224 Published: 2023-10-23
 Radiosity Solution Method

(6.44)

The hidden procedure numerically calculates the view factor in the following conceptual manner.
The hidden-line algorithm is first used to determine which surfaces are visible to every other surface.
Then, each radiating, or "viewing", surface (i) is enclosed with a hemisphere of unit radius. This
hemisphere is oriented in a local coordinate system (x' y' z'), whose center is at the centroid of the
surface with the z axis normal to the surface, the x axis is from node I to node J, and the y axis or-
thogonal to the other axes. The receiving, or "viewed", surface (j) is projected onto the hemisphere
exactly as it would appear to an observer on surface i.

As shown in Figure 6.10: Receiving Surface Projection (p. 225), the projected area is defined by first
extending a line from the center of the hemisphere to each node defining the surface or element.
That node is then projected to the point where the line intersects the hemisphere and transformed
into the local system x' y' z', as described in Kreyszig([24] (p. 926))

Figure 6.10: Receiving Surface Projection

The view factor, Fij, is determined by counting the number of rays striking the projected surface j
and dividing by the total number of rays (Nr) emitted by surface i. This method may violate the
radiation reciprocity rule, that is, AiFi-j ≠ Aj Fj-i.

[Link]. View Factors of Axisymmetric Bodies

When the radiation view factors between the surfaces of axisymmetric bodies are calculated
(GEOM,1,n command), special logic is used. In this logic, the axisymmetric nature of the body is
exploited to reduce the amount of computations. The user, therefore, needs only to build a model
in plane 2D representing the axisymmetric bodies as line "elements."

Consider two axisymmetric bodies A and B as shown in Figure 6.11: Axisymmetric Geometry (p. 225).

Figure 6.11: Axisymmetric Geometry

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 225
Heat Flow

The view factor of body A to body B is calculated by expanding the line "element" model into a
full 3D model of n circumferential segments (GEOM,1,n command) as shown in Figure 6.12: End
View of Showing n = 8 Segments (p. 226).

Figure 6.12: End View of Showing n = 8 Segments

View factor of body A to B is given by

(6.45)

where:

Fk - = view factor of segment k on body A to segment on body B

The form factors between the segments of the axisymmetric bodies are computed using the
method described in the previous section. Since the coefficients are symmetric, the summation
Equation 6.45 (p. 226) may be simplified as:

(6.46)

Both hidden and non-hidden methods are applicable in the computation of axisymmetric view
factors. However, the non-hidden method should be used if and only if there are no blocking surfaces.
For example, if radiation between concentric cylinders are considered, the outer cylinder can not
see part of itself without obstruction from the inner cylinder. For this case, the hidden method must
be used, as the non-hidden method would definitely give rise to inaccurate view factor calculations.

[Link]. Space Node

A space node may be defined (SPACE command) to absorb all energy not radiated to other elements.
Any radiant energy not incident on any other part of the model will be directed to the space node.
If the model is not a closed system, then the user must define a space node with its appropriate
boundary conditions.

6.5. Radiation Matrix Method

In the radiation matrix method, for a system of two radiating surfaces, Equation 6.14 (p. 211) can be ex-
panded as:
(6.47)

or
(6.48)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
226 Published: 2023-10-23
 Radiation Matrix Method

where:

K' cannot be calculated directly since it is a function of the unknowns Ti and Tj. The temperatures from
previous iterations are used to calculate K' and the solution is computed iteratively.

For a more general case, Equation 6.12 (p. 210) can be used to construct a single row in the following
matrix equation:
(6.49)

such that:

(6.50)

(6.51)

Solving for {Q}:

(6.52)

and therefore:
(6.53)

Equation 6.52 (p. 227) is analogous to Equation 6.12 (p. 210) and can be set up for standard matrix
equation solution by the process similar to the steps shown in Equation 6.47 (p. 226) and Equa-
tion 6.48 (p. 226).
(6.54)

[K'] now includes T3 terms and is calculated in the same manner as in Equation 6.48 (p. 226)). To be able
to include radiation effects in elements other than LINK31, MATRIX50 (the substructure element) is used
to bring in the radiation matrix. MATRIX50 has an option that instructs the solution phase to calculate
[K']. The AUX12 utility is used to create the substructure radiation matrix. AUX12 calculates the effective
conductivity matrix, [Kts], in Equation 6.52 (p. 227), as well as the view factors required for finding [Kts].
The user defines flat surfaces to be used in AUX12 by overlaying nodes and elements on the radiating
edge of a 2D model or the radiating face of a 3D model.

6.5.1. View Factor Calculation (2D): Radiation Matrix Method

Two methods are available in the radiation matrix method to calculate the view factors (VTYPE
command): Non-Hidden Method and Hidden Method. The governing equations and theory for calcu-
lating view factors using these methods are described in Non-Hidden Method (p. 224) and Hidden
Method (p. 224). See aslo Considerations for the Non-hidden Method in the Thermal Analysis Guide
and Considerations for the Hidden Method in the Thermal Analysis Guide for a discussion concerning
the use of these two methods.

Details of the theory underlying view factor calculations for axisymetric bodies are found in View
Factors of Axisymmetric Bodies (p. 225).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 227
Heat Flow

The use of a space node to account for radiation to ambient space is described in Space Node (p. 226)
and in Recommendations for Using Space Nodes in the Thermal Analysis Guide.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
228 Published: 2023-10-23
Chapter 7: Thin Fluid Film Flow
Reduced order modeling of fluid flow effects is available in Mechanical APDL. This chapter deals with
the thin film effects which you can model using Mechanical APDL. For detailed fluid flow analysis cap-
abilities, refer to the Ansys Fluent and Ansys CFX products.

A thin film is a small gap of fluid between moving surfaces. This thin layer of fluid can alter the struc-
tural response of the structure by adding stiffness and/or damping to the system. Movement normal
to the gap produces a squeeze film effect (modeled with FLUID136 and FLUID138 elements). Movement
tangential to the gap produces a slide film effect (modeled with FLUID139 elements).

The following thin film topics are available:

7.1. Squeeze Film
7.2. Slide Film
7.3. Hydrodynamic Bearing

7.1. Squeeze Film

Reynolds equations known from lubrication technology and theory of rarified gas physics are the the-
oretical background to analyze fluid structural interactions of microstructures (Blech([335] (p. 944)),
Griffin([336] (p. 944)), Langlois([337] (p. 944))). FLUID136 and FLUID138 can be applied to structures where
a small gap between two plates opens and closes with respect to time. This happens in the case of
accelerometers where the seismic mass moves perpendicular to a fixed wall, in micromirror displays
where the mirror plate tilts around a horizontal axis, and for clamped beams such as RF filters where a
flexible structure moves with respect to a fixed wall. Other examples are published in literature
(Mehner([338] (p. 944))).

FLUID136 and FLUID138 can be used to determine the fluidic response for given wall velocities. Both
elements allow for static, harmonic and transient types of analyses. Static analyses can be used to
compute damping parameter for low driving frequencies (compression effects are neglected). Harmonic
analysis can be used to compute damping and squeeze effects at the higher frequencies. Transient
analysis can be used for non-harmonic load functions. Both elements assume isothermal viscous flow.

7.1.1. Flow Between Flat Surfaces

FLUID136 is used to model the thin-film fluid behavior between flat surfaces and is based on the
generalized nonlinear Reynolds equation known from lubrication theory.

(7.1)

where:

d = local gap separation

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 229
Thin Fluid Film Flow

ρ = density
t = time
= divergence operator
= gradient operator
η = dynamic viscosity
Pabs = absolute pressure

Assuming an ideal gas:

(7.2)

where:

R = gas constant
T = temperature

Substituting Equation 7.2 (p. 230) into Equation 7.1 (p. 229) gives:

(7.3)

After substituting ambient pressure plus the pressure for the absolute pressure (Pabs = P0 + P) this
equation becomes:

(7.4)

Equation 7.4 (p. 230) is valid for large displacements and large pressure changes (KEYOPT(4) = 1).
Pressure and velocity degrees of freedom must be activated (KEYOPT(3) = 1 or 2).

For small pressure changes (P/P0 << 1), Equation 7.4 (p. 230) becomes:

(7.5)

where νz = wall velocity in the normal direction. That is:

(7.6)

Equation 7.5 (p. 230) is valid for large displacements and small pressure changes (KEYOPT(4) = 0).
Pressure and velocity degrees of freedom must be activated (KEYOPT(3) = 1 or 2).

For small displacements (d/d0 << 1) and small pressure changes (P/P0 << 1), Equation 7.5 (p. 230) be-
comes:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
230 Published: 2023-10-23
 Squeeze Film

(7.7)

where

do = nominal gap.

This equation applies when pressure is the only degree of freedom (KEYOPT(3) = 0).

For incompressible flows (ρ is constant), the generalized nonlinear Reynolds equation (Equa-
tion 7.1 (p. 229)) reduces to:

(7.8)

This equation applies for incompressible flow (KEYOPT(4) = 2). Pressure and velocity degrees of freedom
must be activated (KEYOPT(3) = 1 or 2).

Reynolds squeeze film equations are restricted to structures with lateral dimensions much larger than
the gap separation. Futhermore, viscous friction may not cause a significant temperature change.
Continuum theory (KEYOPT(1) = 0) is valid for Knudsen numbers smaller than 0.01.

The Knudsen number Kn of the squeeze film problem can be estimated by:

(7.9)

where:

Lo = mean free path length of the fluid

Pref = reference pressure for the mean free path Lo
Pabs = Po + P

For small pressure changes, Pabs is approximately equal to P0 and the Knudsen number can be estim-
ated by:

(7.10)

For systems that operate at Knudsen numbers <0.01, the continuum theory is valid (KEYOPT(1) = 0).
The effective viscosity ηeff is then equal to the dynamic viscosity η.

For systems which operate at higher Knudsen numbers (KEYOPT(1) = 1), an effective viscosity ηeff
considers slip flow boundary conditions and models derived from Boltzmann equation. This assumption
holds for Knudsen numbers up to 880 (Veijola([339] (p. 944))):
(7.11)

For micromachined surfaces, specular reflection decreases the effective viscosity at high Knudsen
numbers compared to diffuse reflection. Surface accommodation factors, α, distinguish between diffuse
reflection (α = 1), specular reflection (α = 0), and molecular reflection (0 < α < 1) of the molecules at
the walls of the squeeze film. Typical accommodation factors for silicon are reported between 0.8

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 231
Thin Fluid Film Flow

and 0.9, those of metal surfaces are almost 1. Different accommodation factors can be specified for
each wall by using α 1 and α 2 (input as A1 and A2 on R command). α 1 is the coefficient associated
with the top moving surface and α 2 is the coefficient associated with the bottom metallic surface.
Results for high Knudsen numbers with accommodation factors (KEYOPT(1) =2) are not expected to
be the same as those for high Knudsen numbers without accommodation factors (KEYOPT(1) =1).

The effective viscosity equations for high Knudsen numbers are based on empirical correlations. Fit
functions for the effective viscosity of micromachined surfaces are found in Veijola([339] (p. 944)). The
effective viscosity is given by the following equation if α 1 = α 2:

(7.12)

and by the following equation if α 1 ≠α 2:

(7.13)

where D is the inverse Knudsen number:

and Q1, Q2, and Q3 are Poiseuille flow rate coefficients:

for p = 1, 2, or 3.

If both surfaces are the same (α 1 = α 2), the Poiseuille flow rate coefficient is given by:

If the bottom fixed plate is metallic (α 2 = 1) and the top moving plate is not metallic (α 1 ≠1), the
Poiseuille flow rate coefficient is given by:

The general solution valid for arbitrary α 1 and α 2 is a simple linear combination of Q1 and Q2:

7.1.2. Flow in Channels

FLUID138 can be used to model the fluid flow though short circular and rectangular channels of mi-
crometer size. The element assumes isothermal viscous flow at low Reynolds numbers, the channel

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
232 Published: 2023-10-23
 Squeeze Film

length to be small compared to the acoustic wave length, and a small pressure drop with respect to
ambient pressure.

In contrast to FLUID116, FLUID138 considers gas rarefaction, is more accurate for channels of rectan-
gular cross sections, allows channel dimensions to be small compared to the mean free path, allows
evacuated systems, and considers fringe effects at the inlet and outlet which considerably increase
the damping force in case of short channel length. FLUID138 can be used to model the stiffening
and damping effects of fluid flow in channels of micro-electromechanical systems (MEMS).

Using continuum theory (KEYOPT(1) = 0) the flow rate Q of channels with circular cross-section
(KEYOPT(3) = 0) is given by the Hagen-Poiseuille equation:
(7.14)

Q = flow rate in units of volume/time

r = radius
lc = channel length
A = cross-sectional area
ΔP = pressure difference along channel length

This assumption holds for low Reynolds numbers (Re < 2300), for l >> r and r >> Lm where Lm is the
mean free path at the current pressure.
(7.15)

In case of rectangular cross sections (KEYOPT(3) = 1) the channel resistance depends on the aspect
ratio of channel. The flow rate is defined by:

(7.16)

where:

rh = hydraulic radius (defined below)

A = true cross-sectional area (not that corresponding to the hydraulic radius)
χ = so-called friction factor (defined below)

The hydraulic radius is defined by:

(7.17)

and the friction factor χ is approximated by:

(7.18)

where:

H = height of channel
W = width of channel (must be greater than H)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 233
Thin Fluid Film Flow

n = H/W

A special treatment is necessary to consider high Knudsen numbers and short channel length (KEY-
OPT(1) = 1) (Sharipov([340] (p. 944))).

7.2. Slide Film

Slide film damping occurs when surfaces move tangentially with respect to each other. Typical applica-
tions of slide film models are damping between fingers of a comb drive and damping between large
horizontally moving plates (seismic mass) and the silicon substrate. Slide film damping can be described
by a nodal force displacement relationship. FLUID139 is used to model slide film fluid behavior and
assumes isothermal viscous flow.

Slide film problems are defined by:

(7.19)

where:

P = pressure
ν = plate fluid velocity
η = dynamic viscosity
z = normal direction of the laterally moving plates
t = time

Slide film problems can be represented by a series connection of mass-damper elements with internal
nodes where each damper represents the viscous shear stress between two fluid layers and each mass
represents its inertial force. The damper elements are defined by:

(7.20)

where:

C = damping coefficient
A = actual overlapping plate area
di = separation between two internal nodes (not the gap separation)

The mass of each internal node is given by:

(7.21)

where:

ρ = fluid density

In case of slip flow boundary conditions (KEYOPT(3) = 1) the fluid velocity at the moving plate is
somewhat smaller than the plate velocity itself. Slip flow BC can be considered by additional damper
elements which are placed outside the slide film whereby the damping coefficient must be:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
234 Published: 2023-10-23
 Slide Film

(7.22)

where:

Lm = mean free path length of the fluid at the current pressure

In case of second order slip flow (KEYOPT(3) = 2) the damping coefficient is:

(7.23)

where Kn is defined with Equation 7.10 (p. 231)

Note that all internal nodes are placed automatically by FLUID139.

Two node models are sufficient for systems where the operating frequency is below the cut-off frequency
which is defined by:

(7.24)

where:

fc = cut-off frequency
d = gap separation

In this special case, damping coefficients are almost constant, regardless of the frequency, and inertial
effects are negligible. At higher frequencies, the damping ratio increases significantly up to a so-called
maximum frequency, which is defined by:
(7.25)

where:

fmax = maximum frequency

The meaning of both numbers is illustrated below:

Figure 7.1: Flow Theory, Cut-off, and Maximum Frequency Interrelation

In case of large signal damping, the current overlapping plate are as defined by:
(7.26)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 235
Thin Fluid Film Flow

where:

Anew = actual area

Ainit = initial area
ui = nodal displacement in operating direction of the first interface node
un = nodal displacement of the second interface node

For rectangular plates which move parallel to its edge, the area change with respect to the plate dis-
placement (dA/du) is equal to the plate width. These applications are typical for micro-electromechan-
ical systems as comb drives where the overlapping area changes with respect to deflection.

7.3. Hydrodynamic Bearing

The form of the Reynold's equation [428] (p. 949) [429] (p. 949) used to describe hydrodynamic lubrication
is deduced from the Navier-Stokes equations considering the following assumptions:

• The flow is laminar and continuous.

• The flow is isoviscous (see note below).

• The flow is incompressible (see note below).

• The fluid inertia effects are negligible.

• The thickness of the fluid is very small when compared to other dimensions of the fluid domain.

• There is no slipping between the fluid film and the shaft and bearing journal walls.

It can be written in Cartesian coordinates as:

(7.27)

where

and = axes of the film plane. is tangent to the bearing and is along the length of the
bearing.
= fluid thickness
= fluid viscosity
= fluid pressure
= wall (corresponding to the shaft journal) velocity along . The bearing sleeve is fixed.

Note:

In most cases, the lubricant is considered isoviscous and incompressible. However, the effect
of variable viscosity and density can be included using the 3D formulation and tabular ma-
terial properties.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
236 Published: 2023-10-23
 Hydrodynamic Bearing

The first term on the right-hand side of Equation 7.27 (p. 236) represents the shear (or slide) effect, while
the second term represents the squeeze effect.

The equation is solved using a finite-length formulation for 2D elements. A negative pressure reflects
the fluid film rupture or cavitation. Since it has no physical meaning, it is set to zero to calculate related
output results, for example pressure forces.

7.3.1. Finite Length Formulation for 2D Elements

Assuming zero pressure at both ends, a solution of Equation 7.27 (p. 236) can be found using a com-
bination of short and modified long bearing pressure expressions [430] (p. 949). In cylindrical coordinates
using non-dimensional quantities, the pressure is given by:

(7.28)

where

= bearing length (real constant L)

= bearing radial clearance (real constant C)
= non-dimensional coordinate along the bearing length
= non-dimensional eccentricity ratio (eccentricity divided by clearance)
= time derivative of the eccentricity ratio
= time derivative of the attitude angle. The attitude angle is the angle of the line of centers
(shaft center to bearing center).
= average rotational velocity. Because the bearing is fixed, it is equal to half of the shaft
rotational velocity .
= cylindrical coordinate with respect to the line of centers
= non-dimensional film thickness (film thickness divided by clearance)

= bearing radius (real constant R)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 237
Thin Fluid Film Flow

Figure 7.2: Hydrodynamic Bearing

Y

lin
e
of
ce
nt
er
s
R+C
lubricant X

ε ϕ
bearing

h ω
shaft

The pressure is integrated over the shaft journal to obtain the pressure forces. Spatial increment in
defaults to 2 degrees. It can be modified using the ThetaInc real constant (see COMBI214 element).

The stiffness and damping coefficients are obtained from the derivative of the pressure forces based
on a small perturbation increment (real constant PertInc) in displacement and velocity.

(7.29)

where

= stiffness coefficients (see Table 7.214.2: COMBI214 Element Output

Definitions for KEYOPT(1) = 1 or 2)
= damping coefficients (see Table 7.214.2: COMBI214 Element Output
Definitions for KEYOPT(1) = 1 or 2)
= pressure force along direction (1), (2)
= displacement in direction (1), (2)
= velocity in direction (1), (2)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
238 Published: 2023-10-23
 Hydrodynamic Bearing

7.3.2. Finite Element Formulation for 3D Element

In a 3D approach, the fluid domain is meshed using surface elements FLUID218. The derivation of
the finite element matrices is similar to the derivation of the acoustic fluid matrices.

For the pressure-only element, the pressure shape functions are used to determine the fluid stiffness
matrix from the left-hand side of Equation 7.27 (p. 236), giving similar to Equation 8.37 (p. 248).
The film thickness h is approximated with the same shape functions to obtain the element internal
fluid forces from the right hand side of Equation 7.27 (p. 236).

For the coupled-field element (PRES and U degrees of freedom), the fluid-structure interaction matrix
(Equation 8.146 (p. 273)) is introduced. The assembled stiffness matrix is then unsymmetric and
expressed as:

(7.30)

where

= structural stiffness matrix (Equation 2.58 (p. 14))

= interaction matrix (Equation 8.146 (p. 273))

= fluid stiffness matrix (Equation 8.37 (p. 248))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 239
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
240 Published: 2023-10-23
Chapter 8: Acoustics
The following topics related to acoustics are available:
8.1. Acoustic Fundamentals
8.2. Derivation of Acoustic Matrices
8.3. Propagation, Radiation, and Scattering of Acoustic Pressure Waves
8.4. Acoustic Fluid-Structural Interaction (FSI)
8.5. Pure Scattered Pressure Formulation
8.6. Acoustic Output Quantities
[Link] Admittance Matrix
8.8. Random Acoustics
8.9. Room Acoustics
[Link] Full Linear Navier-Stokes (FLNS) Model
8.11. Poroelastic Acoustics
8.12. Nonlinear Acoustics

8.1. Acoustic Fundamentals

The following acoustic fluid fundamentals topics are available:
8.1.1. Governing Equations
8.1.2. Finite Element Formulation of the Wave Equation
8.1.3. Governing Equations with Mean Flow Effect
8.1.4. Finite Element Formulation of the Convective Wave Equation
8.1.5. Governing Equations for Axially Symmetric Elements

8.1.1. Governing Equations

In acoustic fluid-structural interaction (FSI) problems, the structural dynamics equation must be con-
sidered along with the Navier-Stokes equations of fluid momentum and the flow continuity equation.

From the law of conservation of mass the flow continuity equation is:
(8.1)

where:

= the velocity vector in the x-, y-, and z-directions

= density

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 241
Acoustics

= mass source (kg/m3t)

= time

From the law of conservation of momentum, the Navier-Stokes equation is:

(8.2)

where:

= viscous stress tensor

= pressure

= body force

The discretized structural dynamics equation can be formulated using the structural elements as
shown in Equation 15.5 (p. 796). The fluid momentum (Navier-Stokes) equations and continuity equations
are simplified to get the acoustic wave equation using the following assumptions:

1. The fluid is compressible (density changes due to pressure variations).

2. The fluid is irrotational.

3. There is no body force.

4. The pressure disturbance of the fluid is small.

5. There is no mean flow of the fluid.

6. The gas is ideal, adiabatic, and reversible.

The linearized continuity equation is:

(8.3)

The linearized Navier-Stokes equation is:

(8.4)

where:

= acoustic velocity
= acoustic pressure

The acoustic wave equation is given by:

(8.5)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
242 Published: 2023-10-23
 Acoustic Fundamentals

c = speed of sound in fluid medium (input as SONC on the MP command)

ρo = mean fluid density (input as DENS via MP)
K = bulk modulus of fluid
μ = dynamic viscosity (input as VISC via MP)
p = acoustic pressure (=p(x, y, z, t))
Q = mass source in the continuity equation (input as MASS via BF)
t = time

Since the viscous dissipation has been taken in account using the Stokes hypothesis, Equa-
tion 8.5 (p. 242) is referred to as the lossy wave equation for propagation of sound in fluids. The dis-
cretized structural Equation 15.5 (p. 796) and the lossy wave Equation 8.5 (p. 242) must be considered
simultaneously in FSI problems. The wave equation will be discretized in the next subsection, followed
by the derivation of the damping matrix to account for the dissipation at the FSI interface. The
acoustic pressure exerting on the structure at the FSI interface will be considered in Derivation of
Acoustic Matrices (p. 248) to form the coupling stiffness matrix.

Harmonically varying pressure is given by:

(8.6)

where:

p = amplitude of the pressure

ω = 2πf
f = frequency of oscillations of the pressure

Equation 8.5 (p. 242) is reduced to the following inhomogeneous Helmholtz equation:

(8.7)

Note that some theories may not be applicable for 2D acoustic elements using Equation 8.7 (p. 243).
See FLUID29 and FLUID129 for available details.

8.1.2. Finite Element Formulation of the Wave Equation

The finite element formulation is obtained by testing wave Equation 8.5 (p. 242) using the Galerkin
procedure (Bathe [2] (p. 925)). Equation 8.5 (p. 242) is multiplied by testing function w and integrated
over the volume of the domain (Zienkiewicz [87] (p. 929)) with some manipulation to yield the following:

(8.8)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 243
Acoustics

where:

dv = volume differential of acoustic domain ΩF

ds = surface differential of acoustic domain boundary ΓF

From the equation of momentum conservation, the normal velocity on the boundary of the acoustic
domain is given by:

(8.9)

Substituting Equation 8.9 (p. 244) into Equation 8.8 (p. 243) yields the “weak” form of Equation 8.5 (p. 242),
given by:

(8.10)

The normal acceleration of the fluid particle can be presented using the normal displacement of the
fluid particle, given by:

(8.11)

where:

= the displacement of the fluid particle

After using Equation 8.11 (p. 244), Equation 8.10 (p. 244) is expressed as:

(8.12)

8.1.3. Governing Equations with Mean Flow Effect

The fluid momentum (Navier-Stokes) equations and continuity equations are simplified to arrive at
the convective wave equation using the following assumptions:

• The fluid is compressible.

• The fluid is irrotational.

• The pressure disturbance of the fluid is small.

• There is no viscous stress.

• The gas is ideal, adiabatic, and reversible.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
244 Published: 2023-10-23
 Acoustic Fundamentals

• There is mean flow effect.

The linearized continuity equation (Equation 8.1 (p. 241)) is:

(8.13)

The linearized Navier-Stokes equation (Equation 8.2 (p. 242)) is:

(8.14)

where:

= mean flow velocity

= body force

With the introduction of velocity potential and force potential,

(8.15)
(8.16)

where:

= the velocity potential

= the force potential

The second order convective wave equation is given by:

(8.17)

where:

The acoustic pressure is derived by:

(8.18)

As an alternative to the pressure formulation in a transient analysis, the velocity potential formulation
(KEYOPT(1) = 4 on the acoustic element) is cast by setting the mean flow to zero in Equa-
tion 8.17 (p. 245) and multiplying by . Either particle velocity or mass source can be applied as the
excitation source for the velocity potential formulation (rather than the acceleration or mass source
rate used for the pressure formulation).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 245
Acoustics

8.1.4. Finite Element Formulation of the Convective Wave Equation

The finite element formulation is obtained by Equation 8.17 (p. 245) using the Galerkin procedure.
Equation 8.17 (p. 245) is multiplied by testing function w and integrated over the volume of the domain
with some manipulation to yield the following:

(8.19)

Harmonically varying velocity potential is given by:

(8.20)

where:

φ = amplitude of the velocity potential

The static mean flow can be obtained with the assumption of irrotational and incompressible flow.
The Laplacian equation, in regard to the velocity potential, is given by:
(8.21)

The known mean flow values on the domain exterior surface are used as the non-zero Neumann
boundary for the Laplacian equation. The zero Neumann boundary is applied to the rest of the domain
exterior surface, which implies that the mean flow is tangential to the surface. The reference potential
φ = 0 is specified by the program for the pure Neumann boundary value problem.

With the assumptions that the mean flow is tangential to the wall and the convective wave equation
is solved in the frequency domain, the boundary condition on the moving boundary is given by:
(8.22)

where:

= outward normal unit vector of the boundary

un = the normal displacement of the boundary

The boundary integral on the moving boundary in Equation 8.19 (p. 246) is cast by:

(8.23)

If the normal velocity vn on the boundary is known, that is:

(8.24)

the boundary integral in Equation 8.19 (p. 246) is given by:

(8.25)

If the impedance boundary condition exists on the moving boundary, that is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
246 Published: 2023-10-23
 Acoustic Fundamentals

(8.26)

where:

Y = admittance on the boundary surface

The boundary integral in Equation 8.19 (p. 246) is given by:

(8.27)

On an FSI interface, the coupled matrix equation is obtained by substituting Equation 8.18 (p. 245)
into Equation 8.146 (p. 273) and using the moving boundary condition given by Equation 8.23 (p. 246).

8.1.5. Governing Equations for Axially Symmetric Elements

For an axisymmetric configuration, the Helmholtz equation is utilized in cylindrical coordinates. The
rotation direction, , contribution is approximated as in the following equation for a variable in the
Helmholtz equation.

(8.28)

Additionally, the testing function is given by

(8.29)

Based upon the given equilibrium, the gradients of p, Q, and w are obtained. After substituting the
reformed variables into Equation 8.12 (p. 244), the weak formulation for an axially-symmetric element
is expressed as:

(8.30)

where:

= mode number
r = distance to rotation axis

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 247
Acoustics

8.2. Derivation of Acoustic Matrices

Equation 8.12 (p. 244) contains the fluid pressure p and the structural displacement components ux,F,
uy,F, and uz,F as the dependent variables to solve. The finite element approximating shape functions for
the spatial variation of the pressure and displacement components are given by:
(8.31)
(8.32)

where:

{N} = element shape function for pressure

{N'} = element shape function for displacements
{Pe} = nodal pressure vector
{ue} = {uxe},{uye},{uze} = nodal displacement component vectors

From Equation 8.31 (p. 248) and Equation 8.32 (p. 248), the second time derivative of the variables and
the virtual change in the pressure can be expressed as follows:
(8.33)

(8.34)

(8.35)

After substituting Equation 8.31 (p. 248) and Equation 8.32 (p. 248) into Equation 8.12 (p. 244), the finite
element statement of the wave Equation 8.5 (p. 242) is expressed as:

(8.36)

where:

{n} = outward normal vector at the fluid boundary

{q} = nodal mass source vector

Other terms are defined in Acoustic Fundamentals (p. 241). Equation 8.36 (p. 248) can be written in matrix
notation to create the following discretized wave equation:
(8.37)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
248 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

8.3. Propagation, Radiation, and Scattering of Acoustic Pressure Waves

The following topics are available in this section:
8.3.1. Acoustic Boundary Conditions
8.3.2. Absorbing Boundary Condition (ABC)
8.3.3. Artificially Matched Layers
8.3.4. Acoustic Excitation Sources
8.3.5. Sophisticated Acoustic Media

8.3.1. Acoustic Boundary Conditions

The following acoustic boundary condition topics are available:
[Link]. Dirichlet and Neumann Boundary Conditions
[Link]. Robin Boundary Condition
[Link]. Floquet Periodic Boundary Condition

[Link]. Dirichlet and Neumann Boundary Conditions

For Dirichlet or Neumann boundary conditions, that is, on a sound-soft boundary with p = p0 (input
as PRES on the D command) or on a sound-hard boundary with , the surface integration
will be eliminated.

[Link]. Robin Boundary Condition

The Robin boundary condition on impedance boundary ΓZ is given by:
(8.38)

where:

vn,F = normal velocity of fluid particle on the boundary

vn,S = normal velocity of structure surface
Y = boundary admittance

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 249
Acoustics

Z (boundary impedance) = 1/Y (Z is input as IMPD on the SF command)

Substituting Equation 8.38 (p. 249) into Equation 8.10 (p. 244) yields:

(8.39)

For simplicity, the inviscid fluid will be investigated. The matrix forms in Equation 8.37 (p. 248) are
rewritten as:
(8.40)

(8.41)

In acoustic design, the attenuation coefficient (that is, the absorption coefficient) is used to define
the absorption properties of the material. The attenuation coefficient is the ratio of the absorbed
sound power density to the incident sound power density. The ratio is expressed as follows:
(8.42)

where:

Because the module of reflection coefficient is given as , the

equivalent impedance (assumed Im(z)=0) of the attenuating material is expressed as:

(8.43)

where:

z0 = ρ0 c 0 = sound impedance

The impedance value of the Robin boundary condition can be defined by the sound impedance z0
= ρ0 c 0 (input as INF on the SF command), the attenuation coefficient (input as ATTN on the SF
command), or the general complex impedance (admittance for modal analysis) (input as IMPD on
the SF command). The Robin boundary condition can be used for the harmonic response or modal
analysis.

[Link]. Floquet Periodic Boundary Condition

The Floquet principle is satisfied when the acoustic wave propagates in a periodic structure (input
as Lab = FPBC on the BF command or Label = PLAN on the APORT command for plane wave
incidence):
(8.44)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
250 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

where:

= period of infinite periodic structure

= complex propagating constant ( ); = attenuation constant; = phase constant
A modified eigenvalue matrix equation solves the phase shift as eigenvalues for acoustic elements
if the phase shift is not equal to 0 or :

(8.45)

where:

( and are defined in Equation 8.37 (p. 248))

= the internal nodes
= the independent nodes for coupled independent/dependent nodes on the boundary
= dependent nodes for coupled independent/dependent nodes on the boundary
as eigenvalue

8.3.2. Absorbing Boundary Condition (ABC)

Exterior structural acoustics problems typically involve a structure submerged in an infinite, homogen-
eous, inviscid fluid. For a source-free and inviscid fluid, Equation 8.5 (p. 242) is simplified as follows:

(8.46)

where:

The pressure wave must satisfy the Sommerfeld radiation condition (which states that the waves
generated within the fluid are outgoing) at infinity. This is expressed as:
(8.47)

where:

r = distance from the origin

d = dimensionality of the problem (that is, d =3 or d =2 if Ω+ is 3D or 2D, respectively)

A primary difficulty encountered when using finite elements to model an infinite medium arises from
the need to satisfy the Sommerfeld radiation condition, as expressed in Equation 8.47 (p. 251). A typ-
ical approach to this problem involves truncating the unbounded domain Ω+ by introducing an ab-
sorbing (artificial) boundary Γa at some distance from the structure (Equation 8.5 (p. 242)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 251
Acoustics

Figure 8.1: Absorbing Boundary

The homogeneous Helmholtz equation (Equation 8.46 (p. 251)) is then solved in the annular region
Ω+, which is bounded by the absorbing boundary Γa. For the resulting problem in Ω+ to be well-
posed, an appropriate condition must be specified on Γa. To this end, the following second-order
conditions are used on Γa (Kallivokas et al. [220] (p. 937)):

In two dimensions:
(8.48)

where:

n = outward normal to Γa
Pn = pressure derivative in the normal direction
Pλλ = pressure derivative along Γa
k = curvature of Γa
γ = stability parameter

In three dimensions:

(8.49)

where:

n = outward normal
u and v = orthogonal curvilinear surface coordinates (e.g., the meridional and polar angles
in spherical coordinates)
Pu, Pv = pressure derivatives in the Γa surface directions
H and K = mean and Gaussian curvature, respectively
E and G = usual coefficients of the first fundamental form

Following a Galerkin based procedure, Equation 8.46 (p. 251) is multiplied by testing function w and
integrated over the annular domain Ωf. By using the divergence theorem on the resulting equation
it can be shown that:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
252 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

(8.50)

Upon discretization of Equation 8.50 (p. 253), the first term on the left hand side will yield the mass
matrix of the fluid, while the second term will yield the stiffness matrix. The following finite element
approximations for quantities on the absorbing boundary Γa placed at a radius R are introduced:
(8.51)

where:

{N} = vectors of shape functions

p, q(1), q(2) = unknown nodal values (p is output as degree of freedom PRES. q(1) and q(2) are
solved for but not output.)

Taking the nodal shape function as the testing function, the element stiffness and damping matrices
can be further reduced. The two dimensional case is reduced to:

(8.52)

(8.53)

where:

R = radius of absorbing boundary (see Figure 8.1: Absorbing Boundary (p. 252))

dl = arc-length differential

These matrices are 6 x 6 in size, having 2 or 3 nodes per element with 3 degrees of freedom per node
(p, q(1), q(2)). (See FLUID129 for details.)

The three dimensional case is reduced to:

(8.54)

(8.55)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 253
Acoustics

ds = area differential

For the three dimensional case, the surface element can be a quadratic or triangle element with
corner and mid-nodes that have 2 degrees of freedom per node (p, q) (Barry et al. [219] (p. 937)). (See
FLUID130 for details.)

For the axisymmetric case:

(8.56)

(8.57)

where:

r = radius

These matrices are 4x4 in size with 2 or 3 nodes per element and 2 degrees of freedom per node (p,
q).

The local absorbing boundary conditions in this section are applicable to modal, harmonic response,
or transient analyses.

8.3.3. Artificially Matched Layers

Artificially matched layers are artificial anisotropic materials that absorb all incoming waves without
any reflections, except for the grazing wave that parallels the artificially-matched-layers interface in
the propagation direction. Artificially matched layers are currently constructed by the propagation
of acoustic waves in the homogeneous isotropic media for harmonic response analysis (Zhao
[403] (p. 948)).

[Link]. Perfectly Matched Layers (PML)

The perfectly matched layers (PML) are located inside a cubic enclosure. The PML materials are
specified in face, edge, and corner regions in a three-dimensional model.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
254 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

In the frequency domain, the governing lossless and source-free momentum and conservation are
given by:
(8.58)

(8.59)

The anisotropic PML material is assumed as follows:

(8.60)

(8.61)

where:

[Λ] = diagonal tensor diag{sx, sy, sz}, and sx, sy, sz are complex constants.

In order to satisfy the vector operation, the governing acoustic equations for PML are written as:
(8.62)
(8.63)

If the modified operator is expressed as:

(8.64)

Then the wave equation for a PML with a modified differential operator is expressed as:

(8.65)

The interface between the PML and the real acoustic medium can be rendered reflectionless for all
frequencies and angles of incidence of a wave propagating from the acoustic medium to the PML
if the elements of [Λ] are defined as follows:
(8.66)

where:

σi = attenuation constant

Multiplying sxsysz with Equation 8.65 (p. 255) results in the final wave equation needed to construct
the PML:

(8.67)

where:

[Ψ] = diagonal tensor diag{sysz/sx, szsx/sy, sxsy/sz}

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 255
Acoustics

By using the Galerkin method and setting the pressure to zero (input as PRES on the D command)
on the backed boundary, the "weak" formulation of an acoustic pressure wave in PML is expressed
as:

(8.68)

Equation 8.68 (p. 256) can be written in matrix notation to produce the following discretized wave
equation:
(8.69)

where:

SRe = real part of complex constant sxsysz

SIm = imaginary part of complex constant sxsysz
[ΨRe] = real part of complex tensor [Ψ]
[ΨIm] = imaginary part of complex tensor [Ψ]

The PML material is defined using FLUID30, FLUID220, and FLUID221 elements with KEYOPT(4) =
1. Because the PML material is constructed in Cartesian coordinates, the edges of the 3D PML region
must be aligned with the axes of the global Cartesian coordinate system.

More than one 1D PML region may exist in a finite element model. The PML element coordinate
system (PSYS) uniquely identifies each PML region. A parabolic attenuation distribution is used to
minimize numerical reflections in the PML. Ansys, Inc. recommends using at least three PML element
layers to obtain satisfactory accuracy. Some buffer elements between the PML region and objects
should be utilized.

Because a PML acts as an infinite open domain, all boundary conditions and material properties
need to be carried over into the PML region. The pressure must be set to zero (input as PRES on
the D command) on the exterior surface of the PML, excluding Neumann boundaries (symmetric
planes in the model). All types of excitation sources are prohibited in the PML region.

In transient analysis, the additional auxiliary variables are introduced in Equation 8.67 (p. 255), referred
to by Assi [442] (p. 950).

Refer to Artificially Matched Layers in the Acoustic Analysis Guide for more information about using
PML.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
256 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

[Link]. Perfectly Matched Layer (PML) for Axially Symmetric Elements

PML for axisymmetric elements are modeled in two-dimensional acoustic elements. The PML located
in a cylindrical enclosure represents the rotation of two-dimensional elements in the rotation axis.
Therefore, the equations employed for axisymmetric PML are derived in cylindrical coordinates.

For axisymmetric PML, the modified operator, , is defined in cylindrical coordinates (r, θ, i) as
follows:
(8.70)

where and are complex constants, and is given by:

(8.71)

where is the distance to the rotation axis, and is the distance from the interface between
PML and the acoustic medium to the rotation axis. By using Galerkin method, the "weak" formulation
of the acoustic pressure wave in axisymmetric PML is demonstrated as:

(8.72)

where = mode number.

[Link]. Irregular Perfectly Matched Layers (IPML)

Irregular perfectly matched layers (IPML) are located inside a convex enclosure. The IPML material
is specified based on the local normal direction of the IPML interface. The pressure wave is assumed
to propagate in the z'-direction of the local Cartesian coordinates (x', y', z'), which is set on the IPML
interface by the program.

The wave equation for IPML with a modified differential operator is:

(8.73)

The modified operator is defined by:

(8.74)

The complex IPML parameter is constructed by:

(8.75)

Multiplying sz by Equation 8.73 (p. 257) results in the final wave equation for IPML as follows:

(8.76)

where:

= diagonal tensor diag{sz, sz, 1/sz}

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 257
Acoustics

The "weak" formulation of an acoustic pressure wave in IPML can be obtained with the operator
and complex tensor in the local Cartesian coordinate system:

(8.77)

The matrix notation for IPML is similar to Equation 8.69 (p. 256), except for using the local operator
and the complex tensor .

The IPML material is defined using FLUID30, FLUID220, and FLUID221 elements with KEYOPT(4) =
2. The IPML region must be a convex shape. The IPML is not related to any user-defined global or
local coordinate system defined.

More than one 1D IPML region may exist in a finite element model. A parabolic attenuation distri-
bution is used to minimize numerical reflections in the IPML. Ansys, Inc. recommends using at least
three IPML element layers to obtain satisfactory accuracy and the uniform thickness of the IPML
region. Some buffer elements between the IPML region and objects should be utilized.

All boundary conditions and material properties need to be carried over into the IPML region. The
pressure must be set to zero (input as PRES on the D command) on the exterior surface of the IPML,
excluding Neumann boundaries (rigid walls in the model). All types of excitation sources are pro-
hibited in the IPML region. The program can automatically apply zero pressure to the exterior surface
of the IPML region, but the rigid walls (the Neumann boundaries) have to be flagged (with the
SF,,RIGW command), if they exist in the model.

In transient analysis, the additional auxiliary variables are introduced in Equation 8.76 (p. 257) with
the coordinate space transformation. Refer to Assi [442] (p. 950) for details.

Refer to Artificially Matched Layers in the Acoustic Analysis Guide for more information about using
IPML.

8.3.4. Acoustic Excitation Sources

Sound excitation sources are a fundamental part of an acoustic analysis. Various excitation sources,
such as a monopole or plane wave can be imposed in simulations. The following methods can be
used to introduce sound excitation sources:

• Specified pressure or normal velocity on the domain boundary (p. 259)

• Mass source in the wave (p. 259) Equation 8.5 (p. 242)

• Analytic wave sources (p. 260):

– Incident pressure applied on the exterior boundary (Robin boundary condition)

– Incident pressure applied on the PML interface

– Solve the wave equation with pure scattered pressure as a variable (Pure Scattered
Pressure Formulation (p. 275))

• Analytic port modes in a duct (p. 263):

– Obliquely incident plane wave

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
258 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

– Rectangular duct

– Circular duct

• Power Source in the Diffusion Equation (p. 264)

[Link]. Specified Pressure or Normal Velocity on the Domain Boundary

Specified pressure (input as PRES on the D command) or normal velocity on the domain boundary
offers a straightforward method to introduce the excitation source for harmonic response analysis
(normal acceleration for transient analysis) (input as SHLD on the SF command) or the velocity for
harmonic response analysis (acceleration for transient analysis) (input as EF on the BF command)
(Equation 8.10 (p. 244)) .

[Link]. Mass Source in the Wave Equation

A mass source in the wave Equation 8.5 (p. 242) (input as MASS on the BF command) can be used
to introduce excitations for a pressure wave.

A point mass source is given by:

(8.78)

where:

Qp = magnitude of point mass source with unit [kg/s]

(xs, ys, zs) = source origin
δ = Dirac delta function

A line mass source is given by:

(8.79)

where:

Ql = magnitude of line mass source with unit [kg/(ms)]

A surface mass source is given by:

(8.80)

where:

Qs = magnitude of surface mass source with unit [kg/(m2s)]

The volumetric mass source Q in Equation 8.8 (p. 243) has the unit [kg/(m3s)].

Depending on how the load is applied on the finite element, the program will interpret the load
as a point, line, surface, or volumetric mass source. For a transient analysis, the mass source rate
( ) in Equation 8.8 (p. 243) should be defined.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 259
Acoustics

[Link]. Analytic Wave Sources

The various incident wave sources are defined using the AWAVE command for harmonic response
analysis.

A time-harmonic incident planar wave (input as PLAN on the AWAVE command) is given by:
(8.81)

where:

Pinc,0= magnitude of incident planar wave

φ = initial phase angle of incident planar wave (usually ignored)

The particle velocity is given by:

(8.82)

A time-harmonic radiated monopole pressure (input as MONO on the AWAVE command) is intro-
duced by:

(8.83)

where:

k = wave number

If the normal velocity is along the direction, the velocity of particle is given by:

(8.84)

Assuming that the monopole is a pulsating sphere (r=a) with surface pressure pa, the pressure can
be expressed as:

(8.85)

If the normal velocity on the pulsating sphere surface is equal to va, the pressure is expressed as:

(8.86)

Other applicable incident sources (Whitaker [[405] (p. 948)]) are also defined by the AWAVE command.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
260 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

To apply the incident wave to the model, the splitting pressure should be introduced on the exter-
ior excitation surface when the total pressure is solved by FEM. The total pressure consists of the
incident pressure pinc and scattered pressure psc. This can be expressed as:
(8.87)

For a planar wave, the incident wave and scattered wave can be written respectively as:
(8.88)
(8.89)

where:

Assuming the normal scattered wave vector is in the opposite direction of the normal incident
wave vector, this can be expressed as:
(8.90)

After substituting Equation 8.82 (p. 260) into the boundary integration of Equation 8.8 (p. 243), this
can be expressed as:

(8.91)

where:

The damping matrix and source vector in Equation 8.37 (p. 248) for lossless and source-free fluid
with the planar incident wave on the exterior boundary is given by:

(8.92)

(8.93)

A similar procedure can be used for a spherical incident wave, but the scattered wave is assumed
to be a planar wave in the direction. The damping matrix and source vector in Equa-
tion 8.37 (p. 248) for lossless and source-free fluid with the spherical incident wave on the exterior
boundary is given by:

(8.94)

(8.95)

If we assume that the incident wave perpendicularly projects on the transparent exterior surface
of the model, the analogous derivation can be achieved with the Robin boundary condition

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 261
Acoustics

(Equation 8.38 (p. 249)), using the wave admittance Y0 as the impedance boundary. Since the incoming
incident wave is propagating into the model, the Robin boundary condition is expressed as:
(8.96)

where:

Assuming that the scattered wave has the same wave admittance as the incident wave, the Robin
boundary condition for the outgoing scattered wave is expressed as:
(8.97)

where:

Substituting Equation 8.96 (p. 262) into the boundary integration of Equation 8.8 (p. 243) and utilizing
Equation 8.97 (p. 262) leads to:

(8.98)

The damping matrix and source vector in Equation 8.37 (p. 248) for lossless and source-free fluid
with the incident excitation on the impedance boundary is expressed as:
(8.99)

(8.100)

The above equations are useful for the analysis of the wave scattering and the pipe structure, re-
spectively.

The planar scattered pressure wave propagating along the opposite direction of the incident wave
is assumed to be on the exterior surface of the model, while the incident wave projects into the
simulation domain from outside of the model. Obviously, this assumption leads to a numerical error
once the non-planar scattered wave occurs. Hence, the PML is used to absorb the arbitrary scattered
outgoing waves.

Assuming that the PML region is a scattered pressure region and that other regions are total pressure
regions, the "weak" forms for the total pressure and the scattered pressure are respectively written
as:

(8.101)

In PML scattered pressure region:

(8.102)

Total pressure can be split into scattered pressure and incident pressure on the interface between
the general acoustic region and the PML region. This can be expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
262 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

(8.103)

The degrees of freedom on nodes with PML attributes are scattered pressure. The surface integration
of scattered pressure on the interface will be cancelled out in Equation 8.101 (p. 262) and Equa-
tion 8.102 (p. 262). Assuming that the inside model is source free, the general matrix notation in the
normal acoustic region, including the pressure on interface nodes, can be written as:
(8.104)

Substituting Equation 8.103 (p. 263) into Equation 8.104 (p. 263) leads to the matrix equation for the
interior total pressure and the interface scattered pressure, expressed as:

(8.105)

The matrix equation for whole domain is expressed as:

(8.106)

where:

= complex stiffness matrix for interior total pressure from Equation 8.105 (p. 263)
= complex stiffness matrix on interface from Equation 8.105 (p. 263)
= complex stiffness matrix for interface scattered pressure from Equation 8.105 (p. 263)
= complex stiffness matrix for PML interior scattered pressure from Equation 8.69 (p. 256)
finc = complex source vector from Equation 8.105 (p. 263)

[Link]. Analytic Port Modes in a Duct

The various analytic modes in a guided duct are defined using the APORT command for harmonic
response analysis.

A time-harmonic incident planar wave (Label = PLAN on APORT) is given by:

(8.107)

where:

= complex initial pressure

= vector wave number

= position vector

A mode existing in a rectangular duct (Label = RECT on APORT) is given by:

(8.108)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 263
Acoustics

where:

= width of duct
= height of duct
= propagating constant in z-direction

A mode existing in a circular duct (Label = CIRC on APORT) is given by:

(8.109)

where:

= mth order Bessel function of the first kind

= nth zero of the derivative of the mth order Bessel function of the first kind
= radius of the circular duct

A mode existing in a coaxial duct (Label = COAX on APORT) is given by:

(8.110)

where:

= coefficients determined by pressure and rigid wall boundary condition at the

inner radius r = a
= mth order Bessel function of the first kind

= mth order Bessel function of the second kind

= nth zero of the equation
= inner radius of the coaxial duct
= outer radius of the coaxial duct

[Link]. Power Source in the Diffusion Equation

A power source in the wave equation (Equation 8.235 (p. 290)) (input as MASS on the BF command)
can be used to introduce excitations for a diffusion equation solution.

A point power source is given by:

(8.111)

where:

Pp = magnitude of point power source with units of [W]

(xs, ys, zs) = source origin
δ = Dirac delta function

A line power source is given by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
264 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

(8.112)

where:

Pl = magnitude of line power source with units of [W/m]

A surface power source is given by:

(8.113)

where:

Ps = magnitude of surface power source with units of [W/m2]

The volumetric power source P in Equation 8.235 (p. 290) has units of [W/m3].

8.3.5. Sophisticated Acoustic Media

The following topics are available in this section:
[Link]. Non-uniform Acoustic Media
[Link]. Equivalent Fluid of Perforated Materials
[Link]. Impedance Sheet Approximation
[Link]. Viscous-Thermal Media

[Link]. Non-uniform Acoustic Media

In non-uniform acoustic media the mass density and sound speed vary with the spatial position.
Wave Equation 8.5 (p. 242) in lossless media is rewritten as:

(8.114)

According to the ideal gas law, the equation of state and the speed of sound in an ideal gas are
respectively expressed as:
(8.115)
(8.116)

where:

CP = heat coefficient at a constant pressure per unit mass (input as C on the MP command)
CV = heat coefficient at a constant volume per unit mass (input as CVH on the MP command)
R = gas constant
T = temperature
Pstate = the absolute pressure of the gas measured in atmospheres

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 265
Acoustics

Assuming density ρ0 (input as DENS on the MP command) and sound speed c0 (input as SONC on
the MP command) at the reference temperature T0 (input on TREF command), and the reference
static pressure (input as Psref on R command) the density and sound speed in media are cast as
follows:

(8.117)

(8.118)

The spatial temperature is input as TEMP on BF command and the nodal static pressure is input
as SPRE on the BF command.

[Link]. Equivalent Fluid of Perforated Materials

[Link].1. Johnson-Champoux-Allard Model
Assuming that the skeleton of the perforated material is rigid, the perforated material can be
approximated using the Johnson-Champoux-Allard equivalent fluid model, which uses the complex
effective density and velocity (Allard [404] (p. 948)). The wave equation with complex effective
density and velocity is given by:

(8.119)

The effective density is given by:

(8.120)

where:

ω = angular frequency
σ = fluid resistivity
Φ = porosity
α ∞ = tortuosity
Λ = viscous characteristic length
ρ0 = density of fluid
η = dynamic (shear) viscosity

The effective bulk modulus is given by:

(8.121)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
266 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

γ = specific heat ratio

P0 = static reference pressure (input as PSREF on the R command)
P rt=Cpη/κ = Prandtl number
Cp = specific heat
κ = thermal conductivity
Λ’ = thermal characteristic length

The material coefficients used in the Johnson-Champoux-Allard model are input with the TBDATA
command for the TB,PERF material as well as through MP commands.

The complex effective velocity is given by:

(8.122)

The sound impedance is given by:

(8.123)

The “weak” form of the Johnson-Champoux-Allard equivalent fluid model is given by:

(8.124)

The matrix notation is expressed as:

(8.125)

where:

The final matrix equation is expressed as:

(8.126)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 267
Acoustics

[Link].2. Delany-Bazley and Miki Models

Instead of inputting five parameters in the Johnson-Champoux-Allard model, only one parameter,
fluid resistivity, is required in the Delany-Bazley and Miki models. In general, complex impedance
is defined by:
(8.127)

The complex propagation constant s is defined by

(8.128)

where:

R = resistance
X = reactance
α = attenuation constant
β = phase constant

In both the Delany-Bazley and the Miki models, variables are approximated by the functions:

(8.129)

(8.130)

(8.131)

(8.132)

where:

f = frequency
σ = fluid resistivity

Coefficients are provided in the following table.

Table 8.1: Coefficients of the Approximation Functions for Delany-Bazley and Miki Models

Coefficient Delany-Bazley Miki

a 0.0497 0.0699
b -0.754 -0.632
c 0.0758 0.107
d -0.732 -0.632
q 0.169 0.160
r -0.618 -0.618
s 0.0858 0.109
t -0.700 -0.618

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
268 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

The proposed working range for the Delany-Bazley model is 0.01 < < 1.00. The Miki model can
be extended to 0.01 < .

[Link].3. Complex Media Properties

The complex effective mass density and sound velocity in Equation 8.119 (p. 266) may be given
by:
(8.133)
(8.134)

The Helmholtz equation with complex mass density and sound velocity is directly solved with
the defined parameters.

The complex effective mass density and sound velocity are input with the TBDATA command
for the TB,PERF material.

[Link]. Impedance Sheet Approximation

Analogous to Equation 8.38 (p. 249), the boundary condition across an impedance sheet in the
acoustic domain is given by:
(8.135)

where:

Assuming that the acoustic domain is divided into Ω+ and Ω- with the impedance sheet on the
surface, substituting Equation 8.135 (p. 269) into Equation 8.10 (p. 244) casts the damping matrix as
Equation 8.40 (p. 250). For a thin perforated layer, the sheet impedance (input as IMPD on the BF
command) is given by:

(8.136)

where:

ZL = load impedance
d = thickness of perforated layer
= sound impedance of the perforated material defined by Equation 8.123 (p. 267)

= complex wave number of the perforated material (ceff is defined by Equa-

tion 8.122 (p. 267))

For a thin perforated material layer backed with a rigid wall, the impedance is given by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 269
Acoustics

(8.137)

[Link]. Viscous-Thermal Media

An acoustic waves propagating in viscous-thermal media have a complex propagating constant in
the frequency domain. The attenuation of the acoustic wave is proportional to the shear and bulk
viscosity and the thermal conduction coefficient of the media.

[Link].1. Boundary Layer Impedance (BLI) Model

In a rigid-walled, bounded domain, the interaction between he acoustic movement and both the
diffusion of heat and the diffusion of the shear wave leads to a reactive and absorbing process
on the walls. The entropic and vortical perturbations diffuse into the rigid wall in a direction
normal to the boundary and decay to zero depending on the thickness of the boundary layers.
A hybrid numerical and analytic solution using the boundary layer impedance (BLI) model is
proposed with the following assumptions ([411] (p. 948)):

• The thicknesses of the boundary layers are much smaller than the dimensions of the numerical
domain, while the pressure variation is constant over the boundary layers’ thicknesses.

• The acoustic pressure is approximated by its zero expansion in the boundary layers and its
flow is tangential to the wall.

• Spatial variations of both velocity and temperature in the normal direction are much greater
than spatial variations in the tangential directions.

The fundamental Helmholtz governing equation and boundary condition in the BLI model is
given by:

(8.138)

where:

k0 = ω / c0
lvh = lv+(γ-1)lh
kw = the transverse wavenumber

l’v = μ / (ρ0c0)
lh = λ / (ρ0c0Cp)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
270 Published: 2023-10-23
 Propagation, Radiation, and Scattering of Acoustic Pressure Waves

γ = Cp / Cv = the specific heat ratio

ω = the angular frequency
ρ0 = the mass density of the fluid
c0 = the sound speed of the fluid
μ = the dynamic viscosity
η = the bulk viscosity
λ = the thermal conductivity
Cp = the heat coefficient at constant pressure per unit mass
Cv = the heat coefficient at constant volume per unit mass

[Link].2. Low Reduced Frequency (LRF) Model

The low reduced frequency (LRF) model is derived from simplified Navier-Stokes equations with
the following assumptions ([412] (p. 948)):

• The acoustic wavelength is much greater than the length scale of the geometry

• The acoustic wavelength is much greater than the boundary layer thickness

The Helmholtz wave equation with modified density and bulk module for the LRF model is as
follows:

(8.139)

where:

ρ0 = the mass density

K0 = the modified bulk module
l = the length scale

σ = Cpη / κ = Prandtl number

γ = specific heat ratio
The viscous and thermal effects are taken into account by the function B(s/l) and n(sσ/l), respect-
ively.

The analytic solution of both the functions B(s/l) and n(sσ/l) can be derived for three particular
cases: thin layer, tube with rectangular cross-section, and tube with circular cross section.

A thin layer is specified by thickness h. The functions B(s/l) and n(sσ/l) are derived by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 271
Acoustics

(8.140)

(8.141)

where:

A tube with a rectangular cross-section is specified by width w and height h. The functions B(s/l)
and n(sσ/l) are derived by:

(8.142)

(8.143)

where:

A tube with a circular cross section is specified by radius R. The functions B(s/l) and n(sσ/l) are
derived by:

(8.144)

(8.145)

where:

J0 = the zero-order Bessel function

J2 = the second-order Bessel function

8.4. Acoustic Fluid-Structural Interaction (FSI)

The following topics are available in this section:
8.4.1. Coupled Acoustic Fluid-Structural System with an Unsymmetric Matrix Equation

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
272 Published: 2023-10-23
 Acoustic Fluid-Structural Interaction (FSI)

8.4.2. Coupled Acoustic Fluid-Structural System with Symmetric Matrix Equation for Full Harmonic Analysis

8.4.1. Coupled Acoustic Fluid-Structural System with an Unsymmetric Matrix

Equation
The coupling conditions on the interface between the acoustic fluid and the structure are given by:
(8.146)
(8.147)

where:

p = acoustic pressure

Equation 8.146 (p. 273) is a kinetic condition relating the solid stress to the pressure imposed on the
interface by sound. Equation 8.147 (p. 273) is a kinematic condition that assumes that there is no friction
between the solid and acoustic fluid on the interface.

In order to completely describe the FSI problem, the fluid pressure load acting at the interface is added
to Equation 15.6 (p. 797). This effect is included in FLUID29, FLUID30, FLUID220, and FLUID221 only if
KEYOPT(2) ≠ 1. Hence, the structural equation is rewritten as:
(8.148)

The fluid pressure load vector at the interface S is obtained by integrating the pressure over
the area of the surface as follows:
(8.149)

where:

{N'} = shape functions employed to discretize the displacement components u, v, and w

(obtained from the structural element).

Substituting the finite element approximating function for pressure given by Equation 8.34 (p. 248)
into Equation 8.149 (p. 273) leads to:
(8.150)

By comparing the integral in Equation 8.150 (p. 273) with the matrix definition of [R]T in Equa-
tion 8.37 (p. 248), the following relation becomes clear:
(8.151)

Substituting Equation 8.151 (p. 273) into Equation 8.148 (p. 273) results in the dynamic elemental
equation of the structure, expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 273
Acoustics

(8.152)

Equation 8.37 (p. 248) and Equation 8.152 (p. 274) describe the complete finite element discretized
equations for the FSI problem. These equations are written in assembled form as:

(8.153)

The acoustic fluid element in an FSI problem will generate all the submatrices with a superscript F in
addition to the coupling submatrices , [R]T, and [R]. Submatrices with a superscript S will be generated
by the compatible structural element used in the model.

Assuming that the actual surface is at an elevation η relative to the mean surface in z-direction, the
pressure for a sloshing (free) surface is given by:
(8.154)

By utilizing the definition of velocity and the momentum conservation equation in addition to Equa-
tion 8.154 (p. 274), pressure can be expressed as:
(8.155)

The surface integration of the "weak" form (Equation 8.37 (p. 248)) on the sloshing surface is given by:

(8.156)

The acoustic fluid matrix equation with sloshing effect is expressed as:
(8.157)

where:

Substituting Equation 8.157 (p. 274) into Equation 8.153 (p. 274) yields:

(8.158)

If the impedance boundary is exerted on the FSI interface (input as IMPD on the SF command), the
coupling condition expressed in Equation 8.147 (p. 273) is rewritten as:
(8.159)

Substituting Equation 8.159 (p. 274) into Equation 8.37 (p. 248) yields:
(8.160)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
274 Published: 2023-10-23
 Pure Scattered Pressure Formulation

Damping matrix [CFSI] on the impedance FSI interface has been shown to be the same as the damping
matrix in Equation 8.40 (p. 250). Therefore, the coupling matrix expressed in Equation 8.158 (p. 274)
can still be the final matrix equation. In an incompressible fluid the fluid density is independent of
the pressure. This implies the speed of sound equivalently tends toward infinity. The matrix in
Equation 8.158 (p. 274) is set to zero.

8.4.2. Coupled Acoustic Fluid-Structural System with Symmetric Matrix

Equation for Full Harmonic Analysis
The matrix in Equation 8.158 (p. 274) has been shown to be unsymmetric. Solving Equation 8.158 (p. 274)
may consume more computer resources and time than solving the symmetric matrix equation. For
the frequency domain, assume that:
(8.161)

Substituting Equation 8.161 (p. 275) into Equation 8.152 (p. 274) and Equation 8.37 (p. 248) yields:
(8.162)
(8.163)

Dividing coupled Equation 8.163 (p. 275) by yields the acoustic matrix equation, written as:

(8.164)

The coupled matrix equation is given by:

(8.165)

After solving Equation 8.165 (p. 275), the pressure is obtained using Equation 8.161 (p. 275).

8.5. Pure Scattered Pressure Formulation

An alternative method to simulate the pressure wave excited by the analytic incident wave sources
(defined using the AWAVE command), is to use scattered pressure as the solved variable. Pure scattered
pressure formulation may be used in the following situations:

• The scattered pressure is much smaller than the incident pressure wave, and the scattered properties
are under the investigation.

• The analytic wave sources defined using the AWAVE command are inside the model for real acoustic
devices, such as an enclosed-back loudspeaker.

When the scattered pressure is investigated as the degree of freedom, the total pressure is split into
incident pressure and scattered pressure. This can be expressed as:
(8.166)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 275
Acoustics

The pressure boundary conditions are also changed for the scattered pressure.

For the Dirichlet boundary (p =p0):

(8.167)

For the Neumann boundary :

(8.168)

The boundary of each decomposed domain is assumed to consist of the truncation surface Γext, imped-
ance surface ΓZ, and FSI interface ΓFSI. This can be expressed as:
(8.169)

The "weak" form for the scattered pressure derived from Equation 8.37 (p. 248) is written as:

(8.170)

By decomposing Equation 8.170 (p. 276) into the environment medium (Ω0) and non-environment me-
dium (Ωd), and assuming that the incident pressure wave satisfies the wave equation in the environment
medium, we can state that:
(8.171)

The "weak" form of the incident wave including acoustic-structure interaction can be written as:

(8.172)

Substituting Equation 8.170 (p. 276) and Equation 8.172 (p. 276) into Equation 8.37 (p. 248) yields the
"weak" form for pure scattered pressure based formulation, expressed as:

(8.173)

By combining the Neumann boundary condition ΓN (expressed in Equation 8.167 (p. 276) ) with Equa-
tion 8.173 (p. 276), the matrix form is written as:
(8.174)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
276 Published: 2023-10-23
 Acoustic Output Quantities

Since the scattered pressure is solved in the acoustic domain and the total pressure is coupled with the
structural domain, the total pressure in the structural matrix in Equation 8.152 (p. 274) must be split to
ensure a consistent pressure degree of freedom in the solution. This is expressed as:
(8.175)

The coupling matrix equation is given by:

(8.176)

Equation 8.69 (p. 256) is used for the scattered pressure in the PML, because only scattered pressure is
absorbed and the incident is assumed to satisfy the wave equation in the environment medium.

8.6. Acoustic Output Quantities

The following topics related to acoustic output properties are available:
8.6.1. Sound Pressure
8.6.2. Far-field Parameters
8.6.3. Sound Power
8.6.4. Acoustic Surface Quantities
8.6.5. Acoustic Volumetric Quantities
8.6.6. Band Sound Pressure Level
8.6.7. Equivalent Radiated Power

8.6.1. Sound Pressure

Either the total nodal pressure or the scattered nodal pressure is solved and output at the nodes.
Using the calculated nodal pressure values, the pressure gradient is evaluated at the nodes and the
element centroid, which is expressed as follows:
(8.177)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 277
Acoustics

Both the nodal velocity (output as PG on the PRNSOL, PLNSOL, and PLVECT commands) and the
element centroid velocity (output as SMISC on the element table) are calculated for modal (AN-
TYPE,MODAL) and full harmonic (ANTYPE,HARM) analyses as follows:

(8.178)

The nodal and element centroid sound pressure levels (SPL) are output in the post-processor. SPL is
given by:

(8.179)

where:

(input as PREF on the R command, defaults to 20 x 10-6Pa)

In the frequency domain, the time-average root mean square quantities are calculated over one
period of the sinusoidal function. In the time domain, the time-average root mean square quantities
are calculated over the time increment .

The A-weighted sound pressure level (dBA) in the frequency domain is defined by:
(8.180)

where:

In room acoustics, the sound pressure level (SPL) is defined as:

(8.181)

where:

= acoustic energy density

ρ = mass density of fluid
c = speed of sound

The sound intensity is cast by both pressure and velocity:

(8.182)

In the frequency domain, the time-average sound intensity is given by:

(8.183)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
278 Published: 2023-10-23
 Acoustic Output Quantities

p = complex pressure
= complex conjugate velocity

8.6.2. Far-field Parameters

The far-field pressure parameters (output via the PLFAR and PRFAR command) outside of the model
are expressed as:

(8.184)

where:

= enclosed equivalent source surface (defined as MXWF on SF command).

= Green's function in radiation space (defaults to free space). It is defined as follows

for the 3D and 2D cases, respectively:
(8.185)

(8.186)

= far-field observation position

= equivalent source position on the enclosed surface
k = wave number (k is equal to zero for an incompressible fluid)
H0(2) = zeroth order Hankel function of second kind

The Rayleigh integral predicts the radiated far-field pressure from a vibrating panel with an infinite
baffle:

(8.187)

where:

= fluid density
= angular velocity
= normal displacement

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 279
Acoustics

Figure 8.2: Equivalent Source Principle

All acoustic far-field parameters are evaluated in the global spherical coordinate system shown in the
following figure:

Figure 8.3: Spherical Coordinates

8.6.3. Sound Power

The sound power level radiated by the sound source is expressed as:

(8.188)

where:

Wref = reference sound power (defaults to 10-12 w)

Wrad = radiated sound power

The radiated sound power is expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
280 Published: 2023-10-23
 Acoustic Output Quantities

(8.189)

The directivity of the sound source is expressed as:

(8.190)

where:

The directivity of the sound source is also expressed as:

(8.191)

The target strength (TS) is used to describe the reflected acoustic signals. Target strength is given by:
(8.192)

where:

R = radius at the target position

psc = scattered pressure at the target position

The time-averaged sound power through an area S in a harmonic analysis is defined by:
(8.193)

In a network, the return loss and transmission loss are respectively defined as:
(8.194)

(8.195)

where:

Pin = incident sound power at inlet

Pr = reflected sound power at inlet
Pt = transmitted sound power at outlet

The sound power parameters are output via the PRFAR, PLFAR, PRAS or PLAS command.

8.6.4. Acoustic Surface Quantities

The acoustic surface quantities on the selected surface are defined as follows (and are output via the
PRAS and PLAS commands):

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 281
Acoustics

Specified acoustic impedance:

Acoustic impedance:

Mechanical impedance:

where:

= pressure on the surface

= normal component of velocity on the surface
= surface area

8.6.5. Acoustic Volumetric Quantities

The acoustic volumetric quantities on the selected elements are defined as follows (output by the
PRAS and PLAS commands).

Acoustic time-averaged potential energy:

(8.196)

Acoustic time-averaged kinetic energy:

(8.197)

Average square of the L2 norm of pressure:

(8.198)

where:

= root mean square of acoustic pressure

= speed of sound
= mass density

= root mean square of particle velocity

8.6.6. Band Sound Pressure Level

If the frequency range is equally divided into segments and the pressures are calculated
at the center of the frequency segment , the sound pressure level (SPL) of the frequency range is
defined as follows (and output via the PRAS and PLAS commands):

(8.199)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
282 Published: 2023-10-23
 Transfer Admittance Matrix

where:

= SPL of frequency segment .

The A-weighted SPL of the frequency range is written as:

(8.200)

8.6.7. Equivalent Radiated Power

The equivalent radiated power (ERP) that estimates the radiated structure-borne sound power from
the vibrating structural surface is defined as follows:
(8.201)

where:

= speed of sound
= mass density
= radiation factor
= normal velocity of vibrating structural surface

8.7. Transfer Admittance Matrix

A complicated structure may be trimmed and represented by a 2×2 transfer admittance matrix that is
a part of the finite element model.

Figure 8.4: Two-Port Transfer Admittance Matrix

The transfer admittance matrix is given by:

(8.202)

where:

p1 = pressure at port 1
vn1 = normal velocity at port 1

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 283
Acoustics

p2 = pressure at port 2
vn2 = normal velocity at port 2
Y11, Y22 = self-admittances
Y12, Y21 = mutual admittances
α1, α2 = internal sources (usually equal to zero in an acoustic analysis)

The following related topics are available:

[Link] Admittance Matrix Connected to Acoustic Domains
[Link] Admittance Matrix Connected to the Structural and Acoustic Domain

8.7.1. Transfer Admittance Matrix Connected to Acoustic Domains

When two acoustic domains are separated by a transfer admittance matrix, the boundary surface in-
tegrations with the normal velocity on the S1 and S2 surfaces in the “weak” form (Equa-
tion 8.8 (p. 243)-Equation 8.10 (p. 244)) are given by:
(8.203)

(8.204)

where:

[C] = damping matrix

The outward normal vectors o the integration surface point toward the inside of the transfer admittance
matrix. It is consistent with the defined direction of normal velocity in the transfer admittance matrix.
Therefore, the velocity vector can be replaced by the pressure via the transfer admittance matrix,
while the ports of the transfer admittance matrix are connected to the surface S1 and S2 such that:

(8.205)

The equivalent damping and stiffness matrices are given by:

(8.206)

(8.207)

where the indices r and I represent the real and imaginary part of a complex variable.

The internal source is cast by:

(8.208)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
284 Published: 2023-10-23
 Transfer Admittance Matrix

[Link] Admittance Matrix Connected to the Structural and Acoustic

Domain
Assume the transfer admittance matrix to represent porous material on a structural surface, i.e. a port
of the transfer admittance matrix is connected to the FSI interface.

The “weak” form in the structural domain is given by:

(8.209)

The “weak” form in the acoustic domain is given by:

(8.210)

where:

n F,1 is the outward normal unit vector of the porous material domain and points toward the
solid
n F,2 is the outward normal unit vector of the acoustic domain and points toward the porous
material
In the fluid domain

The transfer admittance matrix (Equation 8.202 (p. 283)) is also written as:

(8.211)

where:

n S,1 = the outward normal unit vector of the structural domain

ps = the pressure on the surface S1 of the structural domain
v s = the particle velocity on the surface S1 of the structural domain
pa = the pressure on the surface S2 of the acoustic domain
v a = the particle velocity on the surface S2 of the acoustic domain
The transformation of the transfer admittance matrix in Equation 8.211 (p. 285) is given by:

(8.212)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 285
Acoustics

With and , the transformed transfer admittance matrix is re-written as:

(8.213)

Considering the continuous boundary condition on the FSI interface:

(8.214)

The boundary integration in the structure “weak” form is given by:

(8.215)

Using Equation 8.8 (p. 243) and Equation 8.9 (p. 244), the boundary integration in the acoustic “weak”
form is given by:

Substituting the boundary integrations into Equation 8.209 (p. 285) and Equation 8.210 (p. 285), the
coupled matrix equation is derived as:

(8.216)

where:

It is impossible to evaluate (Nn} on the acoustic surface S2, since the normal vector n F,1 is involved.
An alternative is to evaluate {Nn} on the solid surface S1, for instance:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
286 Published: 2023-10-23
 Random Acoustics

8.8. Random Acoustics

Random acoustics deals with a vibro-acoustic system excited randomly, such as the sound field in a
reverberant room.

The following topics related to random acoustics are available:

8.8.1. Acoustic Diffuse Sound Field
8.8.2. Diffuse Sound Field Power Spectral Density
8.8.3. Diffuse Sound Field Physical Sampling

8.8.1. Acoustic Diffuse Sound Field

A diffuse sound field is assumed to be a weak stationary random process. It is generally defined in
terms of an infinite number of uncorrelated plane waves. The sound field is diffuse only if:

• The amplitudes of plane waves from different angles of incidence are statistically independent.

• The relative phases of the plane wave are uniformly distributed over .

• The amplitudes are modeled with equal intensity.

8.8.2. Diffuse Sound Field Power Spectral Density

Considering plane waves at the two particular points, the first point is located at the origin while
o
the second one is located at the (r, 90 , 0,) in the spherical coordinate system with the position r
= (r, , ). (Rafaely [422] (p. 949)and Van den Nieuwenhof et al. [423] (p. 949))

Figure 8.5: Plane Wave in Spherical Coordinates

th
The instantaneous pressure of the n plane wave at the origin is expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 287
Acoustics

(8.217)

The plane wave at the position (0, 0, z) can be retrieved by the pressure value at the origin:
(8.218)

The diffuse field pressure at the point is the superposition of an infinite number of plane waves from
all directions:

(8.219)

The cross-correlation function of the pressures at the point and is expressed as:
(8.220)

Assuming that plane waves are uncorrelated and all signals have the same auto-correlation
function , the cross-correlation function related to all plane waves directions with the element
area on the unit sphere, which is replacing the element area in the summation, is:

(8.221)

After a change of the variable:

(8.222)

The cross-correlation function is rewritten as:

(8.223)

Because the power spectral density is the Fourier transformation of the auto-correlation
function , the function is written as:

(8.224)

The cross-correlation function is given by:

(8.225)

Because the cross-power spectral density is the Fourier transformation of the cross-correlation
function , that is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
288 Published: 2023-10-23
 Room Acoustics

(8.226)

The cross-power spectral density can be expressed as:

(8.227)

where:

is the acoustic wave number.

8.8.3. Diffuse Sound Field Physical Sampling

The diffuse sound field is approached by the asymptotic model of plane waves. The plane waves with
random phases come from all directions in free space and superposition. The infinite number of plane
waves is replaced with a sufficiently high number of plane waves for simulation purposes.

The reference sphere with the radius and origin are defined. Generally, the radius should be at
least 50x the maximum dimension of the structural panel. The sphere surface is equally divided into
elementary surfaces so that the plane waves carry the energy equally in all incident directions. In
practice, the sphere surface is divided into parallel rings along the z axis of the global Cartesian
coordinate system (default) or a user-specified local Cartesian coordinate system, and the program
generates the elementary surfaces, each having nearly the same area.

Considering an elementary surface with an outward unit normal vector at the center of , the
acoustic pressure at the point generated by the plane wave with the amplitude
and phase propagating along direction, is defined by:
(8.228)

where:
(8.229)

The amplitude of each plane wave is defined by the reference power spectral density, , and ex-
pressed as:

(8.230)

In terms of the definition of diffuse sound field, the N phase angles are uniformly sampled in the
range . The sampled diffuse sound field is given by:

(8.231)

8.9. Room Acoustics

Room acoustics predicts the sound field distribution and the sound decay in rooms (Valeau et al.).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 289
Acoustics

The following topics related to room acoustics are available:

8.9.1. Diffuse Model for Room Acoustics
8.9.2. Mixed Boundary Conditions
8.9.3. Coupled Rooms with a Partition Wall

8.9.1. Diffuse Model for Room Acoustics

In room acoustics, the mean-free path λ of the room is evaluated by:
(8.232)

where:

V = volume of the room

S = surface area of the room

According to the theory of diffusion for particles in a scattering medium, the diffusion coefficient of
a room is defined as:
(8.233)

where:

c = speed of sound

The acoustic energy density flux is defined by the gradient of energy density as:
(8.234)

The acoustic diffusion model is governed by the following diffusion equation:

(8.235)

where:

Dt = the total diffusion coefficient in the room, which is defined by the diffusion coefficient

De in the empty room and the diffusion coefficient of the furniture Df in the room:

ma = coefficient of atmospheric attenuation

αf = absorption coefficient of furniture in the room
λf = mean-free path of furniture in the room, which is calculated by the number of furniture
items per unit volume nf and the furniture’s average scattering-section Qf:

= omnidirectional radiated sound power source

The weak form of the acoustic diffusion model is cast by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
290 Published: 2023-10-23
 The Full Linear Navier-Stokes (FLNS) Model

(8.236)

8.9.2. Mixed Boundary Conditions

The energy exchange on the boundaries in the acoustic diffusion model is described by the mixed
boundary condition:
(8.237)

where:

h = exchange coefficient, which is calculated as:

= outward normal unit vector of the surface

8.9.3. Coupled Rooms with a Partition Wall

The acoustic energy exchanges transferring from one room to another through a partition wall with
coupled area S12 in the room 1 and S21 in the room 2 are described by the following:

(8.238)

where:

= transmission coefficient, which is calculated by the transmission loss R(dB); that is =

-R/10
10

A two-port network admittance matrix is implemented to couple two rooms together (see Transfer
Admittance Matrix (p. 283) for more information on the transfer admittance matrix):

(8.239)

8.10. The Full Linear Navier-Stokes (FLNS) Model

The full linear Navier-Stokes (FLNS) equations are solved for viscous-thermal acoustics, which includes
the viscous and thermal effects for modeling small devices. For this case, the FLNS model achieves more
accurate simulation result compared to the BLI or LRF model. (Kampinga et al. [[436] (p. 950)])

The following topics related to the full linear Navier-Stokes equation solver are available:
[Link] FLNS Model for Viscous-Thermal Acoustics
[Link] Finite Element FLNS Model
8.10.3. Boundary Conditions of the FLNS Model

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 291
Acoustics

8.10.4. Coupling Conditions on the FSI Interface for the FLNS Model
8.10.5. Boundary Layers in the FLNS Model

8.10.1. The FLNS Model for Viscous-Thermal Acoustics

With the assumptions of

• linearization,

• no mean flow, and

• small harmonic perturbation,

the linearized Navier-Stokes equations are derived as follows:

(8.240)
(8.241)

(8.242)

where:

= stress tensor defined as:

= velocity variation
T = temperature variation
p = pressure variation
= dynamic viscosity
= second viscosity
= unit tensor
= volumetric force density
= ambient density
Cp = heat coefficient at constant pressure per unit mass
= heat flow vector defined as:

= heat conduction coefficient

Q = volumetric heat source
p0 = ambient pressure

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
292 Published: 2023-10-23
 The Full Linear Navier-Stokes (FLNS) Model

T0 = ambient temperature

8.10.2. The Finite Element FLNS Model

The finite element formulations are obtained by testing the moment equation (Equation 8.240 (p. 292)),
the continuity equation (Equation 8.241 (p. 292)), and the entropy equation (Equation 8.242 (p. 292))
using the Galerkin procedure. The equations are multiplied by testing function , pw and Tw, then
integrated over the volume of the domain with some manipulation to yield the following:
(8.243)

(8.244)

(8.245)

where:

= volume differential of acoustic domain Ω

= surface differential of acoustic domain boundary
= outward normal unit vector on the boundary

8.10.3. Boundary Conditions of the FLNS Model

Boundary conditions are applied to the finite element model to solve FLNS equations.

Acoustic boundary conditions may be as follows:

Normal velocity (8.246)

Normal stress ≈ (8.247)
pressure
Acoustic impedance (8.248)

On the boundary location, one acoustic boundary condition—either normal velocity, normal stress,
or impedance—should be prescribed.

Thermal boundary conditions may be as follows:

Temperature (8.249)
Heat flux (8.250)
Thermal (8.251)
impedance

On the boundary location, one thermal boundary condition—either temperature, heat flux, or thermal
impedance—should be prescribed.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 293
Acoustics

Viscous boundary conditions may be as follows:

Shear velocity (8.252)

Shear force (8.253)
Viscous (8.254)
impedance

On the boundary location, one viscous boundary condition—either shear velocity, shear force, or
viscous impedance—should be prescribed. The tangential total force is equal to the tangential viscous
force.

8.10.4. Coupling Conditions on the FSI Interface for the FLNS Model
On the FSI interface, the kinetic condition relating to the acting and reaction stress and the kinematic
condition are written as:
(8.255)
(8.256)

where:

= structure stress tensor

= fluid stress tensor
= outward normal unit vector of structure domain
= outward normal unit vector of fluid domain
= velocity of fluid
= displacement of fluid

8.10.5. Boundary Layers in the FLNS Model

There are two boundary layers in the FLNS model, the viscous boundary layer and the thermal
boundary layer. The thickness of the viscous boundary layer can be evaluated for two cases.

In a circular pipe, the viscous boundary layer is cast with:

On the top of a flat plate, the viscous boundary layer is given by:

The thermal boundary layer is related to the viscous boundary layer with the Prandtl number:

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
294 Published: 2023-10-23
 Poroelastic Acoustics

Pr = Prandtl number

8.11. Poroelastic Acoustics

The mixed displacement-pressure finite element model, based on Biot's theory, solves the poroelastic
problem in acoustics (Allard [440] (p. 950)). Sound absorption occurs in the poroelastic materials.

8.11.1. Mixed Displacement and Pressure Formulation for Poroelastic

Acoustic Material
In Biot's poroelasticity theory, the frequency-dependent effective densities are defined as:

(8.257)

where:

= porosity
= fluid mass density (air)
= tortuosity
= the bulk density of the solid phase of the poroelastic material
= solid mass density

= fluid resistivity

= fluid dynamic (shear) viscosity

= viscous characteristic length
ω = angular frequency

In the poroelastic acoustic material, the mixed displacement and pressure (u-P) governing equations
are as follows:
(8.258)

(8.259)

where:

= solid stress tensor (eliminating the dependency of fluid displacement) defined as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 295
Acoustics

= displacement of the solid (skeleton)

= strain tensor of the solid
N = complex shear modulus in vacuo (second Lame coefficient) of the solid
= effective density defined as:

p = pressure in the pores

Ks = bulk modulus of the elastic solid (frame)

Kb = bulk modulus of the porous material in vacuo

Kf = bulk modulus of the air defined as:

P0 = static reference pressure

= specific heat ratio of air
Prt = Prandtl number
= thermal characteristic length

For the majority of poroelastic materials used in acoustics, . The Biot's constant is defined by:

(8.260)

In most cases, the Biot's constant equals 1.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
296 Published: 2023-10-23
 Poroelastic Acoustics

8.11.2. The Finite Element Model

Introducing variable transformation,
(8.261)

the finite element formulations are obtained by testing Equation 8.258 (p. 295) and Equa-
tion 8.259 (p. 295) using the Galerkin procedure. Equation 8.258 (p. 295) and Equation 8.259 (p. 295) are
multiplied by testing function and q, respectively, and integrated over the volume of the domain
with some manipulation to yield the following:

(8.262)

(8.263)

where:

= total stress tensor, defined by:

= fluid phase displacement

The fluid phase displacement is cast by:

(8.264)

In poroelastic material, the total displacement is written as:

(8.265)

8.11.3. Boundary Conditions of Poroelastic Acoustics

Boundary conditions are applied to the finite element model to solve poroelastic acoustic equations
according to the application.

Natural Boundary Condition

(8.266)

where subscript n = normal direction of the interface.

In a finite element model, the boundary integral terms are not considered with the natural boundary
condition.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 297
Acoustics

Fixed Edge
(8.267)

A poroelastic medium is bounded onto a rigid impervious wall. The solid phase displacement and
normal relative displacement vanish at the rigid wall.

Free Porous Surface

(8.268)

On the free porous surface, the pressure is set to zero.

Sliding Edge
The sliding edge condition indicates that the normal displacements on an impervious rigid wall are
constrained, and the tangential displacements are free.

(8.269)

The boundary integrations vanish because the normal solid displacement is constrained.

Pervious Porous Surface with Permeability K

The boundary condition imposed with the difference in displacements between solid phase, fluid
phase, and pressure on the pervious porous surface is written as:
(8.270)

Imposed Pressure

(8.271)

For the case of imposed pressure, the pressure is set to p0.

Imposed Displacement Field

(8.272)

The continuity between the imposed displacement and the solid phase displacement is maintained.
The normal displacements between solid phase and fluid phase continues.

8.11.4. Coupling Conditions of Poroelastic Acoustics

The couplings occur at the interfaces between the poroelastic material and other materials.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
298 Published: 2023-10-23
 Poroelastic Acoustics

Poroelastic-Elastic Coupling Conditions

The coupling conditions at the interface between poroelastic and elastic media are given by:

• Displacement continuity

• Impervious boundary

• Stress continuity

where:

= elastic stress tensor

The coupling between poroelastic and elastic media is natural.

Poroelastic-Acoustic Coupling Conditions

The coupling conditions at the interface between poroelastic and acoustic media are given by:

• Normal stress continuity

• Total displacement continuity

• Pressure continuity

where:

pa = pressure in acoustic media

ρ0 = mass density of acoustic media

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 299
Acoustics

8.12. Nonlinear Acoustics

Nonlinear effects are observed when an ultrasound wave is generated by a high power source in a
thermoviscous fluid. To simulate nonlinear sound wave propagation, either the pressure-based Westervelt
equation (p. 300) or velocity potential-based Kuznetsov equation (p. 300) is solved numerically.

8.12.1. Westervelt Equation

The pressure-based Westervelt equation is given by (Dirkse [444] (p. 950)):

(8.273)

where:

= pressure
= diffusivity of sound, which is related to the dynamic viscosity , the bulk viscosity , the
thermal conductivity , the specific heat capacity at constant volume , and the specific heat
capacity at constant pressure as follows:

= dimensionless nonlinearity coefficient, which is related to the nonlinearity parameter as

follows:
= mass density of fluid
= speed of sound
= mass source rate

8.12.2. Kuznetsov Equation

The velocity potential-based Kuznetsov equation is given by (Hoffelner and Kaltenbacher [445] (p. 950)):

(8.274)

where:

= velocity potential
= mass source

The acoustic pressure and particle velocity are related to the scalar velocity potential by

and

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
300 Published: 2023-10-23
 Nonlinear Acoustics

8.12.3. The Finite Element Model

The finite element formulations are obtained by testing Equation 8.273 (p. 300) and Equa-
tion 8.274 (p. 300) using the Galerkin procedure.

Multiplying the Westervelt equation (Equation 8.273 (p. 300)) by a testing function and integrating
over the volume of the domain with some manipulation yields:

(8.275)

The nonlinear matrix equation is given by:

(8.276)

where:

Since is on the order of 10-10, the second surface integral in Equation 8.275 (p. 301) is neglected if
the surface is not rigid.

Multiplying the Kuznetsov equation Equation 8.274 (p. 300) by a testing function and integrating
over the volume of the domain with some manipulation yield the following:

(8.277)

The nonlinear matrix equation is given by:

(8.278)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 301
Acoustics

where

The nonlinear acoustics equations are iteratively solved using the Newmark method (see Transient
Analysis (p. 796)). The right hand side of Equation 8.276 (p. 301) or Equation 8.278 (p. 301) is iteratively
updated with the solved degree of freedom and its first and second order time derivatives until
convergence is achieved.

8.12.4. Fluid-Structure Interaction in Nonlinear Acoustics

The coupling conditions between the fluid and structure domains are given by Equation 8.146 (p. 273)
and Equation 8.147 (p. 273).

For the Westervelt equation, the coupled matrix equations are:

(8.279)

For the definitions of parameters and matrices, see Equation 8.148 (p. 273) - Equation 8.153 (p. 274)
and Equation 8.276 (p. 301).

For the Kuznetsov equation, the coupled matrix equations are:

(8.280)

For the definitions of parameters and matrices, see Equation 8.278 (p. 301).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
302 Published: 2023-10-23
 Nonlinear Acoustics

In equations Equation 8.279 (p. 302) and Equation 8.280 (p. 302):

is the coupling matrix is given by:

The surface integral with the factor is neglected on the FSI interface.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 303
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
304 Published: 2023-10-23
Chapter 9: Diffusion
The following topics related to diffusion are available:
9.1. Diffusion Fundamentals
9.2. Normalized Concentration Approach
9.3. Derivation of Diffusion Matrices
9.4. Diffusion Analysis Results

9.1. Diffusion Fundamentals

The diffusion process is governed by the first Fick's law:
(9.1)

It is also governed by the mass balance equation:

(9.2)

where:

{J} = diffusion flux vector

Dxx, Dyy, and Dzz = the diffusivity coefficients in the element's X, Y, and Z directions, respectively
(input as DXX, DYY, DZZ on MP command)
C = C(x,y,z,t) = concentration (CONC degree of freedom)
{v} = transport velocity vector (input as VELO on BF command for PLANE238, SOLID239, and
SOLID240 only)
G = diffusing substance generation rate per unit volume (input as DGEN on BF or BFE com-
mands)

= divergence operator
t = time

Substituting Equation 9.1 (p. 305) into Equation 9.2 (p. 305) produces the second Fick’s law augmented
with the transport term:
(9.3)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 305
Diffusion

The applicable boundary conditions and loads are:

1. Specified concentration (input as CONC on D, DA, DK, DL commands). Initial concentration can also
be applied using the CONC label with the IC command.

2. Specified diffusion flux acting on a surface (input as DFLUX on SF or SFE commands):

(9.4)

where:

{n} = unit outward normal vector

{r} = diffusion flow vector

3. Diffusing substance generation rate G (input as DGEN on BF or BFE commands).

4. Transport velocity vector {v} (input as VELO on BF command for PLANE238, SOLID239, and SOLID240
only).

5. Diffusion flow rate (input as RATE on F and FK commands).

9.2. Normalized Concentration Approach

Equation 9.3 (p. 305), along with the boundary conditions and loads, governs the diffusion process in a
homogeneous domain. In an inhomogeneous domain, different materials have different saturated
concentrations. This difference in the saturation levels is responsible for the discontinuity in the concen-
tration across the material interface. To be able to use Equation 9.3 (p. 305) in a finite element analysis,
a continuous variable, normalized concentration , is introduced [401] (p. 947).
(9.5)

where is the saturated concentration of the material (input as CSAT on MP command).

Substituting the concentration from Equation 9.5 (p. 306) into Equation 9.3 (p. 305) produces a
governing equation for the diffusion analysis in terms of normalized concentration :

(9.6)

When is not specified, it defaults to 1.0 and Equation 9.6 (p. 306) becomes Equation 9.3 (p. 305).

can be temperature-dependent for the coupled-field elements PLANE222, PLANE223, SOLID225,

SOLID226, and SOLID227 only with thermal and diffusion DOFs. In this case, applying the chain rule of
differentiation with respect to temperature T to Equation 9.6 (p. 306) without the transport term produces:

(9.7)

For more information, see Thermal-Diffusion Coupling (p. 357).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
306 Published: 2023-10-23
 Diffusion Analysis Results

9.3. Derivation of Diffusion Matrices

The concentration is approximated over the finite element as follows:
(9.8)

where:

{N} = element shape functions

{Ce} = nodal concentration vector (input/output as CONC)

The application of the variational principle and finite element approximation Equation 9.8 (p. 307) to
Equation 9.6 (p. 306) produces the matrix equation as:
(9.9)

where:

= element diffusion damping matrix

= element diffusion conductivity matrix

= element transport conductivity matrix

= element diffusion flux load vector

= element diffusing substance generation load vector

vol = element volume

S = element surface
= nodal diffusion flow rate vector applied to the element (input/output as RATE on F
command)

The finite element Equation 9.9 (p. 307) is unsymmetric. To keep the equation symmetric, the transport
effects can be applied as a load vector by setting KEYOPT(2) = 1 for the diffusion
elements (PLANE238, SOLID239, SOLID240). Using this option activates a nonlinear solution, and at least
two iterations are required to achieve a response to the transport effects.

9.4. Diffusion Analysis Results

After the matrix Equation 9.9 (p. 307) is solved for the nodal concentration vector , the diffusion
gradient {g}, and the diffusion flux {J} are evaluated at the integration points as follows:
(9.10)
(9.11)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 307
Diffusion

Nodal values of concentration gradient {g} (output as CG) and diffusion flux {J} (output as DF) are
computed from the integration point values, as described in Nodal and Centroidal Data Evaluation (p. 442).
With the normalized concentration approach, the element true concentration is calculated at
the element centroid and output as SMISC,1.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
308 Published: 2023-10-23
Chapter 10: Coupling
Coupled-field analyses are useful for solving problems where the coupled interaction of phenomena
from various disciplines of physical science is significant. Several examples of this include: an electric
field interacting with a magnetic field, a magnetic field producing structural forces, a temperature field
influencing fluid flow, a temperature field giving rise to thermal strains and the usual influence of
temperature dependent material properties. The latter two examples can be modeled with most non-
coupled-field elements, as well as with coupled-field elements.

The following coupled-field topics are available:

10.1. Coupled Effects
[Link]
[Link]
[Link]
10.5. Piezoelectrics
10.6. Electroelasticity
10.7. Piezoresistivity
[Link]
10.9. Review of Coupled Electromechanical Methods
10.10. Porous Media Flow
10.11. Structural-Diffusion Coupling
[Link]-Diffusion Coupling
10.13. Electric-Diffusion Coupling
10.14. Magnetoelasticity

10.1. Coupled Effects

The following topics concerning coupled effects are available:
10.1.1. Elements
10.1.2. Coupling Methods

10.1.1. Elements
The following elements have coupled-field capability:

Table 10.1: Elements Used for Coupled Effects

SOLID5 3D Coupled-Field Solid (Derivation of Electromagnetic Matrices (p. 176), Coupled

Effects (p. 309), SOLID5 - 3D Coupled-Field Solid (p. 445))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 309
Coupling

PLANE13 2D Coupled-Field Solid (Derivation of Electromagnetic Matrices (p. 176), Coupled

Effects (p. 309), SOLID5 - 3D Coupled-Field Solid (p. 445))
FLUID29 2D Acoustic Fluid (Derivation of Acoustic Matrices (p. 248), FLUID29 - 2D Acoustic
Fluid (p. 458))
FLUID30 3D 8-Node Acoustic Fluid (Derivation of Acoustic Matrices (p. 248), FLUID30 - 3D
Acoustic Fluid (p. 459))
LINK68 Coupled Thermal-Electric Line (LINK68 - Coupled Thermal-Electric Line (p. 492))
SOLID98 Tetrahedral Coupled-Field Solid (Derivation of Electromagnetic Matrices (p. 176),
Coupled Effects (p. 309), SOLID98 - Tetrahedral Coupled-Field Solid (p. 506))
FLUID116 Coupled Thermal-Fluid Pipe (FLUID116 - Coupled Thermal-Fluid Pipe (p. 513))
CIRCU124 Electric Circuit Element (CIRCU124 - Electric Circuit (p. 521))
TRANS126 Electromechanical Transducer (Capacitance Computation (p. 202), Review of Coupled
Electromechanical Methods (p. 351), TRANS126 - Electromechanical Transducer (p. 526))
SHELL157 Coupled Thermal-Electric Shell (SHELL157 - Thermal-Electric Shell (p. 549))
FLUID220 3D 20-Node Acoustic Fluid
FLUID221 3D 10-Node Acoustic Fluid
PLANE222 2D 4-Node Coupled-Field Solid (PLANE222 - 2D 4-Node Coupled-Field Solid (p. 621))
PLANE223 2D 8-Node Coupled-Field Solid (PLANE223 - 2D 8-Node Coupled-Field Solid (p. 623))
SOLID225 3D 8-Node Coupled-Field Solid (SOLID225 - 3D 8-Node Coupled-Field Solid (p. 625))
SOLID226 3D 20-Node Coupled-Field Solid (SOLID226 - 3D 20-Node Coupled-Field Solid (p. 627))
SOLID227 3D 10-Node Coupled-Field Solid (SOLID227 - 3D 10-Node Coupled-Field Solid (p. 629))
LINK228 3D Coupled-Field Link (LINK228 - 3D Coupled-field Link (p. 631))
PLANE233 2D 8-Node Electromagnetic Solid (Derivation of Electromagnetic Matrices (p. 176),
Electromagnetic Field Evaluations (p. 183), PLANE233 - 2D 8-Node Electromagnetic
Solid (p. 634))
SOLID236 3D 20-Node Electromagnetic Solid (Derivation of Electromagnetic Matrices (p. 176),
Electromagnetic Field Evaluations (p. 183), SOLID236 - 3D 20-Node Electromagnetic
Solid (p. 635))
SOLID237 3D 10-Node Electromagnetic Solid (Derivation of Electromagnetic Matrices (p. 176),
Electromagnetic Field Evaluations (p. 183), SOLID237 - 3D 10-Node Electromagnetic
Solid (p. 635))

There are certain advantages and disadvantages inherent with coupled-field formulations:

[Link]. Advantages
1. Allows for solutions to problems otherwise not possible with usual finite elements.

2. Simplifies modeling of coupled-field problems by permitting one element type to be used in a

single analysis pass.

[Link]. Disadvantages
1. Increases problem size (unless a segregated solver is used).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
310 Published: 2023-10-23
 Coupled Effects

2. Inefficient matrix reformulation (if a section of a matrix associated with one phenomena is re-
formed, the entire matrix will be reformed).

3. Larger storage requirements.

10.1.2. Coupling Methods

There are basically two methods of coupling distinguished by the finite element formulation techniques
used to develop the matrix equations. These are illustrated here with two types of degrees of freedom
({X1}, {X2}):

1. Strong (also matrix, simultaneous, or full) coupling - where the matrix equation is of the form:

(10.1)

and the coupled effect is accounted for by the presence of the off-diagonal submatrices [K12] and
[K21]. This method provides for a coupled response in the solution after one iteration.

2. Weak (also load vector or sequential) coupling - where the coupling in the matrix equation is
shown in the most general form:

(10.2)

and the coupled effect is accounted for in the dependency of [K11] and {F1} on {X2} as well as [K22]
and {F2} on {X1}. At least two iterations are required to achieve a coupled response.

The following is a list of the types of coupled-field analyses including methods of coupling present
in each:

Table 10.2: Coupling Methods

Coupling
Analysis Category Method Example Applications
Used
Thermal-Structural Analysis (p. 313) S, W High temperature turbine
Electromagnetic Analysis (p. 314) S Current fed massive conductors
Stranded Coil Analysis (p. 314) S Stranded coils
Structural-Magnetic Analysis (p. 315) S, W Solenoids, high energy magnets
Structural-Electromagnetic Analysis (p. 316) S (MRI), magnetic actuators
Structural-Stranded Coil Analysis (p. 316) S
Electro-Thermo-Structural Analysis (p. 317) W Electro-thermal MEMS actuators
Electro-Magneto-Thermo-Structural W Direct current electromechanical
Analysis (p. 317) devices in general
Thermal-Electromagnetic Analysis (p. 318)
Piezoelectric Analysis (p. 319) S Transducers, resonators

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 311
Coupling

Coupling
Analysis Category Method Example Applications
Used
Electrostatic-Structural Analysis (p. 320) S, W Dielectric elastomers, air regions in
microelectromechanical systems
(MEMS)
Thermo-Piezoelectric Analysis (p. 320) S, W Sensors and actuators for smart
structures
Piezoresistive Analysis (p. 321) W Pressure and force sensors
Thermo-Pressure Analysis (p. 321) S, W Piping networks
Acoustic-Structural Analysis (p. 322) S, W Acoustics
Thermo-Electric Analysis (p. 322) S, W High temperature electronics, Peltier
coolers, thermoelectric generators
Thermal-Magnetic Analysis (p. 323) W Direct current transients: power
interrupts, surge protection
Circuit-Magnetic Analysis (p. 323) S Circuit-fed solenoids, transformers,
and motors
Structural-Diffusion Analysis (p. 323) S, W Hygroscopic swelling of polymers
in electronics packaging, oxygen or
hydrogen migration in metals
Thermal-Diffusion Analysis (p. 324) S, W Temperature-dependent moisture
migration, thermomigration in
metallic interconnects
Structural-Thermal-Diffusion Analysis (p. 324) S, W Sodium expansion in aluminum
reduction cells
Electric-Diffusion S, W Electromigration in PCB
interconnects
Thermal-Electric-Diffusion S, W Thermomigration and
electromigration in PCB
interconnects
Structural-Electric-Diffusion S, W Hydrostatic stress-migration and
electromigration in PCB
interconnects
Structural-Thermal-Electric-Diffusion S, W Hydrostatic stress-migration,
thermomigration, and
electromigration in PCB
interconnects

where:

S = strong coupling
W = weak coupling

The solution sequence follows the standard finite element methodology. Convergence is achieved
when changes in all unknowns (i.e. DOF) and knowns, regardless of units, are less than the values
specified (on the CNVTOL command). Some of the coupling described above is always or usually

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
312 Published: 2023-10-23
 Coupled Effects

one-way. For example, in Category A, the temperatures affect the displacements of the structure by
way of the thermal strains, but the displacements usually do not affect the temperatures.

The following descriptions of coupled phenomena will include:

1. Applicable element types

2. Basic matrix equation indicating coupling terms in bold print. In addition to the terms indicated
in bold print, any equation with temperature as a degree of freedom can have temperature-de-
pendency in all terms.

3. Applicable analysis types, including the matrix and/or vector terms possible in each analysis type.

The nomenclature used on the following pages is given in Table 10.3: Nomenclature of Coefficient
Matrices (p. 325) at the end of the section. In some cases, element KEYOPTS are used to select the
DOF of the element. DOF will not be fully active unless the appropriate material properties are specified.
Some of the elements listed may not be applicable for a particular use as it may be only 1D, whereas
a 3D element is needed (e.g. FLUID116).

[Link]. Thermal-Structural Analysis

(see Derivation of Structural Matrices (p. 12), Derivation of Heat Flow Matrices (p. 217), and Ther-
moelasticity (p. 329))

1. Element type: SOLID5, PLANE13, SOLID98, PLANE222, PLANE223, SOLID225, SOLID226, SOLID227,
and LINK228

2. Matrix equation

a. Strong coupling:

(10.3)

b. Weak coupling:

(10.4)

where:

[Kt] = [Ktb] + [Ktc]

{F} = {Fnd} + {Fpr} + {Fac}
{Q} = {Qnd} + {Qg} + {Qc}

3. Analysis types:

a. Strong coupling: static, transient, or harmonic

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 313
Coupling

b. Weak coupling: static or transient

Note:

Strong coupling is supported only by PLANE222, PLANE223, SOLID225, SOLID226, and

SOLID227.

{Qted} is applicable only to PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227.

[Link]. Electromagnetic Analysis

(see Derivation of Electromagnetic Matrices (p. 176))

1. Element type: PLANE13, PLANE233, SOLID236, and SOLID237

2. Matrix equation

a. Time-integrated electric potential formulation:

(10.5)

where:

{I} = {Ind}

The above formulation is used with PLANE13. It is also available with KEYOPT(2) = 2 of
PLANE233, SOLID236, and SOLID237.

b. Electric potential formulation:

(10.6)

The above formulation is the default for PLANE233, SOLID236, and SOLID237.

3. Analysis types: Static, harmonic or transient

[Link]. Stranded Coil Analysis

(see Stranded Coil Analysis)

1. Element type: PLANE233, SOLID236, and SOLID237

2. Matrix equation:

A-VOLT-EMF formulation

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
314 Published: 2023-10-23
 Coupled Effects

(10.7)

where:

{I} = {Ind}

The above formulation is used with PLANE233, SOLID236, and SOLID237.

3. Analysis types: Static, harmonic or transient

[Link]. Structural-Magnetic Analysis

(see Derivation of Electromagnetic Matrices (p. 176) and Magnetoelasticity (p. 361))

1. Element type: SOLID5, PLANE13, SOLID98, PLANE223, SOLID226, and SOLID227

2. Matrix equations:

a. Weak coupling with the scalar potential formulation using SOLID5 and SOLID98:

(10.8)

b. Weak coupling with the magnetic vector potential formulation using PLANE13:

(10.9)

c. Strong coupling with the magnetic vector potential formulation using PLANE223 and
the edge-flux formulation using SOLID226 and SOLID227.

i. Maxwell force method (KEYOPT(8) = 0):

(10.10)

ii. Lorentz force method (KEYOPT(8) = 1):

(10.11)

where:

{F} = {Fnd} + {Fpr} + {Fac} +{Fth}

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 315
Coupling

3. Analysis types: static or transient with PLANE13, PLANE223, SOLID226, and SOLID227 elements;
static analysis only with SOLID5 and SOLID98

[Link]. Structural-Electromagnetic Analysis

(see Derivation of Electromagnetic Matrices (p. 176) and Magnetoelasticity (p. 361))

1. Element type: PLANE223, SOLID226, and SOLID227

2. Matrix equation:

a. Strong coupling with the Maxwell force option (KEYOPT(8) = 0):

(10.12)

b. Strong coupling with the Lorentz force option (KEYOPT(8) = 1):

(10.13)

where:

{F} = {Fnd} + {Fpr} + {Fac} +{Fth}

3. Analysis types: static or transient

[Link]. Structural-Stranded Coil Analysis

(see Derivation of Electromagnetic Matrices (p. 176) and Magnetoelasticity (p. 361))

1. Element type: PLANE223, SOLID226, and SOLID227

2. Matrix equation:

a. Strong coupling with the Maxwell force option (KEYOPT(8) = 0):

(10.14)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
316 Published: 2023-10-23
 Coupled Effects

b. Strong coupling with the Lorentz force option (KEYOPT(8) = 1):

(10.15)

where:

{F} = {Fnd} + {Fpr} + {Fac} +{Fth}

3. Analysis types: static or transient

[Link]. Electro-Thermo-Structural Analysis

(see Derivation of Structural Matrices (p. 12), Derivation of Heat Flow Matrices (p. 217), Thermoelasti-
city (p. 329), and Thermoelectrics (p. 348))

1. Element type: PLANE222, PLANE223, SOLID225, SOLID226, SOLID227, and LINK228

2. Matrix equation:

(10.16)

where:

[Kt] = [Ktb] + [Ktc]

{F} = {Fnd} + {Fpr} + {Fac}
{Q} = {Qnd} + {Qg} + {Qc} + {Qj} + {Qp}
{I} = {Ind}

3. Analysis types: static and transient

[Link]. Electro-Magneto-Thermo-Structural Analysis

(see Derivation of Electromagnetic Matrices (p. 176) and Piezoelectrics (p. 334))

1. Element types: SOLID5 and SOLID98

2. Matrix equation:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 317
Coupling

(10.17)

where:

[Kt] = [Ktb] + [Ktc]

{F} = {Fnd} + {Fth} + {Fac} + {Fjb} + {Fpr} + {Fmx}
{Q} = {Qnd} + {Qg} + {Qj} + {Qc}
{I} = {Ind}

3. Analysis types: Static or Transient

[Link]. Thermal-Electromagnetic Analysis

(see Derivation of Electromagnetic Matrices (p. 176) and Derivation of Heat Flow Matrices (p. 217))

1. Element types: SOLID5, SOLID98, PLANE223, SOLID226, and SOLID227

2. Matrix equation

a. Magnetic scalar potential formulation for SOLID5, SOLID98:

(10.18)

b. Magnetic vector potential formulation for PLANE223:

(10.19)

Note:

Electric and magnetic vector potentials are strongly coupled.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
318 Published: 2023-10-23
 Coupled Effects

In this formulation, the VOLT degree of freedom is the actual electric potential
(not the time integrated electric potential as given in Equation 5.92 (p. 177)).
Therefore, matrices and appear in the stiffness matrix in this
unsymmetric strong coupling form.

c. Edge-flux formulation for SOLID226 and SOLID227:

(10.20)

where:

[Kt] = [Ktb] + [Ktc]

{Q} = {Qnd} + {Qg} + {Qj} + {Qc}
{I} = {Ind}

3. Analysis types: Static or Transient

[Link]. Piezoelectric Analysis

(see Piezoelectrics (p. 334))

1. Element types: SOLID5, PLANE13, SOLID98, PLANE222, PLANE223, SOLID225, SOLID226, and
SOLID227.

2. Matrix equation:

(10.21)

where:

{F} = {Fnd} + {Fth} + {Fac} + {Fpr}

{L} = {Lnd} + {Lc} + {Lsc}+{Lth}

Note:

{Lc} and {Lsc} are applicable to only PLANE222, PLANE223, SOLID225, SOLID226,
and SOLID227.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 319
Coupling

3. Analysis types: Static, modal, harmonic, or transient

[Link]. Electrostatic-Structural Analysis

(see Electroelasticity (p. 345))

1. Element types: PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227.

2. Matrix equation

a. Strong coupling:

(10.22)

b. Weak coupling:

(10.23)

where:

{F} = {Fnd} + {Fth} + {Fac} + {Fpr} + {Fe}

{L} = {Lnd} + {Lc} + {Lsc}

3. Analysis types:

a. Strong Coupling: nonlinear static or transient; linear perturbation static, modal, or har-
monic

b. Weak coupling: static or transient

[Link]. Thermo-Piezoelectric Analysis

(see Derivation of Structural Matrices (p. 12), Derivation of Heat Flow Matrices (p. 217), Thermoelasti-
city (p. 329), and Piezoelectrics (p. 334))

1. Element type: PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227

2. Matrix equation

a. Strong coupling:

(10.24)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
320 Published: 2023-10-23
 Coupled Effects

b. Weak coupling:

(10.25)

where:

[Kt] = [Ktb] + [Ktc]

{F} = {Fnd} + {Fpr} + {Fac}
{Q} = {Qnd} + {Qg} + {Qc}
{L} = {Lnd} + {Lc} + {Lsc}

3. Analysis types:

a. Strong coupling: static, transient, harmonic, modal

b. Weak coupling: static or transient

[Link]. Piezoresistive Analysis

(see Piezoresistivity (p. 347))

1. Element type: PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227

2. Matrix equation:

(10.26)

where:

[Kv] = conductivity matrix (see Equation 10.107 (p. 348)) updated for piezoresistive effects
{F} = {Fnd} + {Fth} + {Fpr} + {Fac)
{I} = {Ind}

3. Analysis types: Static or transient

[Link]. Thermo-Pressure Analysis

(see FLUID116 - Coupled Thermal-Fluid Pipe (p. 513))

1. Element type: FLUID116

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 321
Coupling

2. Matrix equation:

(10.27)

where:

[Kt] = [Ktb] + [Ktc] + [Ktm]

{Q} = {Qnd} + {Qc} + {Qg}
{W} = {Wnd} + {Wh}

3. Analysis types: Static or Transient

[Link]. Acoustic-Structural Analysis

(See Derivation of Acoustic Matrices (p. 248).)

1. Element type: FLUID29, FLUID30, FLUID220, and FLUID221 (with other structural elements)

2. Matrix equation:

(10.28)

Values for [M], [C], and [K] are provided by other elements.

3. Analysis types: Transient, harmonic and modal analyses can be performed.

[Link]. Thermo-Electric Analysis

1. Element types: SOLID5, LINK68, SOLID98, SHELL157, PLANE222, PLANE223, SOLID225, SOLID226,
SOLID227, and LINK228

2. Matrix equation:

(10.29)

where:

[Kt] = [Ktb] + [Ktc]

{Q} = {Qnd} + {Qc} + {Qg} + {Qj} + {Qp}
{I} = {Ind}

Note:

{Qp}, [Kvt], and [Cv] are used only for PLANE222, PLANE223, SOLID225, SOLID226, and
SOLID227.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
322 Published: 2023-10-23
 Coupled Effects

3. Analysis types: Static or Transient

[Link]. Thermal-Magnetic Analysis

(see Derivation of Electromagnetic Matrices (p. 176) and Derivation of Heat Flow Matrices (p. 217))

1. Element types: PLANE13, PLANE223, SOLID226, and SOLID227

2. Matrix equation:

(10.30)

where:

[Kt] = [Ktb] + [Ktc]

{Q} = {Qnd} + {Qg} + {Qj} + {Qc}

3. Analysis types: Static or Transient

[Link]. Circuit-Magnetic Analysis

(See Stranded Coil Analyses (p. 193))

1. Element type: CIRCU124

2. Matrix equation:

(10.31)

3. Analysis types: Static, Transient, or Harmonic

[Link]. Structural-Diffusion Analysis

(see Derivation of Structural Matrices (p. 12), Derivation of Diffusion Matrices (p. 307), and Structural-
Diffusion Coupling (p. 354))

1. Element type: PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227

2. Matrix equation

a. Strong coupling:

(10.32)

b. Weak coupling:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 323
Coupling

(10.33)

where:

{F} = {Fnd}+ {Fpr}+ {Fac}

3. Analysis types: static and transient

[Link]. Thermal-Diffusion Analysis

(see Derivation of Heat Flow Matrices (p. 217), Derivation of Diffusion Matrices (p. 307))

1. Element type: PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227

2. Matrix equation

(10.34)

where:

[Kt] = [Ktb] + [Ktc]

{Q} = {Qnd}+ {Qc}+ {Qg}+ {Qp}

3. Analysis types: static and transient

[Link]. Structural-Thermal-Diffusion Analysis

(see Derivation of Structural Matrices (p. 12), Derivation of Heat Flow Matrices (p. 217), Derivation
of Diffusion Matrices (p. 307), Thermoelasticity (p. 329), and Structural-Diffusion Coupling (p. 354))

1. Element type: PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227

2. Matrix equation

a. Strong coupling:

(10.35)

b. Weak coupling:

(10.36)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
324 Published: 2023-10-23
 Coupled Effects

{F} = {Fnd}+ {Fpr}+ {Fac}

{Q} = {Qnd}+ {Qc}+ {Qg}+ {Qp}

3. Analysis type: static and transient

Table 10.3: Nomenclature of Coefficient Matrices

Symbol Meaning Usage

[M] structural mass matrix (discussed in Derivation of Structural Matrices (p. 12)) [1]
fs
[M ] fluid-structure coupling mass matrix (discussed in Derivation of Acoustic [1]
Matrices (p. 248))
[Mp] acoustic mass matrix (discussed in Derivation of Acoustic Matrices (p. 248)) [1]
[C] structural damping matrix (discussed in Derivation of Structural [2]
Matrices (p. 12))
[Ct] thermal specific heat matrix (discussed in Derivation of Heat Flow [2]
Matrices (p. 217))
[Ctu] thermoelastic damping matrix (discussed in Thermoelasticity (p. 329)) [2]
AA
[C ] magnetic damping matrix (discussed in Electromagnetic Field [2]
Evaluations (p. 183))
[Cp] acoustic damping matrix (discussed in Derivation of Acoustic [2]
Matrices (p. 248))
[CAv] magnetic-electric damping matrix (discussed in Derivation of [2]
Electromagnetic Matrices (p. 176))
[Cvv] electric damping matrix (discussed in Derivation of Electromagnetic [2]
Matrices (p. 176))
[CiA] inductive damping matrix (discussed in Stranded Coil Analyses (p. 193)) [2]
v
[C ] dielectric permittivity coefficient matrix (discussed in Quasistatic Electric [2]
Analysis (p. 182))
[Cvh] dielectric damping matrix (discussed in Quasistatic Electric Analysis (p. 182)) [2]
[K] structural stiffness matrix (discussed in Derivation of Structural [3]
Matrices (p. 12))
[Kt] thermal conductivity matrix (may consist of 1, 2, or 3 of the following 3 [3]
matrices) (discussed in Derivation of Heat Flow Matrices (p. 217))
[Ktb] thermal conductivity matrix of material (discussed in Derivation of Heat [3]
Flow Matrices (p. 217))
[Ktc] thermal conductivity matrix of convection surface (discussed in Derivation [3]
of Heat Flow Matrices (p. 217))
[Ktm] thermal conductivity matrix associated with mass transport (discussed in [3]
Derivation of Heat Flow Matrices (p. 217))
[Kut] thermoelastic stiffness matrix (discussed in Thermoelasticity (p. 329)) [3]
m
[K ] scalar magnetic potential coefficient matrix (discussed in Derivation of [3]
Electromagnetic Matrices (p. 176))
[KAA] magnetic reluctivity matrix (discussed in Derivation of Electromagnetic [3]
Matrices (p. 176))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 325
Coupling

Symbol Meaning Usage

VV
[K ] electric conductivity matrix for electromagnetic analysis (discussed in [3]
Derivation of Electromagnetic Matrices (p. 176))
[KAV] magneto-electric coupling matrix (discussed in Derivation of [3]
Electromagnetic Matrices (p. 176))
[KVA] electromagnetic coupling matrix (discussed in Derivation of Electromagnetic [3]
Matrices (p. 176))
[KAi] potential-current coupling stiffness matrix (discussed in Stranded Coil [3]
Analyses (p. 193))
[Kii] resistive stiffness matrix (discussed in Stranded Coil Analyses (p. 193)) [3]
ie
[K ] current-emf coupling stiffness (discussed in Stranded Coil Analyses (p. 193)) [3]
v
[K ] electrical conductivity coefficient matrix for electric analysis (discussed in [3]
Derivation of Electromagnetic Matrices (p. 176))
[Kz] piezoelectric stiffness matrix (discussed in Piezoelectrics (p. 334)) [3]
eu
[K ] electric force stiffness or electrostatic softening matrix (discussed in [3]
Electroelasticity (p. 345))
[KeV] electric force coupling matrix (discussed in Electroelasticity (p. 345)) [3]
zt
[K ] thermo-piezoelectric stiffness matrix (discussed in Piezoelectrics (p. 334)) [3]
d
[K ] dielectric coefficient matrix (discussed in Piezoelectrics (p. 334)) [3]
vt
[K ] Seebeck coefficient coupling matrix [3]
[Kud] Diffision-elastic stiffness matrix (discussed in Structural-Diffusion [3]
Coupling (p. 354))
Maxwell stress-stiffening matrix (discussed in Magnetoelasticity (p. 361)) [3]

Maxwell stress coupling matrix (discussed in Magnetoelasticity (p. 361)) [3]

Lorentz force coupling matrix (discussed in Magnetoelasticity (p. 361)) [3]

1. Coefficient matrices of second time derivatives of unknowns.

2. Coefficient matrices of first time derivative of unknowns

3. Coefficient matrices of unknowns

Vectors of Knowns

Symbol Meaning Associated Input /

Output Label
{Fnd} applied nodal force vector (discussed in Derivation of Structural FX ... MZ
Matrices (p. 12))
{Fnr} Newton-Raphson restoring load vector (discussed in FX ... MZ
Newton-Raphson Procedure (p. 730)
{Fth} thermal strain force vector (discussed in Derivation of Structural FX ... MZ
Matrices (p. 12))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
326 Published: 2023-10-23
 Coupled Effects

Symbol Meaning Associated Input /

Output Label
{Fpr} pressure load vector (discussed in Derivation of Structural FX ... MZ
Matrices (p. 12))
{Fac} force vector due to acceleration effects (that is, gravity) FX ... MZ
(discussed in Derivation of Structural Matrices (p. 12))
{Fjb} Lorentz force vector (discussed in Derivation of Electromagnetic FX ... FZ
Matrices (p. 176))
{Fmx} Maxwell force vector (discussed in Derivation of FX ... FZ
Electromagnetic Matrices (p. 176))
{Fe} electrostatic body-force load vector (discussed in FX ...FZ
Electroelasticity (p. 345))
{Fb} body-force load vector due to non-gravity effects (discussed FX ... MZ
in Derivation of Heat Flow Matrices (p. 217))
{Fdi} diffusion strain force vector (discussed in FX ... MZ
Structural-Diffusion Coupling (p. 354))
{Qnd} applied nodal heat flow rate vector (discussed in Derivation HEAT, HBOT, HE2, ...
of Heat Flow Matrices (p. 217)) HTOP
{Qf} heat flux vector (discussed in Derivation of Heat Flow HEAT, HBOT, HE2, ...
Matrices (p. 217)) HTOP
{Qc} convection surface vector (discussed in Derivation of Heat HEAT, HBOT, HE2, ...
Flow Matrices (p. 217)) HTOP
{Qg} heat generation rate vector for causes other than Joule heating HEAT, HBOT, HE2, ...
(discussed in Derivation of Heat Flow Matrices (p. 217)) HTOP
{Qj} heat generation rate vector for Joule heating (discussed in HEAT
Electromagnetic Field Evaluations (p. 183))
{Qp} Peltier heat flux vector (discussed in Thermoelectrics (p. 348)) HEAT
ted
{Q } heat generation rate vector for thermoelastic damping HEAT
applied nodal source current vector (associated with {A}) CSGZ
(discussed in Derivation of Electromagnetic Matrices (p. 176))
applied nodal flux vector (associated with {ϕ}) (discussed in FLUX
Derivation of Electromagnetic Matrices (p. 176))
source (Biot-Savart) vector (discussed in Derivation of FLUX
Electromagnetic Matrices (p. 176))
coercive force (permanent magnet) vector (discussed in FLUX
Derivation of Electromagnetic Matrices (p. 176))
source current vector (discussed in Derivation of FLUX
Electromagnetic Matrices (p. 176))
{Ind} applied nodal electric current vector (discussed in Derivation AMPS
of Electromagnetic Matrices (p. 176))
{Lnd} applied nodal charge vector (discussed in Piezoelectrics (p. 334))AMPS (CHRG for
PLANE222,
PLANE223,
SOLID225,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 327
Coupling

Symbol Meaning Associated Input /

Output Label
SOLID226, and
SOLID227)
{Lc} charge density load vector (discussed in Derivation of CHRGD
Electromagnetic Matrices (p. 176))
{Lsc} surface charge density load vector (discussed in Derivation of CHRGS
Electromagnetic Matrices (p. 176))
{Lth} thermo-piezoelectric load vector (discussed in TEMP, EPTH
Piezoelectrics (p. 334))
{Wnd} applied nodal fluid flow vector (discussed in FLUID116 - FLOW
Coupled Thermal-Fluid Pipe (p. 513))
{Wh} static head vector (discussed in FLUID116 - Coupled FLOW
Thermal-Fluid Pipe (p. 513))
{R} diffusion flow rate vector RATE

Vectors of Unknowns

{u} displacement vector (discussed in Derivation of Structural UX ... ROTZ

Matrices (p. 12))
{T} thermal potential (temperature) vector (discussed in Derivation TEMP, TBOT, TE2,
of Heat Flow Matrices (p. 217)) ... TTOP
{V} electric potential vector (discussed in Derivation of VOLT
Electromagnetic Matrices (p. 176))
{ΔV} voltage drop in a stranded coil analysis (discussed in Stranded VOLT
Coil Analysis)
{ν} time integrated electric potential vector (discussed in Derivation VOLT
of Electromagnetic Matrices (p. 176))
{φ} magnetic scalar potential vector (discussed in Derivation of MAG
Electromagnetic Matrices (p. 176))
{A} magnetic vector potential or edge-flux (discussed in Derivation AZ
of Electromagnetic Matrices (p. 176))
{i} electric current vector (discussed in Stranded Coil CURR
Analyses (p. 193))
{e} electromagnetic force drop vector (discussed in Stranded Coil EMF
Analyses (p. 193))
{P} pressure vector (discussed in and Derivation of Acoustic PRES
Matrices (p. 248))
{C} concentration vector (discussed in Derivation of Diffusion CONC
Matrices (p. 307))
. time derivative
.. second time derivative

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
328 Published: 2023-10-23
 Thermoelasticity

10.2. Thermoelasticity
The capability to do a thermoelastic analysis exists in the following elements:

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid
LINK228 - 3D Coupled-Field Link

These elements support both the thermal expansion and piezocaloric effects, and use the strong (matrix)
coupling method.

In addition to the above, the following elements support the thermal expansion effect only in the form
of a thermal strain load vector, i.e. use weak coupling method:

SOLID5 - 3D 8-Node Coupled-Field Solid

PLANE13 - 2D 4-Node Coupled-Field Solid
SOLID98 - 3D 10-Node Coupled-Field Solid

Constitutive Equations of Thermoelasticity

The coupled thermoelastic constitutive equations (Nye([356] (p. 945))) are:

(10.37)

(10.38)

where:

{ε} = total strain vector = [εx εy εz εxy εyz εxz]T

S = entropy density
{σ} = stress vector = [σx σy σz σxy σyz σxz]T
ΔT = temperature change = T - Tref
T = current temperature
T0 = absolute reference temperature = Tref + Toff
Tref = reference temperature (input on TREF command or as REFT on MP command)
Toff = offset temperature from absolute zero to zero (input on TOFFST command)
[D] = elastic stiffness matrix (inverse defined in Equation 2.4 (p. 7) or input using TB,ANEL
command)
{α} = vector of coefficients of thermal expansion = [α x α y α z 0 0 0]T (input using, for example,
ALPX, ALPY, ALPZ on MP command)
ρ = density (input as DENS on MP command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 329
Coupling

Cp = specific heat at constant stress or pressure (input as C on MP command)

Using {ε} and ΔT as independent variables, and replacing the entropy density S in Equation 10.38 (p. 329)
by heat density Q using the second law of thermodynamics for a reversible change
(10.39)

we obtain
(10.40)
(10.41)

where:

{β} = vector of thermoelastic coefficients = [D] {α}

Substituting Q from Equation 10.41 (p. 330) into the heat flow equation Equation 6.1 (p. 207) produces:
(10.42)

where:

Kxx, Kyy, Kzz = thermal conductivities (input as KXX, KYY, KZZ on MP command)

Derivation of Thermoelastic Matrices

Applying the variational principle to stress equation of motion and the heat flow conservation equation
coupled by the thermoelastic constitutive equations, produces the following finite element matrix
equation:

(10.43)

where:

[M] = element mass matrix (defined by Equation 2.58 (p. 14))

[C] = element structural damping matrix (discussed in Damping Matrices (p. 694))
[K] = element stiffness matrix (defined by Equation 2.58 (p. 14))
{u} = displacement vector
{F} = sum of the element nodal force (defined by Equation 2.56 (p. 14)) and element pressure
(defined by Equation 2.58 (p. 14)) vectors
[Ct] = element specific heat matrix (defined by Equation 6.28 (p. 218))
[Kt] = element thermal conductivity matrix (defined by Equation 6.28 (p. 218))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
330 Published: 2023-10-23
 Thermoelasticity

{T} = temperature vector

{Q} = sum of the element heat generation load and element convection surface heat flow
vectors (defined by Equation 6.28 (p. 218))

[B] = strain-displacement matrix (see Equation 2.44 (p. 12))

{N} = element shape functions
[Ctu] = element thermoelastic damping matrix = -T0[Kut]T

Energy Calculation

In static and transient thermoelastic analyses, the element instantaneous total strain energy is calculated
as:

(10.44)

where:

Ut = total strain energy (output as an NMISC element item UT).

Note that Equation 10.44 (p. 331) uses the total strain, whereas the standard strain energy (output as
SENE) uses the elastic strain.

In a harmonic thermoelastic analysis, the time-averaged element total strain energy is given by:

(10.45)

where:

{ε}* = complex conjugate of the total strain

The real part of Equation 10.45 (p. 331) represents the average stored strain energy, while its imaginary
part - the average energy loss due to thermoelastic damping.

The thermoelastic damping can be quantified by the quality factor Q derived from the total strain energy
Equation 10.45 (p. 331) using the real and imaginary solution sets:

(10.46)

where:

Ne = number of thermoelastic elements

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 331
Coupling

10.3. Thermoplasticity
The capability to do a thermoplastic analysis exists in the following elements:

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid
LINK228 - 3D Coupled-Field Link

These elements support the thermoplastic effect which manifests itself as an increase in temperature
during plastic deformation due to the conversion of some of the plastic work into heat.

In a thermoplastic analysis, the stress equation of motion (Equation 2.51 (p. 13)) and heat flow conser-
vation equation (Equation 6.1 (p. 207)) are coupled by the plastic heat density rate defined as:
(10.47)

where:

β = fraction of plastic work coefficient (input as QRATE on MP command)

= plastic work rate =

where:

= stress vector =

= plastic strain vector =

The coupled-field finite element matrix equation for the thermoplastic analysis is:

(10.48)

where:

[M] = element mass matrix (defined by Equation 2.58 (p. 14))

[C] = element structural damping matrix (discussed in Damping Matrices (p. 694))
[K] = element stiffness matrix (defined by Equation 2.58 (p. 14))
{u} = displacement vector
{F} = sum of the element nodal force (defined by Equation 2.56 (p. 14)) and element pressure
(defined by Equation 2.58 (p. 14)) vectors
[Ct] = element specific heat matrix (defined by Equation 6.28 (p. 218))
[Kt] = element diffusion conductivity matrix (defined by Equation 6.28 (p. 218))
{T} = temperature vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
332 Published: 2023-10-23
 Thermoviscoelasticity

{Q} = sum of the element heat generation rate load and element convection surface heat flow
vectors (defined by Equation 6.28 (p. 218))

= element plastic heat generation rate load =

where:

= element plastic heat density rate at substep n (output as NMISC,5)

{N} = element shape functions

10.4. Thermoviscoelasticity
The capability to do a thermoviscoelasic analysis exists in the following elements:

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid
LINK228 - 3D Coupled-Field Link

These elements support the thermoviscoelastic effect which manifests itself as an increase in temperature
during viscoelastic deformation due to the conversion of some part of the viscoelastic loss into heat.

In a thermoviscoelastic analysis, the stress equation of motion (Equation 2.51 (p. 13)) and heat flow
conservation equation (Equation 6.1 (p. 207)) are coupled by the viscoelastic heat density rate defined
as:
(10.49)

where:

β = fraction of the energy dissipation density in a viscoelastic material converted to heat

(input as QRATE on the MP command)
= energy dissipation density rate calculated from the incremental energy dissipation
(output as SEND,VDAM); see Equation 4.51 in the Material Reference for the Prony series formu-
lation (TB,PRONY) and Equation 4.232 (p. 132) in the Mechanical APDL Theory Reference (p. 1)
for the Bergstrom-Boyce material model (TB,BB).

The coupled-field finite element matrix equation for the thermoviscoelastic analysis is:

(10.50)

where:

[M] = element mass matrix (defined by Equation 2.58 (p. 14))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 333
Coupling

[C] = element structural damping matrix (discussed in Damping Matrices (p. 694))
[K] = element stiffness matrix (defined by Equation 2.58 (p. 14))
{u} = displacement vector
{F} = sum of the element nodal force (defined by Equation 2.56 (p. 14)) and element pressure
(defined by Equation 2.58 (p. 14)) vectors
[Ct] = element specific heat matrix (defined by Equation 6.28 (p. 218))
[Kt] = element thermal conductivity matrix (defined by Equation Equation 6.28 (p. 218))
{T} = temperature vector
= sum of the element heat generation rate load and element convection surface heat flow
vectors (defined by Equation 6.28 (p. 218))

= element viscoelastic heat generation rate load =

where:
= element viscoelastic heat density rate at substep n (output as NMISC,6)
{N} = element shape functions

10.5. Piezoelectrics
The capability of modeling piezoelectric response exists in the following elements:

SOLID5 - 3D 8-Node Coupled-Field Solid

PLANE13 - 2D 4-Node Coupled-Field Solid
SOLID98 - 3D 10-Node Coupled-Field Solid
PLANE222 - 2D 4-Node Coupled-Field Solid
PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid

10.5.1. Constitutive Equations of Piezoelectricity

In linear piezoelectricity the equations of elasticity are coupled to the charge equation of electrostatics
by means of piezoelectric constants (IEEE Standard on Piezoelectricity([90] (p. 929))):
(10.51)
(10.52)

or equivalently

(10.53)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
334 Published: 2023-10-23
 Piezoelectrics

where:

{T} = stress vector (referred to as {σ} elsewhere in this manual) composed of mechanical {Tm}
and piezoelectric {Tp} stress components
{D} = electric flux density vector composed of electrical {De} and piezoelectric {Dp} flux com-
ponents
{S} = elastic strain vector (referred to as {εel} elsewhere in this manual)
{E} = electric field intensity vector
[cE] = elasticity matrix (evaluated at constant electric field (referred to as [D] elsewhere in
this manual))
[e] = piezoelectric stress matrix
[εS] = dielectric matrix (evaluated at constant mechanical strain)

Equation 10.51 (p. 334) and Equation 10.52 (p. 334) are the usual constitutive equations for structural
and electrical fields, respectively, except for the coupling terms involving the piezoelectric matrix [e].

The elasticity matrix [c] is the usual [D] matrix described in Structural Fundamentals (p. 5) (input
using the MP commands). It can also be input directly in uninverted form [c] or in inverted form [c]-1
as a general anisotropic symmetric matrix (input using TB,ANEL):

(10.54)

The piezoelectric stress matrix [e] (input using TB,PIEZ with TBOPT = 0) relates the electric field vector
{E} in the order X, Y, Z to the stress vector {T} in the order X, Y, Z, XY, YZ, XZ and is of the form:

(10.55)

The piezoelectric matrix can also be input as a piezoelectric strain matrix [d] (input using TB,PIEZ with
TBOPT = 1). Mechanical APDL will automatically convert the piezoelectric strain matrix [d] to a
piezoelectric stress matrix [e] using the elasticity matrix [c] at the first defined temperature:
(10.56)

The orthotropic dielectric matrix [εS] uses the electrical permittivities (input as PERX, PERY and PERZ
on the MP commands) and is of the form:

(10.57)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 335
Coupling

The anisotropic dielectric matrix at constant strain [εS] (input using TB,DPER,,,,0 command) is used
by PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227 and is of the form:

(10.58)

The dielectric matrix can also be input as a dielectric permittivity matrix at constant stress [εT] (input
using TB,DPER,,,,1). The program will automatically convert the dielectric matrix at constant stress to
a dielectric matrix at constant strain:
(10.59)

where:

[εS] = dielectric permittivity matrix at constant strain

[εT] = dielectric permittivity matrix at constant stress
[e] = piezoelectric stress matrix
[d] = piezoelectric strain matrix

10.5.2. Losses in a Piezoelectric Analysis

Structural and electrical losses can be introduced in dynamic piezoelectric analyses using various
material properties. This section describes losses that are included in the piezoelectric constitutive
equations of elements PLANE222, PLANE223, SOLID225, SOLID226, and SOLID227 and result in heat
generation.

Viscous structural damping can be included in a transient or harmonic piezoelectric analysis using a
stiffness matrix multiplier or in the form of a viscosity matrix:
(10.60)

where:

βm = stiffness matrix multiplier (input as BETD on MP or SDAMP on TB)

[η] = anisotropic viscosity matrix (input using TB,AVIS with TBOPT = 0)
= strain rate vector

The stiffness matrix multiplier βm is described in Damping Matrices (p. 694). The symmetric viscosity
matrix [η] is of the form:

(10.61)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
336 Published: 2023-10-23
 Piezoelectrics

Anisotropic damping can also be in the form of fluency matrix [ζ] (input using TB,AVIS with TBOPT
= 1):
(10.62)

where:

[sE] = elastic compliance matrix (evaluated at constant electric field)

The program will convert the fluency matrix [ζ] to viscosity matrix [η] using the elastic stiffness matrix
[cE]:
(10.63)

Hysteretic structural damping can be included in harmonic and damped modal piezoelectric analyses
using several structural damping coefficients or in the form of a loss tangent matrix:
(10.64)

where:

j = imaginary unit
βξ = damping ratio (input as DMPR on MP)
m = constant structural damping coefficient (input as DMPS on MP)
g = structural damping coefficient (input using TB,SDAMP with TBOPT = STRU)
[ψ] = elastic loss tangent matrix (input using TB,ELST with TBOPT = 0)
denotes the Hadamard (or entry-wise) matrix product

The damping coefficients βξ, m, and g are described in Damping Matrices (p. 694). The symmetric
elastic loss tangent matrix [ψ] (input using TB,ELST) is of the form:

(10.65)

In a harmonic and modal piezoelectric analysis, the contributions from all types of viscous and hysteretic
damping can be summarized as follows:
(10.66)

Note that constant structural damping coefficients input using BETAD and DMPSTR commands are
not included in the above piezoelectric constitutive relations since they are not directly associated
with any loss in the material.

Electrical losses can be included in a harmonic piezoelectric analysis by specifying electrical resistivity,
electrical loss tangent coefficient, or a dielectric loss tangent matrix:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 337
Coupling

(10.67)

where:

j = imaginary unit

electrical conductivity matrix

ρxx, ρyy, ρzz = coefficients of electrical resistivity (input as RSVX (Y, Z) on MP command)
tanδ = electric loss tangent coefficient (input as LSST on MP command)
[φ] = dielectric loss tangent matrix (input using TB,DLST with TBOPT = 0)

Electrical resistivity and loss tangent are described in Quasistatic Electric Analysis (p. 175). The symmetric
dielectric loss tangent matrix [φ] (input using TB,DLST) is of the form:

(10.68)

The electric loss tangent coefficient (MP,LSST) and the dielectric loss tangent matrix (TB,DLST) can
also be included in a piezoelectric damped modal analysis (Equation 10.72 (p. 340)).

Electrical resistivity (MP,RSVX/Y/Z) is the only type of electric loss that can be included in a transient
piezoelectric analysis. To account for this loss, the transient piezoelectric analysis must be current-
based (Equation 10.73 (p. 340)).

Different types of losses and their availability in dynamic piezoelectric analyses are summarized in
the table below.

Table 10.4: Losses in Piezoelectric Analysis

Loss Analysis Type

Type Input Modal
Full Full
DAMP UNSYM QRDAMP Harmonic Transient
Structural Viscous MP,BETD x x x x
damping TB,SDAMP,,,,BETD x x x x
TB,AVIS x x x x
Hysteretic MP,DMPR x x x x
damping MP,DMPS x x x x
TB,SDAMP,,,,STRU x x x x
TB,ELST x x x x
Electrical Loss tangent MP,LSST x x x x
TB,DLST x x x x

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
338 Published: 2023-10-23
 Piezoelectrics

Loss Analysis Type

Type Input Modal
Full Full
DAMP UNSYM QRDAMP Harmonic Transient
Resistivity MP,RSVX/Y/Z x x[a]
[a] The transient analysis must be a current-based piezoelectric analysis (KEYOPT(1) = 101).

10.5.3. Derivation of Piezoelectric Matrices

After the application of the variational principle and finite element discretization (Allik([82] (p. 929))),
the coupled finite element matrix equation derived for a one element model is:

(10.69)

where:

[K] = element stiffness matrix (defined by Equation 2.58 (p. 14))

[M] = element mass matrix (defined by Equation 2.58 (p. 14))
[C] = element structural damping matrix (discussed in Damping Matrices (p. 694))
[Cm] = element anisotropic structural damping matrix
{F} = vector of nodal and surface forces (defined by Equation 2.56 (p. 14) and Equa-
tion 2.58 (p. 14))
[Kd] = element dielectric permittivity coefficient matrix ([Kvs] in Equation 5.117 (p. 182) or [Kvh]
in Equation 5.116 (p. 182))
{L} = vector of nodal, surface, and body charges (defined by Equation 5.117 (p. 182))

[B] = strain-displacement matrix (see Equation 2.44 (p. 12))

[Ce] = element electric damping matrix

{εth} = thermal strain vector (as defined by equation Equation 2.3 (p. 6))
{N} = element shape functions

Note:

In a strongly coupled thermo-piezoelectric analysis (see Equation 10.24 (p. 320)), the
electric potential and temperature degrees of freedom are coupled by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 339
Coupling

where:

{α} = vector of coefficient of thermal expansion.

In a harmonic analysis, the structural [Cm] and electric [Ce] damping matrices are formed as follows:
(10.70)

(10.71)

where:

= element viscous structural damping matrix

= element hysteretic structural damping

= element resistive damping matrix

= element hysteretic electric damping

In a modal analysis, when the frequency ω is not known initially, the hysteretic damping terms are
included in the finite element matrix equation as imaginary stiffness:

(10.72)

A current-based piezoelectric analysis is obtained by differentiating the electric charge equation of

the piezoelectric coupled system (Equation 10.69 (p. 339)) with respect to time:

(10.73)

where:

[KV] = element electrical conductivity matrix (defined by Equation 5.115 (p. 182))
{I} = vector of nodal currents (input/output as AMPS)

The current-based piezoelectric analysis can be harmonic or transient. Resistive losses in a piezoelectric
material can be simulated using a transient current-based analysis.

10.5.4. Piezoelectric Results

This section describes the results calculation in a piezoelectric analysis using elements PLANE222,
PLANE223, SOLID225, SOLID226, and SOLID227.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
340 Published: 2023-10-23
 Piezoelectrics

The total stress {T} (output as S) and total electric flux density {D} (output as D) are computed from
elastic strain {S} and electric field {E} as shown in Equation 10.51 (p. 334) and Equation 10.52 (p. 334).
In dynamic piezoelectric analyses, structural and electrical damping is included in the values of stress
and electric flux according to Equation 10.60 (p. 336), Equation 10.64 (p. 337), and Equation 10.67 (p. 338).
Both the stress {T} and electric flux density {D} are evaluated at the integration point locations and
then moved or extrapolated to the corner nodes for output.

In addition to stress {T} and electric flux {D}, piezoelectric results include the total electric current
density {J} (output as JS) calculated as a sum of the total (conduction and displacement) electric current
{Je} (see Quasistatic Electric Analysis (p. 191)) and the piezoelectric displacement current {Jp}:
(10.74)

The total electric current density {J} is calculated at the element centroidal location and output in the
global Cartesian coordinate system. It can be used as a source current density in a subsequent mag-
netic analysis (LDREAD,JS).

In a current-based piezoelectric analysis, the conduction current density {Jc} = [σ]{E} is output (as JC)
instead of the electric flux {D}.

In static and transient piezoelectric analyses, the element instantaneous energies are calculated as:

(10.75)

(10.76)

(10.77)

where:

= stored elastic strain energy (output as an NMISC element item UE).

= dielectric energy (output as an NMISC element item UD)
= electromechanical or mutual energy (output as an NMISC element item UM)

In a transient analysis, the kinetic energy (output as KENE) and the damping energy (output as DENE)
are calculated as shown in Equation 14.340 (p. 781) and Equation 14.352 (p. 782), respectively.

In harmonic and modal piezoelectric analyses, the time-averaged element energies are calculated as:

(10.78)

(10.79)

(10.80)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 341
Coupling

= complex conjugate of the elastic strain

= complex conjugate of the electric flux density

The time-averaged kinetic and damping energies are calculated as shown in Equation 14.342 (p. 781)
and Equation 14.353 (p. 783), respectively.

The real parts of Equation 10.78 (p. 341) and Equation 10.79 (p. 341) represent the average stored
elastic and dielectric energies, respectively. The imaginary parts represent the average elastic and
electric losses. Therefore, the quality factor can be calculated from the total stored energy as:

(10.81)

where:

= number of piezoelectric elements

The total stored energy + is output as SENE. The factor can therefore be derived from the
real and imaginary records of SENE summed over the piezoelectric elements.

In a damped modal analysis, the quality factor can also be calculated using the real and imaginary
part of the complex eigenvalue:
(10.82)

The mutual energy can be used to calculate the electromechanical coupling coefficient as:

(10.83)

The piezoelectric Poynting vector is calculated as a sum of mechanical and electrical (quasi-static
approximation) Poynting vectors:
(10.84)

where:

= stress matrix

{v} = velocity
φ = electric potential

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
342 Published: 2023-10-23
 Piezoelectrics

{J} = total electric current density

In a harmonic analysis, the time-averaged Poynting vector is:

(10.85)

where:

{v}* = complex conjugate of the velocity vector

{J}* = complex conjugate of the total electric current vector

The calculated Poynting vector is output as P. In a harmonic analysis, the real part of the Poynting
vector represents the average power flow.

In a transient piezoelectric analysis, the heat generation rate produced by viscous losses is calculated
as follows:
(10.86)

In a harmonic piezoelectric analysis, the generated heat is due to structural and electrical losses. The
structural part of the time-averaged heat rate per unit volume is calculated from the complex values
of mechanical stress {Tm} and strain {S} as:
(10.87)

The time-averaged electrical heat, or Joule heat, rate is calculated as:

(10.88)

The total heat rate {Q} = {Qm} + {Qe} is stored in both the real and imaginary sets of the JHEAT output
item and can be used as a heat generation load in a subsequent thermal analysis (LDREAD,HGEN).

10.5.5. Perfectly Matched Layers (PML) in Piezoelectric Medium

Perfectly matched layers are artificial anisotropic materials that absorb all incoming waves without
any reflections, except for the graze wave that parallels the PML interface in the propagation direction.
PMLs are currently constructed for the propagation of elastic waves in the elastic media in harmonic
response analysis.

Considering two Cartesian coordinate systems, (1) a global system with respect to an orthogonal
basis and (2) a local system with respect to another orthogonal basis , the displacements
and voltage of the piezoelectric PML medium in the local coordinate system are
governed by the following equations:

(10.89)

(10.90)

(10.91)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 343
Coupling

(10.92)

(10.93)

(10.94)

where:

= components of the infinitesimal strain tensor in the local Cartesian coordinate system
= components of the material stiffness tensor in the local Cartesian coordinate system
= components of the piezoelectric stress matrix in the local Cartesian coordinate system
= components of the dielectric matrix in the local Cartesian coordinate system
= components of the electric flux in the local Cartesian coordinate system
= components of the electric field in the local Cartesian coordinate system
= non-zero complex-valued coordinate stretching functions in the direction of the local
Cartesian coordinate system
= density of the elastic medium

The tensors and vectors in the local coordinate system are transformed into the global coordinate
system using the rotation-of-basis transformation matrix with the component , such as

vector and matrix . On multiplying Equation 10.89 (p. 343) and

Equation 10.90 (p. 343) with s = s1s2s3, and taking the fact that sj is a function of only, the governing
equations of the displacements and voltage in the global coordinate system are
written as:
(10.95)
(10.96)
(10.97)

(10.98)

(10.99)
(10.100)

The PML material is defined using coupled-field elements (PLANE222, PLANE223, SOLID225, SOLID226,
or SOLID227) with KEYOPT(15) = 1. The PML element coordinate system (PSYS command) uniquely
identifies each PML region. For more information about using PMLs, refer to Perfectly Matched Layers
(PML) in the Acoustic Analysis Guide.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
344 Published: 2023-10-23
 Electroelasticity

10.6. Electroelasticity
The capability of modeling electric force coupling in elastic dielectrics exists in the following elements:

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid

Elastic dielectrics exhibit a deformation when subject to an electrostatic field. The electric body force
that causes the deformation can be derived from the Maxwell stress tensor [σM] (Landau and Lif-
shitz([355] (p. 945))).

(10.101)

where:

{E} = electric field intensity vector

{D} = electric flux density vector

Applying the variational principle to the stress equation of motion and to the charge equation of elec-
trostatics coupled by electric force produces the following finite element equation:

(10.102)

where:

[K] = element structural stiffness matrix (see [Ke] in Equation 2.58 (p. 14))
[M] = element mass matrix (see [Me] in Equation 2.58 (p. 14))

[Kd] = element dielectric permittivity coefficient matrix (see [Kvs] in Equation 5.117 (p. 182))
[C] = element structural damping matrix (discussed in Damping Matrices (p. 694))
[Cvh] = element dielectric damping matrix (defined by Equation 5.116 (p. 182))
{F} = vector of nodal and surface forces (defined by Equation 2.56 (p. 14) and Equa-
tion 2.58 (p. 14))
{L} = vector of nodal, surface, and body charges (see {Le} in Equation 5.117 (p. 182))

The electrostatic softening matrix [Keu] and the coupling matrix [KeV] are calculated as derivatives of
the nodal electric force {Fe} with respect to displacement and voltage:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 345
Coupling

These derivatives are obtained by applying the chain rule to the following expression for the electric
force (output as FMAG):

where:
[B] = strain-displacement matrix (see Equation 2.44 (p. 12))

vol = element volume

The strong (matrix) coupling between structural and electric equations in the finite element system
(Equation 10.102 (p. 345)) allows the linear perturbation modal and harmonic analyses to be used following
a nonlinear static or transient analysis.

Note that for KEYOPT(4) = 0, 1, 2, and 4, the finite element system (Equation 10.102 (p. 345)) is symmetric
due to the negative sign assigned to the electric equation.

The electroelastic analysis results are similar to those computed in a piezoelectric analysis ( Piezoelec-
trics (p. 334)) with the coupling matrix derived from the Maxwell stress vector as summarized in the
following table.

Table 10.5: Electroelastic Analysis Results

Equation
Linear perturbation
Nonlinear static and harmonic and modal
Element result Output as transient analyses analyses
Stress {T} S Equation 2.60 (p. 15) Equation 10.51 (p. 334) and
Electric flux density {D} D Equation 5.75 (p. 174) Equation 10.52 (p. 334)

with

Electric current density JS

{J} Equation 10.74 (p. 341) with
Energy
Elastic UE NMISC,1 (UE) Equation 10.75 (p. 341) Equation 10.78 (p. 341)
Dielectric UD NMISC,2 (UD) Equation 10.76 (p. 341) Equation 10.79 (p. 341)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
346 Published: 2023-10-23
 Piezoresistivity

Equation
Mutual UM NMISC,3 (UM) Equation 10.77 (p. 341) Equation 10.80 (p. 341)

with
with

Total stored SENE UE + UD

Kinetic KENE Equation 14.340 (p. 781) Equation 14.342 (p. 781)
Damping DENE Equation 14.352 (p. 782) Equation 14.353 (p. 783)
[a]
Heat generation rate JHEAT Equation 10.86 (p. 343) Equation 10.87 (p. 343) and
Equation 10.88 (p. 343)
[a] Supported losses (see Table 10.4: Losses in Piezoelectric Analysis (p. 338)) are:

• Structural isotropic viscous damping (input as MP,BETD or TB,SDAMP,,,,BETD)

• Structural isotropic hysteretic damping (input as MP,DMPR, MP,DMPS, or

TB,SDAMP,,,,STRU)

• Electric isotropic loss tangent (input as MP,LSST), anisotropic loss tangent (input as
TB,DLST) and electrical resistivity (input as MP,RSVX/Y/Z).

10.7. Piezoresistivity
The capability to model piezoresistive effect exists in the following elements:

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid

In piezoresistive materials, stress or strain cause a change of electric resistivity:

(10.103)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 347
Coupling

(10.104)

where:

{r} = vector of matrix [r] components = [rx ry rz rxy ryz rxz]T

Similarly, for strains:

(10.105)

where:

[m] = piezoresistive strain matrix (input on TB,PZRS command with TBOPT = 1)

{εel} = elastic strain vector

The coupled-field finite element matrix equation for the piezoresistive analysis is given by:

(10.106)

The terms used in the above equation are explained in Piezoresistive Analysis (p. 321) where the con-
ductivity matrix [Kv] is derived as:

(10.107)

10.8. Thermoelectrics
The capability to model thermoelectric effects exists in the following elements:

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
348 Published: 2023-10-23
 Thermoelectrics

SOLID226 - 3D 20-Node Coupled-Field Solid

SOLID227 - 3D 10-Node Coupled-Field Solid
LINK228 - 3D Coupled-Field Link

These elements support the Joule heating effect (irreversible), and the Seebeck, Peltier, and Thomson
effects (reversible).

In addition to the above, the following elements support a basic thermoelectric analysis that takes into
consideration Joule heating effect only:

SOLID5 - 3D 8-Node Coupled-Field Solid

LINK68 - 3D 2-Node Coupled Thermal-Electric Line
SOLID98 - 3D 10-Node Coupled-Field Solid
SHELL157 - 3D 4-Node Thermal-Electric Shell

Constitutive Equations of Thermoelectricity

The coupled thermoelectric constitutive equations (Landau and Lifshitz([355] (p. 945))) are:
(10.108)
(10.109)

Substituting [Π] with T[α] to further demonstrate the coupling between the above two equations,
(10.110)
(10.111)

where:

[Π] = Peltier coefficient matrix = T[α]

T = absolute temperature

{q} = heat flux vector (output as TF)

{J} = electric current density (output as JC for elements that support conduction current calcu-
lation)

{E} = electric field (output as EF)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 349
Coupling

α xx, α yy, α zz = Seebeck coefficients (input as SBKX, SBKY, SBKZ on MP command)

Kxx, Kyy, Kzz = thermal conductivities (input as KXX, KYY, KZZ on MP command)
ρxx, ρyy, ρzz = resistivity coefficients (input as RSVX, RSVY, RSVZ on MP command)

Note that the Thomson effect is associated with the temperature dependencies of the Seebeck coefficients
(MPDATA,SBKX also SBKY, SBKZ).

Derivation of Thermoelectric Matrices

After the application of the variational principle to the equations of heat flow (Equation 6.1 (p. 207)) and
of continuity of electric charge (Equation 5.5 (p. 164)) coupled by Equation 10.108 (p. 349) and Equa-
tion 10.109 (p. 349), the finite element equation of thermoelectricity becomes (Antonova and
Looman([91] (p. 929))):

(10.112)

where:

[Kt] = element thermal conductivity matrix (defined by Equation 6.28 (p. 218))
[Ct] = element specific heat matrix (defined by Equation 6.28 (p. 218))
{Q} = sum of the element heat generation load and element convection surface heat flow
vectors (defined by Equation 6.28 (p. 218))
[Kv] = element electrical conductivity coefficient matrix (defined by Equation 5.115 (p. 182))
[Cv] = element dielectric permittivity coefficient matrix (defined by Equation 5.115 (p. 182))

= Peltier heat contribution to the element thermal

conductivity matrix

= element Seebeck coefficient coupling matrix

= element Peltier heat coupling matrix

= element Joule heat load vector

{N} = element shape functions

{I} = vector of nodal current load

The finite element equation Equation 10.112 (p. 350) is unsymmetric when Seebeck coefficients are
defined by issuing MP,SBKX (also RSVY, RSVZ). To make it symmetric, use KEYOPT(2) = 1 to apply Seebeck
current and Peltier heat coupling as load vectors.

When thermal conductivity, electrical resistivity, and Seebeck coefficients are temperature dependent
(MPTEMP/MPDATA), the stiffness matrix [K] in Equation 10.112 (p. 350) includes terms associated with
the derivatives with respect to temperature of these material properties. These terms that also make
Equation 10.112 (p. 350) unsymmetric can be turned off by setting KEYOPT(2) = 1.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
350 Published: 2023-10-23
 Porous Media Flow

10.9. Review of Coupled Electromechanical Methods

Strong Electromechanical coupling can be performed by the transducer element TRANS126, Gyimesi
and Ostergaard([249] (p. 938)), Gyimesi and Ostergaard([330] (p. 943)), TRANS126 - Electromechanical
Transducer (p. 526).

TRANS126 completely models the fully coupled system, converting electrostatic energy into mechanical
energy and vise versa as well as storing electrostatic energy. Coupling between electrostatic forces and
mechanical forces is obtained from virtual work principles (Gyimesi and Ostergaard([249] (p. 938))).

TRANS126 takes on the form of a 2-node line element with electrical voltage and mechanical displacement
DOFs as across variables and electric current and mechanical force as through variables. Input for the
element consists of a capacitance-stroke relationship that can be derived from electrostatic field solutions
and using the CMATRIX command macro (Gyimesi et al.([288] (p. 941)), Gyimesi and Oster-
gaard([289] (p. 941))).

The element can characterize up to three independent translation degrees of freedom at any point to
simulate 3D coupling. Thus, the electrostatic mesh is removed from the problem domain and replaced
by a set of TRANS126 elements hooked to the mechanical and electrical model providing a reduced
order modeling of a coupled electromechanical system (Gyimesi and Ostergaard ([286] (p. 941)), Gyimesi
et al.([287] (p. 941))).

TRANS126 allows treatment of all kinds of analysis types, including prestressed modal and harmonic
analyses. However, TRANS126 is limited geometrically to problems when the capacitance can be accurately
described as a function of a single degree of freedom, usually the stroke of a comb drive. In a bending
electrode problem, like an optical switch, obviously, a single TRANS126 element can not be applied.
When the gap is small and fringing is not significant, the capacitance between deforming electrodes
can be practically modeled reasonably well by several capacitors connected parallel. The EMTGEN
(electromechanical transducer generator) command macro can be applied to this case.

Convergence issues may be experienced with TRANS126 when applied to the difficult hysteretic pull-
in and release analysis (Avdeev et al.([331] (p. 943))) because of the negative total system stiffness matrix.
The issue is resolved when the augmented stiffness method is applied.

10.10. Porous Media Flow

The coupled pore-pressure thermal elements used in analyses involving porous media are listed in
Coupled Pore-Pressure-Thermal Element Support.

The program models porous media containing fluid by treating the porous media as a multiphase ma-
terial and applying an extended version of Biot's consolidation theory. The flow is considered to be a
single-phase fluid. The porous media can be fully or partially saturated.[[426] (p. 949)][[432] (p. 949)] Op-
tionally, heat transfer in the porous media can also be considered.

Following are the governing equations for Biot consolidation problems with heat transfer:

(10.113)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 351
Coupling

where:

σ = Total Cauchy stress

=

= matrix differentiation operator (3D form shown)

= Bulk density of porous media
= Displacement
= Bulk specific weight of porous media
= Gravity load direction (not to be confused with gravity magnitude)
= Flow flux vector
=
= Gradient operator (3D form shown)
= Biot coefficient
= Volumetric strain of the solid skeleton

= Pore pressure
= Compressibility parameter
= Degree of saturation of fluid
= Free strain
= Temperature
=
Density-specific heat term
= Porosity
= Density of solid skeleton and fluid
= Specific heats of solid skeleton and fluid
= Thermal conductivity

The total stress relates to the effective stress and pore pressure by:

where:

= Second-order identity tensor

The relationship between the effective stress and the elastic strain of solid skeletons is given by:

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
352 Published: 2023-10-23
 Porous Media Flow

= Second-order elastic strain tensor

= Fourth-order elasticity tensor

The relationship between the fluid flow flux and the pore pressure is described by Darcy's Law:

where:

= second-order permeability tensor

= relative permeability
= specific weight of fluid

For displacement , pressure , and temperature as the unknown degrees of freedom, linearizing
the governing equations gives:

(10.114)

The matrices are:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 353
Coupling

where:

= domain
= strain-displacement operator matrix
= displacement interpolation
= pressure interpolation
= temperature interpolation
= thermal load vector

The load force vector includes the body force and surface traction boundary conditions, the vector
includes the flow source, and the vector includes the heat source. ([426] (p. 949))

Combining the linearized equations for porous media with the equation of motion gives the matrix
equation:

(10.115)

where:

= structural damping matrix

The structural damping matrix can be input as Rayleigh damping (TB,SDAMP,,,,ALPD and/or
TB,SDAMP,,,,BETD).

Additional Information

For related information, see the following documentation:

Structural-Pore-Fluid-Diffusion-Thermal Analysis in the Coupled-Field Analysis Guide

Porous Media Material Properties in the Material Reference

10.11. Structural-Diffusion Coupling

The capability to perform a coupled structural-diffusion analysis exists in the following elements:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
354 Published: 2023-10-23
 Structural-Diffusion Coupling

PLANE222 - 2D 4-Node Coupled-Field Solid

PLANE223 - 2D 8-Node Coupled-Field Solid
SOLID225 - 3D 8-Node Coupled-Field Solid
SOLID226 - 3D 20-Node Coupled-Field Solid
SOLID227 - 3D 10-Node Coupled-Field Solid

High-order coupled-field elements PLANE223, SOLID226, and SOLID227 support the effects of diffusion
strain and hydrostatic stress migration (transport of particles due to a hydrostatic stress gradient) when
used in coupled-field analyses having structural and diffusion DOFs. Low-order elements PLANE222 and
SOLID225 support only the diffusion strain effect.

Constitutive Equations

In a coupled structural-diffusion analysis, the total strain is composed of elastic {εel} and diffusion {εd}
parts, respectively:
(10.116)

where:

{ε} = total strain vector = [εx εy εz εxy εyz εxz]T= {εel} + {εdi}

{εel} = elastic strain vector (output as EPEL)

{εdi}= diffusion strain vector (output as EPDI)

{σ} = stress vector = [σx σy σz σxy σyz σxz]T (output as SIG)

ΔC = concentration change = C - Cref

C = concentration =

= normalized concentration (input/output as CONC)

= saturated concentration (input as CSAT on MP command)

Cref = reference concentration (input as CREF on MP command)

[E] = elastic stiffness matrix (inverse defined in Equation 2.4 (p. 7) as [D]-1 or input using TB,ANEL
command)

{β}= vector of coefficients of diffusion expansion = [βx βy βz 0 0 0]T (input using BETX, BETY, BETZ on
MP command)

In addition to the diffusion strain coupling, the structural and diffusion DOFs are coupled by the hydro-
static stress migration effect:
(10.117)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 355
Coupling

{J} = diffusion flux density (input/output as DF)

= diffusivity matrix

Dxx, Dyy, and Dzz = diffusivity coefficients in the element's X, Y, and Z directions, respectively
(input as DXX, DYY, DZZ on MP command)
T = absolute temperature = Tc + Toff
Tc = current temperature (input/output as TEMP on D or BF commands)
Toff = offset temperature from absolute zero to zero (input on TOFFST command)
Ω/k = atomic volume constant (input as C2 on TBDATA command with TB,MIGR)
Ω = atomic volume
k = Boltzmann constant
= hydrostatic stress

Substituting Equation 10.116 (p. 355) into Equation 10.117 (p. 355), we obtain the following expression
for the diffusion flux density {J}:
(10.118)

where:

tr = trace operator

For more information on Equation 10.117 (p. 355) and the related material constant input, see Migration
Model in the Material Reference.

For more information on diffusion analysis, see Diffusion (p. 305).

Derivation of Structural-Diffusion Matrices

Applying the variational principle to the structural equation (discussed in Derivation of Structural
Matrices (p. 12)) and the diffusion equation (Equation 9.6 (p. 306)) coupled by the constitutive equations
(Equation 10.116 (p. 355) and Equation 10.118 (p. 356)), produces the following finite element matrix
equation for the structural-diffusion analysis:

(10.119)

where:

[Mu] = element mass matrix (defined by Equation 2.58 (p. 14))

[Cu] = element structural damping matrix (discussed in Damping Matrices (p. 694))

[Ku] = element stiffness matrix (defined by Equation 2.58 (p. 14))

{u} = nodal displacement vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
356 Published: 2023-10-23
 Thermal-Diffusion Coupling

{F} = sum of the element nodal force (defined by Equation 2.56 (p. 14)) and element pressure (defined
by Equation 2.58 (p. 14)) vectors

[Cd] = element diffusion damping matrix (defined by Equation 9.9 (p. 307))

[Kd] = element diffusion conductivity matrix (defined by Equation 9.9 (p. 307))

{C} = nodal concentration vector

{R} = nodal diffusion flow rate vector (defined by Equation 9.9 (p. 307))

= element diffusion strain stiffness matrix

= element transport conductivity matrix

= nonlinear part of the element diffusion conduct-

ivity matrix

= element conductivity matrix associated with

diffusion strain

= nonlinear part of the element conduct-

ivity matrix associated with diffusion strain

[B] = strain-displacement matrix (see Equation 2.44 (p. 12))

= concentration gradient (output as CG)

{N} = element shape functions

10.12. Thermal-Diffusion Coupling

The capability to perform a coupled thermal-diffusion analysis exists in the following elements:

PLANE222: 2D 4-Node Coupled-Field Solid

PLANE223: 2D 8-Node Coupled-Field Solid
SOLID225: 3D 8-Node Coupled-Field Solid
SOLID226: 3D 20-Node Coupled-Field Solid
SOLID227: 3D 10-Node Coupled-Field Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 357
Coupling

These elements support the effects of thermomigration (the transport of particles due to a temperature
gradient) and temperature-dependent saturated concentration when used in coupled-field analyses
having thermal and diffusion DOFs.

Constitutive Equations
In thermal-diffusion analyses, the diffusion flux {J} is coupled to temperature as follows:
(10.120)

where:

= diffusivity matrix

Dxx, Dyy, and Dzz = diffusivity coefficients in the element's X, Y, and Z directions, respectively
(input as DXX, DYY, DZZ on MP command)
C = concentration =
= normalized concentration (input/output as CONC)
= saturated concentration (input as MP,CSAT)
T = absolute temperature = Tc + Toff
Tc = current temperature (input/output as TEMP on D or BF commands)
Toff = offset temperature from absolute zero to zero (input on TOFFST command)
Q/k = heat transport constant (input as C3 on TBDATA command with TB,MIGR)
Q = particle heat of transport
k = Boltzmann constant

If saturated concentration is a function of temperature Csat = Csat(T), Equation 10.120 (p. 358) takes the
form:

(10.121)

where the derivative of Csat with respect to the temperature is numerically evaluated by the
program.

For more information on Equation 10.120 (p. 358) and the related material constant input, see Migration
Model in the Material Reference.

For more information on diffusion analysis, see Diffusion (p. 305).

Derivation of Thermal-Diffusion Matrices

Applying the variational principle with respect to thermal and diffusion DOFs to the equations of heat
flow Equation 6.1 (p. 207) and mass continuity equations Equation 9.7 (p. 306) coupled by the constitutive

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
358 Published: 2023-10-23
 Thermal-Diffusion Coupling

equation Equation 10.121 (p. 358), we obtain the following finite element matrix equation for the electric-
diffusion analysis:

(10.122)

where:

{T} = nodal temperature vector (input/output as TEMP)

{C} = nodal concentration vector (input/output as CONC)
[Ct] = element specific heat matrix (defined by Equation 6.28 (p. 218))
[Cd] = element diffusion damping matrix (defined by Equation 9.9 (p. 307))
[Kt] = element thermal conductivity matrix (defined by Equation 6.28 (p. 218))
[Kd] = element diffusion conductivity matrix (defined by Equation 9.9 (p. 307))

= element transport conductivity matrix

= nonlinear part of the element transport

conductivity matrix
= thermal gradient (output as TG)

= element thermal-diffusion conductivity matrix

produced by

= nonlinear part of the element diffusion

conductivity matrix associated with thermomigration

nonlinear part

of the element diffusion conductivity matrix produced by

= element thermal-diffusion damping matrix produced by

{Q} = sum of the nodal heat generation and convection loads (input/output as HEAT)
{R} = nodal diffusion flow rate vector (input/output as RATE)
{N} = element shape functions

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 359
Coupling

The finite element equation Equation 10.122 (p. 359) is unsymmetric. If a symmetric equation is desirable,
the thermomigration and temperature-dependent Csat coupling can be applied as a load vector
by setting KEYOPT(2) = 1 for the coupled-field elements
(PLANE222,PLANE223, SOLID225, SOLID226, SOLID227).

10.13. Electric-Diffusion Coupling

The capability to perform a coupled electrical-diffusion analysis exists in the following elements:

PLANE222: 2D 4-Node Coupled-Field Solid

PLANE223: 2D 8-Node Coupled-Field Solid
SOLID225: 3D 8-Node Coupled-Field Solid
SOLID226: 3D 20-Node Coupled-Field Solid
SOLID227: 3D 10-Node Coupled-Field Solid

These elements support electromigration (mass transport of charged particles due to an electric field)
when used in coupled-field analyses having electrical and diffusion DOFs.

Constitutive Equations
In a coupled electric-diffusion analysis, the diffusion flux density {J} is coupled to the electric potential
as follows:
(10.123)

where:

= diffusivity matrix

Dxx, Dyy, and Dzz = diffusivity coefficients in the element's X, Y, and Z directions, respectively
(input as DXX, DYY, DZZ on MP command)
C = concentration =
= normalized concentration (input/output as CONC)
= saturated concentration (input as MP,CSAT)
V = electric potential (input/output as VOLT)
T = absolute temperature = Tc + Toff
Tc = current temperature (input/output as TEMP on D or BF commands)
Toff = offset temperature from absolute zero to zero (input on TOFFST command)
Ze/k = charge constant (input as C4 on TBDATA command with TB,MIGR)
Z = charge number
e = elementary charge

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
360 Published: 2023-10-23
 Magnetoelasticity

k = Boltzmann constant

For more information on Equation 10.123 (p. 360) and related material constant input, see Migration
Model in the Material Reference.

For more information on diffusion analysis, see Diffusion (p. 305).

Derivation of Electric-Diffusion Matrices

Applying the variational principle with respect to electrical and diffusion DOFs to the electric charge
(Equation 5.5 (p. 164)) and mass (Equation 9.6 (p. 306)) continuity equations coupled by the constitutive
equation (Equation 10.123 (p. 360)), we obtain the following finite element matrix equation for the
electric-diffusion analysis:

(10.124)

where:

{V} = nodal electric scalar potential (input/output as VOLT)

{C} = nodal concentration (input/output as CONC)
[Cv] = element electric damping matrix (defined by Equation 5.115 (p. 182))
[Cd] = element diffusion damping matrix (defined by Equation 9.9 (p. 307))
[Kv] = element electric conductivity matrix (defined by Equation 5.115 (p. 182))
[Kd] = element diffusion conductivity matrix (defined by Equation 9.9 (p. 307))

= element transport conductivity matrix

= nonlinear part of the element diffusion

conductivity matrix
{E} = electric field (output as EF)
{I} = nodal current vector (input/output as AMPS)
{R} = nodal diffusion flow rate vector (input/output as RATE)
{N} = element shape functions

The finite element equation Equation 10.124 (p. 361) is unsymmetric. If a symmetric equation is desirable,
the electromigration coupling can be applied as a load vector by setting KEYOPT(2)
= 1 for the coupled-field elements (PLANE222, PLANE223, SOLID225, SOLID226, SOLID227).

10.14. Magnetoelasticity
The capability of modeling magnetic force coupling in current carrying solid conductors and solid
magnetic materials exists in the following current-technology elements:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 361
Coupling

PLANE223: 2D 8-Node Coupled-Field Solid

SOLID226: 3D 20-Node Coupled-Field Solid
SOLID227: 3D 10-Node Coupled-Field Solid

These elements use strong (matrix) coupling with volumetric integral Maxwell (Equation 5.140 (p. 188))
or Lorentz force calculation methods.

In addition to the above elements, the following legacy elements support weak (load vector) coupling
with surface integral Maxwell (Equation 5.136 (p. 187)) force for ferromagnetic regions and Lorentz force
for current carrying conductors:

SOLID5: 3D 8-Node Coupled-Field Solid

PLANE13: 2D 4-Node Coupled-Field Solid
SOLID98: 3D 10-Node Coupled-Field Solid

This discussion addresses strong magnetic force coupling available with the PLANE223, SOLID226, and
SOLID227 elements.

Magnetic solids and current carrying solid conductors deform when subject to a magnetic field. The
magnetic body force {Fm} that causes the deformation can be derived from the Maxwell stress tensor
[σM] (Landau and Lifshitz([355] (p. 945))):

(10.125)

where:

{H} = magnetic field intensity vector

{B} = magnetic flux density vector

As follows:
(10.126)

where:

[B] = strain-displacement matrix (see Equation 2.44 (p. 12))

vol = element volume

Applying the variational principle to the stress equation of motion and to the electromagnetic Equa-
tion 5.1 (p. 163) (Ampere’s law) coupled by magnetic force produces the following finite element equation:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
362 Published: 2023-10-23
 Magnetoelasticity

(10.127)

where:

{u} = structural displacements (UX, UY, also UZ)

{A} = magnetic vector potential (2D) or edge-flux (3D) DOF (AZ)
[K] = element structural stiffness matrix (see [Ke] in Equation 2.58 (p. 14))

[M] = element mass matrix (see [Me] in Equation 2.58 (p. 14))

[KAA] = element magnetic reluctivity matrix (see [KAA] in Equation 5.112 (p. 180))
[C] = element structural damping matrix (discussed in Damping Matrices (p. 694))
{F} = vector of nodal and surface forces (defined by Equation 2.56 (p. 14) and Equa-
tion 2.58 (p. 14))
{JS} = element source current density vector (see {JS} in Equation 5.112 (p. 180)); applies only to
structural-magnetic analysis (KEYOPT(1) = 10001)
{Jpm} = element remnant magnetization load vector (see {Jpm} in Equation 5.112 (p. 180))

The Maxwell force coupling method (KEYOPT(8) = 0) is applicable to both the magnetic solids and
current carrying conductors. In the absence of other coupled-field effects that may affect the symmetry
of Equation 10.127 (p. 363), the resulting system of magneto-structural equations coupled by the Maxwell
force matrices is symmetric for KEYOPT(4) = 0, 1, and 2.

In the case of a current carrying conductor, structural and magnetic equations can be coupled by the
Lorentz magnetic body force calculated from the electric current density {J} and the magnetic flux
density {B} as follows:
(10.128)

where:

{Nu} = vector of shape functions for {u}.

The system of structural and magnetic equations coupled by the Lorentz force coupling matrix is as
follows:

(10.129)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 363
Coupling

{NA} = vector of shape functions for {A}

The Lorentz force coupling method (KEYOPT(8) = 1) is applicable to current carrying conductors only
and produces an unsymmetric system of equations (Equation 10.129 (p. 363)).

The finite element equations for structural-electromagnetic (KEYOPT(1) = 10101) and structural-stranded
coil (KEYOPT(1) = 10201) analyses are the combinations of Equation 10.127 (p. 363) or Equa-
tion 10.129 (p. 363) with Equation 5.112 (p. 180) and Equation 5.162 (p. 194), respectively.

Not that for a conducting solid (KEYOPT(1) = 10101), the electric current density {J} consists of the DC
{Js}, eddy {Je}, and velocity {Jv} current components (Equation 5.123 (p. 184)):

(10.130)

where:

[σ] = electrical conductivity matrix (inverse of the electrical resistivity matrix input as RSVX,
RSVY, RSVZ on the MP command)
V = electric scalar potential (VOLT)

{v} = velocity vector =

Although the eddy and velocity current contributions to the magnetic force acting on a conductor were
illustrated using the Lorentz force method, these effects are included with the Maxwell force method
as well when it is applied to current carrying conductors.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
364 Published: 2023-10-23
Chapter 11: Shape Functions
This chapter provides the shape functions for elements. The shape functions are referred to by the in-
dividual element descriptions in Element Library (p. 445). All subheadings for this chapter are included
in the table of contents to aid in finding a specific type of shape function.

The following shape function topics are available:

11.1. Understanding Shape Function Labels
11.2. 2D Lines
11.3. 3D Lines
11.4. Axisymmetric Shells
11.5. Axisymmetric Harmonic Shells and General Axisymmetric Surfaces
11.6. 3D Shells
11.7. 2D and Axisymmetric Solids
11.8. Axisymmetric Harmonic Solids
11.9. 3D Solids
11.10. Electromagnetic Tangential Vector Elements

11.1. Understanding Shape Function Labels

The given functions are related to the nodal quantities by:

Table 11.1: Shape Function Labels

Input/Output
Variable Meaning
Label
u UX Translation in the x (or s) direction
v UY Translation in the y (or t) direction
w UZ Translation in the z (or r) direction
θx ROTX Rotation about the x direction
θy ROTY Rotation about the y direction
θz ROTZ Rotation about the z direction
Az AZ Z-component of vector magnetic potential
C CONC Concentration
P PRES Pressure
T TEMP, Temperature
TBOT,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 365
Shape Functions

Input/Output
Variable Meaning
Label
TE2, ...
TTOP
V VOLT Electric potential or source current
φ MAG Scalar magnetic potential

The vector correspondences are not exact, since, for example, u, v, and w are in the element coordinate
system, whereas UX, UY, UZ represent motions in the nodal coordinate system. Generally, the element
coordinate system is the same as the global Cartesian system, except for:

1. Line elements (2D Lines (p. 366) to Axisymmetric Harmonic Shells and General Axisymmetric Sur-
faces (p. 369)), where u motions are axial motions, and v and w are transverse motions.

2. Shell elements (3D Shells (p. 372)), where u and v are in-plane motions and w is the out-of-plane
motion.

Subscripted variables such as uJ refer to the u motion at node J. When these same variables have
numbers for subscripts (e.g. u1), nodeless variables for extra shape functions are being referred to. Co-
ordinates s, t, and r are normalized, going from -1.0 on one side of the element to +1.0 on the other,
and are not necessarily orthogonal to one another. L1, L2, L3, and L4 are also normalized coordinates,
going from 1.0 at a vertex to 0.0 at the opposite side or face.

Elements with midside nodes allow those midside nodes to be dropped in most cases. A dropped
midside node implies that the edge is and remains straight, and that any other effects vary linearly
along that edge.

Gaps are left in the equation numbering to allow for additions. Labels given in subsection titles within
parentheses are used to relate the given shape functions to their popular names, where applicable.

Some elements in Element Library (p. 445) (notably the 8-node solids) imply that reduced element
geometries (for example, wedge) are not available. However, the tables in Element Library (p. 445) refer
only to the available shape functions. In other words, the shape functions used for the 8-node brick is
the same as the 6-node wedge.

11.2. 2D Lines
This section contains shape functions for line elements without and with rotational degrees of freedom
(RDOF).

Figure 11.1: 2D Line Element

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
366 Published: 2023-10-23
 3D Lines

11.2.1. 2D Lines Without RDOF

These shape functions are for 2D line elements without RDOF.
(11.1)

(11.2)

(11.3)

11.2.2. 2D Lines With RDOF

These shape functions are for 2D line elements with RDOF.
(11.4)

(11.5)

11.3. 3D Lines
This section contains shape functions for various line elements.

Figure 11.2: 3–D Line Element

11.3.1. 3D 2-Node Lines (Not Combining Translations and Rotations)

These shape functions are for 3D 2-node line elements without RDOF, such as LINK33, LINK68, BEAM188,
or LINK228.
(11.6)

(11.7)

(11.8)

(11.9)

(11.10)

(11.11)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 367
Shape Functions

(11.12)

(11.13)

(11.14)

11.3.2. 3D 2-Node Lines (Combining Translations and Rotations)

These shape functions are for 3D 2-node line elements with RDOF.
(11.15)

(11.16)

(11.17)

(11.18)

11.3.3. 3D 3-Node Lines

These shape functions are for 3D 3-node line elements such as BEAM188 and BEAM189.
(11.19)

(11.20)

(11.21)

(11.22)

(11.23)

(11.24)

(11.25)

(11.26)

11.3.4. 3D 4-Node Lines

These shape functions are for 3D 4-node line elements such as BEAM188.

(11.27)

(11.28)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
368 Published: 2023-10-23
 Axisymmetric Harmonic Shells and General Axisymmetric Surfaces

(11.29)

(11.30)

(11.31)

(11.32)

11.4. Axisymmetric Shells

This section contains shape functions for 2-node axisymmetric shell elements under axisymmetric load.
These elements may have extra shape functions (ESF).

11.4.1. Axisymmetric Shell Without ESF

These shape functions are for 2-node axisymmetric shell elements without extra shape functions,
such as SHELL61.
(11.33)

(11.34)

(11.35)

11.5. Axisymmetric Harmonic Shells and General Axisymmetric Surfaces

This section describes shape functions for axisymmetric shell elements under nonaxisymmetric load
and for general axisymmetric surfaces.
11.5.1. Axisymmetric Harmonic Shells
11.5.2. General Axisymmetric Surfaces

11.5.1. Axisymmetric Harmonic Shells

This section describes shape functions for 2-node axisymmetric shell elements under nonaxisymmetric
(harmonic) load. These elements may or may not have extra shape functions (ESF).
[Link]. Axisymmetric Harmonic Shells Without ESF
[Link]. Axisymmetric Harmonic Shells With ESF

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 369
Shape Functions

Figure 11.3: Axisymmetric Harmonic Shell Element

The shape functions for axisymmetric harmonic shells use the quantities sin β and cos β, where =
input quantity MODE on the MODE command. The sin β and cos β are interchanged if Is = -1, where
Is = input quantity ISYM on the MODE command. If = 0, both sin β and cos β are set equal to 1.0.

[Link]. Axisymmetric Harmonic Shells Without ESF

These shape functions are for 2-node axisymmetric harmonic shell elements without extra shape
functions, such as SHELL61 with KEYOPT(3) = 1.
(11.36)

(11.37)

(11.38)

[Link]. Axisymmetric Harmonic Shells With ESF

These shape functions are for 2-node axisymmetric harmonic shell elements with extra shape
functions, such as SHELL61 with KEYOPT(3) = 0.

(11.39)

(11.40)

(11.41)

11.5.2. General Axisymmetric Surfaces

This section contains shape functions for 2- or 3- node-per-plane general axisymmetric surface elements
such as SURF159. These elements are available in various configurations, including combinations of
the following features:

• With or without midside nodes

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
370 Published: 2023-10-23
 Axisymmetric Harmonic Shells and General Axisymmetric Surfaces

• A varying number of node planes (Nnp) in the circumferential direction (defined via KEYOPT(2)).

The elemental coordinates are cylindrical coordinates and displacements are defined and interpolated
in that coordinate system, as shown in Figure 11.4: General Axisymmetric Surface Elements (when
Nnp = 3) (p. 371).

Figure 11.4: General Axisymmetric Surface Elements (when Nnp = 3)

Z

J2

J1
K2

J3

I2
y
z
K3
x

I3

When Nnp is an odd number, the interpolation function used for displacement is:

(11.42)

where:

i = r, θ, z
hi (s, t) = regular Lagrangian polynominal interpolation functions like Equation 11.6 (p. 367)
or Equation 11.19 (p. 368).
= coefficients for the Fourier terms.

When Nnp is an even number, the interpolation function is:

(11.43)

All of the coefficients in Equation 11.42 (p. 371) and Equation 11.43 (p. 371) can be expressed by nodal
displacements, using ur = u, uj = v, uz = w without midside nodes, and Nnp = 3:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 371
Shape Functions

(11.44)

(11.45)

(11.46)

Similar to the element without midside nodes, the u, v, and w with midside nodes are expressed as:

(11.47)

(11.48)

(11.49)

11.6. 3D Shells
This section contains shape functions for 3D shell elements. These elements are available in a number
of configurations, including certain combinations of the following features:

• triangular or quadrilateral.

- if quadrilateral, with or without extra shape functions (ESF).

• with or without rotational degrees of freedom (RDOF).

- if with RDOF, with or without shear deflections (SD).

• with or without midside nodes.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
372 Published: 2023-10-23
 3D Shells

Figure 11.5: 3D Shell Elements

11.6.1. 3D 3-Node Triangular Shells Without RDOF (Constant Strain Triangle

Elements [CST])
These shape functions are for 3D 3-node triangular shell elements without RDOF, such as SHELL131
or SHELL132:
(11.50)
(11.51)
(11.52)
(11.53)
(11.54)
(11.55)
(11.56)
(11.57)

11.6.2. 3D 6-Node Triangular Shells Without RDOF (Linear Strain Triangle

Elements [LST])
These shape functions are for 3D 6-node triangular shell elements without RDOF, such as SHELL281
when used as a triangle:

(11.58)

(11.59)
(11.60)
(11.61)
(11.62)
(11.63)

11.6.3. 3D 3-Node Triangular Shells With RDOF but Without SD

These shape functions are for the 3D 3-node triangular shell elements with RDOF, but without shear
deflection:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 373
Shape Functions

(11.64)
(11.65)
(11.66)
(11.67)
(11.68)
(11.69)

11.6.4. 3D 4-Node Quadrilateral Shells Without RDOF and Without ESF (Q4)
These shape functions are for 3D 4-node quadrilateral shell elements without RDOF and without extra
displacement shapes:

(11.70)

(11.71)

(11.72)

(11.73)

(11.74)

(11.75)

(11.76)

(11.77)

(11.78)

(11.79)

11.6.5. 3D 4-Node Quadrilateral Shells Without RDOF but With ESF (Quadri-
lateral Modified 6-Node Elements [QM6])
These shape functions are for 3D 4-node quadrilateral shell elements without RDOF but with extra
shape functions:

(11.80)

(11.81)

11.6.6. 3D 8-Node Quadrilateral Shells without RDOF

These shape functions are for 3D 8-node quadrilateral shell elements without RDOF, such as SHELL281:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
374 Published: 2023-10-23
 3D Shells

(11.82)

(11.83)

(11.84)

(11.85)

(11.86)

(11.87)

(11.88)

11.6.7. 3D 4-Node Quadrilateral Shells With RDOF but Without SD and

Without ESF
These shape functions are for 3D 4-node quadrilateral shell elements with RDOF but without shear
deflection and without extra shape functions:

(11.89)

(11.90)
(11.91)
(11.92)

(11.93)

(11.94)

11.6.8. 3D 4-Node Quadrilateral Shells With RDOF but Without SD and With
ESF
These shape functions are for 3D 4-node quadrilateral shell elements with RDOF but without shear
deflection and with extra shape functions:

(11.95)

(11.96)
(11.97)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 375
Shape Functions

(11.98)

(11.99)

(11.100)

11.7. 2D and Axisymmetric Solids

This section contains shape functions for 2D and axisymmetric solid elements. These elements are
available in a number of configurations, including certain combinations of the following features:

• Triangular

• Quadrilateral (with or without extra shape functions [ESF])

• With or without midside nodes

Figure 11.6: 2D and Axisymmetric Solid Element

11.7.1. 2D and Axisymmetric 3-Node Triangular Solids (CST)

These shape functions are for 2D 3-node and axisymmetric triangular solid elements, such as PLANE13,
PLANE182, or PLANE222 with only 3 nodes input:
(11.101)
(11.102)
(11.103)
(11.104)
(11.105)
(11.106)
(11.107)
(11.108)
(11.109)
(11.110)
(11.111)
(11.112)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
376 Published: 2023-10-23
 2D and Axisymmetric Solids

(11.113)

11.7.2. 2D and Axisymmetric 6-Node Triangular Solids (LST)

These shape functions are for 2D 6-node and axisymmetric triangular solids, such as PLANE35:

(11.114)

(11.115)
(11.116)
(11.117)
(11.118)
(11.119)
(11.120)
(11.121)

11.7.3. 2D and Axisymmetric 4-node Quadrilateral Solid Without ESF (Q4)

These shape functions are for the 2D 4-node and axisymmetric quadrilateral solid elements without
extra shape functions, such as PLANE13 with KEYOPT(2) = 1 or PLANE222 with KEYOPT(11) = 0.

(11.122)

(11.123)

(11.124)

(11.125)

(11.126)

(11.127)

(11.128)

(11.129)

(11.130)

(11.131)

(11.132)

(11.133)

(11.134)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 377
Shape Functions

11.7.4. 2D and Axisymmetric 4-node Quadrilateral Solids With ESF (QM6)

These shape functions are for the 2D 4-node and axisymmetric solid elements with extra shape
functions, such as PLANE13 with KEYOPT(2) = 0. (Taylor et al.([50] (p. 927)))

(11.135)

(11.136)

Equation 11.135 (p. 378) is adjusted for axisymmetric situations by removing the u1 or u2 term for
elements near the centerline, in order to avoid holes or "doubled" material at the centerline.

11.7.5. 2D and Axisymmetric 8-Node Quadrilateral Solids (Q8)

These shape functions are for the 2D 8-node and axisymmetric quadrilateral elements such as
PLANE293, PLANE77, and PLANE183:

(11.137)

(11.138)

(11.139)

(11.140)

(11.141)

(11.142)

(11.143)

11.7.6. 2D and Axisymmetric 4-Node Quadrilateral Infinite Solids

Figure 11.7: 4-Node Quadrilateral Infinite Solid Element

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
378 Published: 2023-10-23
 2D and Axisymmetric Solids

These Lagrangian isoparametric shape functions and "mapping" functions are for the 2D and
axisymmetric 4-node quadrilateral solid infinite elements such as INFIN110:

[Link]. Shape Functions

(11.144)

(11.145)

(11.146)

[Link]. Mapping Functions

(11.147)

(11.148)
(11.149)

11.7.7. 2D and Axisymmetric 8-Node Quadrilateral Infinite Solids

Figure 11.8: 8-Node Quadrilateral Infinite Solid Element

These Lagrangian isoparametric shape functions and "mapping" functions are for the 2D and
axisymmetric 8-node quadrilateral infinite solid elements such as INFIN110:

[Link]. Shape Functions

(11.150)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 379
Shape Functions

(11.151)

(11.152)

(11.153)

[Link]. Mapping Functions

(11.154)

(11.155)

The shape and mapping functions for the nodes N, O and P are deliberately set to zero.

11.8. Axisymmetric Harmonic Solids

This section contains shape functions for axisymmetric harmonic solid elements. These elements are
available in a number of configurations, including certain combinations of the following features:

• triangular or quadrilateral.

- if quadrilateral, with or without extra shape functions (ESF).

• with or without midside nodes.

The shape functions of this section use the quantities sin β and cos β (where = input as MODE on
the MODE command). sin β and cos β are interchanged if Is = -1 (where Is = input as ISYM on the
MODE command). If = 0, sin β = cos β = 1.0.

Figure 11.9: Axisymmetric Harmonic Solid Elements

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
380 Published: 2023-10-23
 Axisymmetric Harmonic Solids

11.8.1. Axisymmetric Harmonic 3-Node Triangular Solids

These shape functions are for the 3-node axisymmetric triangular solid elements, such as PLANE25
with only 3 nodes input:
(11.156)
(11.157)
(11.158)
(11.159)

11.8.2. Axisymmetric Harmonic 6-Node Triangular Solids

These shape functions are for the 6-node axisymmetric triangular solids elements, such as PLANE83
input as a triangle:

(11.160)

(11.161)
(11.162)
(11.163)

11.8.3. Axisymmetric Harmonic 4-Node Quadrilateral Solids Without ESF

These shape functions are for the 4-node axisymmetric harmonic quadrilateral solid elements without
extra shape functions, such as PLANE25 with KEYOPT(2) = 1, or PLANE75:

(11.164)

(11.165)

(11.166)

(11.167)

11.8.4. Axisymmetric Harmonic 4-Node Quadrilateral Solids With ESF

These shape functions are for the 4-node axisymmetric harmonic quadrilateral elements with extra
shape functions, such as PLANE25 with KEYOPT(2) = 0.

(11.168)

(11.169)

(11.170)

Unless (MODE) = 1, u1 or u2 and w1 or w2 motions are suppressed for elements near the centerline.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 381
Shape Functions

11.8.5. Axisymmetric Harmonic 8-Node Quadrilateral Solids

These shape functions are for the 8-node axisymmetric harmonic quadrilateral solid elements such
as PLANE78 or PLANE83.

(11.171)

(11.172)

(11.173)

(11.174)

11.9. 3D Solids
This section contains shape functions for 3D solid elements. These elements are available in a number
of configurations, including certain combinations of the following features:

• Element shapes may be tetrahedra, pyramids, wedges, or bricks (hexahedra).

- If wedges or bricks, with or without extra shape functions (ESF)

• With or without rotational degrees of freedom (RDOF)

• With or without midside nodes

The wedge elements with midside nodes (15-node wedges) are either a condensation of the 20-node
brick element or are based on wedge shape functions.

11.9.1. 4-Node Tetrahedra

These shape functions are either a direct 4-node tetrahedral such as SOLID285 or a condensation of
an 8-node brick element such as SOLID5, FLUID30, or SOLID98.

Figure 11.10: 3D Solid Elements

The resulting effective shape functions are:

(11.175)
(11.176)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
382 Published: 2023-10-23
 3D Solids

(11.177)
(11.178)
(11.179)
(11.180)
(11.181)
(11.182)
(11.183)
(11.184)
(11.185)
(11.186)

11.9.2. 10-Node Tetrahedra

These shape functions are for 10-node tetrahedron elements such as SOLID98 and SOLID227, or by
condensation for SOLID90.

Figure 11.11: 10-Node Tetrahedra Element

(11.187)

(11.188)
(11.189)
(11.190)
(11.191)
(11.192)
(11.193)

11.9.3. 5-Node Pyramids

This element is a condensation of an 8-node brick element.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 383
Shape Functions

Figure 11.12: 8-Node Brick Element

The resulting effective shape functions are:

(11.194)

11.9.4. 13-Node Pyramids

These shape functions are for 13-node pyramid elements which are based on a condensation of a
20-node brick element:

Figure 11.13: 13-Node Pyramid Element

(11.195)

(11.196)

(11.197)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
384 Published: 2023-10-23
 3D Solids

(11.198)

(11.199)

(11.200)

11.9.5. 6-Node Wedges Without ESF

Figure 11.14: 6-Node Wedge Element

The 6-node wedge elements are a condensation of an 8-node brick such as SOLID5 or FLUID30. These
shape functions are for 6-node wedge elements without extra shape functions:

(11.201)

(11.202)

(11.203)

(11.204)

(11.205)

(11.206)

(11.207)

11.9.6. 6-Node Wedges With ESF

The 6-node wedge elements are a condensation of an 8-node brick such as SOLID5. (See Figure 11.14: 6-
Node Wedge Element (p. 385).) These shape functions are for 6-node wedge elements with extra shape
functions:

(11.208)

(11.209)

(11.210)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 385
Shape Functions

11.9.7. 15-Node Wedges

These shape functions are for 15-node wedge elements such as SOLID90 that are based on a condens-
ation of a 20-node brick element Equation 11.236 (p. 388). or are computed directly.

Figure 11.15: 15-Node Wedge Element

Elements in a wedge configuration use shape functions based on triangular coordinates and the r
coordinate going from -1.0 to +1.0.

(11.211)

(11.212)

(11.213)

(11.214)

(11.215)

(11.216)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
386 Published: 2023-10-23
 3D Solids

11.9.8. 8-Node Bricks Without ESF

Figure 11.16: 8-Node Brick Element

These shape functions are for 8-node brick elements without extra shape functions such as SOLID5
with KEYOPT(3) = 1, FLUID30, or SOLID225 with KEYOPT(11) = 0:

(11.217)

(11.218)

(11.219)

(11.220)

(11.221)

(11.222)

(11.223)

(11.224)

(11.225)

(11.226)

(11.227)

(11.228)

(11.229)

(11.230)

(11.231)

(11.232)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 387
Shape Functions

11.9.9. 8-Node Bricks With ESF

(Please see Figure 11.16: 8-Node Brick Element (p. 387)) These shape functions are for 8-node brick
elements with extra shape functions such as SOLID5 with KEYOPT(3) = 0:

(11.233)

(11.234)

(11.235)

11.9.10. 20-Node Bricks

Figure 11.17: 20-Node Brick Element

These shape functions are used for 20-node solid elements such as SOLID90:

(11.236)

(11.237)

(11.238)

(11.239)

(11.240)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
388 Published: 2023-10-23
 3D Solids

(11.241)

(11.242)

11.9.11. 8-Node Infinite Bricks

Figure 11.18: 3D 8-Node Brick Element

These shape functions and mapping functions are for the 3D 8-node solid brick infinite elements such
as INFIN111:

[Link]. Shape Functions

(11.243)

(11.244)

(11.245)

(11.246)

(11.247)

(11.248)

(11.249)

(11.250)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 389
Shape Functions

[Link]. Mapping Functions

(11.251)

(11.252)

(11.253)

(11.254)

11.9.12. 20-Node Infinite Bricks

Figure 11.19: 20-Node Solid Brick Infinite Element

These shape functions and mapping functions are for the 3D 20-node solid brick infinite elements
such as INFIN111:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
390 Published: 2023-10-23
 3D Solids

[Link]. Shape Functions

(11.255)

(11.256)

(11.257)

(11.258)

(11.259)

(11.260)

(11.261)

(11.262)

(11.263)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 391
Shape Functions

[Link]. Mapping Functions

(11.264)

(11.265)

(11.266)

The shape and mapping functions for the nodes U, V, W, X, Y, Z, A, and B are deliberately set to
zero.

11.9.13. General Axisymmetric Solids

This section contains shape functions for general axisymmetric solid elements. These elements are
available in a number of configurations, including certain combinations of the following features:

• A quadrilateral, or a degenerated triangle shape to simulate an irregular area, on the master

plane (the plane on which the quadrilaterals or triangles are defined)

• With or without midside nodes

• A varying number of node planes (Nnp) in the circumferential direction (defined via KEYOPT(2))

The elemental coordinates are cylindrical coordinates and displacements are defined and interpolated
in that coordinate system, as shown in Figure 11.20: General Axisymmetric Solid Elements (when Nnp
= 3) (p. 393).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
392 Published: 2023-10-23
 3D Solids

Figure 11.20: General Axisymmetric Solid Elements (when Nnp = 3)

When Nnp is an odd number (except , the interpolation function used for displacement is:

(11.267)

where:

i = r, θ, z
hi (s, t) = regular Lagrangian polynominal interpolation functions like Equation 11.122 (p. 377)
or Equation 11.137 (p. 378).
= coefficients for the Fourier terms.

When Nnp is an even number other than 4, the interpolation function is:

(11.268)

When Nnp = 4, the interpolation function is:

(11.269)

The temperatures are interpolated by Lagrangian polynominal interpolations in s, t plane, and linearly
interpolated with θ in circumferential (θ) direction as:

(11.270)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 393
Shape Functions

= node plane number in circumferential direction

Tn = same as Equation 11.130 (p. 377) and Equation 11.141 (p. 378).

[Link]. General Axisymmetric Solid With 4 Base Nodes

All of the coefficients in Equation 11.267 (p. 393) and Equation 11.268 (p. 393) can be expressed by
node displacements. Using ur = u, uj = v, uz = w, and take Nnp = 3 as an example.

(11.271)

(11.272)

(11.273)

[Link]. General Axisymmetric Solid With 3 Base Nodes

(11.274)

(11.275)
(11.276)

[Link]. General Axisymmetric Solid With 8 Base Nodes

Similar to the element with 4 base nodes, the u, v, and w are expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
394 Published: 2023-10-23
 Electromagnetic Tangential Vector Elements

(11.277)

(11.278)

(11.279)

[Link]. General Axisymmetric Solid With 6 Base Nodes

(11.280)

(11.281)
(11.282)

11.10. Electromagnetic Tangential Vector Elements

In electromagnetics, we encounter serious problems when node-based elements are used to represent
vector electric or magnetic fields. (See Electric Scalar Potential (p. 173).) Node-based elements require
special treatment for enforcing boundary conditions of electromagnetic field at material interfaces,
conducting surfaces and geometric corners. Tangentially continuous vector elements or edge elements,
whose degrees of freedom are associated with the edges of the finite element mesh, have been shown
to be free of such shortcomings. ([410] (p. 948))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 395
Shape Functions

11.10.1. Tetrahedral Elements

The tetrahedral element is the simplest tessellated shape and is able to model arbitrary 3D geometric
structures. It is also well suited for automatic mesh generation. The tetrahedral element, by far, is the
most popular element shape for 3D applications in FEA.

For the 1st-order tetrahedral element (KEYOPT(1) = 1), the degrees of freedom (DOF) are at the edges
of element that is, (DOFs = 6) (Figure 11.21: 1st-Order Tetrahedral Element (p. 396)). In terms of volume
coordinates, the vector basis functions are defined as:
(11.283)
(11.284)
(11.285)
(11.286)
(11.287)
(11.288)

where:

hIJ = edge length between node I and J

λI, λJ, λK, λL = volume coordinates (λK = 1 - λI - λJ - λL)
λI, λJ, λK, λL = the gradient of volume coordinates

Figure 11.21: 1st-Order Tetrahedral Element

The tangential component of the approximated field is constant along the edge. The normal compon-
ent of field varies linearly.

11.10.2. Hexahedral Elements

Tangential vector bases for hexahedral elements can be derived by carrying out the transformation
mapping a hexahedral element in the global xyz coordinate to a brick element in local str coordinates.

For the 1st-order brick element (KEYOPT(1) = 1), the degrees of freedom (DOF) are at the edges of
element (DOFs = 12) (Figure 11.22: 1st-Order Brick Element (p. 397)). The vector basis functions are
cast in the local coordinate
(11.289)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
396 Published: 2023-10-23
 Electromagnetic Tangential Vector Elements

(11.290)

(11.291)

where:

hs, ht, hr = length of element edge

s, t, r = gradient of local coordinates

Figure 11.22: 1st-Order Brick Element

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 397
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
398 Published: 2023-10-23
Chapter 12: Element Tools
The following element tools are available:
12.1. Element Shape Testing
12.2. Integration Point Locations
[Link]-Dependent Material Properties
12.4. Positive Definite Matrices
12.5. Lumped Matrices
12.6. Reuse of Matrices
12.7. Hydrostatic Loads
12.8. Hydrodynamic Loads
12.9. Nodal and Centroidal Data Evaluation

12.1. Element Shape Testing

12.1.1. Overview
All continuum elements (2D and 3D solids, 3D shells) are tested for acceptable shape as they are
defined by the E, EGEN, AMESH, VMESH, or similar commands. This testing, described in the following
sections, is performed by computing shape parameters (such as Jacobian ratio) which are functions
of geometry, then comparing them to element shape limits whose default values are functions of
element type and settings (but can be modified by the user on the SHPP command with Lab =
MODIFY as described below). When defined, an element may generate no warnings, one or more
warnings, or the element may be rejected with an error.

12.1.2. 3D Solid Element Faces and Cross-Sections

Some shape testing of 3D solid elements (bricks [hexahedra], wedges, pyramids, and tetrahedra) is
performed indirectly. Aspect ratio, parallel deviation, and maximum corner angle are computed for
3D solid elements using the following steps:

1. Each of these 3 quantities is computed, as applicable, for each face of the element as though it
were a quadrilateral or triangle in 3D space, by the methods described in sections Aspect Ra-
tio (p. 402), Parallel Deviation (p. 404), and Maximum Corner Angle (p. 406).

2. Because some types of 3D solid element distortion are not revealed by examination of the faces,
cross-sections through the solid are constructed. Then, each of the 3 quantities is computed, as
applicable, for each cross-section as though it were a quadrilateral or triangle in 3D space.

3. The metric for the element is assigned as the worst value computed for any face or cross-section.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 399
Element Tools

A brick element has 6 quadrilateral faces and 3 quadrilateral cross-sections (Figure 12.1: Brick Ele-
ment (p. 400)). The cross-sections are connected to midside nodes, or to edge midpoints where midside
nodes are not defined.

Figure 12.1: Brick Element

A pyramid element has 1 quadrilateral face and 4 triangle faces, as shown on the left side of Fig-
ure 12.2: Pyramid Element (p. 401). The pyramid element has 8 triangle cross-sections, constructed in
4 pairs. One of the 4 pairs is shown on the right side of Figure 12.2: Pyramid Element (p. 401).

For each pair, the closest points (A and B in the figure) between the lines containing a quadrilateral
face edge and the right-ward opposite leg are found. The paired cross-sections are the triangle con-
taining the quadrilateral face edge and point B and the triangle containing the right-ward opposite
leg and point A. The remaining 6 cross-sections are found in a similar manner using the other 3
quadrilateral face edges and their corresponding right-ward legs.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
400 Published: 2023-10-23
 Element Shape Testing

Figure 12.2: Pyramid Element

A wedge element has 3 quadrilateral and 2 triangle faces, and has 3 quadrilateral and 1 triangle cross-
sections. As shown in Figure 12.3: Wedge Element (p. 401), the cross-sections are connected to midside
nodes, or to edge midpoints where midside nodes are not defined.

Figure 12.3: Wedge Element

A tetrahedron element has 4 triangle faces and 6 triangle cross-sections (Figure 12.4: Tetrahedron
Element (p. 402)). Each cross-section is the triangle containing one edge and the closest point on the
line containing the opposite edge. Two of these cross-sections are shown on the right side of Fig-
ure 12.4: Tetrahedron Element (p. 402).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 401
Element Tools

Figure 12.4: Tetrahedron Element

12.1.3. Aspect Ratio

An aspect ratio is computed and tested for all elements except Emag elements (see Table 12.1: Aspect
Ratio Limits (p. 404)). Aspect ratio as a shape measure has been reported in finite element literature
for decades (Robinson [122] (p. 931)) and is one of the easiest ones to understand. For triangles,
equilateral triangles have an aspect ratio of 1 and the aspect ratio increases the more the triangle is
stretched from an equilateral shape. For quadrilaterals, squares have an aspect ratio of 1 and the aspect
ratio increases the more the square is stretched to an oblong shape. Some analysts want to be warned
about high aspect ratio so they can verify that the creation of any stretched elements was intentional.
Many other analysts routinely ignore it.

Unless elements are so stretched that numeric round off could become a factor (aspect ratio > 1000),
aspect ratio alone has little correlation with analysis accuracy. Finite element meshes should be tailored
to the physics of the given problem; that is, fine in the direction of rapidly changing field gradients,
relatively coarse in directions with less rapidly changing fields. Sometimes this calls for elements
having aspect ratios of 10, 100, or in extreme cases 1000. (Examples include shell or thin coating
analyses using solid elements, thermal shock "skin" stress analyses, and fluid boundary layer analyses.)
Attempts to artificially restrict aspect ratio could compromise analysis quality in some cases.

12.1.4. Aspect Ratio Calculation for Triangles

For the benefit of computational efficiency, the aspect ratio of a triangle is computed based on the
angle measures at the corner nodes rather than the height and width of the triangle. For the calculation,
midside nodes are ignored and the triangle is treated as connecting nodes I, J, and K with straight
lines:

1. For the angle θ, at node I, a quantity AMI is calculated as follows:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
402 Published: 2023-10-23
 Element Shape Testing

The quantity AMI is 1 when θ = 60°. As θ deviates from 60° in either direction, AMI increases (to a
maximum of ∞ as θ nears 0° or 180°).

2. The quantity AM is calculated for each of the remaining nodes J and K.

3. A preliminary aspect ratio, ASPRAT, is calculated by averaging the three values of AM, then taking
the square root:

4. Finally, the aspect ratio is modified to produce a distribution of values similar to those that would
be produced if the aspect ratio were calculated directly from the height and width:

The best possible triangle aspect ratio, for an equilateral triangle, is 1. Larger aspect ratios are calculated
as the triangle is stretched. Triangles having aspect ratios of 1 and 20 are shown in Figure 12.5: Aspect
Ratios for Triangles (p. 403).

Figure 12.5: Aspect Ratios for Triangles

12.1.5. Aspect Ratio Calculation for Quadrilaterals

Figure 12.6: Quadrilateral Aspect Ratio Calculation

The aspect ratio for a quadrilateral is computed by the following steps, using only the corner nodes
of the element (Figure 12.6: Quadrilateral Aspect Ratio Calculation (p. 403)):

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 403
Element Tools

1. If the element is not flat, the nodes are projected onto a plane passing through the average of
the corner locations and perpendicular to the average of the corner normals. The remaining steps
are performed on these projected locations.

2. Two lines are constructed that bisect the opposing pairs of element edges and which meet at the
element center. In general, these lines are not perpendicular to each other or to any of the element
edges.

3. Rectangles are constructed centered about each of the 2 lines, with edges passing through the
element edge midpoints. The aspect ratio of the quadrilateral is the ratio of a longer side to a
shorter side of whichever rectangle is most stretched.

4. The best possible quadrilateral aspect ratio, for a square, is one. A quadrilateral having an aspect
ratio of 20 is shown in Figure 12.7: Aspect Ratios for Quadrilaterals (p. 404).

Figure 12.7: Aspect Ratios for Quadrilaterals

Table 12.1: Aspect Ratio Limits

Command to Type of Why default is this Why default is this

Default
modify Limit tight loose
SHPP,MODIFY,1 warning 20 Elements this Disturbance of
stretched look to analysis results
many users like they has not been
deserve warnings. proven

It is difficult to
avoid warnings
even with a
limit of 20.
SHPP,MODIFY,2 error 106 Informal testing has Threshold of
demonstrated round off
solution error problems
attributable to depends on
computer round off what computer
at aspect ratios of is being used.
1,000 to 100,000.
Valid analyses
should not be
blocked.

12.1.6. Parallel Deviation

Parallel deviation is computed and tested for all quadrilaterals or 3D solid elements having quadrilat-
eral faces or cross-sections, except Emag elements (see Table 12.2: Parallel Deviation Limits (p. 405)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
404 Published: 2023-10-23
 Element Shape Testing

Formal testing has demonstrated degradation of stress convergence in linear displacement quadrilat-
erals as opposite edges become less parallel to each other.

12.1.7. Parallel Deviation Calculation

Parallel deviation is computed using the following steps:

1. Ignoring midside nodes, unit vectors are constructed in 3D space along each element edge, ad-
justed for consistent direction, as demonstrated in Figure 12.8: Parallel Deviation Unit Vectors (p. 405).

Figure 12.8: Parallel Deviation Unit Vectors

2. For each pair of opposite edges, the dot product of the unit vectors is computed, then the angle
(in degrees) whose cosine is that dot product. The parallel deviation is the larger of these 2 angles.
(In the illustration above, the dot product of the 2 horizontal unit vectors is 1, and acos (1) = 0°.
The dot product of the 2 vertical vectors is 0.342, and acos (0.342) = 70°. Therefore, this element's
parallel deviation is 70°.)

3. The best possible deviation, for a flat rectangle, is 0°. Figure 12.9: Parallel Deviations for Quadrilat-
erals (p. 405) shows quadrilaterals having deviations of 0°, 70°, 100°, 150°, and 170°.

Figure 12.9: Parallel Deviations for Quadrilaterals

0 70 100

150 170

Table 12.2: Parallel Deviation Limits

Command to Type of Why default is this Why default is this

Default
Modify Limit tight loose
SHPP,MODIFY,11 warning for 70° Testing has shown It is difficult to avoid
elements results are degraded warnings even with
without by this much a limit of 70°
midside distortion
nodes
SHPP,MODIFY,12 error for 150° Pushing the limit Valid analyses
elements further does not should not be
without seem prudent blocked.
midside
nodes
SHPP,MODIFY,13 warning for 100° Elements having Disturbance of
elements deviations > 100° analysis results for

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 405
Element Tools

Command to Type of Why default is this Why default is this

Default
Modify Limit tight loose
with look like they quadratic elements
midside deserve warnings. has not been
nodes proven.
SHPP,MODIFY,14 error for 170° Pushing the limit Valid analyses
elements further does not should not be
with seem prudent blocked.
midside
nodes

12.1.8. Maximum Corner Angle

Maximum corner angle is computed and tested for all except Emag elements (see Table 12.3: Maximum
Corner Angle Limits (p. 406)). Some in the finite element community have reported that large angles
(approaching 180°) degrade element performance, while small angles don't.

12.1.9. Maximum Corner Angle Calculation

The maximum angle between adjacent edges is computed using corner node positions in 3D space.
(Midside nodes, if any, are ignored.) The best possible triangle maximum angle, for an equilateral tri-
angle, is 60°. Figure 12.10: Maximum Corner Angles for Triangles (p. 406) shows a triangle having a
maximum corner angle of 165°. The best possible quadrilateral maximum angle, for a flat rectangle,
is 90°. Figure 12.11: Maximum Corner Angles for Quadrilaterals (p. 406) shows quadrilaterals having
maximum corner angles of 90°, 140° and 180°.

Figure 12.10: Maximum Corner Angles for Triangles

Figure 12.11: Maximum Corner Angles for Quadrilaterals

Table 12.3: Maximum Corner Angle Limits

Command to Type of Why default is this Why default is this

Default
Modify Limit tight loose
SHPP,MODIFY,15 warnings for 165° Any element this Disturbance of
triangles distorted looks like analysis results
it deserves a has not been
warning. proven.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
406 Published: 2023-10-23
 Element Shape Testing

Command to Type of Why default is this Why default is this

Default
Modify Limit tight loose
It is difficult to
avoid warnings
even with a
limit of 165°.
SHPP,MODIFY,16 error for 179.9° We can not allow Valid analyses
triangles 180° should not be
blocked.
SHPP,MODIFY,17 warning for 155° Any element this Disturbance of
quadrilaterals distorted looks like analysis results
without it deserves a has not been
midside warning. proven.
nodes
It is difficult to
avoid warnings
even with a
limit of 155°.
SHPP,MODIFY,18 error for 179.9° We can not allow Valid analyses
quadrilaterals 180° should not be
without blocked.
midside
nodes
SHPP,MODIFY,19 warning for 165° Any element this Disturbance of
quadrilaterals distorted looks like analysis results
with it deserves a has not been
midside warning. proven.
nodes
It is difficult to
avoid warnings
even with a
limit of 165°.
SHPP,MODIFY,20 error for 179.9° We can not allow Valid analyses
quadrilaterals 180° should not be
with blocked.
midside
nodes

12.1.10. Jacobian Ratio

Jacobian ratio is computed and tested for all elements except triangles and tetrahedra that (a) are
linear (have no midside nodes) or (b) have perfectly centered midside nodes (see Table 12.4: Jacobian
Ratio Limits (p. 409)). A high ratio indicates that the mapping between element space and real space
is becoming computationally unreliable.

[Link]. Jacobian Ratio Calculation

An element's Jacobian ratio is computed by the following steps, using the full set of nodes for the
element:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 407
Element Tools

1. At each sampling location listed in the table below, the determinant of the Jacobian matrix is
computed and called RJ. RJ at a given point represents the magnitude of the mapping function
between element natural coordinates and real space. In an ideally-shaped element, RJ is relatively
constant over the element, and does not change sign.

Element Shape RJ Sampling Locations

10-node tetrahedra - corner nodes
SHPP,LSTET,OFF
10-node tetrahedra - integration points
SHPP,LSTET,ON
5-node or 13-node pyramids base corner nodes and near apex node (apex RJ
factored so that a pyramid having all edges the
same length will produce a Jacobian ratio of 1)
8-node quadrilaterals corner nodes and centroid
20-node bricks all nodes and centroid
all other elements corner nodes

2. The Jacobian ratio of the element is the ratio of the maximum to the minimum sampled value
of RJ. If the maximum and minimum have opposite signs, the Jacobian ratio is arbitrarily assigned
to be -100 (and the element is clearly unacceptable).

3. If the element is a midside-node tetrahedron, an additional RJ is computed for a fictitious straight-

sided tetrahedron connected to the 4 corner nodes. If that RJ differs in sign from any nodal RJ
(an extremely rare occurrence), the Jacobian ratio is arbitrarily assigned to be -100.

4. The sampling locations for midside-node tetrahedra depend upon the setting of the linear stress
tetrahedra key on the SHPP command. The default behavior (SHPP,LSTET,OFF) is to sample at
the corner nodes, while the optional behavior (SHPP,LSTET,ON) is to sample at the integration
points. Sampling at the integration points will result in a lower Jacobian ratio than sampling at
the nodes, but that ratio is compared to more restrictive default limits (see Table 12.4: Jacobian
Ratio Limits (p. 409) below). Nevertheless, some elements which pass the LSTET,ON test fail the
LSTET,OFF test - especially those having zero RJ at a corner node. Testing has shown that such
elements have no negative effect on linear elastic stress accuracy. Their effect on other types
of solutions has not been studied, which is why the more conservative test is recommended
for general usage. Brick elements (such as SOLID186) degenerated into tetrahedra are tested in
the same manner as are 'native' tetrahedra (SOLID187). In most cases, this produces conservative
results. However, for SOLID185 and SOLID186 when using the non-recommended tetrahedron
shape, it is possible that such a degenerate element may produce an error during solution, even
though it produced no warnings during shape testing.

5. If the element is a line element having a midside node, the Jacobian matrix is not square (because
the mapping is from one natural coordinate to 2D or 3D space) and has no determinant. For
this case, a vector calculation is used to compute a number which behaves like a Jacobian ratio.
This calculation has the effect of limiting the arc spanned by a single element to about 106°

A triangle or tetrahedron has a Jacobian ratio of 1 if each midside node, if any, is positioned at the
average of the corresponding corner node locations. This is true no matter how otherwise distorted
the element may be. Hence, this calculation is skipped entirely for such elements. Moving a midside

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
408 Published: 2023-10-23
 Element Shape Testing

node away from the edge midpoint position will increase the Jacobian ratio. Eventually, even very
slight further movement will break the element (Figure 12.12: Jacobian Ratios for Triangles (p. 409)).
We describe this as "breaking" the element because it suddenly changes from acceptable to unac-
ceptable- "broken".

Figure 12.12: Jacobian Ratios for Triangles

Any rectangle or rectangular parallelepiped having no midside nodes, or having midside nodes at
the midpoints of its edges, has a Jacobian ratio of 1. Moving midside nodes toward or away from
each other can increase the Jacobian ratio. Eventually, even very slight further movement will break
the element (Figure 12.13: Jacobian Ratios for Quadrilaterals (p. 409)).

Figure 12.13: Jacobian Ratios for Quadrilaterals

A quadrilateral or brick has a Jacobian ratio of 1 if (a) its opposing faces are all parallel to each
other, and (b) each midside node, if any, is positioned at the average of the corresponding corner
node locations. As a corner node moves near the center, the Jacobian ratio climbs. Eventually, any
further movement will break the element (Figure 12.14: Jacobian Ratios for Quadrilaterals (p. 409)).

Figure 12.14: Jacobian Ratios for Quadrilaterals

Table 12.4: Jacobian Ratio Limits

Command to Type of Default Why default is this Why default is this

modify limit tight loose
SHPP,MODIFY,31 warning for 30 if SHPP, A ratio this high Disturbance of
h-elements LSTET,OFF indicates that the analysis results has
10 if SHPP, mapping between not been proven. It
LSTET,ON element and real is difficult to avoid
space is becoming warnings even with
computationally a limit of 30.
unreliable.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 409
Element Tools

Command to Type of Default Why default is this Why default is this

modify limit tight loose
SHPP,MODIFY,32 error for 1,000 if Pushing the limit Valid analyses
h-elements SHPP, further does not should not be
LSTET,OFF seem prudent. blocked.
40 if SHPP,
LSTET,ON

12.1.11. Warping Factor

Warping factor is computed and tested for some quadrilateral shell elements, and the quadrilateral
faces of bricks, wedges, and pyramids. (See Warping Factor (p. 410).) A high factor may indicate a
condition the underlying element formulation cannot handle well, or may simply hint at a mesh
generation flaw.

The following warping factor topics are available:

[Link]. Warping Factor Calculation for Quadrilateral Shell Elements
[Link]. Warping Factor Calculation for 3D Solid Elements

Also see Table 12.5: Warping Factor Limits (p. 412).

[Link]. Warping Factor Calculation for Quadrilateral Shell Elements

A quadrilateral element's warping factor is computed from its corner node positions and other
available data by the following steps:

1. An average element normal is computed as the vector (cross) product of the 2 diagonals (Fig-
ure 12.15: Shell Average Normal Calculation (p. 410)).

Figure 12.15: Shell Average Normal Calculation

2. The projected area of the element is computed on a plane through the average normal (the
dotted outline on Figure 12.16: Shell Element Projected onto a Plane (p. 411)).

3. The difference in height of the ends of an element edge is computed, parallel to the average
normal. In Figure 12.16: Shell Element Projected onto a Plane (p. 411), this distance is 2h. Because
of the way the average normal is constructed, h is the same at all four corners. For a flat quad-
rilateral, the distance is zero.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
410 Published: 2023-10-23
 Element Shape Testing

Figure 12.16: Shell Element Projected onto a Plane

4. The "area warping factor" ( ) for the element is computed as the edge height difference divided
by the square root of the projected area.

5. For all shells except those in the "membrane stiffness only" group, if the thickness is available,
the "thickness warping factor" is computed as the edge height difference divided by the average
element thickness. This could be substantially higher than the area warping factor computed
in 4 (above).

6. The warping factor tested against warning and error limits (and reported in warning and error
messages) is the larger of the area factor and, if available, the thickness factor.

7. The best possible quadrilateral warping factor, for a flat quadrilateral, is zero.

Figure 12.17: Quadrilateral Shell Having Warping Factor (p. 411) shows a "warped" element plotted
on top of a flat one. Only the right-hand node of the upper element is moved. The element is a
unit square, with a real constant thickness of 0.1.

When the upper element is warped by a factor of 0.01, it cannot be visibly distinguished from the
underlying flat one.

When the upper element is warped by a factor of 0.04, it just begins to visibly separate from the
flat one.

Figure 12.17: Quadrilateral Shell Having Warping Factor

Warping of 0.1 is visible given the flat reference, but seems trivial; however, it is well beyond the
error limit for a membrane shell. Warping of 1.0 is visually unappealing. This is the error limit for
most shells.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 411
Element Tools

Warping beyond 1.0 would appear to be obviously unacceptable; however, SHELL181 permits even
this much distortion. Furthermore, the warping factor calculation seems to peak at about 7.0.
Moving the node further off the original plane, even by much larger distances than shown here,
does not further increase the warping factor for this geometry. Users are cautioned that manually
increasing the error limit beyond its default of 5.0 for these elements could mean no real limit on
element distortion.

[Link]. Warping Factor Calculation for 3D Solid Elements

The warping factor for a 3D solid element face is computed as though the 4 nodes make up a
quadrilateral shell element with no real constant thickness available, using the square root of the
projected area of the face as described in 4 (above).

The warping factor for the element is the largest of the warping factors computed for the 6 quad-
rilateral faces of a brick, 3 quadrilateral faces of a wedge, or 1 quadrilateral face of a pyramid.

Any brick element having all flat faces has a warping factor of zero (Figure 12.18: Warping Factor
for Bricks (p. 412)).

Figure 12.18: Warping Factor for Bricks

Twisting the top face of a unit cube by 22.5° and 45° relative to the base produces warping factors
of about 0.2 and 0.4, respectively.

Table 12.5: Warping Factor Limits

Command to modify Type of Default Why default is this Why default is this
limit tight loose
SHPP,MODIFY,51 warning for 1 Elements having Element
"bending warping factors > 1 formulation
with high look like they derived from
warping deserve warnings 8-node solid
limit" shells isn't disturbed
(for by warping.
example,
SHELL181 (p. 586)) Disturbance of
analysis results
has not been
proven
SHPP,MODIFY,52 same as 5 Pushing this limit Valid analyses
above, error further does not should not be
limit seem prudent blocked.
SHPP,MODIFY,53 warning for 0.1 The element It is difficult to avoid
"non-stress" formulation is these warnings even
shells or based on flat with a limit of 0.1.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
412 Published: 2023-10-23
 Element Shape Testing

Command to modify Type of Default Why default is this Why default is this
limit tight loose
"bending shell theory,
stiffness with rigid beam
included" offsets for
shells moment
without compatibility.
geometric
nonlinearities Informal testing
(for has shown that
example, result error
SHELL131 (p. 534)) became
significant for
warping factor
> 0.1.
SHPP,MODIFY,54 same as 1 Pushing this limit Valid analyses
above, error further does not should not be
limit seem prudent. blocked.
SHPP,MODIFY,55 warning for 0.02 The element Informal testing has
"membrane formulation is based shown that the
stiffness on flat shell theory, effect of warping <
only" shells without any 0.02 is negligible.
correction for
moment
compatibility. The
element cannot
handle forces not in
the plane of the
element.
SHPP,MODIFY,56 same as 0.2 Pushing this limit Valid analyses
above, error further does not should not be
limit seem prudent blocked.
SHPP,MODIFY,67 warning for 0.2 A warping factor of Disturbance of
3D solid 0.2 corresponds to analysis results has
element about a 22.5° not been proven.
quadrilateral rotation of the top
face face of a unit cube.
Brick elements
distorted this much
look like they
deserve warnings.
SHPP,MODIFY,68 same as 0.4 A warping factor of Valid analyses
above, error 0.4 corresponds to should not be
limit about a 45° rotation blocked.
of the top face of a
unit cube. Pushing
this limit further
does not seem
prudent.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 413
Element Tools

12.2. Integration Point Locations

Mechanical APDL makes use of both standard and nonstandard numerical integration formulas. The
particular integration scheme used for each matrix or load vector is given with each element description
in Element Library (p. 445). Both standard and nonstandard integration formulas are described in this
section. The numbers after the subsection titles are labels used to identify the integration point rule.
For example, line (1, 2, or 3 points) represents the 1, 2, and 3 point integration schemes along line ele-
ments. Midside nodes, if applicable, are not shown in the figures in this section.

12.2.1. Lines (1, 2, or 3 Points)

The standard 1D numerical integration formulas which are used in the element library are of the form:

(12.1)

where:

f(x) = function to be integrated

Hi = weighting factor (see Table 12.6: Gauss Numerical Integration Constants (p. 414))
xi = locations to evaluate function (see Table 12.6: Gauss Numerical Integration Con-
stants (p. 414); these locations are usually the s, t, or r coordinates)
= number of integration (Gauss) points

Table 12.6: Gauss Numerical Integration Constants

No. Integration
Integration Point Locations (xi) Weighting Factors (Hi)
Points
1 0.00000.00000.00000 2.00000.00000.00000
2 ±0.57735 02691 89626 1.00000.00000.00000
±0.77459 66692 41483 0.55555 55555 55556
3
0.00000.00000.00000 0.88888 88888 88889

For some integrations of multi-dimensional regions, the method of Equation 12.1 (p. 414) is simply
expanded, as shown below.

12.2.2. Quadrilaterals (2 x 2 or 3 x 3 Points)

The numerical integration of 2D quadrilaterals gives:

(12.2)

and the integration point locations are shown in Figure 12.19: Integration Point Locations for Quadri-
laterals (p. 415). The locations and weighting factors can be calculated using Table 12.6: Gauss Numer-
ical Integration Constants (p. 414) two times.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
414 Published: 2023-10-23
 Integration Point Locations

Figure 12.19: Integration Point Locations for Quadrilaterals

One element models with midside nodes (such as PLANE183) using a 2 x 2 mesh of integration points
have been seen to generate spurious zero energy (hourglassing) modes.

12.2.3. Bricks and Pyramids (2 x 2 x 2 Points)

The 3D integration of bricks and pyramids gives:

(12.3)

and the integration point locations are shown in Figure 12.20: Integration Point Locations for Bricks
and Pyramids (p. 415). The locations and weighting factors can be calculated using Table 12.6: Gauss
Numerical Integration Constants (p. 414) three times.

Figure 12.20: Integration Point Locations for Bricks and Pyramids

One element models with midside nodes using a 2 x 2 x 2 mesh of integration points have been seen
to generate spurious zero energy (hourglassing) modes.

12.2.4. Triangles (1, 3, or 6 Points)

The integration points used for these triangles are given in Table 12.7: Numerical Integration for Tri-
angles (p. 415) and appear as shown in Figure 12.21: Integration Point Locations for Triangles (p. 416).
L varies from 0.0 at an edge to 1.0 at the opposite vertex.

Table 12.7: Numerical Integration for Triangles

Type Integration Point Location Weighting Factor

1 Point Rule L1=L2=L3=.3333333 1.000000

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 415
Element Tools

Type Integration Point Location Weighting Factor

L1=.66666 66666 66666 0.33333 33333 33333

L2=L3=.16666 66666 66666

3 Point Rule
Permute L1, L2, and L3 for
other locations)
L1=0.81684 75729 80459 0.10995 17436 55322

Corner L2=L3=0.09157 62135 09661

Points
Permute L1, L2, and L3 for
other locations)
6 Point Rule
L1=0.10810 30181 6807 0.22338 15896 78011
Edge
L2=L3=0.44594 84909 15965
Center
Points Permute L1, L2, and L3 for
other locations)

Figure 12.21: Integration Point Locations for Triangles

12.2.5. Tetrahedra (1, 4, 5, or 11 Points)

The integration points used for tetrahedra are given in Table 12.8: Numerical Integration for Tetra-
hedra (p. 416).

Table 12.8: Numerical Integration for Tetrahedra

Type Integration Point Location Weighting Factor

Center L1=L2=L3=L4=.25000 00000 00000 1.00000 00000 00000
1 Point Rule
Point
L1=.58541 01966 24968 0.25000 00000 00000

L2=L3=L4=.13819 66011
Corner
4 Point Rule 25010
Points
Permute L1, L2, L3, and L4 for
other locations)
Center L1=L2=L3=L4=.25000 00000 00000 -0.80000 00000 00000
5 Point Rule
Point

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
416 Published: 2023-10-23
 Integration Point Locations

Type Integration Point Location Weighting Factor

L1=.50000 00000 00000 0.45000 00000 00000

L2=L3=L4=.16666 66666
Corner
66666
Points
Permute L1, L2, L3, and L4 for
other locations)
Center L1=L2=L3=L4=.25000 00000 00000 -0.01315 55555 55556
Point
Corner L1=L2=L3=.0714285714285714 0.00762 22222 22222
Point
L4=.78571 42857 14286

(Permute L1, L2, L3 and L4 for

other three locations)
11 Point L1=L2=0.39940 35761 66799 0.02488 88888 88888
Rule
L3=L4=0.10059 64238 33201

Edge Permute L1, L2, L3 and L4

Center such that two of L1, L2, L3
Points
and L4 equal 0.39940 35761
66799 and the other two
equal 0.10059 64238 33201
for other five locations

These appear as shown in Figure 12.22: Integration Point Locations for Tetrahedra (p. 417). L varies
from 0.0 at a face to 1.0 at the opposite vertex.

Figure 12.22: Integration Point Locations for Tetrahedra

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 417
Element Tools

12.2.6. Triangles and Tetrahedra (2 x 2 or 2 x 2 x 2 Points)

These elements use the same integration point scheme as for 4-node quadrilaterals and 8-node solids,
as shown in Figure 12.23: Integration Point Locations for Triangles and Tetrahedra (p. 418). The locations
and weighting factors can be calculated using Table 12.6: Gauss Numerical Integration Constants (p. 414)
two or three times.

Figure 12.23: Integration Point Locations for Triangles and Tetrahedra

3x3 and 3x3x3 cases are handled similarly.

12.2.7. Wedges (3 x 2 or 3 x 3 Points)

These wedge elements use an integration scheme that combines linear and triangular integrations,
as shown in Figure 12.24: 6 and 9 Integration Point Locations for Wedges (p. 418). The locations and
weighting factors can be calculated using Table 12.6: Gauss Numerical Integration Constants (p. 414)
and Table 12.7: Numerical Integration for Triangles (p. 415).

Figure 12.24: 6 and 9 Integration Point Locations for Wedges

12.2.8. Wedges (2 x 2 x 2 Points)

These wedge elements use the same integration point scheme as for 8-node solid elements as shown
by two orthogonal views in Figure 12.25: 8 Integration Point Locations for Wedges (p. 419). The locations
and weighting factors can be calculated using Table 12.6: Gauss Numerical Integration Constants (p. 414)
three times.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
418 Published: 2023-10-23
 Integration Point Locations

Figure 12.25: 8 Integration Point Locations for Wedges

12.2.9. Bricks (14 Points)

The 20-node solid uses a different type of integration point scheme. This scheme places points close
to each of the 8 corner nodes and close to the centers of the 6 faces for a total of 14 points. These
locations are given in Table 12.9: Numerical Integration for 20-Node Brick (p. 419):

Table 12.9: Numerical Integration for 20-Node Brick

Type Integration Point Location Weighting Factor

14 Point Corner s = ±.75878 69106 39328 .33518 00554 01662
Rule Points
t = ±.75878 69106 39329

r = ±.75878 69106 39329

Center s = ±.79582 24257 54222, .88642 65927 97784
Points t=r=0.0

t = ±.79582 24257 54222,

s=r=0.0

r = ±.79582 24257 54222,

s=t=0.0

and are shown in Figure 12.26: Integration Point Locations for 14 Point Rule (p. 419).

Figure 12.26: Integration Point Locations for 14 Point Rule

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 419
Element Tools

12.2.10. Nonlinear Bending (5 Points)

Both beam and shell elements that have nonlinear materials must have their effects accumulated
thru the thickness. This uses nonstandard integration point locations, as both the top and bottom
surfaces have an integration point in order to immediately detect the onset of the nonlinear effects.

Table 12.10: Thru-Thickness Numerical Integration

Integration Point
Type Weighting Factor
Location[1]
±0.500 0.1250000
5 ±0.300 0.5787036
0.000 0.5925926

1. Thickness coordinate going from -0.5 to 0.5.

These locations are shown in Figure 12.27: Nonlinear Bending Integration Point Locations (p. 420).

Figure 12.27: Nonlinear Bending Integration Point Locations

12.2.11. General Axisymmetric Elements

The numerical integration of general axisymmetric elements gives:

(12.4)

Hi and Hj are weighting factors on the rz plane, as shown in Figure 11.20: General Axisymmetric Solid
Elements (when Nnp = 3) (p. 393). The values are shown in Table 12.6: Gauss Numerical Integration
Constants (p. 414). In circumferential direction θ:
(12.5)

12.3. Temperature-Dependent Material Properties

Temperature-dependent material properties are evaluated at each integration point. Elements for which
this applies include SHELL181, PLANE182, PLANE183 , SOLID185, SOLID186 , SOLID187, SOLID272, SOL-
ID273, SOLID285, SOLSH190, BEAM188, BEAM189, SHELL208, SHELL209, REINF264, SHELL281, PIPE288,
PIPE289, and ELBOW290.

The LINK180 element uses a closed form solution (without integration points) and has its material
properties evaluated at the average temperature of the element.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
420 Published: 2023-10-23
 Positive Definite Matrices

Other cases:

• For the structural element PLANE13, the nonlinear material properties (TB commands) are evaluated
at the integration points, but the linear material properties (MP commands) are evaluated at the av-
erage element temperature.

• Numerically integrated structural elements PLANE25, SHELL61, and PLANE83 have their linear mater-
ial properties evaluated at the average element temperature.

• Non-structural elements have their material properties evaluated only at the average element tem-
perature, except for the specific heat (Cp) which is evaluated at each integration point.

Whether shape functions are used or not, materials are evaluated at the temperature given, i.e. no account
is made of the temperature offset (TOFFST command).

For a stress analysis, the temperatures used are based directly on the input. As temperature is the un-
known in a heat transfer analysis, the material property evaluation cannot be handled in the same direct
manner. For the first iteration of a heat transfer analysis, the material properties are evaluated at the
uniform temperature (input on the BFUNIF command). The properties of the second iteration are based
on the temperatures of the first iteration. The properties of the third iteration are based on the temper-
atures of the second iteration, etc.

See Temperature-Dependent Coefficient of Thermal Expansion (p. 10) for a special discussion about
the coefficient of thermal expansion.

12.4. Positive Definite Matrices

By definition, a matrix [D] (as well as its inverse [D]-1) is positive definite if the determinants of all sub-
matrices of the series:

(12.6)

including the determinant of the full matrix [D], are positive. The series could have started out at any
other diagonal term and then had row and column sets added in any order. Thus, two necessary (but
not sufficient) conditions for a symmetric matrix to be positive definite are given here for convenience:
(12.7)
(12.8)

If any of the above determinants are zero (and the rest positive), the matrix is said to be positive
semidefinite. If all of the above determinants are negative, the matrix is said to be negative definite.

12.4.1. Matrices Representing the Complete Structure

In virtually all circumstances, matrices representing the complete structure with the appropriate
boundary conditions must be positive definite. If they are not, the message "NEGATIVE PIVOT . . ."
appears. This usually means that insufficient boundary conditions were specified. An exception is a
piezoelectric analysis, which works with negative definite matrices, but does not generate any error
messages.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 421
Element Tools

12.4.2. Element Matrices

Element matrices are often positive semidefinite, but sometimes they are either negative or positive
definite. For most cases where a negative definite matrix could inappropriately be created, the program
will abort with a descriptive message.

12.5. Lumped Matrices

Some of the elements allow their consistent mass or specific heat matrices to be reduced to diagonal
matrices (accessed with the LUMPM,ON command). This is referred to as "lumping".

12.5.1. Diagonalization Procedure

One of two procedures is used for the diagonalization, depending on the order of the element shape
functions. The mass matrix is used as an example.

For lower order elements (linear or bilinear) the diagonalized matrix is computed by summing rows
(or columns). The steps are:

1. Compute the consistent mass matrix in the usual manner.

2. Compute:

(12.9)

where:

n = number of degrees of freedom (DOFs) in the element

3. Set
(12.10)
(12.11)

For higher order elements the procedure suggested by Hinton, et al.([46] (p. 927)), is used. The steps
are:

1. Compute the consistent mass matrix in the usual manner.

2. Compute:

(12.12)

(12.13)

3. Set:
(12.14)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
422 Published: 2023-10-23
 Lumped Matrices

(12.15)

Note that this method ensures that:

1. The element mass is preserved

2. The element mass matrix is positive definite

It may be observed that if the diagonalization is performed by simply summing rows or columns in
higher order elements, the resulting element mass matrix is not always positive definite.

12.5.2. Special Handling of Rotational Degrees of Freedom

Diagonalization of a mass matrix containing rotational degrees of freedom may lead to the following
complications:

1. The lumped mass matrix may lose the property of frame invariance.

2. The coupling between translational and rotational degrees of freedom, generally created by un-
balanced laminate construction or sections with offsets, may be present in the lumped mass
matrix.

Rotational degrees of freedom are excluded in the lumped mass matrix, except for the current tech-
nology elements, which include SHELL181, BEAM188, BEAM189, SHELL208, SHELL209, SHELL281,
PIPE288, PIPE289, and ELBOW290. The following options are available for handling the rotational de-
grees of freedom in current technology elements (see the KeyElt argument of the LUMPM command):

1. Direct diagonalization (see Diagonalization Procedure (p. 422)), in which rotational degrees of
freedom are treated the same as translational degrees of freedom. However, any coupling between
the translational and rotational degrees of freedom is excluded in the process.

2. Translational mass only, in which the mass contributions from rotational degrees of freedom, in-
cluding the coupling with the translational degrees of freedom, are entirely excluded.

3. Frame invariant diagonalization, in which the lumped rotational mass components at each node
are made identical to achieve frame invariance. The procedure is carried out only for BEAM188,
BEAM189, PIPE288, and PIPE289 elements. The value of the lumped mass contribution Me(j,j) of
rotational degree of freedom j is given as follows:
(12.16)

Where Me(i,i) is the lumped translational mass contribution of degree of freedom i from the same
node, Iyy and Izz are the moments of inertia about the local element y and z axes, and A is the
cross section area.

12.5.3. Limitations of Lumped Mass Matrices

Lumped mass matrices have the following limitations:

1. Elements containing both translational and rotational degrees of freedom will have mass contri-
butions only for the translational degrees of freedom. Rotational degrees of freedom are included

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 423
Element Tools

for SHELL181, BEAM188, BEAM189, SHELL208, SHELL209, SHELL281, PIPE288, PIPE289, and EL-
BOW290 (see Special Handling of Rotational Degrees of Freedom (p. 423) for details).

2. Lumping, by its very nature, eliminates the concept of mass coupling between degrees of freedom.
Therefore, the following restrictions exist:

• Lumping is not allowed for FLUID29, FLUID30, or FLUID38 elements.

• Lumping is not allowed for the mass matrix option of MATRIX27 elements if it is defined with
nonzero off-diagonal terms.

Note that coupling due to constraint equations is always included.

12.6. Reuse of Matrices

Matrices are reused automatically as often as possible in order to decrease running time. The information
below is made available for use in running time estimates.

12.6.1. Element Matrices

For static (ANTYPE,STATIC) or full transient dynamic (ANTYPE,TRANS with TRNOPT,FULL) analyses,
element stiffness/conductivity, mass, and damping/specific heat, matrices ([Ke], [Me], [Ce]) are always
reused from iteration to iteration, except when:

1. The full Newton-Raphson option (NROPT,FULL) is used, or for the first equilibrium iteration of a
time step when the modified Newton-Raphson option (NROPT,MODI) is used and the element
has either nonlinear materials or large deformation (NLGEOM,ON) is active.

2. The element is nonlinear (e.g. gap, radiation, or control element) and its status changes.

3. MODE or ISYM (MODE command) have changed from the previous load step for elements PLANE25,
SHELL61, PLANE75, PLANE78, or PLANE83.

4. will be reformulated if a convective film coefficient (input on the SF or SFE commands) on

an element face changes. Such a change could occur as a ramp (KBC,0) within a load step.

5. The materials or real constants are changed by new input, or if the material properties have
changed due to temperature changes for temperature-dependent input.

Element stress stiffness matrices [Se] are never reused, as the stress normally varies from iteration to
iteration.

12.6.2. Structure Matrices

The overall structure matrices are reused from iteration to iteration except when:

1. An included element matrix is reformed (see above).

2. The set of specified degrees of freedom (DOFs) is changed.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
424 Published: 2023-10-23
 Hydrostatic Loads

3. The integration time step size changes from that used in the previous substep for the transient
(ANTYPE,TRANS) analysis.

4. The first iteration of a restart is being performed.

12.6.3. Override Option

The above tests are all performed automatically by the program. The user can select to override the
program's decision with respect to whether the matrices should be reformed or not. For example, if
the user has temperature-dependent input as the only cause which is forcing the reformulation of
the matrices, and there is a load step where the temperature dependency is not significant, the user
can select that the matrices will not be reformed at that load step (KUSE,1). (Normally, the user would
want to return control back to the program for the following load step (KUSE,0)). On the other hand,
the user can select that all element matrices are to be reformed each iteration (KUSE,-1).

12.7. Hydrostatic Loads

The following topics related to hydrostatic loading are available:
12.7.1. Internal and External Pressures
12.7.2. Buoyancy
12.7.3. Effective Tension
12.7.4. Instability Checking
12.7.5. Effect of Water Pressure on Some Elements

12.7.1. Internal and External Pressures

For pipe cross sections, external pressure tends to crush the pipe and internal pressure tends to sta-
bilize it.

The inside static pressure is:

(12.17)

where:

= internal pressure
= input internal pressure (input as face 1 on the SFE command)
= internal fluid density
= acceleration due to gravity (input as ACEL_Z on the ACEL command)
= vertical coordinate of the point of interest
= Z coordinate of free surface of internal fluid (input as face 3 on the SFE command)

The external static pressure is:

(12.18)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 425
Element Tools

= external (crushing) pressure due to hydrostatic effects

= input external pressure (input as face 2 on SFE command)
= external fluid (ocean) density
= vertical offset from the global origin to the mean sea level (input on the OCDATA
command)

12.7.2. Buoyancy
Buoyancy on the outside tends to raise the pipe to the water surface. The buoyant force for a com-
pletely submerged element acting in the positive Z direction is:
(12.19)

where:
= vertical load per-unit-length due to buoyancy

= outside diameter including insulation

= coefficient of buoyancy (input as Cb on the OCDATA command)

Also, an adjustment for the added mass term is made.

12.7.3. Effective Tension

Effective tension, subjected to internal and/or external pressures (effective force), is useful for determ-
ining, among other effects, whether a pipe model will buckle. Effective tension is calculated as:
(12.20)

where:

= effective tension

= force in pipe wall =

= axial stress in pipe wall
= internal pressure

= inside area =
= inside diameter =
= outside diameter (input via SECDATA)
= wall thickness (input via SECDATA)
= external pressure
= outside area =

Exception: If KEYOPT(6) = 1 (PIPE288 and PIPE289), specifying that internal and external pressures
do not cause loads on end caps.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
426 Published: 2023-10-23
 Hydrostatic Loads

12.7.4. Instability Checking

To ensure that the problem is physically possible as input, a check is made at the element midpoint
to see if the pipe cross-section collapses under the hydrostatic effects. The cross-section is assumed
to be unstable if:

(12.21)

where:

= Young's modulus (input as EY on MP command)

= Poisson's ratio (input as PRXY or NUXY on MP command)

12.7.5. Effect of Water Pressure on Some Elements

Elements such as PIPE288 have three direct stress components: axial, hoop, and radial. Elements such
as LINK180 and BEAM188, however, have only an axial stress component; therefore, an adjustment
is necessary to include the effect of the three stress components of a hydrostatic load.

Consider the first line of Equation 2.2 (p. 6), assuming isotropic materials and no thermal effect:
(12.22)

where:

= axial strain
= Young’s modulus (input as EX on the MP command)
= axial stress
= Poisson’s ratio (input as PRXY or NUXY on the MP command)
= = stresses normal to the axial direction

For hydrostatic loads:

(12.23)

where:

= hydrostatic pressure

Combining Equation 12.23 (p. 427) and Equation 12.24 (p. 427):
(12.24)

The factor ( ) is used to give the correct strain. Thus, the following restrictions apply to this ap-
proach:

• Material must be isotropic.

• Material must be elastic.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 427
Element Tools

• Fluid pressure is assumed to be hydrostatic everywhere (that is, CTUBE and HREC beam subtypes
presume that the flooding option is applied, and there is no option for removing it).

It is of course preferable to use the pipe element rather than the beam element, as doing so avoids
the limitations that apply to the beam element with subtype CTUBE.

Because hydrostatic pressure (stress) is normally much lower than the working stress, the restrictions
described above typically need to be considered only for extremely deep applications.

12.8. Hydrodynamic Loads

Hydrodynamic effects may occur because the structure moves in a motionless fluid, the structure is
fixed but fluid motion is present, or both the structure and fluid are moving.

Fluid motion consists of two parts: current and wave motions. Current is input by giving the current
velocity and direction (input as W(i) and θ(i)) at up to eight different vertical stations (input as Z(i)). (All
input quantities referred to in this section not otherwise identified come from the OCTYPE, OCDATA,
and OCTABLE commands.

The velocity and direction are interpolated linearly between stations. The current is assumed to flow
horizontally only. The wave may be input using one of the wave theories (Kw, input as KWAVE on the
OCDATA command) in the following table :

Table 12.11: Wave Theory Table

Wave Theory Kw
Small amplitude wave, unmodified (Airy wave theory), 0
(Wheeler([36] (p. 926)))
Small amplitude wave, modified with empirical depth decay function, 1
(Wheeler([36] (p. 926)))
Stokes fifth order wave, (Skjelbreia([32] (p. 926))) 2
Stream function wave, (Dean([60] (p. 928)), Dalrymple([414] (p. 948))) 3
Random wave (p. 432) 5
Shell new wave (p. 434), (Atkins([388] (p. 947))) 6
Constrained new wave (p. 434), (O'Neill([389] (p. 947)), Pinna([390] (p. 947))) 7
Diffracted wave (p. 436) 8

These related topics follow:

12.8.1. Regular Waves on Line Elements
12.8.2. Irregular Waves on Line Elements (Kw = 5 through 7)
12.8.3. Diffracted Wave on Line and Surface Elements (Kw = 8)
12.8.4. Presence of Both Waves and Current
12.8.5. MacCamy-Fuchs Corrections
12.8.6. Morison's Equation
12.8.7. Dynamic Pressure Head

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
428 Published: 2023-10-23
 Hydrodynamic Loads

12.8.1. Regular Waves on Line Elements

Regular waves include both linear (Airy) and nonlinear (Stoke’s fifth order and Dean’s stream function)
waves.

[Link]. Small Amplitude Linear (Airy) Wave (Kw = 0, 1)

The free surface of the wave is defined by:

(12.25)

where:

= total wave amplitude

= number of waves
= wave amplitude of component i
= surface coefficient (component height) (input quantity H(i) on the
OCTABLE command)

= radial distance to point on element from origin in the X-Y plane in the
direction of the wave
= wave length (input as L(i) on the OCTABLE command if L(i) > 0.0, or
derived from Equation 12.26 (p. 429) otherwise
= time elapsed (input as TIME on the TIME command). The default value
of TIME is usually not desired. If zero is needed, use (1.0E-12) instead.
= wave period (input as T(i) on the OCTABLE command)
= phase shift (input as Ps(i) on the OCDATA command)
= wave location (input as WAVELOC on the OCDATA command)

If Kw = 0 or 1, and Nw > 1, the wave is considered to be irregular (p. 431).

If λi is not input (set to zero) and Kw = 0 or 1, λi is calculated iteratively from:

(12.26)

where:

= output quantity small amplitude wave length

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 429
Element Tools

= output quantity deep water wave length

= acceleration due to gravity (Z direction) (input via the ACEL command)
= water depth (input as DEPTH on the OCDATA command)

Each component of wave height is checked to verify that it satisfies the Miche criterion. The check
ensures that the wave is not a breaking wave, which the included wave theories do not cover. A
breaking wave is one that spills over its crest, normally in shallow water. A warning message is issued
if:
(12.27)

where = height of breaking wave

When using wave loading, an error check occurs to ensure that the input acceleration does not
change after the first load step, as this would imply a change in the wave behavior between load
steps.

The particle velocities at integration points are calculated as a function of depth from:

(12.28)

(12.29)

where:

= radial particle velocity

= vertical particle velocity
ki = 2π/λi
= wave frequency (= 2π/τi)
= height of integration point above the ocean floor = d + Z
= time derivative of ηi

= drift velocity (input as W on the OCTABLE command)

The particle accelerations are calculated by differentiating and with respect to time. Thus:

(12.30)

(12.31)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
430 Published: 2023-10-23
 Hydrodynamic Loads

= radial particle acceleration

= vertical particle acceleration

= wave number (= 2π / λi)
= wave frequency 2π / τi

Expanding Equation 2.29 of the Shore Protection Manual([44] (p. 927)) for a multiple component
wave, the wave hydrodynamic pressure is:

(12.32)

However, use of this equation leads to nonzero total pressure at the surface at the crest or trough
of the wave. Thus, Equation 12.32 (p. 431) is modified to be:

(12.33)

which does result in a total pressure of zero at all points of the free surface. This dynamic pressure,
calculated at the integration points during the stiffness pass, is extrapolated to the nodes for the
stress pass. The hydrodynamic pressure for Stokes fifth order wave theory is:

(12.34)

[Link]. Stoke’s Fifth Order Wave (Kw = 2)

The Stokes fifth order wave theory is discussed by Skjelbreia ([32] (p. 926)). The modification suggested
by Nishimura ([144] (p. 933)) has been included.

[Link]. Dean’s Stream Function Wave (Kw = 3)

The stream function wave theory is described by Dean ([60] (p. 928)). A solution algorithm that uses
the Lagrange multiplier technique suggested by Dalrymple ([414] (p. 948)) is included as a base al-
gorithm.

12.8.2. Irregular Waves on Line Elements (Kw = 5 through 7)

The random wave (p. 432) (Kw = 5), Shell new wave (p. 434) (Kw = 6) and constrained new wave (p. 434)
(Kw = 7) models assume that the wave particle kinematics are generated from a wave spectrum using
linear theory.

A wave spectrum is characterized by a curve of spectral density of wave energy S(ωi). The wave
spectrum can be defined in one of the three forms: Pierson-Moskowitz ([391] (p. 947)), JONSWAP
([392] (p. 947)), or user-defined.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 431
Element Tools

These related topics follow:

[Link]. Random Wave
[Link]. Shell New Wave
[Link]. Constrained New Wave

[Link]. Random Wave

The random wave is created by adding the parameters (wave height, velocity, and acceleration) of
a number of regular Airy waves (wave components) with amplitudes corresponding to the required
spectrum and with random phases.

With the Pierson-Moskowitz and JONSWAP spectrum, the spectrum is divided into a number of
equal energy strips (that is, equal areas) based on the number of wave components specified (input
as NWC on the OCTABLE command). Each of the strips is a wave component, the frequency of
which is defined at the centroid of the strip. The amplitude is given by the square root of twice
the area of the strip and, due to the equal-energy-based strips, remains constant for each wave
component.

For a user-defined spectrum, the frequency and energy parameters to form the spectrum are input,
so the amplitudes may vary for each component. The solution is most stable numerically when
wave component amplitudes are kept constant throughout the definition.

The free surface of the wave can be defined by

but with revised definitions for some of the parameters, as follows:

= number of wave components (input as NWC on the OCTABLE command)

= surface coefficient = where mo is the area under the full spectrum

= wave length derived from Equation 12.26 (p. 429)

= wave period = , where ωi is the angular frequency for wave component i

= phase shift (based on a randomly generated number)

For a given initial seed value (input as SEED on the OCDATA command), the same sequence of
random phases is generated in every run. Changing the initial seed generates a different set of
random phases.

A form of Wheeler stretching, known as delta stretching, is introduced to provide the wave kinematics
under the crest. The depth-mapping function modifies the conventional cosh and sinh terms in the
wave kinematic equations to provide linear extrapolation of the velocities and accelerations above
mean water level so that unrealistic terms are prevented from developing.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
432 Published: 2023-10-23
 Hydrodynamic Loads

Similar to equations Equation 12.28 (p. 430) and Equation 12.29 (p. 430), the radial and vertical fluid
particle velocities are given by:

(12.35)

(12.36)

where:

= radial depth-mapping function, calculated as:

= vertical depth-mapping function, computed as:

= wave kinematics factor (input as KINE on the OCDATA command)

that accounts for the effect of wave-spreading on the radial velocity and
acceleration
= wave number =

The effective depth Z’ is used to compute the depth functions and is given by:

(12.37)

where:

= delta-stretching parameter (input as DELTA on the OCDATA command)

= = depth under the mean water level in which stretching starts to
become effective
= stretching depth factor (input as DSA on the OCDATA command)

= significant wave height (input directly as HS on the OCTABLE command

or calculated internally from the spectrum data on the OCTABLE command)

The particle accelerations are calculated by differentiating and with respect to time.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 433
Element Tools

[Link]. Shell New Wave

The Shell new wave model is similar to random wave (p. 432). It uses a statistically based linear su-
perposition of linear wave components to define the wave profile and associated kinematics rep-
resenting the most likely maximum condition of a real sea.

The free surface of the wave can again be defined by

but with revised definitions for some parameters, as follows:

= Number of wave components (input as NWC on the OCTABLE command)

= surface coefficient = where AC is the wave crest amplitude (input as AMPMAX

on the OCDATA command), S(ωi) is value of the spectral density at angular frequency ωi, Δωi
is the frequency band which is assumed constant for this wave model, and mo is the area under
the full spectrum

= for evolving wave (input as EVOLVING on the OCDATA command), or

for frozen wave

= wave length derived from Equation 12.26 (p. 429)

= wave period = , where ωi is the angular frequency for wave component i

= position offset in the direction of the wave, relative to the structural origin, where the
maximum wave crest occurs (input as ROFF on the OCDATA command)
= time at which the maximum wave crest occurs (input as TOFF on the OCDATA command)

= wave celerity (= ) associated with a wave having a wave period of TZ, where TZ is the
zero-crossing period of the wave spectrum that can be computed from the spectral peak period.

The radial and vertical fluid particle velocities are given by Equation 12.35 (p. 433) and Equa-
tion 12.36 (p. 433) except that the wave kinematics factor fK is now computed from:

where = wave-spreading angle in radians (input as SPANGLE on the OCDATA command)

Delta stretching is available for Shell new wave as described in Random Wave (p. 432). All definitions
are similar except that Ds in Equation 12.37 (p. 433) is set to AC here.

[Link]. Constrained New Wave

Constrained new wave theory embeds a Shell new wave (p. 434) into a random wave (p. 432) so that
the maximum crest amplitude as given by the new wave occurs at a specific time and position
while the statistical nature of the random sea is preserved.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
434 Published: 2023-10-23
 Hydrodynamic Loads

The free surface of a constrained new wave is given by:

(12.38)

where:

= surface elevation equation of a

random wave as previously defined.

= surface elevation equation of a

new wave that yields unit crest amplitude at time t0 and position R0 (input as
TOFF and ROFF, respectively, on the OCDATA command.
= desired constrained wave crest amplitude (input as AMPCONST on the
OCDATA command)
A dot or dots over a variable represent the usual time derivative(s) of the variable.

Equation 12.38 (p. 435) is set up such that it satisfies the conditions of:

Therefore, the desired crest amplitude is produced at (R0, t0) and the corresponding slope of the
free surface is zero.

The height of the new wave components Hn i is related to the random wave component height Hi.
For the equal energy packets that are assumed in the implementation, the expression is simplified
to:

The second and third terms in Equation 12.38 (p. 435) are related to new wave and can be further
manipulated to a more familiar form for linear waves:

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 435
Element Tools

A constrained new wave can thus be considered to have two sets of superimposed random waves
together; therefore, the wave particle kinematics can also be calculated by applying the formula
derived for random waves (p. 432). See Random Wave (p. 432) and refer to Equation 12.35 (p. 433)
through Equation 12.37 (p. 433) for details.

12.8.3. Diffracted Wave on Line and Surface Elements (Kw = 8)

Hydrodynamic pressures and motions obtained from a frequency-based hydrodynamic diffraction
analysis can be imported for further analysis. This feature is enabled by specifying a diffracted wave
type (Kw = 8, input as KWAVE on the OCDATA command). The hydrodynamic data are imported via
the OCREAD command.

Loading can be applied to a line or surface. Both cases are described in the following topics:
[Link]. Diffracted Waves on Line Elements
[Link]. Diffracted Waves on Surface Elements

[Link]. Diffracted Waves on Line Elements

Hydrodynamic loads on structural line elements are calculated based on the results obtained for
hydrodynamic line elements.

For hydrodynamic line elements, the wave data consist of the magnitudes and phase angles of the
wave free surface, dynamic wave pressure, and velocities and accelerations at both ends of the line
elements at a number of defined wave periods and directions.

If the required wave direction θ (input as THETA on the OCDATA command) and/or wave frequency
ωi (2π/τ(i), where τ(i) is the input wave period on the OCTABLE command) differs from the hydro-
dynamic data values, the wave information is obtained via linear interpolations between data in
neighboring hydrodynamic analyzed directions and frequencies.

The position of each structural integration point is mapped to the hydrodynamic mesh and the
hydrodynamic line element containing this point is determined. The wave data at this location are
interpolated linearly from the nodal values of the hydrodynamic line element. The mapping operation
assumes that the hydrodynamic mesh is fixed in the original configuration; therefore, structural
displacements must be small for this procedure to work accurately.

Finally, the wave free surface at the specified phase angle φi (input as φ(i) on the OCTABLE com-
mand) is given by:

(12.39)

where:

Hi = height of component i (input as A(i) on the OCTABLE command)

ηreal(θ,ωi) = real component of the surface elevation obtained from the hydro-
dynamic data
ηimag(θ,ωi) = imaginary component of the surface elevation obtained from the
hydrodynamic data

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
436 Published: 2023-10-23
 Hydrodynamic Loads

The wave height is required because the hydrodynamic diffraction results are for unit wave amplitude
and must be scaled to the required height.

The fluid particle velocities and accelerations are obtained in a similar fashion (that is, with η replaced
by velocity V or acceleration V in Equation 12.39 (p. 436). Current effects, if specified, are added to
the fluid velocities and Morison’s equation is then applied to calculate the hydrodynamic loads.

[Link]. Diffracted Waves on Surface Elements

Hydrodynamic loading acting on a surface is applied by mapping the pressure results on the hydro-
dynamic panel elements onto the structural SURF154 elements. Hydrodynamic loading is introduced
by setting KEYOPT(8) = OCID in the SURF154 element, where OCID is the ocean load ID containing
the hydrodynamic data.

For hydrodynamic panel elements, the wave data are the magnitudes and phase angles of the hy-
drodynamic pressure at the centroids of the panels at a number of defined wave periods and dir-
ections. If the required wave direction θ (input as THETA on the OCDATA command) and/or wave
frequency ωi ( 2π/τ(i), where τ(i) is the input wave period on the OCTABLE command) differs from
the hydrodynamic data values, the wave information is obtained by linear interpolations between
data in neighboring hydrodynamic analyzed directions and frequencies.

The position of each structural integration point is mapped to the hydrodynamic mesh and the
hydrodynamic panel element containing this point is determined. The real and imaginary wave
pressures at this location are interpolated linearly from the nodal values of the hydrodynamic
panel. Because the pressure data are given at the centroidal position, it is necessary to first calculate
the element nodal values from the centroidal values surrounding each node. The mapping operation
assumes that the hydrodynamic mesh is fixed in the original configuration; therefore, structural
displacements must be small for this procedure to work accurately.

Similar to Equation 12.39 (p. 436), the hydrodynamic pressure at the specified phase angle φi (input
as φ(i) on the OCTABLE command) is given by:

(12.40)

where:

Hi = height of component i (input as A(i) on the OCTABLE command)

Preal(θ,ωi) = real component of the pressure obtained from the hydrodynamic
data
Pimag(θ,ωi) = imaginary component of the pressure obtained from the hydro-
dynamic data

The pressure from Equation 12.40 (p. 437) is added to other pressure (PRES) loads applied to the
SURF154 element. Current is not considered in the pressure calculations, and static pressure is also
excluded.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 437
Element Tools

12.8.4. Presence of Both Waves and Current

If both waves and current are present, the question of wave-current interaction must be considered.
Several options are made available via Kcr (input as KCRC on the OCDATA command):

• For Kcr = 0, the current velocity at all points above the mean sea level is simply set equal to Wo,
where Wo is the input current velocity at Z = 0.0. All points below the mean sea level have velocities
selected as though no wave exists.

• For Kcr = 1, the current velocity profile is "stretched" or "compressed" to fit the wave. In equation
form, the Z coordinate location of current measurement is adjusted by

(12.41)

where:

Z(j) = Z coordinate location of current measurement (input as Z(j))

= adjusted value of Z(j)

For Kcr = 2, the same adjustment as for Kcr = 1 is used, as well as a second change that accounts for
"continuity." That is,
(12.42)

where:

W(j) = velocity of current at this location (input as W(j))

= adjusted value of W(j)

The velocity options are shown pictorially in Figure 12.28: Velocity Profiles for Wave-Current Interac-
tions (p. 438).

Figure 12.28: Velocity Profiles for Wave-Current Interactions

An additional Kcr option is available for Kw = 5, 6 or 7. For Kcr = 3, the current velocity is computed
by a nonlinear stretching algorithm as recommended by the API RP2A codes of practice. The relation-
ship is given by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
438 Published: 2023-10-23
 Hydrodynamic Loads

where λp is the wave length of the wave components corresponding to the peak spectral period.

In the presence of current with wave, some codes of practice require the use of apparent wave period
in the computation of wave kinematics. The apparent period can be calculated iteratively from:
(12.43)

(12.44)

where:

T = wave period as seen by a stationary observer

Tapp = wave period as seen by an observer moving at velocity VI
VI = effective in-line current speed as recommended by the American Petroleum

Institute ([393] (p. 947)): , where Uc(z) is the

component of steady current profile at elevation z in the wave direction.

The apparent period calculation is available for Kw = 5, 6 or 7. The calculation is activated by setting
APC = 1 on the OCDATA command

12.8.5. MacCamy-Fuchs Corrections

When Kw = 0 through 3, the MacCamy-Fuchs adjustment can be requested (OCDATA) to account for
diffraction effects, especially for large-diameter objects with shorter wave lengths.

When the MacCamy-Fuchs option is applied, the following two adjustments occur:

• The coefficient of inertia is adjusted:

where:

J0 = zero order Bessel function of the first kind

J1 = first-order Bessel function of the first kind

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 439
Element Tools

Y0 = zero order Bessel function of the second kind

Y1 = first-order Bessel function of the second kind

• The phase shift is added to ϕ i (before the wave-location correction [input via WAVELOC on the
OCDATA command), if used):

12.8.6. Morison's Equation

For line elements, the hydrodynamic loads are calculated using Morison’s equation for a moving body:
(12.45)

where:

{F/L} = vector of loads per-unit-length due to hydrodynamic effects

ρw = water density (mass/length3)

A = cross-sectional area of the body (length2). For PIPE288 or PIPE289, this

value is . For BEAM188 or BEAM189, A is the displaced fluid area (output
via the SLIST command). This value can be augmented by the thickness of
insulation.
= velocity, acceleration of water particle
= velocity, acceleration of the structure (u = solved for unknown)
Ca = added mass coefficient
CD = drag coefficient (expanded below)
De = outer pipe diameter when using PIPE288 or PIPE289. For BEAM188 or
BEAM189, De is based on the average of the maximum height or width (y or
z directions, respectively) (output via the SLIST command). These values can
be augmented by the thicknesses of insulation and biofouling.

The first term in Equation 12.45 (p. 440) is the Froude-Krylov force, the second term is the hydrodynamic
mass force, and the third term is the drag force. The Froude-Krylov force and hydrodynamic mass
force can be rearranged for nonrelative accelerations by using the relationship between Ca and CM:
(12.46)

where CM = inertia coefficient (= Ca +1)

The equations of motion (p. 796) for a submerged structure are expressed as:
(12.47)

where [M], [C], [K] = structural mass, damping, and stiffness matrix per-unit-length, respectively.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
440 Published: 2023-10-23
 Hydrodynamic Loads

The first term on the right side of Equation 12.47 (p. 440) can be added to the mass matrix, as it is a
function of structural acceleration ( ) only.
(12.48)

where [Ma] = added mass matrix per-unit-length, and Ca is input as Cay and Caz
(OCDATA).

Finally, the right side of Equation 12.48 (p. 441) forms the hydrodynamic force vector per-unit-length
{Fhd/L}. The force consists of two normal components and one tangential component in the element
coordinate system. The tangential force includes the drag force only, as no inertial effect exists in
that direction.

(12.49)

where:

{Fhd/L}x,y,z = load vector per-unit-length due to hydrodynamic effects in the element x, y, and z
directions
CDy, CDz= coefficient of normal drag (input as CDy and CDz (OCTABLE))
Dey and Dez = outer pipe diameter when using PIPE288 or PIPE289. For BEAM188 or BEAM189,
this value is based on either the effective height or width (y or z, respectively) (output via the
SLIST command).
CMy, CMz = coefficient of inertia in the element normal (y and z) direction (input as CMy and CMz
(OCTABLE))
CT = coefficient of tangential (x) drag (input as CT (OCTABLE))

= tangential (x) and normal (y,z) particle velocity vector (length/time)

= tangential (x) and normal (y,z) structure velocity vector (length/time)

= normal particle acceleration vector (length/time2)

To calculate the relative velocities and accelerations, the velocity and acceleration of the fluid particle
and the structure must be available so that one can be subtracted from the other. The fluid particle
velocity and acceleration are computed using relationships such as Equation 12.28 (p. 430) through
Equation 12.31 (p. 430), as well as current effects. The structure velocity is available via the Newmark
time integration logic. (See Transient Analysis (p. 796).)

The integration points along the length of the element match the underlying element and are used
to generate the load vector. Integration points below the mud line are simply bypassed. For elements
intersecting the free surface, the integration points are distributed along the wet length only.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 441
Element Tools

The hydrodynamic coefficients (CD, CT, CM) may be defined as fixed numbers (OCTABLE), or as function
of Z level or Reynolds number (via multiple OCTABLE commands).

The definitions of Reynolds number can be expressed in the element coordinate system as:

where:

Ren, Ret = normal and tangential Reynolds number

= vector of normal and tangential particle velocity
μ = viscosity (input as VISC (MP))

Temperature-dependent effects can be included via the MP command, or by specifying the temper-
ature (Te) on the OCTABLE command with ocean current input (OCTYPE,CURR).

Morison's equation assumes that the particle motions are not substantially affected by the presence
of the structure (that is, that the structure diameter is much less than the wavelength).

12.8.7. Dynamic Pressure Head

When Kw = 0 through 3, the dynamic pressure is calculated and output in units of length. It is related
to the external pressure by expanding Equation 12.18 (p. 425) as follows:
(12.50)

where:

= external pressure at point of interest

= input external pressure (input as face 2 on SFE command)
= external fluid (ocean) density
= global Z coordinate at point of interest
= vertical offset from the global origin to the mean sea level (input on the OCDATA
command)
= dynamic pressure head (output as PHDYN) at point of interest

12.9. Nodal and Centroidal Data Evaluation

Area and volume elements normally compute results most accurately at the integration points. The
location of these data, which includes structural stresses, elastic and thermal strains, field gradients,
and fluxes, can then be moved to nodal or centroidal locations for further study. This is done with ex-

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
442 Published: 2023-10-23
 Nodal and Centroidal Data Evaluation

trapolation or interpolation, based on the element shape functions or simplified shape functions given
in Table 12.12: Assumed Data Variation of Stresses (p. 443).

Table 12.12: Assumed Data Variation of Stresses

No. Integration
Geometry Assumed Data Variation
Points
Triangles 3 a + bs + ct
Quadrilaterals 4 a + bs + ct + dst
Tetrahedra 4 a + bs + ct + dr
Hexahedra 8 a + bs + ct + dr + est + ftr + gsr + hstr

where:

a, b, c, d, e, f, g, h = coefficients
s, t, r = element natural coordinates

The extrapolation is done or the integration point results are simply moved to the nodes, based on the
user's request (input on the ERESX command). If material nonlinearities exist in an element, the least
squares fit can cause inaccuracies in the extrapolated nodal data or interpolated centroidal data. These
inaccuracies are normally minor for plasticity, creep, or swelling, but are more pronounced in elements
where an integration point may change status.

Adjustments and special cases exist, as follows:

• SOLID90 uses only the eight corner integration points.

• Uniform stress cases, like a constant stress triangle, do not require the above processing.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 443
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
444 Published: 2023-10-23
Chapter 13: Element Library
This chapter describes the theory underlying each Mechanical APDL element. The explanations are
augmented by references to other sections in this manual as well as external sources.

The table below the introductory figure of each element is complete, except that the Newton-Raphson
load vector is omitted. This load vector always uses the same shape functions and integration points
as the applicable stiffness, conductivity and/or coefficient matrix. Exceptions associated mostly with
some nonlinear line elements are noted with the element description.

13.1. Reserved for Future Use

This section is reserved for future use.

13.2. Not Documented

No detail or element available at this time.

13.3. Reserved for Future Use

This section is reserved for future use.

13.4. Reserved for Future Use

This section is reserved for future use.

13.5. SOLID5 - 3D Coupled-Field Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 445
Element Library

Matrix or Vector Shape Functions Integration

Points (p. 414)
Magnetic Potential Equation 11.232 (p. 387) 2x2x2
Coefficient Matrix
Electrical Conductivity Equation 11.231 (p. 387) 2x2x2
Matrix
Thermal Conductivity Matrix Equation 11.229 (p. 387) 2x2x2
Stiffness Matrix and Thermal Equation 11.217 (p. 387), 2x2x2
Expansion Load Vector Equation 11.218 (p. 387), and
Equation 11.219 (p. 387) or, if modified extra
shapes are included (KEYOPT(3) = 0),
Equation 11.233 (p. 388),
Equation 11.234 (p. 388), and
Equation 11.235 (p. 388)
Piezoelectric Coupling Same as combination of stiffness matrix and 2x2x2
Matrix conductivity matrix.
Specific Heat Matrix Same as conductivity matrix. Matrix is 2x2x2
diagonalized as described in 3D
Lines (p. 367)
Mass and Stress Stiffening Equation 11.217 (p. 387), 2x2x2
Matrices Equation 11.218 (p. 387), and
Equation 11.219 (p. 387)
Load Vector due to Imposed Same as coefficient or conductivity matrix 2x2x2
Thermal and Electric
Gradients, Heat Generation,
Joule Heating, Magnetic
Forces, Magnetism due to
Source Currents and
Permanent Magnets
Load Vector due to Same as stiffness or conductivity matrix 2x2x2
Convection Surfaces and specialized to the surface.
Pressures

References: Wilson([39] (p. 926)), Taylor([50] (p. 927)), Coulomb([77] (p. 929)), Mayergoyz([120] (p. 931)),
Gyimesi([142] (p. 932),[150] (p. 933))

13.5.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. Heat Flow (p. 207) describes the derivation of thermal element matrices and load
vectors as well as heat flux evaluations. Derivation of Electromagnetic Matrices (p. 176) discusses the
scalar potential method, which is used by this element. Piezoelectrics (p. 334) discusses the piezoelectric
capability used by the element.

13.6. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
446 Published: 2023-10-23
 LINK11 - Linear Actuator

13.7. Reserved for Future Use

This section is reserved for future use.

13.8. Reserved for Future Use

This section is reserved for future use.

13.9. Reserved for Future Use

This section is reserved for future use.

13.10. Reserved for Future Use

This section is reserved for future use.

13.11. LINK11 - Linear Actuator

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness and Damping Matrices Equation 11.6 (p. 367) None
Mass Matrix None (lumped mass formulation) None
Stress Stiffness Matrix Equation 11.7 (p. 367) and None
Equation 11.8 (p. 367)

13.11.1. Assumptions and Restrictions

The element is not capable of carrying bending or twist loads. The force is assumed to be constant
over the entire element.

13.11.2. Element Matrices and Load Vector

All element matrices and load vectors are described below. They are generated in the element co-
ordinate system and are then converted to the global coordinate system. The element stiffness matrix
is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 447
Element Library

(13.1)

where:

K = element stiffness (input as K on R command)

The element mass matrix is:

(13.2)

where:

M = total element mass (input as M on R command)

The element damping matrix is:

(13.3)

where:

C = element damping (input as C on R command)

The element stress stiffness matrix is:

(13.4)

where:

F = the axial force in the element (output as FORCE)

L = current element length (output as CLENG)

The element load vector is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
448 Published: 2023-10-23
 Reserved for Future Use

(13.5)

where:

The applied force vector is:

(13.6)

where:

F' = applied force thru surface load input using the PRES label

The Newton-Raphson restoring force vector is:

(13.7)

13.11.3. Force, Stroke, and Length

The element spring force is determined from
(13.8)

where:

F = element spring force (output as FORCE)

SA = applied stroke (output as STROKE) thru surface load input using the PRES label
SM = computed or measured stroke (output as MSTROKE)

The lengths, shown in the figure at the beginning of this section, are:

Lo = initial length (output as ILEN)

Lo + SM = current length (output as CLEN)

13.12. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 449
Element Library

13.13. PLANE13 - 2D Coupled-Field Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Magnetic Potential Quad Equation 11.125 (p. 377) 2x2
Coefficient Matrix; and
Permanent Magnet and 1 if planar
Applied Current Load Triangle Equation 11.104 (p. 376) 3 if axisymmetric
Vector
Thermal Conductivity Quad Equation 11.130 (p. 377) Same as coefficient
Matrix Triangle Equation 11.109 (p. 376) matrix
Equation 11.122 (p. 377) and
Equation 11.123 (p. 377) and, if
modified extra shapes are included
Stiffness Matrix; and Quad (KEYOPT(2) = 0) and element has 4
unique nodes) Same as coefficient
Thermal and Magnetic
Equation 11.135 (p. 378) and matrix
Force Load Vector
Equation 11.136 (p. 378).
Equation 11.101 (p. 376) and
Triangle
Equation 11.102 (p. 376)
Equation 11.122 (p. 377) and
Quad
Mass and Stress Stiffness Equation 11.123 (p. 377) Same as coefficient
Matrices Equation 11.101 (p. 376) and matrix
Triangle
Equation 11.102 (p. 376)
Same as conductivity matrix. Matrix is
Same as coefficient
Specific Heat Matrix diagonalized as described in Lumped
matrix
Matrices (p. 422)
Equation 11.125 (p. 377) and
Quad
Damping (Eddy Current) Equation 11.131 (p. 377) Same as coefficient
Matrix Equation 11.104 (p. 376) and matrix
Triangle
Equation 11.110 (p. 376)
Convection Surface Same as conductivity matrix, specialized to the
2
Matrix and Load Vector surface
Pressure Load Vector Same as mass matrix specialized to the face 2

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
450 Published: 2023-10-23
 COMBIN14 - Spring-Damper

Load Type Distribution

Current Density Bilinear across element
Current Phase Angle Bilinear across element
Heat Generation Bilinear across element
Pressure Linear along each face

References: Wilson([39] (p. 926)), Taylor, et al.([50] (p. 927)), Silvester, et al.([73] (p. 928)),Weiss, et
al.([95] (p. 930)), Garg, et al.([96] (p. 930))

13.13.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. Heat Flow (p. 207) describes the derivation of thermal element matrices and load
vectors as well as heat flux evaluations. Derivation of Electromagnetic Matrices (p. 176) and Electro-
magnetic Field Evaluations (p. 183) discuss the magnetic vector potential method, which is used by
this element. The diagonalization of the specific heat matrix is described in Lumped Matrices (p. 422).

13.14. COMBIN14 - Spring-Damper

Matrix or Vector Shape Functions[1] Integration

Option
Points (p. 414)
Stiffness and Damping Longitudinal Equation 11.6 (p. 367) None
Matrices Torsional Equation 11.18 (p. 368) None
Stress Stiffening Matrix Equation 11.7 (p. 367), and None
Longitudinal
Equation 11.8 (p. 367)

1. There are no shape functions used if the element is input on a one DOF per node basis (KEYOPT(2)
> 0) as the nodes may be coincident.

13.14.1. Types of Input

COMBIN14 essentially offers two types of elements, selected with KEYOPT(2).

1. Single DOF per node (KEYOPT(2) > 0). The orientation is defined by the value of KEYOPT(2) and
the two nodes are usually coincident.

2. Multiple DOFs per node (KEYOPT(2) = 0). The orientation is defined by the location of the two
nodes; therefore, the two nodes must not be coincident.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 451
Element Library

13.14.2. Stiffness Pass

Consider the case of a single DOF per node first. The orientation is selected with KEYOPT(2). If KEY-
OPT(2) = 7 (pressure) or = 8 (temperature), the concept of orientation does not apply. The form of
the element stiffness and damping matrices are:
(13.9)

(13.10)

where:

= stiffness (input as K on R command)

= constant damping coefficient (input as CV1 on R command)

= nonlinear damping coefficient (input as CV2 on R command)
= relative velocity between nodes computed from the nodal Newmark velocities

In full harmonic, full transient, and static analyses, when the stiffness and damping real constants are
input as table parameters, the stiffness ( ) and damping ( ) coefficients are interpolated at each
frequency or time step.

Next, consider the case of multiple DOFs per node. Only the case with three DOFs per node will be
discussed, as the case with two DOFs per node is simply a subset. The stiffness, damping, and stress
stiffness matrices in element coordinates are developed as:

(13.11)

(13.12)

(13.13)

where subscript refers to element coordinates.

and where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
452 Published: 2023-10-23
 COMBIN14 - Spring-Damper

= force in element from previous iteration

= distance between the two nodes

There are some special notes that apply to the torsion case (KEYOPT(3) = 1):

1. Rotations are simply treated as a vector quantity. No other effects (including displacements) are
implied.

2. In a large rotation problem (NLGEOM,ON), the coordinates do not get updated, as the nodes only
rotate. (They may translate on other elements, but this does not affect COMBIN14 with KEYOPT(3)
= 1). Therefore, there are no large rotation effects.

3. Similarly, as there is no axial force computed, no stress stiffness matrix is computed.

13.14.3. Output Quantities

The stretch is computed as:

(13.14)

where:

= coordinates in global Cartesian coordinates

= displacements in global Cartesian coordinates
= displacements in nodal Cartesian coordinates (UX, UY, UZ)
= rotations in nodal Cartesian coordinates (ROTX, ROTY, ROTZ)
= pressure (PRES)
= temperature (TEMP)

If KEYOPT(3) = 1 (torsion), the expression for has rotation instead of translations, and is output
as TWIST. Next, the static force (or torque) is computed:
(13.15)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 453
Element Library

= static force (or torque) (output as FORC (TORQ if KEYOPT(3) = 1))

Finally, if a nonlinear transient dynamic (ANTYPE,TRANS, with TIMINT,ON) analysis is performed, a

damping force is computed:
(13.16)

where:

= damping force (or torque) (output as DAMPING FORCE (DAMPING TORQUE if KEYOPT(3)
= 1))
= relative velocity

relative velocity is computed using Equation 13.14 (p. 453), where the nodal displacements , etc.
are replaced with the nodal Newmark velocities , etc.

13.15. Reserved for Future Use

This section is reserved for future use.

13.16. Reserved for Future Use

This section is reserved for future use.

13.17. Reserved for Future Use

This section is reserved for future use.

13.18. Reserved for Future Use

This section is reserved for future use.

13.19. Reserved for Future Use

This section is reserved for future use.

13.20. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
454 Published: 2023-10-23
 MASS21 - Structural Mass

13.21. MASS21 - Structural Mass

Matrix or Vector Shape Functions Integration

Points (p. 414)
Mass Matrix None None

In the element coordinate system, the element mass matrix is:

(13.17)

where:

ρ = density (input as DENS on MP command)

where a', b', c', d', e', and f' are user input (input on the R command) in the locations shown in the fol-
lowing table:

KEYOPT(3) = 0 KEYOPT(3) = 2 KEYOPT(3) = 3 KEYOPT(3) = 4

a' 1 1 1 1
b' 2 1 1 1
c' 3 1 - -
d' 4 - - -

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 455
Element Library

KEYOPT(3) = 0 KEYOPT(3) = 2 KEYOPT(3) = 3 KEYOPT(3) = 4

e' 5 - - -
f' 6 - 2 -

For the mass summary, only the first real constant is used, regardless of which option of KEYOPT(3) is
used. The precise mass summary used for three-dimensional models as well as analyses with inertial
relief use the complete matrix.

13.22. Reserved for Future Use

This section is reserved for future use.

13.23. Reserved for Future Use

This section is reserved for future use.

13.24. Reserved for Future Use

This section is reserved for future use.

13.25. PLANE25 - Axisymmetric-Harmonic 4-Node Structural Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equations
Equation 11.164 (p. 381),
Equation 11.165 (p. 381) , and
Equation 11.166 (p. 381) or if
Stiffness Matrix and modified extra shape functions
Quad 2x2
Thermal Load Vector are included (KEYOPT(2) = 0)
and element has 4 unique
nodes: Equation 11.168 (p. 381),
Equation 11.169 (p. 381) , and
Equation 11.170 (p. 381)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
456 Published: 2023-10-23
 PLANE25 - Axisymmetric-Harmonic 4-Node Structural Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.156 (p. 381),
Triangle Equation 11.157 (p. 381) , and 3
Equation 11.158 (p. 381)
Equation 11.122 (p. 377),
Quad Equation 11.123 (p. 377) , and 2x2
Mass and Stress Stiffness Equation 11.124 (p. 377)
Matrices Equation 11.101 (p. 376),
Triangle Equation 11.102 (p. 376) , and 3
Equation 11.103 (p. 376)
Same as stress stiffness matrix, specialized to
Pressure Load Vector 2
the surface

Load Type Distribution

Element Temperature Bilinear across element, harmonic around circumference
Nodal Temperature Bilinear across element, harmonic around circumference
Pressure Linear along each face, harmonic around circumference

Reference: Wilson([39] (p. 926)), Zienkiewicz([40] (p. 926)), Taylor([50] (p. 927))

13.25.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.25.2. Assumptions and Restrictions

The material properties are assumed to be constant around the entire circumference, regardless of
temperature dependent material properties or loading. For (input as MODE on MODE command)
> 0, the extreme values for combined stresses are obtained by computing these stresses at every 10/
degrees and selecting the extreme values.

13.25.3. Use of Temperature

In general, temperatures have two effects on a stress analysis:

1. Temperature dependent material properties.

2. Thermal expansion

In the case of = 0, there is no conflict between these two effects. However, if > 0, questions arise.
As stated in the assumptions, the material properties may not vary around the circumference, regardless
of the temperature. That is, one side cannot be soft and the other side hard. The input temperature
for > 0 varies sinusoidally around the circumference. As no other temperatures are available to the
element, the material properties are evaluated at Tref (input on TREF command). The input temperature
can therefore be used to model thermal bending. An approximate application of this would be a
chimney subjected to solar heating on one side only. A variant on this basic procedure is provided

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 457
Element Library

by the temperature KEYOPT (KEYOPT(3) for PLANE25). This variant provides that the input temperatures
be used only for material property evaluation rather than for thermal bending. This second case requires
that α x, α y, and α z (input on MP commands) all be input as zero. An application of the latter case
is a chimney, which is very hot at the bottom and relatively cool at the top, subjected to a wind load.

13.26. Not Documented

No detail or element available at this time.

13.27. MATRIX27 - Stiffness, Damping, or Mass Matrix

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness, Mass, and Damping
None None
Matrices

13.27.1. Assumptions and Restrictions

All MATRIX27 matrices should normally be positive definite or positive semidefinite (see Positive
Definite Matrices (p. 421) for definition) in order to be valid structural matrices. The only exception to
this occurs when other (positive definite) matrices dominate the involved degrees of freedom and/or
sufficient degrees of freedom are removed by way of imposed constraints, so that the total (structure)
matrix is positive definite.

13.28. Reserved for Future Use

This section is reserved for future use.

13.29. FLUID29 - 2D Acoustic Fluid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Fluid Stiffness and Mass
Equation 11.129 (p. 377) 2x2
Matrices

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
458 Published: 2023-10-23
 LINK31 - Radiation Link

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Coupling Stiffness, Mass, Equation 11.122 (p. 377), Equation 11.123 (p. 377)
and Damping Matrices , and Equation 11.129 (p. 377) specialized to the 2
(fluid-structure interface) interface

13.29.1. Other Applicable Sections

Acoustics (p. 241) describes the derivation of acoustic element matrices and load vectors.

13.30. FLUID30 - 3D Acoustic Fluid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Fluid Stiffness and Mass
Equation 11.226 (p. 387) 2x2x2
Matrices
Equation 11.217 (p. 387), Equation 11.218 (p. 387),
Coupling Stiffness and
Equation 11.219 (p. 387), and
Mass Matrices 2x2
Equation 11.226 (p. 387) specialized to the
(fluid-structure interface)
interface
Fluid Damping Matrix No shape functions are used. Instead, the area
(fluid at fluid-structure associated with each node at the interface is None
interface) computed for the damping to act upon.

13.30.1. Other Applicable Sections

Acoustics (p. 241) describes the derivation of acoustic element matrices and load vectors.

13.31. LINK31 - Radiation Link

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 459
Element Library

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix None (nodes may be coincident) None

13.31.1. Standard Radiation (KEYOPT(3) = 0)

The two-surface radiation equation (from Equation 6.14 (p. 211)) that is solved (iteratively) is:
(13.18)

where:

Q = heat flow rate between nodes I and J (output as HEAT RATE)

σ = Stefan-Boltzmann constant (input as SBC on R command)
ε = emissivity (input as EMISSIVITY on R or EMIS on MP command)
F = geometric form factor (input as FORM FACTOR on R command)
A = area of element (input as AREA on R command)
TI, TJ = absolute temperatures at nodes I and J

The program uses a linear equation solver. Therefore, Equation 13.18 (p. 460) is expanded as:
(13.19)

and then rewritten as:

(13.20)

where the subscripts n and n-1 refer to the current and previous iterations, respectively. It is then
recast into finite element form:

(13.21)

with
(13.22)

13.31.2. Empirical Radiation (KEYOPT(3) = 1)

The basic equation is:
(13.23)

instead of Equation 13.18 (p. 460). This form leads to

(13.24)

instead of Equation 13.22 (p. 460). And, hence the matrix Equation 13.21 (p. 460) becomes:

(13.25)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
460 Published: 2023-10-23
 LINK33 - 3D Conduction Bar

13.31.3. Solution
If the emissivity is input on a temperature dependent basis, Equation 13.22 (p. 460) is rewritten to be:
(13.26)

where:

Toff = offset temperature (input on TOFFST command)

Equation 13.24 (p. 460) is handled analogously.

13.32. Reserved for Future Use

This section is reserved for future use.

13.33. LINK33 - 3D Conduction Bar

Matrix or Vector Element configuration Shape Functions Integration Points (p. 414)
Conductivity and Specific Heat 2-node (KEYOPT(4) = 0) Equation 11.13 (p. 368) None
Matrices; and Heat Generation 3-node (KEYOPT(4) = 1) Equation 11.25 (p. 368) 3
Load Vector

13.33.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of thermal element matrices and load vectors as well as
heat flux evaluations.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 461
Element Library

13.33.2. Matrices and Load Vectors for 2-node element

The conductivity matrix is:
(13.27)

where:

A = area (input as AREA on R command)

= conductivity along longitudinal s direction (input as KXX on MP command)
L = distance between nodes

The specific heat matrix is:

(13.28)

where:

ρ = density (input as DENS on MP command)

Cp = specific heat (input as C on MP command)

This specific heat matrix is a diagonal matrix with each diagonal being the sum of the corresponding
row of a consistent specific heat matrix. The heat generation load vector is:
(13.29)

where:

13.33.3. Matrices and Load Vectors for 3-node element

A closed-form representation of the matrices and load vector is not available since integration rules
are used to evaluate them.

For the 3-node element, the specific heat matrix is a full matrix (not a diagonal matrix).

13.33.4. Output
The output is computed as:

(13.30)

and
(13.31)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
462 Published: 2023-10-23
 LINK34 - Convection Link

q = average thermal flux (output as THERMAL FLUX)

TI = temperature at node I
TJ = temperature at node J
Q = average heat rate (output as HEAT RATE)

13.34. LINK34 - Convection Link

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix and
Heat Generation Load None (nodes may be coincident) None
Vector

13.34.1. Conductivity Matrix

The element conductivity (convection) matrix is
(13.32)

where:

A = area over which element acts (input as AREA on R command)

The effective film coefficient is:

(13.33)

where:

Cc = user input constant (input as CC on R command)

The partial film coefficient term is:

(13.34)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 463
Element Library

where:

TB = bulk temperature (input as TBULK on SFE command)

H(x) = alternate film coefficient (input on MP,HF command for material x)

me = material number for this element (input on MAT command)

ΔTp = Tp,J - Tp,I

Tp,J = temperature from previous iteration at node J
n = exponent on temperature change (input as EN on R command)

ΔTp must be thought of as unitless, even though it is obviously derived from temperatures.

The heat generation load vector is:

(13.35)

where:

L = distance between nodes

13.34.2. Output
The output is computed as:
(13.36)

where:

Q = heat rate (output as HEAT RATE)

TI = temperature at node I
TJ = temperature at node J

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
464 Published: 2023-10-23
 SOURC36 - Current Source

13.35. PLANE35 - 2D 6-Node Triangular Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix and
Heat Generation Load Equation 11.119 (p. 377) 6
Vector
Equation 11.119 (p. 377). If KEYOPT(1) = 1, matrix
Specific Heat Matrix is diagonalized as described in Lumped 6
Matrices (p. 422)
Convection Surface
Equation 11.119 (p. 377), specialized to the face 2
Matrix and Load Vector

13.35.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of thermal element matrices and load vectors as well as
heat flux evaluations.

13.36. SOURC36 - Current Source

y y

x
CUR x
J
J
I
DY z
I K
K z
DY CUR
DZ

DZ
a) Type 1 - Coil b) Type 2 - Bar
y DY
CUR
J x

K
DZ z

c) Type 3 - Arc

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 465
Element Library

13.36.1. Description
The functionality of SOURC36 is basically one of user convenience. It provides a means of specifying
the necessary data to evaluate the Biot-Savart integral (Equation 5.18 (p. 167)) for the simple current
source configurations, coil, bar and arc. The magnetic field {Hs} that results from this evaluation in
turn becomes a load for the magnetic scalar potential elements (SOLID5, SOLID96 and SOLID98) as
discussed in Electromagnetics (p. 163).

13.37. COMBIN37 - Control

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix None (nodes may be coincident) None
Mass Matrix None (lumped mass formulation) None
Damping Matrix None None

13.37.1. Element Characteristics

COMBIN37 is a nonlinear, 1D element with two active nodes and one or two control nodes. The element
has spring-damper-sliding capability similar to COMBIN40. The degree of freedom (DOF) for the active
nodes is selected using KEYOPT(3) and the DOF for the control nodes is selected using KEYOPT(2).

The action of the element in the structure is based upon the value of the control parameter (Pcn)
(explained later), On and Of (input as ONVAL and OFFVAL on R command), and the behavior switches
KEYOPT(4) and (5). Figure 13.1: Element Behavior (p. 467) illustrates the behavior of one of the more
common modes of operation of the element. It is analogous to the normal home thermostat during
the winter.

The behavior of all possible combinations of KEYOPT(4) and (5) values is summarized in the following
table. Pcn represents the control parameter (output as CONTROL PARAM). The element is active where
the figure indicates on, and inactive where it indicates off. For some options, the element may be
either on or off for Pcn between On and Of, depending upon the last status change.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
466 Published: 2023-10-23
 COMBIN37 - Control

Figure 13.1: Element Behavior

KEYOPT(4) = 0, KEYOPT(5) = 1, and Of > On

KEYOPT(4) = 0, KEYOPT(5) = 0, Of On: KEYOPT(4) = 0, KEYOPT(5) = 0, Of > On:

KEYOPT(4) = 0, KEYOPT(5) = 1, Of On: KEYOPT(4) = 0, KEYOPT(5) = 1, Of > On:

[1]
KEYOPT(4) = 1, KEYOPT(5) = 0: KEYOPT(4) = 1, KEYOPT(5) = 1:

1. Analogous to Figure 13.1: Element Behavior (p. 467)

13.37.2. Element Matrices

When the element status is ON, the element matrices are:
(13.37)

(13.38)

(13.39)

where:

ko = stiffness (input as STIF on R command)

MI = mass at node I (input as MASI on R command)
MJ = mass at node J (input as MASJ on R command)
Co = damping constant (input as DAMP on R command)

When the element status is OFF, all element matrices are set to zero.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 467
Element Library

13.37.3. Adjustment of Real Constants

If KEYOPT(6) > 0, a real constant is to be adjusted as a function of the control parameter as well as
other real constants. Specifically,
(13.40)
(13.41)
(13.42)
(13.43)
(13.44)
(13.45)
(13.46)
(13.47)

where:

FA = element load (input as AFORCE ON R command)

FS = slider force (input as FSLIDE on RMORE command)
C1, C2, C3, C4 = input constants (input as C1, C2, C3, and C4 on RMORE command)
Pcn = control parameter (defined below)
f1 = function defined by subroutine USERRC

If (or FS, if KEYOPT(6) ≠ 8) is less than zero, it is reset to zero.

13.37.4. Evaluation of Control Parameter

The control parameter is defined as:

(13.48)

where:

t = time (input on TIME command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
468 Published: 2023-10-23
 FLUID38 - Dynamic Fluid Coupling

u = degree of freedom as selected by KEYOPT(2)

The assumed value of the control parameter for the first iteration ( ) is defined as:

(13.49)

where:

St = constant defining starting status where: 1 means ON, -1 means OFF (input as START on
R command)
TUNIF = uniform temperature (input on BFUNIF command)

13.38. FLUID38 - Dynamic Fluid Coupling

Integration
Matrix or Vector Shape Functions
Points (p. 414)

Mass Matrix None

Damping Matrix Not defined None

Reference: Fritz([12] (p. 925))

13.38.1. Description
This element is used to represent a dynamic coupling between two points of a structure. The coupling
is based on the dynamic response of two points connected by a constrained mass of fluid. The points

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 469
Element Library

represent the centerlines of concentric cylinders. The fluid is contained in the annular space between
the cylinders. The cylinders may be circular or have an arbitrary cross-section. The element has two
DOFs per node: translations in the nodal x and z directions. The axes of the cylinders are assumed to
be in the nodal y directions. These orientations may be changed with KEYOPT(6).

13.38.2. Assumptions and Restrictions

1. The motions are assumed to be small with respect to the fluid channel thickness.

2. The fluid is assumed to be incompressible.

3. Fluid velocities should be less than 10% of the speed of sound in the fluid.

4. The flow channel length should be small compared to the wave length for propagating vibratory
disturbances (less than about 10%), in order to avoid the possibility of standing wave effects.

13.38.3. Mass Matrix Formulation

The mass matrix formulation used in the element is of the following form:

(13.50)

The m values are dependent upon the KEYOPT(3) value selected. For KEYOPT(3) = 0 (concentric cylinder
case):
(13.51)
(13.52)
(13.53)

where:

ρ = fluid mass density (input as DENS on MP command)

R1 = radius of inner cylinder (input as R1 on R command)
R2 = radius of outer cylinder (input as R2 on R command)
L = length of cylinders (input as L on R command)

Note that the shape functions are similar to that for PLANE25 with MODE = 1. The element mass used
in the evaluation of the total structure mass is .

For KEYOPT(3) = 2, which is a generalization of the above cylindrical values but for different geometries,
the m values are as follows:
(13.54)
(13.55)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
470 Published: 2023-10-23
 FLUID38 - Dynamic Fluid Coupling

(13.56)
(13.57)
(13.58)
(13.59)

where:

M1 = mass of fluid displaced by the inner boundary (Boundary 1) (input as M1 on R command)

M2 = mass of fluid that could be contained within the outer boundary (Boundary 2) in absence
of the inner boundary (input as M2 on R command)
Mhx, Mhz = hydrodynamic mass for motion in the x and z directions, respectively (input as
MHX and MHZ on R command)

The element mass used in the evaluation of the total structure mass is M2 - M1.

The lumped mass option (LUMPM,ON) is not available.

13.38.4. Damping Matrix Formulation

The damping matrix formulation used in the element is of the following form:

(13.60)

The c values are dependent upon the KEYOPT(3) value selected. For KEYOPT(3) = 0:
(13.61)
(13.62)
(13.63)
(13.64)

where:

Wx, Wz = estimate of resonant frequencies in the x and z response directions, respectively

(input as WX, WZ on RMORE command)
f = Darcy friction factor for turbulent flow (input as F on R command)
Δx, Δz = estimate of peak relative amplitudes between inner and outer boundaries for the
x and z motions, respectively (input as DX, DZ on R command)

For KEYOPT(3) = 2, the c values are as follows:

(13.65)
(13.66)
(13.67)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 471
Element Library

(13.68)

where:

Cx, Cz = flow and geometry constants for the x and z motions, respectively (input as CX, CZ
on RMORE command)

13.39. COMBIN39 - Nonlinear Spring

Integration
Matrix or Vector Option Shape Functions[1]
Points (p. 414)
Longitudinal Equation 11.15 (p. 368) None
Stiffness Matrix
Torsional Equation 11.18 (p. 368) None
Equation 11.7 (p. 367) and
Stress Stiffening Matrix Longitudinal None
Equation 11.8 (p. 367)

1. There are no shape functions used if the element is input as a one DOF per node basis (KEYOPT(4)
= 0) as the nodes are coincident.

13.39.1. Input
The user explicitly defines the force-deflection curve for COMBIN39 by the input of discrete points of
force versus deflection. Up to 250 points on the curve can be defined, and are entered as real constants.
The input curve must pass through the origin and must lie within the unshaded regions, if KEYOPT(1)
= 1.

Figure 13.2: Input Force-Deflection Curve

The input deflections must be given in ascending order, with the minimum change of deflection of:
(13.69)

where:

ui = input deflections (input as D1, D2, ... D250 on R or RMORE commands)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
472 Published: 2023-10-23
 COMBIN39 - Nonlinear Spring

umax = most positive input deflection

umin = most negative input deflection

13.39.2. Element Stiffness Matrix and Load Vector

During the stiffness pass of a given iteration, COMBIN39 will use the results of the previous iteration
to determine which segment of the input force-deflection curve is active. The stiffness matrix and
load vector of the element are then:
(13.70)

(13.71)

where:

Ktg = slope of active segment from previous iteration (output as SLOPE)

F1 = force in element from previous iteration (output as FORCE)

If KEYOPT(4) > 0, Equation 13.70 (p. 473) and Equation 13.71 (p. 473) are expanded to 2 or 3 dimensions.

During the stress pass, the deflections of the current equilibrium iteration will be examined to see
whether a different segment of the force-deflection curve should be used in the next equilibrium it-
eration.

Figure 13.3: Stiffness Computation

13.39.3. Choices for Element Behavior

If KEYOPT(2) = 0 and if no force-deflection points are input for deflection less than zero, the points
in the first quadrant are reflected through the origin (Figure 13.4: Input Force-Deflection Curve Reflected
Through Origin (p. 474)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 473
Element Library

Figure 13.4: Input Force-Deflection Curve Reflected Through Origin

If KEYOPT(2) = 1, there will be no stiffness for the deflection less than zero (Figure 13.5: Force-Deflection
Curve with KEYOPT(2) = 1 (p. 474)).

Figure 13.5: Force-Deflection Curve with KEYOPT(2) = 1

If KEYOPT(1) = 0, COMBIN39 is conservative. This means that regardless of the number of loading re-
versals, the element will remain on the originally defined force-deflection curve, and no energy loss
will occur in the element. This also means that the solution is not path-dependent. If, however, KEY-
OPT(1) = 1, the element is nonconservative. With this option, energy losses can occur in the element,
so that the solution is path-dependent. The resulting behavior is illustrated in Figure 13.6: Nonconser-
vative Unloading (KEYOPT(1) = 1) (p. 474).

Figure 13.6: Nonconservative Unloading (KEYOPT(1) = 1)

When a load reversal occurs, the element will follow a new force-deflection line passing through the
point of reversal and with slope equal to the slope of the original curve on that side of the origin (0+
or 0-). If the reversal does not continue past the force = 0 line, reloading will follow the straight line
back to the original curve (Figure 13.7: No Origin Shift on Reversed Loading (KEYOPT(1) = 1) (p. 475)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
474 Published: 2023-10-23
 COMBIN39 - Nonlinear Spring

Figure 13.7: No Origin Shift on Reversed Loading (KEYOPT(1) = 1)

If the reversal continues past the force = 0 line, a type of origin shift occurs, and reloading will follow
a curve that has been shifted a distance uorig (output as UORIG) (Figure 13.8: Origin Shift on Reversed
Loading (KEYOPT(1) = 1) (p. 475)).

Figure 13.8: Origin Shift on Reversed Loading (KEYOPT(1) = 1)

A special option (KEYOPT(2) = 2) is included to model crushing behavior. With this option, the element
will follow the defined tensile curve if it has never been loaded in compression. Otherwise, it will
follow a reflection through the origin of the defined compressive curve (Figure 13.9: Crush Option
(KEYOPT(2) = 2) (p. 475)).

Figure 13.9: Crush Option (KEYOPT(2) = 2)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 475
Element Library

13.40. COMBIN40 - Combination

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness, Mass, and Damping
None (nodes may be coincident) None
Matrices

13.40.1. Characteristics of the Element

The force-deflection relationship for the combination element under initial loading is as shown below
(for no damping).

Figure 13.10: Force-Deflection Relationship

where:

F1 = force in spring 1 (output as F1)

F2 = force in spring 2 (output as F2)
K1 = stiffness of spring 1 (input as K1 on R command)
K2 = stiffness of spring 2 (input as K2 on R command)
ugap = initial gap size (input as GAP on R command) (if zero, gap capability removed)
uI = displacement at node I
uJ = displacement at node J
FS = force required in spring 1 to cause sliding (input as FSLIDE on R command)

13.40.2. Element Matrices for Structural Applications

The element mass matrix is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
476 Published: 2023-10-23
 COMBIN40 - Combination

(13.72)

(13.73)

(13.74)

where:

M = element mass (input as M on R command)

If the gap is open during the previous iteration, all other matrices and load vectors are null vectors.
Otherwise, the element damping matrix is:
(13.75)

where:

c = damping constant (input as C on R command)

The element stiffness matrix is:

(13.76)

where:

and the element Newton-Raphson load vector is:

(13.77)

F1 and F2 are the current forces in the element.

13.40.3. Determination of F1 and F2 for Structural Applications

1. If the gap is open,
(13.78)

If no sliding has taken place, F1 = F2 = 0.0. However, if sliding has taken place during unidirectional
motion,
(13.79)

and thus
(13.80)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 477
Element Library

us = amount of sliding (output as SLIDE)

2. If the gap is closed and the slider is sliding,

(13.81)

and
(13.82)

where:

u2 = uJ - uI + ugap = output as STR2

3. If the gap is closed and the slider is not sliding, but had slid before,
(13.83)

where:

u1 = u2 - us = output as STR1

and
(13.84)

13.40.4. Thermal Analysis

The above description refers to structural analysis only. When this element is used in a thermal ana-
lysis, the conductivity matrix is [Ke], the specific heat matrix is [Ce] and the Newton-Raphson load
vector is , where F1 and F2 represent heat flow. The mass matrix [M] is not used. The gap size
ugap is the temperature difference. Sliding, Fslide, is the element heat flow limit for conductor K1.

13.41. Reserved for Future Use

This section is reserved for future use.

13.42. Reserved for Future Use

This section is reserved for future use.

13.43. Reserved for Future Use

This section is reserved for future use.

13.44. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
478 Published: 2023-10-23
 INFIN47 - 3D Infinite Boundary

13.45. Reserved for Future Use

This section is reserved for future use.

13.46. Reserved for Future Use

This section is reserved for future use.

13.47. INFIN47 - 3D Infinite Boundary

Matrix or Vector Shape Functions Integration Points (p. 414)

None on the boundary
element IJK itself,
however, 16-point 1D
Gaussian quadrature is
applied for some of the
integration on each of
Magnetic Potential
the edges IJ, JK, and KI
Coefficient Matrix
of the infinite elements
or Thermal
IJML, JKNM, and KILN
Conductivity Matrix
(see Figure 13.11: A
Semi-infinite Boundary
Element Zone and the
Corresponding
Boundary Element
IJK (p. 480))

Reference: Kaljevic', et al.([131] (p. 932))

13.47.1. Introduction
This boundary element (BE) models the exterior infinite domain of the far-field magnetic and thermal
problems. This element is to be used in combination with 3D scalar potential solid elements, and can
have magnetic scalar potential (MAG), or temperature (TEMP) as the DOF.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 479
Element Library

13.47.2. Theory
The formulation of this element is based on a first order triangular infinite boundary element (IBE),
but the element can be used as a 4-node quadrilateral as well. For unbounded field problems, the
model domain is set up to consist of an interior volumetric finite element domain, ΩF, and a series
of exterior volumetric BE subdomains, ΩB, as shown in Figure 13.11: A Semi-infinite Boundary Element
Zone and the Corresponding Boundary Element IJK (p. 480). Each subdomain, ΩB, is treated as an or-
dinary BE domain consisting of five segments: the boundary element IJK, infinite elements IJML, JKNM
and KILN, and element LMN; element LMN is assumed to be located at infinity.

Figure 13.11: A Semi-infinite Boundary Element Zone and the Corresponding Boundary Element
IJK

The approach used here is to write BE equations for ΩB, and then convert them into equivalent load
vectors for the nodes I, J and K. The procedure consists of four steps that are summarized below (see
(Kaljevic', et al.[131] (p. 932)) for details).

First, a set of boundary integral equations is written for ΩB. To achieve this, the potential (or temper-
ature) and its normal derivatives (fluxes) are interpolated on the triangle IJK (Figure 13.11: A Semi-in-
finite Boundary Element Zone and the Corresponding Boundary Element IJK (p. 480)) by linear shape
functions:
(13.85)
(13.86)

where:

φ = potential (or temperature)

NI, NJ, NK = linear shape functions defined earlier

φI, φJ, φK = nodal potentials (or temperatures)
qnI, qnJ, qnK = nodal normal derivatives (or fluxes)
n = normal to the surface IJK

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
480 Published: 2023-10-23
 INFIN47 - 3D Infinite Boundary

Figure 13.12: Infinite Element IJML and the Local Coordinate System

Over an infinite element, such as IJML (Figure 13.12: Infinite Element IJML and the Local Coordinate
System (p. 481)), the dependent variables, that is, potentials (or temperatures) and their normal deriv-
atives (fluxes) are respectively assumed to be (Figure 13.12: Infinite Element IJML and the Local Co-
ordinate System (p. 481)):

(13.87)

(13.88)

where:

qτI, qτJ = nodal (nodes I and J) normal derivatives for infinite element IJML
s = a variable length from node I towards node J
LIJ = length of edge IJ
ρ = radial distance from the origin of the local coordinate system O to the edge IJ
r = radial distance from the edge IJ towards infinity
β = variable angle from x-axis for local polar coordinate system
τ = normal to infinite elements IJML

The boundary integral equations for ΩB are now written as:

(13.89)

where:

c(ξ) = jump term in boundary element method

(x,ξ) = field and source points, respectively

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 481
Element Library

r = distance between field and source points

The integrations in Equation 13.89 (p. 481) are performed in closed form on the boundary element
IJK. The integrations on the infinite elements IJML, JKNM and KILN in the 'r' direction (Figure 13.12: In-
finite Element IJML and the Local Coordinate System (p. 481)) are also performed in closed form.
However, a 16-point Gaussian quadrature rule is used for the integrations on each of the edges IJ, JK
and KI on the infinite elements.

Second, in the absence of a source or sink in ΩB, the flux q(r) is integrated over the boundary ΓB of
ΩB and set to zero:

(13.90)

Third, geometric constraint conditions that exist between the potential φ (or temperature) and its
derivatives and at the nodes I, J and K are written. These conditions would express the
fact that the normal derivative qn at the node I, say, can be decomposed into components along the
normals to the two infinite elements IJML and KILN meeting at I and along OI.

Fourth, the energy flow quantity from ΩB is written as:

(13.91)

This energy flow is equated to that due to an equivalent nodal force vector {F} defined below.

The four steps mentioned above are combined together to yield, after eliminating qn and qτ,
(13.92)

where:

[K] = 3 x 3 equivalent unsymmetric element coefficient matrix

{φ} = 3 x 1 nodal degrees of freedom, MAG or TEMP
{F}eqv = 3 x 1 equivalent nodal force vector

The coefficient matrix [K] multiplied by the nodal DOF's {φ} represents the equivalent nodal load
vector which brings the effects of the semi-infinite domain ΩB onto nodes I, J and K.

As mentioned in the beginning, the INFIN47 can be used with magnetic scalar potential elements to
solve 3D magnetic scalar potential problems (MAG degree of freedom). Magnetic scalar potential
elements incorporate three different scalar potential formulations (see Electromagnetic Field Funda-
mentals (p. 163)) selected with the MAGOPT command:

1. Reduced Scalar Potential (accessed with MAGOPT,0)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
482 Published: 2023-10-23
 INFIN47 - 3D Infinite Boundary

2. Difference Scalar Potential (accessed with MAGOPT,2 and MAGOPT,3)

3. Generalized Scalar Potential (accessed with MAGOPT,1, MAGOPT,2, and then MAGOPT,3)

13.47.3. Reduced Scalar Potential

If there is no "iron" in the problem domain, the reduced scalar potential formulation can be used
both in the FE and the BE regimes. In this case, the potential is continuous across FE-BE interface. If
there is "iron" in the FE domain, the reduced potential formulation is likely to produce "cancellation
errors".

13.47.4. Difference Scalar Potential

If there is "iron" and current in the FE region and the problem domain is singly-connected, we can
use the difference potential formulation in order to avoid cancellation error. The formulation consists
of two-step solution procedures:

1. Preliminary solution in the air domain (MAGOPT, 2)

Here the first step consists of computing a magnetic field {Ho} under the assumption that the
magnetic permeability of iron is infinity, thereby neglecting any saturation. The reduced scalar
potential φ is used in FE region and the total scalar potential ψ is used in BE region. In this case,
the potential will be discontinuous across the FE-BE interface. The continuity condition of the
magnetic field at the interface can be written as:
(13.93)

where:

{τ} = tangent vector at the interface along element edge

{Hs} = magnetic field due to current sources

Integrating the above equation along the interface, we obtain

(13.94)

If we take ψ = φ at a convenient point po on the interface, then the above equation defines the
potential jump at any point p on the interface. Now, the total potential ψ can be eliminated from
the problem using this equation, leading to the computation of the additional load vector,
(13.95)

where:

[K] = coefficient matrix defined with Equation 13.92 (p. 482)

2. Total solution (air and iron) (MAGOPT, 3)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 483
Element Library

In this step the total field, {H} = {Ho} - ψ, is computed where {H} is the actual field and {Ho} is
the field computed in step 1 above. Note that the same relation given in Equation 5.39 (p. 170)
uses φg in place of ψ. The total potential ψ is used in both FE and BE regimes. As a result, no po-
tential discontinuity exists at the interface, but an additional load vector due to the field {Ho} must
be computed. Since the magnetic flux continuity condition at the interface of air and iron is:

(13.96)

where:

μo = magnetic permeability of free space (air)

μI = magnetic permeability of iron

The additional load vector may be computed as

(13.97)

where:

{N} = weighting functions

13.47.5. Generalized Scalar Potential

If there is iron and current in the FE domain and the domain is multiply-connected, the generalized
potential formulation can be used. It consists of three different steps.

1. Preliminary solution in the iron domain (MAGOPT, 1). This step computes a preliminary solution
in the iron only. The boundary elements are not used for this step.

2. Preliminary solution in the air domain (MAGOPT, 2). This step is exactly the same as the step 1
of the difference potential formulation.

3. Total solution (air and iron) (MAGOPT, 3) . This step is exactly the same as the step 2 of the differ-
ence potential formulation.

13.48. Not Documented

No detail or element available at this time.

13.49. Not Documented

No detail or element available at this time.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
484 Published: 2023-10-23
 Reserved for Future Use

13.50. MATRIX50 - Superelement (or Substructure)

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness, Conductivity, Stress
Stiffness (used only when added
to the Stiffness Matrix), Same as the
Convection Surface Matrices; Same as the constituent elements constituent
and Gravity, Thermal and elements
Pressure/Heat Generation and
Convection Surface Load Vectors
Same as the constituent elements Same as the
Mass/Specific Heat and
reduced down to the master degrees of constituent
Damping Matrices
freedom elements

Load Type Distribution

Element Temperature and Heat Generation Rate As input during generation run
Pressure/Convection Surface Distribution As input during generation run

13.50.1. Other Applicable Sections

Superelements are discussed in Substructuring Analysis (p. 828).

13.51. Not Documented

No detail or element available at this time.

13.52. Reserved for Future Use

This section is reserved for future use.

13.53. Reserved for Future Use

This section is reserved for future use.

13.54. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 485
Element Library

13.55. PLANE55 - 2D Thermal Solid

K
t s

L J

Y,v

X,R

Matrix or Vector Geometry Shape Functions Integration

Points (p. 414)
Conductivity Matrix and Heat Quad Equation 11.130 (p. 377) 2x2
Generation Load Vector Triangle Equation 11.109 (p. 376) 1 if planar
3 if axisymmetric

Specific Heat Matrix Same as conductivity matrix. Matrix Same as conductivity

is diagonalized as described in matrix
Lumped Matrices (p. 422).
Convection Surface Matrix and Same as conductivity matrix 2
Load Vector evaluated at the face

13.55.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the element matrices and load vectors as well as heat
flux evaluations. SOLID70 - 3D Thermal Solid (p. 492) describes fluid flow in a porous medium, accessed
in PLANE55 with KEYOPT(9) = 1.

13.55.2. Mass Transport Option

If KEYOPT(8) > 0, the mass transport option is included as described in Heat Flow Fundamentals (p. 207)
with Equation 6.1 (p. 207) and by of Equation 6.28 (p. 218). The solution accuracy is dependent on
the element size. The accuracy is measured in terms of the non-dimensional criteria called the element
Peclet number (Gresho([59] (p. 928))):
(13.98)

where:

V = magnitude of the velocity vector

L = element length dimension along the velocity vector direction
ρ = density of the fluid (input as DENS on MP command)
Cp = specific heat of the fluid (input as C on MP command)
K = equivalent thermal conductivity along the velocity vector direction

The terms V, L, and K are explained more thoroughly below:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
486 Published: 2023-10-23
 Reserved for Future Use

(13.99)

where:

Vx = fluid velocity (mass transport) in x direction (input as VX on R command)

Vy = fluid velocity (mass transport) in y direction (input as VY on R command)

Length L is calculated by finding the intersection points of the velocity vector which passes through
the element origin and intersects at the element boundaries.

For orthotropic materials, the equivalent thermal conductivity K is given by:

(13.100)

where:

Kx, Ky = thermal conductivities in the x and y directions (input as KXX and KYY on MP com-
mand)

For the solution to be physically valid, the following condition must be satisfied (Gresho([59] (p. 928))):
(13.101)

This check is carried out during the element formulation and an error message is printed out if
Equation 13.101 (p. 487) is not satisfied. When this error occurs, the problem should be rerun after
reducing the element size in the direction of the velocity vector.

13.56. Not Documented

No detail or element available at this time.

13.57. Reserved for Future Use

This section is reserved for future use.

13.58. Not Documented

No detail or element available at this time.

13.59. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 487
Element Library

13.60. Reserved for Future Use

This section is reserved for future use.

13.61. SHELL61 - Axisymmetric-Harmonic Structural Shell

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Equation 11.39 (p. 370), Equation 11.40 (p. 370),
Stiffness Matrix; and and Equation 11.41 (p. 370). If extra shape
Thermal and Pressure functions are not included (KEYOPT(3) = 1): 3 along length
Load Vectors Equation 11.36 (p. 370), Equation 11.37 (p. 370),
and Equation 11.38 (p. 370)
Mass and Stress Stiffness Equation 11.33 (p. 369), Equation 11.34 (p. 369), Same as stiffness
Matrices and Equation 11.35 (p. 369) matrix

Load Type Distribution

Element
Linear through thickness and along length, harmonic around circumference
Temperature
Constant through thickness, linear along length, harmonic around
Nodal Temperature
circumference
Pressure Linear along length, harmonic around circumference

Reference: Zienkiewicz([40] (p. 926))

13.61.1. Other Applicable Sections

Structures (p. 5) discusses fundamentals of linear elements. PLANE25 - Axisymmetric-Harmonic 4-
Node Structural Solid (p. 456) has a discussion on temperature, applicable to this element.

13.61.2. Assumptions and Restrictions

The material properties are assumed to be constant around the entire circumference, regardless of
temperature dependent material properties or loading.

13.61.3. Stress, Force, and Moment Calculations

Element output comes in two forms:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
488 Published: 2023-10-23
 SHELL61 - Axisymmetric-Harmonic Structural Shell

1. Stresses as well as forces and moments per unit length: This printout is controlled by the KEYOPT(6).
The thru-the-thickness stress locations are shown in Figure 13.13: Stress Locations (p. 489). The
stresses are computed using standard procedures as given in Structural Strain and Stress Evalu-
ations (p. 15). The stresses may then be integrated thru the thickness to give forces per unit length
and moments per unit length at requested points along the length:
(13.102)
(13.103)
(13.104)

(13.105)

(13.106)

(13.107)

Figure 13.13: Stress Locations

Y xt

y J
xb

xc

Rc
R, X

where:

Tx, Tz, Txz, Mx, Mz, Mxz = resultant forces and moments (output as TX, TZ, TXZ, MX, MZ,
MXZ, respectively)
t = thickness (input as TK(I), TK(J) on R command)
σx, σy, σz, σxz = stresses (output as SX, SY, SZ, and SXZ, respectively)

2. Forces and moments on a circumference basis: This printout is controlled by KEYOPT(4). The values
are computed using:
(13.108)

where:

[TR] = local to global transformation matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 489
Element Library

[Ke] = element stiffness matrix

{ue} = nodal displacements

Another difference between the two types of output are the nomenclature conventions. Since the
first group of output uses a shell nomenclature convention and the second group of output uses a
nodal nomenclature convention, Mz and represent moments in different directions.

The rest of this subsection will describe some of the expected relationships between these two
methods of output at the ends of the element. This is done to give a better understanding of the
terms, and possibly detect poor internal consistency, suggesting that a finer mesh is in order. It is
advised to concentrate on the primary load carrying mechanisms. In order to relate these two types
of output in the printout, they have to be requested with both KEYOPT(6) > 1 and KEYOPT(4) = 1.
Further, care must be taken to ensure that the same end of the element is being considered.

The axial reaction force based on the stress over an angle Δβ is:

(13.109)

or

(13.110)

where:

Rc = radius at midplane
t = thickness

The reaction moment based on the stress over an angle Δβ is:

(13.111)

or

(13.112)

Since SHELL61 computes stiffness matrices and load vectors using the entire circumference for
axisymmetric structures, Δβ = 2π. Using this fact, the definition of , and Equation 13.102 (p. 489)
and Equation 13.105 (p. 489), Equation 13.110 (p. 490) and Equation 13.112 (p. 490) become:
(13.113)

(13.114)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
490 Published: 2023-10-23
 Reserved for Future Use

As the definition of φ is critical for these equations, Figure 13.14: Element Orientations (p. 491) is
provided to show φ in all four quadrants.

Figure 13.14: Element Orientations

x x
J J

y
I
I

y
y
I
I

J J
x
x

In a uniform stress (σx) environment, a reaction moment will be generated to account for the greater
material on the outside side. This is equivalent to moving the reaction point outward a distance yf.
yf is computed by:

(13.115)

Using Equation 13.113 (p. 490) and Equation 13.114 (p. 490) and setting Mx to zero gives:

(13.116)

13.62. Reserved for Future Use

This section is reserved for future use.

13.63. Reserved for Future Use

This section is reserved for future use.

13.64. Not Documented

No detail or element available at this time.

13.65. Reserved for Future Use

This section is reserved for future use.

13.66. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 491
Element Library

13.67. Reserved for Future Use

This section is reserved for future use.

13.68. LINK68 - Coupled Thermal-Electric Line

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Electrical Conductivity Matrices Equation 11.14 (p. 368) None
Thermal Conductivity and Specific
Heat Matrices; and Heat Generation Equation 11.13 (p. 368) None
Load Vector

Reference: Kohnke and Swanson([20] (p. 925))

13.68.1. Other Applicable Sections

Coupling (p. 309) discusses coupled effects.

13.69. Reserved for Future Use

This section is reserved for future use.

13.70. SOLID70 - 3D Thermal Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
492 Published: 2023-10-23
 SOLID70 - 3D Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix and Heat
Equation 11.227 (p. 387) 2x2x2
Generation Load Vector
Equation 11.227 (p. 387). Matrix is
Same as
Specific Heat Matrix diagonalized as described in
conductivity matrix
Lumped Matrices (p. 422)
Convection Surface Matrix and Load Equation 11.227 (p. 387) specialized
2x2
Vector to the face

13.70.1. Other Applicable Sections

Derivation of Heat Flow Matrices (p. 217) has a complete derivation of the matrices and load vectors
of a general thermal analysis element. Mass transport is discussed in PLANE55 - 2D Thermal Sol-
id (p. 486).

13.70.2. Fluid Flow in a Porous Medium

An option (KEYOPT(7) = 1) is available to convert SOLID70 to a nonlinear steady-state fluid flow element.
Pressure is the variable rather than temperature. From Equation 6.28 (p. 218), the element conductivity
matrix is:

(13.117)

[B] is defined by Equation 6.28 (p. 218) and for this option, [D] is defined as:

(13.118)

where:

= absolute permeability of the porous medium in the x direction (input as KXX on MP

command)
ρ = mass density of the fluid (input as DENS on MP command)
μ = viscosity of the fluid (input as VISC on MP command)

β = visco-inertial parameter of the fluid (input as C on MP command)

S = seepage velocity (at centroid from previous iteration, defined below)
α = empirical exponent on S (input as MU on MP command)

For this option, no "specific heat" matrix or "heat generation" load vector is computed.

The pressure gradient components are computed by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 493
Element Library

(13.119)

where:

= pressure gradient in the x-direction (output as PRESSURE GRADIENT (X))

{Te} = vector of element temperatures (pressures)

The pressure gradient is computed from:

(13.120)

where:

gp = total pressure gradient (output as PRESSURE GRADIENT (TOTAL))

The mass flux components are:

(13.121)

The vector sum of the mass flux components is:

(13.122)

where:

f = mass flux (output as MASS FLUX)

The fluid velocity components are:

(13.123)

where:

Sx = fluid velocity in the x-direction (output as FLUID VELOCITY (X))

and the maximum fluid velocity is:

(13.124)

where:

S = total fluid velocity (output as FLUID VELOCITY (TOTAL))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
494 Published: 2023-10-23
 MASS71 - Thermal Mass

13.71. MASS71 - Thermal Mass

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Specific Heat Matrix and Heat
None None
Generation Load Vector

13.71.1. Specific Heat Matrix

The specific heat matrix for this element is simply:
(13.125)

Co is defined as:

(13.126)

where:

ρ = density (input as DENS on MP command)

Cp = specific heat (input as C on MP command)
vol = volume (input as CON1 on R command)
Ca = capacitance (input as CON1 on R command)

13.71.2. Heat Generation Load Vector

The heat generation load vector is:
(13.127)

where:

QR = heat rate (input as QRATE on MP command)

A1, A2, etc. = constants (input as A1, A2, etc. on R command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 495
Element Library

Tunif = uniform temperature (input on BFUNIF command)

= temperature from previous iteration
To = offset temperature (input on TOFFST command)

13.72. Reserved for Future Use

This section is reserved for future use.

13.73. Reserved for Future Use

This section is reserved for future use.

13.74. Not Documented

No detail or element available at this time.

13.75. PLANE75 - Axisymmetric-Harmonic 4-Node Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Conductivity Matrix and Quad Equation 11.167 (p. 381) 2x2
Heat Generation Load
Vector Triangle Equation 11.159 (p. 381) 3
Same as conductivity matrix. Matrix is
Same as
Specific Heat Matrix diagonalized as described in Lumped
conductivity matrix
Matrices (p. 422)
Convection Surface Same as conductivity matrix specialized to the
2
Matrix and Load Vector face

13.75.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the element matrices and load vectors as well as heat
flux evaluations.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
496 Published: 2023-10-23
 PLANE77 - 2D 8-Node Thermal Solid

13.76. Reserved for Future Use

This section is reserved for future use.

13.77. PLANE77 - 2D 8-Node Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Conductivity Matrix and Quad Equation 11.141 (p. 378) 3x3
Heat Generation Load
Vector Triangle Equation 11.119 (p. 377) 6
Same as conductivity matrix. If KEYOPT(1) = 1,
Same as
Specific Heat Matrix matrix is diagonalized as described in Lumped
conductivity matrix
Matrices (p. 422)
Convection Surface Same as conductivity matrix, specialized to the
2
Matrix and Load Vector face

13.77.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the thermal element matrices and load vectors as well
as heat flux evaluations. If KEYOPT(1) = 1, the specific heat matrix is diagonalized as described in
Lumped Matrices (p. 422).

13.77.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the temperature varies linearly
along that edge.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 497
Element Library

13.78. PLANE78 - Axisymmetric-Harmonic 8-Node Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Conductivity Matrix and Quad Equation 11.174 (p. 382) 3x3
Heat Generation Load
Vector Triangle Equation 11.163 (p. 381) 6
Same as conductivity matrix. If KEYOPT(1) = 1,
Same as
Specific Heat Matrix matrix is diagonalized as described in Lumped
conductivity matrix
Matrices (p. 422)
Convection Surface
Same as stiffness matrix, specialized to the face 2
Matrix and Load Vector

13.78.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the thermal element matrices and load vectors as well
as heat flux evaluations.

13.78.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the temperature varies linearly
along that edge.

13.79. Reserved for Future Use

This section is reserved for future use.

13.80. Reserved for Future Use

This section is reserved for future use.

13.81. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
498 Published: 2023-10-23
 PLANE83 - Axisymmetric-Harmonic 8-Node Structural Solid

13.82. Reserved for Future Use

This section is reserved for future use.

13.83. PLANE83 - Axisymmetric-Harmonic 8-Node Structural Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.171 (p. 382),
Stiffness, Mass, and Quad Equation 11.172 (p. 382), and 2x2
Stress Stiffness Matrices; Equation 11.173 (p. 382)
and Thermal Load Equation 11.160 (p. 381),
Vector Triangle Equation 11.161 (p. 381), and 3
Equation 11.162 (p. 381)
Pressure Load Vector Same as stiffness matrix, specialized to the face 2

Load Type Distribution

Element Temperature Same as shape functions across element, harmonic around
circumference
Nodal Temperature Same as element temperature distribution
Pressure Linear along each face, harmonic around circumference

Reference: Zienkiewicz([40] (p. 926))

13.83.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. PLANE25 - Axisymmetric-Harmonic 4-Node Structural Solid (p. 456) has a discussion
of temperature applicable to this element.

13.83.2. Assumptions and Restrictions

A dropped midside node implies that the edge is and remains straight.

The material properties are assumed to be constant around the entire circumference, regardless of
temperature-dependent material properties or loading. For (input as MODE on MODE command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 499
Element Library

> 0, extreme values for combined stresses are obtained by computing these stresses at every 10/
degrees and selecting the extreme values.

13.84. Not Documented

No detail or element available at this time.

13.85. Reserved for Future Use

This section is reserved for future use.

13.86. Not Documented

No detail or element available at this time.

13.87. SOLID87 - 3D 10-Node Tetrahedral Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix and
Heat Generation Load Equation 11.190 (p. 383) Analytical
Vector
Same as conductivity matrix. If KEYOPT(1) = 1,
Specific Heat Matrix the matrix is diagonalized as described in Analytical
Lumped Matrices (p. 422)
Equation 11.190 (p. 383) specialized to the face.
Convection Surface Diagonalized surface matrix if KEYOPT(5) = 0, 6
Matrix and Load Vector consistent surface matrix if KEYOPT(5) = 1

13.87.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of thermal element matrices and load vectors as well as
heat flux evaluations. If KEYOPT(1) = 1, the specific heat matrix is diagonalized as described in Lumped
Matrices (p. 422).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
500 Published: 2023-10-23
 Reserved for Future Use

13.88. Reserved for Future Use

This section is reserved for future use.

13.89. Reserved for Future Use

This section is reserved for future use.

13.90. SOLID90 - 3D 20-Node Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Brick Equation 11.239 (p. 388) 14
Conductivity Matrix and Wedge Equation 11.214 (p. 386) 3x3
Heat Generation Load
Vector Pyramid Equation 11.198 (p. 385) 2x2x2
Tet Equation 11.190 (p. 383) 4
Same as conductivity matrix. If KEYOPT(1) = 1,
Same as
Specific Heat Matrix the matrix is diagonalized as described in
conductivity matrix
Lumped Matrices (p. 422).
Convection Surface Quad Equation 11.86 (p. 375) 3x3
Matrix and Load Vector Triangle Equation 11.61 (p. 373) 6

13.90.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of thermal element matrices and load vectors as well as
heat flux evaluations.

13.91. Reserved for Future Use

This section is reserved for future use.

13.92. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 501
Element Library

13.93. Reserved for Future Use

This section is reserved for future use.

13.94. CIRCU94 - Piezoelectric Circuit

I

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix None (lumped) None
Damping Matrix None (lumped, harmonic analysis only) None
Load Vector None (lumped) None

The piezoelectric circuit element, CIRCU94, simulates basic linear electric circuit components that can
be directly connected to the piezoelectric FEA domain. For details about the underlying theory, see
Wang and Ostergaard([323] (p. 943)). It is suitable for the simulation of circuit-fed piezoelectric transducers,
piezoelectric dampers for vibration control, crystal filters and oscillators etc.

13.94.1. Electric Circuit Elements

CIRCU94 contains 5 linear electric circuit element options:

a. Resistor (KEYOPT(1) = 0)
b. Inductor (KEYOPT(1) = 1)
c. Capacitor (KEYOPT(1) = 2)
d. Current Source (KEYOPT(1) = 3)
e. Voltage Source (KEYOPT(1) = 4)

Options a, b, c, d are defined by two nodes I and J (see figure above), each node having a VOLT DOF.
The voltage source is also characterized by a third node K with CURR DOF to represent an auxiliary
charge variable.

13.94.2. Piezoelectric Circuit Element Matrices and Load Vectors

The finite element equations for the resistor, inductor, capacitor and current source of CIRCU94 are
derived using the nodal analysis method (McCalla([190] (p. 935))) that enforces Kirchhoff's Current Law
(KCL) at each circuit node. To be compatible with the system of piezoelectric finite element equations
(see Piezoelectrics (p. 334)), the nodal analysis method has been adapted to maintain the charge balance
at each node:
(13.128)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
502 Published: 2023-10-23
 CIRCU94 - Piezoelectric Circuit

where:

[K] = stiffness (capacitance) matrix

{V} = vector of nodal voltages (to be determined)
{Q} = load vector of nodal charges

The voltage source is modeled using the modified nodal analysis method (McCalla([190] (p. 935))) in
which the set of unknowns is extended to include electric charge at the auxiliary node K, while the
corresponding entry of the load vector is augmented by the voltage source amplitude. In a transient
analysis, different integration schemes are employed to determine the vector of nodal voltages.

For a resistor, the generalized trapezoidal rule is used to approximate the charge at time step n+1
thus yielding:
(13.129)

(13.130)

(13.131)

where:

θ = first order time integration parameter (input on TINTP command)

Δt = time increment (input on DELTIM command)
R = resistance

The constitutive equation for an inductor is of second order with respect to the charge time-derivative,
and therefore the Newmark integration scheme is used to derive its finite element equation:
(13.132)

(13.133)

where:

L = inductance

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 503
Element Library

α, δ = Newmark integration parameters (input on TINTP command

A capacitor with nodes I and J is represented by

(13.134)

(13.135)

where:

C = capacitance

For a current source, the [K] matrix is a null matrix, while the charge vector is updated at each time
step as

(13.136)

where:

Note that for the first substep of the first load step in a transient analysis, as well as on the transient
analysis restart, all the integration parameters (θ, α, δ) are set to 1. For every subsequent substep/load
step, Mechanical APDL uses either the default integration parameters or their values input using the
TINTP command.

In a harmonic analysis, the time-derivative is replaced by jω, which produces

(13.137)

for a resistor,
(13.138)

for an inductor, and

(13.139)

where:

j = imaginary unit

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
504 Published: 2023-10-23
 SOLID96 - 3D Magnetic Scalar Solid

ω = angular frequency (input on HARFRQ command)

The element charge vector {Q} is a null vector for all of the above components.

For a current source, the [K] matrix is a null matrix and the charge vector is calculated as

(13.140)

where:

IS = source current amplitude

φ = source current phase angle (in radians)

Note that all of the above matrices and load vectors are premultiplied by -1 before being assembled
with the piezoelectric finite element equations that use negative electric charge as a through variable
(reaction "force") for the VOLT degree of freedom.

13.95. Reserved for Future Use

This section is reserved for future use.

13.96. SOLID96 - 3D Magnetic Scalar Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Magnetic Scalar Potential Coefficient Matrix;
and Load Vector of Magnetism due to Equation 11.229 (p. 387) 2x2x2
Permanent Magnets, and Source Currents

References: Coulomb([77] (p. 929)), Mayergoyz([120] (p. 931)), Gyimesi([142] (p. 932),[150] (p. 933))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 505
Element Library

13.96.1. Other Applicable Sections

Derivation of Electromagnetic Matrices (p. 176) discusses the magnetic scalar potential method used
by this element.

13.97. Reserved for Future Use

This section is reserved for future use.

13.98. SOLID98 - Tetrahedral Coupled-Field Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Magnetic Potential
Equation 11.192 (p. 383) 4
Coefficient Matrix
Electric Conductivity Matrix Equation 11.191 (p. 383) 4
Thermal Conductivity Matrix Equation 11.190 (p. 383) 4
Stiffness and Mass Matrices; Equation 11.187 (p. 383),
and Thermal Expansion Equation 11.188 (p. 383), and 4
Load Vector Equation 11.189 (p. 383)
Piezoelectric Coupling Same as combination of stiffness matrix and
4
Matrix conductivity matrix
Same as conductivity matrix. If KEYOPT(3)
Specific Heat Matrix = 1, matrix is diagonalized as described in 11
Lumped Matrices (p. 422)
Load Vector due to Imposed
Thermal and Electric
Gradients, Heat Generation,
Joule Heating, Magnetic Same as coefficient or conductivity matrix 4
Forces, Permanent Magnet
and Magnetism due to
Source Currents
Load Vector due to Same as stiffness or conductivity matrix,
6
Convection and Pressures specialized to the face

References: Zienkiewicz([40] (p. 926)), Coulomb([77] (p. 929)), Mayergoyz([120] (p. 931)), Gyimesi([142] (p. 932))

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
506 Published: 2023-10-23
 Reserved for Future Use

13.98.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. Heat Flow (p. 207) describes the derivation of thermal element matrices and load
vectors as well as heat flux evaluations. Derivation of Electromagnetic Matrices (p. 176) describes the
scalar potential method, which is used by this element. Piezoelectrics (p. 334) discusses the piezoelectric
capability used by the element. If KEYOPT(3) = 1, the specific heat matrix is diagonalized as described
in Lumped Matrices (p. 422). Also, Thermoelectrics (p. 348) discusses the thermoelectric capability.

13.99. Reserved for Future Use

This section is reserved for future use.

13.100. Reserved for Future Use

This section is reserved for future use.

13.101. Reserved for Future Use

This section is reserved for future use.

13.102. Reserved for Future Use

This section is reserved for future use.

13.103. Reserved for Future Use

This section is reserved for future use.

13.104. Reserved for Future Use

This section is reserved for future use.

13.105. Reserved for Future Use

This section is reserved for future use.

13.106. Reserved for Future Use

This section is reserved for future use.

13.107. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 507
Element Library

13.108. Reserved for Future Use

This section is reserved for future use.

13.109. Reserved for Future Use

This section is reserved for future use.

13.110. INFIN110 - 2D Infinite Solid

Integration
Matrix or Vector Mapping and Shape Functions
Points (p. 414)
Equation 11.144 (p. 379),
Magnetic Potential
Equation 11.147 (p. 379), and 2x2
Coefficient Matrix
Equation 11.149 (p. 379)
Equation 11.145 (p. 379),
Thermal Conductivity and
Equation 11.147 (p. 379), and 2x2
Specific Heat Matrices
Equation 11.149 (p. 379)
Dielectric Permittivity and Equation 11.145 (p. 379),
Electrical Conductivity Equation 11.147 (p. 379), and 2x2
Coefficient Matrices Equation 11.149 (p. 379)

References: Zienkiewicz et al.([170] (p. 934)), Damjanic' and Owen([171] (p. 934)), Marques and
Owen([172] (p. 934)), Li et al.([174] (p. 934))

13.110.1. Mapping Functions

The theory for the infinite mapping functions is briefly summarized here. Consider the 1D situation
shown below:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
508 Published: 2023-10-23
 INFIN110 - 2D Infinite Solid

Figure 13.15: Global to Local Mapping of a 1D Infinite Element

x
x
k
xj
xo r
MAP
J K M
O J K M t=-1 t=0 t=1
a a (at infinity) (in local coordinates)
(in global coordinates)

The 1D element may be thought of as one edge of the infinite element of Figure 13.16: Mapping of
2D Solid Infinite Element (p. 509). It extends from node J, through node K to the point M at infinity
and is mapped onto the parent element defined by the local coordinate system in the range -1 t
1.

Figure 13.16: Mapping of 2D Solid Infinite Element

N
8

N
M t
8

I I
Y K M
Map
* J
*
Poles of
Mapping s
X,R J

The position of the "pole", xo, is arbitrary, and once chosen, the location of node K is defined by
(13.141)

The interpolation from local to global positions is performed as

(13.142)

where:

MJ(t) = -2t/(1 - t)
MK(t) = 1 - MJ(t)

Examining the above mapping, it can be seen that t = -1, 0, 1 correspond respectively to the global
positions x = xJ, xK, , respectively.

The basic field variable is:

and can be interpolated using standard shape functions, which when written in polynomial form
becomes
(13.143)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 509
Element Library

Solving Equation 13.142 (p. 509) for t yields

(13.144)

where:

r = distance from the pole, O, to a general point within the element

a = xK - xJ as shown in Figure 13.16: Mapping of 2D Solid Infinite Element (p. 509)

Substituting Equation 13.144 (p. 510) into Equation 13.143 (p. 509) gives
(13.145)

Where c0 = 0 is implied since the variable A is assumed to vanish at infinity.

Equation 13.145 (p. 510) is truncated at the quadratic (r2) term in the present implementation. Equa-
tion 13.145 (p. 510) also shows the role of the pole position, O.

In 2D (Figure 13.16: Mapping of 2D Solid Infinite Element (p. 509)) mapping is achieved by the shape
function products. The mapping functions and the Lagrangian isoparametric shape functions for 2D
and axisymmetric 4 node quadrilaterals are given in 2D and Axisymmetric 4-Node Quadrilateral Infinite
Solids (p. 378). The shape functions for the nodes M and N are not needed as the field variable, A, is
assumed to vanish at infinity.

13.110.2. Matrices
The coefficient matrix can be written as:

(13.146)

with the terms defined below:

1. Magnetic Vector Potential (accessed with KEYOPT(1) = 0)

[Ke] = magnetic potential coefficient matrix

μo = magnetic permeability of free space (input on EMUNIT command)

The infinite elements can be used in magnetodynamic analysis even though these elements do
not compute mass matrices. This is because air has negligible conductivity.

2. Electric Potential (Electric Charge) (accessed with KEYOPT(1) = 1)

[Ke] = dielectric permittivity matrix

εx, εy = dielectric permittivity (input as PERX and PERY on MP command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
510 Published: 2023-10-23
 INFIN110 - 2D Infinite Solid

= effective electrical conductivity (defined by Equation 5.83 (p. 175))

3. Temperature (accessed with KEYOPT(1) = 2)

[Ke] = thermal conductivity matrix

kx, ky = thermal conductivities in the x and y direction (input as KXX and KYY on MP
command)

Cc = ρ Cp
ρ = density of the fluid (input as DENS on MP command)
Cp = specific heat of the fluid (input as C on MP command)
{N} = shape functions given in 2D and Axisymmetric 4-Node Quadrilateral Infinite
Solids (p. 378)

4. Electric Potential (Electric Current) (accessed with KEYOPT(1) = 3)

[Ke] = electrical conductivity matrix

= effective electrical conductivity (defined by Equation 5.83 (p. 175))

εx, εy = dielectric permittivity (input as PERX and PERY on MP command)

Although it is assumed that the nodal DOFs are zero at infinity, it is possible to solve thermal problems
in which the nodal temperatures tend to some constant value, To, rather than zero. In that case, the
temperature differential, θ (= T - To), may be thought to be posed as the nodal DOF. The actual tem-
perature can then be easily found from T = θ + To. For transient analysis, θ must be zero at infinity t
> 0, where t is time. Neumann boundary condition is automatically satisfied at infinity.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 511
Element Library

The {Bi} vectors of the [B] matrix in Equation 13.146 (p. 510) contain the derivatives of Ni with respect
to the global coordinates which are evaluated according to

(13.147)

where:

[J] = Jacobian matrix which defines the geometric mapping

[J] is given by

(13.148)

The mapping functions [M] in terms of s and t are given in 2D and Axisymmetric 4-Node Quadrilateral
Infinite Solids (p. 378). The domain differential d(vol) must also be written in terms of the local coordin-
ates, so that
(13.149)

Subject to the evaluation of {Bi} and d(vol), which involves the mapping functions, the element matrices
[Ke] and [Ce] may now be computed in the standard manner using Gaussian quadrature.

13.111. INFIN111 - 3D Infinite Solid

Integration
Matrix or Vector Mapping and Shape Functions
Points (p. 414)
Magnetic Scalar
Potential Coefficient, Equation 11.151 (p. 380), Equation 11.152 (p. 380),
Dielectric Permittivity, Equation 11.154 (p. 380), Equation 11.155 (p. 380),
2x2x2
Electrical Conductivity Equation 11.156 (p. 381), and
Coefficient, and Thermal Equation 11.157 (p. 381)
Conductivity Matrices
Equation 11.151 (p. 380), Equation 11.155 (p. 380),
Specific Heat Matrix Equation 11.156 (p. 381), and 2x2x2
Equation 11.157 (p. 381)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
512 Published: 2023-10-23
 FLUID116 - Coupled Thermal-Fluid Pipe

13.111.1. Other Applicable Sections

See INFIN110 - 2D Infinite Solid (p. 508) for the theoretical development of infinite solid elements. The
derivation presented in INFIN110 - 2D Infinite Solid (p. 508) for 2D can be extended to 3D in a
straightforward manner.

13.112. Reserved for Future Use

This section is reserved for future use.

13.113. Reserved for Future Use

This section is reserved for future use.

13.114. Reserved for Future Use

This section is reserved for future use.

13.115. Reserved for Future Use

This section is reserved for future use.

13.116. FLUID116 - Coupled Thermal-Fluid Pipe

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Between nodes I
Equation 11.13 (p. 368) None
and J
Thermal Conductivity Convection between
Matrix nodes I and K and
None None
between nodes J
and L (optional)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 513
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Pressure Conductivity Between nodes I
Equation 11.12 (p. 368) None
Matrix and J
Specific Heat Matrix and
Heat Generation Load Equation 11.13 (p. 368) None
Vector

13.116.1. Assumptions and Restrictions

Transient pressure and compressibility effects are also not included.

13.116.2. Combined Equations

The thermal and pressure aspects of the problem have been combined into one element having two
different types of working variables: temperatures and pressures. The equilibrium equations for one
element have the form of:

(13.150)

where:

[Ct] = specific heat matrix for one channel

{T} = nodal temperature vector

{P} = nodal pressure vector

[Kt] = thermal conductivity matrix for one channel (includes effects of convection and mass
transport)
[Kp] = pressure conductivity matrix for one channel
{Q} = nodal heat flow vector (input as HEAT on F command)
{w} = nodal fluid flow vector (input as FLOW on F command)
{Qg} = internal heat generation vector for one channel
{H} = gravity and pumping effects vector for one channel
Nc = number of parallel flow channels (input as Nc on R command)

13.116.3. Thermal Matrix Definitions

Specific Heat Matrix

The specific heat matrix is a diagonal matrix with each term being the sum of the corresponding row
of a consistent specific heat matrix:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
514 Published: 2023-10-23
 FLUID116 - Coupled Thermal-Fluid Pipe

(13.151)

where:

ρ = mass density (input as DENS on MP command

P = pressure (average of first two nodes)
Tabs = T + TOFFST = absolute temperature
T = temperature (average of first two nodes)
TOFFST = offset temperature (input on TOFFST command)
Cp = specific heat (input as C on MP command)
A = flow cross-sectional area (input as A on R command)
Lo = length of member (distance between nodes I and J)
Rgas = gas constant (input as Rgas on R command)

Thermal Conductivity Matrix

The thermal conductivity matrix is given by:

(13.152)

where:

Ks = thermal conductivity (input as KXX on MP command)

B2 = h AI
h = film coefficient (defined below)

B3 = h AJ

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 515
Element Library

D = hydraulic diameter (input as D on R command)

w = mass fluid flow rate in the element

w may be determined by the program or may be input by the user:

(13.153)

The above definitions of B4 and B5, as used by Equation 13.152 (p. 515), cause the energy change due
to mass transport to be lumped at the outlet node.

The film coefficient h is defined as:

(13.154)

Nu, the Nusselt number, is defined for KEYOPT(4) = 1 as:

(13.155)

where:

N1 to N4 = input constants (input on R commands)

μ = viscosity (input as VISC on MP command)

A common usage of Equation 13.155 (p. 516) is the Dittus-Boelter correlation for fully developed tur-
bulent flow in smooth tubes (Holman([56] (p. 927))):
(13.156)

where:

Heat Generation Load Vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
516 Published: 2023-10-23
 FLUID116 - Coupled Thermal-Fluid Pipe

The internal heat generation load vector is due to both average heating effects and viscous damping:

(13.157)

where:

VDF = viscous damping multiplier (input on RMORE command)

Cver = units conversion factor (input on RMORE command)

v = average velocity

The expression for the viscous damping part of Qn is based on fully developed laminar flow.

13.116.4. Fluid Equations

Bernoulli's equation is:

(13.158)

where:

Z = coordinate in the negative acceleration direction

P = pressure
γ = ρg
g = acceleration of gravity
v = velocity
PPMP = pump pressure (input as Pp on R command)
CL = loss coefficient

The loss coefficient is defined as:

(13.159)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 517
Element Library

a = additional length to account for extra flow losses (input as La on R command)

k = loss coefficient for typical fittings (input as K on R command)
f = Moody friction factor, defined below:

For the first iteration of the first load step,

(13.160)

where:

fm = input as MU on MP command

For all subsequent iterations

(13.161)

The smooth pipe empirical correlation is:

(13.162)

Bernoulli's Equation 13.158 (p. 517) may be simplified for this element, since the cross-sectional area
of the pipe does not change. Therefore, continuity requires all velocities not to vary along the length.
Hence v1 = v2 = va, so that Bernoulli's Equation 13.158 (p. 517) reduces to:

(13.163)

Writing Equation 13.163 (p. 518) in terms of mass flow rate (w = ρAv), and rearranging terms to match
the second half of Equation 13.150 (p. 514),

(13.164)

Since the pressure drop (PI - PJ) is not linearly related to the flow (w), a nonlinear solution will be re-
quired. As the w term may not be squared in the solution, the square root of all terms is taken in a
heuristic way:

(13.165)

Defining:

(13.166)

and
(13.167)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
518 Published: 2023-10-23
 Reserved for Future Use

Equation 13.165 (p. 518) reduces to:

(13.168)

Hence, the pressure conductivity matrix is based on the term and the pressure (gravity and
pumping) load vector is based on the term Bc PL.

Two further points:

1. Bc is generalized as:

(13.169)

1. (-ZI + ZJ)g is generalized as:

(13.170)

where:

{Δx} = vector from node I to node J

{at} = translational acceleration vector which includes effects of angular velocities (see
Acceleration Effect (p. 681))

13.117. Reserved for Future Use

This section is reserved for future use.

13.118. Reserved for Future Use

This section is reserved for future use.

13.119. Reserved for Future Use

This section is reserved for future use.

13.120. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 519
Element Library

13.121. PLANE121 - 2D 8-Node Electrostatic Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Dielectric Permittivity and Electrical Quad Equation 11.142 (p. 378) 3x3
Conductivity Coefficient Matrices,
Charge Density Load Vector Triangle Equation 11.120 (p. 377) 3
Surface Charge Density and Load Same as coefficient matrix,
2
Vector specialized to the face

13.121.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of the electrostatic element matrices and load vectors
as well as electric field evaluations.

13.121.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the potential varies linearly along
that edge.

13.122. SOLID122 - 3D 20-Node Electrostatic Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Brick Equation 11.240 (p. 388) 14
Dielectric Permittivity and
Electrical Conductivity Wedge Equation 11.215 (p. 386) 3x3
Coefficient Matrices, Charge Pyramid Equation 11.199 (p. 385) 8
Density Load Vector
Tet Equation 11.191 (p. 383) 4

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
520 Published: 2023-10-23
 CIRCU124 - Electric Circuit

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Surface Charge Density Load Quad Equation 11.87 (p. 375) 3x3
Vector Triangle Equation 11.62 (p. 373) 6

13.122.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of electrostatic element matrices and load vectors
as well as electric field evaluations.

13.123. SOLID123 - 3D 10-Node Tetrahedral Electrostatic Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Dielectric Permittivity and
Electrical Conductivity
Equation 11.191 (p. 383) 4
Coefficient Matrices, Charge
Density Load Vector
Charge Density Surface Load Equation 11.191 (p. 383) specialized to
6
Vector the face

13.123.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of electrostatic element matrices and load vectors
as well as electric field evaluations.

13.124. CIRCU124 - Electric Circuit

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 521
Element Library

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix None (lumped) None
Damping Matrix None (lumped, harmonic analysis only) None
Load Vector None (lumped) None

13.124.1. Electric Circuit Elements

CIRCU124 contains 10 linear electric circuit element options. These may be classified into two groups:

1. Independent Circuit Element options, defined by 2 or 3 nodes:

Resistor (KEYOPT(1) = 0)

Inductor (KEYOPT(1) = 1)

Capacitor (KEYOPT(1) = 2)

Current Source (KEYOPT(1) = 3)

Voltage Source (KEYOPT(1) = 4)

2. Dependent Circuit Element options, defined by 3, 4, 5, or 6 nodes:

Mutual inductor (KEYOPT(1) = 8)

Voltage-controlled current source (KEYOPT(1) = 9)

Voltage-controlled voltage source (KEYOPT(1) = 10)

Current-controlled voltage source (KEYOPT(1) = 11)

Current-controlled current source (KEYOPT(1) = 12)

13.124.2. Electric Circuit Element Matrices

All circuit options in CIRCU124 are based on Kirchhoff's Current Law. These options use stiffness
matrices based on a simple lumped circuit model.

For transient analysis, an inductor with nodes I and J can be presented by:

(13.171)

where:

L = inductance
VI = voltage at node I
VJ = voltage at node J
Δt = time increment

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
522 Published: 2023-10-23
 CIRCU124 - Electric Circuit

θ = time integration parameter

n = time step n

A capacitor with nodes I and J is represented by:

(13.172)

where:

C = capacitance

Similarly, a mutual inductor with nodes I, J, K and L has the following matrix:

(13.173)

where:

L1 = input side inductance

L2 = output side inductance
M = mutual inductance

For harmonic analysis, the above three circuit element options have only a damping matrix. For an
inductor:
(13.174)

for a capacitor:
(13.175)

and for a mutual inductor:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 523
Element Library

(13.176)

13.125. CIRCU125 - Diode

Common Diode Zener Diode
I I

J J
KEYOPT (1) = 0 KEYOPT (1) = 1

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix None (lumped) None
Damping Matrix None None
Load Vector None (lumped) None

13.125.1. Diode Elements

CIRCU125 has two highly nonlinear electric circuit element options:

• Common Diode (KEYOPT(1) = 0)

• Zener Diode (KEYOPT(1) = 1)

The I-V characteristics of common and Zener Diodes are plotted in Figure 13.17: I-V (Current-Voltage)
Characteristics of CIRCU125 (p. 525).

As can be seen, the characteristics of the diodes are approximated by a piece-wise linear curve. The
common diode has two sections corresponding to open and closed states. The Zener diode has three
sections corresponding to open, blocked, and Zener states. The parameters of the piece-wise linear
curves are described by real constants depending on KEYOPT(1) selection.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
524 Published: 2023-10-23
 CIRCU125 - Diode

Figure 13.17: I-V (Current-Voltage) Characteristics of CIRCU125

13.125.2. Norton Equivalents

The behavior of a diode in a given state is described by the Norton equivalent circuit representation
(see Figure 13.18: Norton Current Definition (p. 525)).

The Norton equivalent conductance, G, is the derivative (steepness) of the I-V curve to a pertinent
diode state. The Norton equivalent current generator, I, is the current where the extension of the
linear section of the I-V curve intersects the I-axis.

Figure 13.18: Norton Current Definition

13.125.3. Element Matrix and Load Vector

The element matrix and load vectors are obtained by using the nodal potential formulation, a circuit
analysis technique which is perfectly suited to coupling lumped circuit elements to distributed finite
element models.

The stiffness matrix is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 525
Element Library

(13.177)

The load vector is:

(13.178)

where:

G and I = Norton equivalents of the diode in the pertinent state of operation.

13.126. TRANS126 - Electromechanical Transducer

The line electromechanical transducer element, TRANS126, realizes strong coupling between distributed
and lumped mechanical and electrostatic systems. For details about its theory see Gyimesi and Oster-
gaard([249] (p. 938)). This element is especially suitable for the analysis of Micro Electromechanical Systems
(MEMS): accelerometers, pressure sensors, micro actuators, gyroscopes, torsional actuators, filters, HF
switches, etc.

Figure 13.19: Electromechanical Transducer

+ -

D
Physical representatio
n

I+

EMT
K

D
Finite element representatio
n

See, for example, Figure 13.19: Electromechanical Transducer (p. 526) with a damped spring mass reson-
ator driven by a parallel plate capacitor fed by a voltage generator constituting an electromechanical
system. The top figure shows the physical layout of the transducer connected to the mechanical system,
the bottom figure shows the equivalent electromechanical transducer element connected to the
mechanical system.

TRANS126 is a 2 node element each node having a structural (UX, UY or UZ) and an electrical (VOLT)
DOFs. The force between the plates is attractive:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
526 Published: 2023-10-23
 TRANS126 - Electromechanical Transducer

(13.179)

where:

F = force
C = capacitance
x = gap size
V = voltage between capacitor electrodes

The capacitance can be obtained by using the CMATRIX macro for which the theory is given in Capa-
citance Computation (p. 202).

The current is
(13.180)

where:

I = current
t = time

The first term is the usual capacitive current due to voltage change; the second term is the motion in-
duced current.

For small signal analysis:

(13.181)

(13.182)

where:

F0 = force at the operating point

I0 = current at the operating point
[D] = linearized damping matrices
[K] = linearized stiffness matrices
Δx = gap change between the operating point and the actual solution
ΔV = voltage change between the operating point and the actual solution

The stiffness and damping matrices characterize the transducer for small signal prestressed harmonic,
modal and transient analyses.

For large signal static and transient analyses, the Newton-Raphson algorithm is applied with F0 and I0
constituting the Newton-Raphson restoring force and [K] and [D] the tangent stiffness and damping
matrices.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 527
Element Library

(13.183)

where:

Kxx = electrostatic stiffness (output as ESTIF)

F = electrostatic force between capacitor plates
V = voltage between capacitor electrodes
C'' = second derivative of capacitance with respect to gap displacement
(13.184)

where:

Kvv = motion conductivity (output as CONDUCT)

I = current
C' = first derivative of capacitance with respect to gap displacement
v = velocity of gap opening

Definitions of additional post items for the electromechanical transducer are as follows:
(13.185)

where:

Pm = mechanical power (output as MECHPOWER)

F = force between capacitor plates
v = velocity of gap opening
(13.186)

where:

Pe = electrical power (output as ELECPOWER)

V = voltage between capacitor electrodes
I = current
(13.187)

where:

Wc = electrostatic energy of capacitor (output as CENERGY)

V = voltage between capacitor electrodes
C = capacitance
(13.188)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
528 Published: 2023-10-23
 FLUID129 - 2D Infinite Acoustic

F = electrostatic force between capacitor plates (output as EFORCE)

C = capacitance
x = gap size

V = voltage between capacitor electrodes

13.127. Reserved for Future Use

This section is reserved for future use.

13.128. Reserved for Future Use

This section is reserved for future use.

13.129. FLUID129 - 2D Infinite Acoustic

Matrix or Vector Shape Functions Integration

Points (p. 414)
Fluid Stiffness and Damping Equation 11.12 (p. 368) 2
Matrices

13.129.1. Other Applicable Sections

The mathematical formulation and finite element discretization are presented in FLUID130 - 3D Infinite
Acoustic (p. 530).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 529
Element Library

13.130. FLUID130 - 3D Infinite Acoustic

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Fluid Stiffness and Damping
Equation 11.129 (p. 377) 2x2
Matrices

13.130.1. Mathematical Formulation and F.E. Discretization

The exterior structural acoustics problem typically involves a structure submerged in an infinite, ho-
mogeneous, inviscid fluid. The fluid is considered linear, meaning that there is a linear relationship
between pressure fluctuations and changes in density. Equation 13.189 (p. 530) is the linearized, lossless
wave equation for the propagation of sound in fluids.
(13.189)

where:

P = pressure
c = speed of sound in the fluid (input as SONC on MP command)
= second derivative of pressure with respect to time
Ω+ = unbounded region occupied by the fluid

In addition to Equation 13.189 (p. 530)), the following Sommerfeld radiation condition (which simply
states that the waves generated within the fluid are outgoing) needs to be satisfied at infinity:
(13.190)

where:

r = distance from the origin

Pr = pressure derivative along the radial direction

d = dimensionality of the problem (that is, d =3 or d =2 if Ω+ is 3D or 2D respectively

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
530 Published: 2023-10-23
 FLUID130 - 3D Infinite Acoustic

A primary difficulty associated with the use of finite elements for the modeling of the infinite medium
stems precisely from the need to satisfy the Sommerfeld radiation condition, Equation 13.190 (p. 530).
A typical approach for tackling the difficulty consists of truncating the unbounded domain Ω+ by the
introduction of an absorbing (artificial) boundary Γa at some distance from the structure.

Figure 13.20: Absorbing Boundary

L
I I
n
n
a
a K
J
J
y y
R R

x x
z

The equation of motion Equation 13.189 (p. 530) is then solved in the annular region Ωf which is
bounded by the fluid-structure interface Γ and the absorbing boundary Γa. In order, however, for the
resulting problem in Ωf to be well-posed, an appropriate condition needs to be specified on Γa. Towards
this end, the following second-order conditions are used (Kallivokas et al.([220] (p. 937))) on Γa:

In two dimensions:
(13.191)

where:

n = outward normal to Γa
Pn = pressure derivative in the normal direction
Pλλ = pressure derivative along Γa
k = curvature of Γa
γ = stability parameter

In three dimensions:

(13.192)

where:

n = outward normal
u and v = orthogonal curvilinear surface coordinates (for example, the meridional and polar
angles in spherical coordinates)
Pu, Pv = pressure derivatives in the Γa surface directions
H and K = mean and Gaussian curvature, respectively
E and G = usual coefficients of the first fundamental form

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 531
Element Library

13.130.2. Finite Element Discretization

Following a Galerkin based procedure, Equation 13.189 (p. 530) is multiplied by a virtual quantity δP
and integrated over the annular domain Ωf. By using the divergence theorem on the resulting equation
it can be shown that:

(13.193)

Upon discretization of Equation 13.193 (p. 532), the first term on the left hand side will yield the mass
matrix of the fluid while the second term will yield the stiffness matrix.

Next, the following finite element approximations for quantities on the absorbing boundary Γa placed
at a radius R and their virtual counterparts are introduced:
(13.194)
(13.195)

where:

N1, N2, N3 = vectors of shape functions ( = {N1}, {N2}, {N3})

P, q(1), q(2) = unknown nodal values (P is output as degree of freedom PRES. q(1) and q(2) are
solved for but not output).

Furthermore, the shape functions in Equation 13.194 (p. 532) and Equation 13.195 (p. 532) are set to:
(13.196)

The element stiffness and damping matrices reduce to:

For two dimensional case:

(13.197)

(13.198)

where:

dλe = arc-length differential

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
532 Published: 2023-10-23
 FLUID130 - 3D Infinite Acoustic

These matrices are 6 x 6 in size, having 2 nodes per element with 3 degrees of freedom per node (P,
q(1), q(2)).

For three dimensional case:

(13.199)

(13.200)

where:

dAe = area differential

These matrices are 8 x 8 in size, having 4 nodes per element with 2 degrees of freedom per node (P,
q) (Barry et al.([219] (p. 937))).

For axisymmetric case:

(13.201)

(13.202)

where:

x = radius

These matrices are 4 x 4 in size having 2 nodes per element with 2 degrees of freedom per node (P,
q).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 533
Element Library

13.131. SHELL131 - 4-Node Layered Thermal Shell

K

t
s
L r

Y J

X I
Z

Layer Integration
Matrix or Vector Geometry Layer Shape Functions
Points (p. 414)
In-Plane:
3 unknowns per node In-Plane: 2 x 2
Equation 11.77 (p. 374)
per layer (KEYOPT(3) =
0) Thru Thickness:
Thru Thickness: 2
Conductivity Matrix, Equation 11.25 (p. 368)
Heat Generation In-Plane:
Load Vector, and 2 unknowns per node In-Plane: 2 x 2
Equation 11.77 (p. 374)
Convection Surface per layer (KEYOPT(3) =
Matrix and Load 1) Thru Thickness: Thru Thickness: 1
Vector Equation 11.13 (p. 368)
1 unknown per node In-Plane:
In-Plane: 2 x 2
per layer (KEYOPT(3) = Equation 11.77 (p. 374)
2) Thru Thickness: Constant Thru Thickness: 1
Same as
Same as conductivity matrix. Matrix is diagonalized
Specific Heat Matrix conductivity
as described in Lumped Matrices (p. 422)
matrix

13.131.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the thermal element matrices and load vectors as well
as heat flux evaluations.

13.132. SHELL132 - 8-Node Layered Thermal Shell

O K

r
t
s N
L

Y P
J
M

X
Z I

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
534 Published: 2023-10-23
 FLUID136 - 3D Squeeze Film Fluid Element

Layer
Matrix or
Geometry Layer Shape Functions Integration
Vector
Points (p. 414)
Quad: 3 x 3
3 unknowns per In-Plane Equation 11.86 (p. 375)
Triangle: 3
Conductivity node per layer
Matrix, Heat (KEYOPT(3) = 0)
Generation Thru Thickness Equation 11.25 (p. 368)
2
Load Vector, Quad: 3 x 3
2 unknowns per In-Plane Equation 11.86 (p. 375)
Specific Heat Triangle: 3
node per layer
Matrix and
(KEYOPT(3) = 1)
Convection Thru Thickness Equation 11.13 (p. 368)
1
Surface Matrix Quad: 3 x 3
and Load 1 unknown per In-Plane Equation 11.86 (p. 375)
Triangle: 3
Vector node per layer
(KEYOPT(3) = 2)
Thru Thickness Constant 1

13.132.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the thermal element matrices and load vectors as well
as heat flux evaluations.

13.133. Reserved for Future Use

This section is reserved for future use.

13.134. Reserved for Future Use

This section is reserved for future use.

13.135. Reserved for Future Use

This section is reserved for future use.

13.136. FLUID136 - 3D Squeeze Film Fluid Element

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 535
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Conductivity Matrix and Quad, if KEYOPT(2) = 0 Equation 11.76 (p. 374) 2 x 2 (4-node)
Velocity Load Vector Quad, if KEYOPT(2) = 1 Equation 11.108 (p. 376) 3 x 3 (8-node)
Same as conductivity matrix. If KEYOPT(1) = 1, Same as
Damping Matrix matrix is diagonalized as described in Lumped conductivity
Matrices (p. 422) matrix

13.136.1. Other Applicable Sections

Squeeze Film (p. 229) describes the governing squeeze film equations used as a basis for forming the
element matrices.

13.136.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the pressure varies linearly along
that edge.

13.137. Reserved for Future Use

This section is reserved for future use.

13.138. FLUID138 - 3D Viscous Fluid Link Element

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Pressure and Damping Matrices Equation 11.12 (p. 368) None

13.138.1. Other Applicable Sections

Squeeze Film (p. 229) describes the governing squeeze film equations used as a basis for forming the
element matrices.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
536 Published: 2023-10-23
 Not Documented

13.139. FLUID139 - 3D Slide Film Fluid Element

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Fluid, Stiffness, Mass, and
Analytical Formula None
Damping Matrices

13.139.1. Other Applicable Sections

Slide Film (p. 234) describes the governing slide film equations used as a basis for forming the element
matrices.

13.140. Reserved for Future Use

This section is reserved for future use.

13.141. Reserved for Future Use

This section is reserved for future use.

13.142. Reserved for Future Use

This section is reserved for future use.

13.143. Not Documented

No detail or element available at this time.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 537
Element Library

13.144. Reserved for Future Use

This section is reserved for future use.

13.145. Reserved for Future Use

This section is reserved for future use.

13.146. Reserved for Future Use

This section is reserved for future use.

13.147. Reserved for Future Use

This section is reserved for future use.

13.148. Reserved for Future Use

This section is reserved for future use.

13.149. Reserved for Future Use

This section is reserved for future use.

13.150. Reserved for Future Use

This section is reserved for future use.

13.151. SURF151 - 2D Thermal Surface Effect

(Extra nodes, optional) (Extra nodes, optional)
K L L M
J
J 1
K
1
y x
Y
(or axial) x y
I I

X (or radial)

Integration
Matrix or Vector Midside Nodes [1] Shape Functions
Points (p. 414)
With midside nodes w = C1 + C2x + C3x2 2
All Without midside w = C1 + C2x 2
nodes

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
538 Published: 2023-10-23
 SURF152 - 3D Thermal Surface Effect

1. Midside node setting is controlled by KEYOPT(4).

Load Type Distribution

All Loads Same as shape functions

The logic is very similar to that given for SURF152 - 3D Thermal Surface Effect (p. 539).

13.152. SURF152 - 3D Thermal Surface Effect

(Extra nodes, optional) (Extra nodes, optional)
Q R M N

L
O
L
K
P z K
z
y
y
N
I
x
I
Z J
x

M J
Y
X

Geometry / Midside Nodes Integration

Matrix or Vector Shape Functions
[1] Points (p. 414)
Quad with midside nodes Equation 11.86 (p. 375) 3x3
Convection Surface
Matrix and Load Quad without midside nodes Equation 11.77 (p. 374) 2x2
Vector; and Heat Triangle with midside nodes Equation 11.61 (p. 373) 6
Generation Load
Triangle without midside
Vector Equation 11.107 (p. 376) 3
nodes

1. Midside node setting is controlled by KEYOPT(4).

Load Type Distribution

All Loads Same as shape functions

13.152.1. Matrices and Load Vectors

When the extra node is not present, the logic is the same as given and as described in Derivation of
Heat Flow Matrices (p. 217). The discussion below relates to theory that uses the extra node.

The conductivity matrix is based on one-dimensional flow to and away from the surface. The form is
conceptually the same as for LINK33 (Equation 13.27 (p. 462)) except that the surface has four or eight
nodes instead of only one node. Using the example of convection and no midside nodes are requested
(KEYOPT(4) = 1) (resulting in a 5 x 5 matrix), the first four terms of the main diagonal are:

(13.203)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 539
Element Library

hu = output argument for film coefficient of USRSURF116

{N} = vector of shape functions

which represents the main diagonal of the upper-left corner of the conductivity matrix. The remaining
terms of this corner are all zero. The last main diagonal term is simply the sum of all four terms of
Equation 13.203 (p. 539) and the off-diagonal terms in the fifth column and row are the negative of
the main diagonal of each row and column, respectively.

If midside nodes are present (KEYOPT(4) = 0) (resulting in a 9 x 9 matrix) Equation 13.203 (p. 539) is
replaced by:

(13.204)

which represents the upper-left corner of the conductivity matrix. The last main diagonal is simply
the sum of all 64 terms of Equation 13.204 (p. 540) and the off-diagonal terms in the ninth column
and row are the negative of the sum of each row and column respectively.

Radiation is handled similarly, except that the approach discussed for LINK31 in LINK31 - Radiation
Link (p. 459) is used. A load vector is also generated. The area used is the area of the element. The
form factor is discussed in a subsequent section.

An additional load vector is formed when using the extra node by:
(13.205)

where:

{Qc} = load vector to be formed

[Ktc] = element conductivity matrix due to convection

TEMVEL from USRSURF116 is the difference between the bulk temperature and the temperature of
the extra node.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
540 Published: 2023-10-23
 SURF152 - 3D Thermal Surface Effect

13.152.2. Adiabatic Wall Temperature as Bulk Temperature

There is special logic that accesses FLUID116 information where FLUID116 has had KEYOPT(2) set
equal to 1. This logic uses SURF151 or SURF152 with the extra node present (KEYOPT(5) = 1) and
computes an adiabatic wall temperature (KEYOPT(6) = 1). For this case, Tv, as used above, is defined
as:

(13.206)

where:

FR = recovery factor (see Equation 13.207 (p. 541))

Vabs = absolute value of fluid velocity (input as VABS on R command)

Ω = angular velocity of moving wall (input as OMEGA on R command)

Ωref = reference angular velocity (input as (An)I and (An)J on R command of FLUID116)
Fs = slip factor (input as SLIPFAI, SLIPFAJ on R command of FLUID116)
V116 = velocity of fluid at extra node from FLUID116
gc = gravitational constant used for units consistency (input as GC on R command)
Jc = Joule constant used to convert work units to heat units (input as JC on R command)

The recovery factor is computed as follows:

(13.207)

where:

Cn = constant used for recovery factor calculation (input as NRF on R command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 541
Element Library

μf = viscosity of fluid (from FLUID116)

ρf = density of fluid (from FLUID116)

D = diameter of fluid pipe (from FLUID116)

(13.208)

where:

V = velocity used to compute Reynold's number

The adiabatic wall temperature is reported as:

(13.209)

where:

Taw = adiabatic wall temperature

Tex = temperature of extra node

KEYOPT(1) = 0 or 1 is ordinarily used for turbomachinry analysis, whereas KEYOPT(1) = 2 is ordinarily

used for flow past stationary objects. For turbomachinery analyses Tex is assumed to be the total
temperature, but for flow past stationary objects Tex is assumed to be the static temperature.

13.152.3. Film Coefficient Adjustment

After the first coefficient has been determined, it is adjusted if KEYOPT(7) = 1:
(13.210)

where:

hf = unadjusted film coefficient

TS = surface temperature
TB = bulk temperature (Taw, if defined)
n = real constant (input as ENN on RMORE command)

13.152.4. Radiation Form Factor Calculation

The form factor is computed as:

(13.211)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
542 Published: 2023-10-23
 SURF153 - 2D Structural Surface Effect

also,

F = form factor (output as FORM FACTOR)

Developing B further

α = angle between element z axis at integration point being processed and the line connecting
the integration point and the extra node (see Figure 13.21: Form Factor Calculation (p. 543))

Figure 13.21: Form Factor Calculation

Extra node (Q)

L
K

I J

F is then used in the two-surface radiation equation:

(13.212)

where:

σ = Stefan-Boltzmann constant (input as SBCONST on R command)

ε = emissivity (input as EMIS on MP command)
A = element area

Note that this "form factor" does not have any distance affects. Thus, if distances are to be included,
they must all be similar in size, as in an object on or near the earth being warmed by the sun. For
this case, distance affects can be included by an adjusted value of σ.

13.153. SURF153 - 2D Structural Surface Effect

J
J
K
y
x
Y
x
(or axial) y
I I
X (or radial)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 543
Element Library

Integration
Matrix or Vector Midside Nodes [1] Shape Functions
Points (p. 414)
All With midside nodes w = C1 + C2x + C3x2 3
All Without midside nodes w = C1 + C2x 2

1. Midside node setting is controlled by KEYOPT(4).

Load Type Distribution

All Loads Same as shape functions

The logic is very similar to that given for SURF154 in SURF154 - 3D Structural Surface Effect (p. 545) with
the differences noted below:

1. For surface tension (input as SURT on R command)) on axisymmetric models (KEYOPT(3) = 1), an
average force is used on both end nodes.

2. For surface tension with midside nodes, no load is applied at the middle node, and only the com-
ponent directed towards the other end node is used.

3. When using large deflections, the area on which pressure is applied changes. The updated distance
between the two end nodes is used. For plain strain problems, the thickness (distance normal to
the X-Y plane) remains at 1.0, by definition. For plane stress problems, the thickness is adjusted:
(13.213)

where:

tu = final thickness used.

ti = thickness for user input option (input as TKPS on R command)

εz = strain in thickness direction (normal to X-Y plane)

Using the assumption of constant volume:

(13.214)

where:

εx = strain along the length of the element

εy = strain normal to the underlying solid.

Assuming further that:

(13.215)

yields:
(13.216)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
544 Published: 2023-10-23
 SURF154 - 3D Structural Surface Effect

13.154. SURF154 - 3D Structural Surface Effect

Geometry / Midside Nodes Integration

Matrix or Vector Shape Functions
[1] Points (p. 414)
Quad with midside nodes Equation 11.84 (p. 375) 3x3
Stiffness and Quad without midside nodes Equation 11.72 (p. 374) 2x2
Damping Matrices,
and Pressure Load Triangle with midside nodes Equation 11.116 (p. 377) 6
Vector Triangle without midside
Equation 11.66 (p. 374) 3
nodes
Equation 11.82 (p. 375),
Equation 11.83 (p. 375)
Quad with midside nodes 3x3
and
Equation 11.84 (p. 375)
Equation 11.70 (p. 374),
Equation 11.71 (p. 374)
Mass and Stress Quad without midside nodes 2x2
and
Stiffness Matrices Equation 11.72 (p. 374)
Triangle with midside nodes Equation 11.116 (p. 377) 6
Equation 11.64 (p. 374),
Triangle without midside Equation 11.65 (p. 374)
3
nodes and
Equation 11.66 (p. 374)
Equation 11.82 (p. 375)
Quad with midside nodes and 3x3
Equation 11.83 (p. 375)
Equation 11.70 (p. 374)
Quad without midside nodes and 2x2
Surface Tension Equation 11.71 (p. 374)
Load Vector Equation 11.114 (p. 377)
Triangle with midside nodes and 6
Equation 11.115 (p. 377)
Equation 11.64 (p. 374)
Triangle without midside
and 3
nodes
Equation 11.65 (p. 374)

1. Midside node setting is controlled by KEYOPT(4).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 545
Element Library

Load Type Distribution

All Loads Same as shape functions

The stiffness matrix is:

(13.217)

where:

kf = foundation stiffness (input as EFS on R command)

A = area of element
{Nz} = vector of shape functions representing motions normal to the surface

The mass matrix is:

(13.218)

where:

th = thickness (input as TKI, TKJ, TKK, TKL on RMORE command)

ρ = density (input as DENS on MP command)
{N} = vector of shape functions
Ad = added mass per unit area (input as ADMSUA on R command)

If the command LUMPM,ON is used, [Me] is diagonalized as described in Lumped Matrices (p. 422).

The element damping matrix is:

(13.219)

where:

μ = dissipation (input as VISC on MP command)

The element stress stiffness matrix is:

(13.220)

where:

[Sg] = derivatives of shape functions of normal motions

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
546 Published: 2023-10-23
 SURF154 - 3D Structural Surface Effect

s = in-plane force per unit length (input as SURT on R command)

If pressure is applied to face 1, the pressure load stiffness matrix is computed as described in Pressure
Load Stiffness (p. 45).

The element load vector is:

(13.221)

where:

{Np} = vector of shape functions representing in-plane motions normal to the edge
E = edge of element

{Nx} = vector of shape functions representing motion in element x direction

{Ny} = vector of shape functions representing motion in element y direction

Pv = uniform pressure magnitude

P1 = input (VAL1 with LKEY = 5 on SFE command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 547
Element Library

θ = angle between element normal and applied load direction

Dx, Dy, Dz = vector directions (input as VAL2 thru VAL4 with LKEY = 5 on SFE command)
{NX}, {NY}, {NZ} = vectors of shape functions in global Cartesian coordinates

The integration used to arrive at is the usual numerical integration, even if KEYOPT(6) ≠ 0. The
output quantities "average face pressures" are the average of the pressure values at the integration
points.

13.155. SURF155 - 3D Thermal Surface Line Load

Integration
Matrix or Vector Midside Nodes [1] Shape Functions
Points (p. 414)
With midside nodes w = C1 + C2x + C3x2 2
All Without midside w = C1 + C2x 2
nodes

1. Midside node setting is controlled by KEYOPT(4).

Load Type Distribution

All Loads Same as shape functions

The logic is similar to that given for SURF152 - 3D Thermal Surface Effect (p. 539).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
548 Published: 2023-10-23
 SHELL157 - Thermal-Electric Shell

13.156. SURF156 - 3D Structural Surface Line Load Effect

2 2

x J
y 1 1
J y
K 4
4
x
3
L 3 I
I K
z z

Integration
Matrix or Vector Midside Nodes [1] Shape Functions
Points (p. 414)
Equation 11.19 (p. 368),
With midside nodes Equation 11.20 (p. 368), 3
Pressure Load Equation 11.21 (p. 368)
Vector Equation 11.15 (p. 368),
Without midside nodes Equation 11.16 (p. 368), 2
Equation 11.17 (p. 368)

1. Midside node setting is controlled by KEYOPT(4).

Load Type Distribution

Linear along length for faces 1, 2, and 3; constant along length for
Pressures
face 4

13.157. SHELL157 - Thermal-Electric Shell

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.78 (p. 374). No
Quad 2x2
Electrical Conductivity variation thru thickness
Matrix Equation 11.78 (p. 374). No
Triangle 1
variation thru thickness
Thermal Conductivity Equation 11.77 (p. 374) and
Matrix; Heat Generation Quad Equation 11.78 (p. 374). No 2x2
Load and Convection variation thru thickness
Surface Matrix and Load
Vectors

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 549
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.107 (p. 376) and
Triangle Equation 11.78 (p. 374). No 1
variation thru thickness
Same as conductivity matrix. Matrix is
Same as
Specific Heat Matrix diagonalized as described in Lumped
conductivity matrix
Matrices (p. 422)

13.157.1. Other Applicable Sections

Coupling (p. 309) discusses coupled effects.

13.158. Not Documented

No detail or element available at this time.

13.159. SURF159 - General Axisymmetric Surface with 2 or 3 Nodes

Z

J2

J1
K2

J3

I2
y
z
K3
x

I3

Integration
Matrix or Vector Midside Nodes [1] Shape Functions
Points (p. 414)[2]
Equation 11.47 (p. 372)
With midside nodes through 3 x Nc
Equation 11.49 (p. 372)
Mass Matrix
Equation 11.44 (p. 372)
Without midside nodes through 2 x Nc
Equation 11.46 (p. 372)
With midside nodes
Stress Stiffness Matrix Same as mass matrix. 2 x Nc
Without midside nodes
Pressure Load Vector With midside nodes
Same as mass matrix. 2 x Nc
Without midside nodes

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
550 Published: 2023-10-23
 Reserved for Future Use

1. Midside node setting is controlled by KEYOPT(4).

2. Nc = the number of node planes in the circumferential direction. The Nc integration

points are circumferentially located at:

• the nodal planes, and

• midway between the nodal planes (that is, at the integration planes)

so that Nc = (2 * Nnp), where Nnp = number of nodal planes (KEYOPT(2)).

Exception: If KEYOPT(2) = 1, then Nc = 1.

Load Type Distribution

Pressure Linear along each face in both directions.

13.159.1. Other Applicable Sections

General Element Formulations (p. 49) gives the general element formulations used by this element.

13.159.2. Assumptions and Restrictions

Although the elements are initially axisymmetric, the loads and deformation can be general in non-
axisymmetric 3D. The displacements are interpolated in elemental coordinate system by interpolation
functions, but the user can define the nodal displacements in any direction.

13.160. Reserved for Future Use

This section is reserved for future use.

13.161. Reserved for Future Use

This section is reserved for future use.

13.162. Reserved for Future Use

This section is reserved for future use.

13.163. Reserved for Future Use

This section is reserved for future use.

13.164. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 551
Element Library

13.165. Reserved for Future Use

This section is reserved for future use.

13.166. Reserved for Future Use

This section is reserved for future use.

13.167. Reserved for Future Use

This section is reserved for future use.

13.168. Reserved for Future Use

This section is reserved for future use.

13.169. TARGE169 - 2D Target Segment

13.169.1. Other Applicable Sections

TARGE170 - 3D Target Segment (p. 553) discusses Target Elements.

13.169.2. Segment Types

TARGE169 supports six 2D segment types:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
552 Published: 2023-10-23
 TARGE170 - 3D Target Segment

Figure 13.22: 2D Segment Types

Line
I J

I Arc, clockwise
K J

I K J Arc, counterclockwise

J I Parabola
K

I
Circle

Pilot Node
I

13.170. TARGE170 - 3D Target Segment

Target Segment Element
K TARGE170 K

I J I J

n n n

Y
X
Node-to-Surface
Contact Element
Surface-to-Surface CONTA175
Contact Element
CONTA174
TARGE170 K

I K I J

n
3D Line-to-Surface
Contact Element 3D Line-to-Surface
CONTA177 Contact Element
CONTA177

13.170.1. Introduction
In studying the contact between two bodies, the surface of one body is conventionally taken as a
contact surface and the surface of the other body as a target surface. The "contact-target" pair concept
has been widely used in finite element simulations. For rigid-flexible contact, the contact surface is
associated with the deformable body; and the target surface must be the rigid surface. For flexible-
flexible contact, both contact and target surfaces are associated with deformable bodies. The contact
and target surfaces constitute a "Contact Pair".

TARGE170 is used to represent various 3D target surfaces for the associated contact elements (CON-
TA174, CONTA175, and CONTA177). The contact elements themselves overlay the solid elements, line
elements, or shell element edges describing the boundary of a deformable body that is potentially

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 553
Element Library

in contact with the rigid target surface, defined by TARGE170. Hence, a "target" is simply a geometric
entity in space that senses and responds when one or more contact elements move into a target
segment element.

13.170.2. Segment Types

The target surface is modelled through a set of target segments; typically several target segments
make up one target surface. Each target segment is a single element with a specific shape or segment
type. TARGE170 supports ten 3D segment types; see Figure 13.23: 3D Segment Types (p. 554)

Figure 13.23: 3D Segment Types

K J
L K
3-Node Triangle
TSHAP,TRIA 4-Node Quadrilateral
I TSHAP,QUAD
Cylinder
I J TSHAP,CYLI
R1 = Radius I J
K
Cone
I J TSHAP,CONE
R1 = Radius(I) N M
R2 = Radius (J) 6-Node Triangle
I TSHAP,TRI6
L

Sphere J
I O
TSHAP,SPHE
R1 = Radius L K

P N 8-Node Quadrilateral
Pilot node TSHAP,QUA8
I TSHAP,PILO

I M
J

Line Parabola
I J TSHAP,LINE K TSHAP,PARA
I J

Point
TSHAP,POINT

13.170.3. Reaction Forces

The reaction forces on the entire rigid target surface are obtained by summing all the nodal forces
of the associated contact elements. The reaction forces are accumulated on the pilot node. If the pilot
node has not been explicitly defined by the user, one of the target nodes (generally the one with the
smallest number) will be used to accumulate the reaction forces.

13.171. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
554 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

13.172. CONTA172 - 2D 3-Node Surface-to-Surface Contact

Associated Target Surface

Contact normal Contact Element

Surface of Solid Element

J I
Y

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix W = C1 + C2 x + C3x2 2

13.172.1. Other Applicable Sections

The CONTA172 description is the same as for CONTA174 - 3D 8-Node Surface-to-Surface Contact (p. 555)
except that it is 2D.

13.173. Reserved for Future Use

This section is reserved for future use.

13.174. CONTA174 - 3D 8-Node Surface-to-Surface Contact

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
If KEYOPT(4) = 0 (has midside
nodes) Equation 11.82 (p. 375),
Quad 2x2
Stiffness and Stress Equation 11.83 (p. 375), and
Stiffness Matrices Equation 11.84 (p. 375)
If KEYOPT(4) = 0 (has midside
Triangle 3
nodes) Equation 11.116 (p. 377)

The following CONTA174 topics are available:

13.174.1. Introduction
13.174.2. Contact Kinematics

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 555
Element Library

13.174.3. Frictional Model

13.174.4. Contact Algorithm
13.174.5. Viscous Damping
13.174.6. Energy and Momentum Conserving Contact
13.174.7. Debonding
13.174.8. Contact Surface Wear
[Link]/Structural Contact
13.174.10. Electric Contact
13.174.11. Magnetic Contact
13.174.12. Pore Fluid Contact
13.174.13. Diffusive Contact

13.174.1. Introduction
CONTA174 is an 8-node element that is intended for general rigid-flexible and flexible-flexible contact
analysis. In a general contact analysis, the area of contact between two (or more) bodies is generally
not known in advance. CONTA174 is applicable to 3D geometries. It may be applied for contact
between solid bodies or shells.

13.174.2. Contact Kinematics

Contact Pair

In studying the contact between two bodies, the surface of one body is conventionally taken as a
contact surface and the surface of the other body as a target surface. For rigid-flexible contact, the
contact surface is associated with the deformable body; and the target surface must be the rigid
surface. For flexible-flexible contact, both contact and target surfaces are associated with deformable
bodies. The contact and target surfaces constitute a "Contact Pair".

The CONTA174 contact element is associated with the 3D target segment elements (TARGE170) using
a shared real constant set number. This element is located on the surface of 3D solid, shell elements
(called underlying element). It has the same geometric characteristics as the underlying elements.
The contact surface can be either side or both sides of the shell or beam elements.

Location of Contact Detection

Figure 13.24: Contact Detection Point Location at Gauss Point

Deformable solid

Gauss integration
point Contact segment
Target segment

Rigid
body

CONTA174 is a surface-to-surface contact element. The contact detection points are the integration
points and are located either at nodal points or Gauss points. The contact element is constrained

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
556 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

against penetration into the target surface at its integration points. However, the target surface can,
in principle, penetrate through into the contact surface. See Figure 13.24: Contact Detection Point
Location at Gauss Point (p. 556). CONTA174 uses Gauss integration points by default (Cescotto and
Charlier([215] (p. 937)), Cescotto and Zhu([216] (p. 937))), which generally provides more accurate results
than when using the nodes themselves as the integration points. A disadvantage of using nodal
contact points is that, for a uniform pressure, the kinematically equivalent forces at the nodes are
unrepresentative and indicate release at corners.

The standard surface projection method (based on Puso and Laursen ([379] (p. 946) and [380] (p. 946)))
is available for the 3D element, CONTA174. This option enforces contact constraints on an overlapping
region of contact and target surfaces (see Figure 13.25: Surface-Projection-Based Contact (p. 557))
rather than on individual contact nodes or Gauss points. This method passes patch tests and generally
offers smoother stress distributions (for both contact elements and underlying elements) than the
other contact detection options. The contact forces vary smoothly during large sliding, and the forces
do not jump when contact nodes slide off edges of target surfaces. The surface projection method
also satisfies moment equilibrium for frictional or bonded contact. Other contact detection options
may not satisfy moment equilibrium exactly when an offset exists along the contact normal direction
under certain circumstances (for example, when using a penalty based formulation, when accounting
for shell thickness effects, or when using a user-defined contact offset).

Figure 13.25: Surface-Projection-Based Contact

Overlapping Contact andTarget Element Meshes

Contact element

n
Target element

Initial Element Positions Overlapping Region after

Surface Projection

The dual shape function projection method (based on Popp et al. [443] (p. 950)) is an extension of the
standard surface projection method. It uses a dual shape function for the discretization of contact
traction instead of using a displacement shape function. The dual shape function is derived based
on the biorthogonality relationship from the standard displacement shape function. Therefore, this
method reduces dependent terms from the adjacent nodes for contact constraints built on each
contact node. The resulting gap function and slip increments are closer to the local geometric gap/slip
at the contact nodes.

Penetration Distance

The penetration distance is measured along the normal direction of the contact surface (at integration
points) to the target surface (Cescotto and Charlier([216] (p. 937))). See Figure 13.26: Penetration Dis-
tance (p. 558). It is uniquely defined even when the geometry of the target surface is not smooth.
Such discontinuities may be due to physical corners on the target surface, or may be introduced by
a numerical discretization process (e.g. finite elements). Based on the present way of calculating

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 557
Element Library

penetration distance, there is no restriction on the shape of the rigid target surface. Smoothing is not
always necessary for the concave corner. For the convex corner, it is still recommended to smooth
out the region of abrupt curvature changes (see Figure 13.27: Smoothing Convex Corner (p. 558)).

Figure 13.26: Penetration Distance

Penetration distance Target surface

Contact element Integration point

Figure 13.27: Smoothing Convex Corner

Smoothing Radius

Outward normal

For the surface-projection-based contact method, the contact detection remains at contact nodal
points and the contact normal is perpendicular to the contact surface. However, the contact penetration
distance is computed over the overlapping region (see Figure 13.25: Surface-Projection-Based Con-
tact (p. 557)) in an average sense.

Pinball Algorithm

The position and the motion of a contact element relative to its associated target surface determine
the contact element status. The program monitors each contact element and assigns a status:

STAT = 0 Open far-field contact

STAT = 1 Open near-field contact
STAT = 2 Sliding contact
STAT = 3 Sticking contact

A contact element is considered to be in near-field contact when the element enters a pinball region,
which is centered on the integration point of the contact element. The computational cost of
searching for contact depends on the size of the pinball region. Far-field contact element calculations
are simple and add few computational demands. The near-field calculations (for contact elements
that are nearly or actually in contact) are slower and more complex. The most complex calculations
occur the elements are in actual contact.

Setting a proper pinball region is useful to overcome spurious contact definitions if the target surface
has several convex regions. The current default setting should be appropriate for most contact
problems.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
558 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

13.174.3. Frictional Model

Coulomb's Law

In the basic Coulomb friction model, two contacting surfaces can carry shear stresses. When the
equivalent shear stress is less than a limit frictional stress (τlim), no motion occurs between the two
surfaces. This state is known as sticking. The Coulomb friction model is defined as:
(13.222)
(13.223)

where:

τlim = limit frictional stress

μ = coefficient of friction for isotropic friction (input as MU using either TB command with
Lab = FRIC or MP command; orthotropic friction defined below
P = contact normal pressure
b = contact cohesion (input as COHE on R command)

Once the equivalent frictional stress exceeds τlim, the contact and target surfaces will slide relative
to each other. This state is known as sliding. The sticking/sliding calculations determine when a point
transitions from sticking to sliding or vice versa. The contact cohesion provides sliding resistance even
with zero normal pressure.

CONTA174 provides an option for defining a maximum equivalent frictional stress τmax (input as
TAUMAX on RMORE command) so that, regardless of the magnitude of the contact pressure, sliding
will occur if the magnitude of the equivalent frictional stress reaches this value.

Figure 13.28: Friction Model

Sliding

max
lim

Sticking
b

Contact elements offer two models for Coulomb friction: isotropic friction and orthotropic friction.

Isotropic Friction

The isotropic friction model uses a single coefficient of friction μiso based on the assumption of uniform
stick-slip behavior in all directions. It is available with all 2D and 3D contact elements (CONTA172,
CONTA174, CONTA175, CONTA177, and CONTA178).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 559
Element Library

Orthotropic Friction

The orthotropic friction model is based on two coefficients of friction, μ1 and μ2, to model different
stick-slip behavior in different directions. Use orthotropic friction model in 3D contact only (CONTA174,
CONTA175, and CONTA177). The two coefficients are defined in two orthogonal sliding directions
called the principal directions (see Element Reference for more details). The frictional stress in principal
direction i=1,2 is given by:
(13.224)

By appropriately scaling the frictional stresses in principal directions the expressions for scaled limit
frictional stress ( ) and scaled equivalent frictional stress ( ) for orthotropic friction can be
written in a form similar to isotropic friction (Michalowski and Mroz([358] (p. 945))):
(13.225)

(13.226)

(13.227)
(13.228)

where:

= scaled frictional stress in direction i = 1,2

= scaled equivalent frictional stress
= scaled limit frictional stress
μeq = equivalent coefficient of friction for orthotropic friction
μ1, μ2 = coefficients of friction in first and second principal directions (input as MU1 and MU2
using TB command with Lab = FRIC)

While scaled frictional stresses are used for friction computations, actual frictional stresses are output
after applying the inverse scaling in Equation 13.225 (p. 560).

The coefficient of friction (μ1 and μ2) can have dependence on time, temperature, normal pressure,
sliding distance, or sliding relative velocity (defined as fields on the TBFIELD command). Suitable
combinations of up to two fields can be used to define dependency, for example, temperature or
temperature and sliding distance; see Contact Friction in the Material Reference for details.

Static and Dynamic Friction

CONTA174 provides the exponential friction model, which is used to smooth the transition between
the static coefficient of friction and the dynamic coefficient of friction according to the formula
(Benson and Hallquist([317] (p. 943))):
(13.229)

where:

v = slip rate

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
560 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

μs= Rf μd = static coefficient of friction

Rf = ratio of static and dynamic friction (input as FACT on RMORE command)
c = decay coefficient (input as DC on RMORE command)

Integration of Frictional Law

The integration of the frictional mode is similar to that of nonassociated theory of plasticity (see Rate-
Independent Plasticity (p. 67)). In each substep that sliding friction occurs, an elastic predictor is
computed in contact traction space. The predictor is modified with a radial return mapping function,
providing both a small elastic deformation along with a sliding response as developed by Giannako-
poulos([136] (p. 932)).

The flow rule giving the slip increment for orthotropic friction can be written as:

(13.230)

where:

dui = slip increment in principal direction i = 1, 2

λ = Lagrange multiplier for friction

By appropriately scaling the slip increment, it can be shown that the Lagrange multiplier is equal to
the scaled equivalent slip increment:
(13.231)

(13.232)

and the direction of scaled slip increment is same as that of scaled frictional stress.

(13.233)

Thus, computations for orthotropic friction use the same framework as isotropic friction except for
scaled slip increments and scaled frictional stresses which are converted to actual values for output.

User-defined Friction

For friction models that do not follow Coulomb’s law, you can write a USERFRIC subroutine. Refer to
the Guide to User-Programmable Features in the Programmer's Reference for a detailed description
on writing a USERFRIC subroutine. You can use it with any 2D or 3D contact element (CONTA172,
CONTA174, CONTA175, CONTA177, and CONTA178) with penalty method for tangential contact (select
KEYOPT(2) = 0, 1, or 3). Use the TB,FRIC command with TBOPT = USER to choose the user define
friction option, and specify the friction properties on the TBDATA command.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 561
Element Library

Friction models involve nonlinear material behavior, so only experienced users who have a good
understanding of the theory and finite element programming should attempt to write a USERFRIC
subroutine.

Constitutive Relationship for Frictional Sliding

Regarding frictional sliding, the linearized constitutive relationship between tangential stresses and
sliding increments can be expressed:

(13.234)

where:

= equivalent slip predictor over the current substep (Laursen and Simo[218] (p. 937))
= current equivalent slip rate
ti = the unit vector in the direction of slip (i = 1, 2)

can be obtained from the tabular input (define the friction coefficient to be a function of pressure
via the TB,FRIC command).

can be obtained from the tabular input (define friction coefficient to be a function of sliding ve-
locity via TB,FRIC command) or from the definition of Static and Dynamic Friction (p. 560) via real
constants FACT and DC.

Algorithmic Symmetrization

Contact problems involving friction produce non-symmetric stiffness. Using an unsymmetric solver
(NROPT,UNSYM) is more computationally expensive than a symmetric solver for each iteration. For
this reason, a symmetrization algorithm developed by Laursen and Simo([218] (p. 937)) is used by
which most frictional contact problems can be solved using solvers for symmetric systems. If frictional
stresses have a substantial influence on the overall displacement field and the magnitude of the
frictional stresses is highly solution dependent, any symmetric approximation to the stiffness matrix
may provide a low rate of convergence. In such cases, the use of an unsymmetric stiffness matrix is
more computationally efficient.

Forced Frictional Sliding Using Velocity Input

In a static analysis, you can model steady-state frictional sliding between two flexible bodies or
between a flexible and a rigid body with different velocities. In this case the sliding velocities no
longer follow the nodal displacements, and they are predefined through the CMROTATE command.
This command sets the velocities on the nodes of the element component as an initial condition at
the start of a load step. The program determines the sliding direction based on the direction of the
sliding velocities. The frictional stresses then can be expressed by:
(13.235)

where is obtained from the CMROTATE command.

The linearized frictional stress with respect to the sliding velocity can be expressed in the following
form (Bajer, Belsky, and Kung[394] (p. 947)):

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
562 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

(13.236)

In the complex eigenvalue extraction analysis using the QRDAMP or DAMP method, the above terms
will be included in the damping matrix. The first term is, in general, negative in the case of a negative
friction-sliding velocity gradient, and it can destabilize the system. Therefore, it is referred to as
destabilizing squeal damping. The destabilizing squeal damping is proportional to the contact pressure,
but it follows the sliding direction. The second term is proportional to the equivalent friction stress
but orthogonal to the sliding direction. It may be needed to avoid false unstable modes and is,
therefore, referred to as stabilizing squeal damping (Bajer, Belsky, and Kung[394] (p. 947)). To study
the effects of squeal damping on the complex eigenvalue extraction analysis, the above equation is
rewritten:

(13.237)

where:

cd = a scaling factor for destabilizing squeal damping (real constant FDMD when KEYOPT(16)
= 0)
cs = a scaling factor for stabilizing squeal damping (real constant FDMS when KEYOPT(16) =
0)

= friction-sliding velocity gradient (real constant FDMD when KEYOPT(16) = 1)

= destabilizing damping coefficient (real constant FDMD when KEYOPT(16) = 2)

= stabilizing damping coefficient (real constant FDMS when KEYOPT(16) = 1 or 2)

13.174.4. Contact Algorithm

Four different contact algorithms are implemented in this element (selected by KEYOPT(2)).

• Pure penalty method

• Augmented Lagrangian method (Simo and Laursens([217] (p. 937)))

• Pure Lagrange multiplier method (Bathe([2] (p. 925)))

• Lagrange multiplier on contact normal and penalty on frictional direction

Pure Penalty Method

This method requires both contact normal and tangential stiffness. The main drawback is that the
amount penetration between the two surfaces depends on this stiffness. Higher stiffness values de-
crease the amount of penetration but can lead to ill-conditioning of the global stiffness matrix and
to convergence difficulties. Ideally, you want a high enough stiffness that contact penetration is ac-
ceptably small, but a low enough stiffness that the problem will be well-behaved in terms of conver-
gence or matrix ill-conditioning.

The contact traction vector is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 563
Element Library

(13.238)

where:

P = normal contact pressure

τ1 = frictional stress in direction 1
τ2 = frictional stress in direction 2

The contact pressure is:

(13.239)

where:

Kn = contact normal stiffness

un = contact gap size

The frictional stress for isotropic friction is obtained by Coulomb's law:

(13.240)

where:

Ks = tangential contact stiffness (input as FKT on R command)

Δui = slip increment in direction i over the current substep; In the case of sticking, Δui is
the elastic slip increment, which represents the reversible tangential motion within the current
substep.
= equivalent slip increment over the current substep
μiso = coefficient of friction
= frictional stress in direction i = 1, 2 at the end of previous substep

For orthotropic friction, slip increment and frictional stress are scaled so that

(13.241)

where:

= scaled tangential contact stiffness in principal direction i = 1, 2

= slip increment in principal direction i = 1, 2 over the current substep
= scaled equivalent slip increment over the current substep

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
564 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

μeq = equivalent coefficient of friction

= scaled frictional stress in principal direction i = 1, 2 at the end of previous substep

For consistency between scaled friction stress and scaled slip increment, the scaled tangential contact
stiffness in principal direction i = 1, 2 must be defined as:

(13.242)

As a special case, when contact friction is used in a steady-state rolling analysis, the frictional stress
calculation is changed to a viscous formulation. The frictional stress for isotropic friction given in
Equation 13.240 (p. 564) is modified as:

(13.243)

where is the slip velocity in direction i. All other symbols have the same meaning as in Equa-
tion 13.240 (p. 564).

Similarly, the orthotropic friction calculation given in Equation 13.241 (p. 564) is modified as:

(13.244)

where is the slip velocity in principle direction i.

Augmented Lagrangian Method

The augmented Lagrangian method is an iterative series of penalty updates to find the Lagrange
multipliers (that is, contact tractions). Compared to the penalty method, the augmented Lagrangian
method usually leads to better conditioning and is less sensitive to the magnitude of the contact
stiffness coefficient. However, in some analyses, the augmented Lagrangian method may require ad-
ditional iterations, especially if the deformed mesh becomes excessively distorted.

The contact pressure is defined by:

(13.245)

where:

ε = compatibility tolerance (input as FTOLN on R command)

λi = Lagrange multiplier component at iteration i

The Lagrange multiplier component λi is computed locally (for each element) and iteratively.

Pure Lagrange Multiplier Method

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 565
Element Library

The pure Lagrange multiplier method does not require contact stiffness. Instead it requires chattering
control parameters. Theoretically, the pure Lagrange multiplier method enforces zero penetration
when contact is closed and "zero slip" when sticking contact occurs. However the pure Lagrange
multiplier method adds additional degrees of freedom to the model and requires additional iterations
to stabilize contact conditions. This will increase the computational cost. This algorithm has chattering
problems due to contact status changes between open and closed or between sliding and sticking.
The other main drawback of the Lagrange multiplier method is overconstraint in the model. The
model is overconstrained when a contact constraint condition at a node conflicts with a prescribed
boundary condition on that degree of freedom (for example, D command) at the same node. Over-
constraints can lead to convergence difficulties and/or inaccurate results. The Lagrange multiplier
method also introduces zero diagonal terms in the stiffness matrix, so that iterative solvers (for example,
PCG) can not be used.

The contact traction components (that is, Lagrange multiplier parameters) become unknown DOFs
for each element. The associated Newton-Raphson load vector is:
(13.246)

Lagrange Multiplier on Contact Normal and Penalty on Frictional Direction

In this method only the contact normal pressure is treated as a Lagrange multiplier. The tangential
contact stresses are calculated based on the penalty method (see Equation 13.240 (p. 564)).

This method allows only a very small amount of slip for a sticking contact condition. It overcomes
chattering problems due to contact status change between sliding and sticking which often occurs
in the pure Lagrange Multiplier method. Therefore this algorithm treats frictional sliding contact
problems much better than the pure Lagrange method.

13.174.5. Viscous Damping

Viscous damping is primarily used to damp relative motions between the contact and target surfaces
for open contact. It provides a certain resistance to reduce the risk of rigid body motion. The viscous
damping traction is proportional but opposite to the relative velocities along contact normal or tan-
gential directions:
(13.247)
(13.248)
(13.249)

where:

dn = damping coefficient in normal direction

dt = damping coefficient in tangential direction

The total damping force is calculated by the integration of damping traction over the contact elements.

The damping coefficient is specified as a function of the opening gap:

(13.250)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
566 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

(13.251)

where:

upinb = pinball radius (real constant PINB)

dn0 = normal contact damping coefficient (real constant FDMN)
dt0 = tangential contact damping coefficient (real constant FDMT)

The contact damping energy can be obtained by integrating damping traction over the contacting
area within the time increment:
(13.252)

where:

Ac = area domain of the contact surface

= time increment

13.174.6. Energy and Momentum Conserving Contact

To correctly model the physical interaction between contact and target surfaces in a transient dynamic
analysis, the contact forces must maintain force and energy balance, and ensure proper transfer of
linear momentum. This requires imposing additional constraints on relative velocities between contact
and target surfaces (see Laursen and Chawla ([373] (p. 946)), and Armero and Pet cz ([374] (p. 946))).

Impact Constraints and Contact Forces

In Mechanical APDL, the penetration constraints and the relative velocity constraints between contact
and target surfaces are collectively referred to as impact constraints. These constraints can be selected
by setting KEYOPT(7) = 4 for any of the 2D or 3D contact elements and are valid for all types of contact
interactions (flexible-to-flexible, flexible-to-rigid, and rigid-to-rigid) with and without friction.

An automatic time stepping scheme is used to predict the time of impact and adjust the size of the
time increment to minimize penetration. The resulting penetration may not be within the penetration
tolerance specified using real constant FTOLN. Rather, the allowable penetration is decided such that
energy and momentum are conserved. When contact is detected, the relative velocity constraints are
imposed using one of the four contact algorithms: pure penalty method, augmented Lagrangian
method, pure Lagrange multiplier method, or Lagrange multiplier in contact normal and penalty in
frictional direction method. In the case of rough contact (KEYOPT(12) = 1) the relative velocity constraint
is imposed in the tangential direction also to prevent slip. In the case of standard contact (KEYOPT(12)
= 0) with friction, the slip increment and frictional stress are computed by taking the relative velocity
constraint into consideration.

For the pure penalty method, contact pressure P and friction stresses τi for isotropic friction are
defined as:

(13.253)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 567
Element Library

where:

Kn = contact normal stiffness

un = contact gap size

= algorithmic contact gap size (based on the relative velocity constraint)

Pn-1 = normal contact pressure at the end of previous substep

and:

(13.254)

where:

Ks = tangential contact stiffness (input as FKT on R command)

= slip increment in direction i over the current substep

= equivalent slip increment over the current substep

= algorithmic slip increment in direction i over the current substep

= algorithmic equivalent slip increment over the current substep

μiso = coefficient of friction

= frictional stress in direction i = 1,2 at the end of previous substep

For other contact algorithms, the expressions for contact pressure and frictional stresses are defined
in a similar manner as shown in Equation 13.245 (p. 565) and Equation 13.246 (p. 566) but with addi-
tional variables as shown above in Equation 13.253 (p. 567) and Equation 13.254 (p. 568).

Energy and Momentum Balance

Imposition of the impact constraints at Gauss points of contact elements ensures satisfaction of mo-
mentum and energy balance in a finite element sense. Since the impact constraints act only on the
contact/target interface, energy balance is not enforced for the underlying finite elements used to
model the interior of the contact and target bodies. Total energy at the contact/target interface is
conserved for frictionless or rough contact when relative velocity constraints are satisfied exactly. If
the relative velocity constraints are not satisfied to a tight tolerance there may be some loss of kinetic
energy.

When friction is specified for contact elements, energy is conserved when the contact and target
surfaces are not slipping (STICK) with respect to each other, and energy equal to the work done by
frictional forces is dissipated when the contact and target surfaces are slipping (SLIP) with respect to
each other.

Energy is also lost when numerical damping is used for the time integration scheme.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
568 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

As per the classical theory of impact, exact conservation of energy during impact between rigid
bodies is identified with elastic impact. It corresponds to a coefficient of restitution (COR) of 1. The
impact constraints in Mechanical APDL for impact between rigid bodies satisfy the conditions of
elastic impact when the constraints are satisfied exactly and no numerical damping or friction is
specified. Energy loss during impact between rigid bodies can be modeled by specifying a coefficient
of restitution value (input as real constant COR) smaller than 1.

Time Integration Scheme

The impact constraints are formulated such that they can be used with both methods available for
implicit transient dynamic analysis in Mechanical APDL, the Newmark method and the HHT method.
An important reason for using the impact constraints is that they make the time integration scheme
numerically more stable without using large numerical damping. A small amount of numerical
damping may still be needed to suppress high frequency noise.

13.174.7. Debonding
Debonding refers to separation of bonded contact (KEYOPT(12) = 2, 3, 4, 5 or 6). It is activated by
associating a cohesive zone material model (input with TB,CZM) with contact elements. Debonding
is available only for pure penalty method and augmented Lagrangian method (KEYOPT(2) = 0,1) with
contact elements CONTA172, CONTA174, CONTA175, and CONTA177.

A cohesive zone material model is provided with bilinear behavior (Alfano and Crisfield([362] (p. 945)))
for debonding. The model defines contact stresses as:
(13.255)
(13.256)

and
(13.257)

where:

P = normal contact stress (tension)

τ1 = tangential contact stress in direction 1
τ2 = tangential contact stress in direction 2
Kn = normal contact stiffness
Kt = tangential contact stiffness
un = contact gap
u1 = contact slip distance in direction 1
u2 = contact slip distance in direction 2
d = debonding parameter
direction 1 and direction 2 = principal directions in tangent plane

The debonding parameter is defined as:

(13.258)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 569
Element Library

with d = 0 for Δ 1 and 0 < d 1 for Δ > 1, and Δ and χ are defined below.

Debonding allows three modes of separation: mode I, mode II and mixed mode.

Mode I debonding is defined by setting

(13.259)

and
(13.260)

where:

= contact gap at the maximum normal contact traction (tension)

= contact gap at the completion of debonding (input on TBDATA command as C2 using
TB,CZM)

Mode II debonding is defined by setting

(13.261)

(13.262)

and
(13.263)

where:

= equivalent tangential slip distance at the maximum equivalent tangential stress,

= equivalent tangential slip distance at the completion of debonding (input on TBDATA
command as C4 using TB,CZM)

Mixed mode debonding is defined by setting

(13.264)

and
(13.265)

The constraint on χ that the ratio of the contact gap distances be same as the ratio of tangential slip
distances is enforced automatically by appropriately scaling the contact stiffness values.

For mixed mode, debonding is complete when the energy criterion is satisfied:

(13.266)

with

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
570 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

(13.267)

(13.268)

(13.269)

(13.270)

where:

σmax = maximum normal contact stress (input on TBDATA command as C1 using TB,CZM)
τmax = maximum equivalent tangential contact stress (input on TBDATA command as C3
using TB,CZM)

Verification of satisfaction of energy criterion can be done during postprocessing of results.

The debonding modes are based on input data:

1. Mode I for normal data (input on TBDATA command as C1, C2, and C5).

2. Mode II for tangential data (input on TBDATA command as C3, C4, and C5).

3. Mixed mode for normal and tangential data (input on TBDATA command as C1, C2, C3, C4,
C5 and C6).

Artificial damping can be used to overcome convergence difficulties associated with debonding. It is
activated by specifying the damping coefficient η (input on TBDATA command as C5 using TB,CZM).

Tangential slip under compressive normal contact stress for mixed mode debonding is controlled by
appropriately setting the flag β (input on TBDATA command as C6 using TB,CZM). Settings on β are:

β = 0 (default) indicates no tangential slip

β = 1 indicates tangential slip is allowed

After debonding is completed the surface interaction is governed by standard contact constraints for
normal and tangential directions. Frictional contact is used if friction is specified for contact elements.

13.174.8. Contact Surface Wear

Progressive loss of material from the surface of a solid body when in contact with another body (in
any phase) is called wear. At a micro-scale, the process of wear is quite complex with various mech-
anical and chemical processes resulting in material failure and loss. At a continuum scale, wear is
approximated by phenomenological models that relate various quantities at the contact surface with
material loss (see Meng and Ludema [415] (p. 948)). Mechanical APDL models wear using such a phe-
nomenological approach. Material loss due to wear is approximated by repositioning the contact
nodes at the contact surface. The new coordinates of the nodes are determined by a wear model.
Since the contact nodes are moved to new positions, the contact variables (for example, contact
pressure) change. The underlying continuum elements also experience a loss in material (and volume),
thus simulating wear. After the nodes are moved to a new position, equilibrium may be lost and the
program continues to iterate until equilibrium is achieved.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 571
Element Library

The following options are available for defining a wear model:

• Archard wear model (see Archard [416] (p. 948))

• User-defined wear model

Wear is activated using the TB command with Lab = WEAR for the material assigned to contact ele-
ments. The following contact elements support modeling wear: CONTA172, CONTA174, and CONTA175.

Archard Wear Model

The Archard wear model (see Archard [416] (p. 948)) is a popular sliding wear model with fairly good
results for simulating wear. The original model proposed by Archard assumed that the rate of volume
loss due to wear is linearly proportional to the contact pressure and sliding velocity at the contact
surface. A generalized version of this model allowing proper law dependence on contact pressure
and velocity is implemented in the program. The Archard model is defined using the TB command
with TBOPT = ARCD and Lab = WEAR.

Wear is assumed to occur in the inward normal direction of the surface, taken as the direction opposite
the contact normal direction. The rate of wear at a contact node, , is given by:
(13.271)

where:

K = wear coefficient
H = material hardness
P = contact pressure
vrel = the relative sliding velocity
m = pressure exponent
n = velocity exponent

In the above equation, K, H, m, and n are user-specified values from the TBDATA command. If a value
of 1 is input for the fifth constant (C5) on the TBDATA command, the calculation is based on the
nodal stresses of the underlying solid element instead of the contact pressure. The stress evaluated
at the gauss point of the solid element is first copied to its nearest node. The contact element overlaid
on the solid element to which the gauss point belongs uses this stress, σ, and the contact normal, n,
to calculate the component of the traction vector in the contact normal direction: .
This normal component of the traction vector is used instead of contact pressure in the wear rate
equation.

If C5 on the TBDATA command is set to 10 or 11, the wear increment is averaged over the contact
area of the contact pair such that the total volume lost due to wear is the same as the total volume
lost due to wear when each node wears with a different amount (C5 = 0 or 1). The average wear in-
crement is calculated as:

(13.272)

where A represents the contact area at the point and the summation is over all points in contact in
a contact pair.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
572 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

User-Defined Wear Model (USERWEAR)

If the Archard wear model does not fit your needs, you can define your own wear model by program-
ming the subroutine USERWEAR. Most of the relevant contact results and properties are passed in
the USERWEAR subroutine. You can define the wear increment and direction of the wear increment.
The user-defined wear model is defined using the TB command with TBOPT = USER and Lab = WEAR.

13.174.9. Thermal/Structural Contact

Combined structural and thermal contact is specified if KEYOPT(1) = 1, which indicates that structural
and thermal DOFs are active. Pure thermal contact is specified if KEYOPT(1) = 2. The thermal contact
features (Zhu and Cescotto([280] (p. 940))) are:

Thermal Contact Conduction

(13.273)

where:

q = heat flux (heat flow rate per area)

Kc = thermal contact conductance coefficient (input as TCC on R command)
TT = temperature on target surface
TC = temperature on contact surface

Heat Convection
(13.274)

where:

hf = convection coefficient (input on SFE command with Lab = CONV and KVAL = 1)

Heat Radiation
(13.275)

where:

σ = Stefan-Boltzmann constant (input as SBCT on R command)

ε = emissivity (input using EMIS on MP command)
F = radiation view factor (input as RDVF on R command)
To = temperature offset (input as VALUE on TOFFST command)

Heat Generation Due to Frictional Sliding

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 573
Element Library

(13.276)

where:

qc = amount of frictional dissipation on contact surface

qT = amount of frictional dissipation on target surface
Fw = weight factor for the distribution of heat between two contact and target surfaces (input
as FWGT on R command)
Ff = fractional dissipated energy converted into heat (input on FHTG on R command)
= equivalent frictional stress
v = sliding rate

Note:

When KEYOPT(1) = 2, heat generation due to friction is ignored.

The frictional dissipation energy can be obtained by integrating dissipation density over the contacting
area within the time increment

(13.277)

where:

Ac = area domain of the contact surface

= time increment

13.174.10. Electric Contact

Combined structural, thermal, and electric contact is specified if KEYOPT(1) = 3. Combined thermal
and electric contact is specified if KEYOPT(1) = 4. Combined structural and electric contact is specified
if KEYOPT(1) = 5. Pure electric contact is specified if KEYOPT(1) = 6. The electric contact features are:

Electric Current Conduction (KEYOPT(1) = 3 or 4)

(13.278)

where:

J = current density
σ/L = electric conductivity per unit length (input as ECC on R command)
VT = voltage on target surface
VC = voltage on contact surface

Electrostatic (KEYOPT(1) = 5 or 6)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
574 Published: 2023-10-23
 CONTA174 - 3D 8-Node Surface-to-Surface Contact

(13.279)

where:

13.174.11. Magnetic Contact

The magnetic contact is specified if KEYOPT(1) = 7. Using the magnetic scalar potential approach, the
3D magnetic flux across the contacting interface is defined by:
(13.280)

where:

ψn = magnetic flux
ϕt = magnetic potential at target surface (MAG degree of freedom)
ϕc = magnetic potential at contact surface (MAG degree of freedom)
CM = magnetic contact permeance coefficient
μo = free space permeability
A = contact area
= normal component of the "guess" magnetic field (See Equation 5.16 (p. 166))

The gap permeance is defined as the ratio of the magnetic flux in the gap to the total magnetic po-
tential difference across the gap. The equation for gap permeance is:
(13.281)

where:

t = gap thickness

The magnetic contact permeance coefficient is defined as:

(13.282)

The above equations are only valid for 3D analysis using the Magnetic Scalar Potential (p. 176) approach.

13.174.12. Pore Fluid Contact

Combined structural and pore fluid contact is specified if KEYOPT(1) = 8. Combined structural, pore
fluid, and thermal contact is specified if KEYOPT(1) = 9.

Pore Fluid Flow Permeability (for closed contact and near-field contact):
(13.283)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 575
Element Library

(13.284)

where:

= pore fluid flux density out of contact surface

= pore fluid flux density out of target surface
= pore fluid contact permeability coefficient (input as PCC on R command)
= pore pressure on target surface
= pore pressure on contact surface

Pore Fluid Flow Due to Gap Rate Change (for near-field contact):
(13.285)
(13.286)

where:

= gap distance rate between the contact and target

= weight factor for the distribution of gap pore fluid flow between two contact and target
surfaces (input as FPWT on R command)
= gap pore fluid flow participation factor (input as FPFT on R command)

This pore fluid flow is associated with removing or adding fluid for near field contact. For steady-state
analysis, the changing rate for gap distance is zero so that there is no contribution from this gap-rate-
dependent fluid flow.

Pore Fluid Flow (for far-field contact):

(13.287)

where:

= the pore pressure seepage coefficient (input as PSEE on R command)

= the ambient pore pressure (input as ABPP on R command)

Drainage-Only Flow (for near-field contact):

(13.288)
(13.289)

Drainage-only flow indicates that normal pore fluid flow occurs only from the contact surface to the
environment when the contact surface pore pressures are positive. No fluid flow can enter the contact
surface when the contact surface pore pressures are negative.

13.174.13. Diffusive Contact

Diffusive contact combined with other multiphysics fields is specified if KEYOPT(1) > 10.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
576 Published: 2023-10-23
 CONTA175 - 2D/3D Node-to-Surface Contact

Diffusion Flow for closed contact and near-field contact :

(13.290)
(13.291)

where:

qc = diffusion flux density out of contact surface

qt = diffusion flux density out of target surface
Kd = contact diffusivity coefficient (input as DCC on R command)
Cc = concentration on contact surface
Ct = concentration on target surface

Diffusion Flow for far-field contact :

(13.292)

where:

Kc = diffusive convection coefficient (input as DCON on R command)

C∞ = ambient concentration (input as ABDC on R command)

13.175. CONTA175 - 2D/3D Node-to-Surface Contact

Y Y
I
CONTA175
X I X
CONTA175 Z
target normal target normal

2D associated target 3D associated target

surface (TARGE169) surface (TARGE170)

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Normal Direction None None
Stiffness Matrix
Sliding Direction None None

13.175.1. Other Applicable Sections

The CONTA175 description is the same as for CONTA174 - 3D 8-Node Surface-to-Surface Contact (p. 555)
except that it is a one node contact element.

13.175.2. Contact Models

The contact model can be either contact force based (KEYOPT(3) = 0, default) or contact traction
based (KEYOPT(3) = 1). For a contact traction based model, Mechanical APDL can determine the area

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 577
Element Library

associated with the contact node. For the single point contact case, a unit area will be used which is
equivalent to the contact force based model.

13.175.3. Contact Forces

In order to satisfy contact compatibility, forces are developed in a direction normal (n-direction) to
the target that will tend to reduce the penetration to an acceptable numerical level. In addition to
normal contact forces, friction forces are developed in directions that are tangent to the target plane.

(13.293)

where:

Fn = normal contact force

Kn = contact normal stiffness (input as FKN on R command)
un = contact gap size

(13.294)

where:

FT = tangential contact force

KT = tangential contact stiffness (input as FKT on R command)
uT = contact slip distance

For orthotropic friction, μeq is computed using the expression:

(13.295)

where:

μeq = equivalent coefficient of friction for orthotropic friction

μ1, μ2 = coefficients of friction in first and second principal directions (input as MU1 and MU2
using TB command with Lab = FRIC)

13.176. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
578 Published: 2023-10-23
 CONTA177 - 3D Line-to-Surface Contact

13.177. CONTA177 - 3D Line-to-Surface Contact

Y

CONTA177
X CONTA177
Z J
J
K
K
I
I

I K J
CONTA177 Edge of 3D target
shell elements line segment
3D target surface

Integration
Matrix or Vector Shape Function
Points (p. 414)
Stiffness Matrix W = C1 + C2x + C3x2 None

13.177.1. Other Applicable Sections

The CONTA177 description is the same as for CONTA174 - 3D 8-Node Surface-to-Surface Contact (p. 555)
except that it is a 3D line contact element.

13.177.2. Contact Kinematics

Four different scenarios can be modeled by CONTA177:

• Contact between one beam (or edge) and a surface of solid or shell elements (see above figure)

• Internal contact where one beam (or pipe) slides inside another hollow beam (or pipe) (see Fig-
ure 13.29: Beam Sliding Inside a Hollow Beam (p. 580))

• External contact between two beams that lie next to each other and are roughly parallel (see Fig-
ure 13.30: Parallel Beams in Contact (p. 580))

• External contact between two beams that cross (see Figure 13.31: Crossing Beams in Contact (p. 580))

Use KEYOPT(3) = 0 or 1 for the first three scenarios (internal contact and parallel beams). In these
cases, the contact condition is only checked at contact nodes.

Use KEYOPT(3) = 3 or 4 for the fourth scenario (beams that cross). In these cases, the contact condition
is checked along the entire length of the beams. The beams with circular cross-sections are assumed
to come in contact in a point-wise manner.

Use KEYOPT(3) = 2 for all scenarios. The contact condition is only checked at contact nodes for the
first three scenarios, and on an intersection along the beams for the fourth scenario. The program
reports contact pressure (contact traction-based model).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 579
Element Library

Figure 13.29: Beam Sliding Inside a Hollow Beam

TARGE170 rt

n
rc

CONTA177

Figure 13.30: Parallel Beams in Contact

CONTA177

Figure 13.31: Crossing Beams in Contact

n

CONTA177

rc

rt

TARGE170

Contact is detected when two circular beams touch or overlap each other. The non-penetration
condition for beams with a circular cross section can be defined as follows.

For internal contact:

and for external contact:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
580 Published: 2023-10-23
 CONTA177 - 3D Line-to-Surface Contact

where g is the gap distance; rc and rt are the radii of the cross sections of the beams on the contact
and target sides, respectively; and d is the minimal distance between the two beams which also de-
termines the contact normal direction (see Figure 13.31: Crossing Beams in Contact (p. 580)). Contact
occurs for negative values of g.

13.177.3. Contact Model

The contact model can be either contact force-based (KEYOPT(3) = 0 or 4) or contact traction-based
(KEYOPT(3) = 1, 2, or 3). For the contact traction-based model, the program determines the area
(based on the underlying beam element length and the contact radius, R2) associated with the contact
node.

13.177.4. Contact Forces

In order to satisfy contact compatibility, forces are developed in a direction normal (n-direction) to
the target that will tend to reduce the penetration to an acceptable numerical level. In addition to
normal contact forces, friction forces are developed in directions that are tangent to the target plane.

(13.296)

where:

Fn = normal contact force

Kn = contact normal stiffness (input as FKN on R command)
un = contact gap size

(13.297)

where:

FT = tangential contact force

KT = tangential contact stiffness (input as FKT on R command)
uT = contact slip distance

For orthotropic friction, μeq is computed using the expression:

(13.298)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 581
Element Library

μeq = equivalent coefficient of friction for orthotropic friction

μ1, μ2 = coefficients of friction in first and second principal directions (input as MU1 and MU2
using TB command with Lab = FRIC)

13.178. CONTA178 - 3D Node-to-Node Contact

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Normal Direction None None
Stiffness Matrix
Sliding Direction None None

Load Type Distribution

Element Temperature None - average used for material property evaluation
Nodal Temperature None - average used for material property evaluation

13.178.1. Introduction
CONTA178 represents contact and sliding between any two nodes of any types of elements. This
node-to-node contact element can handle cases when the contact location is known beforehand.

CONTA178 is applicable to 3D geometries. It can also be used in 2D and axisymmetric models by

constraining the UZ degrees of freedom. The element is capable of supporting compression in the
contact normal direction and Coulomb friction in the tangential direction.

13.178.2. Contact Algorithms

Four different contact algorithms are implemented in this element.

• Pure penalty method

• Augmented Lagrange method

• Pure Lagrange multiplier method

• Lagrange multiplier on contact normal penalty on frictional direction

Pure Penalty Method

The Newton-Raphson load vector is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
582 Published: 2023-10-23
 CONTA178 - 3D Node-to-Node Contact

(13.299)

where:

Fn = normal contact force

Fsy = tangential contact force in y direction
Fsz = tangential contact force in z direction

(13.300)

where:

Kn = contact normal stiffness (input FKN on R command)

un = contact gap size

(13.301)

where:

Ks = tangential contact stiffness (input as FKS on R command)

uy = contact slip distance in y direction
μ = coefficient of friction (input as MU on TB command with Lab = FRIC or MP command)

Augmented Lagrange Method

(13.302)

where:

ε = user-defined compatibility tolerance (input as TOLN on R command)

The Lagrange multiplier component of force λ is computed locally (for each element) and iteratively.

Pure Lagrange Multiplier Method

The contact forces (that is, Lagrange multiplier components of forces) become unknown DOFs for
each element. The associated Newton-Raphson load vector is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 583
Element Library

(13.303)

Lagrange Multiplier on Contact Normal Penalty on Frictional Direction

In this method only the contact normal face is treated as a Lagrange multiplier. The tangential forces
are calculated based on penalty method:

(13.304)

13.178.3. Element Damper

The damping capability is only used for modal and transient analyses. Damping is only active in the
contact normal direction when contact is closed. The damping force is computed as:
(13.305)

where:

V = relative velocity between two contact nodes in contact normal direction

Cv1 = constant damping coefficient (input as CV1 on R command)

Cv2 = linear damping coefficient (input as CV2 on R command)

13.178.4. Rigid Coulomb Friction

If you know that a CONTA178 element will be in sliding status throughout the analysis, and that the
relative displacement of the two nodes will be monotonically increasing, the rigid Coulomb friction
option (KEYOPT(10) = 7) can be used to avoid convergence problems. This option removes the stiffness
in the sliding direction, as shown in Figure 13.32: Force-Deflection Relations for Rigid Coulomb Op-
tion (p. 585). Note that if the relative displacement does not increase monotonically, the convergence
characteristics of the rigid Coulomb friction law (KEYOPT(10) = 7) will be worse than for the elastic
Coulomb friction law (KEYOPT(10) = 0).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
584 Published: 2023-10-23
 LINK180 - 3D Spar (or Truss)

Figure 13.32: Force-Deflection Relations for Rigid Coulomb Option

13.179. PRETS179 - Pretension

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix None None

Load Type Distribution

Pretension Force Applied on pretension node K across entire pretension section

13.179.1. Introduction
The element is used to represent a two or three dimensional section for a bolted structure. The pre-
tension section can carry a pretension load. The pretension node (K) on each section is used to control
and monitor the total tension load.

13.179.2. Assumptions and Restrictions

The pretension element is not capable of carrying bending or torsion loads.

13.180. LINK180 - 3D Spar (or Truss)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 585
Element Library

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix; and Equation 11.6 (p. 367) 1
Thermal,
Newton-Raphson
Hydrostatic and Equation 11.6 (p. 367) 2 [1]
Hydrodynamic Load
Vectors
Mass and Stress Equation 11.6 (p. 367), Equation 11.7 (p. 367), and 1
Stiffening Matrices Equation 11.8 (p. 367)

1. For partially submerged elements, the hydrostatic/hydrodynamic load vector integ-

ration point(s) is/are centered on the wetted portion.

Load Type Distribution

Element Temperature Linear along length
Nodal Temperature Linear along length

Reference: Cook et al.([118] (p. 931))

13.180.1. Assumptions and Restrictions

The theory for this element is a reduction of the theory for BEAM189 - 3D 3-Node Beam (p. 601). The
reductions include only 2 nodes, no bending or shear effects, no pressures, and the entire element
as only one integration point.

The element is not capable of carrying bending loads. The stress is assumed to be uniform over the
entire element.

13.181. SHELL181 - 4-Node Shell

Matrix or Vector Shape Functions Integration Points (p. 414)

In-plane:
Equation 11.70 (p. 374),
Stiffness Matrix; and 1 x 1 (KEYOPT(3) = 0)
Equation 11.71 (p. 374), and
Thermal Load Vector 2 x 2 (KEYOPT(3) = 2)
Equation 11.72 (p. 374)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
586 Published: 2023-10-23
 SHELL181 - 4-Node Shell

Matrix or Vector Shape Functions Integration Points (p. 414)

Thru-the-thickness:
1, 3, 5, 7, or 9 per layer for
data input for section general
shell option (KEYOPT(1) = 0)

Thru-the-thickness:
1 per layer for section data
input for membrane shell
option (KEYOPT(1) = 1)

Consistent Mass and Equation 11.70 (p. 374), Closed-form integration

Stress Stiffness Matrices Equation 11.71 (p. 374), and
Equation 11.72 (p. 374)
Lumped Mass Matrix Equation 11.70 (p. 374), Closed-form integration
Equation 11.71 (p. 374),
Equation 11.72 (p. 374)
Transverse Pressure Equation 11.72 (p. 374) 2x2
Load Vector
Edge Pressure Load Equation 11.70 (p. 374) and 2
Vector Equation 11.71 (p. 374) specialized to the
edge

Load Type Distribution

Element Temperature Bilinear in plane of element, linear thru each layer
Nodal Temperature Bilinear in plane of element, constant thru thickness
Pressure Bilinear in plane of element and linear along each edge

References: Ahmad([1] (p. 925)), Cook([5] (p. 925)), Dvorkin([97] (p. 930)), Dvorkin([98] (p. 930)), Bathe and
Dvorkin([99] (p. 930)), Allman([114] (p. 931)), Cook([115] (p. 931)), MacNeal and Harder([116] (p. 931))

13.181.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.181.2. Assumptions and Restrictions

Normals to the centerplane are assumed to remain straight after deformation, but not necessarily
normal to the centerplane.

Each set of integration points thru a layer (in the r direction) is assumed to have the same element
(material) orientation.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 587
Element Library

13.181.3. Assumed Displacement Shape Functions

The assumed displacement and transverse shear strain shape functions are given in Shape Func-
tions (p. 365). The basic functions for the transverse shear strain have been changed to avoid shear
locking (Dvorkin([97] (p. 930)), Dvorkin([98] (p. 930)), Bathe and Dvorkin([99] (p. 930))).

13.181.4. Membrane Option

A membrane option is available for SHELL181 if KEYOPT(1) = 1. For this option, there is no bending
stiffness or rotational degrees of freedom. There is only one integration point per layer, regardless of
other input.

13.181.5. Warping
A warping factor is computed as:
(13.306)

where:

D = component of the vector from the first node to the fourth node parallel to the element
normal
t = average thickness of the element

If φ > 1.0, a warning message is printed.

13.181.6. Shear Correction

The element uses an equivalent energy method to compute shear-correction factors. These factors
are predetermined based on the section lay-up at the start of solution.

13.182. PLANE182 - 2D 4-Node Structural Solid

Matrix or Vector Geometry Shape Functions Integration Points (p. 414)

Equation 11.122 (p. 377) and 2 x 2 if KEYOPT(1) = 0, 2, or 3

Stiffness and Stress Quad 1 if KEYOPT(1) = 1
Equation 11.123 (p. 377)
Stiffness Matrices; and
Thermal Load Vector Equation 11.101 (p. 376) and
Triangle 1
Equation 11.102 (p. 376)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
588 Published: 2023-10-23
 PLANE183 - 2D 8-Node Structural Solid

Matrix or Vector Geometry Shape Functions Integration Points (p. 414)

Quad 2x2
Mass Matrix Same as stiffness matrix
Triangle 1
Same as stiffness matrix, specialized to
Pressure Load Vector 2
face

Load Type Distribution

Bilinear across element, constant thru thickness or around
Element Temperature
circumference
Nodal Temperature Same as element temperature distribution
Pressure Linear along each face

13.182.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.182.2. Theory
If KEYOPT(1) = 0, this element uses method (selective reduced integration technique for volumetric
terms) (Hughes([221] (p. 937)), Nagtegaal et al.([222] (p. 937))).

If KEYOPT(1) = 1, the uniform reduced integration technique (Flanagan and Belytschko([233] (p. 938)))
is used.

If KEYOPT(1) = 2 or 3, the enhanced strain formulations from the work of Simo and Rifai([318] (p. 943)),
Simo and Armero([319] (p. 943)), Simo et al.([320] (p. 943)), Andelfinger and Ramm([321] (p. 943)), and
Nagtegaal and Fox([322] (p. 943)) are used. It introduces 5 internal degrees of freedom to prevent
shear and volumetric locking for KEYOPT(1) = 2, and 4 internal degrees of freedom to prevent shear
locking for KEYOPT(1) = 3. If mixed u-P formulation is employed with the enhanced strain formulations,
only 4 degrees of freedom for overcoming shear locking are activated.

13.183. PLANE183 - 2D 8-Node Structural Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 589
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.137 (p. 378) and
Stiffness and Stress Quad 2x2
Equation 11.138 (p. 378)
Stiffness Matrices; and
Thermal Load Vector Equation 11.114 (p. 377) and
Triangle 3
Equation 11.115 (p. 377)
Quad 3x3
Mass Matrix Same as stiffness matrix
Triangle 3
Pressure Load Vector Same as stiffness matrix, specialized to the face 2 along face

Load Type Distribution

Same as shape functions across element, constant thru thickness
Element Temperature
or around circumference
Nodal Temperature Same as element temperature distribution
Pressure Linear along each face

Reference: Zienkiewicz([40] (p. 926))

13.183.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.183.2. Assumptions and Restrictions

A dropped midside node implies that the face is and remains straight.

13.184. MPC184 - Multipoint Constraint

z

y
I

Y
x
J
X
Z

MPC184 represents a general class of multipoint constraint elements that implement kinematic constraints
using Lagrange multipliers. The elements are loosely classified here as "constraint elements" and "joint
elements". All of these elements are used in situations that require you to impose some kind of constraint
to meet certain requirements. Since these elements are implemented using Lagrange multipliers, the
constraint forces and moments are available for output purposes. The different constraint elements and
joint elements are identified by KEYOPT(1).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
590 Published: 2023-10-23
 MPC184 - Multipoint Constraint

13.184.1. Slider Element

The slider element (KEYOPT(1) = 3) is a 3-node element that allows a dependent node to slide on a
line joining two independent nodes.

Figure 13.33: 184.2 Slider Constraint Geometry

K

Y J

X
Z

The constraints required to maintain the dependent node on the line joining the two independent
nodes are as follows:

Define a unit vector n as:

(13.307)

where:

I J
x ,x = position vectors of nodes I and J in the current configuration

Identify unit vectors l and m such that l, m, and n form an orthonormal set.

The constraints are then defined as:

(13.308)
(13.309)

where:

xk = position vector of the node K in the current configuration

Let i, j, and k be the global base vectors. Then we can define the unit vector l as:
(13.310)

If n = l, then:

(13.311)

Finally, the unit vector m is defined as:

(13.312)

The virtual work contributions are obtained from taking the variations of the above equations.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 591
Element Library

13.184.2. Joint Elements

The equations for the constraints imposed in joint elements are described in the individual element
descriptions:

MPC184-Revolute
MPC184-Universal
MPC184-Slot
MPC184-Point
MPC184-Translational
MPC184-Cylindrical
MPC184-Planar
MPC184-Weld
MPC184-Orient
MPC184-Spherical
MPC184-General
MPC184-Screw

13.185. SOLID185 - 3D 8-Node Structural Solid

SOLID185 is available in two forms:

• Standard (nonlayered) structural solid (KEYOPT(3) = 0, the default) - see SOLID185 - 3D 8-Node
Structural Solid (p. 592).

• Layered structural solid (KEYOPT(3) = 1) - see SOLID185 - 3D 8-Node Layered Solid (p. 593).

13.185.1. SOLID185 - 3D 8-Node Structural Solid

Matrix or Vector Shape Functions Integration Points (p. 414)

Stiffness and Stress 2 x 2 x 2 if KEYOPT(2) = 0, 2, or 3
Equation 11.217 (p. 387),
Stiffness Matrices; and 1 if KEYOPT(2) = 1
Equation 11.218 (p. 387), and
Thermal Load Vector 1 if element shape is tetrahedral
Equation 11.219 (p. 387)

Mass Matrix Same as stiffness matrix 2x2x2

1 if KEYOPT(2) = 1

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
592 Published: 2023-10-23
 SOLID185 - 3D 8-Node Structural Solid

Matrix or Vector Shape Functions Integration Points (p. 414)

Equation 11.70 (p. 374)
Quad and 2x2
Equation 11.71 (p. 374)
Pressure Load Vector
Equation 11.50 (p. 373)
Triangle and 3
Equation 11.51 (p. 373)

Load Type Distribution

Element Temperature Trilinear thru element
Nodal Temperature Trilinear thru element
Pressure Bilinear across each face

13.185.2. SOLID185 - 3D 8-Node Layered Solid

Matrix or Vector Shape Functions Integration Points (p. 414)

Stiffness and Stress In-plane:
Stiffness Matrices; and 2x2
Thermal Load Vector Equation 11.217 (p. 387), Thru-the-thickness:
Equation 11.218 (p. 387), and 2 if no shell section defined.
Equation 11.219 (p. 387) 1, 3, 5, 7, or 9 per layer if a
shell section is defined

Mass Matrix Same as stiffness matrix Same as stiffness matrix

Equation 11.70 (p. 374)
Quad and 2x2
Equation 11.71 (p. 374)
Pressure Load Vector
Equation 11.50 (p. 373)
Triangle and 3
Equation 11.51 (p. 373)

Load Type Distribution

Element Temperature Bilinear in plane of element, linear thru each layer
Nodal Temperature Trilinear thru element
Pressure Bilinear across each face

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 593
Element Library

13.185.3. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.185.4. Theory
If KEYOPT(2) = 0 (not applicable to layered SOLID185), this element uses method (selective reduced
integration technique for volumetric terms) (Hughes([221] (p. 937)), Nagtegaal et al.([222] (p. 937))).

If KEYOPT(2) = 1 (not applicable to layered SOLID185), the uniform reduced integration technique
(Flanagan and Belytschko([233] (p. 938))) is used.

If KEYOPT(2) = 2 or 3, the enhanced strain formulations from the work of Simo and Rifai([318] (p. 943)),
Simo and Armero([319] (p. 943)), Simo et al.([320] (p. 943)), Andelfinger and Ramm([321] (p. 943)), and
Nagtegaal and Fox([322] (p. 943)) are used. It introduces 13 internal degrees of freedom to prevent
shear and volumetric locking for KEYOPT(2) = 2, and 9 degrees of freedom to prevent shear locking
only for KEYOPT(2) = 3. If mixed u-P formulation is employed with the enhanced strain formulations,
only 9 degrees of freedom for overcoming shear locking are activated.

13.185.5. Shear Correction

The element does not perform interlaminar shear correction. Stresses are calculated from strains as
is.

13.186. SOLID186 - 3D 20-Node homogeneous/Layered Structural Solid

SOLID186 is available in two forms:

• homogeneous (nonlayered) structural solid (KEYOPT(3) = 0, the default) - see SOLID186 - 3D 20-
Node Homogeneous Structural Solid (p. 594).

• Layered structural solid (KEYOPT(3) = 1) - see SOLID186 - 3D 20-Node Layered Structural Sol-
id (p. 595).

13.186.1. SOLID186 - 3D 20-Node Homogeneous Structural Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
594 Published: 2023-10-23
 SOLID186 - 3D 20-Node homogeneous/Layered Structural Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.236 (p. 388), 14 if KEYOPT(2) = 1
Brick Equation 11.237 (p. 388), and 2 x 2 x 2 if KEYOPT(2) = 0
Equation 11.238 (p. 388)
Equation 11.211 (p. 386),
Wedge Equation 11.212 (p. 386), and 3x3
Stiffness and Stress Equation 11.213 (p. 386)
Stiffness Matrices; and
Thermal Load Vector Equation 11.195 (p. 384),
Pyramid Equation 11.196 (p. 384), and 2x2x2
Equation 11.197 (p. 384)
Equation 11.187 (p. 383),
Tet Equation 11.188 (p. 383), and 4
Equation 11.189 (p. 383)
14 if brick. If other
Mass Matrix Same as stiffness matrix. shapes, same as stiffness
matrix
Equation 11.82 (p. 375) and
Quad 3x3
Equation 11.83 (p. 375)
Pressure Load Vector
Equation 11.58 (p. 373) and
Triangle 6
Equation 11.59 (p. 373)

Load Type Distribution

Element Temperature Same as shape functions thru element
Nodal Temperature Same as shape functions thru element
Pressure Bilinear across each face

13.186.2. SOLID186 - 3D 20-Node Layered Structural Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
In-plane:
Stiffness and Stress Equation 11.236 (p. 388), 2x2
Stiffness Matrices; and Brick Equation 11.237 (p. 388), and Thru-the-thickness:
Thermal Load Vector Equation 11.238 (p. 388) 2 if no shell section
defined.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 595
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
1, 3, 5, 7, or 9 per
layer if a shell
section is defined

In-plane: 3
Thru-the-thickness:
2 if no shell section
Equation 11.211 (p. 386),
defined.
Wedge Equation 11.212 (p. 386), and
1, 3, 5, 7, or 9 per
Equation 11.213 (p. 386)
layer if a shell
section is defined

In-plane:
3 x 3 if brick
3 if wedge
Mass Matrix Same as stiffness matrix.
Thru-the-thickness:
Same as stiffness matrix

Equation 11.82 (p. 375) and

Quad 3x3
Equation 11.83 (p. 375)
Pressure Load Vector
Equation 11.58 (p. 373) and
Triangle 6
Equation 11.59 (p. 373)

Load Type Distribution

Element Temperature Bilinear in plane of element, linear thru each layer
Nodal Temperature Same as shape functions thru element
Pressure Bilinear across each face

Reference: Zienkiewicz([40] (p. 926))

13.186.3. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.186.4. Shear Correction

The element does not perform interlaminar shear correction. Stresses are calculated from strains as
is.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
596 Published: 2023-10-23
 BEAM188 - 3D 2-Node Beam

13.187. SOLID187 - 3D 10-Node Tetrahedral Structural Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness, Mass, and
Stress Stiffness Matrices; Equation 11.187 (p. 383), Equation 11.188 (p. 383),
4
and Thermal Load and Equation 11.189 (p. 383)
Vector
Equation 11.187 (p. 383), Equation 11.188 (p. 383),
Pressure Load Vector and Equation 11.189 (p. 383) specialized to the 6
face

Load Type Distribution

Element Temperature Same as shape functions
Nodal Temperature Same as shape functions
Pressure Linear over each face

Reference: Zienkiewicz([40] (p. 926))

13.187.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.188. BEAM188 - 3D 2-Node Beam

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 597
Element Library

Matrix or Integration
Option Shape Functions
Vector Points (p. 414)
Stiffness and Linear (KEYOPT(3) = Equation 11.6 (p. 367), Along the length: 1
Stress Stiffness 0) Equation 11.7 (p. 367),
Matrices, and Equation 11.8 (p. 367), Across the section:
Thermal and Equation 11.9 (p. 367), see text below
Newton-Raphson Equation 11.10 (p. 367),
Load Vectors and
Equation 11.11 (p. 367)
Quadratic Equation 11.19 (p. 368),Along the length: 2
(KEYOPT(3) = 2) Equation 11.20 (p. 368),
Equation 11.21 (p. 368),Across the section:
Equation 11.22 (p. 368),see text below.
Equation 11.23 (p. 368),
and
Equation 11.24 (p. 368)
Cubic (KEYOPT(3) = Equation 11.27 (p. 368),Along the length: 3
3) Equation 11.28 (p. 368),
Equation 11.29 (p. 369),Across the section:
Equation 11.30 (p. 369),see text below.
Equation 11.31 (p. 369),
and
Equation 11.32 (p. 369)
Consistent Mass Linear (KEYOPT(3) = Equation 11.6 (p. 367), Along the length: 2
Matrix and 0) Equation 11.7 (p. 367), [1]
Pressure, Equation 11.8 (p. 367),
Hydrostatic, and Equation 11.9 (p. 367), Across the section:
Hydrodynamic Equation 11.10 (p. 367),1
Load Vectors and
Equation 11.11 (p. 367)
Quadratic Equation 11.19 (p. 368),Along the length: 3
(KEYOPT(3) = 2) Equation 11.20 (p. 368),[1]
Equation 11.21 (p. 368),
Equation 11.22 (p. 368),Across the section:
Equation 11.23 (p. 368),1
and
Equation 11.24 (p. 368)
Cubic (KEYOPT(3) = Equation 11.27 (p. 368),Along the length: 4
3) Equation 11.28 (p. 368),[1]
Equation 11.29 (p. 369),
Equation 11.30 (p. 369),Across the section:
Equation 11.31 (p. 369),1
and
Equation 11.32 (p. 369)
Lumped Mass Linear (KEYOPT(3) = Equation 11.6 (p. 367), Along the length: 2
Matrix 0) Equation 11.7 (p. 367),
and Across the section:
Equation 11.8 (p. 367) 1

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
598 Published: 2023-10-23
 BEAM188 - 3D 2-Node Beam

Matrix or Integration
Option Shape Functions
Vector Points (p. 414)
Quadratic Equation 11.19 (p. 368),Along the length: 3
(KEYOPT(3) = 2) Equation 11.20 (p. 368),
and Across the section:
Equation 11.21 (p. 368) 1
Cubic (KEYOPT(3) = Equation 11.27 (p. 368),Along the length: 4
3) Equation 11.28 (p. 368),
and Across the section:
Equation 11.29 (p. 369) 1

1. For partially submerged elements, the hydrostatic/hydrodynamic load vector integ-

ration point(s) is/are centered on the wetted portion.

Load Type Distribution

Element Temperature Bilinear across cross-section and linear along length
Nodal Temperature Constant across cross-section, linear along length
Linear along length. The pressure is assumed to act along the
Pressure
element x-axis.

References: Simo and Vu-Quoc([238] (p. 938)), Ibrahimbegovic([239] (p. 938)).

13.188.1. Assumptions and Restrictions

The element is based on Timoshenko beam theory; therefore, shear deformation effects are included.
It uses three components of strain, one (axial) direct strain and two (transverse) shear strains. The
element is well-suited for linear, large rotation, and/or large strain nonlinear applications. If KEYOPT(2)
= 0, the cross-sectional dimensions are scaled uniformly as a function of axial strain in nonlinear
analysis such that the volume of the element is preserved.

The element includes stress stiffness terms, by default, in any analysis using large deformation
(NLGEOM,ON). The stress stiffness terms provided enable the elements to analyze flexural, lateral and
torsional stability problems (using eigenvalue buckling or collapse studies with arc length methods).
Pressure load stiffness (Pressure Load Stiffness (p. 45)) is included.

Transverse-shear strain is constant through cross-section; that is, cross sections remain plane and
undistorted after deformation. Higher-order theories are not used to account for variation in distribution
of shear stresses. A shear-correction factor is calculated in accordance with in the following references:

• Schramm, U., L. Kitis, W. Kang, and W.D. Pilkey. "On the Shear Deformation Coefficient in Beam
Theory." [Finite Elements in Analysis and Design, The International Journal of Applied Finite Elements
and Computer Aided Engineering]. 16 (1994): 141-162.

• Pilkey, Walter D. [Formulas for Stress, Strain, and Structural Matrices]. New Jersey: Wiley, 1994.

The element can be used for slender or stout beams. Due to the limitations of first order shear de-
formation theory, only moderately "thick" beams may be analyzed. Slenderness ratio of a beam
structure may be used in judging the applicability of the element. It is important to note that this
ratio should be calculated using some global distance measures, and not based on individual element
dimensions. A slenderness ratio greater than 30 is recommended.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 599
Element Library

These elements support only elastic relationships between transverse-shear forces and transverse-
shear strains. Orthotropic elastic material properties with bilinear and multilinear isotropic hardening
plasticity options (BISO, MISO) may be used. Transverse-shear stiffnesses can be specified using real
constants.

The St. Venant warping functions for torsional behavior is determined in the undeformed state, and
is used to define shear strain even after yielding. The element does not provide options to recalculate
the torsional shear distribution on cross sections during the analysis and possible partial plastic
yielding of cross section. As such, large inelastic deformation due to torsional loading should be
treated with caution and carefully verified.

The elements are provided with section relevant quantities (area of integration, position, Poisson
function, function derivatives, etc.) automatically at a number of section points by the use of section
commands. Each section is assumed to be an assembly of predetermined number of nine-node cells
which illustrates a section model of a rectangular section. Each cell has four integration points.

Figure 13.34: Section Model

Section Nodes
Section Integration Points

Rectangular Section

When the material has inelastic behavior or the temperature varies across the section, constitutive
calculations are performed at each of the section integration points. For all other cases, the element
uses the precalculated properties of the section at each element integration point along the length.
The restrained warping formulation used may be found in Timoshenko and Gere([247] (p. 938)) and
Schulz and Fillippou([248] (p. 938)).

13.188.2. Ocean Effects

The element has the same ocean effects as described in section Ocean Effects (p. 673), specialized to
other cross sections, but with the following limitations:

• Sections with internal voids (such as circular tubes (SECTYPE,,BEAM,CTUBE)) cannot have a separate
input for internal pressure; rather, the external pressure is used (similar to setting KFLOOD = 1 on
the OCDATA command).

• Hydrostatic crushing of the cross-section cannot be considered directly as no strains exists in the
element y and z directions. To ensure that the linear elastic materials are handled correctly, axial
loads are adjusted by:

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
600 Published: 2023-10-23
 BEAM189 - 3D 3-Node Beam

= adjustment to the axial term of the element load vector

= external pressure
= Poisson’s ratio (input via PRXY on the MP command)
= cross-section area

Compared to the working stress in the material, is normally very small. In deep-water applications,
however, the approximation is more significant.

13.188.3. Stress Evaluation

Several stress-evaluation options exist. The section strains and generalized stresses are evaluated at
element integration points and then linearly extrapolated to the nodes of the element.

If the material is elastic, stresses and strains are available after extrapolation in cross-section at the
nodes of section mesh. If the material is plastic, stresses and strains are moved without extrapolation
to the section nodes (from section integration points).

13.189. BEAM189 - 3D 3-Node Beam

Matrix or Vector Shape Functions Integration Points (p. 414)

Equation 11.19 (p. 368),
Stiffness and Stress Along the length: 2
Equation 11.20 (p. 368),
Stiffness Matrices; and
Equation 11.21 (p. 368),
Thermal and Across the section: see
Equation 11.22 (p. 368),
Newton-Raphson Load BEAM188 - 3D 2-Node
Equation 11.23 (p. 368), and
Vectors Beam (p. 597)
Equation 11.24 (p. 368)
Consistent Mass Matrix
and Pressure, Along the length: 3 [1]
Hydrostatic, and Same as stiffness matrix
Hydrodynamic Load Across the section: 1
Vectors
Equation 11.19 (p. 368), Along the length: 3
Lumped Mass Matrix Equation 11.20 (p. 368), and
Equation 11.21 (p. 368) Across the section: 1

1. For partially submerged elements, the hydrostatic/hydrodynamic load vector integ-

ration point(s) is/are centered on the wetted portion.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 601
Element Library

Load Type Distribution

Element Temperature Bilinear across cross-section and linear along length
Nodal Temperature Constant across cross-section, linear along length
Linear along length. The pressure is assumed to act along the
Pressure
element x-axis.

References: Simo and Vu-Quoc([238] (p. 938)), Ibrahimbegovic([239] (p. 938)).

The theory for this element is identical to that of BEAM188 - 3D 2-Node Beam (p. 597), except that it is
a 3-node beam element.

13.190. SOLSH190 - 3D 8-Node Layered Solid Shell

P zo

z
O M O, P
L I K, L
y yo K N
x
xo
M
J
I N
Prism Option
Z

J
X Y

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness and Stress In-plane:
Stiffness Matrices; and 2x2
Thermal Load Vector
Thru-the-thickness:
Equation 11.217 (p. 387), Equation 11.218 (p. 387), 2 if no shell section
and Equation 11.219 (p. 387) defined
1, 3, 5, 7, or 9 per layer
if a shell section is
defined

Mass Matrix Same as stiffness matrix Same as stiffness

matrix
Equation 11.70 (p. 374) and
Quad 2x2
Equation 11.71 (p. 374)
Pressure Load Vector
Equation 11.50 (p. 373) and
Triangle 3
Equation 11.51 (p. 373)

Load Type Distribution

Element Temperature Bilinear in-plane of element, linear thru each layer
Nodal Temperature Trilinear thru element
Pressure Bilinear across each face

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
602 Published: 2023-10-23
 INTER192 - 2D 4-Node Gasket

13.190.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.190.2. Theory
SOLSH190 is a 3D solid element free of locking in bending-dominant situations. Unlike shell elements,
SOLSH190 is compatible with general 3D constitutive relations and can be connected directly with
other continuum elements.

SOLSH190 utilizes a suite of special kinematic formulations, including assumed strain method (Bathe
and Dvorkin([99] (p. 930))) to overcome locking when the shell thickness becomes extremely small.

SOLSH190 employs enhanced strain formulations (Simo and Rifai([318] (p. 943)), Simo et al.([320] (p. 943)))
to improve the accuracy in in-plane bending situations. The satisfaction of the in-plane patch test is
ensured. Incompatible shape functions are used to overcome the thickness locking.

13.190.3. Shear Correction

The element allows for parabolic enhanced transverse shear strains. Stresses are then calculated from
enhanced shear strains.

13.191. Reserved for Future Use

This section is reserved for future use.

13.192. INTER192 - 2D 4-Node Gasket

L

I
x

Y y
K
X J

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x and y directions 2
Thermal Load Vector Same as stiffness matrix Same as stiffness
matrix

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 603
Element Library

Load Type Distribution

Nodal temperature Same as element temperature distribution

13.192.1. Other Applicable Sections

The theory for this element is described in INTER194 - 3D 16-Node Gasket (p. 604).

13.193. INTER193 - 2D 6-Node Gasket

L
I

x
O

M y
Y
K

X J

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x, quadratic in y direction 2
Thermal Load Vector Same as stiffness matrix Same as stiffness
matrix

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.193.1. Other Applicable Sections

The theory for this element is described in INTER194 - 3D 16-Node Gasket (p. 604).

13.194. INTER194 - 3D 16-Node Gasket

P

X
L
M T W
x
z S
I

U y
O
Z V K
Q N

R
X Y J

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
604 Published: 2023-10-23
 INTER194 - 3D 16-Node Gasket

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x, quadratic in y and z directions 2x2
Thermal Load Vector Same as stiffness matrix Same as stiffness
matrix

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.194.1. Element Technology

The element is designed specially for simulation of gasket joints, where the primary deformation is
confined to the gasket through-thickness direction. The through-thickness deformation of gasket is
decoupled from the other deformations and the membrane (in-plane) stiffness contribution is neglected.
The element offers a direct means to quantify the through-thickness behavior of the gasket joints.
The pressure-deformation behavior obtained from experimental measurement can be applied to the
gasket material. See Gasket Material (p. 109) for detailed description of gasket material options.

The element is composed of bottom and top surfaces. An element midplane is created by averaging
the coordinates of node pairs from the bottom and top surfaces of the elements. The numerical in-
tegration of interface elements is performed in the element midplane. The element formulation is
based on a corotational procedure. The virtual work in an element is written as:

(13.313)

where:

t = traction force across the element

d = closure across the element
Sint = midplane of the interface surfaces

The integration is performed in the corotational equilibrium configuration and the Gauss integration
procedure is used.

The relative deformation between top and bottom surfaces is defined as:
(13.314)

where, uTOP and uBOTTOM are the displacement of top and bottom surfaces of interface elements in
the local element coordinate system based on the midplane of element.

The thickness direction is defined as the normal direction of the mid plane of the element at the in-
tegration point.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 605
Element Library

13.195. INTER195 - 3D 8-Node Gasket

P

M
x
I
z
y
O
Z
N K

X Y J

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x, bilinear in y and z directions 2x2
Thermal Load Vector Same as stiffness matrix Same as stiffness
matrix

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.195.1. Other Applicable Sections

The theory for this element is described in INTER194 - 3D 16-Node Gasket (p. 604).

13.196. Reserved for Future Use

This section is reserved for future use.

13.197. Reserved for Future Use

This section is reserved for future use.

13.198. Reserved for Future Use

This section is reserved for future use.

13.199. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
606 Published: 2023-10-23
 INTER203 - 2D 6-Node Cohesive

13.200. Reserved for Future Use

This section is reserved for future use.

13.201. Reserved for Future Use

This section is reserved for future use.

13.202. INTER202 - 2D 4-Node Cohesive

L
I
x

y
Y

K
X J

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x and y directions 2

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.202.1. Other Applicable Sections

The theory for this element is described in INTER204 - 3D 16-Node Cohesive (p. 608).

13.203. INTER203 - 2D 6-Node Cohesive

L
I

x
O

Y M y
K
X J

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x, quadratic in y direction 2

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 607
Element Library

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.203.1. Other Applicable Sections

The theory for this element is described in INTER204 - 3D 16-Node Cohesive (p. 608).

13.204. INTER204 - 3D 16-Node Cohesive

P

L
X
T x
M W
z S
I

y
U
O
Q V K
N
Z
R
J

X Y

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x, quadratic in y and z directions 2x2

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.204.1. Element Technology

The element is designed specially for simulation of interface delamination and fracture, where the
interface surfaces are represented by a group of interface elements, in which an interfacial constitutive
relationship characterizes the traction separation behavior of the interface. The element offers a direct
means to quantify the interfacial separation behavior. See Cohesive Zone Material (CZM) Model (p. 143)
for detailed description of interface material options.

The virtual work of the element is written as:

(13.315)

where:

t = traction force across the element

d = separation across the element

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
608 Published: 2023-10-23
 Reserved for Future Use

Sint = midplane of the interface surfaces

The integration is performed in the corotational equilibrium configuration and the Gauss integration
procedure is used.

The separation, d, is defined as the relative deformation between top and bottom surfaces as:
(13.316)

where, uTOP and uBOTTOM are the displacement of top and bottom surfaces of interface elements in
the local element coordinate system based on the midplane of element.

The thickness direction is defined as the normal direction of the midplane of the element at the in-
tegration point.

13.205. INTER205 - 3D 8-Node Cohesive

P

M
x
I
z
y
O
Z K
N

Y J
X

Matrix or Vector Shape Functions Integration

Points (p. 414)
Stiffness Matrix Linear in x, bilinear in y and z directions 2x2

Load Type Distribution

Element temperature Based on element shape function, constant through the direction
perpendicular to element plane
Nodal temperature Same as element temperature distribution

13.205.1. Other Applicable Sections

The theory for this element is described in INTER204 - 3D 16-Node Cohesive (p. 608).

13.206. Reserved for Future Use

This section is reserved for future use.

13.207. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 609
Element Library

13.208. SHELL208 - 2-Node Axisymmetric Shell

KEYOPT(3) = 0 KEYOPT(3) = 2

J J

x x
Axial (Y) internal node

z z
I I
Radial (X)

Matrix or Vector Shape Functions Integration

Points (p. 414)
KEYOPT(3) = 0: Along-the-length:
Equation 11.6 (p. 367), Equation 11.7 (p. 367), 1 (KEYOPT(3) = 0)
Stiffness and Stress Stiffness and Equation 11.11 (p. 367) 2 (KEYOPT(3) = 2)
Matrix; and Thermal and Thru-the-thickness:
Newton-Raphson Load KEYOPT(3) = 2: 1, 3, 5, 7, or 9 per layer
Vectors Equation 11.19 (p. 368), Equation 11.20 (p. 368),
and Equation 11.24 (p. 368)

Along-the-length:
2 (KEYOPT(3) = 0)
Mass Matrix and Pressure 3 (KEYOPT(3) = 2)
Same as stiffness matrix
Load Vector Thru-the-thickness:
1, 3, 5, 7, or 9 per layer

Load Type Distribution

Element Temperature Linear along length and linear thru thickness
Nodal Temperature Linear along length and constant thru thickness
Pressure Linear along length

References: Ahmad([1] (p. 925)), Cook([5] (p. 925))

13.208.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.208.2. Assumptions and Restrictions

Normals to the centerline are assumed to remain straight after deformation, but not necessarily normal
to the centerline.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
610 Published: 2023-10-23
 SHELL209 - 3-Node Axisymmetric Shell

13.208.3. Shear Correction

The element uses an equivalent energy method to compute shear-correction factors. These factors
are predetermined based on the section lay-up at the start of solution.

13.209. SHELL209 - 3-Node Axisymmetric Shell

J

Axial (Y)

x
z
I
Radial (X)

Matrix or Vector Shape Functions Integration Points (p. 414)

Stiffness and Stress Stiffness Equation 11.19 (p. 368), Along-the-length: 2
Matrix; and Thermal and Equation 11.20 (p. 368), and Thru-the-thickness:
Newton-Raphson Load Equation 11.24 (p. 368) 1, 3, 5, 7, or 9 per layer
Vectors
Along-the-length: 3
Mass Matrix and Pressure Thru-the-thickness:
Same as stiffness matrix
Load Vector 1, 3, 5, 7, or 9 per layer

Load Type Distribution

Element Temperature Linear along length and linear thru thickness
Nodal Temperature Linear along length and constant thru thickness
Pressure Linear along length

References: Ahmad([1] (p. 925)), Cook([5] (p. 925))

13.209.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.209.2. Assumptions and Restrictions

Normals to the centerline are assumed to remain straight after deformation, but not necessarily normal
to the centerline.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 611
Element Library

13.209.3. Shear Correction

The element uses an equivalent energy method to compute shear-correction factors. These factors
are predetermined based on the section lay-up at the start of solution.

13.210. Reserved for Future Use

This section is reserved for future use.

13.211. Reserved for Future Use

This section is reserved for future use.

13.212. CPT212 - 2D 4-Node Coupled Pore-Pressure-Thermal Mechanical

Solid
L 3 K

K, L
4
2

Y I
(or axial) I J
1 J (Triangular Option -
X (or radial) not recommended)

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.122 (p. 377) and
Quad 2x2
Stiffness and Stress Equation 11.123 (p. 377)
Stiffness Matrices Equation 11.101 (p. 376) and
Triangle 1
Equation 11.102 (p. 376)
Quad 2x2
Mass Matrix Same as stiffness matrix
Triangle 1
Damping matrices Same as stiffness matrix
Pressure and
Same as stiffness matrix, specialized to
Temperature Load 2
edge
Vector
Flow Load Vector Linear across each edge 2

Load Type Distribution

Bilinear across element, constant thru thickness or around
Element Temperature
circumference
Nodal Temperature Same as element temperature distribution
Surface Pressure Load Linear along each edge
Flow and heat flux Linear along each edge

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
612 Published: 2023-10-23
 CPT213 - 2D 8-Node Coupled Pore-Pressure-Thermal Mechanical Solid

13.212.1. Other Applicable Sections

General Element Formulations (p. 49) gives the general element formulations used by this element.

13.213. CPT213 - 2D 8-Node Coupled Pore-Pressure-Thermal Mechanical

Solid
L
3
K
O
K, L, O
4
P
2
N P N
I I
M J
M
J
1 Degenerated
Y triangle
(or axial) KEYOPT(1) = 0

X (or radial)

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.137 (p. 378) and
Quad 2x2
Stiffness and Stress Equation 11.138 (p. 378)
Stiffness Matrices Equation 11.114 (p. 377) and
Triangle 3
Equation 11.115 (p. 377)
Quad 3x3
Mass Matrix Same as stiffness matrix
Triangle 3
Pressure and
Temperature Load Same as stiffness matrix, specialized to the edge 2
Vector
Flow Load Vector Linear across each edge 2

Load Type Distribution

Same as shape functions across element, constant thru thickness
Element Temperature
or around circumference
Nodal Temperature Same as element temperature distribution
Surface Pressure Load Linear along each edge
Flow and heat flux Linear along each edge

Reference: Zienkiewicz(39 (p. 926))

13.213.1. Other Applicable Sections

General Element Formulations (p. 49) gives the general element formulations used by this element.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 613
Element Library

13.213.2. Assumptions and Restrictions

A dropped midside node implies that the edge is and remains straight.

13.214. COMBI214 - 2D Spring-Damper Bearing

K21
Bearing
C21

Rotor
K11

I J

C11

K12

C12

Y K22 C22

Matrix or Vector Shape Integration Points (p. 414)

Functions
Stiffness and Damping None None
Matrices
Stress Stiffening Matrix None None

13.214.1. Matrices
The following sections describe the different matrices used by the COMBI214 element:

[Link]. Matrices for User-Input Characteristics (KEYOPT(1) = 0)

[Link].1. General Case
If KEYOPT(2) = 0, the element lies in the (XY) plane and the stiffness, damping and stress-stiffness
matrices in nodal coordinates are:

(13.317)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
614 Published: 2023-10-23
 COMBI214 - 2D Spring-Damper Bearing

(13.318)

(13.319)

where:

K11, K12, K21, K22 = stiffness coefficients (input as K11, etc. on the R command)
C11, C12, C21, C22 = damping coefficients (input as C11, etc. on R command)

L1 = distance between the two nodes I and J

L2 = distance between the two nodes K and J

The matrices for KEYOPT(2) equals 1 or 2 are developed analogously.

Stiffness and/or damping matrices may depend upon the rotational velocity (input through
OMEGA or CMOMEGA) if real constants are defined as table parameters.

The mass matrix is:

(13.320)

where:

M11, M12, M21, M22 = mass coefficients (input as M11, etc. on the R command)
s1, s2 = coefficients based on the KEYOPT(6) setting:

KEYOPT(6) s1 s2
1 1 0
2 0.5 0.5
3 0 1

[Link].2. The Case of Rotating Reference Frame Dynamics

In a rotating reference frame analysis (CORIOLIS,ON,,,OFF), additional terms are included if the
bearing is stationary. The equation of the element forces in the stationary reference frame is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 615
Element Library

(13.321)

where:

{Fe} = element forces in the stationary reference frame

{r} = displacement vector in the stationary reference frame

Following the discussion in Rotating Damping Matrix (p. 705), the element forces in the rotating
reference frame can be expressed as:
(13.322)

where:

{Fe'} = element forces in the rotating reference frame

{r'} = displacement vector in the rotating reference frame

In a linear analysis, if the rotating damping effect is activated (RotDamp = ON on the CORIOLIS
command) and C11 = C22 with C12 = C21 = 0, the second term is included and will modify the
apparent stiffness of the element.

In the general case, the element forces are periodic at twice the rotational velocity and are only
included in a transient analysis with NROPT,FULL.

[Link]. Equations for Reynolds Equation Integration (KEYOPT(1) > 0)

For information on equations used for Reynolds equation integration, see Finite Length Formulation
for 2D Elements (p. 237).

13.214.2. Output Quantities

The following sections describe the output quantities from the COMBI214 element:

[Link]. Output for User-Input Characteristics (KEYOPT(1) = 0)

The stretch is computed as:

(13.323)

(13.324)

where:

u', v', w' = displacements in nodal Cartesian coordinates (UX, UY, UZ)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
616 Published: 2023-10-23
 CPT215 - 3D 8-Node Coupled Pore-Pressure-Thermal Mechanical Solid

The static forces are computed as:

(13.325)
(13.326)

Finally, if a nonlinear transient dynamic (ANTYPE,TRANS, with TIMINT,ON) analysis is performed, a

damping force is computed:

The damping forces are computed as:

(13.327)
(13.328)

where:

v1, v2 = relative velocities

Relative velocities are computed using Equation 13.323 (p. 616) and Equation 13.324 (p. 616), where
the nodal displacements u', v', and w' are replaced with the nodal Newmark velocities.

[Link]. Output for Reynolds Equation Integration (KEYOPT(1) > 0)

The pressure forces are calculated by integration of the pressure on the bearing surface. For more
information, see Finite Length Formulation for 2D Elements (p. 237).

13.215. CPT215 - 3D 8-Node Coupled Pore-Pressure-Thermal Mechanical

Solid
4 M O,P
P
5 N
O
I K,L
6
J
M Prism Option
N
3 M,N,O,P
2
L I
K,L
K
J
Z Tetrahedral Option -
I not recommended
1
Y J
X

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness and Stress
Stiffness Matrices Equation 11.217 (p. 387), Equation 11.218 (p. 387),
2x2x2
and Equation 11.219 (p. 387)

Mass Matrix Same as stiffness matrix 2x2x2

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 617
Element Library

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Equation 11.70 (p. 374) and
Pressure and Quad 2x2
Equation 11.71 (p. 374)
Temperature Load
Vector Equation 11.50 (p. 373) and
Triangle 3
Equation 11.51 (p. 373)
Same as
Damping matrices
stiffness
Same as
Flow Load Vector Pressure
Load Vector

Load Type Distribution

Element Temperature Trilinear thru element
Nodal Temperature Trilinear thru element
Surface Pressure Load Bilinear across each face
Flow and heat flux Bilinear across each face

13.215.1. Other Applicable Sections

General Element Formulations (p. 49) gives the general element formulations used by this element.

13.216. CPT216 - 3D 20-Node Coupled Pore-Pressure-Thermal Mechanical

Solid
M,N,O,P,U,V,W,X
Y A,B
Z
K,L,S
I T R
5 P 4 Q
J
W
X Tetrahedral Option
O
6 M,N,O,P,U,V,W,X
M
B V A
Y B Z
U
N T
A L K
2 3 I S
Y R
L S Q J
T Z Pyramid Option
K X
I O,P,W
Z M
R
Q 1 U V A,B
J Y N
Y Z
I K,L,S
X T
Q R
J
Prism Option

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.236 (p. 388),
Stiffness and Stress
Brick Equation 11.237 (p. 388), and 14
Stiffness Matrices
Equation 11.238 (p. 388)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
618 Published: 2023-10-23
 CPT217 - 3D 10-Node Coupled Pore-Pressure-Thermal Mechanical Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Equation 11.211 (p. 386),
Wedge Equation 11.212 (p. 386), and 3x3
Equation 11.213 (p. 386)
Equation 11.195 (p. 384),
Pyramid Equation 11.196 (p. 384), and 2x2x2
Equation 11.197 (p. 384)
Equation 11.187 (p. 383),
Tet Equation 11.188 (p. 383), and 4
Equation 11.189 (p. 383)
3 x 3 x 3 if brick. If
Mass Matrix Same as stiffness matrix. other shapes, same
as stiffness matrix.
Equation 11.82 (p. 375) and
Pressure and Quad 3x3
Equation 11.83 (p. 375)
Temperature Load
Vector Equation 11.58 (p. 373) and
Triangle 6
Equation 11.59 (p. 373)
Equation 11.82 (p. 375) and
Quad 3x3
Equation 11.83 (p. 375)
Flow Load Vector
Equation 11.58 (p. 373) and
Triangle 6
Equation 11.59 (p. 373)
Damping Matrices Same as stiffness matrix

Load Type Distribution

Element Temperature Bilinear in plane of element
Nodal Temperature Same as shape functions thru element
Surface Pressure Load Bilinear across each face
Flow and heat flux Bilinear across each face

13.216.1. Other Applicable Sections

General Element Formulations (p. 49) gives the general element formulations used by this element.

13.217. CPT217 - 3D 10-Node Coupled Pore-Pressure-Thermal Mechanical

Solid
L

4
R
P
Q 3
Y 2
O K

X I N
M
Z
1
J

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 619
Element Library

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Equation 11.187 (p. 383),
Stiffness, Mass, and Stress Stiffness
Equation 11.188 (p. 383), and 4
Matrices
Equation 11.189 (p. 383)
Equation 11.187 (p. 383),
Equation 11.188 (p. 383), and
Pressure and Temperature Load Vector 6
Equation 11.189 (p. 383)
specialized to the face
Equation 11.187 (p. 383),
Equation 11.188 (p. 383), and
Flow Load Vector 6
Equation 11.189 (p. 383)
specialized to the face
Mass Matrix Same as stiffness matrix
Damping Matrices Same as stiffness matrix

Load Type Distribution

Element Temperature Same as shape functions
Nodal Temperature Same as shape functions
Surface Pressure Load Linear over each face
Flow and heat flux Linear across each face

13.217.1. Other Applicable Sections

General Element Formulations (p. 49) gives the general element formulations used by this element.

13.218. FLUID218 - 3D Hydrodynamic Bearing Element

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Fluid Stiffness Matrix, Internal Fluid
Quad Equation 11.77 (p. 374) 2 x 2 (4-node)
Forces Vector
Quad Equation 11.70 (p. 374),
Interaction Matrix Equation 11.71 (p. 374), 2 x 2 (4-node)
Equation 11.77 (p. 374)

13.218.1. Other Applicable Sections

Hydrodynamic Bearing (p. 236) describes the governing hydrodynamic lubrication equations used to
form the element matrices and vectors.

13.218.2. Assumptions and Restrictions

The element normal must lie in the global XY plane as the bearing length is along the global Z-direc-
tion.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
620 Published: 2023-10-23
 PLANE222 - 2D 4-Node Coupled-Field Solid

13.219. Reserved for Future Use

This section is reserved for future use.

13.220. Reserved for Future Use

This section is reserved for future use.

13.221. Reserved for Future Use

This section is reserved for future use.

13.222. PLANE222 - 2D 4-Node Coupled-Field Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Stiffness and Stress Equation 11.122 (p. 377) and
Quad 2x2
Stiffness Matrices; and Equation 11.123 (p. 377)
Thermal Expansion Load Equation 11.101 (p. 376) and
Vector Triangle 1
Equation 11.102 (p. 376)
Quad 2x2
Mass Matrix Same as stiffness matrix
Triangle 1
Pressure Load Vector Same as stiffness matrix, specialized to the face 2
Thermal Conductivity Quad Equation 11.130 (p. 377) 2x2
Matrix and Heat
Generation Load Vector Triangle Equation 11.109 (p. 376) 1
Specific Heat Matrix Same as thermal conductivity matrix.
Convection Surface Same as thermal conductivity matrix, specialized
2
Matrix and Load Vector to the face
Dielectric Permittivity Quad Equation 11.131 (p. 377) 2x2
and Electrical
Conductivity Matrices;
Charge Density, Joule Triangle Equation 11.110 (p. 376) 1[a]
Heating, and Peltier
Heat Flux Load Vectors

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 621
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Diffusivity Matrix and Quad Equation 11.132 (p. 377) 2x2
Diffusing Substance
Generation Load Vector Triangle Equation 11.111 (p. 376) 1[a]
Diffusion Damping Same as diffusivity matrix. If KEYOPT(10)=1, matrix is diagonalized as
Matrix described in Lumped Matrices (p. 422)
Diffusion Flux Load Same as diffusivity matrix, specialized to the
2
Vector face
Thermoelastic Stiffness
Same as combination of stiffness and thermal conductivity matrices
and Damping Matrices
Piezoelectric and Same as combination of stiffness matrix and dielectric matrix
Electroelastic Coupling
Matrix
Seebeck Coefficient Same as combination of electrical conductivity and thermal conductivity
Coupling Matrix matrices
Surface Charge Density Same as dielectric matrix, specialized to the face 2
Load Vector
[a] Axisymmetric coupled-field analyses without structural degrees of freedom use a 3-point
integration rule for all load vector and matrix calculations.

13.222.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. Heat Flow (p. 207) describes the derivation of the thermal element matrices and load
vectors as well as heat flux evaluations. Electromagnetics (p. 163) describes the derivation of the
electric element matrices and load vectors, as well as electric field evaluations. Piezoelectrics (p. 334)
discusses the piezoelectric capability used by the element. Piezoresistivity (p. 347) discusses the
piezoresistive effect. Thermoelectrics (p. 348) discusses the thermoelectric effects. Thermoelasti-
city (p. 329) discusses the thermoelastic effects. Electroelasticity (p. 345) discusses the Maxwell stress
electroelastic coupling. Thermoplasticity (p. 332) discusses the thermoplastic effect. Diffusion (p. 305)
describes the derivation of diffusion matrices and load vectors. Structural-Diffusion Coupling (p. 354)
discusses diffusion strain coupling.

13.222.2. Theory
This element uses the method (selective reduced integration technique for volumetric terms)
(Hughes([221] (p. 937)), Nagtegaal et al. ([222] (p. 937)))

If KEYOPT(6) = 2 or 3, the enhanced strain formulation from the works of Simo and Rifai([318] (p. 943)),
Simo and Armero([319] (p. 943)), Simo et al.([320] (p. 943)), Andelfinger and Ramm([321] (p. 943)), and
Nagtegaal and Fox([322] (p. 943)) is used. It introduces additional internal (user-inaccessible) structural
degrees of freedom (DOFs) to prevent shear and volumetric locking (KEYOPT(6) = 2) or shear locking
only (KEYOPT(6) = 3). These internal DOFs are created and condensed out at the element level.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
622 Published: 2023-10-23
 PLANE223 - 2D 8-Node Coupled-Field Solid

13.223. PLANE223 - 2D 8-Node Coupled-Field Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Stiffness and Stress Equation 11.137 (p. 378) and
Quad 2 x 2[a]
Stiffness Matrices; Equation 11.138 (p. 378)
Thermal Expansion, Equation 11.114 (p. 377) and
Diffusion Expansion Triangle 3
Equation 11.115 (p. 377)
Quad 3x3
Mass Matrix Same as stiffness matrix
Triangle 3
Pressure Load Vector Same as stiffness matrix, specialized to the face 2
Thermal Conductivity Quad Equation 11.141 (p. 378) 2x2
Matrix and Heat
Generation Load Vector Triangle Equation 11.119 (p. 377) 3
Same as thermal conductivity matrix. If KEYOPT(10) = 1 or 2, matrix is
Specific Heat Matrix
diagonalized as described in Lumped Matrices (p. 422).
Convection Surface Same as thermal conductivity matrix, specialized
2
Matrix and Load Vector to the face
Dielectric Permittivity Quad Equation 11.142 (p. 378) 2x2
and Electrical
Conductivity Matrices;
Charge Density, Joule Triangle Equation 11.120 (p. 377) 3
Heating, and Peltier
Heat Flux Load Vectors
Magnetic Reluctivity 2x2
Matrix, Eddy Current
Damping Matrices; Quad Equation 11.140 (p. 378) or
Source Current and
Remnant Magnetization 3x3
Load Vectors Triangle Equation 11.117 (p. 377) 3
Diffusivity Matrix and Quad Equation 11.143 (p. 378) 2x2
Diffusing Substance
Generation Load Vector Triangle Equation 11.121 (p. 377) 3
Diffusion Damping Same as diffusivity matrix. If KEYOPT(10)=1, matrix is diagonalized as
Matrix described in Lumped Matrices (p. 422)
Diffusion Flux Load Same as diffusivity matrix, specialized to the 2
Vector face

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 623
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Thermoelastic stiffness Same as combination of stiffness and thermal conductivity matrices
and Damping Matrices
Piezoelectric and
Electroelastic Coupling Same as combination of stiffness matrix and dielectric matrix
Matrix
Seebeck Coefficient Same as combination of electrical conductivity and thermal conductivity
Coupling Matrix matrices
Magneto-Electric and
Electromagnetic Same as combination of magnetic reluctivity and electric conductivity
(velocity current) matrices
Coupling Matrices
Magneto-Elastic Stiffness
Same as combination of stiffness and magnetic reluctivity matrices
Matrices
Diffusion-Elastic Stiffness Same as combination of stiffness and diffusivity matrices
Matrix
Surface Charge Density
Same as dielectric matrix, specialized to the face 2
Load Vector
[a] For coupled-field analyses with structural DOFs, the element uses 2 x 2- and 3-point integ-
ration rules to calculate the element matrices and load vectors for the quad and triangle
geometries, respectively. All other analyses use 3x3- and 6-point integration rules for the
quad and triangle geometries, respectively.

13.223.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. Electromagnetics (p. 163) describes the derivation of the electric and magnetic element
matrices and load vectors, as well as electric and magnetic field evaluations. Piezoelectrics (p. 334)
discusses the piezoelectric capability used by the element. Piezoresistivity (p. 347) discusses the
piezoresistive effect. Thermoelectrics (p. 348) discusses the thermoelectric effects. Thermoelasti-
city (p. 329) discusses the thermoelastic effects. Electroelasticity (p. 345) discusses the Maxwell stress
electroelastic coupling. Thermoplasticity (p. 332) discusses the thermoplastic effect. Structural-Diffusion
Coupling (p. 354) discusses diffusion strain coupling. Magnetoelasticity (p. 361) discusses magneto-
elastic coupling matrices derived from magnetic Maxwell and Lorentz forces.

13.224. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
624 Published: 2023-10-23
 SOLID225 - 3D 8-Node Coupled-Field Solid

13.225. SOLID225 - 3D 8-Node Coupled-Field Solid

Shape
Matrix or Vector Geometry Integration Points (p. 414)
Functions
Stiffness and Stress
Stiffness Matrices; and
Equation 11.217 (p. 387), 2x2x2
Thermal Expansion,
Equation 11.218 (p. 387), and
Diffusion Expansion, and
Equation 11.219 (p. 387) 1 if element shape is tetrahedral
Electrostatic Force Load
Vector
Mass Matrix Same as stiffness matrix. 2x2x2
Quad Equation 11.70 (p. 374) 2x2
and
Equation 11.71 (p. 374)
Pressure Load Vector
Triangle Equation 11.50 (p. 373) 3
and
Equation 11.51 (p. 373)
Thermal Conductivity
Matrix and Heat Equation 11.227 (p. 387) 2x2x2
Generation Load Vector
Specific Heat Matrix Equation 11.227 (p. 387). 2x2x2
Convection Surface Equation 11.227 (p. 387)
2x2
Matrix and Load Vector specialized to the face
Dielectric Permittivity
and Electrical
Conductivity Matrices;
Equation 11.228 (p. 387) 2x2x2
Charge Density, Joule
Heating, and Peltier
Heat Flux Load Vectors
Surface Charge Density Equation 11.228 (p. 387) 2x2
Load Vector specialized to the face
Diffusivity Matrix and Equation 11.230 (p. 387) 2x2x2
Diffusing Substance
Generation Load Vector
Diffusion Damping Equation 11.230 (p. 387). If 2x2x2
Matrix KEYOPT(10) = 1, matrix is

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 625
Element Library

Shape
Matrix or Vector Geometry Integration Points (p. 414)
Functions
diagonalized as described in
Lumped Matrices (p. 422).
Diffusion Flux Load Equation 11.230 (p. 387) 2x2
Vector specialized to the face
Thermoelastic stiffness
Same as combination of stiffness and thermal conductivity matrices
and Damping Matrices
Piezoelectric Coupling
Same as combination of stiffness matrix and dielectric matrix
Matrix
Seebeck Coefficient Same as combination of electrical conductivity and thermal conductivity
Coupling Matrix matrices
Diffusion-elastic stiffness Same as combination of stiffness and diffusivity matrices
and damping matrices

13.225.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. Electromagnetics (p. 163) describes the derivation of electric element matrices and
load vectors as well as electric field evaluations. Piezoelectrics (p. 334) discusses the piezoelectric
capability used by the element. Piezoresistivity (p. 347) discusses the piezoresistive effect. Thermoelec-
trics (p. 348) discusses the thermoelectric effects. Thermoelasticity (p. 329) discusses the thermoelastic
effects. Electroelasticity (p. 345) discusses the Maxwell stress electroelastic coupling. Thermoplasti-
city (p. 332) discusses the thermoplastic effect. Diffusion (p. 305) describes the derivation of diffusion
matrices and load vectors. Structural-Diffusion Coupling (p. 354) discusses diffusion strain coupling.

13.225.2. Theory
This element uses the method (selective reduced integration technique for volumetric terms)
(Hughes([221] (p. 937)), Nagtegaal et al. ([222] (p. 937)))

If KEYOPT(6) = 2 or 3, the enhanced strain formulation from the works of Simo and Rifai([318] (p. 943)),
Simo and Armero([319] (p. 943)), Simo et al.([320] (p. 943)), Andelfinger and Ramm([321] (p. 943)), and
Nagtegaal and Fox([322] (p. 943)) is used. It introduces additional internal (user-inaccessible) structural
degrees of freedom (DOFs) to prevent shear and volumetric locking (KEYOPT(6) = 2) or shear locking
only (KEYOPT(6) = 3). These internal DOFs are created and condensed out at the element level.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
626 Published: 2023-10-23
 SOLID226 - 3D 20-Node Coupled-Field Solid

13.226. SOLID226 - 3D 20-Node Coupled-Field Solid

Matrix or Vector Geometry Shape Functions Integration Points (p. 414)

Equation 11.236 (p. 388), 14 if KEYOPT(6) = 0
Brick Equation 11.237 (p. 388), and
Equation 11.238 (p. 388) 2 x 2 x 2 if KEYOPT(6) = 1
Stiffness and Stress Equation 11.211 (p. 386),
Stiffness Matrices; and Wedge Equation 11.212 (p. 386), and 3x3
Thermal Expansion, Equation 11.213 (p. 386)
Diffusion Expansion, and Equation 11.195 (p. 384),
Electrostatic Force Load Pyramid Equation 11.196 (p. 384), and 2x2x2
Vector Equation 11.197 (p. 384)
Equation 11.187 (p. 383),
Tet Equation 11.188 (p. 383), and 4
Equation 11.189 (p. 383)
Mass Matrix Same as stiffness matrix. 14 if brick. If other shapes, same as stiffness m
Equation 11.82 (p. 375) and
Quad 3x3
Equation 11.83 (p. 375)
Pressure Load Vector
Equation 11.58 (p. 373) and
Triangle 6
Equation 11.59 (p. 373)
Brick Equation 11.239 (p. 388) 14 [1]
Thermal Conductivity Wedge Equation 11.214 (p. 386) 3x3
Matrix and Heat
Generation Load Vector Pyramid Equation 11.198 (p. 385) 2x2x2
Tet Equation 11.190 (p. 383) 4
Same as thermal conductivity matrix. If KEYOPT(10) = 1 or 2, matrix is diagonalized as descr
Specific Heat Matrix
Matrices (p. 422).
Convection Surface Quad Equation 11.86 (p. 375) 3x3
Matrix and Load Vector Triangle Equation 11.61 (p. 373) 6
Dielectric Permittivity Brick Equation 11.240 (p. 388) 14 [1]
and Electrical Wedge Equation 11.215 (p. 386) 3x3
Conductivity Matrices;
Charge Density, Joule Pyramid Equation 11.199 (p. 385) 2x2x2
Heating, and Peltier
Tet Equation 11.191 (p. 383) 4
Heat Flux Load Vectors

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 627
Element Library

Matrix or Vector Geometry Shape Functions Integration Points (p. 414)

Magnetic Reluctivity Brick Equation 11.289 (p. 396), 14
Matrix, Eddy Current Wedge Equation 11.290 (p. 397), and 3x3
Damping Matrices; Equation 11.291 (p. 397)
Source Current and Pyramid 2x2x2
Remnant Magnetization
or
Load Vectors
3 x 3 x 3 [2]
Tet Equation 11.283 (p. 396), 4
Equation 11.284 (p. 396),
Equation 11.285 (p. 396),
Equation 11.286 (p. 396),
Equation 11.287 (p. 396), and
Equation 11.288 (p. 396),
Diffusivity Matrix and Brick Equation 11.242 (p. 389) 14 [1]
Diffusing Substance Wedge Equation 11.216 (p. 386) 3x3
Generation Load Vector
Pyramid Equation 11.200 (p. 385) 2x2x2
Tet Equation 11.193 (p. 383) 4
Diffusion Damping Same as diffusivity matrix. If KEYOPT(10)=1, matrix is diagonalized as described in Lumped M
Matrix
Diffusion Flux Load Quad Equation 11.88 (p. 375) 3x3
Vector Triangle Equation 11.63 (p. 373) 6
Thermoelastic stiffness
Same as combination of stiffness and thermal conductivity matrices
and Damping Matrices
Piezoelectric Coupling
Same as combination of stiffness matrix and dielectric matrix
Matrix
Seebeck Coefficient
Same as combination of electrical conductivity and thermal conductivity matrices
Coupling Matrix
Magneto-Electric and
Electromagnetic
Same as combination of magnetic reluctivity and electric conductivity matrices
(velocity current)
Coupling Matrices
Magneto-Elastic Stiffness
Same as combination of stiffness and magnetic reluctivity matrices
Matrices
Diffusion-elastic stiffness Same as combination of stiffness and diffusivity matrices
and damping matrices
Surface Charge Density Quad Equation 11.199 (p. 385) 3x3
Load Vector Triangle Equation 11.62 (p. 373) 6

1. When KEYOPT(6) = 1 in a coupled-field analysis with structural DOFs, the unified reduced integration
scheme with 2 x 2 x 2 integration points is used to form the thermal, electric, electrostatic matrices
and load vectors.

2. Thermomagnetic and thermo-electromagnetic analyses use 3 x 3 x 3 integration points for all load
vector and matrix calculations.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
628 Published: 2023-10-23
 SOLID227 - 3D 10-Node Coupled-Field Solid

13.226.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. Electromagnetics (p. 163) describes the derivation of element matrices and load vectors
as well as electromagnetic field evaluations. Piezoelectrics (p. 334) discusses the piezoelectric capability
used by the element. Piezoresistivity (p. 347) discusses the piezoresistive effect. Thermoelectrics (p. 348)
discusses the thermoelectric effects. Thermoelasticity (p. 329) discusses the thermoelastic effects.
Electroelasticity (p. 345) discusses the Maxwell stress electroelastic coupling. Thermoplasticity (p. 332)
discusses the thermoplastic effect. Structural-Diffusion Coupling (p. 354) discusses diffusion strain
coupling. Magnetoelasticity (p. 361) discusses magneto-elastic coupling matrices derived from mag-
netic Maxwell and Lorentz forces.

13.227. SOLID227 - 3D 10-Node Coupled-Field Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness, Mass, and
Stress Stiffness Matrices;
Equation 11.187 (p. 383), Equation 11.188 (p. 383),
and Thermal Expansion 4
and Equation 11.189 (p. 383)
and Electrostatic Force
Load Vectors
Equation 11.187 (p. 383), Equation 11.188 (p. 383),
Pressure Load Vector and Equation 11.189 (p. 383) specialized to the 6
face
Thermal Conductivity
Matrix and Heat Equation 11.190 (p. 383) 4
Generation Load Vector
Same as thermal conductivity matrix. If
Specific Heat Matrix KEYOPT(10) = 1 or 2, matrix is diagonalized as 11
described in Lumped Matrices (p. 422).
Convection Surface Equation 11.190 (p. 383) specialized to the face.
6
Matrix and Load Vector Consistent surface matrix.
Dielectric Permittivity
and Electrical
Equation 11.191 (p. 383) 4
Conductivity Matrices;
Charge Density, Joule

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 629
Element Library

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Heating, and Peltier
Heat Flux Load Vectors
Magnetic Reluctivity
Matrix, Eddy Current Equation 11.283 (p. 396), Equation 11.284 (p. 396),
Damping Matrices; Equation 11.285 (p. 396), Equation 11.286 (p. 396),
4
Source Current and Equation 11.287 (p. 396), and
Remnant Magnetization Equation 11.288 (p. 396),
Load Vectors
Diffusivity Matrix and
Diffusing Substance Equation 11.193 (p. 383) 4
Generation Load Vector
Damping Matrix Same as diffusivity matrix. If KEYOPT(10) = 1, 4
the matrix is diagonalized as described in
Lumped Matrices (p. 422)
Diffusion Flux Load Equation 11.193 (p. 383) specialized to the face 6
Vector
Thermoelastic Stiffness
Same as combination of stiffness and thermal conductivity matrices
and Damping Matrices
Piezoelectric Coupling
Same as combination of stiffness matrix and dielectric matrix
Matrix
Seebeck Coefficient Same as combination of electrical conductivity and thermal conductivity
Coupling Matrix matrices
Magneto-Electric and
Electromagnetic Same as combination of magnetic reluctivity and electric conductivity
(velocity current) matrices
Coupling Matrices
Magneto-Elastic Stiffness
Same as combination of stiffness and magnetic reluctivity matrices
Matrices
Diffusion-Elastic Stiffness
Same as combination of stiffness and diffusivity matrices
and Damping Matrices
Surface Charge Density
Equation 11.191 (p. 383) specialized to the face 6
Load Vector

13.227.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. Electromagnetics (p. 163) describes the derivation of element matrices and load vectors
as well as electromagnetic field evaluations. Piezoelectrics (p. 334) discusses the piezoelectric capability
used by the element. Piezoresistivity (p. 347) discusses the piezoresistive effect. Thermoelectrics (p. 348)
discusses the thermoelectric effects. Thermoelasticity (p. 329) discusses the thermoelastic effects.
Electroelasticity (p. 345) discusses the Maxwell stress electroelastic coupling. Thermoplasticity (p. 332)
discusses the thermoplastic effect. Structural-Diffusion Coupling (p. 354) discusses diffusion strain
coupling. Magnetoelasticity (p. 361) discusses magneto-elastic coupling matrices derived from mag-
netic Maxwell and Lorentz forces.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
630 Published: 2023-10-23
 LINK228 - 3D Coupled-field Link

13.228. LINK228 - 3D Coupled-field Link

Integration
Matrix or Vector Element configuration Shape Functions
Points (p. 414)
Stiffness Matrix; and 2-node (KEYOPT(4) = 0) Equation 11.6 (p. 367) 1
Newton-Raphson 3-node (KEYOPT(4) = 1) Equation 11.19 (p. 368) 2
Mass and Stress Stiffness 2-node (KEYOPT(4) = 0) Equation 11.6 (p. 367), 1
Matrices Equation 11.7 (p. 367),
and
Equation 11.8 (p. 367)
3-node (KEYOPT(4) = 1) Equation 11.19 (p. 368), 2
Equation 11.20 (p. 368),
and
Equation 11.21 (p. 368)
Thermal Conductivity and 2-node (KEYOPT(4) = 0) Equation 11.13 (p. 368) 1
Specific Heat Matrix; and Heat 3-node (KEYOPT(4) = 1) Equation 11.25 (p. 368) 2
Generation Load Vector
Electrical Conductivity Matrices; 2-node (KEYOPT(4) = 0) Equation 11.14 (p. 368) 1
Joule Heating, and Peltier Heat 3-node (KEYOPT(4) = 1) Equation 11.26 (p. 368) 2
Flux Load Vectors
Thermoelastic stiffness and Same as combination of stiffness and thermal conductivity matrices
Damping Matrices
Seebeck Coefficient Coupling Same as combination of electrical conductivity and thermal conductivity
Matrix matrices

13.228.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. Electromagnetics (p. 163) describes the derivation of the electric element matrices
and load vectors, as well as electric field evaluations. Thermoelectrics (p. 348) discusses the thermo-
electric effects. Thermoelasticity (p. 329) discusses the thermoelastic effects. Thermoplasticity (p. 332)
discusses the thermoplastic effect.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 631
Element Library

13.228.2. Assumptions and Restrictions

The theory for this element is a reduction of the theory for 3D beam (BEAM189 - 3D 3-Node
Beam (p. 601)) and 3D coupled-field solid elements. The element is not capable of carrying bending
loads and shear effects.

13.229. Reserved for Future Use

This section is reserved for future use.

13.230. PLANE230 - 2D 8-Node Electric Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Electrical Conductivity and Quad Equation 11.142 (p. 378) 3x3
Dielectric Permittivity Coefficient
Matrices Triangle Equation 11.120 (p. 377) 3

13.230.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of the electric element matrices and load vectors
as well as electric field evaluations.

13.230.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the potential varies linearly along
that edge.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
632 Published: 2023-10-23
 SOLID232 - 3D 10-Node Tetrahedral Electric Solid

13.231. SOLID231 - 3D 20-Node Electric Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Brick Equation 11.240 (p. 388) 14
Electrical Conductivity and Wedge Equation 11.215 (p. 386) 3x3
Dielectric Permittivity Coefficient
Matrices Pyramid Equation 11.199 (p. 385) 8
Tet Equation 11.191 (p. 383) 4

13.231.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of electric element matrices and load vectors as
well as electric field evaluations.

13.232. SOLID232 - 3D 10-Node Tetrahedral Electric Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Dielectric Permittivity and Electrical
Conductivity Coefficient Matrices, Equation 11.191 (p. 383) 4
Charge Density Load Vector

13.232.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of electric element matrices and load vectors as
well as electric field evaluations.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 633
Element Library

13.233. PLANE233 - 2D 8-Node Electromagnetic Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Magnetic Potential Quad Equation 11.140 (p. 378) 3x3
Coefficient, Eddy Current Quad - Axisymmetric
Damping or Displacement Massive Conductor
Current Mass Matrices; and Equation 11.140 (p. 378) 2x2
(KEYOPT(1)=1 and
Permanent Magnet and KEYOPT(3)=1)
Applied Current Load
Vectors Triangle Equation 11.117 (p. 377) 3
Equation 11.140 (p. 378) Same as
Quad and coefficient
Electrical Conductivity and Equation 11.142 (p. 378) matrix
Displacement Current
Damping Matrices Equation 11.117 (p. 377) Same as
Triangle and coefficient
Equation 11.142 (p. 378) matrix
Load Type Distribution
Current Density and Phase Angle Distribution Bilinear across element

13.233.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of the electric element matrices and load vectors
as well as electric field evaluations.

13.233.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the potential varies linearly along
that edge.

13.234. Reserved for Future Use

This section is reserved for future use.

13.235. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
634 Published: 2023-10-23
 SOLID237 - 3D 10-Node Electromagnetic Solid

13.236. SOLID236 - 3D 20-Node Electromagnetic Solid

r t
W
X P
O
M U B V
Y,v N A
Y L s
S
T Z
X,u K
I
Z,w R
Q J

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Brick Equation 11.289 (p. 396), 14
Magnetic Reluctivity, Wedge Equation 11.290 (p. 397), and 3x3
Eddy Current Damping, Pyramid Equation 11.291 (p. 397) 3x3x3
Displacement Current
Equation 11.283 (p. 396),
Mass Matrices, Source
Equation 11.284 (p. 396),
Current Density and
Equation 11.285 (p. 396),
Remnant Magnetization Tet 4
Equation 11.286 (p. 396),
Load Vectors
Equation 11.287 (p. 396), and
Equation 11.288 (p. 396),
Brick Equation 11.240 (p. 388) 14
Electrical Conductivity
and Displacement Wedge Equation 11.215 (p. 386) 3x3
Current Damping Pyramid Equation 11.199 (p. 385) 3x3x3
Matrices
Tet Equation 11.191 (p. 383) 4
Magneto-Electric
Combination of eddy current and electrical conductivity matrices
Coupling Matrix
Magneto-Dielectric
Combination displacement current mass and damping matrices
Coupling Matrix

13.236.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of element matrices and load vectors as well as
electromagnetic field evaluations.

13.237. SOLID237 - 3D 10-Node Electromagnetic Solid

r L

Y,v R
P
Q
O K
X,u
N
Z,w s M
I
t J

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 635
Element Library

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Magnetic Reluctivity,
Equation 11.283 (p. 396),
Eddy Current Damping,
Equation 11.284 (p. 396),
Displacement Current
Equation 11.285 (p. 396),
Mass Matrices, Source 4
Equation 11.286 (p. 396),
Current Density and
Equation 11.287 (p. 396),
Remnant Magnetization
andEquation 11.288 (p. 396),
Load Vectors
Electrical Conductivity
and Displacement
Equation 11.191 (p. 383) 4
Current Damping
Matrices
Magneto-Electric Combination of eddy current and electrical conductivity
Coupling Matrix matrices
Magneto-Dielectric Combination displacement current mass and damping
Coupling Matrix matrices

13.237.1. Other Applicable Sections

Electromagnetics (p. 163) describes the derivation of element matrices and load vectors as well as
electromagnetic field evaluations.

13.238. PLANE238 - 2D 8-Node Diffusion Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Diffusivity Matrix and Diffusing Quad Equation 11.143 (p. 378) 3x3
Substance Generation Load Triangle Equation 11.121 (p. 377) 6
Vector
Damping Matrix Same as diffusivity matrix. If KEYOPT(1) = 1, matrix is
diagonalized as described in Lumped Matrices (p. 422)
Diffusion Flux Load Vector Same as coefficient matrix, specialized 2
to the face

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
636 Published: 2023-10-23
 SOLID239 - 3D 20-Node Diffusion Solid

13.238.1. Other Applicable Sections

Diffusion in the Theory Reference (p. 305) describes the derivation of the diffusion element matrices
and load vectors as well as diffusion flux evaluation. If KEYOPT(1) = 1, the damping matrix is diagon-
alized as described in Lumped Matrices (p. 422).

13.238.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the concentration varies linearly
along that edge.

13.239. SOLID239 - 3D 20-Node Diffusion Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Diffusivity Matrix and Diffusing Brick Equation 11.242 (p. 389) 14
Substance Generation Load Wedge Equation 11.216 (p. 386) 3x3
Vector
Pyramid Equation 11.200 (p. 385) 2x2x2
Tet Equation 11.193 (p. 383) 4
Damping Matrix Same as diffusivity matrix. If KEYOPT(1) = 1, the damping matrix
is diagonalized as described in Lumped Matrices (p. 422).
Diffusion Flux Load Vector Quad Equation 11.88 (p. 375) 3x3
Triangle Equation 11.63 (p. 373) 6

13.239.1. Other Applicable Sections

Diffusion in the Theory Reference (p. 305) describes the derivation of diffusion element matrices and
load vectors as well as diffusion flux evaluation.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 637
Element Library

13.240. SOLID240 - 3D 10-Node Tetrahedral Diffusion Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Diffusivity Matrix and Diffusing Equation 11.193 (p. 383) 4
Substance Generation Load
Vector
Damping Matrix Same as diffusivity 11
matrix. If KEYOPT(1) = 1,
the matrix is
diagonalized as
described in Lumped
Matrices (p. 422)
Diffusion Flux Load Vector Equation 11.193 (p. 383) 6
specialized to the face

13.240.1. Other Applicable Sections

Diffusion in the Theory Reference (p. 305) describes the derivation of diffusion element matrices and
load vectors as well as diffusion flux evaluations. If KEYOPT(1) = 1, the damping matrix is diagonalized
as described in Lumped Matrices (p. 422).

13.241. HSFLD241 - 2D Hydrostatic Fluid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix Same as attached 2D solid element 2 or 3

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
638 Published: 2023-10-23
 HSFLD242 - 3D Hydrostatic Fluid

Load Type Distribution

Fluid Mass Flow Rate Fluid mass flow rate at pressure node
Nodal Temperature Temperature at the pressure node

HSFLD241 is used to model fluids that are fully enclosed by solids (containing vessels). The logic for
HSFLD241 is very similar to that given for HSFLD242 - 3D Hydrostatic Fluid (p. 639), with the differences
noted below:

1. The element thickness must be input on R command.

2. If KEYOPT(3) = 0, the element behaves as 2D. The determination for plane stress or plane strain be-
havior is done based on the attached 2D solid element.

3. If KEYOPT(3) = 1, the element behaves as axisymmetric.

4. The set of orthonormal vectors on the fluid surface for defining fluid volume is given as:
(13.329)

where:

n = outward normal vector (in X-Y plane) to the fluid surface enclosing the fluid volume
t 1 = k = first tangent vector (normal to X-Y plane) to the fluid surface
t 2 = second tangent vector (in X-Y plane) to the fluid surface

Since the model stays in the X-Y plane, the variation of the first tangent vector is always zero. As a
result, no geometric stiffness contributions are made from the rate of change of volume enclosed
by the solid surface:

(13.330)

13.242. HSFLD242 - 3D Hydrostatic Fluid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness Matrix Same as attached 3D solid or shell element 2 x 2 or 3 x 3

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 639
Element Library

Load Type Distribution

Fluid Mass Flow Rate Fluid mass flow rate at pressure node
Nodal Temperature Temperature at the pressure node

13.242.1. Introduction
HSFLD242 is used to model fluids that are fully enclosed by solids (containing vessels). The fluid is
assumed to have no viscosity, rotation or inertial effects (including acceleration loads) such as sloshing.
The fluid fills the containing vessel completely so it has no free surface. Under these conditions, the
fluid can be assumed to be hydrostatic; that is, the fluid pressure is uniform throughout the fluid
volume. The hydrostatic fluid element is pyramid shaped with the base overlaying a 3D solid or shell
element face and the vertex at a pressure node. The element nodes shared with the underlying 3D
solid or shell element have only translation degrees of freedom. The pressure node has a single hy-
drostatic pressure degree of freedom (HDSP) and is shared by all hydrostatic fluid elements defining
the fluid volume. The coupling between the fluid volume and the containing vessel is modeled by
applying the hydrostatic fluid pressure as a surface load on the containing vessel. No fluid is assumed
to flow in or out of the containing vessel unless a hydrostatic pressure value or fluid mass flow rate
is prescribed at the pressure node.

13.242.2. Element Matrices and Load Vectors

Formulating a coupled system of a solid (containing vessel) enclosing a hydrostatic fluid requires
augmenting the internal virtual work for the solid with contributions from the fluid. Starting with the
rate of internal energy expression:

(13.331)

where:

W = internal energy for the solid

Ss = current solid surface enclosing the fluid volume
Sf = current fluid surface enclosing the fluid volume
tsi = component i of surface traction at a point on Ss
tfi = component i of surface traction at a point on Sf
vsi = component i of velocity at a point on Ss
vfi = component i of velocity at a point on Sf

Since the fluid fills the containing vessel completely and the system is in equilibrium, the surface
traction at a point on the solid surface should be equal and opposite to the surface traction at a point
on the fluid surface that it is in contact with. The surface tractions represent the loads exerted by the
fluid and the solid on each other. Assuming no viscosity for the fluid, the surface tractions have only
normal non-zero components, so:
(13.332)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
640 Published: 2023-10-23
 HSFLD242 - 3D Hydrostatic Fluid

P = fluid pressure
ni = component i of outward normal to the fluid surface enclosing the fluid volume

Further, assuming that the fluid has no inertial effects or acceleration loads, Equation 8.2 (p. 242) for
fluid momentum show that the fluid pressure should be uniform. Using this fact with Equa-
tion 13.332 (p. 640) in Equation 13.331 (p. 640) gives:

(13.333)

By taking a variation of Equation 13.333 (p. 641) and then using Gauss' theorem, we get the internal
virtual work expression:

(13.334)

where:

Vf = current fluid volume

vfi,i = div v f = divergence of velocity vector at a point in Vf

The second term on the right hand side in Equation 13.334 (p. 641) is the virtual work done by the
hydrostatic fluid pressure on the solid that contains the fluid and represents the coupling between
the fluid volume and the solid (containing vessel).

Since the fluid volume has uniform pressure, the volumetric strain is also uniform. This allows Equa-
tion 13.334 (p. 641) to be simplified further.

Using the definitions of deformation gradient and its determinant as given in Equation 3.2 (p. 30),
Equation 3.3 (p. 30), and Equation 3.4 (p. 30), and the derivative of the determinant (Gurtin
[378] (p. 946)), it can be seen that:

(13.335)

Similarly, for the solid surface enclosing the fluid volume:

(13.336)

where:

Vs = current volume enclosed by solid surface Ss

Using Equation 13.335 (p. 641) and Equation 13.336 (p. 641) (and its variation) in Equation 13.334 (p. 641)
gives:
(13.337)

The internal virtual work in Equation 13.337 (p. 641) is valid for large as well as small strains. Since the
fluid is assumed to have no acceleration loads or free surface, there are no contributions to external

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 641
Element Library

virtual work from loads. However, if fluid mass flow rate is prescribed at the pressure node, then the
mass, and hence the volume, of the fluid will change and the virtual work expression will be:

(13.338)

where:

w = fluid mass flow rate (use F command)

ρf = current fluid density

The stiffness matrix for the coupled system can be obtained by taking the derivative of the virtual
work expression as given in Equation 13.338 (p. 642):

(13.339)

For completeness, the variational quantities in Equation 13.339 (p. 642) need to be expressed in terms
of the independent variables. To compute , we start by taking a variation of Equation 13.336 (p. 641):

(13.340)

Since the fluid is assumed to be irrotational, there is no contribution from the term containing the
variation of the normal, δni.

For , we start with Equation 13.340 (p. 642), including the term containing the variation of the
normal:

(13.341)

Again, irrotational fluid assumption eliminates terms with variation of the normal. If t 1 and t 2 define
orthonormal tangent vectors on the surface such that:
(13.342)

then, Equation 13.341 (p. 642) can be expanded as:

(13.343)

The terms and Dρf in Equation 13.339 (p. 642) define material stiffness for the fluid volume. This
requires the definition of the pressure-volume relationship as described below.

[Link]. Incompressible Fluid

The incompressibility constraint for the fluid volume can be given by using Equation 13.335 (p. 641)
as:

(13.344)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
642 Published: 2023-10-23
 HSFLD242 - 3D Hydrostatic Fluid

The augmented virtual work expression incorporating the incompressibility constraint can be written
as:

(13.345)

where:

λ = Lagrange multiplier

Comparing Equation 13.345 (p. 643) to Equation 13.338 (p. 642) shows that they are equivalent with
the hydrostatic fluid pressure P acting as a Lagrange multiplier. The stiffness contribution can be
obtained by taking the derivative of the augmented virtual work:
(13.346)

The incompressible fluid may have a change in volume if it has non-zero thermal expansion. Since
temperature is not defined as a degree of freedom, changes in volume due to temperature changes
are computed by specifying temperature loading (use the BF command) and coefficient of thermal
expansion (use the MP command):
(13.347)

where:

α = coefficient of thermal expansion (use MP command)

= rate of change of temperature
V0f = initial fluid volume

When KEYOPT(6) = 1, the change in volume is accommodated by a change in fluid mass (mass
flows into or out of the cavity). When KEYOPT(6) = 2, the change in volume is accommodated by
a change in fluid density such that the fluid mass is held constant.

[Link]. Liquid
Using the constitutive equation for liquid (see Fluid Material Models (p. 159)), the rate of change of
volume and its variation can be given as:
(13.348)

(13.349)

where:

K = bulk modulus (use TB command with Lab = FLUID and TBOPT = LIQUID)

The variation of current density can be defined as:

(13.350)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 643
Element Library

[Link]. Gas
To define the rate of change of volume, we need to look at incremental change in volume:
(13.351)

where:

Δt = time increment for current substep

Vnf = fluid volume at the end of previous substep

Using the Ideal Gas Law, the previous expression can be expanded as:

(13.352)

where:

Pt = Pref + P = total fluid pressure

Pref = reference pressure (specified as a real constant with R command)
Tt = Toff + T = total temperature
Toff = temperature offset from absolute zero to zero (use TOFFST command)
Pnt = total fluid pressure at the end of previous substep
Tnt = total temperature at the end of previous substep

The variation of the rate of change of volume can be easily obtained as:

(13.353)

and the variation of current density can be given as:

(13.354)

[Link]. Pressure-Volume Data

For compressible fluids that do not follow the constitutive equation for liquid or the Ideal Gas Law,
the rate of change of volume can be expressed as:
(13.355)

where:

= slope (interpolated) of the pressure-volume data curve

= slope (interpolated) of the temperature-volume data curve

and the variation of the rate of change of volume can be expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
644 Published: 2023-10-23
 COMBI250 - 3D Bushing Element

(13.356)

The variation of current density is defined as:

(13.357)

See Fluid Material Models (p. 159) for more details.

13.243. Reserved for Future Use

This section is reserved for future use.

13.244. Reserved for Future Use

This section is reserved for future use.

13.245. Reserved for Future Use

This section is reserved for future use.

13.246. Reserved for Future Use

This section is reserved for future use.

13.247. Reserved for Future Use

This section is reserved for future use.

13.248. Reserved for Future Use

This section is reserved for future use.

13.249. Reserved for Future Use

This section is reserved for future use.

13.250. COMBI250 - 3D Bushing Element

13.250.1. Matrices
The following sections describe the different matrices used by the COMBI250 element:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 645
Element Library

[Link]. Stiffness Matrix

In the element coordinate system, the stiffness matrix is equal to:

where

K1 to K6 = stiffness characteristics input on R command.

[Link]. Viscous Damping Matrix

In the element coordinate system, the viscous damping matrix is equal to:

where

CV1 to CV6 = viscous damping characteristics input on R command.

[Link]. Structural Damping Matrix

In the element coordinate system, the structural damping matrix is equal to:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
646 Published: 2023-10-23
 Reserved for Future Use

where

KIMAG1 to KIMAG6 = structural damping characteristics input on R command.

13.251. SURF251 - 2D Radiosity Surface

J

Y
I
X

SURF251 is used only for postprocessing of radiation quantities, such as radiation heat flux. See SURF251
in the Element Reference for details.

13.252. SURF252 - 3D Thermal Radiosity Surface

L K

K,L

I J
I J
Z (Triangular Option)

Y
X

SURF252 is used only for postprocessing of radiation quantities, such as radiation heat flux. See SURF252
in the Element Reference for details.

13.253. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 647
Element Library

13.254. Reserved for Future Use

This section is reserved for future use.

13.255. Reserved for Future Use

This section is reserved for future use.

13.256. Reserved for Future Use

This section is reserved for future use.

13.257. INFIN257 - Structural Infinite Solid

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
648 Published: 2023-10-23
 INFIN257 - Structural Infinite Solid

Matrix or Vector Base Element Mapping and Shape Functions

Equation 11.146 (p. 379), Equation 11.147 (

Linear 2D Solid
Equation 11.148 (p. 379)
Equation 11.153 (p. 380), Equation 11.154 (
Quadratic 2D Solid
Equation 11.155 (p. 380)
Stiffness Matrices
Equation 11.249 (p. 389), Equation 11.250 (
Linear 3D Solid
Equation 11.251 (p. 390), Equation 11.254 (
Equation 11.261 (p. 391), Equation 11.262 (
Quadratic 3D Solid
Equation 11.263 (p. 391), Equation 11.264 (
Mass Matrix N/A
Pressure Load Vector N/A
Load Type Distribution
Element Temperature N/A
Nodal Temperature N/A
Pressure N/A

References: Zienkiewicz et al.([170] (p. 934)), Marques and Owen([172] (p. 934)), Lysmer and Kyhlemeyer
([173] (p. 934))

13.257.1. Structural Infinite Element for Static Analysis

Following is a brief summary of the theory involved in the infinite mapping functions.

In this figure, the 1D element extends from node J, through node K to the point M at infinity and is
mapped onto the parent element defined by the local coordinate system in the range :

Figure 13.35: 1D Infinite Element Mapping

The position of the "pole," xo, is arbitrary; once chosen, the location of node K is defined by:
(13.358)

The interpolation from local to global positions is performed as:

(13.359)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 649
Element Library

where:

As shown, corresponds to the global positions , respectively.

The global-local mapping relationship can be rewritten in terms of , the arbitrary distance
from the pole.

Solving Equation 13.359 (p. 649) for yields:

(13.360)

where:

= distance from the pole O to a general point within the element

as shown in Figure 13.35: 1D Infinite Element Mapping (p. 649)

For the structural analysis, the basic solution variable (displacement component) can be written in
the polynomial forms:
(13.361)

Substituting Equation 13.360 (p. 650) into Equation 13.361 (p. 650) gives:
(13.362)

where is implied, as the variable is assumed to vanish at infinity.

Equation 13.362 (p. 650) is truncated at the quadratic ( ) term in the present implementation. The
equation also shows the role of the pole position O.

The 1D element may be thought of as one edge of the 2D solid infinite element, as shown:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
650 Published: 2023-10-23
 INFIN257 - Structural Infinite Solid

Figure 13.36: 1D Infinite Element Mapping

Mapping is achieved by the shape function products. The mapping functions and the isoparametric
shape functions for 2D and axisymmetric 4 node quadrilaterals are given in 2D and Axisymmetric 4-
Node Quadrilateral Infinite Solids (p. 378). The shape functions for nodes M and N are not needed as
the field variable is assumed to vanish at infinity. 2D mapping can be extended to 3D in a similarly
straightforward manner.

The mapping approach has an important advantage: the original Gauss-Legendre integration technique
can be used with minimal modification. The only change necessary is a new calculation of the Jacobian
matrix with mapping functions.

The coefficient matrix can be written as:

(13.363)

where:

= strain-displacement matrix
= elasticity matrix

The vectors of the matrix in Equation 13.147 (p. 512) contain the derivatives of with respect
to the global coordinates, evaluated according to:

(13.364)

where

= Jacobian matrix which defines the geometric mapping

is given by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 651
Element Library

(13.365)

The mapping functions in terms of and are given in 2D and Axisymmetric 4-Node Quadrilat-
eral Infinite Solids (p. 378). The domain differential must also be written in terms of the local
coordinates, so that:
(13.366)

Subject to the evaluation of and , which involves the mapping functions, the element
matrices can now be calculated in the standard manner using Gaussian quadrature.

13.257.2. Structural Infinite Element for Dynamic Analysis

If sudden loading occurs in an infinite elastic body, the load is not transmitted instantaneously to all
parts of the body. The stresses and displacements radiate in the form of waves. The equations of
equilibrium must be replaced by the equations of motion.

The equations of equilibrium without body force in an elastic medium is given by:
(13.367)

where:

= density
= particle acceleration
= stress
= particle position

The stress-strain relations for an isotropic, linear elastic material can be written in terms of Lame's
constants and :
(13.368)

where:

= Young's modulus
= Poisson's ratio

and the strain-displacement relationship is:

(13.369)

By substituting Equation 13.368 (p. 652) and Equation 13.369 (p. 652) into Equation 13.367 (p. 652), the
equations of motion can be written in terms of displacements and material properties:
(13.370)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
652 Published: 2023-10-23
 INFIN257 - Structural Infinite Solid

If a disturbance is produced at a point in an elastic medium, waves radiate from that point in all dir-
ections. If the medium extends to infinity, the waves can be considered as plane waves at the far
field. Therefore, it may be possible to assume that all particles are moving parallel to the direction of
wave propagation (dilatational wave, longitudinal wave, or p-wave), or perpendicular to that direction
(distortion wave, transverse wave, or s-wave). The governing equation can be solved for the plane
waves.

For the plane waves (dilatational and distortion waves), the equations of motion can be reduced in
the common form:
(13.371)

By comparing Equation 13.370 (p. 652) to Equation 13.371 (p. 653), we obtain:
(13.372)

for the dilatational waves, and:

(13.373)

for the distortion waves, where:

= dilatational wave speed

= distortion wave speed

Any function of form is a solution of Equation 13.371 (p. 653), and any function of form
is also a solution. Therefore the general solution of the governing equation can be repres-
ented by the form:
(13.374)

The function represents "forward" wave motion, and the function represents
"backward" wave motion.

Now consider the dilatational wave that is moving forward along the x-axis. The solution is:
(13.375)

The particle velocity is:

(13.376)

The strain components in x-direction is:

(13.377)

For the stress we have:

(13.378)

The stress can be expressed in terms of wave and particle speed from Equation 13.376 (p. 653) through
Equation 13.378 (p. 653):
(13.379)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 653
Element Library

If we consider the backward-moving wave, the minus sign in Equation 13.379 (p. 653) is replaced by
a plus sign.

The solution of the distortion (shear or s-) wave can be obtained in similar manner using arbitrary
functions of displacements for the forward-moving waves:
(13.380)

In this case, the stress component that is perpendicular to the x-direction is:
(13.381)

If material discontinuity exists at , the impinging stress wave is divided into a refracted wave and
a reflected wave. The reflected wave disturb the solution in the region . Therefore, if the reflected
wave is removed, a convex boundary can be used to model the infinite boundary for the dynamic
analysis.

In typical cases, the waves are not perpendicular to the interface, so the stress components can be
given in terms of normal and shear tractions:

(13.382)

where:

, = normal stress and normal velocity of the particle, respectively

, = tangential stress and tangential velocity of the particle, respectively
and = non-dimensional constants

If the constants and are set as unity, it is called the standard absorbing boundary ([173] (p. 934)),
and a damping matrix is built into the structural infinite elements using the stress components in
Equation 13.382 (p. 654).

13.258. Reserved for Future Use

This section is reserved for future use.

13.259. Reserved for Future Use

This section is reserved for future use.

13.260. Reserved for Future Use

This section is reserved for future use.

13.261. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
654 Published: 2023-10-23
 REINF263 - 2D Smeared Reinforcing

13.262. Reserved for Future Use

This section is reserved for future use.

13.263. REINF263 - 2D Smeared Reinforcing

Base Shape Functions (for each Integration

Matrix or Vector
Element layer) Points (p. 414)
Across the
Equation 11.6 (p. 367), length: 1
Linear 2D
Equation 11.7 (p. 367), and
solid or shell
Equation 11.8 (p. 367) Across the
Stiffness and Stress section: 1
Stiffness Matrices and
Thermal Load Vector Across the
Quadratic Equation 11.19 (p. 368), length: 2
3D solid or Equation 11.20 (p. 368), and
shell Equation 11.21 (p. 368) Across the
section: 1
Same as stiffness
Mass Matrix Same as stiffness matrix.
matrix.
Pressure Load Vector N/A N/A

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 655
Element Library

Load Type Distribution

Element Temperature Linear along the length, constant across the section.
Nodal Temperature N/A
Pressure N/A

13.263.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element. See Stiffness and Mass Matrices of a Reinforcing Layer (p. 659) for the general formu-
lation of the reinforcing stiffness and mass matrices.

13.264. REINF264 - 3D Discrete Reinforcing

P P W
O X
O
M
B V
M N U
KK N
A
Y II JJ
II JJ L S
T Z
L
K K
I
R
Q
I J
J

L L

R
P KK
Q JJ
II JJ II
K O K

I I
N
M
J J

L L O
K K

II JJ JJ
II P KK
N

I I
J J
M

II II
KK
I JJ I JJ
K
J J

II
I
JJ
J

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
656 Published: 2023-10-23
 REINF264 - 3D Discrete Reinforcing

13.264.1. Structural
Base Shape Functions (for each Integration
Matrix or Vector
Element layer) Points (p. 414)
Linear 3D
Equation 11.6 (p. 367), Across the length: 1
spar, beam,
Equation 11.7 (p. 367), and Across the section: 1
solid, or
Stiffness and Stress Equation 11.8 (p. 367)
shell
Stiffness Matrices and
Thermal Load Vector Quadratic
Equation 11.19 (p. 368), Across the length: 2
3D beam,
Equation 11.20 (p. 368), and Across the section: 1
solid, or
Equation 11.21 (p. 368)
shell
Mass Matrix Same as stiffness matrix
Pressure Load Vector N/A N/A

Load Type Distribution

Element Temperature Linear along the length, constant across the section.
Nodal Temperature N/A
Pressure N/A

13.264.2. Thermal
Base Shape Functions (for each Integration
Matrix or Vector
Element layer) Points (p. 414)
Linear 3D
Conductivity Matrix and Equation 11.13 (p. 368) Across the length: 2
spar
Heat Generation Load
Vector Quadratic
Equation 11.25 (p. 368) Across the length: 3
3D
Specific Heat Matrix Same as conduction matrix

Load Type Distribution

Element HGEN Linear along the length, constant across the section.

13.264.3. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general structural element formu-
lations used by this element. See Stiffness and Mass Matrices of a Reinforcing Layer (p. 659) for the
general structural formulation of the reinforcing stiffness and mass matrices. Similar formulations can
be derived for thermal analysis.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 657
Element Library

13.265. REINF265 - 3D Smeared Reinforcing

P P W
O X
O
M
B V
M LL U
N N
KK A
Y PP LL OO KK
L S
II II T Z
L NN
JJ K MM JJ K
I
R
Q
I J
J

L LL
KK
L
K
R
P PP KK, LL, OO
Q
II
O K II
NN JJ
MM I
I JJ N
M J

LL OO KK
L O K
PP
NN
P
Z N
II JJ
MM
I
Y
J
X M

13.265.1. Structural
Base Shape Functions (for each Integration
Matrix or Vector
Element layer) Points (p. 414)
Equation 11.70 (p. 374), In-plane: 1 x 1
Linear 3D
Equation 11.71 (p. 374), and Through-the-thickness: 1
Stiffness and Stress solid or shell
Equation 11.72 (p. 374)
Stiffness Matrices and
Thermal Load Vector Quadratic Equation 11.82 (p. 375), In-plane: 2 x 2
3D solid or Equation 11.83 (p. 375), and Through-the-thickness: 1
shell Equation 11.84 (p. 375)
Same as stiffness
Mass Matrix Same as stiffness matrix
matrix
Pressure Load Vector N/A N/A

Load Type Distribution

Bilinear in plane of each reinforcing layer, constant
Element Temperature
through-the-thickness of each layer.
Nodal Temperature N/A
Pressure N/A

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
658 Published: 2023-10-23
 REINF265 - 3D Smeared Reinforcing

13.265.2. Thermal
Base Shape Functions (for each Integration
Matrix or Vector
Element layer) Points (p. 414)
In-plane: 2 x 2
Linear 3D
Equation 11.77 (p. 374)
solid Through-the-thickness:
Conductivity Matrix and 1
Heat Generation Load
Vector In-plane: 3 x 3
Quadratic
Equation 11.86 (p. 375)
3D solid Through-the-thickness:
1
Specific Heat Matrix Same as conduction matrix

Load Type Distribution

Bilinear in plane of each reinforcing layer, constant
Element HGEN
through-the-thickness of each layer.

13.265.3. Other Applicable Sections

Structures (p. 5) describes the derivation of element matrices and load vectors as well as stress
evaluations. General Element Formulations (p. 49) gives the general structural element formulations
used by this element. Similar formulations can be derived for thermal analysis.

13.265.4. Stiffness and Mass Matrices of a Reinforcing Layer

Each layer of reinforcing fibers is simplified as a membrane with unidirectional stiffness. The equivalent
membrane thickness h is given by:
(13.383)

where:

A = cross-section area of each fiber (input on SECDATA command)

S = distance between two adjacent fibers (input on SECDATA command)

We assume that the reinforcing fibers are firmly attached to the base element (that is, no relative
movement between the base element and the fibers is allowed). Therefore, the degrees of freedom
(DOF) of internal layer nodes (II, JJ, KK, LL, etc.) can be expressed in terms of DOFs of the external
element nodes (I, J, K, L, etc.). Taking a linear 3D solid base element as the example, the DOFs of an
internal layer node II can be shown as:

(13.384)

where:

{uII, vII, wII} = displacements of internal layer node II

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 659
Element Library

{ui, vi, wi} = displacements of base element node i

Ni (ξII, ηII, ζII) = value of trilinear shape function of node i at the location of internal node II

Similar relationships can be established for other type of base elements. The stiffness and mass
matrices of each reinforcing layer are first evaluated with respect to internal layer DOFs. The equivalent
stiffness and mass contributions of this layer to the element is then determined through relationship
(Equation 13.384 (p. 659)).

13.266. Reserved for Future Use

This section is reserved for future use.

13.267. Reserved for Future Use

This section is reserved for future use.

13.268. Reserved for Future Use

This section is reserved for future use.

13.269. Reserved for Future Use

This section is reserved for future use.

13.270. Reserved for Future Use

This section is reserved for future use.

13.271. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
660 Published: 2023-10-23
 SOLID272 - General Axisymmetric Solid with 4 Base Nodes

13.272. SOLID272 - General Axisymmetric Solid with 4 Base Nodes

Matrix or Vector Geometry Shape Functions Integration Points (p. 414)*

Equation 11.267 (p. 393),
Equation 11.268 (p. 393),
Quad Equation 11.271 (p. 394), 2 x 2 x Nc
Equation 11.272 (p. 394), and
Stiffness and Stress Equation 11.273 (p. 394)
Stiffness Matrices; and
Thermal Load Vector Equation 11.267 (p. 393),
Equation 11.268 (p. 393),
Triangle Equation 11.274 (p. 394), 1 x Nc
Equation 11.275 (p. 394), and
Equation 11.276 (p. 394)
Quad 2 x 2 x Nc
Mass Matrix Same as stiffness matrix
Triangle 1 x Nc
Same as stiffness matrix, specialized to
Pressure Load Vector 2 x Nc
face

* Nc = the number of integration points in the circumferential direction. The Nc integration

points are circumferentially located at:

• the nodal planes, and

• midway between the nodal planes (that is, at the integration planes)

so that Nc = (2 * Nnp), where Nnp = number of nodal planes (KEYOPT(2)). Exception: If

KEYOPT(2) = 1, then Nc = 1.

Load Type Distribution

Bilinear across element on rz plane, linear in circumferential
Element Temperature
direction
Nodal Temperature Same as element temperature distribution

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 661
Element Library

Load Type Distribution

Pressure Linear along each face

13.272.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.272.2. Assumptions and Restrictions

Although the elements are initially axisymmetric, the loads and deformation can be general in non-
axisymmetric 3D. The displacements are interpolated in elemental coordinate system by interpolation
functions, but the user can define the nodal displacements in any direction.

13.273. SOLID273 - General Axisymmetric Solid with 8 Base Nodes

Matrix or Vector Geometry Shape Functions Integration Points (p. 414)*

Equation 11.267 (p. 393),
Equation 11.268 (p. 393),
Quad Equation 11.277 (p. 395), 2 x 2 x Nc
Equation 11.278 (p. 395), and
Stiffness and Stress Equation 11.279 (p. 395)
Stiffness Matrices; and
Thermal Load Vector Equation 11.267 (p. 393),
Equation 11.268 (p. 393),
Triangle Equation 11.280 (p. 395), 3 x Nc
Equation 11.281 (p. 395), and
Equation 11.282 (p. 395)
Quad 3 x 3 x Nc
Mass Matrix Same as stiffness matrix
Triangle 3 x Nc
Same as stiffness matrix, specialized to
Pressure Load Vector 2 x Nc
face

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
662 Published: 2023-10-23
 Reserved for Future Use

* Nc = the number of integration points in the circumferential direction. The Nc integration

points are circumferentially located at:

• the nodal planes, and

• midway between the nodal planes (that is, at the integration planes)

so that Nc = (2 * Nnp), where Nnp = number of nodal planes (KEYOPT(2)). Exception: If

KEYOPT(2) = 1, then Nc = 1.

Load Type Distribution

Biquadratic across element on rz plane and linear between nodal
Element Temperature
planes in the circumferential direction
Nodal Temperature Same as element temperature distribution
Pressure Linear along each face

13.273.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.273.2. Assumptions and Restrictions

Although the elements are initially axisymmetric, the loads and deformation can be general in non-
axisymmetric 3D. The displacements are interpolated in elemental coordinate system by interpolation
functions, but the user can define the nodal displacements in any direction.

13.274. Reserved for Future Use

This section is reserved for future use.

13.275. Reserved for Future Use

This section is reserved for future use.

13.276. Reserved for Future Use

This section is reserved for future use.

13.277. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 663
Element Library

13.278. SOLID278 - 3D 8-Node Homogeneous/Layered Thermal Solid

SOLID278 is available in two forms:

• Homogeneous (nonlayered) thermal solid (KEYOPT(3) = 0, the default) - see SOLID278 - 3D 8-

Node Homogeneous Thermal Solid (p. 664).

• Layered thermal solid (KEYOPT(3) = 1) - see SOLID278 - 3D 8-Node Layered Thermal Solid (p. 664).

13.278.1. SOLID278 - 3D 8-Node Homogeneous Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix and Heat
Equation 11.227 (p. 387) 2x2x2
Generation Load Vector
Equation 11.227 (p. 387). Matrix is
Same as
Specific Heat Matrix diagonalized as described in
conductivity matrix
Lumped Matrices (p. 422)
Convection Surface Matrix and Load Equation 11.227 (p. 387) specialized
2x2
Vector to the face

13.278.2. SOLID278 - 3D 8-Node Layered Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
In-plane:
Conductivity Matrix and Heat
Equation 11.227 (p. 387) 2x2
Generation Load Vector
Thru-the-thickness:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
664 Published: 2023-10-23
 SOLID279 - 3D 20-Node Homogeneous/Layered Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
1, 3, 5, 7, or 9 per layer
Defaults to 3

Equation 11.227 (p. 387). Matrix is

Same as
Specific Heat Matrix diagonalized as described in
conductivity matrix
Lumped Matrices (p. 422)
Convection Surface Matrix and Load Equation 11.227 (p. 387) specialized
2x2
Vector to the face

13.278.3. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the element matrices and load vectors as well as heat
flux evaluations for a general thermal analysis element. Mass Transport is discussed in Mass Transport
(Advection) (p. 215) and Mass Transport (Advection) in the Thermal Analysis Guide. The Mass transport
option is only available for KEYOPT(3) = 0 the homogeneous (nonlayered) form.

13.279. SOLID279 - 3D 20-Node Homogeneous/Layered Thermal Solid

SOLID279 is available in two forms:

• Homogeneous (nonlayered) thermal solid (KEYOPT(3) = 0, the default) - see SOLID279 - 3D 20-
Node Homogeneous Thermal Solid (p. 665).

• Layered thermal solid (KEYOPT(3) > 0) - see SOLID279 - 3D 20-Node Layered Thermal Solid (p. 666).

13.279.1. SOLID279 - 3D 20-Node Homogeneous Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Brick Equation 11.239 (p. 388) 14
Conductivity Matrix and Wedge Equation 11.214 (p. 386) 3x3
Heat Generation Load
Vector Pyramid Equation 11.198 (p. 385) 2x2x2
Tet Equation 11.190 (p. 383) 4
Same as
Specific Heat Matrix Same as conductivity matrix.
conductivity matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 665
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Convection Surface Quad Equation 11.86 (p. 375) 3x3
Matrix and Load Vector Triangle Equation 11.61 (p. 373) 6

13.279.2. SOLID279 - 3D 20-Node Layered Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
In-plane:
2x2
Thru-the-thickness:
Brick Equation 11.239 (p. 388)
Conductivity Matrix and 1, 3, 5, 7, or 9 per layer
Heat Generation Load Defaults to 3
Vector
Wedge Equation 11.214 (p. 386) 3x3
Pyramid Equation 11.198 (p. 385) 2x2x2
Tet Equation 11.190 (p. 383) 4
Same as
Specific Heat Matrix Same as conductivity matrix.
conductivity matrix
Convection Surface Quad Equation 11.86 (p. 375) 3x3
Matrix and Load Vector Triangle Equation 11.61 (p. 373) 6

13.280. CABLE280 - 3D 3-Node Cable

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
666 Published: 2023-10-23
 SHELL281 - 8-Node Shell

Matrix or Vector Shape Functions Integration Points

Stiffness and Stress Equation 11.19 (p. 368), Equation 11.20 (p. 368), Along the
Stiffness Matrices; and Equation 11.21 (p. 368) length: 2
Thermal and
Newton-Raphson Load Across the
Vectors section: 1
Consistent Mass Matrix Same as stiffness matrix. Along the
and Pressure length: 3

Across the
section: 1

Load Type Distribution

Temperature Constant across cross-section, linear along length.
Pressure Linear along length. Pressure is assumed to act along the element
x axis.

13.281. SHELL281 - 8-Node Shell

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 667
Element Library

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
In-plane: 2 x 2

Thru-the-thickness:
1, 3, 5, 7, or 9 per layer
for section data input for
general shell option
Equation 11.82 (p. 375),
(KEYOPT(1) = 0)
Quad Equation 11.83 (p. 375), and
Equation 11.84 (p. 375)
Thru-the-thickness:
1 per layer for section
data input for membrane
shell option
Stiffness Matrix and (KEYOPT(1) = 1)
Thermal Load Vector
In-plane: 3

1, 3, 5, 7, or 9 per layer
for section data input for
general shell option
Equation 11.58 (p. 373),
(KEYOPT(1) = 0)
Triangle Equation 11.59 (p. 373), and
Equation 11.60 (p. 373)
1 per layer for section
data input for membrane
shell option
(KEYOPT(1) = 1)

Equation 11.82 (p. 375),

Same as stiffness
Quad Equation 11.83 (p. 375), and
matrix
Consistent Mass and Equation 11.84 (p. 375)
Stress Stiffness Matrices Equation 11.58 (p. 373),
Same as stiffness
Triangle Equation 11.59 (p. 373), and
matrix
Equation 11.60 (p. 373)
Equation 11.82 (p. 375),
Same as stiffness
Quad Equation 11.83 (p. 375), and
matrix
Equation 11.84 (p. 375)
Lumped Mass Matrix
Equation 11.58 (p. 373),
Same as stiffness
Triangle Equation 11.59 (p. 373), and
matrix
Equation 11.60 (p. 373)
Transverse Pressure Quad Equation 11.84 (p. 375) 2x2
Load Vector Triangle Equation 11.60 (p. 373) 3
Edge Pressure Load Same as in-plane mass matrix, specialized to
2
Vector the edge

Load Type Distribution

Element Temperature Linear thru thickness, bilinear in plane of element
Nodal Temperature Constant thru thickness, bilinear in plane of element

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
668 Published: 2023-10-23
 Reserved for Future Use

Load Type Distribution

Pressure Bilinear in plane of element, linear along each edge

References: Ahmad([1] (p. 925)), Cook([5] (p. 925)), Dvorkin([97] (p. 930)), Dvorkin([98] (p. 930)), Bathe and
Dvorkin([99] (p. 930)), Allman([114] (p. 931)), Cook([115] (p. 931)), MacNeal and Harder([116] (p. 931))

13.281.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.281.2. Assumptions and Restrictions

Normals to the centerplane are assumed to remain straight after deformation, but not necessarily
normal to the centerplane.

Each set of integration points thru a layer (in the r direction) is assumed to have the same element
(material) orientation.

13.281.3. Membrane Option

A membrane option is available for SHELL281 if KEYOPT(1) = 1. For this option, there is no bending
stiffness or rotational degrees of freedom. There is only one integration point per layer, regardless of
other input.

13.281.4. Shear Correction

The element uses an equivalent energy method to compute shear-correction factors. These factors
are predetermined based on the section lay-up at the start of solution.

13.282. Reserved for Future Use

This section is reserved for future use.

13.283. Reserved for Future Use

This section is reserved for future use.

13.284. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 669
Element Library

13.285. SOLID285 - 3D 4-Node Tetrahedral Structural Solid

L

Y
K

X I
Z

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Stiffness, Mass, and
Stress Stiffness Matrices; Equation 11.175 (p. 382) through
4
and Thermal Load Equation 11.186 (p. 383)
Vector
Equation 11.50 (p. 373) and
Pressure Load Vector 3
Equation 11.51 (p. 373)

Load Type Distribution

Element Temperature Same as shape functions
Nodal Temperature Same as shape functions
Pressure Linear over each face

13.285.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations. General Element Formulations (p. 49) gives the general element formulations used
by this element.

13.285.2. Theory
Stabilization terms are introduced and condensed out at element as enhanced term (Onate et
al.([92] (p. 929))).

13.286. Reserved for Future Use

This section is reserved for future use.

13.287. Reserved for Future Use

This section is reserved for future use.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
670 Published: 2023-10-23
 PIPE288 - 3D 2-Node Pipe

13.288. PIPE288 - 3D 2-Node Pipe

Integration
Matrix or Vector Option Shape Functions
Points (p. 414)
Stiffness and Stress Linear Equation 11.6 (p. 367), Along the length: 1
Stiffness Matrices; and (KEYOPT(3)=0) Equation 11.7 (p. 367),
Thermal and Equation 11.8 (p. 367), Across the section:
Newton-Raphson Equation 11.9 (p. 367), see text below
Load Vectors Equation 11.10 (p. 367),
and
Equation 11.11 (p. 367)
Quadratic Equation 11.19 (p. 368), Along the length: 2
(KEYOPT(3)=2) Equation 11.20 (p. 368),
Equation 11.21 (p. 368), Across the section:
Equation 11.22 (p. 368), see text below
Equation 11.23 (p. 368),
and
Equation 11.24 (p. 368)
Cubic Equation 11.27 (p. 368), Along the length: 3
(KEYOPT(3)=3) Equation 11.28 (p. 368),
Equation 11.29 (p. 369), Across the section:
Equation 11.30 (p. 369), see text below
Equation 11.31 (p. 369),
and
Equation 11.32 (p. 369)
Consistent Mass Linear Equation 11.6 (p. 367), Along the length: 2
Matrix and Pressure, (KEYOPT(3)=0) Equation 11.7 (p. 367),
Hydrostatic, and Equation 11.8 (p. 367), Across the section:
Hydrodynamic Load Equation 11.9 (p. 367), 1
Vectors Equation 11.10 (p. 367),
and
Equation 11.11 (p. 367)
Quadratic Equation 11.19 (p. 368), Along the length :
(KEYOPT(3)=2) Equation 11.20 (p. 368), 3
Equation 11.21 (p. 368),
Equation 11.22 (p. 368), Across the section:
Equation 11.23 (p. 368), 1
and
Equation 11.24 (p. 368)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 671
Element Library

Integration
Matrix or Vector Option Shape Functions
Points (p. 414)
Cubic Equation 11.27 (p. 368), Along the length: 4
(KEYOPT(3)=3) Equation 11.28 (p. 368),
Equation 11.29 (p. 369), Across the section:
Equation 11.30 (p. 369), 1
Equation 11.31 (p. 369),
and
Equation 11.32 (p. 369)
Lumped Mass Matrix Linear (KEYOPT(3) = Equation 11.6 (p. 367), Along the length: 2
0) Equation 11.7 (p. 367), and [1
Equation 11.8 (p. 367)
Across the section:
1
Quadratic Equation 11.19 (p. 368), Along the length :
(KEYOPT(3) = 2) Equation 11.20 (p. 368), 3 [1]
and
Equation 11.21 (p. 368) Across the section:
1
Cubic (KEYOPT(3) = Equation 11.27 (p. 368), Along the length: 4
3) Equation 11.28 (p. 368), [1]
and
Equation 11.29 (p. 369)` Across the section:
1

1. For partially submerged elements, the hydrostatic/hydrodynamic load vector integ-

ration point(s) is/are centered on the wetted portion.

Load Type Distribution

KEYOPT(1) = 0 Linear through wall and linear along length
Element Temperature KEYOPT(1) = 1 Bilinear across cross-section and linear along
length
Nodal Temperature Constant across cross-section, linear along length
Internal and External Constant, except as adjusted by the affect of internal and external
Pressures fluids.

References: Simo and Vu-Quoc([238] (p. 938)), Ibrahimbegovic([239] (p. 938)).

13.288.1. Assumptions and Restrictions

The element is based on Timoshenko beam theory; therefore, shear deformation effects are included.
It uses six components of strain, three direct strains and three shear strains. The element is well-suited
for linear, large rotation, and/or large strain nonlinear applications.

The element includes stress stiffness terms, by default, in any analysis using large deformation
(NLGEOM,ON). The stress stiffness terms provided enable the elements to analyze flexural, lateral and
torsional stability problems (using eigenvalue buckling or collapse studies with arc length methods).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
672 Published: 2023-10-23
 PIPE288 - 3D 2-Node Pipe

Transverse shear strain is constant through cross-section (that is, cross sections remain plane and
undistorted after deformation). The element can be used for slender or stout beams. Due to the lim-
itations of first-order shear deformation theory, slender to moderately thick beams can be analyzed.
Slenderness ratio of a beam structure may be used in judging the applicability of the element. It is
important to note that this ratio should be calculated using some global distance measures, and not
based on individual element dimensions. A slenderness ratio greater than 30 is recommended.

The elements are provided with section relevant quantities (area of integration, position, Poisson
function, function derivatives, etc.) automatically at a number of section points by the use of section
commands. Each section is assumed to be an assembly of number of nine-node cells. The number of
cells in both the circumferential and thickness directions is controlled by the SECDATA command.
Each cell has six integration points, as shown in the following figure:

Figure 13.37: Section Cell Model

The section includes internal fluid which contributes only mass and applied pressure, and insulation
which contributes only mass, and adds to the effective (hydraulic) diameter.

13.288.2. Ocean Effects

The following ocean effects apply to this element:
[Link]. Location of the Element
[Link]. Mass Matrix
[Link]. Load Vector
[Link]. Hydrostatic Effects
[Link]. Hydrodynamic Effects

[Link]. Location of the Element

The origin for any problem containing PIPE288 using ocean effects must be at the free surface
(mean sea level), although the origin can be offset via the OCDATA command’s Zmsl value. Further,
the Z axis is always the vertical axis, pointing away from the center of the earth.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 673
Element Library

If the element is intended to lie on the ocean floor, gap elements must be provided.

[Link]. Mass Matrix

The element mass matrix has several components, each organized on a mass-per-unit-length basis:
(13.385)

where:

ρ = density of the pipe wall (input as DENS on the MP command)

(13.386)

where:

ρint = density of the internal fluid (input as DENS on the MP command, with the material number
specified via M int on the SECDATA command)
(13.387)

where:

ρi = density of external insulation (input as DENS on the MP command, with the material
number specified via M ins on the SECDATA command)
(13.388)

where:

mah is input as ADDMAS on the SECCONTROL command

(13.389)

where:

Ca = coefficient of added mass of the external fluid (input as Ca on the OCDATA command)
ρw = fluid density (input as DENS on the MP command, with the material number specified via
MATOC on the OCDATA command)

madd only resists submerged lateral acceleration. If the element is above the instantaneous free
surface, madd = 0.0 . For modal (ANTYPE,MODAL) and harmonic (ANTYPE,HARMIC) analyses, the
surface location does not account for ocean waves.

[Link]. Load Vector

The element load vector consists of two parts:

• Distributed force per unit length to account for hydrostatic (buoyancy) effects ({F/L}b) as well as
axial nodal forces due to internal pressure and temperature effects {Fx}.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
674 Published: 2023-10-23
 PIPE289 - 3D 3-Node Pipe

• Distributed force per unit length to account for hydrodynamic effects (current and waves) ({F/L}d).

The hydrostatic and hydrodynamic effects work with the original diameter and length, that is, initial
strain and large deflection effects are not considered.

[Link]. Hydrostatic Effects

See Hydrostatic Loads (p. 425) in the Element Tools section of this document.

[Link]. Hydrodynamic Effects

See Hydrodynamic Loads (p. 428) in the Element Tools section of this document.

13.288.3. Stress Evaluation

Several stress evaluation options exist. The section strains and generalized stresses are evaluated at
element integration points and then linearly extrapolated to the nodes of the element.

If the material is elastic, stresses and strains are available after extrapolation in cross-section at the
nodes of section mesh. If the material is plastic, stresses and strains are moved without extrapolation
to the section nodes (from section integration points).

13.289. PIPE289 - 3D 3-Node Pipe

Matrix or Vector Shape Functions Integration Points (p. 414)

Equation 11.19 (p. 368),
Stiffness and Stress Along the length: 2
Equation 11.20 (p. 368),
Stiffness Matrices; and
Equation 11.21 (p. 368),
Thermal and Across the section: see
Equation 11.22 (p. 368),
Newton-Raphson Load PIPE288 - 3D 2-Node
Equation 11.23 (p. 368), and
Vectors Pipe (p. 671)
Equation 11.24 (p. 368)
Consistent Mass Matrix
Same as stiffness matrix Along the length: 3 [1]
and Pressure,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 675
Element Library

Matrix or Vector Shape Functions Integration Points (p. 414)

Hydrostatic, and
Hydrodynamic Load Across the section: 1
Vectors
Equation 11.19 (p. 368), Along the length: 3
Lumped Mass Matrix Equation 11.20 (p. 368), and
Equation 11.21 (p. 368) Across the section: 1

1. For partially submerged elements, the hydrostatic/hydrodynamic load vector integ-

ration point(s) is/are centered on the wetted portion.

Load Type Distribution

KEYOPT(1) = 0 Linear thru wall and linear along length
Element Temperature KEYOPT(1) = 1 Bilinear across cross-section and linear along
length
Nodal Temperature Constant across cross-section, linear along length
Internal and External Constant, except as adjusted by the affect of internal and external
Pressures fluids.

References: Simo and Vu-Quoc([238] (p. 938)), Ibrahimbegovic([239] (p. 938)).

The theory for this element is identical to that of PIPE288 - 3D 2-Node Pipe (p. 671), except that this
element is a 3-node pipe.

13.290. ELBOW290 - 3D 3-Node Elbow

y y z
L
z
I x x J
K

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Equation 11.19 (p. 368),
Equation 11.20 (p. 368),
Along the length: 2
Stiffness and Stress Equation 11.21 (p. 368),
Around the
Stiffness Matrices; and Equation 11.22 (p. 368),
circumference: 8 or higher
Thermal and Equation 11.23 (p. 368), and
Thru-the-thickness:
Newton-Raphson Load Equation 11.24 (p. 368)
1, 3, 5, 7, or 9 per layer
Vectors
Around the circumference: Fourier
Series
Along the length: 3
Mass Matrix and
Same as stiffness matrix Around the
Pressure Load Vector
circumference: 8 or higher

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
676 Published: 2023-10-23
 ELBOW290 - 3D 3-Node Elbow

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Thru-the-thickness:
1, 3, 5, 7, or 9 per layer

Load Type Distribution

KEYOPT(1) = 0 Linear thru wall and linear along length
Element Temperature
KEYOPT(1) = 1 Bilinear across cross-section and linear along length

Nodal Temperature Constant across cross-section, linear along length

Internal and External
Constant
Pressures

References:

Bathe and Almeida ([366] (p. 945))

Yan, Jospin, and Nguyen ([367] (p. 945))

13.290.1. Other Applicable Sections

Structures (p. 5) describes the derivation of structural element matrices and load vectors as well as
stress evaluations.

13.290.2. Assumptions and Restrictions

Pipe cross-sectional motions (that is, radial expansion, ovalization, and warping) are modeled with
Fourier series. The corresponding unknowns (Fourier magnitudes) are treated as internal degrees of
freedom. A higher number of Fourier modes may be required to achieve an adequate level of accuracy
in cross-sectional motions. Also, a higher number of integration points around the circumference may
be needed for capturing nonlinear material behaviors or ensuring sufficient numerical integration
accuracy.

No slippage is assumed between the element layers. Shear deflections are included in the element;
however, normals to the center wall surface are assumed to remain straight after deformation, but
not necessarily normal to the center surface. Therefore, constant transverse shears through the pipe
wall are allowed.

13.290.3. Shear Correction

The element uses an equivalent energy method to compute shear correction factors. These factors
are predetermined based on the section lay-up at the start of solution

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 677
Element Library

13.291. SOLID291 - 3D 10-Node Tetrahedral Thermal Solid

Integration
Matrix or Vector Shape Functions
Points (p. 414)
Conductivity Matrix and
Heat Generation Load Equation 11.190 (p. 383) 4
Vector
4 if KEYOPT(11) = 0
Specific Heat Matrix Same as conductivity matrix. 11 if KEYOPT(11) >
0
Convection Surface Equation 11.190 (p. 383) specialized to the face. 6
Matrix and Load Vector

13.291.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of thermal element matrices and load vectors as well as
heat flux evaluations.

13.292. PLANE292 - 2D 4-Node Thermal Element

K
t s

L J

Y,v

X,R

Matrix or Vector Geometry Shape Functions Integration

Points (p. 414)
Conductivity Matrix and Heat Quad Equation 11.130 (p. 377) 2x2
Generation Load Vector Triangle Equation 11.109 (p. 376) 1 if planar
3 if axisymmetric

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
678 Published: 2023-10-23
 PLANE293 - 2D 8-Node Thermal Solid

Matrix or Vector Geometry Shape Functions Integration

Points (p. 414)
Specific Heat Matrix Same as conductivity matrix. Matrix Same as conductivity
is diagonalized as described in matrix
Lumped Matrices (p. 422).
Convection Surface Matrix and Same as conductivity matrix 2
Load Vector evaluated at the face

13.292.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the element matrices and load vectors as well as heat
flux evaluations for a general thermal analysis element. Mass Transport is discussed in Mass Transport
(Advection) (p. 215) and Mass Transport (Advection) in the Thermal Analysis Guide.

13.293. PLANE293 - 2D 8-Node Thermal Solid

Integration
Matrix or Vector Geometry Shape Functions
Points (p. 414)
Conductivity Matrix and Quad Equation 11.141 (p. 378) 3x3
Heat Generation Load
Vector Triangle Equation 11.119 (p. 377) 3
Same as conductivity matrix. If KEYOPT(14) = 1,
Same as
Specific Heat Matrix matrix is diagonalized as described in Lumped
conductivity matrix
Matrices (p. 422)
Convection Surface Same as conductivity matrix, specialized to the
2
Matrix and Load Vector face

13.293.1. Other Applicable Sections

Heat Flow (p. 207) describes the derivation of the thermal element matrices and load vectors as well
as heat flux evaluations. If KEYOPT(14) = 1, the specific heat matrix is diagonalized as described in
Lumped Matrices (p. 422).

13.293.2. Assumptions and Restrictions

A dropped midside node implies that the edge is straight and that the temperature varies linearly
along that edge.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 679
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
680 Published: 2023-10-23
Chapter 14: Analysis Tools
The following analysis tools are available:
14.1. Acceleration Effect
14.2. Inertia Relief
14.3. Damping Matrices
14.4. Rotating Structures
14.5. Automatic Time-Stepping
14.6. Solving for Unknowns and Reactions
14.7. Equation Solvers
14.8. Mode-Superposition Method
14.9. Extraction of Modal Damping Parameter for Squeeze Film Problems
14.10. Newton-Raphson Procedure
14.11. Constraint Equations
14.12. Eigenvalue and Eigenvector Extraction
14.13. Analysis of Cyclically Symmetric Structures
14.14. Mass Related Information
14.15. Energies
14.16. Reduced-Order Modeling for State-Space Matrices Export
14.17. Enforced Motion in Structural Analysis

14.1. Acceleration Effect

Accelerations are applicable only to elements with displacement degrees of freedom (DOFs).

The acceleration vector {ac} which causes applied loads consists of a vector with a term for every degree
of freedom in the model. In the description below, a typical node having a specific location and accel-
erations associated with the three translations and three rotations will be considered:

(14.1)

where:

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 681
Analysis Tools

= accelerations in global Cartesian coordinates (input on ACEL command or CMACEL

command for component base acceleration)
= translational acceleration vector due to inertia relief (see Inertia Relief (p. 685))
= rotational acceleration vector due to inertia relief (see Inertia Relief (p. 685))
= translational acceleration vector due to rotations (defined below)
= angular acceleration vector due to input rotational accelerations (defined below)

Mechanical APDL defines three types of rotations:

Rotation 1: The whole structure rotates about each of the global Cartesian axes (input on
OMEGA and DOMEGA commands)
Rotation 2: The element component rotates about an axis defined by user (input on CMOMEGA
and CMDOMEGA commands).
Rotation 3: The global origin rotates about the axis by user if Rotation 1 appears or the rota-
tional axis rotates about the axis defined by user if Rotation 2 appears (input on CGOMGA,
DCGOMG, and CGLOC commands)

Up to two out of the three types of rotations may be applied on a structure at the same time. These
rotations induce accelerations and thus inertia forces when the equations are expressed in a rotating
reference frame.

14.1.1. Acceleration Due to One Rotation

When one rotation is defined (Rotation 1, 2, or 3), the angular acceleration vector is:
(14.2)

The translational acceleration vector used to calculate the inertia force in the rotating reference frame
is deduced from Equation 14.53 (p. 702):
(14.3)

where

= vector cross product

The translational acceleration vector is the difference between the translational acceleration expressed
in a stationary reference frame and the translational acceleration expressed in the rotating reference
frame, ignoring the Coriolis term.

The second term in Equation 14.3 (p. 682) leads to the Euler force and the last term to the centrifugal
force.

14.1.2. Acceleration Due to Two Rotations

The angular acceleration vector due to the rotations is:
(14.4)

The translational acceleration vector used to calculate the inertia force in the rotating reference frame
(second rotation) is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
682 Published: 2023-10-23
 Acceleration Effect

(14.5)

All other quantities are defined below, depending on the types of rotations considered.

[Link]. Rotation 1 and Rotation 3

{ω} = angular velocity vector defined about the global Cartesian origin (input on OMEGA
command)
{Ω} = angular velocity vector of the overall structure about the point CG (input on CGOMGA
command)
= angular acceleration vector defined about the global Cartesian origin (input on
DOMEGA command)
= angular acceleration vector of the overall structure about the point CG (input on
DCGOMG command)
{r} = position vector
{R} = vector from CG to the global Cartesian origin (computed from input on CGLOC
command, with direction opposite).

Figure 14.1: Rotational Coordinate System (Rotations 1 and 3)

Y' {r'}

X'
Z'

[Link]. Rotation 1 and Rotation 2

{ω} = angular velocity vector defined about the rotational axis of the element component
(input on CMOMEGA command)
{Ω} = angular velocity vector defined about the global Cartesian origin (input on OMEGA
command)
= angular acceleration vector defined about the rotational axis of the element com-
ponent (input on CMDOMEGA command)
= angular acceleration vector defined about the global Cartesian origin (input on
DOMEGA command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 683
Analysis Tools

{r} = position vector

{R} = vector from about the global Cartesian origin to the point on the rotational axis of
the component.

Figure 14.2: Rotational Coordinate System (Rotations 1 and 2)

{r'}

Y'

X'
Z'

[Link]. Rotation 2 and Rotation 3

{ω} = angular velocity vector defined about the rotational axis of the element component
(input on CMOMEGA command)
{Ω} = angular velocity vector of the overall structure about the point CG (input on CGOMGA
command)
= angular acceleration vector defined about the rotational axis of the element com-
ponent (input on CMDOMEGA command)
= angular acceleration vector of the overall structure about the point CG (input on
DCGOMG command)
{r} = position vector
{R} = vector from CG to the point on the rotational axis of the component

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
684 Published: 2023-10-23
 Inertia Relief

Figure 14.3: Rotational Coordinate System (Rotations 2 and 3)

{r'}

[Link]. Additional Term for Point Mass Element

For MASS21 with KEYOPT(3) = 0 and MATRIX27 with KEYOPT(3) = 2, additional Euler's equation
terms are considered:
(14.6)

where:

{M} = additional moments generated by the angular velocity

[I] = matrix of input moments of inertia
{ωT} = total applied angular velocities: = {ω} + {Ω}

14.2. Inertia Relief

An equivalent free-body analysis is performed if a static analysis (ANTYPE,STATIC) or a buckling analysis
(ANTYPE,BUCKLE) is used together with an inertia relief (IRLF,1), provided that enough supports are
defined (D). A static analysis performed with automatic inertia relief (AIRL) allows a free-body analysis
without the definition of supports.

Inertia relief is a technique in which the applied forces and torques are balanced by inertial forces induced
by an acceleration field. Consider the application of an acceleration field (to be determined) that precisely
cancels or balances the applied loads:

(14.7)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 685
Analysis Tools

= force components of the applied load vector

= translational acceleration vector due to inertia relief (to be determined)
ρ = density
vol = volume of model
= moment components of the applied load vector

= rotational acceleration vector due to inertia relief (to be determined)

x = vector cross product

In the finite element implementation, the position vector {r} and the moment in the applied load vector
are taken with respect to the center of mass. Equation 14.7 (p. 685) is rewritten in equivalent form
as:

(14.8)

where:

[Mt] = mass tensor for the entire finite element model

[Mr] = mass moments and mass products of the inertia tensor for the entire finite element
model

When applicable, [Mt], [Mr], and other mass-related data are calculated from the total rigid body mass
matrix. For more information, see Precise Calculation of Mass Related Information (p. 773). When not
applicable, these values are calculated using the equations in Mass-Related Information Calculation (p. 687).

Once [Mt] and [Mr] are developed, then and in Equation 14.8 (p. 686) can be solved. The body
forces that correspond to these accelerations are added to the user-imposed load vector, thereby
making the net or resultant reaction forces zero. The output inertia relief summary includes (output
as TRANSLATIONAL ACCELERATIONS) and (output as ROTATIONAL ACCELERATIONS). You may re-
quest only a mass summary for [Mt] and [Mr] (IRLF,-1).

The calculations for [Mt], [Mr], , and are made at every substep for every load step where they
are requested, and reflect changes in material density, applied loads, and deformation.

Several limitations apply:

• Inertia relief is applicable to linear or nonlinear structural analyses or buckling analyses with only
structural degrees of freedom.

• Element mass and/or density must be defined in the model.

• In a model containing both 2D and 3D elements, only Mt(1,1) and Mt(2,2) in [Mt] and Mr(3,3) in [Mr]
are correct in the mass summary. All other terms in [Mt] and [Mr] should be ignored. The acceleration
balance is, however, correct.

• Axisymmetric and generalized plane strain elements are not allowed.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
686 Published: 2023-10-23
 Inertia Relief

• If grounded gap or contact elements are in the model, their status should not change from the ori-
ginal status. Generally, bonded contact is preferred. Otherwise, the exact kinematic constraints stated
above might be violated.

14.2.1. Mass-Related Information Calculation

The computation for [Mt] and [Mr] proceeds on an element-by-element basis:

(14.9)

(14.10)

in which [me] and [Ie] relate to individual elements, and the summations are for all elements in the
model. The output `precision mass summary' (default output) includes components of [Mt] (labeled
as TOTAL MASS) and [Mr] (MOMENTS AND PRODUCTS OF INERTIA TENSOR ABOUT ORIGIN).

The evaluation for components of [me] are simply obtained from a row-by-row summation applied
to the element mass matrix over translational (x, y, z) degrees of freedom. It should be noted that
[me] is a diagonal matrix (mxy = 0, mxz = 0, etc.). The computation for [Ie] is based on the following
equation:
(14.11)

where:

[Me] = element mass matrix

[b] = matrix which consists of nodal positions and unity components. It is a submatrix of the
rigid body motion matrix [D], defined in Equation 14.305 (p. 773).

[Me] is dependent on the type of element under consideration. The description of the element mass
matrices [Me] is given in Derivation of Structural Matrices (p. 12). The derivation for [b] comes about
by comparing Equation 14.7 (p. 685) and Equation 14.8 (p. 686) on a per element basis, and eliminating
to yield:

(14.12)

where:

vol = element volume

If the mass matrix in Equation 14.11 (p. 687) is derived in a consistent manner, the components in [Ie]
are precise. This is demonstrated as follows. Consider the inertia tensor in standard form:

(14.13)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 687
Analysis Tools

which can be rewritten in product form:

(14.14)

The matrix [Q] is the following skew-symmetric matrix:

(14.15)

Next, the shape functions are introduced by way of their basic form,
(14.16)

where:

[N] = matrix containing the shape functions

Equation 14.15 (p. 688) and Equation 14.16 (p. 688) are combined to obtain:
(14.17)

where:

(14.18)

Inserting Equation 14.18 (p. 688) into Equation 14.14 (p. 688) leads to:
(14.19)

Noting that the integral in Equation 14.19 (p. 688) is the consistent mass matrix for a solid element:
(14.20)

It follows that Equation 14.11 (p. 687) is recovered from the combination of Equation 14.19 (p. 688) and
Equation 14.20 (p. 688).

Equation 14.18 (p. 688) and Equation 14.20 (p. 688) apply to all solid elements (in 2D, z = 0). For discrete
elements, such as beams and shells, certain adjustments are made to [b] in order to account for mo-
ments of inertia corresponding to individual rotational degrees of freedom. For 3D beams, for example,
[b] takes the form:

(14.21)

The [Ie] and [Mr] matrices are accurate when consistent mass matrices are used in Equation 14.11 (p. 687).
However, the following limitation applies:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
688 Published: 2023-10-23
 Inertia Relief

• The CENTER OF MASS output does not include the effects of offsets or tapering on current-techno-
logy BEAM, PIPE, or ELBOW elements (BEAM188, BEAM189, PIPE288, PIPE289, and ELBOW290).
Breaking up each tapered element into several individual elements gives a more accurate solution.

14.2.2. Supported Analysis Types for Inertia Relief

Inertia relief (IRLF,1) is supported for the following types of analyses: static (linear or nonlinear), linear
perturbation static, and buckling. Static analyses can be linear or nonlinear.

Automatic inertia relief (AIRL) is only supported for linear static analysis.

The equations used for inertia relief are outlined in the following sections.
[Link]. Linear Static Analysis
[Link]. Linear Static Analysis Without Supports
[Link]. Nonlinear Static Analysis
[Link]. Linear Perturbation Static Analysis with Inertia Relief
[Link]. Linear Perturbation Eigenvalue Buckling Analysis

[Link]. Linear Static Analysis

Assuming total displacement in a linear static analysis is:
(14.22)

where:

= total displacement
= deformable displacement
= rigid body motion
The corresponding acceleration vectors are: , and the general dynamic equation
without damping is:
(14.23)
where:

= total mass matrix

= total stiffness matrix
= total external force
= total acceleration vector, normally

(14.24)

Substituting Equation 14.22 (p. 689) into Equation 14.23 (p. 689) results in
(14.25)

For a static analysis the following terms drop out:

- no deformable acceleration

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 689
Analysis Tools

- rigid body motion generates zero forces

and Equation 14.25 (p. 689) becomes

(14.26)

and can be calculated by solving Equation 14.26 (p. 690). It is worth noting a few facts in this
static analysis. Since in a static analysis,Equation 14.24 (p. 689) can be viewed as:

(14.27)

This implies that the acceleration calculated from inertia relief only represents rigid body
acceleration. If the structure is subject to rigid body motion only (an unconstrained or minimally
constrained structure), deformable displacement is zero, . This condition leads to the
computation of rigid body acceleration (Equation 14.8 (p. 686)). When inertia relief loads are used
with displacement boundary conditions (Equation 14.26 (p. 690)), is no longer zero.

[Link]. Linear Static Analysis Without Supports

While inertia relief (IRLF) requires the use of supports to remove the rigid body modes from the
model, automatic inertia relief (AIRL) allows the analysis of a genuinely free body without any
support. The manual step of support definition and the influence of supports on the stress results
are thus avoided. The results from IRLF and AIRL may exhibit different displacements, but the stress
distribution should be very close, other than the side effects of the artificial supports.

Similar to IRLF, the purpose of AIRL is to balance the external forces applied on the structure by
a free-body acceleration, to be determined. However the static balance equation cannot be solved
directly given the absence of supports. As described here, an augmented linear system is formed,
and Lagrangian multipliers are used to remove the free-body motions.

In the previous section, the total displacement was decomposed into a deformable component
and a rigid-body motion component . Under the assumptions of no deformable accel-
eration and no strain energy induced by the rigid-body motion, it was shown that the general dy-
namic equation can be reformulated as a static problem:
(14.28)

The matrix [K] denotes a positive semi-definite stiffness matrix, and [M] denotes a positive definite
mass matrix. By introducing [R], the matrix containing the nrb rigid-body modes as column vectors,
the free-body acceleration vector may be expressed with respect to the nodal free-body ac-
celeration at a virtual reference point, denoted :
(14.29)

The static equilibrium balance is now expressed as:

(14.30)

Likewise, the free-body momentum balance can be expressed as:

(14.31)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
690 Published: 2023-10-23
 Inertia Relief

According to the previously described decomposition of the total displacement , the deformable
displacement does not contribute to the free-body motion of the structure. We choose to
impose this property on the deformable displacement weighted by the mass:
(14.32)

Using Equation 14.30 (p. 690) and Equation 14.32 (p. 691), the following augmented system is obtained:

(14.33)

The introduction of off-diagonal blocks [M][R] and [R]T[M] eliminates the rigid body modes from
the model and lets us solve simultaneously for and using a sparse direct solver. A
similar formulation is obtained when constraint equations exist in the model. Using the notations
from Constraint Equations (p. 742), the augmented system can be described as:

(14.34)

where:

and the constant unit time τ = 1 is introduced for the sake of dimensional homogeneity. The rigid
body modes [RI ] are extracted from the matrix [K*]. The new augmented system is obtained from
the original Equation 14.30 (p. 690) and Equation 14.32 (p. 691) by applying the constraint equations
according to the derivation made in Derivation of Matrix and Load Vector Operations (p. 742), and
by noticing that if and are respectively in the null space of [K] and
[K*], then .

[Link]. Nonlinear Static Analysis

For a nonlinear static analysis, the first step is to calculate inertia relief acceleration (Equa-
tion 14.27 (p. 690)) on the original geometry based on structural density at the beginning of the
substep of a load step by Equation 14.8 (p. 686). The nonlinear equation to be solved (similar to
Equation 14.16 (p. 688)) is:

(14.35)

where

is the stress vector

is the finite element derivative matrix ( )

The nonlinear Equation 14.35 (p. 691) is solved by Newton-Raphson iterative procedure as previously
described. The nonlinearity can be material or large deformation (NLGEOM,ON) nonlinearity. In the
case of NLGEOM,ON, the inertia force term is re-calculated at every substep to reflect
the effects of updated geometry by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 691
Analysis Tools

(14.36)

while retains the same values as in the first substep, is recalculated by Equation 14.12 (p. 687)
at every substep from the updated geometry to reflect the geometry change in NLGEOM,ON. Also,
is calculated incrementally at every substep to account for follower force effects (if any) if
NLGEOM,ON is issued.

[Link]. Linear Perturbation Static Analysis with Inertia Relief

The analysis process is the same as described in General Procedure for Linear Perturbation Analysis
in the Structural Analysis Guide, but extra attention should be paid to the application of inertia relief
loads.

1. In the base analysis, specify the inertia relief load (IRLF,1). The base analysis is static only. It can
be linear or nonlinear.

2. In the perturbed static analysis, the inertia relief load is recalculated to include the prestressed
matrix, or , by default. To exclude this inertia relief load vector and its effects you
must issue IRLF,0 in the first phase of the linear perturbation static analysis.

3. If large deformation (NLGEOM,ON) is used in the base analysis, the inertia relief loads
will be recalculated from the updated geometry in the perturbation phase.

[Link]. Linear Perturbation Eigenvalue Buckling Analysis

You can include inertia relief in an eigenvalue buckling analysis by following the procedure described
in Linear Perturbation Eigenvalue Buckling Analysis in the Structural Analysis Guide, paying special

attention to the inclusion or exclusion of the inertia relief load (Equation 14.36 (p. 692)).

1. In the base analysis, normally the inertia relief load is included by issuing IRLF,1. The inertia
relief load is used to generate a linear or nonlinear static solution, and consequently, the struc-
tural stresses for the prestressed tangent matrix are used in the perturbation phase.

Note:

For inertia relief, static is the only analysis type allowed for the base analysis. Inertia
relief loads are recalculated from the load vectors for the linear perturbation analysis
(See Application of Perturbation Loads (p. 868)).

2. In the first phase of the perturbed buckling analysis, the SOLVE,ELFORM command is executed.
By default, inertia relief loads are included in the load vector , which is used to
calculate the linearly perturbed stress stiffening matrix .

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
692 Published: 2023-10-23
 Inertia Relief

• When the base analysis is nonlinear, you may exclude this inertia relief load vector and its
effects by issuing IRLF,0 in the second phase of the perturbed buckling analysis so that the
inertia relief loads are included only in the base analysis for the prestressed effect.

• When the base analysis is linear, the inertia load should be included in the perturbation phase
since the inertia loads will be used to calculate the perturbed stress stiffening matrix in the
linear perturbation phase.

3. In the perturbed phase of the analysis, additional loads are allowed so that the buckling solution
is the result of the sum of the inertia loads and the other loads.

4. Critical buckling loads are determined using:

• Equation 15.254 (p. 864) for the case where a linear base analysis is used

• Equation 15.258 (p. 865) for the case where a nonlinear base analysis is used.
The resulting formulas for the critical buckling loads are as follows:

• Linear inertia relief bucking load:

• Nonlinear inertia relief buckling load when IRLF,1 is issued in the linear perturbation
phase:

• Nonlinear inertia relief buckling load when IRLF,0 is issued in the linear perturbation
phase:

where

is the rigid body acceleration calculated to balance out , the

load at the linear perturbation phase.

is the rigid body acceleration calculated to balance out , the

load vector from the base analysis.

is the buckling eigenvalue from the linear perturbation analysis.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 693
Analysis Tools

14.3. Damping Matrices

Damping may be introduced into a transient, harmonic, or damped modal analysis as well as a response
spectrum or PSD analysis. The type of damping allowed depends on the analysis, as follows:
[Link] (FULL) Analysis
14.3.2. Damped Modal Analysis
14.3.3. Harmonic (FULL) Analysis
14.3.4. Mode-Superposition Analysis

14.3.1. Transient (FULL) Analysis

The damping matrix, [C], may be used in transient as well as substructure generation:

• ANTYPE,TRANS with TRNOPT,FULL

• ANTYPE,SUBSTR with SEOPT,,,3

In its most general form, the damping matrix is composed of the following components:

(14.37)

where:

= structural damping matrix

= mass matrix multiplier (input on ALPHAD)
= structural mass matrix
= stiffness matrix multiplier (input on BETAD)

= structural stiffness matrix (see Note (p. 695) below)

= number of materials with MP,ALPD input
= mass matrix multiplier for material (input as ALPD on MP)

= portion of structural mass matrix based on material

= number of elements with mass proportional material damping input (TB,SDAMP,,,,ALPD)
= number of sections in an element with mass proportional material damping input
= mass proportional material damping factor for section point with material
(TB,SDAMP,,,,ALPD); see Material Damping in the Material Reference
= portion of element structural mass matrix based on section

= number of materials with MP,BETD input

= stiffness matrix multiplier for material (input as BETD on the MP)

= portion of structural stiffness matrix based on material

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
694 Published: 2023-10-23
 Damping Matrices

= number of elements with stiffness proportional material damping input

(TB,SDAMP,,,,BETD)
= number of sections in an element with stiffness proportional material damping input
= stiffness proportional material damping factor for section point with material
(TB,SDAMP,,,,BETD); see Material Damping in the Material Reference.
= portion of element structural stiffness matrix based on section

= number of elements with specified damping

= element damping matrix

= number of elements with Coriolis or gyroscopic damping

= element Coriolis or gyroscopic damping matrix; see Rotating Structures in the Theory
Reference (p. 700)
= constant structural damping coefficient (input with DMPSTR)
= number of materials with MP,DMPS input
= constant structural damping coefficient for material j (input with MP,DMPS)

= frequency for the calculation of equivalent viscous damping (DMPSFreq on TRNOPT

or DMPSFreqTab on DMPSTR)

Element damping matrices are available for:

LINK11 Linear Actuator MATRIX50 Superelement

COMBIN14 Spring-Damper SURF153 2D Structural Surface Effect
MATRIX27 Stiffness, Damping, or Mass SURF154 3D Structural Surface Effect
Matrix
COMBIN37 Control SURF159 General Axisymmetric
Surface
FLUID38 Dynamic Fluid Coupling MPC184 Multipoint Constraint (Joint)
Element
COMBIN40 Combination COMBI214 Bearing
CONTA174 Surface-to-Surface Contact COMBI250 Bushing
CONTA178 Node-to-Node Contact

Note:

, the structural stiffness matrix, , the portion of structural matrix based on material
, may include plasticity and/or large-deflection effects (that is, may be the tangent
matrix), but not , the portion of element structural stiffness matrix based on section
. In the case of a rotating structure, it may also include spin softening or rotating damping
effects. Stiffness matrices generated by contact elements and joint elements are not included
in the damping matrix generation.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 695
Analysis Tools

For the special case of thin-film fluid behavior, damping parameters may be computed for structures
and used in a subsequent structural analysis (see Extraction of Modal Damping Parameter for Squeeze
Film Problems (p. 727)).

14.3.2. Damped Modal Analysis

For damped modal analysis (ANTYPE,MODAL with MODOPT,QRDAMP or MODOPT,DAMP), the
damping matrix is complex and is written as:

(14.38)

where:

= the imaginary number

= the circular frequency
= the structural damping matrix defined in Equation 14.37 (p. 694). The equivalent
damping terms based on are not included in a damped modal analysis.
= constant structural damping coefficient (input with DMPSTR command)
= number of materials with MP,DMPS input
= constant structural damping coefficient for material j (input with MP,DMPS)

= portion of structural stiffness matrix based on material j

= number of elements with specified imaginary stiffness matrix

= imaginary stiffness element matrix

Element imaginary stiffness matrices are available for:

COMBIN14 Spring-Damper COMBI250 Bushing

14.3.3. Harmonic (FULL) Analysis

The damping matrix ( ) used in harmonic analyses (ANTYPE,HARM with Method = FULL, AUTO,
or VT on the HROPT command) is composed of the following components:

(14.39)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
696 Published: 2023-10-23
 Damping Matrices

= structural damping matrix

= mass matrix multiplier (input on ALPHAD)
= structural mass matrix
= stiffness matrix multiplier (input on BETAD)
= constant structural damping coefficient (input on DMPSTR)
= excitation circular frequency
= number of materials with MP,ALPD input
= mass matrix multiplier for material (input as ALPD on MP)

= portion of structural mass matrix based on material

= number of elements with mass proportional material damping input (TB,SDAMP,,,,ALPD)
= number of sections in an element with mass proportional material damping input
= mass proportional material damping factor for section point with material (input
on TB,SDAMP,,,,ALPD); see Material Damping in the Material Reference
= portion of element structural mass matrix based on section

= number of materials with MP,BETD, MP,DMPS, or TB,SDAMP,,,,STRU input

= stiffness matrix multiplier for material (input as BETD on MP)

= constant structural damping coefficient for material (input as DMPS on MP)

= structural damping coefficient for material (TB,SDAMP,,,,STRU)

= portion of structural stiffness matrix based on material

= number of elements with stiffness proportional material damping input

(TB,SDAMP,,,,BETD)
= number of sections in an element with stiffness proportional material damping input
= stiffness proportional material damping factor for section point with material
(TB,SDAMP,,,,BETD); see Material Damping in the Material Reference.
= portion of element structural stiffness matrix based on section

= number of elements with specified damping

= element damping matrix

= number of elements with viscoelastic damping
= element viscoelastic damping matrix (input as TB,PRONY)
= number of elements with Coriolis or gyroscopic damping

= element Coriolis or gyroscopic damping matrix; see Rotating Structures in the Theory
Reference (p. 700)
= number of elements with specified imaginary stiffness matrix

= imaginary element stiffness matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 697
Analysis Tools

The input exciting frequency, , is defined in the range between and via:

= beginning frequency (input as FREQB on HARFRQ command)

= end frequency (input as FREQE on HARFRQ command)

Substituting Equation 14.39 (p. 696) into the harmonic response equation of motion (Equa-
tion 15.66 (p. 815)) and rearranging terms yields:

(14.40)

The complex stiffness matrix in the first row of the equation consists of the normal stiffness matrix
augmented by the structural damping terms given by , , , , and which produce an
imaginary contribution. Structural damping is independent of the forcing frequency, , and produces
a damping force proportional to displacement (or strain). The terms , , and are damping coef-
ficients and not damping ratios.

The second row consists of the usual viscous damping terms and is linearly dependent on the forcing
frequency, , and produces forces proportional to velocity.

Viscoelastic damping (see Harmonic Viscoelasticity) introduces a contribution to the complex

stiffness matrix via the loss moduli. Note that the stresses are also computed using the loss moduli,
whereas in the case of structural damping, which is a phenomenological model, the stresses are
computed only using the real material properties and g is not used in the stress calculations.

14.3.4. Mode-Superposition Analysis

For mode-superposition based analyses:

• ANTYPE,HARM with HROPT,MSUP

• ANTYPE,TRANS with TRNOPT,MSUP

• ANTYPE,SPECTRUM with Sptype = SPRS, MPRS, or PSD on the SPOPT command

the damping matrix is not explicitly calculated, but rather the damping is defined directly in terms
of a damping ratio . The damping ratio is the ratio between actual damping and critical damping.

The damping ratio for mode i is the combination of:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
698 Published: 2023-10-23
 Damping Matrices

(14.41)

where:

= constant modal damping ratio (input on DMPRAT)

= modal damping ratio for mode shape i (see below)

= circular natural frequency associated with mode shape

= natural frequency associated with mode shape

= mass matrix multiplier (input on ALPHAD)
= stiffness matrix multiplier (input on BETAD)

The modal damping ratio can be defined for each mode directly via the MDAMP command (un-
damped modal analyses only).

Alternatively, for the case where multiple materials are present whose damping ratios are different,
an effective mode-dependent damping ratio can be defined in the modal analysis if material-de-
pendent damping is defined and the element results are calculated during the expansion and written
to the mode file (MXPAND,,,,YES,,YES). This effective damping ratio is computed from the ratio of the
strain energy in each material in each mode using:

(14.42)

where:

Nm = number of materials

= damping ratio for material j (input as DMPR on MP command); see note below

= strain energy contained in mode i for material j

= displacement vector for mode i
= stiffness matrix of part of structure of material j

These mode-dependent (and material-dependent) ratios, , will be carried over into the subsequent
mode-superposition or spectrum analysis. Note that any manually-defined damping ratios (MDAMP)
will overwrite those computed in the modal analysis via Equation 14.42 (p. 699).

For harmonic analyses (ANTYPE,HARM with HROPT,MSUP), constant structural damping may also be
included. In this case, the harmonic equation of motion in modal coordinates (Equation 15.86 (p. 820))
is:
(14.43)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 699
Analysis Tools

where:

= complex modal coordinate

= natural circular frequency of mode i

= fraction of critical damping for mode i as given in Equation 14.41 (p. 699)
= constant structural damping coefficient (input on DMPSTR command)
= complex force in modal coordinates

[Link]. Mode-Superposition Analyses Following a QR Damp Modal Analysis

When QRDAMP is used for the modal analysis, the modal damping matrix used for the mode super-
position analysis contains the contribution of the global damping, the damping ratios, as well as
the element and material dependent damping:
(14.44)
where:

is the global damping contribution

Note that the second term is not necessarily a diagonal matrix.

is the damping ratios contribution

is the element and material dependent damping contribution

where is the damping matrix defined in Equation 14.39 (p. 696) without the first two
terms for mode superposition analysis and Equation 14.37 (p. 694) without the first two
terms for the mode superposition transient analysis.

14.4. Rotating Structures

When a structure is rotating, inertial forces and moments are observed. Their expressions and physical
meanings are different depending on which reference frame is chosen: a rotating reference frame
(O'X'Y'Z') or a stationary reference frame, the global Cartesian (OXYZ) – see Figure 14.4: Reference
Frames (p. 701).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
700 Published: 2023-10-23
 Rotating Structures

Figure 14.4: Reference Frames

By default (no CORIOLIS command), the systems equations are solved in the rotating reference frame.
In a static linear analysis, the centrifugal load is accounted for, while in a static nonlinear analysis,
centrifugal load as well as all spin softening effects are included. In a nonlinear transient analysis all
nonlinear inertia forces are also added.

Some of these nonlinear inertia terms can be linearized to support linear analyses of a rotating structure
(CORIOLIS command). This is the case for the Coriolis and spin softening effects in the rotating reference
frame and for the gyroscopic effect in the stationary reference frame.

Based on settings defined with the CORIOLIS command, a stationary or rotating reference frame
problem is resolved. The resulting equations are different and are addressed below.

See the Notes section of the CORIOLIS command for a list of supported elements.

Synchronous and asynchronous forces are discussed in Harmonic Analysis with Rotating Forces on Ro-
tating Structures (p. 823).

14.4.1. Rotating Reference Frame Equations

In Figure 14.4: Reference Frames (p. 701) above, a part or component is rotating at angular velocity
{ω}, with components ωx, ωy, and ωz defined in the stationary reference frame. The position of point
P with reference to (OXYZ) is {r}, while its position with reference to the rotating frame of reference
( ) is , and:
(14.45)

The velocity of point P as observed in the stationary and rotating frames is defined as:
(14.46)

and
(14.47)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 701
Analysis Tools

The velocity of point P observed in the stationary frame can be expressed as:
(14.48)

where:

The translational accelerations of point P as observed in the stationary and rotating frames of reference:
(14.49)

and
(14.50)

By substituting Equation 14.48 (p. 702) into Equation 14.49 (p. 702) and using Equation 14.50 (p. 702),
we obtain:
(14.51)

where:

We assume that the origin of the rotating system is fixed, so that:

(14.52)

By substituting Equation 14.52 (p. 702) into Equation 14.51 (p. 702),
(14.53)

The second term , is the rotational acceleration and leads to a load (sometimes called the
Euler force – see Acceleration Effect (p. 681))

The third term is the centrifugal acceleration (see Guo et al.([361] (p. 945)) which
leads to the centrifugal load (Acceleration Effect (p. 681)) and the spin softening effect.

The last term leads to the Coriolis force.

[Link]. Coriolis Matrix, Spin-Softening Matrix, and Centrifugal Load

By applying virtual work from the d'Alembert force, the contribution of the first term {a r} (Equa-
tion 14.53 (p. 702)) to the virtual work introduces the mass matrix of the element (Guo et
al.([361] (p. 945))).

(14.54)

where:

[Me] = element mass matrix

[N] = shape function matrix
ρ = element density

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
702 Published: 2023-10-23
 Rotating Structures

The Coriolis matrix is deduced from the last term of Equation 14.53 (p. 702) (Guo et al.([361] (p. 945))):

(14.55)

where:

[Ge] = element Coriolis damping matrix

and thus [Ge] are skew symmetric matrices.

The governing dynamics equation of motion in a rotating reference frame can be written as,
(14.56)

where:

n = number of elements
[K] = global stiffness matrix
[KSP] = global stiffness matrix due to spin softening (p. 46)
{F} = load vector (sum of external and centrifugal loads)

[Link]. Coriolis Force

In a quasi-static analysis, the Coriolis effect is introduced as a load vector:
(14.57)

where:

{Fc} = Coriolis force

= nodal velocity vector (input using the IC or ICROTATE command).

14.4.2. Stationary Reference Frame Equations

Suppose a structure is spinning around an axis Δ. If a rotation about an axis perpendicular to Δ is
applied to the structure, then a reaction moment appears. It is called the gyroscopic moment. Its axis
is perpendicular to both the spinning axis Δ and the applied rotation axis.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 703
Analysis Tools

The gyroscopic effect is thus coupling rotational degrees of freedom which are perpendicular to the
spinning axis.

Let us consider the spinning axis is along X so:

• The spinning velocity (input using the OMEGA or CMOMEGA commands) is .

• The displacements perpendicular to the spin axis are uy and uz.

• The corresponding rotations are θy and θz, and the angular velocities are .

[Link]. Gyroscopic Matrix

The gyroscopic finite element matrix is calculated from the kinetic energy due to the inertia forces.

The kinetic energy for lumped mass and beam element (Nelson and McVaugh([359] (p. 945))) is de-
tailed in Kinetic Energy for the Gyroscopic Matrix Calculation of Lumped Mass and Beam Ele-
ment (p. 704) below.

The general expression of the kinetic energy used for the development of the gyroscopic matrices
for all other elements (Geradin and Kill [377] (p. 946)) is presented in General Expression of the
Kinetic Energy for the Gyroscopic Matrix Calculation (p. 705)

[Link].1. Kinetic Energy for the Gyroscopic Matrix Calculation of Lumped Mass and
Beam Element
Both mass and beam should be axisymmetric around the spinning axis. The spinning axis is along
one of the principal axis of inertia for lumped mass. For the beam, it is along the length.

Two reference frames are used (see Figure 14.4: Reference Frames (p. 701)) (OXYZ) which is station-
ary and which is attached to the cross-section with axis normal to it.

is defined using 3 successive rotations:

• θz around Z axis to give

• θy around axis to give

• θx around axis to give

Hence for small rotations θy and θz, the instantaneous angular velocity is:

(14.58)

1. For a lumped mass, considering only second order terms, kinetic energy is obtained using
the instantaneous angular velocity vector in Equation 14.58 (p. 704).

(14.59)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
704 Published: 2023-10-23
 Rotating Structures

where:

m = mass
Id = diametral inertia
Ip = polar inertia

The first two terms contribute to the mass matrix of the element and the last term gives
the gyroscopic matrix.

2. The beam element is considered as an infinite number of lumped masses. The gyroscopic
kinetic energy of the element is obtained by integrating the last term of Equa-
tion 14.59 (p. 704) along the length of the beam:

(14.60)

where:

ρ = density
Ix = moment of inertia normal to x
L = length of the beam element

Gyroscopic matrix is deduced based on the element shape functions. (See PIPE288).

[Link].2. General Expression of the Kinetic Energy for the Gyroscopic Matrix Calcu-
lation
A point, in element i, with coordinates (x,y,z) in the stationary reference frame is considered. The
kinetic energy is
(14.61)

where:

EGki = gyroscopic kinetic energy of element i

Vi = volume of element i
dm = elementary mass

The gyroscopic matrix is then calculated using the element shape functions.

[Link]. Rotating Damping Matrix

In a linear approach, the relation between displacements in the stationary reference frame (0XYZ)
and displacements in the rotating reference frame (0X’Y’Z’) can be written as:
(14.62)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 705
Analysis Tools

where:

{r'} = the displacement vector in the rotating reference frame

[R] = the transformation matrix
{r} = the displacement vector in the stationary reference frame

Differentiating Equation 14.62 (p. 705) with respect to time, one obtains the expression for the velocity
vector:
(14.63)

where:

= the velocity vector in the rotating reference frame

= the velocity vector in the stationary reference frame

= the rotational matrix, as defined in Equation 14.55 (p. 703)

If structural damping is present in the rotating structure (proportional damping for example) or if
there is localized viscous damping (as in a damper), damping forces in the rotating reference frame
may be expressed as:
(14.64)

where:

= the damping forces in the rotating reference frame

[C] = the damping matrix

To obtain the damping forces in the stationary reference frame, first apply the transformation of
Equation 14.62 (p. 705):
(14.65)

where:

{Fd}= the damping forces in the stationary reference frame.

Then replace Equation 14.63 (p. 706) in Equation 14.64 (p. 706), the resulting expression in Equa-
tion 14.65 (p. 706) yields:
(14.66)

If the damping is isotropic (implementation assumption):

(14.67)

Where [B] is the rotating damping matrix:

(14.68)

In the particular case of spring-damper element (COMBIN14), general joint (MPC184), and stiffness
matrix (MATRIX27), the following average equation is used:
(14.69)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
706 Published: 2023-10-23
 Rotating Structures

[B] from Equation 14.68 (p. 706) and Equation 14.69 (p. 706) is a non-symmetric matrix which will
modify the apparent stiffness of the structure.

[Link]. Gyroscopic Force

In a quasi-static analysis, the gyroscopic effect is introduced as a load vector:
(14.70)

where

= gyroscopic force
= nodal velocity vector (input using the IC or ICROTATE command)
In this case, an additional centrifugal load is introduced based on the rotational velocity and rota-
tional velocity axis defined with ICROTATE. The translational acceleration used is similar to the
third term of Equation 14.53 (p. 702).

[Link]. Mass Unbalance Transient Forces

The transient forces of a small mass unbalance are expressed as described below.

A body is rotating at around the Z axis. Point A is the center of the body cross-section in plane
(X,Y). m is the mass unbalance located at point B. The distance AB is the eccentricity e, which is
constant. The instantaneous location of A is (ux,uy).

Figure 14.5: Mass Unbalance

In the (X,Y) plane, the mass unbalance coordinates can be expressed as:

(14.71)

Hence, its velocity is:

(14.72)

The kinetic energy of the mass unbalance can be written as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 707
Analysis Tools

(14.73)

The mass unbalance is very small with respect to the mass of the rotating body, hence the first two
terms are neglected. The last term is constant and has no contribution to the dynamics equations.

The kinetic energy reduces to:

(14.74)

Application of Lagrange’s equations leads to the following forces in the (X,Y) plane:

(14.75)

It is the sum of a centrifugal force due to the rotational velocity and a tangential force due to the
rotational acceleration.

14.5. Automatic Time-Stepping

The automatic time-stepping (or automatic loading) method adjusts the time-step size and/or the applied
loads automatically in response to the current state of the analysis. This method (accessed via
AUTOTS,ON) can be applied to structural, thermal, electric, and magnetic analyses performed in the
time domain (TIME), and includes static (or steady-state) (ANTYPE,STATIC) and dynamic (or transient)
(ANTYPE,TRANS) situations.

An important point to be made here is that automatic loading always works through the adjustment
of the time-step size; and that the loads that are applied are automatically adjusted if ramped boundary
conditions are activated (using KBC,0). In other words the time-step size is always subjected to possible
adjustment when automatic loading is engaged. Applied loads and boundary conditions, however, will
vary according to how they are applied and whether the boundary conditions are stepped or ramped.
That is why this method may also be thought of as automatic loading.

The automatic time-stepping algorithm has two key features:

• The capability to estimate the next time-step size. The esitmation is based on current and past
analysis conditions, and the algorithm makes proper load adjustments as necessary. That is, given
conditions at the current time, tn, and the previous time increment, Δtn, the primary aim is to determ-
ine the next time increment, Δtn+1. Since the determination of Δtn+1 is largely predictive, this part
of the automatic time-stepping algorithm is referred to as the time-step prediction component.

• The time-step bisection component. Its purpose is to decide whether or not to reduce the present
time-step size, Δtn, and redo the substep with a smaller step size. For example, working from the
last converged solution at time point tn-1, the present solution begins with a predicted time step,
Δtn. Equilibrium iterations are performed; and if proper convergence is either not achieved or not
anticipated, this time step is reduced to Δtn/2 (that is, it is bisected), and the analysis begins again
from time tn-1. Multiple bisections can occur per substep for various reasons (discussed later).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
708 Published: 2023-10-23
 Automatic Time-Stepping

14.5.1. Time-Step Prediction

At a given converged solution at time tn, and with the previous time increment Δtn, the goal is to
predict the appropriate time-step size to use as the next substep. The predicted time-step size is derived
from the results of several unrelated calculations and is most easily expressed as the minimization
statement:
(14.76)

where:

Δteq = time increment limited by the number of equilibrium iterations needed for convergence
at the last converged time point. The more iterations required for convergence, the smaller
the predicted time step. This is a general measure of all active nonlinearities. Increasing the
maximum number of equilibrium iterations (NEQIT) will tend to promote larger time-step
sizes.
Δt1 = time increment limited by the response eigenvalue computation for first-order systems
(such as thermal transients) (TINTP).
Δt2 = time increment limited by the response frequency computation for second-order systems
(such as structural dynamics). The aim is to maintain 20 points per cycle (described below).
When the middle step criterion is used, this criterion can be disabled.
Δtg = time increment representing the time point at which a gap or a nonlinear (multi-status)
element will change abruptly from one condition to another (status change). KEYOPT(7) en-
ables further control for the contact elements.
Δtc = time increment based on the allowable creep-strain increment.
Δtp = time increment based on the allowable plastic-strain increment. The limit is set at 5
percent per time step.
Δtm = time increment limited by the middle step residual tolerance (described below) for
second-order systems (such as structural dynamics) (MIDTOL). When enabled, the Δt2 criterion
can be disabled.
The program calculates several trial step sizes and selects the minimum step size for the next time
step. The predicted value is further restricted to a range of values expressed by:
(14.77)

and
(14.78)

where:

F = increase/decrease factor. F = 2, if static analysis; F = 3, if dynamic (see the ANTYPE and

TIMINT commands)
Δtmax = maximum time-step size (DTMAX on the DELTIM command or the equivalent
quantity calculated from the NSUBST command)
Δtmin = minimum time-step size (DTMIN on the DELTIM command or the equivalent quantity
calculated from the NSUBST command)
In other words, the current time-step is increased or decreased by a factor of 2 (or 3 if dynamic) at
most, and cannot be less than Δtmin or greater than Δtmax.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 709
Analysis Tools

It is also possible to control the time-step cutback during a nonlinear solution (CUTCONTROL), and
to set a decision parameter for automatically increasing the time-step interval (OPNCONTROL).

14.5.2. Time-Step Bisection

When bisection occurs, the current substep solution (Δtn) is removed, and the time-step size is reduced
by 50 percent. If applied loads are ramped (KBC,0), then the current load increment is also reduced
by the same amount. One or more bisections can take place for several reasons, namely:

1. The number of equilibrium iterations used for this substep exceeds the number allowed (NEQIT
command).

2. It appears likely that all equilibrium iterations will be used.

3. A negative pivot message was encountered in the solution, suggesting instability.

4. The largest calculated displacement DOF exceeds the limit (DLIM on the NCNV command).

5. An illegal element distortion is detected (for example, negative radius in an axisymmetric analysis).

6. For transient structural dynamics, when the middle step residual is greater than the given tolerance.
This check is done only when the middle step residual check is enabled by the MIDTOL command.

More than one bisection may be performed per substep. However, bisection of the time-step size is
limited by the minimum size (defined by DTMIN input on the DELTIM command or the equivalent
NSUBST input).

14.5.3. The Response Eigenvalue for 1st Order Transients

The response eigenvalue is used in the computation of Δt1 and is defined as:

(14.79)

where:

λr = response eigenvalue (item RESEIG for POST26 SOLU command and *GET command)
{Δu} = substep solution vector (tn-1 to tn)

[KT] = the Dirichlet matrix. In a heat transfer or an electrical conduction analysis this matrix
is referred to as the conductivity matrix; in magnetics this is called the magnetic "stiffness".
The superscript T denotes the use of a tangent matrix in nonlinear situations
[C] = the damping matrix. In heat transfer this is called the specific heat matrix.

The product of the response eigenvalue and the previous time step (Δtn) has been employed by
Hughes([146] (p. 933)) for the evaluation of 1st order explicit/implicit systems. In Hughes([146] (p. 933))
the quantity Δtnλ is referred to as the "oscillation limit", where λ is the maximum eigenvalue. For
unconditionally stable systems, the primary restriction on time-step size is that the inequality Δtnλ
>> 1 should be avoided. Hence it is very conservative to propose that Δtnλ = 1.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
710 Published: 2023-10-23
 Automatic Time-Stepping

Since the time integration used employs the trapezoidal rule (Equation 15.42 (p. 808)), all analyses of
1st order systems are unconditionally stable. The response eigenvalue supplied by means of Equa-
tion 14.79 (p. 710) represents the dominate eigenvalue and not the maximum; and the time-step re-
striction above is restated as:
(14.80)

This equation expresses the primary aim of automatic time-stepping for 1st order transient analyses.
The quantity Δtnλr appears as the oscillation limit output during automatic loading. The default is f
= 1/2, and can be changed (using OSLM and TOL on the TINTP command). The quantity Δt1 is ap-
proximated as:
(14.81)

[Link]. Additional Considerations for Thermal Transient Analysis

If the thermal solution reaches steady-state, the time step is allowed to increase by a factor of 2.
Steady-state is determined if the maximum temperature change is less than 3 degrees for 3 time
steps in a row. Use the OPNCONTROL command to modify these defaults, if desired.

For the Quasi option (THOPT command), the time step is also controlled by the property change
tolerance such that the chosen time step keeps the property change below the tolerance. See the
THOPT command for more details.

14.5.4. The Response Frequency for Structural Dynamics

The response frequency is used in the computation of Δt2 and is defined as (Bergan([106] (p. 930))):

(14.82)

where:

fr = response frequency (item RESFRQ for POST26 SOLU command and *GET command)
{Δu} = substep solution vector (tn-1 to tn)

[KT] = tangent stiffness matrix

[M] = mass matrix

This equation is a nonlinear form of Rayleigh's quotient. The related response period is:
(14.83)

Using Tr the time increment limited by the response frequency is:

(14.84)

When the middle step criterion is used, this criterion can be turned off.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 711
Analysis Tools

14.5.5. Creep Time Increment

The time-step size may be increased or decreased by comparing the value of the creep ratio Cmax
(Rate-Dependent Plasticity (Including Creep and Viscoplasticity) (p. 96)) to the creep criterion Ccr. Ccr
is equal to .10 unless it is redefined (using the CRPLIM command). The time-step estimate is computed
as:
(14.85)

Δtc is used in Equation 14.76 (p. 709) only if it differs from Δtn by more than 10%.

14.5.6. Plasticity Time Increment

The time-step size is increased or decreased by comparing the value of the effective plastic strain
increment (Equation 4.26 (p. 74)) to 0.05 (5%). The time-step estimate is computed as:

(14.86)

Δtp is used in Equation 14.76 (p. 709) only if it differs from Δtn by more than 10%.

14.5.7. Midstep Residual for Structural Dynamic Analysis

The midstep residual is used in the computation of Δtm. The midstep residual for the determination
of the time step is based on the following consideration. The solution of the structural dynamic ana-
lysis is carried out at the discrete time points, and the solution at the intermediate time remains un-
known. However, if the time step is small enough, the solution at the intermediate time should be
close enough to an interpolation between the beginning and end of the time step. If so, the unbalanced
residual from the interpolation should be small. On the other hand, if the time step is large, the inter-
polation will be very different from the true solution, which will lead to an unbalanced residual that
is too large. The time step is chosen to satisfy the criterion set by the user (e.g. MIDTOL command).

Refer to the discussion in Newton-Raphson Procedure (p. 730). The residual force at any time between
the time step n and n+1 can be written as:
(14.87)

where:

ν = intermediate state between the time step n and n+1 (0 < ν < 1)
{R} = residual force vector
= vector of the applied load at n + ν
= vector of the restoring load corresponding to the element internal load at n + ν,
which depends on the intermediate state of displacement at n + ν, and also the velocity and
acceleration at n + ν. This intermediate state is approximately calculated based on the
Newmark assumption.

A measure of the magnitude of {R} is established in a manner similar to the convergence check at
the end of the time step. (See Convergence (p. 734)). After the solution has converged at the end of
the time step (n+1), the midstep residual force is compared to the reference value:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
712 Published: 2023-10-23
 Solving for Unknowns and Reactions

(14.88)

where:

||{R}|| = magnitude (vector norm) of residual force vector

Rref = reference force (see Convergence (p. 734))

The convergence criterion for the midstep residual is defined by the value of τb (input as TOLERB
on MIDTOL command):

If τb > 0, the value is used as a tolerance. If τb = 0 is specified or τb is not specified, then a default
positive value is used as a tolerance. The midstep residual is assumed to have converged if its value
is within the desired tolerance (ε τb ). Depending on how well the convergence criterion is satisfied
the time-step size for the next increment is increased or kept unchanged.

If the midstep residual hasn’t converged (ε > τb), the time step is repeated with a smaller increment:
(14.89)

where:

Δtn = old time-step size

τb = midstep residual tolerance(TOLERB on MIDTOL command)

If τb < 0, the value is used as a reference force (reference moment is computed from reference force
value) for midstep convergence check. A procedure similar to the one described above is followed
with modified definition of time-step size:

(14.90)

14.6. Solving for Unknowns and Reactions

In general, the equations that are solved for static linear analyses are:
(14.91)

or
(14.92)

where:

[K] = total stiffness or conductivity matrix =

{u} = nodal degree of freedom (DOF) vector

N = number of elements

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 713
Analysis Tools

[Ke] = element stiffness or conductivity matrix

{Fr} = nodal reaction load vector

{Fa}, the total applied load vector, is defined by:

(14.93)

where:

{Fnd} = applied nodal load vector

{Fe} = total of all element load vector effects (pressure, acceleration, thermal, gravity)

Equation 14.91 (p. 713) thru Equation 14.93 (p. 714) are similar to Equation 15.1 (p. 794) thru Equa-
tion 15.4 (p. 795).

If sufficient boundary conditions are specified on {u} to guarantee a unique solution, Equa-
tion 14.91 (p. 713) can be solved to obtain the node DOF values at each node in the model.

Rewriting Equation 14.92 (p. 713) for linear analyses by separating out the matrix and vectors into those
DOFs with and without imposed values,

(14.94)

where:

s = subscript representing DOFs with imposed values (specified DOFs)

c = subscript representing DOFs without imposed values (computed DOFs)

Note that {us} is known, but not necessarily equal to {0}. Since the reactions at DOFs without imposed
values must be zero, Equation 14.94 (p. 714) can be written as:

(14.95)

The top part of Equation 14.95 (p. 714) may be solved for {uc}:
(14.96)

The actual numerical solution process is not as indicated here but is done more efficiently using one
of the various equation solvers discussed in Equation Solvers (p. 717).

14.6.1. Reaction Forces

The reaction vector , may be developed for linear models from the bottom part of Equa-
tion 14.95 (p. 714):
(14.97)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
714 Published: 2023-10-23
 Solving for Unknowns and Reactions

= reaction forces (output using either OUTPR,RSOL or PRRSOL command)

Alternatively, the nodal reaction load vector may be considered over all DOFs by combining Equa-
tion 14.92 (p. 713) and Equation 14.93 (p. 714) to get:
(14.98)

where only the loads at imposed DOF are output. Where applicable, the transient/dynamic effects
are added:
(14.99)

where:

[M] = total mass matrix

[C] = total damping or conductivity matrix
, = defined below

The element static nodal loads are:

(14.100)

where:

= element nodal loads (output using OUTPR,NLOAD, or PRESOL commands)

e = subscript for element matrices and load vectors

The element damping and inertial loads are:

(14.101)
(14.102)

where:

= element damping nodal load (output using OUTPR,NLOAD, or PRESOL commands)

= element inertial nodal load (output using OUTPR,NLOAD, or PRESOL commands)

Thus,

(14.103)

The derivatives of the nodal DOF with respect to time are:

= first derivative of the nodal DOF with respect to time, for example, velocity
= second derivative of the nodal DOF with respect to time, for example, acceleration

Transient Analysis (p. 796) and Harmonic Analysis (p. 813) discuss the transient and harmonic damping
and inertia loads.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 715
Analysis Tools

If an imposed DOF value is part of a constraint equation, the nodal reaction load vector is further
modified using the appropriate terms of the right hand side of Equation 14.208 (p. 743); that is, the
forces on the non-unique DOFs are summed into the unique DOF (the one with the imposed DOF
value) to give the total reaction force acting on that DOF.

14.6.2. Disequilibrium
The following circumstances could cause a disequilibrium, usually a moment disequilibrium:

Program Option Explanation of Possible Difficulty

non-planar, 4-node If the 4 nodes do not lie in a flat plane moment
membrane shell equilibrium may not be preserved, as no internal
elements corrections are done. However, the program requires
SHELL181 with such elements to be input very close to flat.
KEYOPT(1) = 1
nodal coupling The user can write any form of relationship between
constraint equations the displacements, and these may include fictitious
(CP, CE commands) forces or moments. Thus, the reaction forces printout
can be used to detect input errors.
MATRIX27 The user has the option to input almost any type of
User generated super- erroneous input, so that such input should be checked
element matrix carefully. For example, all terms representing UX
degrees of freedom of one UX row of the matrix should
sum to zero to preserve equilibrium.
COMBIN37 Noncoincident nodes can cause a moment
FLUID38 disequilibrium. (This is usually not a problem if one of
the nodes is attached to a non-rotating ground).
COMBIN39
COMBIN40
COMBIN14 (with Elements with one node having a different nodal
KEYOPT(2) > 0) coordinate system from the other are inconsistent.
MATRIX27
COMBIN37
FLUID38
COMBIN39
COMBIN40

The following circumstances could cause an apparent disequilibrium:

• All nodal coordinate systems are not parallel to the global Cartesian coordinate system. However,
if all nodal forces are rotated to the global Cartesian coordinate system, equilibrium should be seen
to be satisfied.

• The solution is not converged. This applies to the potential discrepancy between applied and in-
ternal element forces in a nonlinear analysis.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
716 Published: 2023-10-23
 Equation Solvers

• The mesh is too coarse. This may manifest itself for elements where there is an element force
printout at the nodes, such as SHELL61 (axisymmetric-harmonic structural shell).

• The "TOTAL" of the moments (MX, MY, MZ) given with the reaction forces does not necessarily
represent equilibrium. It only represents the sum of all applicable moments. Moment equilibrium
would also need the effects of forces taken about an arbitrary point.

• Axisymmetric models are used with forces or pressures with a radial component. These loads will
often be partially equilibrated by hoop stresses, which do not show up in the reaction forces.

• Shell elements have an elastic foundation described. The load carried by the elastic foundation is
not seen in the reaction forces.

• In substructure expansion pass with the resolve method used, the reaction forces at the master
degree of freedom are different from that given by the backsubstitution method (see Substructuring
Analysis (p. 828)).

14.7. Equation Solvers

The system of simultaneous linear equations generated by the finite element procedure is solved either
using a direct elimination process or an iterative method. A direct elimination process is primarily a
Gaussian elimination approach which involves solving for the unknown vector of variables {u} in Equa-
tion 14.104 (p. 717):
(14.104)

where:

[K] = global stiffness/conductivity matrix

{u} = global vector of nodal unknown
{F} = global applied load vector

The direct elimination process involves decomposition (factorization) of the matrix [K] into lower and
upper triangular matrices, [K] = [L][U]. Then forward and back substitutions using [L] and [U] are made
to compute the solution vector {u}.

A typical iterative method involves an initial guess, {u}1, of the solution vector {u} and then a successive
steps of iteration leading to a sequence of vectors {u}2, {u}3, . . . such that, in the limit, {u}n = {u} as n
tends to infinity. The calculation of {u}n + 1 involves [K], {F}, and the {u} vectors from one or two of the
previous iterations. Typically the solution converges to within a specified tolerance after a finite number
of iterations.

In the following sections, all of the solvers are described under two major subsections: Direct Solvers
and Iterative Solvers (all accessed with EQSLV).

14.7.1. Direct Solvers

The direct solver that is available is the Sparse Direct Solver (accessed with EQSLV,SPARSE). The Sparse
Direct Solver makes use of the fact that the finite element matrices are normally sparsely populated.
This sparsity allows the system of simultaneous equations to be solved efficiently by minimizing the
operation counts.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 717
Analysis Tools

14.7.2. Sparse Direct Solver

As described in the introductory section, the linear matrix equation, (Equation 14.104 (p. 717)) is solved
by triangular decomposition of matrix [K] to yield the following equation:
(14.105)

where:

[L] = lower triangular matrix

[U] = upper triangular matrix

By substituting:
(14.106)

we can obtain {u} by first solving the triangular matrix system for {w} by using the forward pass oper-
ation given by:
(14.107)

and then computing {u} using the back substitution operation on a triangular matrix given by:
(14.108)

When [K] is symmetric, the above procedure could use the substitution:
(14.109)

However, it is modified as:

(14.110)

where:

[D] = a diagonal matrix

The diagonal terms of [D] may be negative in the case of some nonlinear finite element analysis. This
allows the generation of [L'] without the consideration of a square root of negative number. Therefore,
Equation 14.105 (p. 718) through Equation 14.108 (p. 718) become:
(14.111)
(14.112)
(14.113)

and
(14.114)

Since [K] is normally sparsely populated with coefficients dominantly located around the main diag-
onal, the Sparse Direct Solver is designed to handle only the nonzero entries in [K]. In general, during
the Cholesky decomposition of [K] shown in Equation 14.105 (p. 718) or Equation 14.111 (p. 718),
nonzero coefficients appear in [L] or [L'] at coefficient locations where [K] matrix had zero entries.
The Sparse Direct Solver algorithm minimizes this fill-in by judiciously reordering the equation numbers
in [K].

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
718 Published: 2023-10-23
 Equation Solvers

The performance of a direct solution method is greatly optimized through the equations reordering
procedure which involves relabeling of the variables in the vector {u}. This simply amounts to permuting
the rows and columns of [K] and the rows of {F} with the objective of minimizing fill-in. So, when the
decomposition step in Equation 14.105 (p. 718) or Equation 14.111 (p. 718) is performed on the reordered
[K] matrix, the fill-in that occurs in [L] or [L'] matrix is kept to a minimum. This enormously contributes
to optimizing the performance of the Sparse Direct Solver.

To achieve minimum fill-in, different matrix coefficient reordering algorithms are available in the lit-
erature (George and Liu([302] (p. 942))). The Sparse Direct Solver uses two different reordering schemes.
They are the Minimum Degree ordering and the METIS ordering. The choice of which reordering
method to use is automated in the solver algorithm in order to yield the least fill-in.

If the stiffness matrix is very ill-conditioned or close to being singular, the Sparse Direct Solver normally
does not break down or stop solving. In this case, some matrix factorized pivots will be close to zero
and the solution may be inaccurate. Also, the resulting solution becomes sensitive to changes in the
product version, the operating system, and/or the number of CPU cores.

14.7.3. Iterative Solver

Mechanical APDL offers a large number of iterative solvers as alternatives to the direct solvers (sparse
solver). These alternatives in many cases can result in less I/O or disk usage, less total elapsed time,
and more scalable parallel performance. However, in general, iterative solvers are not as robust as
the direct solvers. For numerical challenges such as a nearly-singular matrix (matrix with small pivots)
or a matrix that includes Lagrangian multipliers, the direct solver is an effective solution tool, while
an iterative solver is less effective or may even fail.

The first three iterative solvers are based on the conjugate gradient (CG) method. The first of these
three CG solvers is the Jacobi Conjugate Gradient (JCG) solver (Mahinthakumar and Hoole
([145] (p. 933))) (accessed with EQSLV,JCG) which is suitable for well-conditioned problems. Well-
conditioned problems often arise from heat transfer, acoustics, magnetics and solid 2D / 3D structural
analyses. The JCG solver is available for real and complex, symmetric and unsymmetric matrices. The
second solver is the Preconditioned Conjugate Gradient (PCG) solver (accessed with EQSLV,PCG)
which is efficient and reliable for all types of analyses including the ill-conditioned beam/shell struc-
tural analysis. The PCG solver is made available through a license from Computational Applications
and System Integration, Inc. of Champaign, Illinois (USA). The PCG solver is valid for real symmetric
and unsymmetric matrices. The third solver is the Incomplete Cholesky Conjugate Gradient (ICCG)
solver (internally developed, unpublished work) (accessed with EQSLV,ICCG). The ICCG solver is more
robust than the JCG solver for handling ill-conditioned matrices. The ICCG solver is available for real
and complex, symmetric and unsymmetric matrices.

The typical system of equations to be solved iteratively is given as :

(14.115)

where:

[K] = global coefficient matrix

{u} = unknown vector
{F} = global load vector

In the CG method, the solution is found as a series of vectors {pi}:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 719
Analysis Tools

(14.116)

where m is no larger than the matrix size n. The scheme is guaranteed to converge in n or fewer iter-
ations on an infinite precision machine. However, since the scheme is implemented on a machine
with finite precision, it sometimes requires more than n iterations to converge. The solvers allow up
to a maximum of 2n iterations. If it still does not converge after the 2n iterations, the solution will be
abandoned with an error message. The unconverged situation is often due to an inadequate number
of boundary constraints being used (rigid body motion). The rate of convergence of the CG algorithm
is inversely proportional to the square root of the conditioning number of [K] where the condition
number of [K] is equal to the ratio of the maximum eigenvalue of [K] to the minimum eigenvalue of
[K] . A preconditioning procedure is used to reduce the condition number of linear Equa-
tion 14.115 (p. 719). In the JCG algorithm, the diagonal terms of [K] are used as the preconditioner [Q],
while in the ICCG and PCG algorithms, a more sophisticated preconditioner [Q] is used. The CG al-
gorithm with preconditioning is shown collectively as Equation 14.117 (p. 720).

(14.117)

Convergence is achieved when:

(14.118)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
720 Published: 2023-10-23
 Mode-Superposition Method

where:

ε = user supplied tolerance (TOLER on the EQSLV command; output as SPECIFIED TOLERANCE)
{Ri} = {F} - [K] {ui}
{ui} = solution vector at iteration i

also, for the JCG and ICCG solvers:

(14.119)
(14.120)

It is assumed that the initial starting vector {u0} is a zero vector.

If the stiffness matrix is very ill-conditioned or close to being singular, any iterative solver would
normally break down and fail to converge. In this case, the iterative solver will likely iterate for a long
time before it finally fails to converge.

14.8. Mode-Superposition Method

Mode-superposition method is a method of using the natural frequencies and mode shapes from the
modal analysis (ANTYPE,MODAL) to characterize the dynamic response of a structure to transient
(ANTYPE,TRANS with TRNOPT,MSUP, Transient Analysis (p. 796)), or steady harmonic (ANTYPE,HARM
with HROPT,MSUP, Harmonic Analysis (p. 813)) excitations.

The following topics are available:

14.8.1. General Equations
14.8.2. Equations for QR Damped Eigensolver Based Analysis
14.8.3. Equations for Unsymmetric Eigensolver Based Analysis
14.8.4. Modal Damping
14.8.5. Residual Vector Method
14.8.6. Residual Response Method

14.8.1. General Equations

The equations of motion may be expressed as in Equation 15.5 (p. 796):
(14.121)

is the time-varying load vector, given by

(14.122)

where:

= time varying nodal forces

= load vector scale factor (input on LVSCALE command)
= load vector from the modal analysis (see below)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 721
Analysis Tools

The load vector is computed when doing a modal analysis and its generation is the same as for
a substructure load vector, described in Substructuring Analysis (p. 828).

The following development is similar to that given by Bathe ([2] (p. 925)):

Define a set of modal coordinates such that

(14.123)

where:

= the ith mode shape

= the number of modes to be used (input as MAXMODE on TRNOPT or HROPT commands)

Note that Equation 14.123 (p. 722) hinders the use of nonzero displacement input, since defining
in terms of is not straight forward. The inverse relationship does exist (Equation 14.123 (p. 722))
for the case where all the displacements are known, but not when only some are known. Substituting
Equation 14.123 (p. 722) into Equation 14.121 (p. 721),

(14.124)

Premultiply by a typical mode shape :

(14.125)

The orthogonal condition of the natural modes states that

(14.126)

(14.127)

In the mode-superposition method using the Lanczos and other extraction methods, only Rayleigh
or constant damping is allowed so that:
(14.128)

Applying these conditions to Equation 14.125 (p. 722), only the i = j terms remain:
(14.129)

The coefficients of , , and , are derived as follows:

1. Coefficient of :

By the normality condition (Equation 15.54 (p. 812)),

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
722 Published: 2023-10-23
 Mode-Superposition Method

(14.130)

2. Coefficient of :

The damping term is based on treating the modal coordinate as a single DOF system (shown in
Equation 14.121 (p. 721)) for which:
(14.131)

and
(14.132)

Figure 14.6: Single Degree of Freedom Oscillator

Equation 14.132 (p. 723) can give a definition of :

(14.133)

From (Tse [69] (p. 928)),

(14.134)

where:

= fraction of critical damping for mode j

and,
(14.135)

where:

= natural circular frequency of mode j

Combining Equation 14.133 (p. 723) thru Equation 14.130 (p. 723) with Equation 14.131 (p. 723),

(14.136)

3. Coefficient of yj:

From Equation 15.51 (p. 811),

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 723
Analysis Tools

(14.137)

Premultiply by ,

(14.138)

Substituting Equation 14.130 (p. 723) for the mass term,

(14.139)

For convenient notation, let

(14.140)

represent the right-hand side of Equation 14.129 (p. 722). Substituting Equation 14.130 (p. 723),
Equation 14.136 (p. 723), Equation 14.139 (p. 724) and Equation 14.140 (p. 724) into Equa-
tion 14.129 (p. 722), the equation of motion of the modal coordinates is obtained:
(14.141)

Since j represents any mode, Equation 14.141 (p. 724) represents n uncoupled equations in the n
unknowns . The advantage of the uncoupled system (ANTYPE,TRAN with TRNOPT,MSUP) is
that all the computationally expensive matrix algebra has been done in the eigensolver, and long
transients may be analyzed inexpensively in modal coordinates with Equation 14.123 (p. 722).

The are converted back into geometric displacements (the system response to the loading)
by using Equation 14.123 (p. 722). That is, the individual modal responses are superimposed to
obtain the actual response, and hence the name "mode-superposition".

14.8.2. Equations for QR Damped Eigensolver Based Analysis

For the QR damped mode extraction method, the general equations apply except the differential
equations of motion in modal coordinate are deduced from Equation 14.228 (p. 753) with the right
hand side force vector of Equation 14.125 (p. 722). They are written as:
(14.142)

where:

= real eigenvector matrix normalized with respect to mass coming from the LANCZOS
run of QRDAMP (see QR Damped Method (p. 752) for more details).
= diagonal matrix containing the eigenvalues .

= unsymmetric part of the stiffness matrix.

It can be seen that if is arbitrary and/or is unsymmetric, the modal matrices are full so that
the modal equations are coupled.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
724 Published: 2023-10-23
 Mode-Superposition Method

14.8.3. Equations for Unsymmetric Eigensolver Based Analysis

When using the unsymmetric eigensolver, the matrices are unsymmetric. Both left and right normalized
eigenmodes are used to decouple the modal equations as follows.

Equation 14.129 (p. 722) becomes:

(14.143)

where:

= the jth left mode shape

= the jth right mode shape (also called the jth mode shape). For more information, see Unsym-
metric Method (p. 748).

This equations leads to the modal coordinate equation (Equation 14.141 (p. 724)) when Rayleigh
damping is considered.

14.8.4. Modal Damping

The modal damping, , is the combination of several Mechanical APDL damping inputs, as described
in Equation 14.41 (p. 699).

14.8.5. Residual Vector Method

In modal superposition analysis, the dynamic response will be approximate when the applied loading
excites the higher frequency modes of a structure. To improve the accuracy of the dynamic response,
the residual vector method employs additional modal transformation vectors (designated as residual
vectors) in addition to the eigenvectors in the modal transformation (Equation 14.123 (p. 722)).

The residual vector method (RESVEC,ON) uses extra residual vectors computed in the modal analysis
(ANTYPE,MODAL) to characterize the high frequency response of a structure to dynamic loading. It
applies to the following mode-superposition analyses:

• Transient (ANTYPE,TRANS with TRNOPT,MSUP)

• Harmonic (ANTYPE,HARM with HROPT,MSUP)

• PSD (ANTYPE,SPEC with SPOPT,PSD)

• Multiple and single point response spectrum (ANTYPE,SPEC with SPOPT,SPRS or MPRS).

Because of the improved convergence properties of this method, fewer eigenmodes are required
from the eigensolution.

The dynamic response of the structure can be divided into two terms:
(14.144)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 725
Analysis Tools

= lower mode contributions (Equation 14.123 (p. 722))

= higher mode contributions, which can be expressed as the combination of residual
vectors.

First, the flexibility matrix can be expressed as:

(14.145)

where:

= generalized inverse matrix of stiffness matrix [K] (see Geradin and Rixen [365] (p. 945))

= elastic normal modes

= total degree of freedom of the system

The residual flexibility matrix is given by:

(14.146)

Define residual vectors as:

(14.147)

where:

= matrix of force vectors

Orthogonalize the residual vectors with respect to the retained elastic normal modes gives orthogon-
alized residual vectors , that is, pseudo-modes with associated frequencies that are orthogonal
to the mass and stiffness matrices (see Dickens et al. [420] (p. 949)).

Then the basis vectors for modal subspace are formed by:
(14.148)

This method also applies to the component mode synthesis (CMS) generation pass (ANTYPE,SUBSTR
with CMSOPT) when a load is applied to internal degrees of freedom (all DOFs that are not master
DOFs) of the substructure. The basis of normal modes used in the three available CMS methods (p. 831)
is updated following Equation 14.148 (p. 726) to include the resulting residual vector.

14.8.6. Residual Response Method

In mode-superposition analysis, the dynamic response is approximate when the applied loading excites
the higher frequency modes of a structure. To improve the accuracy of the dynamic response, the
residual response method can be used. It is similar to the residual vector method (p. 725) except that
it directly uses the residual static responses calculated with residual flexibility.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
726 Published: 2023-10-23
 Extraction of Modal Damping Parameter for Squeeze Film Problems

The residual response method (RESVEC,,,,,ON) applies to the following mode-superposition analyses:

• Transient (ANTYPE,TRANS with TRNOPT,MSUP)

• Harmonic (ANTYPE,HARM with HROPT,MSUP)

When the equations are symmetric, the residual responses are defined by Equation 14.147 (p. 726).

When the equations are unsymmetric (see Equations for Unsymmetric Eigensolver Based Analys-
is (p. 725)), Equation 14.145 (p. 726) becomes:

(14.149)

The residual flexibility matrix is given by:

(14.150)

The residual responses are calculated with Equation 14.147 (p. 726) with the residual flexibility defined
by Equation 14.150 (p. 727).

14.9. Extraction of Modal Damping Parameter for Squeeze Film Problems

A damping ratio is often applied for harmonic analysis. In practice this approach only leads to satisfying
results if all frequency steps can be represented by the same damping ratio or the operating range
encloses just one eigenmode. Difficulties arise if the damping ratio depends strongly on the excitation
frequency as happens in case of viscous damping in gaseous environment.

A typical damping ratio verse frequency function is shown below. For this example, the damping ratio
is almost constant below the cut-off frequency. Harmonic oscillations at frequencies below cutoff are
strongly damped. Above cut-off the damping ratio decreases. Close to the structural eigenfrequency
the damping ratio dropped down to about 0.25 and a clear resonance peak can be observed.

Figure 14.7: Damping and Amplitude Ratio vs. Frequency

Damping and stiffness coefficients in modal coordinates are defined based on their nodal coordinate
values as:
(14.151)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 727
Analysis Tools

and
(14.152)

where:

[C] = damping coefficient in modal coordinates

{ϕi} = eigenvector i (in modal coordinates)

[C*] = finite element damping matrix in modal coordinates
[K*] = finite element stiffness matrix in nodal coordinates

Unfortunately, both matrices [C*] and [K*] are not directly available for the fluid part of the coupled
domain problem (e.g., squeeze film elements FLUID136). Moreover eigenvectors are derived from the
structural part of the coupled domain problem and consequently neither the modal damping matrix
nor the modal stiffness matrix of the fluidic system are necessarily orthogonal. Essential off-diagonal
elements occur in case of asymmetric film arrangements or asymmetric plate motion as shown below.

Figure 14.8: Fluid Pressure From Modal Excitation Distribution

The goal is to express the viscous damping in modal coordinates as follows:

(14.153)

where:

{F} = modal force vector

{q} = vector of modal displacement amplitudes

[C] = unknown modal damping matrix

[K] = unknown modal squeeze stiffness matrix

The following algorithm is necessary to compute all coefficients of the modal damping and stiffness
matrix:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
728 Published: 2023-10-23
 Extraction of Modal Damping Parameter for Squeeze Film Problems

1. Start with the first mode and excite the fluid elements by wall velocities which correspond to
a unit modal velocity. In fact the nodal velocities become equal to the eigenvector of the ap-
propriate mode.

2. Compute the real and imaginary part of the pressure distribution in a harmonic analyses.

3. Compute modal forces with regard to all other modes. The ith modal force states how much
the pressure distribution of the first mode really acts on the ith mode.

4. The computed modal forces can be used to extract all damping and squeeze stiffness coefficients
of the first column in the [C] and [K] matrices.

5. Repeat step 1 with the next eigenvector and compute the next column of [C] and [K].

The theoretical background is given by the following equations. Each coefficient Cji and Kji is defined
by:
(14.154)

and
(14.155)

where:

F(qi) = complex nodal damping force vector caused by a unit modal velocity of the source
mode i.

Note that the modal forces are complex numbers with a real and imaginary part. The real part, Re,
represents the damping force and the imaginary part, Im, the squeeze force, which is cause by the fluid
compression. The damping and squeeze coefficients are output by DMPEXT and are given by:

(14.156)

and

(14.157)

Assuming the structure is excited by a unit modal velocity we obtain:

(14.158)

and
(14.159)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 729
Analysis Tools

Ω = excitation frequency (input on DMPEXT command)

Modal damping ratios ξ or the squeeze stiffness to structural stiffness ratio KRatio are defined only for
the main diagonal elements and are also output with the DMPEXT command. These numbers are
computed by:
(14.160)

and

(14.161)

where:

mi = modal mass and the eigenfrequency ωi

The damping ratio is necessary to compute α and β (input as ALPHAD and BETAD commands) para-
meters for Rayleigh damping models or to specify constant or modal damping (input by DMPRAT or
MDAMP commands).

The squeeze to stiffness ratio specifies how much the structural stiffness is affected by the squeeze film.
It can not directly be applied to structural elements but is helpful for user defined reduced order
models.

14.10. Newton-Raphson Procedure

The following Newton-Raphson procedure topics are available:
14.10.1. Overview
14.10.2. Convergence
14.10.3. Predictor
14.10.4. Adaptive Descent
14.10.5. Line Search
14.10.6. Arc-Length Method

14.10.1. Overview
The finite element discretization process yields a set of simultaneous equations:
(14.162)

where:

[K] = coefficient matrix

{u} = vector of unknown DOF (degree of freedom) values
{Fa} = vector of applied loads

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
730 Published: 2023-10-23
 Newton-Raphson Procedure

If the coefficient matrix [K] is itself a function of the unknown DOF values (or their derivatives) then
Equation 14.162 (p. 730) is a nonlinear equation. The Newton-Raphson method is an iterative process
of solving the nonlinear equations and can be written as (Bathe [2] (p. 925)):
(14.163)
(14.164)

where:

= Jacobian matrix (tangent matrix)

i = subscript representing the current equilibrium iteration
= vector of restoring loads corresponding to the element internal loads

Both and are evaluated based on the values given by {ui}. The right-hand side of Equa-
tion 14.163 (p. 731) is the residual or out-of-balance load vector; that is, the amount the system is out
of equilibrium. A single solution iteration is depicted graphically in Figure 14.9: Newton-Raphson
Solution - One Iteration (p. 732) for a one DOF model. In a structural analysis, is the tangent
stiffness matrix, {ui} is the displacement vector and is the restoring force vector calculated from
the element stresses. In a thermal analysis, is the conductivity matrix, {ui} is the temperature
vector and is the resisting load vector calculated from the element heat flows. In an electro-
magnetic analysis, is the Dirichlet matrix, {ui} is the magnetic potential vector, and is the
resisting load vector calculated from element magnetic fluxes. In a transient analysis, is the ef-
fective coefficient matrix and is the effective applied load vector which includes the inertia
and damping effects.

As seen in the following figures, more than one Newton-Raphson iteration is needed to obtain a
converged solution. The general algorithm proceeds as follows:

1. Assume {u0}. {u0} is usually the converged solution from the previous time step. On the first time
step, {u0} = {0}.

2. Compute the updated tangent matrix and the restoring load from configuration {ui}.

3. Calculate {Δui} from Equation 14.163 (p. 731).

4. Add {Δui} to {ui} in order to obtain the next approximation {ui + 1} (Equation 14.164 (p. 731)).

5. Repeat steps 2 to 4 until convergence is obtained.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 731
Analysis Tools

Figure 14.9: Newton-Raphson Solution - One Iteration

Figure 14.10: Newton-Raphson Solution - Next Iteration (p. 732) shows the solution of the next iteration
(i + 1) of the example from Figure 14.9: Newton-Raphson Solution - One Iteration (p. 732). The sub-
sequent iterations would proceed in a similar manner.

The solution obtained at the end of the iteration process would correspond to load level {Fa}. The final
converged solution would be in equilibrium, such that the restoring load vector (computed
from the current stress state, heat flows, etc.) would equal the applied load vector {Fa} (or at least to
within some tolerance). None of the intermediate solutions would be in equilibrium.

Figure 14.10: Newton-Raphson Solution - Next Iteration

If the analysis included path-dependent nonlinearities (such as plasticity), then the solution process
requires that some intermediate steps be in equilibrium in order to correctly follow the load path.
This is accomplished effectively by specifying a step-by-step incremental analysis; that is, the final
load vector {Fa} is reached by applying the load in increments and performing the Newton-Raphson
iterations at each step:
(14.165)

where:

[Kn,i] = tangent matrix for time step n, iteration i

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
732 Published: 2023-10-23
 Newton-Raphson Procedure

= total applied force vector at time step n

= restoring force vector for time step n, iteration i

This process is the incremental Newton-Raphson procedure and is shown in Figure 14.11: Incremental
Newton-Raphson Procedure (p. 733). The Newton-Raphson procedure guarantees convergence if and
only if the solution at any iteration {ui} is "near" the exact solution. Therefore, even without a path-
dependent nonlinearity, the incremental approach (that is, applying the loads in increments) is
sometimes required in order to obtain a solution corresponding to the final load level.

Figure 14.11: Incremental Newton-Raphson Procedure

When the stiffness matrix is updated every iteration (as indicated in Equation 14.163 (p. 731) and
Equation 14.165 (p. 732)) the process is termed a full Newton-Raphson solution procedure (NROPT,FULL
or NROPT,UNSYM). Alternatively, the stiffness matrix could be updated less frequently using the
modified Newton-Raphson procedure (NROPT,MODI). Specifically, for static or transient analyses, it
would be updated only during the first or second iteration of each substep, respectively. Use of the
initial-stiffness procedure (NROPT,INIT) prevents any updating of the stiffness matrix, as shown in
Figure 14.12: Initial-Stiffness Newton-Raphson (p. 733). If a multistatus element is in the model, however,
it would be updated at iteration in which it changes status, irrespective of the Newton-Raphson option.
The modified and initial-stiffness Newton-Raphson procedures converge more slowly than the full
Newton-Raphson procedure, but they require fewer matrix reformulations and inversions. A few ele-
ments form an approximate tangent matrix so that the convergence characteristics are somewhat
different.

Figure 14.12: Initial-Stiffness Newton-Raphson

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 733
Analysis Tools

14.10.2. Convergence
The iteration process described in the previous section continues until convergence is achieved. The
maximum number of allowed equilibrium iterations (input on NEQIT command) are performed in
order to obtain convergence.

Convergence is assumed when

(14.166)

and/or
(14.167)

where {R} is the residual vector:

(14.168)

which is the right-hand side of the Newton-Raphson Equation 14.163 (p. 731). {Δui} is the DOF increment
vector, εR and εu are tolerances (CNVTOL,,,TOLER) and Rref and uref are reference values (CN-
VTOL,,VALUE). || || is a vector norm; that is, a scalar measure of the magnitude of the vector (defined
below).

Convergence, therefore, is obtained when the size of the residual (disequilibrium) is less than a toler-
ance times a reference value and/or when the size of the DOF increment is less than a tolerance times
a reference value. The default is to use out-of-balance convergence checking only.

An alternative convergence criterion is based on the incremental energy calculation:

(14.169)

where is the incremental system energy during a Newton interation :

(14.170)

It can be interpreted as the work done by an out-of-balance force on the incremental displacement.
is the tolerance specified by CNVTOL,,,TOLER, and is the reference value specified by CN-
VTOL,,VALUE.

There are three available norms (CNVTOL,,,,NORM) to choose from:

1. Infinite norm

2. L1 norm

3. L2 norm

In the above equation, R can be replaced with different convergence labels, such as Δu for DOF in-
crement convergence. The infinite norm is simply the maximum value in the vector (maximum residual
or maximum DOF increment), the L1 norm is the sum of the absolute value of the terms, and the L2
norm is the square root of the sum of the squares (SRSS) value of the terms, also called the Euclidean
norm. The default is to use the L2 norm.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
734 Published: 2023-10-23
 Newton-Raphson Procedure

The default out-of-balance reference value Rref is ||{Fa}||. For DOFs with imposed displacement con-
straints, the default Rref value is {Fnr} at those DOFs.

If the reference value Rref falls below a minimum reference value specified by CNVTOL,,,,,MINREF,
then Rref is replaced by the minimum reference value. The program also internally calculates a reference
value and uses the maximum of this internally calculated value and the MINREF default. (See CNVTOL
for a list of default minimum reference values for the various convergence criteria.) The internally
calculated minimum reference value depends upon the first iteration residual value and contact forces.
Furthermore, this minimum reference also depends on the displacement convergence value and Rref.

Minimum reference is typically useful for problems with almost no external loading. If the program-
calculated minimum reference is still too low, you should manually specify a minimum reference that
is appropriate for the problem under consideration. Examples where it may be useful to manually
specify the minimum reference include, but are not limited to, problems involving contact initial
penetration resolution, free thermal expansions, and rigid-body motion analyses with stress-free rota-
tion. When the program-calculated internal reference is used, the program may sometimes tighten
the displacement convergence criterion to improve solution accuracy.

Sometimes, the program-calculated minimum reference might be too high. This can lead to premature
convergence, and accuracy of the solution may be affected. This would typically be the case for
problems where very low levels of strain are expected, and bodies are not deforming much. For such
cases, it is recommended you manually define a tight minimum reference (for example, 0.01 or 0.001)
to achieve accurate results.

The default reference value uref is calculated using the historic infinity norm of incremental displace-
ment.

If you do not specify the reference value for the energy norm, , the program calculates it as:
(14.171)

where is the incremental energy from the 1st Newton-Raphson iteration in substep , and || ||
norm represents the absolute value of .

If the residual ({Fa} - {Fnr}) in Equation 14.171 (p. 735) is too small, the program uses the applied force
({Fa}) or the internal force ({Fnr}) to calculate the reference.

In theory, the energy convergence check is equivalent to the combination of the out-of-balance force
and the displacement check together. However, by default the energy convergence check is more
relaxed than this combination. In practice, if only the energy convergence check is invoked with the
default tolerance (the out-of-balance force and displacement convergence are off ), the solution could
converge in a smaller number of cumulative iterations but may not be as accurate as a solution that
includes out-of-balance force and displacement convergence. The energy convergence check provides
an option to achieve faster runtime along with a meaningful solution.

The energy convergence check is more suited to single-field structural applications (UX, UY, UZ, ROTX,
ROTY, ROTZ) or multiphysics applications with structural degrees of freedom because the energy
calculated in the check is the real physical energy of the system. This may not be the case for other
multiphysics applications.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 735
Analysis Tools

14.10.3. Predictor
In static analyses, the solution used for the start of each time step n {un,0} is usually equal to the
current DOF solution {un-1}. The tangent matrix [Kn,0] and restoring load {Fn,0} are based on this con-
figuration. The predictor option (PRED command) extrapolates the DOF solution using the previous
history in order to take a better guess at the next solution. The prediction is based on the displacement
increments accumulated over the previous time step, factored by the time-step size:
(14.172)

where:

{Δun-1} = displacement increment accumulated over the previous time step

n = current time step
n-1 = previous time step

(14.173)

and β is defined as:

(14.174)

where:

= current time-step size

= previous time-step size

β is not allowed to be greater than 5.

Equation 14.172 (p. 736) is called the formulation of the linear predictor. It uses two known solution
values, and , to extrapolate the solution of the next substep in a static analysis.

The linear predictor normally calculates a good prediction for a solution that is smooth, such as a
translatonal displacement (UX, UY, UZ) in a mildly nonlinear problem.

The solution can also be predicted by quadratic extrapolation as described by the formulation shown
in Equation 14.175 (p. 736):
(14.175)

where:

= displacement increment accumulated over the time step prior to the previous
time step.
c1, c2 are scalable parameters determined by the current, previous, and prior to previous
time steps.
The quadradic predictor in Equation 14.175 (p. 736) is normally a better predictor than the linear pre-
dictor for solutions with rotational degrees of freedom (ROTX, ROTY, ROTZ).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
736 Published: 2023-10-23
 Newton-Raphson Procedure

In transient analyses, the prediction is based on the current velocities and accelerations and the time
integration scheme used (Newmark, HHT, Backward Eulter). For example, using the Newmark formulas
for structural DOFs:
(14.176)

where:

= current displacements, velocities and accelerations

Δtn = current time-step size
α = Newmark parameter (input on TINTP command)

For thermal, magnetic and other first order systems, the prediction is based on the trapezoidal formula:
(14.177)

where:

{un - 1} = current temperatures (or magnetic potentials)

= current rates of these quantities
θ = trapezoidal time integration parameter (input on TINTP command)

See Transient Analysis (p. 796) for more details on the transient procedures.

The subsequent equilibrium iterations provide DOF increments {Δu} with respect to the predicted
DOF value {un,0}, hence this is a predictor-corrector algorithm.

14.10.4. Adaptive Descent

Adaptive descent (Adptky on the NROPT command) is a technique which switches to a "stiffer" matrix
if convergence difficulties are encountered, and switches back to the full tangent as the solution
convergences, resulting in the desired rapid convergence rate (Eggert [153] (p. 933)).

The matrix used in the Newton-Raphson equation (Equation 14.163 (p. 731)) is defined as the sum of
two matrices:
(14.178)

where:

[KS] = secant (or most stable) matrix

[KT] = tangent matrix
ξ = descent parameter

The program adaptively adjusts the descent parameter (ξ) during the equilibrium iterations as follows:

1. Start each substep using the tangent matrix (ξ = 0).

2. Monitor the change in the residual ||{R}||2 over the equilibrium iterations:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 737
Analysis Tools

If it increases (indicating possible divergence):

• remove the current solution if ξ < 1, reset ξ to 1 and redo the iteration using the secant matrix

• if already at ξ = 1, continue iterating

If it decreases (indicating converging solution):

• If ξ = 1 (secant matrix) and the residual has decreased for three iterations in a row (or 2 if ξ was
increased to 1 during the equilibrium iteration process by (a.) above), then reduce ξ by a factor
of 1/4 (set it to 0.25) and continue iterating.

• If the ξ < 1, decrease it again by a factor of 1/4 and continue iterating. Once ξ is below 0.0156,
set it to 0.0 (use the tangent matrix).

3. If a negative pivot message is encountered (indicating an ill-conditioned matrix):

• If ξ < 1, remove the current solution, reset ξ = 1 and redo the iteration using the secant matrix.

• If ξ = 1, bisect the time step if automatic time stepping is active, otherwise terminate the exe-
cution.

The nonlinearities which make use of adaptive descent (that is, they form a secant matrix if ξ > 0)
include: plasticity, contact, stress stiffness with large strain, and nonlinear magnetics using the scalar
potential formulation. Adaptive descent is used by default in these cases unless the line search or
arc-length options are on. It is only available with full Newton-Raphson, where the matrix is updated
every iteration. Full Newton-Raphson is also the default for plasticity, contact and large strain nonlin-
earities.

14.10.5. Line Search

The line search option (accessed with LNSRCH command) attempts to improve a Newton-Raphson
solution {Δui} by scaling the solution vector by a scalar value termed the line search parameter.

Consider Equation 14.164 (p. 731) again:

(14.179)

In some solution situations, the use of the full {Δui} leads to solution instabilities. Hence, if the line
search option is used, Equation 14.179 (p. 738) is modified to be:
(14.180)

where:

s = line search parameter, 0.05 < s < 1.0

s is automatically determined by minimizing the energy of the system, which reduces to finding the
zero of the nonlinear equation:
(14.181)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
738 Published: 2023-10-23
 Newton-Raphson Procedure

gs = gradient of the potential energy with respect to s

An iterative solution scheme based on regula falsi is used to solve Equation 14.181 (p. 738) (Schweizerhof
and Wriggers [154] (p. 933)). Iterations are continued until either:

1. gs is less than LSTOL * go, where go is the value of Equation 14.181 (p. 738) at s = 0.0 (that is, using
for {Fnr (s{Δu})}). LSTOL defaults to 0.5 (see LNSRCH).

2. gs is not changing significantly between iterations.

3. Six iterations have been performed.

If go > 0.0, no iterations are performed and s is set to 1.0. s is not allowed below 0.05.

The scaled solution {Δui} is used to update the current DOF values {ui+1} in Equation 14.164 (p. 731)
and the next equilibrium iteration is performed.

The line search algorithm is very susceptible to round-off errors (Crisfield [441] (p. 950)). Truncation
of gs can help solution repeatability (see LStrun on LNSRCH).

14.10.6. Arc-Length Method

The arc-length method (accessed with ARCLEN,ON) is suitable for nonlinear static equilibrium solutions
of unstable problems.

Application of the arc-length method involves the tracing of a complex path in the load-displacement
response into the buckling/post buckling regimes. The arc-length method uses Crisfield’s method
(Crisfield [176] (p. 934)) to prevent any fluctuation of the step size during equilibrium iterations. It is
assumed that all load magnitudes can be controlled by a single scalar parameter (that is, the total
load factor).

An unsmooth or discontinuous load-displacement response, which is sometimes seen in contact

analyses and elastic-perfectly plastic analyses, cannot be traced effectively by the arc-length solution
method.

Mathematically, the arc-length method can be viewed as the trace of a single equilibrium curve in a
space spanned by the nodal displacement variables and the total load factor. Therefore, all options
of the Newton-Raphson method are still the basic method for the arc-length solution. As the displace-
ment vectors and the scalar load factor are treated as unknowns, the arc-length method itself is an
automatic load step method; therefore, AUTOTS,ON is not needed.

For problems with sharp turns in the load-displacement curve or with path-dependent materials, it
is necessary to limit the initial arc-length radius (NSUBST) and the arc-length radius augmentation
(via the MAXARC argument of the ARCLEN command). During the solution, the arc-length method
will vary the arc-length radius at each arc-length substep according to the degree of nonlinearity that
is involved. The range of variation of the arc-length radius is limited by the maximum and minimum
multipliers (MAXARC and MINARC on the ARCLEN command).

In the arc-length procedure, Equation 14.163 (p. 731) is recast and associated with the total load factor
λ:
(14.182)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 739
Analysis Tools

Figure 14.13: Arc-Length Approach with Full Newton-Raphson Method (p. 740) is a graphical represent-
ation of the arc-length method. Writing the proportional loading factor λ in an incremental form at
substep n and iteration i yields:
(14.183)

Figure 14.13: Arc-Length Approach with Full Newton-Raphson Method

Following Equation 14.183 (p. 740), the incremental displacement {Δui} can be written in two parts:
(14.184)

where:
(14.185)

(14.186)

In each arc-length iteration, it is necessary to use Equation 14.185 (p. 740) and Equation 14.186 (p. 740)
to solve for and .

We can define one vector between the previous equilibrium point and a point determined in iteration
i by:
(14.187)

where:

{Δun} = the displacement increment accumulated over the current time step
β = the scaling vector for unit displacements

The norm of this vector is:

(14.188)

At iteration i+1, the vector becomes:

(14.189)

Crisfield’s method assumes that the norm of the vector is constant along the equilibrium iterations,
that is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
740 Published: 2023-10-23
 Newton-Raphson Procedure

(14.190)

Substituting Equation 14.187 (p. 740) and Equation 14.189 (p. 740) and using Equation 14.184 (p. 740),
the following quadratic equation results:
(14.191)

where:

This system has two real roots, Δλ(1) and Δλ(2), which both satisfy the constant arc-length radius.

For each of these roots, we can define the angle between vector tn-1 in the previously converged
substep, and vector ti+1 in the current substep. This angle may be obtained from:

(14.192)

To move the equilibrium path forward, we choose the route for which the cosine of the associated
angle is the closest to 1.

Finally, the solution vectors are updated according to (see Figure 14.13: Arc-Length Approach with
Full Newton-Raphson Method (p. 740)):
(14.193)

and
(14.194)

Values of λn and Δλ are available in POST26 corresponding to labels ALLF and ALDLF, respectively,
on the SOLU command. The normalized arc-length radius label ARCL (also on SOLU) corresponds to
value , where is the initial arc-length radius defined through Equation 14.188 (p. 740) and
Equation 14.195 (p. 741) (an arc-length radius at the first iteration of the first substep).

Defining K0 as the initial tangent matrix, Fa as the full external load, and λ0 as the initial load factor
(specified using the NSUBST command), the initial arc-length radius t0 is determined by:
(14.195)

where:

The factors MAXARC and MINARC (input on the ARCLEN command) are used to define the limits of
the arc-length radius by the following formulas:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 741
Analysis Tools

tMAX = MAXARC * t0

tMIN = MINARC * t0

14.11. Constraint Equations

14.11.1. Derivation of Matrix and Load Vector Operations

Given the set of L linear simultaneous equations in unknowns uj subject to the linear constraint
equation (input on CE command)

(14.196)

where:

Kkj = stiffness term relating the force at degrees of freedom k to the displacement at degrees
of freedom j
uj = nodal displacement of degrees of freedom j
Fk = nodal force of degrees of freedom k
k = equation (row) number
j = column number
L = number of equations

(14.197)

normalize Equation 14.197 (p. 742) with respect to the prime degrees of freedom ui by dividing by Ci
to get:

(14.198)

where:

which is written to a file for backsubstitution. Equation 14.198 (p. 742) is expanded (recall = 1) as:

(14.199)

Equation 14.196 (p. 742) may be similarly expanded as:

(14.200)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
742 Published: 2023-10-23
 Constraint Equations

Multiply Equation 14.199 (p. 742) by Kki and subtract from Equation 14.200 (p. 742) to get:

(14.201)

Specializing Equation 14.201 (p. 743) for k = i allows it to be written as:

(14.202)

This may be considered to be a revised form of the constraint equation. Introducing a Lagrange
multiplier λk, Equation 14.201 (p. 743) and Equation 14.202 (p. 743) may be combined as:

(14.203)

By the standard Lagrange multiplier procedure (see Denn([8] (p. 925))):

(14.204)

Solving Equation 14.199 (p. 742) for ui,

(14.205)

so that
(14.206)

Substituting Equation 14.206 (p. 743) into Equation 14.203 (p. 743) and rearranging terms,

(14.207)

or

(14.208)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 743
Analysis Tools

14.11.2. Constraints: Automatic Selection of Dependent Degrees of Freedom

The constraint equation described by Equation 14.197 (p. 742) can also be written in the following
matrix form:
(14.209)

where [C] can be further composed into the dependent and independent degrees of freedom so that
the direct-elimination method can be used. (In the direct elimination method, Equation 14.207 (p. 743)
is used to solve equation systems Equation 14.196 (p. 742) and Equation 14.197 (p. 742) together.)

Equation 14.209 (p. 744) can be rewritten as:

(14.210)

where {UD} is a displacement of the dependent degrees of freedom, and {UI} is the independent degree
of freedom.

If external CE or CP commands are issued, the user must choose which degrees of freedom are de-
pendent and which are independent (CE/CP). In many applications, Mechanical APDL automatically
creates constraint equations and invokes the method of automatic selection.

When solving the equation with the direct elimination method, the {UD} variables can be removed
from the system by applying the following transformation:
(14.211)

Because the choice of {UD} is not unique, the program selects {UD} dependent degrees of freedom
that ensure that Equation 14.197 (p. 742) has the best possible matrix conditioning (to avoid an ill-
conditioned matrix) and fewer fill-ins when the sparse direct solver is used.

If a value for {UD} cannot be selected that makes Equation 14.196 (p. 742) and Equation 14.197 (p. 742)
solvable together, the equation system is overconstrained. Typically, overconstraint is caused by
contradictory constraint equations or an insufficient number of dependent degrees of freedom.

To calculate the dependent degrees of freedom and detect overconstrained degrees of freedom,
Mechanical APDL uses two different methods automatically:

The Topological Method

This simple method requires little computational time to check for overconstraints and determine
unique dependent degrees of freedom. This method starts by traversing the topology (sparsity)
pattern of the [C] matrix (Equation 14.209 (p. 744)) row-by-row. During this process, the first possible
degrees of freedom visited are considered dependent degrees of freedom. If the dependent degree
of freedom has been used (or appeared) before, then it selects the next possible degree of freedom
visited as a dependent degree of freedom. Although this method traverses the entire matrix, it
is computationally efficient because no numerical computations are involved.

This method does not guarantee that overconstraint or unique dependent degrees of freedom
can be determined for the entire equation system. If this method fails to detect dependent degrees
of freedom, it swtches to the algebreic method automatically.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
744 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

The Algebraic Method

This is a mathematical method that requires more computational time than the topological
method.

The dependent degrees of freedom are chosen using a QR factorization with full pivoting applied
to the [C] matrix. If one pivot is too small (for example, lower than 1.e-08 proportionally), the as-
sociated constraint equation is considered redundant and removed from the set of constraint
equations that must be satisfied.

If redundant constraint equations are detected, the program prints information regarding their
removal.

Note that in cases where a model uses the CE/CP command and u-P elements involving Lagrange
multipliers (for example, mixed u-P formulations (p. 53)), overconstraint could also come from the
Lagrange multipliers (such as pressure degrees of freedom from the mixed u-P formulations (p. 53)),
in addition to CE/CP overconstraint.

If CE/CP values are invalid or redundant, they are removed from equation (Equation 14.196 (p. 742))
automatically, and the invalid or redundant CE/CP will not be used in the entire analysis process,
which includes the Newton-Raphson nonlinear convergence loop. Additionally, if the invalid or re-
dundant CE/CP are user defined, the application stops with an error message.

14.12. Eigenvalue and Eigenvector Extraction

The following extraction methods and related topics are available:
14.12.1. Supernode Method
14.12.2. Block Lanczos
14.12.3. PCG Lanczos
14.12.4. Unsymmetric Method
14.12.5. Subspace Method
14.12.6. Damped Method
14.12.7. QR Damped Method
14.12.8. Shifting
14.12.9. Repeated Eigenvalues
14.12.10. Complex Eigensolutions

The eigenvalue and eigenvector problem needs to be solved for mode-frequency and buckling analyses.
It has the form of:
(14.212)

where:

[K] = structural stiffness matrix

= eigenvector
λi = eigenvalue

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 745
Analysis Tools

[M] = structural mass matrix

For prestressed modal analyses, the [K] matrix includes the stress stiffness matrix [S]. For eigenvalue
buckling analyses, the [M] matrix is replaced with the stress stiffness matrix [S]. The discussions given
in the rest of this section assume a modal analysis (ANTYPE,MODAL) except as noted, but also generally
applies to eigenvalue buckling analyses.

The eigenvalue and eigenvector extraction procedures available include the Block Lanczos, PCG Lanczos,
Supernode, Subspace, unsymmetric, damped, and QR damped methods (MODOPT and BUCOPT com-
mands) outlined in Table 14.1: Procedures Used for Eigenvalue and Eigenvector Extraction (p. 746). Each
method is discussed subsequently. Shifting, applicable to most methods, is discussed in a later section
(see Shifting (p. 754)).

Table 14.1: Procedures Used for Eigenvalue and Eigenvector Extraction

Procedure Input Usages Applicable Reduction Extraction

Matrices++ Technique
Supernode MODOPT, Symmetric (but not [K], [M] None Internally uses node
SNODE applicable to grouping, reduced,
buckling) and Lanczos
methods
Block Lanczos MODOPT, Symmetric [K], [M] None Lanczos which
LANB internally uses QL
algorithm
PCG Lanczos MODOPT, Symmetric (but not [K], [M] None Lanczos which
LANPCG applicable for internally uses QL
buckling) algorithm
Subspace MODOPT, Symmetric [K], [M] None Internally uses
SUBSP Subspace method
with auto-shift
Unsymmetric MODOPT, Unsymmetric [K*], [M*] None Lanczos which
UNSYM matrices internally uses QR
algorithm
Damped MODOPT, Symmetric or [K*], [C*], None Lanczos which
DAMP unsymmetric [M*] internally uses QR
damped systems algorithm
QR Damped MODOPT, Symmetric or [K*], [C*], Modal QR algorithm for
QRDAMP unsymmetric [M] reduced modal
damped systems damping matrix
++ [K] = stiffness matrix, [C] = damping matrix, [M] = mass or stress stiffening matrix, * = can be
unsymmetric

The Block Lanczos and PCG Lanczos methods both use Lanczos iterations to extract the requested ei-
genvalues. However, the Block Lanczos method uses the sparse direct solver, while the PCG Lanczos
method uses the PCG iterative solver internally to solve the necessary system of equations at each
Lanczos iteration.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
746 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

14.12.1. Supernode Method

The Supernode (SNODE) solver is used to solve large, symmetric eigenvalue problems for many modes
(up to 10,000 and beyond) in one solution. A supernode is a group of nodes from a group of elements.
The supernodes for the model are generated automatically by the solver. This method first calculates
eigenmodes for each supernode in the range of 0.0 to FREQE*RangeFact (where RangeFact is specified
by the SNOPTION command and defaults to 2.0), and then uses the supernode eigenmodes to calculate
the global eigenmodes of the model in the range of FREQB to FREQE (where FREQB and FREQE are
specified by the MODOPT command). Typically, this method offers faster solution times than Block
Lanczos or PCG Lanczos if the number of modes requested is more than 200.

The Supernode eigensolver uses an approximate method compared to the Block Lanczos and PCG
Lanczos solutions. The accuracy of the Supernode solution can be controlled by the SNOPTION
command. By default, the eigenmode accuracy is based on the frequency range used (assuming no
rigid body modes are in the model), as shown in the following table.

Frequency Range Accuracy of Supernode

solution
0 - 100 Hz 0.01 percent error
100 - 200 Hz 0.05 percent error
200 - 400 Hz 0.20 percent error
400 - 1000 Hz 1.00 percent error
1000 Hz and higher 3.0 - 5.0 percent error

Typically, the reason for seeking many modes is to perform a subsequent mode-superposition or PSD
analysis to solve for the response in a higher frequency range. The error introduced by the Supernode
eigensolver (shown in the table above, assuming no rigid body modes are in the model) is small
enough for most engineering purposes. You can use the SNOPTION command to increase the accuracy
of the solution, but at the cost of increased computing time. Increasing the value of RangeFact (on
the SNOPTION command) results in a more accurate solution.

In each step of the Supernode eigenvalue calculation, a Sturm check is performed. The occurrence
of missing modes in the Supernode calculation is rare.

The lumped mass matrix option (LUMPM,ON) is not allowed when using the Supernode mode extrac-
tion method. The consistent mass matrix option will be used regardless of the LUMPM setting.

The Supernode eigensolver has the following known limitations:

• If rigid body modes are present in the model and are part of the eigen-solution, the accuracy
of the lower flexible modes is slightly deteriorated due to the rigid body modes present and
the theoretical limitation of the Supernode eigensolver. In this case, you should increase the
value of RangeFact to be higher than 2.0 (default) and smaller than 10.0 to achieve better
accuracy for the lower flexible modes.

• If duplicate modes are present in the eigensolution, the accuracy of the duplicate modes may
be less than desired. In this case, you should increase the value of RangeFact to be higher
than 2.0 (default) to achieve better accuracy for the duplicate modes.

• Since the supernodes are formed arbitrarily and automatically, differences in domain decom-
position (that is, using a different number of cores with distributed-memory parallel processing)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 747
Analysis Tools

slightly affects the solution accuracy. However, the resulting inaccuracy is normally less than
0.1 % in the eigenvalue solution.

• The Supernode eigensolver does not use equation pivoting logic to process or factorize local
equations. Therefore, any finite element formulation that requires pivoting to achieve an
equation solution, such as an element formulation with Lagrangian multipliers (u-P formulation),
is not supported. In this situation, the program automatically switches to the Block Lanczos
modal solution.

14.12.2. Block Lanczos

The Block Lanczos eigenvalue extraction method (accessed with MODOPT,LANB or BUCOPT,LANB)
is available for large symmetric eigenvalue problems.

The block shifted Lanczos algorithm is a variation of the classical Lanczos algorithm, where the Lanczos
recursions are performed using a block of vectors, as opposed to a single vector. Additional theoret-
ical details on the classical Lanczos method can be found in Rajakumar and Rogers([198] (p. 936)).

A block shifted Lanczos algorithm, as found in Grimes et al.([197] (p. 936)) is the theoretical basis of
the eigensolver. The Block Lanczos method employs an automated shift strategy to extract the
number of eigenvalues requested.

At the end of the Block Lanczos calculation, the solver can perform a Sturm sequence check if requested
by the user. (The Sturm check is off by default and is not needed for most applications.) This check
computes the number of negative pivots encountered in the range that the minimum and maximum
eigenvalues encompass. This number will match the number of converged eigenvalues unless some
eigenvalues have been missed. Block Lanczos will report the number of missing eigenvalues, if any.

Use of the Block Lanczos method for solving larger models (1,000,000 DOF, for example) with many
constraint equations (CE) and 3D solid elements can require a significant amount of computer memory.
The alternative method of PCG Lanczos, which internally uses the PCG solver, could result in savings
of memory and computing time.

14.12.3. PCG Lanczos

The theoretical basis of this eigensolver is found in Grimes et al.([197] (p. 936)), which is the same
basis for the Block Lanczos eigenvalue extraction method. However, the implementaion differs
somewhat from the Block Lanczos eigensolver, in that the PCG Lanczos eigensolver:

• does not employ an automated shift strategy during the eigenvalue analysis.

• is only available for modal analyses and is not applicable to buckling analyses.

14.12.4. Unsymmetric Method

The unsymmetric eigensolver (accessed with MODOPT,UNSYM) is applicable whenever the system
matrices are unsymmetric. For example, an acoustic fluid-structure interaction problem using FLUID30
elements results in unsymmetric matrices. Also, certain problems involving the input matrix element
MATRIX27 and/or COMBI214 element, such as in brake squeal problems, can give rise to unsymmetric
system matrices. A generalized eigenvalue problem given by the following equation:
(14.213)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
748 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

can be setup and solved using the mode-frequency analysis (ANTYPE,MODAL). The matrices [K] and
[M] are the system stiffness and mass matrices, respectively. Either or both [K] and [M] can be unsym-
metric. {ϕi} is called the eigenvector. It represents the right eigenvector. The left eigenvector, which
is extracted for mode-superposition harmonic analysis (MODOPT with ModType = BOTH), satisfies
the following equation:
(14.214)

where:

= ith left eigenmode

The method employed to solve the unsymmetric eigenvalue problem is a subspace approach based
on a method designated as Frequency Derivative Method. The FD method uses an orthogonal set of
Krylov sequence of vectors:
(14.215)

To obtain the expression for the sequence of vectors, the generalized eigenvalue Equa-
tion 14.213 (p. 748) is differentiated with respect to λi to get:
(14.216)

Substituting Equation 14.216 (p. 749) into Equation 14.213 (p. 748) and rearranging after applying a
shift s, the starting expression for generating the sequence of vectors is given by:
(14.217)
(14.218)

where:

s = an initial shift

The general expression used for generating the sequence of vectors is given by:
(14.219)

This matrix equation is solved by a sparse matrix solver (EQSLV, SPARSE). However, an explicit spe-
cification of the equation solver (EQSLV command) is not needed.

The subspace made of these derivatives allows the program to find the closest eigensolutions from
this shift point. The shift value s is initially determined using the FREQB value on the MODOPT
command.

For a large number of eigenvalues, the UNSYM extraction algorithm is able to move automatically to
a new shift if the first solve only finds a subset of eigensolutions. This process will be repeated until
all the required eigenvalues are found, unless the algorithm fails several times to find any accurate
eigenvalues.

A subspace transformation of Equation 14.213 (p. 748) is performed using the sequence of orthogonal
vectors which leads to the reduced eigenproblem:
(14.220)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 749
Analysis Tools

where:

[K*] = [Q]T[K] [Q]

[M*] = [Q]T[M] [Q]

The eigenvalues of the reduced eigenproblem (Equation 14.220 (p. 749)) are extracted using a direct
eigenvalue solution procedure. The eigenvalues μi are the approximate eigenvalues of the original
eigenproblem and they converge to λi with increasing subspace size m. The converged eigenvectors
are then computed using the subspace transformation equation:
(14.221)

This method does not perform a Sturm Sequence check for possible missing modes. At the lower
end of the spectrum close to the shift (input as FREQB on MODOPT command), the frequencies
usually converge without missing modes.

If the eigensolutions are real, the frequency (Hz) is output as:

If the eigensolutions are complex, the complex frequency (Hz) is output as:

For more information, see Complex Eigensolutions (p. 756).

14.12.5. Subspace Method

The subspace eigensolver (accessed with BUCOPT,SUBSP and MODOPT,SUBSP) is applicable for
buckling and modal analyses in which the system matrices are symmetric. This eigensolver uses ba-
sically the same algorithm as the unsymmetric eigensolver (accessed with MODOPT,UNSYM) to solve
the generalized eigenvalue problem given by the following equation.

For buckling analysis:

(14.222)

where:

[K] = structural stiffness matrix

[S] = stress stiffness matrix
λi = ith eigenvalue (used to multiply the loads that generated [S])

{Ψi} = ith eigenvector of displacements

For modal analysis:

(14.223)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
750 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

where:

[K] = structural stiffness matrix

[M] = mass matrix
λi = ith eigenvalue

{ϕi} = ith eigenvector of displacements

The advantage of the SUBSP method over the Block Lanczos method is that both the [K] and [S] /
[M] matrices can be indefinite at the same time.

Buckling Analysis

Using the options of the BUCOPT command, the program can find the eigenvalues in one of two
ways:

• In a specified load multiplier range of interest (RangeKey = RANGE), or

• Around an initial shift (RangeKey = CENTER).

Note that when using the CENTER option, the automatic (internal) shifting strategy of the algorithm
is disabled. The eigenvalues found around the initial shift (CENTER value) are available; the solver
may not find eigenvalues far away from the CENTER value.

Modal Analysis

Some options of the subspace algorithm can be set using the SUBOPT command. These include
memory management and Sturm sequence check. By default, the Sturm check is turned off.

14.12.6. Damped Method

The damped eigensolver (accessed with MODOPT,DAMP) is applicable only when the system damping
matrix needs to be included in Equation 14.212 (p. 745), where the eigenproblem becomes a quadratic
eigenvalue problem given by:
(14.224)

where:

[C] = damping matrix

Matrices may be symmetric or unsymmetric.

The method employed to solve the damped eigenvalue problem is the same as for the UNSYM option.
We first transform the initial quadratic equation (Equation 14.224 (p. 751)) in a linear form applying
the variable substitutions:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 751
Analysis Tools

To form the equivalent UNSYM eigenvalue problem.

(14.225)

Solutions of Equation 14.224 (p. 751) and Equation 14.225 (p. 752) are equivalent, except that only the
first-half part of the eigenvectors is considered.

The UNSYM method uses Equation 14.225 (p. 752). The default blocksize value to solve a Quadratic
Damp Eigenproblem is set to four. This value can be controlled using the blocksize parameter of the
MODOPT command.

This method does not perform a Sturm Sequence check for possible missing modes. At the lower
end of the spectrum, close to the shift (input as FREQB on the MODOPT command), the frequencies
usually converge without missing modes. By default, this method does not employ any shift strategy
during the eigenvalue analysis. Use the DAMPOPT,SHIFT command to activate or deactivate the
automated shift strategy.

The complex frequency (Hz) is output as:

For more information, see Complex Eigensolutions (p. 756).

14.12.7. QR Damped Method

The QR damped method (accessed with MODOPT,QRDAMP) is a procedure for determining the
complex eigenvalues and corresponding eigenvectors of damped linear systems. This eigensolver allows
for nonsymmetric [K] and [C] matrices. The eigensolver is computationally efficient compared to the
damped eigensolver (MODOPT,DAMP). This method employs the modal orthogonal coordinate
transformation of system matrices to reduce the eigenvalue problem into the modal subspace. The
QR algorithm is then used to calculate eigenvalues of the resulting quadratic eigenvalue problem in
the modal subspace.

The equations that follow apply to Rayleigh and element-based damping in a QR Damped modal
analysis. For the equations with constant structural damping (input with DMPSTR or MP,DMPS), refer
to QR Damped Method with Constant Structural Damping (p. 753).

The equations of elastic structural systems without external excitation can be written in the following
form:
(14.226)

(See Equation 15.5 (p. 796) for definitions).

It has been recognized that performing computations in the modal subspace is more efficient than
in the full eigenspace. The stiffness matrix [K] can be symmetrized by rearranging the unsymmetric
contributions; that is, the original stiffness matrix [K] can be divided into symmetric and unsymmetric
parts. By dropping the damping matrix [C] and the unsymmetric contributions of [K], the symmetric

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
752 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

eigenvalue problem is first solved to find real eigenvalues and the corresponding eigenvectors. In
the present implementation, the unsymmetric element stiffness matrix is zeroed out for the symmetric
eigenvalue problem. Following is the coordinate transformation (see Equation 14.123 (p. 722)) used
to transform the full eigenvalue problem into modal subspace:
(14.227)

where:

[Φ] = eigenvector matrix normalized with respect to the mass matrix [M]
{y} = vector of modal coordinates

By using Equation 14.227 (p. 753) in Equation 14.226 (p. 752), we can write the differential equations
of motion in the modal subspace as follows:
(14.228)

where:

[Λ2] = a diagonal matrix containing the first n eigen frequencies ωi

For classically damped systems, the modal damping matrix [Φ]T[C][Φ] is a diagonal matrix with the
diagonal terms being 2ξiωi, where ξi is the damping ratio of the i-th mode. For non-classically damped
systems, the modal damping matrix is either symmetric or unsymmetric. Unsymmetric stiffness con-
tributions of the original stiffness are projected onto the modal subspace to compute the reduced
unsymmetric modal stiffness matrix [Φ]T [Kunsym] [Φ].

Introducing the 2n-dimensional state variable vector approach, Equation 14.228 (p. 753) can be written
in reduced form as follows:
(14.229)

where:

(14.230)

(14.231)

The 2n eigenvalues of Equation 14.229 (p. 753) are calculated using the QR algorithm (Press et
al.([254] (p. 939))). The inverse iteration method (Wilkinson and Reinsch([354] (p. 945))) is used to calculate
the complex modal subspace eigenvectors. The full complex eigenvector, {ψ}, of the original system
is recovered using the following equation:
(14.232)

[Link]. QR Damped Method with Constant Structural Damping

When a constant structural damping ratio is defined or an element generating a complex stiffness
is present in a QR damped modal analysis with the complex mode shapes requested (Cpxmod =
YES on MODOPT), the state-space matrix from Equation 14.231 (p. 753) becomes complex. This
can be written as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 753
Analysis Tools

(14.233)

where:

= constant structural damping coefficient (input with DMPSTR)

= number of materials with MP,DMPS input
= constant structural damping coefficient for material j (input with MP,DMPS)

= portion of structural stiffness matrix based on material j

= number of elements with specified imaginary stiffness matrix

= imaginary element stiffness matrix

[Link]. QR Damped Method Output

The QR Damped Method calculates 2n solutions, but only n solutions are output for post-processing:
only the positive imaginary solution (positive frequency) is retained. In the case of high damping,
all overdamped modes are also retained.

The complex frequency (Hz) is output as:

(14.234)

For more information, see Complex Eigensolutions (p. 756).

14.12.8. Shifting
Various shifting strategies are used by most of the extraction methods in an effort to improve the
accuracy, robustness, and efficiency of the algorithms. The logic on how those shift values are chosen
is discussed in this section.

In some cases it is desirable to shift the values of eigenvalues either up or down. These fall in two
categories:

1. Shifting down, so that the solution of problems with rigid body modes does not require working
with a singular matrix.

2. Shifting up, so that the bottom range of eigenvalues will not be computed, because they had ef-
fectively been converted to negative eigenvalues. This will, in general, result in better accuracy
for the higher modes. The shift introduced is:

(14.235)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
754 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

λ = desired eigenvalue
λ0 = eigenvalue shift
λi = eigenvalue that is extracted

λ0, the eigenvalue shift is computed as:

(14.236)

When using the Block Lanczos or PCG Lanczos method, if no user input is given for SHIFT (BUCOPT
command) or FREQB (MODOPT command), the following logic is used:

(14.237)

where Mii and Kii are the diagonals of the [M] and [K] matrices, respectfully. The summation is taken
over all terms where Kii ≠ 0 and where < 10e4. The number of such terms is n.

When using the PCG Lanczos method, if a Lev_Diff value of 1, 2, 3, or 4 is chosen (either automat-
ically or by the user, see PCGOPT), then -λ0 from Equation 14.237 (p. 755) is used when rigid body
modes are detected in order to avoid working with indefinite matrices with the PCG iterative solver.
When using Lev_Diff = 5, λ0 is used as with the Block Lanczos method.

Equation 14.235 (p. 754) is combined with Equation 14.212 (p. 745) to give:
(14.238)

Rearranging,
(14.239)

or
(14.240)

where:

[K]' = [K] - λ0 [M]

It may be seen that if [K] is singular, as in the case of rigid body motion, [K]' will not be singular if
[M] is not totally zero (which is normally true) and if λ0 is input as a non-zero number.

Once λi is computed, λ is computed from Equation 14.235 (p. 754) and reported.

14.12.9. Repeated Eigenvalues

Repeated eigenvalues occur, for example, for a thin, axisymmetric pole. Two independent sets of or-
thogonal motions are possible corresponding to the same frequency.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 755
Analysis Tools

In this case, the eigenvectors are not unique, as there is an infinite number of correct solutions. The
repeated eigenvectors can be computed accurately when all are extracted.

In the special case of two or more identical but disconnected structures run as one analysis, eigen-
vectors may include components from more than one substructure. To reduce confusion in such
cases, you should run a separate analysis for each structure.

14.12.10. Complex Eigensolutions

For problems involving spinning structures with gyroscopic effects, and/or damped structural eigen-
frequencies, the eigensolutions obtained with the Damped Method (p. 751) and QR Damped Meth-
od (p. 752) are complex. For problems involving frictional contact, such as brake squeal analyses, or
problems involving structural damping, the eigensolutions obtained with the Damped Method (p. 751),
QR Damped Method (p. 752), or Unsymmetric Method (p. 748) are also complex. The eigenvalues
are given by:
(14.241)

where:

= complex eigenvalue (rd/s)

σi = real part of the eigenvalue (stability value)
ωi = imaginary part of the eigenvalue (damped frequency)

j=

Complex eigenvalues generally come in complex conjugate pairs, except when system matrices are
unsymmetric and structural damping is present.

The complex eigenvectors can be mass normalized using the following equation:

(14.242)

where:

= complex transpose of

= mass normalized complex eigenvector

The eigenvectors can also be normalized to unity using MODOPT,,,,,,ON. The largest component value
is 1.0 in this case. This is done by first finding the location of the maximum absolute value of the ei-
genvector. The complex component of the eigenvector at this location is denoted as . The unit-
normalized eigenvector is given by:

(14.243)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
756 Published: 2023-10-23
 Eigenvalue and Eigenvector Extraction

= unit normalized complex eigenvector

Note that when normalizing the damped eigenvectors, the normalization is performed on the first
half of the eigenvectors and the original (non-augmented) mass matrix. See Damped Method (p. 751)
for details on the augmented method for damped eigensolutions.

The dynamic response of the system is given by:

(14.244)

where:

t = time

The ith eigenvalue is stable if σi is negative and unstable if σi is positive. This does not apply to the
UNSYM eigensolver when the eigenvalue is purely real. In this case, the sign of σi is not significant.

Complex Frequency
The complex frequency (Hz) is expressed as:
(14.245)

Modal damping ratio

The modal damping ratio is given by:

(14.246)

where:

αi = modal damping ratio of the ith eigenvalue

It is the ratio of the actual damping to the critical damping.

Logarithmic decrement
The logarithmic decrement represents the logarithm of the ratio of two consecutive peaks in the dy-
namic response (Equation 14.244 (p. 757)). It can be expressed as:

(14.247)

where:

δi = logarithmic decrement of the ith eigenvalue

Ti = damped period of the ith eigenvalue defined by:
(14.248)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 757
Analysis Tools

The logarithmic decrement defined in Equation 14.247 (p. 757) is consistent with the API (American
Petroleum Institute) standards for rotating machinery. It is positive for stable motion and obtained
by setting STABVAL = 3 on the PLCAMP and PRCAMP commands.

14.13. Analysis of Cyclically Symmetric Structures

The following topics are available concerning cyclically symmetric structure analysis:
14.13.1. Modal Analysis
14.13.2. Complete Mode Shape Derivation
14.13.3. Mode-Superposition Harmonic Analysis
14.13.4. Aerodynamic Coupling
14.13.5. Expansion to Output Quantities
14.13.6. Mistuning
14.13.7. Cyclic Symmetry Transformations

14.13.1. Modal Analysis

Given a cyclically symmetric (periodic) structure such as a fan wheel, a modal analysis can be performed
for the entire structure by modelling only one sector of it. A proper base sector represents a pattern
that, if repeated n times in cylindrical coordinate space, would yield the complete structure.

Figure 14.14: Typical Cyclically Symmetric Structure

Base Sector

In a flat circular membrane, mode shapes are identified by harmonic indices. For more information,
see the Cyclic Symmetry Analysis Guide.

Constraint relationships (equations) can be defined to relate the lower (θ = 0) and higher (θ = α, where
α = sector angle) angle edges of the base sector to allow calculation of natural frequencies related
to a given number of harmonic indices. The base sector is duplicated in the modal analysis to satisfy
the required constraint relationships and to obtain nodal displacements. This technique was adapted
from Dickens([149] (p. 933)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
758 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

Figure 14.15: Base Sector Definition

Low Component Nodes

High Component Nodes

Z
Y
angle
Sector
X
CSYS =1

Constraint equations relating the lower and higher angle edges of the two sectors are written:

(14.249)

where:

, = calculated displacements on lower angle side of base and duplicated sectors ( and
, respectively)
, = displacements on higher angle side of base and duplicated sectors ( and , respect-
ively) determined from constraint relationships

= sector angle
= number of sectors in 360°

Three basic steps in the procedure are briefly:

1. The CYCLIC command in /PREP7 automatically detects the cyclic symmetry model information,
such as edge components, the number of sectors, the sector angles, and the corresponding cyclic
coordinate system.

2. The CYCOPT command in /SOLU generates a duplicated sector and applies cyclic symmetry con-
straints (Equation 14.249 (p. 759)) between the base and the duplicated sectors.

3. The /CYCEXPAND command in /POST1 expands a cyclically symmetry response by combining

the base and the duplicated sectors results (Equation 14.250 (p. 759)) to the entire structure.

14.13.2. Complete Mode Shape Derivation

The mode shape in each sector is obtained from the eigenvector solution. The displacement compon-
ents (x, y, or z) at any node in sector n for harmonic index k, in the full structure is given by:

where:
(14.250)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 759
Analysis Tools

= sector number, varies from 1 to

= mode number
= base sector displacement
= duplicate sector displacement
= arbitrary phase angle

If the mode shapes are normalized to the mass matrix in the mode analysis (via Nrmkey on the
MODOPT command), the normalized displacement components in the full structure are given by:

(14.251)

If the mode shapes are normalized using the maximum value of the mode (Nrmkey = ON in the
MODOPT command), the normalization equation is:

(14.252)

where

= mode number
= DOF number

The complete procedure addressing static, modal, and prestressed modal analyses of cyclically sym-
metric structures is contained in the Cyclic Symmetry Analysis Guide.

14.13.3. Mode-Superposition Harmonic Analysis

A bladed disk (axial or centrifugal) is cyclically symmetric about its fundamental sector, as shown in
Figure 14.16: Full Model with the Cyclic-Symmetric Sector Highlighted (p. 761). These bladed disks are
typically excited by rotationally symmetric excitations which produce forced vibrations.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
760 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

Figure 14.16: Full Model with the Cyclic-Symmetric Sector Highlighted

The equation of motion for harmonically-varying forced response is:

(14.253)

where:

= stiffness matrix of the bladed disk

= mass matrix of the bladed disk
= load vector
= displacement vector
= forcing frequency, input on the CYCFREQ command
The superscript implies the matrices refer to the entire 360° system. Damping is ignored in this
equation and is discussed in Damping (p. 766).

[Link]. Transform to Modal Coordinates

Equation 14.253 (p. 761) can be transferred into modal space by:
(14.254)

where:

= the set eigenvectors of the full system

= the vector of modal coordinates

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 761
Analysis Tools

Substituting this into Equation 14.253 (p. 761) and pre-multiplying by yields the forced re-
sponse equation in terms of modal coordinates:
(14.255)

where:

= the diagonal matrix for the system natural frequencies squared

= the identity matrix
The number of eigenvectors (modes) used in Equation 14.255 (p. 762) is much less than the total
number of degrees of freedom, so that Equation 14.255 (p. 762) is significantly smaller system than
Equation 14.253 (p. 761).

[Link]. Cyclic Coordinates

A cyclically-symmetric structure has the following transformation from cyclic coordinates to physical
coordinates:
(14.256)

where:

= cyclic (or harmonic) coordinates for harmonic index

= real-valued Fourier matrix
denotes the Kronecker product
The system eigenvectors can also be expressed in terms of their harmonics:
(14.257)

where:

= the real-valued cyclic modes corresponding to harmonic

= a pseudo-block diagonal matrix. A block is 1×1 for = 0 and if is even,

and 2×2 otherwise. is the number of sectors (CYCLIC).
Substituting Equation 14.257 (p. 762) into Equation 14.255 (p. 762) gives the frequency response
equation in cyclic coordinates:
(14.258)

where the modal load is:

(14.259)

[Link]. Properties of Paired Eigenmodes

The Fourier transformation expression:
(14.260)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
762 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

appears in Equation 14.257 (p. 762) and Equation 14.259 (p. 762) (transposed), where are the
harmonic modes. has the structure:

(14.261)

The harmonics (0 < < ) appear as 2×2 sub-blocks, each column being one of the repeated
modes and the rows corresponding to the real and imaginary (base and duplicate sector) solutions.
The first mode in a pair is denoted with subscript and the second mode in the pair is denoted
with subscript .

The real and imaginary terms of each double mode are related and obey the following:
(14.262)

where = 1 if the mode shapes and have the same sign, otherwise = -1. Likewise:

(14.263)

where . This property leads to the .MODE file needing to only contain the real (base) sector
degrees of freedom for each mode shape. Likewise, the stresses and strains for the expanded modes
on the .RST file only contain the results for the base elements (and not the duplicate elements).

[Link]. >Forcing
The forcing function is assumed to be time harmonic as well as spatially harmonic in the circumfer-
ential direction. The force on any sector n can then be related to the force on the base sector by
only a phase shift:
(14.264)

where:

= complex load vector on the base sector

= phase of excitation (inter-blade phase angle, IBPA)

= sector number
= engine order (EO) excitation (CYCFREQ)
= number of sectors

Equation 14.264 (p. 763) defines the distribution of force over the blades using the blade numbering
convention shown in Figure 14.17: Forcing Sign and Numbering Convention (p. 764). With this con-
vention, blade 2 leads blade 1 by the inter-blade phase angle ; it is subjected to the force
first and blade 1 is subjected to the same force after a rotation of radians. The forcing wave

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 763
Analysis Tools

travels in the direction shown in Figure 14.17: Forcing Sign and Numbering Convention (p. 764),
which is a backward traveling wave with respect to the rotation (OMEGA or CMOMEGA).

Figure 14.17: Forcing Sign and Numbering Convention

Equation 14.259 (p. 762) can be written as the sum of each sector, if only blade loads are considered:

(14.265)

where:

= the row of the Fourier matrix

= force on blade

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
764 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

While the engine order can take any integer value, a given engine order will only excite a certain
harmonic index. This aliased engine order is determined from the input engine order (for example,
the number of preceding stators) as outlined in the following table:

Table 14.2: Aliased Engine Order (Excited Harmonic Index)

Engine Order
Aliased Engine Order
Even Odd

… … …

This leads to the well-known zigzag diagram in Figure 14.18: Zigzag Diagram for an Even Number
of Sectors (p. 765). The positive slopes are forward traveling waves and the negative slopes are
backward traveling.

Figure 14.18: Zigzag Diagram for an Even Number of Sectors

Due to the orthogonality of the engine order phasing with the sine and cosine terms of the Fourier
matrix , only the harmonic indices = of are non-zero.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 765
Analysis Tools

The sign of is determined as follows:

1. If the aliased engine order is on a negative slope of Figure 14.18: Zigzag Diagram for an Even
Number of Sectors (p. 765), it is assigned a negative value.

2. If the rotation is opposite that illustrated in Figure 14.17: Forcing Sign and Numbering Conven-
tion (p. 764), in other words a negative in the cyclic coordinate system, the value from the
first step is negated.

3. If the engine order was input as a negative value, the value after the second step is negated
again.

This process ensures that the force is correctly applied to the blades via Equation 14.259 (p. 762),
while respecting the blade numbering, rotation direction, and the relationship between engine
order and nodal diameter.

Returning to the harmonic equation of motion, Equation 14.258 (p. 762), the force is assumed to be
harmonic and of the form:
(14.266)

The excitation frequency , and the corresponding frequency sweep range (HARFRQ) is
typically related to the rotor speed by:
(14.267)

The maximum response will occur when this excitation frequency crosses a natural frequency of
that nodal diameter, for instance when .

[Link]. Damping
Damping may be included in the following forms:

• Global structural damping (DMPSTR)

• Global Rayleigh damping and (ALPHAD and BETAD)

• Mode-dependent damping (MDAMP)

• Global modal damping (DMPRAT)

The damping matrix in cyclic modal coordinates is

(14.268)

where

= diagonal matrix of system frequencies

= diagonal matrix of modal damping terms

It is also possible to apply aerodamping as a part of aerodynamic coupling, which is addressed in

the following section.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
766 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

14.13.4. Aerodynamic Coupling

Interblade aerodynamic coupling can be introduced assuming a first order dependence on the blade
motion by inserting an additional matrix (input using CYCFREQ,AERO) into the frequency response
equation of motion for the whole system. First consider the introduction of the aero matrix in phys-
ical coordinates:
(14.269)

where non-aero damping terms are dropped for conciseness. This equation can be transformed into
modal coordinates using Equation 14.254 (p. 761).
(14.270)

The aerodynamic coupling matrix in physical coordinates can be written in cyclic cantilevered
blade coordinates (see He et. al. [431] (p. 949)):
(14.271)

where is the complex-valued Fourier matrix and is a matrix of cantilevered blade modes.
The resulting matrix is in general complex since the coupling is not in phase with the motion,
but it is presumed to be cyclically symmetric and therefore block diagonal:

(14.272)

Note that a sub-block corresponds to the interblade phase angle , where . It is common
for CFD and other codes to compute the aerodynamic coefficients that are compatible with this
matrix. Each block is made up of coefficients , that capture the coupling between mode i
and the force computed using mode j in the direction of the surface normal. These values can be
obtained using the following area integral:
(14.273)

where is the surface normal and is the area.

The aerodynamic term in Equation 14.270 (p. 767) can be rewritten by assuming that the blade portion
( ) of the system modes can be represented by a linear combination of cantilevered blade modes:
(14.274)

The aero term becomes:

(14.275)

Rearranging Equation 14.271 (p. 767) and substituting into Equation 14.275 (p. 767), it can be seen that
the aero term is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 767
Analysis Tools

(14.276)

14.13.5. Expansion to Output Quantities

Once the vector of modal coordinates is obtained for a given excitation frequency , we can
expand back to the physical solution quantities for the entire bladed disk. The displacements of sector
are determined by:
(14.277)

Stresses and strains can be similarly evaluated by using the stress and strain mode shapes for .

14.13.6. Mistuning
The program employs the CMS-based Component Mode Mistuning (CMM) methodology (see Lim et.
al. [425] (p. 949)) for small stiffness mistuning. This methodology is well suited not only to the simulation
of bladed disks, but also to integrally bladed rotors and impellers where coupled disk-blade motion
is important. From a tuned response of the nominal geometry, mistuning parameters are introduced
into the component modes which are projected onto the system modes, yielding a compact set of
equations for the harmonic frequency response of the system. This methodology builds on the theory
discussed in Mode-Superposition Harmonic Analysis (p. 760), and shares the basic core equations,
damping, forcing, and coordinate transformations.

[Link]. Equation of Motion for a Mistuned System

Stiffness mistuning can be characterized as a variation of the nominal stiffness of a system. Referring
to Equation 14.253 (p. 761), the stiffness matrix can be written as:
(14.278)

where the subscript o indicates the nominal values and indicates deviations from the nominal.
If the mistuning is assumed to be in the blade only, we can assume proportional mistuning:
(14.279)

where is the nominal stiffness matrix for a blade and is the stiffness mistuning deviation
for blade n. For a prestressed analysis, will include any prestress effects. This may alternatively
be considered as deviation in Young's modulus only in the case where there are no prestress effects.
denotes a block diagonal matrix.

[Link]. Transform to Tuned-System Modal Coordinates

Mode-superposition is used to project the equation of motion onto an appropriate basis and reducing
it to a smaller set of degrees of freedom. In contrast to typical mode-superposition techniques
where the equation of motion is projected onto the modes computed from its own system (in this
case, modes of the mistuned system), the program projects onto tuned system modes. Following
the same reasoning as in Transform to Modal Coordinates (p. 761), the forced response equation in
tuned-system modal coordinates is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
768 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

(14.280)

where is the diagonal matrix of the system natural frequencies squared and [I] is the identity
matrix. The left superscript indicates that only the blade DOFs are used from the eigenvectors.

Substituting Equation 14.279 (p. 768) into Equation 14.280 (p. 769) and recognizing that the global
triple product can be replaced with a local sum per sector yields:

(14.281)

where N is the number of sectors.

[Link]. Craig-Bampton Reduction for Mistuning

The blade matrix in Equation 14.281 (p. 769) can be approximated by applying the fixed-inter-
face Craig-Bampton (C-B) transformation:
(14.282)

where are the generalized C-B coordinates (modal participation factors), and the transformation
is given by:

(14.283)

are the set of blade modes derived from assuming the blade-disk interface is fixed (and any
shroud interfaces as well), and are the static shapes:
(14.284)

where the subscripts and indicate the partition into internal and boundary (interface) DOFs of
the blade. Note that b corresponds to master DOFs and corresponds to all remaining DOFs in the
substructure as defined in Component Mode Synthesis (CMS) (p. 831). The boundary DOFs and the
elements defining the blade matrices are both input on CYCFREQ,BLADE.

Applying this transformation to the blade stiffness yields:

(14.285)

The number of blade modes is typically small leading to being a much smaller matrix than
.

Note that this is the typical Component Mode Synthesis (CMS) substructure generation analysis for
the blade.

[Link]. Component Mode Projection

Projecting the Craig-Bampton coordinates onto the system modes for a sector via:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 769
Analysis Tools

(14.286)

and substituting into Equation 14.281 (p. 769) yields for the blade stiffness term:
(14.287)

[Link]. Mistuning Equation of Motion with Cyclic Quantities

The CMM approach to mistuning allows for the use of cyclic quantities within the reduced equations
of motion. The Craig-Bampton coordinates or modal participation factors for the sector can be
expressed in cyclic coordinates using the same transformation as in Equation 14.256 (p. 762):
(14.288)

where is the row of . Substituting Equation 14.287 (p. 770) into Equa-
tion 14.281 (p. 769) and using Equation 14.288 (p. 770) and Equation 14.257 (p. 762) we arrive at:
(14.289)

where the modal load is:

(14.290)

[Link].1. Evaluation of Participation Factors

Rewriting Equation 14.282 (p. 769) in cyclic coordinates:
(14.291)

and segregating into its parts yields:

(14.292)

From the definition of [T] in Equation 14.283 (p. 769), it is clear that:
(14.293)

where are the system modes at the blade interface points.

Premultiplying Equation 14.291 (p. 770) by and using orthogonality of the

modes, we can solve for :

(14.294)

where and are the blade natural frequencies squared and mode shapes from the
Craig-Bampton transformation.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
770 Published: 2023-10-23
 Analysis of Cyclically Symmetric Structures

[Link]. Mistuning Parameters

The stiffness parameter (input on CYCFREQ,MIST) is handled as follows.

The stiffness mistuning parameter represents the deviation of Young’s modulus of each blade
from its nominal value:
(14.295)

This leads to a modification of the blade’s natural frequencies by a common factor. In order to
modify each natural frequency independently, a frequency-dependent mistuning parameter is in-
troduced:

(14.296)

where is the nominal frequency of blade and is the mistuned frequency of blade .
Note that if is constant for all frequencies, then , that is, when all frequencies are
modified by the same ratio that is identical to the modified Young’s modulus.

The application of the effective stiffness mistuning parameter is as follows. The nominal Craig-
Bampton reduced stiffness matrix has the following form, after applying the transformation in
Equation 14.283 (p. 769):

(14.297)

Defining the average frequency mistuning for a blade as:

(14.298)

where is the number of blade frequencies, we can define the stiffness mistuning as:

(14.299)

The constraint modes are modified by the average frequency mistuning, and the frequencies
themselves are modified by the frequency-dependent input values.

14.13.7. Cyclic Symmetry Transformations

The cyclically symmetric solution sequences consist of three basic steps. The first step transforms
applied loads to cyclically symmetric components using finite Fourier theory and enforces cyclic

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 771
Analysis Tools

symmetry constraint equations (see Equation 14.249 (p. 759)) for each harmonic index (nodal diameter)
( = 0, 1, . . ., ).

Any applied load on the full 360° model is treated through a Fourier transformation process and applied
on to the cyclic sector. For each value of harmonic index, , the procedure solves the corresponding
linear equation. The responses in each of the harmonic indices are calculated as separate load steps
at the solution stage. The responses are expanded via the Fourier expansion (Equation 14.250 (p. 759)).
They are then combined to get the complete response of the full structure in postprocessing.

The Fourier transformation from physical components, , to the different harmonic index components,
, is given by the following:

Harmonic Index, = 0 (symmetric mode):

(14.300)

Harmonic Index, 0 < < (degenerate mode)

Base sector:

(14.301)

Duplicate sector:

(14.302)

For even only, Harmonic Index, (antisymmetric mode):

(14.303)

where:

= any physical component, such as displacements, forces, pressure loads, temperatures,

and inertial loads
= cyclically symmetric component

The transformation to physical components, , from the cyclic symmetry, , components is recovered
by the following equation:

(14.304)

The last term, , exists only for even.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
772 Published: 2023-10-23
 Mass Related Information

14.14. Mass Related Information

The mass related information (mass, center of mass, and mass moments of inertia) is printed out in a
mass summary for all analyses that include mass.

Depending on the model, mass related information is calculated using one of two different methods.
If the model is three-dimensional, a precise computation is performed, as detailed in Precise Calculation
of Mass Related Information (p. 773). For all other cases, a lumped calculation is performed, as described
in Lumped Calculation of Mass Related Information (p. 775) along with its limitations.

The mass summary by element type is always based on the basic calculation.

The mass summary is calculated accumulating each element contribution. It does not reflect the
boundary conditions, coupled degrees of freedom (CP), constraint equations (CE, CERIG, RBE3), and
multipoint constraint approach in contact elements with pilot node.

The mass summary is based on unscaled mass properties (see MASCALE command).

14.14.1. Precise Calculation of Mass Related Information

The total rigid body mass properties with respect to the origin is given by:

(14.305)

where:

[Mrig] is the total rigid body mass properties of the model.

N is the number of elements.
[Di] is a matrix containing the six rigid body motion vectors of the ith element. See Equa-
tion 15.171 (p. 840) for more information about these vectors.
[Mi] is the mass matrix of the ith element.

The total rigid body mass matrix can be partitioned as follows:

(14.306)

where:

[Mt] is the translational mass matrix.

[Mtr] and [Mrt] are the coupled translational/rotational mass matrices.
[Mr] is the rotational mass matrix and contains the mass moments of inertia.

The translational mass principal characteristics are obtained from the eigensolution of matrix [Mt].
The eigenvalues are the principal masses and the eigenvectors are the mass principal directions:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 773
Analysis Tools

(14.307)

where:

MX, MY, and MZ are the principal masses.

[Φ] is the matrix of the translational mass eigenvectors representing the mass principal direc-
tions.

If the principal masses are not equal, the center of mass location with respect to the principal axes
is not unique. These locations are computed as:

(14.308)

(14.309)

(14.310)

where:

cdmX, cdmY, and cdmZ are the center of mass locations with respect to the mass principal
axes.

If the principal masses are equal, the center of mass location is unique:

(14.311)

where:

M = MX = MY = MZ is the mass of the model.

The mass moments of inertia with respect to the mass principal axes are calculated as:

(14.312)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
774 Published: 2023-10-23
 Mass Related Information

The inertia principal characteristics are obtained from the eigensolution of matrix . The eigenvalues
are the principal moments of inertia and the eigenvectors are the moment of inertia principal directions.

This precise mass calculation is not used for:

• 1D and 2D models

• Harmonic Axisymmetric models. Note that the precise mass calculation is used for General
Axisymmetric models but not for Harmonic Axisymmetric models.

• Models where UX, UY, and UZ degrees of freedom are not present at all element nodes

• Cyclic symmetry models

• Models with acoustic fluid elements (FLUID29, FLUID30, FLUID220, or FLUID221)

• The restart of a linear perturbation modal analysis.

These cases use the lumped approximation outlined in Lumped Calculation of Mass Related Informa-
tion (p. 775).

14.14.2. Lumped Calculation of Mass Related Information

The computation of the mass moments and products of inertia, as well as the model center of mass,
is described in this section. This approach assumes that the mass is lumped at the center of each
element. The model center of mass is computed as:
(14.313)

(14.314)

(14.315)

where typical terms are:

Xc = X coordinate of model center of mass (output as XC)

ρ = element density, based on average element temperature

Vi = volume of element i

{No} = vector of element shape functions, evaluated at the origin of the element coordinate
system
{Xi} = global X coordinates of the nodes of element i

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 775
Analysis Tools

The moments and products of inertia with respect to the origin are:

(14.316)

(14.317)

(14.318)

(14.319)

(14.320)

(14.321)

where typical terms are:

Ixx = mass moment of inertia about the X axis through the model center of mass (output as
IXX)
Ixy = mass product of inertia with respect to the X and Y axes through the model center of
mass (output as IXY)

Equation 14.316 (p. 776) and Equation 14.318 (p. 776) are adjusted for axisymmetric elements.

The moments and products of inertia with respect to the model center of mass (the components of
the inertia tensor) are:
(14.322)
(14.323)
(14.324)
(14.325)
(14.326)
(14.327)

where typical terms are:

= mass moment of inertia about the X axis through the model center of mass (output as
IXX)
= mass product of inertia with respect to the X and Y axes through the model center of
mass (output as IXY)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
776 Published: 2023-10-23
 Mass Related Information

[Link]. Accuracy of the Lumped Calculation

The above mass calculations are not intended to be precise for all situations, but rather have been
programmed for speed. It may be seen from the above development that only the mass (mi) and
the center of mass (Xi, Yi, and Zi) of each element are included. Effects that are not considered are:

1. The mass being different in different directions.

2. The presence of rotational inertia terms.

3. The mixture of axisymmetric elements with non-axisymmetric elements (can cause negative
moments of inertia).

4. Tapered thicknesses.

5. Offsets used with beams and shells.

6. Trapezoidal-shaped elements.

7. The generalized plane strain option of PLANE182 - 2D 4-Node Structural Solid (p. 588) and
PLANE183 - 2D 8-Node Structural Solid (p. 589). (When these are present, the center of mass and
moment calculations are completely bypassed.)

Thus, if these effects are important, a separate analysis can be performed using inertia relief to find
more precise center of mass and moments of inertia (using IRLF,-1). Inertia relief logic uses the
element mass matrices directly; however, its center of mass calculations also do not include the
effects of offsets.

It should be emphasized that the computations for displacements, stresses, reactions, etc. are correct
with none of the above approximations.

[Link]. Effect of KSUM, LSUM, ASUM, and VSUM Commands

The center of mass and mass moment of inertia calculations for keypoints, lines, areas, and volumes
(accessed by KSUM, LSUM, ASUM, VSUM, and *GET commands) use equations similar to Equa-
tion 14.313 (p. 775) through Equation 14.327 (p. 776) with the following changes:

1. Only selected solid model entities are included.

2. Lines, areas, and volumes are approximated by numerically integrating to account for rotary
inertias.

3. Keypoints are assumed to be unit masses without rotary inertia.

4. Lines are assumed to have unit mass per unit length.

5. Each area uses the thickness as:

(14.328)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 777
Analysis Tools

t = thickness

6. Each area or volume is assumed to have density as:

(14.329)

where:

ρ = density

Composite material elements presume the element material number (defined with the MAT com-
mand).

14.15. Energies
Energies are available in the solution output (by setting Item = ALL, ESOL, or VENG on the OUTPR
command) or by choosing the appropriate items (SENE, TENE, KENE, ASENE, PSENE, AKENE, PKENE,
DENE, and WEXT) via postprocessing commands.

The following energy topics are available:

14.15.1. Potential Energy (or Stiffness Energy)
14.15.2. Kinetic Energy
14.15.3. Damping Energy
14.15.4. Work Done by External Loads
14.15.5. Artificial Energy
14.15.6. Element Support for Energies

Also see Example: Energy Calculations in Transient and Harmonic Analyses in the Structural Analysis
Guide.

14.15.1. Potential Energy (or Stiffness Energy)

For each element, is the potential energy (including strain energy), also called stiffness energy,
accessed by item SENE or TENE:

(14.330)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
778 Published: 2023-10-23
 Energies

NINT = number of integration points

{σ} = stress vector
{εel} = elastic strain vector
vol i = volume of integration point i

= plastic strain energy

Es = stress stiffening energy

[Ke] = element stiffness/conductivity matrix

[Se] = element stress stiffness matrix
{ue} = element displacement vector

[Link]. Potential Energy (or Stiffness Energy) for Complex Results

Complex results (see POST1 and POST26 – Complex Results Postprocessing (p. 920)) are obtained
following a harmonic analysis (ANTYPE,HARM) or a damped modal analysis (ANTYPE,MODAL with
MODOPT,QRDAMP,,,,CPLX, or MODOPT,DAMP).

At each substep, by default, separate calculations are performed for:

• potential energy using the real part of the element stress, strain, and displacement vectors
(postprocessed with KIMG = REAL on the SET command), and

• potential energy using the imaginary part of the element stress, strain, and displacement vectors
(postprocessed with KIMG = IMAG on the SET command).

When EngCalc = YES on the HROUT command (harmonic analysis) or MXPAND command
(modal analysis), the potential energy is calculated as:
(14.331)

It is the average potential energy over the deformation cycle with period , where:

Ω = imposed circular frequency in harmonic analysis (rad/s), or imaginary part of the

complex eigenvalue in damped modal analysis
= element real displacement vector

= element imaginary displacement vector

The amplitude potential energy, accessed with item ASENE, is expressed as:
(14.332)

The peak potential energy, accessed with item PSENE, can be written as:
(14.333)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 779
Analysis Tools

Average and amplitude potential energies define a cosine wave, which is the representation of the
instantaneous potential energy. For the total system, the equation is:
(14.334)

The average potential energy of the total system is:

(14.335)

The amplitude potential energy of the total system is:

(14.336)

where the residual quantity is:

(14.337)

The phase of the cosine wave of the instantaneous potential energy of the total system is:

(14.338)

The peak potential energy of the total system is:

(14.339)

In the above equations:

N = number of elements
i = superscript representing the ith element
j = superscript representing the jth element

In the POST1 postprocessor, , , , and can be printed using the PRENERGY command.
In the POST26 postprocessor, is accessed using the ENERSOL command.

[Link]. Potential Energy Limitations

Average, amplitude, and peak potential energies are available with the following restrictions:

• These quantities are available only for the element listed in Table 14.3: Element Types Sup-
porting Additional Energy Calculations (p. 785).

• The thermal strain vector (Equation 2.3 (p. 6)) is zero.

• The plastic strain energy is zero.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
780 Published: 2023-10-23
 Energies

• For linear perturbation full harmonic and linear perturbation full modal analyses with a
nonlinear base analysis, solutions are approximate and may not be accurate.

14.15.2. Kinetic Energy

In transient, harmonic, and modal analyses, is the kinematic energy accessed by item KENE:
(14.340)

where:

= time derivative of element displacement vector (that is, element velocity vector)
[Me] = element mass matrix

In harmonic and modal analyses, the kinetic energy can also be expressed as:
(14.341)

[Link]. Kinetic Energy for Complex Results

In the case of complex results, the kinetic energy is calculated separately using the real part and
the imaginary part by default (similar to potential energy; see Potential Energy (or Stiffness Energy)
for Complex Results (p. 779)).

For the elements listed in Table 14.3: Element Types Supporting Additional Energy Calculations (p. 785),
if EngCalc = YES on the HROUT or MXPAND commands, the kinetic energy is calculated based
on:
(14.342)

This is the average kinetic energy over the deformation cycle of period . The amplitude kinetic
energy, accessed with item AKENE, is:
(14.343)

Therefore, the peak kinetic energy, accessed with item PSENE, is:
(14.344)

Average and amplitude kinetic energies define a cosine wave, which is the representation of the
instantaneous kinetic energy. For the total system, the equation is:
(14.345)

The average kinetic energy of the total system is:

(14.346)

The amplitude kinetic energy of the total system is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 781
Analysis Tools

(14.347)

where the residual quantity is:

(14.348)

The phase of the cosine wave of the instantaneous kinetic energy of the total system is:

(14.349)

The peak kinetic energy of the total system is:

(14.350)

In the POST1 postprocessor, , , and can be printed using the PRENERGY command. In
the POST26 postprocessor, is accessed using the ENERSOL command.

[Link]. Kinetic Energy Limitations

The following limiations apply:

• Average, amplitude, and peak kinetic energies are available only for the elements listed in
Table 14.3: Element Types Supporting Additional Energy Calculations (p. 785).

• When postprocessing the expansion pass of a substructuring or component mode synthesis

(CMS) analysis, kinetic energy is available only if the OUTRES command with option DSUBres
= ALL was issued in the first load step of the use pass.

14.15.3. Damping Energy

The damping energy, , is the energy dissipated by viscoelastic damping forces over the deform-
ation path:
(14.351)

where:

[Ce] = element damping matrix

In a transient analysis, if EngCalc = YES on the TRNOPT command, Equation 14.351 (p. 782) is ap-
proximated by:

(14.352)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
782 Published: 2023-10-23
 Energies

where:

NSP = Number of substeps

= element velocity vector at substep i
t i = time at substep i
t i-1 = time at substep i-1

Numerical integration done by Equation 14.352 (p. 782) is consistent only if solution data are written
to the database for every substep (OUTRES,ALL,ALL, OUTRES,ESOL,ALL, or OUTRES,VENG, ALL).

[Link]. Damping Energy for Complex Results

In the case of complex results, if EngCalc = YES on the HROUT or MXPAND command, the
damping energy is calculated based on:
(14.353)

[Link]. Damping Energy Limitations

The following limitations apply:

• Damping energy is available only for the element listed in Table 14.3: Element Types Supporting
Additional Energy Calculations (p. 785).

• Damping energy associated with the gyroscopic matrix is not available.

• Damping energy is not available when postprocessing a mode-superposition transient analysis.

• Damping energy is not available when postprocessing an expansion pass following a transient
analysis that uses the substructuring or component mode synthesis (CMS) method.

• In the case of complex results, damping energy is available when postprocessing an expansion
pass of the substructuring or component modes synthesis (CMS) method only if the OUTRES
command with option DSUBres = ALL was issued in the first load step of the use pass. In a
component modes synthesis analysis with Elcalc = NO on the CMSOPT command, damping
energy is available only if damping was introduced in the use pass with the global damping
commands: ALPHAD, BETAD, DMPRAT or DMPSTR. If Elcalc = YES on CMSOPT, damping
energy is available only if damping was introduced in the generation pass and SEMATR = 3 on
the SEOPT command.

14.15.4. Work Done by External Loads

When external loads are applied on nodes or elements, the general form of the work done by external
loads over the deformation path is:
(14.354)

where:

{u(t)} = the displacement vector of the whole system if {F(t)} is the vector of nodal loads

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 783
Analysis Tools

or
{u(t)} = the element displacement vector if {F(t)} is the vector of loads on an element

In a transient analysis, if Item = ALL, ESOL, or VENG on the OUTRES command and EngCalc = YES
on the TRNOPT command, Equation 14.354 (p. 783) is approximated with:
(14.355)

where:

NSP = number of substeps

{F i } = system nodal or element load vector at substep i

Numerical integration done by Equation 14.355 (p. 784) is consistent only if solution data are written
to the database for every substep (OUTRES,ALL,ALL, OUTRES,ESOL,ALL, or OUTRES,VENG, ALL).

[Link]. Work Done by External Loads for Complex Results

In the case of complex results, if Item = ALL, ESOL, or VENG on the OUTRES command and Eng-
Calc = YES on the HROUT or MXPAND command, at each substep the work done by external
loads is expressed as:
(14.356)

where:

{u1} = real part of the system nodal or element displacement vector

{u2} = imaginary part of the system nodal or element displacement vector
{F1} = real part of the system nodal or element load vector
{F2} = imaginary part of the system nodal or element load vector

[Link]. Work Limitations

The following limitations apply:

• Work done by external load is available only for the element listed in Table 14.3: Element Types
Supporting Additional Energy Calculations (p. 785).

• Work due to constraint displacements is not available.

• Work done by external loads is not available when postprocessing a mode-superposition transient
analysis.

• For complex results, work due to element pressure with an imaginary component is supported
only for the SF command (Lab = PRES and non-zero VALUE2) and SFE command (Lab = PRES
and KVAL = 2) applied on surface elements SURF153, SURF154 and SURF159.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
784 Published: 2023-10-23
 Energies

• Work done by external loads is not available when postprocessing a mode-superposition harmonic
analysis if MSUPkey = YES on the MXPAND command. If MSUPkey = NO, work due to element
loads is available in the expansion pass of the real part, but not the work due to nodal loads.

• Work done by external loads is not available when postprocessing an expansion pass following
a transient or harmonic analysis that uses the substructuring or component mode synthesis (CMS)
method.

14.15.5. Artificial Energy

is the artificial energy associated with hourglass control, accessed by AENE (for elements SOLID45,
PLANE182, SOLID185, SHELL181 only):

(14.357)

where:

NCS = total number of converged substeps

= hourglass strain energy defined in Flanagan and Belytschko([243] (p. 938)) due to one
point integrations
[Q] = hourglass control stiffness defined in Flanagan and Belytschko([243] (p. 938))

Artificial energy has no physical meaning. It is used to control the hourglass mode introduced by re-
duced integration. Use this rule to check whether or not the element is stable due to the use of reduced
integration: if < 5% is true, the element is considered stable (that is, it functions the same way
as a fully integrated element).

14.15.6. Element Support for Energies

The following table lists element types that support additional energy calculations (requested via
Item = ALL, ESOL, or VENG on the OUTRES command).

Table 14.3: Element Types Supporting Additional Energy Calculations

Element Type
COMBIN14
MASS21
MATRIX27
COMBIN40
MATRIX50
SURF153
SURF154
SURF156
SURF159
SHELL181

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 785
Analysis Tools

Element Type
PLANE182
PLANE183
SOLID185
SOLID186
SOLID187
BEAM188
BEAM189
SOLSH190
FOLLW201
COMBI214
COMBI250
SOLID272
SOLID273
SHELL281
SOLID285
PIPE288
PIPE289
ELBOW290

The following table list element limitations for energy computations.

Table 14.4: Exceptions for Element Energies

Element Exception
FLUID29 No potential energy
FLUID30 No potential energy
LINK31 No potential energy
LINK34 No potential energy
COMBIN39 No potential energy
SHELL61 Thermal effects not included [1]

1. For example, warping implies that temperatures T1 + T3 ≠ T2 + T4; that is, some thermal strain
is locked in.

14.16. Reduced-Order Modeling for State-Space Matrices Export

The n second order modal equations (Equation 14.141 (p. 724)) are transformed into 2n first order
equations, where n is input as NMODE on the SPMWRITE command, using the following coordinate
transformation:

(14.358)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
786 Published: 2023-10-23
 Reduced-Order Modeling for State-Space Matrices Export

The equation becomes:

(14.359)

is a (2n x 2n) state-space matrix defined by:

(14.360)

Where is the frequency of mode , is the effective modal damping of mode (see Modal Damp-
ing (p. 725)), and is the vector of input forces:

(14.361)

Where ninput is the number of scalar input forces derived from Inputs on the SPMWRITE command.

is a (2n x ninput) state-space matrix defined by:

(14.362)

With
(14.363)

Where is the matrix of eigenvectors and is a unit force matrix with size (ndof x ninput). It has
1 at the degrees of freedom where input forces are active and 0 elsewhere.

Now that the states have been expressed as a function of the input loads, the equation for the
degrees of freedom observed (outputs ) is written as:

(14.364)

is a (3*noutput x 2*n) state-space matrix, where noutput is derived from Outputs on the SPMWRITE
command, and is defined by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 787
Analysis Tools

(14.365)

with
(14.366)

is a unit displacement matrix with size (noutput x ndof ). It has 1 on degrees of freedom where
output is requested and 0 elsewhere.

is a (3*noutput x ninput) state-space matrix defined by:

(14.367)

and are included only if VelAccKey = ON on the SPMWRITE command, otherwise the last
two rows of are not written and is zero so it is not written.

14.17. Enforced Motion in Structural Analysis

In structural analysis, enforced motion is a common excitation. Examples of this behavior include the
response of a building to an earthquake, the vibration of a device carried by a vehicle, etc.

The equations of motion in terms of absolute displacements can be expressed by:

(14.368)

where:

= structural mass matrix

= structural damping matrix

= structural stiffness matrix

= nodal acceleration vector

= nodal velocity vector

= nodal displacement vector

= applied load vector

Partition the degrees of freedom into two sets:

(14.369)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
788 Published: 2023-10-23
 Enforced Motion in Structural Analysis

where:

= displacements remaining free

= displacements with enforced motion

Assuming that the only excitation source is the enforced motion, the load vector applied on degree of
freedom {u1} is zero. Equation 14.368 (p. 788) can be expanded to:

(14.370)

Where {F2} = the reaction force between the structure and its supports.

14.17.1. Full Method for Transient and Harmonic Analyses

In transient and harmonic analyses using the full method (TRNOPT,FULL and HROPT,FULL), the upper
part of Equation 14.370 (p. 789) is rearranged and solved as:
(14.371)

where

is the enforced acceleration. In a full transient analysis, it is input with the D command
and Lab = ACCX, ACCY, ACCZ, DMGX, DMGY, or DMGZ (not supported in harmonic).

is the enforced velocity. In a full transient analysis, it is input with the D command and
Lab = VELX, VELY, VELZ, OMGX, OMGY, or OMGZ (not supported in harmonic).

is the enforced displacement. In full transient and harmonic analyses, it is input with
the D command and Lab = UX, UY, UZ, ROTX, ROTY, or ROTZ.
The solution of Equation 14.371 (p. 789) is the absolute displacement vector .

14.17.2. Enforced Motion Method for Transient and Harmonic Analyses

The nodal displacement due to enforced motion can be separated into the quasi-static response and
the dynamic response, also called relative motion (see Paultre [418] (p. 948)):

(14.372)

where:

{y} is the dynamic response, representing the relative motion of the structure with respect to the base
motion.

is the quasi-static response.

By introducing Equation 14.372 (p. 789) into Equation 14.370 (p. 789), and neglecting time-derivative
terms, the following equation is obtained:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 789
Analysis Tools

(14.373)

From the upper part of Equation 14.373 (p. 790), is derived as:
(14.374)

The absolute displacement can be written as:

(14.375)

Substituting Equation 14.374 (p. 790) and Equation 14.375 (p. 790) into Equation 14.371 (p. 789) gives:
(14.376)

Neglecting the enforced velocity term on the right hand side, the equation reduces to:
(14.377)

This equation is exact in the case of stiffness-based proportional damping (BETAD) and results will
match those obtained with a full method using Equation 14.371 (p. 789). It is approximate for all other
cases of damping.

[Link]. Structure Subjected to Differential Support Motion

When a structure is subjected to excitations from different supports and/or along different axes,
Equation 14.377 (p. 790) can be solved using the mode-superposition method (TRNOPT,MSUP and
HROPT,MSUP). In that case, the degrees of freedom are first fixed for the modal analysis. The
natural frequencies and mode shapes of the conservative system without damping, derived from
Equation 14.377 (p. 790) are used to obtain the uncoupled equations, as explained in Mode-Super-
position Method (p. 721) in this guide.

If different supports are excited, the vector of displacements from enforced motion and the
associated accelerations are composed of subsets with degrees of freedom corresponding to
the different supports. Each subset, also called base, is identified using the Value argument on
the D command during the modal analysis, and for all the degrees of freedom of a subset, the
common value of enforced displacement or enforced acceleration is input using the
DVAL command in the mode-superposition transient and harmonic analyses.

In a harmonic analysis, depending on whether displacements or accelerations are input with the
DVAL command, the following relationship is used internally:

where Ω is the current forced circular frequency.

This equation defines the acceleration of supports in Equation 14.377 (p. 790) when displacements
are specified and defines displacements of supports in Equation 14.374 (p. 790) when accelerations
are specified.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
790 Published: 2023-10-23
 Enforced Motion in Structural Analysis

In a transient analysis, the Newmark time integration method described in Transient Analysis (p. 796)
is internally used to calculate the enforced displacements when enforced accelerations are specified.
Conversely, the same method is used to calculate the enforced accelerations when enforced dis-
placements are specified.

When KeyCal is ON in the DVAL command, the final displacement vector is calculated with
Equation 14.372 (p. 789) and displacements are absolute.

When KeyCal is OFF in the DVAL command, the final displacement vector is {y} and displacements
are relative.

Stresses from the quasi-static solution are not zero and may not be negligible in either of the fol-
lowing scenarios:

• If excitations act in different directions and/or intensity.

• If the quasi-static solution is calculated in the second phase of a linear perturbation modal
analysis, even if only one support is excited, because prestressing a structure modifies the
rigid body motions.

Therefore, stresses obtained with KeyCal set to ON are different from those obtained with
KeyCal set to OFF.

For more information, see Enforced Motion Method for Mode-Superposition Transient and Harmonic
Analyses in the Structural Analysis Guide.

[Link]. Structure Subjected to Global Support Acceleration

When a structure is subjected to global motion of a rigid support (see Geradin and Rixen
[365] (p. 945)), reaction force vector {F2} is zero and the quasi-static response defined in Equa-
tion 14.374 (p. 790) can be rewritten as:
(14.378)

where:

[r1] represents the rigid-body modes of the part of the structure remaining free

= the support motion, the acceleration of which is:

(14.379)

where:

γ = the global acceleration of the support, input on the ACEL command.

In this case, the coupled mass between the free and constrained degrees of freedom ([M12]) is ignored
in Equation 14.377 (p. 790) to obtain:
(14.380)

To solve Equation 14.380 (p. 791), degrees of freedom must be fixed and the full method
transient and harmonic analyses are used. The solution is the relative displacement vector {y}.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 791
Analysis Tools

As stresses of the quasi-static solution are zero, the stresses obtained with a full method working
with absolute displacements (D) or relative displacements (ACEL) are identical.

The relative solution obtained with Equation 14.380 (p. 791) matches the relative results obtained
with the enforced motion method (DVAL,,ACC,,OFF). However, the static element nodal forces and
the static reaction forces do not match as there is no external inertia load in the enforced motion
method.

14.17.3. Large Mass Method

The large mass method is an approximate technique that treats the response to acceleration excitation
as a response to external forces.

Assuming that the masses associated with subsystem are augmented so that [M22 ] becomes
, then Equation 14.370 (p. 789) becomes:

(14.381)

Solve the equations of :

(14.382)

Define and substitute Equation 14.382 (p. 792) into the upper part of Equa-
tion 14.381 (p. 792) to get:
(14.383)

where:
(14.384)

If is large enough, tends to zero, and the following approximation is verified:

(14.385)

Equation 14.383 (p. 792) will be in the same form as Equation 14.371 (p. 789), and Equation 14.381 (p. 792)
is equivalent to Equation 14.370 (p. 789).

A ratio of the large mass to the mass of the entire structure in the range of 104 to 108 keeps the
modeling error small (see Léger et al. [419] (p. 948)).

Large lumped masses can be implemented by using element type MASS21.

The solution of Equation 14.383 (p. 792) is the absolute displacement vector .

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
792 Published: 2023-10-23
Chapter 15: Analysis Procedures
This chapter presents the theoretical basis of the various analysis procedures. The derivation of the in-
dividual element matrices and load vectors is discussed in Derivation of Structural Matrices (p. 12),
Derivation of Electromagnetic Matrices (p. 176), Derivation of Heat Flow Matrices (p. 217), and Derivation
of Acoustic Matrices (p. 248).

In the matrix displacement method of analysis based upon finite element idealization, the structure
being analyzed must be approximated as an assembly of discrete regions (called elements) connected
at a finite number of points (called nodes). If the "force-displacement" relationship for each of these
discrete structural elements is known (the element "stiffness" matrix) then the "force-displacement rela-
tionship" for the entire "structure" can be assembled using standard matrix methods. These methods
are well documented (see, for example, Zienkiewicz([40] (p. 926))) and are also discussed in Analysis
Tools (p. 681). Thermal, fluid flow, and electromagnetic analyses are done on an analogous basis by re-
placing the above words in quotes with the appropriate terms. However, the terms displacement, force,
and stiffness are used frequently throughout this chapter, even though it is understood that the concepts
apply to all valid effects also.

All analysis types for iterative or transient problems automatically reuse the element matrices or the
overall structural matrix whenever it is applicable. See Reuse of Matrices (p. 424) for more details.

Analysis procedure information is available for the following analysis types:

15.1. Static Analysis
[Link] Analysis
15.3. Modal Analysis
15.4. Harmonic Analysis
15.5. Buckling Analysis
15.6. Substructuring Analysis
15.7. Spectrum Analysis
15.8. Linear Perturbation Analysis
15.9. Semi-Implicit Analysis

15.1. Static Analysis

The following static analysis topics are available:
15.1.1. Assumptions and Restrictions
15.1.2. Description of Structural Systems
15.1.3. Description of Thermal, Magnetic and Other First Order Systems

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 793
Analysis Procedures

15.1.1. Assumptions and Restrictions

The static analysis (ANTYPE,STATIC) solution method is valid for all degrees of freedom (DOFs). Inertial
and damping effects are ignored, except for static acceleration fields.

15.1.2. Description of Structural Systems

The overall equilibrium equations for linear structural static analysis are:
(15.1)

or
(15.2)

where:

= total stiffness matrix

= nodal displacement vector

= number of elements
= element stiffness matrix (described in Element Library (p. 445)) (may include the element
stress stiffness matrix (described in Stress Stiffening (p. 40)))
= reaction load vector

, the total applied load vector, is defined by:

(15.3)

where:

= applied nodal load vector

= acceleration load vector

= total mass matrix

= element mass matrix (described in Derivation of Structural Matrices (p. 12))

= total acceleration vector (defined in Acceleration Effect (p. 681))

= element thermal load vector (described in Derivation of Structural Matrices (p. 12))

= element pressure load vector (described in Derivation of Structural Matrices (p. 12))

To illustrate the load vectors in Equation 15.2 (p. 794), consider a one element column model, loaded
only by its own weight, as shown in Figure 15.1: Applied and Reaction Load Vectors (p. 795). Note that
the lower applied gravity load is applied directly to the imposed displacement, and therefore causes

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
794 Published: 2023-10-23
 Static Analysis

no strain; nevertheless, it contributes to the reaction load vector just as much as the upper applied
gravity load. Also, if the stiffness for a certain DOF is zero, any applied loads on that DOF are ignored.

Figure 15.1: Applied and Reaction Load Vectors

Solving for Unknowns and Reactions (p. 713) discusses the solution of Equation 15.2 (p. 794) and the
computation of the reaction loads. Newton-Raphson Procedure (p. 730) describes the global equation
for a nonlinear analysis. Inertia relief is discussed in Inertia Relief (p. 685).

15.1.3. Description of Thermal, Magnetic and Other First Order Systems

The overall equations for linear 1st order systems are the same as for a linear structural static analysis,
Equation 15.1 (p. 794) and Equation 15.2 (p. 794). , though, is the total coefficient matrix (for example,
the conductivity matrix in a thermal analysis) and is the nodal DOF values. , the total applied
load vector, is defined by:

(15.4)

Table 15.1: Nomenclature (p. 795) relates the nomenclature used in Derivation of Heat Flow
Matrices (p. 217) and Derivation of Electromagnetic Matrices (p. 176) for thermal, magnetic and electrical
analyses to Equation 15.2 (p. 794) and Equation 15.4 (p. 795). See Table 10.3: Nomenclature of Coefficient
Matrices (p. 325) for a more detailed nomenclature description.

Table 15.1: Nomenclature

Thermal
temperature heat flow heat flux
heat generation convection
Scalar Magnetic
scalar potential flux coercive force
Vector Magnetic
vector potential current current density and
segment coercive force
Electrical -
voltage current

Solving for Unknowns and Reactions (p. 713) discusses the solution of Equation 15.2 (p. 794) and
Newton-Raphson Procedure (p. 730) describes the global equation for a nonlinear analysis.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 795
Analysis Procedures

15.2. Transient Analysis

The following transient analysis topics are available:
15.2.1. Assumptions and Restrictions
15.2.2. Description of Structural and Other Second Order Systems
15.2.3. Description of Thermal, Magnetic and Other First Order Systems

The transient analysis solution method (ANTYPE,TRANS) used depends on the DOFs involved. Structural,
acoustic, and other second order systems (that is, the systems are second order in time) are solved using
one method and the thermal, magnetic, electrical and other first order systems are solved using another.
Each method is described subsequently. If the analysis contains both first and second order DOFs (for
example, structural and magnetic), then each DOF is solved using the appropriate method. For matrix
coupling between first and second order effects such as for piezoelectric analysis, a combined procedure
is used.

15.2.1. Assumptions and Restrictions

1. Initial conditions are known.

2. Gyroscopic or Coriolis effects are included in a structural analysis when requested (using the
CORIOLIS command).

15.2.2. Description of Structural and Other Second Order Systems

For most structural dynamics problems of a mechanical system, the spatial discretization for the
principle of virtual work using the finite element method gives the finite element semi-discrete
equation of motion as follows:
(15.5)

where:

= structural mass matrix

= structural damping matrix

= nodal acceleration vector

= nodal velocity vector

= nodal displacement vector

= internal load vector

= applied load vector

Three methods are available for solving Equation 15.5 (p. 796):

• Newmark time integration method -- Used for implicit transient analyses as described below. This
method is requested by setting TINTOPT = NMK (which is the default) on the TRNOPT command.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
796 Published: 2023-10-23
 Transient Analysis

• HHT time integration method -- Used for implicit transient analyses as described below. This method
is an extension of the Newmark time integration method and is requested by setting TINTOPT =
HHT on the TRNOPT command.

• Backward Euler time integration method -- Used for implicit transient analyses as described below.
This method can be requested by issuing the command TINTP,QUAS.

The structural dynamics problems concerned with the mechanical behavior governed by the above
differential equation can be classified into two classes; that is, linear and nonlinear problems.

[Link]. Time Integration Scheme for Linear Systems

In linear structural dynamics systems, the internal load is linearly proportional to the nodal displace-
ment, and the structural stiffness matrix remains constant. Therefore, Equation 15.5 (p. 796) can be
rewritten as:
(15.6)

where:

= structural stiffness matrix

Among direct time integration methods for numerically solving the finite element semi-discrete
equation of motion given in Equation 15.6 (p. 797), several methods such as the Newmark method
(Newmark([402] (p. 948))) and the generalized- method (Chung and Hulbert([348] (p. 944))) are in-
corporated in the program. As the generalized- method recovers the Wood-Bosak-Zienkiewicz
method (also called WBZ- method) (Wood et al.([350] (p. 944))), the Hilber-Hughes-Taylor method
(also called HHT- method) (Hilber et al.([349] (p. 944))), and the Newmark family of time integration
algorithms, the program allows you to take advantage of any of the these methods by specifying
different input parameters, as described below.

Newmark Method
The Newmark family of time integration algorithms (Newmark([402] (p. 948))) is one of the most
popular time integration methods as a single step algorithm. The semi-discrete equation of motion
given in Equation 15.6 (p. 797) can be rewritten as (Hughes([166] (p. 934))):
(15.7)

where:

= the nodal acceleration vector at time

= the nodal velocity vector at time

= the nodal displacement vector at time

= the applied load at time

In addition to Equation 15.7 (p. 797), the Newmark family of time integration algorithms requires
the displacement and velocity to be updated as follows:
(15.8)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 797
Analysis Procedures

(15.9)

where:

= Newmark's integration parameters

= nodal acceleration vector at time

= nodal velocity vector at time

= nodal displacement vector at time

Thus, the Newmark family of time integration algorithms can be determined by the Newmark integ-
ration parameters. In the end, the Newmark integration scheme consists of the three finite difference
equations presented in Equation 15.7 (p. 797) through Equation 15.9 (p. 798), and the three unknowns
, , and can be numerically calculated by the three algebraic equations along
with the three known quantities , , and .

By making use of the three algebraic equations given in Equation 15.7 (p. 797) through Equa-
tion 15.9 (p. 798), a single-step time integrator in terms of the unknown and the three known
quantities can be written as:

(15.10)

where:

First, the unknown is calculated using Equation 15.10 (p. 798). Then, the program computes
the two unknowns and by using the following equations:
(15.11)
(15.12)

The most important factors in choosing an appropriate time integration scheme for the finite element
semi-discrete equation of motion given in Equation 15.5 (p. 796) are accuracy, stability, and dissipation.
In conditionally stable time integration algorithms, stability is affected by a chosen size of the time

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
798 Published: 2023-10-23
 Transient Analysis

step; whereas in unconditionally stable time integration algorithms, a time step size can be chosen
independent of stability considerations.

In the Newmark method, the amount of numerical algorithm dissipation can be controlled by one
of Newmark's parameters, , as follows:

(15.13)

With the Newmark parameters satisfying the above conditions, the Newmark family of methods
may be unconditionally stable (Hughes([166] (p. 934))). By introducing the amplitude decay factor
, the above conditions can be written:

(15.14)

Consequently, the program provides the user with the Newmark integration procedure, which is
unconditionally stable via input of the amplitude decay factor on the TINTP command. Alternat-
ively, the and parameters may be input directly using the TINTP command.

Generalized HHT- Method

In the Newmark method, the amount of numerical dissipation can be controlled by one parameter
in Equation 15.13 (p. 799) or in Equation 15.14 (p. 799). However, in low frequency modes the
Newmark method fails to retain the second-order accuracy as . Note that the Newmark implicit
method (constant average method; namely, and ), which is unconditionally stable and
second-order accurate, has no numerical damping. If other sources of numerical damping are not
introduced, the lack of numerical damping can be undesirable so that the higher frequencies of
the structure can produce unacceptable levels of numerical noise (Hughes([166] (p. 934))).

To circumvent the drawbacks of the Newmark family of methods, the program implements the
generalized HHT- method which sufficiently damps out spurious high-frequency response via in-
troducing controllable numerical dissipation in higher frequency modes, while maintaining the
second-order accuracy. It should be noted that the generalized HHT- method incorporated in the
program is capable of recovering the WBZ- method (Wood et al.([350] (p. 944))) and the HHT-
method (Hilber et al.([349] (p. 944))) as well as the Newmark family of time integration algorithms,
depending upon the user's input on the TINTP command.

To solve for the three unknowns , , and , along with Equation 15.8 (p. 797) and
Equation 15.9 (p. 798) the generalized HHT- method uses the algebraic equation:
(15.15)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 799
Analysis Procedures

Equation 15.15 (p. 799) give the finite difference form:

(15.16)

where:

Analogous to the Newmark method, the generalized HHT- method calculates the unknown
at time by making use of Equation 15.16 (p. 800). Then, the program computes the two unknowns
and by using the equations given in Equation 15.11 (p. 798) and Equation 15.12 (p. 798).
Since the generalized HHT- method is also an implicit time scheme, the structural stiffness matrix
must be factorized to solve for at time .

As mentioned in the literature (Chung and Hulbert([348] (p. 944))), the generalized HHT- method
is unconditionally stable and second-order accurate if the parameters meet the following conditions:

(15.17)

where (Wood et al.([350] (p. 944))) and (Hilber et al.([349] (p. 944))). For the generalized
HHT- method, you can input the four parameters on the TINTP command. By introducing the
amplitude decay factor , the program also allows the user to control the amount of numerical
damping if the four parameters on the TINTP command meet the following conditions:

(15.18)

If the WBZ- method is desired, the user can control the amount of numerical damping if the four
parameters on the TINTP command meet the following conditions:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
800 Published: 2023-10-23
 Transient Analysis

(15.19)

Finally, the program also allows a user who wants to use the generalized HHT- method to control
the amount of numerical damping if the four parameters on the TINTP command meet the following
conditions:

(15.20)

It should be noted that the generalized HHT- method is second-order accurate and unconditionally
stable. This method allows you to control the amount of numerical damping. The amplitude decay
factor is recommended to be set as (Hughes([166] (p. 934))), with which any spurious parti-
cipation of the higher modes can be damped out and the lower modes are not affected. A significant
amount of numerical damping may be introduced by setting , but it is not recommended.

Backward Euler Method

The backward Euler method requires displacement and velocity to be updated as follows:

(15.21)

(15.22)

Combining equations Equation 15.21 (p. 801) and Equation 15.22 (p. 801) with the semi-discrete
equation of motion given in Equation 15.7 (p. 797) results in a single-step time integrator in terms
of the unknown {un+1} as:

(15.23)

[Link]. Time Integration Scheme for Nonlinear Systems

In nonlinear structural dynamics problems, the internal load is no longer linearly proportional to
the nodal displacement, and the structural stiffness matrix is dependent on the current displacement.
Therefore, Instead of Equation 15.6 (p. 797), any time integration scheme should be applied to the
nonlinear semi-discrete equation:
(15.24)

Equation 15.24 (p. 801) represents a nonlinear system of simultaneous algebraic equations; hence,
any time integration operator may be used in association with the Newton-Raphson iterative al-
gorithm. For nonlinear structural dynamics problems, both the Newmark method and the generalized
HHT- method are incorporated in the program.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 801
Analysis Procedures

Newmark Method
The Newmark method assumes that at the time , the semi-discrete equation of motion given
in Equation 15.24 (p. 801) can be rewritten as:
(15.25)

Note that is dependent on the current displacement at time . In addition

to Equation 15.25 (p. 802), the Newmark family of time integration algorithms requires the displace-
ment and velocity to be updated as given in Equation 15.8 (p. 797) and Equation 15.9 (p. 798).

By introducing the residual vector , Equation 15.25 (p. 802) can be written as:
(15.26)

It is important to note that the time integration operator given in either Equation 15.25 (p. 802) or
Equation 15.26 (p. 802) represents a nonlinear system of simultaneous algebraic equations. Therefore,
a linearized form of the time integration operator can be obtained by the Newton-Raphson method
as follows:

(15.27)

where:

= the estimate of at the kth iteration

= the displacement increment of at the kth iteration

Equation 15.27 (p. 802) gives:

(15.28)

where:

= the tangent stiffness matrix at time

For nonlinear structural dynamics problems, the program allows a user to input the amplitude decay
factor or the Newmark integration parameters on the TINTP command.

Generalized HHT- Method

The generalized HHT- method for nonlinear structural dynamics problems assumes:
(15.29)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
802 Published: 2023-10-23
 Transient Analysis

By introducing the residual vector , Equation 15.29 (p. 802) can be written as:
(15.30)

The time integration operator given in Equation 15.29 (p. 802) or Equation 15.30 (p. 803) also represents
a nonlinear system of simultaneous algebraic equations. Therefore, a linearized form of the time
integration operator can be obtained by the Newton-Raphson method as follows:

(15.31)

where:

Equation 15.31 (p. 803) gives:

(15.32)

where:

Backward Euler Method

Combining the velocity and acceleration updates given in equations Equation 15.21 (p. 801) and
Equation 15.22 (p. 801) with the semi-discrete equation of motion given in Equation 15.27 (p. 802)
gives:
(15.33)

Equation 15.33 (p. 803) is solved by the Newton-Raphson method. Upon convergence, the displace-
ment is given at substep n+1, {un+1}.

[Link]. Solution
Two methods of solution for the Newmark method (Equation 15.10 (p. 798)) are available: full (p. 804)
and mode-superposition (p. 807) (TRNOPT command). Each are described subsequently. Only the
full solution method is available for HHT (Equation 15.15 (p. 799)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 803
Analysis Procedures

[Link].1. Full Solution Method

The full solution method (TRNOPT,FULL) solves Equation 15.10 (p. 798) directly and makes no
additional assumptions. In a nonlinear analysis, the Newton-Raphson method (Newton-Raphson
Procedure (p. 730)) is employed along with the Newmark assumptions. Automatic Time-Step-
ping (p. 708) discusses the procedure for the program to automatically determine the time step
size required for each time step.

Inherent to the Newmark method is that the values of , , and at the start of the
transient must be known. Nonzero initial conditions are input either directly (with the IC com-
mands) or by performing a static analysis load step (or load steps) prior to the start of the transient
itself. Static load steps are performed in a transient analysis by turning off the transient time in-
tegration effects (with the TIMINT,OFF command). The transient itself can then be started (by
TIMINT,ON). The default with transient analysis (ANTYPE,TRANS) is for the transient to be running
(TIMINT,ON); that is, to start the transient immediately. This implies . The initial
conditions are outlined in the subsequent paragraphs. Cases referring to "no previous load step"
mean that the first load step is transient.

Initial Displacement

The initial displacements are:

(15.34)

where:

= vector of initial displacements

= displacement vector specified by the initial conditions (IC command)

= displacement vector resulting from a static analysis (TIMINT,OFF) of the pre-

vious load step

Initial Velocity

The initial velocities are:

(15.35)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
804 Published: 2023-10-23
 Transient Analysis

where:

= vector of initial velocities

= vector of velocities specified by the initial conditions (IC commands)

= displacements from a static analysis (TIMINT,OFF) of the previous load step

= displacement corresponding to the time point before solution.

is if is the first solution of the analysis (i.e. load step 1 substep 1).
= time increment between and

If the previous load step was run as a static analysis (TIMINT,OFF), initial velocities are calcu-
lated using the previous two displacements and the previous time increment. Using either a
single substep (NSUBST,1) or ramped loading (KBC,0) within the previous load step will result
in nonzero initial velocities (assuming nonzero displacement), as shown in Figure 15.2: Effect
of Number of Substeps (NSUBST) and Ramping (KBC) on Initial Velocity for TIMINT,OFF (p. 805).
Zero initial velocities may be obtained by using multiple substeps (NSUBST,>1) and stepped
loading (KBC,1).

Figure 15.2: Effect of Number of Substeps (NSUBST) and Ramping (KBC) on Initial Velocity
for TIMINT,OFF

Initial Acceleration

The initial acceleration is simply:

(15.36)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 805
Analysis Procedures

= vector of initial accelerations

= vector of accelerations specified by the initial conditions (IC command).

An option is available to calculate initial acceleration that is consistent with the forces applied
during the start of the transient analysis. When activated by the InitialAcc field of the
TRNOPT command, the initial acceleration is calculated as a function of applied force, inertia
force, and velocity at the start time as:
(15.37)

Nodal and Reaction Load

Inertia, damping and static loads on the nodes of each element are computed.

The inertial load part of the element output is computed by:

(15.38)

where:

= vector of element inertial forces

= element mass matrix

= element acceleration vector

The acceleration of a typical DOF is given by Equation 15.12 (p. 798) for time . By default,
the acceleration vector is the average acceleration between time and time , since
the Newmark assumptions (Equation 15.8 (p. 797) and Equation 15.9 (p. 798)) assume the average
acceleration represents the true acceleration. Smoothing can be suppressed using the TINTP
command (AVSMOOTH option).

The damping load part of the element output is computed by:

(15.39)

where:

= vector of element damping forces

= element damping matrix

= element velocity vector

The velocity of a typical DOF is given by Equation 15.11 (p. 798).

The static load is part of the element output computed in the same way as in a static analysis
(Solving for Unknowns and Reactions (p. 713)). The nodal reaction loads are computed as the
negative of the sum of all three types of loads (inertia, damping, and static) over all elements
connected to a given fixed displacement node.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
806 Published: 2023-10-23
 Transient Analysis

[Link].2. Mode-Superposition Method

The mode-superposition method (TRNOPT,MSUP) uses the natural frequencies and mode shapes
of a linear structure to predict the response to transient loads. This solution method imposes the
following additional assumptions and restrictions:

1. Constant and matrices. This implies no large deflections or change of stress stiffening,
as well as no plasticity, creep, or swelling.

2. Constant time step size.

3. There are no element damping matrices. However, various types of system damping are
available.

4. Time varying imposed displacements are not allowed.

The development of the general mode-superposition procedure is described in Mode-Superpos-

ition Method (p. 721). Equation 14.141 (p. 724) and Equation 14.142 (p. 724) are integrated through
time for each mode by the Newmark method.

The initial value of the modal coordinates at time = 0.0 are computed by solving Equa-
tion 14.141 (p. 724) with and assumed to be zero.

(15.40)

where:

= the forces applied at time = 0.0

Expansion Pass

The expansion pass of the mode-superposition transient analysis involves computing element
stresses, element nodal forces, reaction forces, etc.

15.2.3. Description of Thermal, Magnetic and Other First Order Systems

The governing equation of interest is as follows:
(15.41)

where:

= damping matrix

= coefficient matrix

= vector of DOF values

= time rate of the DOF values

= applied load vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 807
Analysis Procedures

In a thermal analysis, is the specific heat matrix, the conductivity matrix, the vector of
nodal temperatures and the applied heat flows. Table 15.2: Nomenclature (p. 808) relates the
nomenclature used in Derivation of Heat Flow Matrices (p. 217) and Derivation of Electromagnetic
Matrices (p. 176) for thermal, magnetic and electrical analyses to Equation 15.41 (p. 807).

Table 15.2: Nomenclature

Thermal
temperature heat flow
Scalar
scalar flux
Magnetic
potential
Vector
vector current
Magnetic
potential segment
Electrical
voltage current

The mode-superposition procedure does not apply to first order systems.

The procedure employed for the solution of Equation 15.41 (p. 807) is the generalized trapezoidal rule
(Hughes([166] (p. 934))):
(15.42)

where:

= transient integration parameter (input as THETA on the TINTP command)

= nodal DOF values at time

= time rate of the nodal DOF values at time (computed at previous time step)

Equation 15.41 (p. 807) can be written at time as:

(15.43)

Substituting from Equation 15.42 (p. 808) into this equation yields:
(15.44)

The solution of Equation 15.44 (p. 808) employs the same solvers used for static analysis in Static
Analysis (p. 793). Once is obtained, is updated using Equation 15.42 (p. 808). In a nonlinear
analysis, the Newton-Raphson method (Newton-Raphson Procedure (p. 730)) is employed along with
the generalized trapezoidal assumption, Equation 15.42 (p. 808).

The transient integration parameter (THETA on the TINTP command) defaults to 1.0 (backward
Euler method). For all , the system equations that follow are said to be implicit. In addition, for
the more limiting case of , the solution of these equations is said to be unconditionally stable;

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
808 Published: 2023-10-23
 Transient Analysis

that is, stability is not a factor in time step ( ) selection. The available range of (using TINTP
command) is therefore limited to
(15.45)

which corresponds to an unconditionally stable, implicit method. is also good at damping out
numerical oscillations that may result when is used.

Since influences , sudden changes in loading need to be handled carefully for values of
. See the Basic Analysis Guide for more details.

The generalized-trapezoidal method requires that the values of and at the start of the
transient must be known. Nonzero initial conditions are input either directly (with the IC command)
(for ) or by performing a static analysis load step (or load steps) prior to the start of the transient
itself. Static load steps are performed in a transient analysis by turning off the transient time integration
effects (with the TIMINT,OFF command). The transient itself can then be started (TIMINT,ON). The
default for transient analysis (ANTYPE,TRANS) is to start the transient immediately (TIMINT,ON). This
implies . The initial conditions are outlined in Initial DOF Values (p. 809).

Coupled-Physics
Some transient coupled-physics analyses involve both a first order system of transient equations and
a second order system of transient equations; for example, a coupled thermal (1st order) – structural
(2nd order) analysis. If the HHT time integration method is chosen in this type of analysis, the first
order system of equations also follows the averaging of the forces used in the HHT time integration
(described in Equation 15.15 (p. 799)). More specifically, the governing equation for the first order
system given in Equation 15.41 (p. 807) is modified to:
(15.46)

where quantities with subscript follow the definitions described in Generalized HHT-α
Method (p. 799).

Both the Newmark and HHT methods used to solve second order equations described in Description
of Structural and Other Second Order Systems (p. 796) are second order accurate. On the other hand,
the backward Euler method ( ) used by default to solve the first order equations is only first
order accurate. If second order accuracy is desired for a coupled physics system such as a piezoelectric
analysis, you can request the Crank-Nicolson method ( ) for the first order systems and constant
average acceleration ( and ) by setting THETA, ALPHA, and DELTA on the TINTP
command.

Initial DOF Values

The initial DOF values for first order systems are:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 809
Analysis Procedures

(15.47)

where:

= vector of initial DOF values

= vector of uniform DOF values

= DOF vector directly specified (IC command)

= DOF vector resulting from a static analysis (TIMINT,OFF) of the previous load step
available

is set to TEMP (BFUNIF command) and/or to the temperature specified by the initial conditions
(IC commands) for thermal DOFs (temperatures) and zero for other DOFs.

Nodal and Reaction Load Computation

Damping and static loads on the nodes of each element are computed.

The damping load part of the element output is computed by:

(15.48)

where:

= vector of element damping loads

= element damping matrix

= element velocity vector

The velocity of a typical DOF is given by Equation 15.42 (p. 808). The velocity vector is the average
velocity between time and time , since the general trapezoidal rule (Equation 15.42 (p. 808))
assumes the average velocity represents the true velocity.

The static load is part of the element output computed in the same way as in a static analysis (Solving
for Unknowns and Reactions (p. 713)). The nodal reaction loads are computed as the negative of the
sum of both types of loads (damping and static) over all elements connected to a given fixed DOF
node.

15.3. Modal Analysis

The following modal analysis topics are available:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
810 Published: 2023-10-23
 Modal Analysis

15.3.1. Assumptions and Restrictions

15.3.2. Description of Analysis for Symmetric Undamped Systems
15.3.3. Participation Factors
15.3.4. Effective Mass and Cumulative Mass Fraction
15.3.5. Modal Mass and Kinetic Energy

15.3.1. Assumptions and Restrictions

1. Valid for structural and fluid degrees of freedom (DOFs). Electrical and thermal DOFs may be
present in the coupled field mode-frequency analysis using structural DOFs.

2. The structure has constant stiffness and mass effects.

3. There is no damping, unless the damped eigensolver (MODOPT,DAMP or MODOPT,QRDAMP) is

selected.

4. The structure has no time varying forces, displacements, pressures, or temperatures applied (free
vibration).

15.3.2. Description of Analysis for Symmetric Undamped Systems

This analysis type (accessed with ANTYPE,MODAL) is used for natural frequency and mode shape
determination. The equation of motion for an undamped system, expressed in matrix notation using
the above assumptions is:
(15.49)

Note that , the structure stiffness matrix, may include prestress effects (PSTRES,ON). For a discussion
of the damped eigensolver (MODOPT,DAMP or MODOPT,QRDAMP) see Eigenvalue and Eigenvector
Extraction (p. 745).

For a linear system, free vibrations will be harmonic of the form:

(15.50)

where:

= eigenvector representing the mode shape of the ith natural frequency

= ith natural circular frequency (radians per unit time)

= time

Thus, Equation 15.49 (p. 811) becomes:

(15.51)

This equality is satisfied if either or if the determinant of is zero. The first

option is the trivial one and, therefore, is not of interest. Thus, the second one gives the solution:
(15.52)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 811
Analysis Procedures

This is an eigenvalue problem which may be solved for up to n values of and n eigenvectors
which satisfy Equation 15.51 (p. 811) where n is the number of DOFs. The eigenvalue and eigenvector
extraction techniques are discussed in Eigenvalue and Eigenvector Extraction (p. 745).

Rather than outputting the natural circular frequencies , the natural frequencies ( ) are output;
where:
(15.53)

where:

= ith natural frequency (cycles per unit time)

If normalization of each eigenvector to the mass matrix is selected (MODOPT,,,,,,OFF):

(15.54)

If normalization of each eigenvector to 1.0 is selected (MODOPT,,,,,,ON), is normalized

such that its largest component is 1.0 (unity).

15.3.3. Participation Factors

The participation factors for a given direction are defined as:
(15.55)

where:

= participation factor for the ith mode

= eigenvector normalized using Equation 15.54 (p. 812)

= vector describing the direction (see Equation 15.171 (p. 840))

The participation factors are independent of the normalization method (Nrmkey on the MODOPT
command has no effect).

If the unsymmetric eigensolver is used (MODOPT,UNSYM), the left eigenvectors, if available, are used
to calculate the participation factors.

15.3.4. Effective Mass and Cumulative Mass Fraction

Details about effective mass and cumulative mass fraction can be found in Effective Mass and Cumu-
lative Mass Fraction (p. 847).

15.3.5. Modal Mass and Kinetic Energy

The modal mass (also called generalized mass) is expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
812 Published: 2023-10-23
 Harmonic Analysis

(15.56)

where:

= modal mass for the ith mode

It is equal to 1 if modes are mass normalized.

The modal kinetic energy is calculated with:

(15.57)

where:

= total kinetic energy for the ith mode

15.4. Harmonic Analysis

The following harmonic analysis topics are available:
15.4.1. Harmonic Analysis Assumptions and Restrictions
15.4.2. Description of Harmonic Analysis
15.4.3. Harmonic Analysis Complex Displacement Output
15.4.4. Nodal and Reaction Load Computation in a Harmonic Analysis
15.4.5. Harmonic Analysis Solution
15.4.6. Automatic Frequency Spacing in a Harmonic Analysis
15.4.7. Logarithm Frequency Spacing in a Harmonic Analysis
15.4.8. Harmonic Analysis with Rotating Forces on Rotating Structures
15.4.9. Harmonic Ocean Wave Procedure (HOWP)

The harmonic analysis (ANTYPE,HARMIC) solves the time-dependent equations of motion (Equa-
tion 15.5 (p. 796)) for linear structures undergoing steady-state vibration.

15.4.1. Harmonic Analysis Assumptions and Restrictions

1. Valid for structural, fluid, magnetic, and electrical degrees of freedom (DOFs). Thermal DOFs may
be present in a coupled field harmonic analysis using structural DOFs.

2. The entire structure has constant or frequency-dependent stiffness, damping, and mass effects.

3. All loads and displacements vary sinusoidally at the same known frequency (although not neces-
sarily in phase).

4. Element loads are assumed to be real (in-phase) only, except for:

a. current density

b. pressures in SURF153, SURF154, SURF156, and SURF159

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 813
Analysis Procedures

15.4.2. Description of Harmonic Analysis

Consider the general equation of motion for a structural system (Equation 15.5 (p. 796)).
(15.58)

where:

= structural mass matrix

= structural damping matrix

= structural stiffness matrix

= nodal acceleration vector

= nodal velocity vector

= nodal displacement vector

= applied load vector

As stated above, all points in the structure are moving at the same known frequency, however, not
necessarily in phase. Also, it is known that the presence of damping causes phase shifts. Therefore,
the displacements may be defined as:
(15.59)

where:

= maximum displacement

= imposed circular frequency (radians/time) =

= imposed frequency (cycles/time) (input as FREQB and FREQE on the HARFRQ command)
= time
= displacement phase shift (radians)

Note that and may be different at each DOF. The use of complex notation allows a compact
and efficient description and solution of the problem. Equation 15.59 (p. 814) can be rewritten as:
(15.60)

or as:
(15.61)

where:

= real displacement vector (input as VALUE on D command, when spe-

cified)
= imaginary displacement vector (input as VALUE2 on D command, when
specified)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
814 Published: 2023-10-23
 Harmonic Analysis

The force vector can be specified analogously to the displacement:

(15.62)
(15.63)
(15.64)

where:

= force amplitude
= force phase shift (radians)

= real force vector (input as VALUE on F command, when specified)

= imaginary force vector (input as on VALUE2 on F command, when

specified)

Substituting Equation 15.61 (p. 814) and Equation 15.64 (p. 815) into Equation 15.58 (p. 814) gives:
(15.65)

The dependence on time ( ) is the same on both sides of the equation and may therefore be re-
moved:
(15.66)

The solution of this equation is discussed later.

15.4.3. Harmonic Analysis Complex Displacement Output

The complex displacement output at each DOF may be given in one of two forms:

1. The same form as and as defined in Equation 15.61 (p. 814) (selected with the command
HROUT,ON).

2. The form and (amplitude and phase angle (in degrees)), as defined in Equation 15.60 (p. 814)
(selected with the command HROUT,OFF). These two terms are computed at each DOF as:

(15.67)

(15.68)

Note that the response lags the excitation by a phase angle of .

15.4.4. Nodal and Reaction Load Computation in a Harmonic Analysis

Inertia, damping and static loads on the nodes of each element are computed.

The inertia load is expressed as:

(15.69)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 815
Analysis Procedures

= vector of element inertia forces

= element mass matrix

= element displacement vector

The real and imaginary inertia load parts of the element output are then computed by:
(15.70)
(15.71)

where:

= vector of element inertia forces (real part)

= element real displacement vector

= vector element inertia forces (imaginary part)

= element imaginary displacement vector

The damping load is expressed as:

(15.72)

where:

= vector of element damping forces

= element damping matrix

The real and imaginary damping loads part of the element output are then computed by:
(15.73)
(15.74)

where:

= vector of element damping forces (real part)

= vector of element damping forces (imaginary part)

The real static load is computed the same way as in a static analysis (Solving for Unknowns and Reac-
tions (p. 713)) using the real part of the displacement solution . The imaginary static load is
computed also the same way, using the imaginary part . Note that the imaginary part of the
element loads (e.g., ) are normally zero, except for current density loads.

The nodal reaction loads are computed as the sum of all three types of loads (inertia, damping, and
static) over all elements connected to a given fixed displacement node.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
816 Published: 2023-10-23
 Harmonic Analysis

15.4.5. Harmonic Analysis Solution

Three methods of solution to Equation 15.66 (p. 815) are available:
[Link]. Full Solution Method
[Link]. Frequency-Sweep Method
[Link]. Mode-Superposition Method

[Link]. Full Solution Method

The full solution method (HROPT,FULL) solves Equation 15.66 (p. 815) directly. Equation 15.66 (p. 815)
may be expressed as:
(15.75)

where denotes a complex matrix or vector. Equation 15.75 (p. 817) is solved using the same sparse
solver used for a static analysis in Equation Solvers (p. 717), except that it is done using complex
arithmetic.

[Link]. Frequency-Sweep Method

A common way to compute the harmonic response of a structure is to compute the normal modes
of the undamped structure, and to use a modal superposition method to evaluate the response,
after determining the modal damping. Determining the modal damping can be based on modal
testing, or by using empirical rules. However, when the structure is non-metallic, the elastic properties
can be highly dependent on the frequency, and the damping can be high enough that the undamped
modes and the damped modes are significantly different, and an approach based on the real, un-
damped modes is not appropriate.

One alternative to straight modal analysis is to build multiple modal bases, for different property
values, and combine them together over the frequency range of the analysis. This technique is
complex, error prone, and does not address the problem of determining the modal damping factors.
Another alternative is a direct frequency response, updating the elastic properties for every frequency
step. This technique gives a much better prediction of the frequency response, but is CPU intensive.

The frequency-sweep method (HROPT,VT) for harmonic analysis, which uses the underlying Vari-
ational Technology method, is a another alternative. With this method, you can define frequency-
dependent material elastic properties (using TB,ELASTIC and TB,SDAMP) and efficiently compute
the frequency response over an entire frequency range. For the underlying theory, see Guillaume
[333] (p. 944) and Beley, Broudiscou, et al. [357] (p. 945).

[Link].1. Viscous or Hysteretic Damping

When using the frequency-sweep method, the user has a choice between viscous and hysteretic
damping.

Viscous Damping

Consider a spring-damper-mass system subjected to a harmonic excitation. The response of the

system is given by:
(15.76)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 817
Analysis Procedures

Due to the damping, the system is not conservative and the energy is dissipated. Using viscous
damping, the energy dissipated by the cycle is proportional to the frequency, . In a single DOF
spring-mass-damper system, with a viscous damper :
(15.77)

where:

= change of energy
= viscous damper

Hysteretic Damping

Experience shows that energy dissipated by internal friction in a real system does not depend
on frequency, and approximately is a function of :
(15.78)

where:

= frequency-dependent damping (input using TB,SDAMP command)

damping is known as structural or hysteretic damping. It can be included in the elastic prop-
erties by using a complex Young's modulus:
(15.79)

where:

= Young's modulus (input using TB,ELASTIC command)

Using this kind of representation, the equations of motion of the system become:
(15.80)

where:

= structural mass matrix

= structural stiffness matrix

= structural damping matrix

= nodal acceleration vector

= nodal displacement vector

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
818 Published: 2023-10-23
 Harmonic Analysis

= applied load vector

Note:

The frequency-sweep method does not support a material with both viscous and
hysteretic damping properties in the same model. In the case, please use the
HROPT,FULL command to analyze the material.

[Link]. Mode-Superposition Method

The mode-superposition method (HROPT,MSUP) uses the natural frequencies and mode shapes to
compute the response to a sinusoidally varying forcing load. This solution method imposes the
following additional assumptions and restrictions:

1. Nonzero imposed harmonic displacements are not allowed.

2. There are no element damping matrices. However, various types of system damping are available.

The development of the general mode-superposition procedure is given in Mode-Superposition

Method (p. 721). The equation of motion (Equation 15.58 (p. 814)) is converted to modal form, as
described in Mode-Superposition Method (p. 721). Equation 14.141 (p. 724) is:
(15.81)

where:

= modal coordinate

= natural circular frequency of mode j

= fraction of critical damping for mode j

= force in modal coordinates

The load vector which is converted to modal coordinates (Equation 14.140 (p. 724)) is given by
(15.82)

where:

= nodal force vector

= load vector scale factor, (input as FACT, LVSCALE command)
= load vector from the modal analysis (see Mode-Superposition Method (p. 721)).

For a steady sinusoidal vibration, has the form

(15.83)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 819
Analysis Procedures

= complex force amplitude

= imposed circular frequency

For Equation 15.81 (p. 819) to be true at all times, must have a similar form as , or
(15.84)

where:

= complex amplitude of the modal coordinate for mode j.

Differentiating Equation 15.84 (p. 820), and substituting Equation 15.83 (p. 819) and Equa-
tion 15.84 (p. 820) into Equation 15.81 (p. 819),
(15.85)

Collecting coefficients of and dividing by ( )

(15.86)

solving for ,

(15.87)

The contribution from each mode is:

(15.88)

where:

= contribution of mode j (output if Mcont = ON, on the HROUT command)

= mode shape for mode j

Finally, the complex displacements are obtained from Equation 14.123 (p. 722) as

(15.89)

where:

= vector of complex displacements

In the case of the QR damped mode extraction method, one substitutes Equation 14.142 (p. 724)
for Equation 14.141 (p. 724), so Equation 15.86 (p. 820) becomes:
(15.90)

Solving the above equation and multiplying by the eigenvectors, one can calculate the complex
displacements shown in Equation 15.89 (p. 820).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
820 Published: 2023-10-23
 Harmonic Analysis

[Link].1. Expansion Pass

The expansion pass of the mode-superposition harmonic analysis involves computing element
stresses, element nodal forces, reaction forces, etc.

15.4.6. Automatic Frequency Spacing in a Harmonic Analysis

In a harmonic analysis, the imposed frequencies that involve the most kinetic energy are those near
the natural frequencies of the structure. The automatic frequency spacing or "cluster" option (Clust
= ON, on the HROUT command) provides an approximate method of choosing suitable imposed
frequencies. The nearness of the imposed frequencies to the natural frequencies depends on damping,
because the resonance peaks narrow when the damping is reduced. Figure 15.3: Frequency Spa-
cing (p. 822) shows two typical resonance peaks and the imposed frequencies chosen by this method,
which are computed from:
(15.91)

(15.92)

Equation 15.91 (p. 821) gives frequencies slightly less than the natural circular frequency . Equa-
tion 15.92 (p. 821) gives slightly higher frequencies. The spacing parameter is defined as:
(15.93)

where:

= modal damping. (If is computed as 0.0, it is redefined to be 0.005 for this equation only).

= integer constant (input as NSBSTP, NSUBST command) which may be between 2 and 20.
Anything above this range defaults to 10 and anything below this range defaults to 4.
= 1, 2, 3, ... N

Each natural frequency, as well as frequencies midway between, are also chosen as imposed frequen-
cies.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 821
Analysis Procedures

Figure 15.3: Frequency Spacing

15.4.7. Logarithm Frequency Spacing in a Harmonic Analysis

The logarithm frequency spacing can be defined over an nth-octave band or a general-frequency
band in a harmonic analysis (the HARFRQ command with LogOpt).

The octave band is defined by the lower-end frequency:

(15.94)

and upper-end frequency:

(15.95)

where the central frequency of the octave band is specified at 16, 31.6, 63, 125, 250, 500, 1000,
2000, 4000, 8000, 16000 Hz for 11 octave bands, and n = 1 (octave band), 2 (1/2 octave band), 3 (1/3
octave band), 6 (1/6 octave band), 12 (1/12 octave band), and 24 (1/24 octave band).

The recursive relationships between the central frequencies of the specified octave bands are:
(15.96)

(15.97)

The logarithm frequency increment is given by:

(15.98)

Where NSBSTP is input by the NSUBST command.

The imposed working frequency is calculated by:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
822 Published: 2023-10-23
 Harmonic Analysis

(15.99)

The imposed working frequency is , if NSBSTP = 1. For a general logarithm frequency span, the
beginning frequency and ending frequency are defined by the FREQB and FREQE parameters
of the HARFRQ command respectively.

15.4.8. Harmonic Analysis with Rotating Forces on Rotating Structures

If a structure is rotating, forces rotating synchronously or asynchronously with the structure are of
interest.

General rotating asynchronous forces in a stationary reference frame are described in General Asyn-
chronous Rotating Force (p. 823). A specific synchronous force: mass unbalance is shown in Specific
Synchronous Forces: Mass Unbalance (p. 824).

In both cases, the equation solved for harmonic analysis is the same as (Equation 15.66 (p. 815)) except
for the coefficients of the damping matrix which will be a function of the rotational velocity of
the structure (see the CORIOLIS command). will be updated for each excitation frequency step
using the following rotational velocity:
(15.100)

where:

= rotational velocity of the structure (rd/s)

= frequency of excitation (rd/s)
= ratio between and (s = 1 for synchronous excitations) (input as RATIO on the
SYNCHRO command).

The right-hand term of the equation is given below depending on the force considered.

All developments below apply to equations expressed in a stationary reference frame. See Rotating
Structures (p. 700) for more information about reference frames. In a rotating reference frame, a gen-
eral rotating asynchronous force is the same as in a stationary reference frame except it is calculated
replacing s by s’ with:

where:

= rotating reference frame ratio ( = 0 for synchronous excitations) (input as RATIO on

the SYNCHRO command).

As a consequence, in this case, a mass unbalance force in a rotating reference frame is a static force.

[Link]. General Asynchronous Rotating Force

If the structure is rotating about X axis, then an asynchronous force having its direction in the plane
perpendicular to the spin axis is expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 823
Analysis Procedures

(15.101)
(15.102)

where:

= amplitude of force

Using complex notations, the equations become:

(15.103)
(15.104)

where:

= real force in Y-direction; also, negative of imaginary force in Z-direction. Input

as VALUE on F, label FY; input as VALUE2 on F, label FZ
= negative of real force in Z-direction; also, negative of imaginary force in Y-
direction. Input as VALUE on F, label FZ; input as VALUE2 on F, label FY

The expression of the forces for structures rotating about another direction than X are developed
analogously.

[Link]. Specific Synchronous Forces: Mass Unbalance

Consider a structure rotating about the X axis. The mass unbalance m situated at node I with the
eccentricity e may be represented as shown in Figure 15.4: Mass Unbalance at Node I (p. 824)

Figure 15.4: Mass Unbalance at Node I

If we only consider the motion in the plane perpendicular to the spin axis (YZ plane), the kinetic
energy of the unbalanced mass is written as:
(15.105)

where:

= mass unbalance
= distance from the mass unbalance to the spin axis

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
824 Published: 2023-10-23
 Harmonic Analysis

= amplitude of the rotational velocity vector of the structure (input as OMEGA or

CMOMEGA command). It is equal to the frequency of excitation .
= phase of the unbalance
= instantaneous velocity along Y
= instantaneous velocity along Z

Because the mass unbalance is much smaller than the weight of the structure, the first two terms
are neglected. The third term being constant, will have no effect on the final equations.

Applying Lagrange's equations, the force vector is:

(15.106)
(15.107)

where:

Using complex notations, it can be written as:

(15.108)
(15.109)

Note:

The multiplication of the forces by is done internally at each frequency step.

15.4.9. Harmonic Ocean Wave Procedure (HOWP)

Standard harmonic analysis with ocean loading includes the added mass of the water outside of line
elements (such as LINK180, BEAM188, BEAM189, PIPE288, and PIPE289). The load vector is calculated
based on the loads at a given time, but the standard analysis method usually misses some important
effects, as all peak wave loads rarely occur at the same time.

All relevant ocean wave loading effects are accounted for via a specialized variation of the harmonic
analysis, the harmonic ocean wave procedure (HOWP).

The HOWP applies to regular waves only (Airy and Wheeler one-component waves, as well as Stokes
and Deans Stream Function waves), and works only with the full-solution harmonic analysis method
(HROPT,FULL).

The HOWP is accessed via the HROCEAN command. The frequency is obtained automatically, directly
from the ocean information (via OCDATA and OCTABLE commands). As with a standard harmonic
analysis, a damping matrix must be added separately if one is needed. Ocean loads are calculated
with the assumption that the structure is stationary.

For every component of every element, HOWP-specific calculations obtain the real and imaginary
loads caused by ocean waves. All other loads (such as those related to pressure, temperature, gravity
and buoyancy) are removed. Stated differently, all loads are sinusoidal at the given frequency, but
the magnitude and phase angle of every component of every element must be determined. To de-

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 825
Analysis Procedures

termine the magnitude and phase angle, a series of static analyses is performed along the phase
angle ranging from 0 to 360 degrees. (The number of these analyses is controlled via the HROCEAN
command.) The result is a roughly sinusoidal force pattern, as shown in this figure:

Figure 15.5: Force History

From this force history, the maximum and minimum forces are calculated by fitting the highest and
lowest points. Then:

(15.110)

(15.111)

(15.112)

so that is the coefficient on the real load vector and is the coefficient on the imaginary
load vector.

15.5. Buckling Analysis

The following buckling analysis topics are available:
15.5.1. Assumptions and Restrictions
15.5.2. Description of Analysis

15.5.1. Assumptions and Restrictions

1. Valid for structural degrees of freedom (DOFs) only.

2. The structure fails suddenly, with a horizontal force-deflection curve (see Figure 15.6: Types of
Buckling Problems (p. 827)).

3. The structure has constant stiffness effects.

4. A static solution with prestress effects included (PSTRES,ON) was run.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
826 Published: 2023-10-23
 Buckling Analysis

Figure 15.6: Types of Buckling Problems

15.5.2. Description of Analysis

This analysis type is for bifurcation buckling using a linearized model of elastic stability. Bifurcation
buckling refers to the unbounded growth of a new deformation pattern. A linear structure with a
force-deflection curve similar to Figure 15.6: Types of Buckling Problems (p. 827) (a) is well modeled
by a linear buckling (ANTYPE,BUCKLE) analysis, whereas a structure with a curve like Figure 15.6: Types
of Buckling Problems (p. 827) (b) is not (a large deflection analysis ( NLGEOM,ON) is appropriate, see
Large Rotation (p. 35)). The buckling problem is formulated as an eigenvalue problem:
(15.113)

where:

= stiffness matrix

= stress stiffness matrix

= ith eigenvalue (used to multiply the loads which generated [S])

= ith eigenvector of displacements

The eigenproblem is solved as discussed in Eigenvalue and Eigenvector Extraction (p. 745). The eigen-
vectors are normalized so that the largest component is 1.0. Thus, the stresses (when output) may
only be interpreted as a relative distribution of stresses.

By default, the Block Lanczos and Subspace Iteration methods find buckling modes in the range of
negative infinity to positive infinity. If the first eigenvalue closest to the shift point is negative (indic-
ating that the loads applied in a reverse direction will cause buckling), the program should find this
eigenvalue.

In the case of an eigenvalue with a very large multiplicity due to symmetries or geometric patterns
in the model, the algorithm may miss some occurrences of repeated eigenvalues. The default block
size is equal to 4 for buckling analysis, and the algorithm may have difficulties finding a higher
number of repeated eigenvalues. Requesting more eigenvalues can solve the problem.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 827
Analysis Procedures

15.6. Substructuring Analysis

The substructure analysis (ANTYPE,SUBSTR) uses the technique of matrix reduction to reduce the system
matrices to a smaller set of degrees of freedom.

The following substructuring analysis topics are available:

15.6.1. Assumptions and Restrictions (Within Superelement)
15.6.2. Description of Analysis
15.6.3. Statics
[Link]
15.6.5. Component Mode Synthesis (CMS)

15.6.1. Assumptions and Restrictions (Within Superelement)

1. Any degree of freedom can be used.

2. The elements have constant stiffness, damping, and mass effects (for example, material properties
do not change with temperature).

3. Coupled-field elements using load-vector coupling and elements with Lagrange multipliers cannot
be used.

15.6.2. Description of Analysis

A superelement simply represents a collection of elements that are reduced to act as one element.
This one (super) element can then be used in the actual analysis (use pass) or be used to generate
more superelements (generation or use pass). To reconstruct the detailed solutions (for example,
displacements and stresses) within the superelement, an expansion pass may be done. See the Sub-
structuring Analysis Guide for loads which are applicable to a substructure analysis.

Although a superelement can be used in any type of structural analysis (except a buckling or explicit
dynamics analysis), only static and transient analyses are considered in the following sections to derive
the reduced forms of the structural matrices (stiffness , mass , damping , load vector ).

For first-order non-structural analyses (thermal, electric, magnetic, diffusion) and second-order non-
structural analyses (fluid), matrices associated with zero order terms ( ), first order terms ( ),
and second order terms ( ) are reduced with the same logic. For coupled-field analyses, only the
reduction of is possible.

15.6.3. Statics
Consider the basic form of the static equations (Equation 15.1 (p. 794)):
(15.114)

includes nodal, pressure, and temperature effects. It does not include (see Newton-Raphson
Procedure (p. 730)). The equations may be partitioned into two groups, the master (retained) DOFs,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
828 Published: 2023-10-23
 Substructuring Analysis

here denoted by the subscript , and the condensed (removed) DOFs, here denoted by the subscript
.

(15.115)

or expanding:
(15.116)
(15.117)

The master DOFs should include all DOFs of all nodes on surfaces that connect to other parts of the
structure. If accelerations are to be used in the use pass or if the use pass will be a transient analysis,
master DOFs throughout the rest of the structure should also be used to characterize the distributed
mass. Solving Equation 15.117 (p. 829) for ,
(15.118)

Substituting into Equation 15.116 (p. 829)

(15.119)

or,
(15.120)

where:
(15.121)
(15.122)
(15.123)

and are the superelement coefficient (for example, stiffness) matrix and load vector, respect-
ively.

This development is equivalent to reduce the nodal displacement vector as:

(15.124)

is a transformation matrix with the form:

(15.125)

where:

= identity matrix
The vectors constituting are also called static constraint modes (see Craig and Bampton
[342] (p. 944)).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 829
Analysis Procedures

It follows that:
(15.126)

(15.127)

Note:

The expression of in Equation 15.118 (p. 829) can be derived from Equa-
tion 15.124 (p. 829) by adding the static correction term to the lower part.

In the preceding development, the load vector for the superelement has been treated as a total load
vector. The same derivation may be applied to any number of independent load vectors, which in
turn may be individually scaled in the superelement use pass. For example, the analyst may wish to
apply thermal, pressure, gravity, and other loading conditions in varying proportions. Expanding the
right-hand sides of Equation 15.116 (p. 829) and Equation 15.117 (p. 829) one gets, respectively:

(15.128)

(15.129)

where:

= number of independent load vectors.

Substituting into Equation 15.122 (p. 829):

(15.130)

To have independently scaled load vectors in the use pass, expand the left-hand side of Equa-
tion 15.130 (p. 830)

(15.131)

Substituting Equation 15.131 (p. 830) into Equation 15.130 (p. 830) :
(15.132)

If the load vectors are scaled in the use pass such that:

(15.133)

where is the scaling factor ( VAL1 on the SFE,,,SELV command), then Equation 15.118 (p. 829) be-
comes:

(15.134)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
830 Published: 2023-10-23
 Substructuring Analysis

Equation 15.134 (p. 830) is used in the expansion pass to obtain the DOF values at the condensed
DOFs if the backsubstitution method is chosen (SEOPT command). If the resolve method is chosen
for expansion pass, then the program will use Equation 15.115 (p. 829) to resolve for . In doing
so, the program makes as the internally prescribed displacement boundary conditions since
are known in expansion pass. As the program treats DOFs associated with as displacement
boundary conditions, the reaction forces by resolve method will be different from that computed at
those master DOFs by the backsubstitution method. However, they are all in self-equilibrium satisfying
Equation 15.115 (p. 829).

The above section Statics (p. 828) is equally applicable at an element level for elements with extra
displacement shapes. The master DOFs become the nodal DOFs and the condensed DOFs become
the nodeless or extra DOFs.

15.6.4. Transients
The general form of the equations for transients is Equation 15.5 (p. 796):
(15.135)

For substructuring, an equation of the form:

(15.136)

is needed. and are computed as they are for the static case (Equation 15.121 (p. 829) and
Equation 15.122 (p. 829) or Equation 15.126 (p. 830) and Equation 15.127 (p. 830)). As suggested by
Guyan ([14] (p. 925)), the mass matrix is also reduced through the transformation matrix given in
Equation 15.125 (p. 829). The reduced mass matrix is calculated by:
(15.137)

where:

The damping matrix is handled similarly:

(15.138)

Equation 15.134 (p. 830) is also used to expand the DOF values to the condensed DOFs in the transient
case if the backsubstitution method is chosen. If the resolve method is chosen, the program will use
Equation 15.115 (p. 829) and make {um} as displacement boundary conditions the same way as the
static expansion method does.

15.6.5. Component Mode Synthesis (CMS)

Component mode synthesis is an option used in substructure analysis (accessed with the CMSOPT
command) when DOFs are structural. It reduces the system matrices to a smaller set of interface DOFs
between substructures (components) and truncated sets of normal mode generalized coordinates
(see Craig [341] (p. 944)).

For a damped system, each CMS substructure is defined by a stiffness matrix, a mass matrix, and a
damping matrix. The matrix equation of motion is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 831
Analysis Procedures

(15.139)

Partitioning the matrix equation into interface and interior DOFs:

(15.140)

where subscripts and refer to:

= master DOFs defined only on interface nodes

= all DOFs that are not master DOFs

Similarly to Equation 15.124 (p. 829), the nodal displacement vector, , may be represented in terms
of master DOFs completed by component generalized coordinates (see Craig [341] (p. 944)) as in
Equation 15.141 (p. 832).

(15.141)

where:

= truncated set of generalized modal coordinates

Fixed-Interface Method

For the fixed-interface method, also commonly referred to as the Craig-Bampton method (see Craig
and Bampton [342] (p. 944)), the transformation matrix has the form:

(15.142)

where:

= fixed-interface normal modes (eigenvectors obtained with interface nodes fixed)

= null matrix

Free-Interface Method

For the free-interface method, also commonly referred to as the Herting method (see Hintz [406] (p. 948),
Herting [409] (p. 948)), the transformation matrix has the form:

(15.143)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
832 Published: 2023-10-23
 Substructuring Analysis

= matrix of the master DOF partition of the free-interface normal modes (eigenvectors
obtained with interface DOFs free)
= matrix of the condensed DOF partition of the free-interface normal modes

= matrix of inertia relief modes

is included only if rigid body modes are present (see CMSOPT„„,Fbddef). Any rigid
body modes present are not included in

where:

= matrix of the master DOF partition of the rigid body modes

For the free-interface method, a subset of master DOFs can be constrained during the mode-extraction
analysis. Let the master DOFs set be separated into subset , the master DOFs constrained during
the mode-extraction analysis, and subset , the master DOFs that remain free. A master DOF is part
of when SUPPORT is set to ON in the M command. Those pseudo-constraints apply only during
the mode-extraction analysis.

Enough master DOFs must be defined with pseudo-constraints so that the truncated set of normal
modes no longer contains rigid body modes. Those mixed-free interface normal modes can be parti-
tioned as:

(15.144)

= matrix of the master DOF partition of the mixed-free interface normal modes that are
constrained during the mode-extraction analysis.

= matrix of the master DOF partition of the mixed-free interface normal modes that remain
free during the mode-extraction analysis.

The transformation matrix takes the form:

(15.145)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 833
Analysis Procedures

Residual-Flexible Free-Interface Method

For the Residual-Flexible Free-Interface (RFFB) method, also commonly referred to as the Martinez
method (see Martinez et al. [407] (p. 948)), the transformation matrix has the form:

(15.146)

where:

= submatrices of the following residual vectors (see Residual Vector Meth-

od (p. 725)):

Any rigid body modes present are included in .

The residual vectors are also called residual attachment modes. Partitioning the residual flexibility
matrix given in Equation 14.146 (p. 726), the residual attachment modes are defined as:

(15.147)

As for the free-interface method, a subset of master DOFs can be constrained during the mode-ex-
traction analysis. With the same partitioning as Equation 15.144 (p. 833) , the transformation matrix
takes the form:

(15.148)

where:

the residual attachment modes calculated with master DOFs, m1, constrained.

Transformation
After applying the transformation in Equation 15.141 (p. 832) into the matrix equation of motion
Equation 15.139 (p. 832) , the equation of motion in the reduced space is obtained. The reduced stiffness,
mass, and damping matrices and the reduced load vector of the CMS substructure will be:
(15.149)
(15.150)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
834 Published: 2023-10-23
 Substructuring Analysis

(15.151)
(15.152)

In the reduced system, master DOFs will be used to couple the CMS superelement to other elements
and/or CMS superelements.

For the fixed-interface method, if the fixed-interface normal modes are mass normalized, the reduced
stiffness, mass, and damping matrices, and the reduced load vector have the final form:

(15.153)

(15.154)

(15.155)

(15.156)

Where:

= a diagonal matrix containing the eigenvalues of retained fixed-interface normal modes.

Recovery of the Condensed Displacements

The displacements at condensed or interior DOFs are recovered from the lower part of Equa-
tion 15.141 (p. 832). For the fixed-interface method, they are:
(15.157)

When reduced load vectors are scaled in the use pass, for instance in a full or mode-superposition
transient analysis or in a full or mode-superposition harmonic analysis, static correction vectors are
added. Equation 15.157 (p. 835) becomes:

(15.158)

with:

where:

= residual flexibility matrix of the condensed DOFs partition of the stiffness matrix

= condensed DOFs partition of the reduced stiffness matrix

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 835
Analysis Procedures

Equation 15.157 (p. 835) and Equation 15.158 (p. 835) are also valid for the free-interface method except
that should be replaced with .

For the RFFB method, is recovered from Equation 15.141 (p. 832) and Equation 15.146 (p. 834)
as:
(15.159)

Default Expansion Pass

The default expansion pass is based on the calculation of element results from the complete displace-
ment solution.

Expansion Pass Preceded by Element Results Calculation in the Generation Pass

If the element results of the component modes were calculated during the generation pass (Elcalc
= YES on CMSOPT), they are combined in the expansion pass in the same way as the displacements.
This is illustrated for the fixed-interface method. Let Equation 15.158 (p. 835) be rewritten in the form:
(15.160)

with:

where:

= alternative generalized modal coordinates

= matrix whose ith column is the static correction vector,

= vector whose ith component is the scaling factor,

For the stresses, if the stress-displacement matrix of element is noted as , the stress vector
of element is:
(15.161)

Partitioning the system into interface and interior DOFs:

(15.162)

For the fixed interface method, introducing Equation 15.160 (p. 836) into Equation 15.162 (p. 836) gives:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
836 Published: 2023-10-23
 Substructuring Analysis

(15.163)

The matrix of static constraint mode stresses of element :

(15.164)

and the matrix of fixed-interface normal mode stresses of element :

(15.165)

and the matrix of static correction vectors stresses of element :

(15.166)

are calculated and stored during the generation pass. Then in the expansion pass, the stress vector
of element is calculated as:
(15.167)

The preceding development is also valid for the free-interface method except that should be
replaced by .

For fixed-interface and free-interface methods, the static correction vectors are taken into account
only if ExpMth was set to MODDIR or BCLV on the SEOPT command, and KeyStat was set to ON
on the EXPASS command. As in the RFFB method, static correction terms are never calculated while
displacement and element results are combined in the following Equation 15.159 (p. 836).

Modal Damping Ratios

Damping ratio commands DMPRAT and MDAMP can be used to include damping terms to elements
of the diagonal of the reduced damping matrix (ANTYPE,SUBSTR with CMSOPT and SEOPT,,3).
The terms are:
(15.168)

where:

= constant modal damping ratio, input with DMPRAT

= modal damping ratio of the ith mode shape (see below), input with MDAMP
= circular frequency of the ith mode shape (see below)

When the fixed-interface method is used (CMSOPT,FIX), terms of Equation 15.168 (p. 837) are added
on the diagonal of in Equation 15.155 (p. 835) . The ith mode shape is the ith fixed-
interface normal mode.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 837
Analysis Procedures

When the free-interface method (CMSOPT,FREE) or the residual-flexible free-interface method (CM-
SOPT,RFFB) is used, terms of Equation 15.168 (p. 837) are added on the lower-left part of the reduced
damping matrix, which is . The ith mode shape is the ith free-interface normal mode.
If modal damping ratios are defined with the MDAMP command, only damping ratios associated
with non-rigid body modes are considered. For example, if there are six rigid-body modes, only data
defined for STLOC ≥ 7 are considered.

Unsymmetric Matrices
The reduction of a skew-symmetric damping matrix due to the effect of gyroscopic damping (CORI-
OLIS,ON,,,ON) is supported. The damping matrix must contain the gyroscopic effect only. For all three
methods, the modes used in the transformation basis are those of the conservative system. The res-
ulting reduced damping matrix is also skew-symmetric.

15.7. Spectrum Analysis

Two types of spectrum analyses (ANTYPE,SPECTR) are supported: the deterministic response spectrum
method and the nondeterministic random vibration method. Both excitation at the support and excitation
away from the support are allowed. The three response spectrum methods are the single-point, multiple-
point and dynamic design analysis method. The random vibration method uses the power spectral
density (PSD) approach.

The following spectrum analysis topics are available:

15.7.1. Assumptions and Restrictions
15.7.2. Description of Analysis
15.7.3. Single-Point Response Spectrum
15.7.4. Damping
15.7.5. Participation Factors and Mode Coefficients
15.7.6. Combination of Modes
15.7.7. Effective Mass and Cumulative Mass Fraction
15.7.8. Dynamic Design Analysis Method
15.7.9. Random Vibration Method
15.7.10. Description of Method
15.7.11. Response Power Spectral Densities and Mean Square Response
15.7.12. Cross Spectral Terms for Partially Correlated Input PSDs
15.7.13. Spatial Correlation
15.7.14. Wave Propagation
15.7.15. Multi-Point Response Spectrum Method
15.7.16. Missing-Mass Response
15.7.17. Rigid Responses

15.7.1. Assumptions and Restrictions

1. The structure is linear.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
838 Published: 2023-10-23
 Spectrum Analysis

2. For single-point response spectrum analysis (SPOPT,SPRS) and dynamic design analysis method
(SPOPT,DDAM), the structure is excited by a spectrum of known direction and frequency compon-
ents, acting uniformly on all support points or on specified unsupported master degrees of freedom
(DOFs).

3. For multi-point response spectrum (SPOPT,MPRS) and power spectral density (SPOPT,PSD) analyses,
the structure may be excited by different input spectra at different support points or unsupported
nodes. Up to twenty different simultaneous input spectra are allowed.

15.7.2. Description of Analysis

The spectrum analysis capability is a separate analysis type (ANTYPE,SPECTR) and it must be preceded
by a mode-frequency analysis. If mode combinations are needed, the required modes must also be
expanded, as described in Modal Analysis (p. 810).

The four options available are the single-point response spectrum method (SPOPT,SPRS), the dynamic
design analysis method (SPOPT,DDAM), the random vibration method (SPOPT,PSD) and the multiple-
point response spectrum method (SPOPT,MPRS). Each option is discussed in detail subsequently.

15.7.3. Single-Point Response Spectrum

Both excitation at the support (base excitation) and excitation away from the support (force excitation)
are allowed for the single-point response spectrum analysis (SPOPT,SPRS). The table below summarizes
these options as well as the input associated with each.

Table 15.3: Types of Spectrum Loading

Excitation Option
Excitation at Support Excitation Away From
Spectrum input Response spectrum table (FREQ and SV commands) Amplitude multiplier table (FREQ
commands)
Orientation of Direction vector (input on SED and ROCK commands) X, Y, Z direction at each node (se
load or FZ on F command)
Distribution of Constant on all support points Amplitude in X, Y, or Z directions
loads on F command)
Type of input Velocity Acceleration Displacement Force
Response 0 2 3 1
spectrum type
(KSV on SVTYP
command)

15.7.4. Damping
Damping is evaluated for each mode and is defined as in Equation 14.41 (p. 699).

Note that the material dependent damping contribution (MP,DMPR) is computed in the modal expan-
sion phase with MXPAND,,,,YES, so that this damping contribution must be included there.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 839
Analysis Procedures

15.7.5. Participation Factors and Mode Coefficients

The participation factors for the given excitation direction are defined as:
(15.169)

(15.170)

where:

= participation factor for the ith mode

= eigenvector normalized using Equation 15.54 (p. 812) (Nrmkey on the MODOPT
command has no effect). If the unsymmetric eigensolver is used, the left eigenvectors are
used to calculate the participation factors.
= vector describing the excitation direction (see Equation 15.171 (p. 840))

= input force vector

The vector describing the excitation direction has the form:

(15.171)

where:

= excitation at DOF j in direction a; a may be either X, Y, Z, or rotations about one of

these axes.

X, Y, Z = global Cartesian coordinates of a point on the geometry

Xo, Yo, Zo = global Cartesian coordinates of point about which rotations are done (reference
point)
= six possible unit vectors

We can calculate the statically equivalent actions at j due to rigid-body displacements of the reference
point using the concept of translation of axes [T] (Weaver and Johnston([279] (p. 940))).

For spectrum analysis, the Da values may be determined in one of two ways:

1. For D values with rocking not included (based on the SED command):
(15.172)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
840 Published: 2023-10-23
 Spectrum Analysis

(15.173)

(15.174)

where:

SX, SY, SZ = components of excitation direction (input as SEDX, SEDY, and SEDZ, respectively,
on SED command)

2. or, for D values with rocking included (based on the SED and ROCK command):
(15.175)
(15.176)
(15.177)

R is defined by:

(15.178)

where:

CX, CY, CZ = angular components of rocking (input as OMX, OMY, and OMZ, respectively, on
ROCK command)
x = vector cross product operator
rX = Xn - LX

rY = Yn - LY

rZ = Zn - LZ

Xn, Yn, Zn = coordinate of node n

LX, LY, LZ = location of center of rotation (input as CGX, CGY, and CGZ on ROCK command)

In a modal analysis, the ratio of each participation factor to the largest participation factor (output
as RATIO) is printed out.

The displacement, velocity or acceleration vector for each mode is computed from the eigenvector
by using a "mode coefficient":
(15.179)

where:

m = 0, 1, or 2, based on whether the displacements, velocities, or accelerations, respectively,

are selected (using label, the third field on the mode combination commands SRSS, CQC,
GRP, DSUM, NRLSUM, ROSE)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 841
Analysis Procedures

Ai = mode coefficient (see below)

The mode coefficient is computed in five different ways, depending on the type of excitation (SVTYP
command).

1. For velocity excitation of base (SVTYP, 0)

(15.180)

where:

= spectral velocity for the ith mode (obtained from the input velocity spectrum at
frequency and effective damping ratio defined in Equation 14.42 (p. 699))

= ith natural frequency (cycles per unit time = )

= ith natural circular frequency (radians per unit time)

2. For force excitation (SVTYP, 1)

(15.181)

where:

= spectral force for the ith mode (obtained from the input amplitude multiplier table
at frequency and effective damping ratio ).

3. For acceleration excitation of base (SVTYP, 2)

(15.182)

where:

= spectral acceleration for the ith mode (obtained from the input acceleration response
spectrum at frequency and effective damping ratio ).

4. For displacement excitation of base (SVTYP, 3)

(15.183)

where:

= spectral displacement for the ith mode (obtained from the input displacement re-
sponse spectrum at frequency and effective damping ratio ).

When , , , or are needed between input frequencies, log-log interpolation is done in the
space as defined.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
842 Published: 2023-10-23
 Spectrum Analysis

The spectral values and the mode coefficients output in the RESPONSE SPECTRUM CALCULATION
SUMMARY table are evaluated at the input curve with the lowest damping ratio, not at the effective
damping ratio .

15.7.6. Combination of Modes

The modal displacements, velocity and acceleration (Equation 15.179 (p. 841)) may be combined in
different ways to obtain the response of the structure. For all excitations but the PSD this would be
the maximum response, and for the PSD excitation, this would be the 1-σ (standard deviation) relative
response. The response includes DOF response as well as element results and reaction forces if com-
puted in the expansion operations (Elcalc = YES on the MXPAND command).

In the case of the single-point response spectrum method (SPOPT,SPRS) or the dynamic-design ana-
lysis method (SPOPT,DDAM) options of the spectrum analysis , it is possible to expand only those
modes whose significance factor exceeds the significant threshold value (SIGNIF value on MXPAND
command). Note that the mode coefficients must be available at the time the modes are expanded.

Only those modes having a significant amplitude (mode coefficient) are chosen for mode combination.
A mode having a coefficient of greater than a given value (input as SIGNIF on the mode combination
commands SRSS, CQC, GRP, DSUM, NRLSUM, ROSE and PSDCOM) of the maximum mode coefficient
(all modes are scanned) is considered significant.

The spectrum option provides the following options for the combination of modes. They are:
[Link]. Complete Quadratic Combination Method
[Link]. Grouping Method
[Link]. Double Sum Method
[Link]. Square Root of the Sum of the Squares (SRSS) Method
[Link]. Naval Research Laboratory Sum (NRL-SUM) Method
[Link]. Closely Spaced Modes (CSM) Method
[Link]. Rosenblueth Method

These methods generate coefficients for the combination of mode shapes. This combination is done
by a generalization of the method of the square root of the sum of the squares which has the form:

(15.184)

where:

= total modal response

= total number of expanded modes
= coupling coefficient. The value of = 0.0 implies modes i and j are independent and
approaches 1.0 as the dependency increases
= modal response in the ith mode (Equation 15.179 (p. 841))

= modal response in the jth mode

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 843
Analysis Procedures

= mode coefficient for the ith mode

= mode coefficient for the jth mode

= the ith mode shape

= the jth mode shape

and may be the DOF response, reactions, or stresses. The DOF response, reactions, or stresses
may be displacement, velocity or acceleration depending on the user request (Label on the mode
combination commands SRSS, CQC, DSUM, GRP, ROSE or NRLSUM).

By default, the mode combination instructions are written to [Link] by the mode combin-
ation command. Inputting this file in POST1 (/INPUT command) automatically performs the mode
combination. If the SPOPT command has been issued with Elcalc = YES, [Link] is written
for possible examination of the combinations, but it is not used for post-processing.

[Link]. Complete Quadratic Combination Method

This method (accessed with the CQC command), is based on Wilson, et al.([66] (p. 928)).

(15.185)

where:

[Link]. Grouping Method

This method (accessed with the GRP command), is from the NRC Regulatory Guide([42] (p. 927)).
For this case, Equation 15.184 (p. 843) specializes to:

(15.186)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
844 Published: 2023-10-23
 Spectrum Analysis

Closely spaced modes are divided into groups that include all modes having frequencies lying
between the lowest frequency in the group and a frequency 10% higher. No one frequency is to
be in more than one group.

[Link]. Double Sum Method

The Double Sum Method (accessed with the DSUM command) also is from the NRC Regulatory
Guide([42] (p. 927)). For this case, Equation 15.184 (p. 843) specializes to:

(15.187)

where:

= damped natural circular frequency of the ith mode

= undamped natural circular frequency of the ith mode

= modified damping ratio of the ith mode

The damped natural frequency is computed as:

(15.188)

The modified damping ratio is defined to account for the earthquake duration time:
(15.189)

where:

= earthquake duration time, fixed at 10 units of time

Modes whose effective damping ratio exceeds or equals 1.0 are not combined.

[Link]. Square Root of the Sum of the Squares (SRSS) Method

The SRSS method, accessed with the SRSS command, is taken from the NRC Regulatory
Guide([42] (p. 927)). For this case, Equation 15.184 (p. 843) reduces to:

(15.190)

[Link]. Naval Research Laboratory Sum (NRL-SUM) Method

The NRL-SUM method (O'Hara and Belsheim([108] (p. 930))), accessed with the NRLSUM command,
calculates the maximum modal response as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 845
Analysis Procedures

(15.191)

where:

= absolute value of the largest modal displacement, stress or reaction at the point
= displacement, stress or reaction contributions of the same point from other modes.

[Link]. Closely Spaced Modes (CSM) Method

The CSM method, accessed with the NRLSUM,,,CSM command, calculates the combined effect of
the modal responses of pairs of closely spaced modes (NAVSEA [408] (p. 948)). Modes are close if
their frequencies are within 10% of the mean frequency. Once each CSM pair combination is de-
termined, it is used in the NRL-SUM (p. 845) as a single effective mode. The modified NRL (or CSM)
sum is:

(15.192)

where

ranges from 1 to N, excluding the closely spaced modes j and k.

is the maximum amplitude of a mode pair response. It is computed as follows.

This procedure first requires a modal response correction to account for the effect of damping
during the first quarter cycle:
(15.193)

where:

= modal response of the ith mode including the quarter cycle correction.
= damping ratio (constant for all modes)
The closed form treatment can be expressed as the envelope of the sum of two decaying
sinusoids (modes j and k). This closed form is a slightly conservative approach, as it is based
on determining the peak of the envelope rather than the peak of the superposed values.
(15.194)

where:

= combined response of two closely spaced modes j and k

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
846 Published: 2023-10-23
 Spectrum Analysis

The times at which this function reaches a minimum or a maximum are t = 0, and:

(15.195)

where:

and

Equation 15.195 (p. 847) has multiple solutions only if S is smaller or equal to 1. If S is greater than
1, the CSM cannot be used to reduce the NRL sum.

[Link]. Rosenblueth Method

The Rosenblueth Method ([372] (p. 946)) is accessed with the ROSE command.

The equations for the Double Sum method (above) apply, except for Equation 15.187 (p. 845). For
the Rosenblueth Method, the sign of the modal responses is retained:

(15.196)

Modes whose effective damping ratio exceeds or equals 1.0 are not combined.

15.7.7. Effective Mass and Cumulative Mass Fraction

The effective mass (output as EFFECTIVE MASS) for the ith mode (which is a function of excitation
direction) is (Clough and Penzien([81] (p. 929))):

(15.197)

Note from Equation 15.54 (p. 812) that

(15.198)

so that the effective mass reduces to . This does not apply to:

• the force spectrum, for which the excitation is independent of the mass distribution

• acoustic fluid-structure interaction (FSI) with unsymmetric formulation

If the unsymmetric eigensolver is used, the left and right eigenvectors are used in Equa-
tion 15.197 (p. 847) and Equation 15.198 (p. 847). For acoustic fluid-structure interaction, the effective
mass is expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 847
Analysis Procedures

(15.199)

where:

is the vector of excitation direction, structural degrees of freedom only

is the stiffness matrix of the structural part (see Equation 8.153 (p. 274))

is the ith right eigenvector, structural degrees of freedom only.

The cumulative mass fraction for the ith mode is:

(15.200)

where N is the total number of modes.

15.7.8. Dynamic Design Analysis Method

For the DDAM (Dynamic Design Analysis Method) procedure (SPOPT,DDAM) (O'Hara and
Belsheim([108] (p. 930))), modal mass in thousands of units are computed from the participation factor:
(15.201)

with

(15.202)

and
(15.203)

For units system BIN or BFT, the modal weights are computed by:
(15.204)

For other unit systems, it is calculated as:

(15.205)

where:

= modal weight and/or modal mass

= generalized mass of the ith mode
= acceleration due to gravity

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
848 Published: 2023-10-23
 Spectrum Analysis

= ith mode shape normalized to unity

The mode coefficients are computed by:

(15.206)

where:

= the greater of or
= minimum acceleration (input as AMIN on the ADDAM command) defaults to 6g)
= the lesser of or
= spectral acceleration

= spectral velocity

= acceleration spectrum computation constants (input as AF, AA, AB, AC, AD

on the ADDAM command)
= velocity spectrum computation constants (input as VF, VA, VB, VC on the
VDDAM command)

The acceleration response, which is required for the nodal acceleration check (refer to Mode Selection
Based on the DDAM Method), is computed by:
(15.207)

where:

= acceleration response for the ith mode

The DDAM procedure is generally used with the NRL-SUM method of mode combination, which was
described in Single-Point Response Spectrum (p. 839). Note that unlike Equation 15.54 (p. 812), O'Hara
and Belsheim ([108] (p. 930)) normalize the mode shapes to the largest modal displacements. As a
result, the NRL-1396 participation factors and mode coefficients will be different.

15.7.9. Random Vibration Method

The random vibration method (SPOPT,PSD) allows multiple power spectral density (PSD) inputs (up
to two hundred) in which these inputs can be:

1. full correlated,

2. uncorrelated, or

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 849
Analysis Procedures

3. partially correlated.

The procedure is based on computing statistics of each modal response and then combining them.
It is assumed that the excitations are stationary random processes.

15.7.10. Description of Method

For partially correlated nodal and base excitations, the complete equations of motions are segregated
into the free and the restrained (support) DOF as:

(15.208)

where are the free DOF and are the restrained DOF that are excited by random loading
(unit value of displacement on D command). Note that the restrained DOF that are not excited are
not included in Equation 15.208 (p. 850) (zero displacement on D command). {F} is the nodal force
excitation activated by a nonzero value of force (on the F command). The value of force can be other
than unity, allowing for scaling of the participation factors.

The free displacements can be decomposed into pseudo-static and dynamic parts as:
(15.209)

The pseudo-static displacements may be obtained from Equation 15.208 (p. 850) by excluding the first
two terms on the left-hand side of the equation and by replacing by :
(15.210)

in which . Physically, the elements along the ith column of [A] are the pseudo-
static displacements due to a unit displacement of the support DOFs excited by the ith base PSD.
These displacements are written as load step 2 on the .RST file. Substituting Equation 15.210 (p. 850)
and Equation 15.209 (p. 850) into Equation 15.208 (p. 850) and assuming light damping yields:
(15.211)

The second term on the right-hand side of the above equation represents the equivalent forces due
to support excitations.

Using the mode-superposition analysis of Mode-Superposition Method (p. 721) and rewriting Equa-
tion 14.123 (p. 722) as:
(15.212)

the above equations are decoupled yielding:

(15.213)

where:

= number of mode shapes chosen for evaluation (input as NMODE on SPOPT command)
= generalized displacements

and = natural circular frequencies and modal damping ratios

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
850 Published: 2023-10-23
 Spectrum Analysis

The modal loads are defined by:

(15.214)

The modal participation factors corresponding to support excitation are given by:
(15.215)

and for nodal excitation:

(15.216)

Note that, for simplicity, equations for nodal excitation problems are developed for a single PSD table.
Multiple nodal excitation PSD tables are, however, allowed in the program.

These factors are calculated (as a result of the PFACT action command) when defining base or nodal
excitation cases and are written to the .psd file. Mode shapes should be normalized with respect
to the mass matrix as in Equation 15.54 (p. 812). If the unsymmetric eigensolver is used, the left eigen-
vectors are used to calculate the participation factors.

The relationship between multiple input spectra are described in Cross Spectral Terms for Partially
Correlated Input PSDs (p. 854).

15.7.11. Response Power Spectral Densities and Mean Square Response

Using the theory of random vibrations, the response PSD's can be computed from the input PSD's
with the help of transfer functions for single DOF systems and by using mode-superposition
techniques (RPSD command in POST26). The displacement response PSD's for ith DOF are given by:

Dynamic Part

(15.217)

Pseudo-Static Part

(15.218)

Covariance Part

(15.219)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 851
Analysis Procedures

= participation factor for mode j (respectively k) corresponding to force excitation

l (respectively m), see Equation 15.216 (p. 851)

= participation factor for mode j (respectively k) corresponding to base excitation

l (respectively m), see Equation 15.215 (p. 851)
= input force PSD

= input acceleration PSD

= number of mode shapes chosen for evaluation (input as NMODE on SPOPT command)
and = number of nodal (away from support) and base PSD tables, respectively

The transfer functions for the single DOF system assume different forms depending on the type (Type
on the PSDUNIT command) of the input PSD and the type of response desired (Lab and Relkey on
the PSDRES command). The forms of the transfer functions for displacement as the output are listed
below for different inputs.

1. Input = force or acceleration (FORC, ACEL, or ACCG on PSDUNIT command):

(15.220)

2. Input = displacement (DISP on PSDUNIT command):

(15.221)

3. Input = velocity (VELO on PSDUNIT command):

(15.222)

where:

= forcing frequency
= natural circular frequency for jth mode

Now, random vibration analysis can be used to show that the absolute value of the mean square re-
sponse of the ith free displacement (ABS option on the PSDRES command) is:

(15.223)

where:

| |Re = denotes the real part of the argument

= variance of the ith relative (dynamic) free displacements (REL option on PSDRES)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
852 Published: 2023-10-23
 Spectrum Analysis

= variance of the ith pseudo-static displacements

= covariance between the static and dynamic displacements

The general formulation described above gives simplified equations for several situations commonly
encountered in practice. For fully correlated nodal excitations and identical support motions, the
subscripts and m would drop out from the Equation 15.217 (p. 851) thru Equation 15.219 (p. 851).
When only nodal excitations exist, the last two terms in Equation 15.223 (p. 852) do not apply, and
only the first term within the large parentheses in Equation 15.217 (p. 851) needs to be evaluated. For
uncorrelated nodal force and base excitations, the cross PSD's (i.e. ≠ m) are zero, and only the terms
for which = m in Equation 15.217 (p. 851) thru Equation 15.219 (p. 851) need to be considered.

Equation 15.217 (p. 851) thru Equation 15.219 (p. 851) can be rewritten as:

(15.224)

(15.225)

(15.226)

where:

= modal PSD's, terms within large parentheses of Equa-

tion 15.217 (p. 851) thru Equation 15.219 (p. 851)

Closed-form solutions for piecewise linear PSD in log-log scale are employed to compute each integ-
ration in Equation 15.223 (p. 852) (Chen and Ali([195] (p. 935)) and Harichandran([196] (p. 935))) .

Subsequently, the variances become:

(15.227)

(15.228)

(15.229)

The modal covariance matrices are available in the .psd file. Note that Equa-
tion 15.227 (p. 853) thru Equation 15.229 (p. 853) represent mode combination (PSDCOM command)
for random vibration analysis.

[Link]. Variance of Element Results and Reaction Forces

The variance of element results (stresses, strains, miscellaneous summable data), nodal forces or
reactions can be computed (Elcalc = YES on SPOPT (if Elcalc = YES on MXPAND)) from equations

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 853
Analysis Procedures

similar to Equation 15.227 (p. 853) thru Equation 15.229 (p. 853). For instance, if the stress variance
is desired, replace the mode shapes ( ) and static displacements with mode stresses
and static stresses . Similarly, if the node force variance is desired, replace the mode shapes
and static displacements with mode nodal forces and static nodal forces . Finally, if re-
action variances are desired, replace the mode shapes and static displacements with mode reaction
and static reactions . Furthermore, the variances of the first and second time derivatives
(VELO and ACEL options respectively on the PSDRES command) of all the quantities mentioned
above can be computed using the following relations:
(15.230)
(15.231)

Finally, the square root of Equations Equation 15.227 (p. 853), Equation 15.228 (p. 853), and Equa-
tion 15.229 (p. 853) are taken to yield the 1-σ results (that is, the standard deviation) for the ith dis-
placement (or strain, stress, or force) quantity, and they are written to the .rst file.

[Link]. Equivalent Stress Mean Square Response

The equivalent stress (SEQV) mean square response is computed as suggested by Segalman et
al([351] (p. 945)) as:

(15.232)

where:

= vector of component "stress shapes" for mode j at node nd

= quadratic operator

Note that the probability distribution for the equivalent stress is neither Gaussian nor is the mean
value zero. However, the"3-σ" rule (multiplying the RMS value by 3) yields a conservative estimate
on the upper bound of the equivalent stress (Reese et al([352] (p. 945))). Since no information on
the distribution of the principal stresses or stress intensity (S1, S2, S3, and SINT) is known, these
values are set to zero.

15.7.12. Cross Spectral Terms for Partially Correlated Input PSDs

For excitation defined by more than a single input PSD, cross terms which determine the degree of
correlation between the various PSDs are defined as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
854 Published: 2023-10-23
 Spectrum Analysis

(15.233)

where:

= input PSD spectra which are related. (Defined by the PSDVAL command and located
as table number (TBLNO) n)
= cospectra which make up the real part of the cross terms. (Defined by the COVAL
command where n and m (TBLNO1 and TBLNO2) identify the matrix location of the cross
term)

= quadspectra which make up the imaginary part of the cross terms. (Defined by
the QDVAL command where n and m (TBLNO1 and TBLNO2) identify the matrix location of
the cross term)

The normalized cross PSD function is called the coherence function and is defined as:

(15.234)

where:

Although the above example demonstrates the cross correlation for 3 input spectra, this matrix may
range in size from 2 x 2 to 200 x 200 (that is, maximum number of tables is 200).

For the special case in which all cross terms are zero, the input spectra are said to be uncorrelated.
Note that correlation between nodal and base excitations is not allowed.

15.7.13. Spatial Correlation

The degree of correlation between excited nodes may also be controlled. Depending upon the distance
between excited nodes and the values of R MIN and R MAX (input as RMIN and RMAX on the PSDSPL
command), an overall excitation PSD can be constructed such that excitation at the nodes may be
uncorrelated, partially correlated or fully correlated. If the distance between excited nodes is less than
R MIN , then the two nodes are fully correlated; if the distance is greater than R MAX , then the two
nodes are uncorrelated; if the distance lies between R MIN and R MAX , excitation is partially correlated
based on the actual distance between nodes. The following figure indicates how R MIN , R MAX and
the correlation are related. Spatial correlation between excited nodes is not allowed for a pressure
PSD analysis (PSDUNIT,PRES).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 855
Analysis Procedures

Figure 15.7: Sphere of Influence Relating Spatially Correlated PSD Excitation

Node i excitation is fully correlated with node j excitation

Node i excitation is partially correlated with node k excitation
Node i excitation is uncorrelated with node excitation

For two excitation points 1 and 2, the PSD would be:

(15.235)

where:

= distance between the two excitation points 1 and 2

= basic input PSD (PSDVAL and PSDFRQ commands)

15.7.14. Wave Propagation

To include wave propagation effects of a random loading, the excitation PSD is constructed as:
(15.236)

where:

= delay

= separation vector between excitations points and m

= velocity of propagation of the wave (input as VX, VY and VZ on PSDWAV command)

= nodal coordinates of excitation point

More than one simultaneous wave or spatially correlated PSD inputs are permitted, in which case the
input excitation [S(ω)] reflects the influence of two or more uncorrelated input spectra. In this case,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
856 Published: 2023-10-23
 Spectrum Analysis

partial correlation among the basic input PSD's is not currently permitted. Wave propagation effects
are not allowed for a pressure PSD analysis (PSDUNIT,PRES).

15.7.15. Multi-Point Response Spectrum Method

The response spectrum analysis due to multi-point support and nodal excitations (SPOPT,MPRS) allows
up to three hundred different excitations (PFACT command) based on up to two hundred different
spectrum tables (SPUNIT, SPFREQ and SPVAL commands). The input spectrum are assumed to be
uncorrelated to each other.

Most of the ingredients for performing multi-point response spectrum analysis are already developed
in the previous subsection of the random vibration method. As with the PSD analysis, the static shapes
corresponding to Equation 15.210 (p. 850) for base excitation are written as load step #2 on the *.rst
th
file. Assuming that the participation factors, , for the input spectrum table have already been
th
computed (by Equation 15.215 (p. 851), for example), the mode coefficients for the table are obtained
as:
(15.237)

where:

th
= interpolated input response spectrum for the table at the jth natural frequency (defined
by the SPFREQ, SPVAL and SPUNIT commands)

For each input spectrum, the mode shapes, mode stresses, etc. are multiplied by the mode coefficients
to compute modal quantities, which can then be combined with the help of any of the available
mode combination techniques (SRSS, CQC, Double Sum, Grouping, NRL-SUM, or Rosenblueth method),
as described in the previous section on the single-point response spectrum method. The Absolute
Sum method (AbsSumKey = yes on the SRSS command) can also be used as described in the section
that follows.

Finally, the response of the structure is obtained by combining the responses to each spectrum using
the SRSS method.

By default, the mode combination instructions are written to the file [Link] by the mode
combination command. Inputting the file in POST1 (/INPUT command) automatically performs the
mode combination. If the SPOPT command has been issued with Elcalc = YES, [Link]
is written for possible examination of the combinations, but it is not used for post-processing.

[Link]. Absolute Sum Combination Method

In a multi-point response spectrum analysis with base excitation, the square root of sum of square
(SRSS) combination method does not take into account the phase relationship between the modal
responses when same direction of excitation is applied at each support. The Absolute Sum combin-
ation method [[417] (p. 948)] takes into account that the peak values for one excitation direction
are occurring at the same time so that the absolute values are summed. The remaining combination
of modes and directions is performed with the SRSS method.

The total response with missing mass included is expressed as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 857
Analysis Procedures

Where:

is the excitation direction index. The excitation direction is specified using SEDX, SEDY, or SEDZ
on the SED command.

is the total number of excitation directions.

is the support (node group) index. The node group is defined with a component based on nodes
(Cname on the SED command).

is the total number of supports.

is the modal response for the jth mode, the direction, and the support.

is the missing mass response for the direction, and the support (MMASS command).
It is optional.

Note:

All supports are not necessarily excited in all directions. For example, if the support
is not excited in the direction then is zero for all modes j.

15.7.16. Missing-Mass Response

The spectrum analysis is based on a mode-superposition approach where the responses of the higher
modes are neglected; therefore, part of the mass of the structure is missing in the dynamic analysis.
The missing-mass response method ([371] (p. 946) and [421] (p. 949)) permits inclusion of the missing
mass effect in a single-point response spectrum (SPOPT, SPRS) or multiple-point response spectrum
analysis (SPOPT,MPRS) when base excitation is considered

Considering a rigid structure, the inertia force due to ground acceleration is:
(15.238)

where:

= total inertia force vector

= spectrum acceleration at zero period (also called the ZPA value), input as ZPA on the MMASS
command.

Mode-superposition can be used to determine the inertia force. For mode j, the modal inertia force
is:
(15.239)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
858 Published: 2023-10-23
 Spectrum Analysis

where:

= modal inertia force for mode j.

Using equations Equation 15.179 (p. 841) and Equation 15.182 (p. 842), this force can be rewritten:
(15.240)

The missing inertia force vector is then the difference between the total inertia force given by Equa-
tion 15.238 (p. 858) and the sum of the modal inertia forces defined by Equation 15.240 (p. 859):

(15.241)

The expression within the parentheses in the equation above is the fraction of degree of freedom
mass missing:

(15.242)

The missing mass response for displacement results is the static shape due to the inertia forces defined
by the equation:
(15.243)

where:

is the missing mass response

For acceleration results, the missing mass corresponds to the residual acceleration load from Equa-
tion 15.241 (p. 859) and is written:

(15.244)

The application of these equations can be extended to flexible structures because the higher truncated
modes are supposed to be mostly rigid and exhibit pseudo-static responses to an acceleration base
excitation.

In a single-point response spectrum analysis, the missing mass response is written as load step 2 in
the .RST file. In a multiple-point response spectrum analysis, it is written as load step 3.

Combination Method

Because the missing-mass response is a pseudo-static response, it is in phase with the imposed accel-
eration but out of phase with the modal responses; therefore, the missing-mass response and the
modal responses defined in Equation 15.184 (p. 843) are combined using the square root of sum of
the squares (SRSS) method.

The total response including the missing mass effect is:

(15.245)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 859
Analysis Procedures

15.7.17. Rigid Responses

For frequencies higher than the amplified acceleration region of the spectrum, the modal responses
consist of both periodic and rigid components. The rigid components are considered separately because
the corresponding responses are all in phase. The combination methods listed in Combination of
Modes (p. 843) do not apply.

The rigid component of a modal response is expressed as:

(15.246)

where:

= the rigid component of the modal response of mode i

= rigid response coefficient in the range of values 0 through 1. See the Gupta and Lindley-Yow
methods below.

= modal response of mode i

The corresponding periodic component is then:

(15.247)

where:

= periodic component of the modal response of mode i

Two methods ([372] (p. 946)) can be used to separate the periodic and the rigid components in each
modal response. Each one has a different definition of the rigid response coefficients .

Gupta Method

(15.248)

where:

= ith frequency value.

and = key frequencies. is input as Val1 and is input as Val2 on RIGRESP command
with Method = GUPTA.

Lindley-Yow Method

(15.249)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
860 Published: 2023-10-23
 Linear Perturbation Analysis

= spectrum acceleration at zero period (ZPA). It is input as Val1 on RIGRESP command with
Method = LINDLEY

= spectrum acceleration corresponding to the ith frequency

Combination Method

The periodic components are combined using the square root of sum of squares (SRSS), the complete
quadratic (CQC) or the Rosenblueth (ROSE) combination methods.

Since the rigid components are all in phase, they are summed algebraically. When the missing mass
response (MMASS) is included in the analysis, since it is a rigid response as well, it is summed with
those components. Finally, periodic and rigid responses are combined using the SRSS method.

The total response with the rigid responses and the missing mass response included is expressed as:

(15.250)

15.8. Linear Perturbation Analysis

The following linear perturbation analysis topics are available:
15.8.1. Assumptions and Restrictions
15.8.2. Description of Analysis
15.8.3. Static Analysis Based on Linear Perturbation
15.8.4. Modal Analysis Based on Linear Perturbation
15.8.5. Eigenvalue Buckling Analysis Based on Linear Perturbation
15.8.6. Harmonic Analysis Based on Linear Perturbation
15.8.7. Substructure or CMS Generation Based on Linear Perturbation
15.8.8. Application of Perturbation Loads
15.8.9. Downstream Analysis Using the Solution of a Linear Perturbation Analysis

15.8.1. Assumptions and Restrictions

The following assumptions and restrictions apply to the linear perturbation method:

• Valid for structural degrees of freedom (DOFs) only.

• Supports only certain structural element types. See Elements Under Linear Perturbation in the Element
Reference for a complete list.

• Performed by first running a linear or nonlinear static or full (TRNOPT,FULL) transient analysis, then
restarting the analysis with the appropriate linear perturbation analysis type.

• For a linear perturbation modal analysis, the mass density and damping information should be
supplied in the analysis phase of the static or full transient procedure (if they are needed later for
the linear perturbation analysis).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 861
Analysis Procedures

• LoadControl options on the PERTURB command, PARKEEP and DZEROKEEP, set the constant
term of any coupling or constant equations to zero in addition to displacement constraints.

15.8.2. Description of Analysis

In many engineering applications, the linear behavior of a structure based on a prior linear or nonlinear
preloaded status is of interest. The linear perturbation analysis procedure is designed to solve a linear
problem from this preloaded case. Typically, the Newton-Raphson procedure is used in the nonlinear
analysis (see Figure 14.9: Newton-Raphson Solution - One Iteration (p. 732)). Without loss of generality,
the nonlinear static problem is discussed here. A similar process can apply to linear static and linear
or nonlinear full transient analyses.

In the case of a nonlinear static analysis, the mode frequency of the structure based on load level
is needed. The tangent matrix can be used in the linear perturbation analysis in order
to obtain the effect of preload, as the linear stiffness matrix without preloading does not give the
correct solution.

Equation 14.163 (p. 731) at iteration i is recast here for convenience of discussion:
(15.251)

Here, is the global tangent matrix which can be symbolically segregated into other matrices
as follows:
(15.252)

where:

= the part of the tangent stiffness contributed from the material property

= the stress stiffening matrix introduced by non-zero stresses from the structure; super-
script N indicates nonlinear (that is, the stress stiffening matrix is obtained from a nonlinear
analysis – not to be confused with the linear stress stiffening matrix obtained from the first
phase of a linear perturbation buckling analysis)
= the load stiffening matrix introduced by external pressure loads or by follower force
effect (element FOLLW201)
= the total stiffness matrix contributed from contact elements of the model

= the spin-softening matrix introduced by rotational velocities

In the preload (or base) analysis, could be generally nonlinear. In the subsequent linear per-
turbation analysis, the material behavior must be linear. The program handles materials in the linear
perturbation analysis differently according to the specified material option (PERTURB) and the mater-
ial properties used. The program uses the consistent material Jacobian of hyperelastic materials at
iteration as the material behavior in the linear perturbation analysis and a harmonic formulation for
hyperviscoelastic material in perturbed full harmonic solutions. For other nonlinear materials, the
program uses the linear portion of the material in the linear perturbation analysis by default. For any
linear materials, the same behaviors are assumed in both the base and the linear perturbation analyses.
For nonlinear materials, an option is available (PERTURB) to use the consistent material tangent Jac-
obian.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
862 Published: 2023-10-23
 Linear Perturbation Analysis

In the case of contact analysis from the static or full transient runs, is by default the consistent
tangent (total) stiffness matrix from the contact elements based on their status at iteration i. However,
you are optionally allowed to change contact behavior and, hence, contact stiffness for all contact
pairs by using the PERTURB command in the linear perturbation analysis or the CNKMOD command
for individual contact pairs. All contact behavior can be changed into the following form (similar to
KEYOPT(12) in the contact element descriptions; for example, CONTA174) from whatever behavior it
was in the previous static or full transient analysis to:

1 -- Rough
2 -- No separation (sliding permitted)
3 -- Bonded
4 -- No separation (always)

15.8.3. Static Analysis Based on Linear Perturbation

A static analysis can be performed by using the linear perturbation analysis procedure. The effect of
the structure from the previous static or full transient analysis is included.

The theoretical steps can be summarized as follows:

1. Perform a static or full transient base analysis; use the RESCONTROL command to create the
restart files for the load points of interest (that is, at various substeps of a load step).

2. Restart the analysis at the load point of interest and issue the commands PERTURB,STATIC
and SOLVE,ELFORM. The program regenerates the total tangent stiffness matrix from
the restart point.

3. For the static analysis, the perturbation load is required and must be defined or modified at
this step (see Application of Perturbation Loads (p. 868) and also the PERTURB command
description).

4. Upon the onset of the second SOLVE command, the nodal coordinates are updated automat-
ically if the base analysis includes large-deflection effects (NLGEOM,ON). The program then
calculates the static solution using the following the equation (similar to Equation 15.1 (p. 794)):
(15.253)

5. Along with the static solution phase, a stress calculation is carried out based on the material
definition specified via the PERTURB command. (See Interpretation of Structural Element
Results After a Linear Perturbation Analysis in the Element Reference for details.)

15.8.4. Modal Analysis Based on Linear Perturbation

A modal analysis can be performed by using the linear perturbation analysis procedure. The effect
of the structure from the previous static or full transient analysis is included. The theoretical steps
can be summarized as follows:

1. Perform a static or full transient base analysis; use the RESCONTROL command to create the
restart files for the load points of interest (that is, at various substeps of a load step).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 863
Analysis Procedures

2. Restart the analysis at the load point of interest and issue the commands PERTURB,MODAL
and SOLVE,ELFORM. The program regenerates the total tangent stiffness matrix from
the restart point.

3. If a downstream analysis is desired using the load vector from the modal analysis, the per-
turbation load can be defined or modified here (see Application of Perturbation Loads (p. 868)
and also the PERTURB command description) for the downstream analyses.

4. Upon the onset of the second SOLVE command, the nodal coordinates are updated automat-
ically if large-deflection was included in the base analysis (NLGEOM,ON). The program then
calculates the eigensolution using the following equation (similar to Equation 14.212 (p. 745)):
(15.254)

where:

= total tangent matrix from the load point of interest

= eigenvector

= structural mass matrix

The structural material behavior in a linear perturbation analysis is discussed above and can
be controlled by the PERTURB command. The UNSYM and DAMP eigensolvers can be used
as well. In the case of the DAMP eigensolution, the following equation (similar to Equa-
tion 14.224 (p. 751)) is solved:
(15.255)

where:

= structural damping matrix

5. Along with the modal solution phase, a stress expansion pass is carried out. (See Interpretation
of Structural Element Results After a Linear Perturbation Analysis in the Element Reference for
details.) To use the appropriate material property and to obtain the total sum of elastic
strain/stress due to the linear perturbation analysis and the base analysis, a stress expansion
pass must be done along with the modal analysis. A separate expansion pass (EXPASS com-
mand) is not allowed after the linear perturbation analysis.

15.8.5. Eigenvalue Buckling Analysis Based on Linear Perturbation

An eigenvalue buckling analysis can be performed by using the linear perturbation analysis procedure.
Unlike the preceding modal analysis description, this procedure is more complicated since the per-
turbation load is a prerequisite for calculating the linearly perturbed stresses; then the perturbed
stresses are used to calculate the linearly perturbed stress stiffening matrix before the buckling ana-
lysis can be started. The ultimate buckling load is calculated differently if the base analysis is linear
or nonlinear.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
864 Published: 2023-10-23
 Linear Perturbation Analysis

Linear base analysis

If the base analysis is linear, the total perturbation load multiplied by the buckling load factor represents
the ultimate buckling load. This procedure is similar to the linear eigenvalue buckling analysis outlined
in Buckling Analysis (p. 826). It can be viewed as restarting the analysis with the tangent matrix
equal to the linear stiffness matrix, and load vector . The prediction of the buckling
load can be written:
(15.256)

Nonlinear base analysis

The theoretical steps for a linear perturbation buckling analysis with a nonlinear base analysis can be
summarized as follows:

1. Perform a static or full transient base analysis; use the RESCONTROL command to create the
restart files for the load points of interest (that is, at various substeps of a load step).

2. Restart the analysis at the load point of interest and issue the commands PERTURB,BUCKLE
and SOLVE,ELFORM. The program regenerates the total tangent stiffness matrix from
the restart point.

3. The perturbation load is required and must be defined or modified at this step (see Application
of Perturbation Loads (p. 868) and also the PERTURB command description) for the calculation
of the linearly perturbed displacements and stresses.

4. Upon the onset of the second SOLVE command, is used to calculate the linearly
perturbed displacement following the equation:
(15.257)

Next, the nodal coordinates are updated automatically if large-deflection was included in the
base analysis (NLGEOM,ON).

Then, internally, is used to calculate the linearly perturbed stresses , and the
linearly perturbed stress stiffening matrix is calculated:

(15.258)

following a formula similar to Equation 3.61 (p. 44).

The program then calculates the eigensolution using the following equation (similar to
Equation 15.113 (p. 827)):
(15.259)

where is the total tangent matrix from the load point of interest. is the
linearly perturbed stress stiffening matrix generated by Equation 15.257 (p. 865) and Equa-
tion 15.258 (p. 865).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 865
Analysis Procedures

5. Along with the eigenvalue buckling solution phase, a stress expansion pass is carried out.
(See Interpretation of Structural Element Results After a Linear Perturbation Analysis in the
Element Reference for details.) To use the appropriate material property and to obtain the
total sum of elastic strain/stress due to the linear perturbation analysis and the base analysis,
a stress expansion pass must be done along with the buckling analysis. A separate expansion
pass (EXPASS command) is not allowed after the linear perturbation analysis.

6. You must perform this step manually. This step determines the total buckling load using the
loads from the restart load point and the additional perturbation load . The user
should manually calculate the total buckling load from the following equation:
(15.260)

where:

= total loads at the current restart load point (load applications are from
the .LDHI file)
= eigenvalues from Equation 15.259 (p. 865). Normally, the lowest eigenvalue (
when j = 1) of Equation 15.259 (p. 865) is of interest.

Equation 15.260 (p. 866) is useful in a way that combines linearly or nonlinearly prestressed
loads with a linearly perturbed load. For example, if the nonlinear static analysis cannot proceed
further due to convergence issues, the buckling loads can still be predicted with this process
under the assumption of linear perturbation.

Generally, eigensolutions for equations Equation 15.113 (p. 827) (purely linear eigenvalue buckling)
and Equation 15.259 (p. 865) (perturbed eigenvalue buckling) can be performed by using the Block
Lanczos eigensolver (BUCOPT,LANB) if stiffness matrix or is non-singular (that is, no rigid
body mechanism is introduced into the model). In the case of purely linear eigenvalue buckling,
is positive definite. However, in linear perturbation, can be indefinite (that is, it contains both
positive and negative pivots if it is factorized by a direct sparse equation solver). The following con-
ditions will likely invoke an indefinite matrix from the base analysis:

1. In the base analysis, the applied load is large enough to trigger nonlinear buckling points in
the nonlinear solution process.

2. In the base analysis, the status of contact elements changes to be open or sliding.

3. In the base and perturbed eigenvalue buckling analyses, mixed u-P formulation (Lagrange
multiplier) elements have been used in the model (see General Element Formulations (p. 49)).

When becomes indefinite (that is, one of the above conditions occurs), the Block Lanczos ei-
gensolver (BUCOPT,LANB) could fail to produce an eigensolution due to the mathematical limitation
of this solver. In such a case, it is recommended that you use the subspace eigensolver (BUCOPT,SUBSP)
to achieve a successful solution. The subspace eigensolver can still give a converged solution even
when both the and matrices are indefinite. However, if the base nonlinear analysis
introduces a singular matrix, then no buckling eigensolutions are possible. This sometimes
happens when the stiffness matrix is highly ill-conditioned.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
866 Published: 2023-10-23
 Linear Perturbation Analysis

15.8.6. Harmonic Analysis Based on Linear Perturbation

Both the full harmonic analysis and the frequency-sweep harmonic analysis can be performed by
using the linear perturbation analysis procedure. The effect of the structure from the previous static
or full transient analysis is included. The theoretical steps can be summarized as follows:

1. Perform a static or full transient base analysis; use the RESCONTROL command to create the
restart files for the load points of interest (that is, at various substeps of a load step).

2. Restart the analysis at the load point of interest and issue the commands PERTURB,HARMONIC
and SOLVE,ELFORM. The program regenerates the total tangent stiffness matrix from
the restart point.

3. For the harmonic analysis, the perturbation load is required and must be defined or modified
here (see Application of Perturbation Loads (p. 868) and also the PERTURB command descrip-
tion).

4. Upon the onset of the second SOLVE command, the nodal coordinates are updated automat-
ically if large-deflection is included in the base analysis (NLGEOM,ON). The program then
calculates the harmonic solution using the following equation (similar to Equa-
tion 15.58 (p. 814)):
(15.261)

where:

= total tangent matrix from the load point of interest

= structural mass matrix

= structural damping matrix

The perturbation load is generally a complex number load vector with frequency
dependency. Also, in the second phase of the linear perturbation analysis, the number of
harmonic solutions (or substeps) is specified by the NSUBST or DELTIM command.

5. Along with the harmonic solution phase, a stress calculation is carried out based on the ma-
terial definition specified via the PERTURB command. (See Interpretation of Structural Element
Results After a Linear Perturbation Analysis in the Element Reference for details.)

Full Harmonic Analysis with Hyperviscoelastic Materials

Set the harmonic frequency (HARFRQ) before solving in the first phase of the linear perturbation
analysis (SOLVE,ELFORM). The tangent stiffness matrix for the hyper viscoelastic materials is reformed
using the specified beginning frequency.

During the perturbed harmonic solution, resetting the harmonic frequency or solving for multiple
frequencies reuses from the solution step but reforms the hyperviscoelastic damping matrix
for the current harmonic frequency.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 867
Analysis Procedures

15.8.7. Substructure or CMS Generation Based on Linear Perturbation

A substructure or CMS (component mode synthesis) generation can be performed by using the linear
perturbation analysis procedure. In this section, the term substructure refers to substructuring and
CMS analyses. The prestressed effect of the structure from the previous static or full transient analysis
is included in the substructure or CMS generation. The theoretical steps can be summarized as follows:

1. Perform a static or full transient base analysis; use the RESCONTROL command to create the
restart files for the load points of interest (that is, at various substeps of a load step).

2. Restart the analysis at the load point of interest and issue the commands PERTURB,SUBSTR
and SOLVE,ELFORM. The program regenerates the total tangent stiffness matrix from
the restart point.

3. For the generation pass, the perturbation load is used to generate the first load vector of the
substructure. The substructure master degrees of freedom (M) must be specified.

4. Upon the onset of the second SOLVE command, the nodal coordinates are updated automat-
ically if large-deflection is included in the base analysis (NLGEOM,ON). The program then
calculates the reduced substructure matrices based on the following equations (similar to
Equation 15.114 (p. 828) to Equation 15.146 (p. 834)).

For a static substructure generation pass:

(15.262)

For a dynamic substructure generation pass:

(15.263)

where:

= total tangent matrix from the load point of interest

= structural mass matrix

= structural damping matrix

In the case of a dynamic analysis, CMS generation is recommended.

5. The generated substructure is assumed to be linear and can be used in a subsequent use
pass. The use pass does not require a linear perturbation procedure.

6. In the expansion pass, all material properties are assumed to be linear, and all contact statuses
are frozen as in a linear application. NLGEOM is turned OFF, even if the base analysis for the
prestressed substructure generation pass is nonlinear. The expansion pass does not require
a linear perturbation procedure.

15.8.8. Application of Perturbation Loads

The total perturbed loads are required for static, eigenvalue buckling, and harmonic analyses and are
optional for modal analyses. One of the obvious reasons they would be needed for a modal analysis
is because a downstream mode-superposition (MSUP) analysis needs loads.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
868 Published: 2023-10-23
 Linear Perturbation Analysis

The total perturbed loads are calculated as follows:

(15.264)

where:

= total loads at the end of the load step of the current restart load point (load applic-
ations are read from the .LDHI file, see Figure 15.8: Linear Perturbation Started from Loadstep
= 2, Substep = 3 (p. 869)). By default, only contains loads contributed by displacement
boundary conditions, inertia loads, and thermal loads (see the PERTURB command).
= additional (new) loads prescribed by the user in the second phase of the linear
perturbation analysis (after the first SOLVE,ELFORM command is invoked). This additional
loading is optional.

Special consideration is required if extra displacement boundary conditions are added or if constraint
equations are added or deleted during the linear perturbation analysis because these modifications
could introduce discontinuities into the structure.

Figure 15.8: Linear Perturbation Started from Loadstep = 2, Substep = 3 (p. 869) provides a graphic
example of the definitions for the different load vectors. Two ramped load steps are shown in the
figure. Since a linear perturbation analysis is started at load step = 2, substep = 3, the load vectors
by definition are = 40 lb and = 25 lb.

Figure 15.8: Linear Perturbation Started from Loadstep = 2, Substep = 3

is the total loads at the current restart point. may be useful for some types of
linear perturbation analysis; for example, to calculate the ultimate buckling loads in an eigenvalue
buckling analysis.

In the first phase of a linear perturbation analysis, the ANTYPE,,RESTART and SOLVE,ELFORM commands
resume the [Link] database and read in the .LDHI file to establish the load.

has a different usage and implication for each perturbed analysis type:

• For Type = STATIC, is the actual external load for the perturbed static analysis.
must be supplied by the user.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 869
Analysis Procedures

• For Type = MODAL, is calculated and stored in the .FULL and .MODE files for
a subsequent mode-superposition, PSD, or other type of modal-based linear dynamic analysis.

• For Type = BUCKLE, is the actual linear buckling load which is used to generate
the linearly perturbed stress stiffening matrix for the eigenvalue buckling analysis.
must be supplied by the user.

• For Type = HARMONIC, is the actual external load for the perturbed full harmonic
analysis. , which must be supplied by the user, can be frequency dependent and
can use complex input.

• For Type = SUBSTR, is used to generate the first reduced external load vector of
the substructure.

15.8.9. Downstream Analysis Using the Solution of a Linear Perturbation

Analysis
Following the linear perturbation analysis, other analysis types can be performed by using the inform-
ation from the linear perturbation analysis.

If the linear perturbation analysis is a modal analysis, the following analysis types are possible by using
the .MODE file generated by the linear perturbation analysis and the database of the model:

• harmonic or transient analysis of the mode-superposition (MSUP) method

• response spectrum analysis

• random vibration analysis

Note that the deformed mesh due to the prior static or full transient analysis is used in the linear
perturbation analysis and in the downstream analysis.

In all the above listed analyses, the first load vector used is from Equation 15.264 (p. 869).
If more loading cases are required, it is required to regenerate a new load vector by using
the MODCONT command. Only linear material properties are supported in these analyses. The program
assumes the analyses are purely linear.

15.9. Semi-Implicit Analysis

The Mechanical APDL solver uses an implicit solution scheme based on Newton-Raphson iterations to
solve the nonlinear mechanical equilibrium equation in static and full transient analyses. Sometimes
the solution does not converge within the convergence criterion and no solution is achieved. The semi-
implicit method (SEMIIMPLICIT command) can alleviate this problem.

The following semi-implicit analysis topics are available:

15.9.1. Assumptions and Restrictions
15.9.2. Description of Analysis

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
870 Published: 2023-10-23
 Semi-Implicit Analysis

15.9.3. Stability and Selective Mass Scaling

15.9.4. Artificial Bulk Viscosity

15.9.1. Assumptions and Restrictions

The following assumptions and restrictions apply to the semi-implicit method:

• This method is available only for static (ANTYPE,STATIC) and full transient (ANTYPE,TRANS) analysis
types.

• Since the semi-implicit method uses the central difference method to solve the transient dynamics
for part of the solution, each node must have a mass associated with it. Therefore, each node must
belong to at least one element that has material density defined.

• The method is available only for structural degrees of freedom. It does not support any element
with non-structural degrees of freedom.

• Some element formulations are not supported. For a complete list, see Semi-Implicit Method Lim-
itations in the Advanced Analysis Guide.

• Nonlinear mesh adaptivity (NLADAPTIVE) and manual rezoning are not supported if the mesh
changes occur before or during the transition to the semi-implicit solve phase, but these capabilities
are supported if the mesh changes occur after the semi-implicit solution has transitioned back to
the implicit solve phase.

15.9.2. Description of Analysis

When an implicit solution fails to converge, the semi-implicit method converts the implicit problem
(static or transient dynamic) into a transient dynamic problem, and the semi-discrete equation of
motion (in the most general form) is given as:
(15.265)

where:

[M] = structural mass matrix

[C] = damping matrix (only present if damping is defined in the initial run)
= nodal acceleration

= nodal velocity

= internal load vector

= applied load vector

The semi-implicit method then solves the mechanical equilibrium using central difference time integ-
ration and, hence, does not require Newton-Raphson iterations (and ensuing convergence checks).

More specifically, the acceleration for substep n is solved as:

(15.266)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 871
Analysis Procedures

And the velocity at the midstep and the displacements for the next substep n+1
are updated as:
(15.267)

(15.268)

(15.269)

After solving the equations in the semi-implicit scheme for some amount of time (which is user-spe-
cified) and overcoming the hurdle that was causing difficulty in convergence, the method transitions
back to the implicit time integration scheme for the remainder of the solution. In the semi-implicit
method, the problem seamlessly transitions between pure implicit and semi-implicit time integration
of the equilibrium equation. The semi-implicit method is a hybrid method with some features of tra-
ditional implicit finite element analysis, including the solution of a coupled system of equations to
obtain nodal acceleration by matrix inversions, the ability to bisect when a solution is unstable, and
the ability to support Lagrange multiplier based constraints and mixed u-P type solid elements. The
method also has some features of traditional explicit finite element analysis, namely the use of central
difference time integration—thus, the name semi-implicit method.

15.9.3. Stability and Selective Mass Scaling

Central difference time integration is conditionally stable. The stability of the central difference scheme
is determined by the Courant condition. For stability, the time increment Δt should be less than the
time associated with the maximum frequency ωmax of the structure, which in turn can be approximated
by the smallest time it takes for sound to traverse through an element; that is:
(15.270)

where:

= element characteristic length

c = speed of sound

The speed of sound c is in turn given by the material properties, density (ρ) and elastic modulus (E).
Thus, the stable time increment depends upon the element size and material density and moduli.
For example, the stable time increment for a 1D element (such as LINK180) with linear elastic moduli
is given by:

(15.271)

where:

me = element mass

The stable time increment increases with increasing mass and increasing mesh size.

With increased mass, the inertia force increases. If the base problem was a quasi-static problem (AN-
TYPE,STATIC), this is especially undesirable as it is assumed that the inertia affects are very small for
the physical problem. However, to be able to solve quasi-static problems using central difference

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
872 Published: 2023-10-23
 Semi-Implicit Analysis

time integration, the stable time increment needs to be reasonably large. This problem is addressed
by a selective mass scaling technique (Olovsson et al. [437] (p. 950)). The selective mass scaling (SMS)
increases the mass such that the stable time increment increases while minimizing the effect of scaled
mass on the inertia force, thus keeping the problem closer to quasi-static analysis as intended.

15.9.4. Artificial Bulk Viscosity

Bulk viscosity is applied by default during the semi-implicit solve phase for all elements that contribute
to the mass matrix, except gasket elements. The force due to bulk viscosity consists of three parts. A
linear term (Landshoff [438] (p. 950)) is introduced to eliminate the unphysical oscillations in the element:
(15.272)

where:

= trace of strain rate tensor

= linear damping ratio (defaults to 1.5)

A quadratic term is used to prevent shock zone collapse under an extremely high velocity gradient
(Von Neumman and Richtmyer [439] (p. 950)):
(15.273)

where:

= quadratic damping ratio (defaults to 0.06)

For shell elements, bulk viscosity force proportional to the rotational DOF is also introduced:
(15.274)

where:

h = shell thickness
= trace of curvature rate tensor
= rotational damping ratio (defaults to 0.06)

An upper limit of bulk viscosity is set such that the force due to bulk viscosity always stays within a
percentage of the element internal force.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 873
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
874 Published: 2023-10-23
Chapter 16: Preprocessing and Postprocessing Tools
The following topics concerning preprocessing and postprocessing tools are available:
16.1. Integration and Differentiation Procedures
16.2. Fourier Coefficient Evaluation
16.3. Statistical Procedures

16.1. Integration and Differentiation Procedures

The following integration and differentiation topics are available:
16.1.1. Single Integration Procedure
16.1.2. Double Integration Procedure
16.1.3. Differentiation Procedure
16.1.4. Double Differentiation Procedure

16.1.1. Single Integration Procedure

(accessed with *VOPER command, INT1 operation; similar capability is in POST26, INT1
command)

Given two vectors Y (parameter Par1) and X (parameter Par2), and an integration constant C1 (input
as CON1), Y* (parameter ParR) is replaced by the accumulated integral of Y over X as follows:

(16.1)

Then for each remaining point in the vector, set:

(16.2)

where:

= integrated value of Y up to point n in the vector

L = length of the vectors

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 875
Preprocessing and Postprocessing Tools

Figure 16.1: Integration Procedure

16.1.2. Double Integration Procedure

(accessed with *VOPER command, INT2 operation)

Given two vectors Y (parameter Par1) and X (parameter Par2), integration constants C1 and C2, (input
as CON1 and CON2) set:

(16.3)

(16.4)

Then, for each remaining point in the vector,

(16.5)

(16.6)

16.1.3. Differentiation Procedure

(accessed with *VOPER Command, DER1 Operation; similar capability is in POST26
DERIV command)

Given two vectors Y (parameter Par1) and X (parameter Par2), the derivative is found by averaging
the slopes of two adjacent intervals (central difference procedure):

(16.7)

A constant second derivative is assumed for the starting and ending intervals.
(16.8)

(16.9)

For DERIV calculation, the first and last terms may differ slightly from that calculated with *VOPER
because DERIV linearly extrapolates these terms from adjacent values.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
876 Published: 2023-10-23
 Fourier Coefficient Evaluation

16.1.4. Double Differentiation Procedure

(accessed by *VOPER command, DER2 Operation)

This is performed by simply repeating the differentiation procedure reported above.

16.2. Fourier Coefficient Evaluation

Fourier coefficients can be evaluated using the *MFOURI command. Given two vectors defining data
points to be fit (parameters CURVE and THETA) and two more vectors defining which terms of the trigo-
nometric series are desired to be computed (parameters MODE and ISYM), the desired coefficients can
be computed (parameter COEFF). The curve fitting cannot be perfect, as there are more data than un-
knowns. Thus, an error Ri will exist at each data point:

(16.10)

where:

Ri = error term (residual) associated with data point i

A = desired coefficients of Fourier series (parameter COEFF)
θi = angular location of data points i (parameter THETA)
L = number of terms in Fourier series
F = sine or cosine, depending on ISYM (parameter ISYM)
M = multiplier on θi (parameter MODE)
Ci = value of data point i (parameter CURVE)
m = number of data points (length of CURVE parameter array)

Equation 16.10 (p. 877) can be reduced to matrix form as:

(16.11)

where:

{R} = vector of error terms

{G} = matrix of sines and cosines, evaluated at the different data points
{A} = vector of desired coefficients

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 877
Preprocessing and Postprocessing Tools

{C} = vector of data points

Note that m > L. If m = L, the coefficients would be uniquely determined with {R} = {O} and Equa-
tion 16.11 (p. 877) being solved for {A} by direct inversion.

The method of least squares is used to determine the coefficients {A}. This means that is to

be minimized. The minimization is represented by

(16.12)

where Aj is the jth component of {A}. Note that

(16.13)

The form on the left-hand side of Equation 16.13 (p. 878) is the more convenient to use. Performing this
operation on Equation 16.11 (p. 877),
(16.14)

Minimizing this with respect to {A}T (Equation 16.12 (p. 878)), it may be shown that:
(16.15)

or
(16.16)

Equation 16.16 (p. 878) is known as the "normal equations" used in statistics. Finally,
(16.17)

[GT] could not have been "cancelled out" of Equation 16.16 (p. 878) because it is not a square matrix.
However, [G]T[G] is square.

In spite of the orthogonal nature of a trigonometric series, the value of each computed coefficient is
dependent on the number of terms requested because of the least squares fitting procedure which
takes place at the input data points. Terms of a true Fourier series are evaluated not by a least squares
fitting procedure, but rather by the integration of a continuous function (for example, Euler formulas,
p. 469 of Kreyszig([24] (p. 926))).

16.3. Statistical Procedures

The following statistical procedures topics are available:
16.3.1. Mean, Covariance, Correlation Coefficient
16.3.2. Random Samples of a Uniform Distribution
16.3.3. Random Samples of a Gaussian Distribution
16.3.4. Random Samples of a Triangular Distribution

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
878 Published: 2023-10-23
 Statistical Procedures

16.3.5. Random Samples of a Beta Distribution

16.3.6. Random Samples of a Gamma Distribution

16.3.1. Mean, Covariance, Correlation Coefficient

The mean, variance, covariance, and correlation coefficients of a multiple subscripted parameter are
computed (using the *MOPER command). Refer to Kreyszig([163] (p. 934)) for the basis of the following
formulas. All operations are performed on columns to conform to the database structure. The covariance
is assumed to be a measure of the association between columns.

The following notation is used:

where:

[x] = starting matrix

i = row index of first array parameter matrix
j = column index of first array parameter matrix
m = number of rows in first array parameter matrix
n = number of columns in first array parameter matrix
subscripts s, t = selected column indices
[S] = covariance matrix n x n
[c] = correlation matrix n x n
= variance

The mean of a column is:

(16.18)

The covariance of the columns s and t is:

(16.19)

The variance, , of column s is the diagonal term Sss of the covariance matrix [S]. The equivalent
common definition of variance is:

(16.20)

The correlation coefficient is a measure of the independence or dependence of one column to the
next. The correlation and mean operations are based on Hoel([164] (p. 934)) (and initiated when CORR
is inserted in the Oper field of the *MOPER command).

Correlation coefficient:

(16.21)

value S of the terms of the coefficient matrix range from -1.0 to 1.0 where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 879
Preprocessing and Postprocessing Tools

-1.0 = fully inversely related

0.0 = fully independent
1.0 = fully directly related

16.3.2. Random Samples of a Uniform Distribution

A vector can be filled with a random sample of real numbers based on a uniform distribution with
given lower and upper bounds (using RAND in the Func field on the *VFILL command) (see Fig-
ure 16.2: Uniform Density (p. 880)):
(16.22)

where:

a = lower bound (input as CON1 on *VFILL command)

b = upper bound (input as CON2 on *VFILL command)

Figure 16.2: Uniform Density

The numbers are generated using the URN algorithm of Swain and Swain([162] (p. 934)). The initial
seed numbers are hard coded into the routine.

16.3.3. Random Samples of a Gaussian Distribution

A vector may be filled with a random sample of real numbers based on a Gaussian distribution with
a known mean and standard deviation (using GDIS in the Func field on the *VFILL command).

First, random numbers P(x), with a uniform distribution from 0.0 to 1.0, are generated using a random
number generator. These numbers are used as probabilities to enter a cumulative standard normal
probability distribution table (Abramowitz and Stegun([161] (p. 934))), which can be represented by
Figure 16.3: Cumulative Probability Function (p. 880) or the Gaussian distribution function:

Figure 16.3: Cumulative Probability Function

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
880 Published: 2023-10-23
 Statistical Procedures

(16.23)

where:

f(t) = Gaussian density function

The table maps values of P(x) into values of x, which are standard Gaussian distributed random
numbers from -5.0 to 5.0, and satisfy the Gaussian density function (Figure 16.4: Gaussian Dens-
ity (p. 881)):

Figure 16.4: Gaussian Density

(16.24)

where:

μ = mean (input as CON1 on *VFILL command)

σ = standard deviation (input as CON2 on *VFILL command)

The x values are transformed into the final Gaussian distributed set of random numbers, with the
given mean and standard deviation, by the transformation equation:
(16.25)

16.3.4. Random Samples of a Triangular Distribution

A vector may be filled with a random sample of real numbers based on a triangular distribution with
a known lower bound, peak value location, and upper bound (using TRIA in the Func field on the
*VFILL command).

First, random numbers P(x) are generated as in the Gaussian example. These P(x) values (probabilities)
are substituted into the triangular cumulative probability distribution function:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 881
Preprocessing and Postprocessing Tools

(16.26)

where:

a = lower bound (input as CON1 on *VFILL command)

c = peak location (input as CON2 on *VFILL command)
b = upper bound (input as CON3 on *VFILL command)

which is solved for values of x. These x values are random numbers with a triangular distribution, and
satisfy the triangular density function (Figure 16.5: Triangular Density (p. 882)):

Figure 16.5: Triangular Density

(16.27)

16.3.5. Random Samples of a Beta Distribution

A vector may be filled with a random sample of real numbers based on a beta distribution with known
lower and upper bounds and α and β parameters (using BETA in the Func field on the *VFILL com-
mand).

First, random numbers P(x) are generated as in the Gaussian example. These random values are used
as probabilities to enter a cumulative beta probability distribution table, generated by the program.
This table can be represented by a curve similar to (Figure 16.3: Cumulative Probability Function (p. 880)),
or the beta cumulative probability distribution function:

(16.28)

The table maps values of P(x) into x values which are random numbers from 0.0 to 1.0. The values of
x have a beta distribution with given α and β values, and satisfy the beta density function (Fig-
ure 16.6: Beta Density (p. 883)):

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
882 Published: 2023-10-23
 Statistical Procedures

Figure 16.6: Beta Density

(16.29)

where:

a = lower bound (input as CON1 on *VFILL command)

b = upper bound (input as CON2 on *VFILL command)
α = alpha parameter (input as CON3 on *VFILL command)
β = beta parameter (input as CON4 on *VFILL command)
B (α, β) = beta function

f(t) = beta density function

The x values are transformed into the final beta distributed set of random numbers, with given lower
and upper bounds, by the transformation equation:
(16.30)

16.3.6. Random Samples of a Gamma Distribution

A vector may be filled with a random sample of real numbers based on a gamma distribution with
a known lower bound for α and β parameters (using GAMM in the Func field on the *VFILL command).

First, random numbers P(x) are generated as in the Gaussian example. These random values are used
as probabilities to enter a cumulative gamma probability distribution table, generated by the program.
This table can be represented by a curve similar to Figure 16.7: Gamma Density (p. 884), or the gamma
cumulative probability distribution function:

(16.31)

where:

f(t) = gamma density function.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 883
Preprocessing and Postprocessing Tools

The table maps values of P(x) into values of x, which are random numbers having a gamma distribution
with given α and β values, and satisfy the gamma distribution density function (Figure 16.7: Gamma
Density (p. 884)):

Figure 16.7: Gamma Density

(16.32)

where:

α = alpha parameter of gamma function (input as CON2 on *VFILL command)

β = beta parameter of gamma density function (input as CON3 on *VFILL command)
a = lower bound (input as CON1 on *VFILL command)

The x values are relocated relative to the given lower bound by the transformation equation:
(16.33)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
884 Published: 2023-10-23
Chapter 17: Postprocessing
The following postprocessing topics are available:
17.1. POST1 - Derived Nodal Data Processing
17.2. POST1 - Vector Operations
17.3. POST1 - Path Operations
17.4. POST1 - Stress Linearization
17.5. POST1 - Electromagnetic Macros
17.6. POST1 - Error Approximation Technique
17.7. POST1 - Harmonic Solid and Shell Element Postprocessing
17.8. POST26 - Data Operations
17.9. POST26 - Response Spectrum Generator (RESP)
17.10. POST1 and POST26 - Interpretation of Equivalent Strains
17.11. POST26 - Response Power Spectral Density
17.12. POST26 - Computation of Covariance
17.13. POST1 and POST26 – Complex Results Postprocessing
17.14. POST1 - Modal Assurance Criterion (MAC)
17.15. POST1 – Frequency response function correlation

17.1. POST1 - Derived Nodal Data Processing

17.1.1. Derived Nodal Data Computation

The computation of derived data (data derived from nodal unknowns) is discussed in Structures with
Geometric Nonlinearities (p. 29) through Acoustics (p. 241). Derived nodal data is available for solid
and shell elements (except SHELL61). Available data include stresses, strains, thermal gradients, thermal
fluxes, pressure gradients, electric fields, electric flux densities, magnetic field intensities, magnetic
flux densities, and magnetic forces. Structural nonlinear data is processed in a similar fashion and in-
cludes equivalent stress, stress state ratio, hydrostatic pressure, accumulated equivalent plastic strain,
plastic state variable, and plastic work.

POST1 averages the component tensor or vector data at corner nodes used by more than one element.

(17.1)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 885
Postprocessing

σik = average derived data component i at node k

σijk = derived data component i of element j at node k
Nk = number of elements connecting to node k

For higher-order elements, component tensor or vector data at midside nodes are calculated by directly
averaging the averaged corner node values, so Equation 17.1 (p. 885) is not used for midside nodes.
Midside node values are printed or plotted only via PowerGraphics (/GRAPHICS,POWER) and /EFACET,2.

Combining principal tensor data (principal stress, principal strain) or vector magnitudes at the nodes
may either be computed using the averaged component data (KEY = 0, AVPRIN command):
(17.2)

where:

f(σik) = function to compute principal data from component data as given in Structures with
Geometric Nonlinearities (p. 29) through Acoustics (p. 241).

or be directly averaged (KEY = 1, AVPRIN command):

(17.3)

where:

σck = averaged combined principal data at node k

σcjk = combined principal data of element j at node k

17.2. POST1 - Vector Operations

The dot product of two vectors and is provided (with the
VDOT command) as:
(17.4)

The cross product of two vectors {A} and {B} is also provided (with the VCROSS command) as:

(17.5)

In both operations, the components of vectors {A} and {B} are transformed to global Cartesian coordinates
before the calculations. The results of the cross product are also in global Cartesian coordinates.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
886 Published: 2023-10-23
 POST1 - Path Operations

17.3. POST1 - Path Operations

General vector calculus may be performed along any arbitrary 2D or 3D path through a solid element
model. Nodal data, element data, and data stored with element output tables (ETABLE command) may
be mapped onto the path and operated on as described below.

17.3.1. Defining the Path

A path is defined by first establishing path parameters (PATH command) and then defining path
points which create the path (PPATH command). The path points may be nodes, or arbitrary points
defined by geometry coordinates. A segment is a line connecting two path points. The number of
path points used to create a path and the number of divisions used to discretize the path are input
(using Npts and the nDiv parameter on the PATH command). The discretized path divisions are inter-
polated between path points in the currently active coordinate system (CSYS command), or as directly
input (on the PPATH command). A typical segment is shown in Figure 17.1: Typical Path Seg-
ment (p. 887) as going from points N1 to N2, for the first segment.

The geometry of each point along the path is stored. The geometry consists of the global Cartesian
coordinates (output label XG, YG, ZG) and the length from the first path point along the path (output
label S). The geometry is available for subsequent operations.

Figure 17.1: Typical Path Segment

17.3.2. Defining Orientation Vectors of the Path

In addition, position (R), unit tangent (T), and unit normal (N) vectors to a path point are available as
shown in Figure 17.2: Position and Unit Vectors of a Path (p. 888). These three vectors are defined in
the active Cartesian coordinate system.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 887
Postprocessing

Figure 17.2: Position and Unit Vectors of a Path

The position vector R (stored with PVECT,RADI command) is defined as:

(17.6)

where:

xn = x coordinate in the active Cartesian system of path point n, etc.

The unit tangent vector T (stored with PVECT,TANG command) is defined as:

(17.7)

(17.8)

(17.9)

where:

x, y, z = coordinate of a path point in the active Cartesian system n = 2 to (L-1)

L = number of points on the path
C = scaling factor so that {T} is a unit vector

The unit normal vector N (PVECT,NORM command) is defined as:

(17.10)

where:

x = cross product operator

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
888 Published: 2023-10-23
 POST1 - Path Operations

{N} is not defined if {T} is parallel to {k}.

17.3.3. Mapping Nodal and Element Data onto the Path

Having defined the path, the nodal or element data (as requested by Item,Comp on the PDEF com-
mand) may be mapped onto the path. For each path point, the selected elements are searched to
find an element containing that geometric location. In the lower order finite element example of
Figure 17.3: Mapping Data (p. 889), point No has been found to be contained by the element described
by nodes Na, Nb, Nc and Nd. Nodal degree of freedom data is directly available at nodes Na, Nb, Nc
and Nd. Element result data may be interpreted either as averaged data over all elements connected
to a node (as described in the Nodal Data Computation topic, see POST1 - Derived Nodal Data Pro-
cessing (p. 885)) or as unaveraged data taken only from the element containing the path interpolation
point (using the Avglab option on the PDEF command). When using the material discontinuity option
(MAT option on the PMAP command) unaveraged data is mapped automatically.

Caution should be used when defining a path for use with the unaveraged data option. Avoid defining
a path (PPATH command) directly along element boundaries since the choice of element for data
interpolation may be unpredictable. Path values at nodes use the element from the immediate pre-
ceding path point for data interpolation.

The value at the point being studied (that is, point No) is determined by using the element shape
functions together with these nodal values. Principal results data (principal stresses, strains, flux
density magnitude, etc.) are mapped onto a path by first interpolating item components to the path
and then calculating the principal value from the interpolated components.

Figure 17.3: Mapping Data

Higher order elements include midside nodal (DOF) data for interpolation. Element data at the midside
nodes are averaged from corner node values before interpolation.

17.3.4. Operating on Path Data

Once nodal or element data are defined as a path item, its associated path data may be operated on
in several ways. Path items may be combined by addition, multiplication, division, or exponentiation

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 889
Postprocessing

(PCALC command). Path items may be differentiated or integrated with respect to any other path
item (PCALC command). Differentiation is based on a central difference method without weighting:
(17.11)

(17.12)

(17.13)

where:

A = values associated with the first labeled path in the operation (LAB1, on the PCALC,DERI
command)
B = values associated with the second labeled path in the operation (LAB2, on the PCALC,DERI
command)
n = 2 to (L-1)
L = number of points on the path
S = scale factor (input as FACT1, on the PCALC,DERI command)

If the denominator is zero for Equation 17.11 (p. 890) through Equation 17.13 (p. 890), then the derivative
is set to zero.

Integration is based on the rectangular rule (see Figure 16.1: Integration Procedure (p. 876) for an illus-
tration):
(17.14)
(17.15)

Path items may also be used in vector dot (PDOT command) or cross (PCROSS command) products.
The calculation is the same as the one described in the Vector Dot and Cross Products Topic, above.
The only difference is that the results are not transformed to be in the global Cartesian coordinate
system.

17.4. POST1 - Stress Linearization

An option is available to allow a separation of stresses through a section into constant (membrane) and
linear (bending) stresses. An approach similar to the one used here is reported by Gordon([64] (p. 928)).
The stress linearization option (accessed using the PRSECT, PLSECT, or FSSECT commands) uses a path
defined by two nodes (with the PPATH command). The section is defined by a path consisting of two
end points (nodes N1 and N2) as shown in Figure 17.4: Coordinates of Cross Section (p. 891) (nodes) and
47 intermediate points (automatically determined by linear interpolation in the active display coordinate
system (DSYS). Nodes N1 and N2 are normally both presumed to be at free surfaces.

Initially, a path must be defined and the results mapped onto that path as defined above. The logic for
most of the remainder of the stress linearization calculation depends on whether the structure is
axisymmetric or not, as indicated by the value of ρ (input as RHO on PRSECT, PLSECT, or FSSECT
commands). For ρ = 0.0, the structure is not axisymmetric (Cartesian case); and for nonzero values of ρ,

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
890 Published: 2023-10-23
 POST1 - Stress Linearization

the structure is axisymmetric. The explicit definition of ρ, as well as the discussion of the treatment of
axisymmetric structures, is discussed later.

Figure 17.4: Coordinates of Cross Section

17.4.1. Cartesian Case

Refer to Figure 17.5: Typical Stress Distribution (p. 891) for a graphical representation of stresses. The
membrane values of the stress components are computed from:

(17.16)

where:

= membrane value of stress component i

t = thickness of section, as shown in Figure 17.4: Coordinates of Cross Section (p. 891)
σi = stress component i along path from results file (`total' stress)
xs = coordinate along path, as shown in Figure 17.4: Coordinates of Cross Section (p. 891)

Figure 17.5: Typical Stress Distribution

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 891
Postprocessing

The subscript i is allowed to vary from 1 to 6, representing σx, σy, σz, σxy, σyz, and σxz, respectively.
These stresses are in global Cartesian coordinates. Strictly speaking, the integrals such as the one
above are not literally performed; rather it is evaluated by numerical integration:

(17.17)

where:

σi,j = total stress component i at point j along path

The integral notation will continue to be used, for ease of reading.

The "bending" values of the stress components at node N1 are computed from:

(17.18)

where:

= bending value of stress component i at node N1

The bending values of the stress components at node N2 are simply

(17.19)

where:

= bending value of the stress component i at node N2

The "peak" value of stress at a point is the difference between the total stress and the sum of the
membrane and bending stresses. Thus, the peak stress at node N1 is:
(17.20)

where:

= peak value of stress component i at node N1

σi1 = value of total stress component i at node N1

Similarly, for node N2,

(17.21)

At the center point (x = 0.0)

(17.22)

where:

= peak value of stress component i at center

σic = computed (total) value of stress component i at center

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
892 Published: 2023-10-23
 POST1 - Stress Linearization

17.4.2. Axisymmetric Case (General)

The axisymmetric case is the same, in principle, as the Cartesian case, except for the fact that there
is more material at a greater radius than at a smaller radius. Thus, the neutral axis is shifted radially
outward a distance xf, as shown in Figure 17.6: Axisymmetric Cross-Section (p. 893). The axes shown
in Figure 17.6: Axisymmetric Cross-Section (p. 893) are Cartesian, that is, the logic presented here is
only valid for structures axisymmetric in the global cylindrical system. As stated above, the axisym-
metric case is selected if ρ ≠ 0.0. ρ is defined as the radius of curvature of the midsurface in the X-Y
plane, as shown in Figure 17.7: Geometry Used for Axisymmetric Evaluations (p. 893). A point on the
centerplane of the torus has its curvatures defined by two radii: ρ and the radial position Rc. Both of
these radii will be used in the forthcoming development. In the case of an axisymmetric straight
section such as a cylinder, cone, or disk, ρ = , so that the input must be a large number (or -1).

Figure 17.6: Axisymmetric Cross-Section

Figure 17.7: Geometry Used for Axisymmetric Evaluations

Each of the components for the axisymmetric case needs to be treated separately. For this case, the
stress components are rotated into section coordinates, so that x stresses are parallel to the path and
y stresses are normal to the path.

Starting with the y direction membrane stress, the force over a small sector is:

(17.23)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 893
Postprocessing

where:

Fy = total force over small sector

σy = actual stress in y (meridional) direction
R = radius to point being integrated
Δθ = angle over a small sector in the hoop direction
t = thickness of section (distance between nodes N1 and N2)

The area over which the force acts is:

(17.24)

where:

Ay = area of small sector

R1 = radius to node N1
R2 = radius to node N2

Thus, the average membrane stress is:

(17.25)

where:

= y membrane stress

To process the bending stresses, the distance from the center surface to the neutral surface is needed.
This distance is shown in Figure 17.6: Axisymmetric Cross-Section (p. 893) and is:

(17.26)

The derivation of Equation 17.26 (p. 894) is the same as for yf given at the end of SHELL61 - Axisym-
metric-Harmonic Structural Shell (p. 488). Thus, the bending moment may be given by:

(17.27)

or

(17.28)

The moment of inertia is:

(17.29)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
894 Published: 2023-10-23
 POST1 - Stress Linearization

The bending stresses are:

(17.30)

where:

c = distance from the neutral axis to the extreme fiber

Combining the above three equations,

(17.31)

or

(17.32)

where:

= y bending stress at node N1

Also,
(17.33)

or

(17.34)

where:

= y bending stress at node N2

σx represents the stress in the direction of the thickness. Thus, σx1 and σx2 are the negative of the
pressure (if any) at the free surface at nodes N1 and N2, respectively. A membrane stress is computed
as:

(17.35)

where:

= the x membrane stress

The treatment of the thickness-direction "bending" stresses is controlled by KB (input as KBR on PRSECT,
PLSECT, or FSSECT commands). When the thickness-direction bending stresses are to be ignored (KB
= 1), bending stresses are equated to zero:
(17.36)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 895
Postprocessing

(17.37)

When the bending stresses are to be included (KB = 0), bending stresses are computed as:
(17.38)
(17.39)

where:

= x bending stress at node N1

σx1 = total x stress at node N1
= x bending stress at node N2
σx2 = total x stress at node N2

and when KB = 2, membrane and bending stresses are computed using Equation 17.25 (p. 894),
Equation 17.32 (p. 895), and Equation 17.34 (p. 895) substituting σx for σy.

The hoop stresses are processed next.

(17.40)

where:

= hoop membrane stress

Fh = total force over small sector
Δφ = angle over small sector in the meridional (y) direction
σh = hoop stress
Ah = area of small sector in the x-y plane
r = radius of curvature of the midsurface of the section (input as RHO)
x = coordinate thru cross-section
t = thickness of cross-section

Equation 17.40 (p. 896) can be reduced to:

(17.41)

Using logic analogous to that needed to derive Equation 17.32 (p. 895) and Equation 17.34 (p. 895),
the hoop bending stresses are computed by:

(17.42)

and

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
896 Published: 2023-10-23
 POST1 - Stress Linearization

(17.43)

where:
(17.44)

for hoop-related calculations of Equation 17.42 (p. 896) and Equation 17.43 (p. 897).

An xy membrane shear stress is computed as:

(17.45)

where:

= xy membrane shear stress

σxy = xy shear stress

Since the shear stress distribution is assumed to be parabolic and equal to zero at the ends, the xy
bending shear stress is set to 0.0. The other two shear stresses (σxz, σyz) are assumed to be zero if KB
= 0 or 1. If KB = 2, the shear membrane and bending stresses are computing using Equa-
tion 17.25 (p. 894), Equation 17.32 (p. 895), and Equation 17.34 (p. 895) substituting σxy for σy

All peak stresses are computed from

(17.46)

where:

σi P = peak value of stress component i

σi = total value of stress of component i

17.4.3. Axisymmetric Case

(Specializations for Centerline)

At this point it is important to mention one exceptional configuration related to the y-direction
membrane and bending stress calculations above. For paths defined on the centerline (X = 0), Rc =
0 and cosΦ = 0, and therefore Equation 17.25 (p. 894), Equation 17.26 (p. 894), Equation 17.32 (p. 895),
and Equation 17.34 (p. 895) are undefined. Since centerline paths are also vertical (φ = 90°), it follows
that R = Rc, and Rc is directly cancelled from stress Equation 17.25 (p. 894), Equation 17.32 (p. 895), and
Equation 17.33 (p. 895). However, xf remains undefined. Figure 17.8: Centerline Sections (p. 898) shows
a centerline path from N1 to N2 in which the inside and outside wall surfaces form perpendicular in-
tersections with the centerline.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 897
Postprocessing

Figure 17.8: Centerline Sections

For this configuration it is evident that cos φ = Rc/ρ as φ approaches 90° (or as approaches
N1 - N2). Thus for any paths very near or exactly on the centerline, Equation 17.26 (p. 894) is generalized
to be:

(17.47)

The second option of Equation 17.47 (p. 898) applied to centerline paths is an accurate representation
for spherical/elliptical heads and flat plates. It is incorrect for axisymmetric shapes that do not form
perpendicular intersections with the centerline (e.g., conical heads). For such shapes (as shown in
Figure 17.9: Non-Perpendicular Intersections (p. 898)) centerline paths must not be selected.

Figure 17.9: Non-Perpendicular Intersections

17.5. POST1 - Electromagnetic Macros

Electromagnetic macros are macro files created to perform specific postprocessing operations for elec-
tromagnetic field analysis. Macros performing computational analysis are detailed in this section.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
898 Published: 2023-10-23
 POST1 - Electromagnetic Macros

17.5.1. Flux Passing Thru a Closed Contour

The flux passing through a surface defined by a closed line contour (PPATH command) is computed
(using the FLUXV command macro). The macro is applicable to 2D and 3D magnetic field analysis
employing the magnetic vector potential A. For 2D planar analyses, the flux value is per unit depth.

The flux passing through a surface S can be calculated as:

(17.48)

where:

φ = flux enclosed by the bounding surface S

{B} = flux density vector
{n} = unit normal vector
area = area of the bounding surface S

Equation 17.48 (p. 899) can be rewritten in terms of the definition of the vector potential as:

(17.49)

where:

{A} = magnetic vector potential

By applying Stokes theorem, the surface integral reduces to a line integral of A around a closed contour;

(17.50)

where:

= length of the bounding contour line

The macro interpolates values of the vector potential, A, to the closed contour path (defined by the
PPATH command) and integrates to obtain the flux using Equation 17.50 (p. 899). In the axisymmetric
case, the vector potential is multiplied by 2πr to obtain the total flux for a full circumferential surface
(where "r" is the x-coordinate location of the interpolation point).

17.5.2. Magnetomotive Forces

The magnetomotive force (current) along a contour or path (defined by the PPATH command) is
calculated (using the MMF command macro) according to Amperes' theorem:

(17.51)

where:

Immf = magnetomotive force

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 899
Postprocessing

{H} = magnetic field intensity vector

The macro interpolates values of magnetic field intensity, H, to the path and integrates to obtain the
Immf as in Equation 17.51 (p. 899). In a static analysis or transverse electromagnetic (TEM) and transverse
electric (TE) wave guide mode computation, Immf can be interpreted as a current passing the surface
bounded by the closed contour.

17.5.3. Power Loss

The power dissipated in a conducting solid body under the influence of a time-harmonic electromag-
netic field is computed (using the POWERH command macro). The r.m.s. power loss is calculated
from the equation (see Harmonic Analysis Using Complex Formalism (p. 171) for further details):

(17.52)

where:

Prms = rms power loss

r = material resistivity
Jt = total current density
~ = complex quantity

The macro evaluates Equation 17.52 (p. 900) by integrating over the selected element set according
to:

(17.53)

where:

n = number of elements
Re{ } = real component of a complex quantity
[ρi] = resistivity tensor (matrix)

= total eddy current density vector for element i

voli = element volume
* = complex conjugate operator

For 2D planar analyses, the resulting power loss is per unit depth.

17.5.4. Energy Supplied

The energy supplied to the coil for a linear system is calculated as:

(17.54)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
900 Published: 2023-10-23
 POST1 - Electromagnetic Macros

W = energy input to coil

{A} = nodal vector potential
{Js} = d.c. source current density
vol = volume of the coil

17.5.5. Terminal Inductance

The inductance as seen by the terminal leads of the coil is calculated as:
(17.55)

where:

L = terminal inductance
i = coil current (per turn)

17.5.6. Flux Linkage

The total flux linkage of a coil can be calculated from the terminal inductance and coil current,
(17.56)

where:

λ = flux linkage

17.5.7. Terminal Voltage

For a coil operating with an a.c. current at frequency ω (Hz), a voltage will appear at the terminal
leads. Neglecting skin effects and saturation, a static analysis gives the correct field distribution. For
the assumed operating frequency, the terminal voltage can be found. From Faraday's law,
(17.57)

where:

u = terminal voltage

Under a sinusoidal current at an operating frequency ω, the flux linkage will vary sinusoidally
(17.58)

where:

λm = zero-to-peak magnitude of the flux linkage

The terminal voltage is therefore:

(17.59)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 901
Postprocessing

U = ωλm = zero-to-peak magnitude of the terminal voltage (parameter VLTG returned by the
macro)

For 2D planar analyses, the results are per unit depth.

17.5.8. Energy in a Magnetic Field

The stored energy and co-energy in a magnetic field are calculated (by the SENERGY command
macro). For the static or transient analysis, the stored magnetic energy is calculated as:

(17.60)

where:

Ws = stored magnetic energy

The magnetic co-energy is calculated as:

(17.61)

where:

Wc = stored magnetic co-energy

Hc = coercive force

For time-harmonic analysis, the r.m.s. stored magnetic energy is calculated as:
(17.62)

where:

Wrms = r.m.s. stored energy

For 2D planar analyses, the results are per unit depth.

17.5.9. Relative Error in Electrostatic or Electromagnetic Field Analysis

The relative error in an electrostatic or electromagnetic field analysis is computed (by the EMAGERR
command macro). The relative error measure is based on the difference in calculated fields between
a nodal-averaged continuous field representation and a discontinuous field represented by each indi-
vidual element's-nodal field values. An average error for each element is calculated. Within a material,
the relative error is calculated as:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
902 Published: 2023-10-23
 POST1 - Electromagnetic Macros

[Link]. Electrostatics
[Link].1. Electric Field

(17.63)

where:

Eei = relative error for the electric field (magnitude) for element i
Ej = nodal averaged electric field (magnitude)
Eij = electric field (magnitude) of element i at node j
n = number of vertex nodes in element i

[Link].2. Electric Flux Density

(17.64)

where:

Dei = relative error for the electric flux density (magnitude) for element i
Dj = nodal averaged electric flux density (magnitude)
Dij = electric flux density (magnitude) of element i at node j

A normalized relative error norm measure is also calculated based on the maximum element
nodal calculated field value in the currently selected element set.
(17.65)

where:

Emax = maximum element nodal electric field (magnitude)

(17.66)

where:

Dmax = maximum element nodal electric flux density (magnitude)

[Link]. Electromagnetics
[Link].1. Magnetic Field Intensity

(17.67)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 903
Postprocessing

Hei = relative error for the magnetic field intensity (magnitude) for element i
Hj = nodal averaged magnetic field intensity (magnitude)
Hij = magnetic field intensity (magnitude) of element i at node j

[Link].2. Magnetic Flux Density

(17.68)

where:

Bei = relative error for the magnetic flux density (magnitude) for element i
Bj = nodal averaged magnetic flux density (magnitude)
Bij = magnetic flue density (magnitude) of element i at node j

A normalized relative error measure is also calculated based on the maximum element nodal
calculated field value in the currently selected element set.
(17.69)

where:

Hmax = maximum element nodal magnetic field intensity (magnitude)

(17.70)

where:

Bmax = maximum nodal averaged magnetic flux density (magnitude)

17.5.10. Electromotive Force

The electromotive force (voltage drop) between two conductors defined along a path contour (PATH
command) is computed (using the EMF command macro):

(17.71)

where:

Vemf = electromotive force (voltage drop)

{E} = electric field vector

The macro interpolates values of the electric field, E, to the path (defined by the PATH command)
and integrates to obtain the electromotive force (voltage drop). The path may span multiple materials
of differing permittivity. At least one path point should reside in each material transversed by the
path. In static analysis or transverse electromagnetic (TEM) and transverse magnetic (TM) wave guide
mode computation, Vemf can be interpreted as a voltage drop.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
904 Published: 2023-10-23
 POST1 - Electromagnetic Macros

17.5.11. Computation of Equivalent Transmission-line Parameters

The equivalent transmission-line parameters for a guiding wave structure are calculated. For a lossless
guiding structure, the total mode voltage, V(Z), and mode current, I(Z), associated with a +Z
propagating field take on the form:
(17.72)

(17.73)

where:

Zo = characteristic impedance for any mode

A = amplitude of the incident voltage wave (see below)
B = amplitude of the backscattered voltage wave (see below)

We can consider the propagating waves in terms on an equivalent two-wire transmission line termin-
ated at Z = by a load impedance .

Figure 17.10: Equivalent Two-Wire Transmission Line

The voltage term "A" in Equation 17.72 (p. 905) can be considered as the amplitude of the incident
wave, and voltage term "B" as the amplitude of the mode voltage wave backscattered off the load
impedance .

Thus,
(17.74)

Rearranging we have,

(17.75)

where:

Γ = voltage reflection coefficient

The voltage standing-wave ratio is calculated as:

(17.76)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 905
Postprocessing

S = voltage standing-wave ratio (output as VSWR)

For a matched load ( = Zo) there is no reflection (Γ = 0) and the S = 1. If is a short circuit, B = -
A, Γ = -1, and the S is infinite. If is an open circuit, B = A, Γ = +1, and the S once again is infinite.

The reflection coefficient is frequently expressed in dB form by introducing the concept of return loss
defined by:
(17.77)

where:

LR = return loss in dB (output as RL)

The macro calculates the above transmission line parameters in terms of the incident, reference and
total voltage.

17.6. POST1 - Error Approximation Technique

17.6.1. Error Approximation Technique for Displacement-Based Problems

The error approximation technique used by POST1 (PRERR command) for displacement-based problems
is similar to that given by Zienkiewicz and Zhu([103] (p. 930)). The essentials of the method are sum-
marized below.

The usual continuity assumption used in many displacement based finite element formulations results
in a continuous displacement field from element to element, but a discontinuous stress field. To obtain
more acceptable stresses, averaging of the element nodal stresses is done. Then, returning to the
element level, the stresses at each node of the element are processed to yield:
(17.78)

where:

= stress error vector at node n of element i

= number of elements connecting to node n

= stress vector of node n of element i

Then, for each element

(17.79)

where:

ei = energy error for element i (accessed with ETABLE (SERR item) command)
vol = volume of the element (accessed with ETABLE (VOLU item) command)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
906 Published: 2023-10-23
 POST1 - Error Approximation Technique

[D] = stress-strain matrix evaluated at reference temperature

{Δσ} = stress error vector at points as needed (evaluated from all {Δσn} of this element)

The energy error over the model is:

(17.80)

where:

e = energy error over the entire (or part of the) model (accessed with *GET (SERSM item)
command)
Nr = number of elements in model or part of model

The energy error can be normalized against the strain energy.

(17.81)

where:

E = percentage error in energy norm (accessed with PRERR, PLDISP, PLNSOL (U item), *GET
(SEPC item) commands)
U = strain energy over the entire (or part of the) model (accessed with *GET (SENSM item)
command)

= strain energy of element i (accessed with ETABLE (SENE item) command) (see Ener-
gies (p. 778))

The ei values can be used for adaptive mesh refinement. It has been shown by Babuska and Rhein-
boldt([104] (p. 930)) that if ei is equal for all elements, then the model using the given number of
elements is the most efficient one. This concept is also referred to as "error equilibration".

At the bottom of all printed nodal stresses (the PRNSOL or PRESOL command), which may consist
of the 6 component stresses, the 5 combined stresses, or both, a summary printout labeled: ESTIMATED
BOUNDS CONSIDERING THE EFFECT OF DISCRETIZATION ERROR gives minimum nodal values and
maximum nodal values. These are:
(17.82)
(17.83)

where min and max are over the selected nodes, and

where:

= nodal minimum of stress quantity (output as VALUE (printout) or SMNB (plot))

= nodal maximum of stress quantity (output as VALUE (printout) or SMXB (plot) )

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 907
Postprocessing

j = subscript to refer to either a particular stress component or a particular combined stress

= average of stress quantity j at node n of element attached to node n

= maximum of stress quantity j at node n of element attached to node n
Δσn = root mean square of all Δσi from elements connecting to node n
Δσi = maximum absolute value of any component of for all nodes connecting to
element (accessed with ETABLE (SDSG item) command)

17.6.2. Error Approximation Technique for Temperature-Based Problems

The error approximation technique used by POST1 (PRERR command) for temperature based problems
is similar to that given by Huang and Lewis([127] (p. 931)). The essentials of the method are summarized
below.

The usual continuity assumption results in a continuous temperature field from element to element,
but a discontinuous thermal flux field. To obtain more acceptable fluxes, averaging of the element
nodal thermal fluxes is done. Then, returning to the element level, the thermal fluxes at each node
of the element are processed to yield:
(17.84)

where:

= thermal flux error vector at node n of element i

= number of elements connecting to node n

= thermal flux vector of node n of element

Then, for each element

(17.85)

where:

ei = error for element i (accessed with ETABLE (TERR item) command)

vol = volume of the element (accessed with ETABLE (VOLU item) command)
[D] = conductivity matrix evaluated at reference temperature
{Δq} = thermal flux error vector at points as needed (evaluated from all {Δqn} of this element)

The error over the model is:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
908 Published: 2023-10-23
 POST1 - Error Approximation Technique

(17.86)

where:

e = error over the entire (or part of the) model (accessed with *GET (TERSM item) command)
Nr = number of elements in model or part of model

The error can be normalized against the thermal dissipation.

(17.87)

where:

E = percentage error in norm (accessed with PRERR, PLNSOL, (TEMP item) or *GET (TEPC
item) commands)
U = thermal dissipation over the entire (or part of the) model (accessed with *GET (TENSM
item) command)

= thermal dissipation of element i (accessed with ETABLE (TENE item) command) (see
Energies (p. 778))

The ei values can be used for adaptive mesh refinement. It has been shown by Babuska and Rhein-
boldt([104] (p. 930)) that if ei is equal for all elements, then the model using the given number of
elements is the most efficient one. This concept is also referred to as "error equilibration".

At the bottom of all printed fluxes (with the PRNSOL command), which consists of the 3 thermal
fluxes, a summary printout labeled: ESTIMATED BOUNDS CONSIDERING THE EFFECT OF DISCRETIZATION
ERROR gives minimum nodal values and maximum nodal values. These are:
(17.88)
(17.89)

where min and max are over the selected nodes, and

where:

= nodal minimum of thermal flux quantity (output as VALUE (printout) or SMNB (plot))

= nodal maximum of thermal flux quantity (output as VALUE (printout) or SMXB (plot))

j = subscript to refer to either a particular thermal flux component or a particular combined

thermal flux

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 909
Postprocessing

= average of thermal flux quantity j at node n of element attached to node n

= maximum of thermal flux quantity j at node n of element attached to node n

Δqn = maximum of all Δqi from elements connecting to node n

Δqi = maximum absolute value of any component of for all nodes connecting to
element (accessed with ETABLE (TDSG item) command)

17.6.3. Error Approximation Technique for Magnetics-Based Problems

The error approximation technique used by POST1 (PRERR command) for magnetics- based problems
is similar to that given by Zienkiewicz and Zhu ([103] (p. 930)) and Huang and Lewis ([127] (p. 931)).
The essentials of the method are summarized below.

The usual continuity assumption results in a continuous temperature field from element to element,
but a discontinuous magnetic flux field. To obtain more acceptable fluxes, averaging of the element
nodal magnetic fluxes is done. Then, returning to the element level, the magnetic fluxes at each node
of the element are processed to yield:
(17.90)

where:

= magnetic flux error vector at node n of element i

= number of elements connecting to node n

= magnetic flux vector of node n of element

Then, for each element

(17.91)

where:

ei = energy error for element i (accessed with ETABLE (BERR item) command)
vol = volume of the element (accessed with ETABLE (VOLU item) command)
[D] = magnetic conductivity matrix evaluated at reference temperature
{ΔB} = magnetic flux error vector at points as needed (evaluated from all {ΔBn} of this element)

The energy error over the model is:

(17.92)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
910 Published: 2023-10-23
 POST1 - Error Approximation Technique

e = energy error over the entire (or part of the) model (accessed with *GET (BERSM item)
command)
Nr = number of elements in model or part of model

The energy error can be normalized against the magnetic energy.

(17.93)

where:

E = percentage error in energy norm (accessed with PRERR, PLNSOL, (TEMP item) or *GET
(BEPC item) commands)
U = magnetic energy over the entire (or part of the) model (accessed with *GET (BENSM
item) command)

= magnetic energy of element i (accessed with ETABLE (SENE item) command) (see En-
ergies (p. 778))

The ei values can be used for adaptive mesh refinement. It has been shown by Babuska and Rhein-
boldt([104] (p. 930)) that if ei is equal for all elements, then the model using the given number of
elements is the most efficient one. This concept is also referred to as "error equilibration".

At the bottom of all printed fluxes (with the PRNSOL command), which consists of the 3 magnetic
fluxes, a summary printout labeled: ESTIMATED BOUNDS CONSIDERING THE EFFECT OF DISCRETIZATION
ERROR gives minimum nodal values and maximum nodal values. These are:
(17.94)
(17.95)

where min and max are over the selected nodes, and

where:

= nodal minimum of magnetic flux quantity (output as VALUE (printout))

= nodal maximum of magnetic flux quantity (output as VALUE (printout))

j = subscript to refer to either a particular magnetic flux component or a particular combined
magnetic flux

= average of magnetic flux quantity j at node n of element attached to node n

= maximum of magnetic flux quantity j at node n of element attached to node n

ΔBn = maximum of all ΔBi from elements connecting to node n

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 911
Postprocessing

ΔBi = maximum absolute value of any component of for all nodes connecting to
element (accessed with ETABLE (BDSG item) command)

17.7. POST1 - Harmonic Solid and Shell Element Postprocessing

As discussed in Axisymmetric Elements with Nonaxisymmetric Loads of the Element Reference, results
from load cases with different values of mode number (input as MODE on MODE command) but at the
same angular location (input as ANGLE on the SET command) can be combined in POST1 (with the
LCOPER command). The below assumes values of the mode number and angle and shows how the
results are extracted.

17.7.1. Thermal Solid Elements (PLANE75, PLANE78)

Data processed in a harmonic fashion includes nodal temperatures, element data stored on a per
node basis (thermal gradient and thermal flux) and nodal heat flow. Nodal temperature is calculated
at harmonic angle θ for each node j.
(17.96)

where:

Tjθ = temperature at node j at angle q

F = scaling factor (input as FACT, SET command)

n = mode number (input as MODE on MODE command)

θ = angle at which harmonic calculation is being made (input as ANGLE, SET command)
Tj = temperature at node j from nodal solution

Thermal gradient are calculated at harmonic angle θ for each node j of element i:
(17.97)
(17.98)
(17.99)

where:

= thermal gradient in x (radial) direction at node j of element i at angle θ

= thermal gradient in x (radial) direction at node j of element i

Nodal heat flow is processed in the same way as temperature. Thermal flux is processed in the same
way as thermal gradient.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
912 Published: 2023-10-23
 POST1 - Harmonic Solid and Shell Element Postprocessing

17.7.2. Structural Solid Elements (PLANE25, PLANE83)

Data processed in a harmonic fashion include nodal displacements, nodal forces, and element data
stored on a per node basis (stress and elastic strain).

Nodal displacement is calculated at harmonic angle θ for each node j:

(17.100)
(17.101)
(17.102)

where:

uxjθ = x (radial) displacement at node j at angle θ

uxj = maximum x (radial) displacement at node j (from nodal solution)

Stress is calculated at harmonic angle θ for each node j of element i:

(17.103)
(17.104)
(17.105)
(17.106)
(17.107)
(17.108)

where:

σxijθ = x (radial) stress at node j of element i at angle θ

σxij = maximum x (radial) stress at node j of element i

Nodal forces are processed in the same way as nodal displacements. Strains are processed in the
same way as stresses.

17.7.3. Structural Shell Element (SHELL61)

Data processed in a harmonic fashion include displacements, nodal forces, member forces, member
moments, in-plane element forces, out-of-plane element moments, stress, and elastic strain.

Nodal displacement is calculated at harmonic angle θ for each node j:

(17.109)
(17.110)
(17.111)
(17.112)

where:

φzjθ = rotation about z (hoop) direction at node j at angle θ

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 913
Postprocessing

φzj = maximum rotation about z (hoop) direction at node j (from nodal solution)

Stress is calculated at harmonic angle θ for each node/interior point j of element i:

(17.113)
(17.114)
(17.115)
(17.116)

where:

σmijθ = meridional stress at point j of element i at angle θ

σmij = meridional stress j of element i

In-plane element forces at harmonic angle θ for each node/interior point j of element i:
(17.117)
(17.118)
(17.119)

where:

= in-plane element force in x (meridional) direction at point j of element i at angle θ

Txij = maximum in-plane element force in x (meridional) direction at point j of element i

Nodal forces, member forces, and member moments are processed in the same way as nodal displace-
ments. Strains are processed in the same way as stresses. Finally, out-of-plane element moments are
processed in the same way as in-plane element forces.

17.8. POST26 - Data Operations

Table 17.1: POST26 Operations (p. 914) shows the operations that can be performed on the time-history
data stored by POST26. (Input quantities FACTA, FACTB, FACTC, and table IC are omitted from
Table 17.1: POST26 Operations (p. 914) for clarity of the fundamental operations.) All operations are
performed in complex variables. The operations create new tables which are also complex numbers.

Table 17.1: POST26 Operations

Description POST26 Real Complex Complex Result

Command Operation Operation
and Result
Addition ADD a+c (a + ib) + (c + id) (a + c) + i(b + d)
Multiplication PROD axc (a + ib) x (c + id) (ac - bd) + i(ad + bc)
Division QUOT a/c

Absolute ABS |a| |a + ib|

Value

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
914 Published: 2023-10-23
 POST26 - Response Spectrum Generator (RESP)

Description POST26 Real Complex Complex Result

Command Operation Operation
and Result
Arc Tangent ATAN 0 atan (a + ib) atan (b / a)
Square Root SQRT

Largest LARGE Maximum of a and c

Variable
Smallest SMALL Minimum of a and c
Variable
Derivative DERIV da/dc d(a + ib)/dc da/dc + i db/dc
Integration INT1

Common CLOG log10a log10(a + ib)

Logarithm
Natural NLOG na n (a + ib)
Logarithm
Exponential EXP ea e(a + ib) ea(cosb + i sinb)
Complex CONJUG a conj (a + ib) a - ib
Conjugate
Real Part REALVAR a real (a + ib) a
Imaginary IMAGIN 0 imag (a + ib) b
Part
Read Data DATA - - -
into Table
Fill Table FILL - - -
with Data
See Response Spectrum Generator Description (POST26 - Response Spectrum Generator
(RESP) (p. 915))
NOTES:

1. a + ib (from Table IA or IX) and c + id (from Table IB or IY) are complex numbers, where
IA and IB are input quantities on above commands.

2.

3. For derivative and integration, see Integration and Differentiation Procedures (p. 875)

17.9. POST26 - Response Spectrum Generator (RESP)

Given a motion as output from a transient dynamic analysis, POST26 generates a response spectrum
in terms of displacement, velocity, or acceleration.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 915
Postprocessing

A response spectrum is generated by imposing the motion of the point of interest on a series of single-
mass oscillators over a period of time and calculating the maximum displacement, velocity, or acceler-
ation. This is illustrated in Figure 17.11: Single Mass Oscillators (p. 916).

In Figure 17.11: Single Mass Oscillators (p. 916), the following definitions are used:

Mi = mass of oscillator i
Ci = damping of oscillator i
Ki = stiffness of oscillator i
ui = motion of oscillator i
ub = motion of point of interest

Figure 17.11: Single Mass Oscillators

In the absence of damping, the natural frequency of an oscillator i is:

(17.120)

The basic equation of motion of the oscillator can be given as a one degree of freedom (DOF) version
of Equation 15.5 (p. 796):
(17.121)

where:

a dot ( ) over a variable = derivative with respect to time

, the relative motion of oscillator i, is defined by:

(17.122)

The damping is given by:

(17.123)

where:

Ccr,i = = critical damping coefficient

Equation 17.120 (p. 916) through Equation 17.123 (p. 916) are combined to give:
(17.124)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
916 Published: 2023-10-23
 POST1 and POST26 - Interpretation of Equivalent Strains

Equation 17.124 (p. 916) is solved when the input time-history is the acceleration (inputType = 1).

For a displacement input ub (inputType = 0), the equation is rewritten as:

(17.125)

where:

is the velocity of the point of interest. It is the derivative of the displacement input.

Both equations are solved using Newmark integration scheme. See Description of Structural and Other
Second Order Systems (p. 796) for more details.

Depending on the spectrum type (specType), the output spectrum values are the following maximum
quantities:

• the relative displacement:

• the relative velocity:

• the absolute acceleration:

• the pseudo-velocity:

• the pseudo-acceleration:

17.9.1. Time Step Size

The time step size (Δt) is selected in the following way. If data is from a full transient analysis (AN-
TYPE,TRANS with TRNOPT,FULL):

Δt = input time step size (input as DTIME on RESP command)

or if no input is provided:
(17.126)

where:

fmax = highest value of frequency table (table input using LFTAB on the RESP command

The transient data from full transient analysis (ANTYPE,TRANS with TRNOPT, FULL analysis) is taken
from the next available time step used in the analysis. This can cause a decrease in accuracy at
higher frequencies if Δt is less than the time step size of the input transient.

17.10. POST1 and POST26 - Interpretation of Equivalent Strains

The equivalent strains for the elastic, plastic, creep, and thermal strains are calculated in postprocessing
using the von Mises equation:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 917
Postprocessing

(17.127)

where:

= appropriate component strain values

= effective Poisson's ratio

The default effective Poisson's ratio for both POST1 and POST26 are given in Combined Strains (p. 18).

The equivalent strain is output with the EQV or PRIN component label in POST1 (via PRNSOL, PLNSOL,
PDEF, or ETABLE) and in POST26 (via ESOL).

17.10.1. Physical Interpretation of Equivalent Strain

The von Mises equation is a measure of the shear strain in the material and does not account for the
hydrostatic straining component. For example, strain values of εx = εy = εz = 0.001 yield an equivalent
strain εeq = 0.0.

17.10.2. Elastic Strain

The equivalent elastic strain is related to the equivalent stress when ν' = ν (input as PRXY or NUXY
on MP command) by:
(17.128)

where:

σeq = equivalent stress (output using SEQV)

= equivalent elastic strain (output using EPEL, EQV)
E = Young's modulus

Note that when ν' = 0 then the equivalent elastic strain is related via
(17.129)

where:

G = shear modulus

17.10.3. Plastic Strain

For plasticity, the accumulated effective plastic strain is defined by (see Equation 4.25 (p. 74) and
Equation 4.42 (p. 77)):
(17.130)

where:

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
918 Published: 2023-10-23
 POST26 - Response Power Spectral Density

= accumulated effective plastic strain (output using NL, EPEQ)

This can be related to (output using EPPL, EQV) only under proportional loading situations during
the initial loading phase and only when ν' is set to 0.5.

17.10.4. Creep Strain

As with the plastic strains, to calculate the equivalent creep strain (EPCR, EQV), use ν' = 0.5.

17.10.5. Total Strain

The equivalent total strains in an analysis with plasticity, creep and thermal strain are:
(17.131)
(17.132)

where:

= equivalent total strain (output using EPTT, EQV)

= equivalent total mechanical strain (output using EPTO, EQV)
= equivalent thermal strain

For line elements, use an appropriate value of ν'. If >> , use ν' = 0.5. For other values, use an
effective Poisson's ratio between n and 0.5. One method of estimating this is through:

(17.133)

This computation of equivalent total strain is only valid for proportional loading, and is approximately
valid for monotonic loading.

17.11. POST26 - Response Power Spectral Density

The cross response PSD between two items is computed using the equation:

(17.134)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 919
Postprocessing

where:

p = reference number of first item (input as IA on RPSD command)

q = reference number of second item (input as IB on RPSD command)
p and q can be displacements, stresses, or reaction forces
denotes the real part of the argument

All other variables are defined in Response Power Spectral Densities and Mean Square Response (p. 851).
When p = q, the above cross response PSD becomes the auto response PSD.

17.12. POST26 - Computation of Covariance

The covariance between two items p and q is computed using the equation:

(17.135)

where:

p = reference number of first item (input as IA on CVAR command)

q= reference number of second item (input as IB on CVAR command)
p and q can be displacements, stresses, or reaction forces.

All other variables in Equation 17.135 (p. 920) are defined in Spectrum Analysis (p. 838). When p = q, the
above covariance becomes the variance.

17.13. POST1 and POST26 – Complex Results Postprocessing

The modal solution obtained using the complex eigensolvers (UNSYM, DAMP, QRDAMP) and the solution
from a harmonic analysis are complex. It can be written as
(17.136)

where:

R = the complex degree of freedom solution (a nodal displacement Ux, a reaction force Fy, etc.).
RR = the real part of the solution R.
RI = the imaginary part of the solution R.

The same complex solution may also be expressed as:

(17.137)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
920 Published: 2023-10-23
 POST1 and POST26 – Complex Results Postprocessing

where:

Rmax = the degree of freedom amplitude.

φ = the degree of freedom phase shift.

The phase shift of the solution is different at each degree of freedom so that the total amplitude at a
node is not the square root of the sum of squares of the degrees of freedom amplitudes (Rmax). More
generally, total amplitudes (SUM), phases and other derived results (principal strains/stresses, equivalent
strain/stress,… for example) at one node do not vary harmonically as degree of freedom solutions do.

The relationship between RR, RI, Rmax and φ is defined as follows:

(17.138)

(17.139)

RR = Rmaxcosφ

RI = Rmaxsinφ

In POST1, use KIMG in the SET command to specify which results are to be stored: the real parts, the
imaginary parts, the amplitudes or the phases.

In POST26, use PRCPLX and PLCPLX to define the output form of the complex variables.

The complete complex solution is harmonic. It is defined as:

(17.140)

where:

Ω = the excitation frequency in a harmonic analysis, or the natural damped frequency in a

complex modal analysis.

In the equations of motion for harmonic and complex modal analyses, the complex notations are used
for ease of use but the time dependant solution at one degree of freedom is real.

The solution of the harmonic analysis is defined as:

(17.141)

The ANHARM command issued after the harmonic analysis is based on Equation 17.141 (p. 921). The
HRCPLXand LCOPER (with Oper2 = CPXMAX) commands are based on Equation 17.141 (p. 921) for
both modal and harmonic analyses.

The complete solution of the complex modal analysis is defined as:

(17.142)

Where σ is the real part of the complex frequency. See Complex Eigensolutions (p. 756) for more details
about complex eigensolutions.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 921
Postprocessing

The ANHARM command issued after a complex modal analysis is based on Equation 17.142 (p. 921)
except if NPERIOD is set to -1. In this case, the exponential term (decay or growth of the oscillation) is
ignored and the equation reduces to Equation 17.141 (p. 921).

In the LCOPER command with Oper2 = CPXMAX, a loop on the phase (Ωt), is performed to calculate
the maximum (that may occur at any phase at each location) of the degree of freedom results, displace-
ments and stresses, the principal stresses (S1, S2, and S3), the stress intensity (SINT), the equivalent
stress (SEQV), the equivalent strain, and the principal strains.

17.14. POST1 - Modal Assurance Criterion (MAC)

The modal assurance criterion (MAC) can compare two real solutions or two complex solutions.

The MAC between two real solutions is computed using the equation:

(17.143)

where:

= the ith solution vector of solution 1. (Solution 1 is read from File1, and index i corres-
ponds to Sbstep1 on the RSTMAC command).

= the jth solution vector of solution 2. For the case of the .unv file, it can be a displacement
or an acceleration vector of solution 2. (Solution 2 is read from File2, and index j corresponds
to Sbstep2 on the RSTMAC command).
m (k) = diagonal of the mass matrix used in obtaining solution k. It is included in the calculation
only if KeyMass is ON (default). k = 1 if nodes are matched ( Option = ABSTOLN and/or
RELTOLN > 0 on the MACOPT command). k = 2 if nodes are mapped and solution 1 is interpol-
ated ( Option = NODMAP on the MACOPT command).

The MAC between two complex solutions is computed using the equation:

(17.144)

where:

= the complex conjugate of a complex vector .

If the diagonal of the mass matrix is not available (for example on a Universal Format file or for non-
structural degrees of freedom), the modal assurance criterion is not weighted with the mass; that is,
the mass is assumed to be equal at all degrees of freedom.

The dot product of the solution vectors is calculated:

• at matched nodes if Option = ABSTOLN or RELTOLN > 0, and at mapped nodes if Option =
NODMAP is activated.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
922 Published: 2023-10-23
 POST1 – Frequency response function correlation

• at matched and selected (MACOPT,DOF) DOFs.

17.15. POST1 – Frequency response function correlation

A frequency response function (FRF) is a function used to quantify the response of a system to an excit-
ation, normalized by the magnitude of this excitation, in the frequency domain. FRF correlation analysis
between pairs of FRFs at each frequency point is evaluated in terms of cross-signature assurance criterion
(CSAC) and cross-signature scale factor (CSF).

Cross Signature Assurance Criterion (CSAC)

The cross signature assurance criterion is a measure of the shape correlation between experimental
and FE generated responses with value ranging from 0 to 1, as described in Equation 17.145 (p. 923).
The CSAC is an adaptation of MAC for the frequency domain. This criterion is sensitive to changes
in mass and stiffness.

(17.145)

where:

is the FE generated FRF, is its conjugate transpose,

is the measured FRF, is its conjugate transpose,
is the number of frequency points.

Cross Signature Scale Factor (CSF)

The cross signature scale factor is a measure of the difference in amplitude between measured
and FE generated responses with value ranging from 0 to 1 as well, as shown in Equa-
tion 17.146 (p. 923). The CSF is sensitive to changes in damping and is often used with the CSAC.

(17.146)

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 923
 Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
924 Published: 2023-10-23
Bibliography
[1] Ahmad, S., Irons, B. M. & Zienkiewicz, O. C. (1970). Analysis of thick and thin shell structures by curved
finite elements. International Journal for Numerical Methods in Engineering. 2(3), 419-451.

[2] Bathe, K. J. (1996). Finite Element Procedures. Upper Saddle River, NJ: Prentice-Hall.

[3] Biot, M. A. (1965). Mechanics of Incremental Deformation. Hoboken, NJ: Wiley.

[4] Chen, L. H. (1959). Piping flexibility analysis by stiffness matrix. ASME, Journal of Applied Mechanics.
New York, NY: ASME Technical.

[5] Cook, R. D. (1981). Concepts and Applications of Finite Element Analysis (2nd ed.). Hoboken, NJ: Wiley.

[6] Cook, R. D. (1972). Two hybrid elements for analysis of thick, thin and sandwich plates. International
Journal for Numerical Methods in Engineering. 5(2), 277-288.

[7] Cuniff, D. F. & O'Hara, G. J. (1965). Normal mode theory for three-directional motion. NRL Report 6170.
Washington, DC: U. S. Naval Research Laboratory.

[8] Denn, M. M. (1969). Optimization by Variational Methods. New York, NY: McGraw-Hill.

[9] Henshell, K. D. & Ong, J. H. (1975). Automatic masters for eigenvalue economization. Earthquake En-
gineering and Structural Dynamics. 3, 375-383.

[10] Imgrund, M. C. (1992). ANSYS® Verification Manual. Pittsburgh, PA: Swanson Analysis Systems.

[11] Flugge, W. (1967). Stresses in Shells. New York, NY: Springer.

[12] Fritz, R. J. (1972). The effect of liquids on the dynamic motions of immersed solids. ASME Journal
of Engineering for Industry. 94(1), 167-173.

[13] Galambos, T. V. (1968). Structural Members and Frames. Englewood Cliffs, NJ: Prentice-Hall.

[14] Guyan, R. J. (1965). Reduction of stiffness and mass matrices. AIAA Journal. 3(2), 380.

[15] Hall, A. S. & Woodhead, R. W. (1961). Frame Analysis. Hoboken, NJ: Wiley.

[16] Rajakumar, C. & Rogers, C. R. (1992). The Lanczos algorithm applied to unsymmetric generalized
eigenvalue problem. International Journal for Numerical Methods in Engineering. 32, 1009-1026.

[17] Irons, B. M. (1970) . A frontal solution program for finite element analysis. International Journal for
Numerical Methods in Engineering. 2(1), 5-23.

[18] Irons, B. M. (1970). A frontal solution program for finite element analysis. International Journal for
Numerical Methods in Engineering. 2(1), 149.

[19] Wilkinson, J. H. (1988). The Algebraic Eigenvalue Problem. (pp. 515-569). Oxford, UK: Clarendon.

[20] Kohnke, P. C. and Swanson, J. A. (1976). Thermo-electric finite elements. Proceedings from the Inter-
national Conference on Numerical Methods in Electrical and Magnetic Field Problems. Santa Mar-
gherita Liqure, Italy.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 925
Bibliography

[21] Kohnke, P. C. (1978). Large deflection analysis of frame structures by fictitious forces. International
Journal of Numerical Methods in Engineering. 12(8), 1278-1294.

[22] Kollbrunner, C. F. and Balser, K. (1969). Torsion in Structures. New York, NY: Springer.

[23] Konopinski, E. J. (1969). Classical Descriptions of Motion. San Francisco, CA: Freeman and Company.

[24] Kreyszig, E. (1962). Advanced Engineering Mathematics. Hoboken, NJ: Wiley.

[25] Lekhnitskii, S. G. (1963). Theory of Elasticity of an Anisotropic Elastic Body. San Francisco, CA: Holden-
Day.

[26] Melosh, R. J. & Bamford, R. M. (1969). Efficient solution of load-deflection equations. ASCE Journal
of the Structural Division. 95, 661-676.

[27] Kanok-Nukulchai, W. (1979). A simple and efficient finite element for general shell analysis. Interna-
tional Journal for Numerical Methods in Engineering. 14, 179-200.

[28] Oden, J. T. (1968). Mechanics of Elastic Structures. New York, NY: McGraw-Hill.

[29] Przemieniecki, J. S. (1968). Theory of Matrix Structural Analysis. New York, NY: McGraw-Hill.

[30] Schnobrich, W. C. & Suidan, M. (1973). Finite element analysis of reinforced concrete. ASCE Journal
of the Structural Division. 99(1), 2109-2122.

[31] Seide, P. (1977). Large deflection of rectangular membranes under uniform pressure. International
Journal of Non-Linear Mechanics. 12(6), 397-406.

[32] Skjelbreia, L. & Hendrickson, J. A. (1961). Fifth order gravity wave theory. (10, 184-196). Proceedings
from the Seventh Conference on Coastal Engineering. The Hague, The Netherlands.

[33] Timoshenko, S. & Woinowskey-Krieger, S. (1959). Theory of Plates and Shells. New York, NY: McGraw-
Hill.

[34] Tracey, D. M. (1974). Finite elements for three dimensional elastic crack analysis. Nuclear Engineering
and Design. 26(2), 282-920.

[35] Vanmarcke, E. H. (1976). Structural response to earthquakes. C. Lomnitz & E. Rosemblueth (Eds.).
Seismic Risk and Engineering Decisions. (287-337). Oxford, UK: Elsvier Scientific.

[36] Wheeler, J. D. (1970). Method of calculating forces produced by irregular waves. Journal of Petroleum
Technology. 22, 359-367.

[37] Willam, K. J. (1982). University of Colorado, Boulder: Private Communication.

[38] Willam, K. J. & Warnke, E. D. (1975). Constitutive model for the triaxial behavior of concrete. Proceed-
ings from the International Association for Bridge and Structural Engineering. Bergamo, Italy.

[39] Wilson, E. L., Taylor, R. L., Doherty, W. P., & Ghaboussi, J. (1973). Incompatible displacement models.
S. J. Fenves, N. Perrone, A. R. Robinson, & W. C. Schnobrich (Eds.). Numerical and Computer
Methods in Structural Mechanics. (43-57). New York, NY: Academic.

[40] Zienkiewicz, O. C. (1977). The Finite Element Method. New York, NY: McGraw-Hill.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
926 Published: 2023-10-23
[41] American Society of Mechanical Engineers. (1974). ASME boiler and pressure vessel code, section
III, division 1, subsection NC, class 2 components. New York, NY: American Society of Mechanical
Engineers, Boiler and Pressure Vessel Committee.

[42] U. S. Nuclear Regulatory Commission. (1976). Regulatory guide 1.92, revision 1. Washington, DC: U.
S. Nuclear Regulatory Commission, Office of Standards Development.

[43] Tamma, K. K. & Namburu, R. R. (1990). Recent advances, trends and new perspectives via enthalpy-
based finite element formulations for applications to solidification problems. International
Journal for Numerical Methods in Engineering. 30, 803-820.

[44] U. S. Army Coastal Engineering Research Center. (1977). Shore Protection Manual. (3rd ed., Vol. 1).
Fort Belvior, VA.

[45] Beer, F. P. & Johnston, R. E. (1962). Vector Mechanics for Engineers, Statics and Dynamics. New York,
NY: McGraw-Hill.

[46] Hinton, E., Rock, A. & Zienkiewicz, O. C. (1976). A note on mass lumping and related processes in
the finite element method. International Journal of Earthquake Engineering and Structural Dynamics.
4, 245-249.

[47] Krieg, R. D. & Krieg, D. B. (1977). Accuracies of numerical solution methods for the elastic-perfectly
plastic model. Journal of Pressure Vessel Technology. 99(4), 510-515.

[48] Thomson, W. T. (1971). Theory of Vibrations with Applications. Upper Saddle River, NJ: Prentice Hall.

[49] Roark, R. J. & Young, W. C. (1975). Formulas for Stress and Strain. New York, NY: McGraw-Hill.

[50] Taylor, R. L., Beresford, P. J., & Wilson, E. L. (1976). A non-conforming element for stress analysis.
International Journal for Numerical Methods in Engineering. 10, 1211-1219.

[51] Hill, R. (1983). The Mathematical Theory of Plasticity. Oxford, UK: Oxford University Press.

[52] Shih, C. F. & Lee, D. (1978). Further developments in anisotropic plasticity. Journal of Engineering
Materials and Technology. 100, 294-302.

[53] Valliappan, S. (1976). Nonlinear analysis for anistropic materials. International Journal for Numerical
Methods in Engineering. 10, 597-606.

[54] Besseling, J. F. (1958). A theory of elastic, plastic, and creep deformations of an initially isotropic
material showing anisotropic strain-hardening creep recovery and secondary creep. Journal of
Applied Mechanics. 529-536.

[55] Owen, R. J., Prakash, A., & Zienkiewicz, O. C. (1974). Finite element analysis of non-linear composite
materials by use of overlay sytems. Computers and Structures. 4(6), 1251-1267.

[56] Holman, J. P. (1976). Heat Transfer (4th ed.). New York, NY: McGraw-Hill.

[57] Batoz, J. L., Bathe, K. J., & Ho, L. W. (1980). A study of three-node triangular plate bending elements.
International Journal of Numerical Methods in Engineering. 15, 1771-1812.

[58] Razzaque, A. (1984). On the four noded discrete kirchhoff shell elements. J. Robinson (Ed.). Accuracy
Reliability Training in FEM Technology. 473-483.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 927
Bibliography

[59] Gresho, P. M. & Lee, R. L. (1979). Don't suppress the wiggles - they're telling you something. Finite
Element Methods for Convection Dominated Flows. 34, 37-61.

[60] Dean, R. G. (1974). Evaluation and development of water wave theories for engineering application:
Special report no. 1. (Vol. 2). Fort Belvoir, VA: Coastal Engineering Research Center.

[61] American Society of Mechanical Engineers. (1974). ASME boiler and pressure vessel code, section
III, division 1, subsection NB, class 1 components. New York, NY: American Society of Mechanical
Engineers, Boiler and Pressure Vessel Committee.

[62] Power Piping. American National Standard Code for Pressure Piping. American Society of Mechanical
Engineers. ANSI B31.1-1977.

[63] Orris, R. M. & Petyt, M. (1974). Finite element study of harmonic wave propagation in periodic
structures. Journal of Sound and Vibration. 33(2), 223-236.

[64] Gordon, J. L. (1976). Outcur: An automated evaluation of two-dimensional finite element stresses
according to ASME Section III stress requirements. Paper No. 76-WA/PVP-16. ASME Winter Annual
Meeting.

[65] Powell, M. J. D. (1964). An efficient method for finding the minimum of a function of several variables
without calculating derivatives. Computer Journal. 7, 155-162.

[66] Wilson, E. L., Der Kiereghian, A., & Bayo, E. (1981). A replacement for the SRSS method in seismic
analysis. Earthquake and Structural Dynamics. 9(2), 187.

[67] Rankin, C. C. & Brogan, F. A. (1986). An element independent corotational procedure for the treatment
of large rotations. Journal of Pressure Vessel Technology. 108, 165-174.

[68] Argyris, J. (1982). An excursion into large rotations. Computer Methods in Applied Mechanics and
Engineering. 32, 85-155.

[69] Tse, S., Morse, I. E., & Hinkle, R. T. (1963). Mechanical Vibrations. Boston, MA: Allyn and Bacon.

[70] Chari, M. V. K. (1974). Finite element solution of the eddy current problem in magnetic structures.
IEEE Transactions on Power Apparatus and Systems. PAS-93, 62-72 .

[71] Brauer, J. R. (1977). Paper A77-122-5, Finite element analysis of electromagnetic induction in
transformers. IEEE Winter Power Meeting.

[72] Tandon, S. C. & Chari, M. V. K. (1981). Transient solution of the diffusion equation by the finite element
method. Journal of Applied Physics. 52(3), 2431-2432.

[73] Silvester, P. P., Cabayan, H. S., & Browne, B. T. (1973). Efficient techniques for finite element analysis
of electric machines. IEEE Transactions on Power Apparatus and Systems. PAS-92, 1274-1281.

[74] Chari, M. V. K. & D'Angelo, J. (1981). Paper No. V1-1, Finite element analysis of magneto-mechanical
devices. Fifth International Workshop in Rare Earth-Cobalt Permanent Magnets and Their Application.
7-10.

[75] Anderson, O. W. (1973). Transform leakage flux program based on the finite element method. IEEE
Transactions on Power Apparatus and Systems. 92(2).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
928 Published: 2023-10-23
[76] Zienkiewicz, O. C., Lyness, J., & Owen, D. R. (1977). Three-dimensional magnetic field determination
using a scalar potential - a finite element solution. IEEE Transactions on Magnetics. 13(5), 1649-
1656.

[77] Coulomb, J. L. & Meunier, G. (1984). Finite element implementation of virtual work principle for
magnetic for electric force and torque computation. IEEE Transactions on Magnetics. 20(5), 1894-
1896.

[78] Moon, F. C. (1984). Magneto-Solid Mechanics. Hoboken, NJ: Wiley.

[79] Baker, A. J. (1983). Finite Element Computational Fluid Mechanics. (266-284). New York, NY: McGraw-
Hill.

[80] Yuan, S. W. (1976). Foundations of Fluid Mechanics. (71-102). Upper Saddle River, NJ: Prentice-Hall.

[81] Clough, R. W. & Penzien, J. (1975). Dynamics of Structures. (559). New York, NY: McGraw-Hill.

[82] Allik, H. & Hughes, J. R. (1970). Finite element for piezoelectric vibration. International Journal Nu-
merical Methods of Engineering. 2, 151-157.

[83] Eer Nisse, N. P. (1967). Variational method for electroelastic vibration analysis. IEEE Transactions on
Sonics and Ultrasonics. 14(4).

[84] Sato, J., Kawabuchi, M. & Fukumoto, A. (1979). Dependence of the electromechanical coupling
coefficient on the width-to-thickness ratio of plant-shaped piezoelectric transducers used for
electronically scanned ultrasound diagnostic systems. Journal of Acoustics Society of America. 66,
1609-1611.

[85] Kinsler, L. E., Frey, A. R., Coppens, A. B., & Sanders, J. V. (1982). Fundamentals of Acoustics. (98-123).
Hoboken, NJ: Wiley.

[86] Craggs, A. (1986). A finite element model for acoustically lined small rooms. Journal of Sound and
Vibration. 108(2), 327-337.

[87] Zienkiewicz, O. C. & Newton, R. E. (1969). Coupled vibrations of a structure submerged in a com-
pressible fluid. Proceedings of the Symposium on Finite Element Techniques. University of Stuttgart,
Germany.

[88] Malvern, Lawrence E. (1969). Introduction to the Mechanics of a Continuous Medium. Upper Saddle
River, NJ: Prentice-Hall.

[89] Siegal, R. & Howell, J. R. (1981). Thermal Radiation Heat Transfer (2nd ed.). New York, NY: Hemisphere
Publishing Corporation.

[90] Institute of Electrical and Electronic Engineers. (1988). ANSI/IEEE std no. 176 on piezoelectricity. New
York, NY: IEEE.

[91] Antonova, E. E. & Looman, D. C. (2005). Finite elements for thermoelectric device analysis in ANSYS.
24th International Conference on Thermoelectrics. Clemson, SC.

[92] Onate, E., Rojek, J., Taylor, R. L., & Zienkiewicz, O. C. (2004). Finite calculus formulation for incom-
pressible solids using linear triangles and tetrahedra. International Journal for Numerical Methods
in Engineering. 59, 1473-1500.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 929
Bibliography

[93] Fossum, A. F. & Fredrich, J. T. (2000). Cap plasticity model and compactive and dilatant prefailure
deformation. Pacific Rocks 200: Rock Around the Rim. 1169-1176.

[94] Tsai, S. W. (1987). Composites Design (3rd ed.). Section 11.6. Dayton, OH: Think Composites.

[95] Weiss, J. (1982). Efficient finite element solution of multipath eddy current problems. IEEE Transactions
on Magnetics. 18(6), 1710-1712.

[96] Garg, V. K. & Weiss, J. (1986). Finite element solution of transient eddy-current problems in multiply-
excited magnetic systems. IEEE Transactions on Magnetics. 22(5), 1257-1259.

[97] Dvorkin, E. N. (1984). On Nonlinear Finite Element Analysis of Shell Structures. Published Ph.D thesis.
Boston, MA: Massachusetts Institute of Technology.

[98] Dvorkin, E. N. & Bathe, K. J. (1984). A continuum mechanics based four-node shell element for
general nonlinear analysis. Engineering Computations. 1, 77-88.

[99] Bathe, K. J. & Dvorkin, E. N. (1986). A formulation of general shell elements - the use of mixed inter-
polation of tensorial components. International Journal for Numerical Methods in Engineering. 22,
697-722.

[100] Hoit, M. & Wilson, E. L. (1983). An equation numbering algorithm based on a minimum front cri-
teria. Computers and Structures. 16, 225-239.

[101] Cuthill, E. & McKee, J. (1969). Reducing the bandwidth of sparse symmetric matrices. Proceedings
of the ACM National Conference. New York.

[102] Georges, A. & McIntyre, D. (1978). On the application of the minimum degree algorithm to finite
element systems. SIAM Journal of Numerical Analysis. 15.

[103] Zinkiewicz, O. C. & Zhu, J. Z. (1987). A simple error estimator and adaptive procedure for practical
engineering analysis. International Journal for Numerical Methods in Engineering. 24, 337-357.

[104] Babuska, I. & Rheinboldt, W. C. (1979). Analysis of optimal finite element meshes in R. Mathematics
of Computation. 33, 431-463.

[105] Carnegie, W. (1959). Vibrations of rotating cantilever blading. Journal of Mechanical Engineering
Science. 1(3), 235-240.

[106] Bergan, P. G. & Mollestad, E. (1985). An automatic time-stepping algorithm for dynamic problems.
Computer Methods in Applied Mechanics and Engineering. 49.

[107] Paris, P. C. & Sih, G. C. (1965). Stress analysis of cracks. Fracture Toughness and Testing and its Ap-
plications. 30-83.

[108] O'Hara, G. J. & Belsheim, R. O. (1963). Interim design values for shock design of shipboard equipment.
NRL Memorandum Report 1396. Washington, DC: U.S. Naval Research Laboratory.

[109] Markovsky, A., Soules, T. F., & Vukcevich, M. R. (1986). Mathematical and computational aspects
of a general viscoelastic theory (Report no. 86-LRL-2021). G. E. Lighting and Research and Technical
Services Operation.

[110] Scherer, G. W. & Rekhson, S. M. (1982). Viscoelastic-elastic composites: I, General theory. Journal
of the American Ceramic Society. 65(7).

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
930 Published: 2023-10-23
[111] Narayanaswamy, O. S. (1971). A model of structural relaxation in glass. Journal of the American
Ceramic Society. 54(10), 491-498.

[112] Zienkiewicz, O. C., Watson, M., & King, I. P. (1968). A numerical method of visco-elastic stress ana-
lysis. International Journal of Mechanical Science. 10, 807-827.

[113] Taylor, R. L., Pister, K. S., & Goudreas, G. L. (1970). Thermochemical analysis of viscoelastic solids.
International Journal for Numerical Methods in Engineering. 2, 45-59.

[114] Allman, D. J. (1984). A compatible triangular element including vertex rotations for plane elasticity
analysis. Computers and Structures. 19, 1-8.

[115] Cook, R. D. (1986). On the allman triangle and a related quadrilateral element. Computers and
Structures. 22, 1065-1067.

[116] MacNeal, R. H. & Harder, R. L. (1988). A refined for-foded membrane element with rotational degrees
of freedom. Computers and Structures. 28(1), 75-84.

[117] Garvey, S. J. (1951). The quadrilateral shear panel. Aircraft Engineering. 134.

[118] Yunus, S. M., Pawlak, T. P., & Cook, R. D. (1991). Solid elements with rotational degrees of freedom
part 1 and part 2. International Journal for Numerical Methods in Engineering. 31, 573-610.

[119] Mohammed, O. A. (1983). Magnetic Vector Potential Based Formulation and Computation of
Nonlinear Magnetostatic Fields and Forces in Electrical Devices by Finite Elements. Ph.D. Disser-
tation. Blacksburg, VA: Virginia Polytechnic Institute and State University.

[120] Mayergoyz, I. D. (1987). A new scalar potential formulation for three-dimensional magnetostatic
problems. IEEE Transactions on Magnetics. 23(6), 3889-3894.

[121] Biro, O. & Preis, K. (1989). On the use of the magnetic vector potential in the finite element ana-
lysis of three-dimensional eddy currents. IEEE Transactions on Magnetics. 25(4), 3145-3159.

[122] Robinson, J. (1985). Basic and Shape Sensivity Tests for Membrane and Plate Bending Finite Elements.
Bridestowe, UK: Robinson and Associates.

[123] Kagawa, Y., Yamabuchi, T. & Kitagami, S. (1986). Infinite boundary element and its application to
a combined finite-boundary element technique for unbounded field problems. C. A. Brebbia
(Ed.). Boundary Elements VIII. New York, NY: Springer-Verlag.

[124] Oden, J. T. & Kikuchi, N. (1982). Finite element methods for constrained problems in elasticity. In-
ternational Journal for Numerical Methods in Engineering. 18(5), 701-725.

[125] Sussman, T. & Bathe, K. J. (1987). A finite element formulation for nonlinear incompressible elastic
and inelastic analysis. Computers and Structures. 26( ½), 357-409.

[126] Zienkiewicz, O. C., Liu, Y. C., & Huang, G. C. (1989). Error estimates and convergence rates for
various incompressible elements. International Journal for Numerical Methods in Engineering.
28(9), 2191-2202.

[127] Huang, H. C. & Lewis, R. W. (1989). Adaptive analysis for heat flow problems using error estimation
techniques. Presented at the 6th International Conference on Numerical Methods in Thermal
Problems. University of Wales. Cardiff, Wales, UK.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 931
Bibliography

[128] Weber, G. G., Lush, A. M., Zavaliangos, A., & Anand, L. (1990). An objective time-integration procedure
for isotropic rate-independent elastic-plastic constitutive equations. International Journal of
Plasticity. 6, 701-749.

[129] Eggert, G. M. & Dawson, P. R. (1988). A viscoplastic formulation with plasticity for transient metal
forming. Computer Methods in Applied Mechanics and Engineering. 70, 165-190.

[130] Narayanaswami, R. & Adelman, H. M. (1974). Inclusion of transverse shear deformation in finite
element displacement formulations. American Institute of Aeronautics and Astronautics Journal.
12(11), 1613-1614.

[131] Kaljevic, I., Saigal, S., & Ali, A. (1992). An infinite boundary element formulation for three-dimen-
sional potential problems. International Journal for Numerical Methods in Engineering. 35(10),
2079-2100.

[132] Simo, J. C. (1986). Finite deformation post-buckling analysis involving inelasticity and contact
constraints. International Journal for Numerical Methods in Engineering. 23, 779-800.

[133] Parisch, H. (1989). A consistent tangent stiffness matrix for three-dimensional non-linear contact
analysis. International Journal for Numerical Methods in Engineering. 28, 1803-1812.

[134] Bayo, E. (1988). A modified Lagrangian formulation for the dynamic analysis of constrained
mechanical systems. Computer Methods in Applied Mechanics and Engineering. 71, 183-195.

[135] Jiang, L. & Rogers, R. J. (1988). Combined Lagrangian multiplier and penalty function finite element
technique for elastic impact analysis. Computers and Structures. 30, 1219-1229.

[136] Giannakopoulos, A. E. (1989). The return mapping method for the integration of friction constitutive
relations. Computers and Structures. 32, 157-167.

[137] Owen, D. R. J. & Rodic, T. (1989). A plasticity theory of friction and joint elements. Proceedings of
the 2nd International Computational Plasticity: Models, Software, and Applications Conference.
Barcelona, Spain.

[138] Wriggers, P., Vu Van, T., & Stein, E. (1990). Finite element formulation of large deformation impact-
contact problems with friction. Computers and Structures. 37, 319-331.

[139] Wriggers, P., Vu Van, T., & Stein, E. (1989). Models of friction, finite-element implementation and
application to large deformation impact-contact problems. Proceedings of the 2nd International
Computational Plasticity: Models, Software, and Applications Conference. Barcelona, Spain.

[140] Yunus, S. M., Kohnke, P. C., & Saigal, S. (1989). An efficient through-thickness integration scheme
in an unlimited layer doubly curved isoparametric composite shell element. International
Journal for Numerical Methods in Engineering. 28, 2777-2793.

[141] Geddes, E. R. (1982). An Analysis of the Low Frequency Sound Field in Non-Rectangular Enclosures
Using the Finite Element Method. Ph.D Thesis. Pennsylvania State University: Centre County, PA.

[142] Gyimesi, M., Lavers, D., Pawlak, T., Ostergaard, D. (1993). Application of the general potential for-
mulation in the ANSYS® program. IEEE Transactions on Magnetics. 29, 1345-1347.

[143] Rajakumar, C. & Ali, A. (1992). A solution method for acoustic boundary element eigenproblem
with sound absorption using Lanczos algorithm. Proceedings of 2nd International Congress on

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
932 Published: 2023-10-23
 Recent Developments in Air-Borne and Structure-Borne Sound and Vibration. Auburn University,
AL.

[144] Nishimura, H., Isobe, M., & Horikawa, K. (1977). Higher order solutions of the stokes and the
cnoidal waves. Journal of the Faculty of Engineering. 34(2), 268.

[145] Mahinthakumar, G. & Hoole, S. R. H. (1990). A parallelized element by element jacobi conjugate
gradients algorithm for field problems and a comparison with other schemes. Applied Electro-
magnetics in Materials. 1, 15-28.

[146] Hughes, T. J. R. (1983). Analysis of transient algorithms with particular reference to stability beha-
vior. T. Belytschko and K. J. Bathe (Eds.). Computation Methods for Transient Analysis. 1, 67-155.

[147] Anand, L. (1982). Constitutive equations for the rate-dependent deformation of metals at elevated
temperatures. Journal of Engineering Materials and Technology. 104, 12-17.

[148] Brown, S. B., Kim, K. H., & Anand, L. (1989). An internal variable constitutive model for hot working
of metals. International Journal of Plasticity. 5, 95-130.

[149] Dickens, J. M. (1980). Numerical Methods for Dynamic Substructure Analysis. Ph.D. Thesis. University
of California, Berkeley.

[150] Gyimesi, M. & Lavers, J. D. (1992). Generalized potential formulation for 3D magnetostatic problems.
IEEE Transactions on Magnetics. 28(4).

[151] Smythe, W. R. (1950). Static and Dynamic Electricity. New York, NY: McGraw-Hill.

[152] Demerdash, N. A., Nehl, T. W., Fouad, F. A., & Mohammed, O. A. (1981). Three dimensional finite
element vector potential formulation of magnetic fields in electrical apparatus. IEEE Transactions
on Power Apparatus and Systems. 100(8), 4104-4111.

[153] Eggert, G. M., Dawson, P. R., & Mathur, K. K. (1991). An adaptive descent method for nonlinear
viscoplasticity. International Journal for Numerical Methods in Engineering. 31, 1031-1054.

[154] Schweizerhof, K. H. & Wriggers, P. (1986). Consistent linearization for path following methods in
nonlinear FE analysis. Computer Methods in Applied Mechanics and Engineering. 59, 261-279.

[155] Zienkiewicz, O. C. & Cormeau, I. C. (1974). Visco-plasticity - Plasticity and creep in elastic solids -
A unified numerical solution approach. International Journal for Numerical Methods in Engineering.
8, 821-845.

[156] Simo, J. C. & Taylor, R. L. (1985). Consistent tangent operators for rate-independent elastoplasticity.
Computer Methods in Applied Mechanics and Engineering. 48, 101-118.

[157] Hughes, T. J. R. (1984). Numerical implementation of constitutive models: Rate-independent devi-

atoric plasticity. Theoretical Foundation for Large-Scale Computations for Nonlinear Material Beha-
vior.

[158] Hughes, T. J. R. & Carnoy, E. (1983). Nonlinear finite element shell formulation accounting for large
membrane strains. Computer Methods in Applied Mechanics and Engineering. 39, 69-82.

[159] Nedelec, J. (1980). Mixed finite elements in R3. Numerische Mathematik. 35, 315-341.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 933
Bibliography

[160] Anand, L. (1985). Constitutive equations for hot-working of metals. International Journal of Plasticity.
1, 213-231.

[161] Abramowitz, M. & Stegun, I. A. (1972). Handbook of mathematical functions with formulas, graphs,
and mathematical tables. (55, 966). Washing, DC: National Bureau of Standards.

[162] Swain, C. G. & Swain, M. S. (1980). A uniform random number generator that is reproducible,
hardware-independent, and fast. Journal of Chemical Information and Computer Sciences. 56-58.

[163] Kreyszig, E. (1972). Advanced Engineering Mathematics (3rd ed.). Hoboken, NJ: Wiley.

[164] Hoel, P. G. (1962). Introduction to Mathematical Statistics (3rd ed.). (196). Hoboken, NJ: Wiley.

[165] Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (1978). Applied Statistics. Boston, MA: Allyn and
Bacon.

[166] Hughes, T. J. R. (1987). The Finite Element Method Linear Static and Dynamic Finite Element Analysis.
Upper Saddle River, NJ: Prentice-Hall.

[167] Wilson, E. L. & Itoh, T. (1983). An eigensolution strategy for large systems. Computers and Structures.
16(1-4), 259-265.

[168] Yokoyama, T. (1990). Vibrations of a hanging Timoshenko beam under gravity. Journal of Sound
and Vibration. 141(2), 245-258.

[169] Coulomb, J. L. (1983). A methodology for the determination of global electromechanical quantities
from a finite element analysis and its application to the evaluation of magnetic forces, torques
and stiffness. IEEE Transactions on Magnetics. 19(6), 2514-2519.

[170] Zienkiewicz, O. C., Emson, C., & Bettess, P. (1983). A novel boundary infinite element. International
Journal for Numerical Methods in Engineering. 19, 393-404.

[171] Damjanic, F. & Owen, D. R. J. (1984). Mapped infinite elements in transient thermal analysis.
Computers and Structures. 19(4), 673-687.

[172] Marques, J. M. M. C. & Owen, D. R. J. (1984). Infinite elements in quasi-static materially nonlinear
problems. Computers and Structures. 18(4), 739-751.

[173] Lysmer, J. & Kuhlemeyer, R. L. (1969). Finite dynamic model for infinite media. Proceedings of the
American Society of Civil Engineers Journal of the Engineering Mechanics Division. 95(4), 859-877.

[174] Li, H., Saigal, S., Ali, A. & Pawlak, T. P. (1994). Mapped infinite elements for 3D vector potential
magnetic problems. International Journal for Numerical Methods in Engineering. 37, 343-356.

[175] Gyimesi, M., Lavers, J., Pawlak, T., & Ostergaard, D. (1993). Biot-Savart integration for bars and arcs.
IEEE Transactions on Magnetics. 29(6), 2389-2391.

[176] Crisfield, M. A. (1981). A fast and incremental/iterative solution procedure that handles snap-
through. Computers & Structures. 13, 55-62.

[177] Nour-Omid, B. & Rankin, C. C. (1991). Finite rotation analysis and consistent linearization using
projectors. Computer Methods in Applied Mechanics and Engineering. 93, 353-384.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
934 Published: 2023-10-23
[178] Emson, C. R. I. & Simkin, J. (1983). An optimal method for 3D eddy currents. IEEE Transactions on
Magnetics. 19(6), 2450-2452.

[179] Viollet, P. L. (1987). The modelling of turbulent recirculating flows for the purpose of reactor
thermal-hydraulic analysis. Nuclear Engineering and Design. 99, 365-377.

[180] Launder, B. E. & Spalding, D. B. (1974). The numerical computation of turbulent flows. Computer
Methods In Applied Mechanics and Engineering. 3, 269-289.

[181] Rice, J. G. & Schnipke, R. J. (1985). A monotone streamline upwind finite element method for
convection-dominated flows. Computer Methods in Applied Mechanics and Engineering. 48, 313-
327.

[182] Harlow, F. H. & Amsden, A. A. (1971). A numerical fluid dynamics calculation method for all flow
speeds. Journal of Computational Physics. 8.

[183] White, F. M. (1991). Viscous Fluid Flow (2nd ed.). New York, NY: McGraw-Hill.

[184] Patankar, S. V. (1980). Numerical Heat Transfer and Fluid Flow. Hemisphere, NY: CRC.

[185] Hestenes, M. R. & Stiefel, E. (1952). Methods of conjugate gradients for solving linear system.
Journal of Research of the National Bureau of Standards. 49(6).

[186] Reid, J. K. (1971). On the method of conjugate gradients for the solution of large sparse sets of
linear equations. J. K. Reid (Ed.). In: Large Sparse Sets of Linear Equations . London, NY: Academic.

[187] Elman, H.C. (1981). Preconditioned conjugate-gradient methods for nonsymmetric systems of linear
equations. V. R. Stepleman (Ed.). Advances In Computer Methods For Partial Differential Equations
IV. 409-413.

[188] More, J. J. & Wright, S. J. (1993). Optimization Software Guide. 13. Philadelphia, PA: SIAM.

[189] Bilger, R. W. (1975). A note on Favre averaging in variable density flows. Combustion Science and
Technology. 11, 215-217.

[190] McCalla, M. C. (1988). Fundamentals of Computer-Aided Circuit Simulation. Kluwer Academic.

[191] Vermeer, P. A. & Verrujit, A. (1981). An accuracy condition for consolidation by finite elements.
International Journal for Numerical and Analytical Methods in Geomechanics. 5, 1-14.

[192] Tsai, S. W. & Hahn, H. T. (1980). Introduction to Composite Materials. (7.2). Lancaster, PA: Technomic
Publishing.

[193] Box, G. E. P., Hunter, W. G., & Hunter, J. S. (1978). Statistics for Experimenters. 10. Hoboken, NJ: Wiley.

[194] Szabo, B. & Babuska, I. (1991). Finite Element Analysis. Hoboken, NJ: Wiley.

[195] Chen, M. T. & Ali, A. (1998). An efficient and robust integration technique for applied random vi-
bration analysis. Computers and Structures. 66(6), 785-798.

[196] Harichandran, R. S. (1992). Random vibration under propagating excitation: Closed-form solutions.
Journal of Engineering Mechanics ASCE. 118(3), 575-586.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 935
Bibliography

[197] Grimes, R. G., Lewis, J. G., & Simon, H. D. (1996). A shifted block Lanczos algorithm for solving
sparse symmetric generalized eigenproblems. SIAM Journal Matrix Analysis Applications. 15(1),
228-272.

[198] Rajakumar, C. & Rogers, C. R. (1991). The Lanczos algorithm applied to unsymmetric generalized
eigenvalue problems. International Journal for Numercial Method in Engineering. 32, 1009-1026.

[199] Gartling, D. K. (1992). Finite element methods for non-Newtonian flows. Report SAND92-0886, CFD
Dept. Albuquerque, NM: Sandia National Laboratories.

[200] Crochet, M. J., Davies, A. R., & Walters, K. (1984). Numerical Simulation of Non-Newtonian Flow.
Amsterdam, The Netherlands: Elsevier Science.

[201] Hallquist, J. O. (1998). LS-DYNA Theoretical Manual. Troy, MI: Livermore Software Technology Cor-
poration.

[202] Biro, O., Preis, K., Magele, C., Renhart, W., Richter, K. R., & Vrist, G. (1991). Numerical analysis of 3D
magnetostatic fields. IEEE Transaction on Magnetics. 27(5), 3798-3803.

[203] Gymesi, M. & Ostergaard, D. (1997). Non-conforming hexahedral edge elements for magnetic
analysis. COMPUMAG. Rio de Janeiro, Brazil: Ansys, internal development.

[204] Gymesi, M. & Lavers, D. (1992). Application of general potential formulation to finite elements.
Honma, T., Sebestyen, I., & Shibata, T. (Eds.). Second Japan Hungarian Joint Seminar on Electro-
magnetics, Applied Electromagnetics in Materials and Computational Technology. Sapporo, Japan.

[205] Preis, K., Bardi, I., Biro, O., Magele, C., Vrisk, G., & Richter, K. R. (1992). Different finite element for-
mulations of 3D magnetostatic fields. IEEE Transactions on Magnetics. 28(2), 1056-1059.

[206] Nedelec, J. C. (1980). Mixed finite elements in R3. Numerical Methods. 35, 315-341.

[207] Van Welij, J. S. (1982). Calculation of eddy currents in terms of H on hexahedra. IEEE Transactions
on Magnetics. 18, 431-435.

[208] Kameari, A. (1990). Calculation of transient 3D eddy current using edge elements. IEEE Transactions
on Magnetics. 26, 466-469.

[209] Jin, J. (1993). The Finite Element Method in Electromagnetics. Hoboken, NJ: Wiley.

[210] Whitney, H. (1957). Geometric Integration Theory. Princeton, NJ: Princeton University Press.

[211] Stratton, J. A. (1941). Electromagnetic Theory. (1.14). New York, NY: McGraw-Hill.

[212] Mitzner, K. M. (1967). An integral equation approach to scattering from a body of finite conduct-
ivity. Radio Science. 2, 1459-1470.

[213] Mittra, R. & Ramahi, O. (1989). Absorbing boundary conditions for the direct solution of partial
differential equations arising in electromagnetic scattering problems. Finite Element Finite Differ-
ence Methods in Electromagnetic Scattering. 2, 133-173.

[214] Peric, D. & Owen, D. J. R. (1992). Computational model for 3D contact problems with friction based
on the penalty method. International Journal for Numercial Method in Engineering. 35, 1289-1309.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
936 Published: 2023-10-23
[215] Cescotto, S. & Charilier, R. (1992). Frictional contact finite elements based on mixed variational
principles. International Journal for Numercial Method in Engineering. 36, 1681-1701.

[216] Cescottot, S. & Zhu, Y. Y. (1994). Large strain dynamic analysis using solid and contact finite elements
based on a mixed formulation - application to metalforming. Journal of Metals Processing Tech-
nology. 45, 657-663.

[217] Simo, J. C. & Laursen, T. A. (1992). An augmented Lagrangian treatment of contact problems in-
volving friction. Computers and Structures. 42(1), 97-116.

[218] Simo, J. C. & Laursen, T. A. (1993). Algorithmic symmetrization of Coulomb frictional problems
using augmented Lagrangians. Computers Methods in Applied Mechanics and Engineering. 10(1-
2), 133-146.

[219] Barry, A., Bielak, J., & MacCamy, R. C. (1988). On absorbing boundary conditions for wave
propagations. Journal of Computational Physics. 792, 449-468.

[220] Kallivokas, L. F., Bielak, J., & MacCamy, R. C. (1991). Symmetric local absorbing boundaries in time
and space. Journal of Engineering Mechanics. 1179, 2027-2048.

[221] Hughes, T. J. R. (1980). Generalization of selective integration procedures to anisotropic and non-
linear media. International Journal for Numerical Methods in Engineering. 15(9), 1413-1418.

[222] Nagtegaal, J. C., Parks, D. M., & Rice, J. R. (1974). On numerically accurate finite element solutions
in the fully plastic range. Computer Methods in Applied Mechanics and Engineering. 4, 153-178.

[223] Gyimesi, M. & Ostergaard, D. (1998). Mixed shape non-conforming edge elements. 1998 Conference
on Electromagnetic Field Computation. Tuscon, AZ.

[224] Gyimesi, M. & Ostergaard, D. (1997). Analysis of benchmark problem TEAM20 with various formu-
lations. Proceedings of the TEAM Workshop, COMPUMAG. (18-20). Rio de Janeiro, Brazil.

[225] Gyimesi, M. & Ostergaard, D. (1997). Magnetic corner: Accurate force computations. Analysis Solutions.
1(2), 10-11.

[226] Brooks, A. N. & Hughes, T. J. R. (1982). Streamline upwind/Petro-Galkerin formulation for convection
dominated flows with particular emphasis on the incompressible Navier-Stokes equations.
Computer Methods in Applied Mechanics and Engineering. 32, 199-259.

[227] Demerdash, N. A. & Arkadan, A. A. (1981). Notes on FEM modeling of permanent magnets in
electrical devices. FEM for Electromagnetic Applications. 3, 2, 6-7, 17, 19.

[228] Demerdash, N. A. & Nehl, T. W. (1982). Determination of inductances in ferrite type magnet electric
machinery by FEM. IEEE Trans. on MAG. 18, 1052-1054.

[229] Nehl, T. W., Faud, F. A., & Demerdash, N. A. (1982). Determination of saturated values of rotation
machinery incremental and apparent inductances by an energy perturbation method. IEEE Trans.
on PAS. 101, 4441-4451.

[230] Gyimesi, M., Zhulin, V., & Ostergaard, D. (1998). Particle trajectory tracing in ANSYS. Fifth Interna-
tional Conference on Charged Particle Optics. Delft University: Delft, Netherlands.

[231] Gyimesi, M. & Ostergaard, D. (1998). Inductance Computation by Incremental Finite Element Analysis.
CEFC '98. Tucson, AZ.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 937
Bibliography

[232] Demerdash, N. A. & Gillott, D. H. (1974). A new approach for determination of eddy currents and
flux penetration in nonlinear ferromagnetic materials. IEEE Trans. on Magnetics. 10, 682-685.

[233] Flanagan, D. P. & Belytschko, T. (1981). A uniform strain hexahedron and quadrilateral with ortho-
gonal hourglass control. International Journal for Numerical Methods in Engineering. 17, 679-706.

[234] Vogel, F. (1997). Topological optimization of linear-elastic structures with ANSYS 5.4. NAFEMS
Conference on Topological Optimization. Aalen, Germany.

[235] Mlejnek, H. P. & Schirrmacher, R. (1993). An engineer's approach to optimal material distribution
and shape finding. Computer Methods in Applied Mechanics and Engineering. 106, 1-26.

[236] Bendsoe, M. P. & Kikucki, N. (1988). Generating optimal topologies in structural design using a
homogenization method. Computer Methods in Applied Mechanics and Engineering. 71, 197-224.

[237] Bonet, J. & Wood, R. D. (1997). Nonlinear Continuum Mechanics for Finite Element Analysis. Cambridge,
England: Cambridge University Press.

[238] Simo, J. C. & Vu-Quoc, L. (1986). A three dimensional finite-strain rod model. Part II: Computational
aspects. Computer Methods in Applied Mechanics and Engineering. 58, 79-116.

[239] Ibrahimbegovic, A. (1995). On finite element implementation of geometrically nonlinear reissner's

beam theory: Three-dimensional Curved Beam Elements". Computer Methods in Applied Mechanics
and Engineering. 122, 11-26.

[240] Vago, I. & Gyimesi, M. (1998). Electromagnetic Fields. Budapest, Hungary: Akademiiai Kiado.

[241] Flügge, S. (1958). Electric fields and waves. In the Encyclopedia of Physics. (Vol. 16). Berlin, DE:
Springer-Verlag.

[242] Lagally, M. (1964). Vorlesungen uber Vektorrechnung. Leipzig, Sachsen: Akademische Verlagsgesellsch.

[243] Flanagan, D. P. & Belytschko, T. (1981). A uniform strain hexahedron and quadrilateral with ortho-
gonal hourglass control. International Journal for Numerical Methods in Engineering. 17, 679-706.

[244] Callen, H. B. (1985). Thermodynamics and Introduction to Thermostatistics (2nd ed.). (84). New York,
NY: Wiley.

[245] Chaboche, J. L. (1989). Constitutive equations for cyclic plasticity and cyclic viscoplasticity. Interna-
tional Journal of Plasticity. 5, 247-302.

[246] Chaboche, J. L. (1991). On some modifications of kinematic hardening to improve the description
of ratcheting effects. International Journal of Plasticity. 7, 661-678.

[247] Timoshenko, S. & Goodier, J. (1961). Theory of Elastic Stability. New York, NY: McGraw-Hill.

[248] Schulz, M. & Fillippou, F. C. (1998). Generalized warping torsion formulation. Journal of Engineering
Mechanics. 339-347.

[249] Gyimesi, M. & Ostergaard, D. (1999). Electro-mechanical capacitor element for MEMS analysis in
ANSYS. Proceedings of Modelling and Simulation of Microsystems Conference. Aalen, Germany.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
938 Published: 2023-10-23
[250] Schulz, M. & Fillippou, F. C. (1993). Capacitance computation with ammeter element. University
of Toronto, Department of Electrical Engineering, unpublished report available upon request
from Ansys.

[251] Wang, C., Zhang, M., Wang, X., Wu, D., & Wang, H. (1999). ANSYS APDL for capacitance. Proceedings
from the Second International Conference on Modeling and Simulation of Microsystems, Semi-
conductors, Sensors and Actuators. San Juan, Puerto Rico.

[252] Simo, J. C. & Hughes, T. J. R. (1997). Computational Inelasticity. Berlin, DE: Springer-Verlag.

[253] Voce, E. (1955). Metallurgica. Col. 51, 219.

[254] Press, W. H. (1993). Numerical Recipes in C: The Art of Scienfitic Computing. Cambridge, England:
Cambridge University Press.

[255] Gyimesi, M., Lavers, D., Ostergaard, D., & Pawlak, T. (1995). Hybrid finite element-Trefftz method
for open boundary anslysis. COMPUMAG, IEEE Transactions on Magnetics. 32(3), 671-674.

[256] Gyimesi, M. & Lavers, D. (1992). Application of the Trefftz Method to Exterior Problems. University
of Toronto, Department of Electrical Engineering, unpublished report. Available upon request
from Ansys.

[257] Gyimesi, M. & Lavers, D. (1993). Application of the Trefftz Method to Exterior Problems. University
of Toronto, Department of Electrical Engineering, unpublished report. Available upon request
from Ansys.

[258] Gyimesi, M. & Lavers, D. (1993). Implementation to the Exterior Trefftz Element. University of
Toronto, Department of Electrical Engineering, unpublished report. Available upon request from
Ansys.

[259] Trefftz, E. (1926). Ein Gegenstuck zum Ritzschen Verfahren. Proceedings of the Second International
Congress on Applied Mechanics. Zurich, Switzerland.

[260] Trefftz, E. (1934). Mechanik det elastischen Korper. Handbuch der Physik, Translated from Matem-
aticheskais teoriia Uprognosti, L. GTTI 1934. 6.

[261] Herrera, I. (1983). Topics in Boundary Element Research. C. A. Brebbia (Ed.). (pp. 225-253). New York,
NY: Springer-Verlag.

[262] Zienkiewicz, O. C. (1978). The generalized finite element method and electromagnetic problems.
Conference on the Computation of Magnetic Fields, Communications Proceedings. Laboratoire
d'Electrotechnique, Grenoble, France.

[263] Zielinski, A. P. & Zienkiewicz, O. C. (1985). Generalized finite element analysis with T-complete
boundary solution function. International Journal for Numerical Methods in Engineering. 21, 509-
528.

[264] Zienkiewicz, O. C., Kelly, D. W., & Bettess, P. (1977). The coupling of the finite element method and
boundary solution procedures. International Journal for Numerical Methods in Engineering. 11,
355-375.

[265] Zienkiewicz, O. C., Kelly, D. W., & Bettess, P. (1979). Marriage a la mode - The best of both worlds
finite element and boundary integrals. Energy Methods in Finite Element Analysis. 81-104.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 939
Bibliography

[266] Jirousek, J. & Guex, L. (1986). The hybrid-Trefftz finite element model and its application to plate
bending. International Journal for Numerical Methods in Engineering. 23, 651-693.

[267] Mayergoyz, I. D., Chari, M. V. K., & Konrad, A. (1983). Boundary Galerkin's method for three-dimen-
sional finite element electromagnetic field computation. IEEE Transactions on Magnetics. 19(6),
2333-2336.

[268] Chari, M. V. K. (1987). Electromagnetic field computation of open boundary problems by semi-
analytic approach. IEEE Transactions on Magnetics. 23(5), 3566-3568.

[269] Chari, M. V. K. & Bedrosian, G. (1987). Hybrid harmonic/finite element method for two-dimensional
open boundary problems. IEEE Transactions on Magnetics. 23(5), 3572-3574.

[270] Arruda, E. M. & Boyce, M. C. (1993). A three-dimensional constitutive model for the large stretch
behavior of rubber elastic materials. Journal of the Mechanics and Physics of Solids. 41(2), 389-
412.

[271] Bergstrom, J. S. & Boyce, M. C. (1998). Constitutive modeling of the large strain time-dependent
behavior of elastomers. Journal of the Mechanics and Physics of Solids. 45(5), 931-954.

[272] Glass, M. W. (1995). Chaparral - A library package for solving large enclosure radiation heat transfer
problems. Sandia National Laboratories Report, SAND95-2049. Albuquerque, NM.

[273] Diaz, A. R. & Kikucki, N. (1992). Solutions to shape and topology eigenvalue optimization problems
using a homogenization method. International Journal for Numerical Methods in Engineering. 35,
1487-1502.

[274] Ladeveze, P. & Leguillon, D. (1983). Error estimation procedure in the finite element method and
applications. SIAM Journal of Numerical Analysis. 20(3), 483-509.

[275] Synge, J. L. (1957). The Hypercircle in Mathematical Physics. Cambridge, England: Cambridge Uni-
versity Press.

[276] Cohen, M. F. & Greenberg, D. P. (1985). The hemi-cube: A radiosity solution for complex environ-
ments. Computer Graphics. 19(3), 31-40.

[277] Williams, M. L., Landel, R. F., & Ferry, J. D. (1955). The temperature dependence of relaxation
mechanisms in amorphous polymers and other glass-forming liquids. Journal of the American
Chemical Society. 77, 3701-3706.

[278] Huerta, A. & Liu, W. K. (1988). Viscous flow with large free surface motion. Computer Methods in
Applied Mechanics and Engineering. 69, 277-324.

[279] Weaver, W. & Johnston, P. R. (1987). Structural Dynamics by Finite Elements. (413-415). Upper Saddle
River, NJ: Prentice-Hall.

[280] Zhu, Y. Y. & Cescotto, S. (1994). Transient thermal and thermomechanical analysis by mixed FEM.
Computers and Structures. 53, 275-304.

[281] Brackbill, J. U., Kothe, D. B., Zemach, C. (1992). A continuum method for modeling surface tension.
Journal of Computational Physics. 100, 335-354.

[282] Kother, D. B. & Mjolsness, R. C. (1992). RIPPLE: A new model for incompressible flows with free
surfaces. AIAA Journal. 30, 2694-2700.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
940 Published: 2023-10-23
[283] Richards, J. R., Lenhoff, A. M., & Beris, A. N. (1994). Dynamic breakup of liquid-liquid jets. Physics
of Fluids. 8, 2640-2655.

[284] Sasmal, G. P. & Hochstein, J. I. (1994). Marangoni convection with a curved and deforming free
surface in a cavity. Transaction of ASME, Journal of Fluid Engineering. 116, 577-582.

[285] Wang, G. (2000). Finite element simulations of gas-liquid flows with surface tension. Presented at
the 2000 International Mechanical Engineering Congress and Exposition. Orlando, FL.

[286] Gyimesi, M. & Ostergaard, D. (2000). Finite element based reduced order modeling of micro electro
mechanical sytems. Presented at the 2000 International Conference on Modeling and Simulation
of Microsystems. San Diego, CA.

[287] Ostergaard, D., Gyimesi, M., Affour, B., Nachtergaele, P., Stirkovich, S. (2000). Efficient reduced order
modeling for system simulation of micro electro mechanical systems from FEM models.
Presented at the Symposium on Design Test Integration and Packaging of MEMS/MOEMS. Paris,
France.

[288] Gyimesi, M., Wang, J. S., Ostergaard, D. (2000). Capacitance Computation by Hybrid P-Element and
Trefftz Method. (37, 3680-3683). Presented at Conference on Electromagnetic Field Computation.
Milwaukee, WI.

[289] Gyimesi, M. & Ostergaard, D. (2000). Capacitance Computation by Hybrid P-Element and Trefftz
Method. Presented at the 2000 Micro Electro Mechanical Systems (MEMS) Conference. San Diego,
CA.

[290] Gyimesi, M. & Ostergaard, D. (1998). Incremental magnetic inductance computation. ANSYS Con-
ference and Exhibition. Pittsburgh, PA.

[291] Hieke, A. (2000). Tiny devices, big problems: Computation of capacitance in microelectronic
structures. ANSYS Solutions. 2(3), 11-15.

[292] Gadala, M. S. & Wang, J. (1999). Simulation of metal forming processes with finite element methods.
International Journal for Numerical Methods in Engineering. 44, 1397-1428.

[293] McMeeking, R. M. & Rice, J. R. (1975). Finite element formulations for problems of large elastic-
plastic deformation. International Journal of Solids and Structures. 121, 601-616.

[294] Crisfield, M. A. (1997). Non-linear finite element analysis of solids and structures - Advanced topics.
(Vol. 2). Hoboken, NJ: Wiley.

[295] Ogden, R. W. (1984). Nonlinear Elastic Deformations. Mineola, NY: Dover Publications.

[296] Perzyna, P. (1968). Fundamental problems in viscoplasticity. Advances in Applied Mechanics. 9, 313-
377.

[297] Peirce, D., Shih, C. F., & Needleman, A. (1984). A tangent modulus method for rate dependent
solids. Computers & Structures. 18, 888-975.

[298] Peri , D. & Owen, D. R. J. (1992). A model for large deformations of elasto-viscoplastic solids at finite
strains: computational issues. Finite Inelastic Deformations: Theory and Applications. Berlin, DE:
Springer-Verlag.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 941
Bibliography

[299] Volakis, J. L., Chatterjee, A., & Kempel, L. C. (1998). Finite Element Method for Electromagnetics: An-
tennas, Microwave Circuits and Scattering Applications. New York, NY: IEEE.

[300] Itoh, T., Pelosi, G., & Silvester, P. P. (1996). Finite Element Software for Microwave Engineering.
Hoboken, NJ: Wiley.

[301] L. Zhao, L. & Cangellaris, A. C. (1996). GT-PML: Generalized theory of perfectly matched layers and
its application to the reflectionless truncation of finite-difference time-domain grids. IEEE Trans.
on Microwave Theory and Techniques. 44, 2555-2563.

[302] George, A. & Liu, J. W. (1981). Computer Solution of Large Sparse Positive Definite Systems. Upper
Saddle River, NJ: Prentice-Hall.

[303] Abramowitz, M. & Stegun, I. A. (1984). Pocketbook of Mathematical Functions -- Abridged Version
of the Handbook of Mathematical Functions. Providence, Rhode Island: American Mathematical
Society.

[304] Ang, A. H. & Tang, W. H. (1975). Probability Concepts in Engineering Planning and Design - Basic
Principles. (Vol. 1). Hoboken, NJ: Wiley.

[305] Ang, A. H. & Tang, W. H. (1990). Probability Concepts in Engineering Planning and Design - Decision,
Risk, and Reliability. (Vol. 2). Hoboken, NJ: Wiley.

[306] Box, G. E. P. & Behnken, D. W. (1960). Some new three level designs for the study of quantitative
variables. Technometrics. 2(4), 455-476.

[307] Box, G. E. P. & Cox, D. R. (1964). An analysis of transformations. Journal of the Royal Statistical Society.
B, 26, 211-252.

[308] Hammersly, J. M. & Handscomb, D. C. (1964). Monte Carlo Methods. Hoboken, NJ: Wiley.

[309] Iman, R. L. & Conover, W. J. (1980). Small sample sensitivity analysis techniques for computer
models, with an application to risk assessment. Communications in Statistics, Part A - Theory and
Methods. A9(17), 1749-1842.

[310] Kececioglu, D. (1991). Reliability Engineering Handbook. (Vol. 1). Englewood Cliffs, NJ: Prentice-Hall.

[311] Liu, P. L. & Der Kiureghian, A. (1986). Multivariate distribution models with prescribed marginals
and covariances. Probabilistic Engineering Mechanics. 1(2), 105-112.

[312] Montgomery, D. C. (1991). Design and Analysis of Experiments. Hoboken, NJ: Wiley.

[313] Myers, R. C. (1971). Response Surface Methodology. Boston, MA: Allyn and Bacon.

[314] Neter, J., Kutner, M. H., Nachtscheim, C. J., & Wasserman, W. (1996). Applied Linear Statistical Models
(4th ed.). New York, NY: McGraw-Hill.

[315] Sheskin, D. J. (1997). Handbook of Parametric and Nonparametric Statistical Procedures. Boca Raton,
FL: CRC.

[316] Hancq, D. A., Walter, A. J., & Beuth, J. L. (2000). Development of an object-oriented fatigue tool.
Engineering with Computers. 16, 131-144.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
942 Published: 2023-10-23
[317] Benson, D. J. & Hallquist, J. (1990). A single surface contact algorithm for the post-buckling analysis
of shell structures. Computer Methods in Applied Mechanics and Engineering. 78(2).

[318] Simo, J. C. & Rifai, M. S. (1990). A class of mixed assumed strain methods and the method of in-
compatible modes. International Journal for Numerical Methods in Engineering. 29, 1595-1638.

[319] Simo, J. C. & Armero, F. (1992). Geometrically non-linear enhanced strain mixed methods and the
method of incompatible modes. International Journal for Numerical Methods in Engineering. 33,
1413-1449.

[320] Simo, J. C., Amero, F., & Taylor, R. L. (1993). Improved versions of assumed enhanced strain tri-linear
elements for 3D finite deformation problems. Computer Methods in Applied Mechanics and En-
gieering. 10, 359-386.

[321] Andelfinger, U. & Ramm, E. (1993). EAS-elements for two-dimensional, three-dimensional, plate
and shell structures and their equivalence to HR-elements. International Journal for Numerical
Methods in Engineering. 36, 1311-1337.

[322] Nagtegaal, J. C. & Fox, D. D. (1996). Using assumed enhanced strain elements for large compressive
deformation. International Journal for Solids and Structures. 33, 3151-3159.

[323] Wang, J. S. & Ostergaard, D. F. (1999). Finite element-electric circuit coupled simulation method
for piezoelectric transducer. Proceedings of the IEEE Ultrasonics Symposium. 2, 1105-1108.

[324] Pipkin, A. C. (1986). Lectures in Viscoelasticity Theory. New York, NY: Springer-Verlag.

[325] Drozdov, D. A. (1996). Finite Elasticity and Viscoelasticity: A Course in the Nonlinear Mechanics of
Solids. New York, NY: World Publishing.

[326] Scherer, G. W. (1986). Relaxation in Glass and Composites. Hoboken, NJ: Wiley.

[327] Simo, J. C. (1987). On fully three-dimensional finite-strain viscoelastic damage model: Formulation
and computational aspects. Computer Methods In Applied Mechechanics and Engineering. 60, 153-
173.

[328] Holzapfel, G. A. (1996). On large strain viscoelasticity: continuum formulation and finite element
applications to elastomeric structures. International Journal for Numerical Methods in Engineering.
39, 3903-3926.

[329] Gyimesi, M., Ostergaard, D., & Avdeev, I. (2002). Triangle Transducer for Micro Electro Mechanical
Systems MEMS Simulation in ANSYS Finite Element Program. 2002 International Conference on
Modeling and Simulation of Microsystems. San Francisco, CA.

[330] Gyimesi, M. & Ostergaard, D. (2001). A transducer finite element for dynamic coupled electrostatic-
structural coupling simulation of microelectromechanical systems devices. MIT Conference.
Cambridge, MA.

[331] Avdeev, I., Gyimesi, M., Lovell, M., & Onipede, D. (2001). Beam modeling for simulation of electro
mechanical transducers using strong coupling approach. 6th U.S. National Congress on Compu-
tational Mechanics. Dearborn, MI.

[332] Chen, W. F. & Han, D. J. (1988). Plasticity for Structural Engineers. New York, NY: Springer-Verlag.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 943
Bibliography

[333] Guilllaume, P. (1994). Derivees d'ordre superieur en conception optimale de forme. These de l'universite
Paul Sabatier de Toulouse. Paul Sabatier University, FR.

[334] Hjelm, H. E. (1994). Yield surface for gray cast iron under biaxial stress. Journal of Engineering Ma-
terials and Technology. 116, 148-154.

[335] Blech, J. J. (1983). On isothermal squeeze films. Journal of Lubrication Technology. 105, 615-620.

[336] Griffen, W. S., Richardson, H. H., & Yamanami, S. (1966). A study of squeeze-film damping. Journal
of Basic Engineering. 451-456.

[337] Langlois, W. E. (1962). Isothermal squeeze films. Quarterly Applied Mathematics. 20(2), 131-150.

[338] Mehner, J. , Kurth, S., Kaufmann, C., Kehr, K., & Dotzel, W. (1998). Simulation of gas film damping
on microstructures with nontrivial geometries. Proceedings of the MEMS Conference, Eleventh
Annual International Workshop of Micro Electro Mechanical Systems. Heidelberg, DE.

[339] Veijola, T. (1999). Equivalent circuit models for micromechanical inertial sensors. Helsinki University
of Technology: Circuit Theory Laboratory Report Series CT-39.

[340] Sharipov, F. (1999). Rarefied gas flow through a long rectangular channel. Journal Vacuum Science
and Technology. A175, 3062-3066.

[341] Craig, R. J. (1987). A review of time domain and frequency domain component mode synthesis
methods. International Journal of Analytical and Experimental Modal Analysis. 2(2), 59-72.

[342] Craig, R. R. & Bampton, M. D. D. (1968). Coupling of substructures for dynamic analysis. American
Institute of Aeronautics and Astronautics Journal. 12, 1313-1319.

[343] Gyimesi, M., Avdeev, I., & Ostergaard, D. (2004). Finite element simulation of micro electro mech-
anical systems (MEMS) by strongly coupled electro mechanical transducers. IEEE Transactions on
Magnetics. 40(2), 557-560.

[344] Auricchio, F., Taylor, R. L., & Lubliner, J. (1997). Shape-memory alloys: macromodeling and numer-
ical simulations of the superelastic behavior. Computational Methods in Applied Mechanical En-
gineering. 146, 281-312.

[345] Belytschko, T., Liu, W. K., Moran, B. (2000). Nonlinear finite elements for continua and structures.
Computational Methods in Applied Mechanical Engineering. Hoboken, NJ: Wiley.

[346] Wilcox, D. C. (1988). Reassessment of the scale-determining equation for advanced turbulence
models. American Institute of Aeuronautics and Astronautics Journal. 26, 1299-1310.

[347] Menter, F. R. (1994). Two-equation eddy-viscosity turbulence models for engineering applications.
American Institute of Aeuronautics and Astronautics Journal. 32, 1598-1605.

[348] Chung, J & Hulbert, G. M. (1993). A time integration algorithm for structural dynamics with improved
numerical dissipation: The generalized-α method. Journal of Applied Mechanics. 60, 371.

[349] Hilber, H. M., Hughes, T. J. R., & Taylor, R. L. (1977). Improved numerical dissipation for time integ-
ration algorithm in structural dynamics. Earthquake Engineering and Structural Dynamics. 5, 283.

[350] Wood, W. L., Bossak, M., & Zienkiewicz, O. C. (1981). An alpha modification of newmark method.
International Journal of Numerical Method in Engineering. 15, 1562.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
944 Published: 2023-10-23
[351] Segalman, D. J., Fulcher, C. W. G., Reese, G. M., & Field, R. V. (1998). An efficient method for calcu-
lating RSM von Mises stress in a random vibration environment. (Vol. 1, Issue 3243, pp. 117-123).
Proceedings of the 16th International Modal Analysis Conference. Santa Barbara, CA.

[352] Reese, G. M., Field Jr., R. V., & Segalman, D. J. (2000). A tutorial on design analysis using von Mises
stress in random vibration environments. The Shock and Vibration Digest. 32(6), 466-474 .

[353] Chapman, A. J. (1984). Heat Transfer (4th ed.). New York, NY: Macmillan.

[354] Wilkinson, J. H. & Reinsch, C. (1971). Linear algebra. Handbook for Automatic Computation. 2, 418-
439.

[355] Landau, L. D. & Lifshitz, E. M. (1984). Course of Theoretical Physics (2nd ed.). (Vol. 8). New York, NY:
Pergamon.

[356] Nye, J. F. (1957). Physical Properties of Crystals: Their Representation by Tensors and Matrices. Oxford,
UK: Clarendon.

[357] Beley, J. D., Broudiscou, C., Guillaume, P., Masmoudi, M., Thevenon, F. (1996). Application de la
methode des derivees d'ordre eleve a l'optimisation des structures. Revue Europeenne des Elements
Finis. 5(5-6), 537-567.

[358] R. Michalowski, R. & Mroz, Z. (1978). Associated and non-associated sliding rules in contact friction
problems. Archives of Mechanics. 30(3), 259-276.

[359] Nelson, H. D, McVaugh, J. M. (1976). The dynamics of rotor-bearing systems using finite elements.
Journal of Engineering for Industry. 98(2), 593-600.

[360] Xu, X. P. & Needleman, A. (1994). Numerical simulations of fast crack-growth in brittle solids.
Journal of the Mechanics and Physics of Solids. 42, 1397-1434.

[361] Guo, D., Chu, F. L., & Zheng, Z. C. (2001). The influence of rotation on vibration of a thick cylindrical
shell. Journal of Sound and Vibration. 242(3), 492.

[362] Alfano, G. & Crisfield, M. A. (2001). Finite element interface models for the delamination anaylsis
of laminated composites: mechanical and computational issues. International Journal for Numer-
ical Methods in Engineering. 50, 1701-1736.

[363] Gurson, A. L. (1977). Continuum theory of ductile rupture by void nucleation and growth: Part I-
yield criterion and flow rules for porous ductile media. Journal of Engineering Materials and
Technology. 1, 2-15.

[364] Needleman, A & Tvergaard, V. (1984). An analysis of ductile rupture in notched bars. Journal of
Mechanical Physics Solids. 32(6), 461-490.

[365] Geradin, M. & Rixen, D. (1984). Mechanical Vibrations: Theory and application to structural dynamics.
(p. 194). Hoboken, NJ: Wiley.

[366] Bathe, K. J. & Almeida, C. A. (1980). A simple and effective pipe elbow element - linear analysis.
Journal of Applied Mechanics. 47(1), 93-100.

[367] Yan, A. M., Jospin, R. J., & Nguyen, D. H. (1999). An enhanced pipe elbow element - Application
in plastic limit analysis of pipe structures. International Journal for Numerical Methods in Engineer-
ing. 46, 409-431.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 945
Bibliography

[368] Bergstrom, J. S. & Boyce, M. C. (1998). Constitutive modeling of the large strain time-dependent
behavior of elastomers. Journal of the Mechanics and Physics of Solids. 46, 931-954.

[369] J.S. Bergstrom, J. S. & Boyce, M. C. (2000). Large strain time-dependent behavior of filled elastomers.
Mechanics of Materials. 32, 627-644.

[370] Dal, H. & Kaliske, M. (2009). Bergstrom-Boyce model for nonlinear finite rubber viscoelasticity:
Theoretical aspects and algorithmic treatment for the FE method. Computational Mechanics. 44,
809-823.

[371] Gupta, A. K. (1990). Response Spectrum Method In Seismic Analysis and Design of Structures. Boca
Raton, FL: CRC.

[372] NRC Regulatory Guide. (2006). Regulatory Guide 1.92, Revision 2. Rockville, MD: Published by the
U.S. Nuclear Regulatory Commission.

[373] Laursen, T. A. & Chawla, V. (1997). Design of energy conserving algorithms for frictionless dynamic
contact problems. International Journal for Numerical Methods in Engineering. 40, 863-886.

[374] Armero, F. & Pet cz, E. (1998). Formulation and analysis of conserving algorithms for dynamic
contact/impact problems. Computer Methods in Applied Mechanics and Engineering. 158, 269-300.

[375] Ogden, R. W. & Roxburgh, D. G. (1999). A pseudo-elastic model for the mullins effect in filled
rubber. Proceedings of the Royal Society of London, Series A (Mathematical and Physical Sciences).
455(1988), 2861-2877.

[376] Bose, K., Hurtado, J., Snyman, M., Mars, W., & Chen, J. (2003). Modeling of stress softening in filled
elastomers. Busfield, J. C. & Muhr, A. H. (Eds.). Constitutive Models for Rubber III: Proceedings of
the 3rd European Conference on Constitutive Models for Rubber (ECCMR). 223-230.

[377] Geradin, M. & Kill, N. (1984). A new approach to finite element modelling of flexible rotors. Engin-
eering Computing. 1.

[378] Gurtin, M. E. (1981). An Introduction to Continuum Mechanics. Orlando, FL: Academic.

[379] Puso, M. A. & Laursen, T. A. (2004). A mortar segment-to-segment contact method for large de-
formation solid mechanics. Computer Methods in Applied Mechanics and Engineering. 193, 601-
629.

[380] Puso, M. A. & Laursen, T. A. (2004). A mortar segment-to-segment frictional contact method for
large deformations. Computer Methods in Applied Mechanics and Engineering. 193, 4891-4913.

[381] Arndt, S., Svenson, B., & Klingbeil, E. (1997). Modellierung der Eigenspannungen an der Rißspitze
mit einem Schädigungsmodell. Technische Mechanik. 4, 323-332.

[382] Besson, J. & Guillemer-Neel, C. (2003). An extension of the green and gurson models to kinematic
hardening. Mechanics of Materials. 35, 1-18.

[383] Sandler, I. S., DiMaggio, F. L., & Baladi, G. Y. (1976). Generalized cap model for geological materials.
Journal of the Geotechnical Engineering Division. 102, 683-699.

[384] Pelessone, D. (1989). A Modified Formulation of the Cap Model. Gulf Atomic Report GA-C19579,
prepared for the Defense Nuclear Agency under Contract DNA-001086-C-0277.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
946 Published: 2023-10-23
[385] Schwer, L. E. & Murray, Y. D. (1994). A three-invariant smooth cap model with mixed hardening.
International Journal for Numerical and Analytical methods in Geomechanics. 18, 657-688.

[386] Foster, C. D., Regueiro, R. A., Fossum, A. F., & Borja, R. I. (2005). Implicit numerical integration of a
three-invariant, isotropic/kinematic hardening cap plasticity model for geomaterials. Computer
Methods in Applied Mechanics and Engineering. 194, 50-52, 5109-5138.

[387] Liu, R., Lin, G., & Bhashyam, G. (2010). Transformation of the Sandler and Rubin nonsmooth cap
model to the Pelessone smooth cap model. ASCE Journal of Engineering Mechanics. 136, 680-685.

[388] WS Atkins PLC. Shell new wave. Engineering Software Report. No. ESR960611.

[389] O'Niell, L. A., Fakas, E., & Cassidy, M. (2004). A novel application of constrained new wave theory
for floatover deck installations. Proceedings of the ASME 2004 23rd International Conference on
Offshore Mechanics and Arctic Engineering. 165-174.

[390] Pinna, R. & Cassidy, M. (2004). Dynamic analysis of a monopod platform using constrained new
wave. Proceedings of the ASME 2004 23rd International Conference on Offshore Mechanics and
Arctic Engineering. 141-148.

[391] Pierson, W. J. & Moskowitz, L. (1964). A proposed spectral form of fully developed wind seas based
on the similarity theory of S. A. Kitaigorodskii. Journal of Geophysical Research. 69, 5181-5190.

[392] Hasselmann, D. E., Dunckel, M., & Ewing, J. A. (1980). Directional wave spectra observed during
Joint North Sea Wave Project (JONSWAP) 1973. Journal Physical Oceanography. 10, 1264-1280.

[393] American Petroleum Institute. (1993). Recommended practice for planning, designing and con-
structing fixed offshore platforms working stress design. Recommended Practice 2A-WSD (20th
ed.). Washington, DC: American Petroleum Institute Publishing Services.

[394] Bajer, A., Belsky, V., Kung, S. W. (2004). Paper Number 2004-01-279: The influence of friction-induced
damping and nonlinear effects on brake squeal analysis. 22nd Annual Brake Colloquium & Ex-
hibition. Anheim, CA: SAE International.

[395] Tool, A. Q. (1946). Relation between inelastic deformability and thermal expansion of glass in its
annealing range. Journal of the American Ceramic Society. 67(10), 240-253.

[396] Markovsky, A. Q., Soules, T. F., & Boyd, D. C. (1984). An efficient and stable algorithm for calculating
fictive temperatures. Journal of the American Ceramic Society. 67, c56-c57.

[397] Hatch, M. R. (2001). Vibration Analysis using Matlab and ANSYS. Boca Raton, FL: Chapman & Hall.

[398] Puck, A., Kopp, J., & Knop, M. (2002). Guidelines for the determination of the parameters in Puck's
action plane strength criterion. Composites Science and Technology. 62, 371-378.

[399] Davila, C. G., Jaunky, N., & Goswami, S. (2003). Failure criteria for FRP laminates in plane stress.
44th AIAA/ASME/ASCE/AHS Structures, Structural Dynamics, and Materials Conference. Norfolk, VA.

[400] Pinho, S. T., Dalvia, C. G., Camanho, P. P., Iannucci, L., & Robinson, P. (2005). Failure Models and
Criteria for FRP Under In-Plane or Three-Dimensional Stress States Including Shear Non-linearity.
NASA/TM-2005-213530.

[401] Wong, E. H., Teo, Y. C., & Lim, T. B. (1998). Moisture diffusion and vapour pressure modeling of IC
packaging. (pp. 1372-1378). 48th Electronics Components and Technology Conference. Seattle, WA.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 947
Bibliography

[402] Newmark, N. M. (1959). Method of computation for structural dynamics. ASCE Journal of Engineering
Mechanics Division. 85, 67-94.

[403] Zhao, L. & Cangerllais, A. C. (1996). GT-PML: Generalized theory of perfectly matched layers and
its application to the reflectionless truncation of finite-difference time-domain grids. IEEE Trans.
Microwave Theory Tech. 44, 2555-2563.

[404] Allard, J. F. & Atalla, N. (2009). Propagation of Sound in Porous Media: Modelling Sound Absorbing
Materials (2nd ed.). Hoboken, NJ: Wiley.

[405] Whitaker, J. C. (2005). The Electronics Handbook. Boca Raton, FL: CRC.

[406] Hintz, R. M. (1975). Analytical methods in component modal synthesis. American Institute of Aero-
nautics and Astronautics Journal. 13(8), 1007-1016.

[407] Martinez, D. R., Carne, T. G,. Gregory, D. L., & Miller, A. K. (1984). Combined experimental/analytical
modeling using component mode synthesis. (pp. 140-152). AIAA/ASME/ASCE/AHS 25th Structures,
Structural Dynamics and Materials Conference. Palm Springs, CA.

[408] NAVSEA. (1995). Shock design criteria for surface ships: 0908-LP-000-3010 REV-1. Naval Sea Systems
Command. Washington, D.C.

[409] Herting, D. N. (1985). A general purpose, multi-stage, component modal synthesis method. Finite
Elements in Analysis and Design. 1, 153-164.

[410] Graglia, R. D., Wilton, D. R., & Peterson, A. F. (1997). Higher order interpolatory vector bases for
computational electromagnetics. IEEE Transactions on Antennas and Propagation. 45, 329-342.

[411] Bossart, R., Joly, N., & Bruneau, M. (2003). Hybrid numerical and analytical systems for acoustic
boundary problems in thermo-viscous fluids. Journal of Sound and Vibration. 263, 69-84.

[412] Beltman, M. (1998). Visco-thermal wave propagating including acousto-elastic interaction. PhD
Thesis. University of Twente, Mechanical Engineering Department. Enschede, Netherlands.

[413] Gao, Y. & Bower, A. F. (2004). A simple technique for avoiding convergence problems in finite
element simulations of crack nucleation and growth on cohesive interfaces. Modelling and Sim-
ulation in Materials Science and Engineering. 12, 453-463.

[414] Dalrymple, R. A. & Cox, J. C. (1976). Symmetric finite-amplitude rotational water waves. Journal of
Physical Oceanography. 6, 847-852.

[415] Meng, H. C. & Ludema, K. C. (1995). Wear models and predictive equations: Their form and content.
Wear. 181-183, 443-457.

[416] Archard, J. F. (1953). Contact and rubbing of flat surfaces. Journal of Applied Physics. 24(1), 18-28.

[417] Hartzman, M., Reich, P., & Bezler, P. (1980). Piping benchmark problems - Dynamics analysis uniform
support motion response method. Nuclear Regulatory Commission - NUREG/CR-1677. Rockville,
MD.

[418] Paultre, P. (2010). Dynamics of Structures. (pp. 784). London, England: ISTE-Wiley.

[419] Léger, P., Idé, I. M., & Paultre, P. (1990). Multiple-support seismic analysis of large structures.
Computers & Structures. 36(6), 1153-1158.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
948 Published: 2023-10-23
[420] Dickens, J. M., Nakagawa, J. M., & Wittbrodt, M. J. (1997). A critique of mode acceleration and
modal truncation augmentation methods for modal response analysis. Computers & Structures.
62(6), 985-998.

[421] Biswas, J. K. & Duff, C. G. (1978). Response spectrum method with residual terms. ASME/CSME
Pressure Vessels and Piping Conference with Nuclear Materials.

[422] Rafaely, B. (2000). Spatial-temporal correlation of a diffuse sound field. Journal of the Acoustical
Society of America. 17, 3254-3258.

[423] Van den Nieuwenhof, B., Lielens, B., & Coyette, J. P. (2010). Modeling acoustic diffuse fields: Updated
sampling procedure and spatial correlation function eliminating grazing incidences. (pp. 4723-
4736). Proceedings of the International Conference on Noise and Vibration Engineering including
USD2010.

[424] Nackenhorst, U. (2010). The ALE-formulation of bodies in rolling contact: Theoretical foundations
and finite element approach. Computer Methods in Applied Mechanics and Engineering. 193(39-
41), 4299-4322.

[425] Lim, S. H., Bladh, R., Castanier, M. P., & Pierre, C. (2007). Compact, generalized component mode
mistuning representation for modeling bladed disk vibration. American Institute of Aeronautics
and Astronautics Journal. 45(9), 2285-2298.

[426] Zienkiewicz, O. C. & Shiomi, T. (1984). Dynamic behaviour of saturated porous media; the generalized
biot formulation and its numerical solution. International Journal for Numerical and Analytical
Methods in Geomechanics. 8(1), 71-96.

[427] M hlich, U. & Brocks, W. (2003). On the numerical integration of a class of pressure-dependent
plasticity models including kinematic hardening. Computational Mechanics. 31(6), 479-488.

[428] Kirk, R. G. & Cunter, E. J. (1976). Short bearing analysis applied to rotor dynamics - Part I theory -
Part II results of journal bearing response. Transactions of ASME - Journal of Lubrication Technology.

[429] Bonneau, D., Fatu, A., & Souchet, D. (2014). Hydrodynamic Bearings. Hoboken, NJ: Wiley.

[430] Hirani, H., Athre, K., & Biswas, S. (1999). Dynamically loaded finite length journal bearings: Analyt-
ical method of solution. 121.

[431] He, Z., Epureanu, B., & Pierre, C. (2007). Fluid-structural coupling effects on the dynamics of mistuned
bladed disks. American Institute of Aeronautics and Astronautics Journal. 45(3), 552-561.

[432] Zienkiewicz, O. C., Xie, Y. M., Schrefler, B. A., Ledesma, A., & Bicanic, N. (1990). Static and dynamic
behaviour of soils: A rational approach to quantitative solutions. II: Semi-saturated problems.
Proceedings of the Royal Society of London. 429(1877), 311-321.

[433] Govindjeet, S. & Mihalic, P. A. (1996). Computational methods for inverse finite elastostatics.
Computer Methods in Applied Mechanics and Engineering. 136, 47-57.

[434] Govindjeet, S. & Mihalic, P. A. (1998). Computational methods for inverse deformations in quasi-
incompressible finite elasticity. International Journal for Numerical Methods in Engineering. 45,
821-838.

[435] Lu, J. & Li, L. (2016). On referential and spatial formulations of inverse elastostatic analysis. Computer
Methods in Applied Mechanics and Engineering. 310, 189-207.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
Published: 2023-10-23 949
Bibliography

[436] Kampinga, W. R., Wijnant, Y. H., & da Boer, A. (2010). Performance of several viscothermal acoustic
finite elements. ACTA Acustica United with Acustica. 96, 115-124.

[437] Olovsson, L., Simonsson, K., & Unosson, M. (2005). Selective mass scaling for explicit finite element
analyses. International Journal for Numerical Methods in Engineering. 63, 1436-1445.

[438] Landshoff, R. (1955). A numerical method for treating fluid flow in the presence of shocks. Los
Alamos National Laboratory. Los Alamos, NM.

[439] Von Neumann, J. & Richtmyer, R. D. (1950). A method for the numerical calculation of hydrodynamic
shocks. Journal of Applied Physics. 21, 232.

[440] Allard, J. F. & Attala N. (2009). Propagation of Sound in Porous Media (2nd ed.). Hoboken, NJ: Wiley.

[441] Crisfield, M. A. (1991). Non-Linear Finite Element Analysis of Solids and Structures. Hoboken, NJ: Wiley.

[442] Assi, H. & Cobbold, R. S.. (2016). A second-order, perfectly matched layer formulation to model 3D
transient wave propagation in anisotropic elastic media. Proceedings of Meetings on Acoustics.
29, 045009.

[443] Popp, A., Gitterle, M., Gee, M.W., & Wall, W.A.. (2010). A dual mortar approach for 3D finite deform-
ation contact with consistent linearization. International Journal for Numerical Methods in Engin-
eering. 83, 1428-1465.

[444] Dirkse, B.. (2014). Finite Element Method Applied to the. One-dimensional Westervelt Equation.
PhD Thesis. Delft University of Technology, Department of Numerical Analysis. Delft, Netherlands.

[445] Hoffelner, J. & Kaltenbacher, M.. (2001). Finite Element Simulation of Nonlinear Wave Propagation
in Thermoviscous Fluids Including Dissipation. IEEE Transactions on Utrasonics, Ferroelectrics, and
Frequency Control. 48(3), 779-786.

Release 2024 R1 - © ANSYS, Inc. All rights reserved. - Contains proprietary and confidential information
of ANSYS, Inc. and its subsidiaries and affiliates.
950 Published: 2023-10-23