Acoustics Module

User’s Guide
Acoustics Module User’s Guide
© 1998–2021 COMSOL
Protected by patents listed on [Link]/patents, or see Help>About COMSOL Multiphysics on the
File menu in the COMSOL Desktop for a less detailed lists of U.S. Patents that may apply. Patents pending.
This Documentation and the Programs described herein are furnished under the COMSOL Software License
Agreement ([Link]/comsol-license-agreement) and may be used or copied only under the terms
of the license agreement.
COMSOL, the COMSOL logo, COMSOL Multiphysics, COMSOL Desktop, COMSOL Compiler,
COMSOL Server, and LiveLink are either registered trademarks or trademarks of COMSOL AB. All other
trademarks are the property of their respective owners, and COMSOL AB and its subsidiaries and products
are not affiliated with, endorsed by, sponsored by, or supported by those trademark owners. For a list of such
trademark owners, see [Link]/trademarks.
Version: COMSOL 6.0

Contact Information
Visit the Contact COMSOL page at [Link]/contact to submit general inquiries
or search for an address and phone number. You can also visit the Worldwide Sales Offices
page at [Link]/contact/offices for address and contact information.

If you need to contact Support, an online request form is located at the COMSOL Access
page at [Link]/support/case. Other useful links include:

• Support Center: [Link]/support

• Product Download: [Link]/product-download
• Product Updates: [Link]/support/updates
• COMSOL Blog: [Link]/blogs
• Discussion Forum: [Link]/forum
• Events: [Link]/events
• COMSOL Video Gallery: [Link]/videos
• Support Knowledge Base: [Link]/support/knowledgebase

Part number: CM020201

C o n t e n t s

Chapter 1: Introduction

Acoustics Module Capabilities 27

What Can the Acoustics Module Do? . . . . . . . . . . . . . . . 27
What Are the Application Areas? . . . . . . . . . . . . . . . . 30
Which Problems Can You Solve?. . . . . . . . . . . . . . . . . 31

Fundamental of Acoustics 33
Acoustics Explained . . . . . . . . . . . . . . . . . . . . . . 33
Mathematical Models for Acoustic Analysis . . . . . . . . . . . . . 34
Damping . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Artificial Boundaries. . . . . . . . . . . . . . . . . . . . . . 38

Acoustics Module Physics Interface Guide 40

Common Physics Interface and Feature Settings and Nodes . . . . . . 46
Where Do I Access the Documentation and Application Libraries? . . . . 47

Overview of the User’s Guide 51

Chapter 2: Pressure Acoustics Interfaces

The Pressure Acoustics, Frequency Domain Interface 59

Domain, Boundary, Edge, Point, and Pair Nodes for the Pressure
Acoustics, Frequency Domain Interface . . . . . . . . . . . . . 64
Pressure Acoustics . . . . . . . . . . . . . . . . . . . . . . 67
Poroacoustics . . . . . . . . . . . . . . . . . . . . . . . . 74
Narrow Region Acoustics . . . . . . . . . . . . . . . . . . . 82
Anisotropic Acoustics . . . . . . . . . . . . . . . . . . . . . 85
Background Pressure Field . . . . . . . . . . . . . . . . . . . 86
Initial Values . . . . . . . . . . . . . . . . . . . . . . . . 89
Monopole Domain Source . . . . . . . . . . . . . . . . . . . 89
Dipole Domain Source . . . . . . . . . . . . . . . . . . . . 90

CONTENTS |3
 Heat Source . . . . . . . . . . . . . . . . . . . . . . . . 90
Aeroacoustic Flow Source . . . . . . . . . . . . . . . . . . . 91
Sound Hard Boundary (Wall) . . . . . . . . . . . . . . . . . . 92
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . . 92
Normal Acceleration . . . . . . . . . . . . . . . . . . . . . 92
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . . 93
Normal Displacement . . . . . . . . . . . . . . . . . . . . . 93
Sound Soft Boundary . . . . . . . . . . . . . . . . . . . . . 94
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . . 94
Impedance . . . . . . . . . . . . . . . . . . . . . . . . . 95
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 100
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 100
Matched Boundary . . . . . . . . . . . . . . . . . . . . . 102
Exterior Field Calculation . . . . . . . . . . . . . . . . . . 103
Port . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
Circular Port Reference Axis . . . . . . . . . . . . . . . . . 113
Lumped Port . . . . . . . . . . . . . . . . . . . . . . . 113
Lumped Speaker Boundary . . . . . . . . . . . . . . . . . . 116
Thermoviscous Boundary Layer Impedance . . . . . . . . . . . . 117
Plane Wave Radiation . . . . . . . . . . . . . . . . . . . . 119
Spherical Wave Radiation . . . . . . . . . . . . . . . . . . 120
Cylindrical Wave Radiation . . . . . . . . . . . . . . . . . . 121
Incident Pressure Field. . . . . . . . . . . . . . . . . . . . 121
Perfectly Matched Boundary . . . . . . . . . . . . . . . . . 124
Interior Sound Hard Boundary (Wall) . . . . . . . . . . . . . . 125
Interior Normal Acceleration . . . . . . . . . . . . . . . . . 126
Interior Normal Velocity . . . . . . . . . . . . . . . . . . . 127
Interior Normal Displacement. . . . . . . . . . . . . . . . . 127
Interior Impedance/Pair Impedance . . . . . . . . . . . . . . . 127
Interior Perforated Plate/Pair Perforated Plate. . . . . . . . . . . 128
Interior Lumped Speaker Boundary . . . . . . . . . . . . . . . 130
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 132
Line Source . . . . . . . . . . . . . . . . . . . . . . . . 132
Line Source on Axis. . . . . . . . . . . . . . . . . . . . . 134
Monopole Point Source . . . . . . . . . . . . . . . . . . . 135
Dipole Point Source. . . . . . . . . . . . . . . . . . . . . 137
Quadrupole Point Source . . . . . . . . . . . . . . . . . . 139
Point Sources (for 2D Components) . . . . . . . . . . . . . . 141

4 | CONTENTS
Circular Source (for 2D Axisymmetric Components) . . . . . . . . 143
Pressure (Point Condition) . . . . . . . . . . . . . . . . . . 144

The Pressure Acoustics, Transient Interface 145

Domain, Boundary, Edge, and Point Nodes for the Pressure
Acoustics, Transient Interface . . . . . . . . . . . . . . . 148
Transient Pressure Acoustics Model . . . . . . . . . . . . . . 148
Nonlinear Acoustics (Westervelt) Contributions. . . . . . . . . . 149
Background Pressure Field (for Transient Models) . . . . . . . . . 152
Incident Pressure Field (for Transient Models). . . . . . . . . . . 154
The Gaussian Pulse Source Type Settings. . . . . . . . . . . . . 155
Normal Acceleration . . . . . . . . . . . . . . . . . . . . 156
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . 156
Normal Displacement . . . . . . . . . . . . . . . . . . . . 157
Exterior Field Calculation (for Transient Models) . . . . . . . . . 157
Interior Normal Acceleration . . . . . . . . . . . . . . . . . 158
Interior Normal Velocity . . . . . . . . . . . . . . . . . . . 159
Interior Normal Displacement. . . . . . . . . . . . . . . . . 159

The Pressure Acoustics, Boundary Mode Interface 160

Initial Values . . . . . . . . . . . . . . . . . . . . . . . 161
Boundary, Edge, Point, and Pair Nodes for the Pressure Acoustics,
Boundary Mode Interface . . . . . . . . . . . . . . . . . 162

The Pressure Acoustics, Boundary Elements Interface 163

Domain, Boundary, Edge, and Pair Nodes for the Pressure
Acoustics, Boundary Elements Interface . . . . . . . . . . . . 171
Pressure Acoustics . . . . . . . . . . . . . . . . . . . . . 171
Background Pressure Field . . . . . . . . . . . . . . . . . . 172
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 173
Sound Hard Boundary (Wall) . . . . . . . . . . . . . . . . . 173
Normal Acceleration . . . . . . . . . . . . . . . . . . . . 173
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . 174
Normal Displacement . . . . . . . . . . . . . . . . . . . . 174
Sound Soft Boundary . . . . . . . . . . . . . . . . . . . . 175
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 175
Impedance . . . . . . . . . . . . . . . . . . . . . . . . 175
Exclude Boundary . . . . . . . . . . . . . . . . . . . . . 175

CONTENTS |5
 Interior Sound Hard Boundary (Wall) . . . . . . . . . . . . . . 176
Interior Normal Acceleration . . . . . . . . . . . . . . . . . 176
Interior Normal Velocity . . . . . . . . . . . . . . . . . . . 177
Interior Normal Displacement. . . . . . . . . . . . . . . . . 178

The Pressure Acoustics, Time Explicit Interface 180

Domain, Boundary, Edge, and Point Nodes for the Pressure
Acoustics, Time Explicit Interface . . . . . . . . . . . . . . 183
Pressure Acoustics, Time Explicit Model . . . . . . . . . . . . . 183
Background Acoustic Field . . . . . . . . . . . . . . . . . . 186
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 188
Mass Source . . . . . . . . . . . . . . . . . . . . . . . 188
Heat Source . . . . . . . . . . . . . . . . . . . . . . . 188
Volume Force Source . . . . . . . . . . . . . . . . . . . . 189
Sound Hard Boundary (Wall) . . . . . . . . . . . . . . . . . 189
Continuity of Total Fields . . . . . . . . . . . . . . . . . . 190
Sound Soft Boundary . . . . . . . . . . . . . . . . . . . . 190
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 190
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 190
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . 190
Impedance . . . . . . . . . . . . . . . . . . . . . . . . 191
Exterior Field Calculation . . . . . . . . . . . . . . . . . . 192
Interior Sound Hard Boundary (Wall) . . . . . . . . . . . . . . 193
Interior Normal Velocity . . . . . . . . . . . . . . . . . . . 193
Material Discontinuity . . . . . . . . . . . . . . . . . . . . 194
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 194
General Flux/Source . . . . . . . . . . . . . . . . . . . . 195
General Interior Flux . . . . . . . . . . . . . . . . . . . . 195

The Pressure Acoustics, Asymptotic Scattering Interface 196

Domain and Boundary Nodes for the Pressure Acoustics, Asymptotic
Scattering Interface. . . . . . . . . . . . . . . . . . . . 198
Pressure Acoustics . . . . . . . . . . . . . . . . . . . . . 199
Background Pressure Field . . . . . . . . . . . . . . . . . . 199
Scattering Object. . . . . . . . . . . . . . . . . . . . . . 199
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 199
Exterior Field Calculation . . . . . . . . . . . . . . . . . . 200

6 | CONTENTS
The Pressure Acoustics, Kirchhoff-Helmholtz Interface 201
Domain and Boundary Nodes for the Pressure Acoustics, Asymptotic
Scattering Interface. . . . . . . . . . . . . . . . . . . . 202
Pressure Acoustics . . . . . . . . . . . . . . . . . . . . . 203
Radiating Object . . . . . . . . . . . . . . . . . . . . . . 203
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 203
Exterior Field Calculation . . . . . . . . . . . . . . . . . . 204

Modeling with the Pressure Acoustics Branch (FEM-Based

Interfaces) 205
Physics-Controlled Mesh for Pressure Acoustics . . . . . . . . . . 205
Meshing: Resolving the Waves in Space . . . . . . . . . . . . . 207
Lagrange and Serendipity Shape Functions . . . . . . . . . . . . 209
Time Stepping: Resolving the Waves in Time . . . . . . . . . . . 210
Frequency Domain, Modal and AWE . . . . . . . . . . . . . . 212
Solving Large Acoustics Problems Using Iterative Solvers. . . . . . . 213
Perfectly Matched Layers (PMLs) . . . . . . . . . . . . . . . . 219
Postprocessing Variables . . . . . . . . . . . . . . . . . . . 224
Evaluating the Acoustic Field in the Exterior: Near- and Far-Field . . . 230
Dedicated Acoustics Plots for Postprocessing . . . . . . . . . . . 232
About the Material Databases for the Acoustics Module . . . . . . . 238
Specifying Frequencies: Logarithmic and ISO Preferred . . . . . . . 238

Modeling with the Pressure Acoustics Branch (BEM-Based

Interface) 239
When to Use BEM . . . . . . . . . . . . . . . . . . . . . 239
Selections: Infinite Void and Finite Voids . . . . . . . . . . . . . 240
Solvers for BEM Models . . . . . . . . . . . . . . . . . . . 241
Meshing BEM Models . . . . . . . . . . . . . . . . . . . . 242
Postprocessing BEM Results. . . . . . . . . . . . . . . . . . 242

Modeling with the Pressure Acoustics Branch (DG-FEM-Based

Interface) 244
Meshing, Discretization, and Solvers . . . . . . . . . . . . . . 244
Postprocessing: Variables and Quality . . . . . . . . . . . . . . 244
Absorbing Layers . . . . . . . . . . . . . . . . . . . . . . 244
Storing Solution on Selections for Large Models . . . . . . . . . . 245
Assemblies and Pair Conditions . . . . . . . . . . . . . . . . 245

CONTENTS |7
 Modeling with the Aeroacoustic Flow Source: Flow-Induced
Noise 246
The Lighthill Analogy . . . . . . . . . . . . . . . . . . . . 246
Hybrid Computational Aeroacoustics (CAA) . . . . . . . . . . . 247
Setting up a Flow-Induced Noise Problem in COMSOL . . . . . . . 248

Theory Background for the Pressure Acoustics Branch 250

The Governing Equations. . . . . . . . . . . . . . . . . . . 250
Pressure Acoustics, Frequency Domain Equations . . . . . . . . . 254
Pressure Acoustics, Transient Equations . . . . . . . . . . . . . 257
The Nonlinear Westervelt Equation . . . . . . . . . . . . . . 258
Pressure Acoustics, Boundary Mode Equations . . . . . . . . . . 259
Theory for the Plane, Spherical, and Cylindrical Radiation Boundary
Conditions . . . . . . . . . . . . . . . . . . . . . . . 260
Theory for the Exterior Field Calculation: The Helmholtz-Kirchhoff
Integral . . . . . . . . . . . . . . . . . . . . . . . . 263

Theory for the Boundary Impedance Models 267

Impedance Conditions . . . . . . . . . . . . . . . . . . . . 267
RCL Models. . . . . . . . . . . . . . . . . . . . . . . . 268
Physiological Models . . . . . . . . . . . . . . . . . . . . 270
Waveguide End Impedance Models . . . . . . . . . . . . . . . 278
Porous Layer Models . . . . . . . . . . . . . . . . . . . . 279
Characteristic Specific Impedance Models . . . . . . . . . . . . 280

Theory for the Interior Impedance Models 282

Interior Perforated Plate Models . . . . . . . . . . . . . . . . 282

Theory for the Equivalent Fluid Models 288

Introduction to the Equivalent Fluid Models. . . . . . . . . . . . 288
About the Pressure Acoustics Fluid Models . . . . . . . . . . . . 290
About the Poroacoustics Models . . . . . . . . . . . . . . . . 292
About the Narrow Region Acoustics Models . . . . . . . . . . . 306

Theory for the Perfectly Matched Layers in the Time

Domain 312
Introduction to Perfectly Matched Layers . . . . . . . . . . . . 312
Perfectly Matched Layers in the Time Domain . . . . . . . . . . . 313

8 | CONTENTS
 References for the Pressure Acoustics Branch 314

Chapter 3: Elastic Waves Interfaces

The Solid Mechanics (Elastic Waves) Interface 322

The Poroelastic Waves Interface 324

Domain, Boundary, and Pair Nodes for the Poroelastic Waves
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 326
Poroelastic Material . . . . . . . . . . . . . . . . . . . . . 326
Porous, Free . . . . . . . . . . . . . . . . . . . . . . . 333
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 334
Fixed Constraint . . . . . . . . . . . . . . . . . . . . . . 334
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 334
Porous, Pressure . . . . . . . . . . . . . . . . . . . . . . 335
Prescribed Displacement . . . . . . . . . . . . . . . . . . . 335
Prescribed Velocity . . . . . . . . . . . . . . . . . . . . . 336
Prescribed Acceleration . . . . . . . . . . . . . . . . . . . 337
Roller . . . . . . . . . . . . . . . . . . . . . . . . . . 338
Septum Boundary Load . . . . . . . . . . . . . . . . . . . 338
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 339

The Elastic Waves, Time Explicit Interface 340

Domain, Boundary, Edge, Point, and Pair Nodes for the Elastic
Waves, Time Explicit Interface . . . . . . . . . . . . . . . 342
Elastic Waves, Time Explicit Model . . . . . . . . . . . . . . . 343
Damping . . . . . . . . . . . . . . . . . . . . . . . . . 348
Compute Displacement . . . . . . . . . . . . . . . . . . . 349
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 349
Body Load . . . . . . . . . . . . . . . . . . . . . . . . 349
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . 350
Free. . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Fixed . . . . . . . . . . . . . . . . . . . . . . . . . . 350
Prescribed Velocity . . . . . . . . . . . . . . . . . . . . . 350
Roller . . . . . . . . . . . . . . . . . . . . . . . . . . 351
Boundary Load . . . . . . . . . . . . . . . . . . . . . . 351

CONTENTS |9
 Low-Reflecting Boundary . . . . . . . . . . . . . . . . . . . 351
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 352
Antisymmetry . . . . . . . . . . . . . . . . . . . . . . . 352
Material Discontinuity . . . . . . . . . . . . . . . . . . . . 352
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 353
General Flux/Source . . . . . . . . . . . . . . . . . . . . 353
General Interior Flux . . . . . . . . . . . . . . . . . . . . 353

The Piezoelectric Waves, Time Explicit Interface 355

Piezoelectric Material . . . . . . . . . . . . . . . . . . . . 356
Mechanical Damping . . . . . . . . . . . . . . . . . . . . 357

Modeling with the Elastic Waves Branch 358

Meshing and Solving Wave Problems Solved with Solid Mechanics . . . 358
Meshing Poroelastic Waves Models . . . . . . . . . . . . . . . 359
Solving Large Poroelastic Wave Models . . . . . . . . . . . . . 359
Meshing and Solving Elastic Waves, Time Explicit Models. . . . . . . 360
Absorbing Layers in Elastic Waves, Time Explicit . . . . . . . . . . 361
Computing the Displacement in the Elastic Waves, Time Explicit. . . . 362

Theory for the Poroelastic Waves Interfaces 364

Elastic Waves Introduction . . . . . . . . . . . . . . . . . . 364
Poroelastic Waves Theory . . . . . . . . . . . . . . . . . . 365
Boundary Conditions for Poroelastic Waves . . . . . . . . . . . 370
Postprocessing Variables . . . . . . . . . . . . . . . . . . . 373
References for the Poroelastic Waves Interfaces . . . . . . . . . . 374

Theory for the Elastic Waves, Time Explicit Interface 376

Governing Equations . . . . . . . . . . . . . . . . . . . . 376
References for the Elastic Waves, Time Explicit Interface . . . . . . 376

10 | C O N T E N T S
Chapter 4: Acoustic-Structure Interaction Interfaces

The Acoustic-Solid Interaction, Frequency Domain Interface 378

The Acoustic-Solid Interaction, Transient Interface 381

The Acoustic-Piezoelectric Interaction, Frequency Domain

Interface 384

The Acoustic-Piezoelectric Interaction, Transient Interface 387

The Acoustic-Poroelastic Waves Interaction Interface 390

The Acoustic-Solid-Poroelastic Waves Interaction Interface 392

The Acoustic-Solid Interaction, Time Explicit Interface 394

The Acoustic-Shell Interaction, Frequency Domain Interface 397

The Acoustic-Shell Interaction, Transient Interface 400

Modeling with the Acoustic-Structure Interaction Branch 403

Prestressed Acoustic-Structure Interaction . . . . . . . . . . . . 403
Solving Large Acoustic-Structure Interaction Models . . . . . . . . 404
Configuration of Perfectly Matched Layers (PMLs) for
Acoustic-Structure Interaction Models . . . . . . . . . . . . 405

Chapter 5: Aeroacoustics Interfaces

The Linearized Euler, Frequency Domain Interface 409

Domain, Boundary, and Pair Nodes for the Linearized Euler,
Frequency Domain Interface . . . . . . . . . . . . . . . . 412
Linearized Euler Model . . . . . . . . . . . . . . . . . . . 413
Rigid Wall . . . . . . . . . . . . . . . . . . . . . . . . 416
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 417

CONTENTS | 11
 Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . 417
Domain Sources . . . . . . . . . . . . . . . . . . . . . . 417
Background Acoustic Fields . . . . . . . . . . . . . . . . . . 418
Pressure (Isentropic) . . . . . . . . . . . . . . . . . . . . 419
Prescribed Acoustic Fields . . . . . . . . . . . . . . . . . . 419
Acoustic Impedance (Isentropic) . . . . . . . . . . . . . . . . 420
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 421
Impedance and Interior Impedance . . . . . . . . . . . . . . . 421
Moving Wall . . . . . . . . . . . . . . . . . . . . . . . 422
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 423
Asymptotic Far-Field Radiation . . . . . . . . . . . . . . . . 423
Outflow Boundary . . . . . . . . . . . . . . . . . . . . . 424
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 424

The Linearized Euler, Transient Interface 426

Domain, Boundary, and Pair Nodes for the Linearized Euler,
Transient Interface . . . . . . . . . . . . . . . . . . . . 428
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 429
Moving Wall . . . . . . . . . . . . . . . . . . . . . . . 429

The Linearized Euler, Boundary Mode Interface 431

Linearized Euler Model . . . . . . . . . . . . . . . . . . . 432
Rigid Wall . . . . . . . . . . . . . . . . . . . . . . . . 432
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 432

The Linearized Navier–Stokes, Frequency Domain Interface 434

Domain, Boundary, and Pair Nodes for the Linearized Navier–Stokes,
Frequency Domain and Transient Interfaces . . . . . . . . . . 438
Linearized Navier–Stokes Model . . . . . . . . . . . . . . . . 439
Domain Sources . . . . . . . . . . . . . . . . . . . . . . 444
First-Order Material Parameters . . . . . . . . . . . . . . . . 444
Background Acoustic Fields . . . . . . . . . . . . . . . . . . 445
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 445
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . 446
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 446
Pressure (Adiabatic). . . . . . . . . . . . . . . . . . . . . 447
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 448
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 448

12 | C O N T E N T S
Interior Normal Impedance . . . . . . . . . . . . . . . . . . 449
No Slip . . . . . . . . . . . . . . . . . . . . . . . . . 450
Slip . . . . . . . . . . . . . . . . . . . . . . . . . . . 450
Prescribed Velocity . . . . . . . . . . . . . . . . . . . . . 451
Prescribed Pressure. . . . . . . . . . . . . . . . . . . . . 452
No Stress . . . . . . . . . . . . . . . . . . . . . . . . 452
Boundary Stress . . . . . . . . . . . . . . . . . . . . . . 452
Normal Impedance . . . . . . . . . . . . . . . . . . . . . 453
Isothermal . . . . . . . . . . . . . . . . . . . . . . . . 453
Adiabatic . . . . . . . . . . . . . . . . . . . . . . . . . 454
Prescribed Temperature . . . . . . . . . . . . . . . . . . . 454
Heat Flux. . . . . . . . . . . . . . . . . . . . . . . . . 454
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 455

The Linearized Navier–Stokes, Transient Interface 456

The Linearized Navier–Stokes, Boundary Mode Interface 459

Linearized Navier-Stokes Model . . . . . . . . . . . . . . . . 460
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 460
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 460

The Linearized Potential Flow, Frequency Domain Interface 462

Domain, Boundary, Edge, Point, and Pair Nodes for the Linearized
Potential Flow, Frequency Domain Interface . . . . . . . . . . 464
Linearized Potential Flow Model . . . . . . . . . . . . . . . . 465
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 466
Sound Hard Boundary (Wall) . . . . . . . . . . . . . . . . . 466
Velocity Potential. . . . . . . . . . . . . . . . . . . . . . 467
Normal Mass Flow . . . . . . . . . . . . . . . . . . . . . 467
Plane Wave Radiation . . . . . . . . . . . . . . . . . . . . 467
Incident Velocity Potential . . . . . . . . . . . . . . . . . . 468
Sound Soft Boundary . . . . . . . . . . . . . . . . . . . . 469
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 469
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . 470
Impedance, Interior Impedance, and Pair Impedance . . . . . . . . 470
Vortex Sheet . . . . . . . . . . . . . . . . . . . . . . . 470
Interior Sound Hard Boundary (Wall) . . . . . . . . . . . . . . 471
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 472

CONTENTS | 13
 Mass Flow Edge Source . . . . . . . . . . . . . . . . . . . 472
Mass Flow Point Source . . . . . . . . . . . . . . . . . . . 472
Mass Flow Circular Source . . . . . . . . . . . . . . . . . . 473
Mass Flow Line Source on Axis . . . . . . . . . . . . . . . . 473
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . 473

The Linearized Potential Flow, Transient Interface 474

Domain, Boundary, Edge, Point, and Pair Nodes for the Linearized
Potential Flow, Transient Interface . . . . . . . . . . . . . . 475

The Linearized Potential Flow, Boundary Mode Interface 477

Boundary, Edge, Point, and Pair Nodes for the Linearized Potential
Flow, Boundary Mode Interface . . . . . . . . . . . . . . . 478

The Compressible Potential Flow Interface 480

Domain, Boundary, and Pair Nodes for the Compressible Potential
Flow Interface . . . . . . . . . . . . . . . . . . . . . 481
Compressible Potential Flow Model. . . . . . . . . . . . . . . 482
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 483
Slip Velocity. . . . . . . . . . . . . . . . . . . . . . . . 483
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 484
Normal Flow . . . . . . . . . . . . . . . . . . . . . . . 484
Mass Flow . . . . . . . . . . . . . . . . . . . . . . . . 484
Mean Flow Velocity Potential . . . . . . . . . . . . . . . . . 484
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 485
Interior Wall (Slip Velocity) . . . . . . . . . . . . . . . . . . 485

Modeling with the Aeroacoustics Branch 486

Selecting an Aeroacoustics Interface . . . . . . . . . . . . . . 486
Meshing . . . . . . . . . . . . . . . . . . . . . . . . . 487
Stabilization . . . . . . . . . . . . . . . . . . . . . . . . 488
Solver Suggestions for Large Aeroacoustic Models . . . . . . . . . 489
Absorbing Layers for the Linearized Euler, Transient Interface . . . . 490
Lagrange and Serendipity Shape Functions . . . . . . . . . . . . 491
Time Stepping in Transient Models . . . . . . . . . . . . . . . 492
Mapping Between Fluid Flow and Acoustics Mesh . . . . . . . . . 492
Coupling to Turbulent Flows (Eddy Viscosity) . . . . . . . . . . . 495
Eigenfrequency Studies . . . . . . . . . . . . . . . . . . . 495

14 | C O N T E N T S
 Suppressing Constraints on Lower Dimensions . . . . . . . . . . 496
Postprocessing Variables . . . . . . . . . . . . . . . . . . . 497

Theory Background for the Aeroacoustics Branch 500

General Governing Equations . . . . . . . . . . . . . . . . . 501
Linearized Navier–Stokes . . . . . . . . . . . . . . . . . . 504
Linearized Euler . . . . . . . . . . . . . . . . . . . . . . 505
Scattered Field Formulation for LE and LNS . . . . . . . . . . . 508
Linearized Potential Flow . . . . . . . . . . . . . . . . . . . 509
Compressible Potential Flow . . . . . . . . . . . . . . . . . 512

References for the Aeroacoustics Branch Interfaces 515

Chapter 6: Thermoviscous Acoustics Interfaces

The Thermoviscous Acoustics, Frequency Domain Interface 518

Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,
Frequency Domain Interface . . . . . . . . . . . . . . . . 524
Thermoviscous Acoustics Model . . . . . . . . . . . . . . . . 526
Background Acoustic Fields . . . . . . . . . . . . . . . . . . 530
Heat Source . . . . . . . . . . . . . . . . . . . . . . . 530
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 531
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . 531
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 531
Pressure (Adiabatic). . . . . . . . . . . . . . . . . . . . . 532
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 533
Port . . . . . . . . . . . . . . . . . . . . . . . . . . . 533
Periodic Condition . . . . . . . . . . . . . . . . . . . . . 539
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 541
Interior Normal Impedance . . . . . . . . . . . . . . . . . . 542
Interior Velocity . . . . . . . . . . . . . . . . . . . . . . 543
Interior Temperature Variation . . . . . . . . . . . . . . . . 544
No Slip . . . . . . . . . . . . . . . . . . . . . . . . . 544
Slip . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
Velocity . . . . . . . . . . . . . . . . . . . . . . . . . 546
No Stress . . . . . . . . . . . . . . . . . . . . . . . . 547

CONTENTS | 15
 Boundary Stress . . . . . . . . . . . . . . . . . . . . . . 547
Normal Impedance . . . . . . . . . . . . . . . . . . . . . 547
Isothermal . . . . . . . . . . . . . . . . . . . . . . . . 548
Adiabatic . . . . . . . . . . . . . . . . . . . . . . . . . 548
Temperature Variation . . . . . . . . . . . . . . . . . . . 549
Heat Flux. . . . . . . . . . . . . . . . . . . . . . . . . 549

The Thermoviscous Acoustics, Transient Interface 550

Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,
Transient Interface . . . . . . . . . . . . . . . . . . . . 553
Thermoviscous Acoustics Model . . . . . . . . . . . . . . . . 555
Nonlinear Thermoviscous Acoustics Contributions . . . . . . . . . 556
Background Acoustic Fields . . . . . . . . . . . . . . . . . . 557

The Thermoviscous Acoustics, Boundary Mode Interface 559

Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,
Boundary Mode Interface . . . . . . . . . . . . . . . . . 562
Thermoviscous Acoustics Model . . . . . . . . . . . . . . . . 563

The Acoustic-Thermoviscous Acoustic Interaction, Frequency

Domain Interface 565

The Thermoviscous Acoustic-Solid Interaction, Frequency

Domain Interface 568

The Thermoviscous Acoustic-Shell Interaction, Frequency

Domain Interface 570

Modeling with the Thermoviscous Acoustics Branch 573

Meshing the Boundary Layer . . . . . . . . . . . . . . . . . 573
Solver Suggestions for Large Thermoviscous Acoustics Models . . . . 575
Lagrange and Serendipity Shape Functions . . . . . . . . . . . . 577
Transient Solver Settings . . . . . . . . . . . . . . . . . . . 578
Postprocessing Variables . . . . . . . . . . . . . . . . . . . 578
Suppressing Constraints on Lower Dimensions . . . . . . . . . . 582

Theory Background for the Thermoviscous Acoustics Branch 584

The Viscous and Thermal Boundary Layers . . . . . . . . . . . . 585

16 | C O N T E N T S
 General Linearized Compressible Flow Equations . . . . . . . . . 586
Acoustic Perturbation and Linearization . . . . . . . . . . . . . 587
Scattered Field Formulation and Background Acoustic Fields . . . . . 592
Formulation for Eigenfrequency Studies . . . . . . . . . . . . . 593
Formulation for Mode Analysis in 2D and 1D Axisymmetry. . . . . . 595
Formulation for the Boundary Mode Interface . . . . . . . . . . . 596
The Nonlinear Thermoviscous Acoustics Equations. . . . . . . . . 597
References for the Thermoviscous Acoustics, Frequency Domain
Interface. . . . . . . . . . . . . . . . . . . . . . . . 598

Chapter 7: Ultrasound Interfaces

The Convected Wave Equation, Time Explicit Interface 602

Domain, Boundary, Edge, Point, and Pair Nodes for the Convected
Wave Equation Interface . . . . . . . . . . . . . . . . . 604
Convected Wave Equation Model . . . . . . . . . . . . . . . 605
Domain Sources . . . . . . . . . . . . . . . . . . . . . . 608
Sound Hard Wall . . . . . . . . . . . . . . . . . . . . . . 609
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 609
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . 609
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 610
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 610
Acoustic Impedance. . . . . . . . . . . . . . . . . . . . . 610
Interior Wall . . . . . . . . . . . . . . . . . . . . . . . 611
Interior Normal Velocity . . . . . . . . . . . . . . . . . . . 611
General Flux/Source . . . . . . . . . . . . . . . . . . . . 611
General Interior Flux . . . . . . . . . . . . . . . . . . . . 612

The Nonlinear Pressure Acoustics, Time Explicit Interface 614

Domain, Boundary, Edge, and Point Nodes for the Nonlinear
Pressure Acoustics, Time Explicit Interface. . . . . . . . . . . 617
Nonlinear Pressure Acoustics, Time Explicit Model . . . . . . . . . 617
Compute Minimum and Maximum Pressure. . . . . . . . . . . . 618
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 619
Mass Source . . . . . . . . . . . . . . . . . . . . . . . 619
Heat Source . . . . . . . . . . . . . . . . . . . . . . . 619

CONTENTS | 17
 Volume Force Source . . . . . . . . . . . . . . . . . . . . 620
Sound Hard Boundary (Wall) . . . . . . . . . . . . . . . . . 620
Sound Soft Boundary . . . . . . . . . . . . . . . . . . . . 621
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 621
Symmetry . . . . . . . . . . . . . . . . . . . . . . . . 621
Normal Velocity . . . . . . . . . . . . . . . . . . . . . . 621
Impedance . . . . . . . . . . . . . . . . . . . . . . . . 622
Interior Sound Hard Boundary (Wall) . . . . . . . . . . . . . . 622
Interior Normal Velocity . . . . . . . . . . . . . . . . . . . 623
Material Discontinuity . . . . . . . . . . . . . . . . . . . . 623
Continuity . . . . . . . . . . . . . . . . . . . . . . . . 624
General Flux/Source . . . . . . . . . . . . . . . . . . . . 624
General Interior Flux . . . . . . . . . . . . . . . . . . . . 625

Modeling with the Convected Wave Equation Interface 626

Meshing, Discretization, and Solvers . . . . . . . . . . . . . . 626
Postprocessing: Variables and Quality . . . . . . . . . . . . . . 628
Absorbing Layers . . . . . . . . . . . . . . . . . . . . . . 629
Stabilizing Physical Instabilities (Filtering) . . . . . . . . . . . . . 630
Storing Solution on Selections for Large Models . . . . . . . . . . 630
Assemblies and Pair Conditions . . . . . . . . . . . . . . . . 630

Modeling with the Nonlinear Pressure Acoustics, Time

Explicit Interface 632
Solving Highly Nonlinear Problems . . . . . . . . . . . . . . . 632
Adaptive Mesh Refinement . . . . . . . . . . . . . . . . . . 633

Theory for the Convected Wave Equation Interface 634

Governing Equations of the Convected Wave Equation . . . . . . . 634
Boundary Conditions . . . . . . . . . . . . . . . . . . . . 636
The Lax–Friedrichs Flux . . . . . . . . . . . . . . . . . . . 637

Theory for the Nonlinear Pressure Acoustics, Time Explicit

Interface 638
Governing Equations for Nonlinear Pressure Acoustics, Time Explicit . . 638

18 | C O N T E N T S
 References for the Ultrasound Interface 639

Chapter 8: Geometrical Acoustics Interfaces

The Ray Acoustics Interface 642

Domain, Boundary, and Global Nodes for the Ray Acoustics
Interface. . . . . . . . . . . . . . . . . . . . . . . . 649
Medium Properties . . . . . . . . . . . . . . . . . . . . . 650
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 653
Axial Symmetry . . . . . . . . . . . . . . . . . . . . . . 659
Accumulator (Boundary) . . . . . . . . . . . . . . . . . . . 659
Material Discontinuity . . . . . . . . . . . . . . . . . . . . 661
Ray Properties . . . . . . . . . . . . . . . . . . . . . . . 663
Release . . . . . . . . . . . . . . . . . . . . . . . . . 663
Sound Pressure Level Calculation . . . . . . . . . . . . . . . 670
Accumulator (Domain) . . . . . . . . . . . . . . . . . . . 671
Nonlocal Accumulator. . . . . . . . . . . . . . . . . . . . 672
Release from Boundary . . . . . . . . . . . . . . . . . . . 673
Release from Symmetry Axis . . . . . . . . . . . . . . . . . 677
Background Velocity . . . . . . . . . . . . . . . . . . . . 678
Auxiliary Dependent Variable . . . . . . . . . . . . . . . . . 679
Release from Edge . . . . . . . . . . . . . . . . . . . . . 679
Release from Point . . . . . . . . . . . . . . . . . . . . . 680
Release from Point on Axis . . . . . . . . . . . . . . . . . . 680
Release from Grid . . . . . . . . . . . . . . . . . . . . . 680
Release from Grid on Axis . . . . . . . . . . . . . . . . . . 684
Release from Data File. . . . . . . . . . . . . . . . . . . . 684
Release from Exterior Field Calculation . . . . . . . . . . . . . 686
Source with Directivity . . . . . . . . . . . . . . . . . . . 689
Ray Continuity. . . . . . . . . . . . . . . . . . . . . . . 692
Ray Termination . . . . . . . . . . . . . . . . . . . . . . 693
Ray Detector . . . . . . . . . . . . . . . . . . . . . . . 694

Modeling with the Ray Acoustics Interface 696

Mixed Diffuse and Specular Wall Conditions . . . . . . . . . . . 696
Assigning Directivity to a Source . . . . . . . . . . . . . . . . 697

CONTENTS | 19
 Impulse Response Plot and Receiver Dataset . . . . . . . . . . . 697
Stopping Rays for a Given Condition . . . . . . . . . . . . . . 705
Mesh Guidelines . . . . . . . . . . . . . . . . . . . . . . 705
Nonlocal Couplings . . . . . . . . . . . . . . . . . . . . . 708
Using Ray Detectors . . . . . . . . . . . . . . . . . . . . 710
Other Results Plots, Datasets, and Derived Values . . . . . . . . . 710

Theory for the Ray Acoustics Interface 712

Introduction to Ray Acoustics . . . . . . . . . . . . . . . . . 712
Initial Conditions: Position . . . . . . . . . . . . . . . . . . 713
Initial Conditions: Direction. . . . . . . . . . . . . . . . . . 714
Material Discontinuity Theory . . . . . . . . . . . . . . . . . 716
Intensity and Wavefront Curvature . . . . . . . . . . . . . . . 717
Intensity and Phase Reinitialization . . . . . . . . . . . . . . . 722
Wavefront Curvature Calculation in Graded Media . . . . . . . . . 725
Attenuation Within Domains . . . . . . . . . . . . . . . . . 730
Ray Termination Theory . . . . . . . . . . . . . . . . . . . 731
Accumulator Theory: Domains . . . . . . . . . . . . . . . . 733
Accumulator Theory: Boundaries . . . . . . . . . . . . . . . 735
Sound Pressure Level Calculation Theory . . . . . . . . . . . . 736
Assigning Spatial Directivity to a Source . . . . . . . . . . . . . 737
References for the Ray Acoustics Interface . . . . . . . . . . . . 738

The Acoustic Diffusion Equation Interface 739

Domain, Boundary, and Global Nodes for the Acoustic Diffusion
Equation Interface . . . . . . . . . . . . . . . . . . . . 742
Acoustic Diffusion Model. . . . . . . . . . . . . . . . . . . 743
Room . . . . . . . . . . . . . . . . . . . . . . . . . . 743
Wall . . . . . . . . . . . . . . . . . . . . . . . . . . 744
Inward Energy Flux . . . . . . . . . . . . . . . . . . . . . 745
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 746
Fitted Domain . . . . . . . . . . . . . . . . . . . . . . . 746
Domain Source . . . . . . . . . . . . . . . . . . . . . . 746
Room Coupling . . . . . . . . . . . . . . . . . . . . . . 747
Mapped Room Coupling . . . . . . . . . . . . . . . . . . . 747
Destination Selection . . . . . . . . . . . . . . . . . . . . 748
Point Source . . . . . . . . . . . . . . . . . . . . . . . 748

20 | C O N T E N T S
 Modeling with the Acoustic Diffusion Equation Interface 749
The Eigenvalue Study Type . . . . . . . . . . . . . . . . . . 749
Combined Stationary and Time Dependent Study . . . . . . . . . 749

Theory for the Acoustic Diffusion Equation Interface 750

Statistical Model of Reverberation Time . . . . . . . . . . . . . 750
The Acoustic Diffusion Equation . . . . . . . . . . . . . . . . 751
References for the Acoustic Diffusion Equation Interface. . . . . . . 757

Chapter 9: Pipe Acoustics Interfaces

The Pipe Acoustics Frequency Domain and Transient

Interfaces 762
The Pipe Acoustics, Frequency Domain Interface. . . . . . . . . . 762
The Pipe Acoustics, Transient Interface . . . . . . . . . . . . . 764
Edge, Boundary, Point, and Pair Nodes for the Pipe Acoustics
Interfaces . . . . . . . . . . . . . . . . . . . . . . . 765
Initial Values . . . . . . . . . . . . . . . . . . . . . . . 766
Fluid Properties . . . . . . . . . . . . . . . . . . . . . . 766
Pipe Properties . . . . . . . . . . . . . . . . . . . . . . 767
Volume Force . . . . . . . . . . . . . . . . . . . . . . . 768
Closed. . . . . . . . . . . . . . . . . . . . . . . . . . 768
Pressure . . . . . . . . . . . . . . . . . . . . . . . . . 769
Velocity . . . . . . . . . . . . . . . . . . . . . . . . . 769
End Impedance . . . . . . . . . . . . . . . . . . . . . . 770

Theory for the Pipe Acoustics Interfaces 772

Governing Equations . . . . . . . . . . . . . . . . . . . . 772
Theory for the Pipe Acoustics Boundary Conditions . . . . . . . . 776
Solving Transient Problems . . . . . . . . . . . . . . . . . . 779
Cutoff Frequency. . . . . . . . . . . . . . . . . . . . . . 780
Flow Profile Correction Factor . . . . . . . . . . . . . . . . 780
References for the Pipe Acoustics Interfaces . . . . . . . . . . . 781

CONTENTS | 21
 Chapter 10: Multiphysics Couplings

Coupling Features 784

Acoustic-Structure Boundary . . . . . . . . . . . . . . . . . 784
Pair Acoustic-Structure Boundary . . . . . . . . . . . . . . . 786
Thermoviscous Acoustic-Structure Boundary . . . . . . . . . . . 787
Pair Thermoviscous Acoustic-Structure Boundary . . . . . . . . . 788
Aeroacoustic-Structure Boundary . . . . . . . . . . . . . . . 789
Acoustic-Thermoviscous Acoustic Boundary . . . . . . . . . . . 790
Acoustic-Porous Boundary . . . . . . . . . . . . . . . . . . 792
Porous-Structure Boundary . . . . . . . . . . . . . . . . . . 793
Background Potential Flow Coupling . . . . . . . . . . . . . . 793
Background Fluid Flow Coupling . . . . . . . . . . . . . . . . 794
Aeroacoustic Flow Source Coupling . . . . . . . . . . . . . . 795
Acoustic FEM-BEM Boundary . . . . . . . . . . . . . . . . . 797
Acoustic-Pipe Acoustic Connection . . . . . . . . . . . . . . . 798
Acoustic-Structure Boundary, Time Explicit. . . . . . . . . . . . 799
Pair Acoustic-Structure Boundary, Time Explicit . . . . . . . . . . 799
Piezoelectric Effect, Time Explicit . . . . . . . . . . . . . . . 800
Piezoelectric Effect . . . . . . . . . . . . . . . . . . . . . 801
Magnetomechanical Forces . . . . . . . . . . . . . . . . . . 803

Predefined Multiphysics Interfaces 804

Modeling with Multiphysics Couplings 806

Use Selections . . . . . . . . . . . . . . . . . . . . . . . 806
The Override Behavior . . . . . . . . . . . . . . . . . . . 807
The Solvers . . . . . . . . . . . . . . . . . . . . . . . . 807
Perfectly Matched Layers (PMLs) . . . . . . . . . . . . . . . . 808

Chapter 11: Structural Mechanics with the Acoustics

Module
Vibroacoustic Applications . . . . . . . . . . . . . . . . . . 812
The Solid Mechanics Interface . . . . . . . . . . . . . . . . . 812
The Piezoelectricity Interface . . . . . . . . . . . . . . . . . 812

22 | C O N T E N T S
 Acoustic-Structure Multiphysics Interaction . . . . . . . . . . . . 813

Chapter 12: Study Types

Acoustics Module Study Types 816

Stationary Study . . . . . . . . . . . . . . . . . . . . . . 817
Frequency Domain Study . . . . . . . . . . . . . . . . . . . 817
Eigenfrequency Study . . . . . . . . . . . . . . . . . . . . 818
Mode Analysis Study . . . . . . . . . . . . . . . . . . . . 820
Boundary Mode Analysis . . . . . . . . . . . . . . . . . . . 821
Time Dependent Study . . . . . . . . . . . . . . . . . . . 821
Frequency Domain, Modal and Time-Dependent, Modal Studies . . . . 822
Ray Tracing . . . . . . . . . . . . . . . . . . . . . . . . 822
Modal Reduced Order Model . . . . . . . . . . . . . . . . . 823
Additional Analysis Capabilities . . . . . . . . . . . . . . . . 823
Mapping . . . . . . . . . . . . . . . . . . . . . . . . . 823
Transient Mapping . . . . . . . . . . . . . . . . . . . . . 824

Chapter 13: Acoustic Properties of Fluids

Material Properties 828

The Bulk Viscosity 832

The Value of the Bulk Viscosity . . . . . . . . . . . . . . . . 832

Attenuation and Loss Models 835

Loss Mechanisms . . . . . . . . . . . . . . . . . . . . . . 835
The Atmosphere and Ocean Attenuation Models . . . . . . . . . 836
Boundary Layer vs. Bulk Losses . . . . . . . . . . . . . . . . 837

CONTENTS | 23
 References for The Acoustic Properties of Fluids 839

Chapter 14: Glossary

Glossary of Terms 842

24 | C O N T E N T S
 1

Introduction

The Acoustics Module is an optional package that extends the COMSOL

Multiphysics® environment with customized interfaces and functionality
optimized for the analysis of acoustics and vibration problems.

This module solves problems in the general areas of acoustics, acoustic-structure

interaction, aeroacoustics, thermoviscous acoustics, linear ultrasound, pressure and
elastic waves in porous materials, vibrations, and geometrical acoustics. The physics
interfaces included are fully multiphysics enabled, making it possible to couple
them to any other physics interface in COMSOL Multiphysics. Explicit
demonstrations of these capabilities are supplied with the product in a library (the
Acoustics Module Application Library) of ready-to-run models and applications
that make it quicker and easier to get introduced to discipline-specific problems.
One example being a model of a loudspeaker involving both electromechanical and
acoustic-structural couplings.

This chapter is an introduction to the capabilities of the Acoustics Module and gives
a short introduction to the fundamentals of acoustics. A summary of the physics
interfaces and where you can find documentation and model examples is also
included. The last section is a brief overview with links to each chapter in this guide.

25
 In this chapter:

• Acoustics Module Capabilities

• Fundamental of Acoustics
• Acoustics Module Physics Interface Guide
• Overview of the User’s Guide

26 | CHAPTER 1: INTRODUCTION
Acoustics Module Capabilities
In this section:

• What Can the Acoustics Module Do?

• What Are the Application Areas?
• Which Problems Can You Solve?

What Can the Acoustics Module Do?

The Acoustics Module is a collection of physics interfaces for COMSOL Multiphysics
adapted to a broad category of acoustics simulations in fluids and solids. This module
is useful even if you are not familiar with computational techniques. It can serve equally
well as an excellent tool for educational purposes.

The Acoustics Module also includes many specialized formulations and material
models that can be used for dedicated application areas like thermoviscous acoustics
used in miniature transducers and mobile devices or Biot’s equations for modeling
poroelastic waves. It also includes many predefined couplings between physics, called
Multiphysics couplings, to model, for example, vibroacoustic problems.

The module supports time-harmonic (frequency domain), eigenfrequency, modal, and

transient studies for all fluids (depending on the acoustic equations solved) as well as
static, transient, eigenfrequency, modal, and frequency-response for the analyses of
wave propagation in structures.

The multiphysics environment is further extended as the module combines several

dedicated numerical methods, including the finite element method (FEM), the
boundary element method (BEM), ray tracing, and the discontinuous Galerkin finite
elements method (dG-FEM).

The available physics interfaces include the following functionality:

• Pressure acoustics: model the propagation of sound waves (pressure waves) in the
frequency domain solving the Helmholtz equation or in the time domain solving
the scalar wave equation. Pressure acoustics comes in different flavors depending on
the numerical formulation used. This includes finite element (FEM) based interfaces
for frequency and transient models, a boundary element (BEM) based interface only
used in the frequency domain, and a discontinuous Galerkin (dG-FEM) formulation
based interface used for transient simulations. The Acoustics Module has built-in

ACOUSTICS MODULE CAPABILITIES | 27

 couplings between BEM and FEM that allows for modeling hybrid FEM-BEM
problems.
• Acoustic-structure interaction: combine pressure waves in the fluid with elastic
waves in the solid to model vibroacoustic problems. The physics interfaces provide
predefined multiphysics couplings at the fluid-solid interface.
• Boundary mode acoustics: find propagating and evanescent modes in ducts and
waveguides. The analysis exists for classical pressure acoustic, elastic waves,
thermoviscous acoustics, and convected acoustic problems.
• Thermoviscous acoustics: model the detailed propagation of sound in geometries
with small length scales. This is acoustics including thermal and viscous losses
explicitly. Also known as microacoustics, viscothermal acoustics, or linearized
compressible Navier-Stokes. In the time domain nonlinear effects can be included.
• Aeroacoustics: model the influence a background mean flow has on the propagation
of sound waves in the flow, so-called, convected acoustics or flow borne noise/
sound. Interfaces exist to solve the linearized potential flow, the linearized Euler
equations, and the linearized Navier-Stokes equations in both time and frequency
domain.
• Flow-induced noise: model the noise generated by a turbulent flow using the
Lighthill analogy to set up an aeroacoustic flow source in pressure acoustics.
• Compressible potential flow: determine the flow of a compressible, irrotational, and
inviscid fluid.
• Solid mechanics and elastic waves: solve structural mechanics problems and the
propagation of elastic waves in solids.
• Piezoelectricity: model the behavior of piezoelectric materials in a multiphysics
environment solving for the electric field and the coupling to the solid structure.
• Poroelastic waves: in porous materials model the coupled propagation of elastic
waves in the solid porous matrix and the pressure waves in the saturation fluid.
Biot’s equations are solved here. Includes options to include both thermal and
viscous losses.
• Ultrasound: in ultrasound problems transient propagation is important and it is also
important to be able to solve models with many wavelengths. These interfaces are
based in the discontinuous Galerkin or dG-FEM formulation.
• Nonlinear acoustics: model nonlinear acoustic phenomena like the cumulative
nonlinear effects by solving the Westervelt equation. Local nonlinear effects such as
vortex shedding can be included using the nonlinear thermoviscous acoustic
functionality.

28 | CHAPTER 1: INTRODUCTION
• Acoustic diffusion equation: solve a diffusion equation for the acoustic energy
density distribution for systems of coupled rooms in room acoustic applications.
• Ray acoustics: compute trajectories and intensity of acoustic rays in room acoustic
as well as underwater acoustic applications. Determine the impulse response and
room acoustic metrics with dedicated features in postprocessing.
• Pipe acoustics: use this physics interface to model the propagation of sound waves
in pipe systems including the elastic properties of the pipe. The equations are
formulated in 1D for fast computation and can include a stationary background
flow.

All the physics interfaces include a large number of boundary conditions. For the
pressure acoustics applications, you can choose to analyze the scattered wave in
addition to the total wave. Impedance conditions can be used to mimic a specific
acoustic behavior at a boundary, for example, the acoustic properties of the human ear
or a mechanical system approximated by a simple RCL circuit. Perfectly matched layers
(PMLs) and absorbing layers provide accurate simulations of open pipes and other
models with unbounded domains. The modeling domain includes support for several
types of damping and losses that occur in porous materials (poroacoustics) or that are
due to viscous and thermal losses (narrow region acoustics). For results evaluation of
pressure acoustics models, you can compute the exterior acoustic field (phase and
magnitude) and plot it in predefined radiation pattern plots.

Figure 1-1: An application example is the modeling of mufflers. Here a plot from the
Absorptive Muffler model from the COMSOL Multiphysics Applications Libraries.

ACOUSTICS MODULE CAPABILITIES | 29

 What Are the Application Areas?
The Acoustics Module can be used in all areas of engineering and physics to model the
propagation of sound waves in fluids and elastic waves in solids. The module also
includes several multiphysics interfaces because it is common for many application
areas involving sound to also have interaction between fluid and solid structures, have
electric fields in piezoelectric materials, have heat generation, or require modeling of
electro-acoustic transducers.

Typical application areas for the Acoustics Module include:

• Automotive applications such as mufflers, particulate filters, and car interiors.

• Sound scattering, absorption, and sound emission problems.
• Civil engineering applications such as characterization of sound insulation and
sound scatterers. Vibration control and sound transmission problems. Pipe acoustics
for HVAC type of systems.

• Modeling of loudspeakers, microphones, and other transducers. Transducers are

devices for transformation of one form of energy to another (electrical, mechanical,
or acoustical). This type of problem is common in acoustics and is a true
multiphysics problem involving electric, structural, and acoustic interfaces.

• Mobile applications such as feedback analysis, optimized transducer placement, and

directivity assessment.
• Convected acoustics for jet engine noise, muffler systems with nonisothermal flow,
and flowmeters.
• Ultrasound piezoelectric transducers.
• Musical instruments.
• Bioacoustic applications with ultrasound and more.
• Underwater acoustics and sonar applications.

• Pressure and elastic waves in geophysics.

• Room acoustics using the ray tracing method or an acoustic diffusion equation
approach.
• Advanced multiphysics applications such as photoacoustics, optoacoustics,
thermoacoustic cooling, acoustofluidics, acoustic streaming and radiation, and
combustion instabilities.

Using the full multiphysics couplings within the COMSOL Multiphysics environment,
you can couple the acoustic waves to, for example, an electromagnetic analysis or a

30 | CHAPTER 1: INTRODUCTION
structural analysis for acoustic-structure interaction. The module smoothly integrates
with all of the COMSOL Multiphysics functionality.

Figure 1-2: Modeling a transducer is a true multiphysics application, comprising

thermoviscous acoustics, electrostatics, and a membrane. Here the displacement of the
microphone diaphragm from the Brüel & Kjær 4134 Condenser Microphone model from
the Acoustics Module Applications Library.

Which Problems Can You Solve?

The Acoustics Module interfaces handle acoustics in fluids (both quiescent and
moving background flows) and solids. The physics interfaces for acoustics in fluids
support transient, eigenfrequency, frequency domain, mode analysis, and boundary
mode analysis in pressure acoustics and linearized potential flow. Thermoacoustic
problems, that involve thermal and viscous losses, have support for eigenfrequency and
frequency domain analysis. The study of elastic and poroelastic waves in solids also has
support for eigenfrequency and frequency domain analysis. The physics interfaces for
solids support static, transient, eigenfrequency, and frequency response analysis.
Further, by using the predefined couplings between fluid and solid interfaces, you can
solve problems involving acoustic-structure interaction including the coupling to
piezoelectric materials.

ACOUSTICS MODULE CAPABILITIES | 31

 All categories are available as 2D, 2D axisymmetric, and 3D models, with the following
differences.

• The Acoustic-Shell Interaction interfaces are supported in 2D axisymmetric and 3D,

and also require the addition of the Structural Mechanics Module.
• The Pipe Acoustics interfaces exist on edges in 2D and 3D.
• In 2D, the module has in-plane physics interfaces for problems with a planar
symmetry as well as axisymmetric physics interfaces for problems with a cylindrical
symmetry.
• Use the fluid acoustics interfaces with 1D and 1D axisymmetric geometries.

When using the axisymmetric models, the horizontal axis represents the r
direction and the vertical axis the z direction. The geometry is in the right
half plane; that is, the geometry must be created and is valid only for
positive r.

32 | CHAPTER 1: INTRODUCTION
Fundamental of Acoustics
This section includes a brief introduction to acoustics and provides a short
introduction to the mathematical formulation of the governing equations. It also
introduces some important concepts like damping and the use of artificial boundaries.

In this section:

• Acoustics Explained
• Mathematical Models for Acoustic Analysis
• Damping
• Artificial Boundaries

Acoustics Explained
Acoustics is the physics of sound. Sound is the sensation, as detected by the ear, of very
small rapid changes in the air pressure above and below a static value. This static value
is the atmospheric pressure (about 100,000 pascals), which varies slowly. Associated
with a sound pressure wave is a flow of energy — the intensity. Physically, sound in air
is a longitudinal wave where the wave motion is in the direction of the movement of
energy. The wave crests are the pressure maxima, while the troughs represent the
pressure minima.

Sound results when the air is disturbed by some source. An example is a vibrating
object, such as a speaker cone in a sound system. It is possible to see the movement of
a bass speaker cone when it generates sound at a very low frequency. As the cone moves
forward it compresses the air in front of it, causing an increase in air pressure. Then it
moves back past its resting position and causes a reduction in air pressure. This process
continues, radiating a wave of alternating high and low pressure propagating at the
speed of sound.

The propagation of sound in solids happens through small-amplitude elastic

oscillations of its shape. These elastic waves are transmitted to surrounding fluids as
ordinary sound waves. The elastic sound waves in the solid are the counterpart to the
pressure waves or compressible waves propagating in the fluid.

FUNDAMENTAL OF ACOUSTICS | 33
 Mathematical Models for Acoustic Analysis
Standard acoustic problems involve solving for the small acoustic pressure variations
(often denoted p’ or p1) on top of the stationary/quiescent background pressure
(often denoted p0 or pA). Mathematically this represents a linearization (small
parameter expansion) around the stationary quiescent values.

The governing equations, for a compressible lossless (no thermal conduction and no
viscosity) fluid flow problem, are the momentum equation (Euler’s equation) and the
continuity equation. These are given by:

∂u 1
------ + ( u ⋅ ∇ )u = – --- ∇p
∂t ρ
∂ρ
------ + ∇ ⋅ ( ρu ) = 0
∂t

where ρ is the total density, p is the total pressure, and u is the total velocity field. In
classical pressure acoustics all thermodynamic processes are assumed reversible and
adiabatic, known as an isentropic process. The small parameter expansion is performed
on a stationary fluid of density ρ0 (SI unit: kg/m3) and at pressure p0 (SI unit: Pa)
such that:

p = p0 + p1 p1 « p0
ρ = ρ0 + ρ1 with ρ1 « ρ0
u = 0 + u1 u1 « c

where the subscript 1 represent the small acoustic variations (sometimes these are
denoted with a prime instead). Inserting these into the governing equations and only
retaining terms linear in the small perturbation variables yields

∂u 1
--------1- = – ------ ∇p 1
∂t ρ0
∂ρ 1
--------- + ρ 0 ( ∇ ⋅ u 1 ) = 0
∂t

One of the dependent variables, the density, is removed by expressing it in terms of the
pressure using the density differential (linearization)

∂ρ 0 1
ρ 1 = --------- p 1 = ----2- p 1
∂p s cs

34 | CHAPTER 1: INTRODUCTION
where cs is recognized as the (isentropic) speed of sound (SI unit: m/s) at constant
entropy s. It should be noted that this equation is valid for constant valued (not space
dependent) background density ρ0 and background pressure p0. The subscripts s is
dropped in the following. From the above expression it also follows that another
requirement for linear acoustics (the perturbation approximation) to be valid is that

2
p1 « ρ0 c

Finally, rearranging the equations (divergence of momentum equation inserted into

the continuity equation) and dropping the subscript 1 yields the wave equation for
sound waves in a lossless medium

2
∂ p-
1 --------
---- – ∇ ⋅ ( ∇p ) = 0 (1-1)
2 2
c ∂t

The speed of sound is related to the compressibility of the fluid where the waves are
propagating. The combination ρ c2 is called the bulk modulus, commonly denoted K
(SI unit: N /m2).

A special case is a time-harmonic wave, for which the pressure varies with time as

iωt
p ( x, t ) = p ( x ) e

where ω = 2π f (SI unit: rad/s) is the angular frequency and f (SI unit: Hz) is denoting
the frequency. The wave equation for acoustic waves reduces to the Helmholtz
equation:

2
ω
∇ ⋅ ( ∇p ) + -----2- p = 0 (1-2)
c

where the ratio ω/c is recognized as the wave number k. This equation can also be
treated as an eigenvalue PDE to solve for eigenmodes and eigenfrequencies.

Typical boundary conditions for the wave equation and the Helmholtz equation are:

• Sound-hard boundaries (walls)

• Sound-soft boundaries
• Impedance boundary conditions
• Radiation boundary conditions

A detailed derivation of the governing equations is given in Theory Background for

the Pressure Acoustics Branch. For the propagation of compressional (acoustic) waves

FUNDAMENTAL OF ACOUSTICS | 35
 in a viscous and thermally conducting fluid the theory is presented in Theory
Background for the Thermoviscous Acoustics Branch and for acoustics in moving
media (aeroacoustics) in Theory Background for the Aeroacoustics Branch.

Damping
Fluids like air or water — by far the most common media in acoustics simulations —
exhibit practically no internal damping (so-called bulk attenuation) over the number
of wavelengths that can be resolved with the finite element method and in the audio
frequency range. However, in ultrasound applications or when using ray tracing to
model room and underwater acoustics, these become important.

In smaller systems, damping takes place through interaction with solids, either because
of friction between the fluid and a porous material filling the domain, or because
acoustic energy is transferred to a surrounding solid where it is absorbed. In systems
with small length scales, significant losses can occur in the viscous and thermal acoustic
boundary layer at walls.

ATMOSPHERE AND OCEAN ATTENUATION

When performing ray acoustics simulations or modeling ultrasound applications, the
bulk or internal attenuation of atmospheric air or the ocean sea water can be modeled
using the built-in Atmosphere attenuation or the Ocean attenuation models. Both
models are semi analytical and fitted to extensive experimental data. The atmosphere
model complies with the ANSI standard.

For further details about material properties and attenuation models see
the Acoustic Properties of Fluids chapter.

POROUS ABSORBING MATERIALS

For frequency-domain modeling, the most convenient and compact description of a
damping material (where material here refers to the homogenization of a fluid and a
porous solid) is given by its complex wave number k and complex impedance Z, both
functions of frequency. Knowing these properties, define a complex speed of sound as
cc = ω/k and a complex density as ρc = kZ/ω. Defining ρc and cc results in a so-called
equivalent-fluid model or fluid model.

It is possible to directly measure the complex wave number and impedance in an

impedance tube in order to produce curves of the real and imaginary parts (the

36 | CHAPTER 1: INTRODUCTION
resistance and reactance, respectively) as functions of frequency. These data can be used
directly as input to COMSOL Multiphysics interpolation functions to define k and Z.

Sometimes acoustic properties cannot be obtained directly for a material you want to
try in a model. In that case you must resort to knowledge about basic material
properties independent of frequency. Several empirical or semi-empirical models exist
in COMSOL Multiphysics and can estimate the complex wave number and impedance
as function of material parameters. These models are defined in the Poroacoustics
domain feature of the Pressure Acoustics interfaces — for example, the
Johnson-Champoux-Allard model and the Delany-Bazley-Miki models; the latter
uses frequency and flow resistivity as input.

The Acoustics Module includes a series of fluid models that are described
in Pressure Acoustics and Theory for the Equivalent Fluid Models. In
addition, The Poroelastic Waves Interface can be used for detailed
modeling of the propagation of coupled pressure and elastic waves in
porous materials.

BOUNDARY LAYER ABSORPTION (THERMOVISCOUS ACOUSTICS)

In systems of small dimensions (or at low frequencies) the size of the acoustic
boundary layer (the viscous and thermal acoustic penetration depth) that exists at all
walls can become comparable to the physical dimensions of the modeled system. In air
the boundary layer thickness is 0.22 mm at 100 Hz. This is typically the case inside
miniature transducers, condenser microphones, in MEMS systems, in tubing for
hearing aids, or in narrow gaps of vibrating structures.

For such systems, it is often necessary to use a more detailed model for the propagation
of the acoustics waves. This model is implemented in the Thermoviscous Acoustics
Interfaces. In simple cases for sound propagating in long ducts of constant cross
sections, the losses occurring at the boundaries can be smeared out on the fluid using
one of the fluid models of the Narrow Region Acoustics domain feature. For
geometries with curved surfaces and non-constant cross sections an alternative is to use

FUNDAMENTAL OF ACOUSTICS | 37
 the Thermoviscous Boundary Layer Impedance boundary condition, also available in
Pressure Acoustics.

More details on the detailed acoustic model for viscous and thermal losses
are described in Thermoviscous Acoustics Interfaces. See the boundary
layer absorption fluid models in Narrow Region Acoustics for simplified
modeling in uniform waveguide structures or the Thermoviscous
Boundary Layer Impedance boundary condition.

DAMPING AT BOUNDARIES
The losses associated with the acoustic field often stem from the interaction with
boundaries, for example, when interacting with a rubber material. In this case, it may
be necessary to include the acoustic-structure interaction using the appropriate
multiphysics coupling. Another way of including the losses is to use an impedance
boundary condition. The Acoustics Module provides a series of impedance models to
model, for example, the human ear, human skin, or a simple mechanical lumped RCL
system.

Artificial Boundaries
In most cases, the acoustic wave pattern that is to be simulated is not contained in a
closed cavity. That is, there are boundaries in the model that do not represent a physical
wall or limit of any kind. Instead, the boundary condition has to represent the
interaction between the wave pattern inside the model and everything outside.
Conditions of this kind are generically referred to as artificial boundary conditions
(ABCs).

Such conditions should ideally contain complete information about the outside world,
but this is not practical. After all, the artificial boundary was introduced to avoid
spending degrees of freedom (DOFs) on modeling whatever is outside. The solution
lies in trying to approximate the behavior of waves outside the domain using only
information from the boundary itself. This is difficult in general for obvious reasons.

One particular case that occurs frequently in acoustics concerns boundaries that can be
assumed to let wave energy propagate out from the domain without reflections. This
leads to the introduction of a particular group of artificial boundary conditions known
as nonreflecting boundary conditions (NRBCs), of which three kinds are available in
this module: port conditions, matched boundary conditions, and radiation boundary
conditions.

38 | CHAPTER 1: INTRODUCTION
Another way to model an open nonreflecting boundary is to add a so-called perfectly
matched layer (PML) domain or an absorbing layer domain. These domains use two
different techniques to dampens all outgoing waves with no or minimal reflections.
See, for example, Perfectly Matched Layers (PMLs) for more information. In pressure
acoustics a special version of the PML exists as a boundary condition, the Perfectly
Matched Boundary.

FUNDAMENTAL OF ACOUSTICS | 39
 Acoustics Module Physics Interface
Guide
The Acoustics Module extends the functionality of the physics interfaces of the
COMSOL Multiphysics base package. The details of the physics interfaces and study
types for the Acoustics Module are listed in the table below.

In the COMSOL Multiphysics Reference Manual:

• Studies and Solvers

• The Physics Interfaces
• For a list of all the core physics interfaces included with a COMSOL
Multiphysics license, see Physics Interface Guide.

PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE

DIMENSION

Acoustics

Pressure Acoustics

Pressure Acoustics, acpr all dimensions eigenfrequency; frequency

Frequency Domain1 domain; frequency domain,
modal; mode analysis (2D
and 1D axisymmetric
models only); boundary
mode analysis (3D and 2D
axisymmetric models only)
Pressure Acoustics, actd all dimensions eigenfrequency; frequency
Transient domain; frequency domain,
modal; time dependent;
time dependent, modal;
mode analysis (2D and 1D
axisymmetric models only)
Pressure Acoustics, acbm 3D, 2D mode analysis
Boundary Mode axisymmetric

40 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Pressure Acoustics, pabe 3D, 2D frequency domain

Boundary Elements
Pressure Acoustics, pate 3D, 2D, 2D time dependent
Time Explicit axisymmetric
Pressure Acoustics, paas 3D frequency domain
Asymptotic Scattering
Pressure Acoustics, pakh 3D frequency domain
Kirchhoff-Helmholtz
Elastic Waves

Solid Mechanics solid 3D, 2D, 2D stationary; eigenfrequency;

(Elastic Waves)1 axisymmetric eigenfrequency,
prestressed; mode analysis;
time dependent; time
dependent, modal; time
dependent, modal
reduced-order model;
frequency domain;
frequency domain, modal;
frequency domain,
prestressed; frequency
domain, prestressed,
modal; frequency domain,
modal reduced-order
model; frequency domain,
AWE reduced-order
model; modal
reduced-order model
Poroelastic Waves pelw 3D, 2D, 2D eigenfrequency; frequency
axisymmetric domain; frequency domain,
modal
Elastic Waves, Time elte 3D, 2D, 2D time dependent
Explicit axisymmetric
Piezoelectric Waves, — 3D, 2D, 2D time dependent
Time Explicit axisymmetric

ACOUSTICS MODULE PHYSICS INTERFACE GUIDE | 41

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Acoustic-Structure Interaction

Acoustic-Solid — 3D, 2D, 2D eigenfrequency; frequency

Interaction, Frequency axisymmetric domain; frequency domain,
Domain3 modal
Acoustic-Solid — 3D, 2D, 2D eigenfrequency; frequency
Interaction, Transient3 axisymmetric domain; frequency domain,
modal; time dependent;
time dependent, modal
Acoustic-Shell — 3D, 2D eigenfrequency; frequency
Interaction, Frequency axisymmetric domain; frequency domain,
Domain2,3 modal
Acoustic-Shell — 3D, 2D eigenfrequency; frequency
Interaction, Transient2, axisymmetric domain; frequency domain,
3 modal; time dependent;
time dependent, modal
Acoustic-Piezoelectric — 3D, 2D, 2D eigenfrequency; frequency
Interaction, Frequency axisymmetric domain; frequency domain,
Domain3 modal
Acoustic-Piezoelectric — 3D, 2D, 2D eigenfrequency; frequency
Interaction, Transient3 axisymmetric domain; frequency domain,
modal; time dependent;
time dependent, modal
Acoustic-Solid-Poroela — 3D, 2D, 2D eigenfrequency; frequency
stic Waves Interaction3 axisymmetric domain; frequency domain,
modal
Acoustic-Poroelastic — 3D, 2D, 2D eigenfrequency; frequency
Waves Interaction3 axisymmetric domain; frequency domain,
modal
Acoustic-Solid — 3D, 2D time domain
Interaction, Time
Explicit
Aeroacoustics

Linearized Euler, lef 3D, 2D, 2D frequency domain;

Frequency Domain axisymmetric, eigenfrequency
1D

42 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Linearized Euler, let 3D, 2D, 2D time dependent

Transient axisymmetric,
1D
Linearized Euler, lebm 3D, 2D mode analysis
Boundary Mode axisymmetric
Linearized Potential ae all dimensions frequency domain; mode
Flow, Frequency analysis (2D and 1D
Domain axisymmetric models only)
Linearized Potential aetd all dimensions frequency domain; time
Flow, Transient dependent; mode analysis
(2D and 1D axisymmetric
models only)
Linearized Potential aebm 3D, 2D mode analysis
Flow, Boundary Mode axisymmetric
Compressible Potential cpf all dimensions stationary; time dependent
Flow
Linearized lnsf 3D, 2D, 2D frequency domain;
Navier-Stokes, axisymmetric, eigenfrequency
Frequency Domain 1D
Linearized lnst 3D, 2D, 2D time dependent
Navier-Stokes, axisymmetric,
Transient and 1D
Linearized lnsbm 3D, 2D mode analysis
Navier-Stokes, axisymmetric
Boundary Mode
Thermoviscous Acoustics

Thermoviscous ta all dimensions eigenfrequency; frequency

Acoustics, Frequency domain; frequency domain,
Domain modal; mode analysis (2D
and 1D axisymmetric
models only)
Thermoviscous tatd all dimensions time dependent
Acoustics, Transient
Thermoviscous tabm 3D, 2D mode analysis
Acoustics, Boundary axisymmetric
Mode

ACOUSTICS MODULE PHYSICS INTERFACE GUIDE | 43

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Acoustic-Thermovisco — 3D, 2D, 2D eigenfrequency; frequency

us Acoustic axisymmetric domain; frequency domain,
Interaction, Frequency modal; boundary mode
Domain3 analysis (3D and 2D
axisymmetric only); mode
analysis (2D only)
Thermoviscous — 3D, 2D, 2D eigenfrequency; frequency
Acoustic-Solid axisymmetric domain; frequency domain,
Interaction, Frequency modal; mode analysis (2D
Domain3 only)
Thermoviscous — 3D eigenfrequency; frequency
Acoustic-Shell domain; frequency domain,
Interaction, Frequency modal
Domain2,3
Ultrasound

Convected Wave cwe 3D, 2D, 2D time dependent

Equation, Time Explicit axisymmetric
Nonlinear Pressure nate 3D, 2D, 2D time dependent
Acoustics, Time axisymmetric
Explicit
Geometrical Acoustics

Ray Acoustics rac 3D, 2D, 2D ray tracing; time dependent

axisymmetric
Acoustic Diffusion ade 3D eigenvalue; stationary; time
Equation dependent
Pipe Acoustics

Pipe Acoustics, pafd 3D, 2D eigenfrequency; frequency

Frequency Domain domain
Pipe Acoustics, patd 3D, 2D time dependent
Transient

44 | CHAPTER 1: INTRODUCTION
PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Fluid-Structure Interaction

Structural Mechanics

Solid Mechanics1 solid 3D, 2D, 2D stationary; eigenfrequency;

axisymmetric eigenfrequency,
prestressed; mode analysis;
time dependent; time
dependent, modal; time
dependent, modal
reduced-order model;
frequency domain;
frequency domain, modal;
frequency domain,
prestressed; frequency
domain, prestressed,
modal; frequency domain,
modal reduced-order
model; frequency domain,
AWE reduced-order model
Piezoelectricity3 — 3D, 2D, 2D stationary; eigenfrequency;
axisymmetric eigenfrequency,
prestressed; time
dependent; time
dependent, modal;
frequency domain;
frequency domain, modal;
frequency domain,
prestressed; frequency
domain, prestressed,
modal; small-signal analysis,
frequency domain
Magnetostriction,3,4 — 3D, 2D, 2D stationary; eigenfrequency;
axisymmetric time dependent; frequency
domain; small-signal
analysis, frequency domain;
eigenfrequency,
prestressed; frequency
domain, prestressed

ACOUSTICS MODULE PHYSICS INTERFACE GUIDE | 45

 PHYSICS INTERFACE ICON TAG SPACE AVAILABLE STUDY TYPE
DIMENSION

Magnetomechanics,3,4 — 3D, 2D, 2D stationary; eigenfrequency;

axisymmetric time dependent; frequency
domain; small-signal
analysis, frequency domain;
eigenfrequency,
prestressed; frequency
domain, prestressed
Magnetomechanics, No — 3D, 2D, 2D stationary; eigenfrequency;
Currents,3,4 axisymmetric time dependent; frequency
domain; small-signal
analysis, frequency domain;
eigenfrequency,
prestressed; frequency
domain, prestressed
1
This physics interface is included with the core COMSOL package but has added
functionality for this module.
2
Requires both the Structural Mechanics Module and the Acoustics Module.
3
This physics interface is a predefined multiphysics coupling that automatically adds all the
physics interfaces and coupling features required.
4
Requires the addition of the AC/DC Module.

Common Physics Interface and Feature Settings and Nodes

There are several common settings and sections available for the physics interfaces and
feature nodes. Some of these sections also have similar settings or are implemented in
the same way no matter the physics interface or feature being used.

In each module’s documentation, only unique or extra information is included;

standard information and procedures are centralized in the COMSOL Multiphysics
Reference Manual.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for

links to common sections and Table 2-5 to common feature nodes.
You can also search for information: press F1 to open the Help
window or Ctrl+F1 to open the Documentation window.

46 | CHAPTER 1: INTRODUCTION
Where Do I Access the Documentation and Application Libraries?
A number of internet resources have more information about COMSOL, including
licensing and technical information. The electronic documentation, topic-based (or
context-based) help, and the application libraries are all accessed through the
COMSOL Desktop.

If you are reading the documentation as a PDF file on your computer,

the blue links do not work to open an application or content
referenced in a different guide. However, if you are using the Help
system in COMSOL Multiphysics, these links work to open other
modules, application examples, and documentation sets.

THE DOCUMENTATION AND ONLINE HELP

The COMSOL Multiphysics Reference Manual describes the core physics interfaces
and functionality included with the COMSOL Multiphysics license. This book also has
instructions about how to use COMSOL Multiphysics and how to access the
electronic Documentation and Help content.

Opening Topic-Based Help

The Help window is useful as it is connected to the features in the COMSOL Desktop.
To learn more about a node in the Model Builder, or a window on the Desktop, click
to highlight a node or window, then press F1 to open the Help window, which then
displays information about that feature (or click a node in the Model Builder followed
by the Help button ( ). This is called topic-based (or context) help.

To open the Help window:

• In the Model Builder, Application Builder, or Physics Builder click a node or

window and then press F1.
• On any toolbar (for example, Home, Definitions, or Geometry), hover the
mouse over a button (for example, Add Physics or Build All) and then
press F1.
• From the File menu, click Help ( ).
• In the upper-right corner of the COMSOL Desktop, click the Help ( )
button.

ACOUSTICS MODULE PHYSICS INTERFACE GUIDE | 47

 To open the Help window:

• In the Model Builder or Physics Builder click a node or window and then
press F1.
• On the main toolbar, click the Help ( ) button.
• From the main menu, select Help>Help.

Opening the Documentation Window

To open the Documentation window:

• Press Ctrl+F1.
• From the File menu select Help>Documentation ( ).

To open the Documentation window:

• Press Ctrl+F1.
• On the main toolbar, click the Documentation ( ) button.
• From the main menu, select Help>Documentation.

THE APPLICATION LIBRARIES WINDOW

Each model or application includes documentation with the theoretical background
and step-by-step instructions to create a model or application. The models and
applications are available in COMSOL Multiphysics as MPH files that you can open
for further investigation. You can use the step-by-step instructions and the actual
models as templates for your own modeling. In most models, SI units are used to
describe the relevant properties, parameters, and dimensions, but other unit systems
are available.

Once the Application Libraries window is opened, you can search by name or browse
under a module folder name. Click to view a summary of the model or application and
its properties, including options to open it or its associated PDF document.

The Application Libraries Window in the COMSOL Multiphysics

Reference Manual.

48 | CHAPTER 1: INTRODUCTION
Opening the Application Libraries Window
To open the Application Libraries window ( ):

• From the Home toolbar, Windows menu, click ( ) Applications

Libraries.
• From the File menu select Application Libraries.

To include the latest versions of model examples, from the File>Help

menu, select ( ) Update COMSOL Application Library.

Select Application Libraries from the main File> or Windows> menus.

To include the latest versions of model examples, from the Help menu
select ( ) Update COMSOL Application Library.

CONTACTING COMSOL BY EMAIL

For general product information, contact COMSOL at info@[Link].

COMSOL ACCESS AND TECHNICAL SUPPORT

To receive technical support from COMSOL for the COMSOL products, please
contact your local COMSOL representative or send your questions to
support@[Link]. An automatic notification and a case number are sent to you by
email. You can also access technical support, software updates, license information, and
other resources by registering for a COMSOL Access account.

ACOUSTICS MODULE PHYSICS INTERFACE GUIDE | 49

 COMSOL ONLINE RESOURCES

COMSOL website [Link]

Contact COMSOL [Link]/contact
COMSOL Access [Link]/access
Support Center [Link]/support
Product Download [Link]/product-download
Product Updates [Link]/support/updates
COMSOL Blog [Link]/blogs
Discussion Forum [Link]/community
Events [Link]/events
COMSOL Application Gallery [Link]/models
COMSOL Video Gallery [Link]/video
Support Knowledge Base [Link]/support/knowledgebase

50 | CHAPTER 1: INTRODUCTION
Overview of the User’s Guide
The Acoustics Module User’s Guide gets you started with modeling acoustics using
COMSOL Multiphysics. The information in this guide is specific to the Acoustics
Module. Instructions on how to use COMSOL in general are included with the
COMSOL Multiphysics Reference Manual.

As detailed in the section Where Do I Access the Documentation and

Application Libraries? this information can also be searched from the
COMSOL Multiphysics software Help menu.

Good practices as well as some tips and tricks are located in dedicated
modeling sections under each branch. They are:

• Modeling with the Pressure Acoustics Branch (FEM-Based Interfaces)

• Modeling with the Pressure Acoustics Branch (BEM-Based Interface)
• Modeling with the Pressure Acoustics Branch (DG-FEM-Based
Interface)
• Modeling with the Aeroacoustic Flow Source: Flow-Induced Noise
• Modeling with the Elastic Waves Branch
• Modeling with the Acoustic-Structure Interaction Branch
• Modeling with the Aeroacoustics Branch
• Modeling with the Thermoviscous Acoustics Branch
• Modeling with the Convected Wave Equation Interface
• Modeling with the Nonlinear Pressure Acoustics, Time Explicit
Interface
• Modeling with the Ray Acoustics Interface
• Modeling with the Acoustic Diffusion Equation Interface
• Modeling with Multiphysics Couplings

Dedicated theory sections also exist under each branch. Here the physics
and mathematics of the governing equations and selected boundary
conditions are discussed.

OVERVIEW OF THE USER’S GUIDE | 51

 TABLE OF CONTENTS, GLOSSARY, AND INDEX
To help you navigate through this guide, see the Contents, Glossary, and Index.

THE PRESSURE ACOUSTICS BRANCH

The Pressure Acoustics Interfaces chapter describes the following interfaces:

The Pressure Acoustics, Frequency Domain Interface is the core physics interface
which models the sound waves in the frequency domain and The Pressure Acoustics,
Transient Interface is the core physics interface which models the sound waves in the
time domain with the possibility to include nonlinear effects. The Pressure Acoustics,
Boundary Mode Interface solves for modes that propagate through a cross section of
waveguides in your geometry.

The Pressure Acoustics, Boundary Elements Interface uses the boundary element
(BEM) method to solve the Helmholtz equation in the frequency domain. It is well
suited for radiation and scattering problems and can be seamlessly coupled to
structures and the finite element based The Pressure Acoustics, Frequency Domain
Interface.

The Pressure Acoustics, Time Explicit Interface uses the discontinuous Galerkin
(dG-FEM) formulation to solve transient models using a time explicit method. The
interface can solve large problems as the method is very computationally efficient.

The Pressure Acoustics, Asymptotic Scattering Interface and The Pressure Acoustics,
Kirchhoff-Helmholtz Interface are dedicated high frequency interfaces. The method
used is also referred to as high frequency BEM or HFB.

THE ELASTIC WAVES BRANCH

The Elastic Waves Interfaces chapter describes the following interfaces:

The Solid Mechanics (Elastic Waves) Interface is a shortcut to add the Solid Mechanics
interface which is used to compute the displacement field in solids with propagating
elastic waves.

The Poroelastic Waves Interface is used to compute the displacement field and acoustic
pressure fluctuation in porous materials with propagating poroelastic waves.

The Elastic Waves, Time Explicit Interface uses the discontinuous Galerkin (dG-FEM)
formulation to solve transient linear elastic wave models using a time explicit method.
The interface can solve large problems as the method is very computationally efficient.

The Piezoelectric Waves, Time Explicit Interface is a multiphysics interface that

couples the The Elastic Waves, Time Explicit Interface and Electrostatics in order to

52 | CHAPTER 1: INTRODUCTION
model the propagation of linear piezoelectric waves in the time domain, using a hybrid
FEM-DG time explicit approach. The method is very computationally efficient.

THE ACOUSTIC-STRUCTURE INTERACTION BRANCH

The Acoustic-Structure Interaction Interfaces chapter describes the following
interfaces:

The Acoustic-Solid Interaction, Frequency Domain Interface is a combination of

pressure acoustics and solid mechanics with predefined couplings and The
Acoustic-Solid Interaction, Transient Interface is a combination of transient pressure
acoustics and solid mechanics with predefined couplings.

The Acoustic-Shell Interaction, Frequency Domain Interface and The Acoustic-Shell

Interaction, Transient Interface requires a Structural Mechanics Module license. This
multiphysics interface combines the Pressure Acoustics, Frequency Domain (or the
Pressure Acoustics, Transient) interface, the Shell interface, and the Acoustic-Structure
Boundary multiphysics coupling. The physics interface is available for 2D axisymmetric
and 3D geometries.

The Acoustic-Piezoelectric Interaction, Frequency Domain Interface is a combination

of Pressure Acoustics, Frequency Domain; solid Mechanics; Electrostatics;
Acoustic-structure boundary; Piezoelectric Effect. The Acoustic-Piezoelectric
Interaction, Transient Interface combines Pressure Acoustics, Transient; Solid
Mechanics; Electrostatics; Acoustic-structure boundary; Piezoelectric Effect.

The Acoustic-Poroelastic Waves Interaction Interface combines Pressure Acoustics,

Frequency Domain and Poroelastic Waves together with the Acoustic-Porous
Boundary multiphysics coupling.

The Acoustic-Solid-Poroelastic Waves Interaction Interface combines Pressure

Acoustics, Frequency Domain and Solid Mechanics together with the
Acoustic-Structure Boundary and Acoustic-Porous Boundary multiphysics coupling.

The Acoustic-Solid Interaction, Time Explicit Interface combines Pressure Acoustics,

Time Explicit and Elastic Waves, Time Explicit together with the Acoustic-Structure
Boundary, Time Explicit multiphysics coupling.

THE AEROACOUSTICS BRANCH

The Aeroacoustics Interfaces chapter describes the following interfaces:

The Linearized Euler, Frequency Domain Interface and The Linearized Euler,
Transient Interface models the acoustic variations in density, velocity, and pressure in

OVERVIEW OF THE USER’S GUIDE | 53

 the presence of a stationary background mean-flow that is well approximated by an
ideal gas flow. This physics interface is used for aeroacoustic simulations that can be
described by the linearized Euler equations.

The Linearized Navier–Stokes, Frequency Domain Interface and The Linearized

Navier–Stokes, Transient Interface models the acoustic variations in pressure, velocity,
and temperature in the presence of any stationary isothermal or nonisothermal
background mean flow. This physics interface is used for aeroacoustic simulations that
can be described by the linearized Navier-Stokes equations.

The The Linearized Navier–Stokes, Boundary Mode Interface and the The Linearized
Euler, Boundary Mode Interface are used to solve for modes that propagate through
a cross section of waveguides in your geometry in the presence of a stationary
mean-flow.

The Linearized Potential Flow, Frequency Domain Interface models acoustic waves in
potential flow in the frequency domain and The Linearized Potential Flow, Transient
Interface models acoustic waves in potential flow in the time domain. The Linearized
Potential Flow, Boundary Mode Interface solves for modes that propagate through a
cross section of your geometry.

The Compressible Potential Flow Interface models irrotational flow used as input for
the background flow in the linearized potential flow interfaces.

THE THERMOVISCOUS ACOUSTICS BRANCH

The Thermoviscous Acoustics Interfaces chapter describes The Thermoviscous
Acoustics, Frequency Domain Interface and The Thermoviscous Acoustics, Transient
Interface which are necessary when modeling acoustics accurately in geometries with
small dimensions. Near walls, viscosity and thermal conduction become important
because they create a viscous and a thermal boundary layer where losses are significant.
Nonlinear effects can be modeled on the time domain. The Thermoviscous Acoustics,
Boundary Mode Interface is used to identify propagating and nonpropagating modes
in waveguides and ducts of small dimensions, including thermal and viscous losses.

The Acoustic-Thermoviscous Acoustic Interaction, Frequency Domain Interface

combines the Thermoviscous Acoustics, Frequency Domain and Pressure Acoustics,
Frequency Domain interfaces together with the Acoustic-Thermoviscous Acoustic
Boundary multiphysics coupling.

The Thermoviscous Acoustic-Solid Interaction, Frequency Domain Interface is also

described here. This physics interface combines the Thermoviscous Acoustics,

54 | CHAPTER 1: INTRODUCTION
Frequency Interface domain and Solid Mechanics Interface together with the
Thermoviscous Acoustic-Structure Boundary multiphysics coupling.

The Thermoviscous Acoustic-Shell Interaction, Frequency Domain Interface requires

a Structural Mechanics Module license. The multiphysics interface combines the
Thermoviscous Acoustics, Frequency domain Interface and the Shell Interface
together with the Thermoviscous Acoustic-Structure Boundary multiphysics coupling.

ULTRASOUND
The Ultrasound Interfaces chapter describes the following interfaces:

The Convected Wave Equation, Time Explicit Interface uses the discontinuous
Galerkin (dG-FEM) formulation to model the propagating of linear ultrasound waves
in the time domain, including the effects of a stationary background flow.

The Nonlinear Pressure Acoustics, Time Explicit Interface uses the discontinuous
Galerkin (dG-FEM) formulation to model the propagation of nonlinear high
amplitude pressure waves in fluids including dissipation effects.

THE GEOMETRICAL ACOUSTICS BRANCH

The Geometrical Acoustics Interfaces includes The Ray Acoustics Interface, used to
compute the trajectories, phase, and intensity of acoustic rays, and The Acoustic
Diffusion Equation Interface, which solves a diffusion equation for the acoustic energy
density. The theory is also discussed for both physics interfaces.

THE PIPE ACOUSTICS BRANCH

The Pipe Acoustics Frequency Domain and Transient Interfaces have the equations
and boundary conditions for modeling the propagation of sound waves in flexible pipe
systems. The equations are formulated in a general way to include the possibility of a
stationary background flow. There are two interfaces, one for transient analysis and
one for frequency domain studies.

THE MULTIPHYSICS COUPLINGS

The Multiphysics Couplings chapter describes all the multiphysics couplings available
with the Acoustics Module. These are the built-in conditions that can couple the
physics together; for example, the Acoustic-Structure Boundary couples pressure
acoustics to any solid boundary.

OVERVIEW OF THE USER’S GUIDE | 55

 STRUCTURAL MECHANICS WITH THE ACOUSTICS MODULE
The Structural Mechanics with the Acoustics Module chapter provides information
about The Solid Mechanics Interface used for modeling, for example, the structural
part of acoustic-structure interaction. This is an extension of the Solid Mechanics
interface in COMSOL Multiphysics, and you find it under the Structural Mechanics
branch, or using the shortcut The Solid Mechanics (Elastic Waves) Interface under the
Elastic Waves branch.

The Piezoelectricity Interface interface is also shortly presented. It combines Solid

Mechanics and Electrostatics together with the constitutive relationships required to
model piezoelectrics. Both the direct and inverse piezoelectric effects can be modeled
and the piezoelectric coupling can be formulated using either the strain-charge or
stress-charge forms. This interfaces is based on the finite element method (FEM) and
can be used in both frequency and time domain.

The theory for the solid mechanics interface as well as for the Piezoelectricity interface
is found in the Structural Mechanics Module User’s Guide.

56 | CHAPTER 1: INTRODUCTION
 2

Pressure Acoustics Interfaces

This chapter describes the Acoustics Module background theory and physics
interfaces found under the Pressure Acoustics branch ( ).

• The Pressure Acoustics, Frequency Domain Interface

• The Pressure Acoustics, Transient Interface
• The Pressure Acoustics, Boundary Mode Interface
• The Pressure Acoustics, Boundary Elements Interface
• The Pressure Acoustics, Time Explicit Interface
• The Pressure Acoustics, Asymptotic Scattering Interface
• The Pressure Acoustics, Kirchhoff-Helmholtz Interface
• Modeling with the Pressure Acoustics Branch (FEM-Based Interfaces)
• Modeling with the Pressure Acoustics Branch (BEM-Based Interface)
• Modeling with the Pressure Acoustics Branch (DG-FEM-Based Interface)
• Modeling with the Aeroacoustic Flow Source: Flow-Induced Noise
• Theory Background for the Pressure Acoustics Branch
• Theory for the Boundary Impedance Models
• Theory for the Interior Impedance Models

57
 • Theory for the Equivalent Fluid Models
• Theory for the Perfectly Matched Layers in the Time Domain
• References for the Pressure Acoustics Branch

58 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The Pressure Acoustics, Frequency
Domain Interface
The Pressure Acoustics, Frequency Domain (acpr) interface ( ), found under the
Pressure Acoustics branch ( ) when adding a physics interface, is used to compute
the pressure variations for the propagation of acoustic waves in fluids at quiescent
background conditions. It is suited for all frequency-domain simulations with
harmonic variations of the pressure field.

The physics interface can be used for linear acoustics described by a scalar pressure
variable. It includes domain conditions to model losses in a homogenized way,
so-called equivalent fluid models, for porous materials as well as losses in narrow
regions (waveguides or slits). The plane wave attenuation behavior of the acoustic
waves may also be entered as a user-defined quantity, or based on the losses in the
atmosphere or ocean. Effective anisotropic material properties can be modeled with an
effective anisotropic density formulation. Domain features also include background
acoustic fields, as well as monopole and dipole domain sources. Aeroacoustic flow
sources or flow-induced noise from turbulence can be included when coupled to an
LES simulation (this requires the CFD Module).

The physics interface solves the Helmholtz equation in the frequency domain for given
frequencies, or as an eigenfrequency or mode analysis study. The harmonic variation of
iωt
all fields and sources is given by e using the +iω convention.

An acoustics model can be part of a larger multiphysics model that describes, for
example, the interactions between structures and acoustic waves. This physics interface
is suitable for modeling acoustics phenomena that do not involve fluid flow (convective
effects).

The sound pressure p, which is solved for in pressure acoustics, represents the acoustic
variations (or acoustic perturbations) to the ambient pressure. In the absence of flow,
the ambient pressure pA is simply the static absolute pressure.

The governing equations and boundary conditions are formulated using the total
pressure pt with a so-called scattered field formulation. In the presence of a
Background Pressure Field defining a background pressure wave pb (this could, for
example, be a plane wave), the total acoustic pressure pt is the sum of the pressure
solved for p (which is then equal to the scattered pressure ps) and the background

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 59

 pressure wave: pt = p+pb. The equations then contain the information about both the
scattered field and the background pressure field.

For good modeling strategies, meshing, solvers, postprocessing

information, acoustics specific plots, as well as tips and tricks, see the
Modeling with the Pressure Acoustics Branch (FEM-Based Interfaces)
section.

When the geometrical dimensions of the acoustic problems are reduced from 3D to
2D (planar symmetry or axisymmetry) or to 1D axisymmetry, it is possible to specify
an out-of-plane wave number kz and an azimuthal mode number m (sometimes
referred to as the circumferential mode number), when applicable. In this case, the
wave number used in the equations keq contains both the ordinary wave number k as
well as the out-of-plane wave number kz and azimuthal wave number km = m/r, when
applicable.

The following table lists the names and SI units for the most important physical
quantities in the Pressure Acoustics, Frequency Domain interface:
TABLE 2-1: PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE PHYSICAL QUANTITIES

QUANTITY SYMBOL SI UNIT ABBREVIATION

Pressure p Pascal Pa
Total pressure pt Pascal Pa
Background pressure pb Pascal Pa
Scattered pressure ps Pascal Pa
Density (quiescent) ρ or ρc kilogram/meter3 kg/m3
Frequency f Hertz Hz
Wave number k 1/meter 1/m
Dipole domain source qd newton/meter3 N/m3
Monopole domain source Qm 1/second2 1/s2
Speed of sound c or cc meter/second m/s
Specific acoustic impedance Z or Zsp pascal-second/meter Pa·s/m
Acoustic impedance Zac pascal-second/meter3 Pa·s/m3
Normal acceleration an meter/second2 m/s2
Normal velocity vn meter/second m/s
Source location x0 meter m
Wave direction ek (dimensionless) 1

60 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

In the following descriptions of the functionality in this physics interface, the subscript
c in ρc and cc (the density and speed of sound, respectively) denotes that these can be
complex-valued quantities in models with damping.

When this physics interface is added, these default nodes are also added to the Model
Builder — Pressure Acoustics Model, Sound Hard Boundary (Wall), and Initial Values.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and point conditions. You can also right-click Pressure Acoustics,
Frequency Domain to select physics features from the context menu.

Physics Nodes — Equation Section in the COMSOL Multiphysics

Reference Manual

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is acpr.

EQUATION
Expand the Equation section to see the equations solved for with the Equation form
specified. The default selection for Equation form is Study controlled. The available
studies are selected under Show equations assuming.

• For Study controlled, the scaling and nonreflecting boundary settings are optimized
for the numerical performance of the different solvers and study types.
• For Frequency domain, enter the settings as described in Scaling Factor and
Nonreflecting Boundary Condition Approximation.

PHYSICS SYMBOLS
Select to Enable physics symbols (selected per default). This shows the physics symbols,
like for example, normals or symmetry planes in the Graphics window when applicable.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 61

 PRESSURE ACOUSTICS EQUATION SETTINGS
In this section you can add out-of-plane information defining an out-of-plane wave
number kz or an azimuthal wave number km = m/r through the mode number m.
Add if applicable:

• For 1D axisymmetric components, the default Out-of-plane wave number kz (SI unit:
rad/m) is 0 rad/m. The default Azimuthal mode number m (dimensionless) is 0. The
pressure has the form:
– i ( k z z + mϕ )
p ( r, ϕ, z ) = p ( r )e

• For 2D axisymmetric components, the default Azimuthal mode number m

(dimensionless) is 0. The pressure has the form:
– imϕ
p ( r, ϕ, z ) = p ( r, z )e

• For 2D components, the default Out-of-plane wave number kz (SI unit: rad/m) is
0 rad/m. The pressure has the form:

– ik z z
p ( x, y, z ) = p ( x, y )e

Note that when performing a mode analysis study in 2D (solving for the out-of-plane
wave number) the Out-of-plane wave number property is not used as this is now the
variable solved for.

Scaling Factor and Nonreflecting Boundary Condition Approximation

For all component dimensions, and if required, click to expand the Equation section,
then select Frequency domain as the Equation form and enter the settings as described
below.

The default Scaling factor Δ is 1/ω2 and Nonreflecting boundary condition approximation
is Second order. These values correspond to the equations for a Frequency Domain
study when the equations are study controlled.

To get the equations corresponding to an Eigenfrequency study, change the Scaling

factor Δ to 1 and the Nonreflecting boundary conditions approximation to First order.

GLOBAL PORT SETTINGS

Select to enable the Activate port sweep option (not selected per default). This option
is used to compute the full scattering matrix when Port conditions are used. For more
details see The Port Sweep Functionality subsection. The section only exists for 3D
and 2D axisymmetry.

62 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Select the Mode shape normalization as Amplitude normalized (the default) or Power
normalized. This setting controls if the mode shapes are normalized to have a unit
maximum amplitude or carry unit power. The selection determines how the scattering
matrix is to be interpreted.

SOUND PRESSURE LEVEL SETTINGS

The zero level on the dB scale varies with the type of fluid. That value is a reference
pressure that corresponds to 0 dB. This variable occurs in calculations of the sound
pressure level Lp based on the root mean square (rms) pressure prms, such that

p rms
L p = 20 log  ----------
1
with p rms = --- p p∗
p ref 2

where pref is the reference pressure and the star (*) represents the complex conjugate.
This is an expression valid for the case of harmonically time-varying acoustic
pressure p.

Select a Reference pressure for the sound pressure level based on the fluid type:

• Use reference pressure for air to use a reference pressure of 20 μPa (20·10−6 Pa).
• Use reference pressure for water to use a reference pressure of 1 μPa (1·10−6 Pa).
• User-defined reference pressure to enter a reference pressure pref, SPL (SI unit: Pa).
The default value is the same as for air, 20 μPa.

TYPICAL WAVE SPEED

Enter a value or expression for the Typical wave speed for perfectly matched layers cref
(SI unit m/s). The default is 343 m/s.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) for the
Pressure, the default is Quadratic Lagrange.

Choosing between Lagrange and Serendipity Shape Functions has

influence on the number of DOFs solved for and on stability for a
distorted mesh.

To see all settings in this section, click the Show More Options button ( ) and select
Advanced Physics Options in the Show More Options dialog box.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 63

 DEPENDENT VARIABLES
This physics interface defines one dependent variable (field), the Pressure p. If
required, edit the name, which changes both the field name and the dependent variable
name. If the new field name coincides with the name of another pressure field in the
model, the interfaces share degrees of freedom and dependent variable name. The new
field name must not coincide with the name of a field of another type, or with a
component name belonging to some other field.

• Domain, Boundary, Edge, Point, and Pair Nodes for the Pressure
Acoustics, Frequency Domain Interface
• Theory Background for the Pressure Acoustics Branch

Eigenmodes of a Room: Application Library path COMSOL_Multiphysics/

Acoustics/eigenmodes_of_room

Acoustic Levitator: Application Library path Acoustics_Module/

Nonlinear_Acoustics/acoustic_levitator. This model also requires the
Particle Tracing Module.

Domain, Boundary, Edge, Point, and Pair Nodes for the Pressure
Acoustics, Frequency Domain Interface
The Pressure Acoustics, Frequency Domain Interface has Domain, Boundary, Edge,
Point, and Pair nodes, listed in alphabetical order, available from the Physics ribbon
toolbar (Windows® users), Physics context menu (macOS or Linux® users), or
right-click to access the context menu (all users). Continuity in the total pressure is the
default condition on interior boundaries.

The Pressure Acoustics, Transient Interface also shares some of these nodes, with some
additional features described in Domain, Boundary, Edge, and Point Nodes for the
Pressure Acoustics, Transient Interface.

The Pressure Acoustics, Boundary Mode Interface also shares some of these nodes,
with one additional feature described in Boundary, Edge, Point, and Pair Nodes for
the Pressure Acoustics, Boundary Mode Interface. For the Pressure Acoustics,

64 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Boundary Mode interface, apply the feature to boundaries instead of domains for 3D
components.

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 65

 • Aeroacoustic Flow Source • Lumped Port
• Anisotropic Acoustics • Lumped Speaker Boundary
• Axial Symmetry • Matched Boundary
• Background Pressure Field • Monopole Point Source
• Circular Source (for 2D • Monopole Domain Source
Axisymmetric Components) • Narrow Region Acoustics
• Continuity • Normal Acceleration
• Cylindrical Wave Radiation • Normal Displacement
• Dipole Point Source • Normal Velocity
• Dipole Domain Source • Quadrupole Point Source
• Exterior Field Calculation • Perfectly Matched Boundary
• Heat Source • Periodic Condition
• Impedance • Plane Wave Radiation
• Incident Pressure Field • Point Sources (for 2D Components)
• Interior Normal Acceleration • Poroacoustics
• Interior Normal Displacement • Port
• Interior Normal Velocity • Pressure Acoustics
• Interior Impedance/Pair Impedance • Pressure
• Interior Perforated Plate/Pair • Pressure (Point Condition)
Perforated Plate
• Sound Hard Boundary (Wall)
• Interior Sound Hard Boundary
• Sound Soft Boundary
(Wall)
• Spherical Wave Radiation
• Interior Lumped Speaker Boundary
• Symmetry
• Initial Values
• Thermoviscous Boundary Layer
• Line Source
Impedance
• Line Source on Axis

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry boundaries (at r = 0) into account and automatically adds an
Axial Symmetry node to the component that is valid on the axial symmetry
boundaries only.

66 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 In the COMSOL Multiphysics Reference Manual, see Table 2-4 for links
to common sections and Table 2-5 for common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Pressure Acoustics
The Pressure Acoustics node adds the equations for time-harmonic (frequency domain)
and eigenfrequency acoustics modeling of classical acoustics solving Helmholtz
equation. In the Settings window, define the properties for the acoustics model such as
material properties and bulk attenuation. Define the so-called Model Inputs to the
properties including the absolute background pressure and temperature, that is the
quiescent ambient properties.

The advanced fluid models are defined using individual physics feature
nodes: Poroacoustics, Narrow Region Acoustics, and Anisotropic
Acoustics. In the time domain nonlinear effects can be included using the
Nonlinear Acoustics (Westervelt) Contributions node.

For details about some of the available fluid models, see Theory for the
Equivalent Fluid Models.

For more information about using variables during the results analysis, see
Postprocessing Variables.

MODEL INPUTS
Select, User defined (the default), Common model input, or an input from another
physics interface, if applicable.

The values of the quiescent (or background) Temperature T (SI unit: K) and Absolute
pressure pA (SI unit: Pa) are entered in the Model Inputs section if required by the
material properties (other inputs like relative humidity or salinity can also appear). The
values of the model inputs can influence the material properties like, for example, the
density and speed of sound, depending on their definition in the Materials node. In

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 67

 general, if the model includes a material property that depends on a model input, the
corresponding text field will be enabled.

Details about the Model Input and the Default Model Inputs are found
in the Global and Local Definitions chapter of the COMSOL
Multiphysics Reference Manual.

If Ideal gas is selected as the Fluid model, both the Temperature T and Absolute pressure
pA fields are enabled. In this case this input option is always available.

In addition, the Temperature T and Absolute pressure pA can be picked up from another
physics interface where the fields have been calculated. For example, select a
temperature field defined by a Heat Transfer interface or a Nonisothermal Flow
interface (if any). Or, if applicable, select a pressure as defined by a Fluid Flow interface
present in the model. For example, select Absolute pressure (spf) to use the absolute
pressure defined by a Laminar Flow interface spf.

The input to the Model Inputs fields influences the value of the material parameters in
the model. Typically, the density ρ and the speed of sound c in the model depend on
the absolute pressure and/or the temperature. Picking up any of those from another
physics interface typically results in space dependent quantities ρ = ρ(x) and c = c(x).

Detailed convected acoustic models that take into account the full
background flow (including the movement of the fluid) are available
among the Aeroacoustics Interfaces. For flow-induced noise simulations
based on the Lighthill analogy see the Aeroacoustic Flow Source.

PRESSURE ACOUSTICS MODEL

To define the properties of the fluid, select a Fluid model from the list:

• Linear elastic (the default) for defining the classical lossless fluid: Go to Linear Elastic
Fluid Model.
• User-defined attenuation for defining losses through an attenuation coefficient: Go
to User-Defined Attenuation Fluid Model.
• Atmosphere attenuation defines the standard attenuation of atmospheric (moist) air:
Go to Atmosphere Attenuation Fluid Model.
• Ocean attenuation defines attenuation in the seawater of the ocean: Go to Ocean
Attenuation Fluid Model.

68 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• Viscous, Thermally conducting, and Thermally conducting and viscous defines the
classical thermoviscous attenuation model: Go to Thermally Conducting and/or
Viscous Fluid Model.
• General dissipation defines losses through the sound diffusivity: Go to General
Dissipation Fluid Model.
• Ideal gas to define the properties of an ideal gas: Go to Ideal Gas Fluid Model.

The theory for the fluid models is in the section Theory for the Equivalent
Fluid Models.

The fluid models represent different bulk loss or attenuation mechanisms

(applied in a homogenized way) or ways of defining the properties of the
fluid. Some of these models are sometimes referred to as equivalent fluid
models. The loss model can be a theoretical model or a model based on
measurement data for the attenuation in the fluid like the atmosphere or
the ocean.

Losses in porous materials are defined in Poroacoustics. Thermoviscous

boundary layer losses in narrow regions of constant cross section (like
waveguides) can be modeled using Narrow Region Acoustics.

Linear Elastic Fluid Model

Linear elastic is the default. When the material parameters are real values, this
corresponds to a lossless compressible fluid. From the Specify list, select Density and
speed of sound (the default), Impedance and wave number, or Bulk modulus and density.
To add user-defined losses, in a general way, specify the properties as complex-valued
data.

For each of the following, the default values (when applicable), are taken From material
or for User defined, enter other values or expressions.

• For Density and speed of sound, define the Speed of sound c (SI unit: m/s) and Density
ρ (SI unit: kg/m3).
• For Impedance and wave number, define the Characteristic acoustic impedance Z
(SI unit: Pa·s/m) and enter a Wave number k (SI unit: rad/m).
• For Bulk modulus and density, define the Effective bulk modulus K (SI unit: Pa) and
Density ρ (SI unit: kg/m3). Selecting User defined is well suited for entering the

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 69

 properties of a user-defined porous material fluid model. Predefined porous models
exist in the Poroacoustics domain feature.

User-Defined Attenuation Fluid Model

The User-defined attenuation model adds a user-defined attenuation to the fluid. This
data is typically based on experimental data for the attenuation coefficient α. Adding
attenuation makes the wave number k complex valued. For example, a plane wave p(x)
moving in the x direction is attenuated according to

ω
k = ---- – iα
c
– ikx – i ( ω ⁄ c )x – αx
p(x) = e = e e

When the attenuation is defined in Np per unit length, the wave has a spatial
exponential decay governed by the attenuation coefficient.

Do not confuse the amplitude attenuation coefficient α with the

intensity attenuation coefficient m most often defined in room
acoustics. The two are simply related through 2α = m. The difference
stems from how the intensity scales in the propagation direction:
I(x) = I0·exp(-2αx) = I0·exp(-mx).

The default Speed of sound c (SI unit: m/s) and Density ρ (SI unit: kg/m3) are taken
From material. For User defined, enter other values or expressions for one or both
options.

Select an Attenuation type: Attenuation coefficient, Np per unit length (the default) to
define an attenuation coefficient α in Np/m (nepers per meter), Attenuation coefficient,
dB per unit length to define an attenuation coefficient α in dB/m (decibel per meter),
or Attenuation coefficient, dB per wavelength to define an attenuation coefficient α in
dB/λ (decibel per wavelength). The last option is often used in underwater acoustics
and uses a slightly different formulation than the first two, see the theory section. For
any selection, enter a value or expression in the Attenuation coefficient α field.

Theory: User-Defined Attenuation Fluid Model

70 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Atmosphere Attenuation Fluid Model
The Atmospheric attenuation model defines attenuation in atmospheric air that follows
the ANSI standard S1.26-2014, see Ref. 6 (Appendix B), 43, 44, and 45 for details.
The model describes attenuation due to thermal and viscous effects (primarily pure
air), the relaxation processes for nitrogen and oxygen, and the dependency on
atmospheric pressure (absolute pressure), temperature, and relative humidity. The
attenuation in air is important for propagation over large distances and for high
frequency processes. This also means that the attenuation effect is more important in
ray tracing simulations where propagation can be simulated over much larger
distances, see The Ray Acoustics Interface.

When the model is selected the Model Input section includes inputs for the Temperature
T (SI unit: K), the Absolute pressure pA (SI unit: Pa), and the Relative humidity φ (SI
unit: 1). In the Pressure Acoustic Model section the default Speed of sound c (SI unit:
m/s) and Density ρ (SI unit: kg/m3) are taken From material. For User defined, enter
other values or expressions for one or both options.

Ocean Attenuation Fluid Model

The Ocean attenuation model defines attenuation in seawater of the ocean. The model
is based on a semianalytical model with parameters that are based on extensive
measurement data. It includes effects due to viscosity in pure water, the relaxation
processes of boric acid and magnesium sulfate, as well as depth, temperature, practical
salinity, and pH value. For further details, see Ref. 46, 47, 48, 49, and 50. The ocean
attenuation model is important in ray tracing simulations where propagation can be
simulated over much larger distances, see The Ray Acoustics Interface.

When the model is selected, the Model Input section includes inputs for the Temperature
T (SI unit: K), the Depth D (SI unit: m), and the Practical salinity Sp (SI unit: 1); the
default is 35. In the Pressure Acoustic Model section, the default Speed of sound c (SI
unit: m/s) and Density ρ (SI unit: kg/m3) are taken From material. For User defined,
enter other values or expressions for one or both options. Enter a value for the pH value
pH (SI unit: 1); the default is 8.

The Practical salinity is defined on the Practical Salinity Scale. It represents

a specific way of measuring salinity through electric conductivity. Practical
salinity is a unit-less quantity, although it can be thought of as given in
units of g/kg. The default value for the Practical salinity Sp is 35. For
details see Ref. 56.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 71

 For a discussion about air and water materials as commonly used in
acoustics, the attenuation in (sea) water and (humid) air, and other
material properties, see the Acoustic Properties of Fluids chapter of this
manual.

Thermally Conducting and/or Viscous Fluid Model

The Viscous, the Thermally conducting, and the Thermally conducting and viscous fluid
models essentially add the same attenuation model. Here, the attenuation is due to
bulk viscous and/or thermal losses. This type of model is relevant in highly viscous
fluids or thermally conducting fluids when acoustic waves are traveling over large
distances (relative to the wavelength). This is not a model for viscous and thermal
boundary layer losses in narrow regions (see Narrow Region Acoustics). The models
define the classical thermoviscous attenuation αtv properties of a fluid and can be
applied in cases when, for example, relaxation processes are not important. The
classical thermoviscous amplitude attenuation is given by

2
ω k(γ – 1)
α tv = ------------3  --- μ + μ B + --------------------
4
 Cp 
2ρc 3

In atmospheric air or salt water where relaxation processes are important use the
Atmosphere attenuation or the Ocean attenuation model as they capture these effects
and also their dependency on, for example, relative humidity or salinity.

For each of the following (when applicable), the default values are taken From material.
For User defined, enter other values or expressions for any or all options.

• Speed of sound c (SI unit: m/s).

• Density ρ (SI unit: kg/m3).
• Heat capacity at constant pressure Cp (SI unit: J/(kg·K)).
• Ratio of specific heats γ (dimensionless).
• Thermal conductivity k (SI unit: W/(m·K)).

72 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• Dynamic viscosity μ (SI unit: Pa·s).
• Bulk viscosity μB (SI unit: Pa·s).

It is possible to assess the magnitude of the losses due to thermal

conduction and viscosity, that is, the power dissipation density
(SI unit: W/m3). This is done during the analysis process by plotting the
variables for:

• Viscous power dissipation density (acpr.diss_visc)

• Thermal power dissipation density (acpr.diss_therm)
• Combined total power dissipation density (acpr.diss_tot)

• Viscous Fluid Model

• Thermally Conducting Fluid Model
• Thermally Conducting and Viscous Fluid Model

General Dissipation Fluid Model

For General dissipation enter the Speed of sound c (SI unit: m/s), the Density ρ (SI unit:
kg/m3), and the Sound diffusivity δ (SI unit: m2/s). By default they are taken From
material and for User defined enter other values or expressions.

The sound diffusivity is automatically calculated for the classical thermoviscous case
using either the Viscous, Thermally conducting or the Thermally conducting and viscous
options (see below). For nonideal fluids like tissue, the sound diffusivity is measured
directly instead and can be entered with the General dissipation.

Ideal Gas Fluid Model

For Ideal Gas, you can edit the Model Inputs section. For each of the following, the
default values are taken From material. For User defined, enter other values or
expressions for any or all options.

• Select a Gas constant type: Specific gas constant Rs (SI unit: J/(kg·K) (the default)
or Mean molar mass Mn (SI unit: kg/mol). For Mean molar mass, the molar gas
constant (universal gas constant) R = 8.314 J/(mol·K) is used as the built-in
physical constant.
• From the Specify Cp or γ list, select Heat capacity at constant pressure Cp (SI unit: J /
(kg·K)) (the default) or Ratio of specific heats γ. For common diatomic gases such as
air, γ = 1.4 is the standard value.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 73

 Poroacoustics
The Poroacoustics node defines a fluid domain with a porous material modeled in a
homogenized way using a so-called equivalent fluid model. Several models exist to
define the attenuation and dispersion experienced by the pressure waves as they
propagate in the porous domain. The different models are described below.

About the Poroacoustics Models

POROACOUSTICS MODEL
Select a Poroacoustics model: Delany-Bazley-Miki (the default), Zwikker-Kosten,
Attenborough, Wilson, Johnson-Champoux-Allard, Johnson-Champoux-Allard-Lafarge,
Johnson-Champoux-Allard-Pride-Lafarge, Three-parameter approximation JCAL model,
Williams EDFM, or Wood.

FLUID PROPERTIES
Enter the properties of the saturating fluid that is inside of the porous material. These
settings are common to most porous models. By default the Fluid material uses the
Domain material.

The following properties are available based on the Poroacoustics model selected above.
The default values are taken From material. For User defined enter a different value or
expression.

• Speed of sound c (SI unit: m/s) of the saturating fluid.

• Density ρf (SI unit: kg/m3) of the saturating fluid (the fluid density).

If any other than the default Delany-Bazley-Miki is selected, the following properties are
also required, depending on the selection. These material parameters are necessary as
the more advanced models include the losses associated with viscosity and thermal
conduction in a more or less detailed way:

• Heat capacity at constant pressure Cp (SI unit: J/(kg·K)).

• Ratio of specific heats γ (dimensionless).
• Thermal conductivity k (SI unit: W/(m·K)).

74 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• Dynamic viscosity μ (SI unit: Pa·s).
• Fluid bulk modulus Kf (SI unit: Pa).

The implemented poroacoustics models are all applicable for any fluid,
except the empirically determined Delany-Bazley-Miki model, which
applies only to air. See Ideal Gas and General Fluid Options for details.

POROUS MATRIX PROPERTIES

In this section, enter the properties that describe the porous material. By default the
Porous elastic material uses the Domain material (the material defined for the domain).
Select another material as needed. For example, create your own material that contains
the properties of a given porous material and refer to it here. Here you also select if
you want to use a rigid frame approximation or a limp frame approximation for the
porous material (where applicable). The poroacoustic model defines complex-valued
frequency-dependent expressions for both the bulk modulus and the equivalent
density.

Based on the Poroacoustics model selected, enter the following settings for the porous
matrix.

• Delany-Bazley-Miki • Johnson-Champoux-Allard-Pride-Lafarg
• Zwikker-Kosten e

• Attenborough • Three-Parameter Approximation JCAL

Model
• Wilson
• Williams Effective Density Fluid Model
• Johnson-Champoux-Allard
(EDFM)
• Johnson-Champoux-Allard-Lafarge
• Wood

Delany-Bazley-Miki
The Delany-Bazley-Miki model is an empirical model used to describe fibrous
materials such as rockwool or glass fiber. The model can be used for materials with a
porosity, εp, close to one. For Delany-Bazley-Miki, the Flow resistivity Rf (SI unit: Pa·s/
m2) uses values From material. For User defined, enter a value or expression.

Select an option from the Constants list: Delany-Bazley (the default), Miki, Qunli,
Mechel, glass fiber, low X, Mechel, glass fiber, high X, Mechel, rock fiber, low X,
Mechel, rock fiber, high X, Komatsu, Modified Allard and Champoux, or User defined. For

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 75

 User defined enter values in the C1 to C8 fields. The models are empirical and based on
fitting to measured data. This means the models have different regions of applicability
with respect to the flow resistivity Rf, the frequency f, the material type, and the
parameter X defined as

f ⋅ ρf
X = -----------
Rf

All the models are applicable for materials with a porosity εp close to 1. The model
informations are listed in Table 2-2. See also Ref. 23 and the relevant section in About
the Poroacoustics Models for further details.

76 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

TABLE 2-2: DELANY-BAZLEY-MIKI MODEL OPTIONS AND APPLICABILITY

MODEL NAME APPLICABILITY

Delany-Bazley Glass and rock wool with:

0.01 ≤ X ≤ 1
3 3 –2
10 ≤ R f ≤ 50 ⋅ 10 Pasm

Miki Glass and rock wool with:

0.01 ≤ X ≤ 1
3 3 –2
10 ≤ R f ≤ 50 ⋅ 10 Pasm

Qunli Porous plastic and open foams:

200 ≤ f ≤ 2000 Hz
3 3 –2
3 ⋅ 10 ≤ R f ≤ 24 ⋅ 10 Pasm

Mechel, glass fiber, low X Glass fiber: X ≤ 0.025

Mechel, glass fiber, high X Glass fiber: X ≥ 0.025
Mechel, rock fiber, low X Rock fiber: X ≤ 0.025
Mechel, rock fiber, high X Rock fiber: X ≥ 0.025
Komatsu Glass and rock wool:

3 3 –2
6 ⋅ 10 ≤ R f ≤ 73 ⋅ 10 Pasm

Modified Allard and Champoux 3

45 ≤ f ≤ 11 ⋅ 10 Hz

Flow resistivity Rf is easy to measure and is independent of frequency.

Absorptive Muffler: Application Library path Acoustics_Module/

Automotive/absorptive_muffler

Zwikker-Kosten
The Zwikker-Kosten model is a two-parameter semiempirical model. It is one of the
earliest equivalent fluid models for porous materials. The model assumes that the pores

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 77

 are cylinder-like with an effective hydraulic radius Hr. See Ref. 15 and About the
Poroacoustics Models for further details.

For Zwikker-Kosten, select a Porous matrix approximation: Rigid (the default) or Limp.
Then based on your choice, the default value for each of the following parameters is
taken From material. For User defined, enter another value or expression.

• Porosity εp (dimensionless).
• Hydraulic radius of pores Hr (SI unit: m).
• Drained density of porous material ρd (SI unit: kg/m3).

Attenborough
The Attenborough model is also based on the cylindrical-like pore assumption. It is a
so-called four-parameter semiempirical model. The model is an extension of the
Zwikker-Kosten model and has two additional input parameters. It accounts for the
tortuosity (high frequency limit) τ∞, which is related to the orientation of the pores
relative to the propagation direction. The hydraulic diameter of the pores is replaced
by an expression that included the flow resistivity Rf and a fitting parameter b (this
parameter is related to the anisotropy of the pore distribution). See Ref. 9, Ref. 16, and
About the Poroacoustics Models for details.

For Attenborough, select a Porous matrix approximation: Rigid (the default) or Limp.
Then based on your choice, the default value for most of the following parameters is
taken From material. For User defined, enter another value or expression.

• Porosity εp (dimensionless).
• Flow resistivity Rf (SI unit: Pa·s/m2).
• Tortuosity factor (high frequency limit) τ∞ (dimensionless). The default is 1.
• Fitting parameter b (dimensionless). The default is 1.
• Drained density of porous material ρd (SI unit: kg/m3).

Wilson
The Wilson model is a generalization of the semianalytical models for porous materials
with constant cross section and parallel pores. This model is intended to match the
middle frequency behavior of a porous material. See Ref. 17, Ref. 9, Ref. 18, and
About the Poroacoustics Models for further details.

78 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

For Wilson, select a Porous matrix approximation: Rigid (the default) or Limp. Then
based on your choice, the default value for the following parameters is taken From
material. For User defined, enter another value or expression.

• Porosity εp (dimensionless).
• Bulk modulus infinite frequency limit K ∞ (SI unit: Pa).
• Density infinite frequency limit ρ ∞ (SI unit: kg/m3).
• Entropy-mode relaxation time τent (SI unit: s).
• Vorticity-mode relaxation time τvor (SI unit: s).

Approximate expressions for the two relaxation times that are based on
intrinsic material properties are given in the theory section for the Wilson
model.

Johnson-Champoux-Allard
The Johnson-Champoux-Allard (or JCA model) is a five-parameter semiempirical
model for describing a large range of porous materials with rigid (or limp) frames. As
input, the model requires the viscous Lv and thermal Lth characteristic lengths
(sometimes denoted Λ and Λ'). These replace the hydraulic radius used in simpler
models and account for the thermal and viscous losses that appear at the acoustic
boundary layer at pore walls. See About the Poroacoustics Models for details.

For Johnson-Champoux-Allard, select a Porous matrix approximation: Rigid (the default)

or Limp. Then based on your choice, the default value for most of the following
parameters is taken From material. For User defined, enter another value or expression.

• Porosity εp (dimensionless).
• Flow resistivity Rf (SI unit: Pa·s/m2).
• From the Specify list:
- Select Viscous characteristic length (the default) to directly enter an expression for
Lv (SI unit: m). The default expression is sqrt([Link]*[Link]*8/
([Link]*acpr.epsilon_p)), which corresponds to s = 1.

- Select Viscous characteristic length parameter to enter a value for s (dimensionless)

(the default is 1).
• Thermal characteristic length Lth (SI unit: m). The default expression is 2*[Link].

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 79

 • Tortuosity factor (high frequency limit) τ∞ (dimensionless). The default is 1.
• Drained density of porous material ρd (SI unit: kg/m3).

The Johnson-Champoux-Allard (JCA) model with a rigid (motionless)

frame (skeleton) is sometimes referred to as a Biot equivalent fluid model
with a rigid porous matrix.

Johnson-Champoux-Allard-Lafarge
The Johnson-Champoux-Allard-Lafarge (or JCAL model) is an extension to the JCA
model. It introduces corrections to the bulk modulus thermal behavior at low
frequencies which is not captured by the JCA model. It introduces the static thermal
permeability parameter κ' 0 and, thus, has one more parameter than the JCA model.
It is a six-parameter semiempirical model. See Ref. 13, Ref. 9, and About the
Poroacoustics Models for further details.

For Johnson-Champoux-Allard-Lafarge, the settings are the same as for

Johnson-Champoux-Allard with the addition of the Static thermal permeability κ' 0
(SI unit: m2) setting.

Johnson-Champoux-Allard-Pride-Lafarge
The Johnson-Champoux-Allard-Pride-Lafarge (or JCAPL model) further extends the
JCAL models by introducing a static viscous τ0 and thermal τ'0 tortuosity.

For Johnson-Champoux-Allard-Pride-Lafarge, the settings are the same as for

Johnson-Champoux-Allard plus these additional parameters:

• Static viscous tortuosity τ0 (dimensionless). The default is 1.

• Static thermal tortuosity τ'0 (dimensionless). The default is 1.
• Static thermal permeability κ' 0 (SI unit: m2).

Three-Parameter Approximation JCAL Model

The Three-parameter approximation JCAL model represents an approximation to the
JCAL model. Instead of requiring 6 parameters as input for the porous matrix it only
requires three. The porosity as well as two parameters that relate to the topology of the
pores are needed (their size and distribution). Two different approximation methods
exist for the density and bulk modulus. Select the Density and bulk modulus
approximation to either Johnson-Champoux-Allard-Pride-Lafarge (the default) or Padé
approximation. Then if necessary enter the values for (default is From material):

• Porosity εp (dimensionless).

80 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• Median pore size s (SI unit: m).
• Standard deviation in pore size distribution σs (dimensionless).

Details about the model can be found in Ref. 57, 58, 59, 60, and 61.

Williams Effective Density Fluid Model (EDFM)

Williams EDFM is an effective density fluid model (EDFM) used to model the
propagation of acoustic waves in sediments. The model assumes that the bulk and
shear moduli of the frame of the porous material are negligible, meaning that the
porous frame is limp. See Ref. 24 and About the Poroacoustics Models for further
details.

The Porous Matrix Properties (the properties for the sediment) for Williams EDFM have
the default values taken From material. For User defined, enter another value or
expression:

• Grain density ρgr (SI unit: kg/m3).

• Grain bulk modulus Kgr (SI unit: Pa).
• Porosity εp (dimensionless).
• Tortuosity factor (high frequency limit) τ∞ (dimensionless). The default is 1.
• Permeability κp (SI unit: m2).
• Hydraulic radius of pores Hr (SI unit: m).

Wood
The Wood model is used for a fluid mixture or a fluid suspension (solid inclusions
completely surrounded by fluid). The Woods formula for the sound velocity is
determined by using the effective bulk modulus of the suspension and the volume
average density. As the Williams EDFM, this model gives effective values for the
mixture. This model is exact for low frequencies, when the wavelength is much larger
than the size of the inclusions. See Ref. 25 and About the Poroacoustics Models for
further details.

The Porous Matrix Properties (the properties for the inclusions) for the Wood model are
entered in the Inclusion properties table. For each Inclusion (any number of inclusions
can be added, the number is automatically incremented) enter the Volume fraction, the
Bulk modulus, and the Density. Add a new row (inclusion) by clicking the plus sign
below the table. The table may be saved or loaded from a file.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 81

 Narrow Region Acoustics
The Narrow Region Acoustics node defines a fluid model for viscous and thermal
boundary-layer-induced losses in channels and ducts of constant cross section. The
losses due to the viscous and thermal dissipation in the acoustic boundary layer are
homogenized and smeared on the fluid. This equivalent-fluid model can be used in
long tubes of constant cross section (or only in a slowly varying cross section) instead
of a fully detailed thermoviscous acoustics model. This type of model has a low
computational cost compared to a thermoviscous acoustic model. The models are
applicable for all fluids, that is, both gases and liquids. In the case of waveguides of
varying thicknesses or in general other structures with curved surfaces, an alternative
is to use the Thermoviscous Boundary Layer Impedance condition.

About the Narrow Region Acoustics Models

DUCT PROPERTIES
Select a Duct type: Wide duct approximation (the default), Very narrow circular duct
(isothermal), Slit, Circular duct, Rectangular duct, Equilateral triangular duct, or User
defined.

The Slit, Circular duct, Rectangular duct, and Equilateral triangular duct
models are applicable as long as the cross-section dimension is much
smaller than the wavelength and the boundary layer thickness is smaller
than the wavelength. The cross-section parameter can be a slowly varying
function of space. These are known as a low reduced frequency (LRF)
model.

• For Wide duct approximation enter a Hydraulic diameter Hd (SI unit: m).
This model can be used in ducts of any cross section as long as the hydraulic
diameter of the duct (four times the cross-section area divided by the circumference)
is much larger than the viscous boundary layer thickness.
• For Very narrow circular duct (isothermal) enter the duct Radius a (SI unit: m).
This model is only valid for very narrow circular ducts where isothermal conditions
apply. The radius of the duct has to be much smaller than the thickness of the
thermal boundary layer. For this model it is assumed that the compressibility (bulk
modulus) of the fluid also takes the isothermal value.

82 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• For Slit enter the slit Height h (SI unit: m). Use this model in narrow slit domains to
include the damping and attenuation that occurs here because of the losses in the
viscous and thermal boundary layer.

h
• For Circular duct enter the duct Radius a (SI unit: m). This model is useful for
modeling the damping and attenuation that occurs when acoustic waves propagate
in all tubing systems of small cross-section dimensions.

• For Rectangular duct enter the duct Side lengths W and H (SI unit: m). Use this
model for waveguides and ducts with a rectangular cross section. Also see Advanced
Physics Options for additional settings.

W
• For Equilateral triangular duct enter the duct Side length d (SI unit: m). Use this
model for waveguides and ducts with an equilateral triangular cross section.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 83

 To determine the complex propagation constants for a waveguide, of
arbitrary cross section, use The Thermoviscous Acoustics, Boundary
Mode Interface. Apply it on the cross-section geometry of the waveguide.
The interface solves for the propagating modes and includes all losses in
detail. The complex wave number kc is then given by the plane wave
mode solved for. This is the variable [Link]. The predefined variable
[Link] gives the (lumped) specific characteristic complex impedance
Zc. Search for the mode nearest to the (lossless) plane wave mode.

FLUID PROPERTIES
Based on your selection in Duct Properties, different fluid properties need to be
entered.

For User defined, enter the value for the Complex wave number kc (SI unit: rad/m) and
(specific characteristic) Complex acoustic impedance Zc (SI unit: Pa·s/m). These values
can be used to model the propagation in ducts of arbitrary cross sections.

For the other duct types, several other fluid properties are needed. The default values
for the following are taken From material.

• Speed of sound c (SI unit: m/s)

• Density ρ (SI unit: kg/m3)
• Ratio of specific heats γ (SI unit: 1). In many liquids the value of γ is close to 1, the
2 –1
exact value can be derived from the expression ( 1 – ( α p T ) ⁄ ( β T ρC p ) ) where the
(isobaric) coefficient of thermal expansion αp and the isothermal compressibility βT
is used.
• Dynamic viscosity μ (SI unit: Pa·s)

The following are available for Wide duct approximation, Slit, Circular duct,
Rectangular duct, and Equilateral triangular duct:
• Heat capacity at constant pressure Cp (SI unit: J/(kg·K))
• Thermal conductivity k (SI unit: W/(m·K))

ADVANCED PHYSICS OPTIONS

To display this section, click the Show More Options button ( ), select Advanced
Physics Options and choose Rectangular duct as the Duct type.

84 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The Number of terms for the sum N (used to describe the rectangular duct model) can
be set. The default is N = 100 and should cover most cases, see under Slits, Circular
Ducts, Rectangular Ducts, and Equilateral triangular Ducts.

• Lumped Receiver Connected to Test Setup with a 0.4-cc Coupler: The

Application Library path Acoustics_Module/Electroacoustic_Transducers/
lumped_receiver_04cc
• Generic 711 Coupler — An Occluded Ear-Canal Simulator: The
Application Library path Acoustics_Module/
Tutorials,_Thermoviscous_Acoustics/generic_711_coupler

The Narrow Region Acoustics models are so-called equivalent fluid models
that have a non-trivial (nonlinear) dependency on the frequency. This
means that performing an eigenfrequency analysis should be considered
carefully, see Eigenfrequency Study. Moreover, the linearization process
(linearization with respect to the frequency) of the underlying
mathematical models can cause numerical problems (an error message is
thrown). For the Rectangular duct one remedy is to decrease the number
of terms used in the sum, for the Circular duct it can be necessary to switch
to another model, like the Very narrow circular duct (isothermal).

Anisotropic Acoustics
The Anisotropic Acoustics node is used to define a fluid with properties that can be
modeled with an effective anisotropic density. The feature can be used to model
metamaterials in a homogenized manner or when modeling fibrous materials with
anisotropic properties. The material properties are defined through an effective bulk
modulus K and an effective density tensor ρ. The density variable can be defined as
being isotropic (a scalar), a diagonal tensor, or a symmetric tensor. The resulting
(generalized) Helmholtz equation solved in the frequency domain is then

2
–1 ω pt
∇ ⋅ ( – ρ ( ∇p t – q d ) ) – ------------ = Q m
K
–1
where it should been emphasized that ρ is here a tensor. The resulting system will
have principal speeds of sound that are (square root of) the eigenvalues and directions
2 –1
that are the eigenvectors of the quantity c = ρ K .

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 85

 COORDINATE SYSTEM SELECTION
Select the Coordinate system in which the density tensor is defined. The default is the
Global coordinate system.

Coordinate systems with directions which change with time should not
be used.

ANISOTROPIC ACOUSTICS
Enter the Effective bulk modulus K (SI unit: Pa) taken From material. For User defined
enter other values or expressions for any or all options.

Enter the Effective density ρ (SI unit: kg/m3) taken From material. For User defined
enter other values or expressions for any or all options. Also select the representation
for the density tensor either Isotropic (the default), Diagonal, or Symmetric.

Acoustic Cloaking: The Application Library path Acoustics_Module/

Tutorials,_Pressure_Acoustics/acoustic_cloaking

Background Pressure Field

Add a Background Pressure Field node to model a background/incident pressure wave
to study the scattered pressure field ps, which is defined as the difference between the
total acoustic pressure pt and the background pressure field pb:

pt = pb + ps

This feature sets up the equations in a so-called scattered field formulation where the
dependent variable is the scattered field p = ps. In a model where the background
pressure field is not defined in all acoustic domains (or it is different), continuity is
automatically applied in the total field pt on interior boundaries between domains.

The background pressure field has built-in options for plane waves, spherical waves,
and cylindrical waves. It can also be defined by a user defined analytical expression, that
is a function of space, or it can be given by the solution to another acoustics problem.

For a Frequency Domain study type, the frequency of the background

pressure field is the same as for the dependent variable p.

86 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 For exciting waveguide structures with a given wave type it is
recommended to use the Port or Lumped Port conditions, instead of a
background pressure field. If still used, the feature should be backed by a
PML or radiation condition to model an infinite waveguide.

BACKGROUND PRESSURE FIELD

Select a Background pressure field type: Plane wave (the default), Cylindrical Wave,
Spherical Wave, or User defined. For 2D axisymmetric models, it is possible to expand
the plane wave in their cylindrical harmonics.

To set up and calculate the intensity variables for both the total, background, and
scattered acoustic fields select Calculate background and scattered field intensity at the
bottom of the settings window. This requires an additional input of the Density ρ (SI
unit: kg/m3) for the defined background pressure field. Selecting this option will also
define the velocity variables for the various fields.

Plane Wave
The Plane wave option defines the background pressure field pb of the type:

x⋅e
– i k s  -------------k
–i ( k ⋅ x ) ek
pb = p0 e = p0 e

where p0 is the wave amplitude, k is the wave vector with amplitude ks = ω/c and wave
direction vector ek, and x is the location on the boundary.

For Plane wave enter values for the Pressure amplitude p0 (SI unit: Pa), the Speed of
sound c (SI unit: m/s) of the fluid defining the field, either From material or User
defined, enter a Wave direction ek (dimensionless), and a Phase φ (SI unit: rad).

In 2D axisymmetric models, the incident fields take a slightly different form due to the
geometrical restrictions. Only the z-component for the Wave direction ek can be
entered. Per default, the wave can only travel in the axial direction since this is the only
axisymmetric form of a plane wave.

By selecting Enable plane wave expansion, the plane wave is expanded in its cylindrical
harmonics and a general Wave direction ek can be selected. This sets up a wave of the
form:

( x ⋅ ( e k ⋅ e r )e r ) ( x ⋅ ( e k ⋅ e z )e z )
i J m  – k s ------------------------------------- exp  – ik s -------------------------------------
iφ sign(m) m
pi = e 2
 ek   ek 

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 87

 where m is the Azimuthal mode number specified in the Pressure Acoustics Equation
Settings. To expand the solution, it is necessary to run a Parametric Sweep over the
mode number from 0 up to the desired resolution.

An example using the Enable plane wave expansion functionality can be

found online in the Application Gallery. The model Plane Wave
Scattering off a 2D Axisymmetric Object: Plane Wave Expansion
Approach is found at [Link]/model/51311

Cylindrical Wave
The Cylindrical Wave option defines the background pressure field pb as a predefined
cylindrical wave:

iφ (2) ( x – x 0 ) × e sa
pb = p0 e H0 ( ks rs ) r s = -----------------------------------
e sa

where p0 is the amplitude given at the reference distance rref = 0.548/ks (the distance
(2)
where the Hankel function is one), ks = ω/c is the wave number, H m is the Hankel
function of the second kind (representing an outgoing cylindrical wave), rs is the
distance from the source axis, esa is the direction of the source axis, x0 is a point on
the source axis, and x is the location on the boundary.

For Cylindrical Wave, enter a Pressure amplitude at reference distance p0 (SI unit: Pa),
the Speed of Sound c (SI unit: m/s) of the fluid defining the field, either From material
or User defined, enter a Source Location x0 (SI unit: m), a Source Axis esa (the vector
does not need to be normalized), and a Phase φ (SI unit: rad).

In 2D axisymmetric models, the incident fields take a slightly different form due to the
geometrical restrictions. Neither source location nor axis is needed as the source is
always located on the z-axis.

Spherical Wave
The Spherical Wave option defines the background pressure field pb as a predefined
spherical wave (only for 2D axisymmetric and 3D):

iφ r ref – ik r
p b = p 0 e -------- e s s r ref = 1 m rs = x – x0
rs

where p0 is the amplitude given at the reference distance of 1 m, ks = ω/c is the wave
number, rs is the distance from the source, x0 is the source location of the spherical
wave, and x is the location on the boundary.

88 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

For Spherical Wave, enter a Pressure amplitude at reference distance p0 (SI unit: Pa), the
Speed of Sound c (SI unit: m/s) of the fluid defining the field, either From material or
User defined, enter a Source Location x0 (SI unit: m), and a Phase φ (SI unit: rad).

In 2D axisymmetric models, the incident fields take a slightly different form due to the
geometrical restrictions. Only enter the Source Location, z coordinate z0 (SI unit: m)
since the source is always located on the z-axis in a 2D axisymmetric model.

For both the cylindrical and the spherical wave options the source and/or
source axis should be located outside the domain where the background
pressure field feature is defined.

User Defined
For the User defined option, enter an expression for the Background pressure field pb
(SI unit: Pa).

• Acoustic Cloaking: Application Library path Acoustics_Module/

Tutorials,_Pressure_Acoustics/acoustic_cloaking
• Acoustic Scattering off an Ellipsoid: Application Library path
Acoustics_Module/Tutorials,_Pressure_Acoustics/acoustic_scattering

Initial Values
The Initial Values node adds an initial value for the pressure. In the time domain, it also
adds an initial value for the pressure time derivative. The initial values can serve as an
initial guess for a nonlinear solver or describe the initial state that will then evolve in
time. If more than one initial value is needed, from the Physics toolbar, click to add
more Initial Values nodes.

INITIAL VALUES
Enter a value or expression for the initial values for the Pressure p (SI unit: Pa) and, in
the time domain, also for the Pressure, first time derivative, ∂p/∂t (SI unit: Pa/s).

Monopole Domain Source

Use the Monopole Domain Source node to add a domain source term Qm to the
governing equation. A monopole domain source added to a domain has a uniform
strength in all directions. In advanced models, this source term can, for example, be

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 89

 used to represent a domain heat source causing pressure variations or a user defined
nonlinear contribution to the equations in the time domain. Add this node from the
More submenu.

MONOPOLE SOURCE
Enter a Monopole domain source Qm (SI unit: 1/s2).

Dipole Domain Source

Use the Dipole Domain Source node to add the domain source term qd to the governing
equation. This source represents a domain volumetric force. This source is typically
stronger in two opposite directions. In advanced models, this term can, for example,
be used to represent a uniform constant background flow convecting the sound field.
Add this node from the More submenu.

DIPOLE SOURCE
Enter coordinates for the Dipole domain source qd (SI unit: N/m3). These are the
individual components of the dipole source vector.

Heat Source
Use the Heat Source feature to add a domain heat source that generates sound. If a fluid
is heated and cooled rapidly, thermal expansion and contraction will generate acoustic
waves. This could be a source representing a pulsating laser beam or a flame source in
a combustion simulation. The feature adds a monopole domain source to the
right-hand side defined as

αp
Q m = ---------- iωQ heat (frequency domain)
ρC p 1 Cp ( γ – 1 )
α p = --- -----------------------
-
α p ∂Q heat c T
Qm = ---------- ---------------- (time domain)
ρC p ∂t

where the (isobaric) coefficient of thermal expansion αp is defined in terms of the

speed of sound c, heat capacity at constant pressure Cp, ambient temperature T (this
value is taken from the model inputs and can be space dependent), and ratio of specific
heats γ.

DOMAIN HEAT SOURCE

Specify the Heat source strength Qheat (SI unit: W/m3).

90 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

FLUID PROPERTIES
Enter the additional fluid properties necessary to define the heat source. Enter the Heat
capacity at constant pressure Cp (SI unit: J/(kg·K)) and the Ratio of specific heats γ
(dimensionless). The default for both is From material. Select User defined to enter a
value or expression directly.

Aeroacoustic Flow Source

The Aeroacoustic Flow Source feature is used to set up aeroacoustic flow-induced noise
sources in the pressure acoustics model. The feature is best used when coupling an LES
fluid flow model to the Aeroacoustic Flow Source feature using the Aeroacoustic Flow
Source Coupling multiphysics coupling and the dedicated Transient Mapping study.
The feature only exists in 3D. The feature can also be used with user-defined
expressions for a Lighthill like source expressed in terms of a stress tensor Tij (SI unit:
N/m2).

AEROACOUSTIC FLOW SOURCE

Select the Acoustic analogy as Lighthill (the default) or Aeroacoustic wave equation. The
selection should match the option in the Aeroacoustic Flow Source Coupling
multiphysics coupling. When the multiphysics coupling is used, the Stress tensor (T),
Normal mass flow on exterior boundaries (mn), or the Flow pressure variation, second
var
time derivative ( p tt ) field are automatically coupled. If user-defined expressions are
selected, enter an expression. The user-defined expressions correspond to the Fourier
components of the different flow sources terms.

Tandem Cylinder Flow Noise tutorial available in the COMSOL

Application Gallery at:

• [Link]

Best practices for modeling with the Aeroacoustic Flow Source and the
associated multiphysics functionality is presented in the Modeling with
the Aeroacoustic Flow Source: Flow-Induced Noise section.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 91

 Sound Hard Boundary (Wall)
The Sound Hard Boundary (Wall) adds a boundary condition for a sound hard boundary
or wall, which is a boundary at which the normal component of the acceleration (and
thus the velocity) is zero:

– n ⋅  – ----- ( ∇p t – q d ) = 0
1
 ρc 

For zero dipole domain source (qd = 0) and constant fluid density ρc, this means that
the normal derivative of the pressure is zero at the boundary:

∂p t
= 0
∂n

Sound-hard boundaries are available for all study types. Note that mathematically this
condition is identical to the Symmetry condition.

Axial Symmetry
The Axial Symmetry node is a default node added for all 2D and 1D axisymmetric
components. The boundary condition is active on all boundaries on the symmetry axis.

Normal Acceleration
The Normal Acceleration adds an inward normal acceleration an:

– n ⋅  – ----- ( ∇p t – q d ) = a n
1
ρc

Alternatively, specify the acceleration a0 of the boundary. The part in the normal
direction is used to define the boundary condition:

– n ⋅  – ----- ( ∇p t – q d ) = – n ⋅ a 0
1
 ρc 

This feature represents an external source term. It can also be used to manually couple
acoustics with a structural analysis for modeling acoustic-structure interaction.

92 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

NORMAL ACCELERATION
Select a Type: Inward Acceleration (the default) or Acceleration.

• For Inward Acceleration enter the value of the Inward acceleration an (SI unit: m/s2).
Use a positive value for inward acceleration or a negative value for outward
acceleration.
• For Acceleration enter values for the components of the Acceleration a0 (SI unit: m/
s2).

Normal Velocity
The Normal Velocity adds an inward normal velocity vn, which is assumed to be
harmonically oscillating:

– n ⋅  – ----- ( ∇p t – q d ) = iωv n
1
 ρc 

Alternatively, specify the acceleration v0 of the boundary. The part in the normal
direction is used to define the boundary condition:

– n ⋅  – ----- ( ∇p t – q d ) = – iωn ⋅ v 0

1
ρc

This feature represents an external source term. It can also be used to manually couple
acoustics with a structural analysis for modeling acoustic-structure interaction.

NORMAL VELOCITY
Select a Type: Inward Velocity (the default) or Velocity.

• For Inward Velocity enter the value of the Inward velocity vn (SI unit: m/s). Use a
positive value for inward velocity or a negative value for outward velocity.
• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s).

Normal Displacement
The Normal Displacement adds an inward normal displacement dn, which is assumed to
be harmonically oscillating:

– n ⋅  – ----- ( ∇p t – q d ) = ( iω ) d n
1 2
ρc

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 93

 Alternatively, specify the displacement d0 of the boundary. The part in the normal
direction is used to define the boundary condition:

– n ⋅  – ----- ( ∇p t – q d ) = – ( iω ) n ⋅ d 0
1 2
 ρc 

This feature represents an external source term. It can also be used to manually couple
acoustics with a structural analysis for modeling acoustic-structure interaction in
vibroacoustic problems.

NORMAL DISPLACEMENT
Select a Type: Inward Displacement (the default) or Displacement.

• For Inward Displacement enter the value of the Inward displacement dn (SI unit: m).
Use a positive value for inward displacement or a negative value for outward
displacement.
• For Displacement enter values for the components of the Displacement d0 (SI unit:
m).

Sound Soft Boundary

The Sound Soft Boundary adds a boundary condition for a sound soft boundary, where
the acoustic pressure vanishes: pt = 0. It is an appropriate approximation for a
liquid-gas interface and in some cases for external waveguide ports. For modeling
detailed port behavior use the Port condition.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Pressure
The Pressure node creates a boundary condition that acts as a pressure source at the
boundary, which means that a constant acoustic pressure p0 is specified and maintained
at the boundary: pt = p0. In the frequency domain, p0 is the amplitude of a harmonic
pressure source. The node is also available from the Pairs submenu as an option at
interfaces between parts in an assembly.

PRESSURE
Enter the value of the Pressure p0 (SI unit: Pa).

94 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Impedance
The Impedance node adds an impedance boundary condition with the option to select
between several built-in impedance models and engineering relations. The impedance
condition is a generalization of the sound-hard and sound-soft boundary conditions:

iωp t
– n ⋅  – ----- ( ∇p t – q d ) = – -----------
1
 ρc  Zi

In the Pressure Acoustics, Transient interface using a Time Dependent study, the
impedance boundary condition is the following:

1 ∂p t
– n ⋅  – --- ( ∇p t – q d ) = ----- --------
1
 ρ  Z i ∂t

Here, Zi is the specific acoustic input impedance of the external domain and it has the
SI unit Pa·s/m — a pressure divided by a velocity. From a physical point of view, the
acoustic input impedance is the ratio between the local pressure and local normal
particle velocity. The Impedance boundary condition is a good approximation of a
locally reacting surface — a surface for which the normal velocity at any point depends
only on the pressure at that exact point.

For plane waves, the specific acoustic impedance Zi is related to the acoustic impedance
Zac (ratio of average surface pressure and flow rate) and the mechanical impedance
Zmech (ratio of force and velocity) via the area A of the boundary, according to

2
Z mech = AZ i = A Z ac

This relation only defines the relation between average or lumped impedance values at
a surface.

All built-in impedance models only exist in the frequency domain. The only exception
is the User defined impedance, which can be applied also in the time domain. In the
frequency domain, the impedance can be any complex-valued number and thus models

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 95

 a surface that is both resistive and reactive. In the time domain, only resistive
(real-valued impedance) effects can be included.

In the two opposite limits Zi → ∞ and Zi → 0, this boundary condition is

identical to the Sound Hard Boundary (Wall) condition and the Sound
Soft Boundary condition, respectively.

COMSOL Multiphysics supports the use of the rayl unit when specifying
a value of the impedance (1 rayl = 1 Pa·s/m). Two variants exist: [rayl]
and [rayl_cgs]. The latter is the definition of the unit in the cgs unit
system. Notice that inconsistent definitions give: 1[rayl_cgs] =
10[rayl] = 10[Pa*s/m].

On an impedance boundary, postprocessing variables exist to evaluate the

equivalent absorption coefficient. They are the absorption coefficient for
normal incidence acpr.imp1.alpha_n and for random incidence
acpr.imp1.alpha_ran (here given for the impedance feature imp1).

Notice that the Impedance boundary condition cannot directly be

combined with a source like Normal Acceleration. In cases where such a
behavior is desired, modeling a source impedance, this can be achieved by
coupling the boundary to an Electrical Circuit model. See, for example,
the Lumped Loudspeaker Driver model under Electroacoustic
Transducers in the Application Library.

IMPEDANCE
A number of different types of impedance boundary conditions are included to address
standard situations in many typical applications of pressure acoustics:

• Simple RCL models define lumped-parameter circuits describing, for example, the
mechanical response of a microphone diaphragm. This model is only fully valid for
plane waves at normal incidence.
• A set of physiological models enable easy modeling of the surface impedance of
human skin and the acoustic loads of the human ear and eardrum. The ear and
eardrum models are based on lumped models and only really valid for plane waves
at normal incidence. They can be used as engineering approximations in other cases.

96 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• The waveguide end impedances define simplified models of the acoustic losses at the
end of pipes and ducts. Like for the RCL impedance, this model is only valid for
plane waves at normal incidence at the end of waveguides. It is not valid when
higher order modes start to propagate.
• A porous layer backed by a sound-hard wall model can be defined.
• Simple wave-type specific characteristic impedance models exist to address open
boundaries in a simplified way.
• An option to define the impedance through the absorption coefficient of the
surface.

The ear impedance, skin impedance, and RCL models provide tools for engineers to
add realistic acoustic loads when, for example, developing and simulating headphones,
hearing aids, headsets, and other mobile devices.

The theory for the impedance models is discussed in detail in Theory for
the Boundary Impedance Models.

Choose an Impedance model — User defined (the default), RCL, Physiological, Waveguide
end impedance, Porous layer, Characteristic specific impedance, or Absorption Coefficient.

User Defined
Allows the user to enter any expression and is the only impedance model that applies
to time-dependent models. It is advantageous to enter complicated user-defined
models as a variable under the Definitions node or use an interpolation function for
measured data.

Enter the value of the Impedance Zi (SI unit: Pa·s/m). The default value is set to the
characteristic-specific impedance of air: 1.2 kg/m3·343 m/s.

RCL
The RCL model includes all possible circuits involving a source of damping (a resistor
Rac), an acoustic mass or inertance (an inductor Lac), and a source of acoustic
compliance (a capacitor Cac). The circuit elements are entered in acoustic units. These
can be used as a simple model of, for example, the input impedance of a microphone,
a loudspeaker cone, or other electromechanical applications. Other applications
include general transmission line/circuit models with applications in materials with
exotic acoustic properties. More advanced circuit models may be entered manually in
the User defined option or by coupling to an Electric Circuit model (this requires the
AC/DC Module).

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 97

 Choose an option from the list: Serial coupling RCL, Parallel coupling RCL, Parallel LC in
series with R, Parallel RC in series with L, Parallel RL in series with C, Serial RC in parallel
with L, Serial LC in parallel with R, or Serial RL in parallel with C.

Notice the matching diagram and Equation section information for each choice. Then
enter the following:

• Equivalent acoustic resistance Rac (SI unit: kg/(m4·s)).

• Equivalent acoustic compliance Cac (SI unit: m4·s2/kg).
• Equivalent acoustic inertance Lac (SI unit: kg/m4).

Generic 711 Coupler — An Occluded Ear-Canal Simulator:

Application Library path Acoustics_Module/
Tutorials,_Thermoviscous_Acoustics/generic_711_coupler

Physiological
This is a set of simple models to address applications involving interactions of acoustics
with the human body. The models comprise human skin, the impedance of the entire
human ear including or excluding the pinna, the outward radiation impedance caused
by the pinna, and the inward impedance experienced at the ear drum comprising the
drum and the entire inner ear. For the two models of the human ear (with/without
pinna), the pressure at the ear drum is automatically calculated. The variable has the
form acpr.imp1.p_ear_drum and is available for postprocessing.

The whole-ear models are based on the geometry of the ear canal and pinna of a
specific ear (see Ref. 28-30), but person-to-person variations are to be expected. For
applications where a specific ear canal geometry can be obtained, better results are
expected by explicitly modeling this and applying the eardrum impedance at the end.

Choose an option from the list: Human skin, Outward human ear radiation, Human ear
drum, Human ear without pinna, or Human ear, full. Then select either From material (the
default) or User defined for the following, as required:

• Ratio of specific heats γ (SI unit: 1).

• Heat capacity at constant pressure Cp (SI unit: J/(kg·K)).
• Thermal conductivity k (SI unit: W/(m·K)).
• Dynamic viscosity μ (SI unit: Pa·s).

98 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

When the From material option is selected, remember to add a material under the
Materials node and assign it to the specific boundary. The boundary will not
automatically assume the physical properties of the domain.

Waveguide End Impedance

This is a set of idealized models for the acoustic losses at the end of pipes opening into
vast domains. The models consider both square and circular cross sections, as well as
flanged and unflanged pipe ends. These models are based on a plane wave assumption
(propagation below the cutoff frequency).

Choose an option from the list: Flanged pipe, circular (the default), Flanged pipe,
rectangular, Unflanged pipe, circular (low ka limit), or Unflanged pipe, circular. Then enter
the following as required:

• Inner radius a (SI unit: m) or

• Inner width wi (SI unit: m) and Inner height hi (SI unit: m).

Open Pipe: Application Library path Acoustics_Module/

Verification_Examples/open_pipe

Porous Layer
This choice models the acoustic losses of a normally incident field on a porous layer of
user-defined thickness d backed by a sound-hard wall. Use this boundary condition as
an alternative to modeling the porous layer explicitly using the Poroacoustics feature.
All material models from Poroacoustics are implemented in this feature.

Enter the Thickness of porous layer d (Si unit: m) and select a Poroacoustic model. The
rest of the settings are the same as for Poroacoustics.

Characteristic Specific Impedance

This is a set of models describing the characteristic impedance associated with three
basic wave types: plane wave, cylindrical wave, and spherical wave. Although mostly of
academic interest, these serve as good first-order and wave-type specific
boundary-condition models of infinite domains (open boundaries). They can be
applied to all cases where the tangential components of the acoustic field at the
boundary may be ignored, that is when the angle of incidence is well defined and the
wave direction is well known. Use the radiation conditions (Plane Wave Radiation,
Spherical Wave Radiation, or Cylindrical Wave Radiation) if a nonreflecting open
boundary is modeled.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 99

 Select a Wave type: Plane wave (the default), Cylindrical wave, or Spherical wave. Then
enter the Wave direction ek for the plane wave (default is normal to the surface); the
Radiating field source location r0 and Radiating field source axis rac for the cylindrical
wave; or the Radiating field source location r0 for the spherical wave.

Absorption Coefficient
With this option the specific impedance of a boundary is defined through the normal
incidence absorption coefficient αn (SI unit: 1) of the boundary. Since the absorption
coefficient carries no phase information, it is also possible to define the phase of the
associated reflection coefficient. If no phase is entered the impedance will be purely
resistive (no reactive component is defined). This is typically an acceptable
approximation at higher frequencies and it is also the assumption in ray tracing models.
Sometimes surfaces are only specified through an absorption coefficient and in such
cases using the Absorption coefficient option can be a first good approximation.

Enter the Normal incidence absorption coefficient αn (SI unit: 1) and the Phase (SI unit:
rad).

Symmetry
The Symmetry node adds a boundary condition where there is symmetry in the
pressure. Use this condition to reduce the size of a model by cutting it in half where
there are symmetries. In pressure acoustics, this boundary condition is mathematically
identical to the Sound Hard Boundary (Wall) condition.

Periodic Condition
The Periodic Condition node adds a periodic boundary condition that can be used to
reduce the model size by using symmetries and periodicities in the geometry and
physics interfaces being modeled. This feature works well for cases like opposing
parallel boundaries. In other cases, use a Destination Selection subnode to control the
destination. By default, it contains the selection that COMSOL Multiphysics identifies.

PERIODICITY SETTINGS
Select a Type of periodicity: Continuity (the default), Antiperiodicity, Floquet periodicity
(Bloch periodicity), or Cyclic symmetry.

• For Floquet periodicity, also known as Bloch periodicity, enter a k-vector for Floquet
periodicity kF (SI unit: rad/m) for the x, y, and z coordinates (3D components), the

100 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 r and z coordinates (2D axisymmetric components), or x and y coordinates (2D
components).
This condition is used to model infinite periodic structures with nonnormal incident
pressure fields or excitations. Use it to model, for example, a large perforated plate
with an oblique incident wave with wave vector k (and set kF = k) by only
analyzing one hole or one subset of holes that is periodic.
• For Cyclic symmetry select a Sector angle: Automatic (the default) or User defined. For
User defined enter a value for θS (SI unit: rad). Enter an Azimuthal mode number m
(dimensionless).
This condition is used to model any geometry that has a cyclic periodic structure
such as a microphone or a loudspeaker driver. Setting the azimuthal mode number
determines which mode that is analyzed. The response of the full system to an
external excitation is in general a linear combination of many different modes.

In the time domain, both the Cyclic symmetry and the Floquet periodicity boundary
conditions reduce to the continuity condition.

To optimize the performance of the Floquet periodicity and the Cyclic

symmetry conditions, it is recommended that the source and destination
meshes are identical. This can be achieved by first meshing the source
boundary or edge and then copying the mesh to the destination boundary
or edge. When the Periodic Condition stretches across regions with a mix
of default material models, PMLs, background pressure fields, or
background acoustic fields, it is recommended to add one Periodic
Condition for each set of such boundaries.

The Porous Absorber model uses Floquet periodic boundary conditions

to model an infinite porous absorber used for sound proofing. The
Application Library path: Acoustics_Module/Building_and_Room_Acoustics/
porous_absorber

CONSTRAINT SETTINGS
To display this section, select Advanced Physics Options in the Show More Options dialog
box.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 101

 EXCLUDED EDGES
To display this section, select Advanced Physics Options in the Show More Options dialog
box.

EXCLUDED POINTS
To display this section, select Advanced Physics Options in the Show More Options dialog
box.

ORIENTATION OF SOURCE
To display this section, select Advanced Physics Options in the Show More Options dialog
box.

• See Orientation of Source and Destination in the COMSOL

Multiphysics Reference Manual.
• See Excluded Surfaces, Excluded Edges, and Excluded Points in the
COMSOL Multiphysics Reference Manual.

Matched Boundary
The Matched Boundary node adds a matched boundary condition. Like the radiation
boundary conditions, it belongs to the class of nonreflecting boundary conditions
(NRBCs). The Incident Pressure Field subnode is available if required.

Properly set up, the matched boundary condition allows one mode with wave number
k1 (set k2 = k1), or two modes with wave numbers k1 and k2, to leave the modeling
domain with minimal reflections. The equation is given by

ω 2
i   ---- + k 1 k 2 p + iΔ || p
  c c 
– n ⋅  – ----- ( ∇p t – q d ) + -------------------------------------------------------------
1
 ρc  ρc ( k1 + k2 )
ω 2
i   ---- + k 1 k 2 p i + iΔ || p i
cc 1
= ---------------------------------------------------------------- + n ⋅ ----- ∇p i
ρc ( k1 + k2 ) ρc

Here, Δ || , for a given point on the boundary, refers to the Laplace operator in the
tangential plane at that point, while pi is the amplitude of an optional incoming plane
wave with wave vector k. In addition to pi, specify the propagation direction nk,
whereas the wave number is defined by keq = ω/cc in 3D and 2D.

102 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 ω 2 2
k eq =  ---- – k z
2
cc

The matched boundary condition is useful for modeling acoustic waves in ducts and
waveguides at frequencies below the cutoff frequency for the second excited transverse
mode. In such situations, set k1 = ω/cc and k2 = ω1/cc, where ω1 = 2πf1, and f1 is the
cutoff frequency for the first excited mode. In general, if several modes exist use the
Port boundary condition. The cutoff frequency or wave number may be found using
a Boundary Mode Acoustics model. When k1 = k2 = ω/cc, the matched boundary
condition reduces to the time-harmonic plane-wave radiation boundary condition.

MATCHED BOUNDARY
From the Match list, select One mode (the default) or Two modes. Then enter a Wave
number (SI unit rad/m) based on the selection: k1 for One mode and k1 and k2 for Two
modes.

Exterior Field Calculation

Use the Exterior Field Calculation node to apply the source boundaries for the exterior
field transformation set up by the Helmholtz-Kirchhoff integral. You also specify a
name for the acoustic exterior field variable (default name is pext) used in subsequent
postprocessing. The feature allows the calculation and visualization of the pressure
field outside the computational domain at any distance including amplitude and phase.
Note that when a Background Pressure Field is present, the feature only operates on
the scattered field variables and is thus also well suited for analyzing the results of a
scattering problem.

The exterior field boundary should be continuous and needs to enclose all sources and
scatterers with the desired symmetry types (see below). The feature should be applied
on an exterior boundary, a boundary on the inside of a PML, or an interior boundary
where the normal direction is continuous along the boundary. In the latter case, the
advanced physics option Reverse normal direction check box should be marked if the
normal is pointing inward (see below). Note also that if an interior boundary is used,
the material properties should be continuous across the boundary.

On exterior radiation boundaries and on the inside boundary of a PML the normals
are always automatically adjusted. In all other cases the direction of the normals can be
visualized with an arrow surface or line plot in postprocessing, plotting nx, ny, and nz.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 103

 The exterior field operator and its associated variables are optimized for
use in postprocessing. This means that they cannot be used for gradient
based optimization studies, like shape and topology optimization.

For gradient based optimization a dedicated operator exists in 3D (only

for the Symmetry plane option) for the pressure pext_opt(x,y,z) and
for the sound pressure level Lp_pext_opt(x,y,z). These two operators
can be used to define objective functions, like specifying a target spatial
response.

An example is given in the Shape Optimization of a Rectangular

Loudspeaker Horn in 3D tutorial. Application Library path
Acoustics_Module/Optimization/rectangular_horn_shape_optimization

EXTERIOR FIELD CALCULATION

Enter an Exterior field variable name for the exterior field acoustic pressure field (the
default is pext).

Select the Symmetry type that applies to the problem. The symmetry type and
subsequent settings should match the conditions used in the underlying acoustics
problem, for example, if using Symmetry or the Periodic Condition.

Select Symmetry planes (the default), Sector symmetry, or Sector symmetry with one
symmetry plane:

• For Symmetry planes, if no symmetry conditions are used (the default) the exterior
field feature boundaries should form a fully closed surface. If necessary, select and
add a condition for one of the Cartesian coordinate planes (with a possible offset)
to model either a symmetry condition in the plane (which is the same as an infinite
sound hard boundary) or an antisymmetry condition in the plane (which is the same
as an infinite sound soft boundary). The infinite sound hard boundary option is
especially useful when modeling system with an infinite baffle configuration.
For each of these planes, select the type for the condition to be applied in the x = x0,
y = y0, or z = z0 planes. Select the type of condition: Off (the default), Symmetry/
Infinite sound hard boundary, or Antisymmetric/Infinite sound soft boundary. Then
enter the value for the plane location x0, y0, or z0 (the default is 0 m). This allows
an offset of the infinite condition planes along the main coordinate axes.

104 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• The Sector symmetry options allows the underlying geometry to represent only a
sector (an subdivision of a full 360o rotation). Select and define the properties of the
sector. First select the Transformation applied to the sector, either Rotation (the
default) or Reflection and rotation. For the latter enter the Reflection plane used for
the initial reflection of the sector. The selection is restricted to the main coordinate
planes (move and rotate the geometry if necessary). The offset distance (x0, y0, z0)
is given by the point a0. Enter the coordinates for the Point on sector symmetry axis,
a0 as well as the Sector symmetry axis direction, adir (default is adir = (0, 0, 1)).
Enter the Number of sectors n, which has to be an integer and for the Reflection and
rotation option an even integer. Finally, if the pure Rotation option is selected enter
an optional Azimuthal mode number m (the default it 0). This option is typically only
relevant when used together with Periodic Condition.
Note that the Sector symmetry option cannot be combined with an infinite baffle
(symmetry plane) used in many transducer models. For this type of setup use the
Sector symmetry with one symmetry plane option.
• The Sector symmetry with one symmetry plane option allows for the combination of
sector symmetry and a single symmetry plane, under certain restrictions. First, select
the single Symmetry plane and symmetry Condition. The offset distance (x0, y0, z0)
is given by the point a0. Then follow the same settings as for the Sector symmetry
option. Note that the sector symmetry axis direction adir is automatically defined as
being normal to the single symmetry plane.
When one of the symmetry plane conditions is enabled the infinite plane, where it
applies, is rendered and can be visualized in the Graphics window. This rendering can
be turned off in the Physics Symbol section by clearing the check box Show physics
symbols.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 105

 Select a Type of integral: Full integral (the default) to compute The
Helmholtz-Kirchhoff Integral Representation or the Far-field integral approximation
for r → ∞ to compute the value in The Far-Field Limit.

The exterior field pressure is evaluated using the exterior field operator
(the name is defined in the Exterior field variable name input field, the
default is pext).

To evaluate the pressure in a point (x0,y0,z0), simply write

pext(x0,y0,z0). To evaluate the sound pressure level in the same point,
it is advantageous to use the subst() operator and write, for example,
subst(acpr.efc1.Lp_pext,x,x0,y,y0,z,z0).

An example of this is given in the Loudspeaker Driver —

Frequency-Domain Analysis tutorial model from the Acoustics
Application Libraries.

In 2D axisymmetry the evaluation of the exterior-field integral

automatically includes the Azimuthal mode number m from the Pressure
Acoustics Equation Settings. For postprocessing purposes, for example in
a Radiation Pattern plot, it is necessary to include the azimuthal
component explicitly by writing: pext(r,z)*exp(-i*acpr.m*phi).

ADVANCED SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

The option Use polynomial-preserving recovery for the normal gradient on interior
boundaries is selected per default on interior boundaries. This means that the exterior
field feature automatically uses the polynomial-preserving recovery operator ppr() to
get an enhanced evaluation of the normal derivative of the pressure. This increases the

106 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

precision of the exterior field calculation. If you click to clear this check box this
removes all instances of the operator from the equations.

The ppr() operator is not added when the exterior field calculation is
performed on an external boundary or a boundary adjacent to a perfectly
matched layer (PML) domain. In the latter case, the down() or up()
operator is automatically added in order to retrieve values of variables
from the physical domain only.

In these cases, use a single boundary layer mesh on the inside of the outer
boundary or on the inside of the PML to enhance the precision of the
exterior field calculation. Note that this layer is automatically added when
using the Physics-Controlled Mesh for Pressure Acoustics.

The option Reverse normal direction on interior boundaries allows reversing the normal
used in the Helmholtz-Kirchhoff integral. To get the correct phase the normal has to
point inward. Typically, if the exterior field is calculated in an interior boundary to the
physics (not a boundary next to a PML), the normals will point outward and the
option should be used.

• Evaluating the Acoustic Field in the Exterior: Near- and Far-Field

• Several dedicated acoustics plots rely on the exterior field calculation,
see Dedicated Acoustics Plots for Postprocessing
• ppr and pprint and up and down (operators) in the COMSOL
Multiphysics Reference Manual

• Acoustic Scattering off an Ellipsoid. The Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/acoustic_scattering

Port
The Port boundary condition is used to excite and absorb acoustic waves that enter or
leave waveguide structures, like a duct or channel, in an acoustic model. A given port
condition supports one specific propagating mode. To provide the full acoustic
description, combine several port conditions on the same boundary. Make sure that all
propagating modes in the studied frequency range are included (all modes that have a
cutoff frequency in the frequency range). By doing this, the combined port conditions

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 107

 provide a superior nonreflecting or radiation condition for waveguides to, for example,
the Plane Wave Radiation condition or a perfectly matched layer (PML) configuration.
The same port boundary condition feature should not be applied to several waveguide
inlets/outlets. The port condition supports S-parameter (scattering parameter)
calculation but it can also be used as a source to just excite a system. The Port boundary
condition exist for 3D, 2D, and 2D axisymmetric models.

• Duct with Right-Angled Bend: Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/duct_right_angled_bend
• Absorptive Muffler: Application Library path Acoustics_Module/
Automotive/absorptive_muffler
• Muffler with Perforates: Application Library path Acoustics_Module/
Automotive/perforated_muffler

On a given boundary, a combination of ports will define the total acoustic field (sum
of incident and outgoing waves) as


in iφ
pt = A e ( S ij + δ ij )p i
i ∈ bnd

where the summation “i” is over all ports on the given boundary “bnd”, Sij is the
scattering parameter, Ain is the amplitude of the incident field (at port “j”), φ is the
phase of the incident field, and pi is the mode shape of the i-th port. The mode shape
pi is normalized to have either a unit maximum amplitude or a unit power (see the
normalization option in the Global Port Settings section). This means that the
scattering parameter Sij defines the amplitude of mode i when a system is exited at
port j (with mode j). This corresponds to a multi-mode expansion of the solution on
the given port. The scattering parameters are automatically calculated when an
acoustic model is set up with just one port exciting the system. To get the full
scattering matrix The Port Sweep Functionality can be used.

Only one port should be excited at a time if the purpose is to compute

S-parameters. The S-parameters are defined as acpr.S11, acpr.S21, etc.
and can be used in postprocessing.

In other cases, having more than one port exciting the system might be
wanted, but the S-parameter variables cannot be correctly computed. So
when several ports are excited, the S-parameter output is turned off.

108 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

PORT PROPERTIES
Enter a unique Port name. Only nonnegative integer numbers can be used as Port name
as it is used to define the elements of the S-parameter matrix. The numeric port names
are also required for port sweep functionality. The port name is automatically
incremented by one every time a port condition is added.

Select a Type of port: User defined (the default), Numeric, Circular, Rectangular or Slit.
Depending on the selection different options appear in the Port Mode Settings section
(see below). Use the Circular and Rectangular for ports with the given cross section in
3D, the Circular option in 2D axisymmetry, and the Slit (plane wave mode) option in
2D. If the port has a different cross section, then either use the User defined option or
the Numeric port.

PORT MODE SETTINGS

Depending on the option selected for the Type of port (see above):

• For User defined enter user defined expressions for the Mode shape pn and the Mode
wave number kn (SI unit: rad/m). The modes shape will automatically be scaled
before it is used in the port condition. The normalized mode shape can be visualized
by plotting [Link] (here for Port 1 etc.). Use the user defined option to
enter a known analytical expression or to use the solution from The Pressure
Acoustics, Boundary Mode Interface. The solutions from the boundary mode
analysis can be referenced using the withsol() operator.
• The Numeric port option is used for waveguide cross sections that are neither circular
nor rectangular. In this case, a boundary mode problem is solved on the port face
to compute the desired propagating mode. This option requires the use of a
Boundary Mode Analysis step in the study. It should be placed before the Frequency
Domain step. In the study, add one Boundary Mode Analysis step for each Numeric port
and make sure to reference the proper Port name in the study step.
Select the Mode wave number from option to decide how the mode wave number kn
is determined:
- In general for a model with losses, use the default Out-of-plane wave number
option; then the wave number is taken from the Boundary Mode Analysis step. In
this case, it is not possible to perform a frequency sweep in the Frequency Domain
study step. Only one frequency can be used and it should correspond to the Mode
analysis frequency entered in the Boundary Mode Analysis step(s). One option is to
add a Parametric Sweep and define a parameter for the frequency used in both the

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 109

 steps. In this case, care should be taken when setting up the search criteria in the
mode analysis.
- For a model without any losses, select the Computed lossless mode cutoff frequency
option. In this case, a frequency sweep is possible. The Boundary Mode Analysis
should only be carried out at the highest frequency in the sweep. The wave
number is computed analytically for all the other frequencies.

When the Numeric port option is used and the boundary mode analysis is
run, the boundary conditions from the Pressure Acoustics model are
automatically inherited in the boundary problem. For this automatic
procedure, there is only support for the Sound Hard Boundary (Wall),
Symmetry, Pressure, and Sound Soft conditions. If you need more complex
behavior, use the Pressure Acoustics, Boundary Mode physics interface in
combination with the User defined port type.

When running a frequency domain sweep, in a model that uses either the
Circular or the Rectangular port options, you can get a solver warning: New
constraint force nodes detected: These are not stored.

This warning can be disregarded. The reason for the warning is that
certain constraints, in underlying equations, only are activated when the
frequency is above the mode cutoff frequency.

• The Circular port option is used for a port with a circular cross section. Enter the
Mode number, azimuthal m and the Mode number, radial n to define the mode
captured by the port. In 3D, also right-click the Port condition to add the Circular
Port Reference Axis when the Circular port type is selected. The cutoff frequency of
the mode can be evaluated in postprocessing using the variable [Link]
(here for Port 1 etc.).

(0,0) (1,0) (0,1) (1,1) (2,0) (0,2)

Figure 2-1: The first 6 modes (m,n) of a waveguide with circular cross section.

110 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• The Rectangular port option, only available in 3D, is used for a port with a
rectangular cross section. Enter the Mode number, longest side m and the Mode
number, shortest side n to define the mode captured by the port.

(0,0) (1,0) (0,1) (1,1) (2,0) (0,2)

Figure 2-2: The first 6 modes (m,n) of a waveguide of rectangular cross section.

• The Slit port option is only valid in 2D geometries. Enter the Mode number m to
define the mode captured by the port.

The cutoff frequency of the port modes can be evaluated in

postprocessing using the variable [Link] (here for Port 1 etc.).

INCIDENT MODE SETTINGS

Activate if the given port is excited by an incident wave of the given mode shape. For
the first Port condition added in a model, the Incident wave excitation at this port is set
to On. For subsequent conditions added, the excitation is set to Off per default. If more
than one port in a model is excited the S-parameter calculation is not performed.

When the Incident wave excitation at this port is set to On, then select how to define
the incident wave. Set Define incident wave to Amplitude (the default) or Power.

• For Amplitude enter the amplitude Ain (SI unit: Pa) of the incident wave. This is in
general defined as the maximum amplitude for a given mode shape.
• For Power enter the power Pin (SI unit: W) of the incident wave. This is in general
defined as the RMS power of the incident wave.
• For both options enter the phase φ (SI unit: rad) of the incident wave. This phase
contribution is multiplied with the amplitude defined by the above options. The
Amplitude input can be a complex number.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 111

 Note that when the Activate port sweep option is selected at the physics level, the
options in the Incident Mode Settings section are deactivated. This is because this option
automatically sends in a mode of unit amplitude, sweeping through one port at a time.

All ports with incident wave excitation turned off have an arrow symbol,
added in the Graphics window, that points outward. When incident wave
excitation is turned on, the arrow symbol points inward.

If a waveguide is cut with one or several symmetry planes, care should be

taken when setting up the port modes and when postprocessing.

For the Circular and Rectangular options make sure to only select modes
that are actually symmetric according to the symmetry planes.

When postprocessing, remember that absolute values like, for example,

the outgoing power at port 1, acpr.port1.P_out, needs to be
multiplied with an appropriate factor. Multiplication with two if one
symmetry plane is used etc.

The port condition is in general not compatible with the Background

Pressure Field domain feature. Combining the two will generate
unphysical results if placed next to each other.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

THE PORT SWEEP FUNCTIONALITY

The port sweep functionality is used to reconstruct the full scattering matrix Sij by
automatically sweeping the port excitation through all the ports included in the model.
When the port sweep is activated, the options in the Incident Mode Settings in the port
conditions are deactivated and COMSOL controls which port that is excited with an
incident mode.

The port sweep functionality is activated at the main physics interface level by selecting
Activate port sweep in the Global Port Settings section. Enter the Sweep parameter
name, the default is PortName. Create a parameter with the same name under Global
Definitions>Parameters 1. This is the name of the parameter to be used in a parametric

112 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

sweep, here it should represent the Port name integer values (defined when adding the
port conditions). Add a parametric sweep study step and run the sweep over the
PortName parameter with an integer number of values representing all the ports in the
model. Once the model is solved the full scattering matrix can be evaluated using the
defined global variables acpr.S11, acpr.S21, acpr.S12 etc. The transmission loss
(TL) between two given ports is also computed, for example, the variable for the TL
loss from port 1 to 2 is given by acpr.TL_12.

Use the Global Matrix Evaluation under Derived Values to evaluate the full
scattering matrix acpr.S.

If only two ports are added to the Pressure Acoustics model, COMSOL also
automatically computes the transfer matrix of the system (variables acpr.T11,
acpr.T12, acpr.T21, acpr.T22) and the impedance matrix of the system (acpr.Z11,
acpr.Z12, acpr.Z21, acpr.Z22). These expressions are only true if plane wave modes
are used.

Circular Port Reference Axis

The Circular Port Reference Axis is necessary when the port type is set to Circular in the
Port feature. Select two points in the geometry to define the azimuthal reference axis
for the port mode orientation.

Lumped Port
The Lumped Port feature connects a waveguide or duct inlet/outlet to a lumped
representation element. This can be an Electrical Circuit, a two-port network defined
through a transfer matrix, or a lumped representation of a waveguide. Basically, it
couples the end of a waveguide to an exterior system that has a given lumped
representation. Several representations and sources exist to describe the lumped system
as well as excite the system. When using the lumped port representation, it is assumed
that only plane waves propagate in the acoustic waveguide.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 113

 LUMPED PROPERTIES
Select the type of lumped representation connected to the port by selecting the
Connection type as Circuit, Two-port (the default), or Waveguide. The selection made
here influences the settings and options in the Connection Type section (see below).

• The Circuit option is used to connect the port boundary to a an Electrical Circuit
interface (this requires the AC/DC module). The connection is set up by adding
the External I vs. U element in the circuit model and selecting the lumped port as the
Electric potential, for example, Voltage from lumped port (acpr/lport1). The feature
automatically introduces a unit transformation from electric to acoustic units. The
classical equivalence between the voltage V (SI units: V) and the pressure p (SI unit:
Pa) as well as the current I (SI unit: A) and the volume velocity Q (SI unit: m3/s)
is made.
• For the Two-port option it is assumed that the port boundary is connected to a
two-port system described through a transfer matrix T. See the Connection Type
section for details.
• For the Waveguide option it is assumed that the port boundary is extended with a
waveguide of constant and identical cross section. The connection is equivalent to
the two-port as the waveguide is defined through the appropriate transfer matrix T,
of a waveguide. See the Connection Type section for details.

For the Two-port and the Waveguide options also select the Exterior condition as Circuit,
Ideal source (the default), Impedance, or Port. The selection made here influences what
section is visible in the user interface. When Port is selected enter the Port name, similar
to the Port condition. See below for a description of the options.

Two electroacoustic tutorial models that couple a circuit (lumped

representation of a microspeaker) to an acoustic domain using the Lumped
Port and the Circuit option are:

• Lumped Receiver Connected to Test Setup with a 0.4-cc Coupler:

Application Library path Acoustics_Module/Electroacoustic_Transducers/
lumped_receiver_04cc
• Lumped Receiver with Full Vibroacoustic Coupling: Application
Library path Acoustics_Module/Electroacoustic_Transducers/
lumped_receiver_vibroacoustic

114 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

CONNECTION TYPE
This section is visible if Two-port or Waveguide is selected as the Connection type.

• For Two-port, select the Two-port representation as Acoustic (the default) or

Electroacoustic. The selection indicates if the transfer matrix is from acoustic to
acoustic variables or from electric to acoustic variables. The first can be used to
include a transfer matrix of an acoustic sub-system, for example, extracted using the
Port condition. The latter should, for example, be used if the transfer matrix
represents a transducer with electric inputs and acoustic outputs.
Enter expressions, for example, calling interpolation functions that depend on
frequency, for the elements of the transfer matrix T11, T12, T21, and T22.
For the Acoustic option, you can choose to invert the transfer matrix (reverse its
direction) by selecting Invert transfer matrix direction. This option will invert and
correct the signs of the transfer matrix T behind the scene.
• For Waveguide, select the Waveguide model as Automatic (the default) or User defined.
The Automatic option will use the cross-section data and fluid properties of the
boundary for the model. If User defined is selected, enter values for the Wave number
k, the Specific characteristic acoustic impedance Z. For both options, enter the
Waveguide length L (SI unit: m) modeled with the lumped representation.

The transfer matrix T couples the inlet and outlet pressure and flow rate,
defined in the feature, according to the following usual convention:

p1 T 11 T 12 p 2
=
Q1 T 21 T 22 Q 2

SOURCE SETTINGS
This section is visible if Ideal source is selected as the Exterior condition.

Select the Input source as Average pressure (the default) or Volume flow. Enter values for
the average pressure pav,in or the volume flow Qin. These represent ideal sources
equivalent to an ideal voltage or ideal current source.

IMPEDANCE SETTINGS
This section is visible if Impedance is selected as the Exterior condition.

In this case the lumped port is not active and an impedance condition can be applied
to terminate the lumped representation. This could, for example, be the acoustic

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 115

 impedance of a microphone membrane. Select the Impedance model as User defined (the
default), RCL, or Waveguide end impedance.

• For the User defined option select the input format as an Acoustic impedance Zac (SI
unit: kg/m4·s) or an Impedance Z (SI unit: Pa·s/m). If the electroacoustic
representation is used for the two-port enter the Electric impedance Z (SI unit: Ω).
• The RCL and Waveguide end impedance options have the same options as the
Impedance boundary condition.

INCIDENT MODE SETTINGS

This section is visible if Port is selected as the Exterior condition.

With this option the exterior behavior of the lumped system is like a Port condition
with only plane wave modes propagating. The system can be excited if Incident wave
excitation at this port is turned On. Enter the amplitude Ain and phase φ of the incident
wave. The units depend on the representation used. For this option, the scattering
parameter acpr.S11, acpr.S21 etc. of the system is also computed automatically.

Lumped Speaker Boundary

The Lumped Speaker Boundary condition is used to model a loudspeaker (dust cap,
cone, and surround assembly) using a lumped representation with a coupling to an
Electrical Circuit interface. The mechanical and electric properties of the speaker are
described through a Thiele-Small parameter representation and associated lumped
parameters. The properties of a back volume are also modeled using an impedance
expression.

If the speaker represents an interior boundary, with air domains on both sides, and the
back volume is modeled explicitly; then use the Interior Lumped Speaker Boundary
condition.

SPEAKER GEOMETRY
Define how the Speaker area (projected) of the speaker is computed by selecting either
Selected boundaries (the default) if all boundaries are present, or Use symmetries if only
the speaker surface is partially represented. Enter the Speaker area multiplication factor
Ascale (default is 1). The settings are required in order to compute the effective area
of the speaker.

Define the Speaker axis direction eax by selecting Automatic (the default) or User
defined. For the automatic option the axis is computed as the average of the surface

116 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

normals, this option is valid if a full speaker surface is selected. The user defined option
should be used if the speaker surface is only partially represented.

BACK VOLUME CORRECTION

The Lumped Speaker Boundary does not model the acoustics on the back side of the
speaker cone explicitly. The properties can be entered either through a volume giving
a simple compliance effect, or through a user defined acoustic impedance of the back
volume.

Select the Acoustic impedance of the back volume as Volume compliance (the default) or
User defined. Enter either a Volume Vback (SI unit: m3) or an Acoustic impedance Zac
(SI unit: kg/(m4·s)).

CIRCUIT
This section gives information about the state of coupling/connection to an Electrical
Circuit interface. If the Interior Lumped Speaker Boundary is not connected the text
Connect to ‘External I vs. U’ in the Electrical Circuit interface text will be displayed. Once
connected a reference with a tag to the associated External I vs. U node, in the Electrical
Circuit interface, is displayed.

Thermoviscous Boundary Layer Impedance

The Thermoviscous Boundary Layer Impedance condition adds the losses due to thermal
and viscous dissipation in the acoustic boundary layers at a wall. The condition is
sometimes known simply as the BLI model. The losses are included in a locally
homogenized manner, where the losses are integrated through the boundary layers
analytically. The condition is applicable in cases where boundary layers are not
overlapping. That is, it is not applicable in a very narrow waveguide (with dimensions
comparable to the boundary layer thickness) or on very curved boundaries. Other than
that, there are no restrictions on the shape of the geometry. This is in contrast to the
Narrow Region Acoustics feature which is applicable only in waveguides of constant
cross section, but also applicable for all frequencies, that is, also the very narrow case
where boundary layers are overlapping. The thickness of the viscous and thermal
boundary layers is given by

2μ 2k -
δv = ------- δ th = --------------
ωρ ωρC p

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 117

 where ω is the angular frequency, μ the dynamic viscosity, ρ the density, k the
coefficient of thermal conductivity, and Cp the (specific) heat capacity at constant
pressure.

The Thermoviscous Boundary Layer Impedance condition adds an impedance-like

boundary condition by defining the inward normal velocity -n·v at the boundary in
terms of the pressure and its tangential derivatives:

α p T δ th α p δv
– n ⋅ v = – iω  T bnd – ----------- p t -------------- – v n + -----------  ∇ || ⋅ v || + ---------- Δ || p t
0 1
 ρC p  1 + i 1 + i iωρ 

where Tbnd is a possible boundary temperature variation source, vn is a possible normal

0
velocity source, and v || is a possible tangential velocity source (normal and tangential
components are computed from a velocity vector). The expression implemented is a
generalization of the expression presented in Ref. 53 including boundary temperature
and velocity sources. Velocity sources are presented in Ref. 54 but without including
the thermal effects. Thermal effects (without sources) are included in Ref. 55. The
general idea of the boundary layer impedance formulation can be found in Pierce,
Ref. 5 (see equation 10-4.12 in section 10-4 about the Acoustic Boundary-Layer
Theory).

The boundary layer impedance is used in the tutorial Piezoelectric MEMS

Speaker. The Application Library path: Acoustics_Module/
Electroacoustic_Transducers/piezo_mems_speaker

MODEL INPUT
Enter a value for the (equilibrium or boundary) Temperature T (SI unit: K). This field
is always necessary as the temperature enters the expression for the boundary
condition. This corresponds to the equilibrium temperature T0 in the thermoviscous
interfaces.

COORDINATE SYSTEM SELECTION

Select a Coordinate system: Global coordinate system (default) or Boundary boundary
System 1 (sys1). If more coordinate systems are present in the model they will also show
in the list. The option is used for entering the velocity vector.

MECHANICAL CONDITION
Select the Mechanical condition that applies for the wall/boundary: No slip (default),
Slip, Velocity, or Normal velocity (slip). These options give most of the relevant

118 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

mechanical conditions that can be set up using the The Thermoviscous Acoustics,
Frequency Domain Interface.

• No slip, default condition at a wall that is not moving.

• Slip, condition at a wall that is not moving and where the viscous boundary layer
effects are disregarded.
• Velocity, a moving/vibrating wall where the viscous boundary layer effects are taken
into account. A pure tangential wall movement will, for example, generate sound.
This is not the case in the lossless case.
• Normal velocity (slip), a moving/vibrating wall where viscous boundary layer effects
are disregarded, only the normal component of the wall movement will contribute
to the sound generation.

THERMAL CONDITION
Select the Thermal condition that applies for the wall/boundary: Isothermal (default),
Adiabatic, or Temperature variation. These options give most of the relevant thermal
conditions that can be set up using the The Thermoviscous Acoustics, Frequency
Domain Interface.

• Isothermal, default condition at a solid wall. Since the thermal conductivity of solids
are orders of magnitude larger than fluids the isothermal condition for the fluid is a
good approximation in most cases.
• Adiabatic, condition at a wall where the thermal boundary effects are disregarded.
• Temperature variation, applies a fluctuating temperature condition at the boundary.

FLUID PROPERTIES
Select where the Fluid material is taken from. The default is the Boundary material
(remember to assign a material to the boundary if this option is used) or select a
specific material if desired.

Enter the necessary material properties for the Speed of sound c, Density ρ, Heat capacity
at constant pressure Cp, Ratio of specific heats γ, Thermal conductivity k, and the
Dynamic viscosity μ. Per default they are taken From material. For User defined enter a
value for the property.

Plane Wave Radiation

The Plane Wave Radiation node adds a radiation boundary condition for a plane wave.
If required, add an Incident Pressure Field to model an incoming wave. This radiation

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 119

 condition allows an outgoing plane wave to leave the modeling domain with minimal
reflections, when the angle of incidence is near normal. Add this node from the
Radiation Conditions submenu.

The plane wave type is suitable for modeling open boundaries and ports (in the
plane-wave regime only). The condition can be used for waveguide structures as long
as they are only interesting in the plane-wave region, but in general the Port boundary
condition should be used. When using the radiation condition on an open far-field
boundary, it is recommended to construct the boundary such that the incidence angle
is near to normal. This of course requires a priori knowledge of the problem and the
solution. See the theory section Theory for the Plane, Spherical, and Cylindrical
Radiation Boundary Conditions for details about the equations and the formulation of
this nonreflecting boundary condition.

• An estimate of the reflection coefficient Rs, for the spurious waves reflecting off the
plane wave radiation boundary, is, for incident plane waves at angle θ, given by the
expression:

cos θ – 1-
R s = ---------------------
N

cos θ + 1

where N is the order of the boundary condition (here 1 or 2). So at normal

incidence (θ = 0), there are no spurious reflections, while, for example, at an
incidence angle of 30o for N = 2 (plane wave radiation in the frequency domain),
the amplitude of the spurious reflected wave is 0.5 % of the incident.

Spherical Wave Radiation

The Spherical Wave Radiation node adds a radiation boundary condition for a spherical
wave, for which you define the source location. If required, add an Incident Pressure
Field to model an incoming wave. This radiation condition allows an outgoing
spherical wave to leave the modeling domain with minimal reflections. The geometry
of the modeling domain should be adapted to have the outgoing spherical waves
coincide with the boundary, in order to minimize spurious reflections. Add this node
from the Radiation Conditions submenu.

See Theory for the Plane, Spherical, and Cylindrical Radiation Boundary Conditions
for details about the equations and the formulation of this nonreflecting boundary
condition.

120 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

SPHERICAL WAVE RADIATION
Enter coordinates for the Radiating field source location r0 (SI unit: m) for 3D and 2D
models. In a 2D axisymmetric model, enter the Radiating field source location, z
coordinate z0 (SI unit: m).

Cylindrical Wave Radiation

The Cylindrical Wave Radiation node adds a radiation boundary condition for a
cylindrical wave, for which you define the source location and the source axis direction.
If required, add an Incident Pressure Field to model an incoming wave. This radiation
condition allows an outgoing cylindrical wave to leave the modeling domain with
minimal reflections. The geometry of the modeling domain should be adapted to have
the outgoing cylindrical waves coincide with the boundary, in order to minimize
spurious reflections. Add this node from the Radiation Conditions submenu.

See Theory for the Plane, Spherical, and Cylindrical Radiation Boundary Conditions
for details about the equations and the formulation of this nonreflecting boundary
condition.

CYLINDRICAL WAVE RADIATION

Enter coordinates for the Radiating field source location r0 (SI unit: m) and the
Radiating field source axis direction raxis (dimensionless).

Acoustic Cloaking. The Application Library path: Acoustics_Module/

Tutorials,_Pressure_Acoustics/acoustic_cloaking

Incident Pressure Field

The Incident Pressure Field node is a subnode to most nonreflecting boundary
conditions (plane, cylindrical, spherical wave radiation, and matched boundary). From
the Physics toolbar, add to Matched Boundary, Plane Wave Radiation, Spherical Wave
Radiation, or Cylindrical Wave Radiation nodes. In the frequency domain, four
options exist for the Pressure field type: Plane wave, Cylindrical wave, Spherical wave, and
User defined.

To set up and calculate the intensity variables for both the total, incident, and reflected
acoustic fields at the boundary select Calculate incident and reflected field intensity at
the bottom of the settings window. This requires and additional input of the Density ρ

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 121

 (SI unit: Pa) for the defined incident pressure field. The variables are available in
postprocessing the results.

The Incident Pressure Field has built-in functionality to perform a plane wave
expansion. This allows modeling scattering problems involving 2D axisymmetry using
a 2D axisymmetric, model even though a plane wave is not in general axisymmetric.

When calculating, for example, a transfer function, use the defined

variable acpr.p_i for the incident field value on the boundary. This
variable also contains phase information that can be difficult to assess
otherwise.

When a perfectly matched layer (PML) is present in the model, do not

apply incident fields on its outer boundaries. The PML is applied to
absorb waves that move out of the computational domain. Defining an
incident field on its boundary will lead to unphysical results.

INCIDENT PRESSURE FIELD

From the Incident pressure field type list, select Plane wave (the default), Cylindrical
wave, Spherical wave, or User defined to define the incident pressure field type.

Plane Wave
If the incident pressure field pi is a predefined plane wave, it is of the type:

x⋅e
– i k s  -------------k
–i ( k ⋅ x )  e 
pi = p0 e = p0 e k

where p0 is the wave amplitude, k is the wave vector with amplitude ks = ω/c and wave
direction vector ek, and x is the location on the boundary.

For Plane Wave, enter a Pressure amplitude p0 (SI unit: Pa), the Speed of Sound c (SI
unit: m/s) of the medium defining the incident wave, the Wave direction ek (SI unit:
m), and a Phase φ (SI unit: rad).

In 2D axisymmetric models, the incident fields take a slightly different form due to the
geometrical restrictions. Only the z-component for the Wave direction ek can be
entered. Per default, the wave can only travel in the axial direction, since this is the only
axisymmetric form of a plane wave.

122 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

By selecting Enable plane wave expansion, the plane wave is expanded in its cylindrical
harmonics and a general Wave direction ek can be selected. This sets up a wave of the
form:

( x ⋅ ( e k ⋅ e r )e r ) ( x ⋅ ( e k ⋅ e r )e r )
i J m  – k s ------------------------------------- exp  – ik s -------------------------------------
iφ sign(m) m
pi = e 2
 ek   ek 

where m is the Azimuthal mode number specified in the Pressure Acoustics Equation
Settings. To expand the solution, it is necessary to run a Parametric Sweep over the
mode number from 0 up to the desired resolution.

Cylindrical Wave
If the incident pressure field pi is a predefined cylindrical wave, it is of the type:

(2) ( x – x 0 ) × e sa
pi = p0 H0 ( ks rs ) r s = -----------------------------------
e sa

where p0 is the amplitude given at the reference distance rref = 0.548/ks (the distance
(2)
where the Hankel function is one), ks = ω/c is the wave number, H m is the Hankel
function of the second kind (representing an outgoing cylindrical wave), rs is the
distance from the source axis, esa is the direction of the source axis, x0 is a point on
the source axis, and x is the location on the boundary.

For Cylindrical wave, enter a Pressure amplitude at reference distance p0 (SI unit: Pa),
the Speed of sound c (SI unit: m/s) of the medium defining the incident wave, a Source
location x0 (SI unit: m), a Source axis esa (the vector is automatically normalized), and
a Phase φ (SI unit: rad).

In 2D axisymmetric models, the incident fields take a slightly different form due to the
geometrical restrictions. No source location nor axis is needed, as the source is always
located on the z-axis.

Spherical Wave
If the incident pressure field pi is a predefined spherical wave (only for 2D
axisymmetric and 3D), it is of the type:

r ref –ik r
p i = p 0 -------- e s s r ref = 1 m rs = x – x0
rs

where p0 is the amplitude given at the reference distance of 1 m, ks = ω/c is the wave
number, rs is the distance from the source, x0 is the source location of the spherical
wave, and x is the location on the boundary.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 123

 For Spherical wave, enter a Pressure amplitude at reference distance p0 (SI unit: Pa), the
Speed of sound c (SI unit: m/s) of the medium defining the incident wave, a Source
location x0 (SI unit: m), and a Phase φ (SI unit: rad).

In 2D axisymmetric models, the incident fields take a slightly different form due to the
geometrical restrictions. Only enter the z-coordinate for the Source location z0 (SI unit:
m) since the source is always located on the z-axis in a 2D axisymmetric model.

For both the cylindrical and the spherical wave options, the source and/
or source axis should be located outside the computational domain where
the radiation condition is applied. The fields should be incident on the
radiation boundary.

User Defined
If User defined is selected, enter the expression for the Incident pressure field pi (SI unit:
Pa) as a function of space.

Perfectly Matched Boundary

The Perfectly Matched Boundary is effectively a perfectly matched layer (PML) that is
applied to the open boundary without the need to define a domain (a layer in the
geometry). The condition automatically applies the PML formulation using the extra
dimension functionality of COMSOL Multiphysics.

The Perfectly Matched Boundary condition only exists for the Pressure
Acoustics, Frequency Domain interface. For other interfaces use the
classical definition and setup for Perfectly Matched Layers (PMLs).

GEOMETRY
Select the Attenuation direction as Radial (the default), Normal, or Custom. For the Radial
option enter the Center coordinate r0, that is a coordinate that approximately gives the
location of the radiation source. The Radial option works well for most open problems
and it will be applicable for most geometries of the radiation boundary. The Normal
option will apply the attenuation in the normal direction to the boundary where the
condition is applied. In this case the shape of the radiating boundary is important. For
the Custom option enter the Custom attenuation direction in the edit fields.

124 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

SCALING AND MESH
Select the Coordinate stretching type as Logarithmic, Polynomial (the default), Rational,
or User defined.

For the Logarithmic, Polynomial, and Rational options enter the PML scaling factor and
the PML scaling curvature parameter. The default values depend on the selected
Coordinate stretching type. These settings correspond to the settings entered for the
classical Perfectly Matched Layers (PMLs). When User defined is selected choose the
functions to be used for the real and imaginary part of the stretching functions,
respectively. These should be Functions defined under the Definitions node.

Enter the Number of mesh points, the default depends on the selected Coordinate
stretching type. This represents the mesh resolution in the PML along the extra
dimension.

Select Use uniform thickness (not selected per default) to enforce an equivalent uniform
thickness in the attenuation direction. This option is only available when the
Attenuation direction is Radial. If selected a slight gain may be achieved in numerical
precision, but the performance/convergence of iterative solvers may be compromised.
For spherical geometries this will make the attenuation behave as a PML with the same
settings and a uniform mesh.

• Piezoacoustic Transducer. The Application Library path:

Acoustics_Module/Piezoelelctric_Devices/piezoacoustic_transducer
• Acoustic Scattering off an Ellipsoid. Application Library path:
Acoustics_Module/Tutorials,_Pressure_Acoustics/acoustic_scattering

Boundary conditions that intersect a Perfectly Matched Boundary are not

continued/extended through the boundary. This setup is possible with a
classical PML where, for example, an interior wall, a periodic, or an
impedance condition can be defined inside the layer. This fact puts some
restrictions on the general use of the Perfectly Matched Boundary.

Interior Sound Hard Boundary (Wall)

The Interior Sound Hard Boundary (Wall) node adds a boundary condition for a sound
hard boundary or wall on interior boundaries. Add this node from the Interior

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 125

 Conditions submenu. A sound-hard boundary is a boundary at which the normal
component of the acceleration is zero:

– n ⋅  – ----- ( ∇p t – q d ) – n ⋅  – ----- ( ∇p t – q d )
1 1
= 0 = 0
 ρc   ρc 
up down

This is applied to the two sides of the boundary (the up and down sides). For zero
dipole source qd and constant fluid density ρc, this means that the normal derivative
of the pressure is zero at the boundary. On an interior sound hard boundary, the
pressure is not continuous but is treated as a so-called slit feature.

Interior Normal Acceleration

The Interior Normal Acceleration node adds a normal acceleration on an interior
boundary and ensures that the pressure is noncontinuous. Add this node from the
Interior Conditions submenu. The pressure has a so-called slit condition on this
boundary. This boundary condition can be used to model sources as, for example, the
movement of a speaker cone modeled as a boundary. The condition adds the normal
part of an acceleration a0:

– n ⋅  – ----- ( ∇p t – q d ) = n ⋅ a 0 – n ⋅  – ----- ( ∇p t – q d )

1 1
= n ⋅ a0
ρc up ρc down

Alternatively, specify the inward acceleration an. The normal of the boundary is
interpreted as pointing outward.

– n ⋅  – ----- ( ∇p t – q d ) = a n – n ⋅  – ----- ( ∇p t – q d )
1 1
= an
 ρc   ρc 
up down

INTERIOR NORMAL ACCELERATION

Select a Type: Acceleration (the default) or Normal acceleration.

• For Acceleration enter values for the components of the Acceleration a0 (SI unit: m/
s2).
• For Normal acceleration enter the value of the Normal acceleration an (SI unit: m/s2)
in the normal direction. The normal is depicted as a red arrow in the graphics
window on the selected boundary.

126 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Interior Normal Velocity
The Interior Normal Velocity node adds a normal velocity on an interior boundary and
ensures that the pressure is noncontinuous here. Add this node from the Interior
Conditions submenu. The pressure has a so-called slit condition on this boundary. The
condition is very similar to the Interior Normal Acceleration condition, simply having
the acceleration given by iω times the velocity.

INTERIOR NORMAL VELOCITY

Select a Type: Velocity (the default) or Normal velocity.

• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s2).
• For Normal velocity enter the value of the Normal velocity vn (SI unit: m/s2) in the
normal direction. The normal is depicted as a red arrow in the graphics window on
the selected boundary.

Interior Normal Displacement

The Interior Normal Displacement node adds a normal displacement on an interior
boundary and ensures that the pressure is noncontinuous. Add this node from the
Interior Conditions submenu. The pressure has a so-called slit condition on this
boundary. The condition is very similar to the Interior Normal Acceleration condition,
simply having the acceleration given by (iω)2 times the displacement.

INTERIOR NORMAL DISPLACEMENT

Select a Type: Displacement (the default) or Normal displacement.

• For Displacement enter values for the components of the Displacement d0 (SI unit:
m/s2).
• For Normal displacement enter the value of the Normal displacement dn (SI unit: m/
s2) in the normal direction. The normal is depicted as a red arrow in the graphics
window on the selected boundary.

Interior Impedance/Pair Impedance

The Interior Impedance and Pair Impedance nodes add an impedance boundary
condition on interior boundaries or boundaries between the parts of pairs. Add the
Interior Impedance node from the Interior Conditions submenu. This condition is a
generalization of the sound-hard and sound-soft boundary conditions. The condition
corresponds to a transfer impedance condition, relating the pressure drop across the

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 127

 boundary p t,down – p t,up to the velocity at the boundary. In the frequency domain, it
imposes the following equations:

–i ω
– n ⋅  – ----- ( ∇p t – q d )
1
= ( p t,down – p t,up ) ---------
 ρc  Zi
up
–i ω
– n ⋅  – ----- ( ∇p t – q d )
1
= ( p t,down – p t,up ) ---------
ρc down Zi

For a Time Dependent study (time domain), the boundary condition uses the
following equations:

1∂
– n ⋅  – --- ( ∇p t – q d ) = ---- ---- ( p t,up – p t,down )
1
ρ up Z i ∂t
∂
– n ⋅  – --- ( ∇p t – q d )
1 1 ----
----
 down = Z i ∂t ( t,up – p t,down )
p
 ρ

On pair (interior) impedance conditions, the up and down, which refers to the up and
down sides of an interior boundary, is replaced by a dst and src subscript in the
equation display. These refer to the destination and source faces in the pair. In the pair
feature the normal on the left-hand side in the equation refers to the destination
normal direction ndst.

Zi is the transfer impedance, which from a physical point, is the ratio between pressure
drop and normal particle velocity.

In the two opposite limits Zi → ∞ and Zi → 0, this boundary condition is

identical to the Sound Hard boundary condition and the Sound Soft
boundary condition, respectively. Additional information is found in
Identity and Contact Pairs in the COMSOL Multiphysics Reference
Manual.

INTERIOR IMPEDANCE/PAIR IMPEDANCE

Enter the value of the Impedance Zi (SI unit: Pa·s/m). The default is 0 Pa·s/m.

Interior Perforated Plate/Pair Perforated Plate

The Interior Perforated Plate and Pair Perforated Plate nodes provide the possibility of
specifying the characteristic properties for a perforated plate. Add the Interior
Perforated Plate node from the Interior Conditions submenu.

128 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

COMSOL Multiphysics has three Model types for calculating the transfer impedance of
a perforated plate. The following options are available:

• Thin plate (the default), the model where the losses due to heat conduction are
negligible.
• Thick plate, the model that takes the thermal effects into account.
• Asymptotic (legacy) model, the model available in COMSOL Multiphysics version
5.2a and earlier.

A detailed description of the implemented models can be found in the Theory for the
Interior Impedance Models section. The model properties are divided into two groups
as shown below.

INTERIOR PERFORATED PLATE/PAIR PERFORATED PLATE

−3
• Hole diameter dh (SI unit: m). The default is 1 mm (10 m).
• Plate thickness tp (SI unit: m). The default is 1.5 mm (1.5·10−3 m).
• Area porosity σ, that is, the holes’ fraction of the boundary surface area, a
dimensionless number between 0 and 1. The default is 0.1, which means 10% of the
plate area consists of holes.
• End correction to the resistance δresist and the reactance δreactt (SI unit: m). The
default built-in sets δresist = δresist = 4dh/3π. Otherwise, select the user defined.
• Hole-hole interaction fint, a dimensionless function that accounts for the influence of
the porosity on the end correction. The default is the built-in Fok function (using
eight terms):
8
n
f int =  an ( σ) .
n=0
( lin )
• Discharge coefficient (linear)CD , that is, a dimensionless coefficient related to the
rate of the real flow through a hole to the theoretical flow. The default is 1.
Enable the Include nonlinear effects check box to include the nonlinear contribution to
the resistance at high sound pressure levels. The following parameters become
available:

• Scaling factor fnl, a dimensionless. Default is 1.

( nl ) ( lin )
• Discharge coefficient (nonlinear) C D , which has the same meaning as C D . The
default is 0.76.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 129

 Enable User-defined contribution check box to, for example, include the effects of a
mean flow on the impedance. The following parameters become available:

• User-defined resistance θ(user) , a contribution to the resistive part of the impedance,

dimensionless. The default is 0.
• User-defined resistance χ(user) , a contribution to the reactive part of the impedance,
dimensionless. The default is 0.

FLUID PROPERTIES
Select the Fluid material from Boundary material (the default) or select a material from
the list. If Boundary material is selected, you need to add a material to the given
boundary under the Materials node.

Select the material properties either From material (the default) or User defined:

• Density ρ (SI unit: kg/m3).

• Speed of sound c (SI unit: m/s).
• Dynamic viscosity μ (SI unit: Pa·s).
• Ratio of specific heats γ (SI unit: 1). Only for the Thick plate model.
• Heat capacity at constant pressure Cp (SI unit: J/(kg·K)). Only for the Thick plate
model.
• Thermal conductivity k (SI unit: W/(m·K)). Only for the Thick plate model.

The transfer impedance models from the list above are only valid for the
perforates with round holes. For other types of perforates — with squared
or slit-shaped holes — the values of some parameters can considerably
differ from that of the suggested built-in perforates. It is recommended
that you use the Interior Impedance/Pair Impedance condition to enter
a user-defined model in such a case. The user-defined impedance can be
obtained from a thermoviscous acoustic submodel (see The
Thermoviscous Acoustics, Frequency Domain Interface) as demonstrated
in the Acoustic Muffler with Thermoviscous Acoustic Impedance
Lumping model.

Interior Lumped Speaker Boundary

The Interior Lumped Speaker Boundary is used to model a loudspeaker (dust cap, cone,
and surround assembly) using a lumped representation with a coupling to an Electrical

130 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Circuit interface. The mechanical and electric properties of the speaker are described
through a Thiele-Small representation and associated lumped parameters. Add this
node from the Interior Conditions submenu. The interior version of the Lumped
Speaker Boundary includes the effects of the fluid domains on both sides of the speaker
boundary explicitly.

SPEAKER GEOMETRY
Define how the Speaker area (projected) of the speaker is computed by selecting either
Selected boundaries (the default) if all boundaries are present, or Use symmetries if only
the speaker surface is partially represented. Enter the Speaker area multiplication factor
Ascale (default is 1). The settings are required in order to compute the effective area
of the speaker.

Define the Speaker axis direction eax by selecting Automatic (the default) or User
defined. For the automatic option the axis is computed as the average of the surface
normals, this option is valid if a full speaker surface is selected. The user defined option
should be used if the speaker surface is only partially represented.

For the Automatic option select Reverse automatic speaker axis direction if necessary.
The axis orientation can be visualized by plotting the axis components
acpr.ilsb1.e_axx, acpr.ilsb1.e_axy, and acpr.ilsb1.e_axz. (using the proper
item and physics scope).

CIRCUIT
This section gives information about the state of coupling/connection to an Electrical
Circuit interface. If the Interior Lumped Speaker Boundary is not connected the text
Connect to ‘External I vs. U’ in the Electrical Circuit interface text will be displayed. Once
connected a reference with a tag to the associated External I vs. U node, in the Electrical
Circuit interface, is displayed.

• Lumped Loudspeaker Driver: Application Library path

Acoustics_Module/Electroacoustic_Transducers/
lumped_loudspeaker_driver
• Headphone on an Artificial Ear: Application Library path
Acoustics_Module/Electroacoustic_Transducers/headphone_artificial_ear

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 131

 Continuity
Continuity is available as an option at interfaces between parts in an assembly, from the
Pairs submenu. The condition gives continuity in total pressure and in the normal
acceleration over the pair:

p t,src = p t,dest

– n ⋅ –  ----- ( ∇p t – q d ) –  – ----- ( ∇p t – q d )
1 1
= 0
 ρc   ρc 
src dest

Subscripts src and dest in the equation refer to the source and destination in the pair;
that is, the two sides in the pair.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

Line Source
Use the Line Source node to add a source on a line/edge in 3D components. This type
of source corresponds to a radially vibrating cylinder in the limit where its radius tends
to zero. The line source adds a source term to the right-hand side of the governing
Helmholtz equation such that:

2
k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = ------ Sδ ( x – x 0 )dl
1 4π
ρc ρc ρc

where δ ( x – x 0 ) is the delta function in 3D that adds the source on the edge where
x = x0 and dl is the line element along the edge (SI unit: m). The monopole amplitude
S (SI unit: N/m2) depends on the source type selected, as discussed below.

• Frequency Domain Study

• Solution (datasets) in the COMSOL Multiphysics Reference Manual

For the Pressure Acoustics, Transient interface, only the Flow (no phase
specification), User defined, and the Gaussian pulse source types are
available. The Gaussian pulse source type has no effect in the frequency
domain. See The Gaussian Pulse Source Type Settings.

132 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

LINE SOURCE
Select a Type: Flow (the default), Intensity, Power, or User defined.

For User defined, enter a Monopole amplitude, S = Suser (SI unit: N/m2). Otherwise,
enter details as follows.

Flow
Select Flow to add an edge source located at r = r0 defined in terms of the volume flow
rate per unit length out from source QS and the phase φ of the source. The flow edge
source defines the following monopole amplitude:

iφ iωρ
S = e -----------c- Q S
4π

A flow edge source with the strength QS represents an area flow out from the source
(the source is a very thin cylinder with a surface that pulsates).

• Enter a Volume flow rate per unit length out from source, QS (SI unit: m2/s) for the
source-strength amplitude in the field.
• Enter a Phase φ (SI unit: rad).

When defining a Solution dataset and plotting the results, specify a

nonzero phase φ to produce a nonzero result when visualizing the
resulting pressure field using the default value (0) in the Solution at
angle (phase).

Intensity
Select Intensity to add an edge source located at r = r0 defined in terms of the source
intensity radiated Irms and the phase φ of the source. Set a desired free space reference
intensity (RMS) Irms at a specified distance dsrc from the source. In a homogeneous
medium, the specified intensity is obtained when the edge is a straight line (this is the
reference). With other objects and boundaries present, or if the edge is curved, the
actual radiated intensity is different. This source type defines the following monopole
amplitude:

iφ d src
S = e -------------- 2I ref ρ c c c
L edge

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 133

 where Ledge is the length of the source line (automatically determined) and dsrc is the
distance from the source where free space reference intensity (RMS) Irms is specified.
Enter values or expressions for:

• Free space reference intensity (RMS), Irms (SI unit: W/m2).

• Distance from source center dsrc (SI unit: m).
• Phase φ (SI unit: rad).

Power
Select Power to add an edge source located at r = r0 specified in terms of the source’s
reference RMS strength by stating the total power Prms a straight line source would
radiate into a homogeneous medium. This source type defines the following monopole
amplitude:

iφ ρ c c c P rms
e
S = ------------- ---------------------
-
L edge 2π

where Ledge is the length of the source line (automatically determined) and Prms
denotes the free space reference RMS power (in the reference homogeneous case) per
unit length measured in W/m. Enter values or expressions for:

• Free space reference power (RMS), Prms (SI unit: W).

• Phase φ (SI unit: rad).

Line Source on Axis

Use the Line Source on Axis node to add a line source on the axis of symmetry in 2D
axisymmetric components. This type of source corresponds a radially vibrating cylinder
in the limit where its radius tends to zero. The line source adds a source term to the
right-hand side of the governing Helmholtz equation such that:

2
k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = ------ Sδ ( z – z 0 )dz
1 4π
ρc ρc ρc

where δ ( z – z 0 ) is the delta function in 3D that adds the source on the axis of
symmetry where z = z0 and r = 0, and dz is the line element along the z-axis

134 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

(SI unit: m). The monopole amplitude S (SI unit: N/m2) depends on the source type
selected and is the same as discussed in the 3D case for a Line Source.

For the Pressure Acoustics, Transient interface, only the Flow (no phase
specification), User defined, and the Gaussian pulse source types are
available. The Gaussian pulse source type has no effect in the frequency
domain. See The Gaussian Pulse Source Type Settings.

LINE SOURCE ON AXIS

Select a Type: Flow (the default), Intensity, Power, or User defined. See the options and
expression for Line Source. The sources are the same but in 2D axisymmetric
components they are only applicable on the symmetry axis at r = 0.

• Frequency Domain Study

• Solution (datasets) in the COMSOL Multiphysics Reference Manual

Monopole Point Source

Use the Monopole Point Source node to add a monopole point source in 3D
components on any point and in 2D axisymmetric components on points on the axis
of symmetry. This is a source that is uniform and equally strong in all directions. A
monopole represents a radially pulsating sphere in the limit where the radius tends to
zero. The monopole point source adds a point source term to the right-hand side of
the governing Helmholtz equation such that:

2
k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = ------ Sδ ( x – x 0 )
1 4π
ρc ρc ρc

where δ ( x – x 0 ) is the delta function in three dimensions and adds the source at the
point where x = x0. The monopole amplitude S (SI unit: N/m2) depends on the
source type selected, as discussed below.

For the Pressure Acoustics, Transient interface, only the Flow (no phase
specification), User defined, and the Gaussian pulse source types are
available. The Gaussian pulse source type has no effect in the frequency
domain. See The Gaussian Pulse Source Type Settings.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 135

 MONOPOLE POINT SOURCE
Select a Type: Flow (the default), Intensity, Power, or User defined.

For User defined enter a Monopole amplitude, S = Suser (SI unit: N/m).

Flow
Select Flow to add an monopole point source located at x = x0 defined in terms of the
volume flow rate out from source QS and the phase φ of the source. The source
defines the following monopole amplitude:

iφ iωρ c
S = e ------------ Q S
4π

Enter values or expressions for:

• Volume flow rate out from source, QS (SI unit: m3/s).

• Phase φ (SI unit: rad).

Intensity
Select Intensity to define the source in terms of the free space reference RMS intensity
Irms it radiates. In a homogeneous medium, the specified intensity is obtained (the
reference), but with other objects and boundaries present the actual intensity is
different. The source defines the following monopole amplitude:

iφ
S = e d src 2ρ c c c I rms

where dsrc is the distance from the source where the intensity Irms is specified and φ
is the phase of the source. Enter values or expressions for:

• Free space reference intensity (RMS), Irms (SI unit: W/m2).

• Distance from source center dsrc (SI unit: m).
• Phase φ (SI unit: rad).

Power
When Power is selected, specify the source’s reference RMS strength by stating the
power it radiates. In a homogeneous medium, the specified power is obtained (the
reference), but with other objects and boundaries present the actual power is different.
The source defines the following monopole amplitude:

iφ ρ c c c P rms
S = e ---------------------
-
2π

136 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

where Pref denotes the radiated RMS power per unit length measured in W/m. Enter
values or expressions for:

• Free space reference power (RMS), Prms (SI unit: W).

• Phase φ (SI unit: rad).

Bessel Panel: Application Library path Acoustics_Module/

Tutorials,_Pressure_Acoustics/bessel_panel

Point sources in 2D components also exist. They actually correspond to

infinite out-of-plane line sources, see Point Sources (for 2D
Components).

Dipole Point Source

Use the Dipole Point Source node to add a dipole point source. A dipole is
mathematically a source that corresponds to two monopoles close to each other that
are completely out of phase. Dipoles appear when there are fluctuating forces in the
medium; for example, a small object that vibrates back and forth. A complex acoustic
source may be expanded and approximated by a collection of point sources (Monopole
Point Source, Dipole Point Source, and Quadrupole Point Source). The dipole point
source adds a point source term to the right-hand side of the governing Helmholtz
equation such that:

2
k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = ------ D ⋅ ∇δ ( x – x 0 )
1 4π
 ρc  ρc ρc

where δ(x − x0) is the delta function in three dimensions and adds the source at the
point where x = x0. The dipole moment vector D (SI unit: N) depends on the source
type selected, as discussed below. In 2D axisymmetric models, the dipole point source
is only added to the z-axis, such that x0 = (0,0,z). See Ref. 5 for details.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 137

 Figure 2-3: Schematic illustration of a dipole source.

DIPOLE POINT SOURCE

Select a Type: Power (the default) or User defined.

User Defined
For User defined enter a Dipole moment vector D (SI unit: N). In 2D axisymmetric
components enter the z-component only.

Power
The Power option defines the following dipole moment vector in terms of the free
space reference power, the dipole direction, and the source phase

3
iφ e D 2 3ρ c c c P rms
D = D e --------- D = --------------------------
eD 2πω
2

Enter the following:

• Free space reference power (RMS), Prms (SI unit: W). In a homogeneous medium,
the specified power is radiated (the reference), but with other objects and
boundaries present the actual power is different.
• The source Phase φ (SI unit: rad).
• Dipole moment direction eD. Defines the direction of the dipole moment vector. In
2D axisymmetric components the direction is aligned along the z-axis such that
eD = ez.

Point sources in 2D components also exist. They actually correspond to

infinite out-of-plane line sources, see Point Sources (for 2D
Components).

138 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Quadrupole Point Source
Use the Quadrupole Point Source node to enter a quadrupole point source. A
quadrupole is mathematically a source that corresponds to two dipoles close to each
other. A complex acoustic source may be expanded and approximated by a collection
of point sources (Monopole Point Source, Dipole Point Source, and Quadrupole
Point Source). The quadrupole point source adds a point source term to the
right-hand side of the governing Helmholtz equation such that:

2
k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = ------ ( D ⋅ ∇ ) ( d ⋅ ∇ )δ ( x – x 0 )
1 4π
ρc ρc ρc

where δ ( x – x 0 ) is the delta function in three dimensions and adds the source at the
point where x = x0. The quadrupole D-vector D (SI unit: N) and the quadrupole
d-vector d (SI unit: m) depend on the source type selected, as discussed below. In 2D
axisymmetric models, the quadrupole point source is only added to the z-axis, such
that x0 = (0,0,z) and only orientations along the axis are possible. Two types of
predefined quadrupole sources exist, a Power (longitudinal) and a Power (lateral)
configuration. In the longitudinal source, the D and d vectors are parallel and point in
the same direction. In the lateral configuration, the D and d vectors are perpendicular
to each other (see Ref. 5 for details).

QUADRUPOLE POINT SOURCE

Select a Type: Power (longitudinal) (the default), Power (lateral), or User defined. The
lateral options do not exist in 2D-axisymmetric models.

User Defined
For User defined enter the Quadrupole d vector (SI unit: m) and Quadrupole D vector (SI
unit: N). In 2D axisymmetric models enter a Quadrupole amplitude Q (SI unit: N⋅m)
as both the D and d vectors are pointed along the z-axis.

Figure 2-4: Schematic illustration of the user-defined quadrupole point source.

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 139

 Power (Longitudinal)
The Power (longitudinal) defines the following values for the two quadrupole vectors

T iφ
ed = eD Q = d⋅D = Qe ( e d ⋅ e D )

2 5ρ c c c P rms
Q = -------------------------
4
-
2πk

Enter the following:

• Free space reference power (RMS), Prms (SI unit: W). In a homogeneous medium,
the specified power is radiated (the reference), but with other objects and
boundaries present the actual power is different.
• The source Phase φ (SI unit: rad).
• Quadrupole direction, e' D (dimensionless). In a 2D axisymmetric model, the
direction is along the z-axis.

Figure 2-5: Schematic illustration of the longitudinal power quadrupole point source.

Power (Lateral)
The Power (lateral), not defined in 2D axisymmetric components, defines the following
values for the two quadrupole vectors:

T iφ
ed = n × eD Q = d⋅D = Qe ( e d ⋅ e D )

2 15ρ c c c P rms
Q = -----------------------------
4
-
2πk

140 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Enter the following:

• Free space reference power (RMS), Prms (SI unit: W). In a homogeneous medium,
the specified power is radiated (the reference), but with other objects and
boundaries present the actual power is different.
• The source Phase φ (SI unit: rad).
• Quadrupole normal, n (dimensionless). This is the normal to the plane in which the
quadrupole is located.
• Quadrupole direction, e' D (dimensionless). The projection of this vector onto the
plane (defined by the normal n) defines the direction of the dipole D-vector.

Figure 2-6: Schematic illustration of lateral quadrupole point source.

Point sources in 2D components also exist. They actually correspond to

infinite out-of-plane line sources, see Point Sources (for 2D
Components).

Point Sources (for 2D Components)

For 2D components, the monopole, dipole, and quadrupole point sources also exist.
All the sources actually correspond to an infinite line source in the out-of-plane
direction. The monopole is uniform and equally strong in all the in-plane directions —
an infinitely long pulsating cylinder. The dipole source represents an infinitely long
cylinder moving back and forth. The quadrupole represents two dipoles close to each
other.

The point sources add the following point source term to the right-hand side of the
governing Helmholtz equation — for monopole, dipole, and quadrupole point
sources, respectively:

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 141

 4π 4π 4π
------ Sδ ( x – x 0 ) ------ D ⋅ ∇δ ( x – x 0 ) ------ ( D ⋅ ∇ ) ( d ⋅ ∇ )δ ( x – x 0 )
ρc ρc ρc

where δ(x − x0) is the delta function in 2D and adds the source at the point where
(x,y) = x = x0. The monopole amplitude S (SI unit: N/m2), the dipole moment
vector D (SI unit: N/m), and the quadrupole d vector (SI unit: m) and D vectors (SI
unit: N/m) depend on the source type selected.

MONOPOLE POINT SOURCE (2D COMPONENTS)

Select a Type: Flow (the default), Intensity, Power, or User defined. For User defined enter
a Monopole amplitude, Suser (SI unit: N/m).

Flow
Select Flow to add a monopole point source located at r = r0 defined in terms of the
volume flow rate per unit length out from source QS and the phase φ of the source.
The source defines the following monopole amplitude:

iφ iωρ c
S = e ------------ Q S
4π

Enter values or expressions for:

• Volume flow rate out from source, QS (SI unit: m3/s).

• Phase φ (SI unit: rad).

Intensity
Select Intensity to define the source in terms of the free space reference RMS intensity
Irms it radiates. In a homogeneous medium, the specified intensity is obtained (the
reference), but with other objects and boundaries present, the actual intensity is
different. The source defines the following monopole amplitude:

iφ 2ρ c ωI rms d src
S = e --------------------------------
-
2π

where dsrc is the distance from the source where the intensity Irms is specified and φ
is the phase of the source. Enter values or expressions for:

• Free space reference intensity (RMS), Irms (SI unit: W/m2).

• Distance from source center dsrc (SI unit: m).
• Phase φ (SI unit: rad).

142 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Power
When Power is selected, specify the source’s reference RMS strength by stating the
power per unit length it radiates. In a homogeneous medium, the specified power is
obtained (the reference), but with other objects and boundaries present, the actual
source power is different. The source defines the following monopole amplitude:

iφ 2ρ c ωP rms
S = e -------------------------
2
( 2π )

where Prms denotes the free space RMS reference power per unit length measured in
W/m and φ is the phase of the source. Enter values or expressions for:

• Free space reference power (RMS) per unit length, Prms (SI unit: W/m).
• Phase φ (SI unit: rad).

DIPOLE POINT SOURCE (2D COMPONENTS)

Enter a user defined Dipole-moment vector, D (SI unit: N/m).

QUADRUPOLE POINT SOURCE (2D COMPONENTS)

Enter a user defined Quadrupole d vector, d (SI unit: m), and a user defined Quadrupole
D vector, D (SI unit: N/m).

For the Pressure Acoustics, Transient interface, only the Flow (no phase
specification), User defined, and the Gaussian pulse source types are
available. The Gaussian pulse source type has no effect in the frequency
domain. See The Gaussian Pulse Source Type Settings.

Circular Source (for 2D Axisymmetric Components)

Use the Circular Source node to add a source in a 2D axisymmetric component on
points off the axis of symmetry. Such points correspond to circular sources or ring
sources. This type of source is, for example, used to mimic source terms from rotors.
The circular source adds a point source term to the right-hand side of the governing
Helmholtz equation such that:

2
k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = ------ Sδ ( x – x 0 )rdϕ
1 4π
 ρc  ρc ρc

THE PRESSURE ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 143

 where δ ( x – x 0 ) is the delta function that adds the source at the point where x = x0
and rdϕ is the line element around the circular source (SI unit: m). The monopole
amplitude S (SI unit: N/m2) depends on the source type selected, as discussed below.

For the Pressure Acoustics, Transient interface the Flow (no phase
specification), User defined, and the Gaussian pulse source types are
available. The Gaussian pulse source type has no effect in the frequency
domain. See The Gaussian Pulse Source Type Settings.

CIRCULAR SOURCE
Select a Type: Flow (the default) or User defined. For User defined enter a Monopole
amplitude, S = Suser (SI unit: N/m2).

Flow
When Flow is selected the source is defined in terms of the volume flow rate QS per
unit length out form the source. The flow circular source defines the following
monopole amplitude:

iφ iωρ
S = e -----------c- Q S
4π

Enter values or expressions for:

• Volume flow rate per unit length out from source, QS (SI unit: m2/s).
• Phase φ (SI unit: rad).

Pressure (Point Condition)

Use the Pressure condition on points to define pressure monopole-like sources with a
specific pressure. The condition is, for example, useful when setting up room acoustic
models where a prescribed pressure is needed (ideal pressure source). The condition
acts as a constraint.

PRESSURE
Enter a value or expression for the Pressure p0 (SI unit: Pa). In the time domain, the
expression should depend on time.

144 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The Pressure Acoustics, Transient
Interface
The Pressure Acoustics, Transient (actd) interface ( ), found under the
Acoustics>Pressure Acoustics branch ( ) when adding a physics interface, is used to
compute the pressure variation when modeling the propagation of acoustic waves in
fluids at quiescent background conditions. It is suited for time-dependent simulations
with arbitrary time-dependent fields and sources.

The physics interface can be used to model linear and nonlinear acoustics that can be
well described by the scalar pressure variable. Domain conditions include the
Nonlinear Acoustics (Westervelt) Contributions to include nonlinear effects, the
Anisotropic Acoustics feature to model fluids with an effective anisotropic density, and
the Background Pressure Field (for Transient Models) for defining a background
acoustic field to model scattering problems or defining incident waves. User-defined
sources can also be added via the Monopole Domain Source or the Dipole Domain
Source. For open problems, Perfectly Matched Layers (PMLs) can be applied, also in
the time domain for Pressure Acoustics, as efficient nonreflecting boundary
conditions.

The physics interface solves the scalar wave equation in the time domain. Studies for
performing time-dependent modal and modal reduced-order models also exist. The
physics interface also solves in the frequency domain with the available boundary
conditions.

For good modeling strategies, meshing, solvers, postprocessing

information, acoustics specific plots, as well as tips and tricks, see the
Modeling with the Pressure Acoustics Branch (FEM-Based Interfaces)
section.

When this physics interface is added, these default nodes are also added to the Model
Builder — Transient Pressure Acoustics Model, Sound Hard Boundary (Wall), and Initial
Values. Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and source. You can also right-click Pressure Acoustics, Transient
to select physics features from the context menu.

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 145

 SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is actd.

TYPICAL WAVE SPEED FOR PERFECTLY MATCHED LAYERS

Enter a value or expression for the Typical wave speed for perfectly matched layers cref
(SI unit m/s). The default is 343 m/s. It is important to set this speed correctly as the
performance of the Perfectly Matched Layers (PMLs) in the time domain depends only
on the wave speed and not the wavelength. For further details see the Time Domain
Perfectly Matched Layers section.

TRANSIENT SOLVER AND MESH SETTINGS

Enter the Maximum frequency to resolve in the model. The value is used for setting up
the solver as well as the physics-controlled mesh. The default frequency is set to
1000[Hz] but should be changed to reflect the frequency content of the sources used
in the model. Select the Time stepping (method) as Fixed (preferred) the default and
recommended or Free. The Free option is in general not recommended for wave
problems. The generated solver will be adequate in most situations if the
computational mesh also resolves the frequency content in the model. Note that any
changes made to these settings (after the model is solved the first time) will only be
reflected in the solver if Show Default Solver or Reset Solver to Defaults is selected in the
study. It is also important to Reset Solver to Defaults if the Nonlinear Acoustics
(Westervelt) Contributions feature is added as special handling of the nonlinear term
is enabled. For highly nonlinear problems set up with user-defined terms, manual
tuning of the solver may be necessary. In nonlinear models the maximum frequency to
resolve should be selected based on the number of harmonics to be resolved.

Further details on Time Stepping: Resolving the Waves in Time in the

Modeling with the Pressure Acoustics Branch (FEM-Based Interfaces)
section.

146 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The remainder of the Settings window is shared with The Pressure
Acoustics, Frequency Domain Interface.

• Domain, Boundary, Edge, and Point Nodes for the Pressure Acoustics,
Transient Interface
• Theory Background for the Pressure Acoustics Branch
• Theory for the Perfectly Matched Layers in the Time Domain

• Transient Gaussian Explosion: Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/gaussian_explosion
• Gaussian Pulse Absorption by Perfectly Matched Layers: Pressure
Acoustics, Transient: Application Library path Acoustics_Module/
Tutorials,_Pressure_Acoustics/gaussian_pulse_perfectly_matched_layers
• Nonlinear Acoustics — Modeling of the 1D Westervelt Equation:
Application Library path Acoustics_Module/Nonlinear_Acoustics/
nonlinear_acoustics_westervelt_1d

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 147

 Domain, Boundary, Edge, and Point Nodes for the Pressure Acoustics,
Transient Interface
The Pressure Acoustics, Transient Interface shares most of its nodes with The Pressure
Acoustics, Frequency Domain Interface, except the following:

• Nonlinear Acoustics (Westervelt) • Interior Normal Displacement

Contributions • Normal Acceleration
• Background Pressure Field (for • Normal Velocity
Transient Models)
• Normal Displacement
• Exterior Field Calculation (for
• Transient Pressure Acoustics Model
Transient Models)
• The Gaussian Pulse Source Type
• Incident Pressure Field (for Transient
Settings
Models)
• Interior Normal Acceleration
• Interior Normal Velocity

Domain, Boundary, Edge, Point, and Pair Nodes for the Pressure
Acoustics, Frequency Domain Interface

Transient Pressure Acoustics Model

The Transient Pressure Acoustics Model node adds the equations for primarily
time-dependent (transient) acoustics modeling. This is the scalar wave equation

2
1 - ∂----------
pt
-------- - + ∇ ⋅  –1
--- ( ∇p t – q d ) = Q m
2 2  ρ 
ρc ∂t

where pt is the total acoustic pressure, ρ is the fluid density, c is the speed of sound, qd
is the Dipole Domain Source, and Qm is the Monopole Domain Source. In this
formulation of the wave equation, the speed of sound and density may in general be
space dependent but only slowly varying in time, that is, at a time scale much slower
than the variations in the acoustic signal.

In the Settings window, define the properties for the acoustics model and model inputs
including temperature.

148 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

TRANSIENT PRESSURE ACOUSTICS MODEL
Select a Fluid model: Linear elastic (the default), Viscous, Thermally conducting, Thermally
conducting or viscous, General dissipation, or Ideal Gas. Then see the descriptions for The
Pressure Acoustics, Frequency Domain Interface:

• Linear Elastic Fluid Model

• Thermally Conducting and/or Viscous Fluid Model
• Ideal Gas Fluid Model

See Pressure Acoustics (frequency domain) for details of the fluid model
equations. Nonlinear effects can be included using the Nonlinear
Acoustics (Westervelt) Contributions node.

MODEL INPUTS (IDEAL GAS ONLY)

When Ideal gas is selected as the Fluid model, both the Temperature T and Absolute
pressure pA fields are always enabled. Select, User defined (the default), Common model
input, or an input from another physics interface, if applicable.

In addition, the Temperature T and Absolute pressure pA can be picked up from another
physics interface where the fields have been calculated. For example, select a
temperature field defined by a Heat Transfer interface or a Nonisothermal Flow
interface (if any). Or, if applicable, select a pressure as defined by a Fluid Flow interface
present in the model. For example, select Absolute pressure (spf) to use the absolute
pressure defined by a Laminar Flow interface spf.

The Nonisothermal Flow interface requires the addition of the Heat

Transfer Module or CFD Module.

Details about the Model Input and the Default Model Inputs are found
in the Global and Local Definitions chapter of the COMSOL
Multiphysics Reference Manual.

Nonlinear Acoustics (Westervelt) Contributions

The propagation of finite-amplitude pressure waves cannot be described by the linear
acoustic wave equation (small perturbation assumption) and requires solving the full
nonlinear second-order wave equation. The latter allows a simplification when

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 149

 cumulative nonlinear effects dominate local nonlinear effects, for example, when the
propagation distance is greater than the wavelength. The simplified equation is
referred to as the Westervelt equation and reads

2 2 2
1 - ∂----------
pt  δ - ∂------------
∇p t β ∂ pt
-------- - + ∇ ⋅ – 1
--- ( ∇p t – q d ) – --------  = ----------- ------------ + Q m ,
ρc ∂t
2 2
 ρ ρc ∂t 
2 2
ρ c ∂t
4 2

where δ is the diffusivity of sound and β is the coefficient of nonlinearity. In general,

the linear assumption is only valid when

2
p « ρc

which has a value of approximately 1.5·105 Pa for air and 2·109 Pa for water. So,
whenever the acoustic pressure is say a factor 100 less than these values, the linear
assumption holds. When this is not the case, the Nonlinear Acoustics (Westervelt)
Contributions feature should be added.

The Nonlinear Acoustics (Westervelt) Contributions node adds the nonlinear

contribution (first term on the right-hand side) to the Transient Pressure Acoustics
Model. That is, the linear acoustic wave equation transforms to the nonlinear Westervelt
equation. To add the dissipation term (last term on the left-hand side) select, for
example, the General dissipation as the Fluid model in the Transient Pressure Acoustics
Model.

The resulting equation is nonlinear and therefore does not allow the
application of the superposition principle. In particular, the Background
Pressure Field (for Transient Models) cannot be defined together with
the Nonlinear Acoustics (Westervelt) Contributions feature.

The presence of the Nonlinear Acoustics (Westervelt) Contributions node

changes the default solver settings for the proper treatment of
nonlinearities. It is therefore required to Reset Solver to Defaults whenever
the feature is added to or removed from the model.

150 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 The frequency spectrum of a nonlinear wave model contains not only the
center frequency f0 but also the harmonics N·f0, N = 1,2,3,… generated.
The Maximum frequency to resolve in the Transient Solver and Mesh
Settings section should be specified to resolve a desired number of
harmonics to achieve a certain precision. This is especially important for
shock waves, where the number of harmonics to resolve can exceed 10.

• For the theory background of this nonlinear feature, see Pressure

Acoustics, Transient Equations.

NONLINEAR ACOUSTICS (WESTERVELT)

Specify the coefficient of nonlinearity. The available options are From parameter of
nonlinearity (the default), From ratio of specific heats (for gases), and User defined.

• For the option From parameter of nonlinearity enter the value of the Parameter of
nonlinearity, B/A. The coefficient of nonlinearity is defined as follows

B
β = 1 + -------
2A

The parameter of nonlinearity quantifies the effect of nonlinearity on the local speed
of sound in the fluid. It is expressed as c = c0 + (B/2A)u in the first-order terms,
where u is the acoustic particle velocity.
• For the option From ratio of specific heats (for gases) enter the Ratio of specific heats,
γ. This option is valid for perfect gases under isentropic conditions. Here,
B/A = γ − 1, which yields
γ+1
β = ------------
2

• Choose the User defined option to specify β explicitly.

SHOCK-CAPTURING STABILIZATION
To display this section, click the Show More Options button ( ) and select
Stabilization. Select Enable q-Laplacian relaxation to add some artificial nonlinear
damping to the model (it is turned off per default). This is typically only necessary in
highly nonlinear models. The dissipative term of physical origin will often balance the
shock formation (remember to define the appropriate Fluid model in the Transient

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 151

 Pressure Acoustics Model). The effective diffusivity is tuned with the stabilization by
adding an extra term that reads

– δκ  1 – ∂ ∇p
q–1

∂t 

The highest artificial damping is achieved where the acoustic pressure increases or
decreases the most rapidly. It reaches its maximal values where the pressure endures
discontinuities, that is, where shocks arise. Thus, this technique provides a
shock-capturing stabilization.

Specify the q-Laplacian exponent q and the q-Laplacian factor κ to get the desirable
amount of artificial damping. Note that the damping must not be too high nor too
low. The particular values of q and κ depend on the material and the input signal
frequency. The two parameters that control the stabilization require manual tuning. A
suggested approach is to use a simple 1D model to tune the parameters based on fluid
material properties and frequency content. Use, for example, the Nonlinear Acoustics
— Modeling of the 1D Westervelt Equation model from the application library to do
so.

Nonlinear Acoustics — Modeling of the 1D Westervelt Equation:

Application Library path Acoustics_Module/Nonlinear_Acoustics/
nonlinear_acoustics_westervelt_1d

If the chosen fluid model has no dissipation (Linear elastic (the default) or
Ideal Gas), a default sound dissipation δ = 2·10-5 m2/s will be used in the
artificial damping term.

Background Pressure Field (for Transient Models)

Add a Background Pressure Field node to model an background/incident pressure wave
to study the scattered pressure field ps, which is defined as the difference between the
total acoustic pressure pt and the background pressure field pb:

pt = pb + ps

This feature sets up the equations in a so-called scattered field formulation where the
dependent variable is the scattered field p = ps. In a model where the background

152 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

pressure field is not defined on all acoustic domains (or it is different), continuity is
automatically applied in the total field pt on interior boundaries between domains.

BACKGROUND PRESSURE FIELD

Select the Pressure field type as Plane wave (monochromatic) (the default), Modulated
Gaussian pulse, or User defined.

• For Plane wave (monochromatic) enter a value for the Pressure amplitude p0 of the
wave, select the Speed of sound as User defined (enter a value for c) or From material,
enter a Wave direction vector ek, enter the Signal frequency f0 of the wave, and a
possible Phase φ .
These settings defined a background pressure field pb as a plane propagating wave
of the type:

x ⋅ ek
p b ( x, t ) = p 0 sin  2πf 0  t – ------------ + φ
  c ek  

• For Modulated Gaussian pulse enter a value for the Pressure amplitude p0 of the wave,
select the Speed of sound as User defined (enter a value for c) or From material, enter
a Wave direction vector ek and enter the Signal emitting plane coordinate x0. The
latter two options define the initial location of the pulse and the propagation
direction. When modeling scattering problems, the initial location should be placed
outside the computational domain. Next define the frequency content of the
Gaussian pulse. Select the Gaussian pulse bandwidth as Full width at half maximum,
power (the default), Full width at half maximum, amplitude, Full duration at half
maximum, power, Full duration at half maximum, amplitude, or Standard deviation.
For the first two options enter a frequency bandwidth Δf and for the latter a time
window Δt. Finally, enter the carrier Signal frequency (for time dependent studies) f0.
These settings defined a background pressure field pb as a propagating modulated
Gaussian pulse of the type:

x ⋅ ek
p b ( x, t ) = p 0 G ( x, t ) sin  2πf 0  t – ------------ 
  c ek  

where the standard deviation σ of the Gaussian G(x,t) depends on the bandwidth
option selected. The modulated Gaussian pulse option only gives contributions
when a time dependent model is solved. It has the advantage of having a defined
frequency content and can be used for limited bandwidth simulations. In
combination with the Exterior Field Calculation (for Transient Models) and an FFT
study, the broadband scattering characteristics of an object can be analyzed.

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 153

 • For User defined enter an analytical expression that can be a function of both space
and time or use a solution from a previous study.

ADVANCED SETTINGS
To display this section, when the Plane wave (monochromatic) option is selected, click
the Show More Options button ( ) and select Advanced Physics Options in the Show
More Options dialog box.

Click to select the Use ramp function option (selected per default). With this option
turned on, a ramp is automatically added to smoothly increase the background
pressure field amplitude over the first period T = 1/f0. This option ensures numerical
stability.

Incident Pressure Field (for Transient Models)

The Incident Pressure Field node is a subnode to all nonreflecting boundary conditions
(plane, cylindrical, and spherical wave radiation). From the Physics toolbar, add to
Plane Wave Radiation, Spherical Wave Radiation, or Cylindrical Wave Radiation nodes.
Three options exist for transient models: Plane wave (monochromatic) (the default),
Modulated Gaussian pulse, and User defined.

Plane Wave (Monochromatic)

The same options apply as for the incident fields as for the Background Pressure Field
(for Transient Models) feature.

Modulated Gaussian Pulse

The same options apply as for the incident fields as for the Background Pressure Field
(for Transient Models) feature.

User Defined
For User defined enter an expression that defined the incident wave. The incident
pressure field needs to be defined as a traveling wave of the form

f ( ωt – k ⋅ x )

where ω is the angular frequency and k is the wave vector. The function f is any
function, for example, a sine function. This spatial and temporal dependency is a
requirement for the radiation boundary condition to work properly.

154 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The Gaussian Pulse Source Type Settings
For the Pressure Acoustics, Transient interface, you can in addition to the User Defined
and Flow option select a predefined Gaussian pulse as a source Type for the Line Source,
Line Source on Axis, Monopole Point Source, Point Sources (for 2D Components),
and Circular Source (for 2D Axisymmetric Components) features. This type adds a
source with a Gaussian time profile defined in terms of its amplitude A, its frequency
bandwidth f0, and the pulse peak time tp. Using this source type results in solving a
wave equation of the type:

2
1 - ∂----------
pt
-------- - + ∇ ⋅ –1
--- p  4π
------
2 2  ρ ( ∇ t – q d ) = ρ Sδ ( x – x 0 )
ρc ∂t
ρ ∂ –π f0 ( t – tp )
2 2 2
π 2 –π f0 ( t – tp )
2 2 2

S = ------ ----- [ Ae ] = – A ρ --- f 0 ( t – t p )e

4π ∂t 2

where S is the source strength.

In 3D components (for Line Source), 2D axisymmetric components (for Line Source

on Axis and Circular Source (for 2D Axisymmetric Components)), and 2D
components (for Point Sources (for 2D Components)), all of which are effectively line
sources, enter the following values or expressions:

• The value of the pulse Amplitude A (SI unit: m2/s).

• Frequency bandwidth f0 (SI unit: Hz).
• Pulse peak time tp (SI unit: s) for the duration of the pulse.

In 3D and 2D axisymmetric components for the Monopole Point Source, enter the
following values or expressions:

• The value of the pulse Amplitude A (SI unit: m3/s).

• Frequency bandwidth f0 (SI unit: Hz).
• Pulse peak time tp (SI unit: s) for the duration of the pulse.

Transient Gaussian Explosion: Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/gaussian_explosion

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 155

 Normal Acceleration
The Normal Acceleration adds an inward normal acceleration an(t). Mathematically, the
condition adds the boundary condition:

– n ⋅  – ------ ( ∇p – q d ) = a n ( t )
1
 ρ0 

Alternatively, specify the acceleration a0 of the boundary. The part in the normal
direction is used to define the boundary condition:

– n ⋅  – --- ( ∇p – q d ) = – n ⋅ a 0 ( t )
1
 ρ 

This feature represents an external source term. It can also be used to manually couple
acoustics with a structural analysis for modeling acoustic-structure interaction.

NORMAL ACCELERATION
Select a Type: Inward Acceleration (the default) or Acceleration.

• For Inward Acceleration enter the value of the Inward acceleration an (SI unit: m/s2).
Use a positive value for inward acceleration or a negative value for outward
acceleration.
• For Acceleration enter values for the components of the Acceleration a0 (SI unit: m/
s2).

In transient models all sources need to be functions of time t in order to

represent an acoustic signal. This can, for example, be a harmonic sine
wave defined as sin(omega*t) or any other signal.

Normal Velocity
The Normal Velocity adds an inward normal velocity vn(t) or specify the velocity v0(t)
of the boundary. The part in the normal direction is used to define the boundary
condition. The condition is very similar to the Normal Acceleration condition having
the acceleration given by the time derivative of the velocity an(t) = ∂vn(t)/∂t. This
feature represents an external source term. It can also be used to manually couple
acoustics with a structural analysis for modeling acoustic-structure interaction.

156 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

NORMAL VELOCITY
Select a Type: Inward Velocity (the default) or Velocity.

• For Inward Velocity enter the value of the Inward velocity vn (SI unit: m/s). Use a
positive value for inward velocity or a negative value for outward velocity.
• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s).

Normal Displacement
The Normal Displacement adds an inward normal displacement dn(t) or specify the
displacement d0(t) of the boundary. The part in the normal direction is used to define
the boundary condition. The condition is very similar to the Normal Acceleration
condition having the acceleration given by the double time derivative of the
displacement an(t) = ∂2dn(t)/∂t2. This feature represents an external source term. It
can also be used to manually couple acoustics with a structural analysis for modeling
acoustic-structure interaction.

NORMAL DISPLACEMENT
Select a Type: Inward Displacement (the default) or Displacement.

• For Inward Displacement enter the value of the Inward displacement dn (SI unit: m).
Use a positive value for inward displacement or a negative value for outward
displacement.
• For Displacement enter values for the components of the Displacement d0 (SI unit:
m).

Exterior Field Calculation (for Transient Models)

The Exterior Field Calculation feature adds the equations necessary to postprocess the
exterior field to the computational domain for open problems. The feature is the same
as the Exterior Field Calculation in The Pressure Acoustics, Frequency Domain
Interface. This feature will only generate variables and results that can be used in
postprocessing when the results of a time-dependent simulation have been
transformed to the frequency domain using a Time to Frequency FFT study.

For more detailed information about the FFT study see the Time to
Frequency FFT section in the COMSOL Multiphysics Reference
Manual.

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 157

 The feature only operates on the scattered field variables and is thus especially well
suited for analyzing the results of a scattering problem. Such a model can be set up by
using the Background Pressure Field (for Transient Models) feature combined with a
radiation condition or PMLs. In this way radiation pattern of, for example, a
loudspeaker modeled in the time can still be analyzed when an FFT of the time domain
solution is performed.

Interior Normal Acceleration

The Interior Normal Acceleration node adds a normal acceleration on an interior
boundary and ensures that the pressure is noncontinuous here. This node from the
Interior Conditions submenu. The pressure has a so-called slit condition on this
boundary. This boundary condition can be used to model sources as, for example, the
movement of a speaker cone modeled as a boundary. The condition adds the normal
part of an acceleration a0(t):

– n ⋅  – ----- ( ∇p t – q d ) – n ⋅  – ----- ( ∇p t – q d )
1 1
= n ⋅ a0 ( t ) = n ⋅ a0 ( t )
 ρc   ρc 
up down

Alternatively, specify the inward acceleration an(t). The normal of the boundary is
interpreted as pointing outward.

– n ⋅  – ----- ( ∇p t – q d ) = a n ( t ) – n ⋅  – ----- ( ∇p t – q d )

1 1
= an ( t )
ρc up ρc down

INTERIOR NORMAL ACCELERATION

Select a Type: Acceleration (the default) or Normal acceleration.

• For Acceleration enter values for the components of the Acceleration a0 (SI unit: m/
s2).
• For Normal acceleration enter the value of the Normal acceleration an (SI unit: m/s2)
in the normal direction. The normal is depicted as a red arrow in the graphics
window on the selected boundary.

In transient models all sources need to be functions of time t in order to

represent an acoustic signal. This can, for example, be a harmonic sine
wave defined as sin(omega*t) or any other signal.

158 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Interior Normal Velocity
The Interior Normal Velocity node adds a normal velocity on an interior boundary and
ensures that the pressure is noncontinuous. Add this node from the Interior Conditions
submenu. The pressure has a so-called slit condition on this boundary. The condition
is very similar to the Interior Normal Acceleration condition having the acceleration
given by the time derivative of the velocity.

INTERIOR NORMAL VELOCITY

Select a Type: Velocity (the default) or Nromal velocity.

• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s2).
• For Normal velocity enter the value of the Normal velocity vn (SI unit: m/s2) in the
normal direction. The normal is depicted as a red arrow in the graphics window on
the selected boundary.

Interior Normal Displacement

The Interior Normal Displacement node adds a normal displacement on an interior
boundary and ensures that the pressure is noncontinuous. Add this node from the
Interior Conditions submenu. The pressure has a so-called slit condition on this
boundary. The condition is very similar to the Interior Normal Acceleration condition
having the acceleration given by the double time derivative of the displacement.

INTERIOR NORMAL DISPLACEMENT

Select a Type: Displacement (the default) or Normal displacement.

• For Displacement enter values for the components of the Displacement d0 (SI unit:
m/s2).
• For Normal displacement enter the value of the Normal displacement dn (SI unit: m/
s2) in the normal direction. The normal is depicted as a red arrow in the graphics
window on the selected boundary.

THE PRESSURE ACOUSTICS, TRANSIENT INTERFACE | 159

 The Pressure Acoustics, Boundary
Mode Interface
The Pressure Acoustics, Boundary Mode (acbm) interface ( ), found under the
Acoustics>Pressure Acoustics branch ( ) when adding a physics interface, is used to
compute and identify propagating and nonpropagating modes in waveguides and
ducts by performing a boundary mode analysis on a given boundary. The study is
useful, for example, when specifying sources at inlets or analyzing transverse acoustic
modes in ducts. It is available for 3D and 2D axisymmetric component models.

The physics interface solves the Helmholtz eigenvalue equation on boundaries,

searching for the out-of-plane wave numbers at a given frequency.

When this physics interface is added, these default nodes are also added to the Model
Builder: Pressure Acoustics Model, Sound Hard Boundary (Wall), and Initial Values. Then,
from the Physics toolbar, add other nodes that implement, for example, boundary
conditions and sources. You can also right-click Pressure Acoustics, Boundary Mode to
select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is acbm.

160 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

EQUATION

For 2D axisymmetric components, the Azimuthal mode number m is by

default 0. It is an integer entering the axisymmetric expression for the
pressure:

– i ( k z z + mϕ )
p ( r, z, ϕ ) = p ( r ) e

Change the value as needed. Also see Physics Nodes — Equation Section
in the COMSOL Multiphysics Reference Manual.

SOUND PRESSURE LEVEL SETTINGS

See the settings for Sound Pressure Level Settings for the Pressure Acoustics,
Frequency Domain interface.

DEPENDENT VARIABLES
This physics interface defines one dependent variable (field), the Pressure p. The name
can be changed but the names of fields and dependent variables must be unique within
a model.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) for the
Pressure, the default is Quadratic Lagrange.

• Initial Values
• Boundary, Edge, Point, and Pair Nodes for the Pressure Acoustics,
Boundary Mode Interface
• Pressure Acoustics, Boundary Mode Equations

Initial Values
The Initial Values node adds initial values for the sound pressure. Add more Initial
Values nodes from the Physics toolbar.

THE PRESSURE ACOUSTICS, BOUNDARY MODE INTERFACE | 161

 INITIAL VALUES
Enter a value or expression for the Pressure p (SI unit: Pa) initial value.

Special postprocessing variables exist for the Boundary Mode Acoustics

interface. They are described in Pressure Acoustics, Boundary Mode
Variables.

Eigenmodes in a Muffler: Application Library path Acoustics_Module/

Automotive/eigenmodes_in_muffler

Boundary, Edge, Point, and Pair Nodes for the Pressure Acoustics,
Boundary Mode Interface

Except for Initial Values, The Pressure Acoustics, Boundary Mode

Interface shares all of its feature nodes with the Pressure Acoustics,
Frequency Domain interface. See Domain, Boundary, Edge, Point, and
Pair Nodes for the Pressure Acoustics, Frequency Domain Interface.

Also, for the Pressure Acoustics, Boundary Mode interface, apply the
features to boundaries instead of domains for 3D components.

162 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The Pressure Acoustics, Boundary
Elements Interface
The Pressure Acoustics, Boundary Elements (pabe) interface ( ), found under the
Pressure Acoustics branch ( ) when adding a physics interface, is used to compute
the pressure variation for the propagation of acoustic waves in gases and liquids. It is
suited for all frequency-domain simulations with harmonic variations of the pressure
field. The formulation is based on the boundary element method (BEM) and the
interface is available in 2D and 3D. The physics interface solves Helmholtz equation
for constant-valued material properties and uses the pressure as the dependent variable.

The interface is fully multiphysics enabled and can be coupled seamlessly with the
physics interfaces that are based on the finite element method (FEM). This includes
coupling to vibrating structures with the Acoustic-Structure Boundary multiphysics
coupling, to FEM acoustic domains, with the Acoustic FEM-BEM Boundary
multiphysics coupling, to thermoviscous acoustic domains with the
Acoustic-Thermoviscous Acoustic Boundary multiphysics coupling, and to poroelastic
waves domains with the Acoustic-Porous Boundary multiphysics coupling. This
approach allows modeling in a FEM-BEM framework, using the strength of each
formulation adequately. The BEM-based interface is especially well suited for radiation
and scattering problems.

The advantage of the boundary element method is that only boundaries need to be
meshed and the degrees of freedom (DOFs) solved for are restricted to the boundaries.
This introduces some clear ease-of-use for handling complex geometries. However, the
BEM technique results in fully populated or dense matrices that need dedicated
numerical methods. The BEM method is so to speak more expensive per DOF than
the FEM method but has fewer DOFs. Assembling and solving these can be very
demanding. This means that when solving acoustic models of small and medium size,
The Pressure Acoustics, Frequency Domain Interface will often be faster than solving
the same problem with the BEM interface. The challenge for the FEM interface is to
set up open boundaries, for example, using PMLs, in an efficient way. When the
geometries are complex or two structures are far apart, large air domains need to be
meshed. This costs a lot on the computational side as the frequency is increased.

For acoustically large models (problems that contain many wavelengths, at high
frequency or for large domains) the stabilized formulation option (see Stabilization)

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 163

 ensures efficient convergence at the cost of some additional degrees of freedom. For
low to medium frequencies (small to medium models), running without stabilization
is more efficient. The stabilized formulation only gives a benefit in computing time for
the acoustically large models.

For good modeling strategies, meshing, solvers, postprocessing

information, as well as tips and tricks, see the Modeling with the Pressure
Acoustics Branch (BEM-Based Interface) section.

• Head and Torso HRTF Computation. The Application Library path:

Acoustics_Module/Tutorials,_Pressure_Acoustics/head_torso_hrtf
• Spherical Scatterer: BEM Benchmark. The Application Library path:
Acoustics_Module/Verification_Examples/
spherical_scatterer_bem_benchmark
• Submarine Target Strength. The Application Library path:
Acoustics_Module/Underwater_Acoustics/submarine_target_strength

The governing Helmholtz equation defined by the Pressure Acoustics, Boundary

Element interface is given by:

2
1 2 k eq
– ----- ∇ p t – -------- p t = 0
ρc ρc
(2-1)
ω 2
=  ----
2
pt = p + pb k eq
 c c

where pt is the total acoustic pressure, keq is the wave number, ρc is the density, and cc
is the speed of sound. The subscript “c” denotes that these can be complex-valued
quantities in models with damping. The governing equations and boundary conditions
are formulated using the total pressure pt with a so-called scattered field formulation.
In the presence of a Background Pressure Field defining a background pressure wave
pb (this could, for example, be a plane wave), the total acoustic pressure pt is the sum
of the pressure solved for p (which is then equal to the scattered pressure ps) and the

164 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

background pressure wave: pt = p+pb. The equations then contain the information
about both the scattered field and the background pressure field.

The boundary element method (BEM) used in the Pressure Acoustics,

Boundary Element interface is based on the direct method with Costabels
symmetric coupling. The adaptive cross approximation (ACA) fast
summation method is used for solving the resulting linear system. This
method uses partial assembly of the matrices where the effect of the
matrix vector multiplication is computed. The default iterative solver used
is GMRES. With the built-in multiphysics couplings, it is easy and
seamless to set up problems that combine FEM- and BEM-based physics.
When solving these coupled models, a hybrid solver is automatically
generated with an appropriate preconditioner for the FEM part of the
problem (direct or multigrid), if applicable.

More details about the BEM formulation can be found in The PDE,
Boundary Elements Interface documentation in the COMSOL
Multiphysics Reference Manual.

For a tutorial that uses a BEM-FEM coupling to model an acoustic

problem see the Bessel Panel. Application Library path Acoustics_Module/
Tutorials,_Pressure_Acoustics/bessel_panel

When this physics interface is added, these default nodes are also added to the Model
Builder — Pressure Acoustics, Sound Hard Boundary (Wall), and Initial Values.

Then, from the Physics toolbar, add other nodes that implement boundary conditions.
You can also right-click Pressure Acoustics, Boundary Elements to select physics features
from the context menu. Infinite conditions like a symmetry plane or an infinite sound
hard boundary are defined in the Symmetry/Infinite Boundary Condition section.

• Domain, Boundary, Edge, and Pair Nodes for the Pressure Acoustics,
Boundary Elements Interface
• Theory Background for the Pressure Acoustics Branch

Physics Nodes — Equation Section in the COMSOL Multiphysics

Reference Manual

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 165

 SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is pabe.

EQUATION
Expand the Equation section to see the equations solved for with the Equation form
specified. The default selection is Equation form is set to Study controlled. The available
studies are selected under Show equations assuming. If Frequency domain is selected as
the Equation form the Frequency of the study can be either taken From solver or User
defined.

PHYSICS SYMBOLS
Select the Enable physics symbols check box to display the infinite boundary condition
lines or planes in the geometry.

SOUND PRESSURE LEVEL SETTINGS

The zero level on the dB scale varies with the type of fluid. That value is a reference
pressure that corresponds to 0 dB. This variable occurs in calculations of the sound
pressure level Lp based on the root mean square (rms) pressure prms, such that

p rms
L p = 20 log  ----------
1
with p rms = --- p p∗
 p ref  2

where pref is the reference pressure and the star (*) represents the complex conjugate.
This is an expression valid for the case of harmonically time-varying acoustic
pressure p.

Select a Reference pressure for the sound pressure level based on the fluid type:

• Use reference pressure for air to use a reference pressure of 20 μPa (20·10−6 Pa).
• Use reference pressure for water to use a reference pressure of 1 μPa (1·10−6 Pa).
• User-defined reference pressure to enter a reference pressure pref, SPL (SI unit: Pa).
The default value is the same as for air, 20 μPa.

166 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

SYMMETRY/INFINITE BOUNDARY CONDITION
In this section you can specify infinite boundary conditions for the boundary element
problem. These are conditions that are not added on a boundary but apply to a line in
2D or a plane in 3D. The planes where the conditions are applied are visualized if the
Enable physics symbols is turned on (the default) in the Physics Symbols section.

For pressure acoustics a symmetry condition is mathematically equivalent to a sound

hard boundary, and an antisymmetry condition is equivalent to a sound soft boundary.
In this way you can include a Symmetry/Infinite sound hard boundary or an
Antisymmetry/Infinite sound soft boundary along “infinite lines” (in 2D) or “infinite
planes” in (3D). Only one condition per Cartesian direction can be applied.

For the infinite conditions to be mathematically valid it is important that all sources
and BEM boundaries are located on the same side of (or on) the infinite conditions.
The condition can, for example, not be used as an infinite baffle with sound radiating
through a hole. In this case, the propagation domain is on a different side than the
source.

An important aspect is how the results are visualized, specifically when setting the
Parameter Bounds in the grid dataset. Say the condition represents a symmetry plane
then visualize the solution in the entire domain. On the other hand if the condition
represents an infinite sound hard boundary, like the ground or an infinite baffle, then
only visualize the solution above that plane by setting the Parameter Bounds in the Grid
3D or Grid 2D dataset in Results.

Choose an option from the Condition for the x = x0 plane, Condition for the y = y0
plane, and Condition for the z = z0 plane lists (when applicable). In 2D, these are
out-of-plane surfaces. Choose one of the following options:

• Off, for no symmetry (the default)

• Symmetric/Infinite sound hard boundary
• Antisymmetric/Infinite sound soft boundary

Then enter the value for the plane location x0, y0, or z0 (the default is 0 m). This allows
an offset of the infinite condition planes along the main coordinate axes.

FAR-FIELD APPROXIMATION
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options.

These settings are used for matrix assembly and postprocessing. They allow
characterization of interactions occurring in the boundary element method as

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 167

 near-field or far-field interactions. While the near-field interactions are represented
explicitly, the far-field interactions can be represented in an approximate way. This
approach results in considerable memory and performance improvements when used
in combination with iterative solvers using matrix-free format or during
postprocessing. The near-field part of the stiffness matrix is used as input by the Direct
and Sparse Approximate Inverse preconditioners.

The Use far-field approximation check box is selected by default in order to accelerate
the solution process. If the check box is cleared, the solution will be slightly more
accurate but the computational time and memory consumption may become
prohibitively high.

The Approximation type can be either ACA+ or ACA. These alternatives correspond to
two different versions of the adaptive-cross-approximation (ACA) method, which is a
fast matrix multiplication method based on far-field approximations.

More details are found under Far-Field Approximation Settings in The

PDE, Boundary Elements Interface documentation in the COMSOL
Multiphysics Reference Manual.

CONDITION AT INFINITY
In this section, you specify the condition to apply at infinity for an unbounded
problem. For the Helmholtz equation, solved here, choose to specify an Outgoing wave
(the default) or an Incoming wave. In almost all cases the Outgoing wave option should
be selected.

QUADRATURE
The quadrature settings are by default set to Automatic. This means that the quadrature
integration order values will follow the element order selection in the Discretization
section. Higher element orders automatically generate higher values for the quadrature
integration orders.

More details are found under Quadrature in The PDE, Boundary

Elements Interface documentation in the COMSOL Multiphysics
Reference Manual.

168 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

STABILIZATION
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box.

For acoustically large models (problems that contain many wavelengths, at high
frequency or for large domains) enable the Stabilized formulation option to ensure
efficient convergence at the cost of some additional degrees of freedom.

When Stabilized formulation is selected, a text field for the Stabilization parameter is
enabled with the default value sqrt(abs(pabe.k[m])). This is a parameter that
should scale inversely with the wavelength. The default gives good performance in
most cases.

If the Stabilized formulation is enabled after a non-stabilized model has

been solved (maybe showing slow convergence) it is important to reset
the solver default. A different solver configuration is used for the
stabilized BEM formulation.

Note that interior boundary conditions like Interior Sound Hard

Boundary (Wall) or Interior Normal Acceleration are not supported with
the Stabilized formulation.

For a model that uses the Stabilized formulation, see the Submarine
Target Strength tutorial. The Application Library path: Acoustics_Module/
Underwater_Acoustics/submarine_target_strength

POSTPROCESSING SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

This setting controls the interpolation that is done close to boundaries (typically in the
last mesh element) between variables on boundaries (these are the variables solved for
in BEM based interfaces) and their value evaluated with the kernel in the domains.
Numerically the kernel evaluation is prone to errors near the boundaries while the
DOFs solved for are exact on the boundaries. Using an interpolation in the inner mesh
element removes singularities in the results when postprocessing values in the domain
near a boundary.

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 169

 Enter an expression for the Interpolation distance from boundary. The entered value
defines the distance beyond which the kernel evaluation is used. The default is 0.5*h
which gives reasonable results. This default corresponds to half the mesh size h. If the
value is set to 0 no interpolation is done.

See also Postprocessing BEM Results in the Modeling with the Pressure
Acoustics Branch (BEM-Based Interface) section.

DISCRETIZATION
From the Dependent variable/Normal boundary flux list, choose from predefined options
for the boundary element discretization order for the dependent variable and the
normal boundary flux. The predefined options represent the suitable combinations of
element orders such as Quadratic/Linear (the default).

The settings under Value types when using splitting of complex variables are important
for sensitivity and optimization computations. See the description of the built-in
operators fsens and fsensimag.

DEPENDENT VARIABLES
This physics interface defines one dependent variable (field), the Pressure p. If
required, edit the name, which changes both the field name and the dependent variable
name. The new field name must not coincide with the name of a field of another type,
or with a component name belonging to some other field.

170 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Domain, Boundary, Edge, and Pair Nodes for the Pressure Acoustics,
Boundary Elements Interface
The Pressure Acoustics, Boundary Elements Interface has the following domain and
boundary conditions (in alphabetic order):

• Background Pressure Field • Normal Acceleration

• Impedance • Normal Displacement
• Initial Values • Normal Velocity
• Interior Normal Acceleration • Pressure
• Interior Normal Displacement • Pressure Acoustics
• Interior Normal Velocity • Sound Hard Boundary (Wall)
• Interior Sound Hard Boundary • Sound Soft Boundary
(Wall)

Pressure Acoustics
The Pressure Acoustics node adds the equations for solving the Helmholtz equation,
defined by Equation 2-1, in the frequency domain using the boundary element
method. The material properties have to be constant valued for each Pressure Acoustics
node added.

Each Pressure Acoustics node adds the necessary BEM equations to the adjacent
boundaries of the selection (with the material properties defined). The domain features
in the BEM interface can be applied to all types of domain selections: solid domains,
finite voids, and the infinite void. Only the exterior boundaries of the selections as well
as applied interior boundary conditions need to be meshed. Finite voids are selections
of domains that have no solids in the geometry but are only defined by a closed set of
boundaries. The Infinite void represents the exterior most infinite domain. This
selection is the typical used for radiation or scattering problems. Remember that the
materials defined in the Materials node need to be applied to the same selections.

See also Selections: Infinite Void and Finite Voids in the Modeling with
the Pressure Acoustics Branch (BEM-Based Interface) section.

In the Settings window, define the properties for the acoustics model and model inputs
including the absolute background pressure and temperature.

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 171

 MODEL INPUTS
The values of the quiescent (or background) Temperature T (SI unit: K) and Absolute
pressure pA (SI unit: Pa) are entered in the Model Inputs section. The values of the
model inputs can influence the material properties like, for example, the density and
speed of sound, depending on their definition in the Materials node. In the BEM
formulation these two values have to be constant valued for each Pressure Acoustics
node added.

PRESSURE ACOUSTICS MODEL

To define the properties of the bulk fluid, select a Fluid model from the list: Linear elastic
(the default), Linear elastic with attenuation, Viscous, Thermally conducting, Thermally
conducting and viscous, or Ideal gas. The settings for the fluid models are the same as
for Pressure Acoustics in The Pressure Acoustics, Frequency Domain Interface.

The fluid models represent different bulk loss or attenuation mechanisms

(applied in a homogenized way) or ways of defining the fluid properties
of the fluid. Some of these models are sometimes referred to as an
equivalent fluid model. The loss model can be a theoretical model or a
model based on measurement data for the attenuation in the fluid.

The theory for the fluid models is in the section Theory for the Equivalent
Fluid Models.

Background Pressure Field

Add a Background Pressure Field node to add a background pressure wave to model
scattering problems solving for the scattered pressure field ps. The total field pt is the
sum of the scattered field and the background pressure field pb:

pt = pb + ps

This feature sets up the equations in a so-called scattered field formulation where the
dependent variable is the scattered field p = ps. When added, the background pressure
field is applied everywhere; it has a locked selection set to all.

In models solved with the boundary element method, it is important that the defined
Background Pressure Field is an exact solution to the governing equation. If this is not

172 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

the case, the results will be unphysical. This means that setting up a user-defined field
needs careful consideration.

BACKGROUND PRESSURE FIELD

Select a Background pressure field type: Plane wave (the default), Cylindrical Wave,
Spherical Wave, or User defined. Most of the options and settings here are the same as
for the Background Pressure Field in The Pressure Acoustics, Frequency Domain
Interface.

The only difference lies in the definition of the material properties for the background
field. In the BEM interface define a Wave number either picked up from a Pressure
Acoustics model in the physics tree (the default), for example, Wave number (pabe/
bpam1), or defined From speed of sound. This option ensures that the background field
is a solution to the governing equation. If several material models are added, they must
have the same speed of sound, to ensure that the background pressure field is a
solution to the underlying Helmholtz equation.

Initial Values
The Initial Values node adds an initial value for the pressure.

Sound Hard Boundary (Wall)

The Sound Hard Boundary (Wall) adds a boundary condition for a sound hard
boundary or wall, which is a boundary at which the normal component of the
acceleration (and thus the velocity) is zero:

– n ⋅  – ----- ∇p t = 0
1
ρc

To apply a sound hard boundary condition to an infinite plane in the boundary

element model, see the Symmetry/Infinite Boundary Condition functionality.

Normal Acceleration
The Normal Acceleration adds a source at an exterior boundary defined by the inward
normal acceleration an or alternatively given through an acceleration a0:

– n ⋅  – ----- ∇p t = a n = – n ⋅ a 0
1
 ρc 

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 173

 NORMAL ACCELERATION
Select a Type: Inward Acceleration (the default) or Acceleration.

• For Inward Acceleration enter the value of the Inward acceleration an (SI unit: m/s2).
Use a positive value for inward acceleration or a negative value for outward
acceleration.
• For Acceleration enter values for the components of the Acceleration a0
(SI unit: m/s2).

Normal Velocity
The Normal Velocity adds a source at an exterior boundary defined by the inward
normal velocity vn or alternatively given through a velocity v0:

– n ⋅  – ----- ∇p t = iωv n = – iω ( n ⋅ v 0 )
1
 ρc 

NORMAL VELOCITY
Select a Type: Inward Velocity (the default) or Velocity.

• For Inward Velocity enter the value of the Inward velocity vn (SI unit: m/s). Use a
positive value for inward velocity or a negative value for outward velocity.
• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s).

Normal Displacement
The Normal Displacement adds a source at an exterior boundary defined by the inward
normal displacement dn or alternatively given through a displacement d0:

– n ⋅  – ----- ∇p t = ( iω ) d n = – ( iω ) n ⋅ d 0
1 2 2
ρc

NORMAL DISPLACEMENT
Select a Type: Inward Displacement (the default) or Displacement.

• For Inward Displacement enter the value of the Inward displacement dn (SI unit: m).
Use a positive value for inward displacement or a negative value for outward
displacement.
• For Displacement enter values for the components of the Displacement d0
(SI unit: m).

174 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Sound Soft Boundary
The Sound Soft Boundary adds a boundary condition for a sound soft boundary, where
the acoustic pressure vanishes: pt = 0. It is an appropriate approximation for a
liquid-gas interface and in some cases for external waveguide ports. To apply a sound
soft condition to an infinite plane in the boundary element model, see the Symmetry/
Infinite Boundary Condition functionality.

Pressure
The Pressure node creates a boundary condition that acts as a pressure source at the
boundary, which means that a constant acoustic pressure p0 is specified and maintained
at the boundary: pt = p0. The pressure can be a complex number and thus includes
phase information.

PRESSURE
Enter the value of the Pressure p0 (SI unit: Pa).

Impedance
The Impedance node adds an impedance boundary condition with the option to select
between several built-in impedance models and engineering relations. All the options
are the same as for the Impedance condition of The Pressure Acoustics, Frequency
Domain Interface. All the settings and models are described there.

Exclude Boundary
The Exclude Boundary feature is used to remove boundaries from the boundary element
problem. The selected boundaries will neither contribute to the assembly of the BEM
problem nor to the solution. The condition can be used to set up models where the
BEM interface is used as an ideal infinite domain in an infinite baffle (infinite half
space) setup when coupled to the Pressure Acoustics, Frequency Domain. In these

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 175

 models, all boundaries of the FEM problem need to be excluded, except for the ones
in the multiphysics coupling.

For a tutorial that uses a BEM-FEM coupling and the Exclude Boundary
feature to model an acoustic problem in an infinite baffle setup see the
tutorial Piezoelectric Tonpilz Transducer with a Prestressed Bolt.
Application Library path Acoustics_Module/Piezoelelctric_Devices/
tonpilz_transducer_prestressed

Interior Sound Hard Boundary (Wall)

The Interior Sound Hard Boundary (Wall) sets up a wall condition on an interior
boundary. Add this node from the Interior Conditions submenu. The pressure is
discontinuous across this boundary.

When postprocessing and plotting variables on boundaries with interior conditions,

the boundary variables (with extension _bnd) give the average value of the two sides.
Use the up (_up) and down (_down) variants to depict the values on both sides of the
boundary.

See also Postprocessing BEM Results in the Modeling with the Pressure
Acoustics Branch (BEM-Based Interface) section.

In the BEM formulation, interior conditions cannot intersect other

(exterior) conditions. It is recommended that this type of model is set up
by using FEM in the region around the interior condition and BEM
outside of that region.

Interior Normal Acceleration

The Interior Normal Acceleration sets up an interior source that can be used to model a
thin vibrating structure, for example, a loudspeaker diaphragm. Add this node from
the Interior Conditions submenu.

The pressure is discontinuous across this boundary. Use the up and down version of
the postprocessing variables to visualize quantities on the two sides of the boundary,
for example, the up side total pressure is pabe.p_t_up.

176 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

INTERIOR NORMAL ACCELERATION
Select a Type: Acceleration (the default) or Normal acceleration.

• For Acceleration enter values for the components of the Acceleration a0

(SI unit: m/s2).
• For Normal acceleration enter the value of the Normal acceleration an (SI unit: m/s2)
in the normal direction.

In the BEM formulation, interior conditions cannot intersect other

(exterior) conditions. It is recommended that this type of model is set up
by using FEM in the region around the interior condition and BEM
outside of that region.

PHYSICS SYMBOLS
Select Show physics symbols to show the normal direction as a symbol (red arrow) in
the Graphics window (selected per default).

Interior Normal Velocity

The Interior Normal Velocity sets up an interior source that can be used to model a thin
vibrating structure, for example, a loudspeaker diaphragm. Add this node from the
Interior Conditions submenu.

The pressure is discontinuous across this boundary. Use the up and down version of
the postprocessing variables to visualize quantities on the two sides of the boundary,
for example, the up side total pressure is pabe.p_t_up.

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 177

 INTERIOR NORMAL VELOCITY
Select a Type: Velocity (the default) or Normal velocity.

• For Velocity enter values for the components of the Velocity v0

(SI unit: m/s).
• For Normal velocity enter the value of the Normal velocity vn (SI unit: m/s) in the
normal direction.

In the BEM formulation, interior conditions cannot intersect other

(exterior) conditions. It is recommended that this type of model is set up
by using FEM in the region around the interior condition and BEM
outside of that region.

PHYSICS SYMBOLS
Select Show physics symbols to show the normal direction as a symbol (red arrow) in
the Graphics window (selected per default).

Interior Normal Displacement

The Interior Normal Displacement sets up an interior source that can be used to model
a thin vibrating structure, for example, a loudspeaker diaphragm. Add this node from
the Interior Conditions submenu.

The pressure is discontinuous across this boundary. Use the up and down version of
the postprocessing variables to visualize quantities on the two sides of the boundary,
for example, the up side total pressure is pabe.p_t_up.

178 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

INTERIOR NORMAL DISPLACEMENT
Select a Type: Displacement (the default) or Normal displacement.

• For Displacement enter values for the components of the Displacement d0

(SI unit: m).
• For Normal displacement enter the value of the Normal displacement dn (SI unit: m)
in the normal direction.

In the BEM formulation, interior conditions cannot intersect other

(exterior) conditions. It is recommended that this type of model is set up
by using FEM in the region around the interior condition and BEM
outside of that region.

PHYSICS SYMBOLS
Select Show physics symbols to show the normal direction as a symbol (red arrow) in
the Graphics window (selected per default).

THE PRESSURE ACOUSTICS, BOUNDARY ELEMENTS INTERFACE | 179

 The Pressure Acoustics, Time Explicit
Interface
The Pressure Acoustics, Time Explicit (pate) interface ( ), found under the
Acoustics>Pressure Acoustics branch ( ) when adding a physics interface, is used to
compute the pressure variation when modeling the propagation of acoustic waves in
fluids at quiescent background conditions. The interface is used to solve large transient
linear acoustic problems containing many wavelengths. It is suited for time-dependent
simulations with arbitrary time-dependent sources and fields. The interface includes a
Background Acoustic Field option for modeling of scattering problems. Absorbing
layers are used to set up effective nonreflecting-like boundary conditions. The exterior
field can be calculated by combining the Exterior Field Calculation feature with a
Time to Frequency FFT study step. The interface exists in 2D, 2D axisymmetry, and
3D.

The interface is based on the discontinuous Galerkin method (dG-FEM) and uses a
time-explicit solver. The method is very memory lean and is well suited for cluster
computing. Application areas include the transient propagation of audio pulses in
room acoustics or modeling scattering phenomena involving large objects relative to
the wavelength.

If the acoustic waves propagate in a background flow, for example, inside a flowmeter,
then use The Convected Wave Equation, Time Explicit Interface.

For modeling vibroacoustic problems that involve acoustic structure interaction (ASI)
the Pressure Acoustics, Time Explicit interface can be combined with The Elastic
Waves, Time Explicit Interface using the Acoustic-Structure Boundary, Time Explicit
multiphysics coupling.

The interface solves the linearized Euler equations assuming an adiabatic equation of
state. The dependent variables are the acoustic pressure and the acoustic velocity
perturbations. Bulk losses (volume attenuation) can be included in the interface using
the classical expression for thermal and viscous attenuation or a general dissipation
term.

Losses at boundaries can be modeled with impedance conditions. The built-in option
includes frequency independent resistive losses. By combining the interface with a user

180 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

defined Boundary ODEs and DAEs mathematics interface general frequency
dependent conditions can be formulated also for the time domain.

When this physics interface is added, these default nodes are also added to the Model
Builder — Pressure Acoustics Time Explicit Model, Sound Hard Boundary (Wall), and Initial
Values. Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and source. You can also right-click Pressure Acoustics, Time
Explicit to select physics features from the context menu.

For good modeling strategies, meshing, solvers, postprocessing

information, acoustics specific plots, as well as tips and tricks, see the
Modeling with the Pressure Acoustics Branch (DG-FEM-Based
Interface) section.

Domain, Boundary, Edge, and Point Nodes for the Pressure Acoustics,
Time Explicit Interface

For a tutorial model that uses the interface to model a scattering problem,
see the Submarine Scattering: Time-Domain Simulation and FFT
tutorial in the COMSOL Application Gallery:
[Link]

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is pate.

FILTER PARAMETERS FOR ABSORBING LAYERS

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options. In the Filter Parameters for Absorbing Layers section you can change
and control the values set for the filter used in the Absorbing Layers. The values of the
filter parameters defined here are used in all absorbing layers added to the model and

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 181

 they override the value of filter parameters enabled in the material model (Pressure
Acoustics, Time Explicit Model). The default values of the filter parameters α, ηc, and
s are set to 0.1, 0.01, and 2, respectively. Inside the absorbing layer it is important to
use a filter that is not too aggressive since this will result in spurious reflections.

For general information about the filter see the Filter Parameters section
under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

SOUND PRESSURE LEVEL SETTINGS

The settings selected here are only used if the transient solution solved is transformed
into the frequency domain using the Time to Frequency FFT study. The zero level on the
dB scale varies with the type of fluid. That value is a reference pressure that corresponds
to 0 dB. This variable occurs in calculations of the sound pressure level Lp based on
the root mean square (rms) pressure prms, such that

p rms
L p = 20 log  ----------
1
with p rms = --- p p∗
 p ref  2

where pref is the reference pressure and the star (*) represents the complex conjugate.
This is an expression valid for the case of harmonically time-varying acoustic
pressure p.

Select a Reference pressure for the sound pressure level based on the fluid type:

• Use reference pressure for air to use a reference pressure of 20 μPa (20·10−6 Pa).
• Use reference pressure for water to use a reference pressure of 1 μPa (1·10−6 Pa).
User-defined reference pressure to enter a reference pressure pref, SPL (SI unit: Pa). The
default value is the same as for air, 20 μPa.

DISCRETIZATION
In this section you can select the discretization for the Acoustic pressure and Acoustic
velocity. Per default both are set to Quartic (4th order). Using quartic elements
together with a mesh size equal to approximately half the wavelength to be resolved,
leads to the best performance when using the DG method. For further details see the
Meshing, Discretization, and Solvers section.

182 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

DEPENDENT VARIABLES
The dependent variables are the Acoustic pressure, and the Acoustic velocity. The names
can be changed, but the names of fields and dependent variables must be unique within
a model. The name for the Acoustic velocity, components can also be selected
individually.

Domain, Boundary, Edge, and Point Nodes for the Pressure Acoustics,
Time Explicit Interface
The Pressure Acoustics, Time Explicit Interface has the following domain and
boundary conditions listed in alphabetic order, available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users). Continuity in velocity and pressure is the default
condition on interior boundaries. If a boundary separates two different materials the
Material Discontinuity or Continuity (the pair version of material discontinuity)
condition has to be applied to ensure the continuity in velocity and pressure.

• Background Acoustic Field • Initial Values

• Continuity • Mass Source
• Continuity of Total Fields • Material Discontinuity
• Exterior Field Calculation • Normal Velocity
• General Flux/Source • Pressure
• General Interior Flux • Pressure Acoustics, Time Explicit
• Heat Source Model

• Impedance • Sound Hard Boundary (Wall)

• Interior Normal Velocity • Sound Soft Boundary

• Interior Sound Hard Boundary • Symmetry

(Wall) • Volume Force Source

Pressure Acoustics, Time Explicit Model

The Pressure Acoustics, Time Explicit Model node adds the equations to model the
transient propagation of linear acoustics waves based on the dG-FEM time explicit
method. For the time explicit method used, the governing equations are formulated
as a first order system, in terms of the linearized continuity equation and the linearized
momentum equation, as

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 183

 1 - ∂p
-------- --------t + ∇ ⋅ u t = Q m
2 ∂t
ρc
(2-2)
∂u t
ρ -------- + ∇ ⋅ ( p t I ) = q d
∂t

where t is time, pt is the total acoustic pressure, ut is the total acoustic velocity, ρ is the
fluid density, c is the speed of sound, and I is the unit matrix. Domain sources like a
Mass Source, Heat Source, or Volume Force Source can be defined through the right
hand sides Qm and qd. These two source terms represent the same quantities as the
Monopole Domain Source and the Dipole Domain Source in The Pressure Acoustics,
Frequency Domain Interface and The Pressure Acoustics, Transient Interface.

The dependent variables are the acoustic pressure p and the acoustic velocity u
(sometimes called particle velocity or perturbation velocity). The equations are
formulated in the total fields pt and ut which is the sum of the scattered field (the
dependent variable solved for) and a possible background field pb and ub. The
background field can be set up using the Background Acoustic Field feature.

The equations solved are the linearized Euler equations in a quiescent setting. The
equations are easily combined to generate the scalar wave equation solved in The
Pressure Acoustics, Transient Interface as discussed in Theory Background for the
Pressure Acoustics Branch.

In the formulation of the wave equation, the speed of sound c and density ρ may in
general be space dependent but only slowly varying in time, that is, at a time scale
much slower than the variations in the acoustic signal.

The equations solved in the Pressure Acoustics, Time Explicit interface

are closely related to the equations solved in The Convected Wave
Equation, Time Explicit Interface. There is no background flow option
in the Pressure Acoustics, Time Explicit interface but, on the other
hand, it uses a scattered field formulation that allows solving scattering
problems. For nonlinear acoustic problems The Nonlinear Pressure
Acoustics, Time Explicit Interface should be used.

In the Settings window, define the properties for the acoustics model and model inputs
including temperature.

184 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

PRESSURE ACOUSTICS MODEL
Select a Fluid model: Linear elastic (the default), Viscous, Thermally conducting, Thermally
conducting and viscous, General dissipation, or Ideal gas.

• If Linear elastic is selected enter the speed of sound c and the density ρ.
• If Viscous, Thermally conducting, or Thermally conducting and viscous is selected enter
the fluid properties (see the Transient Pressure Acoustics Model for details). These
three options will define the sound diffusivity δ through the classical material
properties. The options are equivalent to defining the classical thermoviscous
attenuation factor atv in the frequency domain.
• If General dissipation is selected enter the speed of sound c, the density ρ, and the
sound diffusivity δ. This option can be used for modeling fluids with measured
damping properties. The relation between the sound diffusivity δ and the equivalent
(plane wave) attenuation coefficient α, at a given frequency f, is given by the
expression
2
δω
α = ---------3-
2c

where ω = 2πf and c is the speed of sound.

• If Ideal gas is selected enter the combination of material properties defining the gas.

For all options, the default is to use the material property values from From material,
select User defined from the list to enter a user-defined value or expression in the text
field that appears. For numerical stability reasons it is recommended to use physical
values of attenuation properties.

Selecting any of the dissipations models will modify the governing Equation 2-2 by
adding a right hand side to the momentum equation defined as

q d = ρδ∇ ⋅ ( ∇u ) = ρδ Δu (2-3)

This term has a small cost on the computation time when solving the model. The term
has to be evaluated at every time-step taken by the solver.

LAX-FRIEDRICHS FLUX PARAMETERS

To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box. In this section, you specify the value of the
Lax-Friedrichs flux parameter τLF (default value: 0.2). This value controls the
numerical flux between the elements (nodal discontinuous Lagrange elements) used
with the discontinuous Galerkin (DG) method. The numerical flux defines how

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 185

 adjacent elements are connected and how continuous p and u are. Different
definitions of the numerical flux lead to different variants of the DG method. The flux
implemented here is the so-called global Lax-Friedrichs numerical flux. The value of
the parameter τLF should be between 0 and 0.5. For τLF = 0 a so-called central flux is
obtained. Setting τLF = 0.5 gives a maximally dissipative global Lax-Friedrichs flux.

For general information about the numerical flux see the Numerical Flux
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

FILTER PARAMETERS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. By default, the filter parameters
α, ηc, and s are not active. Select the Activate check box to activate the filter. The filter
provides higher-order smoothing for the DG formulation. Inside absorbing layers the
settings given here are overridden by the Filter Parameters for Absorbing Layers.

For more detailed information about the filter see the Filter Parameters
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

Background Acoustic Field

Add a Background Acoustic Field node to model a background/incident pressure wave
to study the scattered pressure fields ps and us. The scattered fields are defined as the
difference between the total acoustic fields pt and ut and the background acoustic field
pb and ub:

pt = pb + ps
ut = ub + us

This feature sets up the equations in a so-called scattered field formulation where the
dependent variable is the scattered field (p = ps and u = us). In a model where the
background acoustic field is not defined on all acoustic domains (or it is different)
continuity is automatically applied in the total fields pt and ut on interior boundaries
between domains.

186 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

BACKGROUND ACOUSTIC FIELD
Select the Pressure field type as Plane wave (monochromatic) (the default) or User
defined.

• For Plane wave (monochromatic) enter a value for the Pressure amplitude p0 of the
wave, select the Speed of sound and Density as User defined (enter a value for c or ρ)
or From material, enter a Wave direction vector ek, enter the Signal frequency f0 of
the wave, and a possible Phase φ .
These settings defined a background acoustic field as a plane propagating wave of
the type:

x ⋅ ek pb ek
p b ( x, t ) = p 0 sin  2πf 0  t – ------------ + φ u b ( x, t ) = ------ --------
  c ek   ρc e k

• For Modulated Gaussian pulse enter a value for the Pressure amplitude p0 of the wave,
select the Speed of sound as User defined (enter a value for c) or From material, enter
a Wave direction vector ek and enter the Signal emitting plane coordinate x0. The
latter two options define the initial location of the pulse and the propagation
direction. When modeling scattering problems the initial location should outside
the computational domain. Next define the frequency content of the Gaussian
pulse. Select the Gaussian pulse bandwidth as Full width at half maximum, power (the
default), Full width at half maximum, amplitude, Full duration at half maximum, power,
Full duration at half maximum, amplitude, or Standard deviation. For the first two
options enter a frequency bandwidth Δf and for the latter a time window Δt. Finally,
enter the carrier Signal frequency (for time dependent studies) f0.
These settings defined a background pressure field pb as a propagating modulated
Gaussian pulse of the type:

x ⋅ ek pb ek
p b ( x, t ) = p 0 G ( x, t ) sin  2πf 0  t – ------------  u b ( x, t ) = ------ --------
  c ek   ρc e k

where the standard deviation σ of the Gaussian G(x,t) depends on the bandwidth
option selected. The modulated Gaussian pulse option only gives contributions
when a time dependent model is solved. It has the advantage of having a defined
frequency content and can be used for limited bandwidth simulations. In
combination with the Exterior Field Calculation and an FFT study the broadband
scattering characteristics of an object can be analyzed.
• For User defined enter an analytical expression that can be function of both space and
time or use a solution from a previous study.

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 187

 ADVANCED SETTINGS
To display this section, when the Plane wave (monochromatic) option is selected, click
the Show More Options button ( ) and select Advanced Physics Options in the Show
More Options dialog box.

Click to select the Use ramp function option (selected per default). With this option
turned on, a ramp is automatically added to smoothly increase the background
pressure field amplitude over the first period T = 1/f0. This option ensures numerical
stability.

Initial Values
The Initial Values node adds initial values for the acoustic pressure and the acoustic
velocity that can serve as an initial state for the solver. If more than one initial value is
needed, from the Physics toolbar click to add more Initial Values nodes.

INITIAL VALUES
Enter a value or expression for the initial values for the Pressure p (SI unit: Pa) and the
Acoustic velocity, u (SI unit: m/s).

Mass Source
The Mass Source sets up the right-hand side of the continuity equation, in
Equation 2-2, in terms of a mass rate of change. This is a monopole-like source that
acts uniformly in all directions.

1
Q m = --- M
ρ

MASS SOURCE
Enter the Mass source M (SI unit: kg/(m3·s)).

Heat Source
The Heat Source sets up the right-hand side of the continuity equation, in
Equation 2-2, in terms of a heat density source. This is a monopole-like source that
acts uniformly in all directions.

αp
Q m = ---------- Q heat
ρC p

188 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

where αp is the (isobaric) coefficient of thermal expansion, ρ is the density, Cp is the
heat capacity at constant pressure, and Qheat is the heat source.

MODEL INPUT
Enter the ambient Temperature T, the default is 293.15[K].

HEAT SOURCE
Enter the Heat Source Qheat (SI unit: W/m3).

FLUID PROPERTIES
Enter the Heat capacity at constant pressure Cp and the Ratio of specific heats γ either
selected From material or User defined. The ratio of specific heats is used to calculate
the coefficient of thermal expansion through the thermodynamic relation

1 Cp ( γ – 1 )
α p = --- -----------------------
-
c T

Volume Force Source

The Volume Force Source sets up the right-hand side of the momentum equation, in
Equation 2-2, in terms of a volumetric force. This is a dipole-like source that acts
differently in two directions.

qd = F

VOLUME FORCE SOURCE

Enter the Volume force source F (SI unit: N/m3).

Sound Hard Boundary (Wall)

The Sound Hard Boundary (Wall) adds a boundary condition for a sound hard boundary
or wall, which is a boundary at which the normal component of the velocity is zero (a
slip condition)

–n ⋅ u = 0

Note that this condition is not identical to the Symmetry condition. The condition is
also available from the Pairs submenu, as an option at interfaces between parts in an
assembly. This effectively introduces a condition similar to the Interior Sound Hard
Boundary (Wall) but used on an assembly.

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 189

 Continuity of Total Fields
The Continuity of Total Fields feature is a default feature. The feature automatically
ensures continuity of the total acoustic fields when necessary. The feature becomes
active when a Background Acoustic Field feature is applied on only parts of the model
domains. Note that if two different materials are present in the model the Material
Discontinuity feature needs to be applied.

Sound Soft Boundary

The Sound Soft Boundary conditions adds a boundary condition for a sound soft
boundary, where the acoustic pressure vanishes pt = 0. It is an appropriate
approximation, for example, for a liquid-gas interface or in some cases for external
waveguide ports.

Pressure
The Pressure node creates a boundary condition that acts as a pressure source at the
boundary. The total acoustic pressure is given at the boundary pt = p0(t).

PRESSURE
Enter an expression for the Pressure p0(t) (SI unit: Pa). The input can be any analytical
function or interpolation data that can be function of both time and space.

Symmetry
The Symmetry node adds a boundary condition imposing symmetry in the pressure
and velocity field. Use this condition to reduce the size of a model by cutting it in half
where there are known symmetries. The condition implies symmetry in the acoustic
fields

n ⋅ ut = 0

Normal Velocity
The Normal Velocity condition adds a source specified in terms of an inward normal
velocity vn(t) or velocity v0(t) of the boundary. The component in the normal
direction is used to define the boundary condition. This feature represents an external
source term like a vibrating surface. The condition is also available from the Pairs

190 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

submenu, as an option at interfaces between parts in an assembly. This effectively
introduces a condition similar to the Interior Normal Velocity but used on an assembly.

NORMAL VELOCITY
Select a Type: Inward Velocity (the default) or Velocity.

• For Inward Velocity enter the value of the Inward velocity vn (SI unit: m/s).
• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s).

Impedance
The Impedance condition adds a boundary condition defining the relation between the
local total acoustic pressure pt and the normal acoustic velocity n·ut, that is, the
specific acoustic impedance Zi. The condition specifies the normal velocity

pt
n ⋅ u t = – -----
Zi

This condition can be used to model the properties of artificial boundaries by adding
constant resistive losses (no reactive components can be added). When the impedance
is set equal to the characteristic specific impedance of a propagating plane wave ρc, the
condition represent the simplest nonreflecting boundary condition. This is also the
default value of the impedance when the condition is added.

The acoustic impedance condition (with the default value ρc) should be used at the
outer boundary when Absorbing Layers are used in a model.

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 191

 IMPEDANCE
Enter a value or expression for the Impedance Zi (SI unit: Pa·s/m). The default
expression is pate.c*[Link] which s the characteristic specific impedance ρc of a
propagating plane wave.

As mentioned, the built-in impedance boundary condition only supports

frequency independent resistive impedances (they can be space
dependent). To model more advanced general user-defend frequency
dependent impedance conditions, combine the Normal Velocity
boundary condition with the boundary version of The ODE and DAE
Interfaces (see Distributed ODEs and DAEs). The time explicit
formulation and solver are compatible with solving ODEs on boundaries.

For an example, see the COMSOL Application Gallery tutorial: Full

Wave Time Domain Room Acoustics with Frequency Dependent
Impedance. Found at: [Link]

Exterior Field Calculation

The Exterior Field Calculation feature adds the equations necessary to postprocess the
exterior field to the computational domain for open problems. The feature is the same
as the Exterior Field Calculation in The Pressure Acoustics, Frequency Domain
Interface. This feature will only generate variables and results that can be used in
postprocessing when the results of a time-dependent simulation have been
transformed to the frequency domain using a Time to Frequency FFT study.

For more detailed information about the FFT study see the Time to
Frequency FFT section in the COMSOL Multiphysics Reference
Manual.

The feature only operates on the scattered field variables and is thus especially well
suited for analyzing the results of a scattering problem. Such a model can be set up by
using the Background Acoustic Field feature combined with the Absorbing Layers.
Since the time explicit formulation is very memory lean, scattering problems that

192 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

cannot be solved in the frequency domain can be handled with a time domain
approach and an FFT.

Submarine Scattering: Time Domain Simulation and FFT: In the

COMSOL Application Gallery [Link]/model/55921

Interior Sound Hard Boundary (Wall)

The Interior Sound Hard Boundary (Wall) node, found under the Interior Conditions
submenu, adds the boundary condition for a sound hard boundary or wall on interior
boundaries. The condition reads

u t,up ⋅ n = 0
u t,down ⋅ n = 0

The condition ensures that the total normal velocity is zero on both the up and down
side of the boundary, while applying a slit on the pressure (the pressure is
discontinuous across the boundary).

Interior Normal Velocity

The Interior Normal Velocity node, found under the Interior Conditions submenu, adds
the boundary condition for a moving interior boundary. This can, for example,
represent a thin vibrating structure or s speaker membrane. The condition reads

u t,up ⋅ n = v 0 ( t ) ⋅ n = – v n ( t )
u t,down ⋅ n = v 0 ( t ) ⋅ n = – v n ( t )

The condition ensures that the total normal velocity is defined on both the up and
down side of the boundary, while applying a slit on the pressure (the pressure is
discontinuous across the boundary).

NORMAL VELOCITY
Select the Type: Velocity (default) or Normal velocity and specify either the Velocity
(vector) v0 or the normal component through the Normal velocity vn.

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 193

 Material Discontinuity
The Material Discontinuity node, found under the Interior Conditions submenu, adds an
interior boundary condition to handle jumps in material properties. As the interface is
based on the discontinuous Galerkin method special handling is necessary when a
jump exists in the specific characteristic impedance Z0 = ρc of the medium.

The condition implements continuity in total pressure and in the normal velocity over
the interior boundary:

p t, up = p t, down
n ⋅ u t, up = n ⋅ u t, down

Subscripts “up” and “down” in the equation refer to the up and down sides of the
interior boundary, relative to the normal n.

Continuity
The Continuity node, exists as an option at interfaces between parts in an assembly,
from the Pairs submenu. The condition is similar to the Material Discontinuity
condition that applies to interior boundaries. The difference is that the Continuity
condition should also be applied when the same material exists on both sides of a pair.
The advantage of using a pair feature is that the mesh does not need to be conforming
on the two sides of the interface (the two parts pf the assembly). This is especially
advantageous for the time explicit discontinuous Galerkin method as the time step
depends on mesh size and local speed of sound.

The condition implements continuity in total pressure and in the normal velocity over
the pair:

p t, src = p t, dst
n src ⋅ u t, src = – n dst ⋅ u t, dst

Subscripts “src” and “dst” in the equation refer to the source and destination in the
pair; that is, the two sides in the pair.

For more details on using assemblies and the pair features see Assemblies
and Pair Conditions in the Modeling with the Pressure Acoustics Branch
(DG-FEM-Based Interface) section.

194 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

General Flux/Source
The General Flux/Source node, found under the More submenu, adds the most general
boundary condition for the Pressure Acoustics, Time Explicit model, as it is
formulated (in the conservative form) for the discontinuous Galerkin method (dG).
The condition defines the normal flux gp and gu at an exterior boundary, given by

–n ⋅ Γp = n ⋅ u = gp
n1 p
–n ⋅ Γu = n2 p = gu
n3 p

where u = (u, v, w) and n = (nx, ny, nz) in 3D and 2D, or n = (nr, nϕ, nz) in 2D
axisymmetric. Note that this condition is expressed in terms of the dependent variables
and not the total fields. Care should be taken when using this advanced condition as
the method is sensitive to proper formulation of boundary conditions. Only one
characteristic can enter the domain at any boundary at the time, meaning that it is easy
to overspecify the problem. Use the mesh normals (nxmesh, nymesh, and nzmesh; or
nrmesh, nphimesh, and nzmesh) in the expression you define.

GENERAL FLUX/SOURCE
Enter the expression for the components of the Flux vector gp and gu.

General Interior Flux

The General Interior Flux node, found under the More submenu, adds the most general
interior boundary condition for the Pressure Acoustics, Time Explicit model, as it is
formulated (in the conservative form) for the discontinuous Galerkin method (dG).
The condition defines the normal flux gp and gu at an interior boundary. It should
include any necessary numerical flux (as it overrides the Lax-Friedrichs flux). Care
should be taken when using this advanced condition as the method is sensitive to
proper formulation of boundary conditions.

GENERAL INTERIOR FLUX

Enter the expression for the components of the Flux vector gp and gu.

THE PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 195

 The Pressure Acoustics, Asymptotic
Scattering Interface
The Pressure Acoustics, Asymptotic Scattering (paas) interface ( ), found under the
Acoustics>Pressure Acoustics branch ( ) when adding a physics interface, is used to
model scattering problems at high frequencies using the Kirchhoff-Helmholtz integral
formulation. The acoustic field is assumed to be locally plane such that the scattered
field can be expressed analytically. The scattering object has to be acoustically large; the
wavelength should be much smaller than important geometric features as well as the
radius of curvature of important features. The scattering analysis is only valid for open
iωt
domains. The harmonic variation of all fields and sources is given by e using the
+iω convention.

The scattering object surface can be treated as perfectly reflecting or having absorbing
properties by defining a surface normal impedance, a reflection coefficient, or an
absorption coefficient. The latter two can depend on the angle of incidence. The
interface can model scattering from spherical and plane waves in 3D and 2D
axisymmetry, as well as plane and cylindrical waves in 2D. The interface has built-in
functionality to compute the visibility factor.

When this physics interface is added, these default nodes are also added to the Model
Builder — Pressure Acoustics, Background Pressure Field, Scattering Object (with Wall
sub-feature), and Exterior Field Calculation. Then, from the Physics toolbar, add other
nodes that implement, for example, boundary conditions and source. You can also
right-click Pressure Acoustics, Asymptotic Scattering to select physics features from the
context menu.

Domain and Boundary Nodes for the Pressure Acoustics, Asymptotic

Scattering Interface

Submarine High-Frequency Asymptotic Scattering. The Application

Library path: Acoustics_Module/Underwater_Acoustics/
submarine_asymptotic_scattering

196 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is paas.

SOUND PRESSURE LEVEL SETTINGS

The settings selected here are only used if the transient solution solved is transformed
into the frequency domain using the Time to Frequency FFT study. The zero level on the
dB scale varies with the type of fluid. That value is a reference pressure that corresponds
to 0 dB. This variable occurs in calculations of the sound pressure level Lp based on
the root mean square (rms) pressure prms, such that

p rms
L p = 20 log  ----------
1
with p rms = --- p p∗
p ref 2

where pref is the reference pressure and the star (*) represents the complex conjugate.
This is an expression valid for the case of harmonically time-varying acoustic
pressure p.

Select a Reference pressure for the sound pressure level based on the fluid type:

• Use reference pressure for air to use a reference pressure of 20 μPa (20·10−6 Pa).
• Use reference pressure for water to use a reference pressure of 1 μPa (1·10−6 Pa).
User-defined reference pressure to enter a reference pressure pref, SPL (SI unit: Pa). The
default value is the same as for air, 20 μPa.

VISIBILITY
The high frequency assumption used in the asymptotic scattering interface requires the
computation of the visibility. That is the surface of the scattering object “illuminated”
by the background (incident) acoustic field. Select the Method to compute the visibility
either Angle of incidence (the default) or Hemicube.

• Angle of incidence: COMSOL Multiphysics evaluates the angle of incidence between

the background field and the surfaces directly, without considering which face
elements are obstructed by others. This means that shadowing effects (that is,

THE PRESSURE ACOUSTICS, ASYMPTOTIC SCATTERING INTERFACE | 197

 surface elements being obstructed in nonconvex cases) are not taken into account.
Elements facing away from each other are, however, excluded from the integrals.
The Angle of incidence method is fast and accurate for simple geometries with no or
little shadowing.
• Hemicube: The more sophisticated and general hemicube method uses a z-buffered
projection on the sides of a hemicube (with generalizations to 2D and 1D) to
account for shadowing effects. Think of it as rendering digital images of the
geometry in five different directions (in 3D; in 2D only three directions are
needed), and counting the pixels in each mesh element to evaluate its view factor.
Its accuracy can be influenced by setting the Radiation resolution of the virtual
snapshots. The number of z-buffer pixels on each side of the 3D hemicube equals
the specified resolution squared. Thus the time required to evaluate the irradiation
increases quadratically with resolution. In 2D, the number of z-buffer pixels is
proportional to the resolution property, and thus the time is, as well. Visualizing
results will be slower when the Hemicube option is selected.

DISCRETIZATION
In this section you can select the discretization or element order used for resolving the
pressure and its normal derivative at the scattering surfaces. No dependent variable is
solved by the interface, but the representation of curved surfaces is influenced by this
setting as well as the number of elements needed to resolve the problem.

Domain and Boundary Nodes for the Pressure Acoustics, Asymptotic

Scattering Interface
The Pressure Acoustics, Asymptotic Scattering Interface has the following domain and
boundary conditions listed in alphabetic order, available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

• Background Pressure Field • Scattering Object

• Exterior Field Calculation • Wall
• Pressure Acoustics

198 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Pressure Acoustics
The Pressure Acoustics node defines the properties of the fluid surrounding the
scattering object. This is a default feature. The selection of the feature contains all
domains and voids. Select a Fluid Model following the same options as the Pressure
Acoustics node in The Pressure Acoustics, Frequency Domain Interface.

Background Pressure Field

The Background Pressure Field node defines the background or incident pressure field
that is scattered off the Scattering Object. The Pressure field type can be selected as
Plane wave or Spherical wave. Set the wave amplitude and direction, see the
Background Pressure Field node in The Pressure Acoustics, Frequency Domain
Interface.

Several fields can be defined for computing the scattering from a complex incident
sound field. The surface properties defined on the Wall will take the effects of each
background field into account.

Scattering Object
The Scattering Object node defines the domains and/or finite voids that make up the
scattering object(s). This is a so-called singleton feature that is only added once (per
default) to the model. Note that secondary scattering effects are not taken into account
if several not connected objects are selected, that is, the scattered field of one object is
not scattered of the other objects. The properties of the scattering object surfaces are
defined in the Wall sub-node.

Wall
The Wall sub-node defines the surface properties of the scattering object. Several Wall
sub-features can be added to represent different properties of the boundaries of the
surface. Select the Type as Sound hard wall (the default), Reflection coefficient, Normal
impedance, or Absorption coefficient.

• For Sound hard wall, there is perfect specular reflection with no absorption of energy
at the boundary. The condition corresponds to a reflection coefficient R = 1.
• For Reflection coefficient, enter a value for R (SI unit 1). The reflection coefficient is
a complex valued quantity that can depend on the angle of incidence θi. The angle
is defined through the built-in variable [Link].

THE PRESSURE ACOUSTICS, ASYMPTOTIC SCATTERING INTERFACE | 199

 • For Normal impedance, enter a value for the normal impedance Zn (SI unit: Pas/m).
The reflection coefficient of the surface is computed according to the classical plane
wave reflection expression
Z n cos ( θ i ) – Z
R = ------------------------------------
Z n cos ( θ i ) + Z

where Z = ρc is the specific characteristic impedance of the medium and θi is the

(local) angle of incidence. This corresponds to a locally reacting surface. The normal
impedance is typically complex valued.
• For Absorption coefficient, enter a value for the (normal) absorption coefficient αn
(SI unit: 1) and the phase φ (SI unit: rad). Both can depend on the angle of
incidence in the same way as the reflection coefficient. The reflection coefficient of
the surface is computed according to
iφ
R = 1–α e

Angle dependency of the absorption or reflection coefficient can be

introduced using the angle if incidence variable [Link] or the cosine
to the angle of incidence [Link].

Exterior Field Calculation

The Exterior Field Calculation node is the feature that sets up the Kirchhoff-Helmholtz
integral to perform the postprocessing of the scattered field. It is per default added to
all the scattering boundaries. The selection is not locked so it can be modified but it
should in general coincide with all the Wall boundaries of the Scattering Object. Only
one Exterior Field Calculation feature can be added to the interface.

Enter a name for the Exterior field variable name and select the Type of integral. See the
Exterior Field Calculation node in The Pressure Acoustics, Frequency Domain
Interface. Note that no symmetry options exist for this interface.

200 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The Pressure Acoustics,
Kirchhoff-Helmholtz Interface
The Pressure Acoustics, Kirchhoff-Helmholtz (pakh) interface ( ), found under the
Acoustics>Pressure Acoustics branch ( ) when adding a physics interface, is used to
model radiation problems at high frequencies using the Kirchhoff-Helmholtz integral
formulation. The acoustic field is assumed to be locally plane. This technique is also
often referred to as high frequency BEM, HFBEM, or simply HFB. The method is
often used for computing the radiated acoustic field from vibrating structures at high
frequencies, without the need to model the surrounding fluid. The method is valid as
long as the acoustic wavelength in the fluid is smaller then the structure and the
structural modes. The method is only applicable for open radiation problems. For a flat
vibrating surface the method reduces to solving the Rayleigh integral. A user-defined
option gives access to the full Kirchhoff-Helmholtz integral formulation for defining
both pressure and its normal gradient. The harmonic variation of all fields and sources
iωt
is given by e using the +iω convention.

When this physics interface is added, these default nodes are also added to the Model
Builder — Pressure Acoustics, Radiating Object (with Wall sub-feature), and Exterior Field
Calculation. Then, from the Physics toolbar, add other nodes that implement, for
example, boundary conditions and source. You can also right-click Pressure Acoustics,
Asymptotic Scattering to select physics features from the context menu.

Domain and Boundary Nodes for the Pressure Acoustics, Asymptotic

Scattering Interface

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is pakh.

THE PRESSURE ACOUSTICS, KIRCHHOFF-HELMHOLTZ INTERFACE | 201

 SOUND PRESSURE LEVEL SETTINGS
The settings selected here are only used if the transient solution solved is transformed
into the frequency domain using the Time to Frequency FFT study. The zero level on the
dB scale varies with the type of fluid. That value is a reference pressure that corresponds
to 0 dB. This variable occurs in calculations of the sound pressure level Lp based on
the root mean square (rms) pressure prms, such that

p rms
L p = 20 log  ----------
1
with p rms = --- p p∗
 p ref  2

where pref is the reference pressure and the star (*) represents the complex conjugate.
This is an expression valid for the case of harmonically time-varying acoustic
pressure p.

Select a Reference pressure for the sound pressure level based on the fluid type:

• Use reference pressure for air to use a reference pressure of 20 μPa (20·10−6 Pa).
• Use reference pressure for water to use a reference pressure of 1 μPa (1·10−6 Pa).
User-defined reference pressure to enter a reference pressure pref, SPL (SI unit: Pa). The
default value is the same as for air, 20 μPa.

DISCRETIZATION
In this section you can select the discretization or element order used for resolving the
pressure and its normal derivative at the scattering surfaces. No dependent variable is
solved by the interface, but the representation of curved surfaces is influenced by this
setting as well as the number of elements needed to resolve the problem.

Domain and Boundary Nodes for the Pressure Acoustics, Asymptotic

Scattering Interface
The Pressure Acoustics, Kirchhoff-Helmholtz Interface has the following domain and
boundary conditions listed in alphabetic order, available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

• Exterior Field Calculation • Radiating Object

• Pressure Acoustics • Wall

202 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Pressure Acoustics
The Pressure Acoustics node defines the properties of the fluid surrounding the
scattering object. This is a default feature. The selection of the feature contains all
domains and voids. Select a Fluid Model following the same options as the Pressure
Acoustics node in The Pressure Acoustics, Frequency Domain Interface.

Radiating Object
The Radiating Object node defines the domains and/or finite voids that make up the
radiating object(s). This is a so-called singleton feature that is only added once (per
default) to the model. Note that scattering effects are not taken into account. The
sources associated with the radiating object (pressure or velocity) are defined in the
Wall sub-node.

Wall
The Wall sub-node defines the surface properties of the radiating object. That is, how
the surface sources are defined. Several Wall sub-features can be added to represent
different properties of the boundaries of the surface. Note that the term “wall” here is
very generic as it can also represent, for example, an acoustic source boundary. Select
the Type as Velocity (the default), Normal velocity, or User defined.

• For Velocity, enter the components of the vibrating surface velocity v. This could be
the velocity that results from modeling a vibrating surface like the hull of a ship or
and engine. For example, if the vibrating object is modeled by a Solid Mechanics
interface, then enter the velocities as solid.u_tX, solid.u_tY, and solid.u_tZ,
respectively.
• For Normal velocity, enter the normal velocity vn (SI unit: m/s). This is the normal
velocity pointing into the infinite fluid domain. This can be a constant, an expression
that depends on space coordinates, or it can be taken from the solution of another
model.
• For User defined, enter the pressure p (SI unit: Pa) and t the normal velocity vn (SI
unit: m/s) defined at the given boundary. This option gives full user control of the
input to the Kirchoff-Helmholtz integral that is set up by the Exterior Field
Calculation feature.

THE PRESSURE ACOUSTICS, KIRCHHOFF-HELMHOLTZ INTERFACE | 203

 Exterior Field Calculation
The Exterior Field Calculation node is the feature that sets up the Kirchhoff-Helmholtz
integral to perform the postprocessing of the radiated field under the high frequency
assumption. It is per default added to all the source boundaries. The selection is not
locked, so it can be modified but it should in general coincide with all the Wall
boundaries of the Radiating Object. Only one Exterior Field Calculation feature can be
added to the interface.

For the settings see the Exterior Field Calculation node in The Pressure Acoustics,
Frequency Domain Interface. Note that no symmetry options exist for this interface.

204 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Modeling with the Pressure Acoustics
B r a nc h (FE M - Based In t erfaces)
This section contains good practices and modeling tips and tricks for the finite element
method (FEM) based interfaces of the Pressure Acoustics Branch: The Pressure
Acoustics, Frequency Domain Interface, The Pressure Acoustics, Transient Interface,
and The Pressure Acoustics, Boundary Mode Interface. In this section:

• Physics-Controlled Mesh for Pressure Acoustics

• Meshing: Resolving the Waves in Space
• Lagrange and Serendipity Shape Functions
• Time Stepping: Resolving the Waves in Time
• Frequency Domain, Modal and AWE
• Solving Large Acoustics Problems Using Iterative Solvers
• Perfectly Matched Layers (PMLs)
• Postprocessing Variables
• Evaluating the Acoustic Field in the Exterior: Near- and Far-Field
• Dedicated Acoustics Plots for Postprocessing
• About the Material Databases for the Acoustics Module
• Specifying Frequencies: Logarithmic and ISO Preferred

Physics-Controlled Mesh for Pressure Acoustics

When modeling acoustic problems using the Pressure Acoustics, Frequency Domain
or the Pressure Acoustics, Transient interfaces the computational mesh can be
automatically set up using the Physics-controlled mesh functionality. For these two
physics, the generated computational mesh will follow best practices. The mesh will
automatically resolve the wave, perfectly matched layer (PML) domains will have an
appropriate mapped mesh, a boundary layer mesh is added for the exterior field, and
periodic boundaries will be set up with copy mesh operations.

• First, select the Sequence Type as Physics-controlled mesh (the default) or

User-controlled mesh, see Figure 2-7. When the user-controlled option is selected
the mesh sequence nodes, generated by the physics, are shown and can be modified.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 205

 For a fully user-defined mesh proceed and directly right-click the Mesh node, in the
Model Builder tree, and add the first desired mesh element. The physics that
contribute to the physics-controlled mesh generation can be selected in the
Contributor list.

Overall mesh size

Contributing physics

Options to control
the mesh size and wave
resolution

Figure 2-7: User interface of the Physics-controlled mesh options for Pressure Acoustics as
seen on the Mesh node.

Resolving the wave in space (see also Meshing: Resolving the Waves in Space) requires
and input of the maximum frequency solved for in the model. The way the frequency
is picked up and how many elements are used to resolve the wave, can be controlled in
the physics-controlled mesh user interface, see Figure 2-7.

• Select the Maximum mesh element size control parameter as From study (the default)
or Frequency. For the default From study option, the maximum frequency will, for
frequency domain problems, be picked up from the Frequencies list in the frequency
domain study step; while for time domain problems the value is picked up from the
Maximum frequency to resolve setting in the Transient Solver and Mesh Settings
section. In the frequency domain case enter frequency values in the study step before
generating the mesh. Alternatively select the Frequency option and manually enter

206 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 the Maximum frequency. The manual input can be a good option in models with
several studies to ensure that the mesh is the same for all.
• Select the Number of mesh elements per wavelength as Automatic (the default) or User
defined. The Automatic option gives 5 elements per wavelength.

The mesh also needs to resolve the geometry details in the model. The overall
refinement such a minimum mesh size, maximum element growth rate, curvature
factor, and the resolution of narrow region value are controlled by the Element size
option. The default is to use the Normal setting.

• Acoustic Scattering off an Ellipsoid. The Application Library path:

Acoustics_Module/Tutorials,_Pressure_Acoustics/acoustic_scattering
• Absorptive Muffler. The Application Library path: Acoustics_Module/
Automotive/absorptive_muffler

Physics-controlled mesh for the Pressure Acoustics, Frequency Domain

and Pressure Acoustics, Transient was added in COMSOL Multiphysics
version 6.0. Other acoustics physics interfaces do not support the general
physic-controlled mesh functionality. Correct meshing of PMLs is
supported for all acoustics physics, when applicable.

Meshing in the COMSOL Multiphysics Reference Manual

Meshing: Resolving the Waves in Space

Solutions to acoustic problems are wavelike. The waves are characterized by a
wavelength λ in space, whose value depends on the frequency and speed of sound c in
the medium according to λ = c/f. This wavelength has to be resolved by the mesh.

To represent a wave on a discrete grid (the mesh), it is obvious that the mesh elements
must be smaller than the wavelength in order to resolve the wave. That is, there needs
to be several degrees of freedom (DOFs) per wavelength in the direction of
propagation. In reality, the lower limit for a fully reliable solution lies at about ten to
twelve degrees of freedom per wavelength.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 207

 Because the direction of propagation is generally not known beforehand, it is good
practice to aim for an isotropic mesh with about twelve DOFs per wavelength on
average, independently of the direction. Therefore, the number of DOFs in a
sufficiently resolved mesh is about:

• 1728 = 123 times the model volume measured in wavelengths cubed in 3D.
• 144 = 122 times the model area measured in wavelengths squared in 2D.
• 12 times the model length measured in wavelengths in 1D.

Before starting a new model, try to estimate the required number of DOFs using these
guidelines. The maximum number of DOFs that can be solved for differs between
computer systems. See Solving Large Acoustics Problems Using Iterative Solvers for
solver suggestions.

RESOLVING WAVES WITH LAGRANGE AND SERENDIPITY ELEMENTS

When creating an unstructured mesh for use with the default 2nd-order Lagrange
elements or 2nd-order serendipity elements (see Lagrange and Serendipity Shape
Functions), set the maximum element size hmax to about λ/5 or smaller. Note that this
choice is the default for the Physics-Controlled Mesh for Pressure Acoustics. Because
all elements in the constructed mesh are smaller than hmax, the limit is set larger than
the actual required element size. Note that for specific engineering purposes it is
sometimes possible to use a coarser mesh. This is, for example, the case if only the
sound pressure level (SPL) is of interest as a result, and not the exact phase of the
pressure (correct balance between real and imaginary part of the pressure variable in
the frequency domain). In general a mesh sensitivity analysis should be carried out to
investigate the sensitivity on the solutions parameters of interest. The mesh should also
resolve important geometric features and possible gradients in the material parameters
and model inputs. Geometric features are, for example, curved surfaces or narrow gaps
which need to be resolved adequately. Another important example is if a thin structure
is present (possibly modeled as a boundary using an interior condition). In this case it
is important to have a fine mesh near the edge of the structure to resolve gradients in
the pressure.

Unstructured meshes are generally better than structured meshes for

wave problems where the direction of wave propagation is not known
everywhere in advance. The reason is that in a structured mesh, the
average resolution typically differs significantly between directions parallel
to the grid lines and directions rotated 45 degrees about one of the axes.

208 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 Meshing in the COMSOL Multiphysics Reference Manual

MESHING THE PERFECTLY MATCHED LAYER (PML)

When using a perfectly matched layer (PML) to truncate the computational domain,
it is good practice to use a structured mesh inside the PML region. In 3D models, use
a Swept mesh inside the PML and in 2D models use a Mapped mesh. Use at least 5
elements in the thickness when using rational PML scaling and 8 elements when using
the default polynomial scaling in the PML. The automatic Physics-Controlled Mesh
for Pressure Acoustics set the suggested number of layers. Again, make sure the check
for mesh convergence by adding more layers.

When creating the geometry for your model, use the Layers option to create the
geometry of your PML layer or domain. This will ensure that it is suited for proper
meshing using a structured mesh.

Structured Meshes in the COMSOL Multiphysics Reference Manual

Lagrange and Serendipity Shape Functions

In most of the physics interfaces in the Acoustics Module and specifically in the
Pressure Acoustics interfaces, you can choose between two families of shape functions:
Lagrange and serendipity. The current default is to use Lagrange shape functions.

When using a structured mesh it may be advantageous to switch to the serendipity

elements as they generate significantly fewer degrees of freedom (DOFs). The accuracy
is in most cases almost as good as for the Lagrange elements. The Lagrange elements
are however less sensitive to strong mesh distortions.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 209

 The serendipity shape functions differs from the Lagrange shape functions only for the
following element shapes:

• 2D: Quadrilateral elements of discretization order higher than 1.

• 3D: Hexahedral, prism, and pyramid elements of discretization order higher than 1.

In the COMSOL Multiphysics Reference Manual:

• The Lagrange Element (shlag)

• The Nodal Serendipity Element (shnserp)

When coupling two physics interfaces that have the same DOFs like, for example,
displacement, the same type of shape functions should be used in both interfaces to
ensure conformity. Since there is no difference between the two families of shape
functions in 1D, this is not an issue when connecting edges.

Time Stepping: Resolving the Waves in Time

When solving transient models, first decide on the maximal frequency you want to
resolve, say fmax. This frequency translates to a minimal wavelength λmin = c/fmax and
in turn to a maximum element size hmax < λmin/5 as discussed in Meshing: Resolving
the Waves in Space.

The value of this maximum frequency should also be entered into the Maximum
frequency to resolve field in the Transient Solver and Mesh Settings section at the top
physics level. Here it is also possible to select the Time stepping (method) as either
Fixed (preferred) or Free. It is recommended to use the Fixed (preferred) method as this
method is best suited for wave propagation problems. Using these settings the
generated solver will be adequate in most situations if the computational mesh also
resolves the frequency content in the model, see Meshing: Resolving the Waves in
Space. The auto generated suggestion is good for all linear and weakly nonlinear
problems. If the model studied exhibits high nonlinearities the solver may need manual
setup and tuning.

The logic for the automatic choice made is as follows. The mesh resolution imposes a
restriction on the time-step size Δt taken by the solver. The relationship between mesh
size and time-step size is closely related to the Courant number given by the
Courant-Friedrichs-Lewy (CFL) condition (Ref. 33), which is defined as

210 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 cΔt
CFL = ---------
h

where c is the speed of sound and h is the mesh size. This nondimensional number can
be interpreted as the fraction of an element the wave travels in a single (solver) time
step. The Courant number around 1 would correspond to the same resolution in space
and time if the discretization errors were of the same size; however, that is normally
not the case.

By default, COMSOL Multiphysics uses the implicit second-order accurate method

generalized-α to solve transient acoustics problems. In space, the default is 2nd-order
Lagrange elements. Generalized-α introduces some numerical damping of high
frequencies but much less than the BDF method.

Time Dependent in the COMSOL Multiphysics Reference Manual.

The temporal discretization errors for generalized-α are larger than the spatial
discretization errors when 2nd-order elements are used in space. The limiting step size,
where the errors are of roughly the same size, can be found somewhere at CFL < 0.2.
You can get away with a longer time step if the forcing does not make full use of the
mesh resolution; that is, if high frequencies are absent from the outset.

The condition on the Courant number is, for a time implicit method, not necessary for
stability reasons of the method, but it is used to give a good estimate for accuracy.
Basically, you want to ensure the same numerical resolution in time and space.
Defining an good time step for a model with proper spatial resolution.

When the excitation contains all the frequencies the mesh can resolve, there is no point
in using an automatic time-step control which can be provided by the time-dependent
solver (the Free option). The tolerances in the automatic error control are difficult to
tune when there is weak but important high-frequency content. Instead, you can use
your knowledge of the typical mesh size, speed of sound, and Courant number to
calculate and prescribe a fixed time step. This is exactly the default behavior when the
Fixed (preferred) method is chosen in the Transient Solver and Mesh Settings section.
The Free option corresponds to the automatic time-step control but with some tighter
controls of the allowed time-steps. This latter option is still in general not
recommended as the fixed time stepping option typically yields much better results
(and is faster).

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 211

 The internal time step generated by the Fixed (preferred) option and the entered
Maximal frequency to resolve is set by assuming that the user has generated a mesh that
properly resolves the same maximal frequency (minimal wavelength). The following
step is generated

h max CFL CFL 1

Δ t = ------------------------ = --------------- ≈ -----------------
c f max N 60f max

Assuming that N is between 5 and 6 and the Courant number is roughly 0.1, these
values give a good margin of safety. To check that the accuracy is acceptable, it is
recommended that you run a short sequence of typical excitations with progressively
smaller time steps (larger fmax) and check the convergence.

If a Multiphysics coupling is used to couple, for example, acoustics with

structural interfaces, the correct time-stepping method will be used and it
will be based on the settings in the acoustics interface.

When several acoustics interfaces are present in a model, the logic for the
time step is to use the minimal time step dictated by the contributing
physics.

Some physics, like Electrostatics (present when modeling piezoelectric

problems), suggest a BDF time stepping method to be used. If this
physics is present, the BDF method will be used but the maximal time
steps will be taken from the acoustics suggestions. In this case go to the
Time-Dependent Solver node and change the Method in the Time Stepping
section to Generalized alpha. All other settings will be correct.

Transient Gaussian Explosion: Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/gaussian_explosion

Frequency Domain, Modal and AWE

When solving models in the frequency domain and performing a large frequency
sweep, it can sometimes be useful to use the Frequency Domain, Modal study or the
Asymptotic waveform evaluation (AWE) option in the normal Frequency Domain study.
Both approaches can significantly speed up the solution time when sweeping over
many frequencies.

212 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

For the Frequency Domain, Modal consider:

• The modal solver requires that the eigenfrequencies and modes can be easily found
for up to twice the maximal study frequency. If a model contains many resonances
in the desired frequency range, this may not be a good idea.
• Note that since the modal solver assumes linear perturbation, all sources should be
declared using the linper() operator.

For the Asymptotic waveform evaluation consider:

• When using the AWE option under Study Extensions in the Frequency Domain study,
it is necessary to enter an AWE expression. This is an expression that represents the
response of the system. It needs to be a global expression. This can for example be
the sound pressure level (SPL) evaluated using an integration or average operator.
Integration in a point will give the SPL in that point.

Both solvers are used and discussed in the tutorial: Helmholtz Resonator
Analyzed with Different Frequency Domain Solvers: Application
Library path Acoustics_Module/Tutorials,_Pressure_Acoustics/
helmholtz_resonator_solvers

See also the sections Frequency Domain Study and Frequency Domain,
Modal and Time-Dependent, Modal Studies in the Study Types chapter.

Solving Large Acoustics Problems Using Iterative Solvers

This section has some guidance for solving large acoustics problems. For smaller
problems using a direct solver like MUMPS is often the best choice. For larger
problems, especially in 3D, the only option is often to use an iterative method such as
multigrid.

• Direct Solver Tuning

• Automatically Generated Iterative Solver Suggestions
• Complex Shifted Laplacian for Very Large Frequency Domain Models
• Domain Decomposition for Helmholtz on Clusters

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 213

 • Manual Setup of GMG Solver Suggestions and Theory

In large models with structured mesh you can save DOFs by changing the
default Lagrange elements to serendipity elements. For more information
see Lagrange and Serendipity Shape Functions.

DIRECT SOLVER TUNING

When using the default direct solver MUMPS, one option to reduce computation time
and decrease the memory consumption, is to enable the option Block low rank
factorization on the Direct solver node. This can for many pressure acoustics problems
reduce both memory and computation time by about 20 % to 25 %. This option is not
default, as it is not efficient for all types and combinations of boundary conditions. For
example, it is not suited for models with non-local couplings, like models with periodic
conditions.

For further details see the Direct section in the in the COMSOL
Multiphysics Reference Manual.

For an example that uses the Block low rank factorization see: Test Bench
Car Interior. Application Library path: Acoustics_Module/Automotive/
test_bench_car_interior

AUTOMATICALLY GENERATED ITERATIVE SOLVER SUGGESTIONS

If the direct solver runs out of memory a simple first approach is to enable and use one
of the auto generated iterative solver suggestions. A good starting point for this is to
right-click the study node and select Show Default Solver, then expand the Solver
Configuration tree under Stationary Solver or Time-Dependent Solver. Predefined iterative
solver suggestions are automatically generated. Per default, for a pure Pressure
Acoustics, Frequency Domain model, a direct solver is used and four iterative solvers
are suggested (disabled). To turn on one of these, right-click the solver and select
Enable (or press F4). The four suggestions are:

1 Suggested Iterative Solver (GMRES with GMG): uses the GMRES iterative solver
with a geometric multigrid (GMG) preconditioner. This method is typically faster
than the direct solver and uses less memory for medium to large 3D models. For
details see Manual Setup of GMG Solver Suggestions and Theory.

214 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

2 Suggested Iterative Solver (FGMRES with GMG): uses the FGMRES iterative
solver with a geometric multigrid (GMG) preconditioner. This method is more
robust than GMRES, especially for problems that exhibit sharp resonances. If the
GMRES suggestion does not converge try the FGMRES suggestion instead. For
details see Manual Setup of GMG Solver Suggestions and Theory.
3 Suggested Iterative Solver (Shifted Laplace): For increasing frequencies, the first
two suggested iterative solvers, described above, will eventually stop converging.
One solution is to use the complex shifted Laplacian (CSL or SL) method for the
multigrid preconditioner. The SL method will in general speed up convergence for
larger models. For details see Complex Shifted Laplacian for Very Large Frequency
Domain Models.
4 Suggested Iterative Solver (Domain Decomposition): This last suggestion is for
solving very large models that need to run in a cluster (using a distributed
architecture). The performance of the method will be best when used on several
nodes. For details see Domain Decomposition for Helmholtz on Clusters.

If PMLs are present in the model solved with an iterative method, it is

necessary to use the Polynomial scaling option (the default) and the
recommended 8 mesh layers. This option will ensure proper convergence
of the iterative methods. See the Perfectly Matched Layers (PMLs)
section for further details.

For models that run with a very fine frequency step (with a linear
frequency distribution) the Reuse solution from previous step option is
good. The default Auto (or Yes) options will help convergence by
providing a good initial guess for the iterative solvers. However, in models
with large spacing in the solved frequencies and where frequencies are
given on a logarithmic axis the option should be set to Off. This will speed
up convergence.

For an example that solves a model using an iterative solver see:

Headphone on an Artificial Ear. Application Library path:
Acoustics_Module/Electroacoustic_Transducers/headphone_artificial_ear

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 215

 COMPLEX SHIFTED LAPLACIAN FOR VERY LARGE FREQUENCY DOMAIN
MODELS
When the frequency in the solved problem increases, the first two suggested iterative
solvers, described above, will eventually stop converging. One solution is to use the
complex shifted Laplacian (CSL or SL) method for the multigrid preconditioner. The
SL method will in general speed up convergence for larger models. One exception,
where the SL method will not help, is in models solved near resonances with low
damping from the boundaries. Here, the default GMG suggestion described above will
most likely work better. For a pure Pressure Acoustics model the SL solver suggestion
is set up automatically with appropriate settings.

Manual Setup of the SL Solver

To set up the SL method manually, expand the solver tree and select the Multigrid node
under the first iterative suggestion (enable the suggestion). Select the Prefer the free
matrix format check box and then select the Shifted Laplace contribution check box.
Now select the Physics (a pressure acoustics interface present in the list) and select the
Manual option for the Add weak contributions. In the Weak expression field, enter the
following term:

shift*acpr.p_t*test(p)*(-i*abs(acpr.k[m])^1.5)*[Link]/
acpr.rho_c

where shift is a parameter between −1 and 1 (start with 0.5). This parameter acts as
a relaxation factor. If the physics interface tag is different from acpr, update the
expression to match the physics interface tag. The added weak term does not modify
the solution of the original problem. It is only used on the multigrid levels to help with
convergence.

The same should be done by selecting the Shifted Laplace contribution on multigrid
levels check box. Then select the Manual option and add the same term as Weak
expression.

The first option, Shifted Laplace contribution, adds the SL contribution only on the fine
level used for the smoothers. The second option, Shifted Laplace contribution on
multigrid levels, adds the SL contributions also on all multigrid levels.

To further improve the method at very high frequencies the Hierarchy generation
method option (on the multigrid mode) can be changed from the default Lower
element order first (any) to Coarsen mesh. Then enter an appropriate Mesh coarsening
factor. For acoustics models meshed with λ/N the coarsening can be up to about N.
Typically use N = 5 and set a coarsening factor between 2 and 4.

216 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The default iterative solver is GMRES which can become memory consuming for large
problems. When the problem requires many iterations to be solved, it is typically
necessary to increase the Number of iterates before restart to say 1000. In this case the
iterative solver can be changed to TFQMR. This solver will typically require more
iterations to reach convergence, but it is more memory lean.

Another memory saving option is to use the Block low rank factorization option for the
Direct solver (when using MUMPS) under the Coarse Solver node. Expand the
multigrid solver node to locate it.

For further details see Complex Shifted Laplacian for Large Helmholtz
Problems section in the in the COMSOL Multiphysics Reference
Manual.

For an example that solves a pressure acoustics model using the shifted
Laplacian solver see the COMSOL Application Gallery model entry:
[Link]

DOMAIN DECOMPOSITION FOR HELMHOLTZ ON CLUSTERS

For very large models that can be solved on clusters with many nodes you can set up a
special version of domain decomposition for Helmholtz equation. The method can be
set up to use multigrid with the complex shifted Laplacian method as well as absorbing
boundaries for the domains.

For further details see Complex Shifted Laplacian for Large Helmholtz
Problems section in the in the COMSOL Multiphysics Reference
Manual.

MANUAL SETUP OF GMG SOLVER SUGGESTIONS AND THEORY

The underlying equation the problems solved in pressure acoustics is the Helmholtz
equation. For high frequencies (or wave numbers) the matrix resulting from a
finite-element discretization becomes highly indefinite. In such situations, it can be
problematic to use geometric multigrid (GMG) with simple smoothers such as Jacobi
or SOR (the default smoother). Fortunately, there exist robust and memory-efficient
approaches that circumvent many of the difficulties associated with solving the
Helmholtz equation using geometric multigrid.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 217

 When using a geometric multigrid as a linear system solver together with simple
smoothers, the Nyquist criterion must be fulfilled on the coarsest mesh. If the Nyquist
criterion is not satisfied, the geometric multigrid solver might not converge. One way
to get around this problem is to use GMRES or FGMRES as a linear system solver with
geometric multigrid as a preconditioner. This is the strategy used in the default
iterative solver suggestions.

As a good starting point for setting up the solver manually is to select Show Default
Solver on the main study node and expand the tree. Go to the Stationary Solver and add
an Iterative solver node, per default it uses the GMRES method. The default
preconditioner is the incomplete LU (see the subnode to the Iterative node),
right-click the Iterative solver node and select Multigrid. Even if the Nyquist criterion
is not fulfilled for the coarse meshes of the multigrid preconditioner, such a scheme is
more likely to converge. For problems with high frequencies this approach might,
however, lead to a large number of iterations. Then it can be advantageous to use
either:

• Geometric multigrid as a linear system solver (set the Solver selection to Use
preconditioner) with GMRES as a smoother. Under the Multigrid node right-click
the Presmoother and Postsmoother nodes and select the Krylov Preconditioner with
the Solver selection to GMRES.
• FGMRES as a linear system solver (set the Solver selection to FGMRES) with
geometric multigrid as a preconditioner (where GMRES is used as a smoother, as
above).
Using GMRES or FGMRES as an outer iteration and smoother removes the
requirements on the coarsest mesh. When GMRES is used as a smoother for the
multigrid preconditioner, FGMRES must be used for the outer iterations because such
a preconditioner is not constant (see Ref. 34).

Use GMRES as a smoother only if necessary because GMRES smoothing is very time-
and memory-consuming on fine meshes, especially for many smoothing steps.

When solving large acoustics problems, the following options, in increasing order of
robustness and memory requirements, can be of use:

• If the Nyquist criterion is fulfilled on the coarsest mesh, try to use geometric
multigrid as a linear system solver (set Multigrid as preconditioner and set the linear
system solver to Use preconditioner) with default smoothers. The default smoothers
are fast and have small memory requirements.

218 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• An option more robust than the first point is to use GMRES as a linear system solver
with geometric multigrid as a preconditioner (where default SOR smoothers are
used). GMRES requires memory for storing search vectors. This option can
sometimes be used successfully even when the Nyquist criterion is not fulfilled on
coarser meshes. Because GMRES is not used as a smoother, this option might find
a solution faster than the next two options even if a large number of outer iterations
are needed for convergence.
• If the above suggestion does not work, try to use geometric multigrid as a linear
system solver with GMRES as a smoother.
• If the solver still has problems converging, try to use FGMRES as a linear system
solver with geometric multigrid as a preconditioner (where GMRES is used as a
smoother).
• Try to use as many multigrid levels as needed to produce a coarse mesh for which a
direct method can solve the problem without using a substantial amount of
memory.
• If the coarse mesh is still too fine for a direct solver, try using an iterative solver with
5–10 iterations as coarse solver.

Studies and Solvers and Multigrid in the COMSOL Multiphysics

Reference Manual

Perfectly Matched Layers (PMLs)

The perfectly matched layer (PML) is a domain or layer (sometimes called sponge
layer) that is added to an acoustic model to mimic an open and nonreflecting infinite
domain. It sets up a perfectly absorbing domain as an alternative to nonreflecting
boundary conditions. The PML works with all types of waves, not only plane waves. It
is also efficient at very oblique angles of incidence. In the frequency domain the PML
imposes a complex-valued coordinate transformation to the selected domain that
effectively makes it absorbing at a maintained wave impedance, and thus eliminating
reflections at the interface. In the time domain, additional equations are solved in the
PML for the inverse Laplace transformed equations.

A Perfectly Matched Layers ( ) is added to the model in the Definitions node in the
component where the physics is solved. In the frequency domain the PMLs can be
used for the Pressure Acoustics, Acoustic-Structure Interaction, Aeroacoustics, and
Thermoviscous Acoustics interfaces. In the time domain the PMLs only exist for the

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 219

 Pressure Acoustics, Transient interface. Note that for Pressure Acoustics, Frequency
Domain models, the Perfectly Matched Boundary is an alternative to the PML for
radiation and scattering problems. This boundary conditions applies the PML
automatically using the extra dimension machinery of COMSOL Multiphysics.

In this section:

• Geometry of the PML Layer

• Geometry Type (User Defined Option)
• Meshing the PMLs
• Coordinate Stretching, Scaling, and Curvature
• Limitations of the Perfectly Matched Layers
• Time Domain Perfectly Matched Layers

GEOMETRY OF THE PML LAYER

When creating the geometry for your model, it is advantageous to use the Layers
feature in the geometry to create the PML domains. This ensures that the geometry is
suited for a structured mesh. The physical thickness of the layers is not important in
frequency domain models. Here a real stretching is applied to mathematically scale the
thickness relative to the wavelength. The thickness should however be such that the
mesh is more or less regular (avoid too thin mesh elements). In the time domain the
thickness is important, see Time Domain Perfectly Matched Layers for details.

If the PML is located close to a radiating source or a scatterer, evanescent wave

components can interact with the PML stretching and generate unphysical reflections.
This can be avoided by tuning the Coordinate Stretching, Scaling, and Curvature
parameters. To further prevent this, it is also recommended to place the PML more
than λ/8 away from these surfaces, but it is not necessary, if the PML parameters are
tuned correctly.

GEOMETRY TYPE (USER DEFINED OPTION)

When setting up a PML, you select the geometry type of the layer. This is only related
to hoe the layer looks in the geometry. Typically, the predefined options Cartesian,
Cylindrical, or Spherical can apply in most situations. Using these, COMSOL will
automatically detect the layer thickness and define the local coordinates inside the
PML. In some cases the automatic detection can fail (this can, for example, happen for
certain imported CAD geometries). The automatic detection also fails if the domain is
not the outer most entity in the geometry.

220 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

A workaround, when the automatic detection fails, is then to use the User defined
geometry type. This advanced option makes it possible to define the local Distance
functions and layer Thickness manually. For example, for a spherical PML geometry
the typical distance function is sqrt(x^2+y^2+z^2)-r0, where r0 is the radius of the
inner domain. The user-defined option can also be used for special layer shapes.

To verify that the geometry detection is correct, or a user defined geometry type is set
up correctly, it can be useful to plot the normalized distance functions of the PMLs.
The values should lie be between 0 and 1. Select Get Initial Value on the study (it is not
necessary to solve) and then plot the variable <tag>.sDist<i>, where <tag> is the
PML tag (pml1, pml2, etc.) and a number <i> (1, 2, etc.) is added if several stretching
functions are used in the PML, for example in a corner.

Infinite Elements, Perfectly Matched Layers, and Absorbing Layers in the

COMSOL Multiphysics Reference Manual

In the following model a user defined geometry type is used: Headphone

on an Artificial Ear: Application Library path Acoustics_Module/
Electroacoustic_Transducers/headphone_artificial_ear

MESHING THE PMLS

Optimal behavior of the PML is achieved when the mesh inside the PML domain is
structured. Use a mapped mesh in 2D models and a swept mesh in 3D models. Use at
least 8 layers when using the default polynomial stretching option. As a good starting
point for the rational stretching use 5 or 6 mesh layers inside the PML.

COORDINATE STRETCHING, SCALING, AND CURVATURE

The choice of the Coordinate stretching type and the PML scaling factor and the PML
curvature parameter depends on the problem at hand. A detailed description is given
in the PML Implementation section of the COMSOL Multiphysics Reference
Manual. In general, the Rational stretching option is used for open radiation problems
for propagating waves (it is efficient for many angles of incidence). The Polynomial
stretching option is good for systems with a mix of different wave types, for example,
in multiphysics problems involving structural and acoustic waves, or problems
containing a combination of propagating and evanescent waves. For the polynomial
stretching type, the PML scaling curvature parameter can in general be increase to a
value between 3 and 5 for better performance at low frequencies. Note that when

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 221

 solving a model using an iterative solver the Polynomial scaling should always be used
to ensure convergence.

The polynomial stretching should also be used at the end of waveguides. However, in
pressure acoustics models you should use the Port condition as the waveguide
termination as is provides a superior nonreflecting condition.

There is also a User Defined coordinate stretching type which allow users to define
advanced stretching functions to handle special cases. The stretching can in this way be
optimized to a special problem.

To ensure that the PML is working optimally, it is good practice to make a mesh
convergence test by refining (adding more layers to) the structured mesh in the PML
domain. This is especially important at low frequencies, where evanescent waves may
interact with the PML and give erroneous solutions.

The behavior of the PMLs at low frequencies is discussed in the following

model. Plotting the total radiated power can be a good indicator of
possible issues. Lumped Loudspeaker Driver: Application Library path
Acoustics_Module/Electroacoustic_Transducers/lumped_loudspeaker_driver

Infinite Elements, Perfectly Matched Layers, and Absorbing Layers in the

COMSOL Multiphysics Reference Manual

The PMLs damp a certain wavelength existing in the system. The

wavelength is deducted from the frequency and a reference wave speed
cref. The wave speed is defined in the Typical Wave Speed section. Set cref
equal to the speed of sound of the material in the PML.

When multiple physics and materials are involved, be cautious. Depending

on the configuration, use several PMLs and PML domains or a single
PML. For details, see Configuration of Perfectly Matched Layers (PMLs)
for Acoustic-Structure Interaction Models in the Configuration of
Perfectly Matched Layers (PMLs) for Acoustic-Structure Interaction
Models section.

222 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

LIMITATIONS OF THE PERFECTLY MATCHED LAYERS
When a model contains a Background Pressure Field and PMLs, certain configurations
will create incompatibilities that lead to erroneous behavior. The problem arises if a
domain with a background pressure field is next to a domain without the feature (for
example when setting up absorption problems) and the two domains have a common
PML attached to them. Meaning that the PML next to the background pressure field
touches the PML next to the domain without the background pressure field. In this
case, there is an incompatibility at the common edge of the PMLs. In one PML
domain the pressure DOF is interpreted as a scattered field, while it is the total field in
the other. Note that you can set up models that contain this feature configuration as
long as the PMLs do not touch.

When a perfectly matched layer (PML) is present in the model do not

apply an Incident Pressure Field on its outer boundaries. The PML is
applied to absorb waves that move out of the computational domain.
Defining an incident field on its boundary will lead to unphysical results.

TIME DOMAIN PERFECTLY MATCHED LAYERS

In the time domain the PML does not include a real stretching component. This
means that the geometrical thickness L, of the layer in the geometry, needs to be set
adequately. When meshing the PMLs for time domain simulations, it is recommended
to use a structured mesh in the same way as in the frequency domain. Use at least 8
mesh layers for the rational scaling and 6 for the polynomial scaling and the same mesh
element size as that in the adjacent physical domain (a detailed investigation is available
in Ref. 41).

The frequency content of the transient signal should be considered when defining the
layer thickness L. Assume most of th energy is carried in a band from fmin to fmax. This
represents wavelengths from λmax to λmin (remember λ = c/f). The geometrical
thickness of the layer has two considerations to fulfill:

1 The thickness should be larger than λmax/8

2 The thickness should be such that you have a minimum of 6 (or 8 depending on the
scaling) mesh elements in the thickness at fmax. The mesh in the layer has to also
resolve the minimum wavelength.

In the limit of a signal with a narrow band frequency content (or a single frequency),
the two points will be fulfilled by defining a thickness L = λ.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 223

 The recommended values of the PML scaling factor and the PML scaling curvature
parameter are 1, 3 and 1, 1 for the Polynomial and the Rational stretching types,
respectively. For the Polynomial stretching, the PML scaling factor equal to 1
corresponds to the theoretical reflection coefficient R0 = 10-3 from the interface
between the physical domain and the PML for a plane wave.

Note that the absence of a real stretching makes the time domain PMLs unable to
efficiently absorb evanescent waves.

In the settings for the Perfectly Matched Layer it is important that if you
change the Typical wavelength from option to User defined, then it is not
the actual wavelength that should be entered but rather the speed of
sound per Hertz. For example, if User defined is selected in a normal air
domain, then enter 343[m/s]/1[Hz]. The reason is that in the time
domain the PML is not related to wavelength but to speed of sound.
Transient signals typically include many Fourier frequency components.

See Theory for the Perfectly Matched Layers in the Time Domain for
details about the implementation.

Postprocessing Variables
Several specialized variables specific to acoustics are predefined in the Acoustics
Module and can be used when analyzing the results of an acoustic simulation. The
variables are available from the expression selection menus when plotting.

In this section:

• Intensity Variables
• Power Dissipation Variables
• Pressure Acoustics, Boundary Mode Variables

In the COMSOL Multiphysics Reference Manual:

• Results Analysis and Plots

• Operators, Functions, and Constants

224 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

INTENSITY VARIABLES
The propagation of an acoustic wave is associated with a flow of energy in the direction
of the wave motion, the intensity vector I. The sound intensity in a specific direction
(through a specific boundary) is defined as the time average of (sound) power per unit
area in the direction of the normal to that area.

Knowledge of the intensity is important when characterizing the strength of a sound

source — that is, the power emitted by the source. The power is given by the integral
of n·I on a surface surrounding the source, where n is the surface normal. The
intensity is also important when characterizing transmission phenomena, for example,
when determining transmission loss, insertion loss curves, or absorption.

The acoustic intensity vector I (SI unit: W/m2) is defined as the time average of the
instantaneous rate of energy transfer per unit area (sound power) pu, such that

T
1

I = ---- pu dt
T
(2-4)
0

where p is the pressure and u the particle velocity.

Frequency Domain
In the frequency domain (harmonic time dependence), the time averaging integral
Equation 2-4 reduces to

1 1
I = --- Re ( pu∗ ) = --- ( pu∗ + p∗ u ) (2-5)
2 4

where ∗ denotes complex conjugation. In the frequency domain the velocity is

readily expressed in terms of the pressure as u = – 1 ⁄ ( iωρ ) ∇p . Using the specific
characteristic acoustic impedance for plane waves, the intensity can also be expressed
in terms of the root mean square (RMS) pressure as

1 pp∗
I = --- ----------
2 ρc

In the Pressure Acoustics, Frequency Domain interface the general formulation, valid
for all types of waves given in Equation 2-5, is used to define the intensity.

When using the scattered field formulation, by adding a Background Pressure Field
feature, the intensity variables for the total, background, and scattered fields are
available in postprocessing if the Calculate background and scattered field intensity
option is selected.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 225

 Time Domain
For time-dependent problems, the equivalent quantity is the instantaneous value of the
intensity, defined as

i inst = pu

This expression is difficult to recover in pressure acoustics and would require the
solution of an additional PDE to calculate the velocity from the pressure dependent
variable. Only the intensity I (time averaged) is available as postprocessing variable in
the frequency domain and can be selected from the expressions menus when plotting.
The instantaneous intensity does exists as a postprocessing variable in transient
interfaces such as Linearized Euler, Linearized Navier-Stokes, or Thermoviscous
Acoustics where the velocity is solved for explicitly, see Modeling with the
Aeroacoustics Branch or Modeling with the Pressure Acoustics Branch (FEM-Based
Interfaces) for details.

226 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Postprocessing Variables
The variables are defined in Table 2-3, Table 2-4, and Table 2-5. In the variable
names, phys_id represents the interface name, for example, acpr for a Pressure
Acoustics, Frequency Domain interface.
TABLE 2-3: INTENSITY VARIABLES IN 3D

VARIABLE DESCRIPTION

phys_id.I_mag Magnitude of the intensity vector (frequency domain only)

phys_id.Ix x-component of the intensity vector (frequency domain only)
phys_id.Iy y-component of the intensity vector (frequency domain only)
phys_id.Iz z-component of the intensity vector (frequency domain only)

TABLE 2-4: INTENSITY VARIABLES IN 2D AXISYMMETRIC

VARIABLE DESCRIPTION

phys_id.I_mag Magnitude of the intensity vector (frequency domain only)

phys_id.Ir r-component of the intensity vector (frequency domain only)
phys_id.Iz z-component of the intensity vector (frequency domain only)

TABLE 2-5: INTENSITY VARIABLES IN 2D

VARIABLE DESCRIPTION

phys_id.I_mag Magnitude of the intensity vector (frequency domain only)

phys_id.Ix x-component of the intensity vector (frequency domain only)
phys_id.Iy y-component of the intensity vector (frequency domain only)

In the COMSOL Multiphysics Reference Manual:

• Results Analysis and Plots

• Expressions and Predefined Quantities

POWER DISSIPATION VARIABLES

Common to the Pressure Acoustics fluid models (also porous materials) and The
Thermoviscous Acoustics, Frequency Domain Interface is that all the interfaces model
some energy dissipation process, which stem from viscous and thermal dissipation
processes. The amount of dissipated energy can be of interest as a results analysis
variable or as a source term for a multiphysics problem. An example could be to

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 227

 determine the amount of heating in the human tissue when using ultrasound. In the
Acoustics Module special variables exist for the dissipation.

• For the case of a plane wave propagating in the bulk of a fluid (the general thermal
and viscous fluid models described in Thermally Conducting and Viscous Fluid
Model) the dissipation is

4 μb k ( γ – 1 ) ∂p 2
Δ = ----------------2 μ  --- + ------ + --------------------  ------
1
2 3 μ  Cp  ∂t 
( ρc )

and in the frequency domain after averaging over one period

4 μb
2
k(γ – 1) ω
Δ = ----------------2 μ  --- + ------ + -------------------- ------ pp∗
1
(2-6)
2  3 μ  Cp 2
( ρc )

where * in Equation 2-6 is the complex conjugate operator.

• In addition, an approximate expression for the dissipated energy density from a

propagating plane wave exists for the Narrow Region Acoustics, the Poroacoustics
models and attenuation in Pressure Acoustics. This total dissipated power density
Qpw is defined by
Q pw = – 2 I imag ( k )

where I is the magnitude of the intensity vector I, and k is the wave number. This
expression is an approximation and is only valid for traveling plane waves (or waves
that are close to plane); however, it has many uses as a first estimate of the dissipation
since it is easy to calculate in many different situations. The expression is, for
example, not valid for standing waves in resonant systems. When the above
expression is not valid, the dissipated energy should be calculated using an energy
balance approach.

The power dissipation variables are defined in Table 2-6. In the variable names,
phys_id represents the name (acpr, for example, for a pressure acoustics interface).

TABLE 2-6: POWER DISSIPATION VARIABLES

VARIABLE DESCRIPTION

phys_id.diss_therm Thermal power dissipation density

phys_id.diss_visc Viscous power dissipation density
phys_id.diss_tot Total thermal and viscous power dissipation density
phys_id.Q_pw Plane-wave total dissipated power density

228 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

PRESSURE ACOUSTICS, BOUNDARY MODE VARIABLES
A series of special variables exist for postprocessing after solving a boundary mode
acoustics problem. They include in-plane and out-of-plane components of the velocity
v and acceleration a.

The in-plane (ip) and out-of-plane (op) components to the acceleration and velocity
are defined as

a ip = a – ( a ⋅ n )n
v ip = v – ( v ⋅ n )n
a op = ( a ⋅ n )n
v op = ( v ⋅ n )n

where n is the normal to the surface being modeled. The velocity and acceleration are
defined in terms of the gradient of the pressure p as follows

∇p = ∇ || p – ik n pn in 3D

∇p = ∇ || p – ik n p ( n r, 0, n z ) + p  0, – i -----, 0
m
in 2D axisymmetry
r

and

i
v = ------- ∇p and a = iωv
ωρ

where kn is the out-of-plane wave number solved for, m is a possible radial wave mode
number, and ∇ || is the tangential derivative along the boundary.

The boundary mode acoustics variables are defined in Table 2-7. In the variable names,
phys_id represents the name (acbm, for example, for a Boundary Mode Acoustics
interface).
TABLE 2-7: BOUNDARY MODE ACOUSTICS VARIABLES IN 3D

VARIABLE DESCRIPTION

phys_id.vipx In-plane velocity, x-component

phys_id.vipy In-plane velocity, y-component
phys_id.vipz In-plane velocity, z-component
phys_id.vip_rms In-plane velocity RMS value
phys_id.aipx In-plane acceleration, x-component
phys_id.aipy In-plane acceleration, y-component

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 229

 TABLE 2-7: BOUNDARY MODE ACOUSTICS VARIABLES IN 3D

VARIABLE DESCRIPTION

phys_id.aipz In-plane acceleration, z-component

phys_id.aip_rms In-plane acceleration RMS value
phys_id.vopx Out-of-plane velocity, x-component
phys_id.vopy Out-of-plane velocity, y-component
phys_id.vopz Out-of-plane velocity, z-component
phys_id.vop_rms Out-of-plane velocity RMS value
phys_id.aopx Out-of-plane acceleration, x-component
phys_id.aopy Out-of-plane acceleration, y-component
phys_id.aopz Out-of-plane acceleration, z-component
phys_id.aop_rms Out-of-plane acceleration RMS value

Evaluating the Acoustic Field in the Exterior: Near- and Far-Field

The Acoustics Module has functionality to evaluate the acoustic pressure field in the
exterior or far-field region outside of the computational domain. This is the Exterior
Field Calculation feature available for pressure acoustics problems. This section first
gives a definition of the far- and near-field and then some general advice for analyzing
the exterior field.

THE NEAR-FIELD AND FAR-FIELD REGIONS

The solution domain for a scattering or radiation problem can be divided into two
zones, reflecting the behavior of the solution at various distances from objects and
sources. In the far-field region, scattered or emitted waves are locally planar, velocity
and pressure are in phase with each other, and the ratio between pressure and velocity
approaches the free-space characteristic specific impedance of a plane wave rc.

Moving closer to the sources into the near-field region, pressure and velocity
gradually slide out of phase. This means that the acoustic field contains energy that
does not travel outward or radiate. These evanescent wave components are effectively
trapped close to the source. Looking at the sound pressure level, local maxima and
minima are apparent in the near-field region.

Naturally, the boundary between the near-field and far-field regions is not sharp. A
general guideline is that the far-field region is that beyond the last local energy
maximum, that is, the region where the pressure amplitude drops monotonously at a
rate inversely proportional to the distance R from any source or object.

230 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

A similar definition of the far-field region is the region where the radiation pattern —
the locations of local minima and maxima in space — is independent of the distance to
the wave source. This is equivalent to the criterion for Fraunhofer diffraction in optics,
which occurs for Fresnel numbers, F = a2/λR, much smaller than 1. For engineering
purposes, this definition of the far-field region can be applied:

8a 2 8
R > ---------- = ------ ka 2 (2-7)
λ 2π

In Equation 2-7, a is the radius of a sphere enclosing all objects and sources, λ is the
wavelength, and k is the wave number. Another way to write the expression leads to
the useful observation that the size of the near-field region expressed in source-radius
units is proportional to the dimensionless number ka, with a prefactor slightly larger
than one. This relation is known as the Rayleigh radius R0 = S/λ, where S is the
source area, for example for a piston.

Knowing the extent of the near-field region is useful when applying radiation
boundary conditions because these are accurate only in the far-field region. PMLs, on
the other hand, can be used to truncate a domain already inside the near-field region.

THE HELMHOLTZ-KIRCHHOFF INTEGRAL REPRESENTATION

In many cases, solving the acoustic Helmholtz equation everywhere in the domain
where results are requested is neither practical nor necessary. For homogeneous media,
the solution anywhere outside a closed surface containing all sources and scatterers can
be written as a boundary integral in terms of quantities evaluated on the surface. To
evaluate this Helmholtz-Kirchhoff integral, it is necessary to know both Dirichlet and
Neumann values on the surface. Applied to acoustics, this means that if the pressure
and its normal derivative is known on a closed surface, the acoustic field can be
calculated at any point outside, including amplitude and phase. This functionality is
included in the Exterior Field Calculation feature. The feature has two options for the
evaluation, one full integral (the default) and one that only looks in the extreme far
field. See the section Theory for the Exterior Field Calculation: The
Helmholtz-Kirchhoff Integral for further details.

FULL INTEGRAL
To evaluate the full Helmholtz-Kirchhoff integral, use the default Full integral option
in the Type of integral options. The full Helmholtz-Kirchhoff integral gives the pressure
(including phase) at any point at a finite distance from the source surface. This is
necessary in many acoustic applications, for example, when analyzing the sensitivity of

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 231

 transducers. Note that numerical integration tends to lose accuracy at very large
distances. See Exterior Field Calculation.

THE FAR-FIELD LIMIT

In many scattering applications, the quantity of interest is the far-field radiation
pattern, which can be defined as the limit of r|p| when r goes to infinity in a given
direction. To evaluate the pressure in the far-field limit set the Type of integral to
Far-field integral approximation for r → ∞ in the settings for the exterior field variables.
With this option, only the direction of the evaluation point r is of interest not the
distance |r|. See Exterior Field Calculation.

EXTERIOR FIELD OPERATOR

The exterior field pressure is evaluated using the exterior field operator. The name is
defined in the Exterior field variable name input field and the default is pext.

The exterior field pressure can be evaluated in a given point (x0,y0,z0) by simply
writing pext(x0,y0,z0). To evaluate the sound pressure level in the same point, it is
advantageous to use the subst() operator and, for example, write
subst(acpr.efc1.Lp_pext,x,x0,y,y0,z,z0).

For an example, see Loudspeaker Driver: Application Library path

Acoustics_Module/Electroacoustic_Transducers/loudspeaker_driver.

These evaluations are automatically carried out in the dedicated postprocessing

features presented in the section below.

Dedicated Acoustics Plots for Postprocessing

Dedicated acoustics postprocessing functionality exists to plot responses, transfer
functions, directivity patterns and more. They require the Acoustics Module license.

In this section:

• Radiation Pattern Plots (Spatial Response)

• Grid and Parameterized Datasets
• Octave Band Plots
• Directivity Plots

232 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

RADIATION PATTERN PLOTS (SPATIAL RESPONSE)
Evaluating and plotting the acoustic pressure around radiating objects and sources is
essential for the development of several acoustic devices. This is known as the spatial
response or the radiation pattern. Application areas range from underwater acoustic
transducers and loudspeakers, to determining the spatial sensitivity of microphone
systems (for example, using reciprocity). Using the Exterior Field Calculation feature
can help to reduce the computational domain while still being able to determine the
pressure and thus the spatial response in the exterior field (near-field to far-field). The
Radiation Pattern plots are specially designed for easy evaluation of the exterior field
variables, that is, the acoustic exterior pressure and the exterior field sound pressure
level.

In the radiation pattern plots, the exterior field variable (pressure or sound pressure
level) is represented in a polar plot for a selected number of angles. Data is retrieved
on an evaluation circle in 2D, 2D axial symmetry, or 3D. The angle interval and the
number of angles can be manually specified. The evaluation circle origin, orientation,
and radius can be specified as well as the reference direction. The evaluation circle can
be visualized using a Preview Evaluation Plane functionality. There is also a built-in
option to calculate the Beamwidth of the plotted data. The data is stored in a table that
can be plotted using the Table Graph option.

Figure 2-8: Example of a 2D and 3D radiation pattern plot from the Bessel Panel tutorial
model.

The 3D radiation pattern plots create a 3D polar plot (or bubble plot) where the data
is evaluated on a sphere. Here, you can specify a separate expression for the surface
color. In the 3D radiation pattern plots, the directivity can also be calculated by
enabling the Compute directivity option. The result is given in a table and is calculated

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 233

 as the maximum power relative to the average power. The table comprises the maximal
radiation direction (polar angles), the directivity, and the directivity given in dB (also
known as the directivity index).

The main advantage with the radiation pattern plot, as compared to making a line
graph, is that the circle/sphere that you use for defining the plot directions, is not part
of your geometry for the solution. Thus, the number of plotting directions is
decoupled from the discretization of the solution domain.

Variables defined by the Exterior Field Calculation feature like, for example, the pressure
[Link] or sound pressure level acpr.efc1.Lp_pext are globally defined
variables and can be directly evaluated in the Radiation Pattern plot and Directivity
plot. The values plotted only make physical sense if they are evaluated outside of the
Exterior Field Calculation boundary.

To visualize, for example, the sound pressure level inside of the Exterior Field
Calculation boundary plot the expression: at2(x,y,[Link]) in 2D,
at2(r,z,[Link]) in 2D axisymmetry, or at3(x,y,z,[Link]) in 3D. The at2()
and at3() operators ensure the variables are globally defined.

Default exterior field plots (that use the radiation pattern plots) are
automatically added to any model that uses exterior field calculation
feature, when Generate default plots is selected (the default) on the study
node.

In a 3D Radiation Pattern plot, the definition of the polar angle θ and

the azimuthal angle ϕ follows the spherical coordinate system ( r, θ, ϕ )
typically used in physics.

To help visualize the angles, assuming the center is at the origin, simply
plot the following expressions:

• Polar angle θ: acos(z/sqrt(x^2+y^2+z^2))

• Azimuthal angle ϕ : atan2(y,x)

For a 3D example, see Bessel Panel: Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/bessel_panel.

234 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 • Evaluating the Acoustic Field in the Exterior: Near- and Far-Field
• Radiation Pattern and Results Analysis and Plots in the COMSOL
Multiphysics Reference Manual

GRID AND PARAMETERIZED DATASETS

Another way of evaluating and depicting the exterior field is by using either the Grid
2D , the Grid 3D , the Parameterized Curve 2D and Parameterized Curve 3D, or the
Parameterized Surface datasets. All of these datasets allow the evaluation of global
quantities, like the exterior field variables, outside of the computational domain
(outside of the mesh).

The exterior field is depicted in several ways in the model Piezoelectric

Tonpilz Transducer: Application Library path Acoustics_Module/
Piezoelectric_Devices/tonpilz_transducer.

OCTAVE BAND PLOTS

A dedicated Octave Band ( ) plot exists to plot frequency response, transfer
functions, transmission loss, and insertion loss curves. The plot has several built-in
acoustics specific features like predefined weighting (Z, A, C, and User defined) as well

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 235

 as the possibility to plot the response in octaves, 1/3 octaves, 1/6 octaves, or as a
continuous response.

Figure 2-9: Sensitivity curve plotted as 1/3 octave bands and a continuous curve. From
the Loudspeaker Driver model in the Application Library.

For details see Octave Band in the Results Analysis and Plots chapter in
the COMSOL Multiphysics Reference Manual.

The predefined Z, A, and C weightings are defined in IEC 61672-1. See

IEC 61672-1 Electroacoustics - Sound level meters - Part 1:
Specifications for details.

For examples that use an Octave Band plot, see:

• Absorptive Muffler. The Application Library path: Acoustics_Module/

Automotive/absorptive_muffler
• The Brüel & Kjær 4134 Condenser Microphone. The Application Library
path: Acoustics_Module/Electroacoustic_Transducers/bk_4134_microphone.

DIRECTIVITY PLOTS
Another acoustics specific plot, especially used for loudspeakers, is the dedicated
Directivity ( ) plot. The plot allows audio engineers to depict the spatial response of

236 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

a loudspeaker as function of both frequency and spatial angle in a contour-like plot.
Representing the spatial response in this manner is a very common in the loudspeaker
industry. Measured data is often also represented in the same manner. The plot
includes many options to format the plot to achieve maximal insight into the modeled
data, for example, linear and logarithmic frequency axis scaling options, easy switch of
the x- and y-axis (frequency and polar angle axis), as well as options to normalize the
data with respect to a given angle or the maximal value.

Figure 2-10: Example of the directivity plot from the Lumped Loudspeaker driver model
in the Application Library.

For details see Directivity in the Results Analysis and Plots chapter in the
COMSOL Multiphysics Reference Manual.

Lumped Loudspeaker Driver. Application Library path:

Acoustics_Module/Electroacoustic_Transducer/lumped_loudspeaker_driver

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (FEM-BASED INTERFACES) | 237

 About the Material Databases for the Acoustics Module
The Acoustics Module includes two material databases: Liquids and Gases, with
temperature-dependent fluid dynamic and thermal properties, and a Piezoelectric
Materials database with over 20 common piezoelectric materials.

For detailed information about Materials, the Liquids and Gases Materials
Library, and the Piezoelectric Materials Library see the COMSOL
Multiphysics Reference Manual.

For a discussion about the air and water materials as commonly used in
acoustics, as well as attenuation models, and other material properties see
the Acoustic Properties of Fluids chapter of this manual.

Specifying Frequencies: Logarithmic and ISO Preferred

When solving a model in the frequency domain, you need to specify the Frequencies to
solve for in the Frequency Domain study step. Several built-in options exist when
selecting the Range ( ). Specifically for acoustics modeling the Entry method has the
Logarithmic and the ISO preferred frequencies. The latter option is only available with
the Acoustics Module and is activated by clicking the Show More Options button ( )
and selecting Advanced Physics Options in the Show More Options dialog box. For ISO
preferred frequencies, select a start and stop frequency and the interval (Octave, 1/3
Octave, 1/6 Octave, 1/12 Octave, and 1/24 Octave). The 1/3 octave preferred
frequencies are defined based on ISO 266. They are extended to frequencies based on
the preferred numbers of ISO 3 (series R20, R40, and R80).

238 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Modeling with the Pressure Acoustics
B r a nc h (BE M - Based In t erfac e)
This section contains good practices and modeling tips and tricks for the boundary
element method (BEM) based interface of the Pressure Acoustics Branch: The
Pressure Acoustics, Boundary Elements Interface.

In this section:

• When to Use BEM

• Selections: Infinite Void and Finite Voids
• Solvers for BEM Models
• Meshing BEM Models
• Postprocessing BEM Results

When to Use BEM

Solving acoustics problems using the boundary element method (BEM) can be very
advantageous compared to the finite element method (FEM) for some types of
problems. For others problems, FEM has its advantages. The recommended approach
is to use the BEM and FEM methods where they apply best and also to combine them.
Since the BEM interface is fully enabled with multiphysics capabilities, it can be
coupled seamlessly with the physics interfaces that are based on FEM. This
functionality includes coupling to vibrating structures with the Acoustic-Structure
Boundary multiphysics coupling and to other acoustic domains with the Acoustic
FEM-BEM Boundary multiphysics coupling. This approach enables modeling in a
FEM-BEM framework, using the strength of each formulation adequately. The
BEM-based interface is especially well suited for radiation and scattering problems.

The advantage of the boundary element method is that only boundaries need to be
meshed and the DOFs solved for are restricted to the boundaries. This introduces
some clear ease-of-use for handling complex geometries. However, the BEM
technique results in fully populated or dense matrices that need dedicated numerical
methods. Assembling and solving these matrices can be very demanding. This means
that when solving acoustic models of small and medium size, The Pressure Acoustics,
Frequency Domain Interface will often be faster than solving the same problem with
the BEM interface.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (BEM-BASED INTERFACE) | 239

 The challenge with using the FEM interface is to set up open boundaries in an efficient
way, for example, by using PMLs. When PMLs cannot easily be placed snugly around
radiating structures or two structures are far apart, large air domains need to be
meshed. This costs a lot on the computational side, as the frequency is increased. These
problems are well suited for the BEM interface, as only the boundary of the radiating
structure needs to be meshed.

Interior problems can be solved with the BEM interface, but typically it is much more
efficient to solve these problems using the FEM-based interfaces. Remember also that
an iterative solver will struggle to solve interior problems, with no or little damping,
when resonances appear. If you are close to an eigenfrequency with a high Q, the
solution will also be very mesh sensitive.

Note also that at very low frequencies, the BEM method may “break down” for
interior problems and give inaccurate results. This is in practice not an issue, as this
kind of problem should be solved with The Pressure Acoustics, Frequency Domain
Interface.

Selections: Infinite Void and Finite Voids

For modeling with BEM, a geometry model can have multiple domains and multiple
finite voids. However, there can be only one infinite void.

• A domain corresponds to a finite region where the geometry model is of the type
solid. The geometry in most models based on FEM consists of domains only.
• Any other finite region is a finite void. A finite void can be created as a geometric
primitive of type curve (in 2D) or surface (in 3D), for example, or by a Boolean
difference operation, subtracting a closed domain from the model geometry.
• The infinite void is the single unbounded region surrounding the geometry model.
This is the typical region selected when modeling scattering or radiation problems
with the Pressure Acoustics, Boundary Elements interface.
• If the model contains finite voids and an infinite void, those volumetric regions
appear in the domain selection lists for the BEM interface and in the Selection List
window as Finite void 1, and so on, and as Infinite void.
• It is important to also assign the materials under the Materials node to the
appropriate selections of the Pressure Acoustics model. Note that material
parameters need to be constant for each applied Pressure Acoustics model.
• Assigning the Pressure Acoustics model to a selection adds the necessary equations
to the exterior boundaries of the selections. Only these boundaries need to be

240 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 meshed. The condition at infinity (of the infinite void) is selected in the Condition
at Infinity section. The typical and default is an outgoing wave condition.

Solvers for BEM Models

The default for The Pressure Acoustics, Boundary Mode Interface is to use an iterative
solver for 3D models and a direct dense matrix solver for 2D problems. If the Far-Field
Approximation is turned off, the direct dense matrix solver will be used in all situations.
The dense matrix solver is adequate for small BEM problems but quickly runs out of
memory for increasing number of DOFs. It is recommended to use the default
settings.

In 3D two iterative solver suggestions are generated. The default iterative suggestion
uses Sparse Approximate Inverse (SAI) as preconditioner. This approach is memory
efficient but for increasing frequencies the convergence is compromised. In this case
the Stabilized formulation can be used (see below). As an alternative to using the
stabilized formulation the second iterative suggestion can be used (enable the solver
suggestion). This solver uses a Hierarchical LU (HLU) preconditioner which produces
LU factors of high quality. This will ensure better convergence for the iterative
GMRES solver. The memory consumption is higher for this approach.

When the Stabilized formulation is used, a different solver suggestion is generated.

Make sure to reset the solver to default if changes are made. The solver uses PARDISO
as a direct preconditioner (with hybridization splitting the BEM and the stabilized
variables) and GMRES as the iterative method.

The boundary element method (BEM) used in the Pressure Acoustics, Boundary
Elements interface is based on the direct method with Costabel’s symmetric coupling.
The adaptive cross approximation (ACA) fast summation method is used for solving
the resulting linear system. This method uses partial assembly of the matrices where
the effect of the matrix vector multiplication is computed. The default iterative solver
used is GMRES.

For multiphysics problems that combine BEM and FEM interfaces and are set up with
one of the multiphysics couplings; the default approach is to use hybridization with
ACA for BEM and an appropriate preconditioner for the FEM part of the problem

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (BEM-BASED INTERFACE) | 241

 (direct or multigrid). Several solver suggestions are given for the multiphysics
FEM-BEM problems.

Further details about the BEM formulation are found in The PDE,
Boundary Elements Interface documentation in the COMSOL
Multiphysics Reference Manual.

Meshing BEM Models

When solving an acoustics problem with the boundary element method, the same
restrictions exist on the mesh as for the finite-element-based interfaces. In general, the
mesh should resolve the wavelength with 5 to 6 elements. The mesh should also
resolve small details and curved surfaces adequately. For exterior problems, the mesh
can in practice be slightly coarser and it typically suffices to use 4 elements per
wavelength. Using a structured quad surface mesh will result in fewer degrees of
freedom and should be used if possible.

Postprocessing BEM Results

When solving a problem with The Pressure Acoustics, Boundary Elements Interface,
the resulting solution will consist of the dependent variables on boundaries. Evaluating
the solution in a domain is based on a kernel evaluation.

On boundaries, dependent variables exist for the pressure p, up and down

pressure-dependent variables on interior boundaries (pabe.p_up and pabe.p_down),
as well as a normal flux variable [Link].

The Pressure Acoustics, Boundary Elements Interface sets up predefined

postprocessing variables that combine the properties of the boundary variables, when
needed, with variables based on the kernel evaluation.

The BEM solution can be evaluated at a given coordinate using the

at3_spatial(...) operator. In a 3D model you can evaluate the sound
pressure level in the point (x,y,z) = (1 m,0,0) by typing:
at3_spatial(1[m],0,0,[Link],’minc’).

This can, for example, be used to normalize the data in a Radiation Pattern
plot by typing:
[Link]-at3_spatial(1[m],0,0,[Link],’minc’)

242 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Domain Evaluation
The BEM solution can in general be evaluated in domains using the Grid 2D and
Grid 3D datasets. This sets up an underlying grid for the kernel evaluation of the
solution. The grid dataset can, for example, be selected as the input Dataset for a slice
plot. A grid dataset and a multislice plot are generated and used in the default plots
when a BEM model is solved. Note also that the kernel-based variables can be
evaluated on domains that are meshed.

All the Dedicated Acoustics Plots for Postprocessing that exist for the FEM interfaces
can be used with the variables defined by the BEM solution. This is, for example, the
Radiation Pattern plot or the Directivity plot used for analyzing the spatial radiation
pattern of a transducer or a scattered field.

Boundary Evaluation
When evaluating the solution on boundaries, the plot lists are populated with special
boundary variables in a Boundary variables submenu located under, for example, the
Pressure and sound pressure level or Acceleration and velocity menu. The boundary
variables have a “_bnd” added to their names. For example, the total acoustic pressure
on a boundary is pabe.p_t_bnd. On interior boundaries, there are also up and down
variants of the variables, with “_up” and “_down” added, respectively. On interior
boundaries (where an interior boundary condition has been added), the boundary
variable is the average of the up and down variables.

The boundary variables should be used when plotting on surfaces, as they do not rely
on the kernel evaluation but are directly related to the DOFs solved for in the BEM
problem.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (BEM-BASED INTERFACE) | 243

 Modeling with the Pressure Acoustics
Branch (DG-FEM-Based Interface)
This section contains good practices and modeling tips and tricks for the discontinuous
Galerkin finite element method (dG-FEM) based interface of the Pressure Acoustics
Branch: The Pressure Acoustics, Time Explicit Interface. In this section:

• Meshing, Discretization, and Solvers

• Postprocessing: Variables and Quality
• Absorbing Layers
• Storing Solution on Selections for Large Models
• Assemblies and Pair Conditions

Meshing, Discretization, and Solvers

The same considerations for the mesh, discretization, and solvers apply for the Pressure
Acoustics, Time Explicit interface as for The Convected Wave Equation, Time Explicit
Interface. Details are found in Meshing, Discretization, and Solvers under the
modeling section for that physics. The interface is supported on all mesh types and for
all combinations of mesh.

Postprocessing: Variables and Quality

The same considerations for postprocessing apply for the Pressure Acoustics, Time
Explicit interface as for The Convected Wave Equation, Time Explicit Interface.
Details are found in Postprocessing: Variables and Quality under the modeling section
for that physics.

Absorbing Layers
The same considerations for the absorbing layers and their setup apply for the Pressure
Acoustics, Time Explicit interface as for The Convected Wave Equation, Time Explicit
Interface. Details are found in Absorbing Layers under the modeling section for that
physics.

244 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Storing Solution on Selections for Large Models
The same considerations for storing data apply for the Pressure Acoustics, Time
Explicit interface as for The Convected Wave Equation, Time Explicit Interface.
Details are found in Storing Solution on Selections for Large Models under the
modeling section for that physics.

Assemblies and Pair Conditions

The same considerations for using assemblies and pair condition apply for the Pressure
Acoustics, Time Explicit interface as for The Convected Wave Equation, Time Explicit
Interface. Details are found in Assemblies and Pair Conditions under the modeling
section for that physics.

MODELING WITH THE PRESSURE ACOUSTICS BRANCH (DG-FEM-BASED INTERFACE) | 245

 Modeling with the Aeroacoustic Flow
Source: Flow-Induced Noise
This section shortly describes the theory behind the Lighthill acoustic analogy used in
the Aeroacoustic Flow Source. Then the hybrid computational aeroacoustic (CAA)
method used is introduced. Finally, best practices for setting up an aeroacoustic model
in COMSOL using the Aeroacoustic Flow Source, the Aeroacoustic Flow Source
Coupling multiphysics, and the necessary study steps such as Transient Mapping and
Time to Frequency FFT are given.

In this section:

• The Lighthill Analogy

• Hybrid Computational Aeroacoustics (CAA)
• Setting up a Flow-Induced Noise Problem in COMSOL

The Lighthill Analogy

In Lighthill’s seminal work (Ref. 63 and Ref. 64) he transformed (rearranged) the full
set of Navier-Stokes equations (momentum and mass conservation) into an
inhomogeneous wave equation. The source term of the wave equation only being
important within turbulent flow regions and thus representing the flow-induced noise
source. The equation is given by

2
2
∂ p′-
1- ---------- ∂ L ij 2
---- – ∇ ⋅ ∇ ( p′ ) = ---------------- p′ = c 0 ( ρ – ρ 0 )
2
c 0 ∂t
2 ∂x i ∂x j (2-8)
2
L ij = ρv i v j + ( ( p – p 0 ) – c0 ( ρ – ρ 0 ) )δ ij – τ ij

where Lij is called the Lighthill stress tensor, c is the speed of sound, p is the pressure
(total), ρ is the density (total), τij is the viscous stress tensor, and the subscript 0 refers
to the reference properties of the fluid at rest. It is assumed that p’ represents the linear
acoustic pressure variations outside of the source domain. The Lighthill stress tensor
consists of three contributing parts (see Ref. 65): the Reynolds stress (first term), a
term containing the excess pressure generated by nonlinear processes and
compressibility effects in the moving fluid (second term), and the viscous stress tensor
that accounts for attenuation of the sound (third term).

246 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Lighthill’s wave equation is often simplified, specifically by simplifying the source term.
It can be shown that (see Ref. 65):

2
ρv i v j = ρ 0 ( 1 + O ( Ma ) ) ( v i v j ) ≈ ρ 0 v i v j
2 2 2
( p – p 0 ) – c 0 ( ρ – ρ 0 ) ∼ O ( ρ 0 v Ma )

where Ma is the Mach number. This implies that for low Mach numbers the Lighthill
stress tensor is well approximated by

L ij = ρ 0 v i v j – τ ij (2-9)

This is the expression for the source term stress tensor used when the Acoustic analogy
option is set to Lighthill. The viscous dissipation can optionally be included by selecting
Include viscous stress. This is a symmetric tensor that contains 9 elements in 3D of
which 6 are unique.

The Lighthill wave equation can be further approximated at low Mach numbers for an
incompressible flow (see Ref. 65, 67, and 68). This leads to a simplified
inhomogeneous wave equation known as the aeroacoustic wave equation or AWE,
given by

2 2 ic
1- ∂--------p- 1∂ p-
---- – ∇ ⋅ ∇ ( p ) = ----2- ------------ (2-10)
2 2 2
c 0 ∂t c 0 ∂t

where p is the linear acoustic pressure fluctuation and pic is the pressure variation
associated with the incompressible flow simulation. This equation is solved (in the
frequency domain) when the he Acoustic analogy option is set to Aeroacoustic wave
equation. In the Aeroacoustic Flow Source the right hand side is included as the
Fourier components of the second time derivative of the incompressible flow variation,
var
p tt in the Equation section.

Hybrid Computational Aeroacoustics (CAA)

The computational aeroacoustic (CAA) method used here is of the so-called hybrid
type. It is based on a one way coupling between the turbulent flow sources and the
acoustic problem. It is thus assumed that no important back-coupling exists from the
acoustic field to the flow. The computational method is based on the finite element
method (FEM) discretization of Lighthill’s acoustic analogy (wave equation). This
formulation of the equations ensures that any solid (fixed or vibrating) boundaries are

MODELING WITH THE AEROACOUSTIC FLOW SOURCE: FLOW-INDUCED NOISE | 247

 implicitly taken into account. The method is also applicable for interior problems (see
Ref. 65).

The hybrid method is based on a mapping and time to frequency FFT approach, in
order to solve the acoustic problem in the frequency domain and on a mesh
appropriate for the acoustic problem. The mesh used to solve the LES flow problem
has characteristics necessary to resolve the turbulence as well as the flow boundary
layers. The mesh also corresponds to linear discretization for most LES methods. The
acoustic mesh, on the other hand, has to resolve the wavelength and geometry, and it
is typically based on the default second order spatial discretization (see Ref. 65 and
Ref. 66).

Setting up a Flow-Induced Noise Problem in COMSOL

When setting up and solving an aeroacoustic flow-induced noise problem in
COMSOL Multiphysics, using the built-in functionality of the Acoustics Module, four
main steps are necessary:

1 Solve the flow problem using an LES model to resolve the structure of the
turbulence. This problem is solved on an appropriate mesh for the flow with
appropriate time stepping settings. This typically consists of several steps, for
example, first solving a stationary RANS model to get a good initial condition for
the transient LES model. Then an initial LES model has to be solved to get a fully
developed turbulent flow. In this step the solution only needs to be stored at the end
time. Finally, the LES model is run with the fully developed turbulent flow as initial
condition. Here, the solution is stored with an appropriate time resolution defined
in the Output times.
A time resolution Δt (for the output times) will lead to a maximum frequency
resolution of fmax = 1/Δt/2 in the frequency domain acoustics problem. Moreover,
the length of the simulation (Tend) will lead to a frequency resolution of
Δf = 1/Tend. Create parameters for these values.
2 Map the solution from the CFD mesh to the acoustics mesh. This is done by using
the Transient Mapping study step. The input to this study step is only the Time step
Δt used in the LES study. In the Transient Mapping study step it is only the
Aeroacoustic Flow Source Coupling that is solved for.
3 Transform the mapped flow sources to the frequency domain using the Time to
Frequency FFT study step. In the Study Settings for the study step, the End time is set
to Tend and Maximum output frequency is set to fmax. In the Time to Frequency FFT
study step it is also only the Aeroacoustic Flow Source Coupling that is solved for.

248 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

4 Finally, solve the acoustic problem in the frequency domain. In the Frequency
Domain study step it is only the Pressure Acoustics, Frequency Domain physics that is
solved for. The Values of variables not solved for settings should point to the Time to
Frequency FFT study.

Tandem Cylinder Flow Noise tutorial available in the COMSOL

Application Gallery at:

• [Link]

For further details about LES see the CFD Module User’s Guide:

• The Creeping Flow, Laminar Flow, Turbulent Flow, and Large Eddy
Simulation Interfaces
• Theory for the Large Eddy Simulation Interfaces

MODELING WITH THE AEROACOUSTIC FLOW SOURCE: FLOW-INDUCED NOISE | 249

 Theory Background for the Pressure
Acoustics Branch
This section describes the governing equations and the mathematical formulation of
the governing equations as used in the Pressure Acoustics branch of the Acoustics
Module. Details are also given regarding some of the boundary conditions, including
the radiation boundary conditions as well as the exterior field calculation feature. The
section starts with a general introduction to the governing equations used in pressure
acoustics.

In this sections:

• The Governing Equations

• Pressure Acoustics, Frequency Domain Equations
• Pressure Acoustics, Transient Equations
• The Nonlinear Westervelt Equation
• Pressure Acoustics, Boundary Mode Equations
• Theory for the Plane, Spherical, and Cylindrical Radiation Boundary Conditions
• Theory for the Exterior Field Calculation: The Helmholtz-Kirchhoff Integral

The Governing Equations

Pressure acoustic problems involve solving for the small acoustic pressure variations p
(also named p’ or p1) on top of the stationary background pressure p0.
Mathematically, this represents a linearization (small parameter expansion) of the
dependent variables around the stationary quiescent values.

The governing equations for a compressible lossless (no thermal conduction and no
viscosity) fluid flow problem are the mass conservation equation (continuity equation),
the momentum conservation equation (Euler's equation), and the energy equation
(entropy equation). These are given by:

250 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 ∂ρ
------ + ∇ ⋅ ( ρu ) = M
∂t
∂u 1
------ + ( u ⋅ ∇ )u = – --- ∇p + F (2-11)
∂t ρ
∂s
----- + ∇ ⋅ ( su ) = 0
∂t

where ρ is the total density, p is the total pressure, u is the velocity field, s is the
entropy, M and F represent possible source terms. In classical pressure acoustics, all
thermodynamic processes are assumed to be reversible and adiabatic, that is, isentropic
processes. The small parameter expansion is performed on a stationary fluid (u0 = 0)
of density ρ0 (SI unit: kg/m3) and at pressure p0 (SI unit: Pa) such that:

p = p0 + p1
p1 « p0
ρ = ρ0 + ρ1
with ρ1 « ρ0
u = 0 + u1
u1 « c
s = s0 + s1

where the variables with subscript 1 represent the small acoustic variations (the 1st
order expansion). Assuming the initial entropy s0 to be zero, then it will remain and
s0 = 0. Inserting these values into the governing equations and only retaining terms
linear in the acoustic variables yields

∂ρ
--------1- + ∇ ⋅ ( ρ 0 u ) = M
∂t 1

∂u 1
--------1- = – ------ ∇p 1 + F (2-12)
∂t ρ0
∂p 2 ∂ρ
--------1- = c s  --------1- + u 1 ⋅ ∇ρ 0
∂t  ∂t 

where cs is recognized as the (isentropic) speed of sound (SI unit: m/s) at constant
entropy s. The last equation for the pressure time differential is derived from the
entropy equation, for more details see Theory Background for the Thermoviscous
Acoustics Branch. For constant material parameters the last equation reduces to the
usual relation

2
p1 = cs ρ1

This expression gives a useful condition that needs to be fulfilled for the linear acoustic
equations to hold:

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 251

 2
p1 « ρ0 cs

The subscript “s” is dropped in the following along with the subscript “0” on the
background density ρ0. Finally, rearranging Equation 2-12, renaming the source
terms, and dropping the subscript “1” yields the wave equation for pressure waves in
a lossless medium

2
1- ∂ p
+ ∇ ⋅  – --- ( ∇p – q d ) = Q m .
1
------- (2-13)
2 2  ρ 
ρc ∂ t

Here ρ (SI unit: kg/m3) refers to the density, and c (SI unit: m/s) denotes the speed
of sound. In this general formulation of the scalar wave equation, the speed of sound
and density may in general be space dependent, for example, through their dependency
on the background temperature. The equation includes two optional source terms:

• The monopole domain source Qm (SI unit: 1/s2). This source corresponds to a
mass source on the right-hand side of the continuity equation.
• The dipole domain source qd (SI unit: N/m3). This source corresponds to a
domain force source on the right-hand side of the momentum equation.

The combination ρc2 is called the adiabatic bulk modulus, commonly denoted Ks
(SI unit: Pa). The bulk modulus is equal to the one over the adiabatic compressibility
coefficient βs = 1/Ks = 1/ρc2(SI unit: 1/Pa).

In Equation 2-13 both the speed of sound c = c(x) and the density ρ = ρ(x) may be
dependent on the spatial coordinates x while they are independent of time, or only
slowly varying in time (that is, at a time scale much slower than the variations in the
acoustic signal). If both domain sources are set to zero and the density is constant in
space, you can recover the standard wave equation

2
∂--------p- 2 2
2
= c ∇ p
∂t

Some classical references on acoustics for further reading are found in

Ref. 4, Ref. 5, Ref. 6, Ref. 7, and Ref. 8.

An important special case is a time-harmonic wave, for which the pressure varies with
time as

252 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 iω t
p ( x, t ) = p ( x ) e

where ω = 2πf (rad/s) is the angular frequency and f (SI unit: Hz) is the frequency.
Assuming the same harmonic time dependence for the source terms, the wave
equation for acoustic waves reduces to an inhomogeneous Helmholtz equation

2
ω p
∇ ⋅  – ----- ( ∇p – q d ) – ------------2- = Q m .
1
(2-14)
 ρc 
ρc c c

In this equation, the subscript “c” on the density and the speed of sound refers to that
they may be complex valued. Lossy media, like porous materials or highly viscous
fluids, can be modeled by using the complex-valued speed of sound and density. A
selection of such fluid models is available in The Pressure Acoustics, Frequency
Domain Interface. The attenuation in these fluid models is frequency dependent in
different ways, depending on the physical origin of the damping. A description of the
different fluid models is given in Theory for the Equivalent Fluid Models.

In the time domain, only certain frequency dependencies can be modeled, which limits
the number of fluid models that can be used in The Pressure Acoustics, Transient
Interface. One way to model damping in the time domain is to introduce an additional
term of first order in the time derivative to account for attenuation of the sound waves

2
1- ∂ p ∂p
– d a + ∇ ⋅  – --- ( ∇p – q d ) = Q m
------- 1
 ρ  (2-15)
ρ c ∂t
2 2 ∂ t

The damping term in Equation 2-15 is absent from the standard PDE formulations in
the Pressure Acoustics, Transient interface, but it corresponds to a monopole domain
source proportional to the time derivative of the pressure. This approach is, however,
not used in the viscous and thermally conducting fluid models that contain damping
for transient acoustics (see Viscous Fluid Model, Thermally Conducting Fluid Model,
and Thermally Conducting and Viscous Fluid Model). The physical origin of the
damping corresponds here to a dipole-like source.

Even when sound waves propagate in a lossless medium, attenuation can

occur by interaction with the surroundings at the system boundaries. In
particular, this applies to the impedance boundary conditions.

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 253

 Alternatively, treat the Helmholtz Equation 2-14 as an eigenvalue PDE to solve for
eigenmodes and eigenfrequencies, see the Eigenfrequency and Mode Analysis in 2D
and 1D axisymmetric sections below.

In order to solve the governing equations, boundary conditions are necessary. Typical
boundary conditions used in acoustics are:

• Sound Hard Boundary (Wall)

• Sound Soft Boundary (zero acoustic pressure)
• Specified acoustic Pressure
• Specified Normal Acceleration
• Impedance boundary conditions (see also Theory for the Boundary Impedance
Models)
• Plane Wave Radiation
• Spherical Wave Radiation
• Cylindrical Wave Radiation

Pressure Acoustics, Frequency Domain Equations

The Pressure Acoustics, Frequency Domain Interface exists for several types of studies.
Here the equations are presented for the frequency domain, eigenfrequency, and mode
analysis studies. All the interfaces solve for the acoustic pressure p. It is available in all
space dimensions — for 3D, 2D, and 1D Cartesian geometries as well as for 2D and
1D axisymmetric geometries.

FREQUENCY DOMAIN
The frequency domain, or time-harmonic, formulation uses the inhomogeneous
Helmholtz equation:

2
1 k eq p t
∇ ⋅  – ----- ( ∇p t – q d ) – ------------- = Q m (2-16)
 ρc  ρc

This is Equation 2-14 repeated with the introduction of the wave number keq used in
the equations. It contains both the ordinary wave number k as well as out-of-plane and
azimuthal (or circumferential) contributions, when applicable. Note also that the
pressure is here the total pressure pt which is the sum of a possible Background
Pressure Field pb and the scattered field ps. This enables for a so-called scattered field
formulation of the equations. If no background field is present pt = ps = p.

254 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

In this equation, p = p (x,ω) = p(x)eiωt (the dependence on ω is henceforth not
explicitly indicated). Compute the frequency response by doing a parametric sweep
over a frequency range using harmonic loads and sources.

When there is damping, ρc and cc are complex-valued quantities. The available

damping models and how to apply them is described in the sections Pressure Acoustics
and Theory for the Equivalent Fluid Models.

Equation 2-16 is the equation that the software solves for 3D geometries. In
lower-dimensional and axisymmetric cases, restrictions on the coordinate dependence
mean that the equations differ from case to case. Here is a brief summary of the
situation.

2D
In 2D, the pressure is of the form

– i kz z
p ( x ) = p ( x, y )e

which inserted in Equation 2-16 gives

2
1 k eq
∇ ⋅  – ----- ( ∇p t – q d ) – -------- p t = Q m
ρc ρc
(2-17)
2 ω 2- 2
k eq = ------ – kz
c c2

The out-of-plane wave number kz can be set on the Pressure Acoustics page. By
default its value is 0. In the mode analysis study −ikz is used as the eigenvalue λ.

2D Axisymmetry
For 2D axisymmetric geometries the independent variables are the radial coordinate r
and the axial coordinate z. The only dependence allowed on the azimuthal coordinate
ϕ is through a phase factor,

– imϕ
p ( r, ϕ, z ) = p ( r, z )e (2-18)

where m denotes the azimuthal mode number. The mode number defines an
azimuthal wave number km = m/r. Because the azimuthal coordinate is periodic, m
must be an integer. Just like kz in the 2D case, m can be set in the Settings window for
Pressure Acoustics.

As a result of Equation 2-18, the equation to solve for the acoustic pressure in 2D
axisymmetric geometries becomes

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 255

 2
∂ r ∂p ∂ 1 ∂p k eq
– -----  – q r + r – -----  – q z – -------- rp = rQ m
∂ r ρc ∂ r ∂ z ρc ∂ z ρc
ω 2
k eq =  ------ – k m
2 2 m
k m = -----
 c c r

1D Axisymmetry
In 1D axisymmetric geometries,

– i ( kz z + mϕ )
p ( r, φ, z ) = p ( r )e

leading to the radial equation

2
∂ r ∂p k eq
– -----  – q r – -------- rp = rQ m
∂ r ρc  ∂ r ρc
ω 2
k eq =  ------ – k m – k z
2 2 2 m
k m = -----
cc r

where both the azimuthal wave number km and the axial wave number kz, appear
as parameters.

1D
The equation for the 1D case is obtained by letting the pressure depend on a single
Cartesian coordinate x:

2
d  1  dp k eq
– ----- – q d  – -------- p = Q m
d x  ρc  d x ρc
ω 2
k eq =  ------
2
cc

EIGENFREQUENCY
In the eigenfrequency formulation, the source terms are absent; the eigenmodes and
eigenfrequencies are solved for:

2
1 λ p
∇ ⋅  – ----- ∇p + ----------2- = 0 (2-19)
 ρc 
ρc cc

The eigenvalue λ introduced in this equation is related to the eigenfrequency f, and

the angular frequency ω, through λ = i2πf = iω. Because they are independent of the

256 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

pressure, the solver ignores any dipole and monopole sources unless a coupled
eigenvalue problem is being solved.

Equation 2-19 applies to the 3D case. The equations solved in eigenfrequency studies
in lower dimensions and for axisymmetric geometries are obtained from their
time-harmonic counterparts, given in the previous subsection, by the substitution
ω2 → −λ2.

Switch between specifying the eigenvalues, the eigenfrequencies, and the angular
frequencies by selecting from the Eigenvalue transformation list in the solver
sequence’s Settings window for Eigenvalue.

Vibrations of a Disk Backed by an Air-Filled Cylinder: Application

Library path Acoustics_Module/Verification_Examples/coupled_vibrations

MODE ANALYSIS IN 2D AND 1D AXISYMMETRIC

See Mode Analysis Study in the Pressure Acoustics, Boundary Mode Equations
section. The mode analysis study type is only available for the Pressure Acoustics,
Frequency Domain interface in 2D and 1D axisymmetric components, where the
solver solves for the eigenvalues λ = −ikz for a given frequency. Here kz is the
out-of-plane wave number of a given mode and the resulting pressure field p represents
the mode on the cross section of an infinite waveguide or duct.

Pressure Acoustics, Transient Equations

Use the Time Dependent study type to model transient acoustic phenomena in a
stationary fluid and to solve the wave equation

2
∂ p-
+ ∇ ⋅  – --- ( ∇p – q d ) = Q m
1 - -------- 1
--------
ρc 2 ∂t 2  ρ 

for the acoustic pressure, p = p(x, t). Here c is the speed of sound and ρ denotes the
equilibrium density, while qd and Qm are dipole and monopole sources, respectively.
The density and speed of sound can both be nonconstant in space. In contrast, they
are assumed to vary with time on scales much larger than the period for the acoustic
waves and are therefore considered time independent in the previous equation. This
physics interface is available for 3D, 2D, and 1D Cartesian geometries as well as for 2D
and 1D axisymmetric geometries.

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 257

 The Nonlinear Westervelt Equation
To extend the validity of the small perturbation wave equation to finite amplitude
waves, nonlinear effects need to be taken into account. The first idea is to expand the
density pressure relation to second order to include the effect of finite amplitude
waves. This gives

2
2 c0 B 2
p 1 = c 0 ρ 1 + ------ ------- ρ 1 + … (2-20)
ρ 0 2A

where we have kept the subscript “1” to highlight the acoustic perturbation quantities
and a subscript “0” for the quiescent quantities. See Ref. 6 and 42 for details. The
quantity B/A is the parameter of nonlinearity (see Ref. 51). This expression basically
means that the different parts of the acoustic pressure wave travel at different speeds.

Inserting the above into the governing Equation 2-11 (retaining the dissipative terms)
will lead to the full nonlinear second-order wave equation. This is the nonlinear
equivalent to Equation 2-12 (or actually the nonlinear equivalent to the full equations
of thermoviscous acoustics). Next assume that boundary layer effects can be
disregarded and cumulative nonlinear effects dominate local nonlinear effects, for
example, when the propagation distance is greater than the wavelength. This leads to
the well-known Westervelt equation (see Ref. 42)

2 2 2
1 - ∂----------
pt  δ ∂ ∇p  ∂ pt
β -----------
-------- - + ∇ ⋅ – 1
--- ( ∇p t – q d ) – --------- ------------t = ----------- - + Qm (2-21)
ρc ∂t
2 2
 ρ ρc ∂t 
2 2
ρ c ∂t
4 2

Here we have eliminated the density fluctuations and then removed the subscripts.
Hence, pt is the (total) acoustic pressure, ρ is the quiescent density, δ is the diffusivity
of sound, and β is the coefficient of nonlinearity (related to the parameter of
nonlinearity B/A).

The Westervelt equation supports the propagation of shocks. Shocks can be hard to
capture numerically and require the addition of stabilization. This can, for example, be
done by adding diffusivity near the shock location. One method (used in COMSOL)
is to add an effective diffusivity term that reads

– δκ  1 – ∂ ∇p
q–1

∂t 

The highest artificial damping is achieved where the acoustic pressure increases or
decreases the most rapidly. It reaches its maximal values where the pressure endures

258 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

discontinuities, that is, where shocks arise. Thus this technique provides a
shock-capturing stabilization. See Ref. 52 for more details. Specify the so-called
q-Laplacian exponent q and the so-called q-Laplacian factor κ to get the desirable
amount of artificial damping. Note that the damping must not be too high nor too
low. The particular values of q and κ depend on the material and the input signal
frequency. The two parameters that control the stabilization require manual tuning. A
suggested approach is to use a simple 1D model to tune the parameters based on fluid
material properties and frequency content.

Nonlinear Acoustics — Modeling of the 1D Westervelt Equation:

Application Library path Acoustics_Module/Nonlinear_Acoustics/
nonlinear_acoustics_westervelt_1d

Pressure Acoustics, Boundary Mode Equations

When an acoustic wave of a given angular frequency ω is fed into a waveguide or a duct,
only a finite set of shapes, or modes, for the transverse pressure field can propagate over
long distances inside the structure. The higher the frequency, the higher the number
of sustainable modes.

Take, as an example, a uniform straight duct whose axis is in the z direction. The
acoustic field in such a duct can be written as a sum of the form

N
– i k zj z
p( x) =  pj ( x, y )e
j=0

The constant kzj is the axial wave number of the jth propagating transverse mode,
pj(x, y). These transverse modes and their associated axial wave numbers are solutions
to an eigenvalue problem defined on the duct’s cross section. The mode analysis
capabilities in The Pressure Acoustics, Boundary Mode Interface makes it possible to
solve such eigenvalue problems. The physics interface is available for 3D Cartesian and
2D axisymmetric geometries and solves for the transverse eigenmodes for the acoustic
pressure p and the associated propagation constants kz. The Mode Analysis Study is
briefly discussed.

MODE ANALYSIS STUDY

The eigenvalue solver computes a specified number of solutions {pj, λj} to the
equation

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 259

 2
 ω 2 k n
∇ ⋅  – ----- ( ∇p – q d ) –  ----------2- – ----- p = Q m
1
(2-22)
ρc  ρ c c c ρ c

defined on a 2D boundary of the modeling domain (in 3D) or on the 2D domain

itself, with λ = −ikn as the eigenvalue. In this equation, p is the in-plane pressure, ρc is
the density, cc is the speed of sound, ω is the angular frequency, and kn is the
propagation constant in the direction normal to the surface, in this context also
referred to as the out-of-plane wave number.

The out-of-plane wave number is denoted kn, and is in the normal

direction to the two-dimensional surface on which Equation 2-22. As for
a mode analysis study in the frequency domain the propagation direction
does not necessarily have to be normal to the z-axis for 3D geometries.

Special postprocessing variables exist for the Pressure Acoustics,

Boundary Mode interface. They are described in Pressure Acoustics,
Boundary Mode Variables.

Notice that the above equation is identical to the time-harmonic equation for pressure
acoustics, except that kn is interpreted as an eigenvalue and not as a parameter.

For axisymmetric geometries, the relevant eigenvalue equation to solve for the radial
pressure modes and the eigenvalues λ is

r- dp
d  ---- ω 2 2 m 2 rp
+  ------ + λ –  ----- ----- = 0

d r ρc d r   cc   r  ρc

Here m, the azimuthal mode number, is an integer-valued parameter. The equation

is defined on the interval r1 < r < r2. The eigenvalue λ is defined in terms of the axial
wave number kz through the equation λ = −ikz

Theory for the Plane, Spherical, and Cylindrical Radiation Boundary

Conditions
Specify a Plane Wave Radiation, Spherical Wave Radiation, or Cylindrical Wave
Radiation boundary condition to allow an outgoing wave to leave the modeling
domain with minimal reflections. The condition can be adapted to the geometry of the
modeling domain. The plane wave type is suitable for both far-field boundaries and

260 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

ports (for plane waves only). For general radiation boundary conditions for
waveguides (supporting multi-modes) it is recommended to use the Port boundary
condition.

Radiation boundary conditions are available for all types of studies. For the
frequency-domain study, Givoli and Neta’s reformulation of the Higdon conditions
(Ref. 1) for plane waves has been implemented to the second order. For cylindrical and
spherical waves, COMSOL Multiphysics uses the corresponding 2nd-order
expressions from Bayliss, Gunzburger, and Turkel (Ref. 2). The Transient, Mode
analysis, and Eigenfrequency studies implement the same expansions to the first order.

The first-order radiation boundary conditions in the frequency domain read

p pi ∇p i
– n ⋅  – ----- ( ∇p t – q d ) + ( ik + κ ( r ) ) ----- = ( ik + κ ( r ) ) ----- + n ⋅  ---------
1
 ρc  ρc ρc  ρc 

where k is the wave number and κ(r) is a function whose form depends on the wave
type:

• Plane wave: κ(r) = 0

• Cylindrical wave: κ(r) = 1/(2r)
• Spherical wave: κ(r) = 1/r

In the cylindrical and spherical wave cases, r is the shortest distance from the point
r = (x, y, z) on the boundary to the source. The right-hand side of the equation
represents an optional incoming pressure field pi (see Incident Pressure Field).

The second-order radiation boundary conditions in the frequency domain are defined
below. In these equations, Δ || at a given point on the boundary denotes the Laplace
operator in the tangent plane at that particular point.

PLANE WAVE

k k
– n ⋅  – ----- ( ∇p t – q d ) + i ----- p + ------------ Δ || p = ------------ Δ || p i + i ----- p i + n ⋅ ----- ∇p i
1 i i 1
 ρc  ρc 2kρ c 2kρ c ρc ρc

In the notation of Givoli and Neta (Ref. 1), the above expressions correspond to the
parameter choices C0 = C1 = C2 = ω/k. For normally incident waves, this gives a
vanishing reflection coefficient.

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 261

 CYLINDRICAL WAVE

– n ⋅  – ----- ∇p t – q d =
1
 ρc 

 ik + ----- ( pi – p ) ( rΔ || p i – rΔ || p )
– - -----------------
1- --------------------------------
1 1-
- + n ⋅ ---- ∇p + -------------------------------------
 eq 2r 8r ( 1 + ik eq r ) ρ c ρ c i 2 ( 1 + ik eq r )ρ c

The cylindrical wave boundary condition is based on a series expansion of the

outgoing wave in cylindrical coordinates (Ref. 2), and it assumes that the field is
independent of the axial coordinate. Specify the axis of this coordinate system by giving
an orientation (nx, ny, nz) and a point (x0, y0, z0) on the axis. In axisymmetric
geometries, the symmetry axis is the natural and only choice.

SPHERICAL WAVE

1 p rΔ || p
– n ⋅  – ----- ( ∇p t – q d ) +  ik eq + --- ----- – ------------------------------------
1
ρc  
r ρ c 2ρ c ( ik eq r + 1 )
rΔ || p i 1 pi
= – --------------------------------------- +  ik eq + --- ----- + n ⋅ ----- ∇p i
1
2ρ 0c ( ik eq r + 1 )  r  ρc ρc

Use a spherical wave to allow a radiated or scattered wave — emanating from an object
centered at the point (x0, y0, z0) that is specified — to leave the modeling domain
without reflections. The boundary condition is based on an expansion in spherical
coordinates from Bayliss, Gunzburger, and Turkel (Ref. 2), implemented to the
second order.

TRANSIENT ANALYSIS
The transient radiation boundary condition is the first-order expression

1 1 ∂p 1 1 ∂p i
– n ⋅  – --- ( ∇p t – q d ) + ---  --- + κ ( r ) p = ---  --- + κ ( r ) p i + n ⋅ ∇p i
1
ρ ρ c ∂t ρ c ∂t

262 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

where κ(r) is the same wave-type dependent function as for the eigenfrequency case
and pi the optional Incident Pressure Field.

An estimate of the reflection coefficient Rs for spurious waves off the

plane wave radiation boundary, for incident plane waves at angle θ is given
by the expression:

cos θ – 1- N
R s = ---------------------
cos θ + 1

where N is the order of the boundary condition (here 1 or 2). So at

normal incidence (θ = 0) there are no spurious reflections, while, for
example, at an incidence angle of 30o for N = 2 (plane wave radiation in
the frequency domain) the amplitude of the spurious reflected wave is
0.5 % of the incident.

Theory for the Exterior Field Calculation: The Helmholtz-Kirchhoff

Integral
The Acoustics Module has functionality to evaluate the acoustic pressure field in the
exterior field region of the model (outside the computational domain). This section
gives the relevant definitions and mathematical background as well as some general
advice for analyzing the exterior field. Details about how to use the exterior field
functionality is described in Exterior Field Calculation and in the modeling section
Evaluating the Acoustic Field in the Exterior: Near- and Far-Field.

THE HELMHOLTZ-KIRCHHOFF INTEGRAL REPRESENTATION

In many cases, solving the acoustic Helmholtz equation everywhere in the domain
where results are requested is neither practical nor necessary. For homogeneous media,
the solution anywhere outside a closed surface containing all sources and scatterers can
be written as a boundary integral in terms of quantities evaluated on the surface. To
evaluate this Helmholtz-Kirchhoff integral, it is necessary to know both Dirichlet and
Neumann values on the surface. Applied to acoustics, this means that if the pressure
and its normal derivative (which is related to the normal velocity) is known on a closed
surface, the acoustic field can be calculated at any point outside.

In general, the solution p to Helmholtz’ equation

– ∇ ⋅ ∇p – k 2 p = 0

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 263

 in the homogeneous domain exterior to a closed surface, S, can be explicitly expressed
in terms of the values of p and its normal derivative on S:

p(R) =  ( G ( R, r ) ∇p ( r ) – ∇G ( R, r ) p ( r ) ) ⋅ n dS
S

Here the coordinate vector r parameterizes S. The unit vector n is the outward normal
to the exterior infinite domain; thus, n points into the domain that S encloses. The
function G(R, r) is a Green’s function satisfying

(3)
– ∇ ⋅ ∇G ( R, r ) – k 2 G ( R, r ) = δ (R – r)

This essentially means that the Green’s function, seen as a function of r, is an outgoing
traveling wave excited by a simple source at R. In 3D, the Green’s function is
therefore:

e – ik r – R
G ( R, r ) = -----------------------
4π r – R

In 2D, the Green’s function contains a Hankel function instead of the exponential:

i (2)
G ( R, r ) = --- H 0 ( k r – R )
4

Inserting the 3D Green’s function in the general representation formula gives:

1 e – ik r – R ( 1 + ik r – R )
p ( R ) = ------ ----------------------  ∇p ( r ) + p ( r ) -----------------------------------
 - ( r – R ) ⋅ n dS
 (2-23)
4π r–R r–R2
S

while in 2D, the Hankel function leads to a slightly different expression:

(2)
i  (2) H1 ( k r – R ) 
4  
p ( R ) = – ---  H 0 ( k r – R ) ∇p ( r ) + kp ( r ) ------------------------------------ ( r – R ) ⋅ n dS (2-24)
r–R 
S

For axially symmetric geometries, the full 3D integral must be evaluated. The
Acoustics Module uses an adaptive numerical quadrature in the azimuthal direction on
a fictitious revolved geometry in addition to the standard mesh-based quadrature in
the rz-plane.

The default in the Exterior Field Calculation feature is to evaluate the full
Helmholtz-Kirchhoff integral given in Equation 2-23 and Equation 2-24.

264 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

THE FAR-FIELD LIMIT
The full Helmholtz-Kirchhoff integral gives the pressure at any point at a finite
distance from the source surface, but the numerical integration tends to lose accuracy
at very large distances. At the same time, in some applications, the quantity of interest
is the far-field radiation pattern, which can be defined as the limit of r|p| when r goes
to infinity in a given direction.

Taking the limit of Equation 2-23 when |R| goes to infinity and ignoring the rapidly
oscillating phase factor, the far field, pfar is defined as

r⋅R
1 ik ----------- R
p far ( R ) = – ------ e R  ∇p ( r ) – ikp ( r ) ------- ⋅ n dS

4π  R
S

The relevant quantity is |pfar| rather than pfar because the phase of the
latter is undefined. For the same reason, only the direction of R is
important, not its magnitude.

Because Hankel functions asymptotically approach an exponential, the limiting 2D

integral is remarkably similar to that in the 3D case:

r⋅R
1 – i ik ----------- R
p far ( R ) = -------------- e R  ∇p ( r ) – ikp ( r ) ------- ⋅ n dS

4 πk  R
S

For axially symmetric geometries, the azimuthal integral of the limiting 3D case can be
handled analytically, which leads to a rather complicated expression but avoids the
numerical quadrature required in the general case. For zero azimuthal mode number
m = 0, the expression is:

zZ
ik -------
p far ( R ) ≅ – --- re R [ J 0  ----------- ∇p ( r ) ⋅ n –
1 krR
2  R
S (2-25)
( r )- 
in RJ 1  ----------- + n z ZJ 0  -----------  dS
ikp krR krR
----------------
R  r  R   R 

In this integral, r and z are the radial and axial components of r, while R and Z are the
radial and axial components of R.

To evaluate the pressure in the far-field limit according to the equations in this section,
set the Type of integral option to the Far-field integral approximation for r → ∞ option

THEORY BACKGROUND FOR THE PRESSURE ACOUSTICS BRANCH | 265

 in the Exterior Field Calculation section in the Settings window for the feature. See
Exterior Field Calculation.

THE ELKERNEL ELEMENT

These integrals can be implemented as integration coupling variables in COMSOL
Multiphysics. However, such an approach is very inefficient because then the simple
structure of the integration kernels cannot be exploited. In the Acoustics Module,
convolution integrals of this type are therefore evaluated in optimized codes that hides
all details from the user.

266 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Theory for the Boundary Impedance
Models
In this section:

• Impedance Conditions
• RCL Models
• Physiological Models
• Waveguide End Impedance Models
• Porous Layer Models
• Characteristic Specific Impedance Models

Impedance Conditions
An impedance boundary condition relates the acoustic pressure and the acoustic
velocity to each other at a given point. This relationship is provided by knowledge
about the conditions on the boundary, for example, a specific material or an
acoustically active boundary. Therefore, an impedance boundary condition can be
used to impose the properties of the boundary without modeling it explicitly.
Impedance boundary conditions thus generalize the sound-hard and sound-soft
boundary conditions to address a large number of cases between these two extremes.

Mathematically, a specific acoustic impedance Zi is defined on some cross section as the

ratio between the acoustic pressure p and the acoustic velocity perpendicular to the
area v ⊥ (the normal velocity)

p
Z i = ------ .
v⊥

In Pressure Acoustics, Frequency Domain, this boundary condition is imposed as

iωp t
– n ⋅  – ----- ( ∇p t – q d ) = – -----------
1
(2-26)
ρc Zi

while in the Pressure Acoustics, Transient interface using a Time Dependent study, the
impedance boundary condition is the following:

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 267

 1 ∂p t
– n ⋅  – ----- ( ∇p – q d ) = ----- --------
1
(2-27)
 ρc  Z i ∂t

Here Zi is the acoustic input impedance of the external domain and it has the unit of
a specific acoustic impedance. The specific acoustic impedance Zi (SI unit: Pa·s/m) is
related to the acoustic impedance Zac (ratio of pressure and flow rate, SI unit: Pa·s/
m3) and the mechanical impedance Zmech (ratio of force and velocity, SI unit: N·s/m)
via the area A of the boundary, according to

2
Z mech = AZ i = A Z ac (2-28)

Impedance boundary conditions only relate the normal velocity (the velocity
perpendicular to the boundary) to the pressure, but do not consider the tangential
velocity (component parallel to the boundary). This is due to the mathematical
construction of the governing equation and the fact that pressure acoustics solves only
for the scalar pressure. Put differently, the impedance boundary condition only applies
to the normal component of the incident field. Thus, by applying an impedance
boundary condition this tangential velocity component is ignored altogether. For this
reason, impedance boundary conditions are in most cases low-order approximations
to the actual boundary properties. In cases where this is unacceptable, it is
consequently better to either model the boundary explicitly or use a higher-order
model, such as for instance the Plane Wave Radiation at an open boundary.

RCL Models
Circuit models are used as low-order models of acoustically active boundaries
specifying the input impedance of the external domain. For example, modeling
mechanical properties of a loudspeaker cone, a microphone diaphragm, or biological
tissue, as experienced by the incident wave. For such boundaries, the pressure and
velocity are modulated in response to the behavior of the boundary.

An RCL model is intended to provide a simple model to represent the

frequency-dependent mechanical properties of a system that typically has some loss, a
mass, and a compliance. The option includes all combinations of a three-element
circuit consisting of an acoustic damper (acoustic resistance, a resistor Rac), an acoustic
mass (acoustic inertance, an inductor Lac) and an acoustic compliance (a capacitor
Cac), which are the simplest resonating acoustic circuits. While full circuit models tend
to be more sophisticated (see for example Ref. 26), one can typically identify a simple
core circuit responsible for the basic response which consists of such three elements.

268 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

The acoustic impedance Zac is related to the specific impedance Zi implemented in the
weak equations Equation 2-26 and Equation 2-27 by the area of the boundary A as

Z i = AZ ac .

CIRCUIT MODEL OPTIONS

A schematic illustration is provided for each circuit in the Equation Display window.

Serial Coupling of Rac,, Cac, and Lac

1
Z ac = R ac + iωL ac + ---------------
iωC ac

Parallel Coupling of Rac,, Cac, and Lac

1 1 –1
Z ac = iωC ac + -------- + ---------------
R ac iωL ac

Parallel Coupled Lac and Cac in Series with Rac

1 –1
Z ac = R ac + iωC ac + ---------------
iωL ac

Parallel Coupled Rac and Cac in Series with Lac

1 –1
Z ac = iωL ac + -------- + iωC ac
R ac

Parallel Coupled Rac and Lac in Series with Cac

1 1 1 –1
Z ac = --------------- + -------- + ---------------
iωC ac R ac iωL ac

Serial Coupled Rac and Cac in Parallel with Lac

–1 –1
Z ac = --------------- +  --------------- + R ac
1 1
iωL ac iωC ac

Serial Coupled Lac and Cac in Parallel with Rac

–1 –1
Z ac = -------- +  iωL ac + ---------------
1 1
R ac  iωC ac

Serial Coupled Rac and Lac in Parallel with Cac

–1 –1
Z ac = [ iωC ac + ( iωL ac + R ac ) ]

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 269

 LIMITATIONS
A circuit model is one-dimensional and considers only changes in the direction normal
to the boundary; any variations across the boundary are ignored. Therefore, circuit
models provide good results at low frequencies with wavelengths substantially larger
than the dimensions of the boundary.

ACOUSTICAL OR MECHANICAL CIRCUITS

It is sometimes advantageous to use circuit models in mechanical units rather than
acoustical units. In these cases, the overall structure of the impedance boundary
condition equation does not change, but the numerical value of the parameters do.
The mechanical impedance Zmech (the ratio of force F and velocity v) and the acoustic
impedance Zac are related via the boundary surface area A by the expression

2
Z mech = A Z ac .

Physiological Models
The physiological models are simple equivalent circuit models for parts of the human
body which typically are of importance in acoustical applications, namely the skin, the
full human ear, the eardrum and inner ear, and the acoustic impedance from the ear’s
pinna. These models are good approximations to the active, acoustical properties of
these body parts with particular relevance for development of hearing aids, mobile
devices, and head phones.

HUMAN SKIN MODEL

This model has been developed as a lowest-order, reasonable approximation to
experimental data in Ref. 27 (see their Figure 9). It consists of a simple serial RCL
circuit applied on a transducer area of At, and has the impedance

R s iωL s 1
Z i = ------ + ------------ + -------------------
At At iωC s A t

with At = 1.5 cm2, Rs = 9.0 N s m-1, Ls = 0.53·10-3 N s2·m-1 and Cs = 5.3·10-6 m

N-1.

MODELS RELATED TO THE HUMAN EAR

Four models related to the human ear have been included. These detailed and
experimentally-verified circuit models describe parts of the human ear, as well as the
entire human ear, see Ref. 28 to 30 for further details.

270 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Outward Human Ear Radiation
This impedance describes the acoustic radiation losses from the pinna (also known as
the auricle, this is the visible part of the ear which is exterior to the head), see
Figure 2-11. For cases where you model the ear canal explicitly using pressure
acoustics, this boundary condition describes the acoustic losses from the outward
acoustic radiation from the ear canal and into the surrounding air.

3 –1

Z rad = Y par +  Yk ,
(2-29)
k=1

ω ωk
Y k = R k  1 + iQ ------ – ------ 
1 1
Y par = ---------- + ----------------- + iωC par
R par iωL par  ωk ω 

where Rpar = 7.0·106 N·s·m-5, Lpar = 100 N·s2·m-5, Cpar = 1.7·10-12 m5 N-1,
Q = 6, R1 = R2 = Rpar, R3 = 2Rpar, ω1 = 6000·2π Hz, ω1 = 9000·2π Hz, and
ω1 = 13000·2π Hz.

Figure 2-11: Illustration of outward human ear radiation.

Human Ear Drum Impedance

This model describes the impedance of the human ear drum and the entire inner ear,
that is, the acoustic impedance experienced in the ear canal when looking into the ear

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 271

 drum, see Figure 2-12. The model equations are given in Equation 2-30 and
Equation 2-31, and the parameter values in Table 2-8.

Figure 2-12: Illustration of human eardrum impedance.

k 11 Z sc + k 12 k 11 k 12
Z eardrum = Z cav + -------------------------------- K =
k 21 Z sc + k 22 k 21 k 22
Z dmi Y mi Z incus Z dm
k 12 = -----------  1 + -----------------------------------
1
k 11 = -------- ( 1 + Y mi Z dm )
AD AD  Z dmi 
Y ac
k 21 = -------- ( 1 + Y mi Z dm ) + A D Y mi
AD
Y ac Z dmi Y mi Z incus Z dm
k 22 = --------------------  1 + ----------------------------------- + A D ( 1 + Y mi Z incus )
AD  Z dmi 
(2-30)

 A(ω) if ω < ω Aph ω
AD ( ω ) =  Φ A ( ω ) = s Aph log  ------------ + ϕ A
 A ( ω ) e iΦA ( ω ) if ω ≥ ω Aph ω Aph

A0 + A∞
A ( ω ) = ----------------------------------------------- – A∞
ω 2
1 –  ------- + ---------------
iω
ωA QA ωA
–1
1 1 –1
Z cav = Y tcav + ---------------------------- Y tcav = R tcav + -------------------
1 iωC tcav
Z ada + ----------
Y ant

272 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 iωC mac
Z ada = R ada + iωL ada Y ant = iωC ant + -------------------------------------------------------------
-
ω 2
1 –  ------------ + -------------------------
iω
 ω mac Q mac ω mac

1 iΦ ( ω )  ω 
Y ac = ------------------------------------------------------------- e Y L ac ( ω ) = L ac 0  1 + ----------- 
R ac + iωL ac ( ω ) + ---------------
1  ω YL 
ac
iωC ac
ω –1
Φ Y ( ω ) = s Yph log  1 + ------------
1
Y mi = R mi + ----------------
 ω Yph iωC mi
(2-31)
2 1
Z dm = --- Z dmi Z incus = --- Z dmi
3 3
1 –1 1
Z dmi = ----------------- + iωL oss + ( Z cpl + Z free ) Z cpl = R cpl + -----------------
iωC oss iωC cpl
–1 1
Z free = [ R free + iωL free ] Z st = R st + iωL st + --------------
iωC st
1 2
Z c = R c + iωL c + ------------- Z sc = Z st + A F Z c
iωC c

The parameter values are given in Table 2-8. Note that the value of ϕ A is not reported
in the papers Ref. 28 to 30, but has instead been determined during model
implementation. The value ensures a continuity of the phase response.
TABLE 2-8: LIST OF PARAMETERS FOR THE EAR DRUM IMPEDANCE.

SYMBOL VALUE DESCRIPTION

Rtcav 2·106 N s m-5 Tympanic cavity resistance

3
Ctcav ( 0.5cm ) ⁄ ( γp A ) Tympanic cavity compliance
Rada 1.7·106 N s m-5 Aditus ad antrum resistance
Lada 880 N s2 m-5 Aditus ad antrum inertance
3
Cant ( 0.8cm ) ⁄ ( γp A ) Antrum compliance
Qmac 0.4 Mastoid air cell resonance quality factor
3
Cmac ( 8cm ) ⁄ ( γp A ) Mastoid air cell compliance
ωmac 2π3500 rad s-1 Mastoid air cell resonance angular frequency
Rac 4·107 N s m-5 Ear drum shunt resistance
Cac 5·10-12 N-1 m5 Ear drum shunt compliance
L ac0 2.4·103 kg m-4 Zero-frequency ear drum shunt inertance
ωYL 2π·1900 rad s-1 Ear drum shunt inertance cross-over frequency
ac

ωYph 2π·8000 rad s-1 Ear drum shunt admittance cross-over frequency

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 273

 TABLE 2-8: LIST OF PARAMETERS FOR THE EAR DRUM IMPEDANCE.

SYMBOL VALUE DESCRIPTION

sYph 1.4 Ear drum shunt admittance decay slope

A0 38 mm2 Effective ear drum area, low frequency limit
A∞ 2 mm2 Effective ear drum area, infinite frequency limit
ωA 2π·2200 rad s-1 Effective ear drum area resonance angular
frequency
QA 1.3 Effective ear drum area resonance quality factor
sAph -1.2 Effective ear drum area high-frequency decay slope
-1
ωAph 2π·1500 rad s Effective ear drum area cross-over frequency
ϕA -0.8038 Effective ear drum area lag phase constant
Rmi 1 N s m-1 Mechanical resistance of incudomalleal joint
Cmi 0.04 mm N-1 Mechanical compliance of incudomalleal joint
-3 -1
Coss 3·10 mN Mechanical compliance of ossicles
-3
Loss 7·10 g Mechanical inertance of ossicles
Ccpl 0.5·10-3 m N-1 Mechanical compliance of ossicles-drum coupling
-1
Rcpl 0.08 N s m Mechanical resistance of ossicles-drum coupling
-1
Rfree 0.02 N s m Mechanical resistance of peripheral parts of the
drum
Lfree 12·10-3 g Mechanical inertance of peripheral parts of the
drum
Rst 18·10-3 N s m-1 Mechanical resistance of stapes
Lst 3·10-3 g Mechanical inertance of stapes
Cst 1.2·10-3 m N-1 Mechanical compliance of stapes
Rc 70·10-3 N s m-1/ Resistance of cochlea
2
AF
Lc 10·10-3 g / AF2 Inertance of cochlea
Cc 11·10-3 m N-1·AF2 Compliance of cochlea
2
AF 3 mm Stapes footplate area
Human Ear Without Pinna
This model accounts for the acoustic losses associated with the ear canal and the entire
human ear, see Figure 2-13. It does not include the radiation losses associated with the
pinna, the visible part of the ear which is external to the head.

274 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 T 12
T 11 + --------------------
Z eardrum
Z ear w/o pinna = ------------------------------------- (2-32)
T 22
T 21 + --------------------
Z eardrum

Here, Zeardrum is the eardrum impedance defined in Equation 2-31 and Tij are the
components of the transfer matrix T that describes the ear canal as a two-port. The ear
canal is treated as Ntot small segments each with length Δk and radius rk so its full two
port T is given by

N tot cosh ( Γ k Δ k ) sinh ( Γ k Δ k )Z twk

T 11 T 12
T =
T 21 T 22
= ∏ sinh ( Γ k Δ k )
------------------------------
- cosh ( Γ k Δ k )
. (2-33)
k=1 Z twk

where

ρc ω
Z twk = --------2- Γ k = 3α k + i ----
πr k c

ω 1 1 1
α k = ---- ------------------- ζ 1 + -----------2 ζ 2 + ------------------
- ζ3
c 2Wo Wo k Wo k 2
3
k

2 (2-34)
γ–1 γ – 1 γ – 1 + (γ – 1)
ζ 1 = 1 + ----------- ζ 2 = 1 + ----------- – -------------------------------------
Pr Pr 2Pr
2 2 3
7 γ – 1 γ – 1 + (γ – 1) 4 ( γ – 1 ) + 4 ( γ – 1 ) – ( γ – 1 )-
ζ 2 = --- + ----------- – ------------------------------------- + --------------------------------------------------------------------------
8 Pr 2Pr 3
---
2
8Pr

In these expressions, γ is the ratio of specific heats, Γk is the propagation constant

(“wave number”) of the kth ear canal segment which has the segment-specific
attenuation αk,

2
ρωr
Wo k = ------------k-
μ

is the segment-specific Womersley number, and

Cp μ
Pr = ----------
k

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 275

 is the Prandtl number expressed in terms of the specific heat Cp, dynamic viscosity μ,
and thermal conductivity k. Notice that the papers presenting the model (Ref. 28 to
30) do not exactly specify which expression for the attenuation constants αk is being
used, but only refer to Ref. 31. The expression above for αk is the most general
expression taken from this paper. The values for Δk and rk are listed in Table 2-9.

Figure 2-13: Illustration of the impedance of the human ear without pinna.
TABLE 2-9: RADII rk AND LENGTHS Δk OF EAR CANAL SEGMENTS TAKEN FROM Ref. 30.
k rk (mm) Δk (mm)
1 1.0 1
2 1.9 1
3 2.4 1
4 2.6 1
5 2.75 1
6 2.85 1
7 2.95 1
8 3.05 1
9 3.2 2
10 3.35 2
11 3.5 2
12 3.6 2
13 3.7 2
14 3.8 2
15 3.9 2
16 4.0 2

276 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

TABLE 2-9: RADII rk AND LENGTHS Δk OF EAR CANAL SEGMENTS TAKEN FROM Ref. 30.
k rk (mm) Δk (mm)
17 4.1 2
18 4.25 2
The pressure at the eardrum peardrum is calculated whenever this impedance boundary
condition is applied. This pressure is available in postprocessing, and is calculated from
the expression

T 12
T 22 – --------------------------------
Z ear w/o pinna
p eardrum = p t ------------------------------------------------ ,
T 11 T 22 – T 12 T 21

where pt is the pressure on the boundary, Tij are the coefficients of the ear canal
transfer matrix T and Zear w/o pinna is the ear impedance; T and Zear w/o pinna are
defined in Equation 2-32 and Equation 2-33 above.

Human Ear, Full

This model accounts for all acoustic losses associated with the entire human ear, both
the internal parts as well as the pinna, the visible, external part of the ear on the head,
see Figure 2-14. The model does not include any information about the directivity of
the ear (the head related transfer functions, HRTFs) which depends on the ear
geometry at higher frequencies. It is valid in the low frequency limit and for normal
incidence on the ear. It is given by

Z ear full = Z rad + Z ear w/o pinna

where Zrad and Zear w/o pinna are given by Equation 2-29 and Equation 2-32 above.

Figure 2-14: Illustration of the impedance of the full human ear including radiation
losses due to the pinna.

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 277

 The pressure at the eardrum peardrum is calculated whenever this impedance boundary
condition model is applied. This pressure is calculated from Zear full using the
expression

T 12
T 22 – -----------------
Z Z
= p t  1 – ----------------- -------------------------------------------- .
rad ear full
p eardrum
Z ear full T 11 T 22 – T 12 T 21

Waveguide End Impedance Models

Tubes and ducts are acoustic waveguides, and there are acoustic radiation losses when
such a waveguide opens into a large domain. Idealized models for these losses have
been implemented as boundary impedance models. Thus, instead of explicitly
modeling the large domain, an appropriate impedance model Zend can be applied with
Zi = Zend. These models all assume that the domain is infinitely big, that the
propagation is in the direction of the waveguide axis, and that the propagating mode
is a plane wave. As with all other impedance boundary models, only the
boundary-normal velocity component is taken into account.

In the following the term ρc takes different values depending on when the impedance
condition is applied. Specifically, for models with damping they are equal to the
complex valued quantities ρccc, while when applied on an Anisotropic Acoustic domain
the normal direction variables are used ρncn.

Flanged Pipe, Circular

For a pipe of a user-specified radius a, the acoustic losses are given by (see Ref. 6)

2J 1 ( 2ka ) 2H 1 ( 2ka ) ω
Z end = ρc  1 – -------------------------- + i --------------------------- k = ----
2ka 2ka c

where J1(x) is the Bessel function of the first kind of order 1, H1(x) is the Struve
function, and k is the wave number of the wave. This expression is also known as the
impedance from a baffled piston.

Flanged Pipe, Rectangular

For a rectangular duct of user-specified inner width wi and inner height hi, the acoustic
losses are given by (see Ref. 5)

278 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 3
--- w
ρc ω
= ------------------ ( w i h i ) k + ik ( w i h i ) f  ------
2 2 2 i
Z end k = ----
2πw i h i hi c
1 1 3 3 3 .
--- – --- --- – --- ---
2 – 1  1 2 2
2 –1 2 –1 2
f ( x ) = 2x sinh  --- + 2x sinh ( x ) + --- x + x – (x + x )
x 3

This relationship applies provided the following requirements are satisfied

wi ≥ hi kw i « 1 kh i « 1 .

Unflanged Pipe, Circular (Low ka Limit)

For an unflanged circular pipe of a user-specified radius a in the limit of small radius
(low ka), the pipe end impedance is given by the classical expression (see Ref. 6)

2
( ka ) ω
Z end = ρc -------------- + i0.6133ka k = ---- ka « 1 .
4 c

Unflanged Pipe, Circular

For an unflanged pipe of any user-specified radius a relative to the wave number k, an
approximate end impedance is given in Ref. 32. It is

1+R 2ikaδ ( ka ) ω
Z end = ρc -------------- R = –Re k = ----
1–R c

 – (-------------
ka )
2
- 4
 e (--------------
ka )   ------------------------ 19
------
2 1
1 if ka < 1
 log  0.5772  + 12
 +
R =  6 e ka
 3 1
 πkae
– ka
1 + ------ --------------2 if 1 < ka < 3.83
 32 ( ka )


where δ(ka) is a tabulated function reproducing the curve in Fig. 2 in Ref. 32 (where
δ(ka) is referred to by l/a).

Porous Layer Models

A boundary impedance model is implemented to specifically handle cases investigating
the acoustical properties of a porous layer of a given thickness d backed by a
sound-hard wall. This can be applied instead of explicitly modeling the porous layer,
as long as the incident acoustic field is normal to the boundary; as for all the other

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 279

 boundary impedance models, the tangential components of the acoustic field are
ignored by this model.

For a porous layer with a user-specified thickness d, the impedance from the porous
layer backed by a sound-hard wall is given by (see Ref. 9)

ω
Z i = – iρ c c c cot ( kd ) k = ---- .
cc

Here ρc and cc are the equivalent fluid descriptions of the porous model. This
impedance model applies to any type of porous model which can be written as an
equivalent fluid model. All porous models implemented in COMSOL are available for
this impedance boundary condition. See details of the poroacoustics equivalent fluid
models in the section Theory for the Equivalent Fluid Models.

Characteristic Specific Impedance Models

For well-defined wave types in infinite domains, an impedance condition exists in every
point whereby the pressure and normal velocity are related. Thus, these impedances
can be imposed on a boundary to model an infinite, open domain in cases where the
wave source inside the domain is either a direction (plane wave), a point (spherical
wave) or a line (cylindrical wave). However, be aware that only the boundary-normal
component of the velocity is used in the impedance boundary condition while the
tangential component is ignored; in cases of nonnegligible tangential components it is
recommended to instead use the options Plane Wave Radiation, Spherical Wave
Radiation, and Cylindrical Wave Radiation.

Plane Wave
The impedance is given by

Zi = ρc cc ,

see Ref. 6. This is given solely by material parameters and has no user input.

Spherical Wave
This impedance corresponds to the wave from a point source. It is calculated from the
expression (given in Ref. 6)

ρc cc ω
Z i = --------------------- rb = x0 – x k = ----
1 cc
1 + -----------
ikr b

280 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

where x0 is the user-specified location of the point source generating the spherical
waves, x is the position of the boundary, and rb is the distance between the two.

Cylindrical Wave
This classic infinite-domain wave impedance from an acoustic line source with the
user-specified direction n̂ 0 and position x0 is presented in, for example, Ref. 6. The
expression for the impedance is

(2)
H 0 ( kr b ) ω
Z i = iρ c c c -------------------------
(2)
k = ----
H 1 ( kr b ) cc

( x 0 – x ) × e sa
r b = ( x 0 – x ) – { ( x 0 – x ) ⋅ e sa }e sa = -----------------------------------
e sa

(2)
where H m ( x ) = J m ( x ) – iY m ( x ) is the Hankel function of the second kind of order
m given in terms of the Bessel functions of order m of the first and second kind, Jm(x)
and Ym(x), respectively. Notice that the source axis vector esa is automatically
normalized in this implementation.

THEORY FOR THE BOUNDARY IMPEDANCE MODELS | 281

 Theory for the Interior Impedance
Models
In this section:

• Interior Perforated Plate Models

Interior Perforated Plate Models

The transfer impedance of an interior boundary is defined as the ratio of the pressure
drop across the boundary to the velocity on the boundary:

Δp t p t,up – p t,down
Z i = --------- = ----------------------------------- (2-35)
vn vn

The resulting value of Zi can be treated as a superposition of several contributions

which are derived separately. Note that the theory below is only valid for the perforates
with circular-shaped holes. Other types of holes can lead to significantly different
results, which will make the models that are considered here inadequate and unreliable.

TRANSFER IMPEDANCE OF A HOLE

Let the z-coordinate axis be directed along the axis of a cylindrical hole of the height
tp (see Figure 2-15). Let the variation of the pressure and the velocity along the z-axis
have the following pattern:

– ik c z – ik c z
p ( z ) = pe , v ( z ) = ve (2-36)

The substitution of Equation 2-36 into Equation 2-35 results in

p ( z + tp ) – p ( z ) p ( z ) –ik t ⁄ 2 ik t ⁄ 2 kc tp
Z i = ---------------------------------------- = ------------- ( e c p – e c p ) = – 2iZ c sin ---------- ,
vn ( z + tp ⁄ 2 ) vn ( z ) 2

where Zc is the characteristic impedance and kc is the complex wave number defined
according to the low reduced frequency (LRF) models from the Narrow Region
Acoustics for Slits, Circular Ducts, Rectangular Ducts, and Equilateral triangular
Ducts. That is,

282 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 1
1
– --- ---
ω 1 2
k c = ----  [ γ – ( γ – 1 )ϒ th ] ------ .
2
Z c = ρc ( [ γ – ( γ – 1 )ϒ th ]ϒ v ) ,
c ϒ v

tp

dh z

Figure 2-15: To the derivation of the transfer impedance of a one-hole perforate.

The expression for the normalized transfer impedance of an orifice reads

kc tp
2i sin  ----------
Zi  2 
z i = ------ = – ------------------------------------------------- . (2-37)
ρc [ γ – ( γ – 1 )ϒ th ]ϒ v
kc tp kc tp
For thinner plates ( kt p « 1, k = ω ⁄ c ) the approximation sin  ---------- ≈ ---------- is valid,
 2  2
which yields

Zi iω t p
z i = ------ = – ------ ------ . (2-38)
ρc c ϒv

Note that Equation 2-38 coincides with the expression that follows from Crandall’s
formula for an infinite tube with the circular cross section (see Ref. 35). Equation 2-37
accounts for both viscous and thermal effects inside the hole, while the simplified
Equation 2-38 contains the viscous part only (thermal effects are negligible for the
thin plate limit).

THEORY FOR THE INTERIOR IMPEDANCE MODELS | 283

 VENA CONTRACTA AND AREA POROSITY
Equation 2-37 and Equation 2-38 are exact if the streamlines of a flow through the
hole are parallel to the z-axis throughout the orifice area. In reality, there is a radial
component of the flow, which leads to the reduction (contraction) of this area. The
minimum area where the streamlines remain parallel to the z-axis is called vena
contracta. The flow velocity at the vena contracta is also different from that of the ideal
flow. A coefficient that accounts for these effects is called the discharge coefficient, CD.
The value of the discharge coefficient can be obtained from measurements as function
of the plate thickness and the orifice shape and diameter.

Another parameter that used to express the interior impedance of a perforate is the
porosity, σ. The holes are usually uniformly distributed over the plate, and the porosity
accounts for the distribution as the ratio of the hollow area to the area of the plate.
Depending on the pattern used when the holes are strewn over the plate, the porosity
is defined as follows

2 2
πd h πd h
σ = ---------2- and σ = -----------------2
4a 2 3a

for a square and a triangular pattern, respectively. The parameter a determines the hole
spacing.

The resulting expression for the interior impedance of a perforate comes out from
dividing zi by the product σCD:

zi
z orifice = ------------ . (2-39)
σC D

Either Equation 2-37 or Equation 2-38 can be used for the substitution for zi in
Equation 2-39. The resulting models will further be referred to as the thick and the
thin plate models, respectively.

END CORRECTION AND HOLE-HOLE INTERACTION

The subscript “orifice” in Equation 2-39 means that the expression accounts for the
transfer impedance of a perforate caused by the presence of the holes; that is, a piston
of fluid of the length tp. However, the actual mass of fluid affected by an incident wave
is larger than that inside the hole. The effective mass of the fluid can be taken into
account by the piston which is on each side longer by δ than the initial one (see
Figure 2-16). This results in adding two extra terms (for each side of the perforate) of

284 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

the form Equation 2-38 with tp replaced by δ. The choice of Equation 2-38 is due to
the absence of highly conducting walls in the end corrections area.

If two holes are located relatively close to one another, the actual masses of the
attached fluid can become overlapped. This makes the total mass less than just the sum
of those for the separate hole. In order to take the hole-hole interaction into account,
the end

Equivalent
piston

Actual affected
mass of fluid

Figure 2-16: To the definition of end corrections.

correction is reduced by a factor fint. The last is a function of the porosity and the most
often expressed by the Fok function:

N
n
ψ(σ) =  an ( σ) .
n=0

The end correction is usually considered as a function of the hole diameter and can in
practice be different for the resistive and the reactive parts of the transfer impedance.
For this reason, it is useful to split δ into two parts: δresist and δreact. The resulting
term that accounts for the end correction and the hole-hole interaction reads

iω 2δ resist iω 2δ react
z end = – Re  --------------- ---------------- f int – iIm  --------------- ---------------- f int . (2-40)
 cσC D ϒ v   cσC D ϒ v 

THEORY FOR THE INTERIOR IMPEDANCE MODELS | 285

 Equation 2-40 is only acceptable if the media is the same on both sides of
the perforate. If the plate is backed by a porous layer on one side, the
values of δ and fint can differ significantly from those on the other side.
For example, the interaction between holes is hampered by porous media,
which results in neglecting of fint on one side of the plate and leads to the
following correction factor: (δ + 1)fint.

NONLINEAR AND MEAN FLOW EFFECTS

At medium and high sound pressure levels, the displacement of acoustic particles
becomes comparable to the diameter of the holes. This causes flow separation and
vortex shedding at the entrance and the exit of the hole. This results in the acoustic
energy dissipation and increases the acoustic resistance of the perforate Ref. 36. That
is, an extra resistance term should be added to the resulting transfer impedance
expression.

Different forms of the contribution to the resistance are similar in the following sense:

• The resistance term is directly proportional to the acoustic particle velocity, which
makes it nonlinear;
• The term is frequency dependent through the acoustic particle velocity.
The expression incorporated into the Interior Perforated Plate boundary condition
reads

2 f nl
1 – σ -----
z nl = ---------------
2 2 2c n
-v , (2-41)
σ CD

where fnl is a correction factor (equals 1 by default) and vn is the acoustic particle
velocity component normal to the plate. Other expressions for the nonlinear resistance
term can be found in Ref. 36.

The presence of a mean flow also changes of the transfer impedance. In order to
account for these and other possible effects, the Interior Perforated Plate boundary
condition feature includes an option to enable user-defined resistance and reactance:

( user ) ( user )
z user = θ + iχ . (2-42)

286 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Relative to the mean flow, the contribution of a grazing or a bias flow to the resistance
can be expressed through the flow Mach number as shown in Ref. 37.

Combining the expressions Equation 2-39–Equation 2-42 together yields the full
expression for the transfer impedance of a perforate for the thick

kc tp
Zi 2i sin  ----------  iω 2δ resist   iω 2δ react 
2
------ = – ------------------------------------------------- – Re  -------------------- ---------------- f int – iIm  -------------------
- ---------------- f int
ρc [ γ – ( γ – 1 )ϒ th ]ϒ v  cσC D ( lin ) ϒv   cσC (Dlin ) ϒ v 

2
1 – σ f nl ( user ) ( user )
+ -----------------------2 ------ v n + θ + iχ
( nl ) 2c
( σC D )

and the thin

Z  iω t p + 2 δ resist   iω t p + 2 δ react 
------i = – Re  -------------------- ---------------------------
- f int – iIm  -------------------
- --------------------------- f int
ρc  cσC ( lin ) ϒv   cσC ( lin ) ϒv 
D D
2
1 – σ f nl ( user ) ( user )
+ -----------------------2 ------ v n + θ + iχ
( nl ) 2c
( σC D )

plate models. The discharge coefficient CD is taken different for the linear and the
nonlinear parts for the sake of flexibility.

THEORY FOR THE INTERIOR IMPEDANCE MODELS | 287

 Theory for the Equivalent Fluid
Models
In this section:

• Introduction to the Equivalent Fluid Models

• About the Pressure Acoustics Fluid Models
• About the Poroacoustics Models
• About the Narrow Region Acoustics Models

Introduction to the Equivalent Fluid Models

It is possible to define the properties of a fluid in several ways in pressure acoustics. In
a Pressure Acoustics domain feature attenuation properties for the bulk fluid may be
specified. Acoustic losses in porous materials are modeled by homogenizing the porous
matrix and saturating fluid, and are defined by the Poroacoustics domain feature
(frequency domain only). The viscous and thermal losses that occur in the acoustic
boundary layer can be modeled in a homogenized way using the Narrow Region
Acoustics domain feature (frequency domain only). The different ways of defining the
properties of a fluid are called fluid models. They are also often referred to as
equivalent fluid models.

Losses and damping occur when acoustic waves propagate in a porous material (where
material refers to the homogenization of a fluid and a porous solid) because of bulk
viscous and thermal properties, or because of thermal and viscous losses in the acoustic
boundary layer at walls in narrow ducts. The purpose of the fluid model is to mimic a
special loss behavior by defining a complex-valued density ρc and speed of sound cc.
These are often frequency dependent.

In a Pressure Acoustics domain feature, the default Linear elastic fluid model (see
Linear Elastic Fluid Model) enables you to specify a linearly elastic fluid using either
the density ρ and speed of sound c, the impedance Z and wave number k, or the
equivalent bulk modulus K and the density ρ. When any of these material parameters
are complex-valued, damping is introduced.

It is always necessary to specify a set of two parameters (for example Z and k) or

conditions in order to calculate the complex speed of sound and complex density

288 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

needed to specify a fluid model. The choice of parameters typically depend on the
application and which equivalent fluid is being modeled. For example:

• It is possible to determine the complex wave number k and impedance Z from

directly measuring it in an impedance tube in order to produce curves of the real
and imaginary parts (the resistance and reactance, respectively) as functions of
frequency. These data can be used directly as input to COMSOL Multiphysics
interpolation functions to define k and Z.
• The option to define the equivalent bulk modulus K and density ρ is often used
when characterizing the propagation of acoustic waves in a porous material. These
parameters may be determined from measurements or by defining an analytical
model expression.

The linear elastic fluid model enables the user to enter any desired fluid models with
the desired combination of fluid properties. It is the most general fluid model. You
may enter any user-defined analytical expressions for your favorite equivalent fluid
model or use measurement data to represent the lossy behavior of the fluid.

PROPAGATION IN GENERAL FLUID WITH BULK LOSSES

The options are (see About the Pressure Acoustics Fluid Models and the settings for
the Pressure Acoustics node):

• Linear elastic: define density and speed of sound, impedance and wave number, or
equivalent bulk modulus and density.
• Linear elastic with attenuation: define an attenuation parameter for the fluid.
• Ideal gas is also available but not described here. This fluid model is used to specify
the fluid properties by selecting a gas constant type and selecting between entering
the heat capacity at constant pressure or the ratio of specific heats. See Ideal Gas
Fluid Model for details.
• Viscous and/or thermally conducting fluids (bulk losses): the losses are due to
viscous losses, thermal conduction, or the combined thermal conduction and
viscous losses in the bulk of the fluid.

PROPAGATION IN POROUS MATERIALS

A series of fluid models exist for describing the propagation of pressure waves in porous
materials. These range from fully empirical models to semianalytical/empirical models
with varying degree of complexity. See Poroacoustics and About the Poroacoustics
Models for more detail.

THEORY FOR THE EQUIVALENT FLUID MODELS | 289

 PROPAGATION IN NARROW REGIONS, NARROW TUBES, OR WAVEGUIDES
See Narrow Region Acoustics and About the Narrow Region Acoustics Models. The
losses are due to absorption/dissipation in the acoustic boundary layer (thermal and
viscous losses). The losses are smeared on the domain in a homogenized way.

About the Pressure Acoustics Fluid Models

USER-DEFINED ATTENUATION FLUID MODEL

Use the user-defined attenuation fluid model to specify an attenuation defined through
an attenuation coefficient α. The attenuation can be an analytical expression or given
by measurement data imported using an interpolation function. This results in a
complex-valued wave number k. This defines the complex speed of sound.

ω
c c = ----
k

There are different attenuation types to select from: attenuation coefficient Np/m,
attenuation coefficient dB/m, or attenuation coefficient dB/λ. Select Attenuation
coefficient Np/m to define an attenuation coefficient in Np/m (nepers per meter), to
define:

ω c 2
k = ---- – iα ρ c = ρ  ----
c cc

This results in a real valued bulk modulus of the fluid K = ρc2.

Select Attenuation coefficient dB/m to define an attenuation coefficient in dB/m

(decibel per meter), to define:

ω α c 2
k = ---- – i ln ( 10 ) ------ ρ c = ρ  ----
c 20 cc

This results in a real valued bulk modulus of the fluid K = ρc2.

Select Attenuation coefficient dB/λ to define an attenuation coefficient in dB/λ (decibel

per wavelength). This option is often used in under water acoustics. The definition of
the complex valued wavenumber and (real) density follows:

ω α
k = ----  1 – i ln ( 10 ) ------------------ ρc = ρ
c 2π ⋅ 20

290 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Notice the different definition of the complex density. This definition is commonly
used in underwater acoustics, where the density is kept as a real value, see Ref. 62.

Feature description: User-Defined Attenuation Fluid Model

VISCOUS FLUID MODEL

The viscous model is an equivalent-fluid model that mimics the propagation of sound
in a fluid including viscous losses occurring in the bulk of the fluid. The elastic fluid
model with viscous losses is defined by:

1
---
iωδ – 1 iωδ 2
ρ c = ρ  1 + --------
- c c = c  1 + --------
-
 2   2 
c c

δ = ---  --- μ + μ B
1 4
ρ3 

where μ is the dynamic viscosity and μB is the bulk viscosity (see Ref. 4 or Ref. 6). This
choice is only appropriate for situations where the damping takes place in free space
and is not related to interaction between the fluid and a solid skeleton or a wall. These
losses, in most fluids, occur over long distances or at very high frequencies.

THERMALLY CONDUCTING FLUID MODEL

The thermally conducting model is an equivalent-fluid model that mimics the
propagation of sound in a fluid including losses due to thermal conduction in the bulk.
The elastic fluid model with thermal losses is defined by:

--1-
iωδ – 1 iωδ 2
ρ c = ρ  1 + --------
- c c = c  1 + --------
-
 2   2 
c c
1 ( γ – 1 )k
δ = ---  --------------------
ρ  Cp 

where γ is the ratio of specific heats, Cp is the specific heat at constant pressure, and k
is the thermal conductivity (see Ref. 6 chapter 9). This choice is only appropriate for
situations where the damping takes place in free space and is not related to interaction
between the fluid and a solid skeleton or a wall.

THEORY FOR THE EQUIVALENT FLUID MODELS | 291

 THERMALLY CONDUCTING AND VISCOUS FLUID MODEL
The thermally conducting and viscous model is an equivalent-fluid model that mimics
the propagation of sound in a fluid including losses due to thermal conduction and
viscosity in the bulk of the fluid. The elastic fluid model with thermal and viscous losses
is defined by:

1
---
iωδ –1 iωδ 2
ρ c = ρ  1 + --------- c c = c  1 + ---------
 2   2 
c c
( γ – 1 )k
δ = ---   --- μ + μ B + --------------------
1 4
ρ 3 Cp

where μ is the dynamic viscosity and μB is the bulk viscosity, γ is the ratio of specific
heats, Cp is the specific heat at constant pressure, and k is the thermal conductivity (see
Ref. 6 chapter 9). This choice is only appropriate for situations where the damping
takes place in free space and is not related to interaction between the fluid and a solid
skeleton or a wall.

Feature description: Thermally Conducting and/or Viscous Fluid Model

About the Poroacoustics Models

The Poroacoustics node introduces several fluid models for modeling the propagation
of acoustic waves in porous materials. After some general remarks about fluid models
and the rigid and limp regime approximations, these models are discussed in this
section:

• Porous Fluid Models

• Delany-Bazley-Miki
• Zwikker-Kosten
• Attenborough
• Wilson
• Johnson-Champoux-Allard (JCA)
• Johnson-Champoux-Allard-Lafarge (JCAL)
• Johnson-Champoux-Allard-Pride-Lafarge (JCAPL)

292 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

• Williams EDFM (equivalent density fluid model)
• Wood (fluid suspension Model)

POROUS FLUID MODELS

The poroacoustics fluid models are equivalent fluid models that mimic the behaviors
of a full Poroelastic Material model, which is defined by Biot’s theory. A poroacoustics
fluid model is based on describing the frequency-dependent effective fluid density ρ(ω)
and the effective fluid bulk modulus K(ω) of the combined equivalent fluid-solid
system (saturating fluid and porous matrix). The description of these models includes
the losses associated with the propagation of acoustic waves in porous materials. An
equivalent fluid model is computationally less demanding than the full poroelastic
model. However, it is only physically correct for certain choices of material parameters.
Most poroacoustic models are only valid in the rigid or limp porous matrix
approximations.

Limp and Rigid Porous Matrix Models

In the rigid porous matrix limit, the matrix is assumed to be so stiff that it does not
move (sometimes referred to as a motionless skeleton model). In this case it is assumed
that in Biot’s theory u = 0, which yields a wave equation with complex density and
bulk modulus. In a rigid porous material the Biot-Willis coefficient is equal to the
porosity αB = εP. All the poroacoustic fluid models are based on defining the rigid
effective density ρrig (see Ref. 9).

The limp porous matrix limit is the opposite of the rigid assumption. It is used to
model materials where the stiffness of solid phases (the porous matrix) is so weak that
it cannot support free, structure-borne wave propagation (neither longitudinal nor
transverse). That is, the stiffness (in vacuo bulk stiffness) is very small compared to air
such that the solid phase motion becomes acoustically significant. If it is light enough,
the solid phase still moves because it is “dragged along” by the fluid motion; a limp
porous material model is also an equivalent fluid model because it only features a single
longitudinal wave type. Typically, the limp assumption can be applied to very light
weight fibrous materials (less than 10 kg/m3) if these are not stiffened by the injection
of binder material. In the limp case, it is assumed that the stress tensor vanishes and
that in Biot’s theory, the Biot-Willis coefficient is αB = 1. The limp density is related
to the rigid density by a mixture model (see Ref. 10)

2
ρ rig ρ av – ρ f
ρ limp = ----------------------------------- ρ av = ρ d + ε p ρ f
ρ av + ρ rig – 2ρ f

THEORY FOR THE EQUIVALENT FLUID MODELS | 293

 where ρf is the fluid density, ρd is the drained porous matrix density, ρav is the average
effective density, and ρlimp is the resulting effective limp density.

Fluid Parameters
Defining the equivalent density and bulk modulus results in the usual complex-valued
speed of sound and density (the propagation parameters):

ω 2 3
k eq =  ---- – k z –  -----
2 2 m
 c c  r

K K-
cc = -------- or cc = -----------
ρ rig ρ limp
ρ c = ρ rig or ρ c = ρ limp

Ideal Gas and General Fluid Options

We show below that all implemented models of poroacoustics apply for all fluids (gases
and liquids) except the purely empirical Delany-Bazley-Miki model which has been
obtained from fitting to measurement data.

The applicability of the poroacoustics models to general fluids is not described in the
literature, but follows directly from repeating the published derivations without
assuming the saturating fluid to be an ideal gas. The only difference is that one needs
to use the following general expression for the ratio of specific heats γ (express γ in
terms of other material parameters) which is valid for any fluid (for gases γ is typically
a known parameter)

2 2 2
α0 T0 c α0 T0
γ – 1 = -------------------- = -------------------
ρ0 Cp β0 Cp
1 dρ
α 0 = – ------ ------- , (2-43)
ρ 0 dT p
1 dρ 1
β 0 = ------ ------ = -------
ρ 0 dp T
KT

where α0 is the isobaric thermal expansion coefficient, T0 is the background quiescent

temperature, ρ0 is the background quiescent density, Cp is the specific heat at constant
pressure, β0 is the isothermal compressibility (inverse isothermal bulk modulus KT),
and c is the speed of sound. This result comes from thermodynamics.

To show the general applicability of the poroacoustics models, we here rederive the
equivalent bulk modulus for the simplest case of the Zwikker-Kosten model for a
general fluid. All other implemented models (again except the Delany-Bazley-Miki

294 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

model) are extensions of this result (see Ref. 9), and the equivalent bulk moduli can
therefore in general be written as

K0
K ( ω ) = -------------------------------------- (2-44)
γ – ( γ – 1 )ψ ( ω )

for any of these models, where the frequency response Ψ(ω) is specific to each model.
Consequently, the following derivation is sufficient to show the general result.

Derivation of the Equivalent Bulk Modulus Valid for Any Fluid in the Zwikker-Kosten
Theory
In general, the (isentropic) bulk modulus is given by

dp
K 0 = ρ ------ . (2-45)
dρ

A relationship between the pressure p and the density ρ is needed to define the
effective bulk modulus, and the equation of state provides such a relationship. Instead
of relying on the ideal gas equation, we here use the following general equation of state

ρ = ρ ( p, T ) (2-46)

which merely stipulates that the density ρ is a function of both pressure p and
temperature T. The temperature-dependence is nonnegligible since we are dealing
with small pores, and acoustics in porous materials belong to thermoviscous acoustics
(see Thermoviscous Acoustics Interfaces). Assuming the acoustic fields are small
harmonic perturbations (denoted by a subscript 1) about a quiescent reference state
(denoted by subscript 0)

i ωt
ρ = ρ0 + ρ1 e
i ωt (2-47)
p = p0 + p1 e
i ωt
T = T0 + T1 e

and Taylor-expanding the general equation of state Equation 2-46 to first order, we
find

ρ1 = ρ0 ( βT p1 – α0 T1 ) (2-48)

where βT is the isothermal compressibility.

In order to calculate the effective bulk modulus from Equation 2-45 using the
equation of state Equation 2-48, a relationship between T1 and p1 is derived from the

THEORY FOR THE EQUIVALENT FLUID MODELS | 295

 linearized energy equation (see Theory Background for the Thermoviscous Acoustics
Branch)

iω α0 T0
0 =  ∇2– --------- T 1 + ------------- iωp 1 + ----
Q
D th k k
(2-49)
k
D th = -------------
ρ0 Cp

where Dth is the thermal diffusivity, k is the thermal conductivity, and Q is a volumetric
heat source. To proceed with the solution of this equation, it is generally assumed that
(i) the pore can be modeled as a hollow cylinder, (ii) the contributions from radial
variations dominate over axial variations so it suffices to solve the radial problem while
assuming no axial variations Ref. 9, and (iii) the pore radius is much smaller than the
wavelength, whereby the pressure does not change noticeably across the radius. By
these assumptions the pressure p1 can be treated as a constant in Equation 2-49 and
the equation has only radial variations:

2
1∂ ∂
∇2T 1 → --- T 1 + 2 T 1
r ∂r ∂r

The cross-sectional average solution under these assumptions, and with the boundary
condition of zero acoustic temperature T1 on the pore walls at r = R, is

α0 T0 Q 1 2 J 1 ( Wo Pr – i )
T 1 (r) = ------------- p 1 + ------------- ------ 1 – ---------------------------- ---------------------------------------- . (2-50)
ρ0 Cp ρ 0 C p iω Wo Pr – i J ( Wo Pr – i ) 0

Here

2
R ωρ μ
Wo = -----------------0- Pr = ------------- (2-51)
μ ρD th

where Wo is the Womersley number (see the note below), and Pr is the Prandtl
number, measuring the relative width of the viscous boundary layer thickness to the
thermal boundary layer thickness.

With the solution from Equation 2-50, the pressure can now be expressed as a
function of the density using the first-order equation of state Equation 2-48:

296 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 Qα 0 2 J 1 ( Wo Pr – i )
ρ 1 + ----------- 1 – ---------------------------- ----------------------------------------
Cp Wo Pr – i J 0 ( Wo Pr – i )
p 1 ( ρ 1 ) = ---------------------------------------------------------------------------------------------------------------------------
2
- (2-52)
 α0 T0 2 J 1 ( Wo Pr – i ) 
ρ 0  β T – ------------- 1 – ---------------------------- ---------------------------------------- 
 ρ0 Cp Wo Pr – i J ( Wo Pr – i )  0

From this we obtain the following expression for the equivalent bulk modulus (see
Equation 2-45)

dp 1 K0
K ( ω ) = ρ 0 --------- = -------------------------------------------------------------------------------------------------------------- (2-53)
dρ 1 J 1 ( Wo Pr – i )
2
γ – ( γ – 1 ) 1 – ---------------------------- ----------------------------------------
Wo Pr – i J 0 ( Wo Pr – i )

where the ratio of specific heats γ is defined in Equation 2-43 for a general fluid and
K0 is the isentropic bulk modulus of the fluid. This formula Equation 2-53 is
mathematically equivalent to the normal Zwikker-Kosten formula for the equivalent
bulk modulus (see Ref. 15), but it has been derived for a general fluid and not just
an ideal gas.

We emphasize that from this result for a general fluid follows the general formula
Equation 2-44, in this particular case with

2 J 1 ( Wo Pr – i )
ψ ( ω ) = 1 – ---------------------------- ---------------------------------------- (2-54)
Wo Pr – i J 0 ( Wo Pr – i )

with the Womersley number Wo containing the frequency dependence.

The Womersley number Wo measures the influence of viscous effects

relative to the oscillation frequency ω. For Wo « 1 viscosity dominates
and the velocity profile is the well-known Poiseuille parabola, while in the
Helmholtz regime for Wo » 1 the velocity profile is plug-like with a very
small boundary layer close to the walls.

DELANY-BAZLEY-MIKI
The Delany-Bazley-Miki model is an equivalent fluid model that mimics the bulk losses
in certain porous/fibrous materials. The model represents a porous medium with the
following complex propagation constants:

THEORY FOR THE EQUIVALENT FLUID MODELS | 297

 ω –C –C
k c = ---- [ 1 + C 1 X 2 – iC 3 X 4 ]
c
–C 6 –C8
Zc = ρf c [ 1 + C5 X – iC 7 X ]
f ⋅ ρf
X = -----------
Rf

where ρf is the fluid density, f is the frequency, and Rf is the flow resistivity. Several
predefined sets of the coefficients Ci exist. They are the classic Delany-Bazley model,
the Miki model (see Ref. 9, section 2.5 and Ref. 22), the Qunli, several variants of the
Mechel model for different configurations, the Komatsu model, and a so-called
Modified Champoux and Allard model (see Ref. 23). These are all empirical models
based on fitting the two complex functions to measured data for the complex wave
number kc and complex specific acoustic impedance Zc. All the models are applicable
for materials with a porosity εp close to 1. The applicability of the different model
parameters is listed in Table 2-10 and also Ref. 23 for further details.
TABLE 2-10: DELANY-BAZLEY-MIKI MODEL OPTIONS AND APPLICABILITY

MODEL NAME APPLICABILITY

Delany-Bazley Glass and rock wool with:

0.01 ≤ X ≤ 1
3 3 –2
10 ≤ R f ≤ 50 ⋅ 10 Pasm

Miki Glass and rock wool with:

0.01 ≤ X ≤ 1
3 3 –2
10 ≤ R f ≤ 50 ⋅ 10 Pasm

The validity of the model using the Miki

parameters is not well-established for X< 0.01,
but the model is slightly better behaved
mathematically below this limit using the Miki
parameters rather than the Delany-Bazley
parameters, see Ref. 22.
Qunli Porous plastic and open foams:

200 ≤ f ≤ 2000 Hz
3 3 –2
3 ⋅ 10 ≤ R f ≤ 24 ⋅ 10 Pasm

298 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

TABLE 2-10: DELANY-BAZLEY-MIKI MODEL OPTIONS AND APPLICABILITY

MODEL NAME APPLICABILITY

Mechel, glass fiber, low X Glass fiber: X ≤ 0.025

Mechel, glass fiber, high X Glass fiber: X ≥ 0.025
Mechel, rock fiber, low X Rock fiber: X ≤ 0.025
Mechel, rock fiber, high X Rock fiber: X ≥ 0.025
Komatsu Glass and rock wool:

3 3 –2
6 ⋅ 10 ≤ R f ≤ 73 ⋅ 10 Pasm

Modified Champoux and Allard 3

45 ≤ f ≤ 11 ⋅ 10 Hz

ZWIKKER-KOSTEN
Zwikker-Kosten is one of the earliest equivalent fluid models for porous materials
(Ref. 15). It is a rigid frame model defined by the complex (rigid) density

2
ρf 1 ωρ f H r Hr
ρ rig = ----- ----------------------------------------------------------- Wo = ----------------- = 2 ------
εP μ δv
2 J 1 ( Wo – i )
1 – ------------------ ------------------------------
Wo – i J 0 ( Wo – i )

where Hr is the hydraulic radius of the pores (for straight cylindrical pores
Hr = a = radius) and Wo is the Womersley number (see Equation 2-51 and the note
below). Wo is related to the ratio between viscous penetration depth δv and the
hydraulic radius. δv gives the scale of the viscous boundary layer thickness (see Theory
Background for the Thermoviscous Acoustics Branch for details). The bulk modulus
is given by

γp A 1 Cp μ
K = --------- ---------------------------------------------------------------------------------------- Pr = ----------
εP J 1 ( Wo – iPr ) k
2
1 + ( γ – 1 ) ------------------------ ------------------------------------
Wo – iPr J 0 ( Wo – iPr )

where pA denotes the ambient pressure, ρf the fluid density, γ the ratio of specific heat,
Pr the Prandtl number, µ the dynamic viscosity, Cp the heat capacity at constant
pressure, and k the coefficient of thermal conduction. J0 and J1 are Bessel functions
of the first kind. The factor γ pA is the isentropic bulk modulus (K0 = γ pA). The free
parameters of the pores are the porosity εP and the hydraulic radius Hr.

THEORY FOR THE EQUIVALENT FLUID MODELS | 299

 ATTENBOROUGH
The Attenborough model is also based on the cylindrical-like pore assumption. It is a
so-called four parameter semi-empirical model. The model is an extension of the
Zwikker-Kosten model and adds two more input parameters. It accounts for the
tortuosity (high frequency limit) τ ∞ , which is related to the orientation of the pores
relative to the propagation direction. The hydraulic diameter of the pores is replaced
by an expression that includes the flow resistivity Rf, and a fitting parameter b, (this
parameter is related to the anisotropy of the pores). See Ref. 9 and Ref. 16. The
equivalent density and bulk modulus are defined as

ρf τ∞
ρ rig = ----- ------------------------------------------------
εP
2 J 1 ( s' – i )
1 – ------------- -------------------------
s' – i J 0 ( s' – i )

and

γp A 1 Cp μ
K = --------- -------------------------------------------------------------------------------------- Pr = ----------
εP J 1 ( s' Pr – i ) k
2
1 + ( γ – 1 ) ----------------------- -----------------------------------
s' Pr – i J 0 ( s' Pr – i )

where pA denotes the ambient pressure, ρf the fluid density, γ the ratio of specific heat,
Pr the Prandtl number, µ the dynamic viscosity, Cp the heat capacity at constant
pressure, and k the coefficient of thermal conduction. J0 and J1 are Bessel functions
of the first kind. The variable s' (anisotropy factor) is derived from other material
parameters and is related to the Womersley number (see Equation 2-51):

8ωρ f τ ∞
s' = b ------------------
-
εP Rf

Here, ω denotes the angular frequency. The four parameters needed (when the fluid is
air at room temperature) are the porosity εP, the tortuosity τ∞, flow resistivity Rf, and
the fitting parameter b (dimensionless, close to 1). The fitting parameter b is tabulated
for certain well-defined pore cross-sections in Table 2-11.
TABLE 2-11: FITTING FACTOR b FOR DIFFERENT CROSS-SECTIONAL GEOMETRIES (SEE Ref. 9)
CROSS-SECTIONAL SHAPE b
Circle 1
Square 1.07

300 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

TABLE 2-11: FITTING FACTOR b FOR DIFFERENT CROSS-SECTIONAL GEOMETRIES (SEE Ref. 9)
CROSS-SECTIONAL SHAPE b
Equilateral triangle 1.11
Rectangular slit 0.81
In this way, the hydraulic radius of the Attenborough model is formulated in terms of
measurable intrinsic properties of the porous material which alleviates the need to
know the pore radii.

The viscous characteristic length of the model Lv can also be defined by

8μτ δv
μ - = s' ------
L v = b ------------∞- = s' -------- -
εP Rf ωρ f 2

with δv the viscous penetration depth. This length is related to the pores’ circular cross
section radius R (for a cylinder) and the thickness of the viscous boundary layer. The
parameter b = 1/s, where s is the viscous characteristic length parameter. See the
following models:

• Johnson-Champoux-Allard (JCA)
• Johnson-Champoux-Allard-Lafarge (JCAL)
• Johnson-Champoux-Allard-Pride-Lafarge (JCAPL)).

Note that the tortuosity is related to the angle θ between the cylindrical pores and the
direction of propagation of the wave, by

1 -
τ ∞ = ------------------
2
cos ( θ )

WILSON
The Wilson model is a generalization of the analytical models for porous materials with
constant cross section and parallel pores. This model is intended to match the middle
frequency behavior of a porous material (see Ref. 9, Ref. 17, and Ref. 18). It is not a
good model for ω tending to 0 or infinity. The equivalent density and bulk modulus
are given by

THEORY FOR THE EQUIVALENT FLUID MODELS | 301

 1⁄2
( 1 + iωτ vor )
ρ rig = ρ ∞ -----------------------------------------------
1⁄2
-
( 1 + iωτ vor ) –1
1⁄2
( 1 + iωτ ent )
K = K ∞ --------------------------------------------------------
1⁄2
-
( 1 + iωτ ent ) +γ–1

where τvor denotes the vorticity-mode relaxation time, τent the entropy-mode
relaxation time, ρ∞ the infinity frequency limit for the density, K∞ the infinity frequency
limit for the bulk modulus, and γ is the ratio of specific heats. These are the four free
parameters. With appropriate choices for the relaxation parameters, the Wilson model
can be fitted to mimic all the models described here. For example, setting τvor = 2.54/
Rf and τent = 3.75/Rf, the equations mimic the Delany-Bazley model (see Cox and
D’Antonio Sec. 5.4.4, Ref. 18).

Approximate expressions based on nonacoustic parameters (properties of the porous

matrix) also exist for the relaxation times:

2
ρf l 2ρ f τ ∞
τ vor = ---------- ≈ --------------- τ ent ≈ Prτ vor
2μ εp Rf
ρf τ∞ γp A
ρ ∞ ≈ ----------- K ∞ ≈ ---------
εp εp

Here τ∞ denotes the (high frequency limit) tortuosity (it is called q2 in the Wilson’s
paper Ref. 17), εp the porosity, ρf the fluid density, l a characteristic pore dimension,
and Pr is the Prandtl number.

JOHNSON-CHAMPOUX-ALLARD (JCA)
The Johnson-Champoux-Allard (JCA) porous matrix model is defined by the
following equivalent rigid densities ρrig(ω) and equivalent bulk modulus K(ω):

2
τ∞ ρf Rf εp 4iωτ ∞ μρ f
ρ rig = ----------- 1 + ------------------ 1 + -------------------------
εp iωρ f τ ∞ R L ε
2 2 2
f v p
–1 –1
γp A  iωL th Prρ f
2
K = --------- γ – ( γ – 1 )  1 + -------------------------
8μ -
1 + -------------------------
-
εp 2
iωL th Prρ f 16μ 
 

Here τ∞ is the tortuosity factor (high frequency limit), ρf is the fluid density, εp is the
porosity, Rf is the flow resistivity, μ is the dynamic viscosity, pA is the quiescent

302 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

pressure, γ is the ratio of specific heats, Lv is the viscous characteristic length, Lth is the
thermal characteristic length, and Pr is the Prandtl number. The viscous characteristic
length is related to the viscous characteristic length parameter s by

1 8μτ
L v = --- ------------∞-
s εp Rf

Here s is a pore geometry dependent factor between 0.3 and 3.0 (for example 1 for
circular pores, 0.78 for slits)

The expression given for the geometry dependent pore factor s is only
valid for values of s close to 1. If this is not the case, enter the viscous
characteristic length Lv directly into the model (the default selection).

The viscous Lv and thermal Lth characteristic lengths are also sometimes
denoted by Λ and Λ’, respectively.

Porous Absorber: Application Library path Acoustics_Module/

Building_and_Room_Acoustics/porous_absorber

JOHNSON-CHAMPOUX-ALLARD-LAFARGE (JCAL)
The Johnson-Champoux-Allard-Lafarge (JCAL) model introduces corrections to the
bulk modulus thermal behavior at low frequencies that is not captured by the JCA
model (see Ref. 13). The equivalent density is the same as in the JCA model. The
correction is to the bulk modulus and is given by

–1 –1
γp A  εp μ 4iω ( k' 0 ) Prρ f
2
K = --------- γ – ( γ – 1 )  1 + ------------------------- 1 + ------------------------------------
εp  iωk' 0 Prρ f 2 2 
 μL th ε p 

with the introduction of the new parameter, k’0, which is the static thermal
permeability (SI unit: m2). For measurements and details of this parameter see Ref. 11
and Ref. 12 for examples.

JOHNSON-CHAMPOUX-ALLARD-PRIDE-LAFARGE (JCAPL)
The Johnson-Champoux-Allard-Pride-Lafarge (JCAPL) model further extends the
JCAL models by introducing a static viscous τ0 and thermal τ'0 tortuosity, which both

THEORY FOR THE EQUIVALENT FLUID MODELS | 303

 introduce low frequency corrections to the JCAL and JCA models. See Ref. 14 and
Ref. 9.

This model has the complex rigid density given by:

ρ f τ̃ ( ω )
ρ rig = -----------------
εp

1 ˜ ˜ M
τ̃ ( ω ) = τ ∞ 1 + ------ F ( ω ) F ( ω ) = 1 – P + P 1 + ---------2- iω
iω 2P
iωρ f k 0 τ ∞ 8k 0 τ ∞
iω = ------------------------ M = ---------------
-
με p εp Lv
2

2
M 2k 0 τ ∞
P = --------------------------------- = ----------------------------------
-
4 ( τ0 ⁄ τ∞ – 1 ) 2
εp Lv ( τ0 – τ∞ )

where the new parameter is the static viscous tortuosity τ0 (dimensionless). The viscous
permeability is defined as k0 = μ/Rf (SI unit: m2).

In Ref. 9 (equation 5.32), P is called b, k0 is called q0, and Pr is called B2.

The complex bulk modulus K is given by:

γp A 1
K = --------- ------------
ε p β̃ ( ω )
1 ˜ –1
˜ ( ω )–1
β̃ ( ω ) = γ – ( γ – 1 ) 1 + ------- F' ( ω ) = γ – ( γ – 1 )τ'
iω'
˜ M' iωρ f Prk' 0
F' ( ω ) = 1 – P' + P' 1 + ----------2- iω' iω' = -------------------------
2P' με p
8k' 0 M'
M' = -------------
- P' = ----------------------
ε p L th
2 4 ( τ' 0 – 1 )

where the new parameter is the static thermal tortuosity τ'0 (dimensionless).

˜ ( ω ) is called α '(ω), and P’ = 1.

In Ref. 9 (equation 5.35), τ'

304 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 The JCA model is recovered by setting M' = P = P' = 1 and the JCAL is
recovered by setting P = P' = 1.

WILLIAMS EDFM (EQUIVALENT DENSITY FLUID MODEL)

The Williams EDFM model is a so-called equivalent density fluid model. The model is
derived under the assumption that the bulk and shear moduli of the frame of the
porous material are negligible, see Ref. 24 for further details. This is why the model
pertains to sediments. The effective bulk modulus Keff and an effective density ρeff
defined my the model are given by

1 – εp εp –1
K eff =  -------------- + ------
 K gr K f

 iε p ρ mix F ( Wo )μ
 τ ∞ ( 1 – ε p )ρ gr + ε p ( τ ∞ – 1 )ρ f + ---------------------------------------- ρ f ωκ
-
ρ eff ( ω ) = ρ f  ------------------------------------------------------------------------------------------------------------------------
 iε p F ( Wo )μ 
 ε p ( 1 – ε p )ρ gr + ( τ ∞ – 2ε p + ε 2p )ρ f + ----------------------------- -
 ωκ 
ρ mix = ε p ρ f + ( 1 – ε p )ρ gr

where the subscript “gr” pertains to the grains and the subscript “f” to the saturating
fluid. The porosity is denoted εp, the observed mixture density ρmix, the tortuosity
τ ∞ , the dynamic viscosity μ, the angular frequency ω, and the permeability of the
sediments κ. The function F is a function of the Womersley number Wo defined as

Wo
--------- T ( Wo ) iJ 1 ( iWo )
4
F ( Wo ) = ------------------------------------- T ( Wo ) = – ---------------------------------
2i J 0 ( iWo )
1 – --------- T ( Wo )
Wo
2
ρ f ωH r
Wo = ----------------
-
μ

where Jn(x) is the Bessel function of the first kind of order n and Hr is the hydraulic
radius. Based on these the effective complex speed of sound and density are given as

K eff
cc = ----------------- ρ c = ρ eff ( ω )
ρ eff ( ω )

THEORY FOR THE EQUIVALENT FLUID MODELS | 305

 WOOD (FLUID SUSPENSION MODEL)
In a fluid mixture or a fluid suspension (solid inclusions completely surrounded by
fluid), the Wood formula can be used to determine the effective speed of sound for the
mixture. It is determined by calculating the effective bulk modulus of the suspension
and the volume average density. As the Williams EDFM, this model gives effective
values for the mixture. This result is exact for low frequencies (when the wavelength is
much larger than the size of the inclusions) since the effective bulk modulus in the
quasistatic limit. The Wood model defines

N
1- θf θi
--------
K eff
= ------ +
Kf  -----
Ki
i
N

ρ eff = θ f ρ f +  θi ρi
i

where θf, Kf, and ρf, are the fluid’s volume fraction, adiabatic bulk modulus, and
density, respectively; and θi, Ki, and ρi, are the inclusion’s volume fractions, adiabatic
bulk moduli, and densities, respectively. And again the complex speed of sound and
density are defined as

K eff
cc = --------- ρ c = ρ eff
ρ eff

About the Narrow Region Acoustics Models

The Narrow Region Acoustics fluid models are used to mimic the thermal and viscous
losses that exist in narrow tubes where the tube cross-section length-scale is
comparable to the thermal and viscous boundary layer thickness (boundary-layer
absorption). It is essential to include these losses in order to get correct results.

These models are commonly used in situations where solving a full detailed
thermoviscous acoustic model is computationally costly; for example when analyzing
long narrow ducts/tubes of constant cross section. Here it is possible to add or smear
the losses associated with the boundary layer onto the bulk of the fluid: an equivalent
fluid model. For many geometries, analytical expressions exist for the losses associated

306 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

with the acoustic boundary layers. The models can be applied under different
assumptions. The models and assumptions are discussed in this section.

In more complex geometries where thermal and viscous losses are

important, see The Thermoviscous Acoustics, Frequency Domain
Interface, which is more fundamental and detailed.

Narrow Region Acoustics

Several fluid models exist:

• The wide duct approximation can be used for any duct cross section in the limit
where the duct width is significantly larger than the acoustic boundary layer
thickness. See Wide Ducts.
• The very narrow circular ducts (isothermal) can only be used when the duct width
is so small that isothermal conditions apply. This is when the duct width is much
smaller than the acoustic thermal boundary layer thickness. See Very Narrow
Circular Ducts (Isothermal).
• The slit, circular duct, rectangular duct, and equilateral triangular duct models
are based on an analytical solution of the thermoviscous acoustic equations in the
limit where the acoustic wavelength is much larger than both the duct cross section
(below the cutoff frequency) and the boundary layer thickness. This is the case in
most engineering applications. See Slits, Circular Ducts, Rectangular Ducts, and
Equilateral triangular Ducts.
• Finally, selecting the user defined option enables you to enter expressions for the
complex wave number and the complex acoustic impedance. These may be
analytical expressions, interpolated values, or values extracted from a detailed
boundary mode analysis using the full formulation of The Thermoviscous Acoustics,
Boundary Mode Interface.

WIDE DUCTS
For a relatively wide duct, the losses introduced in the acoustic boundary layer may be
studied by adding these as an effective wall shear force. This approach is used in
Blackstock (Ref. 6) and results in equivalent fluid complex wave number kc defined by

THEORY FOR THE EQUIVALENT FLUID MODELS | 307

 ω ω B π
k c = ---- ----------------------------- ≅ ----  1 + ---- -----
1
-
c c  2 iω
π
1 – B ------
iω (2-55)
4 μ γ–1 μC p
1 + -----------
S
B = ------- ------ Pr = ---------- H d = 4 ----
H d πρ  Pr  k C

where Hd is the hydraulic diameter of the duct, S is the duct cross-section area, C is
the duct circumference, μ is the dynamic viscosity, ρ is the fluid density, γ is the ratio
of specific heats, Cp is the specific heat at constant pressure, k is the fluid thermal
conductivity, and Pr is the Prandtl number. For a cylindrical duct, Hd = 2a where a is
the radius.

The approximation in Equation 2-55 is only valid for systems where the effective
radius Hd/2 is larger than the boundary layer, but not so small that mainstream
thermal and viscous losses are important. Thus requiring

Hd c2 1 2μ-
δ v < ------- < ------2 ----- δv = ------
2 ω δv ωρ

where δv is the characteristic thickness of the viscous boundary layer (the viscous
penetration depth), c is the speed of sound, and ω is the angular frequency. The
complex wave number is related to the complex density and speed of sound by the
equation (assuming a real valued bulk modulus)

ω c 2
c c = ----- ρ c = ρ  ----
kc cc

VERY NARROW CIRCULAR DUCTS (ISOTHERMAL)

In the other limit where the duct diameter is sufficiently small or the frequency
sufficiently low, the thermal boundary layer thickness becomes much larger than the
duct cross section a. This is the case when

2
ωρa C p
a « δ th  --------------------- « 1
2k

where δth is the characteristic thickness of the thermal boundary layer (thermal
penetration depth), ρ is the density, Cp is the heat capacity at constant pressure, and k
is the fluid thermal conductivity. In this case see Pierce (Ref. 5); the system may be seen
as isothermal and the acoustic temperature variation is zero everywhere in the duct
T = 0. The fluid complex wave number kc is then defined by

308 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 1 1
--- ---
ω 4μ 2  4μω - 2
k c = ------  -------------2- – i  ---------------
c
 c T = ------ (2-56)
c T  ρωa   ρc 2 a 2 γ
T

where cT is the isothermal speed of sound, a is the duct radius, µ is the dynamic
viscosity, and ω is the angular frequency. The theory is derived for ducts of circular
cross section — the model is therefore only applicable for systems with small variations
away from a circular cross section. The complex wave number is related to the complex
density and speed of sound by the equation (here the bulk modulus is defined in the
isothermal limit)

ω cT 2
c c = ----- ρ c = ρ  ------
kc cc

SLITS, CIRCULAR DUCTS, RECTANGULAR DUCTS, AND EQUILATERAL

TRIANGULAR DUCTS
The slit, circular duct, rectangular duct, and equilateral triangular duct models are
based on the so-called low reduced frequency (LRF) model that describes the
propagation of acoustic waves in small waveguides (ducts and slits) including thermal
and viscous losses. Details about these models are in Ref. 19, Ref. 20, and Ref. 21. The
models cover the range from fully isothermal conditions (very low frequencies or very
narrow tubes) to large ducts where the boundary layer only represents a fraction of the
duct size. The models apply as long as the cross section of the duct is much smaller
than the acoustic wavelength (the model is below the cutoff frequency) and when the
cross section is constant or only very slowly varying.

In a narrow waveguide the complex wave number, kc, and complex specific acoustic
impedance, Zc, are given by

2 γ – ( γ – 1 )ϒ th ω
k c = k 0  ---------------------------------
2
k 0 = ----
 ϒv  c
2
2 Z0
Z c = -------------------------------------------- Z 0 = ρc
ϒ v ( γ – ( γ – 1 )ϒ th )

where Υv and Υth are the mean value (cross section averaged) of the scalar viscous and
thermal field functions, respectively. The functions are geometry and
material-dependent functions (specified below) and γ is the ratio of specific heats. The
fluid density ρ, the speed of sound is c, and the angular frequency ω define the free
space wave number k0 and the specific acoustic impedance Z0. The subscripts “v” and

THEORY FOR THE EQUIVALENT FLUID MODELS | 309

 “th” stand for viscous and thermal (heat) fields, respectively. Once these are known,
the complex speed of sound and complex density are given by

ω
c c = -----
kc
kc Zc k0 Z0 1 ρ
ρ c = ------------ = ------------- ------ = ------
ω ω ϒv ϒv

The values of the Υj functions can be derived by solving the full set of linearized
Navier-Stokes equations (the equations solved by the thermoviscous acoustics
interfaces, see Theory Background for the Thermoviscous Acoustics Branch) by
splitting these into an isentropic (adiabatic), a viscous, and a thermal part. Doing this
introduces the viscous and the thermal wave numbers for the system

2 ρ 2 ρC p
k v = – iω --- k th = – iω ----------
μ k

The equations may be solved analytically under the following assumptions used for the
LRF models:

• The viscous and thermal wave numbers must be much larger than the acoustic wave
number (the acoustic wavelength must be much larger than the boundary layer
thickness):
k
------j » 1  λ0 » δj
k0

• The cross section of the waveguide must be much smaller than the acoustic
wavelength (the model is below the cutoff frequency).
• The cross section of the waveguide must be constant or at most slowly varying in
the propagation direction.
• The length of the waveguide in the propagation direction should be larger than the
boundary layer thickness.

The resulting analytical expressions, for the viscous and thermal Υ functions, are for
the given geometry (these results are reviewed in Ref. 19):

• Slit of height h
tan ( k j h ⁄ 2 )
ϒ j = 1 – -------------------------------
kj h ⁄ 2

• Circular duct of radius a

310 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

 J2 ( kj a )
ϒ i = – --------------------
J0 ( kj a )

• Rectangular duct of side lengths W and H

∞
tan ( α m W ⁄ 2 ) tan ( β m H ⁄ 2 )
 ( α m m' )  1 – ------------------------------------ + ( β m m' )  1 – -----------------------------------
2 –2 –2
ϒi = kj
 αm W ⁄ 2   βm H ⁄ 2 
m=0
m' = ( m + 1 ⁄ 2 )π

k j –  ---------- k j –  ----------
2 2m' 2 2m'
αm = βm =
W H

• Equilateral triangular duct of side length d

tan  --- ( k j l ) – --- ( k j l )

3 3
2 2 d
ϒ i = 1 – 3 -------------------------------------------------------
2
- l = -----------
 --- ( k l ) tan  --- ( k l )
3 3 2 3
2 j  2 j 

User Defined
The user defined option in the Narrow Region Acoustics domain feature can be used
to define a LRF model for a waveguide of an arbitrary cross sections. Enter values for
the complex wave number kc and the characteristic complex impedance Zc. This can
be a user defined analytical expression or values derived from a mode analysis study.

To determine the complex propagation constants for a waveguide, of

arbitrary cross section, use The Thermoviscous Acoustics, Boundary
Mode Interface. Apply it on the cross-section geometry of the waveguide.
The interface solves for the propagating modes and includes all losses in
detail. The complex wave number kc is then given by the plane wave
mode solved for. This is the variable [Link]. The predefined variable
[Link] gives the (lumped) complex characteristic impedance Zc. Search
for the mode nearest to the (lossless) plane wave mode.

THEORY FOR THE EQUIVALENT FLUID MODELS | 311

 Theory for the Perfectly Matched
Layers in the Time Domain
In this section:

• Introduction to Perfectly Matched Layers

• Perfectly Matched Layers in the Time Domain

Introduction to Perfectly Matched Layers

The concept of a perfectly matched layer (PML) as an absorbing boundary condition
was introduced by Bérenger (Ref. 38) with regard to the system of Maxwell’s
equations. The PML provided absorption of propagating waves without introducing
reflections from the interface between the PML and the physical domain. This made
the PML technique attractive for treating open domain problems for acoustic, elastic,
and electromagnetic wave propagation.

Bérenger’s PML formulation is usually referred to as the split PML, because the
unknowns are split into a sum of nonphysical components in PML domains. Another
formulation that does not require splitting the variables is based on the coordinate
transformation in PML domains, where the real coordinate is mapped onto the
complex plane:

x˜ = f ( x ) ∈ C . (2-57)

For example, the mapping Equation 2-57 will transform the 1D Helmholtz equation
as follows:

2
( iω ) - 1 ∂p
p – ----------- ∂  -----------  = 0 .
------------- 1
(2-58)
c
2 f' ( x ) ∂ x f' ( x) ∂x

Infinite Elements, Perfectly Matched Layers, and Absorbing Layers in the

COMSOL Multiphysics Reference Manual

312 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

Perfectly Matched Layers in the Time Domain
For the sake of brevity, consider the 1D Equation 2-58. In order to derive the PML
formulation in the time domain, the following steps are taken (Ref. 39). First, consider
a special form of the mapping Equation 2-57:

x
x˜ = x + ------
1
iω x σ ( ξ ) dξ ,
0
(2-59)

which yields dx˜ = f' ( x )dx = ( 1 + σ ( x ) ⁄ iω )dx .

Then, taking Equation 2-59 into account, multiply Equation 2-58 by 1+σ(x)/iω.
Equation 2-58 transforms to the following form:

 ( iω ) 2 iωσ  
1 - ∂p
- + ---------- p – ∂  ---------------
 ------------- = 0. (2-60)
 c2 c 
2 ∂ x  1 + -----σ- ∂ x
 
iω

The transformation to the time domain is performed according to the rule

iωp → ∂p ⁄ ∂t . Its direct application to Equation 2-60 would result in a time integral
of p. To avoid this, an auxiliary variable u is introduced:

1 - ∂p ∂p ---------------
σ - ∂p ∂p
--------------- = – = +u. (2-61)
σ ∂ x ∂ x σ + iω ∂ x ∂x
1 + ------
iω

Equation 2-60 and Equation 2-61 yield a system of partial differential equations in the
time domain equivalent to the frequency domain Equation 2-58:

2
1-  ∂ p ∂p ∂ ∂p
----  2 + σ  –  u +  = 0
2
c ∂t  ∂ t  ∂ x ∂x
(2-62)
∂u ∂p
+ σ  u +  = 0.
∂t  ∂ x

The derivation of the model in 3D space, where more auxiliary variables are required,
is given in Ref. 39.

In the axisymmetric cases, the coordinate stretching in PML domains is expressed in

curvilinear coordinates. A thorough study of the use of PMLs for such problems in the
frequency and the time domains is given in Ref. 40.

THEORY FOR THE PERFECTLY MATCHED LAYERS IN THE TIME DOMAIN | 313
 References for the Pressure Acoustics
Branch
1. D. Givoli and B. Neta, “High-order Non-reflecting Boundary Scheme for
Time-dependent Waves,” J. Comput. Phys., vol. 186, pp. 24–46, 2004.

2. A. Bayliss, M. Gunzburger, and E. Turkel, “Boundary Conditions for the Numerical

Solution of Elliptic Equations in Exterior Regions,” SIAM J. Appl. Math., vol. 42,
no. 2, pp. 430–451, 1982.

3. A.B. Bauer, “Impedance Theory and Measurements on Porous Acoustic Liners,”

J. Aircr., vol. 14, pp. 720–728, 1977.

4. S. Temkin, Elements of Acoustics, Acoustical Society of America, 2001.

5. A.D. Pierce, Acoustics: An Introduction to its Physical Principles and

Applications, Acoustical Society of America (second print), 1991.

6. D.T. Blackstock, Fundamentals of Physical Acoustics, John Wiley & Sons, 2000.

7. P.M. Morse and K.U. Ignard, Theoretical Acoustics, Princeton University Press,
1986.

8. L.D. Landau and E.M. Lifshitz, Fluid Mechanics, Course of Theoretical Physics,
Volume 6, Butterworth-Heinemann, 2003.

9. J.F. Allard and N. Atalla, Propagation of Sound in Porous Media, Modeling Sound
Absorbing Materials, 2nd Edition, John Wiley & Sons, 2009.

10. R. Panneton, “Comment on the Limp Frame Equivalent Fluid Model for Porous
Media,” J. Acoust. Soc. Am., vol. 122, no. 6, pp. EL217–EL222, 2007.

11. M. Sadouki, M. Fellah, Z.E. Fellah, E. Ogam, N. Sebaa, F.G. Mitri, and
C. Depollier, “Measuring static thermal permeability and inertial factor of rigid porous
materials,” J. Acoust. Soc. Am., vol. 130, p. 2627, 2011.

12. X. Olny and R. Panneton, “Acoustical determination of the parameters governing

thermal dissipation in porous media,” J. Acoust. Soc. Am., vol. 123, 814–824, 2008.

13. D. Lafarge, P. Lemarinier, J.-F. Allard, and V. Tarnow, “Dynamic compressibility

of air in porous structures at audible frequencies,” J. Acoust. Soc. Am., vol. 102, no. 4,
pp. 1994–2006, 1997.

314 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

14. S.R. Pride, F.D. Morgan, and A.F. Gangi, “Drag forces of porous-medium
acoustics,” Phys. Rev. B, vol. 47, pp. 4964–4978, 1993.

15. C. Zwikker and C. W. Kosten, Sound Absorbing Materials, Elsevier Publishing,

New York, 1949.

16. K. Attenborough, “On the acoustic slow wave in air filled granular media,” J.
Acoust. Soc. Am., vol. 81, no. 1, pp. 93–102, 1987.

17. K. Wilson, “Relaxation-matched modeling of propagation through porous media,

including fractal pore structure,” J. Acoust. Soc. Am., vol. 94, no. 2, pp. 1136–1145,
1993.

18. T.J. Cox and P. D’Antonio, Acoustic Absorbers and Diffusers, Taylor and
Francis, 2nd ed., 2009. [Link]

19. R. Kampinga, Viscothermal Acoustics Using Finite Elements, Analysis Tools for
Engineers, PhD thesis, University of Tweente, The Netherlands, 2010.

20. M.R. Stinson, “The propagation of plane sound waves in narrow and wide circular
tubes, and generalization to uniform tubes of arbitrary cross-sectional shapes,” J.
Acoust. Soc. Am., vol. 89, pp. 550, 1990.

21. H. Tijdeman, “On the propagation of sound waves in cylindrical tubes,” J. Sound
Vib., vol. 39, pp 1, 1975.

22. Y. Miki, “Acoustical properties of porous materials - modifications of

Delany-Bazley models,” J. Acoust. Soc. Jpn (E), vol. 11, no. 1, 1990.

23. D. Oliva and V. Hongisto, “Sound absorption of porous materials – Accuracy of

prediction methods,” Appl. Acoust., vol. 74, issue 12, pp. 1473–1479, 2013.

24. K. L Williams, “An effective density fluid model for acoustic propagation in
sediments derived from Biot theory,” J. Acoust. Soc. Am., vol. 110, pp. 2276, 2001.

25. A. B. Wood, A Textbook of Sound, The Macmillan company, New York, 1941.

26. W. M. Leach, Jr, Electroacoustics & Audio Amplifier Design, Kendall Hunt
publishing company, 2010.

27. B. Håkansson, P. Carlsson, and A. Tjellström, “The mechanical point impedance

of the human head, with and without skin penetration”, J. Acoust. Soc. Am., vol. 80,
issue 4, pp.1065–1075 (1986).

REFERENCES FOR THE PRESSURE ACOUSTICS BRANCH | 315

 28. H. Hudde and A. Engel, “Measuring and modeling basic properties of the human
middle ear and ear canal. Part I: Model structure and measure techniques”,
ACOUSTICA acta acoustica, vol. 84, pp. 720–738 (1998).

29. H. Hudde and A. Engel, “Measuring and modeling basic properties of the human
middle ear and ear canal. Part II: Ear canal, middle ear cavities, eardrum, and ossicles”,
ACOUSTICA acta acoustica, vol 84, pp. 894–913 (1998).

30. H. Hudde and A. Engel, “Measuring and modeling basic properties of the human
middle ear and ear canal. Part III: Eardrum impedances, transfer functions and model
calculations”, ACOUSTICA acta acoustica, vol. 84, pp. 1091–1109 (1998).

31. D.H. Keefe, “Acoustical wave propagation in cylindrical ducts: Transmission line
parameter approximations for isothermal and nonisothermal boundary conditions”, J.
Acoust. Soc. Am., vol. 75, issue 1, pp. 58–62 (1984).

32. H. Levine and J. Schwinger, “On the radiation of sound from an unflanged circular
pipe”, Phys. Rev., vol. 73, no. 4, pp. 383–406 (1948).

33. R. Courant, K.O. Friedrichs, and H. Lewy, “On the Partial Difference Equations
of Mathematical Physics,” IBM Journal, vol. 11, pp. 215–234, 1956.

34. Y. Saad, “A Flexible Inner-outer Preconditioned GMRES Algorithm,” SIAM J.

Sci. Statist. Comput., vol. 14, pp. 461–469, 1993.

35. M.R. Stinson and E.A.G. Shaw, “Acoustic impedance of small, circular orifices in
thin plates,” J. Acoust. Soc. Am., vol. 77, issue 6, pp. 2039–2042, 1985.

36. M.A. Temiz, J. Tournadre, and I.L. Arteaga, “Non-linear acoustic transfer
impedance of micro-perforated plates with circular orifices,” J. Sound and Vibration,
vol. 366, pp. 418–428, 2016.

37. T. Elnady, “Modeling and characterization of perforates in lined ducts and

mufflers,” PhD Thesis, The Royal Institute of Technology (KTH), 2004.

38. J.-P. Bérenger, “A perfectly matched layer for the absorption of electromagnetic
waves,” J. Comp. Phys., vol. 114, issue 2, pp. 185–200, 1994.

39. B. Kaltenbacher, M. Kaltenbacher, and I. Sim, “A modified and stable version of a

perfectly matched layer technique for the 3-d second order wave equation in time
domain with an application to aeroacoustics,” J. Comp. Phys., vol. 235, pp. 407–422,
2013.

40. F. Collino and P. Monk, “The perfectly matched layer in curvilinear coordinates,”
SIAM J. Sci. Comput., vol. 19, number 6, pp. 2061–2090, 1998.

316 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

41. M. Kaltenbacher, Numerical Simulation of Mechatronic Sensors and Actuators,
3rd edn., Springer: Berlin, 2015.

42. M.F. Hamilton and D.T. Blackstock, eds., Nonlinear Acoustics, Academic Press,
San Diego, CA, 1998.

43. H.E. Bass, L.C. Sutherland, A.J. Zuckerwar, D.T. Blackstock, and D.M. Hester,
“Atmospheric absorption of sound: Further developments,” J. Acoust. Soc. Am.,
vol. 97, pp. 680-683, 1995; “Erratum,” J. Acoust. Soc. Am., vol. 99, p. 1259. 1996.

44. ANSI S1.26-2014 (supersedes ANSI S1.26-1995) “American National Standard

method for calculation of the absorption of sound by the atmosphere” (Acoustical
Society of America, New York, 2014).

45. National Physics Lab (NPL), “NPL Acoustics: Calculation of absorption of sound
by the atmosphere,” [Link]

46. R.E. Francois and G.R. Garrison, “Sound absorption based on ocean
measurements: Part I: Pure water and magnesium sulfate contributions,” J. Acoust.
Soc. Am., vol. 72, pp. 896–907, 1982.

47. R.E. Francois and G.R. Garrison, “Sound absorption based on ocean
measurements: Part II: Boric acid contribution and equation for total absorption,” J.
Acoust. Soc. Am., vol. 72, pp. 1879–1890, 1982.

48. M.A. Ainslie and J.G. McColm, “A simplified formula for viscous and chemical
absorption in sea water,” J. Acoust. Soc. Am., vol. 103, pp. 1671–1672, 1998.

49. F. H. Fisher and V.P. Simmons, “Sound absorption in seawater,” J. Acoust. Soc.
Am., vol. 62, pp. 558–564, 1977.

50. National Physics Lab (NPL), “Calculation of absorption of sound in seawater,”

[Link]

51. M.F. Hamilton and D.T. Blackstock, “On the coefficient of nonlinearity in
nonlinear acoustics,” J. Acoust. Soc. Am., vol. 83, pp. 74–77, 1988.

52. V. Nikolic, On Certain Mathematical Aspects of Nonlinear Acoustics:

Well-Posedness, Interface Coupling, and Shape Optimization, PhD thesis,
University of Klagenfurt (2015).

53. M. Berggren, A. Bernland, and D. Noreland, “Acoustic Boundary Layers as

Boundary Condition,” J. Comp. Phys., vol. 371, pp. 633-650, 2018.

REFERENCES FOR THE PRESSURE ACOUSTICS BRANCH | 317

 54. J. S. Bach and H. Bruus, “Theory for Acoustics with Viscous Boundary Layers and
Streaming in Curved Elastic Cavities,” J. Acoust. Soc. Am., vol. 144, pp. 766-784
2018.

55. J. H. Joergensen and H. Bruus, “Theory of pressure acoustics with thermoviscous

boundary layers and streaming in elastic cavities,” J. Acoust. Soc. Am., vol. 149, pp.
3599–3610, 2021.

56. IOC, SCOR and IAPSO, 2010: The international thermodynamics equation of
seawater - 2010: Calculation and use of thermodynamic properties.
Intergovernmental Oceanographic Commission, Manuals and Guides No. 56,
UNESCO (English), 196 pp.

57. Horoshenkov, Hurrell, and Groby, “A three-parameter analytical model for the
acoustical properties of porous media,” J. Acoust. Soc. Am., vol. 145 (4), pp.
2512-2517, 2019

58. Horoshenkov, Groby, and Dazel, “Asymptotic limits of some models for sound
propagation in porous media and the assignment of the pore characteristic lengths,” J.
Acoust. Soc. Am., vol. 139 (5), pp. 2463-2474, 2016.

59. Horoshenkov, Attenborough, and Chandler-Wilde, “Padé approximants for the

acoustical properties of rigid frame porous media with pore size distributions,” J.
Acoust. Soc. Am., vol. 104, pp. 1198-1209, 1998.

60. Horoshenkov, Hurrell, and Groby, “Erratum: A three-parameter analytical model

for the acoustical properties of porous media [J. Acoust. Soc. Am. 145(4), 2512–2517
(2019)],” J. Acoust. Soc. Am., vol 147 (1), p. 146, 2020.

61. Horoshenkov, Groby, and Dazel, “Erratum: Asymptotic limits of some models for
sound propagation in porous media and the assignment of the pore characteristic
lengths [J. Acoust. Soc. Am. 139(5), 2463–2474 (2016)],” J. Acoust. Soc. Am., vol
147 (1), p. 205, 2020.

62. G. B. Deane, “Internal friction and boundary conditions in lossy fluid seabeds,” J.
Acoust. Soc. Am., vol 101, p. 233, 1997.

63. M. J. Lighthill, “On sound generated aerodynamically I. General theory,” Proc.

Roy. Soc. Lond., vol. 211, pp. 564-587, 1952.

64. M. J. Lighthill, “On sound generated aerodynamically II. Turbulence as source of

sound,” Proc. Roy. Soc. Lond., vol. 222, pp. 1-22, 1954.

318 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES

65. M. Kaltenbacher, Numerical Simulation of Mechatronic Sensors and Actuators,
Finite Elements for Computational Multiphysics, Springer, Third Edition, 2015.

66. M. Kaltenbacher (Editor), Computational Acoustics, CISM, Courses and

Lectures Volume 579, Springer, 2018.

67. S. Schroder and M. Kaltenbacher, “Hybrid Aeroacoustic Computations: State of

Art and New Achievements,” J. Theor. Comput. Acoust., vol 27, p. 1950020, 2019.

68. S. Glegg and W. Devenport, Aeroacoustics of Low Mach Number Flows, Elsevier,
Academic Press, 2017.

REFERENCES FOR THE PRESSURE ACOUSTICS BRANCH | 319

320 | CHAPTER 2: PRESSURE ACOUSTICS INTERFACES
 3

Elastic Waves Interfaces

This chapter describes the Acoustics Module background theory and physics
interfaces found under the Elastic Waves branch ( ).

• The Solid Mechanics (Elastic Waves) Interface

• The Poroelastic Waves Interface
• The Elastic Waves, Time Explicit Interface
• The Piezoelectric Waves, Time Explicit Interface
• Modeling with the Elastic Waves Branch
• Theory for the Poroelastic Waves Interfaces
• Theory for the Elastic Waves, Time Explicit Interface

321
 The Solid Mechanics (Elastic Waves)
Interface
The Solid Mechanics (Elastic Waves) interface ( ), found under the Acoustics>Elastic
Waves branch ( ) when adding a physics interface, is a shortcut to add the Solid
Mechanics interface which is used to compute the displacement field in solids with
propagating elastic waves. The Solid Mechanics interface supports studies in many
forms, specifically the dynamic Navier’s equation is solved in the solid in the frequency
domain. Dedicated Multiphysics Couplings exist to couple fluid, solid, and porous
domains.

When this physics interface is added, these default nodes are also added to the Model
Builder — Linear Elastic Material, Free, and Initial Values. For 2D axisymmetric
components an Axial Symmetry node is also added. Features and boundary conditions
are described in the Solid Mechanics documentation.

For information about meshing see Meshing and Solving Wave Problems
Solved with Solid Mechanics in the Modeling with the Elastic Waves
Branch section.

THE PORT CONDITION

The Solid Mechanics interface has functionality and features that are specifically tailored
for elastic wave applications. The Port condition allows to set up conditions at the
inlet/outlet of elastic wave guide structures that are mathematically consistent. Use
the Port to set up non-reflecting conditions, that perform better than a PML, as every
propagating modes are captured explicitly.

See the tutorial Mechanical Multiport System: Elastic Wave

Propagation in a Small Aluminum Plate. Application Library path:
Acoustics_Module/Elastic_Waves/mechanical_multiport_system

SOLVING ELASTIC WAVE PROBLEMS IN THE TIME DOMAIN

When solving elastic wave problems with the Solid Mechanics interface in the time
domain it is important to make changes to the Transient Solver Setting section. Expand
the section and select the check-box Maximum frequency to resolve, then enter the

322 | CHAPTER 3: ELASTIC WAVES INTERFACES

maximum frequency to resolve in the model fmax,sol. The maximum frequency is
dictated by the frequency content of the source as well as the eigenmodes of the
structure that can be excited. Setting this correctly will results in a transient solver
tuned specifically for modeling wave problems.

THE SOLID MECHANICS (ELASTIC WAVES) INTERFACE | 323

 The Poroelastic Waves Interface
The Poroelastic Waves (pelw) interface ( ), found under the Acoustics>Elastic Waves
branch ( ) when adding a physics interface, is used to compute the displacement
field and acoustic pressure fluctuation in porous materials with propagating poroelastic
waves. Dedicated Multiphysics Couplings exist that define the couplings between
fluid, solid, and porous domains.

Examples of applications include the propagation of elastic waves in rocks and soils,
modeling the acoustic attenuation properties of particulate filters, characterizing
sound absorbers and liners, or modeling the porous foams in headphones. The physics
interface is valid for modeling the propagation of the coupled linear elastic and linear
acoustic waves in the frequency domain. Harmonic variations of the displacement field
and the sources are assumed. In the porous domains, Biot’s equations are solved
accounting for the coupled propagation of elastic waves in the elastic porous matrix
and pressure waves in the saturating pore fluid. This includes the damping effect of the
pore fluid due to viscous losses only (the Biot model), typically with a saturating liquid
like water or oil, or the combined effect of viscous and thermal losses (the Biot-Allard
model), typically when the saturating fluid is air.

• For information about modeling strategies and meshing see Meshing

Poroelastic Waves Models and Solving Large Poroelastic Wave Models
in the Modeling with the Elastic Waves Branch section.
• See the Theory for the Poroelastic Waves Interfaces for details about
the governing equations. The specifics of the Biot and the Biot-Allard
models are also discussed here.

When the Poroelastic Waves interface is added, these default nodes are also added to
the Model Builder: Poroelastic Material, Porous, Free, and Initial Values. For 2D
axisymmetric components, an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Poroelastic Waves to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

324 | CHAPTER 3: ELASTIC WAVES INTERFACES

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is pelw.

The rest of the physics interface settings are the same as for The Pressure
Acoustics, Frequency Domain Interface and The Solid Mechanics
Interface.

• Acoustics of a Particulate-Filter-Like System: Application Library

path Acoustics_Module/Automotive/acoustics_particulate_filter
• Acoustic Reflections off a Water-Sediment Interface: Application
Library path Acoustics_Module/Underwater_Acoustics/
reflections_water_sediment
• Headphone on an Artificial Ear: Application Library path
Acoustics_Module/Electroacoustic_Transducers/headphone_artificial_ear

THE POROELASTIC WAVES INTERFACE | 325

 Domain, Boundary, and Pair Nodes for the Poroelastic Waves
Interfaces
The Poroelastic Waves Interface has these domain, boundary, and pair nodes available
from the Physics ribbon toolbar (Windows users), Physics context menu (Mac or Linux
users), or right-click to access the context menu (all users)

• Added Mass1 • Prescribed Acceleration

1
• Boundary Load • Prescribed Displacement
• Fixed Constraint • Prescribed Velocity
1
• Free • Rigid Connector1
• Initial Stress and Strain1 • Roller
• Initial Values • Septum Boundary Load
• Periodic Condition • Spring Foundation1
• Poroelastic Material • Symmetry
• Porous, Free • Thin Elastic Layer1
• Porous, Pressure
1
These features are described for the Solid Mechanics interface in the Structural
Mechanics Module User’s Guide

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Poroelastic Material
Use the Poroelastic Material node to define the poroelastic material and fluid
properties, that is the properties of the porous matrix and the saturating [Link]

326 | CHAPTER 3: ELASTIC WAVES INTERFACES

subnode is available from the context menu (right-click the parent node) or from the
Physics toolbar, Attributes menu.

POROELASTIC MODEL
Select the Model used to describe the losses to include in the porous material:

• Biot (viscous losses), this model is primarily used in earth sciences when the
saturating fluid is a liquid like water or oil. The model is based on Biot’s original
work and only includes the effects of viscous losses in the pores.
• Biot-Allard (thermal and viscous losses), this model is intended for simulating porous
materials where the saturating fluid is air. This is for modeling sound absorbers,
liners, foams used in headphones and loudspeakers, cloth and much more.

When selecting a specific model, the required material input will change in order to
align with the data normally available for the intended applications.

Three different waves can propagate in a poroelastic material; fast pressure

waves, slow pressure waves, and shear waves. Their speed can be evaluated
through the variables pelw.cp_fast, pelw.cp_slow, and
pelw.cs_poro, respectively. These should not be confused with the
material properties for the constituting solid ([Link] and [Link]) and
fluid (pelw.c). All three poroelastic wave types should be resolved by the
computational mesh.

POROUS MATRIX PROPERTIES

The default Porous elastic material uses the Domain material (the material defined for
the domain). This setting should be different from the one for the Fluid material
selected in the Fluid Properties section. Typically, select another material directly (a
material defined in the Materials node) for the porous elastic material and use the
default Domain material option for the Fluid material.

Select a Porous model: Drained matrix, isotropic; Drained matrix, orthotropic; or Drained
matrix, anisotropic. Then enter or select the settings as described.

• Porous Model Drained Matrix, Isotropic

• Porous Model for Drained Matrix, Orthotropic
• Porous Model for Drained Matrix, Anisotropic
• Porous Matrix Parameters for Biot (viscous losses) Model
• Porous Matrix Parameters for Biot-Allard (thermal and viscous losses) Model

THE POROELASTIC WAVES INTERFACE | 327

 Porous Model Drained Matrix, Isotropic
If Drained matrix, isotropic is selected from the Porous model list, select a pair of elastic
properties to describe an isotropic drained porous material. The drained parameters are
also sometimes referred to as the in vacuo elastic parameters; they are in principle
measured without the presence of the saturating fluid. From the Specify list, select:

• Young’s modulus and Poisson’s ratio to specify drained Young’s modulus (elastic
modulus) Ed (SI unit: Pa) and Poisson’s ratio νd (dimensionless). For an isotropic
material, Young’s modulus is the spring stiffness in Hooke’s law, which in 1D form
is σ = Edε, where σ is the stress and ε is the strain. Poisson’s ratio defines the normal
strain in the perpendicular direction, generated from a normal strain in the other
direction and follows the equation ε⊥ = −υε||
• Shear modulus and Poisson’s ratio (the default for the Biot-Allard model) to specify
drained shear modulus Gd (SI unit: Pa) and Poisson’s ratio νd (dimensionless).
• Young’s modulus and Shear modulus to specify drained Young’s modulus (elastic
modulus) Ed (SI unit: Pa) and drained shear modulus Gd (SI unit: Pa).
• Bulk modulus and shear modulus (the default for the Biot model) to specify the
drained bulk modulus Kd (SI unit: Pa) and the drained shear modulus Gd (SI unit:
Pa). The bulk drained modulus is a measure of the solid porous matrix’s resistance
to volume changes. The shear modulus is a measure of the solid porous matrix’s
resistance to shear deformations.
• Lamé parameters to specify the drained Lamé parameters λd (SI unit: Pa) and μd
(SI unit: Pa).
• Pressure-wave and shear-wave speeds to specify the drained pressure-wave speed cp
(SI unit: m/s) and the shear-wave speed cs (SI unit: m/s).

For each pair of properties, select from the applicable list to use the value From material
or enter a User defined value or expression. Each of these pairs define the drained elastic
properties and it is possible to convert from one set of properties to another.

Porous Model for Drained Matrix, Orthotropic

When Drained matrix, orthotropic is selected from the Porous model list, the material
properties of the solid porous matrix vary in orthogonal directions only.

The default properties take values From material. For User defined, enter values or
expressions for the drained Young’s modulus E (SI unit: Pa), the drained Poisson’s ratio
ν (dimensionless), and the drained Shear modulus G (SI unit: Pa).

328 | CHAPTER 3: ELASTIC WAVES INTERFACES

Porous Model for Drained Matrix, Anisotropic
When Drained matrix, anisotropic is selected from the Porous model list, the material
properties of the solid porous matrix vary in all directions, and the stiffness comes from
the symmetric Elasticity matrix, D (SI unit: Pa). The default uses values From material.
For User defined enter values in the 6-by-6 symmetric matrix that displays.

Porous Matrix Parameters for Biot (viscous losses) Model

Enter the following (remaining) parameters necessary to define the properties of a Biot
(viscous losses) porous material model. The defaults use values From material. For User
defined enter other values or expressions as needed.

• Bulk modulus and shear modulus (the default for the Biot model as described above)
to specify the drained bulk modulus Kd (SI unit: Pa) and the drained shear modulus
Gd (SI unit: Pa).
• Drained density of porous material to specify the drained density of the porous
material in vacuum ρd (SI unit: kg/m3). The drained density ρd is equal to
(1 − εp) ρs where ρs is the density of the solid material from which the matrix is
made and εp is the porosity.
• Permeability to specify the permeability of the porous material κp (SI unit: m2). The
permeability is a measure of the ability of the porous material to let fluid pass
through it. It hence gives some measure of the pore size and thus correlates to the
viscous damping experienced by pressure waves propagating in the saturating fluid.
• Porosity to specify the porosity of the material εp (dimensionless). It defines the
amount of void volume inside the porous matrix and takes values between 0 (no
porous material only fluid) and 1 (fully solid material no fluid).
• Biot-Willis coefficient to specify the Biot-Willis coefficient αB (dimensionless). This
coefficient relates the bulk modulus (compressibility) of the drained porous matrix
to a block of solid material. It is defined as
Kd
α B = 1 – -------
Ks

where Kd is the drained bulk modulus and Ks is the bulk module of a block of solid
material (made of the matrix material). The drained bulk modulus is related to the
stiffness of the porous matrix, while the solid bulk modulus is related to the
compressibility of the material or grains from which the porous matrix is made. The
Biot-Willis coefficient is bound by ε p ≤ α B ≤ 1 . A rigid porous matrix (Voigt upper

THE POROELASTIC WAVES INTERFACE | 329

 bound) has αB = εp and a soft or limp porous matrix (Reuss lower bound) has
αB = 1.
• Tortuosity factor (high frequency limit) or the structural form factor τ ∞
(dimensionless). This is a purely geometrical factor that depends on the microscopic
geometry and distribution of the pores inside the porous material. It is independent
of the fluid and solid properties and is normally >1. The default is 2. The more
complex the propagation path through the material, the higher is the absorption.
The tortuosity partly represents this complexity.

Porous Matrix Parameters for Biot-Allard (thermal and viscous losses) Model
Enter the following (remaining) parameters necessary to define the properties of a
Biot-Allard (thermal and viscous losses) porous material model. The defaults use values
From material. For User defined enter other values or expressions as needed.

• Shear modulus and Poisson’s ratio (the default for the Biot-Allard model as described
above) to specify drained shear modulus Gd (SI unit: Pa) and Poisson’s ratio νd
(dimensionless).
• Drained density of porous material to specify the drained density of the porous
material in vacuum ρd (SI unit: kg/m3). The drained density ρd is equal to
(1 − εp) ρs where ρs is the density of the solid material from which the matrix is
made and εp is the porosity.
• Porosity to specify the porosity of the material εp (dimensionless). It defines the
amount of void volume inside the porous matrix and takes values between 0 (no
porous material only fluid) and 1 (fully solid material no fluid).
• Flow resistivity to specify the (static) flow resistivity of the porous material Rf (SI
unit: Pa·s/m2). The flow resistivity is a measure of the ability of the porous material
to let fluid pass through it. It hence gives some measure of the pore size and thus
correlates to the viscous damping experienced by pressure waves propagating in the
saturating fluid. The flow resistivity is also sometimes denoted σ (using the unit
N·s/m4) and it is related to the permeability through κp = μ/Rf.
• Isotropic structural loss factor to specify the loss factor of the porous matrix ηs
(dimensionless). This value introduces the damping due to losses in the porous
structure by transform the elastic moduli into complex valued quantities. This
quantity can be frequency dependent if necessary.
• Tortuosity factor (high frequency limit) or the structural form factor τ∞
(dimensionless). This is a purely geometrical factor that depends on the microscopic
geometry and distribution of the pores inside the porous material. It is independent
of the fluid and solid properties and is normally >1. The default is 2. The more

330 | CHAPTER 3: ELASTIC WAVES INTERFACES

 complex the propagation path through the material, the higher the absorption. The
tortuosity partly represents this complexity.
• Viscous characteristic length to specify the viscous length scale Lv (SI unit: m). This
value is sometimes denoted Λ and replaces the hydraulic radius used in simpler
models to account for the viscous losses that appear in the acoustic boundary layer
at pore walls.
• Thermal characteristic length to specify the thermal length scale Lth (SI unit: m). This
value is sometimes denoted Λ’ and replaces the hydraulic radius used in simpler
models to account for the thermal losses that appear in the acoustic boundary layer
at pore walls.
• Biot-Willis coefficient how to calculate the Biot-Willis coefficient αB by selecting
From material, Rigid assumption (the default), General model, or User defined. This
coefficient relates the bulk modulus (compressibility) of the drained porous matrix
to a block of solid material.
- From material to pick up the value from the domain material.
- Rigid assumption (the default) and the model defines a rigid porous matrix (Voigt
upper bound) where αB = εp.
- General model to define the Biot-Willis coefficient αB according to its general
definition
Kd
α B = 1 – -------
Ks

where Kd is the drained bulk modulus and Ks is the bulk module of the skeleton
material (bulk modulus of a block of solid material made of the matrix material).
When this option is selected also enter the Bulk modulus of skeleton material Ks
(taken from material as default). The drained bulk modulus Kd is related to the
stiffness of the porous matrix, while the skeleton bulk modulus Ks is related to
the compressibility of the material or grains from which the porous matrix is
made. The Biot-Willis coefficient is bound by ε p ≤ α B ≤ 1 . A rigid porous matrix
(Voigt upper bound) has αB = εp and a soft or limp porous matrix (Reuss lower
bound) has αB = 1.
- User defined enter a value for the Biot-Willis coefficient αB (dimensionless). In the
limp limit when αB = 1 it is recommended to use the Poroacoustics feature of The
Pressure Acoustics, Frequency Domain Interface instead of the Poroelastic Waves
interface. In poroacoustics the limp limit, when the structure is so “fluffy” that it
moves with the fluid, has the assumption Kd = 0 and αB = 1 included explicitly.

THE POROELASTIC WAVES INTERFACE | 331

 FLUID PROPERTIES
The default Fluid material uses the Domain material. This is the material defined for the
domain in the Materials node.

Define the properties of the saturating fluid in terms of its density, viscosity and
compressibility but also the viscosity model. The defaults use values for the material
parameters are From material. For User defined enter other values or expressions as
needed.

• Fluid Parameters for Biot (Viscous Losses) Model

• Fluid Parameters for Biot-Allard (Thermal and Viscous Losses) Model

Fluid Parameters for Biot (Viscous Losses) Model

• Density defines the density of the saturating fluid ρf (SI unit: kg/m3).
• Dynamic viscosity to define the dynamic viscosity of the saturating fluid μf
(SI unit: Pa·s). The parameter is important for the amount of viscous damping
experienced by the acoustic waves.
• Compressibility of the saturating fluid χf (SI unit: 1/Pa). Remember that the fluid
compressibility χf is related to the fluid bulk modulus Kf (SI unit: Pa) and the speed
of sound c, through the relation
1 1
χ f = ------ = ----------
Kf 2
c ρf

The compressibility of the fluid also enters the expression for Biot’s module M,
given by

Ks Kd
M = ---------------------------------------------------- K s = ----------------
Kd 1 – αB
1 – ε p – ------- + ε p K s χ f
Ks

It should be noted that Biot-Willis coefficient αB only depends on the properties of

the porous matrix while Biot’s module M depends on both fluid and porous matrix
properties.

Select a Viscosity Model, either Biot’s low frequency range or Biot’s high frequency range.

• Biot’s low frequency range models damping at low frequencies where the acoustic
boundary layer (the viscous penetration depth) is assumed to span the full width of
the pores. This is also the so-called Poiseuille limit.
• For Biot’s high frequency range also select Specify as Reference frequency or
Characteristic pore size. Either enter a Reference frequency fr (SI unit: Hz) or a

332 | CHAPTER 3: ELASTIC WAVES INTERFACES

 Characteristic pore size a (SI unit: m). This model implements a correction factor to
the viscosity that accounts for the relative scale difference between a typical pore
diameter and the acoustic boundary layer thickness. The modified viscosity is of the
form

μf
μ̃ f = μ f F  ---f  f r = ------------------
 f 2
r 2πa ρ f

where fr is the reference frequency and a is a characteristic size of the pores. The
expression for fr is one typically used in literature but it is often measured or
empirically determined. The expression for fr corresponds to finding the frequency
at which the viscous boundary layer thickness is of the scale a.

See High Frequency Correction (Biot Model) for more details.

Fluid Parameters for Biot-Allard (Thermal and Viscous Losses) Model

• Density to define the density of the saturating fluid ρf (SI unit: kg/m3).
• Dynamic viscosity to define the dynamic viscosity of the saturating fluid μf
(SI unit: Pa·s). The parameter is important for the amount of viscous damping
experienced by the acoustic waves.
• Ratio of specific heats to define the ratio of specific heats (adiabatic index) of the
saturating fluid γ (dimensionless).
• Heat capacity at constant pressure to define the (specific) heat capacity at constant
pressure of the saturating fluid Cp (SI unit: J/(kg·K)).
• Thermal conductivity to define the thermal conductivity of the saturating fluid k
(SI unit: W/(m·K)). The parameter is important for the amount of thermal
damping experienced by the acoustic waves.

Porous, Free
The Porous, Free node is the default boundary condition for The Poroelastic Waves
Interface. It means that there are no constraints and no loads acting on the porous
matrix, and a sound-soft boundary for the fluid pressure.

Porous, Free (Sound-Soft Boundary) Theory

THE POROELASTIC WAVES INTERFACE | 333

 Initial Values
The Initial Values node adds initial values for the pressure and displacement field. Add
more Initial Values nodes from the Physics toolbar.

INITIAL VALUES
Enter a value or expression for the Pressure p (SI unit: Pa) and Displacement field u
(SI unit: m) initial values. The default is 0 Pa for the pressure and 0 m for the
displacement field.

Fixed Constraint
The Fixed Constraint node adds a condition that makes the porous matrix fixed (fully
constrained); that is, the displacements are zero in all directions. This boundary
condition also sets an impervious (sound-hard) boundary for the fluid pressure.

Fixed (Sound-Hard Boundary) Equations

Periodic Condition
The Periodic Condition node adds a periodic boundary condition that can be used to
reduce the model size by using symmetries and periodicities in the geometry and
physics interfaces being modeled. The condition is found under the Connections
submenu.

PERIODICITY SETTINGS
Select a Type of periodicity: Continuity (the default), Antiperiodicity, Floquet periodicity
(Bloch periodicity), or Cyclic symmetry, or User defined.

For details see the Periodic Condition description in Pressure Acoustics.

334 | CHAPTER 3: ELASTIC WAVES INTERFACES

Porous, Pressure
The Porous, Pressure node creates a boundary condition that acts as a pressure source
at the boundary, which means a constant acoustic pressure p = p0 is specified. In the
frequency domain, p0 is the amplitude of a harmonic pressure source.

PRESSURE
Enter the value of the Pressure p0 (SI unit: Pa) at the boundary.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Pressure Equations

Prescribed Displacement
The Prescribed Displacement node adds a condition where the displacements are
prescribed in one or more directions to the porous matrix boundary.

If a displacement is prescribed in one direction, this leaves the porous matrix free to
deform in the other directions. Also define more general displacements as a linear
combination of the displacements in each direction.

• If a prescribed displacement is not activated in any direction, this is the same as a

Free constraint.
• If a zero displacement is applied in all directions, this is the same as a Fixed Constraint
for the porous matrix and a Sound Hard Wall for the fluid.

PRESCRIBED DISPLACEMENT
Define the prescribed displacements using a Standard notation or a General notation.

THE POROELASTIC WAVES INTERFACE | 335

 Standard Notation
To define the displacements individually, click the Standard notation button (the
default).

To define a prescribed displacement for each space direction (x, y, and z

for 3D), select one or more of the Prescribed in x direction, Prescribed in y
direction, and Prescribed in z direction check boxes. Then enter a value or
expression for the prescribed displacements u0, v0, or w0 (SI unit: m).

For 2D axisymmetric components and to define a prescribed displacement

for each space direction (r and z), select one or both of the Prescribed in r
direction and Prescribed in z direction check boxes. Then enter a value or
expression for the prescribed displacements u0 or w0 (SI unit: m).

General Notation
To specify the displacements using a General notation that includes any linear
combination of displacement components, click the General notation button.

Enter values in the H matrix and R vector fields. For the H matrix, also select an
Isotropic, Diagonal, Symmetric, or Full matrix and enter values as needed.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Prescribed Displacement Equations

Prescribed Velocity
The Prescribed Velocity node adds a boundary condition where the velocity of the
porous matrix is prescribed in one or more directions. With this boundary condition
it is possible to prescribe a velocity in one direction, leaving the porous matrix free in
the other directions. The condition is found under the More Constraints submenu.

336 | CHAPTER 3: ELASTIC WAVES INTERFACES

COORDINATE SYSTEM SELECTION

Coordinate systems with directions which change with time should not
be used.

PRESCRIBED VELOCITY

To define a porous, prescribed velocity for each space direction (x, y, and
z for 3D), select one or all of the Prescribed in x direction, Prescribed in y
direction, and Prescribed in z direction check boxes. Then enter a value or
expression for the components vx, vy, and vz (SI unit: m/s).

For 2D axisymmetric components and to define a prescribed velocity for

each space direction (r and z), select one or both of the Prescribed in r
direction and Prescribed in z direction check boxes. Then enter a value or
expression for vr and vz (SI unit: m/s).

Prescribed Velocity Equations

Prescribed Acceleration
The Prescribed Acceleration node adds a boundary condition, where the acceleration of
the porous matrix is prescribed in one or more directions. With this boundary
condition, it is possible to prescribe a acceleration in one direction, leaving the porous
matrix free in the other directions. The condition is found under the More Constraints
submenu.

COORDINATE SYSTEM SELECTION

Coordinate systems with directions which change with time should not
be used.

THE POROELASTIC WAVES INTERFACE | 337

 PRESCRIBED ACCELERATION

To define a porous, prescribed acceleration for each space direction (x, y,

and z for 3D), select one or all of the Prescribed in x direction, Prescribed
in y direction, and Prescribed in z direction check boxes. Enter a value or
expression for the prescribed acceleration ax, ay, and az (SI unit: m/s2).

For 2D axisymmetric components and to define a porous, prescribed

acceleration for each space direction (r and z), select one or both of the
Prescribed in r direction and Prescribed in z direction check boxes. Then
enter a value or expression for the prescribed acceleration ar and az
(SI unit: m/s2).

Prescribed Acceleration Equations

Roller
The Roller node adds a roller (sliding wall) constraint as the boundary condition; that
is, the porous matrix displacement is zero in the direction perpendicular (normal) to
the boundary, but the porous matrix is free to move in the tangential direction. This
boundary condition also sets an impervious (sound-hard) boundary for the fluid
pressure. The condition is found under the More Constraints submenu.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Roller Equations

Septum Boundary Load

Add a Septum Boundary Load to boundaries for a pressure acting on the porous matrix
through a septum layer.

338 | CHAPTER 3: ELASTIC WAVES INTERFACES

SEPTUM BOUNDARY LOAD
Enter a Surface density ρsep (SI unit: kg/m3). Enter coordinates for the Load FA (SI
unit: N/m2).

Septum Boundary Load Equations

Symmetry
The Symmetry node adds a boundary condition where there is symmetry in the
pressure and displacement field. Use this condition to reduce the size of a model by
cutting it in half where there are symmetries. The condition is found under the More
Constraints submenu.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

THE POROELASTIC WAVES INTERFACE | 339

 T he E la s t i c Wav es, Ti m e E xp l i c i t
Interface
The Elastic Wave, Time Explicit interface ( ), found under the Acoustics>Elastic Waves
branch ( ) when adding a physics interface, is used to model the transient
propagation of elastic waves in linear elastic materials. It is used to compute the velocity
and strain field in solids with propagating elastic waves. It is suited for time-dependent
simulations with arbitrary time-dependent sources and fields. In general, the interface
is suited for modeling the propagation of elastic waves over large distances relative to
the wavelength, for example, ultrasound propagation for nondestructive testing
(NDT), or seismic waves. The interface includes absorbing layers that are used to set
up effective nonreflecting like boundary conditions (sponge layers). The interface
exists in 2D, 2D axisymmetric, and 3D.

The interface is based on the discontinuous Galerkin (dG or dG-FEM) method and
uses a time explicit solver. The method is very memory efficient and can solve problems
with many million degrees of freedom (DOFs). The method is also well suited for
distributed computing on clusters.

The Elastic Waves, Time Explicit interface can be coupled to the The Pressure Acoustics,
Transient Interface and The Nonlinear Pressure Acoustics, Time Explicit Interface,
using either the Acoustic-Structure Boundary, Time Explicit or the Pair
Acoustic-Structure Boundary, Time Explicit multiphysics couplings.

For modeling linear piezoelectric wave problems the Elastic Waves, Time Explicit
interface can be coupled to the Electrostatics interface, see The Piezoelectric Waves,
Time Explicit Interface.

The interface solves the equations of linear elasticity in the velocity-strain formulation.
The dependent variables are the structural velocity v (SI unit: m/s) and the strain E
(SI unit: 1). Structural damping can be introduced using a Rayleigh damping model.
The displacement u (SI unit: m) can be computed by solving additional equations, see
Computing the Displacement in the Elastic Waves, Time Explicit.

When the Elastic Wave, Time Explicit interface is added, these default nodes are also
added to the Model Builder: Elastic Wave, Time Explicit Model, Free, and Initial Values.

340 | CHAPTER 3: ELASTIC WAVES INTERFACES

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Elastic Wave, Time Explicit to
select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is elte.

• Isotropic-Anisotropic Sample: Elastic Wave Propagation. The

Application Library path: Acoustics_Module/Elastic_Waves/
isotropic_anisotropic_sample
• Ground Motion After Seismic Event: Scattering off a Small
Mountain. The Application Library path: Acoustics_Module/
Elastic_Waves/ground_motion_seismic_event
• Propagation of Seismic Waves Through Earth. The Application
Library path: Acoustics_Module/Elastic_Waves/seismic_waves_earth

FILTER PARAMETERS FOR ABSORBING LAYERS

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. In the Filter Parameters for
Absorbing Layers section you can change and control the values set for the filter used in
the Absorbing Layers in Elastic Waves, Time Explicit. The values of the filter
parameters defined here are used in all absorbing layers added to the model and they
override the value of filter parameters enabled in the material model (Elastic Waves,
Time Explicit Model). The default values of the filter parameters α, ηc, and s are set
to 0.2, 0.01, and 2, respectively. Inside the absorbing layer, it is important to use a filter
that is not too aggressive since this will result in spurious reflections.

For general information about the filter see the Filter Parameters section
under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 341

 DISCRETIZATION
In this section you can select the discretization for the Structural velocity and Strain
tensor, Voigt notation. Per default both are set to Quartic (4th order). Using quartic
elements together with a mesh size equal to approximately half the wavelength to be
resolved, leads to the best performance when using the dG method.

DEPENDENT VARIABLES
The dependent variables (field variables) are the Structural velocity and Structural
velocity, components, and the Strain tensor, Voigt notation and Strain tensor, Voigt
notation, components. The names can be changed, but the names of fields and
dependent variables must be unique within a model.

• For information about modeling strategies, meshing, the absorbing

layer and more, see Modeling with the Elastic Waves Branch.
• For the theoretical background of the model solved, see Theory for the
Elastic Waves, Time Explicit Interface.

Domain, Boundary, Edge, Point, and Pair Nodes for the Elastic
Waves, Time Explicit Interface
The Elastic Waves, Time Explicit Interface has these domain, boundary, edge, and pair
nodes, listed in alphabetical order, available from the Physics ribbon toolbar (Windows
users), Physics context menu (Mac or Linux users), or right-click to access the context
menu (all users). Continuity in velocity and stress is the default condition on interior
boundaries. If a boundary separates two different materials the Material Discontinuity
or Continuity (the pair version of material discontinuity) condition has to be applied
to ensure the continuity in velocity and stress.

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

342 | CHAPTER 3: ELASTIC WAVES INTERFACES

 • Axial Symmetry • Fixed
• Antisymmetry • Free
• Body Load • General Flux/Source
• Boundary Load • General Interior Flux
• Compute Displacement • Low-Reflecting Boundary
• Continuity • Material Discontinuity
• Damping • Prescribed Velocity
• Elastic Waves, Time Explicit Model • Roller
• Initial Values • Symmetry

The Piezoelectric Material feature and the Mechanical Damping

sub-feature are describe in the The Elastic Waves, Time Explicit Interface
section.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Elastic Waves, Time Explicit Model

The Elastic Waves, Time Explicit Model node adds the equations for modeling the
propagation of linear elastic waves. Define the properties of a general isotropic or
anisotropic material. The model solves the governing equations for a general linear
elastic material in a velocity-strain formulation

∂v
ρ ------ – ∇ ⋅ S = F v
∂t
∂ε 1 T
----- – --- [ ∇v + ( ∇v ) ] = 0
∂t 2
S = C:ε

where v is the velocity, ρ the density, S the stress tensor, ε the strain tensor, C is the
elasticity tensor (or stiffness tensor), and Fv is a possible body force. The equations are

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 343

 valid for both isotropic and anisotropic material data. Bulk dissipation can be added by
using the Damping subnode.

COORDINATE SYSTEM SELECTION

The Global coordinate system is selected by default. The Coordinate system list contains
all applicable coordinate systems in the component. The coordinate system is used for
interpreting directions of orthotropic and anisotropic material data and when stresses
or strains are presented in a local system.

Coordinate systems with directions which change with time should not
be used.

LINEAR ELASTIC MATERIAL

Define the Solid model representation of the material. Choose:

• Isotropic for a linear elastic material that has the same properties in all directions.
• Orthotropic for a linear elastic material that has different material properties in
orthogonal directions, so that its stiffness depends on the properties Ei, νij, and Gij.
• Anisotropic for a linear elastic material that has different material properties in
different directions, and the stiffness comes from the symmetric elasticity matrix, D
If orthotropic or anisotropic is selected then also select the Material data ordering to
Voigt (the default) or Standard notation. This option defines what numbering and
notation is used for entering user defined material data and when defining material
properties in the Materials node.

Specification of Elastic Properties for Isotropic Materials

For an Isotropic Solid model, from the Specify list select a pair of elastic properties for
an isotropic material — Young’s modulus and Poisson’s ratio, Young’s modulus and shear
modulus, Bulk modulus and shear modulus, Lamé parameters, or Pressure-wave and
shear-wave speeds. For each pair of properties, select from the applicable list to use the
value From material or enter a User defined value or expression.

344 | CHAPTER 3: ELASTIC WAVES INTERFACES

Each of these pairs define the elastic properties and it is possible to convert from one
set of properties to another according to Table 3-1.
TABLE 3-1: EXPRESSIONS FOR THE ELASTIC MODULI.

DESCRIPTION VARIABLE D(E,ν) D(E,G) D(K,G) D(λ,μ)

Young’s E= E E 9KG - 3λ + 2μ
------------------ μ --------------------
modulus 3K + G λ+μ
Poisson’s ν= ν E- λ -
---  1 – ------------------
------- –1 1 3G - --------------------
ratio 2G 2 3K + G 2(λ + μ)

Bulk K= E - EG - K 2μ
----------------------- -------------------------- λ + -------
modulus 3 ( 1 – 2ν ) 3 ( 3G – E ) 3
Shear G= E - G G μ
--------------------
modulus 2(1 + ν)
Lamé λ= Eν G ( E – 2G )- 2G λ
--------------------------------------
- --------------------------- K – --------
parameter λ ( 1 + ν ) ( 1 – 2ν ) 3G – E 3
Lamé μ= E - G G μ
--------------------
parameter μ 2(1 + ν)
Pressure- cp = E(1 – ν) - G ( 4G – E )- K + 4G ⁄ 3- λ + 2μ
wave speed ----------------------------------------- --------------------------- ------------------------- ----------------
ρ ( 1 + ν ) ( 1 – 2ν ) ρ ( 3G – E ) ρ ρ
Shear-wave cs = E - G⁄ρ G⁄ρ μ⁄ρ
speed ------------------------
2ρ ( 1 + ν )

The individual property parameters are:

• Young’s modulus (elastic modulus) E.

• Poisson’s ratio ν.
• Shear modulus G.
• Bulk modulus K.
• Lamé parameter λ and Lamé parameter μ.
• Pressure-wave speed (longitudinal wave speed) cp.
• Shear-wave speed (transverse wave speed) cs. This is the wave speed for a solid
continuum.

Specification of Elastic Properties for Orthotropic Materials

When Orthotropic is selected from the Solid model list, the material properties vary in
orthogonal directions only. The Material data ordering can be specified in either
Standard or Voigt notation. When User defined is selected in 3D, enter three values in

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 345

 the fields for Young’s modulus E, Poisson’s ratio ν, and the Shear modulus G. This
defines the relationship between engineering shear strain and shear stress. It is
applicable only to an orthotropic material and follows the equation

τ ij
ε ij = --------
G ij

νij is defined differently depending on the application field. It is easy to

transform among definitions, but check which one the material uses.

Specification of Elastic Properties for Anisotropic Materials

When Anisotropic is selected from the Solid model list, the material properties vary in all
directions, and the stiffness comes from the symmetric Elasticity matrix, D. The
Material data ordering can be specified in either Standard or Voigt notation. When User
defined is selected, a 6-by-6 symmetric matrix is displayed.

Note that the 6-by-6 Elasticity matrix, D has to be complete regardless of

the spatial dimension of the problem. This is due to the generalized plane
strain formulation of the governing equations solved in 2D.

Density
The default Density ρ uses values From material. For User defined enter another value or
expression. If any material in the model has a temperature dependent mass density, and
From material is selected, the Volume reference temperature list will appear in the Model
Input section. As a default, the value of Tref is obtained from a Common model input.
You can also select User defined to enter a value or expression for the reference
temperature locally.

LAX–FRIEDRICHS FLUX PARAMETER

To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box. In this section, you specify the value of the Lax–
Friedrichs flux parameter τLF (default value: 0.2). This value controls the numerical
flux between the elements (nodal discontinuous Lagrange elements) used with the
discontinuous Galerkin (dG) method. The numerical flux defines how adjacent
elements are connected and how continuous v and E are. Different definitions of the
numerical flux lead to different variants of the dG method. The flux implemented here
is the so-called global Lax–Friedrichs numerical flux. The value of the parameter τLF

346 | CHAPTER 3: ELASTIC WAVES INTERFACES

should be between 0 and 0.5. For τLF = 0 a so-called central flux is obtained. Setting
τLF = 0.5 gives a maximally dissipative global Lax–Friedrichs flux.

For general information about the numerical flux see the Numerical Flux
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

FILTER PARAMETERS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. By default, the filter parameters
α, ηc, and s are not active. Select the Activate check box to activate the filter. The filter
provides higher-order smoothing for the dG formulation and can be used to stabilize
the solution. Inside absorbing layers the settings given here are overridden by the Filter
Parameters for Absorbing Layers.

Enter values for the filter parameters in the corresponding text fields (default values:
36, 0.6, and 3). α must be positive and lie between 0 and 36. α = 0 means no
dissipation and α = 36 means maximal dissipation. ηc should be between 0 and 1,
where 0 means maximal filtering and 1 means no filtering (even if filtering is active).
The s parameter should be larger than 0 and controls the order of the filtering (a
dissipation operator of order 2s). For s = 1, you get a filter that is related to the classical
2nd-order Laplacian. A larger s gives a more pronounced low-pass filter.

For more detailed information about the filter see the Filter Parameters
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

ESTIMATE OF MAXIMUM WAVE SPEED

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. The maximum wave speed enters
the expression for the cell wave time scale used by the solver to estimate a stable
internal time step. Select Automatic (the default) or User defined. With the Automatic
option the maximum speed is set equal to the pressure wave speed, for User defined
enter another value for cmax. The estimate for isotropic materials is correct, but for
certain configurations of anisotropic materials it can be necessary to enter a larger value
to ensue numerical stability.

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 347

 Damping
Use the Damping subnode to add Rayleigh damping to the material model (the Elastic
Waves, Time Explicit Model node). The only option for the Damping type is Rayleigh
damping.

In this damping model, the damping parameter ξ is expressed in terms of the mass m
and the stiffness k as

ξ = α dM m + β dK k

That is, Rayleigh damping is proportional to a linear combination of the stiffness and
mass; there is no direct physical interpretation of the mass damping parameter αdM
and the stiffness damping parameter βdM.

This feature adds a right hand side to the momentum equation of the form:

∂S
– α dM ρv + β dK  ∇ ⋅ ------
 ∂t 

DAMPING SETTINGS
Select Input parameters — Alpha and beta — to enter the damping parameters explicitly,
or Damping ratios to derive the damping parameters from the damping ratio at two
frequencies.

When Alpha and beta is selected, enter the Mass damping parameter αdM and the
Stiffness damping parameter βdK.

When Damping ratios is selected, enter two frequencies, f1 and f2, and the
corresponding damping ratios ζ1 and ζ2 at these frequencies. The Rayleigh damping
parameters are computed as

ς1 f2 – ς2 f1
α dM = 4πf 1 f 2 --------------------------
2 2
-
f2 – f1
ς2 f2 – ς1 f1
β dK = --------------------------
2 2
-
π ( f2 – f1 )

FILTER PARAMETERS
When the Damping subnode is added, filtering is automatically enabled on the selected
domains. To display the section, click the Show More Options button ( ) and select
Advanced Physics Options in the Show More Options dialog box. The filtering is added in

348 | CHAPTER 3: ELASTIC WAVES INTERFACES

order to stabilize the underlying equations. This is necessary because damping is added
as a right-hand-side contribution to the governing equations in the dG-FEM
formulation. In general, the values for the damping parameters should represent
physical behavior. If unrealistically large values are entered, the method becomes
unstable even with the filtering turned on. Note that if only very small amounts of
damning are added, the filter will contribute with an additional small amount of
numerical damping.

Compute Displacement
Use the Compute Displacement subnode (to the Elastic Waves, Time Explicit Model
node) to add additional equations for computing the displacement in a point, along an
edge, on a surface, or in a domain. The feature is a pure postprocessing type and the
variables can be used to visualize displacements. The feature adds and solves an ODE
and thus requires additional computational resources. This is necessary since the
Elastic Waves, Time Explicit interface uses the velocity-strains formulation of the
governing equations. It is recommended to only use the feature where necessary.

Isotropic-Anisotropic Sample: Elastic Wave Propagation. The

Application Library path: Acoustics_Module/Elastic_Waves/
isotropic_anisotropic_sample

Initial Values
The Initial Values node adds initial values for the velocity field and strain field that can
serve as an initial condition for a transient simulation. In addition to the default Initial
Values node always present in the interface, you can add more Initial Values nodes if
needed.

Body Load
The Body Load node adds a general volumetric load to the selected domain. The feature
can be used to approximate sources of complex external processes.

FORCE
Select a Load type — Force per unit volume, Total force, or for 2D components, Force
per unit area.

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 349

 Then enter values or expressions for the components in the force vector based on the
space dimension. These can be general time dependent expressions or data from
interpolation functions.

For Total force COMSOL Multiphysics divides the total force by the
volume of the domains where the load is active. For 2D components, and
if Force per unit area is selected, the body load as force per unit volume is
then the value of F divided by the thickness.

Axial Symmetry
In a 2D axisymmetric geometry the Axial Symmetry condition is automatically added
on the axis of symmetry at r = 0. On this boundary the radial component of the
velocity is set equal to zero as well as the tangential component of the traction:

er ⋅ v = 0
e r × Sn = 0

Free
The Free node is the default boundary condition. It means that there are no constraints
and no loads acting on the boundary. When the physics interfaces is added, a default
Free node is added. If you look at the selections for this node, it will show all
boundaries which do not have any boundary conditions applied.

Fixed
The Fixed node adds a condition that makes the geometric entity fixed; that is, the
velocities are zero in all directions on the selected geometrical entities.

Prescribed Velocity
The Prescribed Velocity node adds a boundary condition where the velocity is
prescribed in one or more directions. It is possible to prescribe a velocity in one or
more directions, leaving the solid fixed in the other directions (the default 0 value).

350 | CHAPTER 3: ELASTIC WAVES INTERFACES

PRESCRIBED VELOCITY
Enter the velocity vector v0 components. These can be general time dependent
expressions or data from interpolation functions.

Roller
The Roller node adds a roller constraint as the boundary condition; that is, the
displacement is zero in the direction perpendicular (normal) to the boundary, but the
boundary is free to move in the tangential direction. A Roller condition is
mathematically similar to a Symmetry condition.

Boundary Load
Use a Boundary Load to apply tractions or pressure to boundaries.

FORCE
Select a Load type — Force per unit area, Pressure, Total force, or for 2D components,
Force per unit length. Then enter values or expressions for the components in the force
vector based on the space dimension.

• For Force per unit length, the boundary load as force per unit area is the value of F
divided by the thickness.
• For Total force, COMSOL Multiphysics divides the total force by the area of the
surfaces where the load is active.
• For Pressure, a scalar input is given, and the orientation of the load is given by the
normal to the boundary. The pressure is positive when directed toward the solid. In
a geometrically nonlinear analysis, the current surface normal and area are used.

Low-Reflecting Boundary
Use the Low-Reflecting Boundary node to let elastic waves pass out from the model with
minimal reflections. As a default, it takes material data from the domain in an attempt
to create a perfect impedance match for both pressure waves and shear waves. The
condition is sensitive to the direction of the incoming wave and works best for normal
incidence.

The Low-Reflecting Boundary should be used in combination with the Absorbing Layers
in Elastic Waves, Time Explicit to create efficient absorbing conditions (sponge layers)

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 351

 that work for most angles of incidence. The condition should be added at the outside
of the absorbing layer.

DAMPING
Select a Damping type — P and S waves or User defined. For User defined enter values or
expressions for the Mechanical impedance di. The defaults for all values are
0.5*[Link]*([Link]+[Link]) (on the diagonal for the Diagonal and Symmetric
options).

See the Low-Reflecting Boundary Condition in the theory chapter of the

Structural Mechanics Module User’s Guide.

Symmetry
The Symmetry node adds a boundary condition that represents symmetry in the
geometry and in the loads. A symmetry condition is free in the plane and fixed in the
out-of-plane direction.

Antisymmetry
The Antisymmetry node adds a boundary condition for an antisymmetry boundary. An
antisymmetry condition is fixed in the plane and free in the out-of-plane direction.

Material Discontinuity
The Material Discontinuity node adds an interior boundary condition to handle jumps
in material properties. As the interface is based on the discontinuous Galerkin method,
special handling is necessary when a jump exists in the material properties.

The condition implements continuity in velocity and normal stress over the interior
boundary:

v down – v up = 0
( S down – S up )n = 0

Subscripts “up” and “down” in the equation refer to the up and down sides of the
interior boundary, relative to the normal n.

352 | CHAPTER 3: ELASTIC WAVES INTERFACES

Continuity
The Continuity node, exists as an option at interfaces between parts in an assembly,
from the Pairs submenu. The condition is similar to the Material Discontinuity
condition that applies to interior boundaries. The difference is that the Continuity
condition should also be applied both when the same material exists on the two sides
of a pair and when the material is different. The advantage of using a pair feature is that
the mesh does not need to be conforming on the two sides of the interface (the two
parts of the assembly). This is especially advantageous for the time explicit
discontinuous Galerkin method as the time step depends on mesh size and local speed
of sound.

The condition implements continuity in velocity and in the normal stress over the pair:

v src = v dst
S src n src = – S dst n dst

Subscripts “src” and “dest” in the equation refer to the source and destination in the
pair; that is, the two sides in the pair.

General Flux/Source
The General Flux/Source node, found under the More submenu, adds the most general
boundary condition for the elastic waves equation, as it is formulated (in the
conservative form) for the discontinuous Galerkin method (dG). The condition
defines the normal flux g at an exterior boundary. Care should be taken when using
this advanced condition as the method is sensitive to proper formulation of boundary
conditions. Only one characteristic can enter the domain at any boundary at the time,
meaning that it is easy to over-specify the problem.

GENERAL FLUX/SOURCE
Enter the expression for the components of the Flux vector g.

General Interior Flux

The General Interior Flux node, found under the More submenu, adds the most general
interior boundary condition for the Elastic Waves, Time Explicit model, as it is
formulated (in the conservative form) for the discontinuous Galerkin method (dG).
The condition defines the normal flux g at an interior boundary. It should include any
necessary numerical flux (as it overrides the Lax–Friedrichs flux). Care should be taken

THE ELASTIC WAVES, TIME EXPLICIT INTERFACE | 353

 when using this advanced condition as the method is sensitive to proper formulation
of boundary conditions.

GENERAL INTERIOR FLUX

Enter the expression for the components of the Flux vector g.

354 | CHAPTER 3: ELASTIC WAVES INTERFACES

The Piezoelectric Waves, Time
Explicit Interface
The Piezoelectric Waves, Time Explicit multiphysics interface ( ), found under the
Acoustics>Elastic Waves branch ( ) when adding a physics interface. This interface
combines The Elastic Waves, Time Explicit Interface and The Electrostatics Interface
together with a dedicated multiphysics coupling and material models. These add the
constitutive relationships required to model piezoelectric phenomena in the time
domain for wave propagation. Both the direct and inverse piezoelectric effects can be
modeled and the piezoelectric coupling can be formulated using the strain-charge or
stress-charge forms.

The interface is based on the discontinuous Galerkin (dG or dG-FEM) method and
uses a time explicit solver. The electrostatics problem is solved at every time step
through an algebraic system of equations solved with classical FEM. The piezoelectric
problem is solved fully coupled. The method is very memory efficient and can solve
problems with many million degrees of freedom (DOFs). The method is also well
suited for distributed computing on clusters.

When adding the Piezoelectric Waves, Time Explicit multiphysics interface a dedicated
time explicit version of the Piezoelectric Effect multiphysics coupling, called
Piezoelectric Effect, Time Explicit, is added. The feature is used to couple the
Piezoelectric Material node in Elastic Waves, Time Explicit to the Charge
Conservation, Piezoelectric node in Electrostatics.

When setting up models using an Assembly geometry and the Pair

multiphysics couplings it is recommended to always use imprints for
assemblies. This will both increase performance and stability.

• Angle Beam Nondestructive Testing. The Application Library path:

Acoustics_Module/Ultrasound/angle_beam_ndt
• Ultrasonic Flowmeter with Piezoelectric Transducers. The
Application Library path: Acoustics_Module/Ultrasound/
flow_meter_piezoelectric_transducers

THE PIEZOELECTRIC WAVES, TIME EXPLICIT INTERFACE | 355

 Piezoelectric Material
The Piezoelectric Material node defines the piezoelectric material properties either in
stress-charge form using the elasticity matrix and the coupling matrix, or in
strain-charge form using the compliance matrix and the coupling matrix. It is normally
used together with a Piezoelectric Effect, Time Explicit multiphysics coupling node
and a corresponding Charge Conservation, Piezoelectric node in the Electrostatics
interface. This node is added by default to the Elastic Waves, Time Explicit interface
when adding a Piezoelectric Waves, Time Explicit multiphysics interface.

This material model requires the Structural Mechanics Module, MEMS Module, or
Acoustics Module.

When the Piezoelectric Material node is added to the Elastic Waves, Time
Explicit interface in the absence of an active Piezoelectric Effect, Time
Explicit multiphysics coupling node, the material behaves similarly to a
Elastic Waves, Time Explicit Model node. The elastic properties
correspond to the elasticity or compliance matrix entered (see below).
The piezoelectric effect is then not included in the equation system.

COORDINATE SYSTEM SELECTION

Select the Coordinate system that defines the polarization orientation of the
piezoelectric crystal. The default is the Global coordinate system. The polarization is
defined in the local x-axis direction in 2D, while it uses the local z-axis in 2D
axisymmetry and 3D.

For entering the material properties matrices (see below), use the
following order (Voigt notation), which is the common convention for
piezoelectric materials: xx, yy, zz, yz, xz, zy.

PIEZOELECTRIC MATERIAL PROPERTIES

Select a Constitutive relation as Stress-charge form (the default) or Strain-charge form.
For each of the following, the default uses values From material. For User defined enter
other values in the matrix or field as needed.

• For Stress-charge form, select an Elasticity matrix, Voigt notation (cE).

• For a Strain-charge form, select a Compliance matrix, Voigt notation (sE).
• Select a Coupling matrix, Voigt notation (eES or dET).

356 | CHAPTER 3: ELASTIC WAVES INTERFACES

• Select a Relative permittivity (εrS or εrT).
• Enter values for the Remanent electric displacement (Dr).
• Select a Density (ρ).

LAX-FRIEDRICH FLUX PARAMETER

See Lax–Friedrichs Flux Parameter.

FILTER PARAMETERS
See Filter Parameters.

ESTIMATE OF MAXIMUM WAVE SPEED

See Estimate of Maximum Wave Speed.

Mechanical Damping
The Mechanical Damping is a sub-node to the Piezoelectric Material used to define the
mechanical damping properties for the parent piezoelectric material (the elastic part).
See the Damping node in The Elastic Waves, Time Explicit Interface.

THE PIEZOELECTRIC WAVES, TIME EXPLICIT INTERFACE | 357

 Modeling with the Elastic Waves
Branch
In this section:

• Meshing and Solving Wave Problems Solved with Solid Mechanics

• Meshing Poroelastic Waves Models
• Solving Large Poroelastic Wave Models
• Meshing and Solving Elastic Waves, Time Explicit Models
• Absorbing Layers in Elastic Waves, Time Explicit
• Computing the Displacement in the Elastic Waves, Time Explicit

Meshing and Solving Wave Problems Solved with Solid Mechanics

In solids several waves propagate depending on the configuration of the model. The
wave with the shortest wavelength (the slowest wave) needs to be resolved with 5 to 6
mesh elements. In general geometry details also need to be resolved by the mesh. For
thin structures use at least two mesh elements in the thickness.

In the bulk, the pressure wave is typically the fastest wave and the shear wave is the
slowest wave. The wave speeds can be evaluated through the variables [Link] and
[Link]. They can be evaluated before solving a model fully, by using the Get Initial
Value on the study, and then plot the variables or evaluate them in a point. If a free
surface or a material discontinuity is present in the model interfacial waves will also
propagate. For example, the classical estimate for the Rayleigh wave speed vR is

v R 0.87 + 1.12ν
------ ≈ ---------------------------------
cs 1+ν

where ν is the Poisson’s ratio and cs is shear wave speed.

When solving elastic wave problems with the Solid Mechanics interface in the time
domain, it is important to make changes to the Transient Solver Setting section (top
interface node). Expand the section and select the check-box Maximum frequency to
resolve, then enter the maximum frequency to resolve in the model fmax,sol (remember
to reset the solver if changes are made to this section). The maximum frequency is
dictated by the frequency content of the source as well as the eigenmodes of the

358 | CHAPTER 3: ELASTIC WAVES INTERFACES

structure that can be excited. Setting this correctly will results in a transient solver
tuned specifically for modeling wave problems.

The correct setup of the solver (specifying the maximum frequency) when
using the Solid Mechanics interface, to model transient wave problems, is
done to ensure adequate resolution in time and space. The transient
solver is based on the time implicit method. This is in contrast to the time
explicit based physics interface, where the internal time step needs to
ensure stability of the method.

Meshing Poroelastic Waves Models

Three different waves can propagate in a poroelastic material; fast pressure waves, slow
pressure waves, and shear waves. Their speed can be evaluated through the variables
pelw.cp_fast, pelw.cp_slow, and pelw.cs_poro, respectively. These should not be
confused with the material properties for the constituting solid ([Link] and
[Link]) and the fluid in the pores (pelw.c). The wave speed values can be evaluated
before solving a model fully, by using the Get Initial Value on the study. Then plot the
variables or evaluate them in a point.

All three poroelastic wave types should be resolved by the computational mesh. This
means that the minimum mesh size is given by the wavelength of the slowest wave (the
shortest wavelength) divided by 5 to 6. The mesh should of course also resolve
geometry details such as highly curved boundaries and narrow regions. In general a
mesh convergence study is always recommended.

Solving Large Poroelastic Wave Models

Large poroelastic waves models and especially large multiphysics models, that involve
poroelastic waves physics, can be solved using an iterative solver (see the application
library example below). For iterative approaches using tight tolerances is
recommended. The preconditioner can be either the multigrid (geometric multigrid
or GMG) solving for both pressure and displacement (as in the example). An
alternative is to use hybridization and a direct preconditioner for the pressure DOFs

MODELING WITH THE ELASTIC WAVES BRANCH | 359

 and another direct preconditioner for the displacement DOFs. Both should use
PARDISO as the direct solver.

Headphone on an Artificial Ear: Application Library path

Acoustics_Module/Electroacoustic_Transducers/headphone_artificial_ear

Meshing and Solving Elastic Waves, Time Explicit Models

Meshing and solution time are closely linked when modeling physics based on the
discontinuous Galerkin (dG) time explicit method. The computational mesh has to
resolve the shortest wavelength in the model (the slowest wave), while it is the fastest
wave speed and the smallest mesh element that dictate the internal time step of the
solver.

In general, the maximal mesh size hmax is dictated by the smallest wavelength (the
slowest wave) such that

c min c min
h max = ----------------------- to ------------------
1.5 ⋅ f max 2 ⋅ f max

for the default 4th (quartic) order discretization, where fmax is the maximal frequency
to resolve in the model, given by the frequency content of the source, and cmin is the
slowest waves speed in the model. This is typically the shear wave speed cs, but for
problems with a free interface or a material discontinuities, interfacial waves also exist.
For example, the classical estimate for the Rayleigh wave speed vR is

v R 0.87 + 1.12ν
------ ≈ ---------------------------------
cs 1+ν

where ν is the Poisson’s ratio and cs is shear wave speed.

Internally, the time step taken by the time explicit method is given by the global
minimum of the local mesh size relative to the maximal local wave speed. This is also
known as the cell wave time scale, the value can be visualized by plotting the variable
[Link]. This means that small mesh elements should be avoided, see Meshing,
Discretization, and Solvers for the Convected Wave Equation documentation, for
more details.

The maximal wave speed used to predict the cell wave time scale uses the pressure wave
speed cp as default. The value is exact for isotropic materials but is an estimate for

360 | CHAPTER 3: ELASTIC WAVES INTERFACES

materials with anisotropic properties. For anisotropic materials it can be necessary to
use a larger value to ensure stability of the numerical method. This can be done in the
Estimate of Maximum Wave Speed section for each material model. To display this
section, click the Show More Options button ( ) and select Advanced Physics Options
in the Show More Options dialog. Select Automatic (the default) or User defined. With
the Automatic option the maximum speed is set equal to the pressure wave speed, for
User defined enter another value for cmax (a larger value may be needed for stability).

When several materials are used in a model, the use of assemblies and pair features is
recommended (Continuity and Pair Acoustic-Structure Boundary, Time Explicit). The
advantage of using a pair feature is that the mesh does not need to be conforming on
the two sides of the interface (the two parts pf the assembly). This is especially
advantageous for the time explicit discontinuous Galerkin method as the time step
depends on mesh size and local speed of sound.

If material properties result in large variations of the local cell wave time scale, then it
can be advantageous to use the Adam-Bashforth 3 (local) method instead of the default
Runge-Kutta method.

Absorbing Layers in Elastic Waves, Time Explicit

The Elastic Waves, Time Explicit Interface includes so-called absorbing layers, also
often referred to as sponge layers. The layers work by combining three techniques: a
scaling system, filtering, and a low-reflecting condition.

The layers are set up by adding the Absorbing Layer under the Definitions node. This
adds a special scaled system. The scaling effectively slows down the propagating waves
and ensures that they “align up” (normal) toward the outer boundary. This means that
the waves will hit the outer boundary in a closer to normal direction. Filtering
attenuates and filters out high frequency components of the wave. The filter
parameters are controlled under the Filter Parameters for Absorbing Layers section at
the main physics level. Finally, at the outer boundary of the layer add a Low-Reflecting
Boundary condition, which will work well to remove all remaining waves, as normal
incidence has been ensured.

For more detailed information about the filter see the Filter Parameters
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

MODELING WITH THE ELASTIC WAVES BRANCH | 361

 The physical thickness of the layer should be such that it contains 2 to 3 mesh
elements. This in practice means that it should roughly have a thickness equal to the
largest wavelength resolved in the model. You want to avoid to create small mesh
elements in the absorbing layers. These can potentially slow down the solution
procedure since the internal time stepping is controlled by the mesh size.

For the Absorbing Layers select the Type (Cartesian, Cylindrical, Spherical, or User defined)
under the Geometry section. Enter values for the Physical Width and Pole Distance under
the Scaling section.

For more details on the Geometry and Scaling see the Infinite Elements,
Perfectly Matched Layers, and Absorbing Layers section in the COMSOL
Multiphysics Reference Manual.

For the layers to work optimally, the filter should not be too aggressive. Moreover, the
scaled coordinates in the layer domain should also vary smoothly. To inspect the scaled
system you can, for example, plot the coordinate variables x_absorb_ab1,
y_absorb_ab1, and z_absorb_ab1. In 2D axisymmetry, they are r_absorb_ab1 and
z_absorb_ab1. Note that the variables are not in the plot menu under the Replace
Expression.

Using the absorbing layers with the three combined techniques will enable the
reduction of spurious reflections by a factor between 100 and 1000 compared to the
incident amplitude.

• Isotropic-Anisotropic Sample: Elastic Wave Propagation. The

Application Library path: Acoustics_Module/Elastic_Waves/
isotropic_anisotropic_sample
• Ground Motion After Seismic Event: Scattering off a Small
Mountain. The Application Library path: Acoustics_Module/
Elastic_Waves/ground_motion_seismic_event

Computing the Displacement in the Elastic Waves, Time Explicit

The Elastic Waves, Time Explicit interface solves for the velocity v and the strain ε. It is
sometimes useful to also compute the displacement u. This comes at an additional
computational cost as an additional equation needs to be solved. The cost depends on

362 | CHAPTER 3: ELASTIC WAVES INTERFACES

where the displacement evaluation is required, in a point, on an edge, on a boundary,
or in a domain. In all cases the simple ODE related the displacement and the velocity:

du
------- = v
dt

This computation is predefined in the Compute Displacement features that can be

added as a subnode to the Elastic Waves, Time Explicit Model node.

For an example where this feature is used in a point see the tutorial
Isotropic-Anisotropic Sample: Elastic Wave Propagation. The
Application Library path: Acoustics_Module/Elastic_Waves/
isotropic_anisotropic_sample

MODELING WITH THE ELASTIC WAVES BRANCH | 363

 Theory for the Poroelastic Waves
Interfaces
The Poroelastic Waves Interface theory is described in this section:

• Elastic Waves Introduction

• Poroelastic Waves Theory
• Boundary Conditions for Poroelastic Waves
• Postprocessing Variables
• References for the Poroelastic Waves Interfaces

Elastic Waves Introduction

The most general linear relation (more details are found in the Structural Mechanics
Theory section of the Structural Mechanics Module User’s Guide) between the stress
and strain tensors in solid materials can be written as

σ ij = c ijkl ε kl

here, σ is the Cauchy’s stress tensor, ε is the strain tensor, and cijkl is a fourth-order
elasticity tensor. For small deformations, the strain tensor is defined as

1
ε ( u ) = --- ( ∇u + ∇u T )
2

where u represents the displacement vector.

The elastic wave equation is then obtained from Newton’s second law

∂2
ρ -------2- u – ∇ ⋅ ( σ ( u ) – s 0 ) = F
∂t

here, ρ is the medium density, and s0 and F represent source terms.

An important case is the time-harmonic wave, for which the displacement varies with
time as

iωt
u ( x, t ) = u ( x ) e

364 | CHAPTER 3: ELASTIC WAVES INTERFACES

with f (SI unit: Hz) denoting the frequency and ω = 2π f (SI unit: rad/s) the angular
frequency. Assuming the same time-harmonic dependency for the source terms s0 and
F, the wave equation for linear elastic waves reduces to an inhomogeneous Helmholtz
equation:

–ρ ω 2 u – ∇ ⋅ ( σ ( u ) –s0 ) = F (3-1)

Alternatively, treat this equation as an eigenvalue PDE to solve for eigenmodes and
eigenfrequencies as described in the Structural Mechanics Module User’s Guide in
the Structural Mechanics Modeling chapter under Eigenfrequency Analysis. Also add
damping as described in Mechanical Damping and Losses.

Poroelastic Waves Theory

In his seminal work, Biot extended the classical theory of linear elasticity to porous
media saturated with fluids (Ref. 1, Ref. 2, and Ref. 3).

In Biot’s theory, the bulk moduli and compressibilities are independent of the wave
frequency, and can be treated as constant parameters. The porous matrix is described
by linear elasticity and damping is introduced by considering the viscosity of the fluid
in the pores, which can be frequency dependent. This description is adequate for the
propagation of poroelastic waves in soils and rocks where the saturating fluid is a liquid,
like oil or water. This formulation is referred to as the Biot model (this is in some sense
the classical formulation).

• High Frequency Correction (Biot Model)

When the considered porous material is saturated by a gas, like air, thermal losses need
to be included in order to properly model its behavior. This is the case when modeling
sound absorbers, car cabin liners, or foams used in headsets or loudspeakers. The
formulation of the equations where both the thermal and viscous losses are included is
sometimes referred to as Biot-Allard model. In this case both the viscosity and the
fluid compressibility are considered to be frequency dependent and complex valued
(Ref. 7, Ref. 8, Ref. 9, Ref. 10, and Ref. 11).

• Biot-Allard Model (Viscous and Thermal Losses)

Consider Biot’s expressions for poroelastic waves (Ref. 3, Ref. 4, and Ref. 6)

THEORY FOR THE POROELASTIC WAVES INTERFACES | 365

 ∂2 ∂2
ρ av -------2- u + ρ f -------2- w – ∇ ⋅ σ = 0
∂t ∂t
(3-2)
∂ 2 μ f∂ τ ∂2
ρ f -------2- u + ----- ----- w + ----- ρ f -------2- w + ∇p f = 0
∂t κ ∂t ε p ∂t

Here, u is the displacement of the porous material, σ is the total stress tensor (fluid
and porous material), w is the fluid displacement with respect to the porous matrix, ρf
and μf are the fluid’s density and viscosity, τ is the tortuosity, εp is the porosity, pf is the
fluid pore pressure, κ is the permeability and ρav the average density. The average
density is the total density (porous material plus pore fluid) ρav = ρd + εpρf.

Assuming a time-harmonic dependency for the variables, u(x,t) = u(x)eiωt,

w(x,t) = w(x)eiωt, the time derivatives can be removed, so the system in Equation 3-2
becomes

– ρ av ω 2 u + ρ f ω 2 w – ∇ ⋅ σ = 0
(3-3)
– ρ f ω 2 u – ω 2 ρ c ( ω )w + ∇p f = 0

Here, the complex density ρc(ω) (Ref. 5) accounts for the tortuosity, porosity, and
fluid density, and the viscous drag on the porous matrix

τ μf
ρ c ( ω ) = ----- ρ f + ---------- (3-4)
εp iωκ

HIGH FREQUENCY CORRECTION (BIOT MODEL)

At low frequencies or for small pore sizes, the flow profile inside the pores can be
assumed to be Poiseuille like. In this case, the viscosity in Equation 3-4 effectively has
a constant value. For increasing frequency, the profile changes and a
frequency-dependent correction factor needs to be taken into account. This is done by
selecting the Biot’s high frequency range option from the Viscosity model list. In this case
Equation 3-4 is implemented with a frequency-dependent viscosity μc(f) (Ref. 2,
Ref. 3, Ref. 5)

μ c ( f ) = μ f ⋅ F c  ---f 
f r

Here, fr is a reference frequency (SI unit: Hz), which determines the low-frequency
range f << fr and the high-frequency range f >> fr.

366 | CHAPTER 3: ELASTIC WAVES INTERFACES

The reference frequency fr can be interpreted as the limit when viscous forces equal
inertial forces in the fluid motion. In a pore with characteristic size a, this happens
when the viscous penetration depth is equal to the pore radius.

μf μf
δv = ---------------
- = a  f r = ------------------
2πf r ρ f 2πa ρ f
2

In the low-frequency limit, viscous effects dominate, while in the high-frequency limit,
inertial effects dominate fluid motion in the pores (losses occur in the viscous
boundary layer). In Biot’s low frequency range, ω → 0 and Fc = 1.

In order to account for a frequency dependence on the viscous drag, Biot defined the
operator Fc(Θ) as

ΘT ( Θ )
F c ( Θ ) = ---  --------------------------------------
1
4 1 + 2iT ( Θ ) ⁄ Θ

where, T(Θ) is related to the Kelvin functions Ber(Θ) and Bei(Θ)

Ber' ( Θ ) + iBei' ( Θ ) – –i J1 ( –i Θ )
T ( Θ ) = ----------------------------------------------- = --------------------------------------
Ber ( Θ ) + iBei ( Θ ) J0 ( –i Θ )

and J0 and J1 are Bessel functions of the first kind. This expression can be recognized
as the loss terms in Zwikker-Kosten like equivalent fluid models (Derivation of the
Equivalent Bulk Modulus Valid for Any Fluid in the Zwikker-Kosten Theory) or the
loss models for cylindrical waveguides in the narrow region acoustics or LRF models
(About the Narrow Region Acoustics Models) models.

U-P FORMULATION
The formulation in terms of the displacements u and w is not optimal from the
numerical viewpoint, since it requires to solve for two displacement fields (Ref. 7,
Ref. 8, Ref. 9). The Poroelastic Waves interface solves for the fluid pore pressure
variable pf instead of the fluid displacement field w.

The second row in Equation 3-3 is simplified to

1
w = ---------------------
- ( ∇p f – ρ f ω 2 u )
ω 2 ρc ( ω )

so the first row in Equation 3-3 becomes

THEORY FOR THE POROELASTIC WAVES INTERFACES | 367

 ρf
– ρ av ω 2 u – -------------- ( ∇p f – ρ f ω 2 u ) – ∇ ⋅ σ = 0 (3-5)
ρc ( ω )

The total stress tensor σ is then divided into the contributions from the elastic porous
(drained) matrix and from the pore fluid

σ ( u, p f ) = σ d ( u ) – α B p f I

here, the identity tensor I means that the pore pressure pf only contributes to the
diagonal of the total stress tensor σ. The parameter αB is the so-called Biot-Willis
coefficient. The drained, elastic stress tensor is written as σd = c:ε when ε is the strain
tensor of the porous matrix, and the elasticity tensor c contains the drained porous
matrix’s elastic properties (see the Linear Elastic Material feature in the Structural
Mechanics Module User’s Guide).

Finally, arrange Equation 3-5 in terms of the variables u and p:

 ρ f2  ρf
–  ρ av – -------------- ω 2 u – ∇ ⋅ ( σ d ( u ) – α B p f I ) = -------------- ∇p f (3-6)
 ρ c ( ω ) ρ c(ω)

The next Biot’s equation comes from taking the divergence of the second row in
Equation 3-3, previously divided by −ρc(ω)

ρf
ω 2 ∇ ⋅  -------------- u + ω 2 ∇ ⋅ w + ∇ ⋅  – -------------- ∇p f = 0
1
(3-7)
ρc ( ω )  ρ c ( ω )

Using the expressions for the volumetric strain εvol = ∇·u and fluid displacement
(Ref. 3, Ref. 4),

pf
– ∇ ⋅ w = ----- + α B ε vol
M

Biot’s modulus M is calculated from the porosity εp, fluid compressibility χf,
Biot-Willis coefficient αB and the drained bulk modulus of the porous matrix Kd

1- αB – εp
---- = ε p χ f + ------------------ ( 1 – α B ) (3-8)
M Kd

so Equation 3-7 simplifies to

ρf
ω 2 ∇ ⋅  -------------- u – ω 2  ----- p f + α B ε vol + ∇ ⋅  – -------------- ∇p f = 0
1 1
(3-9)
 ρc ( ω )  M   ρ c ( ω )

368 | CHAPTER 3: ELASTIC WAVES INTERFACES

and Biot’s wave equations (Equation 3-6 and Equation 3-9) can be written in terms
of the variable u and pf as

 ρ f2  ρf
– ω 2  ρ av – -------------- u – ∇ ⋅ ( σ d ( u ) – α B p f I ) = -------------- ∇p f
 ρ c ( ω ) ρc ( ω )
(3-10)
ω2 1
– ------- p f + ∇ ⋅ – -------------- ( ∇p f – ω 2 ρ f u ) = ω 2 α B ε vol
M ρc ( ω )

The saturated (also called Gassmann) modulus can be obtained from the
drained bulk modulus Kd, Biot modulus M, and Biot-Willis coefficient
αB as Ksat = Kd + αB2M (Ref. 5).

Further arranging the first row in Equation 3-10 to fit the formulation in the Elastic
Waves interface (Equation 3-1) gives

 ρ f2 
– ω 2  ρ av – -------------- u – ∇ ⋅ ( σ d ( u ) – s 0 ) = F (3-11)
 ρ c ( ω )

The body load F depends on the angular frequency and the gradient of fluid pressure
and the fluid pressure acts as a spherical contribution to the diagonal of Cauchy stress
tensor

ρf
F = -------------- ∇p f
ρc ( ω )
s0 = αB pf I

Arranging the second row in Equation 3-10 to fit the implementation of the Pressure
Acoustics, Frequency Domain interface gives (see Theory Background for the Pressure
Acoustics Branch)

ω2
– ------- p f + ∇ ⋅ –  -------------- ( ∇p f – q d ) = Q m
1
(3-12)
M ρc ( ω )

The monopole domain source Qm (SI unit: 1/s2) and the dipole domain source qd
(SI unit: N/m3) depend on the angular frequency ω, the displacement of the porous
matrix u, the fluid density and Biot–Willis coefficient αB

THEORY FOR THE POROELASTIC WAVES INTERFACES | 369

 Q m = ω 2 α B ε vol
qd = ω2 ρf u

BIOT-ALLARD MODEL (VISCOUS AND THERMAL LOSSES)

When both thermal and viscous losses are included, the viscosity in Equation 3-4 and
the fluid compressibility in Equation 3-8 are replaced by frequency-dependent
expressions. The losses due to viscosity are considered by the viscosity expression and
the losses due to thermal conduction by the fluid compressibility expression, see
Ref. 9.

The frequency-dependent complex viscosity is given by

1
2 ---
 4iωτ ∞ μρ f 2
μ ( ω ) = μ  1 + -------------------------
2 2 2

 Rf Lv εp 

where the viscous characteristic length Lv has been introduced (it is sometimes
referred to as Λ). The frequency-dependent complex fluid compressibility is given by

–1
εp  iωL th Prρ f
2
χ f ( ω ) = --------- γ – ( γ – 1 )  1 + -------------------------
8μ -
1 + --------------------------
γp A 2
iωL th Prρ f 16μ 
 

where the thermal characteristic length Lth has been introduced (it is sometimes
referred to as Λ’). The two expressions can be recognized in the JCA equivalent fluid
model (Johnson-Champoux-Allard (JCA)) available in Pressure Acoustics.

Different loss models or formulations for the frequency-dependent

viscosity and fluid compressibility can be entered manually. In order to do
so, select the Biot (viscous losses) model and then set the fluid
Compressibility and the fluid Dynamic viscosity to User defined. In these
two fields, enter the desired model expression. It can, for example,
depend on the frequency, using the variable freq.

Boundary Conditions for Poroelastic Waves

Although boundary conditions can be set up for the porous matrix and fluid
independently of each other, there exist a few common boundary conditions which

370 | CHAPTER 3: ELASTIC WAVES INTERFACES

deserve special attention. The following sections refer to the boundary conditions for
the system written in Equation 3-11 and Equation 3-12. See derivation in Ref. 7,
Ref. 8, and Ref. 9.

POROUS, FREE (SOUND-SOFT BOUNDARY) THEORY

The Porous, Free boundary condition is the default for the porous matrix. It means
that the displacement of the porous matrix in Equation 3-11 is unconstrained, so it can
move freely without experiencing any loads.

The sound soft boundary condition for acoustics creates a boundary condition for
Equation 3-12 where the acoustic pressure vanishes, so it sets pf = 0.

FIXED (SOUND-HARD BOUNDARY) EQUATIONS

For simulating a poroelastic medium bounded by a rigid impervious wall, impose a
Fixed Constraint node for the porous matrix displacement in Equation 3-11, u = 0,
and a sound-hard boundary condition for the pore pressure in Equation 3-12:

n ⋅  -------------- ( ∇p f – ρ f ω 2 u ) = 0
1
ρc ( ω )

PRESSURE EQUATIONS
For a given fluid pressure p0 on the boundary, set the pressure in Equation 3-12 to
p f = p 0 . Since the fluid pressure is set to p0, the normal stress on the porous matrix
in Equation 3-11 reduces to

n ⋅ σ d ( u ) = n ( α B – 1 )p 0

For a rigid porous matrix αB = εp, the load is equivalent to

n ⋅ σ d ( u ) = n ( ε p – 1 )p 0

and for a soft porous matrix αB = 1, there is no load since

n ⋅ σd ( u ) = 0

PRESCRIBED DISPLACEMENT EQUATIONS

For a prescribed displacement u0 at the boundary, set the displacement of the porous
matrix in Equation 3-11 as u = u0 and assume a sound-hard (impervious) boundary
for the fluid pressure in Equation 3-12:

THEORY FOR THE POROELASTIC WAVES INTERFACES | 371

 n ⋅  -------------- ( ∇p f – ρ f ω 2 u ) = 0
1
 ρc ( ω ) 

PRESCRIBED VELOCITY EQUATIONS

For a prescribed velocity v0 at the boundary, set the displacement of the porous matrix
in Equation 3-11 as

1
u = ------ v 0
iω

and assume a sound-hard (impervious) boundary condition for the fluid pressure in
Equation 3-12

n ⋅  -------------- ( ∇p f – ρ f ω 2 u ) = 0
1
 ρc ( ω ) 

PRESCRIBED ACCELERATION EQUATIONS

For a prescribed acceleration a0 at the boundary, set the displacement of the porous
matrix in Equation 3-11 as

1
u = ---------2- a 0
–ω

and assume a sound-hard (impervious) boundary condition for the fluid pressure in
Equation 3-12

n ⋅  -------------- ( ∇p f – ρ f ω 2 u ) = 0
1
 ρc ( ω ) 

ROLLER EQUATIONS
The roller, or sliding wall boundary, means that the boundary is impervious
(sound-hard) to fluid displacements, but it allows tangential displacements of the
porous matrix.

The normal displacement of the porous matrix in Equation 3-11 is constrained, but
the porous matrix is free to move in the tangential direction

n⋅u = 0

The impervious (sound hard) boundary condition for the fluid pressure in
Equation 3-12 is obtained from

372 | CHAPTER 3: ELASTIC WAVES INTERFACES

 n ⋅  -------------- ( ∇p f – ρ f ω 2 u ) = 0
1
 ρc ( ω ) 

SEPTUM BOUNDARY LOAD EQUATIONS

For a prescribed load FA at the boundary, suppose that one side of the septum is fixed
to the porous matrix and the other side bears the load.

A septum is a very limp and thin impervious layer with surface density ρsep. Since the
septum can be seen as a boundary mass density, this boundary condition is achieved by
setting an effective load FS = FA + ρsepω2u on the porous matrix, so the normal stress
in Equation 3-11 reduces to

n ⋅ ( σd ( u ) – αB pf I ) = FS

and a sound-hard (impervious) boundary condition is applied for the fluid pressure in
Equation 3-12

n ⋅  -------------- ( ∇p f – ρ f ω 2 u ) = 0
1
 ρc ( ω ) 

Postprocessing Variables
Other well known formulations of the poroelastic waves equations exist which use
other parameters and variables in their definition (see Ref. 10). To help users that are
familiar with these formulations, some of the most common parameters have been
added as postprocessing variables under the Biot-Allard coefficients menu (when
clicking the Replace Expressions button in plots). The variables included are:
TABLE 3-2: BIOT-ALLARD COEFFICIENTS

COMSOL NAME DEFINITIONS NAME

pelw.rho11 ρ 11 = ρ d – ρ 12 Mass coefficient (11)

pelw.rho22 ρ 22 = ε p ρ f – ρ 12 Mass coefficient (22)
pelw.rho12 ρ 12 = – ρ a = – ε p ρ f ( τ ∞ – 1 ) Inertial interaction mass
coefficient (12)
pelw.rho11_c ρ̃ 11 = ρ 11 + b˜ ( ω ) ⁄ ( iω ) Complex mass coefficient (11)
pelw.rho22_c ρ̃ = ρ + b˜ ( ω ) ⁄ ( iω )
22 22 Complex mass coefficient (11)
pelw.rho12_c ρ̃ 12 = ρ 12 – b˜ ( ω ) ⁄ ( iω ) Complex inertial interaction
mass coefficient (12)
pelw.A_c See Ref. 10 Complex elastic coefficient

THEORY FOR THE POROELASTIC WAVES INTERFACES | 373

 TABLE 3-2: BIOT-ALLARD COEFFICIENTS

COMSOL NAME DEFINITIONS NAME

pelw.Q_c See Ref. 10 Complex elastic coupling

coefficient
pelw.R_c See Ref. 10 Complex bulk modulus
coefficient
pelw.gamma_c See Ref. 10 Diagonal stress coupling
coefficient

References for the Poroelastic Waves Interfaces

1. M.A. Biot, “Theory of Propagation of Elastic Waves in a Fluid-saturated Porous
Solid. I. Low-frequency Range”, J. Acoust. Soc. Am., vol. 28, pp 168–178, 1956.

2. M.A. Biot, “Theory of Propagation of Elastic Waves in a Fluid-saturated Porous

Solid. II. Higher Frequency Range”, J. Acoust. Soc. Am., vol. 28, pp 179–191, 1956.

3. M.A. Biot, “Generalized Theory of Acoustic Propagation in Porous Dissipative

Media”, J. Acoust. Soc. Am., vol. 34, pp 1254–1264, 1962.

4. M.A. Biot, “Mechanics of Deformation and Acoustic Propagation in Porous

Media”, J. Appl. Phys., vol. 33, pp 1482–1498, 1962.

5. G. Mavko et al., The Rock Physics Handbook, 2nd ed., Cambridge University Press,
2009.

6. J.M. Carcione, Wave Fields in Real Media: Wave Propagation in Anisotropic,

Anelastic, Porous and Electromagnetic Media, 2nd ed. Elsevier (Handbook of
Geophysical Exploration, vol. 38, Seismic Exploration), 2007.

7. P. Debergue, R. Panneton, and N. Atalla, “Boundary Conditions for the Weak

Formulation of the Mixed (u,p) Poroelasticity Problem”, J. Acoust. Soc. Am.,
vol. 106, pp 2383–2390, 1999.

8. N. Atalla, M.A. Hamdi, and R. Panneton, “Enhanced Weak Integral Formulation

for the Mixed (u,p) Poroelastic Equations”, J. Acoust. Soc. Am., vol. 109, pp 3065–
3068, 2001.

9. J.F. Allard and N. Atalla, Propagation of Sound in Porous Media, 2nd ed., John
Wiley & Sons, 2009.

10. N. Atalla, R. Panneton, and P. Debergue, “A mixed displacement-pressure

formulation for poroelastic materials”, J. Acoust. Soc. Am., vol. 104, pp 1444, 1998.

374 | CHAPTER 3: ELASTIC WAVES INTERFACES

11. N. Atalla, F. Sgard, and C.K. Amedin, “On the modeling of sound radiation from
poroelastic materials”, J. Acoust. Soc. Am., vol. 120, pp 1990, 2006.

THEORY FOR THE POROELASTIC WAVES INTERFACES | 375

 Theory for the Elastic Waves, Time
Explicit Interface
The Elastic Waves, Time Explicit Interface theory is described in this section:

• Governing Equations
• References for the Elastic Waves, Time Explicit Interface

Governing Equations
The model solves the governing equations for a general linear elastic material in a
velocity-strain formulation

∂v
ρ ------ – ∇ ⋅ S = F v
∂t
∂ε
----- – 1
--- [ ∇v + ( ∇v ) T ] = 0
∂t 2
S = C:ε

where v is the velocity, ρ the density, S the stress tensor, ε the strain tensor, C is the
elasticity tensor (or stiffness tensor), and Fv is a possible body force. The equations are
valid for both isotropic and anisotropic material data.

Independent of the space dimension, the unknowns solved for are three components
of the velocity vector and six components of the strain tensor. This means that the
out-of-plane component of the velocity will be present in 2D geometries. In some
cases, for example for anisotropic materials, the out-of-plane velocity may become
different from zero for in-plane loads. This formulation is ofter referred to as 2.5D.

References for the Elastic Waves, Time Explicit Interface

1. J.S. Hesthaven and T. Warburton, Nodal Discontinuous Galerkin Methods:
Algorithms, Analysis, and Applications, Springer, 2008.

376 | CHAPTER 3: ELASTIC WAVES INTERFACES

 4

Acoustic-Structure Interaction
Interfaces
This chapter describes the physics interfaces found under the Acoustic-Structure
Interaction branch ( ).

• The Acoustic-Solid Interaction, Frequency Domain Interface

• The Acoustic-Solid Interaction, Transient Interface
• The Acoustic-Shell Interaction, Frequency Domain Interface
• The Acoustic-Shell Interaction, Transient Interface
• The Acoustic-Piezoelectric Interaction, Frequency Domain Interface
• The Acoustic-Piezoelectric Interaction, Transient Interface
• The Acoustic-Poroelastic Waves Interaction Interface
• The Acoustic-Solid-Poroelastic Waves Interaction Interface
• The Acoustic-Solid Interaction, Time Explicit Interface
• Modeling with the Acoustic-Structure Interaction Branch

377
 The Acoustic-Solid Interaction,
Frequency Domain Interface
The Acoustic-Solid Interaction, Frequency Domain multiphysics interface ( ), found
under the Acoustics>Acoustic-Structure Interaction branch ( ) when adding a physics
interface, combines the Pressure Acoustics, Frequency Domain and Solid Mechanics
interfaces to connect the acoustics pressure variations in the fluid domain with the
structural deformation in the solid domain. It can, for example, be used to determine
the transmission of sound through an elastic structure or solve for the coupled
vibroacoustics phenomena present in a loudspeaker.

Acoustic-structure interaction or ASI refers to a multiphysics phenomenon where the

acoustic pressure causes a fluid load on the solid domain, and the structural
acceleration acts on the fluid domain as a normal acceleration across the fluid-solid
boundary.

A dedicated multiphysics coupling condition is readily defined for the fluid-solid

boundary and sets up the fluid loads on the solid domain and the effect of the
structural accelerations on the fluid.

The physics interface is available for 3D, 2D, and 2D axisymmetric geometries and has
the capability to model pressure acoustics and solid mechanics in the frequency
domain, including a special acoustic-solid boundary condition for the fluid-solid
interaction.

When a predefined Acoustic-Solid Interaction, Frequency Domain interface is added from

the Acoustics>Acoustic-Structure Interaction branch of the Model Wizard or the Add
Physics windows, the Pressure Acoustics, Frequency Domain and Solid Mechanics
interfaces are added to the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Structure Boundary multiphysics coupling.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

378 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS
When a predefined multiphysics interface is used, for example Acoustic-Solid
Interaction, Frequency Domain, the Selection on the multiphysics coupling is
automatically set to All boundaries. In this way, the multiphysics coupling is
automatically active on all boundaries with Pressure Acoustics, Frequency Domain on one
side and Solid Mechanics on the other.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL Multiphysics adds an empty
Multiphysics Couplings node. You can choose the available multiphysics couplings, but
you need manually to select on which boundaries they need to be applied, or select All
boundaries to recover the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
The Acoustic-Structure Boundary multiphysics coupling is described in the
Multiphysics Couplings chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE ACOUSTIC-SOLID INTERACTION, FREQUENCY DOMAIN INTERFACE | 379

 • The available physics features for The Pressure Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, Edge, Point, and Pair Nodes
for the Pressure Acoustics, Frequency Domain Interface.
• The available physics features for The Solid Mechanics Interface are listed in the
Structural Mechanics Module User’s Guide.

• Theory section in the Structural Mechanics Module User’s Guide

• Theory Background for the Pressure Acoustics Branch

Acoustic-Structure Interaction: Application Library path

Acoustics_Module/Tutorials,_Pressure_Acoustics/acoustic_structure

Suggestions for setting up the solvers for solving large acoustic-structure

interaction problems are given in Solving Large Acoustic-Structure
Interaction Models in the Modeling with the Acoustic-Structure
Interaction Branch section.

380 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

The Acoustic-Solid Interaction,
T r a ns i e nt Interface
The Acoustic-Solid Interaction, Transient multiphysics interface ( ), found under the
Acoustics>Acoustic-Structure Interaction branch ( ) when adding a physics interface,
combines the Pressure Acoustics, Transient and Solid Mechanics interfaces to connect
the acoustics pressure variations in the fluid domain with the structural deformation in
the solid domain. It can, for example, be used to determine the transmission of sound
through an elastic structure or solve for the coupled vibroacoustics phenomena present
in a loudspeaker.

Acoustic-structure interaction or ASI refers to a multiphysics phenomenon where the

acoustic pressure causes a fluid load on the solid domain, and the structural
acceleration acts on the fluid domain as a normal acceleration across the fluid-solid
boundary.

Special physics interface conditions are readily defined at the fluid-solid boundary and
set up the fluid loads on the solid domain and the effect of the structural accelerations
on the fluid. The physics interface allows for transient studies but also solves in the
frequency domain with the available boundary conditions.

When a predefined Acoustic-Solid Interaction, Transient interface is added from the

Acoustics>Acoustic-Structure Interaction branch of the Model Wizard or the Add Physics
windows, the Pressure Acoustics, Transient and Solid Mechanics interfaces are added to
the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Structure Boundary multiphysics coupling.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

For modeling of acoustic-structure interaction in the frequency domain,

The Acoustic-Solid Interaction, Frequency Domain Interface contains
additional functionality that is not applicable for modeling in the time
domain.

THE ACOUSTIC-SOLID INTERACTION, TRANSIENT INTERFACE | 381

 • Theory section in the Structural Mechanics Module User’s Guide
• Theory Background for the Pressure Acoustics Branch

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When the predefined multiphysics interface is used, the Selection on the multiphysics
coupling is automatically set to All boundaries. In this way, the multiphysics coupling is
automatically active on all boundaries with Pressure Acoustics, Transient on one side and
Solid Mechanics on the other.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL Multiphysics adds an empty
Multiphysics Couplings node. You can choose the available multiphysics couplings, but
you need manually to select on which boundaries they need to be applied, or select All
boundaries to recover the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
The Acoustic-Structure Boundary multiphysics coupling is described in the
Multiphysics Couplings chapter.

382 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

Physics Interface Features
Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Pressure Acoustics, Transient Interface are
listed in the section Domain, Boundary, Edge, and Point Nodes for the Pressure
Acoustics, Transient Interface.
• The available physics features for The Solid Mechanics Interface are listed in the
Structural Mechanics Module User’s Guide.

• Theory section in the Structural Mechanics User’s Guide

• Theory Background for the Pressure Acoustics Branch

THE ACOUSTIC-SOLID INTERACTION, TRANSIENT INTERFACE | 383

 The Acoustic-Piezoelectric
Interaction, Frequency Domain
Interface
The Acoustic-Piezoelectric Interaction, Frequency Domain multiphysics interface ( ),
found when adding a physics interface under the Acoustics>Acoustic-Structure
Interaction branch ( ), combines the Pressure Acoustics, Frequency Domain, Solid
Mechanics, and Electrostatics interfaces to connect and solve for the acoustic pressure
variations in fluids with the structural deformation in both solids and piezoelectric
solid domains. The physics interface also includes features from Electrostatics to solve
for the electric field in the piezoelectric material. It may be used for modeling
piezoelectric transducers for sonar or medical applications and, for example, enhancing
the impedance matching layers as well as the far-field radiation patterns of the
transducer.

The Helmholtz equation is solved in the fluid domain and the structural equations in
the solid together with the constitutive relationships required to model piezoelectrics.
Both the direct and inverse piezoelectric effects can be modeled, and the piezoelectric
coupling can be formulated using the strain-charge or stress-charge forms.

When a predefined Acoustic-Piezoelectric Interaction, Frequency Domain interface is

added from the Acoustics>Acoustic-Structure Interaction branch ( ) of the Model
Wizard or the Add Physics windows, the Pressure Acoustics, Frequency Domain, Solid
Mechanics, and Electrostatics interfaces are added to the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Structure Boundary and Piezoelectric Effect multiphysics couplings.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

The equations solved in the solid and fluid domains can be found in the
Structural Mechanics Module User’s Guide and in Theory Background
for the Pressure Acoustics Branch, respectively.

384 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

 Piezoacoustic Transducer: Application Library path Acoustics_Module/
Piezoelectric_Devices/piezoacoustic_transducer

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When the predefined multiphysics interface is used, the Selection on the
Acoustic-Structure Boundary multiphysics coupling is automatically set to All boundaries.
In this way, the multiphysics coupling is automatically active on all boundaries with
Pressure Acoustics, Frequency Domain on one side and Solid Mechanics on the other.
Similarly, the Selection on the Piezoelectric Effect multiphysics coupling is automatically
set to All domains and active in all valid domains.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL Multiphysics adds an empty
Multiphysics Couplings node. You can choose the available multiphysics coupling, but
you need manually to select on which boundaries or domains they need to be applied,
or select All boundaries or All domains to recover the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Couplings
The Acoustic-Structure Boundary multiphysics coupling is described in the
Multiphysics Couplings chapter. The Piezoelectric Effect is described for The
Piezoelectricity Interface.

THE ACOUSTIC-PIEZOELECTRIC INTERACTION, FREQUENCY DOMAIN INTERFACE | 385

 Physics Interface Features
Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Pressure Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, Edge, Point, and Pair Nodes
for the Pressure Acoustics, Frequency Domain Interface.
• The available physics features for The Solid Mechanics Interface that are described
in the Structural Mechanics Module User’s Guide.
• The available physics features for The Electrostatics Interface are listed in the section
Domain, Boundary, Edge, Point, and Pair Nodes for the Electrostatics Interface in
the COMSOL Multiphysics Reference Manual.
• Some additional piezoelectric features for the Solid Mechanics and Electrostatics
interfaces are described for The Piezoelectricity Interface.

• Theory section in the Structural Mechanics Module User’s Guide

• Theory of Electrostatics in the COMSOL Multiphysics Reference
Manual
• Theory Background for the Pressure Acoustics Branch

386 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

The Acoustic-Piezoelectric
Interaction, Transient Interface
The Acoustic-Piezoelectric Interaction, Transient multiphysics interface ( ), found
when adding a physics interface under the Acoustics>Acoustic-Structure Interaction
branch ( ), combines the Pressure Acoustics, Transient and Piezoelectricity
interfaces to connect and solve for the acoustic pressure variations in fluids with the
structural deformation in both solids and piezoelectric solid domains. The physics
interface also includes features from Electrostatics to solve for the electric field in the
piezoelectric material. Examples include modeling piezoelectric transducers for sonar
or medical applications and, for example, enhancing the impedance matching layers.

The pressure wave equation is solved in the fluid domain and the structural dynamic
equations in the solid together with the constitutive relationships required to model
piezoelectrics. Both the direct and inverse piezoelectric effects can be modeled, and the
piezoelectric coupling can be formulated using the strain-charge or stress-charge
forms.

When a predefined Acoustic-Piezoelectric Interaction, Transient interface is added from

the Acoustics>Acoustic-Structure Interaction branch of the Model Wizard or the Add
Physics windows, the Pressure Acoustics, Frequency Domain, Solid Mechanics, and
Electrostatics interfaces are added to the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Structure Boundary and Piezoelectric Effect multiphysics couplings.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When the predefined multiphysics interface is used, the Selection on the
Acoustic-Structure Boundary multiphysics coupling is automatically set to All boundaries.
In this way, the multiphysics coupling is automatically active on all boundaries with
Pressure Acoustics, Frequency Domain on one side and Solid Mechanics on the other.
Similarly, the Selection on the Piezoelectric Effect multiphysics coupling is automatically
set to All domains and active in all valid domains.

THE ACOUSTIC-PIEZOELECTRIC INTERACTION, TRANSIENT INTERFACE | 387

 However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL Multiphysics adds an empty
Multiphysics Couplings node. You can choose the available multiphysics couplings, but
you need manually to select on which boundaries or domains they need to be applied,
or select All boundaries or All domains to recover the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Couplings
The Acoustic-Structure Boundary multiphysics coupling is described in the
Multiphysics Couplings chapter. The Piezoelectric Effect is shortly described in The
Piezoelectricity Interface.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Pressure Acoustics, Transient Interface are
listed in the section Domain, Boundary, Edge, and Point Nodes for the Pressure
Acoustics, Transient Interface.
• The available physics features for The Solid Mechanics Interface are listed in the
Structural Mechanics Module User’s Guide.

388 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

• The available physics features for The Electrostatics Interface are listed in the section
Domain, Boundary, Edge, Point, and Pair Nodes for the Electrostatics Interface in
the COMSOL Multiphysics Reference Manual.
• Some additional piezoelectric features for the Solid Mechanics and Electrostatics
interfaces are described for The Piezoelectricity Interface.

• Theory section in the Structural Mechanics Module User’s Guide

• Theory of Electrostatics in the COMSOL Multiphysics Reference
Manual
• Theory Background for the Pressure Acoustics Branch

THE ACOUSTIC-PIEZOELECTRIC INTERACTION, TRANSIENT INTERFACE | 389

 T he Ac o us t i c - Poroel ast i c W av es
Interaction Interface
The Acoustic-Poroelastic Waves Interaction multiphysics interface ( ) combines
Pressure Acoustics, Frequency Domain and Poroelastic Waves together with the
Acoustic-Porous Boundary multiphysics coupling.

When a predefined Acoustic-Poroelastic Waves Interaction interface is added from the

Acoustics>Acoustic-Structure Interaction branch ( ) of the Model Wizard or the Add
Physics windows, the Poroelastic Waves and Pressure Acoustics, Frequency Domain
interfaces are added to the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Porous Boundary multiphysics coupling.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When the predefined multiphysics interface is used, for example Acoustic-Poroelastic
Waves Interaction, the Selection on the multiphysics coupling is automatically set to All
boundaries. In this way the multiphysics coupling is automatically active on all
boundaries with Pressure Acoustics, Frequency Domain on one side and Poroelastic Waves
on the other.

However, if physics interfaces are added one at a time, followed by the multiphysics
coupling, these modified settings are not automatically included.

For example, if single Poroelastic Waves and Pressure Acoustics, Frequency Domain
interfaces are added, COMSOL Multiphysics adds an empty Multiphysics Couplings
node. You can choose the available multiphysics couplings, but you need manually to

390 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

select on which boundaries they need to be applied, or select All boundaries to recover
the predefined behavior

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
The Acoustic-Porous Boundary multiphysics coupling is described in the Multiphysics
Couplings chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Poroelastic Waves Interface are listed in the
section Domain, Boundary, and Pair Nodes for the Poroelastic Waves Interfaces.
• The available physics features for The Pressure Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, Edge, Point, and Pair Nodes
for the Pressure Acoustics, Frequency Domain Interface.

THE ACOUSTIC-POROELASTIC WAVES INTERACTION INTERFACE | 391

 The Acoustic-Solid-Poroelastic Waves
Interaction Interface
The Acoustic-Solid-Poroelastic Waves Interaction multiphysics interface ( ) combines
Pressure Acoustics, Frequency Domain, Solid Mechanics, and Poroelastic Waves
together with the Acoustic-Structure Boundary, Porous-Structure Boundary, and
Acoustic-Porous Boundary multiphysics couplings.

When a predefined Acoustic-Solid-Poroelastic Waves Interaction interface is added from

the Acoustics>Acoustic-Structure Interaction branch ( ) of the Model Wizard or the
Add Physics windows, the Pressure Acoustics, Frequency Domain, Solid Mechanics, and
Poroelastic Waves interfaces are added to the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Porous Boundary, Porous-Structure Boundary, and Acoustic-Structure Boundary
multiphysics couplings.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When a predefined multiphysics interface is used, for example Acoustic-Solid-Poroelastic
Waves Interaction, the Selection on the multiphysics couplings is automatically set to All
boundaries. In this way, the relevant multiphysics couplings are automatically active on
all relevant boundaries connecting two physics.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if single Poroelastic Waves and Pressure Acoustics, Frequency Domain
interfaces are added, COMSOL Multiphysics adds an empty Multiphysics Couplings
node. You can choose the available multiphysics couplings, but you need manually to

392 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

select on which boundaries they need to be applied, or select All boundaries to recover
the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Couplings
The Acoustic-Porous Boundary, Porous-Structure Boundary, and Acoustic-Structure
Boundary multiphysics couplings are described in the Multiphysics Couplings chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Poroelastic Waves Interface are listed in the
section Domain, Boundary, and Pair Nodes for the Poroelastic Waves Interfaces.
• The available physics features for The Solid Mechanics Interface are listed in the
Structural Mechanics Module User’s Guide.
• The available physics features for The Pressure Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, Edge, Point, and Pair Nodes
for the Pressure Acoustics, Frequency Domain Interface.

THE ACOUSTIC-SOLID-POROELASTIC WAVES INTERACTION INTERFACE | 393

 The Acoustic-Solid Interaction, Time
Explicit Interface
The Acoustic-Solid Interaction, Time Explicit multiphysics interface ( ) combines
Pressure Acoustics, Time Explicit and Elastic Waves, Time Explicit together with the
Acoustic-Structure Boundary, Time Explicit multiphysics coupling. The physics
interface is available for 3D and 2D geometries.

Acoustic-structure interaction or ASI refers to a multiphysics phenomenon where the

acoustic pressure causes a fluid load on the solid domain, and the structural
acceleration acts on the fluid domain as a normal acceleration across the fluid-solid
boundary. The Acoustic-Solid Interaction, Time Explicit interface can, for example, be
used to determine the transmission of sound through an elastic structure or solve for
the coupled vibroacoustics phenomena present in an ultrasonic transducer or
flowmeter.

The discontinuous Galerkin time explicit method enables efficient computation of

acoustically large models (containing many wavelengths) in the time domain. The
formulation of the governing equations, effectively requires the acoustics pressure and
velocity variations in the fluid domain to be coupled with the structural velocity and
strain variations in the solid domain.

A dedicated multiphysics coupling condition is readily defined for the fluid-solid

boundary and sets up the continuity in velocity and normal stress.

When a predefined Acoustic-Solid Interaction, Time Explicit interface is added from the
Acoustics>Acoustic-Structure Interaction branch of the Model Wizard or the Add Physics
windows, the Pressure Acoustics, Time Explicit and Elastic Waves, Time Explicit interfaces
are added to the Model Builder. In addition, the Multiphysics Couplings node is added,
which automatically includes the Acoustic-Structure Boundary, Time Explicit
multiphysics coupling.

See The Multiphysics Branch and Multiphysics Modeling Workflow in the

COMSOL Multiphysics Reference Manual.

394 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS
When a predefined multiphysics interface is used, for example Acoustic-Structure
Boundary, Time Explicit, the Selection on the multiphysics coupling is automatically set
to All boundaries. In this way, the multiphysics coupling is automatically active on all
boundaries with Pressure Acoustics, Time Explicit on one side and Elastic Waves, Time
Explicit on the other.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL Multiphysics adds an empty
Multiphysics Couplings node. You can choose the available multiphysics couplings, but
you need manually to select on which boundaries they need to be applied, or select All
boundaries to recover the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Couplings
The Acoustic-Structure Boundary, Time Explicit multiphysics coupling is described in
the Multiphysics Couplings chapter. Note that a pair version of the multiphysics
coupling exists that can be used for assemblies: Pair Acoustic-Structure Boundary,
Time Explicit.

THE ACOUSTIC-SOLID INTERACTION, TIME EXPLICIT INTERFACE | 395

 Physics Interface Features
Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Pressure Acoustics, Time Explicit Interface are
listed in the section Domain, Boundary, Edge, and Point Nodes for the Pressure
Acoustics, Time Explicit Interface.
• The available physics features for The Elastic Waves, Time Explicit Interface are
listed in the section Domain, Boundary, Edge, Point, and Pair Nodes for the Elastic
Waves, Time Explicit Interface.

• Theory Background for the Pressure Acoustics Branch

• Theory for the Elastic Waves, Time Explicit Interface

396 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

The Acoustic-Shell Interaction,
Frequency Domain Interface
This physics interface requires a Structural Mechanics Module license.
For theory and physics interface feature descriptions relating to the Shell
interface, see the Structural Mechanics Module User’s Guide.

The Acoustic-Shell Interaction, Frequency Domain multiphysics interface ( ), found

under the Acoustics>Acoustic-Structure Interaction branch ( ) when adding a physics
interface, combines features from the Pressure Acoustics, Frequency Domain and Shell
interfaces to connect the acoustics pressure variations in the fluid domain with the
structural deformation of a shell boundary. It may for example be used for determining
the transmission of sound through a thin elastic structure such as a car hood and
analyzing the vibroacoustics of a loudspeaker cone.

Acoustic-structure interaction or ASI refers to a multiphysics phenomenon where the

acoustic pressure causes a fluid load on the solid surface, and the structural acceleration
acts on the fluid domain as a normal acceleration across the fluid-structure boundary.

Special physics interface conditions are readily defined at the fluid-shell boundary and
set up the fluid loads on the shell boundary and the effect of the structural
accelerations on the fluid. The physics interface is available for 3D and 2D
axisymmetric geometries, and it is capable of modeling the coupled pressure acoustics
and shell vibrations in the frequency domain.

When a predefined Acoustic-Shell Interaction, Frequency Domain interface is added from

the Acoustics>Acoustic-Structure Interaction branch of the Model Wizard or the Add
Physics windows, the Pressure Acoustics, Frequency Domain and Shell interfaces are
added to the Model Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
multiphysics coupling Acoustic-Structure Boundary.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When a predefined multiphysics interface is used, for example Acoustic-Shell Interaction,
Frequency Domain, the Selection on the multiphysics coupling is automatically set to All
boundaries. In this way, the multiphysics coupling is automatically active on all Shell
boundaries with Pressure Acoustics, Frequency Domain on one side or both sides.

THE ACOUSTIC-SHELL INTERACTION, FREQUENCY DOMAIN INTERFACE | 397

 However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if single Shell and Pressure Acoustics, Frequency Domain interfaces are
added, COMSOL Multiphysics adds an empty Multiphysics Couplings node. You can
choose the available multiphysics couplings, but you need manually to select on which
boundaries they need to be applied, or select All boundaries to recover the predefined
behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
The Acoustic-Structure Boundary multiphysics coupling is described in the
Multiphysics Couplings chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Pressure Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, Edge, Point, and Pair Nodes
for the Pressure Acoustics, Frequency Domain Interface.
• The available physics features for Results Evaluation are listed in the section
Domain, Boundary, Edge, Point, and Pair Nodes for the Shell and Plate Interfaces
in the Structural Mechanics Module User’s Guide.

398 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

• Theory Background for the Pressure Acoustics Branch
• Theory for the Shell and Plate Interfaces in the Structural Mechanics
Module User’s Guide

THE ACOUSTIC-SHELL INTERACTION, FREQUENCY DOMAIN INTERFACE | 399

 T he Ac o us t i c - S h el l In t eract i on ,
Transient Interface
This physics interface requires a Structural Mechanics license. For theory
and physics interface feature descriptions relating to the Shell interface,
see the Structural Mechanics Module User’s Guide.

The Acoustic-Shell Interaction, Transient multiphysics interface ( ), found under the

Acoustics>Acoustic-Structure Interaction branch ( ) when adding a physics interface,
combines features from the Pressure Acoustics, Transient and Shell interfaces to
connect the acoustics pressure variations in the fluid domain with the structural
deformation of a shell boundary. It may for example be used for determining the
transmission of sound through a thin elastic structure such as a car hood and analyzing
the vibroacoustics of loudspeaker cone.

Acoustic-structure interaction refers to a multiphysics phenomenon where the acoustic

pressure causes a fluid load on the solid surface, and the structural acceleration affects
the fluid domain as a normal acceleration across the fluid-structure boundary.

Special physics interface conditions are readily defined at the fluid-shell boundary and
set up the fluid loads on the shell boundary and the effect of the structural
accelerations on the fluid. The physics interface is only available for 3D geometries, and
it is capable of modeling the coupled pressure acoustics and shell vibrations in the time
domain.

When a predefined Acoustic-Shell Interaction, Transient interface is added from the

Acoustics>Acoustic-Structure Interaction branch of the Model Wizard or the Add Physics
windows, the Pressure Acoustics, Transient and Shell interfaces are added to the Model
Builder.

In addition, the Multiphysics Couplings node is added, which automatically includes the
Acoustic-Structure Boundary multiphysics coupling.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When a predefined multiphysics interface is used, for example Acoustic-Shell Interaction,
Transient, the Selection on the multiphysics coupling is automatically set to All
boundaries. In this way, the multiphysics coupling is automatically active on all Shell
boundaries with Pressure Acoustics, Transient on one side or both sides.

400 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if single Shell and Pressure Acoustics, Transient interfaces are added,
COMSOL Multiphysics adds an empty Multiphysics Couplings node. You can choose
the available multiphysics couplings, but you need manually to select on which
boundaries they need to be applied, or select All boundaries to recover the predefined
behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics Couplings node) or from the Physics toolbar, Multiphysics
menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
The Acoustic-Structure Boundary multiphysics coupling is described in the
Multiphysics Couplings chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Pressure Acoustics, Transient Interface are
listed in the section Domain, Boundary, Edge, and Point Nodes for the Pressure
Acoustics, Transient Interface.
• The available physics features for Results Evaluation are listed in the section
Domain, Boundary, Edge, Point, and Pair Nodes for the Shell and Plate Interfaces
in the Structural Mechanics Module User’s Guide.

THE ACOUSTIC-SHELL INTERACTION, TRANSIENT INTERFACE | 401

 • Theory Background for the Pressure Acoustics Branch
• Theory for the Shell and Plate Interfaces in the Structural Mechanics
Module User’s Guide

402 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

Modeling with the Acoustic-Structure
Interaction Branch
In this section:

• Prestressed Acoustic-Structure Interaction

• Solving Large Acoustic-Structure Interaction Models
• Configuration of Perfectly Matched Layers (PMLs) for Acoustic-Structure
Interaction Models

Prestressed Acoustic-Structure Interaction

When modeling transducers like microphones or ultrasound horns there are often
parts of the structure, like the diaphragm or a bolt, that are under tension or
prestressed. In these cases, it is essential to include this effect when modeling the
transducer. A prestressed part will shift the resonance frequency of the mechanical
system and thus the overall vibroacoustic behavior will change.

• Piezoelectric Tonpilz Transducer with a Prestressed Bolt: Application

Library path Acoustics_Module/Piezoelectric_Devices/
tonpilz_transducer_prestressed.
• The Brüel & Kjær 4134 Condenser Microphone: Application Library
path Acoustics_Module/Electroacoustic_Transducers/
bk_4134_microphone.

Such prestressed acoustic-structure interaction models can be set up and solved fully
coupled including all effects. The procedure is as follows:

1 Set up the model including all relevant physics.

2 Couple the physics using the required multiphysics couplings.
3 All acoustic sources, structural loads, and any forcing that are nonstatic, should be
defined using the linper() operator. This will ensure that they are used only in the
frequency-domain part of the study (that one is set to linear perturbation).
4 Add a Prestressed Analysis, Frequency Domain study.

MODELING WITH THE ACOUSTIC-STRUCTURE INTERACTION BRANCH | 403

 5 Note that in the Stationary step the acoustics physics interfaces have an orange
warning triangle under the Physics and Variables Selection. This simply means that
they will not be used in the stationary study since acoustics is not supporting the
study type. Therefore the Solve for this field will automatically not be marked under
the Dependent Variables in the Solver Configurations.
6 Note also that the Include geometric nonlinearity box is checked in the
Frequency-Domain, Perturbation step. If it is not checked, the prestress effect is lost.
7 Solve the model for the desired frequencies.

A prestressed type of analysis can also be done on pure structural problems in the
frequency domain and when searching for eigenfrequencies of structures.

See also Prestressed Structures, Modeling Pretensioned Bolts, and

Mechanical Damping and Losses in the Structural Mechanics Module
User’s Guide

Solving Large Acoustic-Structure Interaction Models

In models that involve acoustic-structure interaction, the strategy for solving large
problems involves solving the system in a segregated way. That is, the system is not
solved fully coupled in one step but iterations are used solving one physics interface at
a time.

To set up such a solver right-click on the Stationary Solver step and select Segregated.
In the first Segregated Step solve for the structural dependent variables (displacement).
Set up a second segregated step where you select the pressure dependent variable.
Under each of the steps select the solver of choice for solving the single physics
interface problem. For example, in the case where only a small structural domain is
included, use an iterative multigrid approach for the acoustics (see Solving Large
Acoustics Problems Using Iterative Solvers) and a direct solver for the structure.

This strategy is only readily applicable when the coupling between the solid and the
acoustic domain is done via Neumann conditions; this is the case for all models where
the acoustic domain uses Pressure Acoustics. In, for example, models with
thermoviscous acoustic-structure interaction, the coupling is based on a Dirichlet
condition (a pointwise constraint) and required reformulating the continuity
condition using weak constraints. Nor does this approach work for models coupling

404 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

piezoelectric domains, structures, and acoustics; here a fully coupled approach is
necessary.

Studies and Solvers and Multigrid in the COMSOL Multiphysics

Reference Manual

Configuration of Perfectly Matched Layers (PMLs) for

Acoustic-Structure Interaction Models
Care should be taken when setting up perfectly matched layers (PMLs) in models that
include fluid domains and solid domains. In general models with several different
materials, that being different solids or different fluids.

Two configurations exist for the PMLs in these systems:

• The PML regions of two physics (or two different material models) are adjacent and
in contact. In this case, a single PML feature should be used for the two PML
domains. Set the Typical wavelength from to User defined and enter a value based on
the longest wavelength. For example, for an air-steel system enter: 5770[m/s]/freq
The shorter wavelengths should be captured by increasing the mesh resolution or
increasing the PML scaling curvature parameter. This will ensure optimal damping of
all wavelengths in the system. In any case, control the convergence of the results by
increasing the number of mesh layers in the PML.

• Infinite Elements, Perfectly Matched Layers, and Absorbing Layers in

the COMSOL Multiphysics Reference Manual.
• See Perfectly Matched Layers (PMLs) in Modeling with the Pressure
Acoustics Branch (FEM-Based Interfaces) in the Acoustics Module
User’s Guide.

• The PML regions of two physics (or two different material models) do not touch.
In this case, use one PML feature for each PML domain. Keep the default setting of
Typical wavelength from to Physics interface and select the relevant physics interface
under Physics.
The reason that two PML features cannot be used when the PML domains a adjacent
to each other is that the scaling inside the PMLs will differ (because of different speed

MODELING WITH THE ACOUSTIC-STRUCTURE INTERACTION BRANCH | 405

 of sounds). This will lead to an unphysical situation at the interface between the two
PMLs which results in errors in the solution. Spurious waves may be generated.

See the Acoustic-Structure Interaction with a Perfectly Matched Layer

(PML) model in the Application Gallery on the COMSOL website. It is
located here: [Link]
acoustic-structure-interaction-with-a-perfectly-matched-layer-pml-2352
1

406 | CHAPTER 4: ACOUSTIC-STRUCTURE INTERACTION INTERFACES

 5

Aeroacoustics Interfaces

This chapter describes the physics interfaces found under the Aeroacoustics
branch .

The aeroacoustic branch has physics interfaces to solve the fully linearized acoustic
equation for several physical conditions, they are: the linearized Euler (LE)
equations, the linearized Navier–Stokes (LNS) equations, and the linearized
potential flow (LPF) equations. These equations are also referred to as convected
acoustics. Flow-induced noise functionality using the Lighthill analogy exists in
pressure acoustics when using the Aeroacoustic Flow Source feature, associated
multiphysics coupling, and dedicated study.

In this chapter:

• The Linearized Euler, Frequency Domain Interface

• The Linearized Euler, Transient Interface
• The Linearized Euler, Boundary Mode Interface
• The Linearized Navier–Stokes, Frequency Domain Interface
• The Linearized Navier–Stokes, Transient Interface
• The Linearized Navier–Stokes, Boundary Mode Interface
• The Linearized Potential Flow, Frequency Domain Interface

407
 • The Linearized Potential Flow, Transient Interface
• The Linearized Potential Flow, Boundary Mode Interface
• The Compressible Potential Flow Interface
• Modeling with the Aeroacoustics Branch
• Theory Background for the Aeroacoustics Branch
• References for the Aeroacoustics Branch Interfaces

408 | CHAPTER 5: AEROACOUSTICS INTERFACES

The Linearized Euler, Frequency
Domain Interface
The Linearized Euler, Frequency Domain (lef) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations in density, velocity, and pressure in the presence of a
stationary background mean-flow that is well approximated by an ideal gas flow. The
physics interface is used for aeroacoustic simulations that can be described by the
linearized Euler equations.

The equations defined by the Linearized Euler, Frequency Domain interface are the
linearized continuity, momentum (Euler), and energy equations. The physics interface
solves for the acoustic variations in the density ρ, velocity field u, and pressure p. The
equations are formulated in the frequency domain and assume harmonic variation of
all sources and fields. The harmonic variation of all fields and sources is given by eiωt
using the +iω convention. The background mean flow can be any stationary gas flow
that is well approximated by an ideal gas. The coupling between the acoustic field and
the background flow does not include any predefined flow-induced noise. Even
though the equations do not include any loss mechanisms, only acoustic modes exist
in the frequency domain as the driving frequency is predefined and real valued.

Coupling between a background mean flow, computed from a Fluid Flow model, and
the Linearized Euler model is handled by the Background Fluid Flow Coupling
multiphysics coupling and the dedicated Mapping study. Details are also found in the
Mapping Between Fluid Flow and Acoustics Mesh section.

For modeling tips and tricks and good practice, see the Modeling with the
Aeroacoustics Branch section.

Flow-induced noise functionality exists in pressure acoustics when using

the Aeroacoustic Flow Source feature and associated multiphysics
coupling and dedicated study.

The equations are implemented in the so-called scattered field formulation. All
equations and boundary conditions are formulated in the total acoustic fields

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 409

 (ρt, ut, pt). The total fields are in the presence of the Background Acoustic Fields
feature the sum of the background (ρb, ub, pb) and the scattered field (ρ, u, p):

ρt = ρ + ρb ut = u + ub pt = p + pb

The scattered field variables are the variables solved for, that is, the dependent
variables. When no background acoustic field is present, the total field is simply equal
to the scattered field

ρt = ρ ut = u pt = p

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Euler Model, Rigid Wall, and Initial Values. For axisymmetric
components an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Linearized Euler, Frequency
Domain to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is lef.

SOUND PRESSURE LEVEL SETTINGS

See Sound Pressure Level Settings for the Pressure Acoustics, Frequency Domain
interface. Only Use reference pressure for air or User-defined reference pressure are
available selections.

TYPICAL WAVE SPEED

See Typical Wave Speed for the Pressure Acoustics, Frequency Domain interface.

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields), the Density rho,
Velocity field u and its components, and Pressure p. The name can be changed but the
names of fields and dependent variables must be unique within a model.

410 | CHAPTER 5: AEROACOUSTICS INTERFACES

STABILIZATION
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box.

Select the Stabilization Method — No stabilization applied, Galerkin least squares (GLS)
stabilization (the default), Streamline upwind Petrov-Galerkin (SUPG) stabilization, or
Streamline diffusion (legacy method). When stabilization is selected enter a value for the
Stabilization constant αstab (dimensionless). The default value is 1e-5 and should
typically have a numerical value between 1e-3 and 1e-7. In cases where there is no
background flow, set the value to the lower limit 1e-7.

The default GLS stabilization is the most efficient stabilization method as it operates
on the convective and reactive parts of the governing equations. This is also the default
method and the method suggested for most applications. The stabilization constant
αstab can be tuned depending on the problem solved, the nature of the background
mean flow, and on the computational mesh.

• Stabilization in the Modeling with the Aeroacoustics Branch section.

If the stabilization is turned off (when No stabilization applied is selected), it is

recommended to change the discretization (see section below) to ensure a stable
numerical scheme; set the order of the pressure one order lower than the velocity and
density dependent variables, for example, using a P2-P2-P1 discretization.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) for the
density, velocity, and pressure degrees of freedom. The default is Linear for all the
dependent variables.

Choosing between Lagrange and Serendipity Shape Functions has

influence on the number of DOFs solved for and on stability for distorted
meshes.

• Domain, Boundary, and Pair Nodes for the Linearized Euler,

Frequency Domain Interface
• Theory Background for the Aeroacoustics Branch

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 411

 For a tutorial and benchmark model, solving the linearized Euler
equations in both the frequency and the time domains, see the
Application Gallery model: Point Source in 2D Jet: Radiation and
Refraction of Sound Waves Through a 2D Shear Layer

[Link]
-refraction-of-sound-waves-through-a-2d-shea-16685

The model requires the Acoustics Module and the CFD Module.

Domain, Boundary, and Pair Nodes for the Linearized Euler,

Frequency Domain Interface
The Linearized Euler, Frequency Domain Interface has these domain, boundary, and
pair nodes and subnodes, listed in alphabetical order, available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• Acoustic Impedance (Isentropic) • Linearized Euler Model

• Asymptotic Far-Field Radiation • Moving Wall
• Axial Symmetry • Outflow Boundary
• Background Acoustic Fields • Periodic Condition1
• Continuity • Prescribed Acoustic Fields
• Domain Sources • Pressure (Isentropic)
• Impedance and Interior Impedance • Rigid Wall
• Initial Values • Symmetry
• Interior Wall
1
Described for the Thermoviscous Acoustics, Frequency Domain interface

412 | CHAPTER 5: AEROACOUSTICS INTERFACES

 In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Linearized Euler Model

Use the Linearized Euler Model to set up the governing equations, define the
background mean flow, the ideal gas fluid properties, and select gradient term
suppression stabilization (advanced physics option), if needed. The governing
equations solved are (in the time domain):

∂ρ
--------t + ∇ ⋅ ( ρ t u 0 + ρ 0 u t ) = S c
∂t
∂u t  ρt
-------- + u t + ------ u 0 ⋅ ∇ u 0 + ( u 0 ⋅ ∇ )u t + ------ ∇p t = S m
1
(5-1)
∂t  ρ0  ρ0
∂p t
-------- + u t ⋅ ∇p 0 + p t γ ( ∇ ⋅ u 0 ) + u 0 ⋅ ∇p t + γp 0 ( ∇ ⋅ u t ) = S e
∂t

where ρt, ut, and pt are the acoustic perturbations to the density, velocity, and
pressure, respectively. The subscript “t” refers to the fact that the acoustic variables are
the total fields, that is, the sum of possible Background Acoustic Fields and the
scattered fields.

In the frequency domain, the time derivatives of the dependent variables is replaced by
multiplication with iω. The variables with a zero subscript are the background mean
flow values, γ is the ratio of specific heats. The right-hand-side source terms Sc, Sm,
and Se are zero per default. They can be defined in the Domain Sources node. Details
about the physics solved and references are found in the Theory Background for the
Aeroacoustics Branch section.

MODEL INPUTS
In order to model the influence of the background mean flow on the propagation of
the acoustic waves in the fluid, the background mean flow temperature T0, absolute
pressure p0, and velocity field u0 need to be defined.

Select User defined (the default), Common model input, or a variable defined by a flow
simulation performed using the CFD Module. By default, they are set to the quiescent

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 413

 background conditions of air. All the background flow parameters can also be
constants or analytical expressions functions of space.

When modeling aeroacoustics it is important how the Mapping Between

Fluid Flow and Acoustics Mesh is done from a numerical perspective.

Details about the Model Input and the Default Model Inputs are found
in the Global and Local Definitions chapter of the COMSOL
Multiphysics Reference Manual.

Enter User defined values for the Background mean flow temperature T0 (SI unit: K),
Background mean flow pressure p0 (SI unit: Pa), and Background mean flow velocity u0
(SI unit: m/s). The defaults are 293.15 K, 1 atm, and 0 m/s, respectively.

Note that the Background mean flow density also needs to be defined or entered in the
Fluid Properties section below.

FLUID PROPERTIES
Select an option for the Background mean flow density ρ0 (SI unit: kg/m3) — Ideal gas
(the default), From material, User defined (default value 1.2 kg/m3), or it can be picked
up from a flow interface, for example, from a High Mach Number Flow model as
Density (hmnf/fluid1). As the flow is assumed to be an ideal gas, the background density
ρ0 is readily defined as

p0
ρ 0 = -------------
Rs T0

where Rs is the specific gas constant.

Define the remaining fluid properties necessary. Select the Gas constant type: Specific
gas constant (the default) or Mean molar mass. The defaults take values From material
or for User defined enter another value or expression:

• Specific gas constant Rs (SI unit: J/(kg·K)). The default is 287.058 J/(kg·K)).
• Mean molar mass Mn (SI unit: g/mol). The default is 28.97 g/mol), which
calculates Rs = R/Mn, where R is the gas constant.

414 | CHAPTER 5: AEROACOUSTICS INTERFACES

Select an option from the Specify Cp or γ list: Ratio of specific heats (the default) or Heat
capacity at constant pressure. The defaults take values From material or for User defined
enter another value or expression:

• Ratio of specific heats γ (dimensionless). The default is 1.4.

• Heat capacity at constant pressure Cp (SI unit: J/(kg⋅K)). The default is
1005.4 J/(kg·K)), which calculates γ = Cp /(Cp − Rs).

GRADIENT TERM SUPPRESSION STABILIZATION

When the linearized Euler (LE) equations are solved in the time domain (or in the
frequency domain with an iterative solver), linear physical instability waves can
develop, the so-called Kelvin-Helmholtz instabilities. They are instabilities that grow
exponentially because no losses exist in the LE equations (no viscous dissipation and
no heat conduction). They are also naturally limited by nonlinearities in the full
Navier–Stokes flow equations. The terms responsible for the instabilities are typically
the reactive terms in the governing equations. It has been shown that in some
problems the growth of these instabilities can be limited, while the acoustic solution is
retained, by canceling terms involving gradients of the mean flow quantities. This is
known as gradient terms suppression (GTS) stabilization.

More details are available in Ref. 9, Ref. 10, Ref. 11, and in the Theory Background
for the Aeroacoustics Branch section.

All the aeroacoustic specific terms can be disabled. They are grouped into Reactive
terms and Convective terms. Select the following check boxes to activate the applicable
gradient term suppression (GTS):

Reactive terms
• Suppression of mean flow density gradients
This option sets the following reactive term in the continuity equation to zero:
u t ⋅ ∇ρ 0 = 0

• Suppression of mean flow velocity gradients

This option sets the following reactive terms in the three governing equations to
zero:
ρt
 u + -----
ρt ( ∇ ⋅ u0 ) = 0 - u ⋅ ∇ u 0 = 0 γp t ( ∇ ⋅ u 0 ) = 0
 t ρ0 0 

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 415

 • Suppression of mean flow pressure gradients
This option sets the following term in the energy equation to zero:

ut ⋅ ∇ p0 = 0

Convective terms
• Suppression of all convective terms
This option removes all the convective terms in the governing equations. This is a
drastic measure as it removes the convective influence of the flow. It should be
considered carefully before doing this. This option removes all terms of the type:
( u0 ⋅ ∇ ) ( … ) = 0

Rigid Wall
The Rigid Wall condition is used to model a rigid wall, corresponding to the sound hard
wall condition in Pressure Acoustics. In the case of lossless flows, this reduces to the
slip condition u t ⋅ n = 0 where n is the surface normal.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

On highly curved surfaces, it may be necessary to switch the Constraint

method from the default Elemental to Nodal in order to avoid locking
effects.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

For systems with a trailing edge on an airfoil or with an edge singularity

between a liner and a rigid wall, the exclude edges and points
functionality can be used to set up an appropriate Kutta condition. See
Suppressing Constraints on Lower Dimensions.

416 | CHAPTER 5: AEROACOUSTICS INTERFACES

Initial Values
The Initial Values node adds initial values for the density, the velocity field, and pressure
dependent variables. Add more Initial Values nodes from the Physics toolbar.

INITIAL VALUES
Enter a value or expression for the initial values of the Density rho (SI unit: kg/m3),
the Velocity field u (SI unit: m/s), and the Pressure p (SI unit: Pa).

Axial Symmetry
In a 2D axisymmetric geometry the Axial Symmetry condition is automatically added
on the axis of symmetry at r = 0. On this boundary the radial component of the total
velocity is set equal to zero

ut = 0 u t = ( u t, v t, w t )

Domain Sources
Add a Domain Sources node to define the mass source, and momentum and energy
source types. This domain feature adds the right-hand side Sc , Sm, and Se to the
governing Equation 5-1 defined through different options

Sc = M
u0 M 1
S m = A – ------------ or S m = ------ ( F – u 0 M )
ρ0 ρ0
γp 0 M γp 0 M
S e = p rc + --------------- or S e = ( γ – 1 )Q + ---------------
ρ0 ρ0

where M is a mass source, A is an acceleration source, F is a volume force source, prc

is a pressure rate of change source, and Q is a volumetric heat source. Notice that the
mass source term enters both the momentum and the energy equations in a consistent
way. The Domain Sources can be used to create any user defined source by entering
expressions into the fields. The sources can for example be point-like Gaussian sources
to model an idealized single vortex or two interacting vortices.

DOMAIN SOURCES
Enter a Mass source M (SI unit: kg/(m3·s)).

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 417

 Select a Momentum source type: Acceleration (the default) or Volume force. For
Acceleration enter vector expressions or values for the Acceleration source A (SI unit:
m/s2). For Volume force enter vector expressions or values for the Volume force source
F (SI unit: N/m3).

Select an Energy source type: Pressure rate of change (the default) or Heat source. For
Pressure rate of change enter an expression or value for the Pressure rate of change
source prc (SI unit: Pa/s). For Heat source enter vector expressions or values for the
Heat source Q (SI unit: W/m3).

Background Acoustic Fields

The Background Acoustic Fields makes it possible to define a background field in a
domain. This condition can be used to model scattering problems or as an inlet-like
condition. In the latter case, when a perfectly matched layer is also present, you can set
up a model with an incident field that also lets any reflected waves leave the
computational domain. The feature is similar to the Background Pressure Field feature
in Pressure Acoustics and the Background Acoustic Fields in Thermoviscous Acoustics.

This condition defines the background fields at the domain level (ρb, ub, pb). The total
acoustic field is now the sum of the scattered field (the dependent variables solved for)
and the background field, such that

ρt = ρ + ρb ut = u + ub pt = p + pb

On interior boundaries continuity in the total field is automatically applied. All

boundary conditions are expressed in terms if the total fields.

BACKGROUND ACOUSTIC FIELDS

Enter values or expressions for:

• Background acoustic density ρb (SI unit: kg/m3).

• Background acoustic velocity ub (SI unit: m/s).
• Background acoustic pressure pb (SI unit: Pa).

Several predefined variables exist for the total, background, and scattered
field: the temperature, the entropy, the intensity, and the sound pressure
level. They are located in the plot menu group Background and scattered
fields, when postprocessing.

418 | CHAPTER 5: AEROACOUSTICS INTERFACES

Pressure (Isentropic)
The Pressure (Isentropic) node is used for prescribing a pressure pp at a boundary
assuming the isentropic relation between pressure and density (adiabatic relation). This
means defining

pp
pt = pp ρ t = -----2-
c0

PRESSURE
Enter a value or expression for the prescribed pressure pp (SI unit: Pa).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Prescribed Acoustic Fields

The Prescribed Acoustic Fields condition makes it possible to prescribe one or more of
the dependent variables at a boundary. When specifying (constraining) the dependent
variables for the linearized Euler equations, it is necessary to not underconstrain the
system. Typically, this requires defining both the density and pressure, defining the
velocity, or defining all three at the same time.

For systems that can be assumed isentropic, where the acoustic changes
in the entropy are zero, enter an expression for the pressure p_user and
the density p_user/lef.c0^2. This corresponds to the classical relation
2
p = c 0 ρ . This is also achieved using the Pressure (Isentropic) condition.

PRESCRIBED VALUES FOR ACOUSTIC VARIABLES

By default no check boxes are selected. Click to select the following check boxes as
needed.

• Prescribed density ρp (SI unit: kg/m3).

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 419

 • Prescribed velocity (SI unit: m/s). If this check box is selected, choose one or more
of these additional check boxes: Prescribed in x direction upx, Prescribed in y direction
upy, or Prescribed in z direction upz (for 3D components); or Prescribed in r direction
upr and Prescribed in z direction upz (for 2D axisymmetric components).
• Prescribed pressure pp (SI unit: Pa).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Acoustic Impedance (Isentropic)

The Acoustic Impedance (Isentropic) adds a simple acoustic impedance assuming
isentropic (adiabatic) behavior for the fields. This condition should be used with care
as it is not necessary that the solution is isentropic near a boundary. The Ingard-Myers
type impedance condition is defined by the Impedance and Interior Impedance
conditions. The isentropic impedance conditions reads:

pt 2
-------------
- = Zi pt = c0 ρt
n ⋅ ut

In transient models, use this boundary condition in combination with the Absorbing
Layers for the Linearized Euler, Transient Interface to create efficient nonreflection
conditions.

IMPEDANCE
Enter an expression or value for the Impedance Zi (SI unit: Pa·s/m). The default value
is set to let.rho0*let.c0, which is the specific characteristic impedance of a plane
propagating wave.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

420 | CHAPTER 5: AEROACOUSTICS INTERFACES

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Symmetry
The Symmetry node adds a symmetry condition at a boundary; it acts the same as the
Rigid Wall condition.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Impedance and Interior Impedance

Use the Impedance or Interior Impedance condition to specify an acoustic impedance
on an exterior boundary or as a transfer impedance on an interior boundary. The
impedance can be any expression and can, for example, be a function of the frequency
(freq). The condition can be used to model a porous lining in a duct or specifying an
outlet impedance. The condition is based on the Ingard-Myers condition, see Ref. 1
and About the Impedance Boundary Condition located in the Theory Background for
the Aeroacoustics Branch section. The condition is a so-called low frequency
approximation; the viscous boundary layer of the background flow is assumed
infinitely thin at the impedance wall.

IMPEDANCE (INGARD-MYERS)/INTERIOR IMPEDANCE (INGARD-MYERS)

Enter the value of the Normal impedance Zn (SI unit: Pa·s/m). This can be an analytical
expression or data provided through an interpolation function. Click to select the
Grazing flow conditions check box, if the impedance is at a boundary with no normal
velocity component for the background flow, u0·n = 0. This simplifies the equations
that are solved for in the boundary condition.

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 421

 CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

For systems with a trailing edge on an airfoil or with an edge singularity

between a liner and a rigid wall, the exclude edges and points
functionality can be used to set up an appropriate Kutta condition. See
Suppressing Constraints on Lower Dimensions.

Moving Wall
The Moving Wall condition is used to model a vibrating wall with harmonic variations.
The condition is a variant of the Ingard-Myers condition used for the Impedance. See
Ref. 1 and About the Moving Wall Boundary Condition located in the Theory
Background for the Aeroacoustics Branch section for further details.

MOVING WALL
Select a Displacement: Inward normal displacement vn (SI unit: m) (the default),
Displacement field vw (SI unit: m), Inward normal velocity un (SI unit: m/s), or Velocity
field uw(SI unit: m/s). Then enter values or expressions based on the selection.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

422 | CHAPTER 5: AEROACOUSTICS INTERFACES

Interior Wall
The Interior Wall boundary condition is used to model a wall condition on an interior
boundary. It is similar to the Rigid Wall boundary condition available on exterior
boundaries except that it applies on both sides (up and down) of an interior boundary.
It allows discontinuities (in density, velocity, and pressure) across the boundary. Use
the Interior Wall boundary condition to avoid meshing thin structures by applying this
slip-like condition on interior curves and surfaces instead.

Asymptotic Far-Field Radiation

Use the Asymptotic Far-Field Radiation node to prescribe open nonreflecting radiation
conditions. The boundary condition is based on the asymptotic behavior of the
acoustic waves in the far field. The condition is also known as the condition of Tam
and Dong (see Ref. 17). The condition applies at boundaries far from a compact
acoustic source. The waves must be assumed spherical in 3D (and 2D axisymmetric)
or cylindrical in 2D. This means that the condition is valid in the limit where

ω
kr » 1 r = r – r0 k = -----
c0

where r is the distance from the source located at r0 and k is the wave number. The
condition is formulated for the scattered field only, in problems involving an
Background Acoustic Fields feature. The condition of Tam and Dong reads:

ρ
∂- ∂
1 - ---- 1
----------- + ----- + ------ U = 0 U = u
V ( θ ) ∂t ∂r ηr
p
2
V ( θ ) = u0 ⋅ er + c0 – ( u0 ⋅ eθ + u0 ⋅ eφ )

where η = 1 in the 3D (and 2D axisymmetric) spherical case and η = 2 in the 2D

cylindrical case. In the frequency domain the time derivative is replaced by a
multiplication with iω.

The acoustic perturbations behave differently than the entropy and vorticity waves.
The acoustic waves propagate in all directions at the speed of sound (corrected by the
background flow) while the entropy and vorticity waves are only convected by the
background flow. The linearized Euler equations support both these waves so both
have to be taken care off. At boundaries where the background flow leaves the

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 423

 computational domain, add the Outflow Boundary subfeature. A default Outflow
Boundary node is added with a cleared selections.

When using the Asymptotic Far-Field Radiation condition careful meshing near the
boundary needs to be considered as gradients of the degrees of freedom enter the
equation. Because of this challenge, using the absorbing layers (ALs) in the time
domain or the perfectly matched layers (PMLs) in the frequency domain is often more
adequate and results in more accurate solution.

Details about Absorbing Layers for the Linearized Euler, Transient

Interface are found in the Modeling with the Aeroacoustics Branch
section.

ASYMPTOTIC FAR-FIELD RADIATION

Enter a Source location r0 (SI unit: m).

Outflow Boundary
A default Outflow Boundary subnode is added to the Asymptotic Far-Field Radiation.
The outflow condition has to be added where the background mean flow leaves the
computational domain to ensure that vorticity and entropy waves are not created here.

When solving, the interface can return an error if the condition is added where there
is no background mean flow, u0 = 0. Here certain terms evaluate to zero. Adding the
condition at boundaries where the background mean flow enters the computational
domain can result in unphysical results.

Continuity
Continuity is available as an option at interfaces between parts in a pair. This condition
gives continuity in all the dependent variables: density, velocity, and pressure. It
corresponds to a situation where the boundary has no direct effect on the acoustic
fields (subscripts “src” and “dest” in the equation refers to the source and destination,
the two sides of the pair):

T
U src – U dest = 0, U = ( ρ t, u t, p t )

424 | CHAPTER 5: AEROACOUSTICS INTERFACES

PAIR SELECTION
Select an identity pair. A pair is automatically created in the geometry when Create pairs
is checked in the Form Union/Assembly step.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

THE LINEARIZED EULER, FREQUENCY DOMAIN INTERFACE | 425

 T he L i ne a r i z e d E u l er, Tran si en t
Interface
The Linearized Euler, Transient (let) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations in density, velocity, and pressure in the presence of a
stationary background mean-flow that is well approximated by an ideal gas flow. The
physics interface is used for aeroacoustic simulations that can be described by the
linearized Euler equations.

The equations defined by the Linearized Euler, Transient interface are the linearized
continuity, momentum (Euler), and energy equations. The physics interface solves for
the acoustic variations in the density ρ, velocity field u, and pressure p. The equations
are formulated in the time domain. The background mean flow can be any stationary
gas flow that is well approximated by an ideal gas. The coupling between the acoustic
field and the background flow does not include any predefined flow induced noise. As
the equations do not include any loss mechanisms, nonacoustic modes and instabilities
can exist in the time domain.

The equations are defined using a scattered-field formulation just as in The Linearized
Euler, Frequency Domain Interface and allows the use of Background Acoustic Fields.
Open nonreflecting conditions are set up with the Absorbing Layers for the Linearized
Euler, Transient Interface.

Coupling between a background mean flow, computed from a Fluid Flow model, and
the Linearized Euler model is handled by the Background Fluid Flow Coupling
multiphysics coupling and the dedicated Mapping study. Details are also found in the
Mapping Between Fluid Flow and Acoustics Mesh section.

For modeling tips and tricks and good practice, see the Modeling with the
Aeroacoustics Branch section.

Flow-induced noise functionality exists in pressure acoustics when using

the Aeroacoustic Flow Source feature and associated multiphysics
coupling and dedicated study.

426 | CHAPTER 5: AEROACOUSTICS INTERFACES

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Euler Model, Rigid Wall, and Initial Values. For axisymmetric
components an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Linearized Euler to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is let.

STABILIZATION
See Stabilization in the frequency domain interface for details.

DAMPING PARAMETERS FOR ABSORBING LAYERS

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. In the Damping Parameters for
Absorbing Layers section you can change and control the values of the artificial damping
added in the Absorbing Layers for the Linearized Euler, Transient Interface. Enter a
value for the Numerical viscosity parameter μnum (default value is 100 Pa·s) and the
Numerical viscosity curvature n (default value is 2).

TRANSIENT SOLVER SETTINGS

Enter the Maximum frequency to resolve in the model. The default frequency is set to
1000[Hz] but should be changed to reflect the frequency content of the sources used
in the model. Select the Time stepping (method) as Fixed (preferred) the default and
recommended or Free. The Free option is in general not recommended for wave
problems. The generated solver will be adequate in most situations if the
computational mesh also resolves the frequency content in the model. Note that any
changes made to these settings (after the model is solved the first time) will only be

THE LINEARIZED EULER, TRANSIENT INTERFACE | 427

 reflected in the solver if Show Default Solver or Reset Solver to Defaults is selected in the
study.

Details about Time Stepping in Transient Models are found in the

Modeling with the Aeroacoustics Branch section.

DISCRETIZATION
See Discretization in the frequency domain interface for details.

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields), the Density rho,
Velocity field u and its components, and Pressure p. The name can be changed but the
names of fields and dependent variables must be unique within a model.

• Domain, Boundary, and Pair Nodes for the Linearized Euler, Transient
Interface
• Theory Background for the Aeroacoustics Branch

Domain, Boundary, and Pair Nodes for the Linearized Euler,

Transient Interface
The Linearized Euler, Transient Interface has these nodes, listed in alphabetical order,
available from the Physics ribbon toolbar (Windows users), Physics context menu (Mac
or Linux users), or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

428 | CHAPTER 5: AEROACOUSTICS INTERFACES

This physics interface shares some of its node settings with The Linearized Euler,
Frequency Domain Interface.

• Acoustic Impedance (Isentropic)1 • Linearized Euler Model1

• Asymptotic Far-Field Radiation1 • Moving Wall
• Axial Symmetry1 • Rigid Wall1
• Background Acoustic Fields1 • Prescribed Acoustic Fields1
• Continuity1 • Pressure (Isentropic)1
• Domain Sources1 • Symmetry1
• Initial Values
• Interior Wall1
1This feature is described in the Linearized Euler, Frequency Domain interface.

Initial Values
The Initial Values node adds initial values for the density, the velocity field, and pressure
dependent variables. Add more Initial Values nodes from the Physics toolbar.

INITIAL VALUES
Enter a value or expression for the initial values of the Density rho (SI unit: kg/m3),
the Velocity field u (SI unit: m/s), and the Pressure p (SI unit: Pa).

Moving Wall
The Moving Wall condition can in the time domain only be defined in terms of the
normal displacement. It is used to define a vibrating or moving wall. The displacement
can be any time-dependent expression. The condition can, for example, be used to
model an actuator creating sound in a gas flowmeter.

MOVING WALL
Enter a Normal displacement vn (SI unit: m).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

THE LINEARIZED EULER, TRANSIENT INTERFACE | 429

 EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

430 | CHAPTER 5: AEROACOUSTICS INTERFACES

T he L i ne a r i z e d E u l er, Bou n d ary Mod e
Interface
The Linearized Euler, Boundary Mode interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute and identify propagating and nonpropagating modes in waveguides and
ducts in the presence of a stationary background mean-flow that is well approximated
by an ideal gas flow. The interface performs a boundary mode analysis on a boundary,
inlet, or cross section of the waveguide. The interface exists in 3D and 2D
axisymmetric geometries and applies to boundaries.

The interface solves for the acoustic variations of density, velocity, and pressure, as well
as the out-of-plane wave number of the modes. The Linearized Euler, Boundary Mode
interface is, for example, used when setting up sources in waveguide systems in the
presence of a background mean-flow. The background flow is assumed to be uniform
in the propagation direction.

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Euler Model, Rigid Wall, and Initial Values. For axisymmetric
components, an Axial Symmetry node is also added. Then, from the Physics toolbar, add
other nodes that implement, for example, boundary conditions and sources. You can
also right-click Linearized Euler, Boundary Mode to select physics features from the
context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

THE LINEARIZED EULER, BOUNDARY MODE INTERFACE | 431

 The default Name (for the first physics interface in the model) is lebm.

For boundary conditions and domain conditions see:

• Domain, Boundary, and Pair Nodes for the Linearized Euler,

Frequency Domain Interface
• Theory Background for the Aeroacoustics Branch

Stabilization is not implemented for the Linearized Euler, Boundary

Mode interface, which instead uses a default P1-P2-P1 discretization that
ensures stability.

Linearized Euler Model

The Linearized Euler Model node adds the governing equations, and define the
background mean flow and the ideal gas fluid properties. Gradient term suppression
stabilization (advanced physics option) can be selected, if needed. See the Linearized
Euler Model for The Linearized Euler, Frequency Domain Interface for details.

BOUNDARY MODE SETTINGS

Define the Propagation direction of the computed modes. The default is to use the
surface physics normal (components [Link], [Link], and [Link]) that points out
of domains.

Rigid Wall
The Rigid Wall condition is used to model a rigid wall, corresponding to the sound hard
wall condition in Pressure Acoustics. In the case of lossless flows, this reduces to the
slip condition u t ⋅ n = 0 where n is the normal to the edge, that is, the waveguide
wall. See the Rigid Wall condition for The Linearized Euler, Frequency Domain
Interface for further details.

Initial Values
The Initial Values node adds initial values for the density, velocity, and pressure
variables.

432 | CHAPTER 5: AEROACOUSTICS INTERFACES

INITIAL VALUES
Enter a value or expression for the initial values Density rho (SI unit: kg/m3), Velocity
field u (SI unit: m/s), and Pressure p (SI unit: Pa).

THE LINEARIZED EULER, BOUNDARY MODE INTERFACE | 433

 T he L i ne a r i z e d N av i er–St okes,
Frequency Domain Interface
The Linearized Navier–Stokes, Frequency Domain (lnsf) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations in pressure, velocity, and temperature in the presence
of any stationary isothermal or nonisothermal background mean flow. The physics
interface is used for aeroacoustic simulations that can be described by the linearized
Navier–Stokes equations.

The equations are formulated in the frequency domain and assume harmonic variation
of all sources and fields. The equations include viscous losses and thermal conduction
as well as the heat generated by viscous dissipation, if relevant. The coupling between
the acoustic field and the background flow does not include any predefined flow
induced noise.

Flow-induced noise functionality exists in pressure acoustics when using

the Aeroacoustic Flow Source feature and associated multiphysics
coupling and dedicated study.

The equations defined by the Linearized Navier–Stokes, Frequency Domain interface

are the linearized continuity, momentum (Navier–Stokes), and energy equations. The
physics interface solves for the acoustic variations in the pressure p, velocity field u, and
iωt
temperature T. The harmonic variation of all fields and sources is given by e using
the +iω convention. The equations are formulated in the frequency domain for any
fluid including losses due to viscosity and thermal conduction. The background mean
flow can be any stationary flow.

The Linearized Navier–Stokes, Frequency Domain interface is formulated in the

so-called scattered-field formulation where the total acoustic field (subscript t) is the
sum of the scattered field (the field solved for p, u, and T) and a possible background
acoustic field (subscript “b”), such that

pt = p + pb ut = u + ub Tt = T + Tb

434 | CHAPTER 5: AEROACOUSTICS INTERFACES

All governing equations and boundary conditions are formulated in the total field
variables. When no Background Acoustic Fields feature is present (the background
field values are zero per default) the total field is simply the field solved for

pt = p ut = u Tt = T

Coupling the interfaces to structures enables detailed vibration analysis of structures in

the presence of flow, such as FSI in the frequency domain. The coupling in the
frequency domain and time domain is readily performed using the predefined
Aeroacoustic-Structure Boundary multiphysics coupling feature.

Coupling between a background mean flow, computed from a Fluid Flow model, and
the Linearized Euler model is handled by the Background Fluid Flow Coupling
multiphysics coupling and the dedicated Mapping study. Details are also found in the
Mapping Between Fluid Flow and Acoustics Mesh section

For modeling tips and tricks and good practice see the Modeling with the
Aeroacoustics Branch section.

The Helmholtz Resonator with Flow: Interaction of Flow and Acoustics

tutorial model gives an example of how to model the detailed interaction
between flow and acoustics. The model requires both the Acoustics
Module and the CFD Module. The Application Library path is:
Acoustics_Module/Aeroacoustics_and_Noise/helmholtz_resnoator_with_flow

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Navier–Stokes Model, Wall, and Initial Values. For axisymmetric
components, an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Linearized Navier–Stokes,
Frequency Domain to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 435

 different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is lnsf.

EQUATION
Expand the Equation section to see the equations solved for with the Equation form
specified. The default selection is Equation form is set to Study controlled. The available
studies are selected under Show equations assuming.

• For Study controlled, the scaling of the equations is optimized for the numerical
performance of the different solvers and study types.
• For Frequency domain you can manually enter the scaling parameter Δ under
Linearized Navier–Stokes Equation Settings section.

LINEARIZED NAVIER–STOKES EQUATION SETTINGS

Click to select Adiabatic formulation to use an adiabatic equation of state and disable
the temperature degree of freedom for the linearized Navier–Stokes equations. This
formulation is applicable when the thermal losses can be disregarded, this is often the
case in liquids like water. In gases, like air, on the other hand the full formulation is
necessary. When Adiabatic formulation is selected, all temperature conditions and
options are disabled in the user interface.

For all component dimensions, and if required, click to expand the Equation section,
then select Frequency domain as the Equation form and enter the settings as described
below.

The default Scaling factor Δ is 1/(iω). This value correspond to the equations for a
Frequency Domain study when the equations are study controlled. To get the
equations corresponding to an Eigenfrequency study, change the Scaling factor Δ to 1.
Changing the scaling factor influences the coupling to other physics.

SOUND PRESSURE LEVEL SETTINGS

See Sound Pressure Level Settings for the Pressure Acoustics, Frequency Domain
interface.

TYPICAL WAVE SPEED

See Typical Wave Speed for the Pressure Acoustics, Frequency Domain interface.

436 | CHAPTER 5: AEROACOUSTICS INTERFACES

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields), the Pressure p, Velocity
field u and its components, and Temperature variation T. The name can be changed but
the names of fields and dependent variables must be unique within a model.

STABILIZATION
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box.

Select the Stabilization Method — No stabilization applied, Galerkin least squares (GLS)
stabilization (the default), Streamline upwind Petrov-Galerkin (SUPG) stabilization, or
Streamline diffusion (legacy method). When stabilization is selected enter a value for the
Stabilization constant αstab (dimensionless). The default value is 1 and is well suited for
most modeling situations.

The default GLS stabilization is the most efficient stabilization method as it operates
on the convective, reactive, and diffusive parts of the governing equations. This is also
the default method and the method suggested for most applications. The stabilization
constant αstab can be tuned depending on the problem solved, the nature of the
background mean flow, and on the computational mesh.

Note that, if a problem is solved where no background mean flow is present, it is

recommended to either use one of the Thermoviscous Acoustics Interfaces or turn off
stabilization and switch to a P1-P2-P2 discretization.

• Stabilization in the Modeling with the Aeroacoustics Branch section.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) the default
is Linear for all the dependent variables.

Choosing between Lagrange and Serendipity Shape Functions has

influence on the number of DOFs solved for and on stability for distorted
mesh.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 437

 In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

• Domain, Boundary, and Pair Nodes for the Linearized Navier–Stokes,

Frequency Domain and Transient Interfaces
• Theory Background for the Aeroacoustics Branch

As a tutorial model showcasing acoustic-structure interaction in the

presence of flow, that is FSI in the frequency domain, see the Application
Gallery model: Vibrating Plate in a 2D Viscous Parallel Plate Flow.

[Link]
el-plate-flow-18871

An advanced example of a Coriolis flowmeter is found in the Application

Gallery: Coriolis Flowmeter: FSI Simulation in the Frequency Domain.

[Link]
the-frequency-domain-51831

Both the models solve the linearized Navier–Stokes equations coupled to

solid mechanics in the presence of flow. The model requires both the
Acoustics Module and the CFD Module.

Domain, Boundary, and Pair Nodes for the Linearized Navier–

Stokes, Frequency Domain and Transient Interfaces
The Linearized Navier–Stokes, Frequency Domain Interface has these domain,
boundary, and pair nodes, listed in alphabetical order, available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

438 | CHAPTER 5: AEROACOUSTICS INTERFACES

 • Adiabatic • Linearized Navier–Stokes Model
• Axial Symmetry • Periodic Condition1
• Background Acoustic Fields • Prescribed Pressure
• Boundary Stress • Prescribed Temperature
• Continuity • Prescribed Velocity
• Domain Sources • Pressure (Adiabatic)
• First-Order Material Parameters • No Stress
• Heat Flux • Normal Impedance
• Initial Values • No Slip
• Interior Normal Impedance • Slip
• Interior Wall • Symmetry
• Isothermal • Wall
1
Described for the Thermoviscous Acoustics, Frequency Domain interface

Linearized Navier–Stokes Model

The Linearized Navier–Stokes Model sets up the governing equations, defines the
background mean flow, fluid properties, and the compressibility and thermal
expansion properties of the fluid. The governing equations solved are the continuity,
momentum, and energy equations:

∂ρ
--------t + ∇ ⋅ ( ρ 0 u t + ρ t u 0 ) = M
∂t
∂u t
ρ 0  -------- + ( u t ⋅ ∇ )u 0 + ( u 0 ⋅ ∇ )u t + ρ t ( u 0 ⋅ ∇ )u 0 = ∇ ⋅ σ + F – u 0 M
 ∂t 
(5-2)
∂T t
ρ 0 C p  --------- + ( u t ⋅ ∇ )T 0 + ( u 0 ⋅ ∇ )T t + ρC p ( u 0 ⋅ ∇ )T 0
 ∂t 
∂p t
– α p T 0  -------- + ( u t ⋅ ∇ )p 0 + ( u 0 ⋅ ∇ )p t – α p T t ( u 0 ⋅ ∇ )p 0 = ∇ ⋅ ( k∇T t ) + Φ + Q
 ∂t 

where pt, ut, and Tt are the acoustic perturbations to the pressure, velocity, and
temperature, respectively. The subscript “t” refers to the fact that the acoustic variables
are the total fields, that is, the sum of possible Background Acoustic Fields and the
scattered fields.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 439

 In the frequency domain, the time derivatives of the dependent variables are replaced
by multiplication with iω. The stress tensor is σ and Φ is the viscous dissipation
function. The right-hand-side source terms M, F, and Q are initially zero; they can be
defined using the Domain Sources feature. The variables with a zero subscript are the
background mean flow values. The material parameters are defined below. Details
about the physics and references are found in the Theory Background for the
Aeroacoustics Branch section.

The constitutive equations are the stress tensor and the linearized equation of state,
while the Fourier heat conduction law is readily included in the above energy equation,

σ = – p t I + μ ( ∇u t + ( ∇u t ) ) +  μ B – --- μ ( ∇ ⋅ u t )I
T 2
 3  (5-3)
ρt = ρ0 ( βT pt – αp Tt )

when Adiabatic formulation is selected in the Linearized Navier–Stokes Equation

Settings section the equation of state reduces to

βT
ρ t = ρ 0 β s p t = ρ 0 ------ p t
γ

The linearized viscous dissipation function is defined as

Φ = ∇u t :τ ( u 0 ) + ∇u 0 :τ ( u t )

τ ( u t ) = μ ( ∇u t + ( ∇u t ) ) +  μ B – --- μ ( ∇ ⋅ u t )I
T 2
3 (5-4)

τ ( u 0 ) = μ ( ∇u 0 + ( ∇u 0 ) ) +  μ B – --- μ ( ∇ ⋅ u 0 )I
T 2
 3 

MODEL INPUTS
In order to model the influence of the background mean flow on the propagation of
the acoustic waves in the fluid, the background mean flow temperature T0, absolute
pressure p0, and velocity field u0 need to be defined. The density is defined in the Fluid
Properties section below, and is per default taken from the material. It is thus a function
of the model inputs, that is, the background pressure and temperature. Select User
defined (the default) or Common model input. For User defined enter values for:

• Background mean flow temperature T0 (SI unit K). The default is 293.15 K.

440 | CHAPTER 5: AEROACOUSTICS INTERFACES

• Background mean flow pressure p0 (SI unit: Pa). The default is 1 atm.
• Background mean flow velocity u0 (SI unit: m/s). The defaults are 0 m/s.

When modeling aeroacoustics it is important how the Mapping Between

Fluid Flow and Acoustics Mesh is done from a numerical perspective. The
Background Fluid Flow Coupling multiphysics coupling handles this in
an automated manner.

Physically the Coupling to Turbulent Flows (Eddy Viscosity) is also

important to model the attenuation of acoustics waves due to turbulence.

Details about the Model Input and the Default Model Inputs are found
in the Global and Local Definitions chapter of the COMSOL
Multiphysics Reference Manual.

FLUID PROPERTIES
The defaults for the following are taken From material. For User defined edit the default
values:

• Background mean flow density ρ0(p0, T0) (SI unit: kg/m3). The Ideal gas option can
also be selected as the density formulation.
• Dynamic viscosity μ (SI unit: Pa·s).
• Bulk viscosity μB (SI unit: Pa·s).
• Thermal conductivity k (SI unit: W/(m·K)).
• Heat capacity at constant pressure Cp (SI unit: J/(kg·K)).

For a discussion about the air and water materials as commonly used in
acoustics and other material properties, such as the bulk viscosity, see the
Acoustic Properties of Fluids chapter of this manual.

THERMAL EXPANSION AND COMPRESSIBILITY

The Thermal Expansion and Compressibility section is displayed if From material or User
defined is selected as the Background mean flow density under the Fluid Properties
section. For the Ideal gas option the parameters are readily defined.

Select an option from the Coefficient of thermal expansion αp list — From material, From
background mean flow density, From speed of sound (the default), or User defined.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 441

 Select an option from the Isothermal compressibility βT list — From background mean
flow density, From isentropic compressibility, From speed of sound (the default), or User
defined.

The different options for defining the (isobaric) coefficient of thermal expansion and
the isothermal compressibility stem from their respective thermodynamic definitions:

1 ∂ρ 0 1 Cp ( γ – 1 )
α p = – ------ ---------- = ----- -----------------------
-
ρ 0 ∂T 0 p c0 T0

1 ∂ρ 0 1 γ
βT = ------ --------- = ------ ----2- = γβ s
ρ 0 ∂p 0 T ρ0 c
0

• For the From speed of sound (the default for both) options the values for the Speed
of sound c0 (SI unit: m/s) and Ratio of specific heats γ (dimensionless) are taken From
material. For User defined enter a different value or expression. The From speed of
sound option is typically preferred (and the default) as the speed of sound and the
ratio of specific heats are material properties often more readily available (see
definition above).
• For the From background mean flow density the values are computed from the density
expression (see definition above). This option is only valid if the density material
property has the built in dependency on pressure p0 and temperature T0. This is not
always the case.
• For the From material the coefficient of thermal expansion is taken from the
materials node. Not all materials have this property defined. If this is the case a small
warning cross will appear on the materials node.
• For the From isentropic compressibility option the values for the Isentropic
compressibility βs (SI unit: 1/Pa) and Ratio of specific heats γ (dimensionless) are
taken From material (the default). For User defined enter a different values or
expressions.
• For User defined enter a value or expression for the isobaric coefficient of thermal
expansion αp (SI unit: 1/K) and/or the isothermal compressibility βT (SI unit:
1/Pa).

VISCOUS DISSIPATION FUNCTION

Select the Include viscous dissipation function check box if you want to include the heat
source generated by the viscous losses. The viscous dissipation function Φ is defined in
Equation 5-4.

442 | CHAPTER 5: AEROACOUSTICS INTERFACES

GRADIENT TERM SUPPRESSION STABILIZATION
When the linearized Navier–Stokes (LNS) equations are solved in the time domain (or
in the frequency domain with an iterative solver), linear physical instability waves can
develop, the so-called Kelvin-Helmholtz instabilities. They are instabilities that can
grow, but to less degree than in the linearized Euler equations because losses are
present in the LNS equations. They are further naturally limited by nonlinearities in
the full Navier–Stokes flow equations. The terms responsible for the instabilities are
typically the reactive terms in the governing equations. It has been shown that in some
problems the growth of these instabilities can be limited, while the acoustic solution is
retained, by canceling terms involving gradients of the mean-flow quantities. This is
known as gradient terms suppression (GTS) stabilization.

The GTS option can also be used to alter the governing equations to fit a desired
formulation of the linearized Navier–Stokes equations. This can be relevant in certain
applications.

All the aeroacoustic specific terms can be disabled. They are grouped into Reactive
terms and Convective terms. Select the following check boxes to activate the applicable
gradient term suppression (GTS):

Reactive terms
• Suppression of mean flow pressure gradients
This option sets the following term to zero:
u t ⋅ ∇p 0 = 0 u 0 ⋅ ∇p 0 = 0

• Suppression of mean flow velocity gradients

This option sets the following term to zero:
∇ ⋅ u0 = 0 u t ⋅ ∇u 0 = 0 u 0 ⋅ ∇u 0 = 0

• Suppression of mean flow temperature gradients

This option sets the following term to zero:

u t ⋅ ∇T 0 = 0 u 0 ⋅ ∇T 0 = 0

• Suppression of mean flow density gradients

This option sets the following term to zero:

ut ⋅ ∇ ρ0 = 0

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 443

 Convective terms
• Suppression of all convective terms
This option removes all the convective terms in the governing equations. This is a
drastic measure as it removes the convective influence of the flow. It should be
considered carefully before doing this. This option removes all terms of the type:
( u0 ⋅ ∇ ) ( … ) = 0

Domain Sources
Add a Domain Sources node to define the mass source, volume force source, and heat
source, the right-hand-side of Equation 5-2. These can, for example, model varying
thermal sources in a combustion chamber or a pulsating laser. Notice that the mass
source term enters both the continuity and the momentum equation, this results from
a consistent derivation of the governing equations. A mass source term will, when
convected also result in a momentum source.

DOMAIN SOURCES
Enter values or expressions for the following:

• Mass source M (SI unit: kg/(m3·s)).

• Volume force source F (SI unit: N/m3).
• Heat source Q (SI unit: W/m3).

First-Order Material Parameters

To display this node in the context menu, click the Show More Options button ( )
and select Advanced Physics Options in the Show More Options dialog box.

Use the First Order Material Parameters node to include subtle (acoustic) variations in
the material parameters due to the acoustic variations of the dependent variables. The
selected material parameters vary according to a linearization about their background
values (at pressure p0 and temperature T0). The viscosity μ is, for example, replaced by:

μ → μ + p ∂μ
------ ∂μ-
+ T ------
∂p T0 ∂T p0

444 | CHAPTER 5: AEROACOUSTICS INTERFACES

FIRST-ORDER MATERIAL PARAMETERS
By default, no check boxes are selected. Click to select the following check boxes as
needed.

• Derivatives of dynamic viscosity ∂μ/∂p (SI unit: s) and ∂μ/∂T (SI unit: kg/(m·s·Κ)).
• Derivatives of bulk viscosity ∂μB/∂p (SI unit: s) and ∂μB/∂T (SI unit: kg/(m·s·Κ)).
• Derivatives of heat capacity at constant pressure ∂Cp/∂p (SI unit: m3/(kg·K)) and
∂Cp/∂T (SI unit: m2/(s2·K2)).
• Derivatives of thermal conduction ∂k/∂p (SI unit: m2/(s·K)) and ∂k/∂T
(SI unit: m·kg /(s2·K2)).
• Derivatives of coefficient of thermal expansion ∂αp/∂p (SI unit: m·s2/(kg·K)) and
∂αp/∂T (SI unit: 1/K2).

VISCOUS DISSIPATION FUNCTION

Select the Include viscous dissipation function check box to include the effects of the
redefined material parameters on the viscous dissipation function.

Background Acoustic Fields

When the Background Acoustic Fields feature is added to a domain, it is possible to
define the value of the background acoustic field variables pb, ub, and Tb. Using this
feature, it is possible to set up scattering problems as well as defining acoustic fields at
an inlet of a waveguide (using a small domain at the inlet).

BACKGROUND ACOUSTIC FIELDS

Enter expressions for the Background acoustic pressure pb, the Background acoustic
velocity ub, and the Background temperature variation Tb. These can be analytical
expressions or values of dependent variables solved in a previous study that defines the
background field.

Initial Values
The Initial Values node adds initial values for the pressure, the velocity field, and
temperature dependent variables. Add more Initial Values nodes from the Physics
toolbar.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 445

 INITIAL VALUES
Enter a value or expression for the initial values of the Pressure p (SI unit: Pa), the
Velocity field u (SI unit: m/s), and the Temperature T (SI unit: K).

Axial Symmetry
In a 2D axisymmetric geometry the Axial Symmetry condition is automatically added
on the axis of symmetry at r = 0. On this boundary the radial component of the total
velocity is set equal to zero

ut = 0 u t = ( u t, v t, w t )

Wall
The Wall condition is the most common condition at solid surfaces. This condition
contains both a mechanical and a thermal selection. The default is a no-slip and
isothermal condition applicable in most cases. More advanced conditions at
boundaries can be set up by combining any of the Mechanical and Thermal conditions
available.

The Wall condition is also the default condition added when the physics interface is set
up. Note that, the selections in this default condition can be changed (they are not
locked to all boundaries). When used together with a perfectly matched layer (PML)
it is recommended to use the no-slip option at the outer boundary of the layer.
Specifically, using the slip condition in combination with a rational scaling in the PML
will lead to a singularity and an error.

The no-slip condition is the origin of the viscous boundary layer and the isothermal
condition is the origin of the thermal boundary layer. It is within these acoustic
boundary layers that the main dissipation happens. In some applications like for
example, in large muffler systems, it is often not necessary to model these loss effects
and thus switching to slip and adiabatic can be a good approximation. This also means
that the mesh does not need to resolve the acoustic boundary layer which can save
DOFs.

MECHANICAL
Select a Mechanical condition — No slip (the default) or Slip. See the No Slip and Slip
conditions for further details.

446 | CHAPTER 5: AEROACOUSTICS INTERFACES

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic. See the Isothermal
and Adiabatic conditions for further details.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

If Slip is selected for the Mechanical condition and Use weak constraints is enabled then
the default discontinuous Galerkin (penalty like) formulation is switched to a Lagrange
multiplier formulation instead. See the Slip condition for details. For the No-slip and
Isothermal selections a weak formulation is used instead of a pointwise constraint.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Pressure (Adiabatic)
Add a Pressure (Adiabatic) node to give a pressure boundary condition with adiabatic
conditions for the temperature. The condition is given by

pt = pp σn = p p n – n ⋅ ( – k∇ T t ) = 0

where pp is the desired pressure at the boundary, σ is the stress tensor, and n the
surface normal. This is a good approximation when prescribing a pressure at an inlet
or outlet.

PRESSURE (ADIABATIC)
Enter a value or expression for the Pressure pp (SI unit: Pa).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 447

 Symmetry
The Symmetry node adds a boundary condition that represents symmetry. It
corresponds to the Slip condition for the mechanical degrees of freedom and the
Adiabatic condition for the temperature variation. The symmetry condition should
only be applied to flat surfaces.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Interior Wall
The Interior Wall boundary condition is used to model a wall condition on an interior
boundary. It is similar to the Wall boundary condition available on exterior boundaries
except that it applies on both sides (up and down) of an interior boundary. It allows
discontinuities (in pressure, velocity, and temperature) across the boundary. Use the
Interior Wall boundary condition to avoid meshing thin structures by applying this
slip-like condition on interior curves and surfaces instead.

MECHANICAL
Select a Mechanical condition — No slip (the default) or Slip. See the No Slip and Slip
conditions for further details.

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic. See the Isothermal
and Adiabatic conditions for further details.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

If Slip is selected for the Mechanical condition and Use weak constraints is enabled then
the default discontinuous Galerkin (penalty like) formulation is switched to a Lagrange

448 | CHAPTER 5: AEROACOUSTICS INTERFACES

multiplier formulation instead. For the No-slip and Isothermal selections a weak
formulation is used instead of a pointwise constraint.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Interior Normal Impedance

Use the Interior Normal Impedance node to apply a transfer impedance condition on an
interior boundary. The condition relates the normal stress at the two sides of the
boundary (up and down) to the normal velocity via the normal transfer impedance Zn.
The condition allows discontinuities (a slit) in pressure and temperature depending on
the condition chosen for the temperature.

( σ up – σ down )n = Z n ( u t ⋅ n )n

σ = – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I
2μ
 3 
p t,down – p t,up
Z n = -----------------------------------
ut ⋅ n

MECHANICAL
Enter a value or expression for the Normal impedance Zn (SI unit: Pa⋅s/m). This is the
transfer impedance of the interior boundary.

Select a condition for the Tangential velocity — Slip (the default) or No slip. The No Slip
option will apply an additional constraint for the velocity in the tangential direction on
the two sides of the interior boundary.

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 449

 EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

No Slip
The No Slip sets up a no-slip condition for the flow on a hard wall:

ut = 0

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Mechanical and thermal boundary conditions contribute with each other

such that a condition can be set on the velocity and temperature
simultaneously. Thermal type conditions override each other and the
same is true for mechanical type conditions.

Slip
Add a Slip node to define slip for the velocity defined by

n ⋅ ut = 0
σ n – ( σ n ⋅ n )n = 0, σ n = σn

where σ is the stress tensor. The last equation expresses that no tangential stress exists
at that boundary.

This results in a so-called no-penetration condition where no viscous boundary layer

is created. This condition can be used at boundaries where it is not necessary to model
the losses in the viscous boundary layer. Used together with the Adiabatic condition,
no acoustic boundary layer is modeled.

450 | CHAPTER 5: AEROACOUSTICS INTERFACES

When running a model with a perfectly matched layer (PML), it is recommended not
to combine a Slip condition with a rational scaling in the PML, as this will lead to a
singularity and an error.

Per default, the Slip condition uses a so-called discontinuous Galerkin or penalty
formulation to prevent locking problems on curved surfaces.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Select Use weak constraints if you want to switch from the default discontinuous
Galerkin (penalty like) formulation to a Lagrange multiplier formulation for the slip
condition (this formulation is not suited for an iterative solver). The Lagrange
multiplier formulation also prevents locking problems.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Prescribed Velocity
Add a Prescribed Velocity node to define the velocity at a boundary.

PRESCRIBED VELOCITY
Select the components to prescribe and enter a value for the Prescribed in the direction
(SI unit: m/s).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 451

 Prescribed Pressure
Add a Prescribed Pressure node to prescribe the pressure at a boundary (using a
constraint).

PRESCRIBED PRESSURE
Enter a Pressure pp (SI unit: Pa).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

No Stress
Use the No Stress node to set the total surface stress equal to zero:

– p t I + μ ( ∇u t + ∇u t ) –  ------- – μ B ( ∇ ⋅ u t )I n = 0
T 2μ
3

Boundary Stress
Use the Boundary Stress node to define the total surface (boundary) stress through a
stress tensor, a traction force, or a pressure. Add this node from the Mechanical
submenu. The stress condition reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = σ p n
2μ
 3 

where σ p is a symmetric stress tensor, the product σ p n defines a traction force. The
traction condition reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = F tr
2μ
 3 

where Ftr is the surface traction force. The pressure condition reads

452 | CHAPTER 5: AEROACOUSTICS INTERFACES

 – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = – p bnd n
2μ
 3 

where pbnd is the pressure applied to the boundary (acting in the normal direction).

STRESS
Select a Stress type — Stress tensor, Traction (the default), or Pressure. For Stress tensor
enter the (symmetric) Stress tensor σ p (SI unit: N/m2) components, for Traction
enter the traction force Ftr (SI unit: N/m2) components, and for Pressure enter the
boundary pressure pbnd (SI unit: Pa).

Normal Impedance
Use the Normal Impedance node to specify a normal impedance Zn on a boundary. This
feature is useful outside the viscous boundary layer, as this condition mimics the
behavior of a corresponding Pressure Acoustics Model with a normal impedance
condition. The boundary condition reads:

– p t I + μ ( ∇u t + ∇u t ) –  ------- – μ B ( ∇ ⋅ u t )I n = – Z n ( u t ⋅ n )n
T 2μ
3

MECHANICAL
Enter a value or expression for the Normal impedance Zn (SI unit: Pa⋅s/m).

Select a condition for the Tangential velocity — Slip (the default) or No slip. The No Slip
option will apply an additional constraint for the velocity in the tangential direction on
the boundary.

Isothermal
The Isothermal node sets up an isothermal condition:

Tt = 0

This is a good approximation on solid walls, as heat conduction is typically much

higher in solids than in fluids.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 453

 EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Adiabatic
Add an Adiabatic node to set up an adiabatic condition for the temperature, that is, the
natural condition for the temperature

– n ⋅ ( – k∇ T t ) = 0

Prescribed Temperature
Add a Prescribed Temperature node to define the temperature variation Tp at a
boundary.

PRESCRIBED TEMPERATURE
Enter a value or expression for the Temperature Tp (SI unit: K).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Heat Flux
Add a Heat Flux node to define the inward normal heat flux qn at a boundary:

– n ⋅ ( – k ∇T t ) = q n

HEAT FLUX
Enter a value or expression for the Inward normal heat flux qn (SI unit: W/m2).

454 | CHAPTER 5: AEROACOUSTICS INTERFACES

Continuity
Continuity is available as an option at interfaces between parts in a pair. This condition
gives continuity in all the dependent variables: pressure, velocity, and temperature. It
corresponds to a situation where the boundary has no direct effect on the acoustic
fields (subscripts “src” and “dest” in the equation refers to the source and destination,
the two sides of the pair):

T
U src – U dest = 0, U = ( p t, u t, T t )

THE LINEARIZED NAVIER–STOKES, FREQUENCY DOMAIN INTERFACE | 455

 T he L i ne a r i z e d N av i er–St okes,
Transient Interface
The Linearized Navier–Stokes, Transient (lnst) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations in pressure, velocity, and temperature in the presence
of any stationary isothermal or nonisothermal background mean-flow. The physics
interface is used for aeroacoustic simulations that can be described by the linearized
Navier–Stokes equations.

The equations are formulated in the time domain and include viscous losses and
thermal conduction as well as the heat generated by viscous dissipation, if relevant. The
coupling between the acoustic field and the background flow does not include any
predefined flow induced noise.

Flow-induced noise functionality exists in pressure acoustics when using

the Aeroacoustic Flow Source feature and associated multiphysics
coupling and dedicated study.

The equations defined by the Linearized Navier–Stokes, Transient interface are the
linearized continuity, momentum (Navier–Stokes), and energy equations. The physics
interface solves for the acoustic variations in the pressure, velocity, and temperature.
The equations are formulated in the time domain for any fluid including losses due to
viscosity and thermal conduction. The background mean flow can be any stationary
flow.

The governing equations are defined using a scattered field formulation just as in The
Linearized Navier–Stokes, Frequency Domain Interface and can use Background
Acoustic Fields feature.

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Navier–Stokes Model, Wall, and Initial Values. For axisymmetric
components, an Axial Symmetry node is also added. Then, from the Physics toolbar, add
other nodes that implement, for example, boundary conditions and sources. You can
also right-click Linearized Navier–Stokes, Transient to select physics features from the
context menu.

456 | CHAPTER 5: AEROACOUSTICS INTERFACES

Coupling between a background mean flow, computed from a Fluid Flow model, and
the Linearized Euler model is handled by the Background Fluid Flow Coupling
multiphysics coupling and the dedicated Mapping study. Details are also found in the
Mapping Between Fluid Flow and Acoustics Mesh section

For modeling tips and tricks and good practice see the Modeling with the
Aeroacoustics Branch section.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is lnst.

TRANSIENT SOLVER SETTINGS

Enter the Maximum frequency to resolve in the model. The default frequency is set to
1000[Hz] but should be changed to reflect the frequency content of the sources used
in the model. Select the Time stepping (method) as Fixed (preferred) the default and
recommended or Free. The Free option is in general not recommended for wave
problems. The generated solver will be adequate in most situations if the
computational mesh also resolves the frequency content in the model. Note that any
changes made to these settings (after the model is solved the first time) will only be
reflected in the solver if Show Default Solver or Reset Solver to Defaults is selected in the
study.

Details about Time Stepping in Transient Models are found in the

Modeling with the Aeroacoustics Branch section.

The Stabilization, Dependent Variables, and Discretization settings are

the same as for The Linearized Navier–Stokes, Frequency Domain
Interface.

THE LINEARIZED NAVIER–STOKES, TRANSIENT INTERFACE | 457

 For boundary conditions and domain conditions see:

• Domain, Boundary, and Pair Nodes for the Linearized Navier–Stokes,

Frequency Domain and Transient Interfaces
• Theory Background for the Aeroacoustics Branch

458 | CHAPTER 5: AEROACOUSTICS INTERFACES

The Linearized Navier–Stokes,
Boundary Mode Interface
The Linearized Navier–Stokes, Boundary Mode (lnsbm) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute and identify propagating and nonpropagating modes in waveguides and
ducts in the presence of any stationary isothermal or nonisothermal background
mean-flow. The interface performs a boundary mode analysis on a boundary, inlet, or
cross section of the waveguide. The interface considers all thermal and viscous loss
effects and the interaction with the background flow. This includes the acoustic
boundary layer losses, if necessary. The interface exists in 3D and 2D axisymmetric
geometries and applies to boundaries.

The interface solves for the acoustic variations of pressure, velocity, and temperature,
as well as the out-of-plane wave number of the modes. The Linearized Navier-Stokes,
Boundary Mode interface is, for example, used when setting up sources in waveguide
systems in the presence of a background mean-flow. The background flow is assumed
to be uniform in the propagation direction.

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Navier–Stokes Model, Wall, and Initial Values. For axisymmetric
components, an Axial Symmetry node is also added. Then, from the Physics toolbar, add
other nodes that implement, for example, boundary conditions and sources. You can
also right-click Linearized Navier–Stokes, Boundary Mode to select physics features from
the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

THE LINEARIZED NAVIER–STOKES, BOUNDARY MODE INTERFACE | 459

 The default Name (for the first physics interface in the model) is lnsbm.

For boundary conditions and domain conditions see:

• Domain, Boundary, and Pair Nodes for the Linearized Navier–Stokes,

Frequency Domain and Transient Interfaces
• Theory Background for the Aeroacoustics Branch

Stabilization is not implemented for the Linearized Navier-Stokes,

Boundary Mode interface, which instead uses a default P1-P2-P2
discretization that ensures stability.

Linearized Navier-Stokes Model

The Linearized Navier-Stokes Model adds the governing equations, and defines the
background mean flow and the fluid properties. Gradient term suppression
stabilization (advanced physics option) can be selected, if needed. See the Linearized
Navier–Stokes Model for The Linearized Navier–Stokes, Frequency Domain Interface
for details.

BOUNDARY MODE SETTINGS

Define the Propagation direction of the computed modes. The default is to use the
surface physics normal (components [Link], [Link], and [Link]) that points
out of domains.

Wall
The Wall condition is the most common condition at solid surfaces. This condition
contains both a mechanical and a thermal selection. The default is a no-slip and
isothermal condition applicable in most cases. More advanced conditions at
boundaries can be set up by combining any of the Mechanical and Thermal conditions
available. See the Wall condition for The Linearized Navier–Stokes, Frequency
Domain Interface for details.

Initial Values
The Initial Values node adds initial values for the density, velocity, and pressure
variables.

460 | CHAPTER 5: AEROACOUSTICS INTERFACES

INITIAL VALUES
Enter a value or expression for the initial values Pressure p (SI unit: Pa), Velocity field
u (SI unit: m/s), and Temperature variation T (SI unit: K).

THE LINEARIZED NAVIER–STOKES, BOUNDARY MODE INTERFACE | 461

 The Linearized Potential Flow,
Frequency Domain Interface
The Linearized Potential Flow, Frequency Domain (ae) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations in the velocity potential in the presence of an inviscid
and irrotational background mean flow, that is, a potential flow. The background flow
can be modeled using The Compressible Potential Flow Interface. The physics
interface is used for convected acoustic simulations that can be described by the
linearized compressible potential flow equations.

The equations are formulated in the frequency domain and assume harmonic variation
of all sources and fields. The physics interface is limited to flows with a Mach number
Ma < 1, partly due to limitations in potential flow and partly due to the acoustic
boundary settings needed for supersonic flow. The coupling between the acoustic field
and the background flow does not include any predefined flow-induced noise.

When this physics interface is added, these default nodes are also added to the Model
Builder — Aeroacoustics Model, Sound Hard Boundary (Wall), and Initial Values. For
axisymmetric components an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Linearized Potential Flow,
Frequency Domain to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is ae.

462 | CHAPTER 5: AEROACOUSTICS INTERFACES

LINEARIZED POTENTIAL FLOW EQUATION SETTINGS

For 1D axisymmetric components, the Azimuthal wave number m

(dimensionless) is 0 by default. The Out-of-plane wave number kz (SI unit:
rad/m) is 0 rad/m by default.

For 2D components, the Out-of-plane wave number kz (SI unit: rad/m) is

0 rad/m by default.

For 2D axisymmetric components the Azimuthal wave number m

(dimensionless) is 0 by default.

SOUND PRESSURE LEVEL SETTINGS

The settings are the same as Sound Pressure Level Settings for the Pressure Acoustics,
Frequency Domain interface.

TYPICAL WAVE SPEED

The settings are the same as Typical Wave Speed for the Pressure Acoustics, Frequency
Domain interface.

DEPENDENT VARIABLES
This physics interface defines one dependent variable (field), the Velocity potential phi.
The name can be changed but the names of fields and dependent variables must be
unique within a model.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) for the
velocity potential, the default is Quadratic Lagrange.

Choosing between Lagrange and Serendipity Shape Functions has

influence on the number of DOFs solved for and on stability for distorted
mesh.

• Domain, Boundary, Edge, Point, and Pair Nodes for the Linearized
Potential Flow, Frequency Domain Interface
• Theory Background for the Aeroacoustics Branch

THE LINEARIZED POTENTIAL FLOW, FREQUENCY DOMAIN INTERFACE | 463

 • Flow Duct: Application Library path
Acoustics_Module/Aeroacoustics_and_Noise/flow_duct
• Doppler Shift: Application Library path
Acoustics_Module/Aeroacoustics_and_Noise/doppler_shift

Domain, Boundary, Edge, Point, and Pair Nodes for the Linearized
Potential Flow, Frequency Domain Interface
The Linearized Potential Flow, Frequency Domain Interface has these domain,
boundary, edge, point and pair nodes, listed in alphabetical order, available from the
Physics ribbon toolbar (Windows users), Physics context menu (Mac or Linux users),
or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• Continuity • Normal Mass Flow

• Initial Values • Normal Velocity
• Impedance, Interior Impedance, and • Periodic Condition
Pair Impedance • Plane Wave Radiation
• Interior Sound Hard Boundary (Wall) • Sound Hard Boundary (Wall)
• Linearized Potential Flow Model • Sound Soft Boundary
• Mass Flow Circular Source • Velocity Potential
• Mass Flow Edge Source • Vortex Sheet
• Mass Flow Point Source

For axisymmetric components, COMSOL Multiphysics takes the axial

symmetry (at r = 0) into account and automatically adds an Axial
Symmetry node to the component that is valid on the axial symmetry
edges/points only.

464 | CHAPTER 5: AEROACOUSTICS INTERFACES

 In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Linearized Potential Flow Model

The Linearized Potential Flow Model node adds the equations for frequency domain
aeroacoustics modeling. You here need to enter the material properties as well as the
background mean flow information.

LINEARIZED POTENTIAL FLOW MODEL

The default values for the Density ρ0 (SI unit: kg/m3) and the Mean flow speed of sound
c0 (SI unit: m/s) are taken From material. For User defined, enter other values or
expressions. Also enter values or expressions for the Mean flow velocity V (SI unit:
m/s).

It is important for the velocity field to be a solution to a compressible

potential flow simulation. It has to be an irrotational and inviscid flow, for
example, a constant flow field V is of this type or a solution modeled with
The Compressible Potential Flow Interface. Any other type of flow yields
nonphysical solutions for this formulation of the governing equations.

Coupling between the flow form The Compressible Potential Flow

Interface interface to the linearized potential flow model can be set up
using the Background Potential Flow Coupling coupling found under
the Multiphysics Couplings node. The feature appears when both physics
interfaces are present in a model.

Material discontinuities, that is jumps in the material properties, are not

handled by the interface. A jump due to the presence of a shear layer that
separates a stream from the free velocity field is handled using the Vortex
Sheet condition.

THE LINEARIZED POTENTIAL FLOW, FREQUENCY DOMAIN INTERFACE | 465

 Initial Values
The Initial Values node adds initial values for the velocity potential. For The Linearized
Potential Flow, Transient Interface it also adds initial values for the velocity potential,
first time derivative. If needed, add more Initial Values nodes from the Physics toolbar
or the main physics interface node’s context menu.

INITIAL VALUES
Enter a value or expression for the initial value of the Velocity potential phi
(SI unit: m2/s).

For The Linearized Potential Flow, Transient Interface also enter a Velocity potential,
first time derivative, ∂φ ⁄ ∂t (SI unit: m2/s2).

Sound Hard Boundary (Wall)

Use the Sound Hard Boundary (Wall) condition to model rigid boundary surfaces or
walls. It prescribes a vanishing normal component of the particle velocity at the
boundary. Multiplied by the density, it can equivalently be expressed as a no-flow
condition:

 ρ 0 ∂φ 
– n ⋅  ρ 0 ∇φ – V -----2-  + V ⋅ ∇φ  = 0
  ∂ t  
c0

The sound-hard boundary condition is available for all study types. The equation
above applies to the time-domain calculations in The Linearized Potential Flow,
Transient Interface; to obtain the corresponding condition for frequency domain,
simply replace ∂/∂t by iω.

In The Linearized Potential Flow, Boundary Mode Interface the no-flow or wall
condition, known as sound hard, sets the normal acceleration — and thus also the
normal velocity — to zero at the edge.

 ρ0 
– n ⋅  ρ 0 ∇φ – V -----2- ( iωφ + ( ∇φ ⋅ V ) ) = 0
 c0 

466 | CHAPTER 5: AEROACOUSTICS INTERFACES

Velocity Potential
Use the Velocity Potential node when coupling two Linearized Potential Flow,
Frequency Domain interfaces together because it can sometimes be necessary to set
the velocity potential: φ = φ 0 .

VELOCITY POTENTIAL
Enter a Velocity potential φ 0 (SI unit: m2/s).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Normal Mass Flow

Use the Normal Mass Flow node to set the inward mass flow boundary condition.

For The Linearized Potential Flow, Frequency Domain Interface, the natural
boundary condition for the total wave has the meaning of a mass flow through the
boundary surface:

 ρ 0 ∂φ 
– n ⋅  ρ 0 ∇φ – V -----2-  + V ⋅ ∇φ  = m n
 
 c0 ∂ t 

For The Linearized Potential Flow, Boundary Mode Interface, the natural edge
condition for the total wave has the meaning of normal mass flow.

 ρ0 
– n ⋅  ρ 0 ∇φ – V -----2- ( iωφ + ( ∇φ ⋅ V ) + λ φV n ) = m n
 c0 

NORMAL MASS FLOW

Enter an Inward mass flow mn (SI unit: kg/(m2·s)).

Plane Wave Radiation

The Plane Wave Radiation is a class of nonreflecting boundary conditions, which
assume that there is an outgoing plane wave, and optionally also an incoming exciting
wave.

For transient analysis the boundary condition is

THE LINEARIZED POTENTIAL FLOW, FREQUENCY DOMAIN INTERFACE | 467

  ρ 0 ∂φ  ∂φ ρ 0 ∂φ ∂φ
– n ⋅  ρ 0 ∇φ – V -----2-  + ( ∇φ ⋅ V )  – ρ 0 k n – n ⋅ V -----2-  – k n n ⋅ V =
 ∂t  ∂ t ∂t ∂t 
c0 c0

∂ ρ 0 ∂φ 0 ∂φ 0 ρ0 ∂φ 0
ρ0 φ k ( n ⋅ n k ) – n ⋅ V -----2- k k nk ⋅ V – ρ0 kn + n ⋅ V -----2-  k n n ⋅ V
∂t 0 k c ∂ t ∂ t c ∂t
0 0
1 1
k k = --------------------------- kn = ------------------------
c0 + V ⋅ nk c0 + V ⋅ n

while the corresponding time-harmonic equation reads

 ρ0  ρ0
– n ⋅  ρ 0 ∇φ – V -----2- ( iωφ + ∇φ ⋅ V ) – ρ 0 ik n φ – n ⋅ V -----2- ( iω – ik n n ⋅ V )φ =
 c0  c0

 V V  –i k ⋅ r
ρ 0  ik k n ⋅ n k – n ⋅ ----2- ( ik k n k ⋅ V ) – ik n + n ⋅ ----2- ( ik n n ⋅ V ) φ 0 e
 c0 c0 

ω ω ek
k k = --------------------------- k n = ------------------------ k = k nk n k = --------
c0 + V ⋅ nk c0 + V ⋅ n ek

Specify an Incident Velocity Potential (incoming plane wave)

– ik ⋅ r
φ0 e

by supplying its amplitude, φ 0 , and propagation wave direction vector, ek. The vector
nk is the normalized wave direction vector of unit length.

This boundary condition is most relevant for ports, because many

waveguide structures are only interesting in the plane-wave region.

Incident Velocity Potential

The Incident Velocity Potential subnode is available from the context menu (right-click
the Plane Wave Radiation parent node) or from the Physics toolbar, Attributes menu.
Use this to add a velocity potential and wave direction.

INCIDENT VELOCITY POTENTIAL

Enter a Velocity potential φ0 (SI unit: m2/s) and Wave direction ek (SI unit: m). The
default for the wave direction is the inward normal direction of the boundary.

468 | CHAPTER 5: AEROACOUSTICS INTERFACES

Sound Soft Boundary
The Sound Soft Boundary creates a boundary condition for a sound soft boundary,
where the acoustic pressure vanishes and p = 0.

This boundary condition is an appropriate approximation for a liquid-gas

interface and in some cases for external waveguide ports.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Periodic Condition
The Periodic Condition node adds a periodic boundary condition that can be used to
reduce the model size by using symmetries and periodicities in the geometry and
physics interfaces being modeled. This feature works well for cases like opposing
parallel boundaries. In other cases, use a Destination Selection subnode to control the
destination. By default, it contains the selection that COMSOL Multiphysics identifies.

PERIODICITY SETTINGS
Select a Type of periodicity: Continuity (the default) or Antiperiodicity.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

ORIENTATION OF SOURCE
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

See Orientation of Source and Destination in the COMSOL Multiphysics

Reference Manual.

THE LINEARIZED POTENTIAL FLOW, FREQUENCY DOMAIN INTERFACE | 469

 Normal Velocity
Use the Normal Velocity node in time-harmonic analysis to specify the velocity
component normal to the boundary:

 ρ0 
n ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) = ρ 0  v n + ------ V ⋅ ∇ v n
1
 c0  iω

Here vn denotes the outward normal velocity at the boundary surface, which is
specified in the vn text field. The formulation is a variant of the Ingard-Myers type
impedance condition as vn = p/Zi (see Impedance, Interior Impedance, and Pair
Impedance).

NORMAL VELOCITY
Enter a Normal velocity vn (SI unit: m/s).

Impedance, Interior Impedance, and Pair Impedance

Use the Impedance, Interior Impedance, or Pair Impedance node in time-harmonic
analysis to define the input impedance of an external domain, at an interior boundary,
or at the boundary between parts in an assembly as the ratio of pressure (or pressure
drop) to normal velocity, Zi = p/(n·v) at the boundary. The formulation is following
the so-called Ingard-Myers condition. The associated impedance boundary condition
is

 ρ0 
– n ⋅  ρ 0 ∇φ – -----2- ( iω φ + V ⋅ ∇φ )V = ρ 0  ----- + ------ ( V ⋅ ∇ ) -----
p 1 p
i = up, down
 c0  Z i iω Z i

On pair (interior) impedance conditions the up/down, that refers to the up and down
side of an interior boundary, is replaced by a 1 and 2 index in the equation display.
These refer to the two faces in the pair.

IMPEDANCE/INTERIOR IMPEDANCE/PAIR IMPEDANCE

Enter an input Impedance Zi (SI unit: Pa·s/m).

Vortex Sheet
Use the Vortex Sheet boundary condition to model a shear layer that separates a stream
from the free velocity field. Because the velocity potential is discontinuous over this

470 | CHAPTER 5: AEROACOUSTICS INTERFACES

boundary. Vortex sheets are only applicable on interior boundaries or between
assemblies.

The equations defining the vortex sheet boundary condition are

 ρ0 
n ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) = [ ρ 0 ( iω + V ⋅ ∇ ) w ] i i = up, down
 c0 
i
p up = p down w up = – w down

where w denotes the outward normal displacement (SI unit: m) of the vortex sheet
surface. The condition ensures continuity in the normal stress (the pressure) and the
normal displacement, while allowing for a slit (jump) in the tangential component of
the total velocity.

On pair (interior) vortex sheet conditions the up/down, that refers to the up and
down side of an interior boundary, is replaced by a 1 and 2 index in the equation
display (referring to the two faces in the pairs).

This boundary condition adds and additional degree of freedom ae.vs1.w (here using
the default name for the physics interface and boundary condition) which is solved for.

Interior Sound Hard Boundary (Wall)

For The Linearized Potential Flow, Frequency Domain Interface and The Linearized
Potential Flow, Transient Interface, use the Interior Sound Hard Boundary (Wall)
condition to model interior rigid boundary surfaces, or walls. It prescribes a vanishing
normal component of the particle velocity at the boundary. Multiplied by the density,
it can equivalently be expressed as a no-flow condition:

 ρ 0 ∂φ 
– n ⋅  ρ 0 ∇φ – V -----2-  + V ⋅ ∇φ  = 0 i = up, down
  ∂ t  
c0 i

The up/down refers to the up and down side of an interior boundary. The sound-hard
boundary condition is available for all analysis types. The equation above applies to the
time domain calculations in the Linearized Potential Flow, Transient interface; to
obtain the corresponding condition for the frequency domain, simply replace ∂/∂t with
iω.

THE LINEARIZED POTENTIAL FLOW, FREQUENCY DOMAIN INTERFACE | 471

 Continuity
Continuity is available as an option at interfaces between parts in a pair. This condition
gives continuity in the velocity potential as well as continuity in the mass flow. It
corresponds to a situation where the boundary has no direct effect on the acoustic
velocity potential field (subscripts 1 and 2 in the equation refer to the two sides of the
pair):

  V    V 
n ⋅  ρ 0  ∇φ – ----2- ( iω + V ⋅ ∇ )φ  –  ρ 0  ∇φ – ----2- ( iω + V ⋅ ∇ )φ  = 0
  c0    c0 
1 2

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Mass Flow Edge Source

For 3D components, use a Mass Flow Edge Source to specify the mass flow rate on an
edge:

ρ0  ρ0  iθ
– -----2- iω ( iωφ + V ⋅ ∇φ ) + ∇ ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) = m'e δ ( x – x 0 )dl
c0  c0 

In the time domain, factors iω are replaced by partial time derivatives, ∂/∂t.

MASS FLOW EDGE SOURCE

Enter a Mass flow rate m' (SI unit: kg/(m2·s)) and enter a Phase θ (dimensionless).

Mass Flow Point Source

Add a Mass Flow Point Source node to specify the mass flow rate on a point:

ρ0  ρ0  iθ
– -----2- iω ( iωφ + V ⋅ ∇φ ) + ∇ ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) = m'e δ ( x – x 0 )
c0  c0 

In the time domain, factors iω are replaced by partial time derivatives, ∂/∂t.

472 | CHAPTER 5: AEROACOUSTICS INTERFACES

MASS FLOW POINT SOURCE
Enter a Mass flow rate m' (SI unit: kg/s for 3D and 2D axisymmetric components;
kg/(m·s) for 2D components).

Enter a Phase θ (dimensionless).

Doppler Shift: Application Library path

Acoustics_Module/Aeroacoustics_and_Noise/doppler_shift

Mass Flow Circular Source

For 2D axisymmetric components, use a Mass Flow Circular Source node to add a
circular source located at x = x0:

ρ0  ρ0  iθ
– -----2- iω ( iωφ + V ⋅ ∇φ ) + ∇ ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) = m'e δ ( x – x 0 )rdϕ
c0  c0 

In the time domain, factors iω are replaced by partial time derivatives, ∂/∂t.

MASS FLOW CIRCULAR SOURCE

Enter a Mass flow rate m' (SI unit: kg/(m⋅s) and enter a Phase θ (dimensionless).

Mass Flow Line Source on Axis

For 2D axisymmetric components, use a Mass Flow Line Source on Axis node to add a
line source along the symmetry axis:

ρ0  ρ0  iθ
– -----2- iω ( iωφ + V ⋅ ∇φ ) + ∇ ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) = m'e δ ( x – x 0 )dz
c0  c0 

In the time domain, factors iω are replaced by partial time derivatives, ∂/∂t.

MASS FLOW LINE SOURCE ON AXIS

Enter a Mass flow rate m' (SI unit: kg/(m·s)) and enter a Phase θ (dimensionless).

Axial Symmetry
The Axial symmetry feature is a default node added for all axisymmetric components.
The boundary condition is active on all boundaries on the symmetry axis.

THE LINEARIZED POTENTIAL FLOW, FREQUENCY DOMAIN INTERFACE | 473

 The Linearized Potential Flow,
Transient Interface
The Linearized Potential Flow, Transient (aetd) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations in the velocity potential in the presence of an inviscid
and irrotational background mean flow, that is, a potential flow. The background flow
can be modeled using The Compressible Potential Flow Interface. The physics
interface is used for convected acoustic simulations that can be described by the
linearized compressible potential flow equations.

The equations are formulated in the time domain. The physics interface is limited to
flows with a Mach number Ma < 1, partly due to limitations in potential flow and
partly due to the acoustic boundary settings needed for supersonic flow. The coupling
between the acoustic field and the background flow does not include any predefined
flow-induced noise.

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Potential Flow Model, Sound Hard Boundary (Wall), and Initial
Values. For axisymmetric components an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Linearized Potential Flow,
Transient to select physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is aetd.

TRANSIENT SOLVER SETTINGS

Enter the Maximum frequency to resolve in the model. The default frequency is set to
1000[Hz] but should be changed to reflect the frequency content of the sources used

474 | CHAPTER 5: AEROACOUSTICS INTERFACES

in the model. Select the Time stepping (method) as Fixed (preferred) the default and
recommended or Free. The Free option is in general not recommended for wave
problems. The generated solver will be adequate in most situations if the
computational mesh also resolves the frequency content in the model. Note that any
changes made to these settings (after the model is solved the first time) will only be
reflected in the solver if Show Default Solver or Reset Solver to Defaults is selected in the
study.

Details about Time Stepping in Transient Models are found in the

Modeling with the Aeroacoustics Branch section.

The remainder of the settings are shared with The Linearized Potential
Flow, Frequency Domain Interface.

• Domain, Boundary, Edge, Point, and Pair Nodes for the Linearized
Potential Flow, Transient Interface
• Theory Background for the Aeroacoustics Branch

Domain, Boundary, Edge, Point, and Pair Nodes for the Linearized
Potential Flow, Transient Interface
The Linearized Potential Flow, Transient Interface has these domain, boundary, edge,
point, and pair nodes, listed in alphabetical order and described for The Linearized
Potential Flow, Frequency Domain Interface. The nodes are available from the Physics
ribbon toolbar (Windows users), Physics context menu (Mac or Linux users), or
right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE LINEARIZED POTENTIAL FLOW, TRANSIENT INTERFACE | 475

 • Continuity • Mass Flow Edge Source
• Initial Values • Normal Mass Flow
• Interior Sound Hard Boundary (Wall) • Periodic Condition
• Linearized Potential Flow Model • Plane Wave Radiation
• Mass Flow Point Source • Sound Hard Boundary (Wall)
• Mass Flow Circular Source • Velocity Potential

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

476 | CHAPTER 5: AEROACOUSTICS INTERFACES

The Linearized Potential Flow,
Boundary Mode Interface
The Linearized Potential Flow, Boundary Mode (aebm) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ) when adding a physics interface, is used to
compute eigenmodes and out-of-plane wave numbers for the linearized compressible
potential flow equations. This study is used, for example, when specifying sources at
inlets or analyzing transverse acoustic modes in ducts.

The physics interface solves an eigenvalue equation on boundaries, searching for the
out-of-plane wave numbers at a given frequency.

When this physics interface is added, these default nodes are also added to the Model
Builder — Linearized Potential Flow Model, Sound Hard Boundary (Wall), and Initial
Values. For 2D axisymmetric models an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Linearized Potential Flow,
Boundary Mode to select physics features from the context menu.

This physics interface is limited to flows with a Mach number Ma < 1,

partly due to limitations in the potential flow formulation and partly due
to the acoustic boundary settings needed for supersonic flow.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is aebm.

THE LINEARIZED POTENTIAL FLOW, BOUNDARY MODE INTERFACE | 477

 LINEARIZED POTENTIAL FLOW EQUATION SETTINGS
For 2D axisymmetric components, the Azimuthal mode number m (dimensionless) is 0
by default. It is an integer entering the axisymmetric expression for the velocity
potential:

– i ( k z z + mϕ )
φ ( r, z, ϕ ) = φ ( r ) e

SOUND PRESSURE LEVEL SETTINGS

The settings are the same as Sound Pressure Level Settings for the Pressure Acoustics,
Frequency Domain interface.

DEPENDENT VARIABLES
This physics interface defines one dependent variable (field), the Velocity potential phi.
The name can be changed but the names of fields and dependent variables must be
unique within a model.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) the default
is Quadratic Lagrange.

Choosing between Lagrange and Serendipity Shape Functions has

influence on the number of DOFs solved for and on stability for distorted
mesh.

• Boundary, Edge, Point, and Pair Nodes for the Linearized Potential
Flow, Boundary Mode Interface
• Theory Background for the Aeroacoustics Branch

Flow Duct: Application Library path

Acoustics_Module/Aeroacoustics_and_Noise/flow_duct

Boundary, Edge, Point, and Pair Nodes for the Linearized Potential
Flow, Boundary Mode Interface
The Linearized Potential Flow, Boundary Mode Interface has these boundary, edge,
point, and pair nodes, listed in alphabetical order, available from the Physics ribbon

478 | CHAPTER 5: AEROACOUSTICS INTERFACES

toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users).

These features are described for The Linearized Potential Flow, Frequency Domain
Interface. The only difference for the Linearized Potential Flow, Boundary Mode
interface is that you apply the features to boundaries instead of domains for 3D
components. In the Linearized Potential Flow Model you can enter a Unit normal that
defines the positive propagation direction.

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using.

• Continuity • Sound Hard Boundary (Wall)

• Initial Values • Sound Soft Boundary
• Linearized Potential Flow Model • Velocity Potential
• Normal Mass Flow

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

THE LINEARIZED POTENTIAL FLOW, BOUNDARY MODE INTERFACE | 479

 The Compressible Potential Flow
Interface
The Compressible Potential Flow (cpf) interface ( ), found under the
Acoustics>Aeroacoustics branch ( ), is used to compute the velocity potential and
density in a compressible potential flow model. Derived values include the associated
pressure, velocity, and temperature in the flow. In a compressible potential flow model,
the flow is assumed to be represented by an ideal barotropic, irrotational fluid at
constant entropy, that is, the fluid is also inviscid. The physics interface is used for
modeling the background mean flow used as input to The Linearized Potential Flow,
Frequency Domain Interface or The Linearized Potential Flow, Transient Interface.

When this physics interface is added, these default nodes are also added to the Model
Builder — Compressible Potential Flow, Slip Velocity, and Initial Values. For axisymmetric
models an Axial Symmetry node is also added.

Then, from the Physics toolbar, add other nodes that implement, for example,
boundary conditions and sources. You can also right-click Compressible Potential Flow
to select physics features from the context menu.

The potential flow formulation for steady compressible flow is in general

not suited for modeling shocks. In the region after a shock the flow is
typically rotational, hence it is only suited for problems with a Mach
number Ma < 1.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is cpf.

480 | CHAPTER 5: AEROACOUSTICS INTERFACES

REFERENCE VALUES
Edit or enter the values as needed:

• Reference pressure pref (SI unit: Pa). The default is 1 atm.

• Reference density ρref (SI unit: kg/m3). The default is 1.2 kg/m3.
• Reference velocity νref (SI unit: m/s).
• Reference force potential Ψref (SI unit: J/kg).

DEPENDENT VARIABLES
This physics interface defines two dependent variables (field), the Mean flow velocity
potential Phi and the Density rho. The name can be changed but the names of fields
and dependent variables must be unique within a model.

DISCRETIZATION
Select the discretization order for the Mean flow velocity potential and the Density
dependent variables. The default is Quadratic for both.

• Domain, Boundary, and Pair Nodes for the Compressible Potential

Flow Interface
• Theory Background for the Aeroacoustics Branch

Flow Duct: Application Library path

Acoustics_Module/Aeroacoustics_and_Noise/flow_duct

Domain, Boundary, and Pair Nodes for the Compressible Potential

Flow Interface
The Compressible Potential Flow Interface has these domain, boundary, and pair
nodes, listed in alphabetical order, available from the Physics ribbon toolbar (Windows

THE COMPRESSIBLE POTENTIAL FLOW INTERFACE | 481

 users), Physics context menu (Mac or Linux users), or right-click to access the context
menu (all users).

• Compressible Potential Flow Model • Normal Flow

• Initial Values • Periodic Condition
• Interior Wall (Slip Velocity) • Slip Velocity
• Mass Flow • Symmetry
• Mean Flow Velocity Potential

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Compressible Potential Flow Model

The Compressible Potential Flow Model node adds equations for time-dependent or
stationary modeling of compressible potential flow. The model will solve for the mean
flow velocity potential and the density. The pressure and the velocity field are defined
from these dependent variables.

COMPRESSIBLE POTENTIAL FLOW MODEL

For the Ratio of specific heats γ (dimensionless) select From material (the default) or
User defined. For the User defined option, the default value is 1.4.

Enter a Force potential ψ (SI unit: J/kg). To model a domain force or domain source
acting on the fluid.

Select the Calculate temperature option it you need to also calculate the temperature
field in the fluid. If this is selected, you also need to specify the Specific gas constant Rs
(SI unit: J/(kg·K)). Select From material (the default) or User defined. Enabling this
calculates the temperature according to

482 | CHAPTER 5: AEROACOUSTICS INTERFACES

 pA
T = ----------
ρR s

where pA is the pressure and ρ the density in the fluid.

Depending on the material selected in the Materials node for the fluid (typically Air)
the Model Inputs may be active. Per default the Absolute pressure is taken from the
model itself (unavailable). If the Calculate temperature option is enabled the
Temperature is also taken from the model (unavailable). This will give a physically
consistent solution. In both cases, you can click the Make All Model Inputs Editable icon
to enter a user defined expression.

Note that the interface can be solved for in domains that have a PML
defined. The PML does not have any influence in the solution, as it does
not represent an acoustic perturbation. This means that the background
flow solution can be used directly as input, also in PML domains, when
solving a model with The Linearized Potential Flow, Frequency Domain
Interface.

Initial Values
The Initial Values node adds initial values for the mean flow velocity potential and
density variables. Add more Initial Values nodes from the Physics toolbar.

INITIAL VALUES
Enter a value or expression for the initial values Mean flow velocity potential Phi
(SI unit: m2/s). Enter a Density rho (SI unit: kg/m3) the default is the reference
density [Link].

Slip Velocity
The Slip Velocity node is the natural condition at a boundary impervious to the flow,
meaning that the velocity normal to the boundary is zero. By multiplying with the
density, this condition can alternatively be expressed as a vanishing mass flow through
the boundary

– n ⋅ ρ∇Φ = 0

THE COMPRESSIBLE POTENTIAL FLOW INTERFACE | 483

 Symmetry
The Symmetry condition is applied at planes where the model has symmetries. Using
the intrinsic symmetries of a problem is a simple way to reduce the model size.

Normal Flow
The Normal Flow node implies that the flow is normal to the boundary and thus that
the tangential velocity is zero. This corresponds to a constant velocity potential along
the boundary. Because the velocity potential is determined only up to a constant,
imposing this condition fixes the arbitrary constant to zero where φ = 0 .

Setting the Normal Flow condition on two or more disjoint boundaries

can result in the wrong features unless symmetry implies that the velocity
potential is equal on the boundaries in question.

Mass Flow
The Mass Flow node specifies the mass flow through the boundary. The mass flow is
given by the product of the normal velocity vn and the density at the boundary ρbnd.
This results in a flux condition given by

– n ⋅ ρ∇Φ = v n ρ bnd

MASS FLOW
Enter the Normal velocity vn (SI unit: m/s) and Fluid density at the boundary ρbnd
(SI unit: kg/m3). The defaults are [Link] and [Link], respectively. The
values given in the Reference Values.

Mean Flow Velocity Potential

Use the Mean Flow Velocity Potential condition to prescribe the flow potential Φ at a
boundary, Φ = Φ0. This condition can be used to set up theoretical flow conditions at
a boundary or couple to other physics interfaces.

MEAN FLOW VELOCITY POTENTIAL

Enter the Mean flow velocity potential Φ0 (SI unit: m2/s).

484 | CHAPTER 5: AEROACOUSTICS INTERFACES

Periodic Condition
Use the Periodic Condition to define periodicities between boundaries in the model.
From the Type of periodicity menu select Continuity (default) or Antiperiodicity.

Interior Wall (Slip Velocity)

The Interior Wall (Slip Velocity) is used to set up interior walls on interior boundaries.
This simplifies modeling and meshing steps as the wall does not need to have a
thickness. The condition is simply applied to an interior boundary in the model. The
condition enforces a vanishing mass flow on the up and down sides of the wall.

( – n ⋅ ρ∇Φ ) up = 0 ( – n ⋅ ρ∇Φ ) down = 0

The dependent variables are defined as being discontinuous (so called slits) at this
boundary.

THE COMPRESSIBLE POTENTIAL FLOW INTERFACE | 485

 Modeling with the Aeroacoustics
Branch
In this section:

• Selecting an Aeroacoustics Interface

• Meshing
• Stabilization
• Solver Suggestions for Large Aeroacoustic Models
• Absorbing Layers for the Linearized Euler, Transient Interface
• Lagrange and Serendipity Shape Functions
• Time Stepping in Transient Models
• Mapping Between Fluid Flow and Acoustics Mesh
• Coupling to Turbulent Flows (Eddy Viscosity)
• Eigenfrequency Studies
• Suppressing Constraints on Lower Dimensions
• Postprocessing Variables

Selecting an Aeroacoustics Interface

When modeling a muffler with an internal nonisothermal flow, a jet engine, or a flow
sensor you should consider which physics interface to use. The influence the
background mean flow has on the acoustic behavior in an aeroacoustic model can be
modeled in several ways. The effects that need to be included typically depend on the
Mach number (Ma). A rule of thumb says that for a Mach number below 0.1
(Ma < 0.1) the convective effects of the background flow need not to be included,
above they do. Other considerations are of course also important; for example,
whether viscous and thermal losses are important, if the background flow has large
gradients, or if the flow is turbulent or has vorticity.

• Mach number less than 0.1 (Ma < 0.1)

In this situation the convective flow effects can normally be neglected and Pressure
Acoustics can be used. Only the background temperature distribution T = T(x) and
background pressure distribution pA = pA(x) need to be included. This can be done

486 | CHAPTER 5: AEROACOUSTICS INTERFACES

 directly in the Pressure Acoustics interface as a Model Inputs. The effects are included
by making sure that the material properties depend on the local pressure and
temperature; the spatial variations in the speed of sound is c = c(pA,T) = c(x) and in
the density ρ = ρ(pA,T) = ρ(x), respectively.
• Mach number greater than 0.1 (Ma > 0.1):
In this case the convective effects of the flow can probably not be neglected.
Modeling this type of system requires the use of one of the Aeroacoustics interfaces:
linearized potential flow, linearized Euler, or linearized Navier–Stokes. The choice
depends on the assumptions about the flow that can be made.

Note that when the background mean flow velocity u0 is set to zero, the
linearized Navier–Stokes equations reduce to the thermoviscous acoustic
equations. However, in thermoviscous acoustics the default discretization
is P1, P2, P2 and no stabilization is applied.

Meshing
When solving a model using one of the Aeroacoustic interfaces, it is important to
remember that this is a wave problem. This means that the wavelength should be
resolved by an appropriate number of mesh elements. Using the same guidelines as for
Pressure Acoustics is reasonable, that is, a first good mesh should use at least 5
elements per wavelength for second order shape functions. In the linearized Euler
interface, the default shape functions are first order and here using at least 15 to 20
elements per wavelength is a first good choice.

Meshing guidelines for the Pressure Acoustics interface is given in

Meshing: Resolving the Waves in Space under the Modeling with the
Pressure Acoustics Branch (FEM-Based Interfaces) section.

The Background Flow

Another important parameter to consider is resolving details in the background mean
flow field. The acoustic mesh should capture gradients that exist in the background
flow as these have a large influence on the acoustic propagation. Waves may, for
example, be reflected and refracted in shear layers.

MODELING WITH THE AEROACOUSTICS BRANCH | 487

 Meshing the Acoustic Boundary Layers in LNS
The Linearized Navier–Stokes interfaces just as the Thermoviscous Acoustics interface
captures the physics of the acoustic boundary layer. When no-slip and isothermal
conditions are used on walls a viscous and a thermal boundary layer will exist. In order
for the solution to be well behaved and capture losses correctly, it is important to mesh
this layer, for example, by using a boundary layer mesh. If the model is large using a
single boundary layer mesh, with the approximate extend of the acoustic boundary
layer, is the minimum requirement for a good solution.

Resolving Vorticity and Physics in LE and LNS

In the LE and LNS interfaces the GLS stabilization is very efficient and can ensure
smooth and converged solution even without resolving details like vorticity generation
(the propagation of vorticity waves). If these are important processes in the model the
mesh should of course be able to resolve these details. Either refine the mesh or switch
to (P2, P2, P2) discretization (keep the stabilization turned on). For LNS models
these phenomena are typically generated at walls where the no-slip condition generates
vorticity when an acoustic wave is interacting with the background flow.

Stabilization
When solving the linearized Euler (LE) and linearized Navier–Stokes (LNS) equations
using the finite element method (the standard Galerkin formulation). It can be shown
that the method looses its good approximation characteristics when convective terms
are present and when these terms locally dominate (Ref. 19). This can lead to spurious
numerical oscillations. To remedy these oscillations stabilization is used.

It can be shown that if the cell Péclet number has a value larger than one
Pec > 1 the solution will oscillate. In the linearized Navier–Stokes
interface the variables lnsf.CellPe_th and lnsf.CellPe_v can be
plotted to assess the characteristic number when comparing convection to
viscous and thermal diffusion, respectively.

THE STABILIZATION METHODS

In the linearized Navier–Stokes (LNS) and linearized Euler (LE) interfaces two
stabilization methods are implemented. The Galerkin least squares (GLS) and the
streamline upwind Petrov-Galerkin (SUPG) stabilization. It is in general
recommended to use the GLS method. The SUPG method is implemented for
completeness and can be used by experienced users. There is also the option to not use

488 | CHAPTER 5: AEROACOUSTICS INTERFACES

stabilization or to use the legacy method from version 5.2a and older. The legacy
method for the LE interface is based on the implementation discussed in Ref. 10.

The implementation of the GLS stabilization methods follows the one discussed in
Ref. 14, 19, 21, and 23. A general overview about stabilization methods can be found
in Ref. 22. The GLS method combines stability and accuracy and the method order of
accuracy is O(h2p+1) where p is the shape function order and h is the local mesh size.
The SUPG method has an accuracy order of O(hp+1/2) for convection dominated
problems and O(hp+1) for diffusion dominated problems.

The default discretization for the LNS and LE interface is to use linear elements for all
the dependent variables (P1-P1-P1). This effectively removes the stabilization on the
diffusive parts of the equation. The GLS method is still superior to the SUPG method
as it also stabilizes the reactive terms. These are the terms where gradients of the
background mean flow fields enter. If no stabilization is used then set a P1-P2-P2
discretization for the dependent variables as it ensures a more stable numerical scheme.

The equation residuals can be visualized in the LE and the LNS interfaces
by plotting, for example, lnsf.res_e (equation residual for the energy
equation), lnsf.res_mx (equation residual for the momentum equation
x-component), lnsf.res_my, lnsf.res_mz, and lnsf.res_e (equation
residual for the energy equation).

See for example the model Point Source in 2D Jet: Radiation and
refraction of sound waves through a 2D shear layer found in the Model
Gallery:
[Link]/model/point-source-in-2d-jet-radiation-and-refractio
n-of-sound-waves-through-a-2d-shea-16685

Solver Suggestions for Large Aeroacoustic Models

Studying aeroacoustic applications often lead to very large models that can contain
many degrees of freedom (DOFs). Different steps can be taken to reduce the size of
the models, for example, consider the use of slip instead of no-slip conditions in the
LNS interface (avoiding the details in the acoustic boundary layer) or consider the
choice between Lagrange and Serendipity Shape Functions.

MODELING WITH THE AEROACOUSTICS BRANCH | 489

 When a model reaches a size where it can no longer fit in memory (when solved using
the default direct solver) or if the solution procedure is slow with a direct solver, then
it can be advantageous to switch to an iterative solver approach.

For the Linearized Navier–Stokes interfaces iterative solver suggestions are

automatically generated when the default solver is generated. These can be seen by
expanding the Solver Configuration tree under Stationary Solver or Time-Dependent
Solver. Per default a direct solver is used and two iterative solvers are suggested and
disabled. To turn on one of these approached right-click the solver and select Enable
(or press F4). The first suggestion (GMRES with Direct Precon.) uses an iterative
solver with a direct preconditioner. This method is typically faster than the direct solver
and uses 20% less memory. The second suggestion (GMRES with DD) uses an
iterative solver with the domain decomposition method. This method is very robust
(also for multiphysics applications) and very memory efficient, but it can be slow.

For the Linearized Euler interfaces the default solver is PARDISO, which is also the
most robust option. Iterative solver will often struggle for the linearized Euler
equations because there is not dissipation involved and instabilities are inherent.

Further details on solving large models can be found in the Solver

Suggestions for Large Thermoviscous Acoustics Models section. These
suggestions also apply to the linearized Navier–Stokes physics.

SOLVING IN THE ADIABATIC CASE

In certain cases, it is a good assumption to not include thermal conduction in the
model and treat all processes as adiabatic (isentropic). This is, for example, relevant for
fluids where the thermal boundary layer is much thinner than the viscous, like in water.
Not solving for the temperature field T also saves some degrees of freedom (DOFs).

This is achieved by selecting the Adiabatic formulation option under the Linearized
Navier–Stokes Equation Settings section. When Adiabatic formulation is selected all
temperature options and conditions are disabled in the user interface.

Absorbing Layers for the Linearized Euler, Transient Interface

The Linearized Euler, Transient Interface includes so-called absorbing layers, also
often referred to as sponge layers. The layers work by combining three techniques: a
scaling system, damping, and a simple nonreflecting condition.

490 | CHAPTER 5: AEROACOUSTICS INTERFACES

The layers are set up by adding the Absorbing Layer under the Definitions node. This
adds a special scaled system. The scaling effectively slows down the propagating waves
and ensures that they “align up” (normal) toward the outer boundary. This means that
the waves will hit the outer boundary in a closer to normal direction. Damping in the
form of artificial numerical viscosity, is automatically applied in the layer in a matched
manner (increasing from zero at the boundary to the physical domain). The artificial
viscosity attenuates or damps high-frequency components of the wave in the layer. The
amount of damping is controlled under the Damping Parameters for Absorbing Layers
section at the main physics level. Finally, at the outer boundary of the layer add a simple
Acoustic Impedance (Isentropic) boundary condition, which will work well to remove
all remaining waves, as normal incidence has been ensured.

The physical thickness of the layer should be such that it contains a minimum of 10
structured mesh elements. This in practice means that it should roughly have a
thickness equal to the largest wavelength resolved in the model.

For the Absorbing Layers select the Type (Cartesian, cylindrical, spherical, or user
defined) under the Geometry section. Enter values for the Physical Width and Pole
Distance under the Scaling section.

For more details on the Geometry and Scaling see the Infinite Elements,
Perfectly Matched Layers, and Absorbing Layers section in the COMSOL
Multiphysics Reference Manual.

Lagrange and Serendipity Shape Functions

In most of the physics interfaces in the Acoustics Module and specifically in the
Aeroacoustics interfaces, you can choose between two families of shape functions in
the Discretization section: Lagrange and serendipity. The current default is to use
Lagrange shape functions.

When using a structured mesh, it may be advantageous to switch to the serendipity

elements as they generate significantly fewer degrees of freedom (DOFs). The accuracy
is in most cases almost as good as for the Lagrange elements. The Lagrange elements
are however less sensitive to strong mesh distortions.

MODELING WITH THE AEROACOUSTICS BRANCH | 491

 The serendipity shape functions differ from the Lagrange shape functions only for the
following element shapes:

• 2D: Quadrilateral elements of discretization order higher than 1.

• 3D: Hexahedral, prism, and pyramid elements of discretization order higher than 1.

In the COMSOL Multiphysics Reference Manual:

• The Lagrange Element (shlag)

• The Nodal Serendipity Element (shnserp)

When coupling two physics interfaces that have the same DOFs like, for example,
displacement, the same type of shape functions should be used in both interfaces to
ensure conformity. Since there is no difference between the two families of shape
functions in 1D, this is not an issue when connecting edges.

Time Stepping in Transient Models

When solving transient wave problems, it is recommended to use a fixed time stepping
in the solver (this is the manual option in the solver settings). In general, there is no
point in using an automatic time-step control which can be provided by the
time-dependent solver. The tolerances in the automatic error control are difficult to
tune in wave problems when there is weak but important high-frequency content. It
is recommended to use the suggestion generated when the Fixed (preferred) method is
chosen (the default) in the Transient Solver Settings section. The internal time step of
the solver that is generated, when the Maximal frequency to resolve is set, will in most
cases produce a solver that is adequate. It is assuming that the user has generated a
mesh that properly resolves the same maximal frequency (minimal wavelength).

For further details see the Time Stepping: Resolving the Waves in Time
section in Modeling with the Thermoviscous Acoustics Branch.

Mapping Between Fluid Flow and Acoustics Mesh

When the aeroacoustic and the Fluid Flow (CFD) model are solved on different
computational mesh, careful mapping of the CFD solution from the CFD mesh onto
the acoustics mesh should be done. This step is crucial in order not to introduce
nonphysical numerical noise into the acoustic solution (Ref. 20). If the solution is not

492 | CHAPTER 5: AEROACOUSTICS INTERFACES

properly mapped terms containing gradients of the background mean flow variables
(especially the reactive terms) become very noisy and are typically the source of the
numerical error. This mapping is handled by the multiphysics coupling Background
Fluid Flow Coupling and the dedicated Mapping study.

If the acoustic problem is solved on the same mesh as the CFD then the mapping
problem is less of an issue, especially if both physics use the same shape function order
(they are both linear per default). Using the Background Fluid Flow Coupling is also
necessary, if perfectly matched layers (PMLs) are present in the model, in order to get
the flow field inside the PML regions. In any case, the coupling can still be useful as
the CFD solution will be smoothed by the mapping (the degree of smoothing can be
controlled).

The multiphysics coupling of turbulent flow and acoustics is described in

the application library model Helmholtz Resonator with Flow:
Interaction of Flow and Acoustics. The Application Library path
Acoustics_Module/Aeroacoustics_and_Noise/helmholtz_resnoator_with_flow

MANUAL MAPPING
The multiphysics coupling Background Fluid Flow Coupling can couple all
Single-Phase Flow and Nonisothermal Flow interfaces including the high Mach
Number Flow interfaces. If such a coupling and mapping is necessary, manual mapping
can be set up using the procedure described here.

The manual mapping is defined by a study that solves an additional set of equations
which maps and smooths the background flow variables onto the acoustics mesh. The
mapping equations can be set up using the Weak Form PDE interface from the
Mathematics branch. In the most general case the mean background flow pressure p0,
velocity field u0, density ρ0, temperature T0, and turbulent viscosity μT variables
should be mapped onto corresponding variables on the acoustics mesh, for example,
p0,aco, u0,aco, ρ0,aco, [Link], and μT,aco. These new variables should then be used as
the model inputs in the aeroacoustics model.

A simple but efficient mapping and smoothing is achieved by solving

MODELING WITH THE AEROACOUSTICS BRANCH | 493

 2
p 0, aco – p 0 = δh ∇ ⋅ ( ∇p 0, aco )
2
u i, 0, aco – u i, 0 = δh ∇ ⋅ ( ∇u i, 0, aco )
2
ρ 0, aco – ρ 0 = δh ∇ ⋅ ( ∇ρ 0, aco )
2
T 0, aco – T 0 = δh ∇ ⋅ ( ∇T 0, aco )
2
μ T, aco – μ T = δh ∇ ⋅ ( ∇μ T, aco )

where the term on the right-hand side adds smoothing using isotropic diffusion. The
amount of diffusion is controlled by the parameter δ (a constant that can be tuned with
a typical value of 0.01) and the mesh size squared h2. The term corresponds to
so-called source term stabilization as known from CFD.

In the Weak Form PDE interface define as many dependent variables as necessary and
give them the same shape order as the order used for the acoustics (typically all linear).
Assume that the source CFD variable for the x-velocity component is u and the new
destination variable is U0, then the above mapping is achieved with the following Weak
Expression, entered in the user interface:

(U0-withsol('sol1',u,setval(Ma,0.1)))*test(U0)+delta*h^2*(U0x*tes
t(U0x)+U0y*test(U0y)+U0z*test(U0z)

For the pressure map the total pressure [Link] to P0 using:

(P0-withsol('sol1',[Link],setval(Ma,0.1)))*test(P0)+delta*h^2*(P0
x*test(P0x)+P0y*test(P0y)+P0z*test(P0z))

Notice the use of the withsol() operator, which is an extrusion-like coupling

operator that can refer directly to a solution object and parameter value. In this
example it is used to fetch data from the CFD mesh, in the solution generated by solver
sol1 and for the parameter value Ma = 0.1. The setval() statement is optional. It is
important to use such an operator such that the CFD solution is mapped and
interpolated correctly to the integration (Gauss) points on the acoustics mesh.

In the COMSOL Multiphysics Reference Manual:

• The Weak Form PDE

• Operators, Functions, and Constants and Built-In Operators for details
about the withsol() operator
• Numerical Stabilization (general introduction)

494 | CHAPTER 5: AEROACOUSTICS INTERFACES

In some cases it can be necessary to extend the mapping equation with boundary
conditions for the mapped variables. This should be done at symmetry planes for the
mapped velocity field. This is also the cases when flow details near walls are important,
for example, resolving the acoustic boundary layer. In this case add a no-slip condition
on u0,aco. In general constraining the value of the mapped variables to the value in the
CFD model values at boundaries can be necessary.

Coupling to Turbulent Flows (Eddy Viscosity)

When acoustic waves propagate through a turbulent flow they will experience
attenuation due to the turbulence. It can be shown theoretically (see Ref. 18 and
Ref. 19) that the experienced attenuation corresponds to extending the value of the
dynamic viscosity μ to include the eddy viscosity μτ, such that the total effective
viscosity is:

μt = μ + μτ

This effect can be included in the linearized Navier–Stokes model by using the effective
dynamic viscosity as taken from a CFD model spf.mu_eff. This option is available in
the Background Fluid Flow Coupling. This variable includes the dynamic viscosity and
the turbulent viscosity ([Link]). Notice, however, that the value of the turbulent
viscosity can be much larger than the true eddy viscosity and that it depends on the
turbulence model selected. This means that the experienced attenuation can be too
large. Typically, more advanced turbulence models like the SST model give less
numerical diffusion and thus a better estimate of the eddy viscosity.

Eigenfrequency Studies
Both the linearized Euler and the linearized Navier–Stokes interfaces have the
Eigenfrequency study as a predefined study type. It should be noted that this analysis
type is challenging when a nonzero background mean flow is present. Mathematically
it can be very difficult to identify the pure acoustic eigenvalues.

In order to identify acoustic modes in the eigenvalue spectrum it can be a good idea
to use the Region option in the Eigenfrequency search method when setting up an
eigenfrequency study step. In this way the nonacoustic modes that consist of entropy
and vorticity waves (eigenfrequencies with large imaginary components) can be
filtered out. The pure acoustic modes typically have eigenfrequencies with a small
complex component can then more easily be identified. The finals identification needs
to be done with a visual inspection of the mode shape in postprocessing. Set the search

MODELING WITH THE AEROACOUSTICS BRANCH | 495

 region for the real part as desired and then set a small interval for the imaginary part,
for example, from −0.02 to 0.02.

For more details see the Eigenfrequency section in the COMSOL

Multiphysics Reference Manual

Suppressing Constraints on Lower Dimensions

Sometimes, boundary conditions on two adjacent objects can come into conflict on a
shared object (point or edge). For the all the boundary conditions in the
thermoviscous physics that are based on constraints, you have the possibility to select
that certain objects of lower dimensions should be excluded from the main selection.
To do this, you must first click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. In the settings a new sections
named Excluded Edges and/or Excluded Points will then appear (when applicable). In
these sections, you can select geometrical objects which should be excluded from the
main selection when the constraint is applied.

Kutta Like Conditions for Linearized Euler

A Kutta condition can be set up to handle edge singularities that appear, for example,
at the trailing edge of an airfoil or at the transition from a liner (modeled with an
Ingard-Myers type impedance) to a hard wall. By using Exclude Edges and Exclude
Points the appropriate Kutta condition can be set up. The problematic edges or points
are excluded from the adjacent conditions and extra conditions can be set up here
using the user defined Pointwise Constraint.

Velocity Condition for Linearized Navier–Stokes

When prescribing a velocity condition to you may get conflicts if it is located next to a
no-slip condition (a wall). To remedy this, you can exclude the common edge or point.

Periodic Condition
When using the periodic condition, you may get conflicts or duplicates which makes
the model either overconstrained, or problematic for the automatic constraint
elimination algorithm. If you are aware of such situations, it is good practice to remove
one of the potentially conflicting constraints. This could be in multiphysics models
where you combine solid mechanics with thermoviscous acoustics and have edges
where you both apply a periodic condition (solid and thermoviscous acoustics) and the
Thermoviscous Acoustic- Structure Boundary multiphysics coupling.

496 | CHAPTER 5: AEROACOUSTICS INTERFACES

Particularly if the geometry is curved, there is a risk that these constraints are not
identical from a numerical point of view. In this case, excluding the conflicting edge
from the selected boundary will make the behavior unique and fully predictable.

Another example where constraints will come in conflict is if you want to, for example,
use periodic condition on parts of the geometry using weak constraints, while keeping
the default pointwise constraints on other parts. If the same mesh node has both types
of constraints, the solution will fail, so you must exclude any common geometrical
objects from the selection in one of the constraints.

See also Excluded Surfaces, Excluded Edges, and Excluded Points in the
COMSOL Multiphysics Reference Manual.

Postprocessing Variables

Intensity
The flow of energy is expressed by the acoustic intensity I, which is formally defined
by the time-averaged sound power per unit area (unit: W/m2). This quantity is
available as a built-in postprocessing variable.

For a general fluid, including thermal and viscous losses (see Ref. 15 and Ref. 16 for
details), the time-averaged intensity is given by

T
1

I = ---- pudt =
T
0

ρ-  ∗
T
∗ T 1
--- Re ( ρ 0 u + ρu 0 )  -----
- + u 0 ⋅ u – ( ρ 0 u + ρu 0 ) ⋅  -----
1 p
- τ – ----- τ 
2  ρ0   ρ0 2 0
ρ 
0

∇T T ∗
+ ρ 0 T ( su 0 )∗ + Tk – -------- + ------2 ∇T 0
T0 T
0

where the viscous stress tensors for the acoustic and background fields, τ and τ0
respectively, are given by (in index notation)

MODELING WITH THE AEROACOUSTICS BRANCH | 497

  ∂u j ∂u i  2  ∂u k
τ ij = μ  +  –  μ B – --- μ  δ
 ∂ xi ∂ xj 3  ∂ x k ij

 ∂u 0 j ∂u 0 i 2  ∂u 0 k
τ 0 ij = μ  +  – --- μ  δ
 ∂ xi ∂ x j  3  ∂ x k  ij

It is understood that all dependent variables in these equations are the complex
amplitudes.

In the time domain, the equivalent quantity is the instantaneous intensity i given by

T 1 ρ  T
i = pv = ( ρ 0 u + ρu 0 )  ------ + u 0 ⋅ u – ( ρ 0 u + ρu 0 ) ⋅  ------ τ – -----2- τ 0
p
ρ0  ρ0 ρ 0

∇T T
+ ρ 0 Tsu 0 + Tk – -------- + ------2 ∇T 0
T0 T
0

where the dependent variables now include the explicit time dependence. The
instantaneous expression is not defined in the frequency domain since it would
represent effects happening at the double frequency.

See also the theory section The Energy Corollary for the Linearized Euler
interface.

The intensity variables are defined for most Aeroacoustics physics where it makes sense.
That is, the instantaneous quantities are defined in the time domain and the (time
averaged) intensity variables are defined in the frequency domain. It is understood that
the intensity vector is in general the time averaged quantity. For the Linearized
Potential Flow interface, the variables are not defined in the time domain since it
would require solving for an extra variable.
TABLE 5-1: INTENSITY VARIABLES IN 3D

VARIABLE DESCRIPTION

phys_id.I_mag Magnitude of the intensity vector (frequency domain only)

phys_id.Ix x-component of the intensity vector (frequency domain only)
phys_id.Iy y-component of the intensity vector (frequency domain only)
phys_id.Iz z-component of the intensity vector (frequency domain only)

498 | CHAPTER 5: AEROACOUSTICS INTERFACES

TABLE 5-1: INTENSITY VARIABLES IN 3D

VARIABLE DESCRIPTION

phys_id.Ii_mag Magnitude of the instantaneous intensity vector (time domain

only)
phys_id.Iix x-component of the instantaneous intensity vector (time
domain only)
phys_id.Iiy y-component of the instantaneous intensity vector (time domain
only)
phys_id.Iiz z-component of the instantaneous intensity vector (time domain
only)
TABLE 5-2: INTENSITY VARIABLES IN 2D AXISYMMETRIC

VARIABLE DESCRIPTION

phys_id.I_mag Magnitude of the intensity vector (frequency domain only)

phys_id.Ir r-component of the intensity vector (frequency domain only)
phys_id.Iz z-component of the intensity vector (frequency domain only)
phys_id.Ii_mag Magnitude of the instantaneous intensity vector (time domain
only)
phys_id.Iir r-component of the instantaneous intensity vector (time domain
only)
phys_id.Iiz z-component of the instantaneous intensity vector (time domain
only)
TABLE 5-3: INTENSITY VARIABLES IN 2D

VARIABLE DESCRIPTION

phys_id.I_mag Magnitude of the intensity vector (frequency domain only)

phys_id.Ix x-component of the intensity vector (frequency domain only)
phys_id.Iy y-component of the intensity vector (frequency domain only)
phys_id.Ii_mag Magnitude of the instantaneous intensity vector (time domain
only)
phys_id.Iix x-component of the instantaneous intensity vector (time
domain only)
phys_id.Iiy y-component of the instantaneous intensity vector (time domain
only)

MODELING WITH THE AEROACOUSTICS BRANCH | 499

 T he o r y B a c k g rou n d for t h e
Aeroacoustics Branch
The scientific field of aeroacoustics deals with the interaction between a background
mean flow and an acoustic field propagating in this flow. In general, this concerns both
the very complex description of the creation of sound by turbulence in the background
flow, that is, flow induced noise and the influence the background mean flow has on
the propagation of an externally created sound field, that is, flow borne noise or
convected acoustics. The computational aeroacoustics (CAA) capabilities of the
aeroacoustics interfaces in COMSOL Multiphysics only cover the convected acoustics
situation.

Aeroacoustic simulations would ideally involve solving the fully compressible

continuity, momentum (Navier–Stokes equations), and energy equations in the time
domain. The acoustic pressure waves would then form a subset of the fluid solution.
This approach is often impractical for real-world CAA applications due to the required
computational time and memory resources. Instead, for solving many practical
engineering problems, a decoupled two-step approach is used: first solve for the fluid
flow, then the acoustic perturbations of the flow.

For solving the acoustic problem, the governing equations are linearized around the
background mean flow and only solved for the acoustic perturbation. Acoustic
variables are assumed to be small and perturbation theory can be used, for example,
the total pressure

p tot = p 0 + p

is the sum of the background mean pressure p0 and the acoustic pressure variations p
(sometimes labeled p’ or p1).

This section presents the basic mathematical framework for the aeroacoustic equations
solved in the aeroacoustic interfaces, starting with the general governing equations for
fluid flow, that is, conservation equations, constitutive equations, and equations of
state. Then the linearized Navier–Stokes equations, the linearized Euler equations, the
linearized potential flow equations, and finally the equations for the compressional
potential flow are presented.

500 | CHAPTER 5: AEROACOUSTICS INTERFACES

In this section the theory background for:

• General Governing Equations

• Linearized Navier–Stokes
• Linearized Euler
• Scattered Field Formulation for LE and LNS
• Linearized Potential Flow
• Compressible Potential Flow

General Governing Equations

The equations governing the physics in any fluid are the foundation for deriving the
linearized aeroacoustic equations and in general any acoustics equations. The
governing equations for the motion of a compressible fluid are the continuity equation
(mass conservation), the Navier–Stokes equation (momentum conservation), and the
general heat transfer equation (energy conservation). In order to close the system of
equations, constitutive equations are needed, along with the equation of state, and
thermodynamic relations. See, for example, Ref. 3, Ref. 4, Ref. 5, Ref. 6, or Ref. 7 for
details and further reading.

CONSERVATION EQUATIONS
The conservation equations are for mass, momentum, and energy:

∂ρ
------ + ∇ ⋅ ( ρu ) = M
∂t
∂u
ρ -------- = ρ  ------ + ( u ⋅ ∇ )u = ∇ ⋅ σ + F
Du
Dt ∂t (5-5)

∂s
ρT ------- = ρT  ----- + ( u ⋅ ∇ )s = – ∇ ⋅ q + Φ + Q
Ds
Dt  ∂t 

where u is the velocity field, ρ is the density, T is the temperature, s is the specific
entropy, σ is the stress tensor, q is the local heat flux, Φ is the viscous dissipation
function, and M, F, and Q are source terms. The operator D/Dt is the material
derivative (or advection operator) defined as

D- ∂
------ = ----- + ( u ⋅ ∇ )
Dt ∂t

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 501

 THERMODYNAMIC RELATIONS
Some thermodynamic relations are necessary when reformulating the energy equations
in terms of other sets of thermodynamic variables, like (p, T) or (ρ, p). They are the
density differential, the specific energy relation, a relation due to Helmholtz, and the
fundamental entropy relation:

dρ
------- = – α p dT + β T dp
ρ
∂u ∂u
du =  ------- +  ------
 ∂T ρ  ∂ρ T

 ∂u α
------ 1- 
----- -----p- 
 ∂ρ T = 2  p – β T T
ρ
1 p –1
ds = ---- du + ---- d ( ρ )
T T

where u is the specific internal energy, αp is the coefficient of thermal expansion

(isobaric), and βT is the isothermal compressibility. See, for example, Ref. 5 and Ref. 6
for details. They are defined together with the specific heat at constant pressure Cp and
specific heat at constant volume Cv as

1 ∂ρ 1 ∂ρ
α p = – ---  ------- β T = ---  ------
ρ  ∂T p ρ  ∂p T
∂u ∂u
C v =  ------- C p =  -------
∂T ρ ∂T p

Using the above thermodynamic relations, the entropy differential can be expressed as
(used for the linearized Navier–Stokes equations)

ρTds = ρC p dT – α p Tdp

while for an ideal gas it can be given as (used for the linearized Euler equations)

dp dρ
ds = C v ------- – C p -------
ρ p

CONSTITUTIVE EQUATIONS
The constitutive equations are the equations of state (density expressed in terms of any
set of thermodynamic variables), the Stokes expression for the stress tensor, and the
Fourier heat conduction law

502 | CHAPTER 5: AEROACOUSTICS INTERFACES

 σ = – pI + τ = – pI + μ ( ∇u + ( ∇u ) ) +  μ B – --- μ ( ∇ ⋅ u )I
T 2
 3 
q = – k∇T

where k is the thermal conduction, μ is the dynamic viscosity, and μB is the bulk
viscosity. This then also defines the viscous dissipation function

Φ = ∇u:τ

PERTURBATION THEORY
In the following, the governing equations are linearized and expanded to first order in
the small parameters around the average stationary background solution. For details
about perturbation theory see Ref. 3, Ref. 4, and Ref. 8. The small parameter variables
(1’st order) represent the acoustic variations on top of the stationary background mean
(or average) flow (0’th order solution). Note that, when solving the equations, the
value of the acoustic field variables can also represent nonacoustic waves like thermal
waves (entropy waves) and vorticity waves. In the time domain, these can be linear
instabilities and can actually represent the onset of turbulence.

The dependent variables and sources are expanded according to

A = A 0 ( x ) + A 1 ( x, t )

where A is any of the dependent variables or sources. In the frequency domain, the first
order variables are assumed to be harmonic and expanded into Fourier components,
such that

iωt
A = A 0 ( x ) + A 1 ( x )e

The first order variation to material parameters, that are not treated as dependent
variables like the density ρ in the Linearized Navier–Stokes interface, is expressed using
the density differential and related to perturbations in the temperature and the
pressure.

The above perturbation schemes are inserted into the governing equations and the
linearized acoustic equations are derived retaining only first order linear terms.

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 503

 Linearized Navier–Stokes
The linearized Navier–Stokes equations are derived by linearizing the full set of fluid
flow equations given in General Governing Equations. After some manipulation, the
continuity, momentum, and energy equations and are:

∂ρ
------ + ∇ ⋅ ( ρ 0 u + ρu 0 ) = M
∂t
∂u
ρ 0  ------ + ( u ⋅ ∇ )u 0 + ( u 0 ⋅ ∇ )u + ρ ( u 0 ⋅ ∇ )u 0 = ∇ ⋅ σ + F – u 0 M
∂t
(5-6)
∂T
ρ 0 C p  ------- + u ⋅ ∇ T 0 + u 0 ⋅ ∇ T + ρC p ( u 0 ⋅ ∇ T 0 )
 ∂t 
∂p
– α p T 0  ------ + u ⋅ ∇ p 0 + u 0 ⋅ ∇ p – α p T ( u 0 ⋅ ∇ p 0 ) = ∇ ⋅ ( k∇T ) + Φ + Q
 ∂t 

where p, u, and T are the acoustic perturbations to the pressure, velocity, and
temperature, respectively. In the frequency domain, the time derivatives of the
dependent variables are replaced by multiplication with iω. The stress tensor is σ and
Φ is the viscous dissipation function. The variables with a zero subscript are the
background mean flow values and the subscript “1” is dropped on the acoustic
variables.

The constitutive equations are the stress tensor and the linearized equation of state,
while the Fourier heat conduction law is readily included in the above energy equation,

σ = – pI + μ ( ∇u + ( ∇u ) ) +  μ B – --- μ ( ∇ ⋅ u )I
T 2
 3 
ρ = ρ0 ( βT p – αp T )

The linearized viscous dissipation function is defined as

Φ = ∇u:τ ( u 0 ) + ∇u 0 :τ ( u )

τ ( u ) = μ ( ∇u + ( ∇u ) ) +  μ B – --- μ ( ∇ ⋅ u )I
T 2
 3 

τ ( u 0 ) = μ ( ∇u 0 + ( ∇u 0 ) ) +  μ B – --- μ ( ∇ ⋅ u 0 )I
T 2
3

The terms in the governing equations presented in Equation 5-6 can be divided into
four categories. The time derivative (or frequency dependent) term, convective terms
like u0 ⋅ ∇p, reactive terms like u ⋅ ∇p0, diffusive terms, and source terms. In many
aeroacoustic formulations the reactive terms are removed (or simplified) from the

504 | CHAPTER 5: AEROACOUSTICS INTERFACES

governing equations in order to avoid the Kelvin-Helmholtz instabilities. This is
sometimes referred to as gradient term stabilization (GTS).

Linearized Euler
The linearized Euler equations are derived from Euler’s equations, that is
Equation 5-5 with no thermal conduction and no viscous losses. The fluid in the
linearized Euler physics interface is assumed to be an ideal gas. The energy equation is
also often written in terms of the pressure. This manipulation is possible using the
thermodynamic differential for the entropy valid for an ideal gas. This is the common
approach in literature. A review of the linearized Euler equations is found in, for
example, Ref. 12 and Ref. 13.

GOVERNING EQUATIONS
A linearization of the governing equations yields after some manipulation

∂ρ
--------t + ∇ ⋅ ( ρ t u 0 + ρ 0 u t ) = M
∂t
∂u t  ρ
--------- + u t + -----t- u 0 ⋅ ∇ u 0 + ( u 0 ⋅ ∇ )u t + ------ ∇p t = ------ ( F – u 0 M )
1 1
(5-7)
∂t  ρ0  ρ0 ρ0
∂p γp
--------t + u t ⋅ ∇p 0 + p t γ ( ∇ ⋅ u 0 ) + u 0 ⋅ ∇p t + γp 0 ( ∇ ⋅ u t ) = ( γ – 1 )Q + --------0- M
∂t ρ0

Here, the subscript “1” has been dropped on the acoustic perturbation variables. The
time derivatives are replaced with multiplication by iω in the frequency domain. And
source terms include a mass source M, a volume force F, and a heat source Q.

Some ideal gas relations of interest are the equation of state, specific heat capacity and
specific gas constant relations, and the isothermal compressibility and isobaric
coefficients of thermal expansion:

2 γp 0
c 0 = --------- = R s γT 0
ρ0
Cp
γ = ------- Cp = Cv + Rs
Cv
1 1
α p = ------ β T = ------
T0 p0

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 505

 It also follows from the governing equations and the thermodynamic relations that the
acoustic variations in the specific entropy s and in the temperature T are given by

Cv Cp
s = ------ p – ------- ρ
p0 ρ0
ρ
T = T 0  ------ – ------
p
 p 0 ρ 0

The specific heat capacity at constant volume and constant pressure are
often labeled with a lowercase c, here we use uppercase C’s but they are
the specific quantities.

INSTABILITIES
When the linearized Euler (LE) equations are solved in the time domain (or in the
frequency domain with an iterative solver), linear physical instability waves can
develop, the so-called Kelvin-Helmholtz instabilities. They are instabilities that grow
exponentially because no losses exist in the LE equations (no viscous dissipation and
no heat conduction). Furthermore, they are limited by nonlinearities in the full
Navier–Stokes flow equations. It has been shown that in certain cases the growth of
these instabilities can be limited, while the acoustic solution is retained, by canceling
terms involving gradients of the mean flow quantities. This is known as gradient terms
suppression (GTS) stabilization. See more details in Ref. 9, Ref. 10, and Ref. 11.

Gradient terms suppression (GTS) stabilization is available for the linearized Euler
equations. The reactive terms involving gradients of the background flow can be
disabled individually. These terms are responsible for the generation of the instabilities.
It is also possible to disable all convective terms if desired.

THE ENERGY COROLLARY

Expressions for the energy flux, that is, the acoustic intensity vector, are often referred
to as Myers’ energy corollary, see Ref. 15 and Ref. 16. The instantaneous intensity
vector Ii is defined for both transient and frequency domain models as

I i = ( ρ 0 u + ρu 0 )  ------ + u 0 ⋅ u + ρ 0 u 0 Ts
p
 ρ0 

The (time averaged) intensity vector I is given in the frequency domain by

506 | CHAPTER 5: AEROACOUSTICS INTERFACES

 ∗
I = --- Re  ( ρ 0 u + ρu 0 )  ------ + u 0 ⋅ u + ρ 0 u 0 Ts∗
1 p
2 ρ0

ABOUT THE IMPEDANCE BOUNDARY CONDITION

The Linearized Euler, Frequency Domain Interface and The Linearized Euler,
Transient Interface have an Impedance and Interior Impedance physics feature and its
theory is included here.

In the frequency domain the Ingard-Myers equation (Ref. 1) gives an expression for
the normal velocity at a boundary with a normal impedance condition. It is a so-called
low-frequency approximation condition in the limit of very thin flow boundary layers
(compared to the wavelength). Such conditions are used, for example, for porous
lining conditions in ducts (Ref. 2). The condition is given by:

u ⋅ n = ------ + ------ u 0 ⋅ ∇  ------ – ------------- n ⋅ ( n ⋅ ∇u 0 )

p 1 p p
(5-8)
Z n iω Zn iωZ n

where the surface normal n here points out of the domain and is the surface normal
impedance.

If the flow is parallel to the impedance boundary condition u0·n = 0, for example, slip
flow over a mechanical impedance boundary condition (the same is true for the
moving wall boundary condition described below), one can use a formulation with the
tangential derivative ( ∇ || ) for the second term on the right-hand side:

u 0 ⋅ ∇A = u 0 ⋅ ∇ || A + u 0 ⋅ n ( n ⋅ ∇A ) = u 0 ⋅ ∇ || A

where A is an arbitrary scalar.

The last term on the right-hand side of Equation 5-8 can be reformulated as follows:

n ⋅ ( n ⋅ ∇u 0 ) = ( n ⋅ ∇ ) ( n ⋅ u 0 ) – u 0 ⋅ ( ( n ⋅ ∇ )n )

Again, these terms reduce significantly for the case where n⋅u0 = 0. If the boundary
does not have curvature (planar boundary) then it is equal to zero. If the boundary is
planar and the impedance condition is used inside the flow, for example at an outflow
condition, then it reduces to the normal gradient of the velocity normal to the surface.

ABOUT THE MOVING WALL BOUNDARY CONDITION

The Linearized Euler, Frequency Domain Interface and The Linearized Euler,
Transient Interface have a Moving Wall physics feature and its theory is included here.

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 507

 Myers’ equation (Ref. 1) gives the expressions used for a boundary condition at a
moving wall. In the frequency domain it is given by

u ⋅ n = – iωv n – u 0 ⋅ ∇ ( v n ) + v n n ⋅ ( n ⋅ ∇u 0 )

The inward normal displacement is given by

–1
v n = – n ⋅ v n = ( iω ) u n un = –n ⋅ ub

where un is the inward normal velocity is.

In the time domain, the condition on the normal velocity is given by

∂v n
u ⋅ n = – --------- – u 0 ⋅ ∇ ( v n ) + v n n ⋅ ( n ⋅ ∇u 0 )
∂t

where the inward normal displacement is given by vn = −n⋅v (SI unit: m).

Scattered Field Formulation for LE and LNS

The governing equations, given above in Equation 5-7 and Equation 5-6, are defined
in the general scattered field formulation. Assuming that all the acoustic fields are the
sum of a background (prescribed) field and the resulting scattered field, the total fields
can be written as

pt = ps + pb ut = us + ub Tt = Ts + Tb ρt = ρs + ρb
p ≡ ps u ≡ us T ≡ Ts ρ ≡ ρs

where subscript “t” stands for total, subscript “s” for scattered, and subscript “b” for
background. The scattered field is the field solved for, that is the dependent variables
p, u, T, and ρ (depending on the interface). If no background acoustic field is defined
(the default) the scattered field is equal to the total field. The governing equations and
all boundary conditions are expressed in terms of the total fields. The background
acoustic fields are defined by adding the Background Acoustic Fields feature in the LE
interface or the Background Acoustic Fields feature in the LNS interface. The features
have a user defined option where all fields can be entered, either as analytical
expressions or defined in terms of a solution to another study or model.

508 | CHAPTER 5: AEROACOUSTICS INTERFACES

Linearized Potential Flow
The equations presented here are the linearized potential flow equations. This restricts
the applications of the physics interface to systems where the background flow is well
described by a compressible potential flow, that is, a flow that is inviscid, barotropic,
and irrotational. The sound sources also need to be external to the flow or at least they
need to be represented by simple well defined sources. Application areas typically
include modeling of how jet engine noise is influenced by the mean flow.

The basic dependent variable is the velocity potential φ conventionally defined by the
relationship

u = ∇φ

where u = u(r, t) is the particle velocity associated with the acoustic wave motion. The
total particle velocity is given by

u tot ( r, t ) = V ( x ) + u ( r, t ) (5-9)

where V denotes the local mean velocity for the fluid motion (the mean flow is labeled
u0 in the linearized Euler and Navier–Stokes interfaces). The dynamic equations for
this mean-flow field are described in the next subsection. For now, just assume V to be
a given irrotational background velocity field; hence, also the mean-flow velocity can
be defined in terms of a potential field Φ, by V = ∇Φ.

The linearized equation for the velocity potential φ , governing acoustic waves in a
background flow with mean background velocity V, mean background density ρ0, and
mean background speed of sound c0, is

ρ 0 ∂ ∂φ ρ 0 ∂φ
– -----2-  + V ⋅ ∇φ + ∇ ⋅ ρ 0 ∇φ – -----2-  + V ⋅ ∇φ V = 0 (5-10)
∂ ∂  
c0 t t c0 ∂ t

In deriving this equation, all variables appearing in the full nonlinear fluid-dynamics
equations were first split in time-independent and acoustic parts, in the manner of
Equation 5-9. Then, linearizing the resulting equations in the acoustic perturbation
and eliminating all acoustic variables except the velocity potential gives Equation 5-10.
Thus, the density ρ in this equation is the time-independent part. The corresponding
acoustic part is ρ(r, t) = p(r, t)/c02 where p is the acoustic pressure, given by

∂φ
p ( r, t ) = – ρ 0  + V ⋅ ∇φ
∂t 

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 509

 Hence, once Equation 5-10 has been solved for the velocity potential, the acoustic
pressure can easily be calculated.

When transformed to the frequency domain, the wave Equation 5-10 reads

ρ0 ρ0
– -----2- iω ( iωφ + V ⋅ ∇φ ) + ∇ ⋅ ρ 0 ∇φ – -----2- ( iωφ + V ⋅ ∇φ )V = 0
c0 c0

while the acoustic pressure is

p ( r ) = – ρ 0 ( iωφ + V ⋅ ∇φ )

Typical boundary conditions include:

• Sound-hard boundaries or walls

• Sound-soft boundaries
• Impedance boundary conditions
• Radiation boundary conditions

FREQUENCY DOMAIN EQUATIONS

In the frequency domain the velocity potential φ is assumed to be a harmonic wave of
the form

iωt
φ ( r, t ) = φ ( r )e

The governing frequency domain — or time-harmonic — equation is

ρ0 ρ0
– -----2- iω ( iωφ + V ⋅ ∇φ ) + ∇ ⋅ ρ 0 ∇φ – -----2- ( iωφ + V ⋅ ∇φ )V = 0
c0 c0

In 2D, where

i ( ωt – kz z )
φ ( r, t ) = φ ( x, y )e

the out-of-plane wave number kz enters the equations when the ∇ operators are
expanded:

510 | CHAPTER 5: AEROACOUSTICS INTERFACES

 ρ0  ρ0 
– i ω -----2- ( iωφ + V ⋅ ∇φ – ik z V z φ ) + ∇ ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ – ik zV z φ )
c0  c0 

2 ρ0
+ ρ 0 k z φ + ik z V z -----2- ( iωφ + V ⋅ ∇φ – ik z V z φ ) = 0
c0

The default value of the out-of-plane wave number is 0, that is, no wave propagation
perpendicular to the 2D plane. In a mode analysis the equations are solved for kz.

For 2D axisymmetric components

i ( ωt – mϕ )
φ ( r, t ) = φ ( r, z ) e

the azimuthal wave number m similarly appears in the equation as a parameter:

ρ0  ρ0  m2
– i ω -----2- ( iωφ + V ⋅ ∇φ ) + ∇ ⋅  ρ 0 ∇φ – V -----2- ( iωφ + V ⋅ ∇φ ) + ρ 0 ----2- φ = 0
c0  c0  r

The background velocity field V cannot have an azimuthal component

because the flow is irrotational.

TIME-DEPENDENT EQUATION
In the time domain, the physics interface solves for the velocity potential φ with an
arbitrary transient dependency. The following equation governs the acoustic waves in
a mean potential flow:

ρ 0 ∂ ∂φ ρ 0 ∂φ
– -----2-  + V ⋅ ∇φ + ∇ ⋅ ρ 0 ∇φ – -----2-  + V ⋅ ∇φ V = 0 (5-11)
c0 ∂ t ∂ t c0 ∂ t

Here ρ0 (SI unit: kg/m3) is the background mean flow density, V (SI unit: m/s)
denotes the background mean velocity, and c0 (SI unit: m/s) refers to the speed of
sound. The software solves the equation for the velocity potential φ , with SI unit
m2/s. The validity of this equation relies on the assumption that ρ0, V, and c0 are
approximately constant in time, while they can be functions of the spatial coordinates.

The background velocity field V cannot have an azimuthal component

because the flow is irrotational.

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 511

 BOUNDARY MODE ANALYSIS
The boundary mode analysis type in 3D uses the eigenvalue solver to solve the
equation

ρ0  ρ0 
– i ω -----2- ( iωφ + ( V t ⋅ ∇φ ) – ik zV n φ ) + ∇ ⋅  ρ 0 ∇φ – V t -----2- ( iωφ + ( V t ⋅ ∇φ ) – ik zV n φ )
c0  c0 
(5-12)
2 ρ0
+ ρ 0 k z φ + ik zV n -----2- ( iωφ + ( V ⋅ ∇φ ) – ik zV n φ ) = 0
c0

for the eigenmodes, φ , and eigenvalues, λ = −ikz, on a bounded two-dimensional

domain, Ω, given well-posed edge conditions on ∂Ω. In this equation, φ is the velocity
potential, ρ0 is the background mean flow density, c0 is the speed of sound, ω is the
angular frequency, and kz is the out-of-plane wave number or propagation constant.
Furthermore, Vt denotes the background mean velocity in the tangential plane while
Vn is the background mean velocity component in the normal direction.

Although the out-of-plane wave number is called kz, the two-dimensional

surface on which Equation 5-12 is defined does not necessarily have to be
normal to the z-axis for 3D geometries.

Compressible Potential Flow

Consider a compressible and inviscid fluid in some domain Ω. The motion and state of
the fluid is described by its velocity V, density ρ, pressure p, and total energy per unit
volume e. Its dynamics is governed by the Euler equations, expressing the conservation
of mass, momentum, and energy (see General Governing Equations):

∂ρ
+ ∇ ⋅ ( ρV ) = 0
∂t
∂V
ρ + ( V ⋅ ∇ )V + ∇p = F (5-13)
∂t 
∂e
----- + ∇ ⋅ ( ( e + p )V ) = 0
∂t

Here a volume force F has been included on the right-hand side of the momentum
equation, whereas a possible heat-source term on the right-hand side of the energy
equation (the last one) has been set to zero.

512 | CHAPTER 5: AEROACOUSTICS INTERFACES

To close this system of five equations with six unknowns, an equation of state is
required. Here, this is taken to be the equation for an ideal barotropic fluid,

ρ γ
p = p ref  --------
 ρ ref

where γ = Cp /CV is the ratio of specific heats, Cp is the specific heat at constant pressure
and CV at constant volume, while pref and ρref are reference quantities for the pressure
and the density, respectively. An alternative form of the ideal-fluid state equation is

p = ρ ( γ – 1 )e

The assumption that the fluid is barotropic means that p = p(ρ). Taking the total time
derivative and using the chain rule, leads to the relation

dp dp dρ 2 dρ
= ≡c
dt dρ dt dt

where, using the equation of state,

p
c = γ ---
ρ

defines the speed of sound in the ideal fluid.

Assuming the flow to be irrotational, there exists a velocity potential field Φ, such that
V = ∇Φ. If, in addition, the volume force is assumed to be given by F = −ρ∇Ψ, where
Ψ is referred to as the force potential, the second of Equation 5-13 can be integrated
to yield the Bernoulli equation

∂Φ γ p ref  ------- γ γ p ref

------- +  1 -------- ρ - + Ψ = --- v ref + ---------------- --------
1 2
--- ∇Φ 2 + ----------- + Ψ ref
∂t  2 γ – 1 ρ  ρ ref  2 ( γ – 1 ) ρ ref

In this equation, two additional reference quantities have entered: the velocity vref,
and the force potential Ψref, both valid at the same reference point as pref and ρref.
Note, in particular, that neither the pressure p, nor the energy per unit volume e,
appears in this equation.

TIME DEPENDENT STUDY

Collecting the results, the equations governing the compressible, inviscid, irrotational
flow of an ideal fluid are

THEORY BACKGROUND FOR THE AEROACOUSTICS BRANCH | 513

 ∂Φ γ p ref  ------- ρ γ γ p ref
------- +  --- ∇Φ + ----------- -------- - + Ψ = 1
1 2 2
--- v ref + ---------------- -------- + Ψ ref
∂t  2 γ – 1 ρ ρ ref    2 ( γ – 1 ) ρ ref

∂ρ p
+ ∇ ⋅ ( ρ ∇Φ ) = 0 c = γ --- γ ≡ Cp ⁄ CV
∂t ρ

where γ is the ratio of specific heats and Ψ denotes a possible domain force potential
(SI unit: J/kg). In this equation, subscript ref signifies reference quantities that apply
at a specific point or surface. Thus, pref is a reference pressure, ρref is a reference
density, vref is a reference velocity, and Ψref is a reference force potential.

STATIONARY STUDY
In a stationary study, the same equation is used, but all time derivatives are set to zero,
such that

1 p ref  ρ  γ
γ -------- γ p ref
--- ∇Φ 2 + ----------- -------- + Ψ = 1
2
--- v ref + ---------------- -------- + Ψ ref
2 γ – 1 ρ  ρ ref  2 ( γ – 1 ) ρ ref
∇ ⋅ ( ρ∇Φ ) = 0

514 | CHAPTER 5: AEROACOUSTICS INTERFACES

References for the Aeroacoustics
B r a nc h I nte r f a c es
1. M. K. Myers, “On the basic boundary condition in the presence of flow”, J. of
Sound and Vibration, vol. 71, p. 429, 1980.

2. W. Eversman, “The boundary condition at an impedance wall in a non-uniform duct

with potential mean flow”, J. of Sound and Vibration, vol. 246, p. 63, 2001.

3. D.T. Blackstock, Fundamentals of Physical Acoustics, John Wiley & Sons, 2000.

4. H. Bruus, Theoretical Microfluidics, Oxford University Press, 2010.

5. G.K. Bachelor, An Introduction to Fluid Dynamics, Cambridge University Press,

2000.

6. L. D. Landau and E. M. Lifshitz, Fluid Mechanics, Course on Theoretical Physics

volume 6, Butterworth-Heinemann, 2003.

7. B. Lautrup, Physics of Continuous Matter, Exotic and Every Day Phenomena in

the Macroscopic World, 2nd ed., CRC Press, 2011.

8. C. M. Bender and S. A. Orszag, Advanced Mathematical Methods for Scientists

and Engineers: Asymptotic Methods and Perturbation Theory, Springer, 1999.

9. C. Bogey, C. Bailly, and D. Juvé, “Computation of Flow Using Source Terms in

Linearized Euler’s Equations”, AIAA Journal, vol. 40, p. 235, 2002.

10. P. P. Rao and P. J. Morris, “Use of Finite Element Methods in Frequency Domain
Aeroacoustics”, AIAA Journal, vol. 44, p. 1643, 2006.

11. A. Agarwal, P. J. Morris, and R. Mani, “Calculation of Sound Propagation in

Nonuniform Flows: Suppression of Instability Waves”, AIAA Journal, vol. 42, p.80,
2004.

12. C. Bailly and D. Juvé, “Numerical Solution of Acoustic Propagation Problems

Using Linearized Euler Equations”, AIAA Journal, vol. 38, p. 22, 2000.

13. C. W. Tam, “Computational Aeroacoustics: Issues and Methods”, AIAA Journal,

vol. 33, 1995.

REFERENCES FOR THE AEROACOUSTICS BRANCH INTERFACES | 515

 14. G. Hauke and T. J. R. Hughers, “A comparative study of different sets of variables
for solving compressible and incompressible flows”, Comput. Methods Appl. Mech.
Engrg., vol. 153, pp. 1–44, 1998.

15. M. K. Myers, “An Exact Energy Corollary for Homentropic Flows”, J. Sound.
Vib., vol. 109, pp. 277–284, 1986.

16. M. K. Myers, “Transport of energy disturbances in arbitrary steady flows”, J. Fluid

Mech., vol. 226, pp. 383–400, 1991.

17. C. K. W. Tam and Z. Dong, “Radiation and Outflow Boundary Conditions for
Direct Computation of Acoustic and Flow Disturbances in Nonuniform Mean Flow”,
J. of Comp. Acoustics 4, pp. 175–201, 1996.

18. L. Du, A. Holmberg, M. Karlsson, and M Åbom, “Sound amplification at a

rectangular T-junction with merging mean flows”, J. Sound Vib., vol. 367, pp. 69–83
2016.

19. J. Gikadi, S. Föller, T. Sattelmayer, “Impact of turbulence on the prediction of

linear aerodynamic interactions: Acoustic response of a turbulent shear layer”, J.
Sound Vib., vol. 333, pp. 6548–6559, 2014.

20. S. Redonnet and G. Cunha, “An advanced hybrid method for the acoustic
prediction”, Adv. Eng. Softw., vol. 88, pp. 30–52, 2015.

21. J. Gikadi, “Prediction of Acoustic Modes in Cumbustors using Linearized Navier–

Stokes Equations in the Frequency Domain”, Ph.D. thesis, TUM, 2013.

22. O. C. Zienkiewicz, R. L. Taylor, and P. Nithiarasu, “The Finite Element Method

for Fluid Dynamics”, Butterworth-Heinemann, 7th edition, 2014.

23. G, Hauke, “Simple stabilizing matrices for the computation of compressible flows
in primitive variables,” Comp. Methods. Appl. Mech. Engrg., vol. 190, pp. 6881–
6893, 2001.

516 | CHAPTER 5: AEROACOUSTICS INTERFACES

 6

Thermoviscous Acoustics Interfaces

This chapter describes the physics interfaces found under the Thermoviscous
Acoustics branch ( ).

• The Thermoviscous Acoustics, Frequency Domain Interface

• The Thermoviscous Acoustics, Transient Interface
• The Thermoviscous Acoustics, Boundary Mode Interface
• The Acoustic-Thermoviscous Acoustic Interaction, Frequency Domain
Interface
• The Thermoviscous Acoustic-Solid Interaction, Frequency Domain Interface
• The Thermoviscous Acoustic-Shell Interaction, Frequency Domain Interface
• Modeling with the Thermoviscous Acoustics Branch
• Theory Background for the Thermoviscous Acoustics Branch

517
 The Thermoviscous Acoustics,
Frequency Domain Interface
The Thermoviscous Acoustics, Frequency Domain (ta) interface ( ), found under the
Thermoviscous Acoustics branch ( ) when adding a physics interface, is used to
compute the acoustic variations of pressure, velocity, and temperature. This physics
interface is required to accurately model acoustics in geometries of small dimensions,
often referred to as microacoustics. Near walls, viscous losses and thermal conduction
become important because a boundary layers exists. The thicknesses of these boundary
layers are also known as the viscous and thermal penetration depth. For this reason, it
is necessary to include thermal conduction effects and viscous losses explicitly in the
governing equations. It is, for example, used when modeling the response of
transducers like microphones, miniature loudspeakers, and receivers. Other
applications include analyzing feedback in hearing aids and in mobile devices, or
studying the damped vibrations of MEMS structures.

The physics interface solves the equations in the frequency domain assuming all fields
and sources to be harmonic. The harmonic variation of all fields and sources is given
by eiωt using the +iω convention. Linear acoustics is assumed. Nonlinear effects can
be included when modeling in the time domain using The Thermoviscous Acoustics,
Transient Interface and the Nonlinear Thermoviscous Acoustics Contributions
feature.

The equations defined by the Thermoviscous Acoustics, Frequency Domain interface

are the linearized Navier–Stokes equations in quiescent background conditions solving
the continuity, momentum, and energy equations (first order perturbation equations).
Thermoviscous acoustics is also known as viscothermal acoustics or sometimes
thermoacoustics (not to be confused with the field discussing heating and cooling
using acoustics). Due to the detailed description necessary when modeling
thermoviscous acoustics, the model simultaneously solves for the acoustic pressure p,
the acoustic velocity variation u (particle velocity), and the acoustic temperature
variations T. It is available for 3D, 2D, and 1D Cartesian geometries as well as for 2D
and 1D axisymmetric geometries.

The Thermoviscous Acoustics, Frequency Domain interface is formulated in the

so-called scattered field formulation where the total field (subscript t) is the sum of the

518 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

scattered field (the field solved for, p, u, and T) and a possible background acoustic
field (subscript b), such that

pt = p + pb ut = u + ub Tt = T + Tb

When no Background Acoustic Fields feature is present (the background field values
are zero per default) the total field is simply the field solved for, pt = p, ut = u, and
Tt = T. All governing equations and boundary conditions are formulated in the total
field variables.

When this physics interface is added, these default nodes are also added to the Model
Builder — Thermoviscous Acoustics Model, Wall, and Initial Values. Then, from the
Physics toolbar, add other nodes that implement, for example, boundary conditions
and sources. You can also right-click Thermoviscous Acoustics to select physics features
from the context menu.

For good modeling strategies, solver suggestions, postprocessing

information, as well as tips and tricks, see the Modeling with the
Thermoviscous Acoustics Branch section.

On the Thermoviscous Acoustics Physics Interface

The Thermoviscous Acoustics, Frequency Domain interface solves, as mentioned, the
full linearized Navier–Stokes (momentum), continuity, and energy equations. It solves
for the propagation of compressible linear waves in a general viscous and thermally
conductive fluid. The length scale at which the thermoviscous acoustic description is
necessary is given by the thickness of the viscous boundary layer (the viscous
penetration depth), which is

μ-
δv = -----------
πfρ 0

and the thickness of the thermal boundary layer (the thermal penetration depth)

k -
δ th = ------------------
πfρ 0 C p

where the definition of the symbols f, μ, ρ0, k, and Cp may be found in Table 6-1. The
thickness of both boundary layers depends on the frequency f and decreases with
increasing frequency. The ratio of the two length scales is related to the
nondimensional Prandtl number Pr, by

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 519

 δ μC p
------v- = ----------- = Pr
δ th k

which define the relative importance of the thermal and viscous effects for a given
material. In air at 20 oC and 1 atm, the viscous boundary layer thickness is 0.22 mm
at 100 Hz while it is only 55 μm in water under the same conditions. The Prandtl
number is 0.7 in air and 7 in water.

Evaluate the value of the viscous and thermal boundary layer thickness as
well as the Prandtl number in postprocessing. They are defined by the
variables ta.d_visc, ta.d_therm, and [Link], respectively.

The physical quantities commonly used in the thermoviscous acoustics interfaces are
defined in Table 6-1 below.
TABLE 6-1: THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN PHYSICAL QUANTITIES.

QUANTITY SYMBOL SI UNIT ABBREVIATION

Pressure (acoustic) p pascal Pa

Total acoustic pressure pt pascal Pa
Scattered acoustic pressure ps pascal Pa
Temperature variation T kelvin K
(acoustic)
Total temperature variation Tt kelvin K
Scattered temperature Ts kelvin K
variation
Acoustic velocity field u = (u, v, w) meter/second m/s
Total acoustic velocity field ut meter/second m/s
Scattered acoustic velocity us meter/second m/s
field
Dynamic viscosity μ pascal-second Pa·s
Bulk viscosity μB pascal-second Pa·s
Thermal conductivity k watt/meter-kelvin W/(m·K)
3
Heat capacity at constant Cp joule/meter -kelvin J/(m3·K)
pressure
Isothermal compressibility βT 1/pascal 1/Pa
Coefficient of thermal αp 1/kelvin 1/K
expansion (isobaric)

520 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

TABLE 6-1: THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN PHYSICAL QUANTITIES.

QUANTITY SYMBOL SI UNIT ABBREVIATION

Ratio of specific heats γ (dimensionless) 1

Frequency f hertz Hz
Wave number k 1/meter 1/m
Equilibrium pressure p0 pascal Pa
Equilibrium density ρ0 kilogram/meter3 kg/m3
Equilibrium temperature T0 kelvin K
Speed of sound c meter/second m/s
Acoustic impedance Z pascal-second/meter Pa·s/m

As the thermoviscous acoustics physics interface solves for pressure,

velocity, and temperature, models can easily become large and contain
many DOFs. See Solver Suggestions for Large Thermoviscous Acoustics
Models for suggestions on how to solve large thermoviscous acoustic
models.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is ta.

EQUATION
Expand the Equation section to see the equations solved for with the Equation form
specified. The default selection for Equation form is set to Study controlled. The
available studies are selected under Show equations assuming.

• For Study controlled, the scaling of the equations is optimized for the numerical
performance of the different solvers and study types.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 521

 • For Frequency domain you can manually enter the scaling parameter Δ under the
Thermoviscous Acoustics Equation Settings section.
• For Mode analysis you can manually enter the Mode analysis frequency (applicable in
2D and 2D axisymmetry).

THERMOVISCOUS ACOUSTICS EQUATION SETTINGS

Adiabatic Formulation
Click to select Adiabatic formulation to use an adiabatic equation of state and disable
the temperature degree of freedom for the thermoviscous acoustic equations. This
formulation is applicable when the thermal losses can be disregarded, this is often the
case in liquids, like water. In gases, like air, on the other hand the full formulation is
necessary. When Adiabatic formulation is selected all temperature conditions and
options are disabled in the user interface.

Out-of-Plane Mode Extension

In 2D and 2D axisymmetric models, click to select Out-of-plane mode extension in order
to model the third dimension implicitly. When selected, an additional degree of
freedom, for the velocity, is solved for in the out-of-plane direction. The behavior is
determined by entering the Out-of-plane wave number kz in 2D models; or by entering
the Azimuthal mode number m in 2D axisymmetric models. The latter defines an
azimuthal wave number km = m/r. All out-of-plane gradients of the fields are then
defined in terms of the wave number. The solved equations closely follow the ones
used in the Formulation for the Boundary Mode Interface.

Scaling Factor
For all component dimensions, and if required, click to expand the Equation section,
then select Frequency domain as the Equation form and enter the settings as described
below.

The default Scaling factor Δ is 1/(iω). This value corresponds to the equations for a
Frequency Domain study when the equations are study controlled. To get the
equations corresponding to an Eigenfrequency study, change the Scaling factor Δ to 1.
Changing the scaling factor influences the coupling to other physics.

SOUND PRESSURE LEVEL SETTINGS

See the settings for Sound Pressure Level Settings for the Pressure Acoustics,
Frequency Domain interface.

522 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

TYPICAL WAVE SPEED FOR PERFECTLY MATCHED LAYERS
Enter a value or expression for the typical wave speed for perfectly matched layers cref
(SI unit: m/s). The default is 343 m/s.

STABILIZATION
To display this section, click the Show More Options button ( ) and select
Stabilization. Select No stabilization applied (the default), Galerkin least-squares (GLS)
stabilization, or Streamline upwind Petrov-Galerkin (SUPG) stabilization. When linear
thermoviscous acoustic problems are solved (like in the frequency domain) the
numerical problem is stable with the default P1-P2-P2 discretization. Enabling
stabilization will ensure stability also for other combinations of discretization orders.

GLOBAL PORT SETTINGS

Select to enable the Activate port sweep option (not selected per default). This option
is used to compute the full scattering matrix when Port conditions are used. For more
details see The Port Sweep Functionality subsection. The section only exists for 3D,
2D, and 2D axisymmetry.

Select the Mode shape normalization as Amplitude normalized (the default) or Power
normalized. This setting controls if the mode shapes are normalized to have a unit
maximum pressure amplitude or carry unit power. The selection determines how the
scattering matrix is to be interpreted.

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) for the
Pressure, the Velocity field, and the Temperature variation, respectively. The default is
Linear for the pressure and Quadratic serendipity for the velocity and the temperature.

• For numerical stability reasons, the element order for the pressure
should be one less than the element order for the velocity. Unless
Stabilization is enabled, in which case, for example, a P1-P1-P1
discretization can be used.
• In fluids where the thermal and viscous boundary layer thickness are of
the same order of magnitude (where the Prandtl number Pr is of the
order 1, like in air), it is recommended to use the same element order
for the temperature and the velocity. Both fields vary equally over the
same length scale in the acoustic boundary layers near walls.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 523

 Choosing between Lagrange and Serendipity Shape Functions has
influence on the number of DOFs solved for and on stability for distorted
meshes.

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields): the Pressure p, the
Velocity field u and its components, and the Temperature variation T. The names can
be changed but the names of fields and dependent variables must be unique within a
model.

• Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,

Frequency Domain Interface
• Theory Background for the Thermoviscous Acoustics Branch

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

• Uniform Layer Waveguide: Application Library path

Acoustics_Module/Verification_Examples/uniform_layer_waveguide
• Transfer Impedance of a Perforate : Application Library path
Acoustics_Module/Tutorials,_Thermoviscous_Acoustics/
transfer_impedance_perforate
• Generic 711 Coupler — An Occluded Ear-Canal Simulator:
Application Library path Acoustics_Module/
Tutorials,_Thermoviscous_Acoustics/generic_711_coupler

Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,

Frequency Domain Interface
The Thermoviscous Acoustics, Frequency Domain Interface has these domain,
boundary, and pair nodes available from the Physics ribbon toolbar (Windows users),

524 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Physics context menu (Mac or Linux users), or right-click to access the context menu
(all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

The Continuity node with this physics interface is available as a pair

boundary condition. This gives continuity in pressure, temperature
variation, velocity, and in the flux on a pair boundary between
thermoviscous acoustic domains.

Some of the following nodes are available from the Mechanical and Thermal submenus
(listed in alphabetical order):

• Adiabatic • Normal Impedance

• Axial Symmetry • No Slip
• Background Acoustic Fields • No Stress
• Boundary Stress • Periodic Condition
• Heat Flux • Port
• Heat Source • Pressure (Adiabatic)
• Initial Values • Slip
• Interior Normal Impedance • Symmetry
• Interior Temperature Variation • Temperature Variation
• Interior Velocity • Thermoviscous Acoustics Model
• Interior Wall • Velocity
• Isothermal • Wall

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 525

 Thermoviscous Acoustics Model
Use the Thermoviscous Acoustics Model node to define the model inputs (the
background equilibrium temperature and pressure) and the material properties of the
fluid (dynamic viscosity, bulk viscosity, thermal conductivity, heat capacity at constant
pressure, and equilibrium density) necessary to model the propagation of acoustic
compressible waves in a thermoviscous acoustic context. Extended inputs are available
for the coefficient of thermal expansion and the compressibility, which enables
modeling of any constitutive relation for the fluid.

MODEL INPUTS
This section contains field variables that appear as model inputs. The fields for the
Equilibrium pressure p0 and the Equilibrium temperature T0 are always active as they
enter the governing equations explicitly. If material properties depend on other model
inputs they will automatically appear in this section.

Select User defined (the default), Common model input, or an existing variable from
another physics interface.

The Equilibrium pressure p0 (SI unit: Pa) has the default value set to 1 atm. The
Equilibrium temperature T0 (SI unit: K) has the default value set to 293.15 K (that is,
20oC).

Details about the Model Input and the Default Model Inputs are found
in the Global and Local Definitions chapter of the COMSOL
Multiphysics Reference Manual.

FLUID PROPERTIES
Define the material properties of the fluid by selecting an Equilibrium density — Ideal
gas, From material, or User defined.

• If From material is selected (the default), the equilibrium density, and its dependence
on the equilibrium pressure p0 and temperature T0, is taken from the defined
material. Make sure that the Thermal Expansion and Compressibility settings are
correct.

526 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

• For Ideal gas also select the Gas constant type — select Specific gas constant Rs (SI
unit: J/(kg·K) or Mean molar mass Mn (SI unit: kg/mol)
• For User defined enter a value or expression for the Equilibrium density ρ0(p0, T0) (SI
unit: kg/m3). The default is ta.p0/(287[J/kg/K]*ta.T0), which is the ideal gas
law.

The other thermoviscous acoustic model parameters defaults use values From material.
For User defined enter another value or expression for:

• Dynamic viscosity μ (SI unit: Pa·s).

• Bulk viscosity μB (SI unit: Pa·s). The bulk viscosity parameter describes the
difference between the mechanical and thermodynamic pressures. It is associated
with losses due to expansion and compression.
• Thermal conductivity k (SI unit: W/(m·K)).
• Heat capacity at constant pressure Cp (SI unit: J/(kg·K)). This is the specific heat
capacity or heat capacity per unit mass.

For a discussion about the air and water materials as commonly used in
acoustics and other material properties, such as the bulk viscosity, see the
Acoustic Properties of Fluids chapter of this manual.

THERMAL EXPANSION AND COMPRESSIBILITY

One of the main characteristics of an acoustic wave is that it is a compressional wave.
In the detailed thermoviscous acoustic description, this property is closely related to
the constitutive relation between the density, the pressure, and the temperature. This
results in the important (linear) relation for the acoustic density variation

ρt = ρ0 ( βT pt – α p T t )

where ρt is the total density variation, pt is the total acoustic pressure, Tt is the total
acoustic temperature variations, βT is the (isothermal) compressibility of the fluid, and
αp the (isobaric) coefficient of thermal expansion (sometimes named α0). If this
constitutive relation is not correct, then no waves propagate or possibly they propagate
at an erroneous speed of sound. The default behavior is to define both quantities from
the speed of sound and the ratio of specific heats (using the From speed of sound
option) which are material properties often more readily available.

Note that, when the Adiabatic formulation is selected option under the Thermoviscous
Acoustics Equation Settings section the equation of stated reduces to

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 527

 βT
ρ t = ρ 0 β s p t = ρ 0 ------ p t .
γ

When the From equilibrium density option is selected for the coefficient of thermal
expansion and the compressibility, both values are derived from the equilibrium
density ρ0(p0,T0) using their defining relations

1 ∂ρ 0 1 ∂ρ 0
β T = ------ --------- α p = – ------ ----------
ρ 0 ∂p 0 T0 ρ 0 ∂T 0 p0

If the equilibrium density ρ0 is a user-defined constant value, is picked up from another

physics interface, or the material model does not define both a pressure and
temperature dependence for ρ0, the coefficient of thermal expansion and/or the
compressibility will evaluate to 0. Then the default From speed of sound option or the
User defined options should be used.

If the material is air, the From equilibrium density option works well as the
equilibrium density ρ0 = ρ0(p0,T0) is defined as a function of both
pressure and temperature.

For the water material the coefficient of thermal expansion is well defined
as ρ0 = ρ0(T0), while the compressibility should be defined using the
default From speed of sound option.

The Thermal Expansion and Compressibility section is displayed if From material or User
defined is selected as the Equilibrium density under Thermoviscous Acoustics Model. For
the Ideal gas option, the parameters are readily defined.

Select an option from the Coefficient of thermal expansion αp list — From material, From
equilibrium density, From speed of sound (the default), or User defined. For User defined
enter a value for αp (SI unit: 1/K = K-1).

Select an option from the Isothermal compressibility βT list — From equilibrium density,
From isentropic compressibility, From speed of sound (the default), or User defined. For
User defined, enter a value for βT (SI unit: 1/Pa = Pa-1).

The different options for defining the coefficient of thermal expansion and the
compressibility stem from their thermodynamic definitions:

528 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

 1 ∂ρ 0 1 γ
β T = ------ --------- = ------ ----2- = γβ s
ρ 0 ∂p 0 T ρ0 c

1 ∂ρ 0 1 Cp ( γ – 1 )
α p = – ------ ---------- = --- -----------------------
-
ρ 0 ∂T 0 p c T0

For each of the following, and based on the above selection, the default is taken From
material. For User defined enter another value or expression in the text field.

• Speed of sound c (SI unit: m/s).

• Ratio of specific heats γ (dimensionless). The default is 1.
• Isentropic compressibility βs (SI unit: 1/Pa = Pa-1) .

See the Theory Background for the Thermoviscous Acoustics Branch

section for a detailed description of the governing equations and the
constitutive relations.

To visualize the dissipated energy due to viscosity and thermal conduction

in postprocessing. Three postprocessing variables exist:

• The viscous power dissipation density ta.diss_visc.

• The thermal power dissipation density ta.diss_therm.
• The total thermo-viscous power dissipation density ta.diss_tot.

In certain cases, it can be interesting not to include thermal conduction

in the model and treat all processes as adiabatic (isentropic). This is, for
example, relevant for most liquids where the thermal boundary layer is
much thinner than the viscous. Not solving for the temperature field T
also saves some degrees of freedom (DOFs). This is achieved by selecting
the Adiabatic formulation option under the Thermoviscous Acoustics
Equation Settings section.

See also Solver Suggestions for Large Thermoviscous Acoustics Models

for suggestions on how to select an iterative solver for large problems.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 529

 Background Acoustic Fields
When the Background Acoustic Fields feature is added to a domain, it is possible to
define the value of the background acoustic field variables pb, ub, and Tb. Use this
feature to set up scattering problems. To define fields at the inlet of a waveguide, it is
recommended to use the Port condition.

MODEL INPUTS
This section contains field variables that appear as model inputs. The fields are always
active as the equilibrium (background) temperature T0 enters the governing energy
equation explicitly. It is used for the Plane wave option.

BACKGROUND ACOUSTIC FIELDS

Select the Acoustic field type — User defined (the default) or Plane wave.

When User defined is selected, enter expressions for the Background acoustic pressure
pb, the Background acoustic velocity ub, and the Background temperature variation Tb.
These can be analytical expressions or values of dependent variables solved in a
previous study that defines the background field.

When Plane wave is selected, enter the Pressure Amplitude |pb| (given at x = 0) and the
wave direction vector ek (this vector is automatically normalized). Finally, select where
the Material data should be taken from; the default is From material model (the same as
the domain model material) or select a material from the list. This will set up the
background pressure, velocity, and temperature variations for a plane wave with the
correct thermoviscous attenuation compatible with the governing equations.

For many applications the Plane wave option available in the Port
condition is a good alternative to the Background Acoustic Fields feature.
The port automatically ensures a non-reflective behavior for the reflected
outgoing waves.

Heat Source
Use the Heat Source node to define the heat source for the thermoviscous acoustics
model. This adds a domain heat source Q to the right-hand side of the energy
equation.

530 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

HEAT SOURCE
Enter a value for the Heat source Q (SI unit: W/m3).

Initial Values
The Initial Values node adds initial values for the sound pressure, velocity field, and
temperature variation. If necessary, add more Initial Values nodes from the Physics
toolbar to give initial values in, for example, other domains.

INITIAL VALUES
Enter values or expressions for the Pressure p (SI unit: Pa), Velocity field u (SI unit: m/
s), and Temperature variation T (SI unit: K).

Axial Symmetry
In a 2D axisymmetric geometry the Axial Symmetry condition is automatically added
on the axis of symmetry at r = 0. On this boundary the radial component of the total
velocity is set equal to zero

ut = 0 u t = ( u t, v t, w t )

Wall
Use the Wall node to model the most common conditions at solid surfaces. This is the
default boundary condition. This condition contains both a mechanical and a thermal
selection. The default is a no-slip and isothermal condition applicable in most cases.
More advanced conditions at boundaries can be set up by combining any of the
Mechanical and Thermal conditions available.

The Wall condition is also the default condition added when the physics interface is set
up. Note that the selections in this default condition can be changed (they are not
locked to all boundaries). When used together with a perfectly matched layer (PML)
it is recommended to use the no-slip option at the outer boundary of the layer.
Specifically, using the slip condition in combination with a rational scaling in the PML
will lead to a singularity and an error.

The no-slip condition is the origin of the viscous boundary layer and the isothermal
condition is the origin of the thermal boundary layer. It is within these acoustic
boundary layers that the main dissipation happens. See the Theory Background for the
Thermoviscous Acoustics Branch section for more details.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 531

 MECHANICAL
Select a Mechanical condition — No slip (the default) or Slip. See the No Slip and Slip
conditions for further details.

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic. See the Isothermal
and Adiabatic conditions for further details.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. If Slip is selected for the Mechanical
condition and Use weak constraints is enabled, then the default discontinuous Galerkin
(penalty like) formulation is switched to a Lagrange multiplier formulation instead. See
the Slip condition for details. For the No-slip and Isothermal selections a weak
formulation is used instead of a pointwise constraint.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Pressure (Adiabatic)
Use the Pressure (Adiabatic) node to specify a prescribed pressure pbnd, that acts as a
pressure source at the boundary, typically an inlet or outlet. In the frequency domain
pbnd is the amplitude of a harmonic pressure source. The adiabatic condition states
that no heat flows into or out of the boundary:

p t = p bnd

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = – p bnd n

2μ
3
– n ⋅ ( – k ∇T t ) = 0

This condition is in general not physically correct on a solid wall because solids are
generally better thermal conductors than air.

PRESSURE
Enter the value of the Pressure pbnd (SI unit: Pa) at the boundary. The default is 0 Pa.

532 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Symmetry
The Symmetry node for The Thermoviscous Acoustics, Frequency Domain Interface
adds a boundary condition that represents symmetry. It corresponds to the Slip
condition for the mechanical degrees of freedom and the Adiabatic condition for the
temperature variation.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Port
The Port boundary condition is used to excite and absorb acoustic waves that enter or
leave waveguide structures, like small ducts or channels, in an acoustic model. The
thermoviscous port formulation ensures that the nontrivial mode shapes of the
acoustic velocity and thermal fields are captured correctly.

A given port condition supports one specific propagating mode. To provide the full
acoustic description, combine several port conditions on the same boundary. Typically,
only the plane wave mode is propagating in small structures where the thermoviscous
representation is necessary. The port condition provides a superior nonreflecting or
radiation condition for waveguides compared to a simple impedance condition or a
perfectly matched layer (PML) configuration. The same port boundary condition
feature should not be applied to several waveguide inlets and outlets. The port

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 533

 condition supports S-parameter (scattering parameter) calculation but it can also be
used as a source to just excite a system.

The Port boundary condition exists for 3D, 2D, and 2D axisymmetric models.

Only one port should be excited at a time if the purpose is to compute

S-parameters. The S-parameters are defined as ta.S11, ta.S21, etc. and
can be used in postprocessing.

In other cases, having more than one port exciting the system might be
wanted, but the S-parameter variables cannot be correctly computed. So,
when several ports are excited, the S-parameter output is turned off.

If only two ports are added to the system and a port sweep is performed,
the system transfer matrix and system impedance matrices are also
computed.

On a given boundary, a combination of ports will define the total acoustic fields (sum
of incident and outgoing pressure, temperature, and velocity waves) as


in iφ
pt = A e ( S ij + δ ij )p i
i ∈ bnd


in iφ
ut = A e ( S ij u i + δ ij ( u i – 2n ( n ⋅ u i ) ) )
i ∈ bnd


in iφ
Tt = A e ( S ij + δ ij )T i
i ∈ bnd

where the summation “i” is over all ports on the given boundary “bnd”, Sij is the
scattering parameter, Ain is the amplitude of the incident field and φ the phase (at port
“j”), and pi, ui, and Ti are the mode shape of the i-th port. The mode shape is
normalized to have either a unit maximum amplitude (for the pressure pi) or carry unit
power (see the normalization option in the Global Port Settings section). For both
definitions the scattering parameter Sij defines the amplitude of mode i when a system
is exited at port j (with mode j). For the power scaling, |Sij|2 directly gives the power
of the given mode. This corresponds to a multimode expansion of the solution on the
given boundary. The scattering parameters are automatically calculated when an
acoustic model is set up with just one port exciting the system. To get the full
scattering matrix The Port Sweep Functionality can be used.

534 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

PORT PROPERTIES
Enter a unique Port name. Only nonnegative integer numbers can be used as Port name
as it is used to define the elements of the S-parameter matrix. The numeric port names
are also required for the port sweep functionality. The port name is automatically
incremented by one every time a port condition is added.

Select a Type of port: User defined (the default), Numeric (0,0)-mode, Circular (0,0)-mode,
Slit (0,0)-mode, or Plane wave. Depending on the selection, different options appear in
the Port Mode Settings section (see below). Use the Circular (0,0)-mode for a port with
a circular cross section in 3D or 2D axisymmetry and the Slit (0,0)-mode option in 2D.
If the port has a different cross section than either of these, use the User defined option
or the Numeric (0,0)-mode port. The Plane wave option represents a situation where the
boundary layers are not included, this is for example for a wave propagating in free
space or when slip and adiabatic (or symmetry) conditions are applied to all adjacent
boundaries.

PORT MODE SETTINGS

Depending on the option selected in the Type of port (see above):

• For User defined, enter user defined expressions for the Mode shape pn, un, Tn, and
the Mode wave number kn (SI unit: rad/m). The mode shape will automatically be
scaled before it is used in the port condition. Use the user-defined option to enter
a known analytical expression or to use the solution from The Thermoviscous
Acoustics, Boundary Mode Interface. The solutions from the boundary mode
analysis can be referenced using the withsol() operator.
• The Numeric (0,0)-mode port options is used for waveguides of arbitrary cross
sections. In this case, the shape of the propagating plane-wave mode (0,0) is solved
on the port face. The boundary conditions for the mode are taken from the adjacent
waveguide boundaries. This automatic detection works for slip, no-slip, adiabatic,
isothermal, and symmetry conditions (including the same options when selected in
the wall condition). If all the adjacent wave guide boundaries are slip and adiabatic
(or symmetry) then use the Plane wave option.
For this option, a special solver sequence is automatically generated since the port
mode shape (the port variables psi, Psi_th, Psi_v, and vip) should be solved
before the domain problem (the main degrees of freedom p, u, T, and Sparam1). If
an iterative solver suggestion is to be used, keep the linear solver for the first
segregated step and then select the iterative suggestion for the second segregated
step.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 535

 • The Circular (0,0)-mode port option is used for waveguides of circular cross section.
The analytical mode is a plane-wave mode (0,0) given by a constant cross section
pressure, no-slip condition for the velocity, and isothermal condition for the
temperature. An example of the propagating mode shape in a cylindrical waveguide
is seen below.

pn |un| Tn
Figure 6-1: Plane wave mode for a circular duct of 1 mm diameter at f = 250 Hz.

Select how the Circle radius of the cross section is defined, either Automatic (the
default) or User defined. The latter option can for some geometry configurations
increase the numerical precision of the computed mode.
• The Slit (0,0)-mode port option only exists in 2D on a boundary. In 2D, the
geometry is assumed infinite in the out-of-plane direction and represents a slit. The
analytical mode is a plane-wave mode (0,0) given by a constant cross section
pressure, no-slip condition for the velocity, and isothermal condition for the
temperature.
• The Plane wave port option represents a situation where the boundary layers are not
included in the mode shape, this is, for example, for a wave propagating in free space
or when slip and adiabatic (or symmetry) conditions are applied to all adjacent
boundaries.

The solution for the numeric port is obtained using the generalized low
reduced frequency (LRF) approximation commonly used in
thermoviscous acoustics. The theory is based on an order of magnitude
analysis of the full thermoviscous equations and on the assumption of a
plane propagating mode (constant pressure in cross section). The circular
and slit options use an analytical solution for the given cross section.

The mode shapes based on LRF approximation are valid as long as the
wavelength is much larger than the waveguide cross section (λ >> a) and
the wavelength is much larger than the boundary layer thickness
(λ >> δvisc and λ >> δtherm).

536 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

INCIDENT MODE SETTINGS
Activate if the given port is excited by an incident wave of the given mode shape. For
the first Port condition added in a model, the Incident wave excitation at this port is set
to On. For subsequent conditions added the excitation is set to Off per default. If more
than one port in a model is excited, the S-parameter calculation is not performed.

When the Incident wave excitation at this port is set to On, then select how to define
the incident wave. Select the Define incident wave: Amplitude (the default) or Power
(Power per unit length in 2D models)

• For Amplitude enter the amplitude Ain (SI unit: Pa) of the incident wave. This is in
general defined as the maximum pressure amplitude for a given mode shape.
• For Power enter the power Pin (SI unit: W) of the incident mode. In 2D models this
will be a Power per unit length (SI unit: W/m).
• Enter the phase φ (SI unit: rad) of the incident wave. This phase contribution is
multiplied with the amplitude defined through the above two options. The
Amplitude input can be a complex number.

Note, that when the Activate port sweep option is selected at the physics level, the
options in the Incident Mode Settings section are deactivated. This is because this option
automatically sends in a mode of unit amplitude, sweeping through one port at the
time.

All ports with incident wave excitation turned off have an arrow symbol,
added in the Graphics window, that points outward. When incident wave
excitation is turned on, the arrow symbol points inward.

If a waveguide is cut with one or several symmetry planes, care should be

taken when setting up the port modes and when postprocessing.

For the Circular and Slit options, make sure to only select modes that are
actually symmetric according to the symmetry planes.

When postprocessing, remember that absolute values like, for example,

the outgoing power at port 1, ta.port1.P_out, need to be multiplied
with an appropriate factor. Multiplication with two if one symmetry plane
is used, for example.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 537

 CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details

THE PORT SWEEP FUNCTIONALITY

The port sweep functionality is used to reconstruct the full scattering matrix Sij by
automatically sweeping the port excitation through all the ports included in the model.
When the port sweep is activated, the options in the Incident Mode Settings in the port
conditions are deactivated and COMSOL controls which port is excited with an
incident mode.

The port sweep functionality is activated at the main physics interface level by selecting
Activate port sweep in the Global Port Settings section. Enter the Sweep parameter
name, the default is PortName. Create a parameter with the same name under Global
Definitions>Parameters 1. This is the name of the parameter to be used in a parametric
sweep; it should represent the Port name integer values (defined when adding the port
conditions). Add a parametric sweep study step and run the sweep over the PortName
parameter with an integer number of values representing all the ports in the model.
Once the model is solved, the full scattering matrix can be evaluated using the defined
global variables ta.S11, ta.S21, ta.S12, and so on. The transmission loss (TL)
between two given ports is also computed, for example, the variable for the TL loss
from port 1 to 2 is given by ta.TL_12.

Use the Global Matrix Evaluation under Derived Values to evaluate the full
scattering matrix ta.S.

If only two ports are added to the thermoviscous model, COMSOL also automatically
computes the transfer matrix of the system (variables ta.T11, ta.T12, ta.T21,
ta.T22) and the impedance matrix of the system (ta.Z11, ta.Z12, ta.Z21, ta.Z22).
These expressions are only true if plane wave modes are used. This is nearly the case in
all configurations when working with microacoustic systems. For ports in
thermoviscous acoustics, the Numeric (0,0)-mode, Circular (0,0)-mode, Slit (0,0)-mode,
and Plane wave options are for plane waves only, that is, the (0,0) mode with varying
boundary conditions. Higher-order modes can only be introduced with the User

538 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

defined option. The transfer matrix representation is often used in electroacoustic
modeling.

Wax Guard Acoustics: Transfer Matrix Computation. The Application

Library path: Acoustics_Module/Tutorials,_Thermoviscous_Acoustics/
wax_guard_acoustics

Transfer Impedance of a Perforate. The Application Library path:

Acoustics_Module/Tutorials,_Thermoviscous_Acoustics/
transfer_impedance_perforate

Periodic Condition
The Periodic Condition node adds a periodic boundary condition that can be used to
reduce the model size by using symmetries and periodicities in the geometry and
physics interfaces being modeled. This feature works well for cases like opposing
parallel boundaries. In other cases, use a Destination Selection subnode to control the
destination. By default, it contains the source and destination selection that COMSOL
Multiphysics identifies. The periodic condition automatically defines a mapping
between the source xs and the destination xd points on the boundaries.

PERIODICITY SETTINGS
Select a Type of periodicity: Continuity (the default), Antiperiodicity, Floquet periodicity
(Bloch periodicity), Cyclic symmetry, or User Defined.

• For Continuity the values of the field variables at destination are set equal to the
source: p(xd) = p(xs), u(xd) = u(xs), and T(xd) = T(xs). If the source and
destination boundaries are rotated with respect to each other, a transformation is
automatically performed, so that corresponding velocity components are
connected.
• For Antiperiodicity the values of the field variables on the destination are set equal
to the values on the source with the sign reversed: p(xd) = -p(xs), u(xd) = -u(xs),
and T(xd) = -T(xs). If the source and destination boundaries are rotated with
respect to each other, a transformation is automatically performed, so that
corresponding velocity components are connected.
• For Floquet periodicity, also known as Bloch periodicity, enter a k-vector for Floquet
periodicity kF (SI unit: rad/m) for the x, y, and z coordinates (3D components), the

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 539

 r and z coordinates (2D axisymmetric components), or x and y coordinates (2D
components). This is the wave number of the excitation.
This condition is used to model infinite periodic structures with nonnormal incident
pressure fields or excitations. Use it to model, for example, a large perforated plate
with an oblique incident wave with wave vector k (and set kF = k) by only
analyzing one hole or one subset of holes that is periodic.
• For Cyclic symmetry select a Sector angle: Automatic (the default) or User defined. For
User defined enter a value for θS (SI unit: rad). Enter an Azimuthal mode number m
(dimensionless).
This condition is used to model any geometry that has a cyclic periodic structure
such as a microphone or a loudspeaker driver. Setting the azimuthal mode number
determines what mode is analyzed. The response of the full system to an external
excitation is in general a linear combination of many different modes.
• For User defined select the check box for any of the field variables as needed. Then
for each selection, choose the Type of periodicity — Continuity or Antiperiodicity. If
the source and destination boundaries are rotated with respect to each other, a
transformation is automatically performed, so that corresponding velocity
components are connected.
In the time domain, both the Cyclic symmetry and the Floquet periodicity boundary
conditions reduce to the continuity condition.

To optimize the performance of the Floquet periodicity and the Cyclic

symmetry conditions, it is recommended that the source and destination
meshes are identical. This can be achieved by first meshing the source
boundary or edge and then copying the mesh to the destination boundary
or edge. When the Periodic Condition stretches across regions with a mix
of default material models, PMLs, background pressure fields, or
background acoustic fields, it is recommended to add one Periodic
Condition for each set of such boundaries.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

540 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

ORIENTATION OF SOURCE
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

See Orientation of Source and Destination in the COMSOL Multiphysics

Reference Manual.

Interior Wall
The Interior Wall boundary condition is used to model a wall condition on an interior
boundary. It is similar to the Wall boundary condition available on exterior boundaries
except that it applies on both sides (up and down) of an internal boundary. It allows
discontinuities (in pressure, velocity, and temperature) across the boundary. Use the
Interior Wall boundary condition to avoid meshing thin structures by applying this
slip-like condition on interior curves and surfaces instead.

MECHANICAL
Select a Mechanical condition — No slip (the default) or Slip. See the No Slip and Slip
conditions for further details.

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic. See the Isothermal
and Adiabatic conditions for further details.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

If Slip is selected for the Mechanical condition and Use weak constraints is enabled, the
default discontinuous Galerkin (penalty like) formulation is switched to a Lagrange
multiplier formulation instead. For the No-slip and Isothermal selections a weak
formulation is used instead of a pointwise constraint.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 541

 EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Interior Normal Impedance

Use the Interior Normal Impedance node to apply a transfer impedance condition on an
interior boundary. The condition relates the normal stress at the two sides of the
boundary (up and down) to the normal velocity via the normal transfer impedance Zn.
The condition allows discontinuities (a slit) in pressure and temperature depending on
the condition chosen for the temperature.

( σ up – σ down )n = Z n ( u t ⋅ n )n

σ = – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I
2μ
 3 
p t,down – p t,up
Z n = -----------------------------------
ut ⋅ n

MECHANICAL
Enter a value or expression for the Normal impedance Zn (SI unit: Pa⋅s/m). This is the
transfer impedance of the interior boundary.

Select a condition for the Tangential velocity — Slip (the default) or No slip. The No Slip
option will apply an additional constraint for the velocity in the tangential direction on
the two sides of the interior boundary.

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic. See the Isothermal
and Adiabatic conditions for further details.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

542 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Interior Velocity
Use the Interior Velocity node to specify a velocity on an interior boundary in
thermoviscous acoustics. The condition can be used to specify sources, for example,
the velocity of a diaphragm in a miniature transducer that is modeled using a lumped
circuit model. The velocity components can be prescribed independently, and there is
an option to force continuity in pressure across the boundary, as well as options for the
thermal conditions.

MECHANICAL
Select the Continuous pressure check box if the velocity condition is to be used as a
source where the pressure should not be slit.

To define a prescribed velocity for each space direction (x and y, plus z for
3D), select one or more of the Prescribed in x direction, Prescribed in y
direction, and Prescribed in z direction check boxes. Then enter a value or
expression for the prescribed velocities u0x, u0y, or u0z (SI unit: m/s).

To define a prescribed velocity for each space direction (r and z), select
one or both of the Prescribed in r direction and Prescribed in z direction
check boxes. Then enter a value or expression for the prescribed velocities
u0r or u0z (SI unit: m/s).

THERMAL
Select a Thermal condition — Isothermal (the default) or Adiabatic. See the Isothermal
and Adiabatic conditions for further details.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. For the Interior Velocity condition
the Constraint method is set to Nodal.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 543

 Interior Temperature Variation
Use the Interior Temperature Variation condition to model a thermal type of source on
an interior boundaries, Tt = Tbnd. This can, for example, be a thin sheet that has a
harmonically varying temperature, given by Tbnd, for example, due to ohmic heating.
This temperature fluctuation will give rise to acoustic waves that can be modeled using
the Thermoviscous Acoustics interface and this boundary condition.

MECHANICAL
Select a Mechanical condition — No slip (the default) or Slip. See the No Slip and Slip
conditions for further details.

THERMAL
Enter a value or expression for the Temperature variation Tbnd (SI unit: K) that has to
be applied at the interior boundary.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

No Slip
The No Slip node sets up a no-slip condition for the acoustic velocity fluctuations on a
hard wall. Add this node from the Mechanical submenu. The condition reads ut = 0.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

544 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Mechanical and thermal boundary conditions contribute such that a

condition can be set on the velocity and temperature simultaneously (they
do not override). Thermal type condition override each other and the
same is true for mechanical type conditions.

Slip
Use the Slip node to prescribe zero normal velocity, and therefore also zero tangential
stress, condition on the boundary. Add this node from the Mechanical submenu. The
condition reads

n ⋅ ut = 0
σ n – ( σ n ⋅ n )n = 0

σ n = – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n
2μ
3

This results in a so-called no-penetration condition where no viscous boundary layer

is created. Use the condition in places where the viscous losses in the boundary layer
are nonimportant. Here it is not necessary to mesh the boundary layer resulting in
fewer mesh elements.

When running a model with a perfectly matched layer (PML), it is recommended not
to combine a Slip condition with a rational scaling in the PML, as this will lead to a
singularity and an error.

Per default, the Slip condition uses a so-called discontinuous Galerkin or penalty
formulation to prevent locking problems on curved surfaces.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Select Use weak constraints if you want to switch from the default discontinuous
Galerkin (penalty like) formulation to a Lagrange multiplier formulation for the slip

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 545

 condition (this formulation is not suited for an iterative solver). The Lagrange
multiplier formulation also prevents locking problems.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Velocity
Use the Velocity node to define the prescribed velocities u0 on the boundary. Add this
node from the Mechanical submenu. The condition reads ut = u0. This condition is
useful, for example, when modeling a vibrating wall.

VELOCITY

To define a prescribed velocity for each space direction (x and y, plus z for
3D), select one or more of the Prescribed in x direction, Prescribed in y
direction, and Prescribed in z direction check boxes. Then enter a value or
expression for the prescribed velocities u0x, u0y, or u0z (SI unit: m/s).

To define a prescribed velocity for each space direction (r and z), select
one or both of the Prescribed in r direction and Prescribed in z direction
check boxes. Then enter a value or expression for the prescribed velocities
u0r or u0z (SI unit: m/s).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. For the Velocity condition the
Constraint method is set to Nodal.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

546 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

No Stress
Use the No Stress node to set the total surface stress equal to zero. Add this node from
the Mechanical submenu. The condition reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = 0
2μ
 3 

Boundary Stress
Use the Boundary Stress node to define the total surface (boundary) stress through a
stress tensor, a traction force, or a pressure. Add this node from the Mechanical
submenu. The stress condition reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = σ p n
2μ
 3 

where σ p is a symmetric stress tensor, the product σ p n defines a traction force. The
traction condition reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = F tr
2μ
3

where Ftr is the surface traction force. The pressure condition reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = – p bnd n

2μ
3

where pbnd is the pressure applied to the boundary (acting in the normal direction).

STRESS
Select a Stress type — Stress tensor, Traction (the default), or Pressure. For Stress tensor
enter the (symmetric) Stress tensor σ p (SI unit: N/m2) components, for Traction
enter the traction force Ftr (SI unit: N/m2) components, and for Pressure enter the
boundary pressure pbnd (SI unit: Pa).

Normal Impedance
Use the Normal Impedance node to specify a normal specific impedance Z0 on a
boundary. Add this node from the Mechanical submenu. This feature is useful outside
the viscous boundary layer, as this condition mimics the behavior of a corresponding

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 547

 Pressure Acoustics Model with a normal impedance condition. The boundary condition
reads

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = – Z 0 ( u t ⋅ n )n
2μ
 3 

MECHANICAL
Enter a value or expression for the Normal impedance Z0 (SI unit: Pa⋅s/m).

Select a condition for the Tangential velocity — Slip (the default) or No slip. The No Slip
option will apply an additional constraint for the velocity in the tangential direction on
the boundary.

Isothermal
Use the Isothermal node to model a wall that is assumed to be a good thermal
conductor and backed by a large heat reservoir kept at constant temperature. This
implies that the harmonic temperature variations vanish: Tt = 0. Add this node from
the Thermal submenu.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Adiabatic
Use the Adiabatic node to define a situation with no heat flow into or out of the
boundary. Add this node from the Thermal submenu. The condition reads

– n ⋅ ( – k ∇T t ) = 0

548 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Temperature Variation
Use the Temperature Variation node to define the temperature variation on the
boundary Tbnd. Add this node from the Thermal submenu. In the frequency domain,
this is the amplitude of a harmonic temperature variation, Tt = Tbnd.

TEMPERATURE VARIATION
Enter a value or expression for the Temperature variation Tbnd (SI unit: K).

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

EXCLUDED EDGES/POINTS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. See Suppressing Constraints on
Lower Dimensions for details.

Heat Flux
Use the Heat Flux node to define a thermal source given by an inward normal heat flux
qn = -n·q through the boundary. Add this node from the Thermal submenu. The
condition reads

– n ⋅ ( – k ∇T t ) = q n

HEAT FLUX
Enter a value for the Inward normal heat flux qn (SI unit: W/m2).

THE THERMOVISCOUS ACOUSTICS, FREQUENCY DOMAIN INTERFACE | 549

 The Thermoviscous Acoustics,
Transient Interface
The Thermoviscous Acoustics, Transient (tatd) interface ( ), found under the
Thermoviscous Acoustics branch ( ) when adding a physics interface, is used to
compute the transient evolution of the acoustic variations in pressure, velocity, and
temperature. The interface is the time domain equivalent of The Thermoviscous
Acoustics, Frequency Domain Interface. This physics interface is required to accurately
model acoustics in geometries of small dimensions. Near walls, viscous losses and
dissipation due to thermal conduction become important because boundary layers
exists. The thicknesses of these boundary layers are known as the viscous and thermal
penetration depth. For this reason, it is necessary to include thermal conduction effects
and viscous losses explicitly in the governing equations. It is, for example, used when
modeling the response of transducers like microphones, miniature loudspeakers and
receivers. Other applications include analyzing feedback in hearing aids, smart phones
and in mobile devices, or studying the damped vibrations of MEMS structures.

The physics interface solves the equations in the time domain. The model can be
extended to model nonlinear effects by adding the Nonlinear Thermoviscous
Acoustics Contributions feature. In the time domain it is also possible to model
nonlinear effects due to topology changes, like nonlinear squeeze film damping. This
is achieved when combining the interface with the Moving Mesh functionality.

The equations defined by the Thermoviscous Acoustics, Transient interface are the
first and second order perturbation formulation of the Navier-Stokes equations in
quiescent background conditions solving the continuity, momentum, and energy
equations. Due to the detailed description necessary when modeling thermoviscous
acoustics, the model simultaneously solves for the acoustic pressure p, the acoustic
velocity variation u (particle velocity), and the acoustic temperature variations T. It is
available for 3D, 2D, and 1D Cartesian geometries as well as for 2D and 1D
axisymmetric geometries.

The Thermoviscous Acoustics, Transient interface is, as the frequency domain variant,
formulated in the so-called scattered field formulation where the total field (subscript
t) is the sum of the scattered field (the field solved for, p, u, and T) and a possible
background acoustic field (subscript b), such that

550 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

 pt = p + pb ut = u + ub Tt = T + Tb

The scattered field formulation is not applicable in domains where the Nonlinear
Thermoviscous Acoustics Contributions are included. When no Background Acoustic
Fields feature is present (the background field values are zero per default) the total field
is simply the acoustic field solved for, pt = p, ut = u, and Tt = T. All governing
equations and boundary conditions are formulated in the total field variables.

When this physics interface is added, these default nodes are also added to the Model
Builder — Thermoviscous Acoustics Model, Wall, and Initial Values. Then, from the
Physics toolbar, add other nodes that implement, for example, boundary conditions
and sources. You can also right-click Thermoviscous Acoustics to select physics features
from the context menu.

For good modeling strategies, solver suggestions, postprocessing

information, as well as tips and tricks, see the Modeling with the
Thermoviscous Acoustics Branch section.

For more details about the physics interface see On the Thermoviscous
Acoustics Physics Interface subsection, under The Thermoviscous
Acoustics, Frequency Domain Interface section. Details about the
governing equations are found in the Theory Background for the
Thermoviscous Acoustics Branch section.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is tatd.

THE THERMOVISCOUS ACOUSTICS, TRANSIENT INTERFACE | 551

 EQUATION
Expand the Equation section to see the equations solved for with the Equation form
specified. The default selection for Equation form is set to Study controlled. The
available studies are selected under Show equations assuming.

• For Study controlled, the scaling of the equations is optimized for the numerical
performance of the different solvers and study types.
• For Frequency domain you can manually enter the scaling factor Δ under the
Thermoviscous Acoustics Equation Settings section.

THERMOVISCOUS ACOUSTICS EQUATIONS SETTINGS

See Thermoviscous Acoustics Equation Settings.

STABILIZATION
To display this section, click the Show More Options button ( ) and select
Stabilization. Select No stabilization applied (the default), Galerkin least-squares (GLS)
stabilization, or Streamline upwind Petrov-Galerkin (SUPG) stabilization. When linear
thermoviscous acoustic problems are solved, the problem is stable (with the default
P1-P2-P2 discretization), but as soon as the Nonlinear Thermoviscous Acoustics
Contributions feature is used, stabilization may be required. For weakly nonlinear
problems, no stabilization is necessary, but for moderate and highly nonlinear
problems using stabilization is essential. In most of those cases, use the Galerkin
least-squares (GLS) stabilization option.

TRANSIENT SOLVER SETTINGS

Enter the Maximum frequency to resolve in the model. The default frequency is set to
1000[Hz] but should be changed to reflect the frequency content of the sources used
in the model. Select the Time stepping (method) as Fixed (preferred) the default and
recommended or Free. The Free option is in general not recommended for wave
problems. The generated solver will be adequate in most situations if the
computational mesh also resolves the frequency content in the model. Note that any
changes made to these settings (after the model is solved the first time) will only be
reflected in the solver if Show Default Solver or Reset Solver to Defaults is selected in the
study.

Details about Transient Solver Settings are found in the section Modeling
with the Thermoviscous Acoustics Branch.

552 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

DISCRETIZATION
From the list select the element order and type (Lagrange or serendipity) for the
Pressure, the Velocity field, and the Temperature variation, respectively. The default is
Linear for the pressure and Quadratic serendipity for the velocity and the temperature.

• For numerical stability reasons, the element order for the pressure
should one less than the element order for the velocity. Unless
Stabilization is enabled, in which case, for example, a P1-P1-P1
discretization can be used.
• In fluids where the thermal and viscous boundary layer thickness are of
the same order of magnitude (where the Prandtl number Pr is of the
order 1, like in air), it is recommended to use the same shape order for
the temperature and the velocity. Both fields vary equally over the same
length scale in the acoustic boundary layers near walls.

Choosing between Lagrange and Serendipity Shape Functions has

influence on the number of DOFs solved for and on stability for distorted
mesh.

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields), the Pressure p, the
Velocity field u and its components, and the Temperature variation T. The names can
be changed but the names of fields and dependent variables must be unique within a
model.

Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,

Transient Interface
The Thermoviscous Acoustics, Transient Interface has these domain, boundary, and
pair nodes available from the Physics ribbon toolbar (Windows users), Physics context
menu (Mac or Linux users), or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE THERMOVISCOUS ACOUSTICS, TRANSIENT INTERFACE | 553

 In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

The Continuity node with this physics interface is available as a pair

boundary condition. This gives continuity in pressure, temperature
variation, velocity, and in the flux on a pair boundary between
thermoviscous acoustic domains.

All of the boundary conditions available with the Thermoviscous Acoustics, Transient
interface are the same as for the Frequency Domain version, and are described there
(see the following list). Some of the following nodes are available from the Mechanical
and Thermal submenus (listed in alphabetical order):

• Adiabatic1 • Normal Impedance1

• Axial Symmetry1 • No Slip1
• Background Acoustic Fields • No Stress1
• Boundary Stress1 • Periodic Condition1
• Heat Flux1 • Pressure (Adiabatic)1
• Heat Source1 • Slip1
• Initial Values1 • Symmetry1
• Interior Normal Impedance1 • Temperature Variation1
• Interior Temperature Variation1 • Thermoviscous Acoustics Model
• Interior Velocity1 • Velocity1
• Interior Wall1 • Wall1
• Isothermal1
• Nonlinear Thermoviscous Acoustics
Contributions
1
Described for the Thermoviscous Acoustics, Frequency Domain interface

554 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Thermoviscous Acoustics Model
Use the Thermoviscous Acoustics Model node to define the model inputs (the
background equilibrium temperature and pressure) and the material properties of the
fluid (equilibrium density, dynamic viscosity, bulk viscosity, thermal conductivity, and
heat capacity at constant pressure) necessary to model the transient propagation of
acoustic compressible waves in a thermoviscous acoustic context. Extended inputs are
available for the coefficient of thermal expansion and the compressibility, which
enables modeling of any constitutive relation for the fluid.

The equations solved are the linear (first order) perturbation equations for the
conservation of mass, momentum, and energy. The model also assumes a linear
equation of state relating acoustic density, pressure, and temperature fluctuations. The
governing equations are given by:

∂ρ
--------t + ∇ ⋅ ( ρ 0 u t ) = 0
∂t
∂u t
σ = – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I
2μ
ρ 0 --------- = ∇ ⋅ σ + F
∂t  3  (6-1)
∂T t ∂p t
ρ 0 C p  --------- + u t ⋅ ∇T 0 – T 0 α p  -------- + u t ⋅ ∇p 0 = – ∇ ⋅ ( – k ∇T t ) + Q
 ∂t   ∂t 
ρt = ρ0 ( βT pt – αp Tt )

the equations are formulated in terms of the total acoustic fields pt, ut, and Tt. If the
Background Acoustic Fields feature is added, for modeling scattering problems or as a
source, the total fields are the sum of the background acoustic fields and the scattered
acoustic fields. The equations can be extended to include nonlinear effects by adding
the Nonlinear Thermoviscous Acoustics Contributions.

The Model Inputs, the Thermoviscous Acoustics Model, and the Thermal Expansion and
Compressibility sections are the same as for the frequency domain interface. They are
described under the Thermoviscous Acoustics Model section in The Thermoviscous
Acoustics, Frequency Domain Interface.

Note that for the transient interface, it is equally important as for the frequency domain
interface, to set the material parameters correctly. The values of the coefficient of
thermal expansion and the isothermal compressibility are prone to errors, so make sure
to take a thorough look at the Thermal Expansion and Compressibility section.

THE THERMOVISCOUS ACOUSTICS, TRANSIENT INTERFACE | 555

 Nonlinear Thermoviscous Acoustics Contributions
The Nonlinear Thermoviscous Acoustics Contributions feature adds the necessary
contributions to the governing Equation 6-1 in order to model nonlinear effects in a
transient thermoviscous simulation. The contributions allow modeling of vortex
shedding that may happen at sudden expansions, like in a perforate, a grill, or at a
miniature sound port. Vortex shedding will in general introduce distortion in the
measured response of a system, with the generation of harmonics. The feature can also
capture the nonlinear effect associated with high sound pressure levels that require a
nonlinear representation of the equation of state (pressure, density, and temperature
relation). The nonlinear contributions to the left-hand side of Equation 6-1 are:

… + ∇ ⋅ ( ρt ut ) = …
∂u t
… + ρ t --------- + ρ 0 ( u t ⋅ ∇u t ) = …
∂t
∂T t
… + C p  ρ t --------- + ρ 0 ( u t ⋅ ∇T t ) + ρ t ( u t ⋅ ∇T 0 )
 ∂t 
∂p t
– α p  T t -------- + T 0 ( u t ⋅ ∇p t ) + T t ( u t ⋅ ∇p 0 ) = …
∂t

These represent all the second order nonlinear contributions, of the acoustic
perturbation fields, to the governing equations (keeping a linear equation of state).
This includes the nonlinear convective terms that are necessary to model vortex
shedding and other similar effects with detachment. The contributions are important
for modeling local high particle velocity situations, when the linearity condition
|ut| << c is no longer fulfilled. The contributions to the energy equation are not added
when the Adiabatic formulation is used.

Nonlinear Slit Resonator. Application Library path: Acoustics_Module/

Nonlinear_Acoustics/nonlinear_slit_resonator

For systems where the linearity condition on the density ρt << ρ0 is no longer fulfilled,
the density expansion can be changed to include second order terms (see below). This
will allow modeling of so-called cumulative effects.

The Nonlinear Thermoviscous Acoustics Contributions feature is not compatible with the
Background Acoustic Fields feature. The superposition principle is not valid in a
nonlinear model.

556 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Note that when solving nonlinear models, it is often necessary to use numerical
stabilization. Turn it on in the Stabilization section. Per default no stabilization is used.
Using stabilization also allows to switch to a P1-P1-P1 discretization which can be
more efficient in transient models. Remember to use an adequate mesh for a lower
order discretization, especially in the acoustic boundary layers.

MODEL INPUTS
The model inputs for the Equilibrium pressure p0 and the Equilibrium temperature T0
are always visible as they contribute to the governing equations.

DENSITY REPRESENTATION
In models with high local sound pressure levels, the linear equation of state may no
longer be valid. This happens when the linearity condition on the density ρt << ρ0 is
no longer fulfilled. If necessary, change the default Density expansion from First order
to Second order.

When Second order is selected, additional inputs to the model are necessary. For the
general case, the second order derivatives of the equilibrium density ρ0 = ρ0(p0,T0)
with respect to pressure p0 an temperature are necessary T0 are needed. They
contribute to the second-order Taylor expansion of the density. Per default, they are
taken From equilibrium density; this implies that the dependency of the density on
pressure and temperature should be correct.

If the Adiabatic formulation is used (see Thermoviscous Acoustics Equation Settings),

the user interface inputs correspond to the Nonlinear Acoustics (Westervelt)
Contributions equation in Pressure Acoustics, Transient. Select to specify the
Parameter of nonlinearity (default), the Ratio of specific heats (for gases), or the
Coefficient of nonlinearity.

VISCOUS HEATING
Click to select Include viscous dissipation (disabled per default). This will add a
right-hand side heat source to the energy equation. The viscous dissipation is a
nonlinear (second order) effect and can only be included in the nonlinear model.

Background Acoustic Fields

When the Background Acoustic Fields feature is added to a domain, it is possible to
define the value of the background acoustic field variables pb, ub, and Tb. Using this
feature, it is possible to set up scattering problems as well as defining acoustic fields at
an inlet of a waveguide (using a small domain at the inlet). The feature cannot be used

THE THERMOVISCOUS ACOUSTICS, TRANSIENT INTERFACE | 557

 together with the Nonlinear Thermoviscous Acoustics Contributions, as splitting into
background and scattered field is not possible for nonlinear equations.

BACKGROUND ACOUSTIC FIELDS

Enter expressions for the Background acoustic pressure pb, the Background acoustic
velocity ub, and the Background temperature variation Tb. These can be analytical
expressions or values of dependent variables solved in a previous study that defines the
background field. The time-domain version of the background acoustics field feature
does not include the plane wave option available for the frequency domain interface.

558 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

The Thermoviscous Acoustics,
Boundary Mode Interface
The Thermoviscous Acoustics, Boundary Mode (tabm) interface ( ), found under the
Thermoviscous Acoustics branch ( ) when adding a physics interface, is used to
compute and identify propagating and nonpropagating modes in waveguides and
ducts. The interface performs a boundary mode analysis on a given boundary,
including the thermal and viscous loss effects that are important in the acoustic
boundary layer near walls.

The interface is applied at boundaries which represent the cross section or the inlet of
a waveguide or duct of small dimensions. It solves for the acoustic variations of
pressure p, velocity u, and temperature T, as well as the out-of-plane wave number kn
of the modes. Near walls, viscous losses and thermal conduction become important
because boundary layers exists. The thickness of these layers is known as the viscous
and thermal penetration depth. For this reason, it is necessary to include thermal
conduction effects and viscous losses explicitly in the governing equations. The
Thermoviscous Acoustics, Boundary Mode interface is, for example, used when setting
up sources in systems with small ducts like hearing aids or mobile devices. It can also
be used to identify the propagating wave number and characteristic impedance of a
duct cross section and use that information in the homogenized Narrow Region
Acoustics model in The Pressure Acoustics, Frequency Domain Interface.

The propagation wave number is defined by the postprocessing variable

[Link] and the (lumped) characteristic impedance by the variable
[Link]. Both are global variables.

The Thermoviscous Acoustics, Boundary Mode interface solves the equations defined
by the linearized Navier-Stokes equations (linearized continuity, momentum, and
energy equations), in quiescent background conditions, on boundaries, searching for
the out-of-plane wave numbers at a given frequency. All gradients in the governing
equations are expressed in terms of the in-plane gradient and the out-of-plane wave
number that is being solved for. Due to the detailed description necessary when
modeling thermoviscous acoustics, the model simultaneously solves for the acoustic
pressure p, the velocity variation u (particle velocity), and the acoustic temperature

THE THERMOVISCOUS ACOUSTICS, BOUNDARY MODE INTERFACE | 559

 variations T. The interface is available on boundaries for 3D and on edges for 2D
axisymmetric geometries.

For details about the governing equations, see the Theory Background
for the Thermoviscous Acoustics Branch.

When this physics interface is added, these default nodes are also added to the Model
Builder — Thermoviscous Acoustics Model, Wall, and Initial Values. Then, from the
Physics toolbar, add other nodes that implement, for example, boundary conditions
and sources. You can also right-click Thermoviscous Acoustics, Boundary Mode to select
physics features from the context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is tabm.

EQUATION
Expand the Equation section to see the equations solved for with the Equation form
specified. The default selection is Equation form is set to Study controlled. The available
studies are selected under Show equations assuming.

• For Study controlled, the frequency used for the mode analysis study is given in the
study.
• For Mode analysis you can set the frequency manually. The default Mode analysis
frequency f is 100 Hz.

SOUND PRESSURE LEVEL SETTINGS

See the settings for Sound Pressure Level Settings for the Pressure Acoustics,
Frequency Domain interface.

560 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

DISCRETIZATION
From the list, select the element order for the Pressure, the Velocity field, and the
Temperature variation. The default uses Linear elements for the pressure and Quadratic
serendipity for the velocity field and the temperature variations.

In order for the system to be numerically stable, it is important that the order for the
pressure degree of freedom (DOF) is one lower than the velocity field. Per default, the
velocity components and the temperature share the same element order as they vary
similarly over the same length scale in the acoustic boundary layer. Therefore, both
typically require the same spatial accuracy. There is no stabilization option for the
boundary mode interface.

Increase the element order to, for example, quadratic for the pressure and
cubic for the velocity and temperature dofs. This can add additional
accuracy but it also adds additional degrees of freedom compared to the
default setting and will be computationally more expensive.

DEPENDENT VARIABLES
This physics interface defines these dependent variables (fields), the Pressure p, the
Velocity field u and its components, and the Temperature variation T. The names can
be changed but the names of fields and dependent variables must be unique within a
model.

• Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,

Frequency Domain Interface
• Theory Background for the Thermoviscous Acoustics Branch

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

THE THERMOVISCOUS ACOUSTICS, BOUNDARY MODE INTERFACE | 561

 Domain, Boundary, and Pair Nodes for the Thermoviscous Acoustics,
Boundary Mode Interface
The Thermoviscous Acoustics, Boundary Mode Interface has these boundary and pair
nodes available from the Physics ribbon toolbar (Windows users), Physics context menu
(Mac or Linux users), or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

The Continuity node with this physics interface is available as a pair

boundary condition. This gives continuity in pressure, temperature
variation, velocity and in the flux on a pair boundary between
thermoviscous acoustic domains.

All of the boundary conditions available with the Thermoviscous Acoustics, Boundary
Mode interface are the same as for the Frequency Domain and Transient version, and
are described there (see the following list). The Thermoviscous Acoustics Model

562 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

feature is described here. Some of the following nodes are available from the Mechanical
and Thermal submenus (listed in alphabetical order):

• Adiabatic1 • No Stress1
• Boundary Stress1 • Pressure (Adiabatic)1
• Heat Flux1 • Slip1
• Heat Source1 • Symmetry1
• Initial Values1 • Temperature Variation1
• Isothermal1 • Thermoviscous Acoustics Model
1
• Normal Impedance • Velocity1
• No Slip1 • Wall1
1Described for the Thermoviscous Acoustics, Frequency Domain interface

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Thermoviscous Acoustics Model

Use the Thermoviscous Acoustics Model node to define the model inputs (the
background equilibrium temperature and pressure) and the material properties of the
fluid (dynamic viscosity, bulk viscosity, thermal conductivity, heat capacity at constant
pressure, and equilibrium density) necessary to model the propagation of acoustic
compressible waves in a thermoviscous acoustic context. Extended inputs are available
for the coefficient of thermal expansion and the compressibility, which enables
modeling of any constitutive relation for the fluid.

The Model Inputs, the Thermoviscous Acoustics Model, and the Thermal Expansion and
Compressibility sections are the same as for the frequency domain and transient
interfaces. They are described under the Thermoviscous Acoustics Model section in
The Thermoviscous Acoustics, Frequency Domain Interface.

Note that for the boundary mode interface it is equally important as for the frequency
domain and transient interfaces to set the material parameters correctly. The values of
the coefficient of thermal expansion and the isothermal compressibility are prone to

THE THERMOVISCOUS ACOUSTICS, BOUNDARY MODE INTERFACE | 563

 errors, so make sure to take a thorough look at the Thermal Expansion and
Compressibility section.

BOUNDARY MODE SETTINGS

This section contains an input for the Propagation direction n of the modes analyzed
with the interface. Typically, this direction is equal to the surface or boundary normal.
The default value is ([Link], tabm,ny, tabm,nz) in 3D and ([Link], [Link]) in
2D axisymmetric. When setting up models, this vector defines the positive propagation
direction for the waves and thus the sign of the wave number.

564 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

The Acoustic-Thermoviscous Acoustic
Interaction, Frequency Domain
Interface
The Acoustic-Thermoviscous Acoustic Interaction, Frequency Domain multiphysics
interface ( ), found under the Thermoviscous Acoustics branch ( ) when adding a
physics interface, combines the Thermoviscous Acoustics, Frequency Domain and
Pressure Acoustics, Frequency Domain interfaces together with the
Acoustic-Thermoviscous Acoustic Boundary multiphysics coupling. To couple the two
physics and use Thermoviscous Acoustics only where necessary can reduce the
computational cost of a model.

The physics interface solves the equations in the frequency domain assuming all fields
and sources to be harmonic. Linear acoustics is assumed.

When a predefined Acoustic-Thermoviscous Acoustic Interaction, Frequency Domain

interface is added from the Acoustics>Thermoviscous Acoustics branch of the Model
Wizard or the Add Physics windows, the Thermoviscous Acoustics, Frequency Domain and
Pressure Acoustics, Frequency Domain interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling Acoustic-Thermoviscous Acoustic Boundary.

For details about all the multiphysics couplings in the Acoustics Module,
see Multiphysics Couplings.

The use of multiphysics couplings helps improve the flexibility and design
options for your modeling. For for details, see The Multiphysics Branch
and Multiphysics Modeling Workflow in the COMSOL Multiphysics
Reference Manual.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When a predefined multiphysics interface is used, for example Acoustic-Thermoviscous
Acoustic Interaction, Frequency Domain, the Selection on the multiphysics couplings is
automatically set to All boundaries. In this way, the multiphysics coupling is

THE ACOUSTIC-THERMOVISCOUS ACOUSTIC INTERACTION, FREQUENCY DOMAIN INTERFACE | 565

 automatically active on all boundaries with Thermoviscous Acoustics, Frequency Domain
on one side and Pressure Acoustics, Frequency Domain on the other.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL adds an empty Multiphysics
node. You can choose the available multiphysics couplings, but you need to manually
select on which boundaries they need to be applied, or select All boundaries to recover
the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics node) or from the Physics toolbar, Multiphysics menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
See Acoustic-Thermoviscous Acoustic Boundary in the Multiphysics Couplings
chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

566 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

 • The available physics features for The Thermoviscous Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, and Pair Nodes for the
Thermoviscous Acoustics, Frequency Domain Interface.
• The available physics features for The Pressure Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, Edge, Point, and Pair Nodes
for the Pressure Acoustics, Frequency Domain Interface

• Theory Background for the Pressure Acoustics Branch

• Theory Background for the Thermoviscous Acoustics Branch

THE ACOUSTIC-THERMOVISCOUS ACOUSTIC INTERACTION, FREQUENCY DOMAIN INTERFACE | 567

 The Thermoviscous Acoustic-Solid
Interaction, Frequency Domain
Interface
The Thermoviscous Acoustic-Solid Interaction, Frequency Domain multiphysics
interface ( ), found under the Thermoviscous Acoustics branch ( ) when adding a
physics interface, combines the Thermoviscous Acoustics, Frequency Domain and
Solid Mechanics interfaces. The physics interface solves for, and has predefined
couplings between, the displacement field in the solid and the acoustic variations in the
fluid domains.

The physics interface solves the equations in the frequency domain assuming all fields
and sources to be harmonic. Linear acoustics is assumed.

When a predefined Thermoviscous Acoustic-Solid Interaction, Frequency Domain interface

is added from the Acoustics>Thermoviscous Acoustics branch of the Model Wizard or the
Add Physics windows, the Thermoviscous Acoustics, Frequency Domain and Solid
Mechanics interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling Thermoviscous Acoustic-Structure Boundary.

For details about all the multiphysics couplings in the Acoustics Module,
see Multiphysics Couplings.

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

When a predefined multiphysics interface is used, for example Thermoviscous
Acoustic-Solid Interaction, Frequency Domain, the Selection on the multiphysics coupling
is automatically set to All boundaries. In this way, the multiphysics coupling is
automatically active on all boundaries with Thermoviscous Acoustics, Frequency Domain
on one side and Solid Mechanics on the other.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

568 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

For example, if the single interfaces are added, COMSOL Multiphysics adds an empty
Multiphysics node. You can choose the available multiphysics couplings, but you need
to manually select on which boundaries they need to be applied, or select All boundaries
to recover the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics node) or from the Physics toolbar, Multiphysics menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
See Thermoviscous Acoustic-Structure Boundary in the Multiphysics Couplings
chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• The available physics features for The Thermoviscous Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, and Pair Nodes for the
Thermoviscous Acoustics, Frequency Domain Interface.

• Theory Background for the Thermoviscous Acoustics Branch

• Regarding the available structural-mechanics features and their theory background,

see The Solid Mechanics Interface.

THE THERMOVISCOUS ACOUSTIC-SOLID INTERACTION, FREQUENCY DOMAIN INTERFACE | 569

 The Thermoviscous Acoustic-Shell
Interaction, Frequency Domain
Interface
This physics interface requires a Structural Mechanics Module license.
For theory and physics interface feature descriptions relating to the Shell
interface, see the Structural Mechanics Module User’s Guide.

The Thermoviscous Acoustic-Shell Interaction, Frequency Domain multiphysics interface

( ), found under the Acoustics>Thermoviscous Acoustics branch ( ) when adding
a physics interface, combines the Thermoviscous Acoustics, Frequency Domain and
Shell interfaces. The physics interface solves for and has a predefined coupling between
the displacement field of the shell and the acoustic variations in the fluid domains. It
can be used, for example, for modeling the vibrating response of micromirrors in
MEMS applications. The physics interface is available for 3D and 2D axisymmetric
geometries, and it is capable of modeling the coupled thermoviscous acoustics and
shell vibrations in the frequency domain.

The physics interface solves the equations in the frequency domain assuming all fields
and sources to be harmonic. Linear acoustics is assumed.

When a predefined Thermoviscous Acoustic-Shell Interaction, Frequency Domain interface

is added from the Acoustics>Thermoviscous Acoustics branch of the Model Wizard or the
Add Physics windows, the Thermoviscous Acoustics, Frequency Domain and Shell
interfaces are added to the Model Builder.

In addition, the Multiphysics node is added, which automatically includes the

multiphysics coupling Thermoviscous Acoustic-Structure Boundary. This multiphysics
coupling automatically couples the thermoviscous acoustic domain to exterior shells
(on exterior boundaries) and to interior shells (on interior boundaries).

For details about all the multiphysics couplings in the Acoustics Module,
see Multiphysics Couplings.

570 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

SETTINGS FOR PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS
When a predefined multiphysics interface is used, for example Thermoviscous
Acoustic-Shell Interaction, Frequency Domain, the Selection on the multiphysics coupling
is automatically set to All boundaries. In this way, the multiphysics coupling is
automatically active on all boundaries with Thermoviscous Acoustics, Frequency Domain
on one side (exterior shells) or on both sides (interior shells) of Shell boundaries.

However, if physics interfaces are added one at a time, followed by the multiphysics
couplings, these modified settings are not automatically included.

For example, if the single interfaces are added, COMSOL adds an empty Multiphysics
node. You can choose the available multiphysics couplings, but you need to manually
select on which boundaries they need to be applied, or select All boundaries to recover
the predefined behavior.

Multiphysics couplings are available from the context menu (right-click

the Multiphysics node) or from the Physics toolbar, Multiphysics menu.

PHYSICS INTERFACES AND MULTIPHYSICS COUPLINGS

Use the online help in COMSOL Multiphysics to locate and search all the
documentation. All these links also work directly in COMSOL
Multiphysics when using the Help system.

Multiphysics Coupling
See Thermoviscous Acoustic-Structure Boundary in the Multiphysics Couplings
chapter.

Physics Interface Features

Physics nodes are available from the Physics ribbon toolbar (Windows users), Physics
context menu (Mac or Linux users), or right-click to access the context menu (all
users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE THERMOVISCOUS ACOUSTIC-SHELL INTERACTION, FREQUENCY DOMAIN INTERFACE | 571

 • The available physics features for The Thermoviscous Acoustics, Frequency Domain
Interface are listed in the section Domain, Boundary, and Pair Nodes for the
Thermoviscous Acoustics, Frequency Domain Interface.
• The available physics features for Results Evaluation are listed in the section
Domain, Boundary, Edge, Point, and Pair Nodes for the Shell and Plate Interfaces
in the Structural Mechanics Module User’s Guide

• Theory Background for the Thermoviscous Acoustics Branch

• Theory for the Shell and Plate Interfaces in the Structural Mechanics
Module User’s Guide

572 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Modeling with the Thermoviscous
Ac o us t i c s Br a nch
This section contains modeling tips and tricks, information about solvers,
postprocessing variables, and meshing.

In this section:

• Meshing the Boundary Layer

• Solver Suggestions for Large Thermoviscous Acoustics Models
• Lagrange and Serendipity Shape Functions
• Transient Solver Settings
• Postprocessing Variables
• Suppressing Constraints on Lower Dimensions

Meshing the Boundary Layer

When modeling using the Thermoviscous Acoustics interfaces, several length scales
become important when setting up the mesh.

• First, there is the wavelength which should still be resolved as in pressure acoustics,
see Meshing: Resolving the Waves in Space in the Pressure Acoustics Interfaces
chapter. This length scale is often resolved as the geometries are small when
modeling thermoviscous acoustic phenomena in microacoustic applications.
• Secondly, there is the thickness of the viscous and thermal boundary layers. In order
for the model to include the correct amount of damping, the boundary layers need
to be resolved. Ideally this is done using a Boundary Layers mesh. The Thickness of
first boundary layer and the Number of boundary layers should be set such that they
resolve the boundary layer at the specific modeling frequency. Remember that the
boundary layer thickness scales as one over the square root of the frequency.

MODELING WITH THE THERMOVISCOUS ACOUSTICS BRANCH | 573

 A great deal of information about meshing and modeling with the
Thermoviscous acoustics interface can be found in two COMSOL Blog
posts. They are:

• [Link]/blogs/
theory-thermoacoustics-acoustics-thermal-viscous-losses/
• [Link]/blogs/
modeling-thermoacoustic-interface-comsol/

See the Boundary Layers section in the COMSOL Multiphysics Reference

Manual for more details.

• Finally, it is important to consider the thickness of the boundary layer compared to

the physical dimensions of the model, for example, the viscous boundary layer
thickness δv compared to the tube radius a in a waveguide. This is sometimes known
as the Womersley number

2
ωρa a 2 a 2
-------------- = ----------- = ------- ------
-
Wo =
μ δv δ th Pr

where δv is the viscous boundary layer thickness, δth is the thermal boundary layer
thickness, and Pr is the Prandtl number.
When the boundary layer is small compared to the geometry, the Womersley
number is large, say Wo > 10. Then the effect associated with the losses in the
viscous boundary layer can normally be disregarded. In this case, the boundary layer
do not need to be meshed and a Slip condition can be used instead of a No-slip
condition. The same is true for the thermal boundary layer thickness compared to
the tube radius. Here, an Isothermal condition can be replaced by an Adiabatic
condition. If the No-slip or Isothermal conditions are still kept, then remember to
add at least one boundary layer mesh that is of roughly the size of the acoustic
boundary layer. If this is not done, erroneous losses can be introduced in the model.
On the other hand whenever Wo is O(1) or smaller, the boundary layer effects need
to be included and the boundary layer needs to be meshed adequately as mentioned
above.

574 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Solver Suggestions for Large Thermoviscous Acoustics Models
Solving thermoviscous acoustic problems can easily involve solving for many degrees
of freedom (DOFs) as the model solves for the acoustic variations in both pressure,
velocity field (2 or 3 components), and temperature. First of all, it is important to
restrict the use of the thermoviscous acoustics model to domains and regions of the
model where it is necessary. Couple to Pressure Acoustics using the multiphysics
coupling. In the frequency domain use the Narrow Region Acoustics feature or the
Thermoviscous Boundary Layer Impedance feature of Pressure Acoustics to reduce
the number of DOFs. Secondly, care should be taken when meshing the computational
domain. If these two things have been carefully considered, the solver can be changed
from its default Direct setting to use an Iterative solver suggestion. Depending on the
model size and involved physics two options are described below.

In both cases, a good starting point for setting up a new solver configuration is to
right-click the study node and select Show Default Solver, then expand the Solver
Configuration tree under Stationary Solver or Time-Dependent Solver. Predefined iterative
solver suggestions are automatically generated. Per default, a direct solver is used and
two iterative solvers are suggested and disabled. To turn on one of these approaches,
right-click the solver and select Enable (or press F4).

• The first suggestion (GMRES with Direct Precon.) uses an iterative solver with a
direct preconditioner. This method is typically faster than the direct solver and uses
20% less memory.
• The second suggestion (GMRES with DD) uses an iterative solver with the domain
decomposition method. This method is very robust (also for multiphysics
applications) and very memory efficient, but it can be slow.

Both suggestions are described below as well as how to set them up manually (in most
cases the default suggestions should be used). In liquids where thermal effects can be
neglected, the model can be solved in the Adiabatic formulation and DOFs saved.
Choosing different shape functions can also reduce the memory consumption, for
example, switching to the default P1-P2-P2 discretization to an all linear discretization
if Stabilization is enabled.

• Iterative Solver with Direct Preconditioner using Hybridization

• Iterative Solver with Domain Decomposition
• Solving in the Adiabatic Case
• Choosing Shape Functions

MODELING WITH THE THERMOVISCOUS ACOUSTICS BRANCH | 575

 ITERATIVE SOLVER WITH DIRECT PRECONDITIONER USING
HYBRIDIZATION
For large 2D problems and 3D problems, that only involve thermoviscous acoustics,
using the following approach will save around 20% memory and can speed up the
solution procedure by a factor 2 or 3. Under the Stationary Solver node take the
following steps: Add an Iterative solver with the GMRES solver. As preconditioner add
the Direct Preconditioner and switch the solver to PARDISO. Expand the Hybridization
section and select Multi preconditioner in the Preconditioner variables list add the
Pressure and Velocity field. Add a second direct preconditioner with the same settings
but now select only the Temperature as preconditioner variables. Set the Pivoting
Perturbation to 1e-13 for the pressure-velocity group. The reason for splitting the
equations up in this manner is that the energy equation is only loosely coupled to the
momentum and continuity equations.

See the Direct Preconditioner section in the COMSOL Multiphysics

Reference Manual for more details.

The direct preconditioner approach is used in the model: Transfer

Impedance of a Perforate . Application Library path Acoustics_Module/
Tutorials/transfer_impedance_perforate

ITERATIVE SOLVER WITH DOMAIN DECOMPOSITION

A more advanced approach, to handle very large 3D models, is again to use the
GMRES iterative solver but now with the domain decomposition preconditioner. This
approach can also be used for multiphysics problems involving several physics.

Start by adding an Iterative solver and select GMRES as the solver. Then right-click the
iterative node and select Domain Decomposition (Schwarz). A good starting point for this
solver, is to use the default settings with only a few changes:

• If you set Recompute and clear subdomain data to On you will get a very memory lean
solver (but the solution may take longer). Recompute and clear subdomain data is a
list with the options Automatic, Off, and On. If the option Automatic is chosen, the
recompute and clear mechanism is activated if there is an out-of-memory error
during the domain decomposition setup phase. The setup is then repeated with
recompute and clear activated (which can be costly in terms of time but is better
than failure). A warning is given in this case.
• In the Coarse Level section change the Use coarse level to Aggregation.

576 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

• In the direct solver under the Coarse Solver and Domain Solvers subnodes use the
PARDISO solver and set the Pivoting perturbation to 1e-13.
This type of approach should make it possible to solve large thermoviscous acoustics
models, also including multiphysics interactions, using a minimum of RAM. Possibly
increase the value of the Maximum number of DOFs per subdomain option to use a larger
amount of the RAM at your disposition.

See the Domain Decomposition (Schwarz) section in the COMSOL

Multiphysics Reference Manual for more details

SOLVING IN THE ADIABATIC CASE

In certain cases, it is a good assumption to not include thermal conduction in the
model and treat all processes as adiabatic (isentropic). This is, for example, relevant for
fluids where the thermal boundary layer is much thinner than the viscous, like in water.
Not solving for the temperature field T also saves some degrees of freedom (DOFs).

This is achieved by selecting the Adiabatic formulation option under the Thermoviscous
Acoustics Equation Settings section. When Adiabatic formulation is selected, all
temperature options and conditions are disabled in the user interface.

CHOOSING SHAPE FUNCTIONS

In models with a structured mesh, it can be advantageous to switch to the serendipity
shape functions instead of the default Lagrange shape functions; see Lagrange and
Serendipity Shape Functions below. In general, if a boundary layer mesh is used (to
resolve the thermal and viscous boundary layers) or if a PML is used in the model, the
mesh contains structured mesh regions.

Lagrange and Serendipity Shape Functions

In most of the physics interfaces in the Acoustics Module, and specifically in the
Thermoviscous Acoustic interfaces, you can choose between two families of shape
functions in the Discretization section: Lagrange and serendipity. The current default
is to use serendipity shape functions.

When using a structured mesh, it is advantageous to use the serendipity elements as

they generate significantly fewer degrees of freedom (DOFs). The accuracy is in most
cases almost as good as for the Lagrange elements. The Lagrange elements are,
however, less sensitive to strong mesh distortions.

MODELING WITH THE THERMOVISCOUS ACOUSTICS BRANCH | 577

 The serendipity shape functions differ from the Lagrange shape functions only for the
following element shapes:

• 2D: Quadrilateral elements of discretization order higher than 1.

• 3D: Hexahedral, prism, and pyramid elements of discretization order higher than 1.

In the COMSOL Multiphysics Reference Manual:

• The Lagrange Element (shlag)

• The Nodal Serendipity Element (shnserp)

When coupling two physics interfaces that have the same DOFs like, for example,
displacement, the same type of shape functions should be used in both interfaces to
ensure conformity. Since there is no difference between the two families of shape
functions in 1D, this is not an issue when connecting edges.

Transient Solver Settings

When solving transient wave problems, it is recommended to use fixed time stepping
in the solver (the manual option in the solver configuration). In general, there is no
point in using an automatic time-step control, which can be provided by the
time-dependent solver. The tolerances in the automatic error control are difficult to
tune in wave problems when there is weak but important high-frequency content. It
is recommended to use the suggestion generated when the Fixed (preferred) method is
chosen (the default) in the Transient Solver Settings section. The internal time step of
the solver that is generated, when the Maximal frequency to resolve is set, will in most
cases produce a solver that is adequate. It is assuming that the user has generated a
mesh that properly resolves the same maximal frequency (minimal wavelength).

For further details see the Time Stepping: Resolving the Waves in Time
section in Modeling with the Thermoviscous Acoustics Branch.

Postprocessing Variables
This section contains information about variables for:

• Viscous and Thermal Boundary Layer Variables

• Material Properties

578 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

• Stress Tensor Components
• Power Dissipation and Intensity Variables
• In and Out-of-Plane Variables for the Boundary Mode Interface

In the variable names, in the tables below, phys_id represents the

interface name. For example, ta for the frequency domain interface, tatd
for the transient interface, and tabm for the boundary mode interface.

VISCOUS AND THERMAL BOUNDARY LAYER VARIABLES

The viscous and thermal boundary layer thickness (thermal and viscous penetration
depth) as defined in Theory Background for the Thermoviscous Acoustics Branch can
be evaluated in postprocessing for the frequency-domain models. The same is the case
for the Prandtl number relating the two length scales (available in both frequency and
time domain).
TABLE 6-2: BOUNDARY LAYER VARIABLES

VARIABLE DESCRIPTION

phys_id.d_visc Viscous boundary layer thickness

phys_id.d_therm Thermal boundary layer thickness
phys_id.Pr Prandtl number

MATERIAL PROPERTIES
Material properties are readily available for postprocessing in plots. In plots, click the
Replace Expression icon and browse to the Material properties list under the
thermoviscous acoustics interface. Important parameters to plot are the coefficient of
thermal expansion phys_id.alpha0 and the isothermal compressibility
phys_id.betaT. These should not evaluate to zero.

STRESS TENSOR COMPONENTS

The stress tensor components are defined as variables and can be evaluated in
postprocessing or used to create exotic couplings between physics. In the table below
only the xx component of the stress tensor and the x component of the stress are
shown. Change the spatial reference accordingly,
TABLE 6-3: STRESS VARIABLES ONE COMPONENT IN 3D

VARIABLE DESCRIPTION

phys_id.K_stress_tensorxx Viscous stress tensor, xx-component

phys_id.T_stress_tensorxx Total stress tensor, xx-component

MODELING WITH THE THERMOVISCOUS ACOUSTICS BRANCH | 579

 TABLE 6-3: STRESS VARIABLES ONE COMPONENT IN 3D

VARIABLE DESCRIPTION

phys_id.K_stressx Viscous stress, x-component (on boundaries)

phys_id.T_stressx Total stress, x-component (on boundaries)

POWER DISSIPATION AND INTENSITY VARIABLES

The thermoviscous acoustics interface models the energy dissipation process which
stem from viscous and thermal dissipation processes. The amount of dissipated energy
can be of interest as a results analysis variable or as a source term for a multiphysics
problem.

The energy conservation-dissipation corollary describes the transport and dissipation

of energy in a system (see Ref. 1 p. 516 or Ref. 6). In linear acoustics, this equation is
derived by taking the dot product (scalar product) of the momentum and the velocity
v, adding it to the continuity equation, and then adding the entropy. After some
manipulation and integration, the use of the divergence theorem yields Equation 6-2

∂ w dV ∂w

∂t
+  ( i ⋅ n ) dA = –  Δ dV or -------
∂t
+ ∇ ⋅ i = –Δ
Ω ∂Ω Ω

1 ρ0 T0 2
2
1 2 1 p
w = --- ρ 0 u + --- -----------2 + --- ------------- s
2 2ρ c 2 Cp
0 0 (6-2)
T k T
i = pu – [ u ⋅ τ ] – ------ T ∇T
T0
k 2
Δ = τ:∇u + ------ ( ∇T ) = Δ v + Δ t
T0

where w is the disturbance energy of the control volume, u = |u| is the velocity, T is
the temperature variation, p is the acoustic pressure variations, p0 is the background
equilibrium pressure, T0 the background equilibrium temperature, ρ0 the background
density, c0 the (isentropic) speed of sound, Cp the heat capacity at constant pressure
(per unit mass), k the coefficient of thermal conduction, i is the instantaneous intensity
(flux of energy out of a control volume), Δ is the dissipated energy per unit volume
and time (SI unit: Pa/s = J/(m3s) = W/m3), s is the entropy, τ is the viscous stress
tensor, τ:∇u is the viscous dissipation function, and T indicates transpose of vector.
Δv and Δt are the viscous and thermal contributions to the dissipation function. In
Equation 6-2 we have made use of Ref. 6 for the expression for the intensity I.

In the Thermoviscous Acoustics, Frequency Domain interface, the dissipation term Δ

is directly given by the RMS value of the tensor expression

580 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

 1
Δ v =  τ:∇u = --- ( τ∗ :∇u + τ: ( ∇u )∗ ) (6-3)
4

where “:” in Equation 6-3 is the double dot operator (or total inner product) and *
is the complex conjugate. In the above expressions, the time-averaged expressions for
a product in the frequency domain is defined as:

iωt iωt 1
 AB =  Re ( Ae )Re ( Be ) = --- ( A∗ B + AB∗ )
4

The power dissipation variables are defined in Table 6-4.

TABLE 6-4: POWER DISSIPATION VARIABLES

VARIABLE DESCRIPTION

phys_id.diss_therm Thermal power dissipation density

phys_id.diss_visc Viscous power dissipation density
phys_id.diss_tot Total thermal and viscous power dissipation density

In the Thermoviscous Acoustics, Frequency Domain interface the (time averaged)

intensity I is given by averaging the instantaneous intensity vector i in Equation 6-2
using the same time averaged products defined above. The intensity and intensity
magnitude are defined in Table 6-5.
TABLE 6-5: INTENSITY VARIABLES

VARIABLE DESCRIPTION

phys_id.Ix Intensity x-component (in 1D, 2D, and 3D)

phys_id.Iy Intensity y-component (in 2D, and 3D)
phys_id.Iz Intensity z-component (2D axisymmetric and 3D)
phys_id.Ir Intensity r-component (2D axisymmetric)
phys_id.I_mag Intensity magnitude

In the Thermoviscous Acoustics, Transient interface the instantaneous intensity i

variables is available for postprocessing. The instantaneous intensity and instantaneous
intensity magnitude are defined in Table 6-6.
TABLE 6-6: INSTANTANEOUS INTENSITY VARIABLES

VARIABLE DESCRIPTION

phys_id.Iix Instantaneous intensity x-component (in 1D, 2D, and 3D)

phys_id.Iiy Instantaneous intensity y-component (in 2D, and 3D)
phys_id.Iiz Instantaneous intensity z-component (2D axisymmetric and 3D)

MODELING WITH THE THERMOVISCOUS ACOUSTICS BRANCH | 581

 TABLE 6-6: INSTANTANEOUS INTENSITY VARIABLES

VARIABLE DESCRIPTION

phys_id.Iir Instantaneous intensity r-component (2D axisymmetric)

phys_id.Ii_mag Instantaneous intensity magnitude

IN AND OUT-OF-PLANE VARIABLES FOR THE BOUNDARY MODE

INTERFACE
Several dedicated variables exist for The Thermoviscous Acoustics, Boundary Mode
Interface where quantities are defined in terms of their in-plane and out-of-plane
values. For example, the intensity variable I has the following derived values

I ip = I – ( I ⋅ n )n
I op = ( I ⋅ n )n

where ip stands for in-plane and op for out-of-plane. These two variables are named
[Link] and [Link] (with spatial components x, y, and z). The magnitude of
these two variables is given by tabm.Iip_mag and tabm.Iop_mag. In the same manner
variables exist for the acceleration and the velocity.

Suppressing Constraints on Lower Dimensions

Sometimes, boundary conditions on two adjacent objects can come into conflict on a
shared object (point or edge). For the boundary conditions in the thermoviscous
physics that are based on constraints, you have the possibility to select that certain
objects of lower dimensions should be excluded from the main selection. To do this,
you must first click the Show More Options button ( ) and select Advanced Physics
Options in the Show More Options dialog box. In the settings, a new section named
Excluded Edges and/or Excluded Points will then appear (when applicable). In these
sections, you can select geometrical objects which should be excluded from the main
selection when the constraint is applied.

Velocity Condition
When using the mechanical Velocity condition to you may get conflicts if it is located
next to a no-slip condition (a wall). To remedy this, you can exclude the common edge
or point.

Periodic Condition
When using the Periodic Condition, you may get conflicts or duplicates which makes
the model either overconstrained or problematic for the automatic constraint
elimination algorithm. If you are aware of such situations, it is good practice to remove

582 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

one of the potentially conflicting constraints. This could be in multiphysics models
where you combine solid mechanics with thermoviscous acoustics and have edges
where you both apply a periodic condition (solid and thermoviscous acoustics) and the
Thermoviscous Acoustic-Structure Boundary multiphysics coupling.

Particularly if the geometry is curved, there is a risk that these constraints are not
identical from a numerical point of view. In this case, excluding the conflicting edge
from the selected boundary will make the behavior unique and fully predictable.

Another example where constraints will come in conflict is if you want to, for example,
use periodic condition on parts of the geometry using weak constraints, while keeping
the default pointwise constraints on other parts. If the same mesh node has both types
of constraints, the solution will fail, so you must exclude any common geometrical
objects from the selection in one of the constraints.

See also Excluded Surfaces, Excluded Edges, and Excluded Points in the
COMSOL Multiphysics Reference Manual.

MODELING WITH THE THERMOVISCOUS ACOUSTICS BRANCH | 583

 T he o r y B a c k g rou n d for t h e
Thermoviscous Acoustics Branch
The Thermoviscous Acoustics, Frequency Domain Interface, The Thermoviscous
Acoustics, Transient Interface, and The Thermoviscous Acoustics, Boundary Mode
Interface are designed for the analysis of acoustics in viscous and thermally conducting,
compressible Newtonian fluids. The physics interface solves the linearized Navier–
Stokes equation, the continuity equation, and the energy equation. This corresponds
to a small parameter expansion of the dependent variables. The physics interface solves
for the acoustic pressure variations p, the fluid velocity variations u, and the acoustic
temperature variations T. The interface uses a scattered field formulation where it is
possible to define background acoustic fields. All equations and boundary conditions
are formulated in the total fields and solve for the scattered fields.

The Thermoviscous Acoustics, Frequency Domain Interface is available for 3D, 2D,
and 1D Cartesian geometries as well as for 2D and 1D axisymmetric geometries. The
physics interface solves problems in the frequency domain, that is, Frequency Domain,
Frequency-Domain Modal, and Eigenfrequency type analysis. In 2D and 1D
axisymmetric systems, a Mode Analysis study is also available for the out-of-plane
component.

The Thermoviscous Acoustics, Transient Interface is available for 3D, 2D, and 1D
Cartesian geometries as well as for 2D and 1D axisymmetric geometries. The physics
interface solves problems in the time domain using the Time Dependent analysis type.

The Thermoviscous Acoustics, Boundary Mode Interface is available on boundaries in

3D and 2D axisymmetric geometries. The physics interface helps identify and solve for
propagating and nonpropagating modes using the Mode Analysis study.

In this section:

• The Viscous and Thermal Boundary Layers

• General Linearized Compressible Flow Equations
• Acoustic Perturbation and Linearization
• Scattered Field Formulation and Background Acoustic Fields
• Formulation for Eigenfrequency Studies
• Formulation for Mode Analysis in 2D and 1D Axisymmetry

584 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

• Formulation for the Boundary Mode Interface
• References for the Thermoviscous Acoustics, Frequency Domain Interface

The Viscous and Thermal Boundary Layers

In general, a tangential harmonic oscillation of amplitude u0 and frequency f applied
to a wall at z = 0 creates a viscous wave of the form

πfρ
– -----------0 ( 1 + i )z
u ( z ) = u0 e μ

where, f is the frequency, ρ0 is the static density, and μ is the dynamic viscosity. The
viscous shear waves are therefore dispersive with wavelength Lv and the associated
wavenumber kv

μ - = 2πδ 2 iωρ 0
L v = 2π ----------- k v = – ------------
πρ 0 f v μ

The viscous waves are highly damped since their amplitude decays exponentially with
distance from the boundary (see Ref. 3). In fact, in just one wavelength, the amplitude
decreases to about 1/500 of its value at the boundary. Therefore, the viscous boundary
layer thickness can for most purposes be considered to be less than Lv. The length scale
δv is the so-called viscous penetration depth or viscous boundary layer thickness.

Similarly, a harmonically oscillating temperature with amplitude T0 and frequency f at

z = 0 creates a thermal wave of the form

πfρ 0 C p
- ( 1 + i )z
– -----------------
T ( z ) = T0 e k

where Cp is the (specific) heat capacity at constant pressure and k is the thermal
conductivity. The wavelength and wavenumber is here

k 2 iωρ 0 C p
L th = 2π ------------------- = 2πδ th k th = – --------------------
πρ 0 fC p k

and a decay behavior similar to the viscous waves. The length scale δth is the thermal
penetration depth.

The ratio of viscous wavelength to thermal wavelength is a nondimensional number

related to the Prandtl number Pr, as

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 585

 L μC p
-------v- = ----------- = Pr
L th k

In air, this ratio is roughly 0.8, while in water, it is closer to 2.7. Thus, at least in these
important cases, the viscous and thermal boundary layers are of the same order of
magnitude. Therefore, if one effect is important for a particular geometry, so is
probably the other.

General Linearized Compressible Flow Equations

In general, the motion of a viscous compressible Newtonian fluid, including the energy
equation, is governed by the set of equations listed in Equation 6-4 below.

dρ
+ ρ(∇ ⋅ u) = 0
dt
du
ρ = ∇⋅σ+F
dt
dT dp
ρC p – αp T = –∇⋅q+φ+Q
dt dt (6-4)

σ = – pI + τ = – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I
2μ
 3 
q = – k ∇T
ρ = ρ ( p, T )

where the dependent variables are pressure p, velocity u, temperature T, and density
ρ. The first three equations are the continuity equation, the momentum equation (the
Navier–Stokes equation), and the energy equation, respectively. The last three
equation are the constitutive equations. They define the total stress tensor σ and the
viscous stress tensor τ through Stokes expression, the Fourier heat conduction law, and
an equation of state. See, for example, Ref. 1 to 7 for further details.

The material time derivatives (or total derivatives) d/dt are in the following expanded
according to

dA ( x, t )- ∂A
-------------------- = ------- + u ⋅ ∇A
dt ∂t

where A is a dummy variable. The equation of state relates local values of pressure,
density, and temperature and is therefore an algebraic equation or an ODE, rather than

586 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

a PDE. A common form of state equations is to know the density as function of
pressure and temperature, ρ = ρ(p, T). In the following, it is assumed that the state
equation has this form.

The basic properties of the fluid are the dynamic viscosity μ and thermal conductivity
k. The coefficient μB is the bulk (or second) viscosity and describes losses due to
compressibility (expansion and contraction of the fluid), where μ describes losses due
to shear friction. The bulk viscosity can in some cases be used to model an empirically
observed deviation from Stokes’ assumption but is usually negligible compared to μ
unless the motion is irrotational; see Ref. 3 and Ref. 4. These three properties are
taken to be constant or at most weakly temperature-dependent. The specific heat at
constant pressure Cp (per unit mass) and the (isobaric) coefficient of volumetric
thermal expansion αp (sometimes denoted α0),

1 ∂ρ
α p = – ---
ρ ∂T p

are both possibly functions of pressure and temperature.

In the energy equation

φ = τ ( u ):S ( u ) = τ ( u ): ∇u

is the viscous dissipation function — that is, the scalar contraction of the viscous stress
tensor τ with the rate of strain tensor S. Both tensors are seen as functions of a velocity
vector. If the mean velocity is zero, this term vanishes in the following linearization
because it is homogeneous of second order in the velocity gradients. Otherwise, it acts
as an oscillating source or sink (the term is included in the full Linearized
Navier-Stokes interfaces). In the right sides of Equation 6-4, F and Q are a volume
force and a heat source, respectively.

Acoustic Perturbation and Linearization

For small perturbations around steady-state solution, the dependent variables and
sources can be assumed to take on the following form:

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 587

 u = u 0 ( x ) + u 1 ( t, x )
p = p 0 ( x ) + p 1 ( t, x )
T = T 0 ( x ) + T 1 ( t, x )
ρ = ρ 0 ( x ) + ρ 1 ( t, x )
F = F 0 ( x ) + F 1 ( t, x )
Q = Q 0 ( x ) + Q 1 ( t, x )

where the subscript “1” variables represent the acoustic perturbations (first-order
perturbation) and subscript “0” the background mean flow quantities. Assuming zero
mean flow u0 = 0 and after inserting into the governing Equation 6-4, the steady-state
equations can be subtracted from the system, which is subsequently linearized to first
order by ignoring terms quadratic in the acoustic variables. Dropping the subscript “1”
for readability yields the thermoviscous acoustic equations:

∂ρ-
----- + ∇ ⋅ ( ρ0 u ) = 0
∂t
∂u
ρ 0 ------- = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I + F
2μ
(6-5)
∂t 3
∂T ∂p
ρ 0 C p  ------- + u ⋅ ∇T 0 – T 0 α p  ------ + u ⋅ ∇p 0 = – ∇ ⋅ ( – k ∇T ) + Q
 ∂t   ∂t 

where the unprimed variables are now the acoustic deviation from the steady state.

The density ρ is expressed in terms of the pressure and the temperature variations using
the density differential (Taylor expansion about the steady quiescent values)

∂ρ ∂ρ
ρ = p ------ + T ------- = ρ 0 ( pβ T – Tα p ) (6-6)
∂p T0 ∂T p0

The two thermodynamic quantities (the coefficient terms in square brackets) define
the isobaric coefficient of thermal expansion αp (sometimes named α0) and the
isothermal compressibility βT, according to the following relations

1 ∂ρ 0 1 1 γ
β T = ------ --------- = ------- = ------ ----2- = γβ s
ρ 0 ∂p T KT ρ0 c
Cp Ks
γ = ------- = ------- (6-7)
Cv KT

1 ∂ρ 0
α p = – ------ ---------
ρ 0 ∂T p

588 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

where Ks is the isentropic bulk modulus (sometimes named K0), KT the isothermal
bulk modulus, Cv is the heat capacity at constant volume (per unit mass), c is the
(isentropic) speed of sound, and γ is the ratio of specific heats (the adiabatic index).
The isothermal compressibility βT is related to the isentropic (or adiabatic)
compressibility βs (sometimes named β0) and the coefficient of thermal expansion αp
via the thermodynamic relations

2
αp T0
β s = β T – ------------- β T = γβ s (6-8)
ρ0 Cp

It is derived using the Maxwell relations; see, for example, Ref. 5 and Ref. 7.

From Equation 6-7 and Equation 6-8 the isothermal compressibility and the isobaric
coefficient of thermal expansion can be expressed in terms of the speed of sound as

1 γ 1 Cp ( γ – 1 )
β T = ------ ----2- α p = --- -----------------------
- (6-9)
ρ0 c c T0

The equations presented in Equation 6-5 and Equation 6-6 are the ones solved in the
time domain in The Thermoviscous Acoustics, Transient Interface. Assuming small
harmonic oscillations about a steady-state solution, the dependent variables and
sources can be assumed to take on the following form

u = u 0 + u 1 e iωt p = p 0 + p 1 e iωt
T = T 0 + T 1 e iωt ρ = ρ 0 + ρ 1 e iωt
F = F 0 + F 1 e iωt Q = Q 0 + Q 1 e iωt

Inserting this into the governing equations and performing the linearization yield the
equations solved in the frequency domain in The Thermoviscous Acoustics, Frequency
Domain Interface:

iωρ + ∇ ⋅ ( ρ 0 u ) = 0

iωρ 0 u = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I + F
2μ
 3  (6-10)
ρ 0 C p ( iωT + u ⋅ ∇T 0 ) – T 0 α p ( iωp + u ⋅ ∇p 0 ) = – ∇ ⋅ ( – k ∇T ) + Q
ρ = ρ 0 ( pβ T – Tα p )

The system of equations implemented in the thermoviscous acoustics interfaces is

further given in a scattered field formulation, as described below in Scattered Field
Formulation and Background Acoustic Fields.

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 589

 When the Adiabatic formulation option under the Thermoviscous Acoustics Equation
Settings section the solved equations reduce to:

iωρ + ∇ ⋅ ( ρ 0 u ) = 0

iωρ 0 u = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I + F
2μ
 3  (6-11)
αp T0
ρ = ρ0 βs p T = ------------- p
ρ0 Cp

where the temperature T depends directly on the acoustic pressure p.

IDEAL GAS
For an ideal gas, the equation of state p = ρRT, where R is the specific gas constant,
leads to

1 1
β T = ------ α 0 = ------
p0 T0

and the density

ρ = ρ 0  ------ – ------
p T
 p 0 T 0

Inserting these expressions and dividing the continuity equation by the reference
density, Equation 6-10 in the frequency domain take on the following simplified form

iω  ------ – ------ + ∇ ⋅ u = 0
p T
 p 0 T 0

iωρ 0 u = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I
2μ
 3 
iω ( ρ 0 C p T – p ) + u ⋅ ∇p 0 + ρ 0 C p ( u ⋅ ∇T 0 ) = – ∇ ⋅ ( – k ∇T ) + Q

This is, for example, the system of equations implemented in the Thermoviscous
Acoustics, Frequency Domain interface when the ideal gas law is selected.

ISENTROPIC (ADIABATIC) IDEAL GAS CASE

If the process is assumed to be adiabatic and reversible — that is, isentropic — the
thermal conductivity is effectively zero. Then also the temperature can be eliminated,
giving for the ideal gas case:

590 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

 iω  ------ – -------------------- p + ∇ ⋅ u = 0
1 1
 p0 ρ0 Cp T 
0 (6-12)
iωρ 0 u = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I
2μ
 3 

where we have assumed constant background properties. Defining the speed of sound
c in analogy with the standard assumptions for linear acoustics (term in front of the
pressure in the continuity equation), it is found that

= ------ – -------------------- = ------  1 – ------- = ---------

1- 1 1 1 R 1
-----------
ρ0 c 2 p0 ρ0 Cp T0 p0 Cp γp 0

or

γp
c = --------0-
ρ0

In the case with a general fluid, the corresponding relation is using Equation 6-7 and
Equation 6-8:

2
1 ∂ρ 0 T 0 1 ∂ρ 0 2 1 T0 αp
– -------------  ------ ---------  = ------- – -------------
1- 1
----------- = ------- = ------ ---------
ρ0 c 2 K0 ρ 0 ∂p T ρ 0 C p ρ 0 ∂T p KT ρ0 Cp

where K0 is the adiabatic bulk modulus, KT the isothermal bulk modulus, and α0 the
coefficient of thermal expansion.

ISOTHERMAL CASE
If, on the other hand, the thermal conductivity is high, or the thermoviscous acoustic
waves propagate in a narrow space between highly conductive walls, the temperature
can be assumed to be constant (isothermal assumption) and the system of equations
for an ideal gas becomes:

iωρ 0 u = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I

2μ
3

iω  ------ p + ∇ ⋅ u = 0
1
p0

which, again comparing to standard assumptions, gives

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 591

 1- 1
----------- = ------
ρ0 c 2 p0

or equivalently

p0
c = ------
ρ0

Therefore, thermal conductivity and/or conducting walls decrease the apparent speed
of sound in narrow domains.

THE HELMHOLTZ EQUATION

If the thermodynamic processes in the system are assumed to be adiabatic and viscosity
can be neglected Equation 6-12 reduces, for constant background properties, to

iωρ 0 u = ∇ ⋅ ( – pI ) = ∇p
1
iω -----------2- p + ∇ ⋅ u = 0
ρ0 c

Now, taking the divergence of the momentum equations and inserting the expression
for the divergence of the velocity, taken from the continuity equations, yields the
Helmholtz equation for constant material properties:

2
2 ω
∇ p + ------
2
p = 0
c

Scattered Field Formulation and Background Acoustic Fields

The governing equations, given above in Equation 6-7, are defined in the general
scattered field formulation. Assuming that all the acoustic fields are the sum of a
background (prescribed) field and the resulting scattered field, the total fields can be
written as

pt = ps + pb ut = us + ub Tt = Ts + Tb
p ≡ ps u ≡ us T ≡ Ts

where subscript t stands for total, subscript s for scattered, and subscript b for
background. The scattered field is the field solved for (that is, the dependent variables
p, u, and T). If no background acoustic field is defined (the default), the scattered field
is equal to the total field. The governing equations and all boundary conditions are

592 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

expressed in terms of the total fields. The equations solved for are thus, in general, for
the frequency domain case given by

iωρ t + ∇ ⋅ ( ρ 0 u t ) = 0

iωρ 0 u t = ∇ ⋅ – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I
2μ
 3  (6-13)
ρ 0 C p ( iωT t + u t ⋅ ∇T 0 ) – T 0 α p ( iωp t + u t ⋅ ∇p 0 ) = – ∇ ⋅ ( – k ∇T t ) + Q
ρt = ρ0 ( βT pt – αp Tt )

The background acoustic fields are defined by adding the Background Acoustic Fields
feature. The feature has a user-defined option where all fields can be entered, either as
analytical expressions or defined in terms of a solution to another study or model. The
background acoustic fields feature also has a plane-wave option (in the frequency
domain interface) that defines a plane propagating attenuated wave.

PLANE WAVE SOLUTION

The plane wave option in the Background Acoustic Fields feature (in the frequency
domain only) defines the following fields

–kb ( nk ⋅ x )
pb = pb e
ω ( βT pb – αp Tb )
u b = ------------------------------------------ n k
kb
iωα p T 0 p b
T b = -----------------------------------2- (6-14)
iωρ 0 C p + kk b
1
---
ω iωb tv 2 4 ( γ – 1 )k
k b = ----  1 + -------------
- b tv = --- μ + μ B + --------------------
c ρ c 
2 3 Cp
0

where kb is the is the wave number of a plane propagating wave (in the direction nk)
with viscous and thermal attenuation; see Ref. 3. The expressions given in
Equation 6-14 are an exact solution to the governing equations Equation 6-13.

Formulation for Eigenfrequency Studies

When performing an eigenfrequency study, the governing equations (Equation 6-7)
are on the form:

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 593

 – λρ + ∇ ⋅ ( ρ 0 u ) = 0

– λρ 0 u = ∇ ⋅ – pI + μ ( ∇u + ( ∇u ) T ) –  ------- – μ B ( ∇ ⋅ u )I
2μ
3
– ρ 0 C p ( λT t + u t ⋅ ∇T 0 ) + T 0 α p ( λp t + u t ⋅ ∇p 0 ) = – ∇ ⋅ ( – k ∇T ) + Q
ρ = ρ0 ( βT p – αp T )

where the eigenvalue is λ = −iω. It is important to note that there is a difference

between regular pressure acoustics and thermoviscous acoustics in terms of what
modes can exist and which modes are found during an eigenfrequency study. In
pressure acoustics only the pure acoustic modes exist; the equations and assumptions
made ensure this. In thermoviscous acoustics, on the other hand, the equations are
formulated for all small-signal components that can exist. This means that other
nonacoustic modes also exist — that is, thermal and vorticity modes.

VORTICITY AND THERMAL MODES

When solving an eigenfrequency problem in thermoviscous acoustics, it is important
to take a close look at the obtained eigenfrequencies and assess if they are acoustic or
not. The nature of the solution is of the form

– λt iωt
p ( x, t ) ∼ p ( x )e = p ( x )e λ = α + iβ ω = – β + iα

where λ is the eigenvalue. Typically, eigenvalues exist near the positive real axis, where
β ≈ 0. These are exponentially decaying nonacoustic (nonoscillating) modes that stem
from the thermal equation or the deviatoric part of the momentum equation (the
nonpressure and nonvolume part of the stress tensor) also called the vorticity modes.
The acoustic eigenvalues on the other hand lie close to the imaginary axis and are
oscillating and slightly damped.

OTHER SPURIOUS MODES

Note that other spurious and nonacoustic modes can also exist when, for example, a
PML layer is used to model an open boundary. These modes stem from nonphysical
phenomena and the scaling inside the PML layer. In all cases, it is a good idea to have
an a priori knowledge of the location and type of the eigenvalues, maybe from solving
an lossless pressure acoustics model, and also to look at the modes in terms of, for
example, the pressure field.

594 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

Formulation for Mode Analysis in 2D and 1D Axisymmetry
The Mode Analysis study type is available for The Thermoviscous Acoustics,
Frequency Domain Interface in 2D and 1D axisymmetric models. This type of study
is used to determine the form of the propagating acoustic modes in waveguide
structures. The analyzed 2D and 1D axisymmetric geometries can be thought of as the
cross sections of a waveguide. This study is really restricted to the cross section of the
waveguides. To calculate the modes in 3D structures at the inlet of the waveguide use
The Thermoviscous Acoustics, Boundary Mode Interface, described below.

The spatial dependency in the (out-of-plane) axial z direction along the waveguide is
assumed to be of the form of a traveling wave with wave number kz. The dependent
variables in 2D are rewritten as

– ik z z
p = p ( x, y )e
– ik z z
u = u ( x, y )e
– ik z z
T = T ( x, y )e

and in 1D axisymmetric as

– ik z z
p = p ( r )e
– ik z z
u = u ( r )e
– ik z z
T = T ( r )e

Using this form of the dependent variables, differentiation with respect to z reduces to
a multiplication with −ikz. The propagating modes are determined by solving an
eigenvalue problem in the variable λ = −ikz.

The expression for the pressure can now be written retaining the harmonic time
dependency, as

k z = β + iα
– ik z z iωt αz i ( ωt – βz )
p = p ( x )e e = p ( x )e e

where x is the in plane coordinate(s). The axial wave number is split into a real and an
imaginary part. The imaginary part α of the wave number describes how fast the
propagating modes decay along the waveguide, it is often referred to as the attenuation
coefficient. The real part β is related to the phase speed cph of the propagating mode
by cph = β/ω. In thermoviscous acoustics, the obtained wave numbers always have an

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 595

 imaginary part as the modeled system always includes losses. The relation between the
angular frequency ω and the axial wave number kz is called the dispersion relation.

• Mode Analysis in the COMSOL Multiphysics Reference Manual

Formulation for the Boundary Mode Interface

The Thermoviscous Acoustics, Boundary Mode Interface interface adds extended
functionality to the abovementioned mode analysis available in 2D and 1D
axisymmetric geometries. The interface can be applied on boundaries in 3D (and 2D
axisymmetric) geometries and can be used to compute the propagating modes and
out-of-plane wave number kn on a (flat) surface of any orientation in 3D. This makes
it possible to more easily set up boundary conditions at inlets of waveguides using the
mode information.

The equations solved are the same as for The Thermoviscous Acoustics, Frequency
Domain Interface but using modified expressions for the gradients. The
implementation assumes propagation in the out-of-plane direction (that is, in the
normal direction n), and the fields can then be written as

– ik n x op
p = p ( x ip )e
– ik n x op
u = u ( x ip )e
– ik n x op
T = T ( x ip )e

where xip is the in-plane coordinate, xop is the out-of-plane coordinate, and kn is the
wave number that this formulation computes. Because of the above assumption about
the acoustic fields, the gradient operator can be split into a tangential in-plane
component (||) and an normal out-of-plane component (⊥), such that

∇p = ∇ || p – ik n pn
T
∇u = ∇ || u – ik n ( u ⋅ n )
∇T = ∇ || T – ik n Tn

The in-plane gradient is defined in COMSOL Multiphysics by the tangential derivative

operator dtang(). The divergence of the velocity field is defined as the trace of the
Jacobian, that is

596 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

 ∇ ⋅ u = trace ( ∇u )

With the above modifications and redefinition of the gradient operators, solving for
the propagating modes reduces to an eigenvalue problem solving for the acoustic field
and the out-of-plane wave number for a given frequency.

The Nonlinear Thermoviscous Acoustics Equations

The governing equations solved in a transient thermoviscous acoustics model that uses
the Nonlinear Thermoviscous Acoustics Contributions feature are also derived from
Equation 6-4 using a first order perturbation approach. However nonlinear terms in
the perturbations are retained. This yields the set of nonlinear thermo acoustics
equations

∂ρ
--------t + ∇ ⋅ ( ρ 0 u t ) + ∇ ⋅ ( ρ t u t ) = 0
∂t
∂u t ∂u t
ρ 0 --------- + ρ t --------- + ρ 0 ( u t ⋅ ∇u t ) = ∇ ⋅ σ + F
∂t ∂t
∂T t ∂p t
ρ 0 C p  --------- + u t ⋅ ∇T 0 – T 0 α p  -------- + u t ⋅ ∇p 0
 ∂t   ∂t 
∂T t
… + C p  ρ t --------- + ρ 0 ( u t ⋅ ∇T t ) + ρ t ( u t ⋅ ∇T 0 )
 ∂t 
∂p t
… – α p  T t -------- + T 0 ( u t ⋅ ∇p t ) + T t ( u t ⋅ ∇p 0 ) = – ∇ ⋅ ( – k ∇T t ) + Q
 ∂t 

ρt = ρ0 ( βT pt – α p T t )

σ = – p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I
2μ
 3 

In the default formulation only the linear equation of state is retained. This is a good
approximation in most cases as cumulative nonlinear effects typically can be neglected
in microacoustics. The nonlinear equations will capture local nonlinear effects like
vortex shedding. These can occur when the linearity condition |ut| << c is no longer
fulfilled. Note that the vortices are generated by the acoustic field itself.

In models with high local sound pressure levels, the linear equation of state may no
longer be valid. This happens when the linearity condition on the density ρt << ρ0 is
no longer fulfilled. If necessary, the feature has an option to change the default Density

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 597

 expansion from First order to Second order. This corresponds to the following second
order equation of state

2 2 2
1 2 ∂ ρ0 ∂ ρ0 2∂ ρ
ρ t = ρ 0 ( β T p t – α p T t ) + --- p t ----------- -+1
- + p t T t ------------------- --- T t -----------0-
2 ∂p 2 ∂p 0 ∂T 0 2 ∂T 2
0 0

Additional inputs to the model are necessary. For the general case, the second order
derivatives of the equilibrium density ρ0 = ρ0(p0,T0) with respect to pressure p0 and
temperature T0 are needed. They contribute to the second-order Taylor expansion of
the density. Per default, they are taken From equilibrium density; this implies that the
dependency of the density on pressure and temperature should be correct. Note that
this dependency is included if a material is defined using the functionality of the Liquid
and Gas Properties Module (see the Material Properties section in the Acoustic
Properties of Fluids chapter for details).

References for the Thermoviscous Acoustics, Frequency Domain

Interface
1. W.M. Beltman, P.J.M. van der Hoogt, R.M.E.J. Spiering, and H. Tijdeman,
“Implementation and Experimental Validation of a New Viscothermal Acoustic Finite
Element for Acousto-Elastic Problems”, J. Sound and Vibration, vol. 216, pp. 159–
185, 1998.

2. M. Malinen, M. Lyly, and others, “A Finite Element Method for the Modeling of
Thermo-Viscous Effects in Acoustics”, Proc. ECCOMAS 2004, Jyväskylä, 2004.

3. D.T. Blackstock, Fundamentals of Physical Acoustics, John Wiley & Sons, 2000.

4. H. Bruus, Theoretical Microfluidics, Oxford University Press, 2010.

5. G.K. Bachelor, An Introduction to Fluid Dynamics, Cambridge University Press,

2000.

6. M.K. Myers, “Transport of energy by disturbances in arbitrary steady flows,”

J. Fluid Mech., vol. 226, pp. 383–400, 1991.

7. B. Lautrup, Physics of Continuous Matter, Exotic and Every Day Phenomena in

the Macroscopic World, 2nd ed., CRC Press, 2011.

8. A.S. Dukhin and P.J. Goetz, “Bulk viscosity and compressibility measurements
using acoustic spectroscopy,” J. Chem. Phys., vol. 130, p. 124519-1, 2009.

598 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES

9. M. S. Cramer, “Numerical estimate for the bulk viscosity of ideal gases,” Phys.
Fluids, vol. 30, pp. 066102, 2012.

THEORY BACKGROUND FOR THE THERMOVISCOUS ACOUSTICS BRANCH | 599

600 | CHAPTER 6: THERMOVISCOUS ACOUSTICS INTERFACES
 7

Ultrasound Interfaces

This chapter describes the physics interfaces found under the Acoustics>Ultrasound
branch ( )

• The Convected Wave Equation, Time Explicit Interface

• The Nonlinear Pressure Acoustics, Time Explicit Interface
• Modeling with the Convected Wave Equation Interface
• Modeling with the Nonlinear Pressure Acoustics, Time Explicit Interface
• Theory for the Convected Wave Equation Interface
• Theory for the Nonlinear Pressure Acoustics, Time Explicit Interface
• References for the Ultrasound Interface

601
 The Convected Wave Equation, Time
Explicit Interface
The Convected Wave Equation, Time Explicit (cwe) interface ( ), found under the
Acoustics>Ultrasound subbranch ( ) when adding a physics interface, is used to solve
large transient linear acoustic problems containing many wavelengths in a stationary
background flow. It is suited for time-dependent simulations with arbitrary
time-dependent sources and fields. In general, the interface is suited for modeling the
propagation of acoustic signals over large distances relative to the wavelength, for
example, linear ultrasound problems. Application areas include ultrasound flowmeters
and other ultrasound sensors where time of flight is an important parameter. The
applications are not restricted to ultrasound. The interface includes absorbing layers
that are used to set up effective nonreflecting like boundary conditions. The interface
exists in 2D, 2D axisymmetric, and 3D.

The interface is based on the discontinuous Galerkin (dG or dG-FEM) method and
uses a time explicit solver. The method is very memory efficient and can solve problems
with many million degrees of freedom (DOFs). The method is also well suited for
distributed computing on clusters.

In the no-flow case it can be advantageous to use The Pressure Acoustics, Time
Explicit Interface instead. The equations are the same but the latter interface has more
options to, for example, model scattering phenomena using a scattered field
formulation; or for handling sharp material discontinuities.

The interface solves the linearized Euler equations assuming an adiabatic equation of
state. The dependent variables are the acoustic pressure p and the acoustic velocity
perturbation u. The background mean flow can be any stationary flow with small to
moderate velocity gradients. General bulk dissipation (volumetric damping) can be
added to model real fluids.

When solving models that are based on the dG method, optimizing the
mesh is important. For further details see Optimizing the Mesh for DG
in the Meshing, Discretization, and Solvers section.

When this physics interface is added, these default nodes are also added to the Model
Builder — Convected Wave Equation Model, Sound Hard Wall, Initial Values, and Axial

602 | CHAPTER 7: ULTRASOUND INTERFACES

Symmetry (in case of 2D axisymmetric). Then, from the Physics toolbar, add other
nodes that implement, for example, boundary conditions. You can also right-click
Convected Wave Equation, Time Explicit to select physics features from the context
menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is cwe.

FILTER PARAMETERS FOR ABSORBING LAYERS

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. In the Filter Parameters for
Absorbing Layers section you can change and control the values set for the filter used in
the Absorbing Layers. The values of the filter parameters defined here are used in all
absorbing layers added to the model and they override the value of filter parameters
enabled in the material model (Convected Wave Equation Model). The default values
of the filter parameters α, ηc, and s are set to 0.1, 0.01, and 2, respectively. Inside the
absorbing layer it is important to use a filter that is not too aggressive since this will
result in spurious reflections.

For general information about the filter see the Filter Parameters section
under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

EQUATION SETTINGS (2D AXISYMMETRIC ONLY)

This section is only available in 2D Axisymmetric. Per default the acoustic velocity
perturbation u is assumed to have only two components: the radial component ur and
the axial component uz. You can select the Include out-of-plane components check box
to enable the third — circumferential component u ϕ , which physically can only be
nonzero if the background mean flow has the corresponding component.

THE CONVECTED WAVE EQUATION, TIME EXPLICIT INTERFACE | 603

 DISCRETIZATION
In this section you can select the discretization for the Acoustic pressure and Acoustic
velocity. Per default both are set to Quartic (4th order). Using quartic elements
together with a mesh size equal to approximately half the wavelength to be resolved,
leads to the best performance when using the dG method. For further details see the
Meshing, Discretization, and Solvers section.

DEPENDENT VARIABLES
The dependent variables (field variables) are the Acoustic pressure, Acoustic velocity, and
Acoustic velocity, components. The names can be changed, but the names of fields and
dependent variables must be unique within a model.

• For information about modeling strategies, meshing, the absorbing

layer and more, see Modeling with the Convected Wave Equation
Interface.
• For the theoretical background of the model solved, see Theory for the
Convected Wave Equation Interface.

• Gaussian Pulse in 2D Uniform Flow: Convected Wave Equation and

Absorbing Layers. The Application Library path:
Acoustics_Module/Tutorials,_Pressure_Acoustics/
gaussian_pulse_absorbing_layers
• Ultrasound Flowmeter with Generic Time-of-Flight Configuration .
The Application Library path:
Acoustics_Module/Ultrasound/ultrasound_flow_meter_generic

Domain, Boundary, Edge, Point, and Pair Nodes for the Convected
Wave Equation Interface
The Convected Wave Equation, Time Explicit Interface has these domain, boundary,
edge, point, and pair nodes, listed in alphabetical order, available from the Physics
ribbon toolbar (Windows users), Physics context menu (Mac or Linux users), or

604 | CHAPTER 7: ULTRASOUND INTERFACES

right-click to access the context menu (all users). Continuity in pressure and velocity
is the default condition on interior boundaries.

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

• Acoustic Impedance • Interior Wall

• Convected Wave Equation Model • Normal Velocity
• Domain Sources • Pressure
• General Flux/Source • Sound Hard Wall
• Initial Values • Symmetry
• Interior Normal Velocity

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Convected Wave Equation Model

The Convected Wave Equations node adds the equations for modeling the propagation
of acoustic waves in a stationary background flow. Adiabatic behavior is assumed,
meaning that this in some sense represents pressure acoustics in the presence of flow.
The convected wave equation model solves the linearized Euler equations also referred
to as linear acoustic equations for moving media. The equations are valid for any
stationary background mean flow as long as there are not too large gradients in the
background properties.

The linear continuity equation, momentum equation, and equation of state solved are
given by:

THE CONVECTED WAVE EQUATION, TIME EXPLICIT INTERFACE | 605

 ∂ρ
------ + ( u 0 ⋅ ∇ )ρ + ( u ⋅ ∇ )ρ 0 + ρ ( ∇ ⋅ u 0 ) + ρ 0 ( ∇ ⋅ u ) = f p
∂t
∂u- 1 ρ
------ + ( u 0 ⋅ ∇ )u + ( u ⋅ ∇ )u 0 + ------ ∇p – -----2- ∇p 0 = f v
∂t ρ0 ρ
0
p
ρ = ----2-
c0

where u0 is the background mean flow velocity, p0 is the background mean flow
pressure, ρ0 is the background mean flow density, and c0 is the speed of sound. All
background properties can vary in space (the variation has to be smooth), but cannot
be discontinuous. The right-hand side sources fp and fv can be defined by the Domain
Sources. More information about the governing equations is given in the Theory for
the Convected Wave Equation Interface section.

In the Settings window, define the properties for the acoustics model and model inputs
including the background mean flow, pressure, and velocity.

MODEL INPUTS
In order to model the influence the background mean flow has on the propagation of
the acoustic waves in the fluid, the Background mean flow pressure p0 and Background
mean flow velocity u0 need to be defined. If a material that is temperature dependent
is used the Background mean flow temperature T0 field is also present.

All the background mean flow parameters can be functions of space. They can be either
analytical expressions (user defined) or they can be picked up from a fluid flow
simulation performed using the CFD Module. In this case use the Background Fluid
Flow Coupling multiphysics feature to ensure correct mapping of the flow data. By
default they are set to the quiescent constant background conditions of air.

FLUID PROPERTIES
Select a Fluid model either as Linear elastic (the default) or General dissipation. For both
options define the background mean flow density ρ0 (SI unit: kg/m3) and the
background mean flow speed of sound c0 (SI unit: m/s). The default is to use the
property values from the material (From material). Select User defined from the list to
enter a user-defined value in the text field that appears. For the General dissipation
option also define the Sound diffusivity δ (SI unit: m2/s). For numerical stability
reasons it is recommended to use physical values of attenuation properties.

606 | CHAPTER 7: ULTRASOUND INTERFACES

The density can also be picked up from a fluid flow simulation when the Background
Fluid Flow Coupling multiphysics feature is used.

The relation between the sound diffusivity δ and the equivalent (plane
wave) attenuation coefficient α, at a given frequency f, is given by the
2 3
expression α = δω ⁄ ( 2c ) , where ω = 2πf and c is the speed of sound.

LAX–FRIEDRICHS FLUX PARAMETER

To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box. In this section, you specify the value of the Lax–
Friedrichs flux parameter τLF (default value: 0.2). This value controls the numerical
flux between the elements (nodal discontinuous Lagrange elements) used with the
discontinuous Galerkin (dG) method. The numerical flux defines how adjacent
elements are connected and how continuous p and u are. Different definitions of the
numerical flux lead to different variants of the dG method. The flux implemented here
is the so-called global Lax–Friedrichs numerical flux. The value of the parameter τLF
should be between 0 and 0.5. For τLF = 0 a so-called central flux is obtained. Setting
τLF = 0.5 gives a maximally dissipative global Lax–Friedrichs flux.

For general information about the numerical flux see the Numerical Flux
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

FILTER PARAMETERS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. By default, the filter parameters
α, ηc, and s are not active. Select the Activate check box to activate the filter. The filter
provides higher-order smoothing for the dG formulation and can be used to stabilize
the solution, for example, when a large background flow is present or large gradients
are present. Inside absorbing layers the settings given here are overridden by the Filter
Parameters for Absorbing Layers.

Enter values for the filter parameters in the corresponding text fields (default values:
36, 0.6, and 3). α must be positive and lie between 0 and 36. α = 0 means no
dissipation and α = 36 means maximal dissipation. ηc should be between 0 and 1,
where 0 means maximal filtering and 1 means no filtering (even if filtering is active).
The s parameter should be larger than 0 and controls the order of the filtering (a

THE CONVECTED WAVE EQUATION, TIME EXPLICIT INTERFACE | 607

 dissipation operator of order 2s). For s = 1, you get a filter that is related to the classical
2nd-order Laplacian. A larger s gives a more pronounced low-pass filter.

For more detailed information about the filter see the Filter Parameters
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

Domain Sources
Use a Domain Sources node to add various sources to a domain. The feature adds
right-hand side sources to either the continuity or the momentum equation.

The Domain mass source qsource and the Domain pressure source psource both contribute
to the source term fp of the continuity equation. This is a monopole like source that
acts uniformly in all directions. The two represent different formulations of the same
source type.

1 ∂p source
f p = q source + ----2- -------------------
c 0 ∂t

The Domain velocity source usource and Domain force source fsource both contribute to
the source term fv of the momentum equation. This is a dipole like source that acts in
the direction of its vector. The two represent different formulations of the same source
type.

∂u source
f v = ρ 0 -------------------- + f source
∂t

All source terms can be any analytical or interpolation function of the time. The
variable t is reserved to represent time in expressions.

DOMAIN SOURCES
In this section, add one or more of the following sources:

• Add a Domain mass source qsource (SI unit: kg/(m3·s))

• Add a Domain pressure source psource (SI unit: Pa)
• Add a Domain velocity source usource (SI unit: m/s).
• Add a Domain force source fsource (SI unit: N/m3).

608 | CHAPTER 7: ULTRASOUND INTERFACES

Sound Hard Wall
The Sound Hard Wall adds a boundary condition for a sound hard boundary or wall,
which is a boundary at which the normal component of the velocity is zero (a slip
condition). This is assumed true for both the acoustic velocity and the background
velocity

–n ⋅ u = 0 –n ⋅ u0 = 0

Note that this condition is not identical to the Symmetry condition where only the
normal component of the acoustic velocity is assumed to be zero.

Initial Values
The Initial Values node adds initial values for the acoustic pressure and the acoustic
velocity that can serve as an initial state for the solver. If more than one initial value is
needed, from the Physics toolbar click to add more Initial Values nodes.

INITIAL VALUES
Enter a value or expression for the initial values for the Pressure p (SI unit: Pa) and the
Acoustic velocity, u (SI unit: m/s).

Normal Velocity
The Normal Velocity adds an inward normal velocity either given as a scalar vn or as a
velocity vector vb, the condition given is

–n ⋅ u = vn ( t ) = –n ⋅ vb ( t )

where n is the surface normal. Both expressions should be functions of time. This
feature represents an external source term. It can also be used to model a vibrating
transducer surface or other vibrating surfaces.

NORMAL VELOCITY
Select a Type — Inward Velocity (the default) or Velocity.

• For Inward Velocity enter an expression for the Inward velocity vn(t) (SI unit: m/s).
Use a positive value for inward velocity or a negative value for outward velocity.
• For Velocity enter expressions for the components of the Velocity vb(t) (SI unit:
m/s).

THE CONVECTED WAVE EQUATION, TIME EXPLICIT INTERFACE | 609

 Pressure
The Pressure node creates a boundary condition that acts as a pressure source at the
boundary. The acoustic pressure is given at the boundary p = pb(t).

PRESSURE
Enter an expression for the Boundary pressure pb (SI unit: Pa).

Symmetry
The Symmetry node adds a boundary condition imposing symmetry in the pressure
and velocity field. Use this condition to reduce the size of a model by cutting it in half
where there are known symmetries. The condition implies symmetry in the acoustic
fields

–n ⋅ u = 0

Acoustic Impedance
The Acoustic Impedance node adds a boundary condition defining a relation between
the local acoustic pressure p and the normal acoustic velocity n·u, that is, the acoustic
impedance Z. The condition specifies the inward normal velocity

p
– n ⋅ u = – ----
Z

This condition can be used to model the properties of artificial boundaries. When the
impedance is set equal to the characteristic specific impedance of a propagating plane
wave ρ0c0, the condition represent the simplest nonreflecting boundary condition.
This is also the default value of the impedance when the impedance condition is added.

Note that this condition is not equivalent to the general impedance condition, as given
by Myers (see Ref. 4), when a flow is present. This means that, for example, it cannot
be used at a reacting surface which has a tangential flow. For this kind of detailed
conditions a frequency domain representation is needed.

The acoustic impedance condition (with the default value ρ0c0) should be used at the
outer boundary when Absorbing Layers are used in a model.

610 | CHAPTER 7: ULTRASOUND INTERFACES

ACOUSTIC IMPEDANCE
Enter a value or expression for the Acoustic Impedance Z (SI unit: Pa·s/m). The default
expression is cwe.c0*cwe.rho0 which is the characteristic specific impedance ρ0c0 of
a propagating plane wave.

Interior Wall
The Interior Wall node adds the boundary condition for a sound hard boundary or wall
on interior boundaries. The condition reads

u up ⋅ n = u down ⋅ n = 0
u 0, up ⋅ n = u 0, down ⋅ n = 0

The condition ensures that the normal velocity is zero on both the up and down side
of the boundary, while applying a slit on the pressure (the pressure is discontinuous
across the boundary). The condition also assumes that the background flow has a
similar condition applied at the boundary.

Interior Normal Velocity

The Interior Normal Velocity node adds the boundary condition for a moving interior
boundary. This can, for example, represent a thin vibrating structure or s speaker
membrane. The condition reads

u t,up ⋅ n = v b ( t ) ⋅ n = – v n ( t )
u t,down ⋅ n = v b ( t ) ⋅ n = – v n ( t )

The condition ensures that the normal velocity is defined on both the up and down
side of the boundary, while applying a slit on the pressure (the pressure is
discontinuous across the boundary).

NORMAL VELOCITY
Select the Type: Velocity (default) or Inward velocity and specify either the Velocity
(vector) vb(t) or the normal component through the Inward normal velocity vn(t).

General Flux/Source
The General Flux/Source node, found under the More submenu, adds the most general
boundary condition for the convected wave equation, as it is formulated (in the

THE CONVECTED WAVE EQUATION, TIME EXPLICIT INTERFACE | 611

 conservative form) for the discontinuous Galerkin method (dG). The condition
defines the normal flux g at an exterior boundary, given by

p
ρ 0 ( n ⋅ u ) + ----2- ( n ⋅ u 0 )
c0
p-
---- u ( n ⋅ u0 ) + ρ0 ( u ( n ⋅ u0 ) + u0 ( n ⋅ u ) ) + n1 p
2 0
c0
–n ⋅ Γ = = g
p-
---- v ( n ⋅ u0 ) + ρ0 ( v ( n ⋅ u0 ) + v0 ( n ⋅ u ) ) + n2 p
2 0
c0
p-
---- w ( n ⋅ u0 ) + ρ0 ( w ( n ⋅ u0 ) + w0 ( n ⋅ u ) ) + n3 p
2 0
c0

where u = (u, v, w), u0 = (u0, v0, w0), and n = (nx, ny, nz) (n = (nr, n ϕ , nz) in 2D
axisymmetric) is the surface normal. Care should be taken when using this advanced
condition as the method is sensitive to proper formulation of boundary conditions.
Only one characteristic can enter the domain at any boundary at the time, meaning
that it is easy to over-specify the problem. Use the mesh normals (nxmesh, nymesh,
and nzmeshor, respectively, nrmesh, nphimesh, and nzmesh) in the expression you
define.

GENERAL FLUX/SOURCE
Enter the expression for the components of the Flux vector g.

General Interior Flux

The General Interior Flux node, found under the More submenu, adds a the most
general interior boundary condition for the convected wave equation, as it is
formulated (in the conservative form) for the discontinuous Galerkin method (dG).
The condition defines the normal flux g on an interior boundary by

– n ⋅ Γ∗ = g

here Γ* represents the total numerical flux. This means that if this condition is added
on an interior boundary it overrides the existing Lax–Friedrichs numerical flux. So care
should be taken when specifying this condition as it needs to include the necessary
numerical flux contribution to ensure stability of the method.

On an interior boundary you can use the up() and down() operators to access values
from both sides of the boundary. If a dependent variable is used in this expression

612 | CHAPTER 7: ULTRASOUND INTERFACES

(without up() or down() operators) an implicit mean() operation is invoked taking
the average of the up and down side values. Use the mesh normals (nxmesh, nymesh,
and nzmesh or, respectively, nrmesh, nphimesh, and nzmesh) in the expression you
define.

GENERAL INTERIOR FLUX

Enter the expression for the components of the Flux vector g.

THE CONVECTED WAVE EQUATION, TIME EXPLICIT INTERFACE | 613

 T he No nli ne a r Pressu re A cou st i c s,
Time Explicit Interface
The Nonlinear Pressure Acoustics, Time Explicit (nate) interface ( ), found under the
Acoustics>Ultrasound subbranch ( ) when adding a physics interface, is used to
model the propagation of nonlinear finite amplitude acoustic waves in computational
domains that contain many wavelengths. It is suited for time-dependent simulations
with arbitrary time-dependent sources and fields. Absorbing layers are used to set up
effective nonreflecting-like boundary conditions. Application areas include biomedical
applications, for instance, ultrasonic imaging and high-intensity focused ultrasound
(HIFU). The applications are not restricted to ultrasound. The interface exists in 2D,
2D axisymmetric, and 3D.

The interface is based on the discontinuous Galerkin (dG or dG-FEM) method and
uses a time explicit solver. The method is very memory efficient and can solve problems
with many million degrees of freedom (DOFs). The method is also well suited for
distributed computing on clusters.

In the linear case it can be advantageous to use The Pressure Acoustics, Time Explicit
Interface instead, for example, when modeling scattering phenomena using a scattered
field formulation and a background acoustic field. This split is not possible for
nonlinear problems. In the presence of a stationary background flow and linear
propagation the The Convected Wave Equation, Time Explicit Interface should be
used, for example, when modeling ultrasonic flow meters.

For modeling acoustic-structure interaction (ASI) or vibroacoustic problems the

interface is fully multiphysics enabled and can be coupled to the The Elastic Waves,
Time Explicit Interface, using either the Acoustic-Structure Boundary, Time Explicit
or the Pair Acoustic-Structure Boundary, Time Explicit multiphysics couplings.

The interface solves the second-order nonlinear governing equations for the acoustic
pressure p and the acoustic velocity perturbations u. The interface is suited for
modeling progressive wave propagation phenomena when the cumulative nonlinear
effects surpass the local nonlinear effects. Thus the model is consistent with the
second-order Westervelt equation for the acoustic pressure (see also the Nonlinear
Acoustics (Westervelt) Contributions feature available with The Pressure Acoustics,
Transient Interface). General bulk dissipation (volumetric damping) can be added to
model real fluids.

614 | CHAPTER 7: ULTRASOUND INTERFACES

Several features are available to help solve the nonlinear and highly nonlinear problems
including the use of a numerical Limiter (to capture shocks) but also the use of
Adaptive Mesh Refinement.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is nate.

FILTER PARAMETERS FOR ABSORBING LAYERS

To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. In the Filter Parameters for
Absorbing Layers section you can change and control the values set for the filter used in
the Absorbing Layers. The values of the filter parameters defined here are used in all
absorbing layers added to the model and they override the value of filter parameters
enabled in the material model (Nonlinear Pressure Acoustics, Time Explicit Model).
The default values of the filter parameters α, ηc, and s are set to 0.1, 0.01, and 2,
respectively. Inside the absorbing layer it is important to use a filter that is not too
aggressive since this will result in spurious reflections.

For general information about the filter see the Filter Parameters section
under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

NUMERICAL FLUX
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box. Only one option exists for the Numerical flux
formulation used in the numerical scheme when solving the dG problem. For Lax–
Friedrichs (the only option) enter the Lax–Friedrichs flux parameter τLF (default is 0.2).
The value of the parameter τLF should be between 0 and 0.5. A parameter value of 0
represents a central flux, which is the least dissipative but also the least stable numerical
flux. A parameter value of 0.5 gives a maximally dissipative global Lax–Friedrichs flux
(see The Lax–Friedrichs Flux).

THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 615

 LIMITER
To display this section, click the Show More Options button ( ) and select Stabilization
in the Show More Options dialog box. The limiter is used to control and stabilize highly
nonlinear problems with shock formation. The limiter can only be used with linear
discretization. For details see the Solving Highly Nonlinear Problems section in
Modeling with the Nonlinear Pressure Acoustics, Time Explicit Interface.

• For an example of a highly nonlinear problem that uses the WENO

limiter see the Nonlinear Propagation of a Cylindrical Wave —
Verification Model tutorial model. The Application Library path
Acoustics_Module/Nonlinear_Acoustics/ nonlinear_cylindrical_wave
• For an example of a nonlinear problem with pulse propagation that
uses mesh adaptation see the High-Intensity Focused Ultrasound
(HIFU) Propagation Through a Tissue Phantom tutorial model. The
Application Library path Acoustics_Module/
Nonlinear_Acoustics/hifu_tissue_sample

DISCRETIZATION
In this section you can select the discretization for the Acoustic pressure and Acoustic
velocity (the same is used for both). Per default both are set to Quartic (4th order).
Using quartic elements together with a mesh size equal to approximately one and a
half of the wavelength to be resolved, leads to the best performance when using the
dG method.

DEPENDENT VARIABLES
The dependent variables are the Acoustic pressure, and the Acoustic velocity. The names
can be changed, but the names of fields and dependent variables must be unique within
a model. The name for the Acoustic velocity, components can also be selected
individually.

• For information specific to modeling nonlinear problems using the

time explicit interface, see Modeling with the Nonlinear Pressure
Acoustics, Time Explicit Interface.
• For information about modeling strategies, meshing, the absorbing
layer and more, see the relevant information in the Modeling with the
Convected Wave Equation Interface.

616 | CHAPTER 7: ULTRASOUND INTERFACES

Domain, Boundary, Edge, and Point Nodes for the Nonlinear
Pressure Acoustics, Time Explicit Interface
The Nonlinear Pressure Acoustics, Time Explicit Interface has the following domain
and boundary conditions listed in alphabetic order, available from the Physics ribbon
toolbar (Windows users), Physics context menu (Mac or Linux users), or right-click to
access the context menu (all users). Continuity in velocity and pressure is the default
condition on interior boundaries. If a boundary separates two different materials the
Material Discontinuity or Continuity condition (the pair version of material
discontinuity condition) has to be applied to ensure the continuity in velocity and
pressure.

• Continuity • Mass Source

• General Flux/Source • Material Discontinuity
• General Interior Flux • Nonlinear Pressure Acoustics, Time
• Heat Source Explicit Model

• Impedance • Normal Velocity

• Initial Values • Pressure

• Interior Normal Velocity • Sound Hard Boundary (Wall)

• Interior Sound Hard Boundary • Sound Soft Boundary

(Wall) • Symmetry
• Volume Force Source

Nonlinear Pressure Acoustics, Time Explicit Model

The Nonlinear Pressure Acoustics, Time Explicit Model node adds the equations for
modeling the propagation of nonlinear acoustic waves. The nonlinear continuity
equation and momentum equation solved are given by:

1 - ∂p t  βp 
-------- -------- + ∇ ⋅  1 + ---------t u t = Q m
2 ∂t 2
ρc  ρc 
∂u t (7-1)
ρ --------- + ∇ ⋅ ( p t I ) = q d
∂t
pt = p ut = u

THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 617

 where pt is the total acoustic pressure and ut is the total acoustic velocity fluctuations.
Because the problem solved is nonlinear the superposition principle does not apply
which means that the “total” fields are always equal to the dependent variables.

In the Settings window, define the properties for the acoustics model, model inputs,
and material properties.

MODEL INPUTS
Depending on the material selected enter values for the Model Inputs, typically the
Temperature T, the Pressure pA, and possibly the Relative Humidity φ .

PRESSURE ACOUSTICS MODEL

Select the Fluid model — Linear elastic, Viscous, Thermally conducting, Thermally
conducting and viscous, General dissipation, or Ideal gas. The settings are the same as for
the Pressure Acoustics, Time Explicit Model.

COEFFICIENT OF NONLINEARITY
Select how to define the Coefficient of nonlinearity β — From parameter of nonlinearity
(default) or User defined. The Parameter of nonlinearity B/A can be taken From material
or entered as a User defined expression. For the Ideal gas fluid model, it is also possible
to compute the Coefficient of nonlinearity β From ratio of specific heats.

FILTER PARAMETERS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box. By default, the filter parameters
α, ηc, and s are not active. Select the Activate check box to activate the filter. The filter
provides higher-order smoothing for the DG formulation. Inside absorbing layers the
settings given here are overridden by the Filter Parameters for Absorbing Layers.

For more detailed information about the filter see the Filter Parameters
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

Compute Minimum and Maximum Pressure

The Compute Minimum and Maximum Pressure subfeature can be added to the
Nonlinear Pressure Acoustics, Time Explicit Model as either a domain or a boundary
feature. The minimum and maximum value of the pressure, over time and space, will
be computed on the selected entity. It will create two variables nate.p_min and

618 | CHAPTER 7: ULTRASOUND INTERFACES

nate.p_max, these can be used in postprocessing to, for example, evaluate the size of
a focal zone.

For an example, see the High-Intensity Focused Ultrasound (HIFU)

Propagation Through a Tissue Phantom tutorial model. The Application
Library path Acoustics_Module/ Nonlinear_Acoustics/hifu_tissue_sample

Initial Values
The Initial Values node adds initial values for the acoustic pressure and the acoustic
velocity that can serve as an initial state for the solver. If more than one initial value is
needed, from the Physics toolbar click to add more Initial Values nodes.

INITIAL VALUES
Enter a value or expression for the initial values for the Pressure p (SI unit: Pa) and the
Acoustic velocity, u (SI unit: m/s).

Mass Source
The Mass Source sets up the right-hand side of the continuity equation, in
Equation 7-1, in terms of a mass rate of change. This is a monopole-like source that
acts uniformly in all directions.

1
Q m = --- M
ρ

MASS SOURCE
Enter the Mass source M (SI unit: kg/(m3·s)).

Heat Source
The Heat Source sets up the right-hand side of the continuity equation, in
Equation 7-1, in terms of a heat density source. This is a monopole-like source that
acts uniformly in all directions.

αp
Q m = ---------- Q heat
ρC p

THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 619

 where αp is the (isobaric) coefficient of thermal expansion, ρ is the density, Cp is the
heat capacity at constant pressure, and Qheat is the heat source.

MODEL INPUT
Enter the ambient Temperature T, the default is 293.15[K].

HEAT SOURCE
Enter the Heat Source Qheat (SI unit: W/m3).

FLUID PROPERTIES
Enter the Heat capacity at constant pressure Cp and the Ratio of specific heats γ either
selected From material or User defined. The ratio of specific heats is used to calculate
the coefficient of thermal expansion through the thermodynamic relation

1 Cp ( γ – 1 )
α p = --- -----------------------
-
c T

Volume Force Source

The Volume Force Source sets up the right-hand side of the momentum equation, in
Equation 7-1, in terms of a volumetric force. This is a dipole-like source that acts
differently in two directions.

qd = F

VOLUME FORCE SOURCE

Enter the Volume force source F (SI unit: N/m3).

Sound Hard Boundary (Wall)

The Sound Hard Boundary (Wall) adds a boundary condition for a sound hard boundary
or wall, which is a boundary at which the normal component of the velocity is zero (a
slip condition)

n ⋅ ut = 0

The condition is also available from the Pairs submenu, as an option at interfaces
between parts in an assembly. This effectively introduces a condition similar to the
Interior Sound Hard Boundary (Wall) but used on an assembly.

620 | CHAPTER 7: ULTRASOUND INTERFACES

Sound Soft Boundary
The Sound Soft Boundary conditions adds a boundary condition for a sound soft
boundary, where the acoustic pressure vanishes pt = 0. It is an appropriate
approximation, for example, for a liquid-gas interface or in some cases for external
waveguide ports.

Pressure
The Pressure node creates a boundary condition that acts as a pressure source at the
boundary. The total acoustic pressure is given at the boundary pt = p0(t).

PRESSURE
Enter an expression for the Pressure p0(t) (SI unit: Pa). The input can be any analytical
function or interpolation data that can be function of both time and space.

Symmetry
The Symmetry node adds a boundary condition imposing symmetry in the pressure
and velocity field. Use this condition to reduce the size of a model by cutting it in half
where there are known symmetries. The condition implies symmetry in the acoustic
fields

n ⋅ ut = 0

Normal Velocity
The Normal Velocity condition adds a source specified in terms of an inward normal
velocity vn(t) or velocity v0(t) of the boundary. The component in the normal
direction is used to define the boundary condition. This feature represents an external
source term like a vibrating surface. The condition is also available from the Pairs
submenu, as an option at interfaces between parts in an assembly. This effectively
introduces a condition similar to the Interior Normal Velocity but used on an assembly.

NORMAL VELOCITY
Select a Type: Inward Velocity (the default) or Velocity.

• For Inward Velocity enter the value of the Inward velocity vn (SI unit: m/s).
• For Velocity enter values for the components of the Velocity v0 (SI unit: m/s).

THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 621

 Impedance
The Impedance condition adds a boundary condition defining the relation between the
local acoustic pressure p and the normal acoustic velocity n·ut, that is, the specific
acoustic impedance Zi. The model includes the option to define the nonlinear or
second order pressure-velocity relation to extend the validity of the impedance
condition to higher pressure amplitudes.

The second order impedance expression specifies how the normal velocity relates to
the pressure through the impedance, including a correction factor, according to

pt  βp t 
n ⋅ u t = -----  1 – ------------2
Zi  2ρc 

This condition can be used to model the properties of artificial boundaries by adding
resistive losses (no reactive components can be added). When the impedance is set
equal to the characteristic specific impedance of a propagating plane wave ρc, the
condition represent the simplest nonreflecting boundary condition. This is also the
default value of the impedance when the condition is added.

The acoustic impedance condition (with the default value ρc) should be used at the
outer boundary when Absorbing Layers are used in a model.

IMPEDANCE
Select the Pressure-particle velocity relation used for the impedance formulation First
order (default) or Second order. Enter a value or expression for the Impedance Zi (SI
unit: Pa·s/m). The default expression is nate.c*[Link] which is the characteristic
specific impedance ρc of a propagating plane wave (linear theory).

Interior Sound Hard Boundary (Wall)

The Interior Sound Hard Boundary (Wall) node, found under the Interior Conditions
submenu, adds the boundary condition for a sound hard boundary or wall on interior
boundaries. The condition reads

u t, up ⋅ n = 0
u t, down ⋅ n = 0

622 | CHAPTER 7: ULTRASOUND INTERFACES

The condition ensures that the total normal velocity is zero on both the up and down
side of the boundary, while applying a slit on the pressure (the pressure is
discontinuous across the boundary).

Interior Normal Velocity

The Interior Normal Velocity node, found under the Interior Conditions submenu, adds
the boundary condition for a moving interior boundary. This can, for example,
represent a thin vibrating structure or s speaker membrane. The condition reads

u t, up ⋅ n = v 0 ( t ) ⋅ n = – v n ( t )
u t, down ⋅ n = v 0 ( t ) ⋅ n = – v n ( t )

The condition ensures that the total normal velocity is defined on both the up and
down side of the boundary, while applying a slit on the pressure (the pressure is
discontinuous across the boundary).

NORMAL VELOCITY
Select the Type: Velocity (default) or Normal velocity and specify either the Velocity
(vector) v0 or the normal component through the Normal velocity vn.

Material Discontinuity
The Material Discontinuity node, found under the Interior Conditions submenu, adds an
interior boundary condition to handle jumps in material properties. As the interface is
based on the discontinuous Galerkin method special handling is necessary when a
jump exists in the specific characteristic impedance Z0 = ρc of the medium.

The condition implements continuity in total pressure and in the normal velocity over
the interior boundary:

p up = p down
n ⋅ u up = n ⋅ u down

Subscripts “up” and “down” in the equation refer to the up and down sides of the
interior boundary, relative to the normal n.

THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 623

 Continuity
The Continuity node, exists as an option at interfaces between parts in an assembly,
from the Pairs submenu. The condition is similar to the Material Discontinuity
condition that applies to interior boundaries. The difference is that the Continuity
condition should also be applied when the same material exists on both sides of a pair.
The advantage of using a pair feature is that the mesh does not need to be conforming
on the two sides of the interface (the two parts pf the assembly). This is especially
advantageous for the time explicit discontinuous Galerkin method as the time step
depends on mesh size and local speed of sound.

The condition implements continuity in total pressure and in the normal velocity over
the pair:

p t, src = p t, dst
n src ⋅ u t, src = – n dst ⋅ u t, dst

Subscripts “src” and “dst” in the equation refer to the source and destination in the
pair; that is, the two sides in the pair.

General Flux/Source
The General Flux/Source node, found under the More submenu, adds the most general
boundary condition for the Nonlinear Pressure Acoustics, Time Explicit model, as it is
formulated (in the conservative form) for the discontinuous Galerkin method (dG).
The condition defines the normal pressure and velocity fluxes gp and gu at an exterior
boundary, given by

βp
– n ⋅ Γ p = –  1 + --------2- n ⋅ u = g p
 
ρc
– n ⋅ Γ u = – np = g u

where u = (u, v, w) and n = (nx, ny, nz) in 3D and 2D, or n = (nr, nϕ, nz) in 2D
axisymmetric. Note that this condition is expressed in terms of the dependent variables
and not the total fields. Care should be taken when using this advanced condition as
the method is sensitive to proper formulation of boundary conditions. Only one
characteristic can enter the domain at any boundary at the time, meaning that it is easy
to overspecify the problem. Use the mesh normals (nxmesh, nymesh, and nzmesh; or
nrmesh, nphimesh, and nzmesh) in the expression you define.

624 | CHAPTER 7: ULTRASOUND INTERFACES

GENERAL FLUX/SOURCE
Enter the expression for the components of the Flux vector gp and gu.

General Interior Flux

The General Interior Flux node, found under the More submenu, adds the most general
interior boundary condition for the Nonlinear Pressure Acoustics, Time Explicit
model, as it is formulated (in the conservative form) for the discontinuous Galerkin
method (dG). The condition defines the normal flux g = [gp gu] at an interior
boundary. It should include any necessary numerical flux (as it overrides the Numerical
Flux). Care should be taken when using this advanced condition as the method is
sensitive to proper formulation of boundary conditions.

GENERAL INTERIOR FLUX

Enter the expression for the components of the Flux vector gp and gu.

THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 625

 Modeling with the Convected Wave
Equation Interface
This section contains information about:

• Meshing, Discretization, and Solvers

• Postprocessing: Variables and Quality
• Absorbing Layers
• Stabilizing Physical Instabilities (Filtering)
• Storing Solution on Selections for Large Models
• Assemblies and Pair Conditions

Meshing, Discretization, and Solvers

The Convected Wave Equation, Time Explicit Interface (CWE) is based on the
discontinuous Galerkin method also known as dG-FEM or simply dG. The method is
very memory efficient and is based on a time explicit formulation. This means that it
is not necessary to invert a full system matrix when stepping forward in time. Inversion
of this matrix is necessary in time implicit methods and is very memory consuming for
large problems. Because the CWE interface is not based on the classical FEM
formulation, used in most of the other acoustics interfaces, other strategies apply for
meshing and discretization.

The internal time stepping size of a time explicit method is strictly controlled by the
CFL condition and thus the mesh size. Meaning that the smallest mesh elements will
restrict the time steps (see Optimizing the Mesh for DG below). It turns out that the
dG formulation has a sweet spot for speed and efficiency for wave problems. This is
achieved by using fourth order (quartic) shape functions (the default in the interface)
and a mesh with the element size of about half the wavelength of the highest frequency
component that needs to be resolved. In practice a mesh with size set to λmin/2 to
λmin/1.5 can usually be used.

The interface is supported on all mesh types and for all combinations of mesh. The
internal time step, used by the solver, is automatically calculated based on mesh size
(inscribed circle metric), the wave speed and background mean flow speed, and the
specifics of the solver selected. By default the time explicit solver uses the Runge-Kutta

626 | CHAPTER 7: ULTRASOUND INTERFACES

4th order the method (RK4). This method is good as long as the mesh has a uniform
mesh element size. If a mesh includes a large distribution of mesh element sizes (in
particular in cases with only a few small elements) it can be advantageous to switch to
the Adam-Bashforth 3 (local) method (AB3loc). This method uses intermediate local
time steps in the region where there are small mesh elements. It is not expected that
AB3loc will outperform RK4 for problems with a relatively uniform distribution of
elements. For these cases RK4 is be the best choice.

For general information about the two methods see The Time-Explicit
Solver Algorithms in the COMSOL Multiphysics Reference Manual.

OPTIMIZING THE MESH FOR DG

As mentioned above, the internal time step, used by the solver, is controlled by the
smallest mesh element in the model (relative to the speed of sound in the domain). So
when meshing care should be taken to avoid small elements.

To visualize a metric for the time step used in elements, plot the variable [Link]. The
variable is known as the cell wave time scale and is proportional to the solver time step.
The actual time step depends on the solver chosen. The global minimal value exists as
the variable cwe.wtc_min. When plotting this variable set the Resolution to No
refinement and the Smoothing to None, both settings in the Quality section of the plot.
This can help identify problematic mesh element in the model. To filter out elements
with the large add a Filter subnode to the plot and add a logic expression of the type:
[Link]<1.1*cwe.wtc_min.

Some important meshing and geometry strategies apply:

1 In 3D mesh, always use the Avoid too small elements option in the Element Quality
Optimization settings on the Free Tetrahedral nodes in the mesh. The Optimization
Level can be set to different degrees Basic, Medium, or High. Using this option can
greatly improve the mesh for dG applications and thus speed up the computation
significantly. It is recommended to avoid mesh elements with short edges, since
these are bad for the dG method
2 In 2D mesh, a possible strategy is to create a mapped mesh. This will give a uniform
mesh distribution. This however only works if a mapped mesh can be constructed.
In general the mesh can be optimized by adding an Adapt node after meshing (one
for each region with a given mesh size). In the Adaptation section set Solution to
None, Type of expression to Absolute size, and then in the Size expression enter the
mesh size for the domain. Then set the Maximum number of refinements to 0. This

MODELING WITH THE CONVECTED WAVE EQUATION INTERFACE | 627

 operation will move mesh nodes around to try to maximize the mesh size and try to
remove small mesh elements.
3 Avoid small edges and surfaces in the geometry as these control the mesh. Several
tools exist to remedy this by using the Virtual Operations in the geometry.
4 Curved surfaces and boundaries need to be resolved adequately to ensure numerical
stability. Typically using a Curvature factor of 0.3 to 0.4 will work well. If a curved
boundary is not adequately resolved it may result in too large internal time-steps. At
the same time too small elements should of course be avoided.

For general information about optimizing quality see Element Quality

Optimization in the COMSOL Multiphysics Reference Manual.

For information about virtual operations see Virtual Geometry and Mesh
Control Operations in the COMSOL Multiphysics Reference Manual.

Postprocessing: Variables and Quality

POSTPROCESSING VARIABLES
Several variables are defined to be used when postprocessing. They include the
pressure, velocity components, and the norm of the velocity, as well as material
parameters. The instantaneous intensity also exists as a variable ([Link]). It is defined
as

i = ( ρ 0 u + ρu 0 )  ------ + u ⋅ u 0
p
 ρ0 

Notice that the instantaneous intensity differs from the usual intensity which is an
averaged value.

QUALITY
When analyzing the results from a simulation with the CWE interface, be aware of the
fact that fourth order elements are used to discretize the dependent variables (as
discussed in the section above). This, in some sense, means that within a mesh element,
the shape function has a lot of freedom and can contain a lot of details. These details
are revealed by setting a high Resolution in the Quality section in the plots. The default
plots generated already have that option set. If you add more user defined plots,
remember to set the resolution.

628 | CHAPTER 7: ULTRASOUND INTERFACES

Absorbing Layers
The Convected Wave Equation, Time Explicit Interface includes so-called absorbing
layers, also often referred to as sponge layers. The layers work by combining three
techniques: a scaling system, filtering, and simple nonreflecting conditions. For a
review of the method see, for example, Ref. 5.

The layers are set up by adding the Absorbing Layer under the Definitions node. This
adds a special scaled system. The scaling effectively slows down the propagating waves
and ensures that they “align up” (normal) toward the outer boundary. This means that
the waves will hit the outer boundary in a closer to normal direction. Filtering
attenuates and filters out high frequency components of the wave. The filter
parameters are controlled under the Filter Parameters for Absorbing Layers section at
the main physics level. Finally, at the outer boundary of the layer add a simple Acoustic
Impedance condition, which will work well to remove all remaining waves, as normal
incidence has been ensured.

For more detailed information about the filter see the Filter Parameters
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

The physical thickness of the layer should be such that it contains 2 to 3 mesh
elements. This in practice means that it should roughly have a thickness equal to the
largest wavelength resolved in the model. This is due to the meshing requirements
described in the Meshing, Discretization, and Solvers section. You want to avoid to
create small mesh elements in the absorbing layers. These can potentially slow down
the solution procedure since the internal time stepping is controlled by the mesh size.

For the Absorbing Layers select the Type (Cartesian, cylindrical, spherical, or user
defined) under the Geometry section. Enter values for the Physical Width and Pole
Distance under the Scaling section.

For more details on the Geometry and Scaling see the Infinite Elements,
Perfectly Matched Layers, and Absorbing Layers section in the COMSOL
Multiphysics Reference Manual.

For the layers to work optimally the filter should not be too aggressive. Moreover, the
scaled coordinates in the layer domain should also vary smoothly. To inspect the scaled
system you can, for example, plot the coordinate variables x_absorb_ab1,

MODELING WITH THE CONVECTED WAVE EQUATION INTERFACE | 629

 y_absorb_ab1, and z_absorb_ab1. In 2D axisymmetric they are r_absorb_ab1 and
z_absorb_ab1. Note that the variables are not in the plot menu under the Replace
Expression.

Using the absorbing layers with the three combined techniques will enable the
reduction of spurious reflections by a factor between 100 and 1000 compared to the
incident amplitude.

See the Gaussian Pulse in 2D Uniform Flow: Convected Wave

Equation and Absorbing Layers tutorial model for an example. The
Application Library path
Acoustics_Module/Tutorials/gaussian_pulse_absorbing_layers

Stabilizing Physical Instabilities (Filtering)

In the Convected Wave Equation Model (at the domain level), it is possible to activate
a filter similar to the one used in the absorbing layer. This filter can be used to stabilize
and suppress physical instabilities that can occur in the solution. These are well known
in linearize-Euler-like equations, see Ref. 6. In the presence of a background flow,
vorticity waves can occur; these are propagated with the background flow and are not
acoustic waves.

Storing Solution on Selections for Large Models

Since the CWE interface is suited for solving large acoustic problems (measured in the
number of wavelengths per geometry unit it can handle), the model solved can easily
contain many million degrees of freedom (DOFs). Storing the solution with a desired
time resolution can thus results in huge data files. To circumvent this, a good practice
is to use the Store fields in output option found under the Values of Dependent Variables
section in the main Time Dependent solver (for example, in the Study 1>Step 1:
Time Dependent node). Using the For selections option, it is possible to only store data
on predefined selections. This can, for example, be on a symmetry plane, along an
edge, or at the location of a receiver in a model. When postprocessing the data, best
results are obtained by adding the same selections to the dataset used.

Assemblies and Pair Conditions

When setting up model based on the discontinuous Galerkin time explicit interfaces it
can be advantageous to use assemblies in the geometry and pair conditions in the

630 | CHAPTER 7: ULTRASOUND INTERFACES

physics interface. This allows to use non-conforming meshes at the interface between,
for example, two materials. Different materials are used in the different parts of the
assembly. To couple the parts, at the physics level, the pair boundary conditions should
be used. They are in general available from the Pairs submenu.

For more detailed information about assemblies see the The Form
Union/Assembly Node — Uniting the Geometry section in the
COMSOL Multiphysics Reference Manual.

For more detailed information about the pair features see the Continuity
on Interior Boundaries and Identity and Contact Pairs sections in the
COMSOL Multiphysics Reference Manual.

MODELING WITH THE CONVECTED WAVE EQUATION INTERFACE | 631

 Modeling with the Nonlinear Pressure
Acoustics, Time Explicit Interface
This section contains information about:

• Solving Highly Nonlinear Problems

• Adaptive Mesh Refinement

Solving Highly Nonlinear Problems

When solving nonlinear problems the degree of nonlinearity, the number of harmonics
generated and resolved, as well as the possibility of shock formation needs to be taken
into account.

• For nonlinear problems where only a few harmonics are generated the default
physics and solver settings will work well.
• For highly nonlinear problems (with no shock formation), where more than a few
harmonics are generated, it is recommended to make some changes to the
Time-Explicit Solver. Change the Update time step to Manual, this will ensure that
the local speed of sound is reevaluated and the internal time step is updated to
ensure numerical stability.
• For models that are highly nonlinear and exhibit shock formation, the use of the
WENO limiter is necessary. This functionality is only available for linear
discretization. Change the Element order to Linear in the Discretization section on
the physics interface level. Then, in the Limiter section select WENO. When limiters
are used and the problem is highly nonlinear changes in the Time-Explicit Solver are
also necessary. The computation of discontinuous solutions requires that a Strong
Stability Preserving (SSP) Runge–Kutta method be used. The third order SSP
Runge–Kutta method is achievable by changing the Order of the Runge–Kutta
method from the default 4 to 3. Moreover change the Update time step to Manual,

632 | CHAPTER 7: ULTRASOUND INTERFACES

 this will ensure that the local speed of sound is reevaluated and the internal time step
is update to ensure numerical stability.

For an example of a highly nonlinear problem with shocks see the

Nonlinear Propagation of a Cylindrical Wave — Verification Model
tutorial model. The Application Library path:
Acoustics_Module/Nonlinear_Acoustics/nonlinear_cylindrical_wave

Adaptive Mesh Refinement

One challenge when solving nonlinear problems is that the mesh should potentially
resolve many harmonics to get accurate solutions. To remedy this, adaptive mesh
refinement technology can be used. The method will automatically refine the mesh to
resolve large gradients, that is, sharp signal details that include several harmonics. The
approach is useful for modeling the propagation of spatially localized signals like tone
bursts or Gaussian pulses.

To use the adaptive mesh refinement in a model follow these steps:

• Mesh to resolve only the fundamental frequency in the model.

• In the Time Dependent study step, under the Adaptation section, select Adaptive
mesh refinement. Then generate the default solver sequence.
• Expand the solver tree and go to the Adaptive Mesh Refinement node. Some changes
need to be done here. For the Adaptation method select General modifications. Do
not enable Allow coarsening as the mesh needs to resolve the fundamental frequency.
Finally, update the Error indicator expression to use the pressure gradient norm. In
a 2D axisymmetric model the expression is sqrt([Link]^2+[Link]^2) and
in a 3D model the expression is sqrt([Link]^2+[Link]^2+[Link]^2).
• For highly localized signals, where the mesh adaptation generates a localized region
with small elements, it can also be advantageous to switch from the default
Runge-Kutta (RK4) solver method to the Adams-Bashforth 3 (local) method.

For an example of a nonlinear problem with pulse propagation that uses

mesh adaptation see the High-Intensity Focused Ultrasound (HIFU)
Propagation Through a Tissue Phantom tutorial model. The Application
Library path: Acoustics_Module/Nonlinear_Acoustics/hifu_tissue_sample

MODELING WITH THE NONLINEAR PRESSURE ACOUSTICS, TIME EXPLICIT INTERFACE | 633
 Theory for the Convected Wave
Equation Interface
The Convected Wave Equation, Time Explicit Interface theory is described in this
section:

• Governing Equations of the Convected Wave Equation

• Boundary Conditions
• The Lax–Friedrichs Flux

Governing Equations of the Convected Wave Equation

The governing equations solved in the Convected Wave Equation (CWE) interface
describe the propagation of linear acoustic waves (assuming an adiabatic equation of
state) in the presence of a background flow. These equation are derived by Pierce in
Ref. 1 (section 8.6) and Ref. 2 and are the equations solved by this interface,

∂ρ
------ + ( u 0 ⋅ ∇ )ρ + ( u ⋅ ∇ )ρ 0 + ρ ( ∇ ⋅ u 0 ) + ρ 0 ( ∇ ⋅ u ) = f p
∂t
∂u- 1 ρ
------ + ( u 0 ⋅ ∇ )u + ( u ⋅ ∇ )u 0 + ------ ∇p – -----2- ∇p 0 = f v
∂t ρ0 ρ (7-2)
0
p
ρ = ----2-
c0

Pierce argues for the use of the adiabatic assumption for the acoustic processes
(perturbations in the entropy, s = 0) but also for not retaining the 0th order entropy
variable s0 (background mean flow entropy). The argument is that the entropy s only
varies because of variations in the background fields (it is zero in a homogeneous
medium). This leads to a term in the momentum equation that is second order in
2
gradients of the background field, for example, ( ∂u 0 ⁄ ∂x ) and so forth. These terms
are disregarded. This also means that the equations are not valid when these terms are
large, meaning when gradients in the background fields are large.

In order to fit into the discontinuous Galerkin (dG) formulation, the governing
equations need to be put on a general conservative form of the type

634 | CHAPTER 7: ULTRASOUND INTERFACES

 ∂
d a ----- U + ∇ ⋅ Γ ( U ) = S (7-3)
∂t

where U is the vector containing the dependent variables (p, u), da is the mass matrix
of the system, Γ is the flux matrix, and S is the right-hand-side (RHS) source vector.
The conservative form of Equation 7-2 is derived as follows. Start with Euler’s
equations on a conservative form (omitting the RHS). For now, the dependent
variables represent the full fields (not the acoustic perturbations). The continuity,
momentum, and equation of state can be written

∂ρ
------ + ∇ ⋅ ( ρu ) = 0
∂t
∂ρu T (7-4)
----------- + ∇ ⋅ ( pI + ρuu ) = 0
∂t
ρ = ρ(p)

The equations describe the conservation of mass ρ and momentum flux ρu. Linearize
these equations according to the usual scheme, using

p = p 0 ( x ) + p 1 ( x, t ) u = u 0 ( x ) + u 1 ( x, t ) ρ = ρ 0 ( x ) + ρ 1 ( x, t )

Now, insert these into Equation 7-4 and retain only 1st order terms (the acoustic
perturbations)

∂ρ 1
--------- + ∇ ⋅ ( ρ 1 u 0 + ρ 0 u 1 ) = 0
∂t
∂u 1 ∂ρ 1 T T T
ρ 0 ---------- + u 0 --------- + ∇ ⋅ ( p 1 I + ρ 1 u 0 u 0 + ρ 0 u 1 u 0 + ρ 0 u 0 u 1 ) = 0 (7-5)
∂t ∂t
p1
ρ 1 = -----2-
c0

This is the conservative form of the equations implemented in the CWE interface. In
the remaining of this section, the subscript 1 will be omitted from the acoustic fields.
The subscript 0 is kept on the variables that represent the background mean properties.

Equation 7-5 can now be put on the form given in Equation 7-3 yielding the following
components

THEORY FOR THE CONVECTED WAVE EQUATION INTERFACE | 635

 2
1 ⁄ c0 0 0 0 Γ ( U )p
p
u
2
u0 ⁄ c0 ρ0 0 0 Γ ( U )u
U = da = Γ(U) = (7-6)
v 2 Γ ( U )v
v0 ⁄ c0 0 ρ0 0
w Γ ( U )w
2
w0 ⁄ c0 0 0 ρ0

with u = [u, v, w]T and u0 = [u0, v0, w0]T, and the flux components are

p
Γ ( U ) p = ρ 0 u + ----2- u 0
c0

p
p
Γ ( U ) u = ----2- u 0 u 0 + ρ 0 ( uu 0 + u 0 u ) + 0
c0
0
(7-7)
0
p
Γ ( U ) v = ----2- v 0 u 0 + ρ 0 ( vu 0 + v 0 u ) + p
c0
0

0
p
Γ ( U ) w = ----2- w 0 u 0 + ρ 0 ( wu 0 + w 0 u ) + 0
c0
p

Boundary Conditions
Within the dG formulation all boundary conditions are given in terms of the normal
flux -n·Γ at a given (exterior) boundary. Thus prescribing a boundary condition means
specifying the vector g in the expression

p
ρ 0 ( n ⋅ u ) + ----2- ( n ⋅ u 0 )
c0
p-
---- u ( n ⋅ u0 ) + ρ0 ( u ( n ⋅ u0 ) + u0 ( n ⋅ u ) ) + n1 p
2 0
c0
–n ⋅ Γ = = g (7-8)
p-
---- v ( n ⋅ u0 ) + ρ0 ( v ( n ⋅ u0 ) + v 0 ( n ⋅ u ) ) + n2 p
2 0
c0
p-
---- w ( n ⋅ u0 ) + ρ0 ( w ( n ⋅ u0 ) + w0 ( n ⋅ u ) ) + n3 p
2 0
c0

636 | CHAPTER 7: ULTRASOUND INTERFACES

This very general condition is the one available in the General Flux/Source condition.
Details about the proper formulation of the flux terms entering the vector g can be
found in Ref. 3.

The Lax–Friedrichs Flux

The Lax–Friedrichs (LF) flux parameter used in the definition of the Lax–Friedrichs
numerical flux is given, element wise for each dependent variable, by the expression

c0 + u0
------------------
2
c0
τ∗ = τ LF ( c 0 + u 0 )ρ 0
(7-9)
( c 0 + u 0 )ρ 0
( c 0 + u 0 )ρ 0

where u 0 = u 0 is the norm of the background mean flow velocity. In the CWE
interface we use the local value of the velocity. See Ref. 3 for further details on the
numerical fluxes in general.

The value of the parameter τLF (the parameter that can be edited in the user interface)
controls the numerical flux between the elements (nodal discontinuous Lagrange
elements) used with the dG method. The numerical flux defines how adjacent
elements are connected and how continuous p and u are. Different definitions of the
numerical flux lead to different variants of the dG method. The flux implemented here
is the so-called global Lax–Friedrichs numerical flux. The value of the parameter τLF
should be between 0 and 0.5. For τLF = 0 a so-called central flux is obtained. Setting
τLF = 0.5 gives a maximally dissipative global Lax–Friedrichs flux.

For general information about the numerical flux see the Numerical Flux
section under Wave Form PDE in the COMSOL Multiphysics Reference
Manual.

THEORY FOR THE CONVECTED WAVE EQUATION INTERFACE | 637

 Theory for the Nonlinear Pressure
Acoustics, Time Explicit Interface
The The Nonlinear Pressure Acoustics, Time Explicit Interface theory is described in
this section.

• Governing Equations for Nonlinear Pressure Acoustics, Time Explicit

Governing Equations for Nonlinear Pressure Acoustics, Time Explicit

The Nonlinear Pressure Acoustics, Time Explicit interface solves the nonlinear
continuity equation and momentum equation that represent the physics described by
the Westervelt equation (see Ref. 7). It is derived assuming that the cumulative
nonlinear effects surpass the local nonlinear effects. That is, the contribution of the
Lagrangian energy density is negligible:

2
1 2 pt
L = --- ρ u t – ------------2 = 0
2 2ρc

The governing equations are given by:

1 - ∂p  βp 
-------- --------t + ∇ ⋅  1 + ---------t u t = Q m
2 ∂t 2
ρc  ρc 
∂u t (7-10)
ρ --------- + ∇ ⋅ ( p t I ) = q d
∂t
pt = p ut = u

where pt is the total acoustic pressure, ut is the total acoustic velocity fluctuations, and
β is the coefficient of nonlinearity. Because the problem solved is nonlinear the
superposition principle does not apply which means that the “total” fields are always
equal to the dependent variables.

Equation 7-10 is a nonlinear hyperbolic system written in the conservative form. The
system allows solution discontinuities, and the conservative formulation is the only one
that gives correct discontinuous (shock) solutions, see Ref. 8.

638 | CHAPTER 7: ULTRASOUND INTERFACES

References for the Ultrasound
Interface
1. A.D. Pierce, Acoustics and Introduction to its Physical Principles and
Applications, Acoustical society of America, 1991.

2. A.D. Pierce, “Wave equation for sound in fluids with unsteady inhomogeneous
flow,” J. Acoust. Soc. Am., vol. 87, p. 2293, 1990.

3. J.S. Hesthaven and T. Warburton, Nodal Discontinuous Galerkin Methods:

Algorithms, Analysis, and Applications, Springer, 2008.

4. M.K. Myers, “On the Acoustic Boundary Condition in the Presence of Flow,”
J. Sound Vibration, vol. 73, pp. 429–434, 1980.

5. P.G. Petropoulos, L. Zhao, and A.C. Cangellaris, “A Reflectionless Sponge Layer

Absorbing Boundary Condition for the Solution of Maxwell’s Equations with
High-Order Staggered Finite Difference Schemes,” J. Comp. Phys., vol. 139,
pp. 184–208, 1998.

6. C.W. Tam, “Computational Aeroacoustics: Issues and Methods,” AIAA Journal,

vol. 33, 1995.

7. M.D. Diaz, M.A. Solovchuk, and T.W.H. Sheu, “A conservative numerical scheme
for modeling nonlinear acoustic propagation in thermoviscous homogeneous media,”
J. Comp. Phys., vol. 363, pp. 200–230, 2018.

8. E.F. Toro, Riemann Solvers and Numerical Methods for Fluid Dynamics. A
Practical Introduction, 3rd Ed., Springer, 2009.

REFERENCES FOR THE ULTRASOUND INTERFACE | 639

640 | CHAPTER 7: ULTRASOUND INTERFACES
 8

Geometrical Acoustics Interfaces

This chapter describes the physics interfaces found under the

Acoustics > Geometrical Acoustics branch ( ).

• The Ray Acoustics Interface

• Modeling with the Ray Acoustics Interface
• Theory for the Ray Acoustics Interface
• The Acoustic Diffusion Equation Interface
• Modeling with the Acoustic Diffusion Equation Interface
• Theory for the Acoustic Diffusion Equation Interface

641
 The Ray Acoustics Interface
The Ray Acoustics (rac) interface ( ), found under the Acoustics>Geometrical
Acoustics subbranch ( ) when adding a physics interface, is used to compute the
trajectories, phase, and intensity of acoustic rays. Ray acoustics is valid in the
high-frequency limit where the acoustic wavelength is much smaller than the
characteristic geometric features. The interface can be used to model room acoustics
in concert halls, theaters, class rooms, or car cabins, but also for underwater acoustics,
atmospheric acoustics, and other outdoor environments.

The properties of the media in which the rays propagate can be constant within
domains, change continuously within domains (graded media) or discontinuously at
boundaries. At exterior boundaries it is possible to assign a variety of wall conditions,
including combinations of specular and diffuse reflection. Impedance and absorption
can depend on the frequency, intensity, and direction of incident rays. Transmission
and reflection are also modeled at material discontinuities. A background velocity may
also be assigned to any medium to, for example, model the effects of wind in
atmospheric acoustics.

• For a best practice tutorial on room acoustics and impulse response

computation, see the Small Concert Hall Acoustics. The Application
Library path: Acoustics_Module/Building_and_Room_Acoustics/
small_concert_hall
• For a model of an actual chamber music hall model located in Berlin
with comparison to measurements, see the Chamber Music Hall. The
Application Library path:
Acoustics_Module/Building_and_Room_Acoustics/chamber_music_hall
• For an underwater acoustics tutorial, see the Underwater Ray
Tracing Tutorial in a 2D Axisymmetric Geometry tutorial in the
COMSOL Application Gallery:
[Link]

For room acoustics applications the impulse response can be determined in

postprocessing by combining the information of the simulated rays collected by a
Receiver dataset and postprocess using the dedicated Impulse Response plot. The energy
and level decay curves as well as room acoustic metrics can be computed with the

642 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Energy Decay subfeature. For details see the Impulse Response Plot and Receiver
Dataset section in Modeling with the Ray Acoustics Interface.

When this physics interface is added, these default nodes are also added to the Model
Builder: Medium Properties, Wall, Material Discontinuity, and Ray Properties. Then, from
the Physics toolbar, add other nodes that implement, for example, boundary
conditions. You can also right-click Ray Acoustics to select physics features from the
context menu.

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is rac.

RAY RELEASE AND PROPAGATION

By default, the Allow frequency distributions at release features check box is cleared.
Select this check box to model propagation of rays of different frequencies
simultaneously. The ray frequency can be specified at release features by entering a
value or expression, sampling the frequency from a distribution, or entering a list of
values. If this check box is cleared, the frequency is the same for all rays and is specified
in the settings for the Ray Properties node.

Regardless of whether the Allow frequency distributions at release features check box is
selected, properties at walls can depend on the frequency (variable rac.f) of each ray
released. Domain material properties, like the speed of sound, can also depend on the
frequency. Entering a frequency distribution of the released rays also enables emitting
more complex source signals (decomposed in their Fourier components).

To define the speed of sound or another medium property as a function

of ray properties such as the ray frequency or intensity, the ray variable
must be enclosed in the noenv() operator. For example, to use the ray
frequency rac.f in an expression for the speed of sound, it must be
included as part of the expression noenv(rac.f). See also Medium
Properties.

THE RAY ACOUSTICS INTERFACE | 643

 The Maximum number of secondary rays prevents an inordinate number of rays from
being generated by capping them at the number supplied in the text field. The default
is 500. Rather than being produced directly by release features such as the Release from
Grid node, secondary rays are released when an existing ray is subjected to certain
boundary conditions. For example, when a ray undergoes refraction at a Material
Discontinuity between different media, the incident ray is refracted and a reflected ray
is created; the degrees of freedom for this reflected ray are taken from one of the
available secondary rays, which are preallocated when the study begins.

If an insufficient number of secondary rays are preallocated, a reflected ray may not be
released when an existing ray undergoes refraction. Thus, the Maximum number of
secondary rays should be large enough that all reflected rays which significantly affect
the solution can be released.

MATERIAL PROPERTIES OF EXTERIOR AND UNMESHED DOMAINS

Enter a value or expression for the Speed of sound cext (SI unit: m/s). The default value
is 343[m/s]. Then enter a value or expression for the Density ρext (SI unit: kg/m3).
The default value is 1.2[kg/m^3].

When one of the Intensity Computation options is selected (see below), select an option
from the Attenuation type list:

• For Attenuation coefficient, Np per unit length (the default), enter a value or
expression for the Attenuation coefficient αext (SI unit: Np/m). The default is 0.
• For Intensity attenuation coefficient, Np per unit length, enter a value or expression
for the Intensity attenuation coefficient mext (SI unit: Np/m). The default is 0.
• For Attenuation coefficient, dB per unit length, enter a value or expression for the
Attenuation coefficient α’ext (SI unit: dB/m). The default is 0.
• For Attenuation coefficient, dB per wavelength, enter a value or expression for the
( λ ) (SI unit: dB). The default is 0.
Attenuation coefficient α ext
A more detailed explanation of each definition of the attenuation coefficient is given
in the Medium Properties section.

These material properties are used when tracing rays outside the geometry or when
tracing rays through domains that are not included in the selection of the Ray
Acoustics physics interface. That is, meshless ray tracing. The medium properties in the
deselected and exterior void domains must be constant and cannot be functions of any
field variables, such as temperature. The properties like the attenuation can depend on
a parameter like the band center frequency when modeling room acoustics
applications, see the Impulse Response Plot and Receiver Dataset section.

644 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

INTENSITY COMPUTATION
Select an option from the Intensity computation list: None (the default), Compute
intensity, Compute power, Compute intensity and power, Compute intensity in graded
media, or Compute intensity and power in graded media. For None the ray intensity and
power is not computed.

• For Compute intensity, auxiliary dependent variables are used to compute the
intensity of each ray. For a complete list of the auxiliary dependent variables that are
defined, see Intensity and Wavefront Curvature in Theory for the Ray Acoustics
Interface. This option is more accurate and is usually less computationally
demanding than Compute intensity in graded media, but is only valid for computing
intensity in homogeneous (constant speed of sound) media.
• For Compute power, the total power transmitted by each ray is defined as an auxiliary
dependent variable, in addition to the auxiliary dependent variables that are declared
when Compute intensity is selected. Like intensity, the power is affected by absorbing
media and by reflection or refraction at boundaries. Unlike intensity, ray power does
not change due to the convergence or divergence of a bundle of rays. The Sound
Pressure Level Calculation subnode is available for the Wall feature. The Compute
power option is necessary for the impulse response postprocessing, see Impulse
Response Plot and Receiver Dataset.
• The option Compute intensity and power combines the dependent variables created
by the Compute intensity and Compute power options.
• For Compute intensity in graded media, auxiliary dependent variables are used to
compute the intensity of each ray. The intensity is affected by gradients in the speed
of sound. Graded media may be present, for example, when modeling the effect of
the saline climb in underwater acoustics. For a complete list of the auxiliary
dependent variables that are defined, see Intensity and Wavefront Curvature in
Theory for the Ray Acoustics Interface. This intensity computation method is valid
for both homogeneous and graded media. If all media are homogeneous, meaning
that the gradient of the speed of sound is zero everywhere except at material
discontinuities, then it is recommended to select Compute intensity instead, since it
is the more accurate intensity computation method for such cases.
• For Compute intensity and power in graded media, the total power transmitted by
each ray is defined as an auxiliary dependent variable, in addition to the auxiliary
dependent variables that are declared when Compute intensity in graded media is
selected. The Sound Pressure Level Calculation subnode is available for the Wall
feature. If all media are homogeneous then it is recommended to select Compute
intensity and power instead.

THE RAY ACOUSTICS INTERFACE | 645

 When the ray intensity or power is solved for (one of the options Compute intensity,
Compute power, Compute intensity and power, Compute intensity in graded media, or
Compute intensity and power in graded media is selected), a list for the Reference pressure
for the sound pressure level appears. Select Use reference pressure for air (the default) to
set the reference pressure to 20 μPa, Use reference pressure for water to set the
reference pressure to 1 μPa, or User-defined reference pressure to enter a value or
expression in the text field. The reference pressure is used to calculate the sound
pressure level (variable [Link]) and sound intensity level (variable [Link]) along rays.

If the ray intensity is solved for (one of the options Compute intensity, Compute intensity
and power, Compute intensity in graded media, or Compute intensity and power in graded
media is selected), the Compute phase check box is available. By default, the Compute
phase check box is cleared. Select this check box to compute the phase along each ray.
An auxiliary dependent variable for the phase [Link] is created. The phase changes
along the ray trajectory based on its frequency and is also subjected to phase shifts that
happen at walls with specular reflection and absorption.

When Compute intensity in graded media or Compute intensity and power in graded
media is selected from the Intensity computation list, enter a Tolerance for curvature
tensor computation (dimensionless). This tolerance is used internally when computing
the principal radii of curvature in a graded medium, and is used to define a criterion
by which both principal radii of curvature are declared approximately equal. A larger
tolerance makes the solution less accurate but more stable.

ADDITIONAL VARIABLES
Select the Count reflections check box to allocate an auxiliary dependent variable for the
number of reflections undergone by each ray, including reflections by the Wall and
Material Discontinuity features. The auxiliary variable begins at 0 when rays are
released and is incremented by 1 every time a ray is reflected at a boundary. The Count
reflections option is necessary for the impulse response postprocessing, see Impulse
Response Plot and Receiver Dataset.

By default, the Store ray status data check box is cleared. Select this check box to add
new variables for quantities that cannot necessarily be recovered from the ray trajectory
data alone. This is especially true if automatic remeshing has been used in a model. The
following variables are created:

• The release time of a given ray (variable name [Link]).

646 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

• The time at which a ray stopped at a boundary (variable name [Link]).
• The final status of the ray (variable name [Link]). This indicates the status of a ray
at the final time step. The value is an integer which has one of the following values:

- 0 for unreleased rays

- 1 for rays that are still in the modeling domain
- 2 for frozen rays
- 3 for stuck rays
- 4 for rays that have disappeared.

To summarize the total number of rays having each final status, the following global
variables are also defined.
TABLE 8-1: GLOBAL VARIABLES BASED ON RAY STATUS

NAME DESCRIPTION

fac Fraction of active rays at final time step

fds Fraction of disappeared rays at final time step
ffr Fraction of frozen rays at final time step
fse Fraction of secondary rays released
fst Fraction of stuck rays at final time step
nsr Number of released secondary rays
nsrf Number of released secondary rays at final time step
nsu Number of unreleased secondary rays
nsuf Number of unreleased secondary rays at final time step

The global variable names in Table 8-1 all take the unreleased secondary rays into
account. For example, suppose an instance of the Ray Acoustics interface includes 100
primary rays and 100 allocated secondary rays. At the last time step, suppose that 80
of the primary rays have disappeared at boundaries and that 40 secondary rays have
been emitted, all of which are still active. Then the variable [Link], the fraction of
active rays at the final time step, would have the value (20 + 40)/(100 + 100) or 0.3.

ADVANCED SETTINGS
This section is only shown when Advanced Physics Options are enabled (click the Show
More Options button in the Model Builder toolbar).

The Wall accuracy order sets the accuracy order of the time stepping used for time steps
during which a ray-wall interaction happens. Select an order of 1 to use a forward Euler
step and compute the motion both before and after the wall collision. Select an order

THE RAY ACOUSTICS INTERFACE | 647

 of 2 (the default) to use a second-order Taylor method and compute the motion before
the wall collision. After the collision a second-order Runge-Kutta method is used.

Select an option from the Arguments for random number generation list: Generate unique
arguments, Generate random arguments, or User defined. This setting determines how
the additional argument to random functions is defined in features such as the Wall
boundary condition with the Diffuse scattering wall condition. Typically, the random
numbers are functions of the ray index, position, time, and another argument i, which
is defined as follows:

• For Generate unique arguments, the additional argument is based on the position of
each node in the Model Builder. As a result, random numbers generated in different
nodes are created independently of each other, but the same result can be
reproduced by running the same study several times.
• For Generate random arguments, the additional argument is randomly created,
causing the random functions to return different results each time the study is run.
• For User defined, the additional argument is defined by a user input in the Settings
window each feature. Independent solutions can be obtained by running a
parametric sweep for different values of i.

The Allow multiple release times check box, which is cleared by default, allows an array
of release times for the rays to be specified in any of the ray release features. If the check
box is cleared, all rays are released at time t = 0.

By default the Exclude Jacobian contributions from ray equations check box is cleared. If
this check box is selected, then expressions for the time derivatives of the degrees of
freedom on rays are not included in the Jacobian matrix. Excluding contributions to
the Jacobian can significantly improve solution time. The drawback is that the Jacobian
is not exact, so smaller time steps or path length intervals may be needed to obtain an
accurate solution. This drawback is most noticeable when computing the ray intensity
or power in attenuating media.

By default, the Treat unmeshed releasing entities as errors check box is selected. Certain
ray release features, such as the Release and Release from Boundary features, compute
the ray release positions based on a geometric entity, which must be meshed. While this
check box is selected, failure to mesh the selections of such features will result in an
error message when running any study in the model. If this check box is cleared, failure
to mesh the selections of such features will result in them releasing zero rays, but will
otherwise not interrupt the computation. Other physics features will still be able to
release rays normally.

648 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Enter a value for the Maximum number of wall interactions per time step. The default
value is 1000. If a ray undergoes more than the specified number of boundary
interactions in a single time step taken by the solver, the ray will disappear. This is
included as a safeguard to prevent rays from getting stuck in infinite loops if the time
between successive ray-wall interactions becomes infinitesimally small.

DEPENDENT VARIABLES
The dependent variables (field variables) are the Ray position components and Wave
number components. The name can be changed but the names of fields and dependent
variables must be unique within a model.

Theory for the Ray Acoustics Interface

Domain, Boundary, and Global Nodes for the Ray Acoustics Interface
The Ray Acoustics Interface has these domain, boundary, and global nodes available
(listed in alphabetical order).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

THE RAY ACOUSTICS INTERFACE | 649

 • Accumulator (Boundary) • Ray Termination
• Accumulator (Domain) • Release
• Auxiliary Dependent Variable • Release from Data File
• Axial Symmetry • Release from Edge
• Background Velocity • Release from Exterior Field
• Release from Boundary Calculation

• Release from Symmetry Axis • Release from Grid

• Material Discontinuity • Release from Grid on Axis

• Medium Properties • Release from Point

• Nonlocal Accumulator • Release from Point on Axis

• Ray Continuity • Sound Pressure Level Calculation

• Ray Detector • Source with Directivity

• Ray Properties • Wall

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Medium Properties
Use the Medium Properties node to specify the speed of sound and density of the
medium. This node is created by default and includes all of the selected domains for
the physics interface.

PRESSURE ACOUSTICS MODEL

If the ray intensity or power is computed in the model, select an option from the Fluid
model list: Linear elastic (the default), User-defined attenuation, Atmosphere attenuation,
Ocean attenuation, or Thermally conducting and viscous. This section determines which
additional fluid properties can be specified in this section.

The Speed of sound c (Si unit: m/s) is always required. By default it uses values From
material. For User defined enter a value or expression. The default is 343 m/s.

650 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

If the ray intensity or power is solved for, both the density and the speed of sound are
required to compute the transmittance and reflectance at boundaries between different
media. The default Density ρ (SI unit: kg/m3) uses values From material. For User
defined, enter a value or expression. The default is 1.2 kg/m3.

Additional inputs are shown in this section depending on the option selected from the
Fluid model list:

• If Linear elastic is selected, absorption by the medium is neglected, so no other

inputs are required.
• If User-defined attenuation is selected, select an option from the Attenuation type list.
The default is Attenuation coefficient, Np per unit length.
- For Attenuation coefficient, Np per unit length enter a value or expression for the
Attenuation coefficient α (SI unit: Np/m). The default is 0. Note that this is the
attenuation coefficient of the pressure amplitude rather than the intensity. The
ray intensity (neglecting wavefront curvature terms) actually scales as
I(x) = I0exp(−2αx), where x is the propagation distance.
- For Intensity attenuation coefficient, Np per unit length enter a value or expression
for the Intensity attenuation coefficient m (SI unit: Np/m). The default is 0.
Neglecting curvature terms, the ray intensity decreases as I(x) = I0exp(−mx).

In most room acoustics references the attenuation coefficient given is the

intensity attenuation coefficient m. When setting up models it is
important to know which definition is used and choose between the
preceding two options accordingly.

- For Attenuation coefficient, dB per unit length, enter a value or expression for the
Attenuation coefficient α’ (SI unit: dB/m). The default is 0.
- For Attenuation coefficient, dB per wavelength, enter a value or expression for the
Attenuation coefficient α(λ) (SI unit: dB). The default is 0.

• For Atmosphere attenuation enter the following in the Model Input section:
- Temperature T (SI unit: K, default 293.15 K)
- Absolute pressure pA (SI unit: Pa, default 1 atm)
- Relative humidity φw (dimensionless, default 0)
By default, these model inputs are all User defined, but they can also be taken from
a Common model input.

THE RAY ACOUSTICS INTERFACE | 651

 • For Ocean attenuation enter the following in the Model Input section:
- Temperature T (SI unit: K, default 293.15 K)
- Practical salinity Sp (dimensionless, default 35)
- Depth D (SI unit: m, default 0)
By default, these model inputs are all User defined, but they can also be taken from
a Common model input. Also, in the Pressure Acoustics Model section, enter a value or
expression for the pH value pH (dimensionless). The default is 8.

The Practical salinity is defined on the Practical Salinity Scale. It represents

a specific way of measuring salinity through electric conductivity. Practical
salinity is a unitless quantity, although it can be thought of as given in
units of g/kg. The default value for the Practical salinity Sp is 35. For
details, see Ref. 22.

• If Thermally conducting and viscous is selected, specify the following material

properties. By default, all fluid properties take default values From material. Choose
User defined to enter a value or expression for the:
- Density ρ (SI unit: kg/m3). The default is 1.2 kg/m3.
- Heat capacity at constant pressure Cp (SI unit: J/(kg·K)). The default value is
1005.4 J/(kg·K).
- Ratio of specific heats γ (dimensionless). The default value is 1.4.
- Thermal conductivity k (SI unit: W/(m·K)). The default value is
0.0257 W/(m·K). For ray acoustics, the thermal conductivity is assumed to be
isotropic.
- Dynamic viscosity μ (SI unit: Pa·s). The default value is 0.0181 mPa·s.
- Bulk viscosity μB (SI unit: Pa·s). The default is 0.

To define the attenuation coefficient or another medium property as a function of ray

properties such as the ray frequency or intensity, the ray variable must be enclosed in
the noenv() operator. For example, to use the ray frequency rac.f in an expression
for the speed of sound, it must be included as part of the expression noenv(rac.f).

Attenuation Within Domains

652 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Wall
Use the Wall node to determine what happens to the rays when contact with a
boundary is made. The Wall node is the default boundary condition on all exterior
boundaries.

The Accumulator (Boundary) subnode is available from the context menu (right-click
the parent node) or from the Physics toolbar, Attributes menu.

WALL CONDITION
Select a Wall condition: Freeze (the default), Specular Reflection, Stick, Disappear, Pass
through, Diffuse scattering, Isotropic scattering, Mixed diffuse and specular reflection, or
General reflection.

TABLE 8-2: WALL CONDITION OPTIONS

OPTIONS DESCRIPTION

Freeze Select to fix the ray position and wave vector at the instant a wall is
struck. So, the ray position no longer changes after contact with the
wall and the wave vector of the ray remains at the same value as when
the ray struck the wall. This boundary condition is typically used to
recover the ray intensity or phase at the instant contact was made with
the wall.
Specular Select to specularly reflect from the wall, so that the angle of incidence
Reflection and the angle of reflection are equal.
Stick Select to fix the ray position at the instant the wall is struck. The wave
vector components are set to zero.
Disappear This option means that the ray is not displayed once it has made
contact with the wall. Use it if the ray location after contact with the
wall is not of interest.
Pass through This option allows rays to cross the boundary unimpeded. It does not
cause rays to be refracted at the boundary between different materials,
though; to model refraction, instead use the Material Discontinuity
feature.
Diffuse Select to reflect rays at a wall according to Lambert’s cosine law. That
scattering is, the probability of a reflected ray propagating in a given direction
within a solid angle dω is given by cos(θ)dω where θ is the angle
between the direction of the ray and the wall normal.
Isotropic Select to reflect rays at a wall isotropically in random directions.
scattering

THE RAY ACOUSTICS INTERFACE | 653

 TABLE 8-2: WALL CONDITION OPTIONS

OPTIONS DESCRIPTION

Mixed Select to reflect rays at a wall either specularly or according to

diffuse and Lambert’s cosine law, based on a user-defined probability.
specular
reflection
General Select to allow an arbitrary direction vector to be specified after a ray
reflection makes contact with the wall. The ray direction vector components can
be functions of the incident ray direction, phase, or any other quantity.

GENERAL REFLECTION SETTINGS

This section is available when General reflection (see Table 8-2) is selected as the Wall
condition.

Enter values for the Reflected ray direction vector Lp (dimensionless) either in Cartesian
coordinates (x, y, z) (the default) or select the Specify tangential and normal direction
vector components check box to enter coordinates in the tangent-normal coordinate
system (t1, t2, n). In this case, the normal direction is selected so that an incident ray
is reflected back into the domain it previously occupied if the specified normal
direction vector component is positive. The tangential directions are oriented so that
they form a right-handed coordinate system, together with the normal direction.

PRIMARY RAY CONDITION

Use this section to apply two different wall conditions on the same set of boundaries
based on a probability or logical expression.

Select a Primary ray condition: None (the default), Probability, or Expression. When the
default, None, is kept, it means that the Wall condition is always respected by the
incident rays.

Probability
If Probability is selected, the Wall condition is applied with a certain probability. Enter
a value for the Probability, γ (dimensionless). If the Wall condition is not used, the ray
instead behaves according to the Otherwise setting.

The value of γ should always be between 0 and 1.

654 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

For example, if the Wall condition is set to:

• Freeze and γ is set to 0.1, then for every 10 rays that strike the wall, on average one
freezes and the remaining 9 rays behave according to the Otherwise setting.
• Stick and γ is set to 0.5 then on average half of the rays stick to the wall and the other
half behave according to the Otherwise setting.

Expression
If Expression is selected, the Evaluation expression e (dimensionless) is evaluated
whenever a ray strikes the wall. The default expression is 1. If the Evaluation expression
is nonzero, the ray behaves according to the Wall condition, otherwise the ray behaves
according to the Otherwise setting.

Otherwise
The options available for the Otherwise setting are the same as for the Wall Condition,
except that General reflection and Mixed diffuse and specular reflection are not available.
The Otherwise setting can be used to make rays interact with a wall differently with a
certain probability or when a certain condition is satisfied. For example, to model
reflection at a partially specular surface in which 50% of the ray intensity is absorbed
and 30% is reflected specularly:

• Select Mixed diffuse and specular reflection as the Wall condition.

• Set the Probability of specular reflection to 0.3/(1-0.5).
• Set the Primary ray condition to Probability.
• Set the Probability to 1-0.5.
• Select Freeze as the Otherwise option.

PHASE SHIFT
This section is available when:

• The ray intensity is solved for in the model.

• The Compute phase check box is selected under the physics interface Advanced
settings section.
• The Wall condition is set to Specular Reflection, Diffuse scattering, Isotropic scattering,
General reflection, or Mixed diffuse and specular reflection.

Select the Apply manual phase shift check box to specify the phase shift applied to
reflected rays. By default this check box is cleared, causing the phase shift be derived
from the reflection coefficient.

THE RAY ACOUSTICS INTERFACE | 655

 If the Apply manual phase shift check box is selected, enter a Phase shift ΔΨ (SI unit:
rad). The default is 0.

If the Wall condition is set to Mixed diffuse and specular reflection, this section is instead
called Phase shifts and the phase shifts for the diffusely and specularly reflected rays are
specified separately. All of the text fields are given subscripts s or d for specularly and
diffusely reflected rays, respectively.

ABSORPTION COEFFICIENT
This section is available when the ray intensity or power is solved for in the model, and
the Wall condition is set to any option that can reflect rays, including the following:

• Specular Reflection,
• Diffuse scattering,
• Isotropic scattering,
• General reflection, or
• Mixed diffuse and specular reflection.
Select one of the following from the Compute reflected intensity using list: Absorption
coefficients, Absorber, specified impedance, Reflection coefficients, Fluid-fluid interface,
Fluid-solid interface, or Layered fluid half space.

• For Absorption coefficients enter the Absorption coefficient α (dimensionless). The

default is 0. The intensity or power of the reflected ray will be proportional to 1 − α.
• For Absorber, specified impedance enter the Characteristic impedance of absorber Z1
(SI unit: Pa·s/m). The default value is 1 GPa·s/m, corresponding to a hard wall.

656 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

• For Reflection coefficients enter the Reflection coefficient R (dimensionless). The
default value is 1. The intensity or power of the reflected ray will be proportional to
R2.
• For Fluid-fluid interface enter the following:
- Speed of sound, adjacent fluid c1 (SI unit: m/s). The default is 1500 m/s.
- Density, adjacent fluid ρ1 (SI unit: kg/m3). The default is 1000 kg/m3.
- Attenuation coefficient, adjacent fluid α1 (SI unit: 1/m). The default is 0.

• For Fluid-solid interface enter the following:

- Compressional speed of sound, adjacent solid cp,1 (SI unit: m/s). The default is
1500 m/s.
- Shear speed of sound, adjacent solid cs,1 (SI unit: m/s). The default is 1500 m/s.
- Density, adjacent solid ρ1 (SI unit: kg/m3). The default is 1000 kg/m3.
- Compressional attenuation coefficient, adjacent solid αp,1 (SI unit: 1/m). The
default is 0.
- Shear attenuation coefficient, adjacent solid αs,1 (SI unit: 1/m). The default is 0.

• For Layered fluid half-space enter the following:

- Thickness, adjacent fluid layer h (SI unit: m). The default is 1 μm.
- Speed of sound, adjacent fluid layer c1 (SI unit: m/s). The default is 1500 m/s.
- Density, adjacent fluid layer ρ1 (SI unit: kg/m3). The default is 1000 kg/m3.
- Attenuation coefficient, adjacent fluid layer α1 (SI unit: 1/m). The default is 0.
- Speed of sound, adjacent fluid domain c2 (SI unit: m/s). The default is 1500 m/s.
- Density, adjacent fluid domain ρ2 (SI unit: kg/m3). The default is 1000 kg/m3.
- Attenuation coefficient, adjacent fluid domain α2 (SI unit: 1/m). The default is 0.

If the Wall condition is set to Mixed diffuse and specular reflection, this section is instead
called Absorption Coefficients and the absorption coefficients for the diffusely and

THE RAY ACOUSTICS INTERFACE | 657

 specularly reflected rays are specified separately. All of the text fields are given
subscripts s and d for specularly and diffusely reflected rays, respectively.

To create detailed models the wall properties, for example, the surface
normal impedance or the absorption coefficient, must depend on the
properties of the incident ray. Doing this is possible as all wall properties
may depend on:

• The frequency of the ray rac.f

• The direction of the ray, the acute angle of incidence [Link]
• Even the intensity of the ray rac.I

SURFACE ROUGHNESS
This section is available when

• the ray intensity or power is solved for in the model, and

• the Wall condition is set to Specular Reflection, General reflection, or Mixed diffuse and
specular reflection. For Mixed diffuse and specular reflection the roughness only
applies to the specularly reflected rays.

Select an option from the Surface roughness model list: None (the default) or Rayleigh
roughness. If Rayleigh roughness is selected, enter the RMS roughness σ (SI unit: m). The
default is 0. Use this setting to reduce the intensity of rays that are reflected by a rough
wall.

NEW VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model.

When a ray crosses or touches a boundary, the values of the user-defined auxiliary
dependent variables can be changed. The value can be a function of any combinations
of ray variables and variables defined on the boundary.

Select the Assign new value to auxiliary variable check box or boxes based on the
number of auxiliary variables in the model. Then enter the new value or expression in
the field. For example, if there is an auxiliary variable, psi, then enter a value for psinew
in the field. So, to increment the value of psi by 1 when a ray touches or crosses a
boundary, enter psi+1 in the text field for psinew.

658 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

ADVANCED SETTINGS
If the Primary ray condition is set to Probability, or if the Diffuse scattering, Isotropic
scattering, or Mixed diffuse and specular reflection wall condition is used, then the Wall
feature generates random numbers.

If, in addition, the Arguments for random number generation setting is set to User defined
in the physics interface Advanced Settings section, the Advanced Settings section is
available.

Enter the Additional input argument to random number generator i (dimensionless). The
default value is 1.

Axial Symmetry
The Axial Symmetry node is automatically added to 2D axisymmetric models. The
options available for the feature are the same as for Wall, except as follows. The
Boundary Selection is locked and is only applicable to edges on the symmetry axis. The
only wall conditions available are Specular reflection (the default), Freeze, Stick, and
Disappear. In addition, the Primary Ray Condition section is not included.

The Axial Symmetry condition only applies to straight boundaries that are
both on the axis of symmetry and adjacent to a domain in the selection of
the physics interface. If rays are propagating in the void region outside of
the selected domains, it is possible for the rays to enter the region where
r < 0, even though no geometry is defined there.

To ensure that rays cannot propagate into the region where r < 0, draw
line segments on the parts of the symmetry axis where no geometric
entities exist, then add the Wall node to these edges. Keep in mind that
rays can only be subjected to boundary conditions if the corresponding
boundaries are meshed.

Accumulator (Boundary)
The Accumulator subnode is available from the context menu (right-click the Wall or
Material Discontinuity parent node) or from the Physics toolbar, Attributes menu. Each
Accumulator subnode defines a variable, called the accumulated variable, on each

THE RAY ACOUSTICS INTERFACE | 659

 boundary element in the selection of the parent node. The accumulated variables are
incremented when rays interact with the boundaries where they are defined.

ACCUMULATOR SETTINGS
Select an option from the Accumulator type list: Density (default) or Count.

• For Density the accumulated variable is divided by the surface area of the boundary
element where it is defined.
• For Count the accumulated variable is the sum of the source terms of all rays that hit
the boundary element, and is unaffected by the boundary element size.

Select an option from the Accumulate over list: Ray-wall interactions (the default) or
Rays in boundary elements.

• For Ray-wall interactions the accumulated variable is affected by all rays that hit the
boundary element.
• For Rays in boundary elements the accumulated variable is only affected by rays that
freeze or stick to the boundary element.

Enter the Accumulated variable name. The default is rpb. The accumulated variable is
defined as <scope>.<name>, where <scope> includes the name of the physics
interface node, parent boundary condition, and the Accumulator node, and <name> is
the accumulated variable name.

For example, if the Accumulator subnode is added to a Wall node in an instance of the
Ray Acoustics interface using the default variable name rpb, the accumulated variable
name might be [Link].

Enter a Source R. The unit of the source term depends on the settings in the Units
section. Whenever a ray collides with a boundary element in the selection of the parent
node, the accumulated variable in that element is incremented by the source term. If
the Accumulator type is set to Density, the source term is divided by the area of the
boundary element (in 3D) or the length of the boundary element (in 2D).

For example, if the source is 1 and Count is selected from the Accumulator type list, a
variable is created to store the number of times each boundary element is hit by a ray.

UNITS
Select a Dependent variable quantity from the list; the default is Dimensionless [1]. To
enter a unit, select None from the list and in the Unit field enter a value, for example, K,
m/s, or mol/m^3.

660 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

SMOOTHING
The accumulated variables are computed using discontinuous shape functions. Select
the Compute smoothed accumulated variable check box to compute a smoothed
accumulated variable by computing the average value of the variable within a sphere of
a user-defined radius. Then enter a Smoothing radius r (SI unit: m). The default is 1 m.

Accumulator Theory: Boundaries

Material Discontinuity
The Material Discontinuity node is the default feature on interior boundaries.

The Ray Acoustics interface always applies reflection and refraction at boundaries
between different media using a deterministic ray splitting approach. The direction of
the refracted ray is computed using Snell’s law, based on the speed of sound on either
side. If extra degrees of freedom have been allocated for secondary rays, a reflected ray
is also released. If the incident ray undergoes total internal reflection, no refracted ray
is produced and no secondary rays are needed to release the reflected ray.

If the ray intensity or power is solved for in the model, the Material Discontinuity feature
computes the new values of these variables for the reflected and refracted rays.

The maximum number of reflected rays can be controlled via the

Maximum number of secondary rays text field, which is found under the Ray
Release and Propagation section for the physics interface.

The Accumulator (Boundary) subnode is available from the context menu (right-click
the parent node) or from the Physics toolbar, Attributes menu.

RAYS TO RELEASE
Use the inputs in this section to control whether the incident ray is split into a reflected
and refracted ray at the material discontinuity. Select an option from the Release
reflected rays list: Always (the default), Never, or Based on logical expression.

• If Always is selected, whenever a ray hits the material discontinuity, the incident ray
is refracted across the boundary, and a reflected ray is produces using some of the
preallocated degrees of freedom for secondary rays.

THE RAY ACOUSTICS INTERFACE | 661

 • If Never is selected, the incident ray is refracted, but no reflected ray is produced.
This does not prevent a ray from undergoing total internal reflection at the
boundary.
• If Based on logical expression is selected, enter a value or expression for the Evaluation
expression e (dimensionless). The default is 1. A reflected ray is produced only if the
value of this expression is nonzero for the incident ray.

If the ray intensity is solved for in the model, enter a Threshold intensity Ith
(SI unit: W/m2). The default is 1 mW/m2. If the interaction of a ray with a material
discontinuity would create a reflected ray of intensity less than the threshold intensity,
the release of this reflected ray is suppressed.

If the ray power is solved for in the model, enter a Threshold power Qth (SI unit: W).
The default is 10-3 mW. If the interaction of a ray with a material discontinuity would
create a reflected ray of power less than the threshold power, the release of this
reflected ray is suppressed.

If both the Evaluation expression and the Threshold intensity are specified, a reflected
ray is only released if the value of the expression is nonzero and the intensity of the
reflected ray exceeds the threshold. In other words, a ray must satisfy all of the criteria
in the Rays to Release section in order to be released, not just one of them.

The release of reflected rays at a material discontinuity can be suppressed to prevent an

arbitrarily large number of degrees of freedom from being used to model rays of
extremely low intensity. Keep in mind that this can lead to a reduction in the total
energy of the system.

AUXILIARY DEPENDENT VARIABLES, REFRACTED RAY

If an Auxiliary Dependent Variable has been added to the model then there is an
option to reinitialize the values of auxiliary dependent variables for the refracted ray.
These settings are the same as in the New Value of Auxiliary Dependent Variables
section for the Wall feature.

AUXILIARY DEPENDENT VARIABLES, REFLECTED RAY

If an Auxiliary Dependent Variable has been added to the model then there is an
option to reinitialize the values of auxiliary dependent variables for the reflected ray.
These settings are the same as in the New Value of Auxiliary Dependent Variables
section for the Wall feature.

662 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

ADVANCED SETTINGS
Select the Show boundary normal check box to view the boundary normal in the
Graphics window.

Material Discontinuity Theory

Ray Properties
The Ray Properties node is added by default. Only one instance of this node can exist
for each physics interface. The Ray Properties node defines the equations for ray
propagation and may also include inputs for some ray properties.

RAY PROPERTIES
This section is shown when the Allow frequency distributions at release features check
box is cleared in the physics interface Ray Release and Propagation section. Otherwise
the ray frequency is controlled separately by each ray release feature.

Enter a Ray frequency f (SI unit: Hz) here. The default is 1000 Hz.

Release
Use the Release node to release rays within domains based on arbitrary expressions or
based on the positions of the mesh elements.

RELEASE TIMES
This section is only available when the Allow multiple release times check box has been
selected in the physics interface Advanced Settings section. Enter Release times
(SI unit: s) or click the Range button ( ) to select and define a range of specific times.
At each release time, rays are released with initial position and ray direction vector as
defined next.

INITIAL POSITION
Select an Initial position: Density (the default) or Mesh based.

Density
For Density enter a value for the Number of rays per release N (dimensionless). The
default is 100. Then enter a value or expression for the Density proportional to ρ
(dimensionless). The default is 1.

THE RAY ACOUSTICS INTERFACE | 663

 The Density proportional to ρ can be an expression rather than a number; the resulting
ray distribution approximately has a number density that is proportional to this
expression. The resulting distribution looks a bit random, and it depends on the order
in which the mesh elements are numbered. The distribution is probably not exactly the
same in different COMSOL Multiphysics versions, but the total number of rays
released is always N.

The Density proportional to expression must be strictly positive.

Advanced Options for Density-Based Release

The following options can be adjusted to make the number density of rays more closely
conform to the Density proportional to expression.

Select a Release distribution accuracy order between 1 and 5 (the default is 5), which
determines the integration order that is used when computing the number of rays to
release within each mesh element. The higher the accuracy order, the more accurately
rays will be distributed among the mesh elements.

The Position refinement factor (default 0) must be a nonnegative integer. When the
refinement factor is 0, each ray is always assigned a unique position, but the density is
taken as a uniform value over each mesh element. If the refinement factor is a positive
integer, the distribution of rays within each mesh element is weighted according to the
density, but it is possible for some rays to occupy the same initial position. Further
increasing the Position refinement factor increases the number of evaluation points
within each mesh element to reduce the probability of multiple rays occupying the
same initial position.

Mesh Based
For Mesh based the rays are released from a set of positions determined by a selection
of geometric entities (of arbitrary dimension) in the mesh. Given a Refinement factor
between 1 and 5, the centers of the refined mesh elements are used. Thus, the number
of positions per mesh element is refine^dim, except for pyramids, where it is
2
(4*refine -1)*refine/3.

664 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

RAY DIRECTION VECTOR
Select an option from the Ray direction vector list: Expression (the default), Spherical,
Hemispherical, Conical, or Lambertian (3D only).

• For Expression a single ray is released in the specified direction. Enter coordinates
for the Ray direction vector L0 (dimensionless) based on space dimension.
• For Spherical a number of rays are released at each point, sampled from a spherical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50.
• For Hemispherical a number of rays are released at each point, sampled from a
hemispherical distribution in wave vector space. Enter the Number of rays in wave
vector space Nw (dimensionless). The default is 50. Then enter coordinates for the
Hemisphere axis r based on space dimension.
• For Conical a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Cone axis r based
on space dimension. Then enter the Cone angle α (SI unit: rad; default π/3).
• The Lambertian option is only available in 3D. A number of rays are released at each
point, sampled from a hemisphere in wave vector space with probability density
based on the cosine law. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Hemisphere axis r
based on space dimension.

If Conical is selected in a 3D model, select an option from the Conical distribution list:

• Uniform density (the default): rays are released with polar angles from 0 to the
specified cone angle. The rays are distributed in wave vector space so that each ray
subtends approximately the same solid angle.
• Specify polar and azimuthal distributions: specify the Number of polar angles Nθ
(dimensionless) and the Number of azimuthal angles Nϕ (dimensionless). Rays are
released at uniformly distributed polar angles from 0 to the specified cone angle. A
single axial ray (θ = 0) is also released. For each value of the polar angle, rays are
released at uniformly distributed azimuthal angles from 0 to 2π. Unlike other
options for specifying the conical distribution, it is not necessary to directly specify
the Number of rays in wave vector space Nw (dimensionless), which is instead derived
from the relation Nw = Nθ × Nϕ + 1.
• Hexapolar: specify the Number of polar angles Nθ (dimensionless). In this
distribution, for each release point, one ray will be released along the cone axis. Six

THE RAY ACOUSTICS INTERFACE | 665

 rays are released at an angle α/Nθ from the cone axis, then 12 rays at an angle of
2α/Nθ, and so on. The total number of directions is Nw = 3Nθ(Nθ + 1) + 1.
• Flat: rays are released in a flat fan shape within the specified angle.
• Marginal rays only: the rays are all released at an angle α with respect to the cone axis.
The rays are released at uniformly distributed azimuthal angles from 0 to 2π.
• Marginal and axial rays only: the rays are all released at an angle α with respect to the
cone axis, except for one ray which is released along the cone axis. The marginal rays
are released at uniformly distributed azimuthal angles from 0 to 2π.
Hexapolar Uniform density Polar and azimuthal

Marginal and axial Marginal only Flat

Figure 8-1: Comparison of the available cone-based release distributions.

In 3D for the Conical distribution you can also let the Transverse direction be Automatic
(the default) or User defined. For User defined enter the components of et. This
controls, for example, the orientation of the ray fan when Flat is selected.

For Spherical, Hemispherical, Conical, and Lambertian, select an option from the
Sampling from Distribution list: Deterministic (the default) or Random. If Deterministic is
selected, the initial ray direction vectors are computed using an algorithm that seeks to
distribute the rays as evenly as possible in wave vector space. This algorithm will give
the same initial ray directions whenever the study is run. If Random is selected, the
initial direction of each ray is sampled from a probability distribution in wave vector
space using pseudorandom numbers. The result may be the same when rerunning the
study multiple times on the same computer, but the solution is likely to be different
on different architectures.

666 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

For Expression it is also possible to initialize the ray direction vector either in the global
coordinate system or in a coordinate system that moves with the same velocity as the
background medium. Select an option from the Initial wave vector specification list:
With respect to fluid (the default) or With respect to coordinate system.

• For With respect to fluid the initial wave vector is computed with respect to a
coordinate system that moves at the background velocity, so the initial ray direction
might not be parallel to the vector entered in the Ray direction vector text field if the
medium is moving.
• For With respect to coordinate system the initial ray direction is parallel to the vector
entered in the Ray direction vector text field as long as a ray could reasonably
propagate in that direction. For example, rays cannot be released in certain
directions if the background fluid is moving with a supersonic velocity.

INITIAL RAY FREQUENCY

This section is available when the Allow frequency distributions at release features check
box is selected under the physics interface Ray Release and Propagation section.

Select a Distribution function: None (the default), Normal, Lognormal, Uniform, or List of
values.

When None is selected, enter an initial value f0 (SI unit: Hz). The default value is
1000 Hz.

Select Normal to create a normal distribution function, Lognormal to create a

log-normal distribution function, or Uniform to create a uniform distribution function.
For any of these distributions, select an option from the Sampling from distribution
check box: Deterministic (the default) or Random. For Random sampling the mean and
standard deviation may not be exactly equal to the specified values but will statistically
converge as the number of rays is increased. The Number of values sets the number of
values that are sampled from the distribution function at each release point.

For the Normal or Lognormal distribution enter a user-defined Mean (default 1000 Hz)
and Standard deviation (default 100 Hz). For the Uniform distribution enter the
Minimum ray frequency fmin (default 1000 Hz) and Maximum ray frequency fmax
(default 2000 Hz).Select List of values to enter a list of distinct frequency values.

INITIAL PHASE
This section is available when the Compute phase check box is selected under the
physics interface Intensity Computation section. Enter an Initial phase Ψ0 (SI unit: rad).
The default value is 0.

THE RAY ACOUSTICS INTERFACE | 667

 INITIAL INTENSITY
This section is available when the ray intensity is solved for in the model and Expression
is selected as the Ray direction vector. Enter a value for the Initial intensity I0
(SI unit: W/m2). The default is 1000 W/m2.

INITIAL RADII OF CURVATURE

This section is available when the ray intensity is solved for in the model and Expression
is selected as the Ray direction vector. Select a Wavefront shape. In 3D the available
options are Plane wave (the default), Spherical wave, and Ellipsoid. In 2D the available
options are Plane wave (the default) and Cylindrical wave.

• For an idealized plane wave the radii of curvature would be infinite. However,
because the algorithm used to compute intensity requires finite values, when Plane
wave is selected the initial radii of curvature are instead given an initial value that is
108 times greater than the characteristic size of the geometry.
• For a Spherical wave or Cylindrical wave, enter the Initial radius of curvature r0
(SI unit: m).
• For an Ellipsoid, enter the Initial radius of curvature, 1 r1,0 (SI unit: m) and the Initial
radius of curvature, 2 r2,0 (SI unit: m). Also enter the Initial principal curvature
direction, 1 e1,0 (dimensionless).

For spherical and cylindrical waves the Initial radius of curvature must be
nonzero. To release a ray such that the initial wavefront radius of
curvature is zero, instead select a different option such as Conical from the
Ray direction vector list.

Principal Radii of Curvature

TOTAL SOURCE POWER

This section is available:

• when the ray intensity is solved for, and

• when Spherical, Hemispherical, or Conical is selected as the Ray direction vector.
• It is also available when the ray power is solved for, and then any choice of Ray
direction vector displays this section.

668 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Select an option from the Intensity initialization list: Uniform distribution (the default)
or Weighted distribution.

If Uniform distribution or Weighted distribution is selected, enter a Total source power

Psrc (SI unit: W). The default is 1 W. In 2D, instead enter the Total source power per
unit thickness Psrc (SI unit: W/m). The default is 1 W/m. For Weighted distribution,
also enter an expression for the Power weighting factor Pwt. The weighting factor may
have any unit. The released rays will have initial intensity and power proportional to
the weighting factor, while still adding up to the specified source power.

For example, if you release 1000 rays in a Spherical distribution with an

initial power of 10 W, and enter a weighting factor of [Link]+1, then
the sum of the power over all released rays, [Link](rac.Q), will equal
10 W. Since [Link] is the z-component of the ray direction vector, rays
close to the z-axis will each have power of about 0.02 W, those around
the negative z direction will have almost no power, and those in the
xy-plane will each have power of about 0.01 W.

INITIAL VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model.

For each of the Auxiliary Dependent Variable nodes added to the model, select a
Distribution function for the initial value of the auxiliary dependent variables and
whether the initial value of the auxiliary dependent variables should be a scalar value
or sampled from a distribution function.

The number of rays simulated can increase substantially and the following options are
available for each Auxiliary Dependent Variable added to the model.

When None is selected, enter an initial value. The symbol for the initial value is the
auxiliary variable name followed by a subscript 0, so for the default name rp the initial
value has symbol rp0.

For the initial value of the auxiliary dependent variables, select Normal to create a
normal distribution function, Lognormal to create a log-normal distribution function,
or Uniform to create a uniform distribution function. For any of these distributions,
select an option from the Sampling from distribution check box: Deterministic (the
default) or Random. For Random sampling the mean and standard deviation may not be

THE RAY ACOUSTICS INTERFACE | 669

 exactly equal to the specified values but will statistically converge as the number of rays
is increased. The Number of values sets the number of values that are sampled from the
distribution function at each release point.

For the Normal or Lognormal distribution enter the Mean (default 0) and Standard
deviation (default 1). For the Uniform distribution enter the Minimum (default 0) and
Maximum (default 1). Select List of values to enter a set of numerical values directly.

By default auxiliary dependent variables are initialized after all other degrees of
freedom. Select the Initialize before wave vector check box to compute the initial value
of the auxiliary dependent variable immediately after computing the initial wave
vectors of the rays. By selecting this check box it is possible to define the initial ray
direction as a function of the auxiliary dependent variables.

Sound Pressure Level Calculation

When the ray power is solved for, the Sound Pressure Level Calculation subnode is
available from the context menu (right-click the Wall, parent node) or from the Physics
toolbar, Attributes menu. The Sound Pressure Level Calculation subnode computes the
sound pressure level resulting from the interaction of the incident and reflected
acoustic waves at the wall.

SMOOTHING
Select the Compute smoothed accumulated variable check box to enter a Smoothing
radius r (SI unit: m) The smoothing radius functions like the corresponding setting for
the Accumulator (Boundary) node.

When the Compute smoothed accumulated variable is selected, smoothed

variables are defined for the intensity and sound pressure level. They can
be plotted in postprocessing and are, for example, named
rac.wall1.spl1.Iw_sm and rac.wall1.spl1.Lp_sm.

Sound Pressure Level Calculation Theory

670 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Accumulator (Domain)
Use the Accumulator node to define additional degrees of freedom on a domain. Each
Accumulator defines a variable, called the accumulated variable, on each domain
element in the selection list. The values of the accumulated variables are determined
by the properties of rays in each domain element.

ACCUMULATOR SETTINGS
Select an Accumulator type: Density (default) or Count.

• For Density, the accumulated variable is divided by the volume of the mesh element
where it is defined.
• For Count, the accumulated variable is unaffected by the element size.

Select an option from the Accumulate over list: Elements (the default) or Elements and
time.

• For Elements, the value of the accumulated variable in an element is the sum of the
source terms of all rays in that element. If the Accumulator type is set to Density, this
sum is divided by the mesh element volume.
• For Elements and time, the time derivative of the accumulated variable in an element
is the sum of the source terms of all rays in that element. If the Accumulator type is
set to Density, this sum is divided by the mesh element volume. As each ray
propagates through a series of mesh elements, it leaves behind a contribution to the
accumulated variable that remains even after the ray has moved on.

Enter the Accumulated variable name. The default is rpd. The accumulated variable is
defined as <name>.<varname>, where <name> is the physics interface name and
<varname> is the accumulated variable name. For example, in an instance of the Ray
Acoustics interface with default name rac and default accumulated variable name rpd,
the variable would be named [Link].

Enter a Source R. The unit of the source depends on the settings in the Units section.
The source term is used to calculate the accumulated variable in a manner specified by
the Accumulate over and Accumulator type settings.

If Elements and time is selected from the Accumulate over list, select an option from the
Source interpolation list: Constant, Linear (the default), Quadratic, or Exponential. The
Source interpolation determines what functional form the Source is assumed to follow
during each time step taken by the solver. This information is used to compute the

THE RAY ACOUSTICS INTERFACE | 671

 accumulated variable in mesh elements that the rays pass through during each time
step.

UNITS
Select a Dependent variable quantity from the list; the default is Dimensionless [1]. To
enter a unit, select None from the list and in the Unit field enter a value, for example, K,
m/s, or mol/m^3.

Nonlocal Accumulator
Use the Nonlocal Accumulator subnode to communicate information from a ray’s
current position to the position from which it was released.

The subnode is available from the context menu (right-click the Release from
Boundary parent node) or from the Physics toolbar, Attributes menu.

Each Nonlocal accumulator subnode defines a variable, called the accumulated variable,
that is computed using variables defined on rays released by the parent Release from
Boundary node or on domains and boundaries encountered by such rays.

ACCUMULATOR SETTINGS
Select an Accumulator type: Density (default) or Count.

• For Density, the accumulated variable is divided by the volume of the mesh element
where it is defined.
• For Count, the accumulated variable is unaffected by the element size.

Select an option from the Accumulate over list: Elements (default) or Elements and
time.

• For Elements, the accumulated variable is proportional to the instantaneous value of

the Source term R for all applicable rays.
• For Elements and time, the time derivative of the accumulated variable is
proportional to the instantaneous value of the Source term R for all applicable rays,
and thus the accumulated variable considers the time history of rays in the modeling
domain instead of just their current values.

Enter the Accumulated variable name. The default is rpi.

Enter a Source R. The unit of the source depends on the settings in the Units section.
The source term is used to calculate the accumulated variable in a manner specified by
the Accumulate over and Accumulator type settings.

672 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Select a Source geometric entity level: Domains, Boundaries, or Domains and boundaries.

• If Domains is selected, rays only contribute to the accumulated variable on their

releasing surface if they are still active; that is, they are still propagating through a
domain.
• If Boundaries is selected, the rays only contribute to the accumulated variable if they
have become stuck or frozen to a boundary somewhere in the model.
• If Domains and boundaries is selected, all of the active, stuck, and frozen rays released
by a feature can contribute to the accumulated variable.

UNITS
Select a Dependent variable quantity from the list; the default is Dimensionless [1]. To
enter a unit, select None from the list and in the Unit field enter a value, for example, K,
m/s, or mol/m^3.

SMOOTHING
Select the Compute smoothed accumulated variable check box to enter a Smoothing
radius r (SI unit: m). The default is 0.1 m.

Release from Boundary

Use the Release from Boundary node to determine how to release rays on an interior or
exterior boundary in a specific direction.

The Nonlocal Accumulator subnode is available from the context menu (right-click
the parent node) or from the Physics toolbar, Attributes menu.

See Release for information on the following sections: Release Times, Initial Ray
Frequency, Initial Phase, Initial Intensity, Total Source Power, and Initial Value of Auxiliary
Dependent Variables.

COORDINATE SYSTEM SELECTION

It is possible to specify the initial ray direction in terms of the global coordinates or in
another coordinate system defined for the model Component. Select an option from
the Coordinate system list. By default Global coordinate system is selected. If other
coordinate systems are defined, they can also be selected from the list. When specifying
the initial ray direction (see the Ray Direction Vector section), direction components can
be specified using the basis vectors of whichever coordinate system has been selected
from the list.

THE RAY ACOUSTICS INTERFACE | 673

 When a coordinate system other than Global coordinate system is selected from the
Coordinate system list, arrows will appear in the Graphics window to indicate the
orientation of the basis vectors of the coordinate system on the selected boundaries.

INITIAL POSITION
Select an Initial position: Uniform distribution (the default for 2D components)
Projected plane grid (the default for 3D components), Mesh based, or Density. Mesh
based and Density have the same settings as described for the Release node.

For 2D components, if Uniform distribution is selected, enter the Number

of rays per release N (dimensionless). The union of the selected
boundaries is divided into N segments of approximately equal length, and
a ray is placed in the middle of each segment.

For 3D components, if Projected plane grid is selected, enter the Number

of rays per release N (dimensionless). The rays are distributed on a plane
grid in planes that are approximately tangential to the selected boundaries
(for a plane boundary, you get a uniform distribution).

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Expression (the default),
Hemispherical, Conical, or Lambertian (3D only).

• For Expression, a single ray is released in the specified direction. Enter coordinates
for the Ray direction vector L0 (dimensionless) based on space dimension. Select the
Specify tangential and normal vector components check box to specify the initial
direction using a coordinate system based on the directions tangential and normal
to the surface (t1, t2, n).
• For Hemispherical, a number of rays are released at each point, sampled from a
hemispherical distribution in wave vector space. Enter the Number of rays in wave
vector space Nw (dimensionless). The default is 50. Then enter coordinates for the
Hemisphere axis r based on space dimension. Select the Specify tangential and normal
vector components check box to specify the hemisphere axis using a coordinate
system based on the directions tangential and normal to the surface (t1, t2, n).
• For Conical, a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Cone axis r based

674 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 on space dimension. Then enter the Cone angle α (SI unit: rad). The default is
π/3 radians. Select the Specify tangential and normal vector components check box to
specify the cone axis using a coordinate system based on the directions tangential
and normal to the surface (t1, t2, n).
• The Lambertian option is only available in 3D. A number of rays are released at each
point, sampled from a hemisphere in wave vector space with probability density
based on the cosine law. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Hemisphere axis r
based on space dimension. Select the Specify tangential and normal vector
components check box to specify the hemisphere axis using a coordinate system
based on the directions tangential and normal to the surface (t1, t2, n)

When the Specify tangential and normal vector components check box is
selected, arrows indicating the normal direction on the selected
boundaries will appear in the Graphics window.

Note that the normal direction may be opposite the built-in variable for
the boundary normal (for example, nx, ny, and nz) to ensure that a
positive value causes rays to be released into the simulation domain. This
often occurs when the release is applied to exterior boundaries.

When the normal direction used by the Release from Boundary feature is
opposite the normal vector defined by the geometry, the tangential
directions are similarly inverted to ensure that the boundary coordinate
system is right-handed.

If Conical is selected in a 3D model, select an option from the Conical distribution list:

• Uniform density (the default): rays are released with polar angles from 0 to the
specified cone angle. The rays are distributed in wave vector space so that each ray
subtends approximately the same solid angle.
• Specify polar and azimuthal distributions: specify the Number of polar angles Nθ
(dimensionless) and the Number of azimuthal angles Nϕ (dimensionless). Rays are
released at uniformly distributed polar angles from 0 to the specified cone angle. A
single axial ray (θ = 0) is also released. For each value of the polar angle, rays are
released at uniformly distributed azimuthal angles from 0 to 2π. Unlike other
options for specifying the conical distribution, it is not necessary to directly specify
the Number of rays in wave vector space Nw (dimensionless), which is instead derived
from the relation Nw = Nθ × Nϕ + 1.

THE RAY ACOUSTICS INTERFACE | 675

 • Hexapolar: specify the Number of polar angles Nθ (dimensionless). In this
distribution, for each release point, one ray will be released along the cone axis. Six
rays are released at an angle α/Nθ from the cone axis, then 12 rays at an angle of
2α/Nθ, and so on. The total number of ray directions in the distribution is
Nw = 3Nθ(Nθ + 1) + 1.
• Flat: rays are released in a flat fan shape within the specified angle.
• Marginal rays only: the rays are all released at an angle α with respect to the cone axis.
The rays are released at uniformly distributed azimuthal angles from 0 to 2π.
• Marginal and axial rays only: the rays are all released at an angle α with respect to the
cone axis, except for one ray which is released along the cone axis. The marginal rays
are released at uniformly distributed azimuthal angles from 0 to 2π.

The available Conical distribution options are illustrated in Figure 8-1.

In 3D for the Conical distribution you can also let the Transverse direction be Automatic
(the default) or User defined. For User defined enter the components of et. This
controls, for example, the orientation of the ray fan when Flat is selected.

For Hemispherical, Conical, and Lambertian, select an option from the Sampling from
Distribution list: Deterministic (the default) or Random. If Deterministic is selected, the
initial ray direction vectors are computed using the same algorithm, which seeks to
distribute the rays as evenly as possible in wave vector space, whenever the study is run.
If Random is selected, the initial direction of each ray is sampled from a probability
distribution in wave vector space using pseudorandom numbers.

INITIAL RADII OF CURVATURE

This section is available when the ray intensity is solved for in the model and Expression
is selected as the Ray direction vector. Select a Wavefront shape. In 3D the available

676 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

options are From surface curvature (the default), Plane wave, Spherical wave, and
Ellipsoid. In 2D the available options are Plane wave (the default) and Cylindrical wave.

• For a Spherical wave or Cylindrical wave, enter the Initial radius of curvature r0
(SI unit: m).
• For an Ellipsoid (3D only), enter the Initial radius of curvature, 1 r1,0 (SI unit: m)
and the Initial radius of curvature, 2 r2,0 (SI unit: m). Also enter the Initial principal
curvature direction, 1 e1,0 (dimensionless).

For spherical and cylindrical waves the Initial radius of curvature must be
nonzero. To release a ray such that the initial wavefront radius of
curvature is zero, instead select a different option such as Conical from the
Ray direction vector list.

Principal Radii of Curvature

Release from Symmetry Axis

Use the Release from Symmetry Axis node to release rays from specified edges along the
axis of symmetry in a 2D axisymmetric geometry. It functions like the Release from
Boundary node, except that it can only be applied to the axis of symmetry and that the
options in the Ray Direction Vector section are slightly different.

See Release for information on the following sections: Release Times, Initial Ray
Frequency, Initial Phase, Initial Intensity, Initial Radii of Curvature, Total Source Power,
and Initial Value of Auxiliary Dependent Variables.

INITIAL POSITION
Select an Initial position: Uniform distribution (the default), Mesh based, or Density. Mesh
based and Density have the same settings as described for the Release node.

If Uniform distribution is selected, enter the Number of rays per release N

(dimensionless). The default is 100. The union of the selected boundaries is divided
into N segments of approximately equal length, and a ray is placed in the middle of
each segment.

THE RAY ACOUSTICS INTERFACE | 677

 RAY DIRECTION VECTOR
Select an option from the Ray direction vector list: Expression (the default), Spherical,
or Conical.

• For Expression, a single ray is released in the specified direction. Enter the radial and
axial components of the Ray direction vector L0 (dimensionless).
• For Spherical, a number of rays are released at each point, from the negative
z direction to the positive z direction in the half plane with r > 0. These rays are
sampled from an anisotropic hemispherical distribution in wave vector space, so that
rays are more heavily weighted toward the r direction rather than the ±z directions.
In 3D this would correspond to a spherical distribution of rays such that each ray
subtends approximately the same solid angle in wave vector space.
Enter the Number of rays in wave vector space Nw (dimensionless). The default is 50.
Then enter coordinates for the Hemisphere axis r based on space dimension. Select
the Specify tangential and normal vector components check box to specify the
hemisphere axis using a coordinate system based on the directions tangential and
normal to the surface (t1, t2, n).
• For Conical, a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then enter coordinates for the Cone axis r based
on space dimension. Then enter the Cone angle α (SI unit: rad). The default is
π/3 radians. Unlike the Spherical distribution, the rays in the Conical distribution
each subtend the same angle in 2D, not the same solid angle in 3D.

For Hemispherical and Conical, select an option from the Sampling from Distribution list:
Deterministic (the default) or Random. If Deterministic is selected, the initial ray
direction vectors are computed using the same algorithm, which seeks to distribute the
rays as evenly as possible in wave vector space, whenever the study is run. If Random is
selected, the initial direction of each ray is sampled from a probability distribution in
wave vector space using pseudorandom numbers.

Background Velocity
The Background Velocity node is used to assign a velocity to the medium in which the
rays propagate. When no Background velocity node is present, the medium is assumed
to be stationary.

678 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

BACKGROUND VELOCITY
The default Velocity field u (SI unit: m/s) is User defined. Enter values or expressions
for the vector components.

Auxiliary Dependent Variable

Use the Auxiliary Dependent Variable node to solve additional first-order ordinary
differential equations for each ray released.

AUXILIARY DEPENDENT VARIABLE

Enter a Field variable name. The default is rp and can be changed to anything provided
it does not conflict with the name of the variables for the position or momentum
degrees of freedom. The name should not conflict with other auxiliary dependent
variables.

Enter a Source R. The unit of the source depends on the unit of the auxiliary dependent
variable and the option selected from the Integrate list described below.

Select a way to Integrate the equation you have defined: With respect to time or Along
ray trajectory. If you are interested in the residence time of a group of rays in a given
system then you can set the Source to 1 and set Integrate to With respect to time. To
compute the length of the ray trajectory you can set the Source to 1 and set Integrate
to Along ray trajectories.

UNITS
These settings are the same as for Accumulator (Domain).

Release from Edge

Use the Release from Edge node to release rays from positions along a set of edges in a
3D geometry.

See Release for information on the following sections: Release Times, Ray Direction
Vector, Initial Ray Frequency, Initial Phase, Initial Intensity, Initial Radii of Curvature, Total
Source Power, and Initial Value of Auxiliary Dependent Variables.

INITIAL POSITION
Select an Initial position: Mesh based (the default), Uniform distribution, or Density. Mesh
based and Density have the same settings as described for the Release node. If Uniform
distribution is selected, enter the Number of rays per release N (dimensionless). The

THE RAY ACOUSTICS INTERFACE | 679

 union of the selected edges is divided into N segments of approximately equal length,
and a ray is released from the middle of each segment.

Release from Point

Use the Release from Point node to release rays from a set of points in a geometry. By
default, one ray is released at every selected point, although it is possible to release
multiple rays at each point if the initial direction or an auxiliary dependent variable is
sampled from a distribution.

See Release for information on the following sections: Release Times, Ray Direction
Vector, Initial Ray Frequency, Initial Phase, Initial Intensity, Initial Radii of Curvature, Total
Source Power, and Initial Value of Auxiliary Dependent Variables.

Release from Point on Axis

The Release from Point on Axis feature is similar to the Release from Point feature,
except that it can only release rays from points along the axis of symmetry in a 2D
axisymmetric geometry.

See Release from Boundary for information on the Ray Direction Vector section.

For all other sections, see Release from Grid.

Release from Grid

Use the Release from Grid node to release rays from a user-defined grid of points.

See Release for information on the following sections: Release Times, Ray Direction
Vector, Initial Ray Frequency, Initial Phase, Initial Intensity, Initial Radii of Curvature, Total
Source Power, and Initial Value of Auxiliary Dependent Variables.

INITIAL COORDINATES
Select an option from the Grid type list: All combinations (the default), Specified
combinations, Cylindrical (3D only), or Hexapolar (3D only).

Linear, Rectangular, and Arbitrary Grids

For All combinations and Specified combinations enter Initial coordinates based on space
dimension (qx,0, qy,0, and qz,0 for 3D components) for the ray positions or click the
Range button ( ) to select and define a range of specific coordinates.

680 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

If Specified combinations is selected, the number of initial coordinates entered for each
space dimension must be equal, and the total number of rays released is equal to the
length of one of the lists of initial coordinates. If All combinations is selected, the total
number of rays released is equal to the product of the lengths of each list of initial
coordinates.

For example, suppose a 2D component includes a Release from Grid node with the
following initial coordinates:

• qx,0 = range(0,1,3)
• qy,0 = range(2,2,8)

If All combinations is selected, a total of 16 rays will be released, including every possible
combination of the initial x- and y-coordinates. If Specified combinations is selected, 4
rays will be released with initial positions (0,2), (1,4), (2,6), and (3,8).

Cylindrical Grids
For Cylindrical, enter coordinates for the Center location qc (SI unit: m). By default, the
distribution is centered at the origin. Then enter the components of the Cylinder axis
direction rc (dimensionless). The rays will be released at specified radial distances and
angles in the plane containing the point qc and orthogonal to the direction rc.

Select an option from the Radial distribution list: Uniform radius intervals (the default),
Uniform number density, or User defined.

For Uniform radius intervals or Uniform number density, enter a value or expression for
the Radius Rc (SI unit: m). The default is 1 m. Then enter a positive integer for the
Number of radial positions Nc (dimensionless). The default is 5.

For User defined, enter a list of Radial coordinates qr (SI unit: m) directly. An arbitrary
number of radial coordinates can be entered in the list. The default is 1 m.

The effect of the Radial distribution setting on the resulting ray placement is illustrated
in Figure 8-2. For Uniform radius intervals, the distances between the concentric rings
of rays are all equal, but the number density of rays is greater at the center of the
distribution than at the outer edge. The option Uniform number density corrects this
imbalance by defining nonuniform increments in the radial position between the
concentric rings.

Enter a positive integer for the Number of angles N φ (dimensionless). The default is 10.

THE RAY ACOUSTICS INTERFACE | 681

 The total number of rays released (before accounting for distributions of ray direction
or auxiliary dependent variables at each release point) is N c × N φ + 1 because a single
ray is also released at the center of the distribution.

Figure 8-2: Comparison of the cylindrical release in which rays are positioned at uniform
radius intervals (left), with uniform number density (center), or at user-defined radii
(right).

Hexapolar Grids
For Hexapolar, enter coordinates for the Center location qc (SI unit: m). By default, the
distribution is centered at the origin. Then enter the components of the Cylinder axis
direction rc (dimensionless). The rays will be released at specified radial distances and
angles in the plane containing the point qc and orthogonal to the direction rc.

Enter a value or expression for the Radius Rc (SI unit: m). The default is 1 m. Then
enter a positive integer for the Number of radial positions Nc (dimensionless). The
default is 5.

Unlike the Cylindrical grid types described in the previous section, each concentric ring
of the Hexapolar grid releases a different number of rays, as shown in Figure 8-3. One
ray is always released at the center. The first ring of rays surrounding the center has 6
rays arranged in a regular hexagon. Each ring of rays beyond the first has 6 more rays
than the ring preceding it, with the grid points arranged in a regular polygon. The
radius increments between consecutive rings are uniform.

The total number of grid points generated is

N grid = 1 + 3N c ( N c + 1 )

Thus the relationship between the number of radius intervals and the total number of
grid points is quadratic.

682 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Figure 8-3: Comparison of hexapolar grids with 2 rings (left), 5 rings (middle), and 10
rings (right).

Previewing Grid Points

In the Initial Coordinates section, you can click the Preview Initial Coordinates and
Preview Initial Extents buttons to visualize the ray release positions. Clicking
Preview Initial Coordinates will cause a point to appear in the Graphics window for every
release position. Clicking Preview Initial Extents will cause a bounding box to appear,
indicating the spatial extents of the released rays. Examples are shown in Figure 8-4
and Figure 8-5.

Figure 8-4: Graphics window after clicking the Preview Initial Coordinates button.

THE RAY ACOUSTICS INTERFACE | 683

 Figure 8-5: Graphics window after clicking the Preview Initial Extents button.

Release from Grid on Axis

The Release from Grid on Axis feature is similar to the Release from Grid feature, except
that it can only release rays from a grid of points along the axis of symmetry in a 2D
axisymmetric geometry; that is, qr,0 = 0 for all released rays.

See Release from Boundary for information on the Ray Direction Vector section.

For all other sections, see Release from Grid.

Release from Data File

Use a Release from Data File node to specify the initial ray position, direction, and values
of auxiliary dependent variables using data from a text file.

See Release for information on the following sections: Release Times, Initial Phase,
Initial Intensity, Initial Radii of Curvature, and Total Source Power.

Formatting Guidelines for Files Containing Ray Data

The imported data file should be a text file (*.txt) arranged in a spreadsheet format;
that is, each row corresponds to a distinct ray and should have the same number of
columns as all other rows. Columns can be separated by spaces, tabs, or a combination
of the two. Begin a line with the percent (“%”) character to include comments or
empty lines in the data file.

684 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

For example, a data file containing the following text would release rays at the positions
(0.1, 0.2, 0.6) and (0.2, 0.4, 0.8) in a three-dimensional geometry:

% Ray release positions

% qx0 qy0 qz0
0.1 0.2 0.6
0.2 0.4 0.8

INITIAL POSITION
Browse your computer to select a text file, then click Import to import the data. To
remove the imported data, click Discard. Enter the Index of first column containing
position data i to indicate which column represents the first coordinate of the ray
position vectors. The default value, 0, indicates the first column.

TRANSFORMATIONS
The distribution of loaded ray release positions may be scaled, rotated, and translated
before the rays are released.

To scale the distribution of release positions, enter a value or expression for the Scale
factor R (dimensionless). The default is 1. This scale factor can be used to correct unit
discrepancies between the data file and the model geometry. For example, if the
geometry length unit is in meters but the data file lists coordinates in millimeters, enter
a scale factor of 0.001.

To rotate the distribution, enter the Euler angles (Z-X-Z) α, β, and γ (in 3D) or the
Rotation angle α (in 2D). The default values are all 0.

In 3D, α is the rotation angle about the space-fixed z-axis, then β is the rotation angle
about the transformed x-axis (or x’-axis), and finally γ is the rotation angle about the
transformed z-axis (or z’’-axis). Positive values indicate counterclockwise rotations.

Enter values or expressions for the components of the Displacement vector Δq

(SI unit: m). The default is not to apply any translation.

If a translation and another type of transformation (scaling or rotation) are applied,

then the translation is always applied after the other transformations.

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Expression (the default), From file,
Spherical, Hemispherical, Conical, or Lambertian (3D only).

• For From file, enter the Index of first column containing direction data i. The default
is 3. The columns are zero-indexed; that is, an index of 0 corresponds to the first

THE RAY ACOUSTICS INTERFACE | 685

 column. If you chose to rotate the loaded ray release positions in the Initial Position
section, select the Rotate initial directions check box to apply the same rotation to
the initial direction of the released rays. The check box is cleared by default.
• For all other settings, the settings are the same as for the Release node.

INITIAL RAY FREQUENCY

This section is available when the Allow frequency distributions at release features check
box is selected under the physics interface Ray Release and Propagation section.

Select a Distribution function: None (the default), Normal, Lognormal, Uniform, List of
values, or From file.

• When From file is selected, enter the Index of column containing frequency data. The
default value is 3. The columns are zero-indexed; that is, an index of 0 corresponds
to the first column.
• For all other options, the settings are the same as for the Release node.

INITIAL VALUE OF AUXILIARY DEPENDENT VARIABLES

This section is available if an Auxiliary Dependent Variable has been added to the
model.

For each of the active Auxiliary Dependent Variable nodes in the model, choose an
option from the Distribution function list: From file, None (the default), Normal,
Lognormal, Uniform, or List of Values.

• When From file is selected, enter the Index of column containing data. The default
value is 3. The columns are zero-indexed; that is, an index of 0 corresponds to the
first column.
For all other options, the settings are the same as for the Release node.

Release from Exterior Field Calculation

This feature is only available in 3D.

686 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Use a Release from Exterior Field Calculation node to launch rays with an initial intensity
and phase distribution matching the exterior field solved for in another study.

To add this feature, either the ray intensity or power must be solved for.
Select an appropriate option from the Intensity computation list in the
physics interface Intensity Computation section.

The Release from Exterior Field Calculation node can be used for multiscale acoustics
modeling. First, the Pressure Acoustics, Frequency Domain interface should be used
to compute the near-field acoustic pressure distribution, such as the first few
wavelengths surrounding a speaker or another source. The Exterior Field Calculation
node should be added to this instance of the Pressure Acoustics, Frequency Domain
interface, and Far-field integral approximation for r → ∞. should be selected from the
Type of integral list. Because the near-field calculation is solving for the instantaneous
pressure amplitude in individual waves, the Nyquist criterion must be fulfilled.
Typically this requires 10 linear elements or 5 second-order elements per wavelength.

Then, in the same model component or a different component, the Ray Acoustics
interface can be used to extend the exterior field over an arbitrarily large number of
wavelengths, because the ray tracing approach does not require a finite element mesh
fine enough to resolve individual oscillations of the acoustic pressure field.

The Release from Exterior Field Calculation node is only available in 3D. However, the
exterior field function it uses to initialize the ray intensity and phase can be defined
either in a 3D model component or a 2D axisymmetric model component. If the
previous model component was a 2D axisymmetric component, the intensity
distribution of released rays will also be assumed to be axisymmetric.

See Release for information on the Initial Values of Auxiliary Dependent Variables section
and the optional Release Times section.

INITIAL COORDINATES
Select an option from the Grid type list: All combinations (the default) or Specified
combinations.

Linear, Rectangular, and Arbitrary Grids

Enter Initial coordinates based on space dimension (qx,0, qy,0, and qz,0) for the ray
positions or click the Range button ( ) to select and define a range of specific
coordinates.

THE RAY ACOUSTICS INTERFACE | 687

 If Specified combinations is selected, the number of initial coordinates entered for each
space dimension must be equal, and the total number of rays released is equal to the
length of one of the lists of initial coordinates. If All combinations is selected, the total
number of rays released is equal to the product of the lengths of each list of initial
coordinates.

Previewing Grid Points

In the Initial Coordinates section, you can click the Preview Initial Coordinates and
Preview Initial Extents buttons to visualize the ray release positions. Clicking
Preview Initial Coordinates will cause a point to appear in the Graphics window for every
release position. Clicking Preview Initial Extents will cause a bounding box to appear,
indicating the spatial extents of the released rays.

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Spherical (the default), Hemispherical,
or Conical.

• For Spherical a number of rays are released at each point, sampled from a spherical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50.
• For Hemispherical a number of rays are released at each point, sampled from a
hemispherical distribution in wave vector space. Enter the Number of rays in wave
vector space Nw (dimensionless). The default is 50. Then select an option from the
Hemisphere axis list: From Euler angles or User Defined. For User Defined enter
coordinates for the Hemisphere axis r based on space dimension.
• For Conical a number of rays are released at each point, sampled from a conical
distribution in wave vector space. Enter the Number of rays in wave vector space Nw
(dimensionless). The default is 50. Then select an option from the Cone axis list:
From Euler angles or User Defined. For User Defined enter coordinates for the Cone
axis r based on space dimension. Then enter the Cone angle α (SI unit: rad). The
default is π/3 radians.

Note that unlike the Release from Grid feature, the options Expression and Lambertian
are not available here. In addition, if Conical is selected, the cone must be isotropic.
This is because each ray is assumed to subtend approximately the same solid angle, so
direction distributions that give unequal weight to some directions (like Lambertian)
or options with no direction distribution at all (like Expression) cannot initialize the ray
intensity and power in a consistent and physically meaningful way.

688 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

EXTERIOR FIELD VARIABLE NAME
Enter an expression for the Exterior field variable name. The default expression is pext.
This expression must match the variable name in a corresponding Exterior Field
Calculation node.

SOURCE ORIENTATION
Use the settings in this section to orient the ray source. This allows, for example, the
release of rays from a speaker to be modeled with the speaker oriented in many
different directions, while only solving for the exterior field once.

Enter the Euler angles (Z-X-Z). The labels Z-X-Z indicate the order in which rotation
about the different axes is performed. First, the local coordinate system of the exterior
field is rotated about its Z-axis by the angle α. Then it is rotated about its X-axis (which
is now at an angle α to the global x-axis) by the angle β. Finally, this local coordinate
system is rotated about its new Z-axis by the angle γ. All three inputs are plane angles
(SI unit: rad) with default values of 0.

ADVANCED SETTINGS
This section is only shown if the Allow frequency distributions at release features check
box is cleared in the physics interface Ray Release and Propagation section.

The Use frequency from the far-field calculation as the ray frequency check box is selected
by default. While this check box is selected and the Release from Exterior Field
Calculation node is active, the frequency of released rays will be determined by the
Frequency Domain or Wavelength Domain study in the Values of Variables not Solved For
section of the Time Dependent solver settings. Usually, this previous study was used to
solve for the exterior field using the Pressure Acoustics, Frequency Domain interface.

While this check box is selected and the Release from Exterior Field Calculation node is
active, the frequency specified in the Ray Properties node is ignored.

Source with Directivity

This feature is only available in 3D.

Use the Source with Directivity node to release rays with initial intensity or power
following a spatial directivity function. This is useful for modeling loudspeakers or
human talker sources in a room acoustics context. The rays will emanate outward from

THE RAY ACOUSTICS INTERFACE | 689

 a point in either a spherical or hemispherical distribution. Predefined variables for the
azimuthal and polar angles within this distribution can be used to define the spatial
directivity function, either as a sound pressure level or in terms of absolute power.

COORDINATE SYSTEM SELECTION

By default, Global coordinate system is selected from the Coordinate system list. Other
user defined coordinate systems can also be selected from the list. However, boundary
coordinate systems are not included in the list because there is no requirement for the
ray release position to be on a boundary.

There are two main reasons to choose a transformed coordinate system for this feature:

1 When using a Hemispherical distribution of rays (see the Ray Direction Vector
section), the first basis vector of the selected coordinate system defines the
hemisphere axis. However, if the Specify local coordinates check box is selected in the
Local Coordinates section, then the Forward direction ef defines the hemisphere axis.
(This Forward direction can be defined in either global coordinates or in a
transformed coordinate system.)
2 The Source with Directivity node defines built-in variables for the azimuthal and
polar angle. For a physics interface with the default name rac and a Source with
Directivity node with the default name swd1, these variables are called
[Link] (azimuthal angle) and [Link] (polar angle). These angles
are defined with respect to the selected coordinate system; for example, if the Global
coordinate system is used, then [Link] is the angle between the ray
direction vector and the positive z-axis, while [Link] is the angle of the
projected ray direction vector in the xy-plane, measured counterclockwise from the
positive x-axis.

No matter which coordinate system is selected, the Ray release position (see the Initial
Position section) is always specified in global Cartesian coordinates. Thus the selected
coordinate system controls the orientation of the source while the global Cartesian
system always controls its position.

RELEASE TIMES
This section is only shown if the Allow multiple release times check box is selected in
the physics interface Advanced Settings section. Furthermore, this Advanced Settings
section is only shown when Advanced Physics Options are enabled (click the Show More
Options button in the Model Builder toolbar).

690 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Enter Release times (SI unit: s) or click the Range button ( ) to select and define a
range of specific times. At each release time, rays are released with initial position and
ray direction vector as defined in the following sections.

INITIAL POSITION
The rays are released in a spherical or hemispherical distribution of directions
emanating outward from a single point. Here, enter the components of the Ray release
position vector q0 (SI unit: m). By default the ray release position is at the origin.

RAY DIRECTION VECTOR

Select an option from the Ray direction vector list: Spherical (the default) or
Hemispherical. If Hemispherical is selected, the hemisphere axis is not defined in this
section; the axis is either the first basis vector in the selected coordinate system (see the
Coordinate System Selection section) or the Forward direction ef specified in the Local
Coordinates section.

INTENSITY AND POWER

This section is shown when either ray intensity or ray power, or both, are solved for,
meaning that any option except None has been selected from the Intensity computation
list in the physics interface Intensity Computation section.

Select an option from the Directivity list: Specify spatial directivity (the default) or
Specify total source power.

• For Specify spatial directivity, enter a value or expression for the Spatial directivity
D(ϕ,θ) in dB. The default expression is
0[dB]*cos([Link])*sin([Link]). The expression is a
placeholder that shows the variable names for the two polar angles. The directivity
will typically be defined by an interpolation function that represents a given source
like a loudspeaker. Also specify the Reference level Lref in dB (default 65 dB) and the
Reference distance Rref (default 1 m). The ray intensity or power is initialized such
that the sound pressure level of each ray equals the sum of the Spatial directivity and
the Reference level, evaluated for that ray’s initial direction, when the ray reaches a
distance from the source equal to the Reference distance.
• For Specify total source power, enter a value or expression for the Total source power
Psrc (SI unit: W). The default is 1 W. Enter a Power weighting factor Pwt, which can
have any unit. The default expression is 1. The initial intensity and power of each
released ray will be proportional to the Power weighting factor evaluated for that ray,
normalized such that the sum of the power over all rays equals the Total source

THE RAY ACOUSTICS INTERFACE | 691

 power. Much like the Spatial directivity, the Power weighting factor can be a function
of ray variables such as the azimuthal and polar angles.

INITIAL RAY FREQUENCY

This section is shown when the Allow frequency distributions at release features check
box is selected in the physics interface Ray Release and Propagation list. Unlike some
other ray release features, only a single value of the Initial ray frequency f0 (SI unit: Hz)
may be entered, not a distribution or a list of values. The default is 1 kHz.

INITIAL PHASE
This section is shown when ray intensity is solved for, and in addition the Compute
phase check box is selected in the physics interface Intensity Computation section. Enter
a value or expression for the Initial phase Ψ0 (SI unit: rad). The default is 0.

LOCAL COORDINATES
The Specify local coordinates check box is cleared by default. When it is selected, enter
the components of the Forward direction ef and the Transverse direction et
(dimensionless). By default these directions are the first and third basis vectors,
respectively, in whichever coordinate system was selected from the Coordinate System
Selection section. These vectors are used to set the hemisphere axis (if Hemispherical is
selected from the Ray Direction Vector list). They also control the definition of built-in
variables for the azimuthal and polar angles, which can be used to define a spatial
directivity function for the released rays.

Assigning Spatial Directivity to a Source

Ray Continuity
Use the Ray Continuity node to specify that rays should cross a pair boundary as if it
were invisible. Pair boundaries appear when the geometry sequence ends in Form
Assembly instead of Form Union. Such boundaries require special handling because the
mesh elements on either side of the pair boundary aren’t required to match up exactly.

PAIR SELECTION
Select one or more identity pairs to allow rays to cross between the source and
destination boundaries of these pairs. Such identity pairs are typically created

692 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

automatically on interior boundaries when the geometry sequence ends in a Form
Assembly node instead of a Form Union node.

The Ray Continuity node does not cause the ray position components to
change discontinuously; the source and destination boundaries for the
identity pairs must be overlapping.

If the geometry sequence ends in a Form Union node, the Ray Continuity
node usually isn’t necessary because rays can freely cross interior
boundaries where no boundary condition has been applied.

Ray Termination
Use the Ray Termination feature to remove rays from the model as they exit the
geometry, or to annihilate rays of sufficiently low intensity or power. Only one Ray
Termination node can be added to the model, but it is possible to combine
position-based and intensity-based criteria for removing rays if the ray intensity or
power is computed by the model.

TERMINATION CRITERIA
Select an option from the Spatial extents of ray propagation list: Bounding box, from
geometry (the default), Bounding box, user defined, or None. This list controls whether
rays are terminated based on their positions.

• For Bounding box, from geometry, a ray is immediately terminated if it reaches the
surface of the bounding box of the geometry, defined as the smallest rectangle or
rectangular prism with edges parallel to the axes of the global coordinate system that
encloses all of the geometric entities. This bounding box is extended by 5% in each
direction so that rays can still interact with boundaries that coincide with the
perimeter of the bounding box. Therefore, rays can propagate a short distance away
from the geometry before disappearing.
• For Bounding box, user defined, a ray is immediately terminated if it reaches the
surface of a box of user-specified dimensions. To define the box, enter maximum
and minimum values of each spatial coordinate.
• For None, the rays that propagate outward from the geometry will continue to
propagate until the study is complete. If the maximum path length for the
Ray Tracing study step is too large, this may result in trajectories that are extremely
long relative to the characteristic geometry size.

THE RAY ACOUSTICS INTERFACE | 693

 If rays are released from outside the bounding box, they won’t immediately disappear;
instead, they will only disappear if they enter the bounding box and then exit it again.

Intensity- and Power-Based Termination Criteria

If the ray intensity or power is being computed in the model, you can then terminate
rays based on either or both of these quantities.

Select an option from the Additional termination criteria list. None (the default) is always
available. If ray intensity is solved for, then Intensity is available. If ray power is solved
for, then Power is available. If both ray intensity and power are solved for, the Intensity
and Power option is also shown.

• For Intensity or Intensity and power, enter the Threshold intensity Ith
(SI unit: W/m2). The default is 1e-12[W/m^2]. Rays will be removed from the
model if their intensity decreases below the specified value.
• For Power or Intensity and power, enter the Threshold power Qth (SI unit: W). The
default is 1e-3[mW].

The Ray Termination feature can terminate rays at the exact time at which
their intensity or power reaches the specified threshold (to within
numerical precision), if there exists a closed-form analytic expression for
this time. Such a closed-form analytic solution exists if:

• The rays represent planar wavefronts in an absorbing medium, or

• The rays represent spherical or ellipsoidal wavefronts in a nonabsorbing
medium.

Otherwise, the rays are stopped at the first discrete time step taken by the
solver for which the intensity or power is less than the threshold value.

Ray Termination Theory

Ray Detector
Use the Ray Detector feature to compute information about rays that are located in a
set of selected domains or on a set of selected boundaries. The detector may detect all
rays or only the rays released by a specified release feature. Computed variables are the

694 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

number of rays transmitted, the transmission probability, and a logical expression
which can be used to filter the rendered rays during results processing.

The Ray Detector feature is also capable of computing the total transmitted ray power.
If the ray power is solved for, then a global variable for the total power transmitted to
the detector is available.

RAY DETECTOR
Select an option from the Release feature list: All (the default), or any of the features
which are capable of releasing rays. When All is selected, the variables computed are
summed over all release features in the model.

THE RAY ACOUSTICS INTERFACE | 695

 Modeling with the Ray Acoustics
Interface
In this section:

• Mixed Diffuse and Specular Wall Conditions

• Assigning Directivity to a Source
• Impulse Response Plot and Receiver Dataset
• Stopping Rays for a Given Condition
• Mesh Guidelines
• Nonlocal Couplings
• Using Ray Detectors
• Other Results Plots, Datasets, and Derived Values

Mixed Diffuse and Specular Wall Conditions

At a wall where both specular reflections and diffuse reflections occur it is common to
define the scattering coefficient s. This coefficient relates the amount of energy that is
specularly Es and diffusely Ed reflected to the (total) absorption coefficient α. The
normalized total reflected energy is Et. The relations between these are

Et = 1 – α = Es + Ed
Es = ( 1 – α ) ( 1 – s ) = 1 – αs
E d = ( 1 – α )s = 1 – α d

where αs is the specular absorption coefficient and αd is the diffuse absorption

coefficient.

In the Ray Acoustics interface the mixed reflection condition is treated using a Monte
Carlo like approach. When a ray hits a mixed wall condition the ray is either diffusely
or specularly reflected according to the Probability of specular reflection γs. This means
that the probability plays the role of the scattering coefficient, by setting γs = 1 − s. The
Specular absorption coefficient and the Diffuse absorption coefficient should both be set
equal to the total absorption α.

696 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Assigning Directivity to a Source
Acoustic sources often have a certain directivity, for example, the directivity of a
loudspeaker or even the directivity of a human speaker. When setting up sources in a
ray acoustics model this directivity can be specified using the Source with Directivity
feature. The spatial directivity can be imported from measured data using an
interpolation function that depends on the two polar angles as well as the band center
frequency.

As an alternative the source directivity, for a single frequency simulation, can be

automatically coupled from a pressure acoustics model using the Release from Exterior
Field Calculation feature.

Impulse Response Plot and Receiver Dataset

The impulse response (IR) in a room acoustics simulation can be analyzed by
collecting the ray information using the Receiver dataset and then postprocess using the
Impulse Response plot. Room acoustic metrics like reverberation times, early energy
metrics, and speech intelligibility can also be postprocessed using the Energy Decay
subfeature. The IR can be determined for sources and boundary data (absorption
coefficients, scattering parameters, source directivity, volume attenuation etc.) given in
octave, 1/3 octave, or 1/6 octave bands.

PREPARING A ROOM ACOUSTICS SIMULATION

When setting up a Ray Acoustics simulation and preparing it for a room acoustics
simulation where the impulse response (IR) is an important output, certain
considerations and steps need to be taken. They are as follows:

• Decide on the interpretation of the frequency in the ray acoustics simulation. To be

able to reconstruct the IR with the dedicated plot, the frequency should represent
the center frequency of octave, 1/3 octave, or 1/6 octave bands. The center
frequency is also used as input for boundary conditions (absorption coefficients and
scattering parameter), for sources (source strength and directivity), and for the
volume absorption in the medium, typically air.
• Create parameters under Global Definitions > Parameters for the band center
frequency and the source power, for example, f0 and P0 respectively.
• Set up interpolation functions for the properties that depend on the frequency. The
data can easily be stored in a Spreadsheet formatted file and interpolation can be set
to Nearest neighbor. An example of the format for a .txt file that specifies

MODELING WITH THE RAY ACOUSTICS INTERFACE | 697

 absorption coefficients for some boundaries in a simulation could be (data is
invented):
%f0, wall, entrance, window, floor, diffusers, seats
125 0.1 0.4 0.01 0.02 0.1 0.05
250 0.07 0.2 0.01 0.03 0.1 0.06
500 0.05 0.12 0.02 0.05 0.1 0.07
1000 0.04 0.07 0.02 0.1 0.1 0.15
2000 0.04 0.05 0.02 0.3 0.1 0.35
4000 0.05 0.05 0.03 0.5 0.1 0.4
8000 0.08 0.05 0.04 0.5 0.1 0.6
16000 0.1 0.05 0.05 0.5 0.1 0.6

• Create parameters under Global Definitions > Parameters for the receiver radius and
number of released rays. These two quantities are linked by the expected error in a
given time interval Δt of the impulse response, see Ref. 6 for details. With the room
volume V, the receiver radius R, and for an expected error of 1 dB, the number of
released rays N needed is
2 V -
N = 4.34 -------------------
2
πR cΔt

The receiver radius can be taken as R = 0.3 m to match the width of a common seat,
and a standard time interval is Δt = 0.01 s. For small rooms or car cabins it can be
useful to set a receiver radius that represents the actual size of a microphone or
slightly larger. The value obtained from this calculation should be rounded up to
ensure an integer number of released rays. A higher value can also be chosen and
will result in increased resolution.
• Make sure that either the Compute intensity and power or the Compute power option
is selected in the Intensity Computation section. The Compute power option
requires fewer degrees of freedom and is more robust. If it is selected the ray
intensity (rac.I and [Link]) cannot be visualized in ray trajectories plots, but
both the Sound Pressure Level Calculation on boundaries and the impulse response
can be computed.
• Make sure that Count reflections is selected in the Additional Variables section.
• Also make sure that the parameter that you created for the band center frequency
(say f0) is used as the Ray frequency under the Ray Properties node.
• Set up the ray acoustics model with appropriate boundary conditions and sources.
Generally, all Wall Conditions should be set to Mixed diffuse and specular reflection,
with a default scattering coefficient s = 0.05 for flat surfaces.

698 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

• Create a Ray Termination node and choose Bounding box, from geometry as Spatial
extents of ray propagation. Select Power as Additional termination criteria and enter
the threshold P0/N*1e-6. For an omnidirectional source, this expression makes rays
disappear once the power they carry has dropped by 60 dB from its initial value.
• To be able to postprocess the IR, a Parametric Sweep needs to be used in the study
around the Ray Tracing study step. The sweep parameter should be the frequency
defined in the parameters. To easily create a Parameter value list representing the
center frequencies of the bands use the ISO preferred frequencies entry method.
• For the Times, specified under the Ray Tracing study step, only enter 0 and the final
simulation time (the final time should be slightly larger than the estimated
reverberation time). COMSOL uses small internal steps that accurately account for
all reflections. The so-called Extra Time Steps are used when reconstructing the IR
and other data (see below).
• Run the simulation.

RECEIVER DATASET
Use a Receiver 2D ( ) or Receiver 3D dataset ( ), selected from the More 2D Datasets
submenu and the More 3D Datasets submenu, respectively, to collect the data necessary
to visualize the impulse response using an Impulse Response Plot. The Impulse
Response plot uses the data from a Receiver dataset as input

Data
Select the appropriate ray dataset in the Dataset list. Then select the frequency
parameters that should be used by the receiver; select one frequency for a single band
analysis or all for the broad band analysis. This is the typical behavior if a parametric
sweep has been set up, as described in Preparing a Room Acoustics Simulation.

Receiver
The receiver behaves as a (transparent) sphere with a given radius and located at a
center position (the microphone needs to have a certain finite size such that there is a
reasonable probability of rays interacting with it). By changing the center position the
receiver can be moved around without the need to run the simulation again.

• Under Center, specify the x, y, and (3D only) z coordinates for the center of the
receiver.
• Under Radius, specify the radius of the receiver. From the Radius input list, choose
Expression to determine the radius using an expression (see below) or choose User
defined to enter a value for the radius in the Radius field (SI unit: m). Different
theories exist for the appropriate size of the receiver. For room acoustic applications,

MODELING WITH THE RAY ACOUSTICS INTERFACE | 699

 it is recommended to use a fixed radius from which the number of rays can be
derived, as described in Preparing a Room Acoustics Simulation.
When selected, the following built-in Expression determines the radius R of the
receiver:

4
R = log 10 ( V )d SR ----
N

where you enter the values for the Number of rays, N; Room volume, V (SI unit: m2);
and Source-receiver distance, dSR (SI unit: m), to determine the radius. See Ref. 19
for details on the expression. This computed receiver size is optimized for modeling
of large rooms. However, it will return a different value for each source-receiver pair
if there are multiple sources and receivers present in the simulation, and it will not
necessarily correspond to the dimensions of a listener.

Directivity
From the Directivity type list, choose Omnidirectional (the default), User defined (dB),
or User defined (linear) and enter a directivity expression in the Expression field.

• The User defined (dB) option applies the given gain to the received signal (positive
or negative gain).
• The User defined (linear) option applies a linear gain. If a negative value is used the
phase of the arriving ray is inverted.

The expression can depend on the ray direction vector components ([Link],
[Link], [Link]) in order to set up a complex directivity pattern.

Extra Time Steps

Use the default All option. The impulse response plot has been tuned to use this option
in order to get accurate and fast evaluation of arrival time and ray power.

Advanced
For the Interpolation between time steps list always use the default Linear interpolation
option. The Cubic interpolation is not applicable for the impulse response computation
and can lead to erroneous results.

Under Normal variables and Other variables, if desired, you can change the default
names of the created variables:

• The normal direction of the receiver surface (nx, ny, and, in 3D, nz).
• The Distance traveled, that is the distance traveled by a ray inside the receiver.
• The receiver Volume.

700 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

• The Directivity, that can be used to visualize the expression for the directivity.
• The First ray arrival time.

Evaluation and Export

The main purpose for the Receiver dataset is to be used as input for the Impulse
Response Plot. However, the data generated by the dataset can also be exported and
used in an external analysis tool or software (ray power and arrival times).

Right-click the dataset and select Add Data to Export, then enter the desired data. For
example, export the arrival time of the rays t, the frequency of the rays rac.f, the
power of the rays rac.Q, the distance traveled by the rays in receiver re1dist, and so
forth. The exported data can be stored and sorted in several formats.

When working with phase alignment of loudspeakers, it can be useful to evaluate the
arrival time if the first ray. This is easily done in a Derived Values > Global Evaluation
feature that points to the receiver dataset. Simply evaluate the special variable
re1first.

See also the Receiver 2D and Receiver 3D section in the COMSOL

Multiphysics Reference Manual.

IMPULSE RESPONSE PLOT

In room acoustics applications, the impulse response (IR) of a source-receiver
configuration represents one of the most important postprocessing results. To create
the plot based on the data collected by the Receiver dataset, add an Impulse Response
subnode ( ) to a 1D Plot Group to create the impulse response plot. After defining
the characteristics of the impulse response, click Plot ( ) to create the plot. The
Energy Decay ( ) subnode can be added to perform an analysis of the impulse
response to get room acoustic metrics and the energy/level decay curves. The first
rendering of the IR plot can take some time. As the data is cached any subsequent
changes to the plot will be fast.

The Small Concert Hall Acoustics model is a tutorial on how to compute

the impulse response using the Ray Acoustics interface. Application
Library path: Acoustics_Module/Building_and_Room_Acoustics/
small_concert_hall

MODELING WITH THE RAY ACOUSTICS INTERFACE | 701

 In the ray tracing methodology it is, as discussed above, typical to characterize sources
and boundaries in frequency bands. This also means that some of the frequency
content is lost. This frequency content is reconstructed when determining or
reconstructing the IR. For each ray that interacts with the receiver, a small piece of
signal with appropriate amplitude, arrival time, phase, and frequency content is added
to the IR (the filter bank kernel). The sum of all the contributions reconstruct the full
IR.

Data
From the Dataset list, choose the appropriate Receiver Dataset. Then from the
Frequency interpretation list choose an interpretation of the frequency: Octave (the
default), 1/3 octave, or 1/6 octave. This selections should coincide with the frequency
interpretation used in boundary conditions and sources in the underlying ray acoustics
simulation, as described in Preparing a Room Acoustics Simulation. The frequency
content and resolution of the impulse response is based on this selection.

Expressions
Set up the variables that are necessary for reconstructing the impulse response
(typically use the defaults). These variables are used to define and add the contribution
of each ray, that intersects the receiver dataset, to the impulse response.

• Define the frequency variable for the rays (SI unit: Hz) in the Frequency field. The
default is rac.f.
• Define the power variable for the rays (SI unit: kg·m2/s2) in the Power field. The
default is rac.Q.
• Define the density (SI unit: kg/m3) in the Density field. The default is [Link].
• Define the speed of sound (SI unit: m/s) in the Speed of Sound field. The default is
rac.c.

• Define the number of reflections undergone by the rays (SI unit: 1) in the Number
of reflections field. The default is [Link].

x-Axis Data
From the Transformation list, choose a transformation of the data on the x-axis: None
(the default), for no transformation, to show the time domain signal, or Frequency
spectrum. The latter will perform an FFT of the IR and depict the power spectra as
function of frequency.

702 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Advanced
The following advanced settings are available to control some aspects of the IR
reconstruction as well as the underlying filter kernel used.

• Define a sampling frequency fs (SI unit: Hz). The default is 44100.

• Define the Zero passing length. The default is 11025.

The next two options are legacy options (to achieve behavior for COMSOL
Multiphysics releases older than version 5.6).

• Select Remove noncausal signal to remove the signal (force it to zero) prior to the
arrival of the first ray. Using this option will alter the energy content of the early
energy component of the IR.
• Select Use fully randomized phase to use the legacy method for reconstructing the
IR. The fully randomized phase method (see Ref. 20) can lead to unwanted
cancellations of the direct sound and early energy contributions.

The following options control the filter kernel used to reconstruct the impulse
response. Advanced options allow to modify the kernel, enter a user-defined analytical
kernel, or a user-defined kernel based on an interpolation function. The default is to
use a Brick-wall with Kaiser window filter kernel (see Ref. 21).

• Select Show the filters to show and inspect the filter bank in the Graphics window,
click Plot ( ) to create the plot. To switch between time domain and frequency
domain representation change the Transformation option in the x-Axis Data section.
• Select the Filter kernel to use Brick-wall with Kaiser window (the default) or User
defined. When User defined is selected enter the expression for the filter kernel hfc[t].
Use and set up parameters in the Parameters table (as you would use variables). The
parameter fc is reserved for the band center frequency; fs is reserved for the
sampling frequency; t is for time; and Np is the padding length. Note that per default
the User defined option sets up the brick-wall with Kaiser window filter kernel for
octaves.
• If the default Brick-wall with Kaiser window is selected, you can control the ripple
factor δ (the default is 0.05). The effect on the filter can be seen using the Show the
filters option.

MODELING WITH THE RAY ACOUSTICS INTERFACE | 703

 Coloring and Style, and Legends
You can also make change in the Coloring and Style section or add and format a legend
in the Legends section.

See also the Impulse Response section tn the COMSOL Multiphysics

Reference Manual.

ENERGY DECAY AND OBJECTIVE ROOM ACOUSTIC METRICS

For further analysis of the impulse response, add the Energy Decay ( ) subnode to
the Impulse Response plot. Once selections have been made, click Plot ( ) to create
the plot.

Display
Select the Band type as Broadband (the default) or Individual bands. For the Individual
bands option select the Band frequency, either an individual band or All. For All, all the
level or energy decay curves are plotted.

Select the Plot as Energy decay (the default), Level decay, or Modulation transfer function,
to define the type of graphical analysis of the impulse response to be plotted in the
Graphics window.

Table
Select the room acoustic metrics to be included in the results table, per default all
objective quality metrics are selected. The default name of the table is Objective Quality
Metrics. The table displays the results in columns with the band center frequency fc in
the first column. The results table can also be located under the Results > Tables node.
A plot of the results located in the table can be created using the Table Graph plot.

The speech transmission index (STI) metric is defined as a single value

that combines information from 14 modulation transfer function bands
and 7 octave bands. To get the value specified by the standard solve the
model for octaves from 125 Hz to 8 kHz. When evaluating use the Band
type set to the default Broadband option.

CUSTOM PLOTS
The energy response depicted as a reflectogram or echogram can be plotted using a
Ray (Plot) and depict the expression re1dist*rac.Q/re1vol. This will give the
intensity received by the “microphone” for each ray.

704 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Stopping Rays for a Given Condition
For some ray acoustics applications it can be advantageous to stop rays once their
intensity falls below a certain threshold. This can be simply achieved at walls using the
Primary ray condition with an Expression. Here you can enter a logical expression that
must be satisfied for a ray to undergo a certain type of ray-wall interaction. For the
intensity specify a condition of the type rac.I<threshold. Then the ray is either
reflected according the selected condition or it can Freeze or Disappear, depending on
your selection.

The number of reflections can be used in such an Expression. This variable is readily set
up under the Additional Variables section.

The Ray Termination feature can also be used in a similar manner to remove rays for
certain conditions, without the need for a wall interaction.

Mesh Guidelines
When setting up a mesh for a Ray Acoustics simulation the following guidelines
should be followed:

• For modeling ray propagation within domains and interaction with flat surfaces, the
mesh can be very coarse.
• When rays may interact with curved boundaries (either reflection or refraction), the
accuracy of the reinitialized ray trajectory (that is, the reinitialized wave vector k)
depends on the mesh. To create a very fine mesh on curved surfaces without making
the mesh unnecessarily fine elsewhere, use the Curvature factor in the Size settings
window. The closer this number is to zero, the finer the mesh will be on a curved
surface. A value of 0.1 seems adequate for most applications. If there are small,
tightly curved surfaces in the model, it might also be necessary to reduce the
Minimum element size to allow the small curvature factor to do its job.
• When computing ray intensity, the accuracy of the reinitialized intensity at a curved
wall or material discontinuity usually doesn't depend on the level of mesh
refinement because it uses surface curvature variables that are based directly on the
geometry, not on the mesh. However, if ALE is included in the model, the accuracy
of the surface curvature does depend on the mesh.
• Domains that are not included in the selection for the Ray Acoustics interface do
not need to be meshed. However, a boundary mesh must always be defined
wherever any boundary condition is applied. In addition, it is assumed that the
speed of sound and density in all deselected domains are the Speed of sound in

MODELING WITH THE RAY ACOUSTICS INTERFACE | 705

 exterior domains and Density of exterior domains, which are specified in the physics
interface Ray Release and Propagation section. Thus, media with graded or
temperature-dependent material properties must always be included in the selection
of the Ray Acoustics interface, and they must always be meshed.
• Accumulator features (both on surfaces and on domains) work by defining additional
dependent variables using constant shape functions. When using an Accumulator on
a Wall, for example, the contribution from each ray is distributed uniformly over
whatever boundary element it hits. If the mesh on the boundary is very fine, this
will make the value of the accumulated variable look noisy unless the number of rays
hitting the boundary is much larger than the number of boundary elements.

REPRESENTATION OF CURVED SURFACES

When rays reach the boundaries of geometric entities in a model, they do not interact
with an exact parameterized representation of the geometry. Rather, they propagate
through the mesh elements that discretize the modeling domain and interact with the
boundary elements that cover the surfaces of the geometric entities.

When the surfaces of the geometry are flat, the shape of the surface mesh is
indistinguishable from the shape of the geometric entities themselves. Therefore, the
fact that rays interact with the mesh instead of the geometry does not introduce any
discretization error, and it is possible to accurately compute ray trajectories even when
the mesh is extremely coarse.

Curved surfaces in the geometry, however, usually incur a significant amount of

discretization error when predicting how rays will interact with them. The time and
location at which the ray interacts with the boundary mesh element might be slightly
different from the time at which it would have interacted with an exact representation
of the surface. In addition, the tangential and normal directions on the boundary mesh
element may differ from the tangential and normal directions on the surface, affecting
the accuracy of boundary conditions that involve the tangential and normal directions,
such as the Specular reflection condition.

The order of the curved mesh elements used to determine the geometry shape is
controlled by the Geometry shape order list in the Model Settings section of the Settings
window for the main Component node. If Automatic, the default, is selected, the curved
mesh elements are usually represented by quadratic curves; in some cases, linear
functions are used to prevent inverted mesh elements from being created.

The effect of the geometry shape order is most notable on a coarse mesh, as shown in
Figure 8-6. The mesh elements are shown as pale gray lines in the background and the

706 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

ray trajectories are represented as thick red arrows. The rays initially propagate
downward and are specularly reflected by a parabolic surface. If Linear is selected from
the Geometry shape order list, all rays that hit the same boundary element are specularly
reflected in the same direction, as shown on the left. Even though the bottom surface
is parabolic, the rays don’t all intersect at a single focus due to the discretization error.
If Quadratic or Automatic is selected, rays that hit the same boundary element can still
be reflected in different directions because the tangential and normal directions can
vary along the surface of the curved element. As a result, the rays reflected by the
parabolic surface all intersect at a well-defined focal point as expected.

Figure 8-6: Comparison of rays being specularly reflected at a curved boundary

represented using linear elements (left) and quadratic elements (right).

RAY TRACING IN AN IMPORTED MESH

It is also possible to compute ray trajectories in an imported mesh. The mesh can be
imported from a COMSOL Multiphysics file (.mphbin for a binary file format
or .mphtxt for a text file format) or from a NASTRAN file (.nas, .bdf, .nastran,
or .dat).

If the mesh is imported from a COMSOL Multiphysics file, the imported mesh always
uses linear geometry shape order for the purpose of modeling ray-boundary
interactions, even if the model used to generate the mesh had a higher geometry shape
order.

If the mesh is imported from a NASTRAN file, the ray-boundary interactions may be
modeled using either linear or higher geometry shape order. If Export as linear elements

MODELING WITH THE RAY ACOUSTICS INTERFACE | 707

 is selected when generating the NASTRAN file, or if Import as linear elements is
selected when importing the file, then linear geometry shape order will be used.

Nonlocal Couplings
The purpose of a model is often to compute the sum, average, maximum value, or
minimum value of a quantity over a group of rays, such as the average intensity or the
maximum path length. An instance of the Ray Acoustics interface with Name rac
creates the following four nonlocal couplings:

The built-in nonlocal coupling names have recently changed, but the
couplings with the old names are still available. In their descriptions the
couplings with the old names are marked as “deprecated”. Both the old
and new names are listed in Table 8-3.

• [Link](expr) evaluates the sum of the expression expr over the rays. The sum
includes all rays that are active, frozen, or stuck to boundaries. It excludes rays that
have not yet been released and those that have disappeared.
• rac.sum_all(expr) evaluates the sum of the expression expr over all rays,
including those rays that are not yet released or have disappeared. Since the
coordinates of unreleased and disappeared rays are not-a-number (NaN), the sum
may return NaN if the model includes unreleased or disappeared rays. An expression
such as rac.sum_all(isnan(qx)) can be used to compute the total number of
unreleased and disappeared rays.
• <name>.<name>aveop1(expr) evaluates the average of the expression expr over
the active, frozen, and stuck rays. Unreleased and disappeared rays contribute to
neither the numerator nor the denominator of the arithmetic mean.
• <name>.<name>aveop_all1(expr) evaluates the average of the expression expr
over all rays. It is likely to return NaN if the model includes unreleased or disappeared
rays.
• [Link](expr) evaluates the maximum value of the expression expr over all active,
frozen, and stuck rays.
• rac.max_all(expr) evaluates the maximum value of the expression expr over all
rays.

708 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

• [Link](expr, evalExpr) evaluates the expression evalExpr for the ray that has
the maximum value of the expression expr out of all active, frozen, and stuck rays.
For example, the expression [Link](rac.I, qx) would evaluate the x-coordinate
qx of the ray with the greatest intensity rac.I.
• rac.max_all(expr, evalExpr) evaluates the expression evalExpr for the ray
that has the maximum value of the expression expr for all rays, including
disappeared and unreleased rays.
• [Link](expr) evaluates the minimum value of the expression expr over the
active, frozen, and stuck rays.
• rac.min_all(expr) evaluates the minimum value of the expression expr over all
rays.
• [Link](expr, evalExpr) evaluates the expression evalExpr for the ray that has
the minimum value of the expression expr out of all active, frozen, and stuck rays.
For example, the expression [Link](rac.I, qx) would evaluate the x-coordinate
qx of the ray with the minimum intensity rac.I.

• rac.min_all(expr, evalExpr) evaluates the expression evalExpr for the ray

that has the minimum value of the expression expr for all rays, including
disappeared and unreleased rays.

An instance of the Ray Acoustics interface with the default name rac defines the
built-in nonlocal couplings shown in Table 8-3.
TABLE 8-3: BUILT-IN NONLOCAL COUPLINGS FOR THE RAY ACOUSTICS INTERFACE

NAME NAME DESCRIPTION

[Link](a) rac.racop1(a) Sum over rays

rac.sum_all(a) rac.racop_all1(a) Sum over all rays
[Link](a) rac.racaveop1(a) Average over rays
rac.ave_all(a) rac.racaveop_all1(a) Average over all rays
[Link](a) rac.racmaxop1(a) Maximum over rays
rac.max_all(a) rac.racmaxop_all1(a) Maximum over all rays
[Link](a) rac.racminop1(a) Minimum over rays
rac.min_all(a) rac.racminop_all1(a) Minimum over all rays
[Link](a,b) rac.racmaxop1(a,b) At maximum over rays
rac.max_all(a,b) rac.racmaxop_all1(a,b) At maximum over all rays
[Link](a,b) rac.racminop1(a,b) At minimum over rays
rac.min_all(a,b) rac.racminop_all1(a,b) At minimum over all rays

MODELING WITH THE RAY ACOUSTICS INTERFACE | 709

 Using Ray Detectors
A Ray Detector feature is a domain or boundary feature that provides information
about rays arriving on a set of selected domains or surfaces from a release feature. Such
quantities include the number of rays transmitted, the transmission probability, and a
logical expression for ray inclusion. The feature provides convenient expressions that
can be used in the Filter node of the Ray Trajectories plot, which allows only the rays
which reach the ray detector selection to be visualized. The following variables are
provided by the Ray Detector feature, with the feature tag <tag>:

• <tag>.Ntf is number of transmitted rays from the release feature to the ray detector
at the end of the simulation.
• <tag>.alpha is the transmission probability from the release feature to the ray
detector.
• <tag>.rL is a logical expression for ray inclusion. This can be set in the Filter node
of the Ray Trajectories plot in order to visualize the rays which connect the release
feature to the detector.

The Ray Detector feature only creates variables, which do not affect the
solution. Therefore, they can be added to a model without the need to
re-compute the solution, it just needs to be updated. To do this, right
click on the Study node and select Update Solution. The new variables
described above will be immediately available for results processing.

Other Results Plots, Datasets, and Derived Values

A variety of dedicated postprocessing tools are available with this physics interface.
They are listed below and referred to in the COMSOL Multiphysics Reference
Manual.

SOUND PRESSURE LEVEL ON SURFACES AND IN CROSS SECTIONS

The sound pressure level on surfaces is plotted using the Sound Pressure Level
Calculation subfeature to a Wall node.

The SPL can also be visualized on an interior cut-like plane by adding a plane in the
geometry, meshing it, and then apply a wall with the Wall condition set to Pass through.
In this way the boundary will only be used to map the SPL and will have no influence
on the ray propagation.

710 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

PLOTS

In the COMSOL Multiphysics Reference Manual:

• Ray (Plot)
• Ray Trajectories
• Filter for Ray and Ray Trajectories
• Particle (Plot)
• Impulse Response
• Plot Groups and Plots

RAY DATASET

In the COMSOL Multiphysics Reference Manual:

• Ray (Dataset)
• Receiver 2D and Receiver 3D
• Datasets

RAY EVALUATION DERIVED VALUES

In the COMSOL Multiphysics Reference Manual:

• Ray Evaluation
• Derived Values, Evaluation Groups, and Tables

MODELING WITH THE RAY ACOUSTICS INTERFACE | 711

 Theory for the Ray Acoustics
Interface
The Ray Acoustics Interface theory is described in this section:

• Introduction to Ray Acoustics

• Initial Conditions: Position
• Initial Conditions: Direction
• Material Discontinuity Theory
• Intensity and Wavefront Curvature
• Intensity and Phase Reinitialization
• Wavefront Curvature Calculation in Graded Media
• Attenuation Within Domains
• Ray Termination Theory
• Accumulator Theory: Domains
• Accumulator Theory: Boundaries
• Sound Pressure Level Calculation Theory
• Assigning Spatial Directivity to a Source
• References for the Ray Acoustics Interface

Introduction to Ray Acoustics

It is often convenient to express an acoustic wave in a stationary fluid as a plane wave
of the form

iΨ
φ = ae

where φ (SI unit: m2/s) is the velocity potential of the fluid, a is its amplitude of the
velocity potential, and Ψ (SI unit: rad) is the phase. When the distance from any source
is many orders of magnitude larger than the wavelength of the acoustic wave, the wave
may be assumed to be locally plane; that is,

Ψ = k ⋅ q – kct

712 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

where k (SI unit: rad/m) is the wave vector, q (SI unit: m) is the position vector,
c (SI unit: m/s) is the speed of sound in the medium, and t (SI unit: s) is the time.

While the velocity potential is defined with a stationary fluid in mind, it is possible to
apply the same treatment to a homogeneous fluid moving at velocity u (SI unit: m/s)
by first formulating the equation of the acoustic wave in a coordinate system in which
the fluid is stationary. This results in a more general form of the acoustic wave
equation,

Ψ = k ⋅ q – ( kc + k ⋅ u )t

By defining the angular frequency of the wave as

ω = ck + u ⋅ k

it follows that the equation of the acoustic wave may be expressed in the Hamiltonian
form

dk- ∂ω
------ = – -------
dt ∂q
dq- ∂ω
------ = -------
dt ∂k

This is a mixed time-frequency formulation that has many advantages. See, for
example, Ref. 1 and Ref. 5 for further details. It allows a ray acoustic problem to be
broken down into its Fourier components, solving for the propagation of one
frequency component per ray. Another benefit is the simple use of frequency
dependent boundary conditions.

Initial Conditions: Position

The ray release positions may be mesh based, from a user-specified grid, uniformly
distributed, loaded from a file, or based on an analytic expression.

LOADING INITIAL COORDINATES FROM FILE

If the initial ray positions are loaded from a text file using the Release from Data File
node, the initial positions may be scaled, rotated, translated, or some combination of
these. Assuming q0 to be the loaded ray release position, q0 can be rotated using the
rotation matrix A

q 0' = Aq 0

where A in 3D is given by

THEORY FOR THE RAY ACOUSTICS INTERFACE | 713

 cos α cos γ – sin α cos β sin γ – cos α sin γ – sin α cos β cos γ sin α sin β
A = sin α cos γ + cos α cos β sin γ – sin α sin γ + cos α cos β cos γ – cos α sin β
sin β sin γ sin β cos γ cos β

and in 2D by

A = cos α – sin α
sin α cos α

In 3D, α, β, and γ are the Euler angles:

• First α rotates the position about the space-fixed z-axis.

• Then β rotates the position about the x’-axis, i.e. the body-fixed X-axis after rotation
about the z-axis.
• Finally γ rotates the position about the z’’-axis, i.e. the body-fixed Z-axis after
rotation about the z-axis and x’-axis.

In 2D, α is the angle of counterclockwise rotation about the origin.

All components of the ray release position vector are multiplied by the Scale factor,

q 0'' = Rq 0'

Finally, the initial positions may be translated by a displacement vector Δq,

q 0''' = q 0'' + Δq

The primes indicate that translation is applied after any rotation or dilation (scaling).
However, because the dilation applies equally to all vector components, the dilation
and rotation commute with each other, so their relative order does not matter.

Initial Conditions: Direction

There are several options available for specifying the initial direction of rays. These
options work by assigning values to the degrees of freedom corresponding to the wave
vector k (SI unit: rad/m) of each ray.

EXPRESSION
The default is to enter an expression for the Ray direction vector L0 (dimensionless).
This vector need not have a magnitude of unity because it is always normalized
automatically. The initial wave vector is

714 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 ωL 0
k = -------------
c L0

where ω (SI unit: rad/s) is the angular frequency and c (SI unit: m/s) is the speed of
sound in the medium.

LOADING INITIAL DIRECTION FROM FILE

If the initial ray positions are loaded from a text file, then optionally the initial ray
direction can also be loaded from the same file. If the direction is loaded from the file
and a rotation has been applied to the ray release positions, then optionally the same
rotation can be applied to the initial direction.

SPHERICAL
When Spherical is selected the initial wave vectors are sampled from a distribution in
wave vector space at each release point. The number of rays released from each point
is usually equal to the specified value Nw (dimensionless), although it may be larger if
the initial values of any auxiliary dependent variables are also sampled from a
distribution.

In 2D the initial wave vector components are

ω
k x = ---- cos θ
c
ω
k y = ---- sin θ
c

where θ goes from 0 to 2π in Nw steps. In 3D the initial wave vector components are
sampled according to the expressions

ω
k x = ---- sin θ cos ϕ
c
ω
k y = ---- sin θ sin ϕ
c
ω
k z = ---- cos θ
c

The azimuthal angle ϕ is uniformly distributed from 0 to 2π. The polar angle θ is
sampled from the interval [0, π] with probability density proportional to sin θ. The
polar angle is arbitrarily chosen as the angle that the initial wave vector makes with the
positive z-axis, but any direction could be chosen because the sphere is isotropic.
Therefore, each ray subtends approximately the same solid angle in wave vector space.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 715

 HEMISPHERICAL
The Hemispherical option is the same as the Spherical option, except that in 2D θ goes
from 0 to π and in 3D θ goes from 0 to π/2. The angle is measured from the direction
given by the Hemisphere axis setting.

CONICAL
The Conical option is the same as the Spherical option, except that θ goes from 0 to α.
The angle is measured from the direction given by the Cone axis setting.

LAMBERTIAN
The Lambertian option releases rays within a hemisphere in 3D, but the probability
distribution function is different from that of the Hemispherical option. Recall that for
an isotropic hemispherical distribution the polar angle θ has a probability density
proportional to sin θ; for the Lambertian distribution the probability density is instead
proportional to sin θ cos θ. Because of this extra cosine term, distributions following
this probability density are said to follow Lambert’s cosine law.

Material Discontinuity Theory

At a Material Discontinuity between two different media, the wave vector is
reinitialized using Snell’s law. First, the angle of incidence θi is computed:

ni ⋅ ns
θ i = acos  ------------------
 ni ns 

where ni is a unit vector in the direction of the incident ray and ns is a unit vector
normal to the material discontinuity.

At a boundary between two isotropic, nonabsorbing media, the refracted ray

propagates in the direction nt given by the following relations:

n t = ηn i + γn s

γ = – η cos θ i + cos θ t

c2
η = -----
c1

θ t = asin ( η sin θ i )

716 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

where the ray propagates from the medium with speed of sound c1 into the medium
with speed of sound c2.

If the maximum number of secondary rays has not yet been released, a reflected ray is
also released at the material discontinuity with the initial direction

n r = n i – 2n s cos θ i

If the ray intensity is computed, the release of secondary rays may be suppressed when
the intensity of the reflected ray is less than the user-defined Threshold intensity Ith
(SI unit: W/m2). Similarly, when ray power is computed, a Threshold power Qth
(SI unit: W) can be specified.

Intensity and Wavefront Curvature

The following describes the algorithm used to compute the ray intensity when Intensity
computation is set to Compute intensity or Compute intensity and power in the physics
interface Intensity Computation section. In this algorithm, each ray is treated as a
propagating wavefront subtending a small solid angle. The algorithm is only valid
when the rays propagate in homogeneous media. The following auxiliary dependent
variables are defined for each ray:

In 3D:

• The initial ray intensity I0, which is reinitialized at material discontinuities and walls.
• The integral A of the attenuation coefficient along the ray path.
• Two initial principal radii of curvature, r1,i and r2,i.
• Two principal radii of curvature, r1 and r2, which represent the maximum and
minimum radii of curvature of the intersection of the wavefront with an arbitrary
plane.
• Three components of a unit vector e1 in the direction corresponding to the first
principal radius of curvature. This information is used to reinitialize the principal
radii of curvature at curved boundaries.

In 2D:

• The initial ray intensity I0.

• The integral A of the attenuation coefficient along the ray path.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 717

 • One initial principal radius of curvature, r1,i.
• One radius of curvature r1 of the wavefront. All wavefronts are assumed to be
converging or diverging cylindrical waves, so it is not necessary to define a second
radius of curvature.

A different algorithm is used to compute the intensity of each ray if the Intensity
computation is instead set to Compute intensity in graded media or Compute intensity and
power in graded media. This algorithm accounts for the effect of graded media on the
ray intensity, but it is also slower and less accurate for homogeneous media. The
following auxiliary dependent variables are defined for each ray:

In 3D:

• The intensity help variable Γ.

• The integral A of the attenuation coefficient along the ray path.
• Two principal curvature calculation help variables α1 and α2 and the rotation angle
φ , which indicates the orientation of the principal curvature directions. An
additional help variable is used internally to detect poles in the local coordinate
system definition and to redefine the local coordinate system accordingly.
In 2D:

• The intensity help variable Γ.

• The integral A of the attenuation coefficient along the ray path.
• One principal curvature calculation help variable α1.

For the options Compute Power, Compute intensity and power, and Compute intensity
and power in graded media an additional auxiliary dependent variable is defined to
indicate the total power transmitted by each ray. Like intensity, power is affected by
absorbing media and by reflection or refraction at boundaries. Unlike intensity, power
doesn’t change due to the focusing or divergence of a ray bundle.

PRINCIPAL RADII OF CURVATURE

In 3D models, each ray is treated as a wavefront for which two principal radii of
curvature, r1 and r2, are defined. In addition, the values of the two principal radii of
curvature are stored as r1,i and r2,i whenever the ray reaches a boundary.

Within domains, the equations

dr 1 dr
--------- = – 1 --------2- = – 1
ds ds

718 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

are solved. Negative radii of curvature indicate that the wavefront is expanding as the
ray propagates, while positive radii of curvature indicate that the wavefront is
converging. A continuous locus of points at which either of the principal radii of
curvature equals zero is called a caustic surface. The unit vector e1 is defined so that
r1 is the radius of curvature of the intersection of the wavefront with the plane tangent
to e1 and the wave vector k. Similarly, if a vector e2 is defined so that

k × e1
e 2 = ------------------
k × e1

Then r2 is the radius of curvature of the intersection of the wavefront with a plane
tangent to e2 and k. The components of e1 are stored as auxiliary dependent variables
for each ray. The components of e2 can then be derived from e1 and k at any time.

The principal radii of curvature are reinitialized at material discontinuities, and the
orientation of the unit vectors e1 and e2 may change. If the unit vector in the direction
of the incident ray ni is not parallel to the surface normal ns, then it is possible to
define a unique tangent plane, called the plane of incidence, that contains the incident,
reflected, and refracted rays. The unit vector normal to this plane, u0, is defined as

ni × ns
u 0 = ---------------------
ni × ns

If the ray is normal to the surface, then the incident ray is parallel to the refracted ray
and antiparallel to the reflected ray, and u0 can be any arbitrary unit vector orthogonal
to ni. In addition to the unit vector normal to the plane of incidence, the following
unit vectors tangent to the incident wavefront, refracted wavefront, reflected
wavefront, and surface normal are defined:

ui = ni × u0
ut = nt × u0
ur = nr × u0
us = ns × u0

where the subscripts i, t, and r denote the incident, refracted, and reflected wavefronts,
respectively. For a wavefront propagating in a direction n, with principal curvatures
k1 = 1/r1 and k2 = 1/r2 defined for directions e1 and e2, respectively, the principal
curvatures in two other orthogonal directions e1' and e2' (both orthogonal to n) are

THEORY FOR THE RAY ACOUSTICS INTERFACE | 719

 k 1' = k 1 cos2 θ + k 2 sin2 θ (8-1)
k 2' = k 1 sin2 θ + k 2 cos2 θ
k 12' = ( k 1 – k 2 ) cos θ sin θ

where θ is the angle of rotation about n which transforms the vectors e1 and e2 to e1'
and e2', respectively. Because k1' and k2' are not principal curvatures, it is necessary to
include off-diagonal elements of the wavefront curvature tensor equal to k12'.

The following algorithm is used to reinitialize the principal radii of curvature of the
wavefront and their orientations. The reinitialization of the curvature variables follows
the method of Stavroudis in Ref. 2.

1 Given ni and ns, compute unit vectors in the directions of the reflected and
refracted rays, nr and nt.
2 Compute the vectors u0, ui, ut, ur, and us.
3 Compute the angle of rotation θ(i) needed to transform the local coordinate system
with axes parallel to e1, e2, and ni to a local coordinate system with axes parallel to
u0, ui, and ni. (Note: the superscript (i) is used to distinguish this coordinate system
rotation angle from the angle of incidence of the ray, often denoted θi.)
4 Transform the principal curvatures to the local coordinate system defined by u0, ui,
and ni by using Equation 8-1, substituting θ(i) for θ.
5 Given the two principal curvatures of the surface, k1,s and k2,s with directions e1,s
and e2,s, transform the curvature variables to a local coordinate system with axes
parallel to u0, us, and ns. Let the new curvature variables be denoted by k1,s', k2,s',
and k12,s'.
6 The curvatures of the wavefront and the surface have now been defined in
coordinate systems that share the axis u0 and only differ by a rotation by the angle
of incidence θi about u0. Defining the variables η and γ as in Material Discontinuity
Theory, compute the curvature variables of the refracted ray in a coordinate system
defined by u0, ut, and nt using the equations

k 1, t' = ηk 1' + γk 1, s'

η cos θ i γ
k 12, t' = ------------------ k 12' + -------------- k 12, s'
cos θ t cos θ t
η cos2 θ i γ
k 2, t' = -------------------- k ' + ----------------- k '
cos2 θ t 2 cos2 θ t 2, s

720 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

7 Obtain the principal curvatures of the refracted ray by rotating the coordinate
system defined by u0, ut, and nt by an angle θ(t) about nt. The angle θ(t) is defined
as
2k 12, t'
θ ( t ) = --- atan  ----------------------------
1
2  k 1, t' – k 2, t'

8 Reinitialize the principal curvatures of the refracted ray:

(t) (t) (t) (t)
k 1, t = k 1, t' cos2 θ + 2k 12, t' cos θ sin θ + k 2, t' sin2 θ
(t) (t) (t)
k 2, t = k 1, t' sin2 θ – 2k 12, t' cos θ sin θ + k 2, t' cos2 θ ( t )

9 Invert the principal curvatures to obtain the principal radii of curvature of the
refracted ray. Rotate u0 about nt by the angle θ(t) to obtain the reinitialized
principal curvature direction e1,t.
10 If a reflected ray is released, compute the curvature variables of the reflected ray in
a coordinate system defined by u0, ur, and nr using the equations

k 1, r' = k 1' – 2k 1, s' cos θ i

k 12, r' = – k 12' + 2k 12, s'

2
k 2, r' = k 2' – -------------- k 2, s'
cos θ i

11 Obtain the principal curvatures of the reflected ray by rotating the coordinate
system defined by u0, ur, and nr by an angle θ(r) about nr. The angle θ(r) is defined
as
2k 12, r'
θ ( r ) = --- atan  -----------------------------
1
2  k 2, r' – k 1, r'

12 Initialize the principal curvatures of the reflected ray:

(r) (r) (r) (r)
k 1, r = k 1, r' cos2 θ + 2k 12, r' cos θ sin θ + k 2, r' sin2 θ
(r) (r) (r)
k 2, r = k 1, r' sin2 θ – 2k 12, r' cos θ sin θ + k 2, r' cos2 θ ( r )

13 Invert the principal curvatures to obtain the principal radii of curvature of the
reflected ray. Rotate u0 about nr by the angle θ(r) to obtain the initialized principal
curvature direction e1,r.

In 2D components, only one principal radius of curvature is computed; each ray is

treated as a cylindrical wave. In addition, no auxiliary degrees of freedom for the

THEORY FOR THE RAY ACOUSTICS INTERFACE | 721

 principal curvature direction are required because the out-of-plane direction can
always be treated as one of the axes of the local coordinate system that defines the
orientation of the wavefront.

In 2D axisymmetric model components, radii of curvature are computed for the

in-plane direction (that is, the rz-plane) as well as the out-of-plane or azimuthal
directions. During ray-boundary interactions, these radii of curvature are reinitialized
as if the ray interacted with a 3D surface of revolution, which may have finite radii of
curvature in both the in-plane and out-of-plane directions.

INTENSITY CALCULATION
The value of the ray intensity is stored as the auxiliary dependent variable I0 when a ray
is released. At any point along the ray’s trajectory, the intensity is equal to

r 1, i r 2, i
I = I 0 ------------------
r1 r2

In 2D, the second principal radius of curvature r2 is treated as an arbitrarily large

distance that remains constant for each ray.

Intensity and Phase Reinitialization

At a material discontinuity, the intensity of the reflected and refracted rays can be
computed using the material properties of the media in the adjacent domains:

2
Ir = I R

2
It = I ( 1 – R )

Z 2 cos θ i – Z 1 cos θ t
R = ------------------------------------------------- (8-2)
Z 2 cos θ i + Z 1 cos θ t

Z = ρc c

where ρ is the density of the medium and cc is the complex speed of sound,

c
c c = --------------------
c
1 – i ---- α
ω

722 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

where

• c (SI unit: m/s) is the real speed of sound in the medium,

• ω (SI unit: rad/s) is the angular frequency of the acoustic ray, and
• α (SI unit: 1/m) is the attenuation coefficient.

The subscripts 1 and 2 denote the sides of the boundary into which the reflected and
refracted ray propagate, respectively; see Ref. 4.

At a Wall, the reflection coefficient R can be defined explicitly or in terms of the

absorption coefficient α or characteristic impedance of absorber Z1. In terms of the
absorption coefficient,

R = 1–α

Where the principal square root is used for purposes of phase reinitialization. In terms
of the characteristic impedance of absorber,

Z 1 cos θ ⁄ Z 0 – 1
R = ----------------------------------------
Z 1 cos θ ⁄ Z 0 + 1

Z 0 = ρc c

The Wall feature also includes built-in options to compute the reflection coefficient for
fluid-fluid interfaces, fluid-solid interfaces, and a fluid layer adjacent to a semi-infinite
fluid domain. When treating the Wall boundary as a Fluid-fluid interface, the reflection
coefficient is computed using Equation 8-2 as if the wall were a material discontinuity
where the fluid properties of the adjacent medium are specified, except that no
refracted ray is produced.

If the Wall boundary is modeled as a Fluid-solid interface, the reflection coefficient is

instead defined as

Z0
Z tot – --------------
cos θ i
R = -------------------------------
Z0
Z tot + --------------
cos θ i

where

Z p, 1 2 Z s, 1 2
Z tot = ------------------- cos ( 2θ s, t ) + ------------------- sin ( 2θ s, t )
cos θ p, t cos θ s, t

THEORY FOR THE RAY ACOUSTICS INTERFACE | 723

 where the subscripts p and s refer to the propagation of compressional and shear waves
in the adjacent solid domain, respectively. For example, θp,t is the refraction angle
computed using the compressional complex speed of sound in the adjacent solid
domain.

If the Wall boundary is modeled as a Layered fluid half-space, the boundary is treated as
a thin layer of one fluid backed by a semi-infinite domain of a second fluid. The
reflection coefficient is

ζ 1 ( ζ 2 – ζ 0 ) – i ( ζ 12 – ζ 0 ζ 2 ) tan φ
R = -------------------------------------------------------------------------------
-
ζ 1 ( ζ 2 + ζ 0 ) – i ( ζ 12 – ζ 0 ζ 2 ) tan φ

Z0
ζ 0 = --------------
cos θ i
Z1
ζ 1 = -------------------
cos θ t, 1
Z2
ζ 2 = -------------------
cos θ t, 2

where the subscripts 1 and 2 refer to the properties of the thin layer and the
semi-infinite fluid domain, respectively. For example, θt,1 is the angle of refraction in
the thin layer. The angle φ is the phase delay in the thin layer,

ω
φ = ----- h 1 cos θ t, 1
c1

where h1 (SI unit: m) is the layer thickness.

If the Rayleigh roughness model is used, the reflected intensity is multiplied by an

additional factor to account for surface roughness:

2
I r = I R exp ( – 0.5Γ 2 )

Γ = 2kσ cos θ i

where k (SI unit: rad/m) is the wave vector magnitude of the ray and σ (SI unit: m) is
the RMS roughness of the surface.

If the phase of the ray is also computed, the phase shift of the reflected ray at a
boundary is

ΔΨ r = arg ( R )

724 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

At material discontinuities, the phase of the refracted ray is unchanged.

Wavefront Curvature Calculation in Graded Media

It is possible to compute the ray intensity by changing Intensity computation to
Compute intensity or Compute intensity in graded media in the Settings window for the
Ray Acoustics interface. The options Compute intensity and power and Compute
intensity and power in graded media can also be used to compute intensity, with the only
difference being that these options define an additional auxiliary dependent variable
for the total power transferred by the ray. The setting Compute intensity is more robust
and accurate than Compute intensity in graded media, but is only applicable to
homogeneous media. The setting Compute intensity in graded media can be used for
both homogeneous and graded media, but it introduces more numerical error than
Compute intensity.

ASSUMPTIONS FOR COMPUTING INTENSITY IN HOMOGENEOUS MEDIA

When Compute intensity is selected, the ray intensity is computed along each ray path
using the following assumptions:

1 The wavefront represented by each ray subtends a constant plane angle (2D) or solid
angle (3D), which can change only when rays are reflected or refracted at
boundaries.
2 Given the radii of curvature at two positions within the same domain and the
intensity at one of these two positions, the intensity at the other position can be
computed using the relationship
r 1, 0 r 2, 0
I = I 0 ---------------------
r1 r2

3 Except when rays are reflected or refracted at boundaries, the principal radii of
curvature change at a constant rate:
dr 1 dr
--------- = --------2- = – 1
ds ds

4 The principal curvature directions do not change, except when rays are reflected or
refracted at boundaries.

These assumptions are not valid in graded media, in which the speed of sound changes
continuously as a function of position. It is possible for the angle between two different
rays to change as they propagate through the medium, so the solid angle subtended by

THEORY FOR THE RAY ACOUSTICS INTERFACE | 725

 the wavefront is no longer constant. As a result, the ray intensity cannot be expressed
using the ratio of initial and final principal radii of curvature.

CURVATURE TENSOR DEFINITION IN GRADED MEDIA

The calculation of ray intensity in graded media is based on the concept of a curvature
tensor K, defined in terms of the principal curvatures κ1 and κ2 and the corresponding
principal curvature directions e1 and e2:

K = κ 1 e 1 e 1T + κ 2 e 2 e 2T

Because e1 and e2 are orthogonal, it follows that κ1 and κ2 are eigenvalues of K. It

also follows that K is singular because there is no contribution that is orthogonal to
both e1 and e2.

The signs of the principal curvatures are chosen so that positive curvature indicates that
the wavefront is converging, whereas negative curvature indicates that the wavefront
is diverging.

The selection of the coordinate system in which the curvature tensor is defined is
crucial. It is convenient to describe the coordinate system so that two of the coordinate
axes lie in the plane containing e1 and e2, since this reduces the number of nonzero
terms in K. Further reduction in the number of nonzero terms can be achieved if K
can be defined in a coordinate system in which e1 and e2 are basis vectors. This is
possible in 2D because one of the principal curvature directions is always parallel to the
out-of-plane direction, but in 3D it is not feasible because the principal curvature
directions can change as rays propagate through a graded medium.

In the most general 3D case, the curvature tensor is described using the following
symbols, each of which corresponds to a different orthonormal basis:

• KX: curvature tensor defined in the Cartesian coordinate system with basis vectors
x, y, and z.
• KW: curvature tensor defined in the coordinate system consisting of the two
principal curvature directions e1 and e2 and the direction of propagation t.
• KS: curvature tensor defined in a coordinate system in which one basis vector is the
direction of propagation t. The other two basis vectors s1 and s2 can be determined
arbitrarily, as long as they are orthogonal to each other and to t, such that
s 1 × s 2 = t . It is more convenient to operate in this coordinate system if the basis
vectors can be expressed strictly in terms of x, y, z, and t.

The basis vectors s1 and s2 are defined as follows:

726 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 t×z
s 1 = ---------------
t×z

s2 = t × s1

To avoid poles in the definition of the basis vectors, the following alternative
definitions are used when the rays propagate nearly parallel to the z-axis:

t×x
s 1 = ---------------
t×x

s2 = t × s1

The relationship between s1, s2, e1, and e2 is given by

e 1 = s 1 cos φ – s 2 sin φ

e 2 = s 1 sin φ + s 2 cos φ

where the rotation angle φ is an auxiliary dependent variable that is stored for each ray.

The relationship between KS and KW is given by

T
KS = Q0 KW Q0

where Q0 is the rotation matrix:

s1 ⋅ e1 s1 ⋅ e2 s1 ⋅ t cos φ sin φ 0
Q 0 = s 2 ⋅ e 1 s 2 ⋅ e 2 s 2 ⋅ t = – sin φ cos φ 0
t ⋅ e1 t ⋅ e2 t ⋅ t 0 0 0

The relationship between KX and KS is given by

T
K X = QK S Q

where Q is the rotation matrix:

x ⋅ s1 x ⋅ s2 x ⋅ t
Q = y ⋅ s1 y ⋅ s2 y ⋅ t
z ⋅ s1 z ⋅ s2 z ⋅ t

THEORY FOR THE RAY ACOUSTICS INTERFACE | 727

 In 2D, one of the principal curvature directions is always known, so it is possible to
define s1 and s2 so that they coincide with e1 and e2. The rotation matrix Q0 is the
identity, and the auxiliary dependent variable for the angle φ can be omitted.

DERIVATIVES OF THE CURVATURE TENSOR

Following Ref. 3, the derivative of the curvature tensor along the ray trajectory in a
graded medium is given by the expression

5
dK
-----------S- =
ds  Mi
i=1

where the terms Mi correspond to the different ways in which the curvature can
depend on the medium properties. They are defined using the following expressions:

M1 = KS ⋅ KS

M 2 = – c  t ⋅ ∇ S  ---  K S
1
  c 

M 3 = – cΠ  ∇ S ⊗ ∇ S  --- 
1
  c 

M 4 = 2c 2 Π  ∇ S  --- ⊗ ∇ S  ---  Π
1 1
c c
M 5 = – c  K∇ S  ---  ⊗ t + t ⊗ ∇ S  --- K
1 1
  c   c

where Π is the projection matrix:

Π = I – (t ⊗ t)

The gradient operator ∇S consists of derivatives that are taken with respect to the local
coordinates with basis vectors s1, s2, and t.

DERIVATIVES OF THE PRINCIPAL CURVATURES

By application of the chain rule, the derivatives of the nonzero elements of KS can be
expressed in terms of the principal curvatures that occupy the diagonal elements of KW
and the rotation angle φ :

∂κ 1 ∂K S, 11 ∂K S, 12 ∂K S, 22
--------- = -----------------
- cos 2 φ – 2 -----------------
- sin φ cos φ + -----------------
- sin 2 φ
∂s ∂s ∂s ∂s

728 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 ∂κ ∂K S, 11 ∂K S, 12 ∂K S, 22
--------2- = -----------------
- sin 2 φ + 2 -----------------
- sin φ cos φ + -----------------
- cos 2 φ
∂s ∂s ∂s ∂s

∂φ ∂K S, 12 ∂K S, 11 ∂K S, 22
---  -----------------
1 - cos ( 2φ ) + 1
------ = ------------------ ----------------- - – ------------------ sin ( 2φ )
∂s κ2 – κ1 ∂s 2  ∂s ∂s 

NUMERICAL STABILIZATION
The principal curvatures κ1 and κ2 are not ideal choices for the auxiliary dependent
variables stored by each ray because their values can become arbitrarily large as rays
approach caustics. Similarly, the principal radii of curvature r1 and r2 are not ideal
choices because their values can become arbitrarily large when a diverging wavefront
begins to converge while propagating through a graded medium.

Instead, the auxiliary dependent variables stored by each ray are the help variables α1
and α2. In 2D, only one help variable is allocated because the out-of-plane principal
radius of curvature is assumed to be infinite. The principal curvatures are expressed in
terms of the help variables using the expression

κ i = κ 0 tan α

where κ0 = 1 1/m and i ∈ [1, 2]. The derivatives are then related by the expression

dα 1 dκ i 2
---------i = -----
- --------- cos α i
ds κ 0 ds

Similarly, the intensity may become infinitely large at caustics, and its reciprocal
becomes infinitely large as rays undergo attenuation, so a help variable Γ is used to
represent the intensity of the ray:

I = I 0 tan Γ

where I0 = 1 W/m2. The relationship between the ray intensity help variable and the
principal curvatures is

dΓ
------- = 1
--- sin ( 2Γ ) [ κ 1 + κ 2 – 2α ]
ds 2

where α is the attenuation coefficient of the medium.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 729

 Attenuation Within Domains
Rays can gradually lose energy as they propagate through absorbing media. For a plane
wave propagating through an absorbing medium with attenuation coefficient α
(SI unit: 1/m), the intensity decreases exponentially:

dI-
----- = – 2αI (8-3)
ds

The Ray Acoustics interface defines an additional dependent variable A

(dimensionless) for the path integral of the attenuation coefficient,

dI 0
--------- = 0 (8-4)
ds
I = I 0 exp ( A )
dA
-------- = – 2α
ds

Despite requiring an extra degree of freedom per ray, the advantage of Equation 8-4
over Equation 8-3 is that the dependent variables in Equation 8-4 all vary linearly
within a homogeneous medium. This makes Equation 8-4 much less prone to
numerical stiffness; that is, it remains more accurate when the solver takes long time
steps, compared to Equation 8-3, potentially reducing solution time and improving
accuracy.

The attenuation coefficient is controlled by the option selected from the Fluid model
list in the Pressure Acoustics Model section of the Medium Properties.

If Linear elastic is selected, no attenuation occurs.

If Linear elastic with attenuation is selected, the attenuation factor is user-defined. Note
that the attenuation coefficient α is a pressure amplitude attenuation coefficient in
nepers per meter (Np/m). Alternatively, you can specify the amplitude attenuation
coefficient in decibels per meter α’ or decibels per wavelength α(λ). Another option is
to specify the intensity attenuation coefficient m directly. The relationships between
these definitions of the attenuation coefficient may be summarized as follows:

m α' α(λ)
α = ----- = ---------------------------- = --------------------------------
2 20log 10 ( e ) 20log 10 ( e )λ

If Thermally conducting and viscous is selected, the attenuation factor is defined in terms
of the medium properties:

730 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 2
ω b
α = ------------3
2ρc

4 ( γ – 1 )k
b ≡ --- μ + μ B + --------------------
3 Cp

where

• c (SI unit: m/s) is the speed of sound in the medium,

• ρ (SI unit: kg/m3) is the density.
• Cp (SI unit: J/(kg·K)) is the heat capacity at constant pressure.
• γ (dimensionless) is the ratio of specific heats.
• k (SI unit: W/(m·K)) is the thermal conductivity.
• μ (SI unit: Pa·s) is the dynamic viscosity.
• μB (SI unit: Pa·s) is the bulk viscosity.

Built-in models for attenuation in the ocean and atmosphere are also
available. The theory for these features can be found in the Attenuation
and Loss Models section in the Acoustic Properties of Fluids chapter.

Ray Termination Theory

The Ray Termination feature can annihilate rays when their intensity or power
decreases below a specified threshold.

If the ray intensity is computed, then the threshold ray intensity Ith (SI unit: W/m2)
can be specified. Then the termination criterion is

I < I th

If the ray power is computed, then the threshold ray power Qth (SI unit: W) can be
specified. Then the termination criterion is

Q < Q th

It is also possible to specify termination criteria based on both intensity and power; in
this case, the rays terminate if either criterion is satisfied.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 731

 Whenever possible, the Ray Termination feature attempts to stop rays at the exact
instant when the ray intensity or power reaches a specified threshold value. A
closed-form analytic solution for the exact stop time is available under the following
circumstances:

• The rays represent planar wavefronts in an absorbing medium, or

• The rays represent spherical or ellipsoidal wavefronts in a nonabsorbing medium.

If neither of these special conditions is met, the ray can only terminate if the
termination criterion is met at the beginning of a time step taken by the solver; the ray
intensity or power at the instant the ray is stopped may be less than the threshold.

PLANAR WAVEFRONTS IN AN ABSORBING MEDIUM

In this special case, the principal wavefront radii of curvature are so large that the ray
intensity can only change significantly in an attenuating medium. The exact time ts
(SI unit: s) at which the ray intensity reaches the threshold value is

t s = t 0 + τ log  -------
I
I th

where

• t0 (SI unit: s) is the previous time step, reflection time, or refraction time,
• I (SI unit: W/m2) is the intensity at t0, and
• Ith (SI unit: W/m2) is the threshold intensity.

If the termination criterion is based on the ray power Q (SI unit: W) instead of ray
intensity, the analogous expression for the stop time is

t s = t 0 + τ log  ---------
Q
 Q th

The characteristic time for ray attenuation τ (SI unit: s) is

1
τ = --------------
2αV g

where

• α (SI unit: 1/m) is the attenuation coefficient of the medium and

• Vg (SI unit: m/s) is the group velocity magnitude.

732 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

DIVERGING WAVEFRONTS IN A NONABSORBING MEDIUM
In a nonabsorbing medium in which the wavefront is nonplanar, the ray can only
terminate due to the intensity decreasing below the threshold, since the power does
not change. The time ts (SI unit: s) at which the ray intensity reaches the threshold in
a nonattenuating, homogeneous medium is

t s = t 0 + ----------  r 1 + r 2 + ( r 1 – r 2 ) 2 + 4 ------- r 1 r 2 
1 I
2V g  I th 

where

• t0 (SI unit: s) is the previous time step, reflection time, or refraction time,
• Vg (SI unit: m/s) is the group velocity magnitude,
• r1 (SI unit: m) is the first principal radius of curvature of the wavefront at t0,
• r2 (SI unit: m) is the second principal radius of curvature of the wavefront at t0,
• I (SI unit: W/m2) is the intensity at t0, and
• Ith (SI unit: W/m2) is the threshold intensity.

In 2D models, or in 3D models where the wavefronts are cylindrical (one extremely

large radius of curvature and one finite radius of curvature), the stop time is instead

t s = t 0 + ------  r 1 + ------- r 1 
1 I
Vg I th

where r1 (SI unit: m) is the finite principal radius of curvature of the cylindrical
wavefront.

Accumulator Theory: Domains

The Accumulator (Domain) node is used to transfer information from rays to the
domains they occupy or pass through. Each Accumulator defines a variable, called the
accumulated variable, in the selected domains. The accumulated variable is discretized
using constant shape functions, so its value is uniform over every mesh element and
may be discontinuous between adjacent mesh elements.

The name of the accumulated variable is specified in the Accumulated variable name
field in the Accumulator Settings section of the Settings window. The default variable
name, rpd, will be used in the remainder of this section when referring to the
accumulated variable.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 733

 ACCUMULATOR TYPE
The options in the Accumulator type list are Density and Count. If Density is selected, the
source term is divided by the area or volume of the mesh element when calculating
each ray’s contribution to the accumulated variable. If Count is selected, no division by
the area or volume of the mesh element occurs.

The equations in the following section are valid for the Density type. The
corresponding value of the accumulated variable for the Count type is

rpd count = rpd density × V

where V is the mesh element volume (in 3D) or area (in 2D).

ACCUMULATION OVER ELEMENTS

When Elements is selected from the Accumulate over list, the value of the accumulated
variable in a mesh element is the sum of the source terms Ri evaluated for all rays in
that mesh element:

N
1
rpd = ----
V  Ri
i=1

where N is the total number of rays in the element and V is the area or volume of the
mesh element. In other words, the contribution of each ray to the accumulated
variable is distributed uniformly over the mesh element the ray is in, regardless of the
ray’s exact position within the element.

If Elements and time is selected from the Accumulate over list, then the sum of the
source terms for rays in the mesh element is used to define the time derivative of the
accumulated variable, rather than its instantaneous value:

N
d ( rpd ) 1
------------------
dt
= ----
V  Ri
i=1

Thus the value of the accumulated variable depends on the time history of rays in the
mesh element, instead of the instantaneous positions of the rays. As each ray
propagates, it will leave behind a trail based on its contributions to the accumulated
variables in mesh elements it has traversed. The algorithm for accumulating over time
takes into account the fraction of a time step taken by the solver that the ray spends in
each mesh element, even if it crosses between elements during the time step.

734 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Accumulator Theory: Boundaries
The Accumulator (Boundary) feature transfers information from rays to the
boundaries they hit or pass through. Each Accumulator defines a variable, called the
accumulated variable, on the selected boundaries. The accumulated variable is
discretized using constant shape functions, so its value if uniform over every mesh
element and may be discontinuous between adjacent mesh elements.

The name of the accumulated variable is specified in the Accumulated variable name
field in the Accumulator Settings section of the settings window. The default variable
name, rpb, will be used in the remainder of this section when referring to the
accumulated variable.

The options in the Accumulator type list are Density and Count. If Density is selected, the
source term is divided by the surface area or length of the boundary mesh element
when calculating each ray’s contribution to the accumulated variable. If Count is
selected, no division by the surface area or length of the boundary element occurs.

The equations in the following section are valid for the Density type. The
corresponding value of the accumulated variable for the Count type is

rpb count = rpb density × V

where V is the boundary element surface area (in 3D) or length (in 2D).

When Rays in boundary elements is selected from the Accumulate over list, the
accumulated variable in a boundary element gets incremented by the source term R
whenever a ray freezes or sticks to the boundary:

R
rpb new = rpb + ----
V

where division by the mesh element area or length occurs because the accumulator is
assumed to be of type Density. Thus the source term evaluated for an incident ray is
uniformly distributed over the boundary element it freezes or sticks to.

If instead Ray-wall interactions is selected from the Accumulate over list, then the
accumulated variable gets incremented regardless of what type of ray-wall interaction
occurs. Thus, it is possible for the same ray to increment the accumulated variable in
many different boundary elements, or even in the same element multiple times.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 735

 BUILT-IN GLOBAL VARIABLES
By default, the boundary Accumulator defines the following global variables:
TABLE 8-4: BUILT-IN GLOBAL VARIABLES FOR BOUNDARY ACCUMULATORS

NAME EXPRESSION

<scope>.<name>_ave Average of accumulated variable

<scope>.<name>_int Integral of accumulated variable
<scope>.<name>_max Maximum of accumulated variable
<scope>.<name>_min Minimum of accumulated variable
<scope>.<name>_sum Sum of accumulated variable over elements

Here, <scope> includes the physics interface name and the names the Accumulator
and parent feature. For example, the average of the accumulated variable over a
boundary may be called rac.wall1.bacc1.rpb_ave, where rac is the name of the
Ray Acoustics interface, wall1 is the name of the parent Wall node, bacc1 is the name
of the Accumulator node, and rpb is the accumulated variable name. These variables
are all available in the Add/Replace Expression menus during results evaluation.

These global variables are computed by defining a set of nonlocal couplings on the
selection of the parent physics feature, such as the Wall feature to which the
Accumulators are added. The following expressions for the global variables are used.
TABLE 8-5: BUILT-IN GLOBAL VARIABLE DEFINITIONS FOR BOUNDARY ACCUMULATORS

NAME EXPRESSION

<scope>.<name>_ave <wscope>.aveop(<scope>.<name>)
<scope>.<name>_int <wscope>.intop(<scope>.<name>)
<scope>.<name>_max <wscope>.maxop(<scope>.<name>)
<scope>.<name>_min <wscope>.minop(<scope>.<name>)
<scope>.<name>_sum <wscope>.intop(<scope>.<name>/<scope>.meshVol)

Here, <wscope> is the scope of the parent boundary feature, for example
goprac.wall1.

Sound Pressure Level Calculation Theory

It is possible to compute the sound pressure level due to incident and reflected rays at
a surface. To do so, the ray power must be solved for. Then the Sound Pressure Level
Calculation node can be added as a subnode to the Wall node.

The sound pressure level is computed by first computing the Wall intensity Iw
(SI unit: W/m2). The wall intensity on a given boundary element is

736 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 1
I w, i = ------
Ai  Qj ( R 2 + 2 R cos ( arg ( R ) ) + 1 )
j

where the subscript i denotes the ith boundary element, the subscript j indicates the
jth ray, Qj (SI unit: W) is the power transferred by the jth ray, and R (dimensionless)
is the reflection coefficient. The sum is taken over all rays that hit the ith boundary
element.

The sound pressure level Lp,i in the ith element is

 ρcI w, i 
L p, i = 10 log  ---------------------
-
 p 2ref, SPL

where ρ is the density of the fluid, c is the speed of sound, and pref,SPL is the reference
pressure corresponding to 0 dB. Note that the base-10 logarithm is used.

Assigning Spatial Directivity to a Source

The Source with Directivity node can be used to release rays in a distribution with user
defined spatial directivity. This node sets up a point source where rays emanate
outward in a spherical or hemispherical distribution.

When the ray intensity or power is solved for, you may either specify the spatial
directivity in terms of sound pressure level or enter the total source power. When
specifying a spatial directivity function, the intensity of each released ray at a distance
r (SI unit: m) from the release point is

2
p RMS R ref 2
I = ---------------  -----------
ρc r

where pRMS is the root mean square pressure at a distance r = Rref from the source,

p RMS
D ( ϕ, θ ) + L ref = 20log 10  ----------------------
p ref, SPL

• ρ (SI unit: kg/m3) is the density of the surrounding fluid,

• c (SI unit: m/s) is the speed of sound in the medium,
• Rref (SI unit: m) is the reference distance,
• D(ϕ,θ) is the spatial directivity in dB, and
• Lref is the reference level in dB.

THEORY FOR THE RAY ACOUSTICS INTERFACE | 737

 Strictly speaking, the ray acoustics approximation gives infinite ray intensity for a
spherical or hemispherical wavefront as it is first released from an infinitesimally small
point source. In reality, diffraction effects would make the ray acoustics approximation
inaccurate within a few wavelengths of the release position.

If the ray power is solved for, then the total power of all released rays is

2
p RMS
P src = Ω --------------- R ref
2
ρc

where Ω (SI unit: rad) is the total solid angle subtended by the released rays: 2π for a
hemispherical release or 4π for a spherical release.

Alternatively you can specify the total source power Psrc directly, along with a power
weighting factor Pwt. Then the intensity and power of each released ray is proportional
to the value of Pwt evaluated on that ray, while the sum of the initial power over all rays
is Psrc.

References for the Ray Acoustics Interface

1. L.D. Landau and E.M. Lifshitz, Fluid Mechanics, Course of Theoretical Physics,
Volume 6, Butterworth-Heinemann, 2003.

2. O.N. Stavroudis, The Optics of Rays, Wavefronts, and Caustics, Academic, 1972.

3. R.M. More and K. Kosaka, “Wave-front curvature in geometrical optics”, Phys.

Rev. E, vol. 57, no. 5, pp. 6127–6134, 1998.

4. A. D. Pierce, Acoustics: An Introduction to Its Physical Principles and

Applications, Acoustical Society of America, 1991.

5. F. B. Jensen, W. A. Kuperman, M. B. Porter, and H. Schmidt, Computational

Ocean Acoustics, Second Edition, Springer, 2011.

6. M. Vorländer, Auralization: Fundamentals of Acoustics, Modelling, Simulation,

Algorithms and Acoustic Virtual Reality, Springer, 2008.

738 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

The Acoustic Diffusion Equation
Interface
The Acoustic Diffusion Equation (ade) interface ( ), found under the
Acoustics>Geometrical Acoustics subbranch ( ), solves a diffusion equation for the
acoustic energy density. This type of analysis also known as energy finite elements or
EFEM. It is applicable for high-frequency acoustics inside coupled rooms when the
acoustic fields are diffuse. The interface only exists in 3D. The diffusion of the acoustic
energy density depends on the mean free acoustic path and thus on the individual
room geometry. Absorption may be applied at walls and a transmission loss may be
applied when coupling rooms together. Increased diffusion due to room fitting can be
added. Material properties and sources may be specified in frequency bands.

The interface is well suited for quick assessment of sound pressure level distribution
inside buildings and other large structures. Compared to a ray acoustics simulation this
interface does not include any phase information, direct sound, and early reflections.
The interface supports stationary studies for modeling a steady state sound energy or
sound pressure level distribution. A time dependent study can be used to determine
energy decay curves and reverberation times. The reverberation time of coupled and
uncoupled rooms can also be determined using the eigenvalue study. The eigenvalue
is directly related to the exponential decay time.

When this physics interface is added, these default nodes are also added to the Model
Builder — Acoustic Diffusion Model, Room, and Initial Values. Then, from the Physics
toolbar, add other nodes that implement, for example, boundary conditions. You can
also right-click Acoustic Diffusion Equation to select physics features from the context
menu.

For tutorial models that use the Acoustic Diffusion Equation, see the
Application Gallery models:

• One-Family House Acoustics: [Link]/model/20169

• Acoustics of Coupled Rooms Using the Acoustic Diffusion Equation:
[Link]/model/20171
• Acoustic Mean Free Path in a Room:
[Link]/model/72491

THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 739

 SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is ade.

SOUND PRESSURE LEVEL SETTINGS

See Sound Pressure Level Settings for the Pressure Acoustics, Frequency Domain
interface. Note that only Use reference pressure for air or User-defined reference pressure
are available selections.

BAND STUDIED
This section cannot be edited when the Band Input Type is Flat (continuous). For all
other options under Band Input Type, the default Nstudy (dimensionless) is 1, meaning
that material properties etc. that are defined in the first band are used. To sweep the
bands enter a parameter (defined under Global Definitions > Parameters) and solve the
model using a parametric sweep in the study. In this way solving the model for the
desired number of bands.

BAND INPUT TYPE

Select a Type — Flat (continuous) (the default), Octave bands, 1/3 octave bands, or User
defined bands. The frequency content of a model solved using the acoustic diffusion
equation all lies implicitly in material parameters, absorption properties, sources and so
forth. These can be given in a broadband (or for a single frequency) by selecting Flat
(continuous) or in bands when given in this way.

• For Octave bands the table that displays defaults to 11 rows under Band number with
the associated Lower band limit (Hz), Center frequency (Hz), and Upper band limit (Hz)
values.

740 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

• For 1/3 octave bands the table that displays defaults to 32 rows under Band number
with the associated Lower band limit (Hz), Center frequency (Hz), and Upper band limit
(Hz) values.
• For User defined bands the table that displays defaults to 1 row under Band number
with default associated values in the Lower band limit (Hz), Center frequency (Hz), and
Upper band limit (Hz) columns.
- Under the table click the Add ( ) or Delete ( ) buttons to edit the table
contents. Or right-click a table cell and select Add or Delete.
- To save the contents of a table, click the Save to File button ( ) and enter a File
name in the Save to File dialog box, including the extension .txt. Click to Save
the text file. The information is saved in space-separated columns in the same
order as displayed on screen.
- Use the Load from File button ( ) and Load from File dialog box to import data
in text files, generated by, for example, a spreadsheet program. Data must be
separated by spaces or tabs (or be in a Microsoft Excel Workbook spreadsheet if
the license includes LiveLink™ for Excel®).

The lower band limit, center frequency, and upper band limit in the
studied band are defined as the global variables [Link], [Link], and
[Link], respectively. These can, for example, be used when plotting data
in postprocessing.

Note that the tables that are imported, using the Load from File
button ( ), will replace the existing data. You can only import a table
that has the expected size. The band number column is disregarded when
importing. To create a data file to load, it is recommended to save the
existing table (octaves or 1/ octaves) and then edit and change the values
in that file.

DEPENDENT VARIABLES
This physics interface defines one dependent variable (field), the Sound energy
density w. If required, edit the name, which changes both the field name and the
dependent variable name. If the new field name coincides with the name of another
pressure field in the model, the interfaces share degrees of freedom and dependent
variable name. The new field name must not coincide with the name of a field of
another type, or with a component name belonging to some other field.

THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 741

 DISCRETIZATION
From the list select the element order for the Sound energy density, the default is
Quadratic.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

• Domain, Boundary, and Global Nodes for the Acoustic Diffusion

Equation Interface
• Theory for the Acoustic Diffusion Equation Interface

Domain, Boundary, and Global Nodes for the Acoustic Diffusion

Equation Interface
The Acoustic Diffusion Equation Interface has these domain, boundary, and point
nodes available (listed in alphabetical order).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

742 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 • Acoustic Diffusion Model • Inward Energy Flux
• Continuity1 • Mapped Room Coupling
• Destination Selection • Point Source
• Domain Source • Room
• Fitted Domain • Room Coupling
• Initial Values • Wall
1
Described for the Pressure Acoustics, Frequency Domain interface

Acoustic Diffusion Model

The Acoustic Diffusion Model is added by default and is used to define the material
properties and the governing diffusion equations.

MATERIAL PROPERTIES
For each of the following properties the default takes values From material, or select
User defined to enter a different value or expression in the text field.

• Density ρ (SI unit: kg/m3). The default is 1.2 kg/m3.

• Speed of sound c (SI unit: m/s). The default is 343 m/s.
• Volume absorption coefficient ma (SI unit: 1/m). The default is 0 1/m. This
parameter determines the amount of absorption due to losses in the air. The effects
are usually of interest in very large rooms/volumes or at very high frequencies. The
built-in option Classical viscous and thermal absorption defines the frequency
absorption coefficient due to viscous and thermal losses. For the Flat (continuous)
band option it is evaluated at the studied frequency feval (the default is 1000 Hz)
while it is averaged over the studied band for the other Band Input Type.

Room
The Room node is added by default and is used to define the mean free path model of
each room on a model. A Wall subnode is also added by default. Additional Wall
subnodes and the Inward Energy Flux subnode are available from the context menu
(right-click the parent node) or from the Physics toolbar.

Each room in a model needs a Room node with associated definitions of all wall
properties. The volume and surface area of the room is calculated and used to
determine the mean free path (depending on the model selected).

THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 743

 MEAN FREE PATH MODEL
Select a Mean free path model — Regular cubic room (the default), Long room, or User
defined. For User defined enter a Mean free path λ (SI unit: m). In most cases the regular
cubic room is applicable and valid for an ideal diffuse field. For the user defined case
the mean free path can be computed using a ray acoustics model.

For a tutorial model that computes the mean free path, see the
Application Gallery model: Acoustic Mean Free Path in a Room
[Link]/model/acoustic-mean-free-path-in-a-room-724915

Classical statistical room acoustic variables that are based on the room
geometry and absorption properties of the walls are defined and can be
used for postprocessing. See Statistical Model of Reverberation Time for
details. They are:

• The total room absorption ade.A

• The Sabine reverberation time ade.T60_Sabine
• The Eyring-Norris reverberation time ade.T60_EN
• The Schroeder frequency of the room ade.f_Schroeder

Each room feature (say ade.room1, ade.room2, etc.) also has several
room variables associated that may be plotted in postprocessing. They
are:

• The room volume ade.room1.V

• The room surface area ade.room1.S
• The mean free path of the room [Link]

Wall
The Wall subnode is available from the context menu (right-click the Room parent
node) or from the Physics toolbar, Attributes menu. It is used to define the wall
absorption properties.

WALL ABSORPTION PROPERTIES

The absorption coefficient properties to define depend on the choice of Band Input
Type for The Acoustic Diffusion Equation Interface. The absorption is defined

744 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

through the random incidence absorption coefficient α (the value is different from the
normal incidence absorption typically used in, for example, pressure acoustics).

• If the Band Input Type is Flat (continuous) enter an Absorption coefficient α

(dimensionless).
• For Octave bands, 1/3 octave bands, or User defined bands, enter an Absorption
coefficient α (dimensionless) in each row corresponding to the Band Number.

Select an Exchange coefficient model — Sabine type (the default), Eyring type, Modified
Eyring type, or User defined. For User defined enter an Exchange coefficient h (SI unit:
m/s). The choice gives different statistical models which are applicable for different
ranges of absorption coefficients. The Sabine type is applicable for α < 0.2 while the
two others are valid up to an absorption of 1.

Note that the tables that are imported, using the Load from File
button ( ), will replace the existing data. You can only import a table
that has the expected size. The band number column is disregarded when
importing. To create a data file to load, it is recommended to save the
existing table and then edit and change the values in that file.

Inward Energy Flux

The Inward Energy Flux subnode is available from the context menu (right-click the
Room parent node) or from the Physics toolbar, Attributes menu. It is used to define
the energy flux that stems from wall sources. This could be a vibrating wall of a
structure or the energy flux as measured from external sources to a building.

ENERGY FLUX
The inward energy flux properties to define depend on the choice of Band Input Type
for The Acoustic Diffusion Equation Interface.

• If the Band Input Type is Flat (continuous) enter an Inward energy flux Jin (SI unit:
W/m2).
• For Octave bands, 1/3 octave bands, or User defined bands, enter an Inward energy flux
Jin (SI unit: W/m2) in each row corresponding to the Band Number.

THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 745

 Initial Values
The Initial Values node adds initial values for the sound energy density. Add more Initial
Values nodes from the Physics toolbar. When calculating the energy impulse response
of a system using a time dependent study set the initial value inside a small volume
(small domain) or define it in terms of a Gaussian shape.

INITIAL VALUES
Enter a value or expression for the Sound energy density w (SI unit: J/m3) initial value.
The default is 0 J/m3.

Fitted Domain
Use the Fitted Domain node to define the fitting parameters. These are parameters that
define a region with, for example, furniture inside a room. The domain is characterized
in terms if its average statistical properties.

FITTED DOMAIN PARAMETERS

Enter values or expressions for the following:

• Number density of fitting nf (SI unit: 1/m3).

• Average cross-section of fitting Qf (SI unit: m2).

How you define the Absorption coefficient of fitting is based on the choice of Band
Input Type for The Acoustic Diffusion Equation Interface.

• If the Band Input Type is Flat (continuous) enter a value in the text field for the
Absorption coefficient of fitting α f (dimensionless). For Octave bands, 1/3 octave
bands, or User defined bands, enter an Absorption coefficient of fitting α f
(dimensionless) in each row corresponding to the Band Number.

Domain Source
Use the Domain Source node to define a source, the right-hand side of the governing
diffusion equation. The source can be a constant value defined inside a small volume,
it can be a function of time, or it can a function of space. The value corresponds to the
total emitted power Q of the source it is automatically transformed into a power
density q by dividing with the source domain volume (calculated automatically).

746 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

DOMAIN SOURCE
How you define the domain source depends on the choice of Band Input Type for The
Acoustic Diffusion Equation Interface.

• If the Band Input Type is Flat (continuous) enter a Domain source Q(SI unit: W).
• For Octave bands, 1/3 octave bands, or User defined bands, enter a Domain source Q(SI
unit: W) in each row corresponding to the Band Number.

Room Coupling
Use the Room Coupling node to couple rooms together allowing an acoustic energy flux
back and forth. This can be used to model transmission through walls or openings
defined in terms of their transmission loss. The feature is also used to couple rooms
that are connected through a large open surface. It is also available as the Pair Room
Coupling.

TRANSMISSION LOSS
How you define the transmission loss depends on the choice of Band Input Type for
The Acoustic Diffusion Equation Interface.

• If the Band Input Type is Flat (continuous) enter a Transmission loss TL (SI unit: dB).
• For Octave bands, 1/3 octave bands, or User defined bands, enter a Transmission loss
TL (SI unit: dB) in each row corresponding to the Band Number.

CONTINUITY
Select the Force continuity check box to constrain the dependent variable resulting in a
zero transmission loss. Use this option when rooms are coupled through large open
surfaces. Note that the absorption properties of these surfaces need to be set correctly
independent of this option.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Mapped Room Coupling

Use the Mapped Room Coupling node to couple two rooms when they do not share a
common inner boundary. This is, for example, necessary if mid-planes have not been
added in the CAD drawing. The condition has the same functionality as the Room

THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 747

 Coupling feature except that a mapping is used between the source and destination
walls. The Destination Selection subnode is available from the context menu
(right-click the parent node) or from the Physics toolbar, Attributes menu.

TRANSMISSION LOSS
How you define the transmission loss depends on the choice of Band Input Type for
The Acoustic Diffusion Equation Interface.

• If the Band Input Type is Flat (continuous) enter a Transmission loss TL (SI unit: dB).
• For Octave bands, 1/3 octave bands, or User defined bands, enter a Transmission loss
TL (SI unit: dB) in each row corresponding to the Band Number.

CONTINUITY
Select the Force continuity check box to ensure continuity resulting in a zero
transmission loss.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Destination Selection
The Destination Selection subnode is available from the context menu (right-click the
Mapped Room Coupling parent node) or from the Physics toolbar, Attributes menu.

Point Source
Use the Point Source node to add a point source to model a source of small dimensions.
For a time dependent model it is possible to enter an expression that depend on time.

POINT SOURCE
Enter a Point source power qp (SI unit: W). The default is 0 W.

748 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

Modeling with the Acoustic Diffusion
E qua t i o n Inte r face
In this section:

• The Eigenvalue Study Type

• Combined Stationary and Time Dependent Study

The Eigenvalue Study Type

The Acoustic Diffusion Equation interface and eigenvalue study will produce the
modes that decay exponentially for each room including the effect of couplings to
adjacent rooms and wall absorption. The modes are of the form

–λi t
w ( x, t ) = w i ( x )e

where each mode wi has an associated eigenvalue λi (this is the exponential time scale).
The eigenvalue directly gives the reverberation time T60 of that subsystem

60 55.3
T 60 = --------------------------------- ≈ -----------
10log 10 ( e )λ i 4λ i

Typically, there will be as many modes as there are rooms in the building.

A list relating the eigenvalues to the reverberation times is automatically shown under
Derived Values in the Results node. Identify the mode associated with a given
reverberation time by looking at the default plot Sound Energy Density Mode.

Combined Stationary and Time Dependent Study

By combining a Stationary and a Time Dependent study you can readily model the
so-called steady state decay behavior of the system. Use the stationary solution as an
initial condition for the transient model. Remember to disable all sources in the
transient study step by enabling the Modify physics tree and variables for study step
option. Evaluating the sound pressure level, [Link], in different points will result in
the energy decay curves.

MODELING WITH THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 749

 Theory for the Acoustic Diffusion
Equation Interface
This section contains a small review of the classical models for the prediction of
reverberation time in room acoustics followed by the theory background of the
acoustic diffusion equation.

In this section:

• Statistical Model of Reverberation Time

• The Acoustic Diffusion Equation
• References for the Acoustic Diffusion Equation Interface

Statistical Model of Reverberation Time

In classical statistical acoustics (see Ref. 9 and Ref. 16) the reverberation time T60 (RT)
is the time taken for the sound pressure level (created by an impulse source) to decay
60 dB. The reverberation time can be calculated using the Sabine formula

55.3V
T 60 = ---------------
cA

where V is the room volume, c is the speed of sound, and A is the total absorption of
all surfaces. The factor 55.3 is a roundoff of 24·ln(10). The total absorption is
calculated by

A =  Si αi = Sα
where Si and αi denote the area and absorption of the i’th area. S is the total surface
area and α is the average absorption of the room. For large rooms this equation is
corrected by the absorption of air

55.3V
T 60 = -----------------------------
cA + 4m a V

where ma is the air absorption constant (see Table 3.1 in Ref. 9). Other versions of the
T60 reverberation time exist one is the Eyring–Norris given by

750 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 55.3V
T 60 = -----------------------------------
– cS ln ( 1 – α )

The models for the reverberation time given above are statistical models for diffuse
sound field (large number of reflections). At low frequencies the modal behavior of the
room becomes important. The lower bound for the application of these statistical
models is the Schroeder frequency (see Ref. 10 and Ref. 16) given by

3⁄2
f > 2000 ( m/s ) T 60 ⁄ V

The classical expressions given above are available for postprocessing in the Acoustic
Diffusion Equation interface.

The Acoustic Diffusion Equation

The acoustic diffusion model is based on the assumption that the volumes (rooms)
studied contain scatterers that uniformly scatter the sound field and that the sound
field is diffuse (large number of reflections). Using the diffusion of light in a scattering
environment as an analogy one can express a diffusion equation for the sound-energy
density w = w(x,t) (SI unit: J/m3). The diffusion equation describes the energy flow
from high to low energy regions. Further details about the model equations and
boundary conditions are found in papers by Xiang and others (see Ref. 3, Ref. 7, and
Ref. 8) and in papers by Billon, Valeau, and more (see Ref. 5, Ref. 6, Ref. 11, and
Ref. 15).

DOMAIN EQUATIONS
The domain diffusion equation for the sound-energy density w = w(x,t) is given by

∂w
------- + ∇ ⋅ J + cm a w = q ( x, t )
∂t

where the local energy flux vector J (SI unit: J/m2/s = W/m2) is defined in the usual
way, as

J = – D t ∇w

The total diffusion coefficient is Dt = D = λc/3 (SI unit: m2/s), λ is the mean free path
(SI unit: m), c is the speed of sound (SI unit: m/s), and ma is the volumetric
absorption coefficient of air (SI unit: 1/m). The volumetric absorption coefficient (or
attenuation coefficient) should not be confused with the energy absorption coefficient
α used in boundary conditions which is dimensionless. The source term q represents

THEORY FOR THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 751

 the spatial sound source (SI unit: J/m3/s = W/m3). The term cmaw accounts for
volume absorption in air (dissipation). Note that in certain models the cmaw term
accounts for the total absorption at boundaries in a “mean” sense (only using no flux
boundary conditions), this approach is not used here.

In the interface, ma can be given as a user input or it can be defined as the “classical
viscous and thermal (volumetric) absorption” coefficient (as in pressure acoustics)
given by

2
ω μ 4 γ–1
m a, classical = ------------3  --- + -----------
 Pr 
2ρc 3

The absorption coefficient is either integrated over the band or it is given at a certain
frequency. The volumetric absorption is only important for very large domains/rooms.
It is furthermore assumed that ma λ << 1 such that the diffusion coefficient remains
unchanged.

The mean free path λ is the distance a sound particle on average travels between
reflections. It is related to the average reflections frequency n by λ = c/ n . The mean
free path is a property of each room and for a regular cubic like room it can be
calculated by (the usual convention)

4V
λ = -------
S

where V is the room volume and S the total room surface area (see Ref. 18). The mean
free path is defined for every room as it depends on the room geometry and shape. For
long rooms like corridors Visentin (see Ref. 13 and Ref. 14) proposes to use another
definition of the mean free path given by

S
λ = ------
4π

In general, the mean free path varies for different room shapes. Entering a user defined
mean free path based on for example measurements is also possible (see Ref. 1, Ref. 2,
and Ref. 17).

Fitted rooms/zones with obstacles (chairs, machines etc.):

In zones with fittings, like furniture or other absorbers, the scatterers are modeled
statistically by their number density nf (SI unit: 1/m3), their average cross-section Qf
(SI unit: m2), and their absorption coefficient αf (dimensionless). For the domain with
scatterers the mean free path becomes

752 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

 1
λ f = -----------
nf Qf

defining the fitting diffusion constant

cλ f
D f = -------
3

With these definitions in place, the governing diffusion equation can be modified to
take the fitting into account (see, for example, Ref. 15)

∂w α Df D
------- + ∇ ⋅ ( – D t ∇w ) + c  m a + -----f w = q ( x, t ) D t = -----------------
∂t  λf Df + D

VALIDITY
The time limit after which the acoustic diffusion model leads to correct results has been
discussed by many authors see, for example, Ref. 6, Ref. 12, or Ref. 15. They suggest
that a limit of one mean free time λ/c could be considered for the diffusion equation
to have physical meaning. Before this time, the high probability of the particles of not
having hit a scatterer/surface yet leads to nonvalid results. Typically, this type of
models only apply at frequencies above the Schroeder frequency. This is a good “rule
of thumb” measure — in several publications results have been seen to match
measurements also at lower frequencies. Below the Schroeder frequency the room
eigenmodes are important and can be modeled using Pressure Acoustics.

SOURCES
The source term q (x,t) (SI unit of power density J/m3/s = W/m3) can be defined as
a point source or a volume source.

• Point source located at xs are of the type q (x,t) = P(t)δ(x - xs) where P is the power
of the source (SI unit: W).
• Volume/domain sources of volume Vs emitting the power Q (SI unit: W) yielding
a power density Q/Vs.
• An impulse source in a transient model is typically modeled as an initial value for
w(x,0) in a given small domain (constant in space). The source (initial value) could
also be Gaussian shaped in space, here a larger domain.

THEORY FOR THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 753

 BOUNDARY CONDITIONS

Wall (with different absorption properties):

At walls, a mixed boundary condition accounting for absorption losses is used (the
surface normal n being outward to the volume Ω)

∂w
J ⋅ n = – D t ------- = J out = hw
∂n

where h is the exchange coefficient. Different models exist for this coefficient as given
by Xiang and others (see Ref. 3 and Ref. 4). The Sabine type exchange coefficient is

α
h = cA S = c --- for α < 0.2
4

This expression is valid for surfaces where the absorption coefficient is smaller than 0.2.
The Eyring type model is defined by

– ln ( 1 – α )
h = cA E = c + --------------------------- for α<1
4

This model is not suited for absorption coefficients close to 1. Finally, there is the
(modified) model by Jing and Xiang (see Ref. 4)

α
h = cA M = c --------------------- for α≤1
2(2 – α)

In this model, the boundary condition is theoretically grounded and can model high
absorption for a small portion of surfaces. In addition, the diffusion-equation model
inherently assumes that overall absorption in rooms under test must not be “high”.

Interior BCs (coupling between rooms/volumes):

On interior boundaries between rooms (thin walls, doors, grills, thin panels etc.) there
can be a transmission loss (TL) associated. The transmission loss is related to the
transmission coefficient τ of the boundary as

TL = 10 log  ---
1
τ

This results in a condition where the field w is discontinuous across the boundary (a
slit). Using the usual up/down conventions the condition, including possible
absorption. The TL for an opening is typically given for the boundary in which the
opening is located. That is the combined transmission and absorption is smeared on a
boundary.

754 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

With the surface normal pointing out of the domains, the flux on the down side of the
surface/wall (including absorption) is

τc τc
J ⋅ n down = down  hw + ----- w – up  ----- w
4 4

on the up side of the surface/wall (including absorption) the total flux is

τc τc
J ⋅ n up = up  hw + ----- w – down  ----- w
4 4

The transmission coefficient is due to reciprocity the same from both sides. This results
in the combined condition

τc
J ⋅ n up – J ⋅ n down = [ hw ] up – [ hw ] down – ----- ( w up – w down )
4

The “room coupling” condition adds the fluxes related to the TL. Note also that the
terms of the type τcw/4 should probably be modified for small TL because τ then
becomes larger than 0.2 (at around 7 dB). An option exists to force continuity by
constraining

w up = w down

which basically corresponds to a null transmission loss.

Inflow and Outflow Boundary Conditions:

Some general boundary conditions defining a total inflow or outflow of energy are
added as

∂w
J ⋅ n = – D t ------- = J out
∂n
∂w
J ⋅ n = – D t ------- = – J in
∂n

POSTPROCESSING VARIABLES/EXPRESSIONS
The square pressure (valid for spatially averaged quantities and volumes that are large
compared to the wavelength but small compared to the mean free path):

2 2
p ( x, t ) = ρc w ( x, t )

Solution of the energy equation yields (with an impulse source) the energy room
impulse response (excluding the direct sound, as the sound field is assumed diffuse).

THEORY FOR THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 755

 It is also known as the energy-time function (ETF) or energy-time curve (ETC). The
SPL response is given by

 ρc 2 w ( x, t )
L p ( x, t ) = 10 log  ---------------------------
2
-
 P ref 

this is in general the local SPL for any source. For a steady state model this is the spatial
SPL distribution Lp(x).

The steady-state sound-energy decay d(x,t) can be calculated from an energy impulse
response using so-called Schroeder integration. This corresponds to first finding the
steady state (using a stationary solver) and then use this solution as initial condition
(source turned off) in a time-dependent model. The same response can be calculated
using Schroeder integration of an impulse response model

∞ ∞
1

d ( x, t ) = ------------- w ( x, τ ) dτ
N(x)
N(x) =  w ( x, τ ) dτ
t 0

This integration can be set up in COMSOL Multiphysics postprocessing using the

built-in timeint() operator.

A variable for the a posteriori assessment of whether the diffuse field assumption is met
is given by (named “A posteriori diffuse field evaluation norm” in the postprocessing
list)

J D t ∇w
L diffuse = ------- = -------------------
cw cw

If this variable is small, then the acoustic diffusion assumption is met. Another variable
that can be assessed is the Schroeder frequency as given above.

DIFFERENT STUDY TYPES

Transient
A transient simulation is used to model the energy impulse response (the response to
a source of the type E0λ(t), typically approximated by a step function of a shirt time)
of a room or the steady-state energy decay response (decay from a steady state
solution). In a single room the response will be the same as measured in all points
(except for a small time lag). This will result in identical reverberation times (RT)
estimates. The interest of the transient model is more evident when several rooms are

756 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

coupled. Here several time scales exist because of the interaction between the rooms
and these can be seen in a transient energy impulse response.

Steady State
The steady-state model is used to find the spatial SPL distribution inside a room or
collection of rooms and volumes when subject to a steady noise source. This could be
an engine noise or the noise generated by other machinery.

Eigenvalue
An eigenvalue analysis of the transient equation results in all the “basis functions” of
the problem. That is, modes that do not change shape in space but only decay in time
with the exponential time scale given by the eigenvalue. They are of the type

–λ i t
w ( x, t ) = w i ( x )e

where λi is the i:th eigenvalue and wi the associated mode. The eigenvalues are a direct
measures of the reverberation time (RT) of the different rooms. T60 is for example
simply given by

60 55.3
T 60 = ------------------------------------ = -----------
10 ⋅ log ( e ) ⋅ λ i 4λ i

where the last equation is the classical expression. Evaluating the eigenvalues will gives
an easy measures of the slopes of the energy-time functions (ETFs) and can be
combined with a transient model to give the full picture of the reverberation in the
different volumes (rooms) and their coupled behavior. When performing an
eigenvalue study the variable ade.T60 gives the reverberation time associated with the
given eigenvalue. Inspect the sound energy density modes plot to determine see to
which room the reverberation time is associated.

References for the Acoustic Diffusion Equation Interface

1. J. Picaut, L. Simon, and J.D. Polack, “Sound field in long rooms with diffusely
reflecting boundaries”, Appl. Acoust., vol. 56, pp. 217–240, 1999.

2. J. Pujolle, “Les différentes définitions du libre parcours moyen du son dans une
salle”, Revue d’Acoustique, vol. 36, pp. 44–50, 1976 (in French).

3. N. Xiang, Y. Jing, and A.C. Bockman, “Investigation of acoustically coupled

enclosures using a diffusion-equation model”, J. Acoust. Soc. Am., vol. 126, 2009.

THEORY FOR THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 757

 4. Y. Jing and N. Xiang, “On boundary conditions for the diffusion equation in
room-acoustic prediction”, J. Acoust. Soc. Am. vol. 123, pp. 145–153, 2008.

5. A. Billon and others, “Modeling the sound transmission between rooms coupled
through partition walls by using a diffusion model”, J. Acoust. Soc. Am., vol. 123,
2008.

6. V. Valeau and others, “On the use of a diffusion equation for room acoustic
prediction”, J. Acoust. Soc. Am., vol. 119, 2006.

7. Y. Jing and N. Xiang, “On The Use of a Diffusion Equation Model for Sound
Energy Flow Prediction in Acoustically Coupled Spaces”, COMSOL Conference
Boston 2008.
[Link]/paper/on-the-use-of-a-diffusion-equation-model-for-sound-energ
y-flow-prediction-in-aco-5124

8. Y. Jing and N. Xiang, “Boundary Conditions for the Diffusion Equation Model in
Room-acoustic Prediction”, COMSOL Conference Boston 2007.

9. T. J. Cox and P. D’Antonio, Acoustic, Absorbers and Diffusers, Taylor and Francis,
2009.

10. R. Schroeder, “New method of measuring reverberation time”, J. Acoust. Soc.

Am., vol. 37, pp. 409–412, 1965.

11. C. Foy, V. Valeau, A. Billon, J. Picaut, and M. Hodgson, “An Empirical Diffusion
Model for Acoustic Prediction in Rooms with Mixed Diffuse and Specular
Reflections”, Acta Acustica united with Acustica, vol. 95, pp. 97–105, 2009.

12. J. Escolano, J.M. Navarro, and J.L. Lopes, “On the limitation of a diffusion
equation model for acoustic predictions of rooms with homogeneous dimensions
(L)”, J. Acoust. Soc. Am., vol. 128, p. 1586, 2010.

13. C. Visentin, N. Prodi, V. Valeau, and J. Picaut, “A numerical investigation of the

Fick’s law of diffusion in room acoustics”, J. Acoust. Soc. Am., vol. 132, p. 3180,
2012.

14. C. Visentin, A numerical and experimental validation of the diffusion equation

applied to room acoustics, Ph.D. thesis, 2012.

15. V. Valeau, M. Hodgson, and J. Picaut, “A Diffusion-Based Analogy for the

Prediction of Sound Fields in Fitted Rooms”, Acta Acustica united with Acustica,
vol. 93, pp. 94–105, 2007.

758 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES

16. A. D. Pierce, Acoustics, An Introduction to Its Physical Principles and
Applications, Chapter SIX, Acoustical Society of America, 1991.

17. L. Beranek and N. Nishihara, “Mean-free-paths in concert and chamber music

halls and the correct method for calibrating dodecahedral sound sources”, J. Acoust.
Soc. Am., vol. 135, p. 223, 2014.

18. H. Kuttruff, Room Acoustics, Fifth Edition, Spon Press, 2009.

19. Z. Xiangyang, C. Ke’an, and S. Jincai, “On the accuracy of the ray-tracing
algorithms based on various sound receiver models”, Appl. Acoust. vol. 64, pp. 433–
441, 2003.

20. F. Pind, “Broadband Hybrid Room Acoustic Simulations,” Master Thesis,

Technical University of Denmark, DTU, 2013.

21. J. Soares Augusto, “The Io-sinh function, calculation of Kaiser windows and
design of FIR filters”, 10.13140/RG.2.1.3247.0489, 2015.

22. IOC, SCOR and IAPSO, 2010: The international thermodynamics equation of
seawater - 2010: Calculation and use of thermodynamic properties.
Intergovernmental Oceanographic Commission, Manuals and Guides No. 56,
UNESCO (English), 196 pp.

THEORY FOR THE ACOUSTIC DIFFUSION EQUATION INTERFACE | 759

760 | CHAPTER 8: GEOMETRICAL ACOUSTICS INTERFACES
 9

Pipe Acoustics Interfaces

This chapter describes the physics interfaces found under the Pipe Acoustics branch
( ) and background theory.

• The Pipe Acoustics Frequency Domain and Transient Interfaces

• Theory for the Pipe Acoustics Interfaces

761
 T he P i pe Ac o us t i cs Freq u en c y D omai n
and Transient Interfaces
In this section:

• The Pipe Acoustics, Frequency Domain Interface

• The Pipe Acoustics, Transient Interface
• Edge, Boundary, Point, and Pair Nodes for the Pipe Acoustics Interfaces

The Pipe Acoustics, Frequency Domain Interface

The Pipe Acoustics, Frequency Domain (pafd) interface ( ), found under the
Acoustics>Pipe Acoustics branch ( ) when adding a physics interface, is used to
compute the acoustic pressure and velocity variations when modeling the propagation
of sound waves in flexible pipe systems. The governing equations are formulated in a
general way to include the possibility of a stationary background flow. The physics
interface can for example be used to compute the propagation of sound waves in
HVAC systems, other large piping systems, or simply in an organ pipe.

In the frequency domain all sources and variations are assumed to be harmonic. The
solved equations assume that the propagating waves are plane. The propagation of
higher-order modes that exist above their cutoff frequency, dictated by the pipe cross
section, is not modeled.

The equations governing the propagation of sound in pipes stem from considering
momentum, mass, and energy balances for a control volume of a piece of pipe. The
resulting equations are expressed in the cross-sectional averaged variables and reduce
the equations to a 1D component with scalar dependent variables. The physics
interface is available in 3D on edges and points, and in 2D on boundaries and points.

In the no-background flow case, the 1D pipe domain can be coupled to a 3D pressure
acoustics model using the Acoustic-Pipe Acoustic Connection multiphysics coupling.

When this physics interface is added, these default nodes are also added to the Model
Builder: Fluid Properties, Pipe Properties, Closed, and Initial Values. Then, from the
Physics toolbar, add other nodes that implement, for example, boundary conditions
and point conditions. You can also right-click Pipe Acoustics, Frequency Domain to select
physics features from the context menu.

762 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

SETTINGS
The Label is the default physics interface name.

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is pafd.

SOUND PRESSURE LEVEL SETTINGS

The zero level on the dB scale varies with the type of fluid. That value is a reference
pressure that corresponds to 0 dB. This variable occurs in calculations of the sound
pressure level Lp based on the root mean square (rms) pressure prms, such that

p rms
L p = 20 log  ----------
1
with p rms = --- p p∗
 p ref  2

where pref is the reference pressure and the star (*) represents the complex conjugate.
This is an expression valid for the case of harmonically time-varying acoustic
pressure p.

Select a Reference pressure for the sound pressure level based on the fluid type:

• Use reference pressure for air to use a reference pressure of 20 μPa (20·10−6 Pa).
• Use reference pressure for water to use a reference pressure of 1 μPa (1·10−6 Pa).
• User-defined reference pressure to enter a reference pressure pref, SPL (SI unit: Pa).
The default value is the same as for air, 20 μPa.

For postprocessing, plot the sound pressure level [Link], which

depends on the selected reference pressure. You can also plot the intensity
magnitude pafd.I_mag (this is an RMS value) in the frequency domain
or the instantaneous intensity magnitude patd.Ii_mag.in the time
domain.

DISCRETIZATION
Select the discretization of the element types used in the finite element formulation.
Select the Pressure and Tangential velocity discretization orders. The default is to use a
Quadratic/Linear formulation which also ensures numerical stability of the problem.

THE PIPE ACOUSTICS FREQUENCY DOMAIN AND TRANSIENT INTERFACES | 763

 DEPENDENT VARIABLES
This section is used to define the dependent variables (fields) for Pressure p
(SI unit: Pa) and Tangential velocity u (SI unit: m/s). If required, edit the name, but
dependent variables must be unique within a model.

• Edge, Boundary, Point, and Pair Nodes for the Pipe Acoustics
Interfaces
• Theory for the Pipe Acoustics Interfaces

The Pipe Acoustics, Transient Interface

The Pipe Acoustics, Transient (patd) interface ( ), found under the Acoustics>Pipe
Acoustics branch ( ) when adding a physics interface, is used to compute the acoustic
pressure and velocity variations when modeling the propagation of sound waves in
flexible pipe systems. The governing equations are formulated in a general way to
include the possibility of a stationary background flow. The physics interface can for
example be used to compute the propagation of sound waves in HVAC systems, other
large piping systems, or simply in an organ pipe.

The solved equations assume that the propagating waves are plane. The propagation
of higher-order modes that exist above their cutoff frequency, dictated by the pipe
cross section, is not modeled.

The equations governing the propagation of sound in pipes stem from considering
momentum, mass, and energy balances for a control volume of a piece of pipe. The
resulting equations are expressed in the cross-sectional averaged variables and reduce
the equations to a 1D component with scalar dependent variables. The physics
interface is available in 3D on edges and points, and in 2D on boundaries and points.

In the no-background flow case, the 1D pipe domain can be coupled to a 3D pressure
acoustics model using the Acoustic-Pipe Acoustic Connection multiphysics coupling.

When this physics interface is added, these default nodes are also added to the Model
Builder: Fluid Properties, Pipe Properties, Closed, and Initial Values. Then, from the
Physics toolbar, add other nodes that implement, for example, boundary conditions
and point conditions. You can also right-click Pipe Acoustics, Transient to select physics
features from the context menu.

SETTINGS
The Label is the default physics interface name.

764 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

The Name is used primarily as a scope prefix for variables defined by the physics
interface. Refer to such physics interface variables in expressions using the pattern
<name>.<variable_name>. In order to distinguish between variables belonging to
different physics interfaces, the name string must be unique. Only letters, numbers, and
underscores (_) are permitted in the Name field. The first character must be a letter.

The default Name (for the first physics interface in the model) is patd.

TRANSIENT SOLVER SETTINGS

Select the Time stepping (method) as Manual (default and recommended) or Automatic/
free and then enter the Maximum frequency to resolve in the model. The default
frequency is set to 1000[Hz] but should be changed to reflect the frequency content
of the sources used in the model. The generated solver will be adequate in most
situations if the computational mesh also resolves the frequency content in the model.
Note that any changes made to these settings (after the model is solved the first time)
will only be reflected in the solver if Show Default Solver or Reset Solver to Defaults is
selected in the study.

The rest of the settings are the same as for The Pipe Acoustics, Frequency Domain
Interface.

• Edge, Boundary, Point, and Pair Nodes for the Pipe Acoustics
Interfaces
• Theory for the Pipe Acoustics Interfaces

Edge, Boundary, Point, and Pair Nodes for the Pipe Acoustics
Interfaces
The Pipe Acoustics, Frequency Domain Interface and The Pipe Acoustics, Transient
Interface have these edge, boundary, point, and pair nodes, listed in alphabetical order,
available from the Physics ribbon toolbar (Windows users), Physics context menu (Mac
or Linux users), or right-click to access the context menu (all users).

In general, to add a node, go to the Physics toolbar, no matter what

operating system you are using. Subnodes are available by clicking the
parent node and selecting it from the Attributes menu.

THE PIPE ACOUSTICS FREQUENCY DOMAIN AND TRANSIENT INTERFACES | 765

 • Closed • Pipe Properties
• End Impedance • Pressure
• Fluid Properties • Velocity
• Initial Values • Volume Force

In the COMSOL Multiphysics Reference Manual see Table 2-4 for links
to common sections and Table 2-5 to common feature nodes. You can
also search for information: press F1 to open the Help window or Ctrl+F1
to open the Documentation window.

Initial Values
The Initial Values node adds initial values for the pressure and tangential velocity that
can serve as an initial condition for a transient simulation or as an initial guess for a
nonlinear solver.

INITIAL VALUES
Enter values or expressions for the initial value of the Pressure p (SI unit: Pa) and the
Tangential Velocity u (SI unit: m/s).

Fluid Properties
The Fluid Properties node adds the momentum and continuity equations solved by the
physics interface, except for volume forces which are added by the Volume Force node.
The node also provides an interface for defining the material properties of the fluid.

MODEL INPUTS
Enter a value for the and Background mean flow pressure p0 (SI unit: Pa) and the
Background mean flow temperature T0 (SI unit: K). These will also serve as input the
material parameters that depend on the pressure and temperature.

BACKGROUND VELOCITY
Enter a value or expression for the Background mean flow velocity u0 (SI unit: m/s).
Values for the background variables for the pressure p0 and velocity u0 can be obtained
by solving a Pipe Flow model on the same geometry. The Pipe Flow interface requires
the Pipe Flow Module.

766 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

PHYSICAL PROPERTIES
Select a Fluid model as Linear elastic (the default). Then select the Fluid Material as Edge
Material (the default) or select another material if added to the Materials node.

The default Density ρ (SI unit: kg/m3) and Speed of sound cs (SI unit: m/s) use the
values From material. For User defined enter different values or expressions.

Pipe Properties
The Pipe Properties node is used to define the pipe shape, pipe model, wall drag force,
and flow profile correction factor.

PIPE SHAPE
Select a pipe shape from the list — Not set (the default), Circular, Square, Rectangular,
or User defined.

• For Circular enter a value or expression for the Inner diameter di (SI unit: m). The
default is 10 cm (0.1 m).
• For Square enter a value or expression for the Inner width wi (SI unit: m). The
default is 5 cm (0.05 m).
• For Rectangular enter a value or expression for the Inner width wi (SI unit: m; the
default is 5 cm) and Inner height hi (SI unit: m; the default is 10 cm).
• For User defined enter a value or expression for the Cross-sectional area A (SI unit:
m2; the default is 0.01 m2) and Wetted perimeter Z (SI unit: m; the default is
0.4 m).

PIPE MODEL
Select a Pipe model — Incompressible cross section (the default), Zero axial stress,
Anchored at one end, or Anchored at both ends.

When Zero axial stress, Anchored at one end, or Anchored at both ends is chosen, select
an option from the Young’s modulus E (SI unit: Pa) and Wall thickness Δw lists — Not
set (the default) or User defined. For User defined in either case, enter different values
or expressions.

For Anchored at one end or Anchored at both ends also select an option from the
Poisson’s ratio v (dimensionless) list — Not set (the default) or User defined. For User
defined enter a value or expression for Poisson’s ratio.

THE PIPE ACOUSTICS FREQUENCY DOMAIN AND TRANSIENT INTERFACES | 767

 WALL DRAG FORCE
Enter a value or expression for τw (SI unit: N/m2). The default is 0 N/m2.

FLOW PROFILE CORRECTION FACTOR

Enter a value or expression for β (dimensionless). The default is 1. For most practical
applications this correction factor is 1 as the propagating waves are assumed plane and
uniform. This value should typically be changed if a wall drag force is introduces or if
a non-plug-flow background flow field is used. The flow profile correction factor is
defined as

β =  ( u ˜ + u˜ )dA 2
˜ + u˜ ) 2 dA ⁄  ( u
 0    0 

where the velocity field is the actual one that exists in the pipe cross section. The factor
thus measures the deviation from a flat background flow profile (plug flow) and a plane
propagating acoustic wave.

Flow Profile Correction Factor

Volume Force
Use the Volume Force node to add a volumetric force acting on the fluid in the pipe.
The feature can be used to model sources that generate propagating waves in a pipe
system. The force is projected in the tangential direction of the pipe,

VOLUME FORCE
Enter the three components of the Volume Force F (SI unit: N/m3).

Closed
Use the Closed node to impose zero velocity. This is the default condition added on all
end points.

Theory for the Pipe Acoustics Boundary Conditions

768 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

Pressure
Use the Pressure node to define the boundary pressure at the pipe ends.

PRESSURE
Enter a value or expression for the Pressure p (SI unit: Pa). The default is 0 Pa.

In the frequency domain p represents the amplitude and phase (if it is complex-valued)
of a harmonic pressure source.

In the time domain enter an expression for the pressure p, for example, a forward
moving sinusoidal wave of amplitude 1 Pa can be written as
1[Pa]*sin(omega*t-k*x), where omega and k are parameters defining the angular
frequency and wave number, respectively.

CONSTRAINT SETTINGS
To display this section, click the Show More Options button ( ) and select Advanced
Physics Options in the Show More Options dialog box.

Theory for the Pipe Acoustics Boundary Conditions

Velocity
Use the Velocity node to prescribe a velocity at the pipe ends.

VELOCITY
Enter a value or expression for the velocity uin (SI unit: m/s) at the inlet and/or outlet
of a pipe. The default is 0 m/s. The velocity uin is defined relative to background flow
u0 and thus in the tangential coordinate system. Enable the Show physics symbols from
the Graphics and Plot Windows menu on the Preference dialog box in order to visualize
the boundary or edge tangent direction. Click the Fluid Properties node to see the
tangents as a red arrows.

• Theory for the Pipe Acoustics Boundary Conditions

• The Preferences Dialog Box in the COMSOL Multiphysics Reference
Manual

THE PIPE ACOUSTICS FREQUENCY DOMAIN AND TRANSIENT INTERFACES | 769

 End Impedance
Use the End Impedance node to model conditions at the end of a pipe. The condition
can either model an infinite pipe and thus represent the characteristic impedance of the
pipe system at that point. This results in a zero reflection condition. Alternatively, the
condition can represent the radiation impedance of an open pipe in either a flanged (in
an infinite baffle) or unflanged (a pipe ending in free open space). The end impedance
can also be user defined and could then represent modeled or experimental values for
a specific pipe configuration.

The wave speed c in the pipe can be different from the speed of sound cs
in an open space. It then depends on the elastic properties of the pipe
structure. It is defined in Equation 9-6 in the Governing Equations
section.

The wave speed can be evaluated as sqrt(1/patd.invc2) or sqrt(1/

pafd.invc2) during the analysis and results stage.

END IMPEDANCE
Select an Impedance model.

For The Pipe Acoustics, Transient Interface select Infinite pipe (low Mach number limit)
(the default) or User defined. The Infinite pipe (low Mach number limit) models and
infinite pipe by specifying the characteristic impedance at that point. This condition
creates a non-reflecting boundary. The expression is valid for small values of the Mach
number Ma = u0/c. For User defined enter an End impedance Zend (SI unit: Pa·s/m).
The default is [Link]*(sqrt(1/patd.invc2)), which is ρ·c.

For The Pipe Acoustics, Frequency Domain Interface select from the following —
Infinite pipe (low Mach number limit) (the default), Infinite pipe, Flanged pipe, circular,
Flanged pipe, rectangular, Unflanged pipe, circular (low ka limit), Unflanged pipe, circular,
or User defined.

• For Infinite pipe enter a Wave number k (SI unit: rad/m). The default expression is
[Link]*(sqrt(pafd.invc2)). This end impedance models the infinite pipe
using the full (nonlinear) dispersion relation. It is valid for all Mach numbers but
requires the additional input of the wave number k.
• For Flanged pipe, circular enter an Inner radius a (SI unit: m). The default expression
is [Link]/2. This end impedance models the radiation impedance of a circular
pipe terminated in an infinite baffle. It is an exact analytical result valid for all

770 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

 frequencies and pipe radii. In the low frequency limit it reduces to the classical
results:

Z end = ρc  --- ( ka ) + i ( 0.8216 ⋅ ka )

1 2
2

• For Flanged pipe, rectangular enter an Inner width wi (SI unit: m). The default is 5 cm
(0.05 m). Also enter an Inner height hi (SI unit: m). The default is 10 cm (0.1 m).
This end impedance models the radiation impedance of a pipe of rectangular cross
section terminated in an infinite baffle. The model is only valid in the low frequency
range where kw i « 1 and kh i « 1 .
• For Unflanged pipe, circular (low ka limit) or Unflanged pipe, circular enter an Inner
radius a (SI unit: m). The default expression is [Link]/2. These two end
impedance models prescribe the radiation impedance of an unflanged circular pipe
(a pipe ending in free open space). The first model is the classical low-frequency
approximation valid for ka « 1 . While the second model extends the frequency
range to ka < 3.83.
For User defined enter an End impedance Zend (SI unit: Pa⋅s/m). The default expression
is [Link]*(sqrt(1/pafd.invc2)), which is ρ·c.

For a detailed review of the end impedance models see: Theory for the
Pipe Acoustics Boundary Conditions.

THE PIPE ACOUSTICS FREQUENCY DOMAIN AND TRANSIENT INTERFACES | 771

 Theory for the Pipe Acoustics
Interfaces
The equations governing the propagation of sound in pipes stem from considering
momentum, mass, and energy balances for a control volume of a piece of pipe. The
resulting equations are expressed in the cross-sectional averaged variables and reduces
the equations to a 1D component with scalar dependent variables. The present theory
assumes no thermal conduction and thus no losses due to thermal conduction
(isentropic sound propagation).

In this section:

• Governing Equations
• Theory for the Pipe Acoustics Boundary Conditions
• Solving Transient Problems
• Cutoff Frequency
• Flow Profile Correction Factor
• References for the Pipe Acoustics Interfaces

Governing Equations
The continuity equation derived for a control volume is given by

∂---------------
( Aρ )-
+ ∇ ⋅ ( Aρu ) = 0 (9-1)
∂t

and the corresponding momentum balance equation is

∂-------------------
( ρAu )- 2
+ ∇ ( ρAβu ) = – A∇p – τ w Z + AF (9-2)
∂t

where Z is the inner perimeter of the pipe and A = A(x, p,...) is the inner wetted
cross-sectional area, u is the area-averaged mean velocity, which is also defined in the
tangential direction u = uet, p is the mean pressure along the pipe, τw is the wall drag
force, and F is a volume force. The gradient is taken in the tangential direction et. The
term β is a flow profile correction factor relating the mean of the squared total velocity
to the square of the mean velocity. Such that

772 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

 β =  ----  u˜ dA  ⁄ u
1 ˜
p = ---- p˜ dA
1 1 2
  
2
u = ---- u ⋅ dA (9-3)
A A A 

where

p˜ = p˜ ( x ) and u˜ = u˜ ( x )

are the local non-averaged parameters. Again p and u are the area-averaged dependent
variables.

LINEARIZATION
The governing equations are now linearized; that is, all variables are expanded to first
order assuming stationary zero-order values (steady-state background properties). The
acoustic variations of the dependent variables are assumed small and on top of the
background values. This is done according to the following scheme:

u ( x, t ) = u 0 ( x ) + u 1 ( x, t )
p ( x, t ) = p 0 ( x ) + p 1 ( x, t )
ρ ( x, t ) = ρ 0 ( x ) + ρ 1 ( x, t )
A ( x, t ) = A 0 ( x ) + A 1 ( x, t )

where A0 is often only function of x; however, A0 can be changed by external factors

such as heating or structural deformation, thus the time dependency. The first-order
terms represent small perturbations on top of the background values (zero order).
They are valid for

2
ρ1 « ρ0 p1 « ρ0 c 0 u1 « c0 A1 « A0

Moreover, the perturbations for the fluid density and cross-sectional area are expanded
to first order in p0 in a Taylor series such that

∂ρ
ρ 1 = ρ – ρ 0 = ( p – p 0 ) ------
∂p s
0

∂A
A 1 = A – A 0 = ( p – p 0 ) -------
∂p s
0

where the subscript s refers to constant entropy; that is, the processes are isentropic.
The relations for the fluid compressibility and the cross-sectional area compressibility
are

THEORY FOR THE PIPE ACOUSTICS INTERFACES | 773

 1 1 1 ∂ρ 1 1
β 0 = ------- = ------ = ------ ------ = ------ ----2-
K0 Ks ρ 0 ∂p s ρ0 c
0 s

1 ∂A 1
β A = ------- ------- = -------
A 0 ∂p s KA
0

Here, β0 is the fluid compressibility at the given reference pressure p0, the isentropic
bulk speed of sound is denoted cs, and ρ0 is the fluid density at the given reference
temperature and reference pressure. βA is the effective compressibility of the pipe’s
cross-sectional A0 due to changes in the inner fluid pressure. The bulk modulus K is
equal to one over the compressibility.

Inserting the above expansions into the governing equations (Equation 9-1 and
Equation 9-2) and retaining only first-order terms yield the pipe acoustics equations
including background flow. These are:

1 ∂p 1   u0 
A 0 ----2- --------- + ∇ ⋅  A 0 ρ 0  u 1 + -----------2 p 1  = 0
c ∂t   ρ0 c 
2
 ∂u 1 u 0 ∂p 1  u0 
ρ 0 A 0  ---------- + -----------2 --------- + ∇  A 0 β ------ p 1 + 2ρ 0 A 0 βu 0 u 1
 ∂t ρ 0 c ∂t 
2
 c  (9-4)

+ A 0 ( ∇p 1 + p 1 β A ∇p 0 ) + τ w Z – A F = 0
ρ0
----- = ρ 0 ( β 0 + β A ) = ρ 0  ------
1 1
1- ------- 1- -------
----
2  K 0 + K A = 2 + K A
c cs

where c is the effective speed of sound in the pipe (it includes the effect due to the
elastic properties of the pipe defined through KA). The bulk modulus for the
cross-sectional area KA is given by the pipe material properties according to the
so-called Korteweg formula (see Ref. 2). For a system with rigid pipe walls cs = c as KA
tends to infinity.

Using the fact that the velocity is taken along the tangential direction et, the governing
equations are rewritten in terms of the scalar values u and p and projected onto the
tangent. The 0 subscript is dropped on the density and area and the 1 subscript is also
dropped on the dependent variables.

774 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

 1 ∂p   u0  
A ----2- ------ + ∇ t  Aρ  u + --------2- p  ⋅ e t = 0
c ∂t   ρc  
2
 ∂u u 0 ∂p  u0 
ρA  ------ + --------2- ------ + ∇ t  Aβ -----2- p + 2ρAβu 0 u ⋅ e t
 ∂t ρc ∂t   c  (9-5)

+ A ( ∇t p + p βA ∇t p0 ) ⋅ et + τw Z –A ( F ⋅ et ) = 0
ρ
= ρ ( β 0 + β A ) = ρ  ------- + ------- = ----2- + -------
1-
----
1 1 1
2  K 0 K A
c cs KA

where ∇ t is the tangential derivative, τw is the tangential wall drag force (SI unit: N/
m2) and F is a volume force (SI unit: N/m3).

GOVERNING EQUATIONS

Pipe Acoustics, Transient Interface

Finally, the expression for the time derivative of the pressure in the momentum
equation is replaced by spatial derivatives using the continuity equation. This yields the
equations solved in the Pipe Acoustics, Transient interface:

1 ∂p   u0  
A ----2- ------ + ∇ t  Aρ  u + --------2- p  ⋅ e t = 0
c ∂t   ρc  
2
∂u  u0    u0  
ρA ------ + ∇ t  Aβ -----2- p + 2ρAβu 0 u ⋅ e t + u 0 ∇ t  Aρ  u + --------2- p  ⋅ e t
∂t  c    ρc   (9-6)

+ A ( ∇t p + p βA ∇t p0 ) ⋅ et + τw Z –A ( F ⋅ et ) = 0
ρ
= ρ ( β 0 + β A ) = ρ  ------- + ------- = ----2- + -------
1-
----
1 1 1
2  K 0 K A K
c cs A

Pipe Acoustics, Frequency Domain Interface

In the frequency domain all variables are assumed to be time harmonic such that

p = p˜ ( x )e
iωt
(9-7)
u = u˜ ( x )e
iωt

inserting this into the governing Equation 9-6 (and dropping the tilde) yields the
equations solved in the Pipe Acoustics, Frequency Domain interface:

THEORY FOR THE PIPE ACOUSTICS INTERFACES | 775

 A   u0  
iω ----2- p + ∇ t  Aρ  u + --------2- p  ⋅ e t = 0
c   ρc  
2
 u0    u0  
iωρAu + ∇ t  Aβ -----2- p + 2ρAβu 0 u ⋅ e t + u 0 ∇ t  Aρ  u + --------2- p  ⋅ e t
 c    ρc   (9-8)

+ A ( ∇ t p + p βA ∇ t p0 ) ⋅ et + τ w Z –A ( F ⋅ et ) = 0
ρ
= ρ ( β 0 + β A ) = ρ  ------- + ------- = ----2- + -------
1-
----
1 1 1
2  K 0 K A
c cs KA

where ω = 2πf is the angular frequency and f is the frequency.

Theory for the Pipe Acoustics Boundary Conditions

PRESSURE, OPEN, AND CLOSED CONDITIONS

The simplest boundary conditions to specify are to prescribe the pressure or the
velocity at the pipe ends. These result in the Pressure condition

p = p in

and the Velocity condition

u = u in

and can be set independently of each other leaving the other dependent variable free.
A special subclass of the velocity condition is the Closed condition, where

u = 0

corresponds to the sound-hard wall condition in pressure acoustics. It is also assumed

here that u0 = 0 at a closed boundary.

END IMPEDANCE CONDITION

At the end of pipes the relation between the pressure and the velocity can be defined
in terms of an end impedance Zend. The End Impedance condition is in the Pipe
Acoustics interface given by

 u0   1 u 0
A  ρu + -----2- p = A  ρ ----------- + -----2- p (9-9)
 c   Z end c 

776 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

where Zend = p/u (SI unit: (Pa⋅s)/m). Different models for the end impedance exist
in the Pipe Acoustics interfaces. The variety depend on if the transient or the frequency
domain equations are solved.

Transient End-Impedance Models

In the transient version of the physics interface the end impedance can be user-defined
or set to mimic an infinite long pipe for low Mach number background flow
conditions. In this case it is assumed that the pipe continues with constant cross section
A and that there is no external body force F and drag τw. Because the acoustic waves
are, by design, always normal to the pipe ends. In order to define the relation between
the pressure and the velocity (the impedance) the dispersion relation for a plane wave
needs to be determined.

In order to do so insert the assumed plane waveform

p = Re ( p˜ e
i ( ωt – kx )
)
˜
u = Re ( u e
i ( ωt – kx )
)

into the governing Equation 9-6 and solve for the desired relations. After some
manipulation this results in

11 ω
= --- = ----2- ---  ---- – u 0
1- u
----------
Z end p c ρ k

with the dispersion relation

ω u 2
---- = βu 0 ± c β ( β – 1 )  -----0- +  1 – --- β A ∇p 0
1
(9-10)
k  c  k 

This dispersion relation is nonlinear in k. In the limit where βA tends to zero and for
small Mach numbers M (= u0/c), the expression is expanded to

ω u 2
--- β ( β – 1 )  -----0- 
---- ≅ βu 0 ± c  1 + 1
k  2  c 

Hence, the infinite pipe (low Mach number limit) end impedance relation reads

u0 2
= ----2- --- ( β – 1 )u 0 ± c  1 + --- β ( β – 1 )  ------ 
1- 11 1
---------- (9-11)
Z end c ρ 2 c

where the sign in front of c depends on the direction of propagation of the wave.

THEORY FOR THE PIPE ACOUSTICS INTERFACES | 777

 Frequency Domain End-Impedance Models
In the frequency domain many engineering relations exist for the end impedance or
radiation impedance of a pipe or waveguide. Most of the relations apply only to a
specific geometry or frequency range. The relations available in the Pipe Acoustics,
Frequency Domain interface are:

• Infinite pipe (low Mach number limit): This is the same relation as for the
transient study and the end impedance is given by Equation 9-11. This can be
thought of as the characteristic impedance of the tube.
• Infinite pipe: This relation uses the full dispersion relation given in Equation 9-10
and yields the expression

u0 2
= ----2- --- ( β – 1 )u 0 ± c β ( β – 1 )  ------ +  1 – --- β A ∇p 0
1- 11 1
---------- (9-12)
Z end c ρ c  k 

where the wave number k at the right hand side is a user input. In the frequency
domain a good estimate for this quantity is simply ω/c.

• Flanged pipe, circular: In the case of a circular pipe terminated in an infinite baffle
(a flanged pipe) an analytical expression exists for the radiation impedance (see
Ref. 1),

2J 1 ( 2ka ) 2H 1 ( 2ka )
Z end = ρc  1 – -------------------------- + i --------------------------- (9-13)
 2ka 2ka 

where J1 is the Bessel function of order 1, H1 is the Struve function of order 1, a is

the pipe radius, and k is the wave number. The Struve function is approximated
according to Ref. 3 by

H 1 ( x ) ≈ --- – J 0 ( x ) +  ------ – 5 ------------ +  12 – ------ --------------------

2 16 sin x 36 1 – cos x-
(9-14)
π π  x  π 2
x

In the low frequency limit (small ka) Equation 9-13 reduces to the classical
expression for the radiation impedance

Z end = ρc  --- ( ka ) + i ( 0.8216 ⋅ ka )

1 2
(9-15)
2

• Flanged pipe, rectangular: In the case of a pipe of rectangular cross-section (with

sides wi and hi) terminated in an infinite baffle (a flanged pipe) the radiation
impedance can be approximated by

778 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

 ρc 2 3 ⁄ 2 wi
Z end = ------  k ( w i h i ) + ik ( w i h i ) f  ------ 
2
kw i « 1, kh i « 1
2π   hi  
(9-16)
1⁄2 –1 –1 –1 ⁄ 2 2 3 ⁄ 2 2 –3 ⁄ 2 2
–1 –1 3 ⁄ 2
f ( x ) = 2x sinh x + 2x sinh x + --- x + --- x – --- ( x + x )
3 3 3

see Ref. 4 and Ref. 5.

• Unflanged pipe, circular (low ka limit): In the case of a circular pipe of radius a
ending in free air the classical low ka limit for the radiation impedance is given by

Z end = ρc  --- ( ka ) + i ( 0.6133 ⋅ ka )

1 2
ka « 1 (9-17)
4

see Ref. 1 and Ref. 5.

• Unflanged pipe, circular: A solution for the unflanged pipe exists for the case when
ka « 3.83 = 1.22π , it is presented in Ref. and is based on solving the
Wiener-Hopf integral, it reads
1+R 2ikaδ
Z end = ρc -------------- R = Re
1–R
– ( ka ) ⁄ 2 
1 + --- ( ka ) ln   --------- + ------ 
2
1 4 1 19 0.5772
R = e γ=e ka < 1 (9-18)
 6   γka 12 
– ka 
1 + ------ --------------2
3 1
R = πkae 1 < ka < 3.83
 32 ( ka ) 

where δ is an interpolation function found by numerical integration for ka = 0,

δ = 0.6133.
Common for the last four radiation impedance relations is that they do only apply
when there is no background flow present u0 = 0 (or at least when it is very small).

Solving Transient Problems

When solving transient acoustic problems where the wave shape is not necessarily
harmonic it might be necessary to resolve its spatial variations with a fine mesh, say
with a minimal scale dx. Now, in order for the numerical solution of the temporal
development of the acoustic field to be good it is necessary to restrict the maximal time
steps dt taken by the solver. The condition is known as the CFL condition (Courant–
Friedrichs–Lewy condition). For transient acoustic problems it is defined as

dt
C = c⋅
dx

THEORY FOR THE PIPE ACOUSTICS INTERFACES | 779

 where C is the Courant number, and c is the velocity.

For applications where all the shape functions are quadratic the Courant number
should be around 0.2. This condition restricts any acoustic disturbances to propagate
more than 20% of the mesh size dx during one time step dt. In the Pipe Acoustics
interface where a mixed formulation exists, with linear elements for the pressure and
quadratic elements for the velocity, the condition might have to be tightened such that
C < 0.2.

For an example where the CFL condition is used see Water Hammer:
Application Library path Pipe_Flow_Module/Verification_Examples/
water_hammer_verification.

Cutoff Frequency
The Pipe Acoustics interface assumes plane wave propagation. This means that it
cannot model the propagation of the higher order modes that can propagate above
their cutoff frequency fc. In a rectangular pipe of cross-section width wi and height hi,
the cutoff frequency is

f mn = --- c  ------ +  -----

c 1 m 2 n 2
2  w i  h i

In a pipe of circular cross section (with radius a), the cutoff frequency is

c α' mn c
f mn = ---------------
2πa

where α'mn is the n’th zero of the differential of the Bessel function J’m(x) or order
m. The first few values are α'01 = 0, α'02 = 3.83, α'11 = 1.84, and α'21 = 3.05 (see
Ref. 1 and Ref. 5 for further details).

Flow Profile Correction Factor

The flow profile correction factor, β, accounts for the ratio of the integrated local
square velocity field to the square of the integrated local velocity field (see
Equation 9-3). It is defined in terms of the total velocity field (background plus
acoustic variations).

780 | CHAPTER 9: PIPE ACOUSTICS INTERFACES

In the case of no-background flow (u0 = 0) β is 1 in the absence of a wall drag
coefficient, as only plane wave modes propagate. If a wall drag force is introduced, to
model some loss mechanism, β starts to differ slightly from 1. This can for example be
losses introduced to model viscous and thermal effects in narrow pipes.

In the presence of a background u0 the factor β can be set different 1 in order to model
a non-flat velocity profile inside the tube. The value of b (and the actual shape of the
background field) influences the convective momentum transfer balances. The places
where β enter the governing equations are multiplied with either the Mach number or
the Mach number squared, indicating that the effects become important for an
increasing background flow.

References for the Pipe Acoustics Interfaces

1. D.T. Blackstock, Fundamentals of Physical Acoustics, John Wiley & Sons, 2000.

2. M.S. Ghidaoui, M. Zhao, D.A. McInnis, and D.H. Axworthy, “A Review of Water
Hammer Theory and Practice”, Applied Mechanics Reviews, ASME, 2005.

3. R.M. Aarts and A.J.E.M. Janssen, “Approximation of the Struve Function H1

Occurring in Impedance Calculations”, J. Acoust. Soc. Am., vol. 113, pp. 2635–2637,
2003.

4. O.A. Lindemann, “Radiation Impedance of a Rectangular Piston at Very Low

Frequencies”, J. Acoust. Soc. Am., vol 44, pp. 1738–1739, 1968.

5. A.D. Pierce, Acoustics: An Introduction to its Physical Principles and

Applications, Acoustics Society of America, 1994.

H. Levine and J. Schwinger, “On the Radiation of Sound from an Unflanged Circular
Pipe”, Phys. Rev., vol. 73, pp. 383–406, 1948.

THEORY FOR THE PIPE ACOUSTICS INTERFACES | 781

782 | CHAPTER 9: PIPE ACOUSTICS INTERFACES
 10

Multiphysics Couplings

This chapter contains a description of all the Multiphysics coupling features of the
Acoustics Module. Several of the coupling features are used in predefined
multiphysics interfaces but can as well be added manually when modeling
multiphysics systems.

• Coupling Features
• Predefined Multiphysics Interfaces
• Modeling with Multiphysics Couplings

783
 Coupling Features
The Acoustics Module contains a number of Multiphysics Coupling features that can
couple acoustic domains to structures and porous materials, thermoviscous acoustics
to pressure acoustics, a flow to an aeroacoustic interface (convected acoustics), or an
aeroacoustic flow source to pressure acoustics. The couplings features are described
below.

The couplings features that exist are:

• Acoustic-Structure Boundary
• Pair Acoustic-Structure Boundary
• Thermoviscous Acoustic-Structure Boundary
• Pair Thermoviscous Acoustic-Structure Boundary
• Aeroacoustic-Structure Boundary
• Acoustic-Thermoviscous Acoustic Boundary
• Acoustic-Porous Boundary
• Porous-Structure Boundary
• Background Potential Flow Coupling
• Background Fluid Flow Coupling
• Aeroacoustic Flow Source Coupling
• Acoustic FEM-BEM Boundary
• Acoustic-Pipe Acoustic Connection
• Acoustic-Structure Boundary, Time Explicit
• Pair Acoustic-Structure Boundary, Time Explicit
• Piezoelectric Effect, Time Explicit
• Piezoelectric Effect
• Lorentz Coupling
• Magnetomechanical Forces

Acoustic-Structure Boundary
The Acoustic-Structure Boundary coupling ( ) is used to couple a Pressure Acoustics
model to any structural component. This includes both the FEM based acoustics

784 | CHAPTER 10: MULTIPHYSICS COUPLINGS

interfaces and the BEM based acoustics interface. The feature couples to Solid
Mechanics, Poroelastic Waves, Shell, Layered Shell, Membrane, and Multibody Dynamics
interfaces.

The coupling includes the fluid load on the structure and the structural acceleration as
experienced by the fluid. For thin interior structures like shells or membranes with
fluid on both sides, a slit is added to the pressure variable and care is taken to couple
the up and down sides.

Mathematically the condition on exterior boundaries reads

– n ⋅  – ----- ( ∇p t – q d ) = – n ⋅ u tt
1
ρc
FA = pt n

where utt is the structural acceleration, n is the surface normal, pt is the total acoustic
pressure and FA is the load (force per unit area) experienced by the structure. On
interior boundaries the condition reads

– n ⋅  – ----- ( ∇p t – q d ) = – n ⋅ u tt
1
 ρc 
up

– n ⋅  – ----- ( ∇p t – q d )
1
= – n ⋅ u tt
ρc down
F A = ( p t,down – p t,up )n

such that the acoustic load is given by the pressure drop across the thin structure. The
up and down subscripts refer to the two sides of the interior boundary.

In 2D components it is possible to define the out-of-plane thickness for

the structural interfaces while in acoustics this thickness effectively is unity
(set to 1 m). A rescaling is automatically performed for the pressure load
to account for this difference in interpretation in 2D.

In models with large deformations of the structure, for example, if a

membrane is prestressed and deforms, the resulting area change is
automatically included when using the multiphysics coupling between the
structure and the acoustic domain.

COUPLING FEATURES | 785

 SETTINGS
The Label is the default multiphysics coupling feature name.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is asb1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Structure lists include all applicable physics interfaces.

The default values depend on how this node is created.

• If it is added from the Physics ribbon (Windows users), Physics contextual toolbar
(Mac and Linux users), or context menu (all users), then the first physics interface
of each type in the component is selected as the default.
• If it is added automatically when a multiphysics interface is selected in the Model
Wizard or Add Physics window, then the two participating interfaces are selected.

You can also select None from either list to uncouple this node from a physics interface.
If the physics interface is removed from the Model Builder, for example Pressure
Acoustics, Frequency Domain is deleted, then the Acoustics list defaults to None as there
is nothing to couple to.

If a physics interface is deleted and then added to the model again, then
in order to re-establish the coupling, you need to choose the physics
interface again from the Acoustics or Structure lists. See Multiphysics
Modeling Workflow in the COMSOL Multiphysics Reference Manual.

Pair Acoustic-Structure Boundary

The Pair Acoustic-Structure Boundary coupling ( ) is used to couple the Pressure
Acoustics, Frequency Domain or the Pressure Acoustics, Transient interface to the Solid
Mechanics interface in an assembly geometry where identity pairs have been created.
This allows the use of non-confirming mesh at the acoustic-structure boundary. As the

786 | CHAPTER 10: MULTIPHYSICS COUPLINGS

wave speeds differ in solids and fluids, the computational mesh can take advantage of
this when resolving the waves. In this way saving degrees of freedom when solving.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Structure lists include all applicable physics interfaces.

PAIR SELECTION
Add and select the identity pairs where the pair multiphysics coupling feature is
applied.

Thermoviscous Acoustic-Structure Boundary

The Thermoviscous Acoustic-Structure Boundary coupling ( ) is used to couple a
Thermoviscous Acoustics interface to any structural component. The feature couples
to Solid Mechanics, Shell, Layered Shell, Membrane, and Multibody Dynamics interfaces.

The coupling prescribes continuity in the displacement field

∂u solid
u t,fluid = iωu solid or u t,fluid = ----------------
∂t

where ut,fluid is the total fluid velocity (including a background component if

applicable) and usolid is the solid displacement. The first equation is in the frequency
domain and the second in the time domain, respectively. This coupling results in the
stress also being continuous across the boundary. The condition for the total
temperature Tt can be set to either isothermal or adiabatic. In the case where a shell or
membrane is interior to the thermoviscous acoustics domain, a slit is automatically
applied to the pressure p and temperature T degrees of freedom (DOFs).

The coupling has a special formulation for eigenfrequency studies, which allows the
program to find the eigenvalue for coupled systems.

For an example using this multiphysics coupling see the Vibrating

Micromirror with Viscous and Thermal Damping model: Application
Library path Acoustics_Module/Vibrations_and_FSI/vibrating_micromirror

SETTINGS
See Settings for further details about Label and Name.

COUPLING FEATURES | 787

 The default Name (for the first multiphysics coupling feature in the model) is tsb1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The
Thermoviscous Acoustics and Structure lists include all applicable physics interfaces. See
the Coupled Interfaces in Acoustic-Structure Boundary for details.

THERMAL
Select the Thermal condition that applies on the solid surface to be either Isothermal (the
default selected) or Adiabatic.

CONSTRAINT SETTINGS
These settings are shown when the Advanced Physics Options is selected. It controls how
the constraints (continuity in displacement and the thermal condition) are handled.
For the Mechanical constraints type, select either Study controlled (the default), Weak
constraints, or Pointwise constraints. For the Thermal constraint type, select either
Pointwise constraint (the default) or Weak constraints. Finally, select the Constraint
method to apply to the pointwise constraints, either Elemental or Nodal (the default).

When the Mechanical constraints type is set to Study controlled, the weak constraints are
automatically selected when an eigenfrequency study is performed. This type of
coupling is necessary as the eigenvalue (the angular frequency ω) enters the coupling
expression. This will create extra variables at the boundary (so-called Lagrange
multipliers), ensuring the correct behavior and solution. For a normal frequency
domain study, the pointwise constraint is automatically selected. In the time domain
there are no options, and the weak formulation is always used.

Pair Thermoviscous Acoustic-Structure Boundary

The Pair Thermoviscous Acoustic-Structure Boundary coupling ( ) is used to couple
the Thermoviscous Acoustics, Frequency Domain or the Thermoviscous Acoustics, Transient
interface to the Solid Mechanics interface in an assembly geometry where identity pairs
have been created. This allows the use of non-confirming mesh at the
acoustic-structure boundary. As the wave speeds differ in solids and fluids, the
computational mesh can take advantage of this when resolving the waves. In this way
saving degrees of freedom when solving.

788 | CHAPTER 10: MULTIPHYSICS COUPLINGS

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Structure lists include all applicable physics interfaces.

THERMAL
Select a condition for the temperature variations either Isothermal (the default) or
Adiabatic.

CONSTRAINT SETTINGS
These settings are shown when the Advanced Physics Options is selected. It controls how
the constraints (continuity in displacement and the thermal condition) are handled.
For the Mechanical constraints type, select either Study controlled (the default), Weak
constraints, or Pointwise constraints. For the Thermal constraint type, select either
Pointwise constraint (the default) or Weak constraints. Finally, select the Constraint
method to apply to the pointwise constraints, either Elemental or Nodal (the default).

When the Mechanical constraints type is set to Study controlled, the weak constraints are
automatically selected when an eigenfrequencies study is performed. This type of
coupling is necessary as the eigenvalue (the angular frequency ω) enters the coupling
expression. This will create extra variables at the boundary (so-called Lagrange
multipliers), ensuring the correct behavior and solution. For a normal frequency
domain study, the pointwise constraint is automatically selected. In the time domain a
penalty/Nitsche formulation is always used automatically.

PAIR SELECTION
Add and select the identity pairs where the pair multiphysics coupling feature is
applied.

Aeroacoustic-Structure Boundary
The Aeroacoustic-Structure Boundary coupling ( ) is used to couple an Aeroacoustic
interface (it only applies to the Linearized Navier–Stokes interfaces) to any structural
component. The feature couples to Solid Mechanics, Shell, Layered Shell, Membrane, and
Multibody Dynamics interfaces. The coupling can be used to, for example, model
fluid-structure interaction (FSI) in the frequency domain.

The coupling prescribes continuity in the displacement field

∂u solid
u t,fluid = iωu solid or u t,fluid = ----------------
∂t

COUPLING FEATURES | 789

 where ut,fluid is the total fluid velocity (including a background component if
applicable) and usolid is the solid displacement. The first equation is in the frequency
domain and the second is in the time domain. This coupling results in the stress also
being continuous across the boundary. The condition for the total temperature Tt can
be set to either isothermal or adiabatic. In the case where a shell or membrane is
interior to the thermoviscous acoustics domain, a slit is automatically applied to the
pressure p and temperature T degrees of freedom (DOFs).

The coupling has a special formulation for eigenfrequency studies, which allows the
user to find the eigenvalue for coupled systems.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is aesb1.

THERMAL
Select the Thermal condition that applies on the solid surface to be either Isothermal (the
default selected) or Adiabatic.

CONSTRAINT SETTINGS
See Constraint Settings under the Thermoviscous Acoustic-Structure Boundary.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The
Aeroacoustics and Structure lists include all applicable physics interfaces. See the
Coupled Interfaces in Acoustic-Structure Boundary for details.

Acoustic-Thermoviscous Acoustic Boundary

The Acoustic-Thermoviscous Acoustic Boundary coupling ( ) is used to couple a
Thermoviscous Acoustics interface to a Pressure Acoustics interface (in both frequency
and time domain). This includes both the FEM-based acoustics interfaces and the
BEM-based acoustics interface. As it is only necessary to solve the full detailed
thermally conducting and viscous model near walls in the boundary layer region, it
makes sense to switch to classical pressure acoustics outside this region. This approach
saves a lot of memory and solution time due to the reduced number of degrees of
freedom.

790 | CHAPTER 10: MULTIPHYSICS COUPLINGS

Note that, when using the coupling at places where pressure acoustics and
thermoviscous acoustics interact at a common wall, the coupling becomes unphysical
if no-slip and isothermal conditions are used in the thermoviscous domain. The reason
is that the conditions at the wall must match between the physics. The coupling
assumes adiabatic behavior and normal-stress coupling only, which is not valid in the
thermal and viscous boundary layers. The introduced error depends on the model size
and the thickness of the acoustic boundary layer. If it makes physical sense then, use
the slip and adiabatic options on that specific wall (add an extra wall boundary
condition). In general using the coupling in a cross section of a waveguide will not be
physically correct if the boundary layer effects are of importance.

The coupling prescribes continuity in the total normal stress (dynamic condition) and
the total normal acceleration for the mechanical part (kinematic condition). An
adiabatic condition is prescribed for the total temperature to match the physical
assumptions of the pressure acoustics formulation. The coupling reads

– n ⋅  – ----- ( ∇p t – q d ) = – n ⋅ iωu t

1 pa
(frequncy domain)
ρc
∂u t
– n ⋅  – ----- ( ∇p t – q d ) = – n ⋅ --------
1 pa
(time domain)
 ρc  ∂t

– p t I + μ ( ∇u t + ( ∇u t ) T ) –  ------- – μ B ( ∇ ⋅ u t )I n = – p t n
2μ pa
 3 
– n ⋅ ( – k∇T t ) = 0

The thermoviscous acoustic pressure variable is pt and the pressure acoustic pressure is
pa
here denoted p t for clarity.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is atb1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Thermoviscous Acoustics lists include all applicable physics interfaces. See the
Coupled Interfaces in Acoustic-Structure Boundary for details.

COUPLING FEATURES | 791

 Acoustic-Porous Boundary
The Acoustic-Porous Boundary coupling ( ) is used to couple a Pressure Acoustics,
Frequency Domain interface to a porous material (Biot’s model) domain from the
Poroelastic Waves interface. This includes both the FEM based acoustics interfaces and
the BEM based acoustics interface. The boundary coupling feature includes the
following interaction between the fluid and the porous domains:

• Continuity of the fluid pressure on the boundaries where the fluid interacts with the
porous domain. The pore pressure in the porous domain is set equal to the total
pressure in the fluid domain
p pore = p t

• A pressure load from the fluid pressure is experienced by the elastic waves in the
porous material
n ⋅ ( σ d – α B pI ) = – np t

where pt is the total acoustic pressure in the fluid domain and the left hand side
represents the total stress for the saturated porous domain.
• The pressure acoustic domain also experiences a normal acceleration that depends
both on the acceleration of the porous matrix skeleton but also on the pore pressure.
Because of the pressure boundary condition, which is a bidirectional constraint, this
condition reduces to the fluid experiencing a normal acceleration
2
a n = ( iω ) u

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is apb1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Porous media lists include all applicable physics interfaces. See the Coupled
Interfaces in Acoustic-Structure Boundary for details.

Headphone on an Artificial Ear. The Application Library path

Acoustics_Module/Electroacoustic_Transducers/headphone_artificial_ear

792 | CHAPTER 10: MULTIPHYSICS COUPLINGS

Porous-Structure Boundary
The Porous-Structure Boundary coupling ( ) is used to couple a porous domain
(Biot’s model) domain from the Poroelastic Waves interface to a Solid Mechanics, Shell,
Layered Shell, Membrane, or Multibody Dynamics interface. At the solid-porous
boundaries continuity of the displacement field is applied.

In 2D components it is possible to define the out-of-plane thickness for

the structural interfaces while in acoustics this thickness effectively is unity
(set to 1 m). A rescaling is automatically performed for the pressure load
to account for this difference in interpretation in 2D.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is psb1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Porous
media and Structure lists include all applicable physics interfaces. See the Coupled
Interfaces in Acoustic-Structure Boundary for details.

Acoustics of a Particulate-Filter-Like System. The Application Library

path Acoustics_Module/Automotive/acoustics_particulate_filter

Background Potential Flow Coupling

The Background Potential Flow Coupling ( ) is a unidirectional multiphysics coupling
that couples the background flow calculated by a Compressible Potential Flow interface
to the Linearized Potential Flow interfaces. The coupling provides the flow inputs for
the Linearized Potential Flow Model.

For good numerical performance and accuracy, the Compressible Potential Flow
model and the linearized acoustics model need to be solved on the same mesh. This
multiphysics coupling does not rely on a mapping strategy like the Background Fluid
Flow Coupling used to couple a fluid flow solution to the linearized Euler and
linearized Navier-Stokes interfaces. The coupling is also applicable in PML domains
where the Compressible Flow model can be solved.

COUPLING FEATURES | 793

 SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is pfc1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Source and
Destinations lists include all applicable physics interfaces. See the Coupled Interfaces in
Acoustic-Structure Boundary for details.

Flow Duct. The Application Library path: Acoustics_Module/

Aeroacoustics_and_Noise/flow_duct

Background Fluid Flow Coupling

The Background Fluid Flow Coupling ( ) is a unidirectional multiphysics coupling
between a Fluid Flow interface and either the Linearized Navier–Stokes, Linearized
Euler, or Convected Wave Equation interfaces. The coupling should be solved for in
a separate study using the dedicated Mapping ( ) study.

The multiphysics coupling feature and mapping study ensures that the fluid flow
variables are mapped correctly from the fluid flow mesh to the acoustics mesh. The two
meshes are typically different because of different resolution requirements, or in the
case of the Convected Wave Equation interface because different shape function orders
are used. The mapping also per default adds a small amount of smoothing of the
mapped flow solution. This ensures that no unphysical numerical noise is introduced
to the acoustics model through the reactive terms in the governing equations.

Helmholtz Resonator with Flow: Interaction of Flow and Acoustics.

The Application Library path: Acoustics_Module/Aeroacoustics_and_Noise/
helmholtz_resnoator_with_flow

COUPLED INTERFACES
Select the Source fluid flow interface and the Destination acoustics interface.

VARIABLES TO MAP
Select the variables that need to be mapped, this depends on the physics selected. For
most cases the background mean flow velocity and pressure need to be mapped. If the

794 | CHAPTER 10: MULTIPHYSICS COUPLINGS

flow is compressible also map the density. If the flow is nonisothermal also map the
temperature. For linearized Navier-Stokes models the turbulent viscosity can also be
mapped, this ensures that acoustic waves are correctly attenuated as they propagate
through regions of high turbulence.

• Select Map the pressure (selected per default), Map the velocity (selected per default),
Map the density, Map the temperature, or Map the turbulent viscosity as necessary.
• For the velocity select Use no slip boundary on no slip walls if the mapped variables
should have a true/forced no-slip condition applied. This may modify the flow
solution slightly near walls and should be used with care. When combined with the
linearized Navier–Stokes physics this may be necessary to get consistent no-slip
conditions at walls.
• The Use symmetry on symmetry walls option, selected per default, ensures that the
mapped flow field is symmetric for symmetry conditions.
• For the pressure, velocity, density, temperature, and turbulent viscosity variables the
option Constrain the (variable) on exterior boundaries option sets up a constrain on
all exterior boundaries, where the mapped variable is set equal to the source fluid
flow variable.

SMOOTHING
Select the Smoothing method as Isotropic diffusion (the default) or None. For the
isotropic diffusion option set the (numerical) Diffusion constant (the default: 1e-2). It
is recommended to always use a small amount of smoothing to ensure smooth
gradients of the background mean flow variables. To verify that the amount added is
adequate, for example, compare the mapped variables and the fluid flow variables in a
plot.

Special care should be taken if perfectly matched layers (PMLs) are

present in the acoustic model. In this case make sure to disable the PMLs
(either in the model builder tree or in the study) when setting up and
solving the fluid flow problem. Remember to enable the PMLs again (if
disabled in the model tree) when solving the acoustics problem.

Aeroacoustic Flow Source Coupling

The Aeroacoustic Flow Source Coupling ( ) is a unidirectional multiphysics coupling
between an LES (large eddy simulations) Fluid Flow interface and the Aeroacoustic
Flow Source feature in the Pressure Acoustics, Frequency Domain interface.

COUPLING FEATURES | 795

 The coupling feature should be solved for in a separate study using the dedicated
Transient Mapping ( ) study. This step will map necessary data from the CFD
source mesh to an acoustics mesh. Following the Transient Mapping a separate Time
to Frequency FFT study has to be used to transform the flow source data from the time
domain to the frequency domain. Then the Pressure Acoustics, Frequency Domain
model is solved.

For more details about theory and setting up an aeroacoustic model with
flow-induced noise, see Modeling with the Aeroacoustic Flow Source:
Flow-Induced Noise.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is afsc1.

DOMAIN SELECTION
Select the domains where the Lighthill stress tensor is computed, mapped to the
acoustic mesh, and used in the Aeroacoustic Flow Source feature as a source term.

BOUNDARY SELECTION
If the Acoustic analogy option is set to Lighthill (see below), select boundaries where the
normal mass flow is mapped. This corresponds to sources generated by moving/
vibrating boundaries in the flow simulation. If all boundaries are rigid no boundary
selection is necessary.

COUPLED INTERFACES
Select the Source and Destination physics. The source can only be an LES physics
interface and the destination the Pressure Acoustics, Frequency Domain physic
interface.

AEROACOUSTIC FLOW SOURCE

Select the Acoustic analogy, used for the computational aeroacoustics (CAA)
simulation, as Lighthill (the default) or Aeroacoustic wave equation. For the coupling to
work the Acoustic analogy selection in the Aeroacoustic Flow Source has to match the
options selected in the multiphysics coupling.

• For Lighthill, select Include viscous stress to include the viscous stress components of
the tensor. This is, for example, necessary for models with small dimensions where

796 | CHAPTER 10: MULTIPHYSICS COUPLINGS

 damping can be important. The data mapped/coupled to the Aeroacoustic Flow
Source feature will contain the necessary 6 elements of the Lighthill stress tensor (as
the tensor is symmetric). The full tensor is used in the subsequent acoustic
simulation.
• For Aeroacoustic wave equation, only the scalar data associated with the second time
derivative of the pressure is mapped/coupled to the Aeroacoustic Flow Source
feature. This option includes neither viscous effects nor the effects of vibrating
boundaries.

SMOOTHING
Select the Smoothing method as Isotropic diffusion (the default) or None. For the
isotropic diffusion option set the (numerical) Diffusion constant (the default: 1e-4). It
is recommended to always use a small amount of smoothing to ensure smooth
gradients of the background mean flow variables.

Acoustic FEM-BEM Boundary

The Acoustic FEM-BEM Boundary coupling ( ) is used to couple the FEM based
Pressure Acoustics, Frequency Domain interface to the BEM based Pressure Acoustics,
Boundary Elements interface. The coupling works as a bidirectional coupling as most
other multiphysics couplings in the Acoustics Module. The coupling prescribes
continuity in the total pressure

p t,bem = p t,fem

where [Link] is the total pressure in the BEM based pressure acoustics interface and
pt,fem is the total pressure in the FEM based pressure acoustics interface.

Using this coupling the strengths of the FEM and BEM methods can be used where
they apply best. For example, the FEM based interface should be used in interior
domains or to model porous domains with the Poroacoustics feature. When coupled
to an open space where radiation occurs it can be advantageous to use BEM here. This
is, for example, the case when modeling loudspeaker cabinets and enclosures that have
a vent.

Another powerful application of the FEM-BEM coupling is to use the BEM based
interface, when coupled to Pressure Acoustics, Frequency Domain, as a combined
radiation condition and far-field calculating feature. The radiation condition will be
perfect as the BEM problem is solved exactly and the BEM variables can be used in

COUPLING FEATURES | 797

 postprocessing instead of the usual far-field variables. The BEM variables like, for
example, the sound pressure level [Link] can be used in the existing far-field plots.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is apb1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Boundary
elements acoustics and Finite elements acoustics lists include all applicable physics
interfaces. See the Coupled Interfaces in Acoustic-Structure Boundary for details.

Acoustic-Pipe Acoustic Connection

The Acoustic-Pipe Acoustic Connection multiphysics coupling ( ) is use used to couple
the Pipe Acoustics interfaces to the Pressure Acoustics interface. It applies in quiescent
conditions, that is, for zero background mean flow in the pipe acoustic interface. The
coupling is between a point (the end of a 1D pipe) in the pipe acoustics interface and
a boundary in the pressure acoustics interface.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is apc1.

BOUNDARY SELECTION, ACOUSTICS

Select the boundary in the pressure acoustics model that corresponds to the inlet/
outlet of the 3D model coupled to pipe acoustics.

POINT SELECTION, PIPE ACOUSTICS

Select the point in the pipe acoustics model that corresponds to the inlet/outlet of the
pipe model coupled to the 3D pressure acoustics domain.

• Acoustics of a Pipe System with 3D Bend and Junction. The

Application Library path: Acoustics_Module/Tutorials,_Pipe_Acoustics/
acoustics_pipe_system
• Probe Tube Microphone. The Application Library path:
Acoustics_Module/Tutorials,_Pipe_Acoustics/probe_tube_microphone

798 | CHAPTER 10: MULTIPHYSICS COUPLINGS

Acoustic-Structure Boundary, Time Explicit
The Acoustic-Structure Boundary, Time Explicit multiphysics coupling ( ) is used to
couple the Pressure Acoustics, Time Explicit or the Nonlinear Pressure Acoustics, Time
Explicit interface with the Elastic Waves, Time Explicit interface. The condition is added
on the boundary between the fluid domain and the solid domain for modeling acoustic
structure interaction (ASI) or vibroacoustic problems. If possible, the Pair
Acoustic-Structure Boundary, Time Explicit condition should be used as it will
improve the performance of the time explicit method.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is asbte1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Structure lists include all applicable physics interfaces. That is The Pressure
Acoustics, Time Explicit Interface and The Nonlinear Pressure Acoustics, Time
Explicit Interface, and The Elastic Waves, Time Explicit Interface can be selected here.

Propagation of Seismic Waves Through Earth. The Application Library

path: Acoustics_Module/Elastic_Waves/seismic_waves_earth

Pair Acoustic-Structure Boundary, Time Explicit

The Pair Acoustic-Structure Boundary, Time Explicit multiphysics coupling ( ) is used
to couple the Pressure Acoustics, Time Explicit or the Nonlinear Pressure Acoustics, Time
Explicit interface with the Elastic Waves, Time Explicit interface, in an assembly
geometry where identity pairs have been created.

This pair version of the coupling makes it possible to take full advantage of the time
explicit formulation, where the global time step is controlled by the minimum of all
local mesh based time steps (complying with the local CFL condition). That is using a
mesh size based on the local minimum wave speed cmin, as the mesh sizes do not need
to be conforming where the fluid and the solid domains meet. The local mesh size
hmax should always comply with

COUPLING FEATURES | 799

 c min
h max = ---------------
Nf max

where N depends on the element approximation order (1.5-2 for the default settings)
and fmax is the maximum frequency to resolve. There will often be a clear advantage
in the computation time using the Pair Acoustic-Structure Boundary, Time Explicit over
the Acoustic-Structure Boundary, Time Explicit feature due to a larger internal time step
and in general fewer degrees of freedom solved for.

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is aspte1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Acoustics
and Structure lists include all applicable physics interfaces. That is The Pressure
Acoustics, Time Explicit Interface and The Nonlinear Pressure Acoustics, Time
Explicit Interface, and The Elastic Waves, Time Explicit Interface can be selected here.

PAIR SELECTION
Add and select the identity pairs where the pair multiphysics coupling feature is
applied.

Create separate Identity boundary Pairs for components of the boundary

(where the coupling feature is used) if they meet at a corner or sharp edge.
One identity pair for each boundary. By default, identity pairs are created
automatically when forming an assembly. Use the automatically created
identity boundary pair, duplicate it and then edit them removing
boundaries, where necessary.

Piezoelectric Effect, Time Explicit

The Piezoelectric Effect, Time Explicit ( ) is a multiphysics coupling feature between
the Elastic Waves, Time Explicit interface and the Electrostatics interface used to model
a piezoelectric material. It couples the Piezoelectric Material node in elastic waves to
the Charge Conservation, Piezoelectric node in Electrostatics. The multiphysics
coupling is automatically added when the The Piezoelectric Waves, Time Explicit
Interface is added as a new multiphysics interface.

800 | CHAPTER 10: MULTIPHYSICS COUPLINGS

SETTINGS
See Settings for further details about Label and Name.

The default Name (for the first multiphysics coupling feature in the model) is pzete1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. Only the Elastic
Waves, Time Explicit and Electrostatics physics interfaces can be selected.

Piezoelectric Effect
The Piezoelectric Effect ( ) is a multiphysics coupling feature between the Solid
Mechanics interface and the Electrostatics interface used to model a piezoelectric
material. This feature is used for the finite element method (FEM) based interfaces and
is applicable for transient, frequency domain, and stationary studies. The Piezoelectric
Effect is described for The Piezoelectricity Interface.

Lorentz Coupling
The Lorentz Coupling ( ) is a multiphysics coupling feature between the Magnetic
Fields interface and the Solid Mechanics interface. It passes the Lorentz force F = J × B
from the Magnetic Fields interface to the Solid Mechanics interface and, passes the
induced electric field E = v × B from the Solid Mechanics interface to the Magnetic Fields
interface. The Lorentz force effect on a moving structure is similar to that of the
Rayleigh alpha-damping, where the damping coefficient is proportional to the
background magnetic field and the material electric conductivity.

The Lorentz force and induced electric field can also be added separately from the
interfaces, for instance, by using the Body Load feature from the Solid Mechanics
interface and the Velocity (Lorentz term) feature (in 2D and 2D axisymmetric) from the
Magnetic Fields interface. Once the Lorentz Coupling feature is applied, those interface
features with the same purpose should not be used to avoid double counting. When
modeling electroacoustic transducers, like loudspeakers, the coupling is typically added
in the voice coil domain.

The Lorentz Coupling feature is available in 2D, 2D axisymmetric, and 3D geometries

and is allowed for both time-dependent and frequency-domain studies (including
frequency domain perturbation and eigenfrequency studies). The coupling feature
handles frames as well as the correct formulation for transient and frequency domain
problems. For time-dependent studies, it only accounts for the Lorentz force
contribution when the Include geometric nonlinearity check box (in Study settings) is

COUPLING FEATURES | 801

 selected. When the displacement of the vibrating or moving part of the model cannot
be neglected (when topology changes are important), the Moving Mesh feature (added
from the Definitions) is usually used together with the Lorentz Coupling feature.

• When the Lorentz Coupling feature is applied to a domain with magnetic

materials such as a magnetizable armature, the Lorentz force is no
longer the only force contribution. In this case, the force contribution
from the magnetic polarization might need to be considered. This can
be done by adding a Force Calculation feature from the Magnetic Fields
interface and adding a Body Load feature from the Solid Mechanics
interface. Note that the Force Calculation feature includes the force
contribution not only from magnetic polarization but also from the
Lorentz force, that is, the total electromagnetic force. Therefore, when
the coupling between the Magnetic Fields interface and Solid Mechanics
interface is implemented with the Force Calculation feature and Body
Load feature, the Lorentz Coupling feature should not be used at the
same domain.
• The Magnetomechanics multiphysics interface ( ) can be used for
modeling deformation of magnetically actuated structures, which
includes interaction of magnetic fields with magnetic materials and
current carrying elements such as coils and wires.

SETTINGS
The Label is the multiphysics coupling feature name. The default Label (for the first
multiphysics coupling feature in the model) is Lorentz Coupling 1.

The Name is used primarily as a scope prefix for variables defined by the coupling node.
Refer to such variables in expressions using the pattern <name>.<variable_name>. In
order to distinguish between variables belonging to different coupling nodes or physics
interfaces, the name string must be unique. Only letters, numbers, and underscores (_)
are permitted in the Name field. The first character must be a letter.

The default Name (for the first multiphysics coupling feature in the model) is ltzc1.

COUPLED INTERFACES
This section defines the physics involved in the multiphysics coupling. The Magnetic
fields and Solid mechanics lists include all applicable physics interfaces.

802 | CHAPTER 10: MULTIPHYSICS COUPLINGS

When the Lorentz Coupling node is added from the Physics ribbon (Windows users),
Physics contextual toolbar (macOS and Linux users), or context menu (all users), then
the first physics interface of each type in the component is selected as the default.

You can also select None from either list to uncouple the Lorentz Coupling node from
a physics interface. If the physics interface is removed from the Model Builder, for
example Solid Mechanics is deleted, then the list defaults to None as there is nothing to
couple to.

If a physics interface is deleted and then added to the model again, then
in order to reestablish the coupling, you need to choose the physics
interface again from the lists. This is applicable to all multiphysics
coupling nodes that would normally default to the once present interface.
See Multiphysics Modeling Workflow in the COMSOL Multiphysics
Reference Manual.

There are several library examples that demonstrate its usage:

Loudspeaker Driver — Frequency-Domain Analysis: Application

Library path Acoustics_Module/Electroacoustic_Transducers/
loudspeaker_driver

Loudspeaker Driver — Transient Analysis: Application Library path

Acoustics_Module/Electroacoustic_Transducers/loudspeaker_driver_transient

Magnetic Damping of Vibrating Conducting Solids: Application

Library path ACDC_Module/Motors_and_Actuators/magnetic_damping

Magnetomechanical Forces
In models where the Lorentz Coupling is not applicable, for example, for
magnetizable armatures in transducers, the Magnetomechanical Forces multiphysics
coupling should be used. For details see the Magnetomechanical Forces and the The
Magnetomechanics Interface documentation of the Structural Mechanics User’s
Guide.

COUPLING FEATURES | 803

 Predefined Multiphysics Interfaces
The Acoustics Module includes several predefined multiphysics interfaces. When these
interfaces are selected from the Model Wizard or the Add Physics window both
contributing physics and the multiphysics couplings are added to the Model Builder. It
is always possible to build a model sequentially adding one physics at the time and then
coupling them. Note that far from all possible couplings are represented in the
predefined physics interfaces. All the couplings are listed in the Coupling Features
section.

When physics interfaces are added using the predefined multiphysics interfaces, for
example Thermoviscous Acoustic-Structure Interaction, the Selection of the coupling
feature, in this case the Thermoviscous Acoustic-Structure Boundary, is automatically set
to All boundaries. In this way the multiphysics coupling is automatically active on all
boundaries with Thermoviscous Acoustics on one side and Solid Mechanics on the other.

If physics interfaces are added one at a time, followed by the coupling features, the
selections are not automatically set. For example, if the single interfaces are added,
COMSOL Multiphysics adds an empty Multiphysics Couplings node. You can choose
the available coupling features, but you need manually to select on which boundaries
they need to be applied, or select All boundaries to recover the predefined behavior.

The Acoustics Module includes the following predefined multiphysics interfaces:

• The Acoustic-Solid Interaction, Frequency Domain Interface

• The Acoustic-Solid Interaction, Transient Interface
• The Acoustic-Piezoelectric Interaction, Frequency Domain Interface
• The Acoustic-Piezoelectric Interaction, Transient Interface
• The Acoustic-Solid-Poroelastic Waves Interaction Interface
• The Acoustic-Poroelastic Waves Interaction Interface
• The Acoustic-Solid Interaction, Time Explicit Interface
• The Acoustic-Shell Interaction, Frequency Domain Interface
• The Acoustic-Shell Interaction, Transient Interface
• The Acoustic-Thermoviscous Acoustic Interaction, Frequency Domain Interface
• The Thermoviscous Acoustic-Solid Interaction, Frequency Domain Interface
• The Thermoviscous Acoustic-Shell Interaction, Frequency Domain Interface

804 | CHAPTER 10: MULTIPHYSICS COUPLINGS

• The Piezoelectric Waves, Time Explicit Interface
Another option is to create the coupled predefined multiphysics interface by adding it
from the Add Multiphysics window after adding the separate physics interfaces.

PREDEFINED MULTIPHYSICS INTERFACES | 805

 Modeling with Multiphysics Couplings
Using the predefined multiphysics couplings has many advantages. It improves the
flexibility and enables sequential modeling, increasing the complexity of a model step
by step. Care is also taken to make the correct couplings both on the physics level but
also from a numerics point of view. It is of course possible to couple physics manually
and this is maybe also necessary in some cases when more advanced and exotic
couplings are necessary.

The predefined multiphysics coupling approach used improves the

flexibility and design options for your modeling. For for details, see The
Multiphysics Branch and Multiphysics Modeling Workflow in the
COMSOL Multiphysics Reference Manual.

In this section a few tips and tricks for modeling acoustics and vibration problems using
the multiphysics couplings are presented.

• Use Selections
• The Override Behavior
• The Solvers
• Perfectly Matched Layers (PMLs)

Use Selections
When modeling it is always recommended to use selections to predefine and give
names to several domains or boundaries. Define the selections under the Definitions
node. Setting up these selections will for multiphysics modeling avoid adding two
incompatible physics to the same domain. If acoustics and structures are coupled this
happens at boundaries, where two distinct physics should meet. For, for example,
piezoelectric devices the coupling between the electric and mechanical effect happen
on the domain level. Operations can also be performed on the selections if the
geometry is complex to find intersections, complement sets and so forth.

See Creating Named Selections in the COMSOL Multiphysics Reference

Manual.

806 | CHAPTER 10: MULTIPHYSICS COUPLINGS

The Override Behavior
When coupling acoustics, both Pressure Acoustics, Thermoviscous Acoustics, and
Aeroacoustics, to a structure the behavior of the multiphysics coupling is such that it
will override all boundary condition given on the acoustic side. Loads and constraints
defined on the common boundaries in the structural physics interface will contribute
with the coupling. In this way external forces that drive the system can be added here.
It is also possible to constrain the structure on these boundaries while it will still
experience the acoustic loads.

The Solvers
When solving a Multiphysics problem it is necessary to ensure that the solvers behave
correctly. It is especially important to check if a model is solved Fully Coupled or in a
Segregated way. COMSOL has built in logic to handle most multiphysics problems
when the predefined multiphysics couplings are used ensuring that the model solves
correctly. If many physics or couplings are added at the same time this logic can fail and
manual settings should be performed. For large models COMSOL generates Iterative
Solver Suggestions (see below).

See Introduction to Solvers and Studies and Solution Operation Nodes

and Solvers in the COMSOL Multiphysics Reference Manual.

It is also possible to enable and disable both physics and couplings from within a Study
Step. This can be done in the Physics and Variables Selection section in Settings. This
functionality can be used to solve a model both coupled and uncoupled by adding two
Studies : one where the Multiphysics coupling is active and one where it has been
deactivated.

ITERATIVE SOLVER SUGGESTIONS

Multiphysics models can in 3D easily grow rapidly in size and become difficult to solve
with a direct solver, simply because you run out of RAM. In this case it can be necessary
to switch to an iterative solver approach. For most multiphysics couplings iterative
solver suggestions are automatically generated. Set up a new solver configuration (first
time the model is solved) or right click the study node and select Show Default Solver,
then expand the Solver Configuration tree under Stationary Solver or Time-Dependent
Solver. The predefined iterative solver suggestions that are automatically generated can
be seen here. In the predefined cases a direct solver is used as default and iterative

MODELING WITH MULTIPHYSICS COUPLINGS | 807

 solvers are suggested and disabled. To turn on one of these approaches right-click the
solver and select Enable (or press F4). Predefined suggestions exist for:

• Acoustic-Structure Boundary, gives one iterative solver suggestion. The suggestion

depends on which physics are coupled with the multiphysics coupling. Different
suggestions are given when coupling acoustics to shell/membrane and solid. The
suggestion will also be different depending on if the acoustics interface is based on
the FEM or BEM formulation.
• Thermoviscous Acoustic-Structure Boundary, generates two suggestions. The first is
fast and can save 20% of memory while the second that uses domain decomposition
(DD) is very robust and memory lean, but it can be slow.
• Combining Acoustic-Structure Boundary and Piezoelectric Effect yields a suggestion
that can be used for modeling large piezoelectric transducer assemblies.
• The Acoustic BEM-FEM Boundary generates a dedicated hybrid iterative solver to solve
the combined FEM and BEM problems.

All solver suggestions have a tag that makes it possible to see which physics feature that
has generated them, for example, (asb1) means that the suggestion comes from the
Acoustic-Structure Boundary 1 feature.

Some tutorial models from the library that use a solver suggestions:

• Electroacoustic_Transducers/vented_loudspeaker_enclosure
• Tutorials,_Thermoviscous_Acoustics/transfer_impedance_perforate
• Vibrations_and_FSI/vibrating_micromirror

TIME STEPPING IN TRANSIENT MODELS

When a multiphysics model is solves in the time domain it is important to update the
Transient Solver Settings section of the acoustics interface used in the model. The time
stepping settings generated by these settings will also be applied to the coupled
problem when a Multiphysics couplings is used.

Perfectly Matched Layers (PMLs)

Care should be taken when setting up perfectly matched layers (PMLs) in multiphysics
models where different waves propagate. A discussion about the Acoustic-Structure
interaction situation is given in Configuration of Perfectly Matched Layers (PMLs) for
Acoustic-Structure Interaction Models in Modeling with the Acoustic-Structure

808 | CHAPTER 10: MULTIPHYSICS COUPLINGS

Interaction Branch. The approach described is valid in general for models where
several sound speeds are present.

MODELING WITH MULTIPHYSICS COUPLINGS | 809

810 | CHAPTER 10: MULTIPHYSICS COUPLINGS
 11

Structural Mechanics with the Acoustics

Module

Elastic waves and vibrations are an important part of acoustics. The Acoustics
Module is therefore equipped with an extended version of the base package Solid
Mechanics interface shipped with COMSOL Multiphysics. The interface contain
less features when compared to the full Solid Mechanics interface available with the
Structural Mechanics Module. This chapter describes the features and applications
of the Solid Mechanics and Piezoelectricity interfaces available with the Acoustics
Module. The physics interfaces is found under the Structural Mechanics branch
( ).

For a detailed overview of the functionality available in each product, visit https:/
/[Link]/products/specifications/

The full documentation of all features in the full Solid Mechanics interface and the
Piezoelectricity interface is found in the Structural Mechanics Module User’s
Guide, specifically in these sections:

• Solid Mechanics
• Structural Mechanics Modeling

811
 • Structural Mechanics Theory
• Multiphysics Interfaces and Couplings and specifically The Piezoelectricity Interface

In this chapter:

• Vibroacoustic Applications
• The Solid Mechanics Interface
• The Piezoelectricity Interface
• Acoustic-Structure Multiphysics Interaction

Vibroacoustic Applications
The Solid Mechanics and Piezoelectricity interfaces are available with the Acoustics
Module such that vibroacoustic problems involving solids, piezoelectric materials, and
acoustic domains can be modeled. These multiphysics applications include piezo
transducers, loudspeaker drivers, loudspeaker cabinets, automotive applications, and
sound transmission/isolation problems. Pure elastic wave problems can also be
modeled as the Solid Mechanics equations are formulated include the full dynamic
behavior.

Additional multiphysics capabilities of the Acoustics Module are enabled when it is

combined with the Structural Mechanics Module. The Shell interface, which is part of
the Structural Mechanics Module, enables predefined acoustic-shell couplings. The
Acoustic-Shell and Thermoacoustic-Shell interaction multiphysics interfaces are
enabled. The option to couple to shells, membranes and multibody dynamics using the
Acoustic-Structure Boundary is also enabled, see the Multiphysics Couplings chapter
for further details.

The Solid Mechanics Interface

Included in the Acoustics Module is an extended version of the Solid Mechanics
interface shipped with COMSOL Multiphysics. The interface contain less features
when compared to the full Solid Mechanics interface available with the Structural
Mechanics Module.

The Piezoelectricity Interface

The Piezoelectricity interface is also included in the Acoustics Module. This is a
Multiphysics interface that couples the Electrostatics interface with the Solid

812 | CHAPTER 11: STRUCTURAL MECHANICS WITH THE ACOUSTICS MODULE

Mechanics interface. See Modeling Piezoelectric Problems section in the Structural
Mechanics Module User’s Guide or further details.

Acoustic-Structure Multiphysics Interaction

As soon as an acoustics interface and a structural mechanics interface are present in the
same model the Multiphysics node appear, if there exist a predefined coupling between
the two interfaces. Under the Multiphysics node you can add the predefined
multiphysics couplings that couple the fluid domain and to the solid domain. The
couplings are described in detail in the Multiphysics Couplings chapter.

You can also find a set of predefined multiphysics interface that will automatically sets
up the involved single physics and the multiphysics coupling. These are described in
the Acoustic-Structure Interaction Interfaces chapter. Here you will also find modeling
tips for acoustic-structure interaction problems in the Modeling with the
Acoustic-Structure Interaction Branch section.

| 813
814 | CHAPTER 11: STRUCTURAL MECHANICS WITH THE ACOUSTICS MODULE
 12

Study Types

The Acoustics Module contains a number of study types suited for acoustics
simulations.

This chapter provides information about those study types — including stationary,
frequency-domain, and time-dependent study types — and how you can use them
for different types of acoustics simulations.

In this chapter:

• Acoustics Module Study Types

815
 Acoustics Module Study Types
The Acoustics Module is designed for both frequency-domain simulations and
transient analysis. The frequency domain simulations include the related eigenvalue
and mode analysis problems. The Pressure Acoustics interfaces support most study
types with certain exceptions; the Pressure Acoustics, Boundary Element interface only
supports frequency domain and the Pressure Acoustics, Time Explicit only the time
domain analysis. The Thermoviscous Acoustics and the linearized Aeroacoustic (or
convected acoustics) interfaces all support both transient and frequency-domain
analysis. The Compressible Potential Flow interface is tailored to model a stationary
background flow to be used in a subsequent time-harmonic aeroacoustics simulation.
In the Solid Mechanics interface, the static analysis type is also included and can be
used to model the stationary state of prestressed systems subject to time-harmonic
vibrations.

The analysis types require different solvers and equations. The following study types,
briefly discussed in this section, can help you find good candidates for the application:

• Stationary Study • Frequency Domain, Modal and

• Frequency Domain Study Time-Dependent, Modal Studies

• Eigenfrequency Study • Ray Tracing

• Mode Analysis Study • Modal Reduced Order Model

• Boundary Mode Analysis • Additional Analysis Capabilities

• Time Dependent Study • Mapping

• Transient Mapping

Studies and Solvers and Harmonic Perturbation, Prestressed Analysis, and

Small-Signal Analysis in the COMSOL Multiphysics Reference Manual.

Prestressed Acoustic-Structure Interaction type of problems are also

sported using a combination of a stationary solver and a frequency
domain perturbation solver.

816 | CHAPTER 12: STUDY TYPES

Stationary Study
A stationary analysis solves for stationary displacements or a steady-state condition. All
loads and constraints are constant. For a stationary analysis, use a Stationary study
type ( ). For all pure acoustic and vibration problems, this type of analysis yields the
zero solution as, by definition, these represent and describe propagating varying fields
— either time dependent or time harmonic in the frequency domain.

Stationary in the COMSOL Multiphysics Reference Manual

Frequency Domain Study

Wave propagation is modeled by equations from linearized fluid dynamics (pressure
waves) and structural dynamics (elastic waves). The full equations are time dependent,
but noting that a harmonic excitation of the field p has a time dependence of the form

p ( t ) = pe iωt

gives rise to an equally harmonic response with the same frequency; the time can be
eliminated completely from the equations. Instead the angular frequency ω = 2πf,
enters as a parameter where f is the frequency. In COMSOL, the +iω convection is
used for the time harmonic formulations.

This procedure is often referred to as working in the frequency domain or Fourier

domain as opposed to the time domain. From the mathematical point of view, the
time-harmonic equation is a Fourier transform of the original time-dependent
equations and its solution as function of ω is the Fourier transform of a full transient
solution. It is therefore possible to synthesize a time-dependent solution from a
frequency-domain simulation by applying an inverse Fourier transform.

For the interfaces in the Acoustics Module that are based on the finite element method
(FEM) or the boundary element method (BEM), a frequency domain simulation is
well suited. Therefore, select the Frequency Domain study type ( ) over a Time
Dependent study whenever possible. Certain important software features, like
damping due to porous materials or advanced impedance boundary conditions, are
only present when using the frequency domain physics interfaces.

ACOUSTICS MODULE STUDY TYPES | 817

 The result of a frequency domain analysis is a complex time-dependent field p, which
can be interpreted as an amplitude pamp = abs(p) and a phase angle pphase = arg(p).
The actual pressure at any point in time is the real part of the solution

p ( t ) = p amp cos ( 2πf ⋅ t + p phase )

Visualize the amplitudes and phases as well as the solution at a specific angle (time).
When using the Solution datasets, the solution at angle (phase) parameter makes this
task easy. When plotting the solution, COMSOL Multiphysics multiplies it by eiϕ,
where ϕ is the angle in radians that corresponds to the angle (specified in degrees) in
the Solution at angle field. The plot shows the real part of the evaluated expression:

p = p amp cos ( ϕ + p phase )

The angle ϕ is available as the variable phase (in radians) and is allowed in plot
expressions. Both the frequency freq and angular frequency omega are available
variables.

In a frequency domain study, almost everything is treated as harmonic —

prescribed pressures, displacements, velocities, and accelerations — not
only the forces and dependent fields. Notable exceptions are certain
quantities, such as the sound pressure level or intensity, which by
definition are time averages.

Frequency Domain and Solution (datasets) in the COMSOL Multiphysics

Reference Manual

Eigenfrequency Study
If all sources are removed from a frequency-domain equation, its solution becomes
zero for all but a discrete set of angular frequencies ω, where the solution has a
well-defined shape but undefined magnitude. These solutions are known as
eigenmodes and their corresponding frequencies as eigenfrequencies.

The eigenmodes and eigenfrequencies have many interesting mathematical properties,

but also direct physical significance because they identify the resonance frequency (or
frequencies) of the structure. When approaching a resonance frequency in a
harmonically-driven problem, a weaker and weaker source is needed to maintain a

818 | CHAPTER 12: STUDY TYPES

given response level. At the actual eigenfrequency, the time-harmonic problem loses
the uniqueness of the solution for a nonzero excitation.

Select the Eigenfrequency study type ( ) when you are interested in the resonance
frequencies of the acoustic domain or the structure, whether you want to exploit them,
as in a musical instrument, or avoid them, as in a reactive muffler or inside a hi-fi
speaker system. To an engineer, the distribution of eigenfrequencies and the shape of
eigenmodes can also give a good first impression about the behavior of a system.

An eigenfrequency analysis solves for the eigenfrequencies and the shape of the
eigenmodes. When performing an eigenfrequency analysis, specify whether to look at
the mathematically more fundamental eigenvalue λ (available as the variable lambda)
or the eigenfrequency f, which is more commonly used in an acoustics context:

–λ
f = ---------
2πi

NONLINEAR EIGENFREQUENCY ANALYSIS

In certain circumstances the material properties or boundary conditions can be
frequency dependent. This is, for example, the case if a model set up with The Pressure
Acoustics, Frequency Domain Interface contains a Narrow Region Acoustics or a
Poroacoustics equivalent fluid model. The same is true if the Impedance boundary
condition is added: most of the options listed in Theory for the Boundary Impedance
Models define the acoustic impedance as a function of frequency. If the frequency
dependence in these models is nonlinear, this will lead to a so-called nonlinear
eigenvalue problem, which is more complex and therefore often must be treated
carefully.

While solving a nonlinear eigenvalue problem, the linearization around some

linearization point takes place behind the scenes. Expand the Study node, right-click
the Solver Configurations node, and select Show Default Solver. In the generated solver
sequence, navigate to the Eigenvalue Solver node and find the Transform point field. The
value specified there will be used as the linearization point in the eigenvalue study, and
therefore it is very important to set up for nonlinear eigenvalue problems.

The default Transform point value used in the acoustics interfaces is 100 Hz. If the
eigenfrequencies of interest lie in kHz or MHz range, the default value of the
linearization point may not ensure an accurate solution of a nonlinear eigenvalue
problem — depending on its nonlinear properties. In this case, it is good practice to
move the linearization point closer to the range of desired eigenfrequencies. For
example, it can be the mean of the range. The other option is to define the Transform

ACOUSTICS MODULE STUDY TYPES | 819

 point as a parameter, for example TP, add a Parametric Sweep over this parameter to the
study, and search for one eigenfrequency around TP. This approach will be more
accurate if the model exhibits strong nonlinear behavior within the range of desired
eigenfrequencies, where a constant linearization point in not enough.

Eigenfrequency in the COMSOL Multiphysics Reference Manual

Mode Analysis Study

The Mode Analysis study ( ) is available with The Pressure Acoustics, Frequency
Domain Interface, The Linearized Potential Flow, Frequency Domain Interface, and
The Thermoviscous Acoustics, Frequency Domain Interface in plane 2D and axially
symmetric 1D acoustics interfaces.

The Pressure Acoustics, Boundary Mode Interface, The Thermoviscous Acoustics,

Boundary Mode Interface, and The Linearized Potential Flow, Boundary Mode
Interface are special interfaces for more advanced Mode Analysis studies on boundaries
in 3D and 2D axisymmetry. Acoustic waves can propagate over large distances in ducts
and pipes, with a generic name referred to as waveguides. After some distance of
propagation in a waveguide of uniform cross section, such guided waves can be
described as a sum of just a few discrete propagating modes, each with its own shape
and phase speed. The equation governing these modes can be obtained as a spatial
Fourier transform of the time-harmonic equation in the waveguide axial direction
(here assuming propagation in the z direction) or by inserting the assumption that the
mode is harmonic in space,

p ( x ) = p ( x, y )e – ikz z

and eliminating all out-of-plane z dependence. Here p(x,y) is the in-plane mode shape.

The axial wave number kz is a parameter in the 2D acoustics

interfaces.

Similar to the full time-harmonic equation, the transformed equation can be solved at
a given frequency with a nonzero excitation for most axial wave numbers kz. But at
certain discrete values the equation breaks down. These values are the propagation
constants or wave numbers of the propagating or evanescent waveguide modes. The

820 | CHAPTER 12: STUDY TYPES

eigenvalue solver can solve for these propagation constants together with the
corresponding mode shapes.

The propagating wave number is a function of the frequency. The relation

between the two is commonly referred to as a dispersion curve.

The most common use for the Mode Analysis is to define sources for a subsequent
time-harmonic simulation. If there is a component with one or more waveguide
connections, its behavior can be described by simulating its response to the discrete set
of propagating modes on the waveguide port cross sections. The Port boundary
condition in Pressure Acoustics uses this concept to define sources and absorb
outgoing modes. In thermoviscous acoustics, a Mode Analysis study also provides
information about the absorption coefficient for the propagating modes. The complex
wave number (solved for) and characteristic impedance (defined in postprocessing)
define the homogenized propagation variables that can be used in Narrow Region
Acoustics.

Mode Analysis in the COMSOL Multiphysics Reference Manual

Jet Pipe: Application Library path Acoustics_Module/

Aeroacoustics_and_Noise/jet_pipe

Boundary Mode Analysis

The Boundary Mode Analysis study ( ) is used when solving for the propagating
modes used in a Port boundary condition in The Pressure Acoustics, Frequency
Domain Interface. The study has special settings that can match the shape and wave
number of the mode to a given port condition.

Time Dependent Study

The complete equations behind the theory of acoustic wave propagation are time
dependent, as discussed in the Frequency Domain Study section. Solving time-domain
equations is more complicated from a numerical point of view. They are necessary
when studying systems where the time of flight is of importance, like in a flowmeter

ACOUSTICS MODULE STUDY TYPES | 821

 application. Short-term transient processes like step and impulse responses can benefit
from modeling in the time domain, if not for efficiency so for convenience.

A dedicated Transient Solver Settings settings section exist at the physics level for all the
transient acoustics interfaces where the Maximum frequency to resolve should be
entered.

Some central modeling techniques, such as the use of PMLs, are not
available for all physics with the Time Dependent study type. PMLs only
exist in the time domain for the Pressure Acoustics, Time Domain
interface. Be careful when defining your sources to avoid, as far as
possible, exciting waves at frequencies that the mesh cannot resolve.

Time Dependent in the COMSOL Multiphysics Reference Manual.

Further details on Time Stepping: Resolving the Waves in Time in the

Modeling with the Pressure Acoustics Branch (FEM-Based Interfaces)
section.

Frequency Domain, Modal and Time-Dependent, Modal Studies

The Frequency Domain, Modal study type ( ) is used to do modal analysis in the
frequency domain and the Time-Dependent, Modal study ( ) is used to do
time-dependent modal analysis.

Frequency Domain, Modal and Time Dependent, Modal in the

COMSOL Multiphysics Reference Manual

Helmholtz Resonator Analyzed with Different Frequency Domain

Solvers: Application Library path Acoustics_Module/
Tutorials,_Pressure_Acoustics/helmholtz_resonator_solvers

Ray Tracing
The Ray Tracing study type ( ) is used when solving acoustic problems using The
Ray Acoustics Interface. It is a special case of the Time Dependent study and is used
to compute the trajectories of rays. Special built-in stop conditions can be used to stop

822 | CHAPTER 12: STUDY TYPES

the solver when no active rays remain or the intensity of the rays is below a certain
criterion.

Modal Reduced Order Model

The Model Reduction and Modal Reduced Order Model study steps have the property that
they can assemble the modal matrices and make them available for output. In the Model
Reduction node, the Store reduced matrices check box must be selected.

After solving the model, right-click the Derived Values node (under Results) and select
System Matrix. In the output section, select the Matrix to display and the Format. Using
the Matrix settings it is possible to access the stiffness, damping, and mass matrices of
the system, for example.

In the COMSOL Multiphysics Reference Manual:

• Modal Reduced-Order Model and Introduction to Solvers and Studies

• System Matrix and Results Analysis and Plots

Additional Analysis Capabilities

In a multiphysics interface you might want to use different analysis types for the
different dependent variables. This can be done by adding an Empty Study ( ), and
then adding different study steps to this study. Also perform parametric analyses by
using the Parametric Sweep study node ( ). Typical parameters to vary include
geometric properties, the frequency band in ray tracing, and the out-of-plane or axial
wave number.

Parametric Sweep in the COMSOL Multiphysics Reference Manual

Mapping
The Mapping ( ) study is a dedicated study to use when the Background Fluid Flow
Coupling couples a Fluid Flow interface and an Aeroacoustics interface. The study
should only solve for the multiphysics coupling. The study uses a Stationary solver.
Together with the multiphysics coupling, the study handles the mapping of dependent

ACOUSTICS MODULE STUDY TYPES | 823

 variables between the fluid flow mesh and the acoustics mesh for consistent simulation
results.

PHYSICS AND VARIABLES SELECTION

Only select the Background Fluid Flow Coupling as being Solved for. No other physics
will be solved by the mapping study.

SOLUTION TO MAP
From the Study list select the study containing the coupled fluid flow interface to map.
Depending on the type of study, used for the fluid flow simulation, you will get a
different type of list to choose from: time steps or parameter values. From the list,
select the appropriate solution or solutions to map. The option All will automatically
add a parameter sweep and map all parameter values from the selected solution. The
Mapping study step will automatically ensure that those parameter values are used.

Note that the mapping does not benefit from running continuation for any parameter.
When the solution is mapped for all parameter values, for a multiple parameter sweep
where continuation was used in the fluid flow study, the resulting dataset structure of
the two studies will therefore be slightly different.

DESTINATION MESH SELECTION

Select the destination/target mesh of the mapping, that is, the acoustics mesh used for
the aeroacoustics simulation.

Transient Mapping
The Transient Mapping ( ) study is a dedicated study to use when the Aeroacoustic
Flow Source Coupling couples a LES Fluid Flow interface with the Aeroacoustic Flow
Source in pressure acoustics. The study should only solve for the multiphysics
coupling. The study uses a tailored Transient solver. Together with the multiphysics
coupling, the study handles the mapping of flow source variables (the Lighthill stress
tensor) between the fluid flow mesh and the acoustics mesh for consistent simulation
results.

SOLUTION TO MAP
From the Study list select the study containing the coupled fluid flow interface to map.

824 | CHAPTER 12: STUDY TYPES

DESTINATION MESH SELECTION
Select the destination/target mesh of the mapping, that is, the acoustics mesh used for
the aeroacoustics simulation.

ACOUSTICS MODULE STUDY TYPES | 825

826 | CHAPTER 12: STUDY TYPES
 13

Acoustic Properties of Fluids

The Acoustics Module contains a number of models describing attenuation

processes and loss mechanisms, and they have various inputs for the acoustic
relevant material properties of fluids. This chapter provides information about
those properties and puts them all in context.

In this chapter:

• Material Properties
• The Bulk Viscosity
• Attenuation and Loss Models
• References for The Acoustic Properties of Fluids

827
 Material Properties
The Acoustics Module includes two material databases: Liquids and Gases, with
temperature-dependent fluid dynamic and thermal properties, and a Piezoelectric
Materials database with common piezoelectric materials.

Probably, the two most common materials used in acoustics are air and water. In this
section we shortly discuss the definition and properties of the built-in Air and Water,
liquid materials from the Material Library.

The built-in air and water materials are general and to some extent simplified models.
For example, the air represents dry air and the material properties do not depend on
moisture. This means that for high-precision absolute-value simulations, detailed
material data need to be entered. Two options exist:

1 Using the Predefined System for Moist air, Dry air, or Water-steam defined in the
Thermodynamics feature available with the Liquid & Gas Properties Module. Once
set up, a material can be automatically generated.
2 Create your own material, for example, following the IEC standard for moist air (see
Ref. 13). Set up a material and let it depend on the necessary model inputs
(pressure, temperature, or relative humidity). Then, set up the necessary
interpolation functions or analytical expressions and store the material under the
User-Defined Library for future use.

Moist air is generated from Thermodynamics in the Pressure Reciprocity

Calibration Coupler with Detailed Moist Air Material Properties
tutorial. Application Library path
Acoustics_Module/Tutorials,_Thermoviscous_Acoustics/
pressure_reciprocity_calibration_coupler

The Air and Water, liquid materials were updated for the COMSOL
version 5.4 release. Some material data has been added to simplify
modeling, and inconsistencies due to rounding have been removed.

To take advantage of the new and updated material data, when opening
an existing model (older than version 5.4), delete the material and add it
again from the Materials window.

828 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

 For detailed information about Materials, the Liquids and Gases Materials
Library, and the Piezoelectric Materials Library see the COMSOL
Multiphysics Reference Manual.

For detailed information about the Predefined system see the

Thermodynamics chapter in the Liquid & Gas Properties User’s Guide.

THE (DRY) AIR MATERIAL

The built-in Air material, located both in the Built-In and the Liquids and Gases
library, is commonly used in applications and models. The material properties
represent those of dry air without the inclusion of moisture.

The air material defines common material parameters and their dependency on the
ambient pressure pA and temperature T. The relations are simplified and not all
thermodynamic dependencies are taken into account. The speed of sound c and the
density ρ are defined through the ideal gas law (assuming adiabatic behavior) following

2 R Mn pA
c = γ -------- T ρ = ---------------
Mn RT

with the ratio of specific heats γ = 1.4, the gas constant R (COMSOL has a built-in
constant called R_const), and the molar mass Mn = 0.02897 kg/mol. This is an
idealization of air valid in many cases; see Ref. 2. This means that the speed of sound
is not a function of the ambient pressure for this built-in Air material.

Some material properties are only temperature dependent and are given by polynomial
fit curves like:

• The dynamic viscosity μ = μ(T)

• The (specific) heat capacity at constant pressure Cp = Cp(T)
• The thermal conductivity k = k(T)

Some properties are defined as constant:

• The ratio of specific heats γ = 1.4

• The molar mass Mn = 0.02897 kg/mol

Furthermore, there are material parameters and properties that are derived from their
definitions and the above material data:

MATERIAL PROPERTIES | 829

 • The bulk viscosity μB is defined as μB = 0.6·μ(T) in order to comply with the
absorption behavior (see discussion in The Bulk Viscosity section).
• The (isobaric) coefficient of thermal expansion is defined as αp = 1/T, following the
ideal gas law.
• The parameter of nonlinearity B/A is also defined for gases following the ideal gas
law as B/A = (γ+1)/2 (see Ref. 12).

THE (CLEAN) WATER MATERIAL

The built-in Water, liquid material, located both in the Built-In and the Liquids and
Gases library, is also commonly used in applications and models. These material
properties represent those of clean water without the dependency on, for example,
salinity or pH value necessary to describe salt water in the ocean.

Specifically for acoustics applications, the water material properties include properties
that are temperature dependent like:

• The speed of sound c = c(T)

• The density ρ = ρ(T)
• The dynamic viscosity μ = μ(T)
• Thermal conductivity k = k(T)
• The heat capacity at constant pressure Cp = Cp(T)

Furthermore, there are detailed material parameters that are derived from the above
definitions:

• The bulk viscosity is defined as μB = 2.79·μ(T). In order to comply with the

absorption behavior, (see the discussion in The Bulk Viscosity section).
• The (isobaric) coefficient of thermal expansion is defined through the
thermodynamic definition as

1 ∂ρ
α p = – --- -------
ρ ∂T pA

• The ratio of specific heats γ = γ(Τ) is based on the thermodynamic relation

830 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

 2 2
Tα p c
γ = 1 + ---------------- .
Cp

This definition is necessary in detailed models solving the thermoviscous acoustics

or the linearized Navier-Stokes equations in water. The definition ensures that it is
not just set to γ = 1.

MATERIAL PROPERTIES | 831

 T he Bulk V i s c osi t y
The bulk viscosity μB of a fluid is also known as the volume viscosity or the second
viscosity. It is often not used in classical CFD applications but it plays a role in acoustics
as well as in high-velocity compressible flows. The bulk viscosity is associated with
losses due to expansion and compression, that is, the compressional resistance. The
parameter describes the difference between the mechanical and thermodynamic
pressures. It is often described as representing all the effects that describe the time it
takes to establish thermodynamic equilibrium for vibrational and rotational motion of
the fluid molecules (see Ref. 18). In some sense, it can be said to describe low
frequency relaxation processes.

In acoustics, the losses associated with the bulk viscosity only really play a role in
high-frequency ultrasound propagation. The bulk viscosity term in the thermoviscous
acoustics or the linearized Navier-Stokes equations is sometimes used to model the
frequency dependent losses due to relaxation processes. Various models exist to
include these effects; details can be found in Ref. 17.

The Value of the Bulk Viscosity

The value of the bulk viscosity is difficult to measure and typically requires absorption
experiments to be determined. For most fluids, its numerical value is of the same order
as the dynamic viscosity. For some further details see, for example, Ref. 14 for fluids,
Ref. 16 for water values, and Ref. 15 for gases.

The bulk viscosity enters the classical (pure) thermoviscous attenuation coefficient
defined as

2
ω k(γ – 1)
α tv = ------------3  --- μ + μ B + --------------------
4
3 Cp 
2ρc

This is the attenuation experienced by plane propagating waves away from boundaries.
The expression is derived directly from the full thermoviscous acoustics equations. In
the Air and Water, liquid materials in the Material Library, the bulk viscosity is defined
to match the high-frequency absorption properties of the fluids.

In Figure 13-1 standardized atmosphere attenuation curves (see the Atmosphere

Attenuation Fluid Model in pressure acoustics) are depicted for various relative
humidities together with the classical thermoviscous attention. The graph also shows

832 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

the case for zero bulk viscosity. The bulk viscosity is defined such that the
thermoviscous absorption curve matches the standard at high frequencies, this
happens for μB = 0.6·μ. This corresponds to how the value is determined in an
experimental setup.

The same graphs are depicted in Figure 13-2 for water using the Ocean Attenuation
Fluid Model, here for various values of the practical salinity (with a pH value of 7 and
evaluated at the sea surface). The attenuation match gives a value for the bulk viscosity
of μB = 2.79·μ. This is also the value reported in literature (see Ref. 14 and 16).

Figure 13-1: Atmosphere attenuation for various relative humidities (phi), the full
classical thermoviscous attenuation, and thermoviscous attenuation for zero bulk viscosity.

THE BULK VISCOSITY | 833

 Figure 13-2: Ocean attenuation for various practical salinities (S·1000), the full classical
thermoviscous attenuation, and thermoviscous attenuation for zero bulk viscosity.

834 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

At t e nua ti o n a n d Loss Mod el s
The Acoustics Module includes several different physics interfaces and various features
that introduce acoustic losses. These describe everything from thermoviscous
boundary layer losses to the attenuation experienced when acoustic waves propagate
over long distances in the ocean. The models have different applicability depending on
the frequency regime and the system modeled.

This section first gives a short description of the different loss mechanism, then the
applicability and the regime where boundary layer and bulk losses apply is discussed,
and finally some remarks are given on the bulk viscosity and how it is related to
attenuation.

In this section:

• Loss Mechanisms
• The Atmosphere and Ocean Attenuation Models
• Boundary Layer vs. Bulk Losses

Loss Mechanisms
Loss mechanisms in acoustics due to viscosity, thermal conduction, relaxation
processes, and other processes cause absorption and dissipation of the acoustic energy.
These result in a reduction in the pressure wave amplitude. This is not due to
geometrical spreading where no energy is lost, but because heating actually takes place.
When an acoustic wave undergoes absorption, this process is in general also
accompanied by dispersion, that is, the dependence of the speed of sound on
frequency.

Loss mechanism in acoustics can roughly be divided into four categories, but can of
course also happen simultaneously. The division is mostly conceptual:

1 Bulk or volume losses are associated with the propagation of waves over long
distances or at very high frequencies (also known as internal damping). The (plane
wave) attenuation coefficient α (SI unit: 1/m) is often associated with this
mechanism. Several loss models are included in the Pressure Acoustics model (see
also Theory for the Equivalent Fluid Models) including: User-Defined Attenuation
Fluid Model, Atmosphere Attenuation Fluid Model, Ocean Attenuation Fluid
Model, or Thermally Conducting and/or Viscous Fluid Model. Note that the

ATTENUATION AND LOSS MODELS | 835

 Thermally Conducting and/or Viscous Fluid Model should not be confused with
boundary layer losses (see next point). Bulk losses are due to several different
mechanisms including viscous and thermal dissipation, relaxation processes, and
other loss mechanism.
2 Viscous and thermal boundary-layer losses occur at hard walls because of the
effective no-slip and isothermal conditions. These are most important in geometries
of small dimensions comparable to the boundary layer thickness. The losses can be
included in a very general manner using one of the Thermoviscous Acoustics
Interfaces. The Narrow Region Acoustics feature can be used in narrow waveguides
of constant cross section, while the Thermoviscous Boundary Layer Impedance
condition can be used an any boundary as long as no boundary layers are
overlapping. The latter two features are great engineering approximations (often
yielding exact results) to include the losses at a lower computational cost using the
The Pressure Acoustics, Frequency Domain Interface.
3 In porous materials, losses are again due to viscous and thermal boundary layer
losses, here inside the channels of the porous matrix. The losses are also caused by
damping because of the coupling to the porous matrix structure. Detailed modeling
is done using The Poroelastic Waves Interface that solves the full Biot model for the
coupled pressure and structural waves. A simplified so-called equivalent fluid
description can be done using the Poroacoustics feature (see also Theory for the
Equivalent Fluid Models).
• Losses can also occur due to interaction with the surroundings like solids and
membranes. This is best modeled using multiphysics, see Acoustic-Structure
Boundary in the Multiphysics Couplings chapter. Here, simplified models exist
using the many options available with the Impedance boundary conditions (see
Theory for the Boundary Impedance Models).

The Atmosphere and Ocean Attenuation Models

The Pressure Acoustics, Frequency Domain Interface and The Ray Acoustics Interface
include two semi analytical fluid models that are calibrated with extensive
measurement data. This is the Atmospheric attenuation and the Ocean attenuation
models. Both include effects due to viscosity and thermal conduction and relaxation
processes of molecules.

The Atmospheric attenuation model defines attenuation in atmospheric air that follows
the ANSI standard S1.26-2014, see Ref. 3 (Appendix B), 4, 5, and 6 for details. The
model describes attenuation due to thermal and viscous effects (primarily pure air), the

836 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

relaxation processes for nitrogen and oxygen, and the dependency on atmospheric
pressure (absolute pressure), temperature, and relative humidity. The attenuation in air
is important for propagation over large distances and for high frequency processes.
This also means that the attenuation effect is more important in ray tracing simulations
where propagation can be simulated over much larger distances and especially at higher
frequencies. In the next section, we will see when the full details of this model are
necessary when compared to boundary layer losses.

The Ocean attenuation model defines attenuation in seawater of the ocean. The model
is based on a semianalytical model with parameters that are based on extensive
measurement data. No standard exists for the attenuation in sea water. It includes
effects due to viscosity in pure water, the relaxation processes of boric acid and
magnesium sulfate, as well as depth, temperature, practical salinity, and pH value. For
further details see Ref. 7, 8, 9, 10, and 11. Like the atmosphere model, the ocean
attenuation model is important in ray tracing simulations where propagation can be
simulated over much larger distances.

Boundary Layer vs. Bulk Losses

As discussed in the above sections, the Acoustics Module includes models for both
bulk losses and boundary layer losses. In pressure acoustics, you can include either
form of losses. For example, the Atmosphere attenuation fluid model in pressure
acoustics is a detailed bulk loss model that includes all dissipation effects and it
complies with the ANSI standard. On the other hand, the Narrow Region Acoustics
feature includes the boundary layer losses for narrow waveguide structures.

The importance of the two loss mechanisms (bulk and boundary layer) depend on the
geometry scale and the frequency range studied. As an example, the losses experienced
by a plane wave propagating in an air filled narrow circular duct is depicted in
Figure 13-3. The graph includes:

• The bulk atmosphere attenuation for different relative humidities ( φ = 0 % , 20 %

and 80 %).
• The classical (bulk) thermoviscous attenuation.
• The thermoviscous boundary layer attenuation for different duct radii (a = 0.5 mm,
2.0 mm, and 10.0 mm) modeled using the Narrow Region Acoustic feature.
• The solution using the full thermoviscous acoustics equations in the cylindrical duct
of radius a = 2.0 mm. This can be seen as a reference solution since it models both
boundary layer and thermoviscous bulk losses.

ATTENUATION AND LOSS MODELS | 837

 From the graph, it is evident that in relatively narrow waveguide structures the
boundary layer losses far surpass the bulk losses in a large frequency range. The
transition happens in the ultrasound regime. The graph also shows that (for air) the
real life relaxation effects captured by the atmosphere attenuation model can be
disregarded in most waveguides and really only are important away from boundaries
and for propagation over large distances.

The transition between the two loss regimes is captured by the full thermoviscous
acoustics models. For the vast majority of practical applications, when using the
Thermoviscous Acoustics physics interfaces, it is not necessary to include the “true”
bulk behavior captured by the atmosphere model. This is simply because it is not
computationally possible to model large scale acoustics with the Thermoviscous
Acoustics interfaces - their true application is in complex narrow geometries.

Figure 13-3: The importance of various attenuation mechanisms for the propagation of
plane waves in a cylindrical duct of radius a. The plot shows the (bulk) atmosphere
attenuation contribution for different relative humidities φ, the classical (bulk)
thermoviscous attenuation, the thermoviscous boundary layer attenuation for different
duct radii a, and reference solution using the full thermoviscous acoustics model (in the
duct with a radius of 2.0 mm).

838 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

References for The Acoustic
Properties of Fluids
1. S. Temkin, Elements of Acoustics, Acoustical Society of America, 2001.

2. A.D. Pierce, Acoustics: An Introduction to its Physical Principles and

Applications, Acoustical Society of America (second print), 1991.

3. D.T. Blackstock, Fundamentals of Physical Acoustics, John Wiley & Sons, 2000.

4. . E. Bass, L. C. Sutherland, A. J. Zuckerwar, D. T. Blackstock, and D. M. Hester,

“Atmospheric absorption of sound: Further developments,” J. Acoust. Soc. Am., vol.
97, pp. 680-683, 1995; “Erratum” J. Acoust. Soc. Am., vol. 99, p. 1259. 1996.

5. ANSI S1.26-2014 (supersedes ANSI S1.26-1995) “American National Standard

method for calculation of the absorption of sound by the atmosphere” (Acoustical
Society of America, New York, 2014).

6. National Physics Lab (NPL), “NPL Acoustics: Calculation of absorption of sound

by the atmosphere,” [Link]

7. R. E. Francois and G. R. Garrison, “Sound absorption based on ocean

measurements: Part I: Pure water and magnesium sulfate contributions,” J. Acoust.
Soc. Am., vol. 72, pp. 896-907, 1982.

8. R. E. Francois and G. R. Garrison, “Sound absorption based on ocean

measurements: Part II: Boric acid contribution and equation for total absorption,” J.
Acoust. Soc. Am., vol. 72, pp. 1879-1890, 1982.

9. M. A Ainslie and J. G. McColm, “A simplified formula for viscous and chemical

absorption in sea water,” J. Acoust. Soc. Am., vol. 103, pp. 1671-1672, 1998.

10. Fisher F. H., Simmons V. P., “Sound absorption in seawater,” J. Acoust. Soc. Am.,
vol. 62, pp. 558-564, 1977.

11. National Physics Lab (NPL), “Calculation of absorption of sound in seawater,”

[Link]

12. M. F. Hamilton and D. T. Blackstock, “On the coefficient of nonlinearity in

nonlinear acoustics”, J. Acoust. Soc. Am., vol. 83, pp. 74-77, 1988.

REFERENCES FOR THE ACOUSTIC PROPERTIES OF FLUIDS | 839

 13. IEC 61094-2:2009, “Electroacoustics - Measurement microphones - Part 2:
Primary method for pressure calibration of laboratory standard microphones by the
reciprocity technique,” International Standard.

14. A. S. Dukhin and P. J. Goetz, “Bulk viscosity and compressibility measurements

using acoustic spectroscopy”, J. Chem. Phys., vol. 130, p. 124519-1, 2009.

15. M. S. Cramer, “Numerical estimate for the bulk viscosity of ideal gases”, Phys.
Fluids, vol. 30, p. 066102, 2012.

16. P. B. Muller and H. Bruus, “Numerical study of thermoviscous effects in

ultrasound-induced acoustic streaming in microchannels,” Phys. Rev. E, vol. 90, pp.
043016, 2014.

17. E. M. Viggen, “The lattice Boltzmann method: Fundamentals and acoustics,”

Ph.D. Thesis, NTNU-Trondheim, 2014.

18. P. M Morse and K. U. Ingard, Theoretical Acoustics, Princeton University Press,

1986.

840 | CHAPTER 13: ACOUSTIC PROPERTIES OF FLUIDS

 14

Glossary

This Glossary of Terms contains finite element modeling terms in an acoustics

context. For mathematical terms as well as geometry and CAD terms specific to the
COMSOL Multiphysics® software and documentation, see the glossary in the
COMSOL Multiphysics Reference Manual. For references to more information
about a term, see the index.

841
 Glossary of Terms
acoustic impedance At a specified surface, the complex quotient of acoustic pressure
by normal fluid velocity. SI unit: Pa/(m/s).

acoustic reactance The imaginary part of the acoustic impedance.

acoustic resistance The real part of the acoustic impedance.

acoustic-structure interaction A multiphysics phenomenon where the fluid’s pressure

causes a fluid load on the solid domain, and the structural acceleration affects the fluid
domain as a normal acceleration across the fluid-solid boundary.

adiabatic bulk modulus One over the compressibility βs measured at constant entropy.
The adiabatic bulk modulus is denoted Ks and gives a measure of the compressibility
of the fluid and is directly related to the speed of sound cs in the fluid. SI unit: Pa.

1 2
K s = ----- = ρ 0 c s
βs

admittance The reciprocal of impedance.

aeroacoustics The scientific field of study used to couple acoustics and fluid dynamics.

anisotropy Variation of material properties with direction.

arbitrary Lagrangian-Eulerian (ALE) method A technique to formulate equations in a

mixed kinematical description. An ALE referential coordinate system is typically a mix
between the material (Lagrangian) and spatial (Eulerian) coordinate systems.

Bernoulli equation An integrated form of Euler’s momentum equation along a line of

flow. The equation gives an expression for an invariant quantity in an inviscid fluid. A
decrease in the speed of the fluid translates to an increase in the fluid pressure and/or
potential energy.

bulk modulus One over the compressibility. It gives a measure of the compressibility
of the fluid and is related to the speed of sound in the fluid. SI unit: Pa. See also
adiabatic bulk modulus.

842 | CHAPTER 14: GLOSSARY

characteristic impedance The product of the equilibrium density and the speed of
sound in a medium. SI unit: Pa/(m/s).

compliance Reciprocal of stiffness.

compliance matrix The inverse of the elasticity matrix. See elasticity matrix.

Cauchy stress The most fundamental stress measure defined as force/deformed area
in fixed directions not following the body.

constitutive equations Equations that relate two physical quantities. In thermoviscous

acoustics both the stress tensor (relating velocity to stress) and Fourier’s law of heat
conduction (relating heat conduction to temperature) are constitutive relations. In
structural mechanics this is the equation formulating the stress-strain relationship of a
material. Constitutive equations are supplemented by equilibrium equations (mass,
momentum, and energy) and an equation of state to provide a full physical
description.

creep Time-dependent material nonlinearity that usually occurs in metals at high

temperatures in which the effect of the variation of stress and strain with time is of
interest.

damping Dissipation of energy in the fluid or a vibrating structure. The damping is

typically due to viscous losses or thermal conduction. In acoustics this happens in
structures with small geometrical dimensions, for example, small pipes or porous
materials. In structures a common assumption is viscous damping where the damping
is proportional to the velocity. See also Rayleigh damping.

decibel (dB) Logarithmic unit that indicates the ratio of a physical quantity relative to
a reference value.

dipole source An acoustic source that behaves as a translational oscillating sphere.

Doppler effect Change in the observed frequency of a wave caused by a time rate of
change in the effective length of the path of travel between the source and the
observation point.

effective sound pressure RMS instantaneous sound pressure at a point during a time
interval, T, long enough that the measured value is effectively independent of small
changes in T. SI unit: Pa = N/m2.

GLOSSARY OF TERMS | 843

 equation of state The thermodynamic relation between three independent
thermodynamic variables. Typically in acoustics it is the density ρ = ρ(p,s) given as
function of the entropy s and the pressure p.

eigenmode A possible propagating mode of an acoustic wave.

elasticity matrix The matrix D relating strain to stresses:

σ = Dε

Eulerian Model described and solved in a coordinate system that is fixed (spatial). See
also Lagrangian and arbitrary Lagrangian-Eulerian method.

Green-Lagrange strain Nonlinear strain measure used in large-deformation analysis.

In a small strain, large rotation analysis, the Green-Lagrange strain corresponds to
the engineering strain, with the strain values interpreted in the original directions. The
Green-Lagrange strain is a natural choice when formulating a problem in the
undeformed state. The conjugate stress is the second Piola-Kirchhoff stress.

impedance At a specified frequency, the quotient of a dynamic field quantity (such as

force, sound, or pressure) by a kinematic field quantity (such as vibration velocity or
particle velocity).

instantaneous sound pressure Total instantaneous pressure at a point in a medium

2
minus the static pressure at the same point. SI unit: Pa = N/m .

irrotational background velocity field A velocity field u that has the property of having
rotation ∇ × u = 0 everywhere, where the first term is the vorticity of the fluid. In
such a fluid the viscous stress does not contribute to the acceleration of the fluid. The
mean pressure in this fluid is described by Bernoulli’s equation.

Lagrangian Model described and solved in a coordinate system that moves with the
material. See also Eulerian and arbitrary Lagrangian-Eulerian method.

monopole source An acoustic source that behaves as a radially oscillating sphere.

particle velocity In a sound field, the velocity caused by a sound wave of a given
infinitesimal part of the medium relative to the medium as a whole.

844 | CHAPTER 14: GLOSSARY

PML (perfectly matched layer) Domain adjoined at a system boundary designed to
emulate a nonreflecting boundary condition independently of the shape and frequency
of the incident wavefront.

principle of virtual work States that the variation in internal strain energy is equal to
the work done by external forces.

propagating acoustic modes The acoustic modes or wave shapes that propagate with
no significant damping for a given frequency in a duct of a given cross-section.

Rayleigh damping A viscous damping model where the damping is proportional to the
mass and stiffness through the mass and stiffness damping parameters.

reference sound pressure See definition in the entry for sound pressure level.

resonance frequency A frequency at which the system has the tendency to oscillate at
a greater amplitude than at nonresonance frequencies. At the resonance frequencies
the system can easily transfer energy from the actuation to the vibrating structure or
acoustic wave.

RMS value Root-mean-square value; for the (complex) sound pressure, p(t), over the
time interval T1 < t < T2 defined as

1
T Re [ p ( t ) ]
T2 2
p RMS = ------------------- dt
T2 – T1 1

iωt
For a harmonic pressure wave, p ( t ) = p 0 e , the time interval is taken to be a
complete period, resulting in pRMS = p0 /√2.

second Piola-Kirchhoff stress Conjugate stress to Green-Lagrange strain used in

large deformation analysis.

sound energy Total energy in a given part of a medium minus the energy that would
exist at the same part in the absence of sound waves. SI unit: J.

sound-energy flux density See sound intensity.

sound intensity Average rate of sound energy transmitted in a specified direction at a

2
point through a unit area normal to this direction. SI unit: W/m .

sound pressure See effective sound pressure.

GLOSSARY OF TERMS | 845

 sound pressure amplitude Absolute instantaneous sound pressure in any given cycle
of a sound wave at some specified time. SI unit: W/m2.

sound power density See sound intensity.

sound pressure level Ten times the logarithm (to the base ten) of the ratio of the
time-mean-square pressure of a sound, in a stated frequency band, to the square of a
reference sound pressure, pref. For gases, pref = 20 μPa, for other media (unless
otherwise specified) pref = 1 μPa. Unit: dB (decibel).

sound source strength Maximum instantaneous rate of volume displacement

produced by a source when emitting a harmonic sound wave. SI unit: m3/s.

specific acoustic impedance At a point in a sound field, the quotient of sound pressure
by particle velocity. SI unit: Pa/(m/s).

speed of sound The rate of change of particle displacement with distance for a sound
wave. SI unit: m/s.

spin tensor The skew-symmetric part of the velocity gradient tensor.

static pressure Pressure that would exist at a point in the absence of a sound wave.

stiffness Ratio of change of force (or torque) to the corresponding change in

translational (or rotational) displacement of an elastic element.

thermoviscous acoustics The interaction between thermodynamic and acoustic

phenomena, which takes into account the temperature oscillations that accompany the
acoustic pressure oscillations. The combination of these oscillations produces
thermoviscous acoustic effects. Thermoviscous acoustic phenomena are modeled by
solving the full linearized Navier-Stokes equation (momentum equation), the
continuity equation, and the energy equation. Thermoviscous acoustics is also known
as viscothermal acoustics or thermoacoustics.

velocity potential When a flow is irrotational ∇ × u = 0 the vector field (velocity

field) can always be derived from a scalar potential φ(x) as u = ∇φ , where φ is the
velocity potential. See also irrotational background velocity field.

waveguide structures Structures that have the property of guiding sound waves. See
also propagating acoustic modes.

846 | CHAPTER 14: GLOSSARY

GLOSSARY OF TERMS | 847
848 | CHAPTER 14: GLOSSARY
I n d e x
A accumulator (node), boundaries 659 Application Libraries window 48
accumulator (node), domains 671 application library examples
accuracy order, of time stepping 647 acpr interface 64, 792–793, 798, 803,
acoustic diffusion equation interface 739 828
acoustic diffusion model (node) 743 actd interface 147, 152, 259
acoustic intensity vector 225 ae interface 464
acoustic-elastic waves interaction inter- aebm interface 478
face 392 background pressure field 89
acoustic-piezoelectric interaction, fre- Biot equivalent fluid model 303
quency domain (acpz) interface 384 cpf interface 481
acoustic-piezoelectric interaction, tran- cylindrical wave radiation 121
sient (acpztd) interface 387 eigenfrequency study 257
acoustic-poroelastic waves interaction Eigenmodes in a Muffler 162
interface 390 far field calculation 107, 114, 131
acoustic-porous boundary (multiphysics far field plots 234–235
coupling) 792 flow point source 137
acoustic-shell interaction, frequency do- fluid model 77
main interface 397 Gaussian pulse point source 155
acoustic-shell interaction, transient in- Generic 711 Coupler — An Occluded
terface 400 Ear-Canal Simulator 98
acoustic-solid interaction, frequency do- Lumped Receiver Connected to Test
main interface 378 Setup with an 0.4-cc Coupler 85
acoustic-solid interaction, transient in- mass flow point source 473
terface 381 mode analysis study 821
acoustic-structure boundary (multiphys- pipe acoustics, transient 780
ics coupling) 784 point source 99
acoustic-thermoacoustic boundary (mul- poroelastic waves 325
tiphysics coupling) 790 pressure acoustics, frequency domain
acoustic-thermoacoustic interaction, 64, 792–793, 798, 803, 828
frequency domain 565 spherical wave radiation 165
adiabatic (node) 454, 548 ta interface 524, 787
adiabatic bulk modulus 252 artificial boundary conditions 38
adiabatic, reversible processes 590 asymptotic far-field radiation (node) 423
aeroacoustic-structure boundary (multi- attenuation coefficient 70, 595
physics coupling) 789 auxiliary dependent variable 658
analysis. see study types. auxiliary dependent variable (node) 679

INDEX| 849
 axial symmetry (node) 92, 473 complex conjugate operator 228
complex impedance 36
B background potential flow (multiphysics
complex wave numbers 36
coupling) 793
compressible potential flow (cpf) inter-
background pressure field (node) 86, 152
face 480
background pressure wave 59
compressible potential flow model
background velocity (node) 678
(node) 482
barotropic fluids 513
compute intensity 645
Bernoulli equation 513
compute intensity and power 645
Biot theory 365
compute intensity and power in graded
Biot’s high frequency range 366
media 645
Bloch periodicity 539
compute intensity in graded media 645
boundary mode acoustics (acbm) inter-
continuity (node)
face 160
acpr interface 132
theory 259
ae interface 424, 472
boundary nodes
lns interface 455
acbm interface 162
convected wave equation interface 602
acpr interface 64
count reflections 646
actd interface 148
Coupling Operators 708
ae interface 464
Courant number 780
aebm interface 478
cwe nodes
aetd interface 475
acpr interface 604
cpf interface 481
cylindrical wave radiation (node) 121
cwe interface 604
lef and let interfaces 412 D damping
lnsf and lnst interfaces 438 boundaries 38
pelw interfaces 326 degrees of freedom 207
ta interface 524, 553 Density of exterior domains 644, 706
tabm interface 562 diffuse scattering, wall condition 653
boundary-layer absorption 306 dipole point source (node) 137
bulk modulus 774 dipole source (node) 90
elastic moduli 345 dipole sources 252
pressure acoustics 35 disappear, wall condition 653
dispersion curves 821
C caustic surface 719
dispersion relation 596
CFL condition 779
documentation 47
circular duct 309
domain nodes
circular source (node) 143
acbm interface 162
circumferential wave number 255, 260
acpr interface 64
common settings 46

850 | I N D E X
 actd interface 148 first order material parameters (node)
ae interface 464 444
aetd interface 475 fitted domain (node) 746
cpf interface 481 fixed (node) 334
cwe interface 604 Floquet periodicity 539
lef and let interfaces 412 fluid models 288
lnsf and lnst interfaces 438 fluid models, pressure acoustics 68, 172
pelw interfaces 326 Fraunhofer diffraction 231
ta interface 524, 553 freeze, wall condition 653
domain source (node) 746 frequency domain modal study 822
domain sources (node) 417, 444 frequency domain study 817
double dot operator 581 ae interface 510
Fresnel numbers 231
E edge nodes
acbm interface 162 G general reflection, wall condition 654
acpr interface 64 Green’s function 264
actd interface 148
H Hankel function 264
ae interface 464
harmonic time dependence 225
aebm interface 478
heat flux (node) 454
aetd interface 475
heat source (node) 530
eigenfrequency study 818
Helmholtz-Kirchhoff integral 231, 263
pressure acoustics 256
Higdon conditions 261
elastic waves (elw) interface
high frequency range 366
theory 364
I ideal barotropic fluids 513
elkernel element 266
ideal gas fluid model 73
emailing COMSOL 49
impedance (node) 421
empty study 823
acpr interface 95
end impedance (node) 770
ae interface 470
equilateral triangular duct 309
impedance, complex 36
equivalent fluid model 69, 172, 293
incident acoustic fields (node) 418
equivalent fluid models 288
incident pressure field (node) 121, 154
Euler equations 512
incident velocity potential (node) 468
evanescent modes 820
initial values (node)
evanescent wave components 230
abe interface 746
F far field variables 233
acbm interface 161
far-field calculation (node) 103
acpr interface 89, 609
far-field limits 232, 265
ae interface 466
far-field regions 230
cpf interface 483
far-field variables 232

INDEX| 851
 elw interface 334 K knowledge base, COMSOL 50
lem interface 417 Korteweg formula 774
let interface 429 L Lagrange elements 208
lnsf and lnst interfaces 445 Lamé parameters 345
pafd interface 766 large acoustics problems 213
patd interface 766 Lax-Friedrichs flux parameter 185, 346,
ta interface 531 607
inlet on axis (node) 677 limp frame 75
insertion loss curves 225 limp porous matrix model 293
intensity 645 line source (node) 132
intensity attenuation coefficient 70 line source on axis (node) 134
intensity computation 717–718 linear elastic fluid model 288
intensity variables 225 linearized Euler model (node) 413
interior impedance (node) 127, 421 linearized Euler, frequency domain inter-
interior normal acceleration (node) 126 face 409
interior normal displacement (node) 127, linearized Euler, transient interface 426
159 linearized Navier-Stokes model (node)
interior normal impedance (node) 449, 439
542 linearized Navier-Stokes, boundary
interior normal velocity (node) 127, 159 mode interface 459
interior perforated plate (node) 128 linearized Navier-Stokes, frequency do-
interior sound hard boundary (wall) main interface 434
(node) linearized Navier-Stokes, transient inter-
acpr interface 125 face 456
ae interface 471 linearized potential flow model (node)
interior velocity (node) 543 465
interior wall (node) 423, 448, 541 linearized potential flow, boundary mode
internet resources 47 interface 477
inward energy flux (node) 745 linearized potential flow, frequency do-
irrotational velocity fields 509 main interface 462
isentropic processes 590 linearized potential flow, transient inter-
isentropic speed of sound 580 face 474
isothermal (node) 453, 548 liquids and gases materials 238, 828
isotropic meshes 208 low reduced frequency (LRF) 309
isotropic scattering, wall condition 653
M mapped room coupling (node) 747
J JCA model 79 mass flow (node) 484
JCAL model 80 mass flow circular source (node) 473
JCAPL model 80 mass flow edge source (node) 472

852 | I N D E X
 mass flow line source on axis (node) 473 no slip (node) 450, 544
mass flow point source (node) 472 no stress (node) 452, 547
mass flows 467 nodes, common settings 46
matched boundary (node) 102 no-flow conditions 466, 471
matched boundary conditions 38 nonlinear pressure acoustics time explic-
material discontinuity (node) 661 it interface 614
materials nonlocal accumulator (node) 672
porous absorbing 36 nonreflecting boundary conditions 38
medium properties (node) 650 normal acceleration (node) 92–93, 156–
microacoustics 518 157, 190, 609
mixed diffuse and specular reflection, normal flow (node) 484
wall condition 654 normal impedance (node) 453, 547
mode analysis study 820 normal mass flow (node) 467
boundary mode acoustics 259 normal velocity (node) 470
theory 595 NRBC. see nonreflecting boundary con-
monopole point source (node) 135 dition. 102
monopole source (node) 89 Nyquist criterion 218
monopole sources 252
O outflow boundary (node) 424
motionless skeleton model 293
out-of-plane wave number 255, 510
moving wall (node) 422, 429
P pair impedance (node)
MPH-files 48
acpr interface 127
multigrid solvers 213
ae interface 470
multiphysics 378, 381, 397, 400, 565, 568
pair nodes
multiphysics coupling
acpr interface 64
acoustic-porous boundary (node) 792
ae interface 464
acoustic-structure boundary (node)
aebm interface 478
784
aetd interface 475
acoustic-thermoacoustic boundary
cpf interface 481
(node) 790
cwe interface 604
aeroacoustic-structure boundary
lef and let interfaces 412
(node) 789
lnsf and lnst interfaces 438
background potential flow (node) 793
pelw interfaces 326
porous-structure boundary (node) 793
ta interface 524, 553
thermoviscous acoustic-structure
tabm interface 562
boundary (node) 787
pair perforated plate (node) 128
multiscale acoustics modeling 687
pair room coupling (node) 747
N narrow region acoustics (node) 82
parameter of nonlinearity 258
near-field regions 230
particle velocity 225

INDEX| 853
 pass through, wall condition 653 power dissipation variables 227
perfectly matched layers (node) 219 ppr() operator 106
periodic condition (node) 100 Prandtl number 519
phase factors, pressure acoustics 255 prescribed acceleration (node) 337
phase variables 818 prescribed acoustic fields (node) 419
physics interfaces, common settings 46 prescribed displacement (node) 335
piezoelectric waves, time explicit inter- prescribed pressure (node) 452
face 355 prescribed temperature (node) 454
pipe acoustics, frequency domain (pafd) prescribed velocity (node) 336, 451
interface 762 pressure (adiabatic) (node) 447
pipe acoustics, transient (patd) interface pressure (isentropic) (node) 419
764 pressure (node) 94, 610
theory 772 pressure acoustics (node) 67, 605
plane of incidence 719 pressure acoustics interfaces 250
plane wave radiation (node) pressure acoustics model, fluid models
acpr interface 119 289
ae interface 467 pressure acoustics models, theory 288
point nodes pressure acoustics theory 250
acbm interface 162 pressure acoustics, asymptotic scattering
acpr interface 64 (paas) interface 196, 201
actd interface 148 pressure acoustics, frequency domain
ae interface 464 (acpr) interface 59, 163
aebm interface 478 pressure acoustics, time explicit (pate)
aetd interface 475 interface 180
cwe interface 604 pressure acoustics, time explicit model
point source (node) (node) 183
ade interface 748 pressure acoustics, transient (actd) in-
Poisson’s ratio 345 terface 145
poroacoustics (node) 74 pressure, adiabatic (node) 532
poroelastic material (node) 326 pressure-wave speeds 345
poroelastic waves (elw) interface 324 principal radii of curvature 718
poroelastic waves (pelw) interface propagating acoustic modes 595
theory 364 propagating modes 820
porous absorbing materials 36
Q quadrupole point source (node) 139
porous, free (node) 333
R radiation boundary conditions 38
porous, pressure (node) 335
random number generation 659
porous-structure boundary (multiphys-
ray acoustics interface 642
ics coupling) 793
ray continuity (node) 692

854 | I N D E X
 ray properties (node) 663 ae interface 469
ray termination (node) 693 source term 227, 580
ray termination, theory 731 source with directivity (node) 689
Rayl unit 96 specular reflection, wall condition 653
rectangular duct 309 Speed of sound in exterior domains 644,
release (node) 663 705
release from boundary (node) spherical wave radiation (node) 120
ray acoustics 673 standard settings 46
release from data file (node) 684 stationary study 514, 817
release from edge (node) 679 stick, wall condition 653
release from exterior field calculation study types
(node) 687 acbm interface 259
release from grid (node) 680 actd interface 257
release from grid on axis (node) 684 ae interface 510
release from point (node) 680 swirl-correction factor 772
release from point on axis (node) 680 symmetry (node) 100, 421, 610
resonance frequency 818 ta interface 533
results analysis and variables 224 symmetry planes, far-field analysis 104
rigid frame 75
T Taylor series 773
rigid porous matrix model 293
technical support, COMSOL 49
rigid wall (node) 416
temperature variation (node) 549
roller (node) 338
theory 250
room (node) 743
acbm interface 259
room coupling (node) 747
elw and pelw interfaces 364
S septum boundary load (node) 338 patd interface 772
shear modulus expression 345 pressure acoustics models 288
shear-wave speeds 345 ta interface 584
slip (node) 450, 545 tas interface 584
slip velocity (node) 483 thermal dissipation 227, 580
slit 309 thermally conducting and viscous fluid
solving large problems 213 model 292
sound hard boundary (wall) (node) thermally conducting fluid model 291
acpr interface 92, 609 thermoacoustics 518
ae interface 466 thermoviscous acoustics model (node)
sound pressure level calculation (node), 526, 563
boundaries 670 thermoviscous acoustics, frequency do-
sound soft boundary (node) main (ta) interface 518, 559
acpr interface 94 theory 584

INDEX| 855
 thermoviscous acoustic-shell interac- W wall (node) 446, 531, 744
tion, frequency domain (tash) inter- ray optics 653
face 570 wall condition 653
thermoviscous acoustic-solid interac- wave numbers 36, 255, 510
tion, frequency domain 568 wave radiation equations 261
thermoviscous acoustic-solid interac- wave speeds 345
tion, frequency domain (tas) inter- waveguide 820
face waveguide structures 595
theory 584 websites, COMSOL 50
thermoviscous acoustic-structure wide ducts 307
boundary (multiphysics coupling) Womersley number 574
787 Y Young’s modulus expression 345
time dependent study 821
pressure acoustics, transient 257
time-dependent, modal study 822
time-harmonic study
aeroacoustics 510
transient pressure acoustics model
(node) 148
transmission loss 225

U U-P formulation 367

user-defined attenuation fluid model 290

V variables
for far fields 233
intensity 225
phase 818
power dissipation 227
results analysis 224
velocity (node) 546, 769
velocity potential (node) 467
velocity potential, compressible poten-
tial flow, and 513
very narrow circular ducts (isothermal)
308
viscothermal acoustics 518
viscous dissipation 227, 580
viscous fluid model 72, 291
vortex sheet (node) 470

856 | I N D E X