Crack Bandwidth

Outline

1 Description 3
2 Finite element model 4
2.1 Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
2.3 Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.5 Mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3 Structural Nonlinear Analysis 13
3.1 Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2.1 Crack Bandwidth – Rots Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
3.2.2 Crack Bandwidth – Govindjee Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.2.3 Crack Bandwidth – User Specified 0.0447 m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.2.4 Crack Bandwidth – User Specified 0.1 m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2.5 Comparison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Appendix A Additional Information 24

Crack Bandwidth | [Link] 2/25

1 Description
This tutorial demonstrates the function of the different crack bandwidths models available in Diana. A concrete two-dimensional beam loaded in tensile failure is modelled using a rotating
total strain model [Fig. 1]. The model has the following characteristics:

• the beam is discretized with elongated elements with a dimension of 0.1mm x 0.02 mm
• the model runs four times, each time with a different crack bandwidth definition
• a force versus displacement graph is made for each analysis in order to compare the results
• only the right half of the beam is modeled since symmetry conditions are used

Figure 1: Characteristics of the model

Crack Bandwidth | [Link] 3/25

2 Finite element model
We create a new project for a two dimensional structural model with a model size of 1m m, hexa/quad mesher type and quadratic mesh order with linear interpolation. The used units are SI
units.
DianaIE

Main menu File New [Fig. 2]

Geometry browser Reference system Units [Fig. 3]
Property Panel [Fig. 4]

Figure 2: New project dialog Figure 3: Geometry browser - units Figure 4: Properties panel - units

Crack Bandwidth | [Link] 4/25

2.1 Geometry
The geometry of the model consists of two polygon sheets. One sheet is for the beam [Fig. 5] and the other for the loading plate [Fig. 6].
DianaIE

Main menu Geometry Create Polygon sheet [Fig. 5] [Fig. 7]

Main menu Geometry Create Polygon sheet [Fig. 6] [Fig. 8]

Figure 7: Geometry view - Beam

Figure 5: Add polygon sheet Figure 6: Add polygon sheet Figure 8: Geometry view - Plate

Crack Bandwidth | [Link] 5/25

2.2 Properties
We define the material and geometry properties of the shapes. We start with the concrete beam.
DianaIE

Main menu Geometry Assign Shape properties [Fig. 9]

Shape properties Material Add material [Fig. 10] Edit material [Fig. 11]– [Fig. 14]

Figure 9: Properties assignment Beam Figure 10: Add material Concrete Figure 11: Edit material Concrete

Crack Bandwidth | [Link] 6/25

Note that initially Rots model is chosen for the specification of the crack bandwidth.

DianaIE

Shape properties Geometry Add new geometry [Fig. 15]

Figure 12: Edit material Concrete Figure 13: Edit material Concrete Figure 14: Edit material Concrete Figure 15: Edit geometry Thickness

Crack Bandwidth | [Link] 7/25

We set the properties of the steel loading plate. Here we can use the geometry thickness created previously for the concrete beam and assign it to the loading plate as well.
DianaIE

Main menu Geometry Assign Shape properties [Fig. 16]

Shape properties Material Add material [Fig. 17] Edit material [Fig. 18]

Figure 16: Properties assignment Plate Figure 17: Add material Steel Figure 18: Edit material Steel

Crack Bandwidth | [Link] 8/25

2.3 Boundary Conditions
We add three supports to the model: i) a point support at the right-hand side [Fig. 19], ii) a line support for the symmetry condition [Fig. 20] and iii) a point support for the prescribed
deformation [Fig. 23].
DianaIE

Main menu Geometry Assign Supports [Fig. 19] [Fig. 21]

Main menu Geometry Assign Supports [Fig. 20] [Fig. 22]

Figure 21: View support: Boundary condition

Figure 19: Attach support: Boundary

condition Figure 20: Attach support: Symmetry Figure 22: View support: Symmetry

Crack Bandwidth | [Link] 9/25

 DianaIE

Main menu Geometry Assign Supports [Fig. 23] [Fig. 24]

Figure 23: Attach support: Deform Figure 24: View support: Deform

Crack Bandwidth | [Link] 10/25

2.4 Loads
We define the load applied to the beam: a prescribed deformation of 1 mm in negative Y direction.
DianaIE

Main menu Geometry Assign Loads [Fig. 25] [Fig. 26]

Figure 25: Application of the load Figure 26: View of the load

Crack Bandwidth | [Link] 11/25

2.5 Mesh
For the mesh properties we use the divisions seeding method. We use one division for the steel loading plate [Fig. 27] and ten divisions for the beam [Fig. 28]. This leads to an element size of
0.1 × 0.02 mm for all the elements [Fig. 29].
DianaIE

Main menu Geometry Mesh Mesh properties [Fig. 27]

Main menu Geometry Mesh Mesh properties [Fig. 28]
Main menu Geometry Mesh Generate mesh [Fig. 29]

Figure 27: Mesh properties - Plate Figure 28: Mesh properties - Beam Figure 29: Finite element mesh

Crack Bandwidth | [Link] 12/25

3 Structural Nonlinear Analysis
3.1 Commands
We setup a structural nonlinear analysis which we will run four times, one for each model considered for the crack bandwidth. We specify the load steps and use arc length control.
DianaIE

Main menu Analysis Add analysis [Fig. 30]

Analysis browser Nonlinear failure Add command Structural nonlinear [Fig. 31]
Analysis browser Nonlinear failure Structural nonlinear new execute block Load steps Edit properties [Fig. 32] [Fig. 33]

Figure 30: Analysis browser Figure 31: Analysis browser Figure 32: Analysis browser Figure 33: Properties - load steps

Crack Bandwidth | [Link] 13/25

Within the load step properties, we define the arc length control. We increase the maximum number of iterations per step to 30, and select a force convergence norm only. For this convergence
norm we specify that the analysis should continue when no convergence has occurred.
DianaIE

Arc length setting Control type - Regular Settings [Fig. 34]

Regular arc length control settings Control sets Add new [Fig. 35]
Analysis browser Nonlinear failure Structural nonlinear new execute block Equilibrium iteration Edit properties [Fig. 36]

Figure 37: Iteration properties

Figure 35: Regular arc length control

Figure 34: Regular arc length control settings settings Figure 36: Analysis browser Figure 38: Force convergence norm

Crack Bandwidth | [Link] 14/25

We define the output properties of the analysis. We select total displacements, force reactions and crack strains. We then we run the analysis.
DianaIE

Analysis browser Nonlinear failure Structural nonlinear Output Edit properties [Fig. 39] Result User selection Modify
Main menu Analysis Run analysis

Figure 39: Analysis browser Figure 40: Output properties Figure 41: Output selection

Crack Bandwidth | [Link] 15/25

3.2 Results
3.2.1 Crack Bandwidth – Rots Model
We present the crack strains Eknn for the model with Rots crack bandwidth specification [Fig. 13].
We copy the vertical support reaction in the loaded node and past it in an Excel sheet to create a graph with the results from all the different crack bandwidth models.
DianaIE

Results browser Case Load-step 40, Load-factor ...

Results browser Nonlinear failure Output Element results Crack Strains Eknn [Fig. 42]
Results browser Nonlinear failure Output Nodal results Reaction Forces FBY Show table [Fig. 49]

Figure 42: Crack strains with Rots crack bandwidth model

Crack Bandwidth | [Link] 16/25

3.2.2 Crack Bandwidth – Govindjee Model
We now run the same analysis again, after we have changed the definition of the crack bandwidth from Rots to Govindjee [Fig. 43].
DianaIE

Geometry browser Materials Concrete Edit material [Fig. 43]

Main menu Geometry Mesh Generate mesh
Main menu Analysis Run analysis

Figure 43: Edit material - tensile behavior

Crack Bandwidth | [Link] 17/25

We present the crack strains Eknn for this model.
We also copy the vertical support reaction in the loaded node and past it in the Excel sheet for the comparison graph.
DianaIE

Results browser Case Load-step 40, Load-factor ...

Results browser Nonlinear failure Output Element results Crack Strains Eknn [Fig. 44]
Results browser Nonlinear failure Output Nodal results Reaction Forces FBY Show table [Fig. 49]

Figure 44: Crack strains with Govindjee crack bandwidth model

Crack Bandwidth | [Link] 18/25

3.2.3 Crack Bandwidth – User Specified 0.0447 m
We run the same analysis again, after we have changed the definition of the crack bandwidth from Govindjee to User specified with a crack bandwidth of 0.0447 [Fig. 45].
DianaIE

Geometry browser Materials Concrete Edit material [Fig. 45]

Main menu Geometry Mesh Generate mesh
Main menu Analysis Run analysis

Figure 45: Edit material - tensile behavior

Crack Bandwidth | [Link] 19/25

We present the crack strains Eknn for this model.
We also copy the vertical support reaction in the loaded node and past it in the Excel sheet for the comparison graph.
DianaIE

Results browser Case Load-step 40, Load-factor 1.0486

Results browser Nonlinear failure Output Element results Crack Strains Eknn [Fig. 46]
Results browser Nonlinear failure Output Nodal results Reaction Forces FBY Show table [Fig. 49]

Figure 46: Crack strains with a crack bandwidth of 0.0447 m

Crack Bandwidth | [Link] 20/25

3.2.4 Crack Bandwidth – User Specified 0.1 m
We now run the same analysis after we have changed the crack bandwidth from 0.0447 to 0.1 m [Fig. 47].
DianaIE

Geometry browser Materials Concrete Edit material [Fig. 45]

Main menu Analysis Run analysis

Figure 47: Edit material: Concrete

Crack Bandwidth | [Link] 21/25

We present the crack strains Eknn for this model.
We also copy the vertical support reaction in the loaded node and past it in the Excel sheet for the comparison graph.
DianaIE

Results browser Case Load-step 40, Load-factor 1.0335

Results browser Nonlinear failure Output Element results Crack Strains Eknn [Fig. 48]
Results browser Nonlinear failure Output Nodal results Reaction Forces FBY Show table [Fig. 49]

Figure 48: Crack strains with a crack bandwidth of 0.1 m

Crack Bandwidth | [Link] 22/25

3.2.5 Comparison

With the results that are exported to Excel, we can make a graph with the vertical reaction force versus the displacement.

In Figure 49 we can observe that using Rots crack bandwidth leads to almost exactly the same results as using a user specified crack bandwidth of 0.0477 m. Similarly, the results of using
Govindjee crack bandwidth leads to almost exactly the same results as using a user specified crack bandwidth of 0.1 m.
The difference between the results using Rots and Govindjee came from the different approaches used in the two models: Rots 1 approach uses the size of the elements to determine the crack
bandwidth while Govindjee 2 approach uses the orientation of the crack to determine the crack bandwidth.

Figure 49: Force vs. displacement graph

1
Rots, Computational Modeling of Concrete Fracture, 1988
2
Govindjee et al., Anisotropic Modelling and Numerical Simulation of Brittle Damage in Concrete, 1995

Crack Bandwidth | [Link] 23/25

Appendix A Additional Information

Folder: Tutorials/CrackBandWidth

Number of elements ≈ 100

Keywords:
analys: nonlin physic.
constr: suppor.
elemen: cq16m pstres.
load: deform.
materi: crack elasti harden isotro linear rotati soften thoren totstr.
option: arclen direct newton normal regula select update.
post: binary ndiana.
pre: dianai.
result: crack displa force green reacti strain total.

References:

[1] J. G. Rots. Computational Modeling of Concrete Fracture. PhD thesis, Delft University of Technology, 1988.
[2] S. Govindjee, G. J. Kay, and J. C. Simo. Anisotropic Modelling and Numerical Simulation of Brittle Damage in Concrete. Int. J. Num. Meth. Eng., 38:3611–3633, 1995.

Crack Bandwidth | [Link] 24/25

 DIANA FEA BV DIANA FEA BV
Thijsseweg 11 Vlamoven 34
2629JA Delft 6826 TN Arnhem
The Netherlands The Netherlands
T +31 (0) 88 34262 00 T +31 (0) 88 34262 00

[Link]
© DIANA FEA BV
Disclaimer: The aim of this technical tutorial is to illustrate various tools, modelling techniques and analysis workflows in DIANA.
DIANA FEA BV does not accept any responsibility regarding the presented cases, used parameters, and presented results.