Crack Propagation problem in ABAQUS

[GauthamBP]

Hello there!

I have been trying to model a crack propagation problem for quite some time now. I have taken a simple double cantilever beam as my problem of consideration. I am using a direct cyclic step and using VCCT debonding for modelling the crack. I have applied loads such that the double cantilever 'opens up'. I have also been following this sample problem solved for the double cantilever beam as seen in the link below. I have modelled the double cantilever beam using shell elements, just like the sample problem.

http://dsk.ippt.pan.pl/docs/abaqus/v6.13/books/exa/default.htm?startat=ch01s04aex55.html

The input file for the sample solved problem is given below.

https://files.engineering.com/getfi...dc3-9672cfdf9209&file=dcb_vcct_fatigue_3d.inp

My setup of the problem is the input file given below.

https://files.engineering.com/getfi...aad1-e5b6277c82ec&file=Double-Cantilever2.inp

The problem is that my simulation doesn't show any propagation of the crack. Moreover, when I run the sample problem input file, it directly skips to the time where the fracture takes place. I am not able to see how we are supposed to make that happen. It would be helpful if you could tell me what I am missing!

 

[FEA way]

Unfortunately I can’t run your analysis now but I will try to help anyway. Did you use only CAE to set this up or was it also prepared by editing .inp file ? Do you get any warnings after your simulation is completed ? The fact that there’s no crack propagation in your case means that either the criterion was not met or interface was not properly defined. The problem with visualization of results from exemplary analysis may be caused by incrementation/output writing frequency settings (try adding more increments and points at which data is written to odb) or the way you display results. Create animation and see what happens.

 

[GauthamBP]

I did modify the input file too. I had to include the fatigue fracture criterion. I did get a warning but that was associated with the nlgeom being on in the previous step and not being on for the direct cyclic step. What stumps me is that when you run the sample input file, it directly jumps to the step where the fracture takes place. I even relaxed the critical energy release rates(to meet the criterion) to allow the crack to propagate but I don't see it happening. Increasing the time points also doesn't seem to make a difference. I can't see why the sample problem works and mine doesn't. Could you compare the input files and see if I may have missed anything out?