Plane Stress assumption in Chaboche Viscoplasticity

[VaishnavBhaskaran]

Hello!
I am using ChabocheViscoplasticity material behaviour to model a simple plate. I have two issues in implementing Plane Stress using this library.

Issue 1:
I am using legacy FEniCS. I am using an older version of Chaboche (please refer Old_ChabocheViscoplasticity.mfront). Everything works fine with this if Plane Strain is assumed but it doesn't have Plane Stress behaviour in it. Is there a way to include Plane Stress in the same code?

Issue 2:
I downloaded the latest version of Chaboche (please refer New_ChabocheViscoplasticity.mfront) which has the following libraries

If I use this in my code (with legacy FEniCS), I get the following error after some increments for Plane Strain.

I change it to Plane Stress and still it cannot load the behaviour

Traceback (most recent call last): File "/home/vaishnav27/Fenics_Simulation/test.py", line 59, in <module> b = mgis_bv.load('/home/vaishnav27/Fenics_Simulation/New_folder/src/libBehaviour.so','ChabocheViscoplasticity',h) RuntimeError: LibrariesManager::getBehaviour: can't load behaviour 'ChabocheViscoplasticity' in library '/home/vaishnav27/Fenics_Simulation/New_folder/src/libBehaviour.so' for hypothesis 'PlaneStress'

Is FEniCS version causing this issue or is it something else?

It's been only a few weeks since I started using FEniCS and mfront. I apologize if I have missed anything fundamental.
New folder.zip

[thelfer]

Hi @VaishnavBhaskaran,
I am currently on vacation, I can't read your zip file on my phone. Can you copy/paste the Mfront implémentations ?
However, support for plane stress must bé explicitly requêtes using:

@ModellingHypotheses ".+";

But this requires that thé behaviour Can Indeed support plane stress

[maxence-wz]

@VaishnavBhaskaran, as you use the StandardElasticity brick, plane stress is automatically handled by the generated law (see). So no need for the changes in the example here, sorry for the confusion...

[VaishnavBhaskaran]

No problem @maxence-wz!

I tried replacing the IsotropicLinearHardeningPlasticity.mfront from this tutorial with both the versions of ChabocheViscoplasticity.mfront but still the newer version resulted in the same convergence error. The older version was running fine.

[maxence-wz]

@VaishnavBhaskaran the convergence error only occurs for plane stress hypothesis?

[VaishnavBhaskaran]

No, I tried with Plane Strain. I don't think the convergence is related to the Plane Stress hypothesis

[maxence-wz]

@VaishnavBhaskaran are you sure of your material property values in your test file? In particular, the value of m.

[thelfer]

Hi @VaishnavBhaskaran,
I checked the old and new versions. The new one seems much better IMHO, although quite similar to the old one. Both shall work as expected, but you do have to use a very low convergence criterion (typically 1e-13 is fine).
I need to investigate further.
Thomas

[thelfer]

So everything is ok, right ?

[VaishnavBhaskaran]

Yes. Thank you very much for your timely help!
Just a small question. Are there any tutorials for implementing a dynamic analysis with MFront?

[thelfer]

Dynamic aspects are handled by the solver. From the behaviour point of view, nothing changes...

P.S.: This is not exaclty true: the behaviour can provide additional information to help the solver, but it's quite advanced

[thelfer]

BTW @bleyerj has published some nice tutorial to make dynamic computations with FEniCs (probably FeniCsX)

[VaishnavBhaskaran]

Okay. Thank you so much! I'll have a look at those.