[mgis.fenics] tutorial for small strain and finite strain crystal plasticity

[jeanmichelscherer]

Hi @thelfer and @bleyerj,

I wrote a demo for crystal plasticity while getting familiar with mgis.fenics (see archive below).
If you want to add it here, I can write a short tutorial inspired by those already available. Let me know which file format you would prefer.

mgis_fenics_single_crystal.tar.gz

[bleyerj]

Hi @jeanmichelscherer
if you can write a short demo that would be indeed perfect, thank you!
I wonder if doing it for a polycrystal rather than a single crystal is not preferable though. I have some piece of code somewhere but I never took time to write the demo. I can send it to you if you want.
I think that the main points to illustrate, compared to the already existing demos are the fact that the behaviour is rate-dependent (hence the use of problem.dt) and the handling of rotation matrices for global/material frame rotation.
Regarding format, Jupyter notebook is preferable. We can then convert everything to Markdown or HTML from this.

[thelfer]

@jeanmichelscherer Thanks for the proposal. All the documentation is currently in markdown (pandoc flavor), but as @bleyerj said, do with what is the easiest for you as long if it is not in Word on something similar.

But if you use LaTeX or something similar, maybe could you use the commands used everywhere else in the documentation (for tensors, etc...). See this file for example: https://github.com/thelfer/tfel/blob/master/docs/web/FeFpImplicitPlasticity.md

[jeanmichelscherer]

Ok for a polycrystal tutorial. I have a working polycrystal example for which I can write a demo. I will post the jupyter notebook (using your LaTeX shortcuts) here when it is ready.

[thelfer]

Great !

[jeanmichelscherer]

Hi,
Attached is the polycrystal tutorial and associated files for the crystal plasticity constitutive behaviour, the polycrystal mesh and the list of crystallographic orientations. I left two questions for you in the Jupyter notebook. The first concerns a trick in the original crystal plasticity implementation that seems to be a good idea at first sight, but which makes the global newton converge poorly or even diverge. The second is a question regarding the definition of the eel tensor used by Mfront for a finite strain behaviour.

Let me know if you have suggestions regarding these files.

CrystalPlasticityTutorial.tar.gz

[thelfer]

Thanks @jeanmichelscherer !
For the first question, I have no clue. @bleyerj is the divergence of the behaviour not correctly handled ?
For the second question, the FiniteStrainSingleCrystal brick does define eel as the Green-Lagrange strain in the intermediate configuration.