Extending Mohr Coulomb state variables

[jduriez]

Description
This PR includes new state variables for the Mohr-Coulomb material:

• 4 (binary) flags specifying the type of failure (if any). It can now be distinguished whether the material is currently (or has ever) failed in shear and/or tension

• stored elastic energy. Note that it is denoted E_el, should it be rather eEl ?

• cumulated plastic work, i.e. dissipation, given as a positive number. (well, let's not argue here whether this really is a state variable from a theoretical perspective ;-) ). Which is currently W_pl (should it be rather wPl ?)

Additional context

• I was about to modify mpm-doc as well to document this change, but have not yet fully decided about the most appropriate place to do so. Probably in source corresponding to https://mpm.cb-geo.com/#/theory/material/mohr-coulomb, in a new (final) section "Output state variables" ?

• This code comes from @SchDvr work during his PhD thesis. There is quite a number of other updates that could be proposed to be applied following his work, but this should probably be further discussed elsewhere.

[jduriez]

Sorry for the initial check failure, my local test took place in a more compliant Ubuntu 22.04 environment (where brace initializing Vector6d was possible)

[codecov]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (53fcb76) 96.75% compared to head (bc7cfa9) 96.75%.

Files	Patch %	Lines
include/materials/mohr_coulomb.tcc	96.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff            @@
##           develop     #742   +/-   ##
========================================
  Coverage    96.75%   96.75%           
========================================
  Files          130      130           
  Lines        25932    25972   +40     
========================================
+ Hits         25090    25129   +39     
- Misses         842      843    +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[kks32]

Thank you so much @jduriez for the suggested updates. Could we add some mpm-benchmark tests to make sure the previous version will still work and any new tests for this modified MC?

[jduriez]

Thank you so much @jduriez for the suggested updates. Could we add some mpm-benchmark tests to make sure the previous version will still work and any new tests for this modified MC?

Thanks for feedback @kks32 ! Do you mean "mpm-benchmark tests" like in mpm-benchmark ? If yes, I'm afraid that would be too much for us at the moment (and for possibly a long time due to lack of time..) but I just tried to include a greater consideration of these new state variables in the catch test tests/materials/mohr_coulomb_test.cc as here:

mpm/tests/materials/mohr_coulomb_test.cc

Line 3392 in f4d183b

REQUIRE(state_variables.at("shear_fail") == 1);

to L3398

As for whether previous version would still work, we somehow relied on the fact existing MC tests still pass and I would mostly stress out we are (aim at, at least ! ;-) ) touching here only post-processing state variables.

Let me know what you think (I should anyway try to come with a mpm-doc update about MC state variables next week..)

[jduriez]

For the record, the incremental elastic energy expression is actually to improve. We currently multiply here the updated stress value with elastic strain increment, which is an overestimation. Using a mid-point rule between updated_stress and stress should be much better, so maybe it is indeed better not to merge currently