Thomas Nagel (c42aeb73) at 15 Mar 17:44
Merge branch 'THM-RemovePrevStrain' into 'master'
... and 214 more commits
A few remarks:
F_\text{Dil}
implementations, I see the same equations whether effective is true or false. For true, you might want to add primes.Made some changes that allow larger \Delta t
Thomas Nagel (aa3d992d) at 22 Feb 19:25
Removed test files. Use Training Course Repo for now.
Thomas Nagel (8f4bb2ed) at 22 Feb 15:59
Merge branch ogs:master into SimpleMaterialModels
... and 185 more commits
Just FYI: @joergbuchwald will see if we can use an enhanced Newton scheme to achieve convergence with larger \Delta t
.
In principle it works, as the following picture shows. But convergence requires very small time steps in this case. I think the implementation can be improved (the return mapping on each weak plane may be possible directly via analytical integration due to linearity) both in robustness and speed. But don't have time for that right now. You can proceed with testing first if the basic implementation is correct.
Thomas Nagel (a8e7a4aa) at 21 Feb 17:14
WIP change Dirichlet procedure for Newton
... and 1 more commit
Thomas Nagel (3f10661b) at 21 Feb 14:54
UBI JN fix.
This seems to work. You can start testing. triax_UBI.prj contains an exemplary input. The notebook UBI.ipynb produces the following results (in 2D) by activating 2 joint sets.
Should the model HoekBrownC2UBI3Ortho fail, you can try HoekBrownC2UBI3OrthoAS, but it's less strict (pull back can lead to violation of other YS)
Thomas Nagel (64b9b537) at 21 Feb 14:49
Bugfix
Thomas Nagel (2afaa855) at 21 Feb 14:40
Independent material properties and two versions.
yes, would be nice to have this as a jupyter for the website ...
@endJunction When running triax_UBI.prj in Tests/Data/Mechanics/HoekBrown I get "error: map::at" ... I assume that's related to MFront state handling. The corresponding MTest runs: HoekBrownUBI3Ortho_uniaxial.mtest
Let me know if that's an easy fix or something bigger. Might be related to vector valued states in MFront : @StateVariable real lamWP[3];
If it's difficult, I could make the implementation via three independend variables and more code redundancy.