Open mangerij opened 6 years ago
Can these high order tensors just be put in the phase_field module? The reason we have rank 2 and 4 in MOOSE is because those tensors are used by both mechanics modules and phase_field. They are pretty general. I can see why you might just want to lump all of these tensors together but it's my job to always try to keep the framework from taking on too much bloat. If these belong in a module, lets think about that instead.
They can go in modules 👍 Gives me a reason to keep the dependency on phase field. I'll do this eventually
@permcody and @mangerij can we keep the Rank-3 tensor in framework?
@sapitts - Are we using Rank-3 in two modules? Tensor mechanics and PF?
....not in currently checked in code, but we have some devel work with the Rank-3 tensors in tensor mechanics crystal plasticity. At some point I will get said code checked in, too
Yeah, I'm not sure where it is used in the moose repo but we use it in Ferret (and its covered by one of our tests). I don't care where it lives, but I think rank 2-4 stuff is generic enough to stay in framework.
But I agree with Cody's suggestion to dump rank 6 and 8 to PF.
Rationale
In Ferret, we take a "discontinuous" approach to the grain boundary or material interface that may arise in a structure. In other words, our grain structures or interfaces between two materials uses the block restriction system. Changing the orientation is simply a rotation of the materials constants (C_{ijkl}, etc). For elasticity, rank four is sufficient.
However, for the ferroelectric material, the Landau order parameter expansion of the free energy is out to sixth and eighth powers depending on the material. It would be nice to be able to have the capability to study polycrystalline effects.
Description
Need two new classes in /utils
Impact
New support for new capability in the Ferret application.