Create interface to ensure strain formulation consistency in solid mechanics simulations

[dschwen]

Motivation

We currently have no enforcement of consistency for the use of small and finite strain formulations. Turns out we've messed this up a bunch as some objects (like the PressureBC) default to running on teh displaced mesh, which is not appropriate for small strain models.

Design

• Create an interface class with one virtual method getStrainFormulation() that returns the {SMALL, FINITE} enum.
• Have solid mechanics implicitly add an action to every simulation that fetches all objects with the new interface and ensures that on a given block (and its adjacent boundaries) the objects use a consistent formulation

Impact

Better error checking

[maxnezdyur]

Would this error out if I was using a small strain formulation but have a boundary condition that is using the displaced mesh? I can see examples where that would be useful, I am thinking of a solid dynamic simulation of a moving object where the boundary condition has a function term that is location dependent. Is this inappropriate for small strain models since the displacement gradients of the object is still small?

[bwspenc]

@maxnezdyur There are cases where it would be valid to do that. Contact is a good example -- you could potentially do small-strain contact. This is going to be tricky to get right.