Lagrangian mechanics kernels with correct large deformation Jacobians

[reverendbedford]

Reason

The Jacobian for the current StressDivergenceTensors kernel is wrong. Moreover, it would be nice to have a large deformations option that doesn't use_displaced_mesh (i.e. a total Lagrangian option).

Design

We should have:

• Mechanics kernels that have correct Jacobians for large displacements.
• Options for both updated and total Lagrangian formulations.
• Compatibility with the existing tensor_mechanics material system.

Impact

• At least for now I suggest adding new kernels, not replacing the current ones, so that we can gradually build up the functionality until we can replace the current StressDivergenceTensors which cover a lot of options (cylindrical and spherical coordinates, beams and bars, etc.).
• So the new mechanics kernels would supplement those in tensor_mechanics.

[dschwen]

Yeah, for example the derivative of the Jacobian weights wrt the displacements is missing. I tried adding that at one point, but never finished it, and then came... automatic differentiation :-)

[dschwen]

Also, I want to point out ADComputeGreenLagrangeStrain, but that might not cover your needs.