vertexCentredNonLinTotalLagGeometry
This solid model is currently disabled and does not compile into the library. See the commented-out entry in src/solids4FoamModels/Make/files.*.
Why
No tutorial in the repository selects it, so nothing has ever exercised it. An attempt to give it a regression case, on a hyperelastic cantilever beam with a neoHookeanElastic material, did not succeed:
useGeometricStiffnessandcompactImplicitStencilare looked up with no default, and no case sets either.- The material tangent path needs
tangentEpsin the material sub-dictionary, which no tutorial in the repository sets. It is an absolute perturbation, and so has the scale-dependence problem that the equivalent in the mechanical constitutive law framework was fixed for. - The scalar Jacobian arm additionally needs an
interpolate(impK)entry insystem/dualMesh/fvSchemes, which no case provides.
With all of those supplied, the run still fails with a floating point exception inside the PETSc solve, on the first Newton step.
What would re-enable it
A case that runs it, and whatever fix the floating point exception turns out to need. This is independent of the mechanical constitutive law framework: the model was already in this state before that work began, and the framework neither caused nor depends on it.
Its linear-geometry sibling, vertexCentredLinearGeometry, is unaffected, compiles, and is covered by the cantilever2d tutorial.