Major update in grid deformation subroutines

[ghost]

Several bugs have been fixed in the grid deformation and the config file have been updated.

In particular, the inverse volume method uses now an element volume computed in ComputeDeforming_Element_Volume (not the incomplete determinant computed in FEM). fabs have been added to the determinant in the final evaluation of the FEM stiffness matrix (depending on the element orientation, the determinant was negative). The formulation Mu = E / (2.0_(1.0 + Nu)), Lambda = Nu_E/((1.0+Nu)_(1.0-2.0_Nu)) is used instead of the basic simplification Mu = E, Lambda = -E. We can recover that formulation using a large number for Nu (DEFORM_COEFF). The code has been simplified and that IO is now more clear.

I have performed some tests and large deformations are now possible.

However, systematic regression tests for grid deformation is required using all the different kinds of elements. I remember that T. Economon put together some basic grids to test the mesh deformation, we should use those to check that everything is working fine before closing the release.

[economon]

Thanks for the fixes, Francisco. I'll be working to fix up the failing tests. I still have the old meshes you mention, so I can add those to the tests that we already have for mesh deformation with the RAE 2822 and ONERA M6.