porosity calculation is way off for LNE; also inaccurate for MCC

[jacklinzoho]

I'm running triaxial compression for Oka CD5.

Currently Lewis is calculated using _total_porosity (search for "// Update Lewis number" in redbackMaterial.C), but this varies a huge amount from the volumetric strain. Especially bad for LNE where _total_porosity is actually increasing under compression.

I'm not sure if it's overwritten elsewhere, but it might also affect compressibility and thermal_pressurization (also in redbackMaterial.C).

Probably easiest way to fix it at least for triaxial is just set _total_porosity = _p0_param - volumetric_strain? Not sure how that affects other things though.