Add some extra check

[JiZur]

It would be good to have an initial check that overburden pressure >= fluid pressure and overburden pressure >= reference pressure (i.e effective pore pressure and effective reference pressure are positive!) with understandable error message. That would avoid the current case where Hertz-Mindlin equation crashes due to "invalid value encountered in power function" (due to negative pressure, which is not said) with an error message which does not make the debugging easy.

Since many users use the patchy cement model for dry rock with pressure correction, they neglect to input a reasonable reference pressure (as it is later overwritten by the pressure adjustment), resulting in the intermediate step of dry rock at reference pressure propagating NaN value due to the negative pressure.