Possibly wrong gravity implementation for cctpfa
The current implementation of the gravity term in flux/cctpfa/Darcy's law.hh seems to be unsuitable for problems with heterogeneous distribution of petrophysical properties unsuitable on grids with convex cells (e.g. corner-point grids) which can yield negative transmissiblities (edited Timo). Previous versions (<3.0) have a more straightforward implementation and seem to work fine. Potential Solution: darcyslaw.hh (old Darcy's law implementation but incomplete e.g. on network grids)