INS: need to check boundary conditions

[tcew]

The bc handling may need attention in the INS solver. Some elements might be contributing Neumann contributions even if they are interior and only touch the boundary at an edge or vertex
I couldn't quite trace the logic and there were some commented out ogs calls so it likely that a comprehensive solutionhas not been fully fleshed out.

I think two distinct things have to happen in the construction of the volume bc maps - one for Dirichlet and one for Neumann.