Fix/old staggered fluxes

Open Mathis Kelm requested to merge fix/old-staggered-fluxes into master

What this MR does / why does DuMux need it:

In the old-staggered implementation of Fick's law the fluxes at the boundary use only the (molar) density of the insideVolVars, while other for other discretizations use the average. We have decided that when replacing a component balance with a total balance it should always be the total mass and never the total mole balance, even if useMoles is true.

With these two changes the results of the test freeflow/navierstokesnc/densitydriven agree with those of the portet test in !2986.

Furthermore the enthalpy fluxes due to molecular diffusion compared the sign of the flux to the scvf direction for upwinding purposes but we always compute fluxes out of the scv (fickslaw.hh).

Notes for the reviewer

TODO: insert text here

Before you request a review from someone, make sure to revise the following points:

Edited by Mathis Kelm

Merge request reports