I was wondering why OpenFOAM does not need the parameter and ELMER does. Before merging here, it would be interesting to see whether a coupling with ELMER using the density parameter delivers the same results as other solver combinations do.

From my current understanding (see #226 (comment)), I would propose to provide a density for solid and fluid. For the solid the density is technically a "don't care", but

• this is something we can and should verify (not sure if the Elmer adapter is already far enough for this kind of verification) and
• I would still define a density (even if it is arbitrary), because this will avoid discussions in the future and make it easier for everybody who does not want to dig deep into the technical details.

For this PR this means:

• Leave it as it is
• Do some experiments with densities for more security

From my current understanding (see #226 (comment)), I would propose to provide a density for solid and fluid. For the solid the density is technically a "don't care", but

Agree

For this PR this means:

* Leave it as it is

* Do some experiments with densities for more security

I would vote for the second option because once we merge here nobody cares any more and we might painfully realize an issue later on.

I just added the ambient pressure as an additional parameter. From my perspective this closes the system and allows us to derive the molar mass (if the solver requires it). See #226 (comment) for the derivation.

I think this PR is ready to merge. We could still perform some experiments, but since the derivation is quite simple I don't think that this is necessary anymore. @davidscn could you review and merge/comment?