[Artur]

Hi

I want to explore the capabilities of interPhaseChangeFoam and for that I created a 2D test case based on NACA0012. I have successfully ran the very same case (same mesh, same settings, etc.) using simpleFoam, pimpleFoam and interFoam, all of which yielded very similar results in terms of flow features and forces.

However, when I run the case with interPhaseChangeFoam, the force values are completely wrong (Cd and Cl approx. twice as big as they should be). I set the reference pressure to a very high value (5e8 Pa) by using a fixedValue BC at the outlet to ensure there is just water in the domain and no cavities are present, which is indeed the case (alpha1 field = 1.0 everywhere).

Here's what I've tried:
- using different schemes on U, k, omega: (Gauss linearUpwind grad(U), Gauss upwind)
- changing the reference pressure values
- changing the transportProperties based on the (admittedly, a bit vague) guidance given in this forum
- modifying the fvSolution file (changing tolerances, no. various correctors)
- changing the Courant number (in range from 0.5 to 2.5)
- having done some research on this forum I also compiled the versions of the solver from 1.6-ext and 2.1.0 but they yielded similar results
- changing the two phase model (SchnerrSauer, Kunz, Merkle) - again, no change, although the Merkle model was prone to crashing

A bit more about my setup:
- k-omega SST model
- OpenFOAM 2.2.0
- Re 3e+6 (chord 1 m at 30 ms-1)
- fully turbulent flow assumed (I know this is not quite true but the results between different solvers did agree pretty well despite this)

I would really appreciate any guidance as to what I could be doing wrong.

I am not sure where the problem is and the case is too big to be uploaded in one go so please tell me and I will post any bits you think might be relevant.

Thanks in advance

-------
EDIT: I managed to squeeze some stuff and fit the case within the allowed limit. Please see the attachment.

EDIT 2: Just to clarify something. In the attached case I first solve using simpleFoam and then map the converged solution as an initial condition to the interPhaseChangeFoam to speed up the convergence. Of course I tried without doing it, i.e. starting from a "blank" solution, with the same end results.