Hello,

I am working on a two-phase-flow problem (gas/liquid with similar properties as air/water). I have one outflow boundary, where only one phase is a allowed to pass (for the other phase it is like a wall). Orthogonal to that I have an outlet for both phases (distance between the outlets is about 3 mm (or 3 cells). The flow is incompressible, laminar and steady. I think I should use zero-gradient boundary conditions with fixed pressure correction at the last interior cell row (at one outlet? at both outlets?). Does anyone have suggestions or experience with that kind of problem?

Thanks. Christian

Hi Christian,

I don't know anything about two-phase flows. Anyhow, if the flow is incompressible, and if you have several outlets, then the average pressures you enforce at each outlet will directly drive how the inlet mass flux is splitted onto the several outlets. Convertly, if you would enforce the mass flux at each outlet (which could then be considered as inlets...), it would drive the pressure you would find afterwards at these "outlets". That sounds quite physical, doesn't it?... but the mathematics say the same thing (of course)

As for the actual BC you enforce for the pressure, I think it should be of Neumann-type. Remains the problem of the ill-posedness of a pure Neumann problem (see one of the few previous thread). Enforcing the mean pressure on the(each) outlet(s) aleviates the problem.

See the following papers:

Heywood, Rannacher, Turek, "Artificial boundaries and flux and pressure conditions for the incompressible N-S equations", IJNMF, 22, 325-352, 1996.

or

Bertagnolio, "Solution of the incompressible N-S equations on domains with one or several open boundaries", IJNMF, 31, 1061-1085, 1999.

Hope this helps. Regards,

franck