Backflow in Open Channel VOF simulations

Kalous

Hello,

I am simulating a side-weir combined sewer overflow using VOF model. There's an inlet sewer with free surface flow and two outlets - a throttle pipe and an overflow sewer (for water that overflows the side weir, free surface flow). Air should be insignificant. It is supposed to flow from inlet direcly to overflow sewer. However, I cannot get rid of backflow of air from the overflow sewer into my computational domain. It is not just a few cells near the boundary. Air blows all the way in. I have tried several boundary condition settings without success. I suspect the problem to be in the way Fluent treats hydrostatic pressure.

Description of my setup follows. Thanks for any suggestions.

Kalous

My setup: grid â€" mostly hexagonal, small tetrahedral regions models - unsteady, standard k-epsilon, VOF (primary phase = air, secondary phase = water) operating conditions: gravity enabled, operating density not specified (I also tried to specify operating density as the density of air, but the problem persisted.)

inlet BC - mass flow inlet (water+air), open channel throlle pipe BC - pressure outlet overflow sewer BC - pressure outlet, open channel (I have tried pressure outlet with 0 pressure and outflow BC as well, but the problem persisted.)

solver: coupled scheme pressure - PRESTO! volume fraction â€" Geo-Reconstruct other â€" 1st order upwind