[sudheer717]

Dear Foamers,

I am trying to use open boundaries where my flow can leave the boundaries but restricted to backflow. For that I am using inletOutlet velocity boundary condition and fixedFluxPressure condition for p_rgh.

I can see unwanted/unrealistic velocities popping up somewhere in the boundaries right from the first time step onwards. These are increasing my Co num and so the time step is falling drastically. I have read many reports on this kind of error but unable to find the main cause or a solution.

Does any one know why this kind of velocities occur? Is it because of boundary conditions or is it just random ?!

The attached files are from two different cases.

Thank you
-Sudheer

[sudheer717]

Well, after spending some time on this issue, I learned that this is caused by boundary conditions + initial time step size.

I tried to ramp-up the simulation with very small initial time step values (order of 1e-6) and once I see the velocity build up near the inlet region I increase the time step and eventually impose the Courant number criteria.

One more thing I observed is, if I have wall boundaries the time spent in this ramping process is a bit large than it would take when I have totalPressure or atmospheric boundaries.