[Dano62]

Hello,

I am trying to model an outlet boundary which functions similarly to a loss coefficient. In other words, the pressure drop at the boundary is proportional to the velocity squared i.e. where C is an experimentally determined constant. This boundary condition has worked well for me in ANSYS for simulating the flow through an orifice beneath a free surface, but I haven't been able to find an equivalent boundary condition in OpenFOAM. Any help would be much appreciated.

[piu58]

I don't believe that you think in the right direction. The pressure loss is a function of the flow in it's geometry. It can be calculated, but not used as a boundary condition.

It may be that from a theoretical view of point this pressure drop has to be expected. But you cannot use it as bc. If you give two pressures as bc this will change the flow in each time step, mostly increasing. Often such simulation blow up because the flow is not limited.

In your case the rise of the flow may be limited, but if you don't have the fortune of finding an exact balance between your flow and the b.c. the flow moves to anywhere but not to a physical acceptable result.

[Dano62]

Thanks Piu58. Perhaps I should rephrase description. I'm looking to set a boundary pressure to a function of velocity. To properly simulate the orifice pressure drop would require substantially more nodes than I can spare.

As an example, if I have an orifice at the bottom of a bucket that's discharging into the atmosphere, I can calculate it's pressure loss experimentally as a function of its velocity. I would then use this function to specify the pressure of an outlet boundary. I suspect that this may be doable if I use the results from the previous timestep.

In my case, I have numerous orifices spread around the bottom of a trough. Fully resolving these orifices and the air beneath them is not feasible.

Edit: I'm looking for a boundary condition comparable to Fluent's Outlet Vent Boundary condition.

[piu58]

If I understand you right you want to have some kind of pressure resistance at the oulet, where the resistance itself is velocity dependent?

I think this may be implemented with a Robin b.c. which contains a linear connection between velocity and pressure.