[Richard Renaud]

I am attempting to model flow in a gravity separation vessel. The vessel takes in slurry through a central feedwell and separates it by density to overflow and underflow streams. I tried to attach a simple sketch and I hope it works.

I would like to avoid the computational complexity of using a VOF model for the free surfaces by modeling them as BCs. My first instinct would be to model them as free slip walls, but when I looked into the parameters of the VOF model, I noticed that it included a drag coefficient between the two phases that the interface is modeled between. Is there a more accurate way to approximate a free surface? Perhaps using specified shear?

I'm developing the model from the ground up, comparing CFD results to a measured velocity field, starting with a water only simulation. The preliminary model using free slip walls did not produce results that I was completely happy with, and I'm looking for ways to increase the accuracy without increasing the computational complexity.

Thank you,
Richard

[ghorrocks]

If you use a pressure boundary set to atmospheric pressure you will get a small flow in or out of the domain, but it will set up the correct hydrostatic pressure (which the wall approach cannot).

[Richard Renaud]

That's an interesting idea. Do you mean to use an opening BC with pressure set to atmospheric?

[ghorrocks]

Yes, that is correct.

[Richard Renaud]

Thanks, I'll add that to my list of ideas to improve the simulation. Something else has come up so that I have to put this on the backburner for a while, but I'll let you know how it works out when I get the chance to run it.