problem when simulating overflow at a sharp-crested weir
 

Hi guys:

I am trying to reproduce the overflow at a sharp-crested vertical weir but experiencing problem. The computational domain is sketched in the attached picture. When the overflow jet hit the right bottom, the air region (identified as D in the pic) enclosed by weir, water jet and ground will be disconnected from the atmosphere in 2D case. At the mean time, the jet will split into two part basically, one of them (the left-going part) will climb the weir and collapse finally. The problem is, the total area of the closed air region D will not be likely to change, but as the water increases, the area occupied by air will have to decrease, but this is not quite possible neither for interFoam nor physical facts. Therefore, I changed the boundary condition of the edge I to pressureInletOutlet for U and Totalpressure = 0 for P. And I applied non-slip wall condition on all other wall surfaces. It seems doesn't solve the problem and the solver will always crash soon after the moment the snapshot shows. Please help to provide some suggestions on addressing this problem, many thanks!

I have a related Problem, in my case the LES turbulence can't handle the air at the overfall. Could you get it fixed?

I am confused -- presumably the top boundary is open with an 'atmosphere' type BC. What is happening on the right surface? If you have set this as a wall, then I agree with your analysis and you effectively have a 'tank' that will certainly fill up and push air out the top. Shouldn't the right side be open as well for the liquid to flow out?