[russel60]

Hi everyone,

I have been developing a fluid solid interaction (FSI) code with a compressible fluid solver with cavitation, and I am looking for some advice on proper boundary conditions to use.

The flow that I am simulating is entirely viscous driven - there is no set pressure differential or across the domain, and the only fluid movement comes from two moving walls (one translatingWallVelocity and one rotatingWallVelocity).

I have done very similar cases with an incompressible fluid solver, and have had success with setting the inlet as a 'zeroGradient' and outlet as 'fixedValue' of 0. I cannot use these boundary conditions for the compressible flow case because of the cavitation model - the minimum pressure when accounting for cavitation is the liquid's vapor pressure (around 4500 pa).

So far, I have tried using two 'totalPressure' boundary conditions that set both the inlet and the outlet to 1.1e5 pa, around atmospheric pressure. I have also tried setting both the inlet and the outlet to 'zeroGradient'. Both of these cases gave very similar results - very little pressure build up in the narrow region, and a large pressure differential between the sides. I have attached an image of my domain with the pressure contours from the dual 'zeroGradient' boundary condition case.

I am wondering what the best boundary conditions would be for this case. Reading through the B.C. documentation on openFoam.com and DOxygen, I think I might try using a 'freeStream' condition or a combination of 'inletOutlet' / 'outletInlet' conditions.

Does anyone have any advice / suggestions for what to try next? Any help is appreciated!!

Thanks!

[russel60]

Update for anyone who finds this thread:

It turned out that having the two outlets have a totalPressure boundary condition set to be atmospheric pressure worked for this case. The max pressure in this case comes at the very narrow region, so setting both outlets to be at ~1atm worked. I must have been doing something wrong when I tried it and it didn't work the first time.