[lcbuijs]

Dear Foamers,

I'm trying to implement variable density and viscosity in buoyantPimpleFoam (bPF) for a simple Rayleigh-Benard problem in a cubical cavity. My simulations are working fine for buoyantBoussinesqPimpleFoam (bBPF), but switching to bPF causes some problems. See my previous thread for a more detailed explanation.

I've tried changing the standard hotRoom tutorial for bPF to my own problem step by step, and I've run into something that surprises me. Running the standard hotRoom without a temperature gradient, generates physically significant (~20cm/s) velocity values. This seems very strange to me, because I don't see any other forcing active. Running the exact same tutorial in bBPF results in a nice 0 velocity field, as I would expect. Trying to trace back why this is happening looking at the UEqn-file in bPF, I figured this is coming from the values of p_rgh, which are quite large. I don't have a physical explanation for this though. Can anybody tell me what is happening?

Just for clarity: this is the completely standard hotRoom tutorial, copied from the bPF folder in OF.

Many thanks!
Luuk