[Blue8655]

Hello Foamers,

I have been stuck with a simulation of mine for a few days now. I am trying to simulate a heated small 3D cavity with buoyantSimpleFoam (OF-3.0.1). Cavity is small (0.4m x 0.15 x 0.6m) There is one inlet and one outlet, and a uniform heated floor (see image). Reynolds is around 7100, and there is a big recirculation inside the cavity. I built the mesh with blockHexMesh. It is 415,000 cells. Note that in my particular variant, the inlet is full width of cavity (w = W).

It seems to work during first iterations, however a pressure instability appears in the corners and eventually causes a crash. I did stumble upon a mesh once where the oscillation eventually disappeared...but I want to understand...

I used to simulate this by adding a tube inlet and using wall functions. However, I dropped the wall functions to improve accuracy (the grid study was not giving good results although some meshes did find a very accurate answer, but not the finest ones) (Limane et al, 2015) did reproduce results from (Nielsen,1976) by not adding any inlet, just a cavity. So I thought I could do it to. But imposing uniform U (10 to 15 m/s) on the inlet seems very hard on the solver. I understand that some cells are imposed 15m/s while touching a boundary with zeroVelocity but yet... some people succeed.

My y+ post-treatment seems to have let me down for reasons yet unknown, but they are low (<10). I would expect poor results but not a crash at iteration 50.

I joined my OF case, removing the mesh to make it small. Any help appreciated. Tell me if more info needed.

(Limane et al. 2015) "Thermo-ventilation study by OpenFOAM of the airflow in a cavity with heated floor"
(Nielsen,76) Flow in air conditionned rooms.

Fred