Discussion Forum

Hi there,

I have experience dealing with CFD problems with the finite volume method. I use my own code, starccm+ and openFOAM very often.

I am trying to make some animations with a multiphase flow for the breaking dam problem (a tutorial from openFOAM). I find very difficult to convergence these solutions and almost random. (The problems divergences easily)

www.openfoam.org/docs/user/damBreak.php a example of the tutorial

Notes:

- I can only convergence if the mesh is triangular, if I used mapped grid it never convergences.

- The gravity must be used with a ramp function in time.

- I used pressure constrain point and bernoulli function for the initial value of pressure.

- I added a outlet in small portion of the square doamin bottom (case without obstacle). When the water falls in this boundary it just bounces back. I tried the very options of outlet for testing. Sometimes this case divergences before the water even reaches the outlet boundary.

- I reduced sometimes the discretization for P1+P1 but does not seem to solve the problem (as better convergence properties).

- I refine the mesh everywhere for a case that worked previously and the case divergences during computation.

Does anyone have similar difficulties? This physics seems to work only for the tutorial case. I need some advice on this model with comsol.