Hello everybody,

I'm simulating some kind of mixing tank which is filled with water (70%) and air (30%). Solution using sliding meshes is OK but needs a lot of computational time to achieve stationary solution. In order to get first impression of the flow field Moving Reference Frame (MRF) method can be used. In my case transient interaction between moving and non-moving parts is week. The problem is as follows: when i set up simulation with only one phase (water) with MRF solution achieve convergence without any problems. Using two-phase implicit VOF model leads to some strange problems, i.e. volume (and also mass) of the water is increasing and decreasing. At the begging I have 223 l and at the end 300 1, whereas total volume is about 500 l. Simulation settings: PISO, PRESTO! for pressure, Second Order for Momentum, Modified HRIC for VOF. All under-relaxation coefficients set to 0.2 (recommended in FLUENT User's Guide). I have made many attempts with different algorithms and discretization schemes. Seems that implicit VOF does not conserve mass. Any suggestion? Or maybe I schould forget about MRF? Thanks in advance.

You should forget the MRF model. It does not work with VOF.

Thank you for your response. Your opinion is as I expected.

MRFndoesn't work with VOF?why?i didn't know! But did you try to change you axis of rotation in the MRF?