[30na]

Hi everybody,
I am modeling a 3-D flow in a rotating CD using openFoam solver MRFInterFoam. The results seem to be reasonable up to a certain time but after reaching an unknown time the time step decreases dramatically and then increases. And while viewing in parafoam the results are logical up to that point (the fluid moves outward because of the rotation), but then the fluid shows a strange manner and kinda explodes all over the channel losing it's integrity.
I noticed that this also happens in 2-D. And happens only when the capillary force is active. When there is no capillary force the results are as predicted!
So can anyone help me? Has anyone encountered a simillar situation?
What is wrong with capillary and rotation?( because i have used capillary before with no rotation and had no problem)