Hi, we have a system consisting on air and water inlets into a t-junction. The resultant biphasic mixture flows in a pipe into a cubic cavity with a pressure outlet b.c at the opposite face. Diameter of pipes is of the order of mm while side of the cube is 20cm. We're using mixture and k-epsilon models. With turbulence intensity around 1% and hydraulic diameter 1mm we've got convergence problems and error message about reverse flow at the outlet. Do you have any idea about where does the problem come from? Thank you.

It is not an error. Reverse flow at the out may be due to high back pressure.

I'm not sure I understand your geometry, but the backflow and the convergence issues may be the same issue. In my experience, the backflow is usually due to the boundary being too close (to the region of interest) and the backflow causes convergence issues. I would try moving the outflow further downstream.

Jason

I would try to reduce turbulence intesnity. And wait for some time to converge. If not , I would again "initialize" the solution. If there is no other source if error, it should converge.

with a mixture model being used, i suggest tht turbulence intensity is not realistic in this case. also flow from 1mm hydraulic dia (a micro channel)into a 20 cm cubic region??? i think ricard, u need to make sure of the dimensions u r using to model the flow