Hi my name is joseph, I am currently undertaking a computaional model on a swirling conical diffuser with 20 degree angle. The swirl is sufficiently small to avoid separation. My main problem rise when the calculation start safter fue iterations fluent tells me that there is reversed flow X faces. I would like to know if there is anyway of avoiding this situation i read an article made by armfield saying that they did have the same problem and used a novel way to treat the pressure. they pointed out the fact that the axial velocity is nearly reversing at exit and the turbulent kinetic energy is very low at the centerline. I wonder if any of you had this problem.

I thank you in advance for any kind of help you could give me.

In general, there is 'reverse flow' message because 'there may be reverse flow in real situation'.

I have many situations of 'reverse flow'. Sometimes, the message disappears as iteration continues. Otherwise, the only way, not to see the message, is to extend the computational domain, sometimes, including very wide range of external region.

The other way is, accept the result if it converges with appropriate backward flow boundary conditions.

Sincerely, Jinwook