problem with convergence
 

Hi Everybody,

Well my problem may be very simple to solve, but i'm currently out of solution...

I'm running a calculation concerning air curtains, the focus is put in the centre of a large domain, where a jet is blown and then aspire. I use the K-E turbulence model.

I put a pressure outlet condition on the external boundary (pressure gauge=0), where normally "nothing" should happen, but after several thousands of iterations, there is still the message "reverse flow" for this area and when I check the turbulence rate, I can see that there is a lot of turbulence in this zone. Moreover my calculation doesn't converge very well (max. 10-3 for the residual of the continuity equation)

Does anyone know it be due to? Is it a problem of boundary condition? Of grid (i used thetrahedrical elements)?

Thank you in advance for your help!