I am a total newbie so please understand... Trying to simulate sprays. So I set up a simple box 1mx0.6m 2D domain as sprays don't travel that far with a surface opening at the top at 0.1m. Inlet velocity =1m/s entering at the surface opening. I set turbulence model k=1.0, e=0.5 This works fine, and then I try vel = 20m/s and get reverse flow. Do I have to extend the domain or change to unsteady flow to simulate the spray.

I think before to make unsteady flow, try to extend the domain in the downstream and for precison change the options of k-e model. You use now k=1 (Default value by fluent) so try to use (I and Dh ). For I use an estimate value from the expression in doc of fluent ( Imax=10%) and Dh = 2* Length in you case 2m. And let us know if it will be OK

extended the domain Dh (Hydraulic Diameter?) = 2. That actually made it worse than the original k-e settings.

It converges at k=1, e=0.05. Whenever I reduce the opening i get reverse flow. Is it because the flow is entering an open flow field causing reverse flow which rebounds off the adjacent wall (wall parallel to the opening)

geometery

--------ooo--------

I I

I I

I I

I I

I I

I I

-------------------

I hope you know to set different Dh depending on the hydraulic dia of pipe inlet/outlet. for example, in your case, spray hole dia is, 0.1m or so. Means, you can't define Dh=2 there. -US

Also, I is aproaximately 4% for Re=50000. So, adjust that value accordingly.