Combatting reversed flow
 

I'm wondering if any of you have any suggestions for combatting reversed flow that is occuring on one of my faces (pressure-outlet). My setup is a small 60 m/s jet (16 square-centimeter cross-sectional area) blowing into a rather large test chamber (6 square-meter cross-sectional area), large enough that the turbulence and energy from the jet should dissipate by the time the flow reaches the chamber outlet (which is about 3.6 meters away from the jet). The velocity contour plots show that there's very little flow velocity at the outlet (on the order of 0.01 m/s). But my problem is that, as Fluent marches through time steps, I'm getting reversed flow on more and more outlet faces; by the end of the iterations, half of my 6000 faces are experiencing reversed flow. This cannot be.

So I'm thinking that some numerical error is creeping in and am wondering if you all know some tricks (like changing some of the solution controls) that could help to fix this seemingly unstable solution.

Thanks!

Tucker

Re: Combatting reversed flow
 

If you do not expect reversed flow in your solution, you may try to shift your boundary farther away. it may so happen that you have internal geometric details that allow reversed flow or your outlet is smaller than chamber so that jet is hitting chamber walls. I read somewhere that using zero pressure at outlet stops reverse flow after some iterations. I do not know how far this is applicable to your problem. Also, there is nothing wrong with reversed flow when you are using pressure outlet. Alarm should be sounded however when you find reversed flow in outflow boundary condition.

Swarup.