reversed flow
 

hi during solution i found reversed flow in some faces still my program runs and residual for continuity is of the order 1e-2. so can it be harmful or misleading the solution? how can it be cured?

thanks in advance

Re: reversed flow
 

Try to avoid backflow in areas you want to know the flowsolution. If the in/outlets that causes this backflow are far enough away then don´t worry. If they are not than enlarge the calculation area. P.s.: Sometimes the reversed flow disappears after some calculation time - I would try to get a better continuity than 1e-2. hope that helps

Re: reversed flow
 

Hi,

I too get reversed flow warnings on a few faces on my pressure-outlet boundary. Problem is that the residuals for continuity, x, y and z velocities oscillates / almost stabilizes around 1, 0.01, 0.01 and 0.1 respectively. How do I improve the residual values? I have checked the flow conditions and they appear reasonable. Do I need to change the relaxation parameters or make an improved gauge pressure estimate during initialization?

How do I enlarge the calculation area? The backflow outlet is a region of interest as I would need the pressure drop across the inlet and outlet boundaries.

Please advise.

Thanks, Vivek

Re: reversed flow
 

To enlarge the calculation area is easy ( in most cases ). Where does your fluid go after leaving the outlet? Apend another volume to simulate this and make this volume that big that the backflow at the new outlet(s) does not influence your old one. Set the old outlet as interior and you can get all the values you need out of it. I know that this is also a question of cell quantity and there are always facts that might make this impossible but i think in most cases it works. If you really make it like i told you I can help you to save time for the new calculation by using the `old´ results as starting conditions.( if you´r interested in ) mfg Gernot

Re: reversed flow
 

Attach similar extension volumes to the inlet and outlet boundaries, set-up the case, "INTERPOLATE" the data from previous case and re-run the case again. Right Gernot ?

;-) Nak.

Re: reversed flow
 

Yes thats exactly what I mean.

Re: reversed flow
 

hello: this case is normal in flow pattern and is not harmful.

Re: reversed flow
 

ya but after running so many interations still revrsed flow message exists and problem does not converge. my momentum underrelaxation factor is 0.3.