Backflow
 

hello all I want to know how to impose a pressure outlet boundary condition without backflow. Rgds

Re: Backflow
 

I don´t know if this is possible with a pressure outlet but if you really want to avoid backflow, you can take a negative velocity inlet ( but only if its an incompressible fluid ).

Re: Backflow
 

If backflow occurs only in the intermediate solution, take it easy. Otherwise, pls check the pressure setting of your boundary condtion.

Re: Backflow
 

it seems impossible to eliminate backflows in the current Fluent versions if the fluid is imconpressible. Non-reflecting boundary conditions may be applied to compressible flows.

Re: Backflow
 

I have understood that using a pressure outlet boundary condition without backflow is impossible. I have tried to use outflow boundary condition, but the pressure at the outlet was not reasonable. So, who can tell me how to impose pressures with a given flow rate. Thanks

Re: Backflow
 

No it is not impossible - it depends on your geometry. Just at the beginning of a calculation it is very hard to avoid it but when the flow conditions are stable why not ? But as it was often discussed here if the result of the outflow- area is important and you get a backflow then you have to expand your calculation area to regions behind the outflow so that the `new´ outflow there does not influence your `old´ one you are interested in. hope that helps

Re: Backflow
 

Hello,

How can you say it's not reasonable? Do you have experimental results for the case?

I don't think you should think about imposing the elimination of the backflow, as it comes as the result of the simulation you'd set. If you are sure that this backflow is unreal, you should look for problems in your simulation, and not just "impose" the solution that you think is real.

My best regards, Victor

Re: Backflow
 

The following note is mentioned in the Fluent doc

Even if no backflow is expected in the converged solution, you should always set realistic values to minimize convergence difficulties in the event that backflow does occur during the calculation.

In the literature no back flow has been mentioned in all experimental works concerning the subject I am studying (hydrocyclones). What does "realistic values" mean in the note above. Thanks.

i have meet this problem.i set smaller gauge pressure at pressure outlet boundary condition and get rid of backflow.

Hi Khar Nab,

Is it possible to post a pic of your model? Can't advise specifically where to look at without seeing your setup.

Just for starters:
1. The backflow may indicate turbulence in that area; you can extend the domain to see if this is the case and whether it dissipates.
2. the CFD solution is strictly based on the condition you've applied. I'll advise care when forcing the flow regime; easiest way is to set the boundary as an outlet instead of opening (ur flow will be uni-directional).
3. Is experimental/empirical data your validating against, flow-field or measured results?
4. What turbulence model are you using? =)
5. Hard bit (optional): Run your simulation in steady-state, and then transient for 1 sec. Post the output plots of variables, including the wall and turbulence plot for comparison. It'll help to see if your solution's stable.

-dazza

backflow
 

hi all,

ı have an rectangular geometry (lenght:76, width:27 mm). heat flux is applied from one of the vertical walls. ı have velocity inlet BC at lower part and also pressure outlet BC at upper part. velocity magnitute is zero (0) (no forced convection), initial gauge pressure is also zero and temperature is 298 K at velocity inlet BC. also, for pressure outlet BC gauge pressure is 0 and backflow temperature is set 298 K. initial temperature is also 298K.

the problem is that reversed flow occured.
"reversed flow in 34 faces on pressure-outlet 17"

the solution is converged but ı am not sure it is right or not.

any suggestion. thanks for you time.

Asking the same question all over the place in multiple threads is very annoying, not helping your cause at all, good luck.

sometimes backflow is unreal in problems. its there is for invalid bc in outlet! for example in outlet pressure gradiant is not zero but used from "pressure outlet" in this problem you can increase geometry lengths(adding virtual length) in outlet but in reporting compute pressure in initial(real) outlet!;)