CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Siemens (https://www.cfd-online.com/Forums/siemens/)
-   -   Recirculation problems (https://www.cfd-online.com/Forums/siemens/54275-recirculation-problems.html)

Simon December 30, 2004 16:05

Recirculation problems
 
I am simulating a simple diffusion flame. I am using a 2D square mesh. One side of the square is the two inlets (fuel stream in the center, the rest is the oxidizer stream) and the opoosite side is the outlet. The velocities of the streams are O(cm/s). The problem is that I don't get a converged solution, because I keep getting inflows at the outlet boundary, along with recilculation regions. Could the boundary conditions be faulty? Any suggestions are more than welcome.

Brian December 31, 2004 01:35

Re: Recirculation problems
 
0cm/s? Zero? That would be a problem indeed. Perhaps a typo?

However, it is entirely likely you would get recirculations with such a setup - and the associated inflows on the outlet. Try extending the length of the mesh or contstrain the outlet to a smaller area.

Simon December 31, 2004 12:59

Re: Recirculation problems
 
Yes, a typo indeed. I meant that the velocities are order of cm/s. The problem is that I tried extending the mesh length-wise, but the recirculations persist. If the required increase is more than, say, three times the given length, then there has to be some other way.I was wondering whether I could impose some sort of velocity profile at the outlet (steep hyperbolic, perhaps) but STAR-CD doesn't allow something like that, as far as I know. Any ideas about which boundary is the proper one in such a case? Thank you in advance.

Brian January 1, 2005 05:50

Re: Recirculation problems
 
You most certainly can impose a velocity profile at the outlet, using boundary type inlet and a table or user coding to specify the velocities.

But it could be tricky balancing the mass flow closely enough.

I would suggest a test case where you make the outlet boundary smaller, like just a portion of the existing boundary. This may not be the case you want, but it would most likely allow a sufficient degree of convergence for you to see what is going on.

The other alternative is to use a pressure boundary at the outlet, but this has its own stability related issues.

Vladimir January 1, 2005 06:40

Re: Recirculation problems
 
Hi Simon

I noticed that extendsion of the domain did not help so what I would do is get rid of the type of BC at the outlet you used (I mean OUTLET). I would recommend you to switch the OUTLET to PRESSURE bc.

Hope it'll help.

Simon January 1, 2005 18:58

Re: Recirculation problems
 
Thank you everyone for the suggestions. I will try them out asap. And a happy new year to you all.


All times are GMT -4. The time now is 10:40.