CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Vorticity plot in grid refinement study (http://www.cfd-online.com/Forums/main/13498-vorticity-plot-grid-refinement-study.html)

CH May 21, 2007 19:38

Vorticity plot in grid refinement study
 
Hi,

I am trying to a grid refinement study and I tried to compare the different vorticity plot of a moving airfoil at the same instant using successive higher grid resolution.

So when can I stop? As I go higher up, the main vortex are captured but some smaller vortex appeared at different places when I increase the resolution. Do I have to make sure they all appear at the same location to prove that my grid is sufficiently refined?

Thanks.


to May 22, 2007 04:50

Re: Vorticity plot in grid refinement study
 
Hi

Most probably, you will stop when you will not have computational ressources anymore, or when you will not have time enough to use another mesh.

The limit of your grid refinement study is the full DNS. I hope your case is not a high-Re flow ...

Except if you use a turbulence model, that will probably inhibate the appearance of small scale vortices.

Regards

CH May 22, 2007 12:21

Re: Vorticity plot in grid refinement study
 
Well, it's mid level, Re~1e4.

the lift/drag coeffcient doesn't change as i reduce the y+ level so it's kept constant. i wonder if it'll affects the vorticity plot...

Mani May 22, 2007 12:35

Re: Vorticity plot in grid refinement study
 
Something tells me you already have your own suspicions that the purpose of a grid resolution study is not to find the resolution above which the vorticity plot doesn't change. You obviously understand that this is not going to happen until you reach the smallest turbulence scales.

The fact that the picture keeps changing doesn't mean it's not converging, though. If everything goes well, the changes appear on ever smaller scales.

A grid resolution study is applied to find the convergent solution to your problem. You need to know what the parameters are that you are interested in (lift, drag, ...). Those are the parameters you should observe for grid convergence. If the vorticity field itself is the subject of your studies, you need to define a scale of interest.


All times are GMT -4. The time now is 18:05.