CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Avoiding Skew in Tight Corners (http://www.cfd-online.com/Forums/fluent/45427-avoiding-skew-tight-corners.html)

Marc July 18, 2007 10:56

Avoiding Skew in Tight Corners
 
Hi,

I was wondering if anyone could recommend how to avoid high skew in tight corners? I've minimized my skew by using a Quad/Tri Map meshing for my faces, but still have 2 elements with skew > 0.98 on one face which creates an entire problem region when I mesh the volume.

I have tried decomposing my geometry to avoid these corners but the geometry has been designed the way it is to maximize the hexahedral elements (as that is an important design criteria).

Any help is much appreciated.

red lemon July 18, 2007 16:18

Re: Avoiding Skew in Tight Corners
 
Use Gambit node adjust tool to move nodes individually to reduce localised skewness.

red lemon July 18, 2007 16:45

Re: Avoiding Skew in Tight Corners
 
or increase node count in these regions or delete the cells in solver if very flat and use robust solver for low quality cells such as pressure based coupled solver. Or convert them to polyhedral cells in Fluent. Just a few ideas here!

Marc July 18, 2007 20:39

Re: Avoiding Skew in Tight Corners
 
These are all techniques that I have not yet tried. I will try them all between tonight and tomorrow. Though, I'm not familiar with how to delete the problem cells using FLUENT. Having said that, I think that this problem needs to be fixed at the source (GAMBIT).

I appreciate the advice!

red lemon July 23, 2007 15:21

Re: Avoiding Skew in Tight Corners
 
yep sort them at source is best (smoothing/remeshing etc). To delete problem cells in Fluent, use adaption and mark cells of high skewness, separate them out then delete the cell zone. Boundary mods will occur which need repairing (inc. non contiguous zones) but most of the removed cells will be flat.


All times are GMT -4. The time now is 08:46.