CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Mesh Generation & Pre-Processing (http://www.cfd-online.com/Forums/mesh-generation/)
-   -   Mesh cells type most suitable for CFD (http://www.cfd-online.com/Forums/mesh-generation/115774-mesh-cells-type-most-suitable-cfd.html)

rv82 April 6, 2013 10:43

Mesh cells type most suitable for CFD
 
1 Attachment(s)
Hi friends! What is the most suitable and optimal mesh cells type for CFD? I read somewhere that tri-prisms and tet elements are worst for CFD computation. But my case contains that elements, about 20% of all mesh. Is it bad? I can exchange them by hex elements but they will be not orthogonal.

And my second question: what flow directions support tri-prism elements? Can I use these elements as shown on picture?

Thank you!

Roman

jchawner April 8, 2013 09:45

The general rule of thumb is that hex cells are superior to tet cells for CFD, especially in high gradient regions like boundary layers and wakes. The exceptions are that the hex cells must be of decent quality (e.g. low skew). Another exception is that a hybrid mesh with prisms in the boundary layer and tets elsewhere can support accurate CFD. You can find many examples of good CFD meshes in any of the reports from the AIAA's Drag Prediction Workshop.

ztdep April 12, 2013 09:44

we care the mesh quality. you can use any cell type. Tet, Hex, Prism, wedge, etc.
orthogonal mesh will increase the accuracy, and is more easy to solve.


All times are GMT -4. The time now is 00:54.