Why do we need prism layers??
 

How do prism layers resolve the boundary layer flow phenomena? Can tri/hex resolve the same efficiently if mesh is fine at the boundaries?

Hex refinement at boundaries is generally preferred to prisms (aka wedges), but often the surface mesh is triangular. The goal of boundary layer refinement is to resolved the flow details. Those details are (usually) present only in the wall normal direction. So, while it is (in principle) possible to tet a boundary layer, the cell counts will be enormous and the mesh skewness, especially in the cells close to the wall, will be crippling.

On the other hand, simply extruding the tri-surface mesh into the flow domain using prisms avoids all of those issues. The flow can be well resolved in the wall normal direction while keep cell count and skewness under control.

Thanks s lot ! It is useful information

John C. Chien, who were you?!?

http://www.cfd-online.com/Forums/mai...shing-cfd.html

bye