Hi,guys.

I would appreciate your help.I found sometimes my CFD grid have a big deal of nonplanarity due to the deformation/twist of the grid. I mean sometime rectangular cell faces are not not plane surfaces. So how this will affect the solution of fluid flow.

Thanks in advance

Regards,

gonski

what is needed is the cell surface area, so how much is the difference between the real surface area and the calculated surface area ? (which depends on the coordinates of the vertices)

What kind of trouble can be observed in the solution of fluid flow?

Interesting, I try to search some proofs for this non-planar problem in this whole day, and get nothing. There are one or two papers in finite element method discussing the problem. However their explanations are not convincing. Now everything push me to believe that non-planar faces should be no problem in finite volume method as no one talk about this problem. I would appreciate your comments.

Best regards,

Gonki

The usual trick is to split the face into two triangles and define the normal to the face to be the average of the normals to the two triangles. If these two normals are very different (so that it effects your answer) then your grid needs smoothing (to the point where the error induced is comparable to the discretization error).

Thanks a lot! Tom and Ahmed