Boundary layer treatment of wavy or sawtooth-shaped meshes

[MWTWM M]

Dear community,

This is a boundary layer treatment problem for a wavy or sawtooth-shaped mesh, as shown in the figure. Is it better to generate the mesh using third-party software such as Gmsh and process the data with avg_all, or is it more recommended to handle the boundary layer treatment in the .usr file?

Any suggestions and criticisms are appreciated.

Student W. T. Ma

[Bastien B. - LHEEA - Centrale Nantes]

Hi,

I would recommend to deform the domain in the usrdat2 subroutine, it will provide a much smoother surface because each GLL point of the mesh will be placed exactly on the wavy surface.
If the mesh is deformed in Gmsh, Nek5000 will interpolate the GLL points using 2nd order reconstruction methods in each element: you can loose a lot of information, especially for low grid resolutions, or high frequency wavy patterns.

I attached 2 helical-domains generated using Gmsh. The first (meshNoGLL) is deformed in Gmsh, the second  (meshGLL) in usrdat2. Both have the same grid resolution. It can give you an idea of what you could get with Gmsh.

Best regards,

Bastien

[MWTWM M]

Dear  Bastien

Thanks for your answer, it really helps.

Student Ma