grid refinement in turbulent flow
 

Hi CFD-cats,

What might be a good rule of tumb for grid refinement near walls in turbulent room flows (combined convection), modelled with RANS models? Of course a first restriction is y+>11 for high Re formulations. However, I found that predictions are quite different if I compare grids with 50<y+<100 and 200<y+<300. Any comments?

Cheers,

pepe

Re: grid refinement in turbulent flow
 

Hello Pepe,

This is a known problems when wall functions are used. It is not possible to get mesh independent solutions when you use wall functions. The so called y+ is a function of the mesh.

For disretized form of the equations to be consistent with the governing equations, we need to get mesh independent solutions.

This is the one big reason why it is recommended to use turbulence models without wall functions whenever possible.

Thanks,

Thomas