AMG divergence and boundary layer mesh
 

Hi, all

Recently I occurred the problem of AMG divergence: pressure correction. Firstly I checked my boundary condition by other model, it was fine. Then I changed the URF to 0.1 and 0.5 for pressure and others, but it was still diverged. I tried to use density based solver, it is no problem, but it needed more memory. So I tested the meshes. I found that if I deleted boundary layer mesh around walls, it also did work well.

My question is why the boundary layer mesh( y+ =1)can cause the pressure correction error, while no problem for coupled solver?