equiangle skew with tetrahedral cells

Wouter

Hi, I am experiencing convergence issues with a cylinder like geometry in which a swirly flow occurs, so I need to use RSM turbulence modeling. The convergence issues seem to be associated with cells with an equiangleskew close to 0.8. The problems appear as small vortices in the flow that do not disappear.

However, with Gambit I find it hard to generate meshes with no cells with an equiangle skew around 0.8: even for a simple geometry such as a cylinder volumetric meshing (tet/hex, Tgrid method) many cells with higher skewness values occur. Probably this is not an issue for flows with simpler physics, but does anyone have an advice how to generate tet-meshes with small skewness value? Thanks