LES Grid - multiblock-structured or snappy?
 

Hi,

Am wondering which approach is more appropriate for LES in external aerodynamics. I currently have 2 mesh for my geometry - a multiblock-structured mesh made in GAMBIT (due to the geometry or my inability, checkMesh complains about nonorthogonality, about 0.3% of total number of faces is too nonortho). The second mesh I have is snappyHexMesh polyhedral.

Multiblock structured mesh contains cells that have aspect ratio of around 10, which makes me concerned about whether SGS model is still valid under the circumstances. I'm starting out with the Smagorinsky model.

Meanwhile, polyhedral mesh cells have aspect ratio closer to 1. However, it does have some 5-sided triangular pyramids.

As far as I know (sorry if I'm wrong, my knowledge in LES is limited), SGS models carry the assumption that grid is uniform. The question is which is a better compromise? Any reference I can check out?

My supervisor seems to be favor with the multiblock-structured grid. Looking for differential diagnosis.

Thanks!