[LJA_Koers]

Hi all,

For a burner design project, I'm initially trying to run a steady state, mixing problem of low caloric natural gas, 80% methane 20% N2, and air. The burner has 9 spuds and is axisymmetric. Because of the symmetry, I only model 1/9th of the domain. I want to use Fluent v21 R1's mesh adaption features to refine the mesh in the mixing region.

I'm however not sure what a proper mesh refinement criterion would be. Gradient or curvature based? And for what variable? Velocity, molar fraction or temperature?

I initially want it to refine for a steady case, and use transient mesh adaption later for a URANS and LES case. Any recommendation is appreciated! Figure with contour and mesh attached.
mole-frac_and_mesh.jpg