Wall Burning Along Inflated Boundaries

John

I've encountered a phenomena while running combustion in CFX using a tet mesh with an inflated boundary where the flame burns along the inflated boundary. Specifically when the flame front encounters the first prism layer it will follow that layer some finite distance before stepping into the next layer. It continues this pattern until it burns to the wall.

At first I thought it might've been an issue with the turbulence model but varying the model or specifying various properties of the turbulence had no affect on the flame behavior. The solution exhibits large gradients in temperature, shear strain, velocity, eddy viscosity, K, omega, epsilon, etc... within the inflated boundary but smooth out as soon as it transitions to the tet mesh. In fact the gradients exhibited vary greatly layer-to-layer, i.e. one can view a plot of velocity gradients and the individual prism layer boundaries will be easily recognizable.

I'm starting to wonder if what I'm seeing isn't a discretization error with how CFX handles tets and hexes. I've tried varying the diffusion schemes but so far nothing seems to have any effect.

I've run this same geometry using a 100% hex mesh from ICEM and it shows none of the irregular behavior as the hybrid tet-prism mesh. Has anyone encountered this type of behavior before or have any insights on what the code could be doing here?

Regards, John