Diverging radiation model in engine flow
 

Hello everyone,

I'm simulating flow in a Stirling engine (no combustion) with discrete ordinates radiation turned on. It works very well up until around the 180 degree crank angle, the DO radiation suddenly begins to diverge in the regenerator region of the stirling engine (which has been defined as a porous zone).

What I tried initially was I turned off the porous zone but the divergence was still there. When I turned off the radiation model completely, there were no issues converging. I have generated an isosurface which shows that this divergence is a in a very small localised region that is far away from the dynamic mesh zone of the engine. The average element quality in this localised region is almost 1 (I used cutcell meshing). I've also lowered the URF for DO only to 0.8, 0.6 and 0.1 where all configurations leads to divergence.

Does anyone have any experience with this issue? I'd be happy to provide more details if you need any.