Combustion using composition PDF transport model with reynolds stress 5 eqn model
 

I'm trying to model stratified Premixed methane-air combustion using composition PDF transport model and using 17 species skeletal mechanism. When I patch the region to initiate combustion, either the hot region slowly moves away to be replaced by room temperature (300K) flow as iterations progress or the whole region stays hot at the flame temperature (uniform hot temperature over the entire domain) and the velocity contours go haywire. I was wondering what could be the reason for the same and how to solve the issue. I appreciate your help.

Composition PDF Transport
 

Are you initiating the simulation with a patch of high temperature? That's not a good idea when using PDF transport, particularly if you have chemistry agglomeration enabled. You should solve the case first with finite-rate or EDC and then switch over to PDF transport using results from finite-rate or EDC as initial solution.