sharp temperature decrease near the wall
 

I have read in the manual that adiabatic boundary condition in case radiation heat transfer is activated means that temperature is adapted to ensure the sum of all heat fluxes to be zero q_rad+q_cond=0. Does anyone could help me to understand wheather obtained result is physical: there is a steady combustion process (eddy dissipation and k-epsilon) - the hottest flame is in the center, temperature slowly lowers when approaching the wall and then a sharp (~70 K) decrease of the temperature occurs in the 1st wall element thickness. I tried a finer mesh and the same situation occurs - temperature sharply decreases in the one element thickness near the wall (element is ~0.5 mm thin and velocity below 1 m/s). In this situation as far as I understand q_rad and q_cond has got the same direction near the wall what is not in agreement with q_rad+q_cond=0, am I wrong? Balances and residuals are low enough for corvergence. In case radiation is not activated everything seems ok.

thank you in advance for any comment on this issue