Problem with continuity
 

Hello everyone. I have the below geometry. i try to model combustion but always something goes wrong. for this reason i put cold air to the inlets to have some results. the result was that in approximately in the 2500 iteration continuity and e started to diverge (turbulence viscocity in * cells etc). i dont know what is going wrong. where do you believe in this geometry should i have better mesh if i want to model combustion and with which model should i use?