buoyancy modeling/turbulence

tony

I'm modeling gas downflow in a heated pipe. The flow rate will be such that it is laminar. From previous studies I expect the flow to have flow reversal (upflow) near the heated wall section due to buoyancy. Also it has been found that this flow reversal occurs in oscillatory behavior, where reverse flow separates off one side, then the other side.

So I was wondering on any suggestions on turbulence model. Do i do it laminar or turbulent. With secondary flow I'd suspect turbulent (to capture this), but from tutorials I've modeled vortex shedding with laminar. And if I use turbulence, any suggestions on which model.

Thanks so much in advance!

tony