Problem in Buoyancy Model?
 

The working fluid is water and the model is about natural convection. So CFX itself imposes boussinesq's approximation as buoyancy model to solve the problem. But the simulated result is far away from the experimental result. It seems that there is some problem in simulation. I guess, the problem may lies in the buoyanocy model. But i am not sure. Any suggestion/advice will be highly appreciated. Thanks.

Could be many things. Can you describe what you are modelling? What Ra number? What temperature range?

The model is for cooling purpose. A circle shaped loop where heat flux is imposed in the lower half and a fixed surface temperature is imposed on the upper half.

Unless you tell us the Ra number and temperature range I won't be able to help you much.