I have recently performed a CHT simulation of the natural convection observed in a heated heat sink immersed in air. The results I have achieved were good. Now, I am interested in doing the same but with a heat sink immersed in an oil medium. In order to do a small test, I have already set up a simulation and have run it for some hours. Apparently, the results would not converge, as the temperature measured on one of the heat sink surfaces (which is the project goal) kept raising. This is the link:
I guess the main problem is related to the boundary conditions. As the oil is not an “infinite” medium as the air and, actually, is surrounded by well-defined walls (the heat sink is submerged in a box filled with oil), I am not sure if I can set, for example, a pressure-inlet-outlet velocity boundary condition for all the lateral surfaces. And, even if I could, I guess the pressure value should be taken as dependent of the oil depth.
Is there anyone who has ever done something similar to this or has an idea of what can be done?