Hi to all, I try to model a steady laminar natural convection heat transfer over a hot plate using Boussinesq approach. I specify the BC as a wall at the hot surface, and the bottom surface as a pressure inlet, the top surface as a pressure outlet, and side as a pressure inlet. Also i try to specify the bottom and side as an inlet vent, these boundary condition gave me the same results for heat flux, but i get a reverse flow while iterating the solution. If any one face this problem could give me a hint please.

Make the domain bigger. Fluent calculates the cell pressure just inside the boundary as just lower than the boundary condition causing inflow at an outlet boundary. Also use gravity on, ideal gas and PRESTO or BFW for pressure discretisation. Also ensure grid is resolved to capture plumes and hot plate may need prism layers for convection.

Thanks for your help. I used gravity and presto as a pressure discretization and my grid are fine close to the boundary layer. Let me try to make my domain bigger and i will respond to you. thanks again for your help.

Hi red lemon, Still i face the reverse flow even when i increase the domain size. I think there is something else i have to consider it. if you have any comments let me know. thanks

keep increasing the domain size until no reverse flow