[anonymousbach]

Hi All,

I'm trying determine the surface nusselt numbers on a single 2D cylinder in a rectangular domain. The domain has constant pressure and temperature on all 4 sides. The gauge pressure is assumed to be 0.

I'm having trouble coming up with the proper boundary conditions. Initially I thought I could treat two walls as inlets and 2 walls as outlets but the nusselt numbers change significantly depending on which walls are inlets or outlets. Is there a way to model an "ambient" atmosphere in Fluent, with just constant pressure all around?

Thanks for any help, it would be much appreciated!

[LuckyTran]

Usually a constant pressure inlets and outlets are used.

Naturally convection problems are tricky to model because of contamination of the boundary condition onto the modeled domain. The rectangular box domain needs to be large enough that the constant pressure inlets act as ambient boundary conditions. If your Nusselt numbers are changing depending on which is where, that is a hint that your domain is not large enough.

If possible it's better to specify stagnation pressure than static pressure, since the stagnation pressure is more physically correct.

[anonymousbach]

Thank you LuckyTran.

Follow up question, and I apologize because I am largely self taught at Fluent and therefore there's quite a few gaps in my knowledge, are there boundary conditions that tell Fluent these lines are just conceptual and in no way affect the flow?

My thinking is this. Because I am only interested in the temperature distribution in my smaller domain, and the nusselt numbers on the surface of my cylinder,, I'll create a much bigger domain to encapsulate it, but still just look at the smaller domain.