[dpitz]

Hi everyone,

I'm running a unsteady 3D calculation of buoyant flow in an annular sealed rotating cavity using a 45º sector. No turbulence modelling is used and the boundary conditions are as follows:

- Adiabatic side walls;
- Inner (cylindrical) wall maintained at a cold temperature;
- Outer wall maintained at a hot temperature;
- The other two walls are periodic.
- No slip condition at all walls.

The problem is solved using the relative velocity formulation, and a rotational speed around the Z axis is imposed to the whole domain (under Cell Zone Conditions). The Boussinesq approximation is used for changes in density (air is the working fluid).

As the solution advances in time, the heat flux at the outer wall keeps decreasing and does not seem to stabilize at a given value. I think that there might be something wrong with my setup. When using the Boussinesq approximation, the force field to be considered is the centrifugal force instead of gravity, so do I need to impose the centrifugal force as a source term or Fluent automatically considers it since the domain is rotating?

I have tried different solution methods without success. I'm using a time step of 0.001 s (Rayleigh number is 10^8) and doing 20 iterations per time step.

Has anyone faced a similar problem before or have any idea about why the heat flux is always decreasing?

Thank you.

[dpitz]

Anyone?