[natalialissette]

I hope someone can help me.

I'm trying to simulate natural convection in a tall vertical annuli cavity. Height is 0.215 m, outer cylinder is 0.15 m, and inner cylinder is 0.03 m in diameter.

My mesh is al hexahedral 2D, with 70700 elements. My model is 2D axysimetric. There are more than 30 elements in the region of the boundary layer.

Gravity is -9.81 in X. I'm using double precision in Fluent. Pressure-based solver, steady. I'm using SST k-omega with low Re-correction. I'm doing the boussinesq approximation stating operating density and a compresibility coefficient. Fluid is water.

The boundary conditions are uniform heat flux 440 w/m^2K in the inner wall and -176 W/m^2K for the outer wall. The Rayleigh number is 10^8. Scheme SIMPLE, Spatial discretization for pressure is PRESTO!, second order upwind for momentum, turbulence, dissipation and energy. All relaxation factors are default. I doesn't converge. k is constanlty diverging.

I also tried transient approaches and it seems to converge only at low Ra numbers. I have tried other solvers, and other models for viscosity and nothing seems to work.

Can someone give me suggestions??

Thanks in advance