Reply To: Thermal analysis problem

[Rob]

The lower image shows the buoyant plume nicely, assuming that's a contour of temperature. I can't read the text as it's too small. 

Convergence is not just residuals. It's also monitor points (is the solution changing) and fluxes (does what goes in come out). So in this case it's very likely the residuals won't look good. You may well have a transient effect in the flow, additionally the answer may not be too far off the initial condition (for some fields) so the residual maths can't reduce the error.