CFD Online Discussion Forums - The convergence problem of buoyantBoussinesqSimpleFoam in the tutorial hotRoom

Hi,
Recently I found a convergence problem of buoyantBoussinesqSimpleFoam. Let's start with the tutorial hotRoom. The classical case starts with some hot spot in the patch and the heat could radiated over the room. Meanwhile the air in the room flows. Now let's consider the stagnant case: there is no hot spot initially and there should be no flow at all. In the case settings, i switch off turbulent models. It turns out that the case is never converged. However we can see a stagnant flow if we analyse the data by paraview. So what's wrong here? Is there anything to do with the settings in the dictionary? Or is it related to parasitic currents?

Code:

    residualControl

    {

        p_rgh           1e-2;

        U               1e-3;   //1e-4;

        T               1e-3;



        // possibly check turbulence fields

        "(k|epsilon|omega)" 1e-3;

    }

The output residual is attached in the file.