CFD Online Discussion Forums - buoyantPimpleFoam diverges with the negative temperature

Hi,

I want to simulate a 3-D transient of external wind around buildings. The turbulent model was set K-epsilon. I tried to use the buoyantPimpleFoam solver in OpenFOAM 8, and after a few time steps, the solution diverges with the negative temperature error. I want to use buoyantPimpleFoam because I need to apply the Boussinesq approximation. The grid of the computational domain is shown in Figure. The output from checkMesh was OK. I think there is something wrong with pressure; I am not sure! I am new to OpenFoam, and any help would be highly appreciated.
The BCs are attached in the 0 folder, and the other case setups are presented below:

** thermophysical properties:

PHP Code:


		
			
thermoType
{
    type            heRhoThermo;
    mixture         pureMixture;
    transport       const;
    thermo          eConst;
    equationOfState Boussinesq;
    specie          specie;
    energy          sensibleInternalEnergy;
}

mixture
{
    specie
    {
        molWeight       28.9;
    }
    equationOfState
    {
        rho0            1;
        T0              290;
        beta            3e-03;
    }
    thermodynamics
    {
        Cv              712;
        Hf              0;
    }
    transport
    {
        mu              1.8e-05;
        Pr              0.7;
    }
}

fvSchemes:

PHP Code:


		
			
ddtSchemes
{
    default         Euler;
}

gradSchemes
{
    default         leastSquares;
}

divSchemes
{
    default         none;

    div(phi,U)      Gauss linearUpwind grad(U);
    div(phi,e)      Gauss upwind;

    div(phi,k)      Gauss upwind;
    div(phi,epsilon) Gauss upwind;

    div(phiv,p)     Gauss linear;
    div(phi,K)      Gauss linear;

    div(phi,R)          Gauss upwind;
    div(R)              Gauss linear;
    div(phi,nuTilda)    Gauss upwind;

    div(((rho*nuEff)*dev2(T(grad(U))))) Gauss linear;
}

laplacianSchemes
{
    default         Gauss linear corrected;
}

interpolationSchemes
{
    default         linear;
}

snGradSchemes
{
    default         corrected;
}

fvSolution:

PHP Code:


		
			
solvers
{
    "rho.*"
    {
        solver          diagonal;
    }

    p_rgh
    {
        solver           GAMG;
        tolerance        1e-7;
        relTol           0.01;

        smoother         DICGaussSeidel;
    }

    p_rghFinal
    {
        $p_rgh;
        relTol          0;
    }

    "(U|e|k|epsilon)"
    {
        solver          smoothSolver;
        smoother        symGaussSeidel;
        tolerance       1e-05;
        relTol          0.1;
    }

    "(U|e|k|epsilon)Final"
    {
        $U;
        relTol          0;
    }
}

PIMPLE
{
    momentumPredictor no;
    nOuterCorrectors 1;
    nCorrectors     2;
    nNonOrthogonalCorrectors 2;

    pRefCell        0;
    pRefValue       0;
}

Thank you!Attachment 98393

Attachment 98394

Attachment 98395

Attachment 98396