[P_P]

Hello,

currently I am trying to calculate the heat transfer in a bubble column with two phases. In order to do so I have to solve a couple of enthalpie equations.

Unfortunately the calculation stops at the very beginning with the following report:

fehler 8

#0 Foam::error:rintStack(Foam::Ostream&) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#1 Foam::sigFpe::sigHandler(int) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#2
at sigaction.c:0
#3 Foam:ICPreconditioner::calcReciprocalD(Foam::Fie ld<double>&, Foam::lduMatrix const&) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#4 Foam:ICSmoother:ICSmoother(Foam::word const&, Foam::lduMatrix const&, Foam::FieldField<Foam::Field, double> const&, Foam::FieldField<Foam::Field, double> const&, Foam::UPtrList<Foam::lduInterfaceField const> const&) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#5 Foam::lduMatrix::smoother::addsymMatrixConstructor ToTable<Foam:ICSmoother>::New(Foam::word const&, Foam::lduMatrix const&, Foam::FieldField<Foam::Field, double> const&, Foam::FieldField<Foam::Field, double> const&, Foam::UPtrList<Foam::lduInterfaceField const> const&) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#6 Foam::lduMatrix::smoother::New(Foam::word const&, Foam::lduMatrix const&, Foam::FieldField<Foam::Field, double> const&, Foam::FieldField<Foam::Field, double> const&, Foam::UPtrList<Foam::lduInterfaceField const> const&, Foam::dictionary const&) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#7 Foam::GAMGSolver::initVcycle(Foam::PtrList<Foam::F ield<double> >&, Foam::PtrList<Foam::Field<double> >&, Foam::PtrList<Foam::lduMatrix::smoother>&) const in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#8 Foam::GAMGSolver::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#9 Foam::fvMatrix<double>::solveSegregated(Foam::dict ionary const&) in "/software/openfoam/2.2.2/OpenFOAM-2.2.2/platforms/linux64GccDPOpt/lib/libfiniteVolume.so"
#10 Foam::fvMatrix<double>::solve(Foam::dictionary const&) in "/home/rafii/OpenFOAM/rafii-2.2.2/platforms/linux64GccDPOpt/bin/multiphaseTempEulerFoam_CTM_enthalpy_zeroModel"
#11 Foam::fvMatrix<double>::solve() in "/home/rafii/OpenFOAM/rafii-2.2.2/platforms/linux64GccDPOpt/bin/multiphaseTempEulerFoam_CTM_enthalpy_zeroModel"
#12
in "/home/rafii/OpenFOAM/rafii-2.2.2/platforms/linux64GccDPOpt/bin/multiphaseTempEulerFoam_CTM_enthalpy_zeroModel"
#13 __libc_start_main in "/lib64/libc.so.6"
#14
in "/home/rafii/OpenFOAM/rafii-2.2.2/platforms/linux64GccDPOpt/bin/multiphaseTempEulerFoam_CTM_enthalpy_zeroModel"
Floating point exception


The first line saying "fehler 8" gives me the possibility to track the exact position of the problem.
The piece of code which causes this problem is as follows:

Info<< "fehler 7 " << nl << endl;

volScalarField mut1 = nut*rho1;


dimensionedScalar null("null",dimensionSet(1,-1,-3,0,0,0,0), scalar(0.0));

Qint1 = min((heatTransferCoeff*(T1-T2)),null);//änderung
Info<< "fehler 8 " << nl << endl;

solve
(
fvm::laplacian((alpha1*mut1), he1_turb)
);

Since there exists no possibility that a division by 0 occurs, I wonder where this problem may come from. Can you help me out ?

Greatz P_P

[alexeym]

Hi,

Originally Posted by P_P

Since there exists no possibility that a division by 0 occurs, I wonder where this problem may come from. Can you help me out ?

I won't be so sure, as the error occurs here:

void Foam::DICPreconditioner::calcReciprocalD
(
    scalarField& rD,
    const lduMatrix& matrix
)
{
    scalar* __restrict__ rDPtr = rD.begin();

...

    // Calculate the DIC diagonal
    register const label nFaces = matrix.upper().size();
    for (register label face=0; face<nFaces; face++)
    {
        rDPtr[uPtr[face]] -= upperPtr[face]*upperPtr[face]/rDPtr[lPtr[face]];
    }


    // Calculate the reciprocal of the preconditioned diagonal
    register const label nCells = rD.size();

    for (register label cell=0; cell<nCells; cell++)
    {
        rDPtr[cell] = 1.0/rDPtr[cell];
    }
}

where rD is "The reciprocal preconditioned diagonal". If you've got infinity on the diagonal, then reciprocal is zero.

[P_P]

Thank you very much for your reply,

do you have an idea how to solve this problem? How can i prevent infinity on the diagonal? If this is not possible would another preconditioner be the best solution?

[alexeym]

In general the error means diverging solution. Surely you can try to use GAMG (you can find syntax for it in tutorial cases) as preconditioner but the change in precoditioner will not remove inconsistency of initial and boundary conditions, will not correct the model etc.