[Shibi]

Hello to all,


I would like to know to following:
Why do transient solvers (e.g., rhoPimpleFoam) have the steady-state option available in fvSchemes?


Can I run rhoPimpleFoam in steady-state (although the solver is used for transient simulations..)? If the time derivatives are set to 0, the transient formulation should become the steady-state formulation. This is correct, right?


For a case study, I am running the aerofoilNACA0012 tutorial with rhoPimpleFoam with ddt schemes as steadyState (just to see if it runs). However, at the first iteration the solver diverges.


Giving:

Creating finite volume options from "system/fvOptions"

Selecting finite volume options type limitTemperature
    Source: limitT
    - selecting all cells
    - selected 16000 cell(s) with volume 4.1993135
Courant Number mean: 5467.174 max: 121506.46

Starting time loop

Courant Number mean: 5467.174 max: 121506.46
Time = 1
PIMPLE: iteration 1
#0  Foam::error::printStack(Foam::Ostream&) at ~/OpenFOAM/OpenFOAM-v2012/src/OSspecific/POSIX/printStack/printStack.C:237
#1  Foam::sigFpe::sigHandler(int) at ~/OpenFOAM/OpenFOAM-v2012/src/OSspecific/POSIX/signals/sigFpe.C:126
#2  ? in /lib/x86_64-linux-gnu/libpthread.so.0
#3  Foam::divide(Foam::Field<double>&, Foam::UList<double> const&, Foam::UList<double> const&) at ~/OpenFOAM/OpenFOAM-v2012/src/OpenFOAM/fields/Fields/scalarField/scalarField.C:125 (discriminator 3)
#4  Foam::operator/(Foam::UList<double> const&, Foam::UList<double> const&) at ~/OpenFOAM/OpenFOAM-v2012/src/OpenFOAM/fields/Fields/scalarField/scalarField.C:125
#5  Foam::diagonalSolver::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const at ~/OpenFOAM/OpenFOAM-v2012/src/OpenFOAM/matrices/lduMatrix/solvers/diagonalSolver/diagonalSolver.C:71
#6  Foam::fvMatrix<double>::solveSegregated(Foam::dictionary const&) at ~/OpenFOAM/OpenFOAM-v2012/src/finiteVolume/fvMatrices/fvScalarMatrix/fvScalarMatrix.C:181 (discriminator 1)
#7  Foam::fvMatrix<double>::solveSegregatedOrCoupled(Foam::dictionary const&) at ~/OpenFOAM/OpenFOAM-v2012/src/finiteVolume/lnInclude/fvMatrixSolve.C:95
#8  Foam::fvMesh::solve(Foam::fvMatrix<double>&, Foam::dictionary const&) const at ~/OpenFOAM/OpenFOAM-v2012/src/finiteVolume/fvMesh/fvMesh.C:552
#9  Foam::fvMatrix<double>::solve(Foam::dictionary const&) in ~/OpenFOAM/OpenFOAM-v2012/platforms/linux64GccDPInt32Debug/bin/rhoPimpleFoam
#10  Foam::fvMatrix<double>::solve() in ~/OpenFOAM/OpenFOAM-v2012/platforms/linux64GccDPInt32Debug/bin/rhoPimpleFoam
#11  ? in ~/OpenFOAM/OpenFOAM-v2012/platforms/linux64GccDPInt32Debug/bin/rhoPimpleFoam
#12  __libc_start_main in /lib/x86_64-linux-gnu/libc.so.6
#13  ? in ~/OpenFOAM/OpenFOAM-v2012/platforms/linux64GccDPInt32Debug/bin/rhoPimpleFoam
Floating point exception (core dumped)

I only created the PIMPLE dictionary with:

PIMPLE
{
    nNonOrthogonalCorrectors 0;
    nOuterCorrectors 10;
    nCorrectors 3;

    pMinFactor      0.1;
    pMaxFactor      2;
}

Can anyone clarify this for me?


If I to run a couple of transient simulations with rhoPimpleFoam and then another to check the steady solution, can't I use the same solver for both?!


Best Regards

[Tobermory]

Check out this earlier thread:

Steady-State with PIMPLE

Yes you can run the transient solvers to solve for a steady state solution ... either with localEuler (LTS) or by choosing steadyState for the ddtScheme. Be aware that you will probably need to underrelax since you lose the stabilization of the transient term (see the discussion in the above thread).