[Jury]

Hey guys,

when I try to switch from EDC v2005 to v2016 in the DLR_A_LTS case, the pressure flucuates in the flame zone and the calculation doesn't converge.

I'm using the Jones-Lindsted mechanism and these are my settings:

chemistryProperties:

chemistryType
{
    solver            EulerImplicit;
    method            standard;
}

EulerImplicitCoeffs
{
    cTauChem        1;
    equilibriumRateLimiter on;
}

combustionProperties:

combustionModel  EDC;

EDCCoeffs
{
    version v2016;
}

fvSolution:

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

    p
    {
        solver           PCG;
        preconditioner   DIC;
        tolerance        1e-6;
        relTol           0.01;
    }

    pFinal
    {
        $p;
        relTol           0;
    }

    "(U|h|k|epsilon)"
    {
        solver          PBiCGStab;
        preconditioner  DILU;
        tolerance       1e-6;
        relTol          0.1;
    }

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

    "Yi.*"
    {
        solver          PBiCG;
        preconditioner  DILU;
        tolerance       1e-8;
        relTol          0;
    }
}

PIMPLE
{
    momentumPredictor   yes;
    nOuterCorrectors    1;
    nCorrectors         2;
    nNonOrthogonalCorrectors 0;

    maxDeltaT           1e-7;
    maxCo               0.025;
    alphaTemp           0.005;
    alphaY              0.005;
    Yref
    {
        O2                  0.1;
        CH4                 0.1;
    }
    rDeltaTSmoothingCoeff   0.05;
    rDeltaTDampingCoeff     0.2;
}

relaxationFactors
{
    equations
    {
        ".*"     1.0;
    }

}

fvSchemes:

ddtSchemes
{
    default         localEuler;
}

gradSchemes
{
    default         Gauss linear;
}

divSchemes
{
    default             none;

    div(phi,U)          Gauss limitedLinearV 1;
    div(phi,Yi)         Gauss limitedLinear01 1;
    div(phi,h)          Gauss limitedLinear 1;
    div(phi,K)          Gauss limitedLinear 1;
    div(phid,p)         Gauss limitedLinear 1;
    div(phi,epsilon)    Gauss limitedLinear 1;
    div(phi,Yi_h)       Gauss limitedLinear01 1;
    div(phi,k)          Gauss limitedLinear 1;
    div(((rho*nuEff)*dev2(T(grad(U)))))     Gauss linear;
}

laplacianSchemes
{
    default         Gauss linear orthogonal;
}

interpolationSchemes
{
    default         linear;
}

snGradSchemes
{
    default         orthogonal;
}

chemkin:

ELEMENTS
H O C N
END

SPECIES
CH4 O2 CO2 CO H2O H2 N2 H O OH
END

REACTIONS CAL/MOLE

CH4  + 0.5O2         =>  CO+2H2              7.82E+13  0.0     30000 !1
FORD/CH4 0.5/
FORD/O2 1.25/

CH4  + H2O           =>  CO+3H2              3.0E11    0.0     30000 !2


CO    + H2O          <=>  CO2+H2             2.75E+12  0.0     20000 !3

H2    + 0.5O2        <=>  H2O                1.209E+18 -1.0    40000 !4
FORD/H2 0.25/
FORD/O2 1.5/

O2                   <=> 2O                   1.5E+09  0.0    113000 !5

H2O                  <=> H + OH               2.3E+22  -3.0   120000 !6

END

Reducing the maxCo, alphaT and alphaTemp Numbers didn't fix the problem so far.

Does anyone have an Idea what could lead to this instability?

[piu58]

1) Check your boundary conditions. Are they physical correct?
2) check your mesh
3) During the run: check maxCo and time step continuity error

If all above is ok:

4) start with a more stable simulation:
- first order schemes
- underrelaxation

If even that does not help:

5) start with a more stable physics
- start with low Reynolds number (small velocity or high viscosity)
- start with steady state calculation (simpleFoam)
- start with laminar or even potential flow

If you set runTimeModifialble you may change most of it during the run.

[Jury]

Hello Uwe,

thank you very much for your fast reply.

The case is part of the tutorials, why I assume that the BC's and mesh shouldn't be any problem.

I tried to obey the other steps you suggested above and this is the result for the residual of the static pressure p.

Despite getting smaller, the residual is still oscillating and doesn't converge.

[piu58]

I only can repeat my advices. You switched to another model and this means you have a different case. Please increase complexity step by step.