[Clau.77]

Hey guys!


I am trying to simulate an air flow around a Geometry. I have attached a foto of the geometry as well as a close up of the mesh. I need to use chtMultiRegion because later on I will study the heat transfer. Right now i am using k-epsilon turbulence and steadyState. The objective is to use this solution as a start for a transient simulation with heat exchange.



Problem: My solution does not converge, especially p_rgh. I tried playing around with the mesh, the relaxation factor, the nonOrthogonalCorrectors... but everything is just making it worse.


Question:

1. Is it a problem with my mesh? I have some non-orthogonal an skew faces.

2. How can I improve my simulation?

3. In my log file you can see that OF calculates p_rgh multiple times during 1 timestep. Is that an error?




the fvSolution file for the air:

FoamFile
{
    version     2.0;
    format      ascii;
    class       dictionary;
    location    "system/Luft";
    object      fvSolution;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

solvers
{
  "h.*"
    {
      solver           PBiCGStab;
      preconditioner   DILU;
      tolerance        1e-7;
      relTol           1e-7;
      minIter          5;
      maxIter          50;
    }
  "hFinal.*"
    {
      $h;
    }
  "e.*"
    {
      $h;
    }
  "eFinal.*"
    {
      $h;
    }

    rho
    {
        solver           GAMG;
        smoother         GaussSeidel;
        tolerance        0.0001;
        relTol           0.0001;
    // maxIter         10;
    }

    rhoFinal
    {
        $rho;
    relTol           0;
    // maxIter       20;
    }

    p_rgh
    {
        solver           GAMG;
        smoother         GaussSeidel;
        tolerance        0.0001;
        relTol           0.0001;
    maxIter         20;
    }

    p_rghFinal
    {
        $p_rgh;
        tolerance        0.0001;
    relTol         0.0001; 
    }


    "(U|k|epsilon|omega|R).*"
    {
        solver           PBiCGStab;
        preconditioner   DILU;
        tolerance        0.0001;
        relTol           0.0001;
    }

    "(U|k|epsilon|omega|R)Final.*"
    {
        $U;
    tolerance     0.0001;
        relTol           0;
    //    maxIter   20;
    }
}

PIMPLE
{
    momentumPredictor   yes;
    nOuterCorrectors    1;
    nCorrectors         2;
    nNonOrthogonalCorrectors 0;//1 leads to extreme fluctuation of p_rgh;
    
    outerCorrectorResidualControl
      {
    "(U|k|epsilon|omega|p_rgh).*"
      {
        tolerance   1e-4;
        relTol      0;
      }
    "h.*"
      {
        tolerance   1e-7;
        relTol      0;
      }
    "e.*"
      {
        tolerance   1e-7;
        relTol      0;
      }
      }

}

relaxationFactors
{
    fields
    {
      p_rgh              1;
      p_rghFinal        1;
      rho               1;
      rhoFinal          1;

    }       

 equations
    {
      h                0.001;
      e                0.001;
      hFinal            0.001;
      eFinal            0.001;
      U                 0.1;
      UFinal            0.1;
      k             0.1;
      omega        0.1;
      epsilon        0.1;
      kFinal            0.1;
      epsilonFinal      0.1;
      omegaFinal        0.1; 
    }
}

FoamFile
{
    version     2.0;
    format      ascii;
    class       dictionary;
    location    "system/Luft";
    object      fvSchemes;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

ddtSchemes
{
    default         steadyState;//CrankNicolson 0.9;
}

gradSchemes
{
    default         Gauss linear;
}

divSchemes
{
    default         none;

    div(phi,U)      bounded Gauss upwind;  //bounded weg für transient?
    div(phi,K)      bounded Gauss upwind;;
    div(phi,h)      bounded Gauss upwind;
    div(phi,e)      bounded Gauss upwind;
    div(phi,k)      bounded Gauss upwind;
    div(phi,epsilon)    bounded Gauss upwind;
    div(phi,omega)  bounded Gauss upwind;
    div(phi,R)      bounded Gauss upwind;
    div(phi,K)      bounded Gauss linear;
    div(phi,Ekp)    bounded Gauss linear;
    div(R)          Gauss linear;
    div(U)          Gauss linear;
    div(((rho*nuEff)*dev2(T(grad(U))))) Gauss linear;
}

laplacianSchemes
{
    default         Gauss linear uncorrected;
}

interpolationSchemes
{
    default         linear;
}

snGradSchemes
{
    default         corrected;
}

wallDist
{
method        meshWave;
}

i also attached the first steps of my log file.


Can someone give me advice?

[HosamAlrefaie]

I have the same problem with p_rgh, it has high residuals and too many iterations each time.

I am using fixedFluxPressure BC for the inner pipe walls.

[piu58]

I recommend using relaxationFactors in fvSolution. You should at least some kind of result then.

I don't like your mesh, I found the steps of refinement to steep. May be a coarser mesh is at least a better starting point. But this is hard to say without a result.

I recommend simplifying your geometry and investigating the "bump" and the "pits" isolated, at least at the beginning.