[Pat84]

Dear all,

I would like to test the kOmegaSST low reynolds turbulence model with a t-junction and have generated my mesh in icem. I´ve tested the mesh in fluent and have planned to compare the result of fluent and openfoam, but when I use the icem mesh ( I convert the .msh file with fluent3DMeshToFoam ) I get very high residuals for the pressure - in order of 0.6 - 1.0 and up to 1000 iterations. After a while the simulation diverges. I use the same BC in fluent and OF, but in fluent the simulation works - y+ max is ~0.6. I think the error lies in the conversion of the mesh from .msh to openfoam mesh, since there are two warnings while converting:

 fluent3DMeshToFoam mixingtee.msh 
/*---------------------------------------------------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM Extend Project: Open source CFD        |
|  \\    /   O peration     | Version:  1.6-ext                               |
|   \\  /    A nd           | Web:      www.extend-project.de                 |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
Build  : 1.6-ext-bd38c3b48291
Exec   : fluent3DMeshToFoam mixingtee.msh
Date   : Aug 12 2013
Time   : 14:56:49
Host   : Knecht.site
PID    : 8419
Case   : /home/patrick/OpenFOAM/patrick-1.6-ext/run/mixingtee_fineWall
nProcs : 1
SigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Dimension of grid: 3
Number of points: 399592
PointGroup: 15 start: 0 end: 399591 nComponents: 3.  Reading points...done.
Number of cells: 391417
CellGroup: 16 start: 0 end: 391416 type: 1
Number of faces: 1182246
FaceGroup: 17 start: 0 end: 1166255.  Reading uniform faces...done.
FaceGroup: 18 start: 1166256 end: 1168348.  Reading uniform faces...done.
FaceGroup: 19 start: 1168349 end: 1170441.  Reading uniform faces...done.
FaceGroup: 20 start: 1170442 end: 1172534.  Reading uniform faces...done.
FaceGroup: 21 start: 1172535 end: 1182245.  Reading uniform faces...done.
Zone: 16 name: FLUID type: fluid.  Reading zone data...done.
Zone: 17 name: int_FLUID type: interior.  Reading zone data...done.
Zone: 18 name: INLET-Y type: velocity-inlet.  Reading zone data...done.
Zone: 19 name: INLET-Z type: velocity-inlet.  Reading zone data...done.
Zone: 20 name: OUTLET type: outlet-vent.  Reading zone data...done.
Zone: 21 name: WALL type: wall.  Reading zone data...done.

FINISHED LEXING

--> FOAM Warning : 
    From function min(const UList<Type>&)
    in file lnInclude/FieldFunctions.C at line 342
    empty field, returning zero
--> FOAM Warning : 
    From function min(const UList<Type>&)
    in file lnInclude/FieldFunctions.C at line 342
    empty field, returning zero
Creating patch 0 for zone: 18 name: INLET-Y type: velocity-inlet
Creating patch 1 for zone: 19 name: INLET-Z type: velocity-inlet
Creating patch 2 for zone: 20 name: OUTLET type: outlet-vent
Creating patch 3 for zone: 21 name: WALL type: wall
Creating cellZone 0 name: FLUID type: fluid
Creating faceZone 0 name: int_FLUID type: interior
faceZone from Fluent indices: 0 to: 1166255 type: interior
patch 0 from Fluent indices: 1166256 to: 1168348 type: velocity-inlet
patch 1 from Fluent indices: 1168349 to: 1170441 type: velocity-inlet
patch 2 from Fluent indices: 1170442 to: 1172534 type: outlet-vent
patch 3 from Fluent indices: 1172535 to: 1182245 type: wall

    From function void polyMesh::initMesh()
    in file meshes/polyMesh/polyMeshInitMesh.C at line 82
    Truncating neighbour list at 1166256 for backward compatibility

Writing mesh to "/home/patrick/OpenFOAM/patrick-1.6-ext/run/mixingtee_fineWall/constant/region0"

End

The problem is, that checkMesh returns no error, but if you have a look at the attached pichtures, the skewness is a bit small, or? :

checkMesh 
/*---------------------------------------------------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM Extend Project: Open source CFD        |
|  \\    /   O peration     | Version:  1.6-ext                               |
|   \\  /    A nd           | Web:      www.extend-project.de                 |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
Build  : 1.6-ext-bd38c3b48291
Exec   : checkMesh
Date   : Aug 12 2013
Time   : 15:01:12
Host   : Knecht.site
PID    : 8731
Case   : /home/patrick/OpenFOAM/patrick-1.6-ext/run/mixingtee_fineWall
nProcs : 1
SigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).

// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
Create time

Create polyMesh for time = 0

Time = 0

Mesh stats
    all points:           399592
    live points:           399592
    all faces:            1182246
    live faces:            1182246
    internal faces:   1166256
    cells:            391417
    boundary patches: 4
    point zones:      0
    face zones:       1
    cell zones:       1

Overall number of cells of each type:
    hexahedra:     391417
    prisms:        0
    wedges:        0
    pyramids:      0
    tet wedges:    0
    tetrahedra:    0
    polyhedra:     0

Checking topology...
    Boundary definition OK.
    Point usage OK.
    Upper triangular ordering OK.
    Face vertices OK.
    Number of regions: 1 (OK).

Checking patch topology for multiply connected surfaces ...
    Patch               Faces    Points   Surface topology                  
    INLET-Y             2093     2120     ok (non-closed singly connected)  
    INLET-Z             2093     2120     ok (non-closed singly connected)  
    OUTLET              2093     2120     ok (non-closed singly connected)  
    WALL                9711     9788     ok (non-closed singly connected)  

Checking geometry...
    This is a 3-D mesh
    Overall domain bounding box (-0.0761695 -0.3556 -0.0760639) (0.0761887 0.3556 0.37465)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Mesh (non-empty, non-wedge) dimensions 3
    Boundary openness (-1.40907e-15 -1.33355e-16 2.53797e-16) Threshold = 1e-06 OK.
    Max cell openness = 3.84166e-14 OK.
    Max aspect ratio = 813.001 OK.
    Minumum face area = 1.92001e-08. Maximum face area = 8.03386e-05.  Face area magnitudes OK.
    Min volume = 1.01453e-10. Max volume = 2.12297e-07.  Total volume = 0.0153058.  Cell volumes OK.
    Mesh non-orthogonality Max: 58.654 average: 15.2185 Threshold = 70
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 2.6457 OK.

Mesh OK.

End

I have also tried the same mesh with standard kOmegaSST and kEpsilon with wall functions, but all calculations show the same behavior.
The mesh is a full hexa mesh with o-grid. OF version is OF-1.6 extend.
The BC for k and Omega in the low reynolds SST case are zeroGradient.
What can be the reason for the pressure divergence?

Best regards,
Patrick

[Pat84]

I have uploaded the case here:

http://uploaded.net/file/wdodjyce

The key is: cfd-online

[Pat84]

I have the reason for a smaller mesh then the attached one:

The behavior is caused by the GAMG solver for the pressure. My settings were:

p
    {
    solver          GAMG;
    smoother    GaussSeidel;
    agglomerator    faceAreaPair;
    nCellsInCoarsestLevel 100;
    mergeLevels    1;
    cacheAgglomeration false; 
    
        tolerance       1e-06;
        relTol          0.001;
    }

With conjugate gradients the convergence is not perfect but it works:

p
    {
        solver           PCG;
        preconditioner   DIC;
        tolerance        1e-08;
        relTol           0.0001;
    };

The question is why GAMG for pressure worked with a coarser grid with about 50k cells but not for this mesh with nearly 280k cells and why the attached mesh with ~400 does not work. Even with PCG and not GAMG for all values. PotentionalFoam gave a strange result with the icem mesh (attached 400k cells mesh) which looks like there are some internal walls