checkMesh Ok but bad cells still present

Hiroshiman · August 14, 2013, 12:02

Hi,
I'm meshing a complex 3D geometry using snappyHex mesh for transient flow simulation using pimpleFoam. When I run checkMesh on my case I get :

Code:

Time = 0

Mesh stats
    points:           241926
    faces:            585183
    internal faces:   523327
    cells:            179081
    faces per cell:   6.18999
    boundary patches: 7
    point zones:      0
    face zones:       0
    cell zones:       0

Overall number of cells of each type:
    hexahedra:     128793
    prisms:        9069
    wedges:        0
    pyramids:      0
    tet wedges:    925
    tetrahedra:    30
    polyhedra:     40264
    Breakdown of polyhedra by number of faces:
        faces   number of cells
            4   12216
            5   6431
            6   6864
            7   386
            8   415
            9   6591
           10   139
           11   134
           12   3677
           13   43
           14   40
           15   3021
           16   7
           17   7
           18   258
           20   1
           21   30
           24   4

Checking topology...
    Boundary definition OK.
    Cell to face addressing 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
                     il1      398      475  ok (non-closed singly connected)
            defaultFaces    52012    66767  ok (non-closed singly connected)
                     ren      180      293  ok (non-closed singly connected)
                   inlet     5979     6306  ok (non-closed singly connected)
                     il2     1317     1469  ok (non-closed singly connected)
                     il3      933     1032  ok (non-closed singly connected)
                     il4     1037     1148  ok (non-closed singly connected)

Checking geometry...
    Overall domain bounding box (-0.0738687 -0.236453 0.724997) (0.0819296 -0.12425 1.085)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Boundary openness (5.27378e-16 -5.96184e-16 3.35583e-17) OK.
    Max cell openness = 3.42542e-16 OK.
    Max aspect ratio = 9.17275 OK.
    Minimum face area = 4.24086e-09. Maximum face area = 1.61588e-05.  Face area magnitudes OK.
    Min volume = 1.24757e-12. Max volume = 5.47073e-08.  Total volume = 0.000215698.  Cell volumes OK.
    Mesh non-orthogonality Max: 64.62 average: 13.9596
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 2.96129 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

End

After approximately 50% on my run the simulation diverges and when I look at the last available results I can spot a bad looking cell (see picture).
crash_U_mesh.jpg
It's a crop of the outlet (visible on the left) on which a pressure is applied using codedFixValue.

Are there anything appart from refining the mesh to avoid this ? Which error parameter in the snappyHexMeshDict should be modified ? Usually checkmesh is talkative enough to give hints...

Thanks,

Florian

l_r_mcglashan · August 14, 2013, 12:05

Could you run checkMesh with the options -allGeometry and -allTopology?

Hiroshiman · August 14, 2013, 12:13

Thank you for your help, here is the result of checkMesh -allGeometry -allTopology :

Code:

Enabling all (cell, face, edge, point) topology checks.

Enabling all geometry checks.

Time = 0

Mesh stats
    points:           241926
    faces:            585183
    internal faces:   523327
    cells:            179081
    faces per cell:   6.18999
    boundary patches: 7
    point zones:      0
    face zones:       0
    cell zones:       0

Overall number of cells of each type:
    hexahedra:     128793
    prisms:        9069
    wedges:        0
    pyramids:      0
    tet wedges:    925
    tetrahedra:    30
    polyhedra:     40264
    Breakdown of polyhedra by number of faces:
        faces   number of cells
            4   12216
            5   6431
            6   6864
            7   386
            8   415
            9   6591
           10   139
           11   134
           12   3677
           13   43
           14   40
           15   3021
           16   7
           17   7
           18   258
           20   1
           21   30
           24   4

Checking topology...
    Boundary definition OK.
    Cell to face addressing OK.
    Point usage OK.
    Upper triangular ordering OK.
    Face vertices OK.
    Topological cell zip-up check OK.
    Face-face connectivity OK.
    Number of regions: 1 (OK).

Checking patch topology for multiply connected surfaces...
                   Patch    Faces   Points                  Surface topology Bounding box
                     il1      398      475  ok (non-closed singly connected) (-0.0371782 -0.174729 0.820446) (-0.0315461 -0.170118 0.824443)
            defaultFaces    52012    66767  ok (non-closed singly connected) (-0.0738687 -0.236453 0.725) (0.0819296 -0.12425 1.085)
                     ren      180      293  ok (non-closed singly connected) (0.0329212 -0.167896 0.997608) (0.0355591 -0.164201 1.00142)
                   inlet     5979     6306  ok (non-closed singly connected) (-0.0247617 -0.155035 1.085) (0.00522895 -0.12425 1.085)
                     il2     1317     1469  ok (non-closed singly connected) (0.0336118 -0.175452 0.813477) (0.0444973 -0.165195 0.8188)
                     il3      933     1032  ok (non-closed singly connected) (0.0733771 -0.212006 0.725) (0.0819296 -0.20347 0.725029)
                     il4     1037     1148  ok (non-closed singly connected) (-0.0723222 -0.218663 0.724997) (-0.0639713 -0.208937 0.725006)

Checking geometry...
    Overall domain bounding box (-0.0738687 -0.236453 0.724997) (0.0819296 -0.12425 1.085)
    Mesh (non-empty, non-wedge) directions (1 1 1)
    Mesh (non-empty) directions (1 1 1)
    Boundary openness (5.27378e-16 -5.96184e-16 3.35583e-17) OK.
    Max cell openness = 3.42542e-16 OK.
    Max aspect ratio = 9.17275 OK.
    Minimum face area = 4.24086e-09. Maximum face area = 1.61588e-05.  Face area magnitudes OK.
    Min volume = 1.24757e-12. Max volume = 5.47073e-08.  Total volume = 0.000215698.  Cell volumes OK.
    Mesh non-orthogonality Max: 64.62 average: 13.9596
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 2.96129 OK.
    Coupled point location match (average 0) OK.
 ***Error in face tets: 4 faces with low quality or negative volume decomposition tets.
  <<Writing 4 faces with low quality or negative volume decomposition tets to set lowQualityTetFaces
    Min/max edge length = 1.3e-05 0.00451136 OK.
   *There are 2416 faces with concave angles between consecutive edges. Max concave angle = 67.4995 degrees.
  <<Writing 2416 faces with concave angles to set concaveFaces
    Face flatness (1 = flat, 0 = butterfly) : average = 0.996275  min = 0.693467
   *There are 13 faces with ratio between projected and actual area < 0.8
    Minimum ratio (minimum flatness, maximum warpage) = 0.693467
  <<Writing 13 warped faces to set warpedFaces
    Cell determinant (wellposedness) : minimum: 0.053324 average: 19.6533
    Cell determinant check OK.
 ***Concave cells (using face planes) found, number of cells: 12208
  <<Writing 12208 concave cells to set concaveCells

Failed 2 mesh checks.

End

I can now see the mesh isn't as good as thought... I guess I'll have to be more strict on the meshQualityControls.

l_r_mcglashan · August 14, 2013, 12:22

You can view the warpedFaces and lowQualityTetFaces sets by selecting "Include sets" in paraview. Are those faces where your bad cell is?

Hiroshiman · August 14, 2013, 12:40

There is one of the warpedFaces and 0 lowQualityTetFaces on the outlet where the simulation diverged.

August 14, 2013, 12:05		#2
l_r_mcglashan Senior Member Laurence R. McGlashan Join Date: Mar 2009 Posts: 370 Rep Power: 23	Could you run checkMesh with the options -allGeometry and -allTopology? cfdonline2mohsen and wht like this. __________________ Laurence R. McGlashan :: Website

August 14, 2013, 12:22		#4
l_r_mcglashan Senior Member Laurence R. McGlashan Join Date: Mar 2009 Posts: 370 Rep Power: 23	You can view the warpedFaces and lowQualityTetFaces sets by selecting "Include sets" in paraview. Are those faces where your bad cell is? __________________ Laurence R. McGlashan :: Website

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Problem with divergence	TDK	FLUENT	13	December 14, 2018 07:00
[Commercial meshers] failed checkMesh after converting from .msh: non closed cells	phsieh2005	OpenFOAM Meshing & Mesh Conversion	11	September 20, 2012 09:51
[snappyHexMesh] Layers:problem with curvature	giulio.topazio	OpenFOAM Meshing & Mesh Conversion	10	August 22, 2012 10:03
physical boundary error!!	kris	Siemens	2	August 3, 2005 01:32
Ignore cells on partition boundary	Karl	FLUENT	7	May 11, 2002 23:12

August 14, 2013, 12:40		#5
Hiroshiman Member Join Date: Sep 2012 Posts: 51 Rep Power: 14	There is one of the warpedFaces and 0 lowQualityTetFaces on the outlet where the simulation diverged.