[jet_engine]

Dear all, I'm running snappyHexMesh to generate a mesh around a plate and a cylinder, but I'm apparently obtaining a bad resulting mesh. After running blockMesh, checkMesh's report looks fine:

Checking geometry...
    Overall domain bounding box (-0.225 -0.35 -0.125) (1.12 0.35 0.84)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (3.71949e-18 -6.24427e-17 1.8895e-17) OK.
    Max cell openness = 1.849e-16 OK.
    Max aspect ratio = 1.07621 OK.
    Minimum face area = 0.0032691. Maximum face area = 0.00351823.  Face area magnitudes OK.
    Min volume = 0.000197167. Max volume = 0.000197167.  Total volume = 0.908547.  Cell volumes OK.
    Mesh non-orthogonality Max: 0 average: 0
    Non-orthogonality check OK.
    Face pyramids OK.
    Max skewness = 2.02586e-14 OK.
    Coupled point location match (average 0) OK.

Mesh OK.

End

But after applying snappyHexMesh, the checkMesh output is quite different though, showing up to 10 errors...

Enabling all geometry checks.

Time = 0

Mesh stats
    points:           26716253
    faces:            73469248
    internal faces:   70259631
    cells:            23392159
    faces per cell:   6.14432
    boundary patches: 7
    point zones:      0
    face zones:       0
    cell zones:       0

Overall number of cells of each type:
    hexahedra:     21925085
    prisms:        164543
    wedges:        0
    pyramids:      0
    tet wedges:    769
    tetrahedra:    0
    polyhedra:     1301762
    Breakdown of polyhedra by number of faces:
        faces   number of cells
            4   2466
            5   2193
            6   163374
            7   141807
            8   6875
            9   857584
           10   314
           11   136
           12   108408
           14   87
           15   18158
           18   360

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
                   inlet    22488    24182  ok (non-closed singly connected) (-0.225 -0.35 -0.125) (-0.225 0.35 0.84)
                  outlet    11398    12527  ok (non-closed singly connected) (1.12 -0.35 -0.125) (1.12 0.35 0.84)
             lateralWall   255501   280238  ok (non-closed singly connected) (-0.225 -0.35 -0.125) (1.12 0.35 0.84)
               upperWall      288      325  ok (non-closed singly connected) (-0.225 -0.35 0.84) (1.12 0.35 0.84)
               lowerWall     1152     1225  ok (non-closed singly connected) (-0.225 -0.35 -0.125) (1.12 0.35 -0.125)
                Cylinder   280699   285958  ok (non-closed singly connected) (-0.0601417 -0.0599639 -3.35276e-10) (0.0600005 0.0599639 0.24)
             Plate_round  2638091  2665903  ok (non-closed singly connected) (-0.223998 -0.35 -0.015) (1.12 0.35 0.000998369)

Checking geometry...
    Overall domain bounding box (-0.225 -0.35 -0.125) (1.12 0.35 0.84)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (-5.87433e-16 6.97082e-17 -1.48306e-13) OK.
 ***High aspect ratio cells found, Max aspect ratio: 1.24904e+94, number of cells 35
  <<Writing 35 cells with high aspect ratio to set highAspectRatioCells
    Minimum face area = 5.21919e-10. Maximum face area = 0.00357495.  Face area magnitudes OK.
 ***Zero or negative cell volume detected.  Minimum negative volume: -6.5735e-10, Number of negative volume cells: 35
  <<Writing 35 zero volume cells to set zeroVolumeCells
    Mesh non-orthogonality Max: 179.989 average: 7.91616
   *Number of severely non-orthogonal (> 70 degrees) faces: 115.
 ***Number of non-orthogonality errors: 109.
  <<Writing 224 non-orthogonal faces to set nonOrthoFaces
 ***Error in face pyramids: 240 faces are incorrectly oriented.
  <<Writing 183 faces with incorrect orientation to set wrongOrientedFaces
 ***Max skewness = 16.2622, 20 highly skew faces detected which may impair the quality of the results
  <<Writing 20 skew faces to set skewFaces
    Coupled point location match (average 0) OK.
 ***Error in face tets: 717 faces with low quality or negative volume decomposition tets.
  <<Writing 318 faces with low quality or negative volume decomposition tets to set lowQualityTetFaces
    Min/max edge length = 2.07766e-05 0.0607762 OK.
   *There are 66069 faces with concave angles between consecutive edges. Max concave angle = 89.3598 degrees.
  <<Writing 66069 faces with concave angles to set concaveFaces
    Face flatness (1 = flat, 0 = butterfly) : min = 0.0400623  average = 0.999966
   *There are 86 faces with ratio between projected and actual area < 0.8
    Minimum ratio (minimum flatness, maximum warpage) = 0.0400623
  <<Writing 86 warped faces to set warpedFaces
    Cell determinant (wellposedness) : minimum: 9.01638e-05 average: 10.1297
 ***Cells with small determinant (< 0.001) found, number of cells: 4
  <<Writing 4 under-determined cells to set underdeterminedCells
 ***Concave cells (using face planes) found, number of cells: 590857
  <<Writing 590857 concave cells to set concaveCells
    Face interpolation weight : minimum: 0.01328 average: 0.478489
 ***Faces with small interpolation weight (< 0.05) found, number of faces: 11
  <<Writing 11 faces with low interpolation weights to set lowWeightFaces
    Face volume ratio : minimum: -10.9825 average: 0.902224
 ***Faces with small volume ratio (< 0.01) found, number of faces: 142
  <<Writing 142 faces with low volume ratio cells to set lowVolRatioFaces

Failed 10 mesh checks.

End

The snappyHexMesh dict and a picture of the case around which I want to do the meshing are attached to this post as well. Does anybody know what might be causing these checks to fail? Thanks in advance!!

PS: An UPDATE--> I have tried separately the 3 steps and the errors just appear after the addLayers step, the others work well, hope that helps!

[Atzori]

Hi!

With the snappyHexMeshDict only I cannot reproduce the case, so I tryed to guess, but I cannot be sure. Anyway:

Try for instance to increase the minThickness:

minThickness 0.05; => minThickness 0.3;

With the option relativeSizes true; you are asking to produce the first layer that is 0.05 of a cell, maybe "he" gives up.
I don't know the size of the underlying blocks, but considering the refinment levels and nCellsBetweenLevels 3; my hypothesis is that the cells in the layers are simply too small.

Best Regards!

[jet_engine]

Originally Posted by Atzori

Hi!

With the snappyHexMeshDict only I cannot reproduce the case, so I tryed to guess, but I cannot be sure. Anyway:

Try for instance to increase the minThickness:

minThickness 0.05; => minThickness 0.3;

With the option relativeSizes true; you are asking to produce the first layer that is 0.05 of a cell, maybe "he" gives up.
I don't know the size of the underlying blocks, but considering the refinment levels and nCellsBetweenLevels 3; my hypothesis is that the cells in the layers are simply too small.

Best Regards!

Dear Marco,

Thank you for your answer, I've proceded to try it and here is the output:

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
                   inlet    22488    24182  ok (non-closed singly connected) (-0.225 -0.35 -0.125) (-0.225 0.35 0.84)
                  outlet     9294    10419  ok (non-closed singly connected) (1.12 -0.35 -0.125) (1.12 0.35 0.84)
             lateralWall   210533   235262  ok (non-closed singly connected) (-0.225 -0.35 -0.125) (1.12 0.35 0.84)
               upperWall      288      325  ok (non-closed singly connected) (-0.225 -0.35 0.84) (1.12 0.35 0.84)
               lowerWall     1152     1225  ok (non-closed singly connected) (-0.225 -0.35 -0.125) (1.12 0.35 -0.125)
                Cylinder   280699   285958  ok (non-closed singly connected) (-0.0601417 -0.0599639 -3.35276e-10) (0.0600005 0.0599639 0.24)
             Plate_round  2638091  2665903  ok (non-closed singly connected) (-0.223998 -0.35 -0.015) (1.12 0.35 0.000998369)

Checking geometry...
    Overall domain bounding box (-0.225 -0.35 -0.125) (1.12 0.35 0.84)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (-6.15461e-16 -2.68982e-15 -7.85125e-14) OK.
 ***High aspect ratio cells found, Max aspect ratio: 4.6537e+92, number of cells 2
  <<Writing 2 cells with high aspect ratio to set highAspectRatioCells
    Minimum face area = 1.95889e-09. Maximum face area = 0.00355656.  Face area magnitudes OK.
 ***Zero or negative cell volume detected.  Minimum negative volume: -6.44127e-13, Number of negative volume cells: 2
  <<Writing 2 zero volume cells to set zeroVolumeCells
    Mesh non-orthogonality Max: 169.126 average: 8.14453
   *Number of severely non-orthogonal (> 70 degrees) faces: 16.
 ***Number of non-orthogonality errors: 32.
  <<Writing 48 non-orthogonal faces to set nonOrthoFaces
 ***Error in face pyramids: 64 faces are incorrectly oriented.
  <<Writing 40 faces with incorrect orientation to set wrongOrientedFaces
 ***Max skewness = 31.6768, 2 highly skew faces detected which may impair the quality of the results
  <<Writing 2 skew faces to set skewFaces
    Coupled point location match (average 0) OK.
 ***Error in face tets: 240 faces with low quality or negative volume decomposition tets.
  <<Writing 106 faces with low quality or negative volume decomposition tets to set lowQualityTetFaces
    Min/max edge length = 2.99844e-05 0.0606271 OK.
   *There are 59376 faces with concave angles between consecutive edges. Max concave angle = 89.8234 degrees.
  <<Writing 59376 faces with concave angles to set concaveFaces
    Face flatness (1 = flat, 0 = butterfly) : min = 0.11079  average = 0.999975
   *There are 85 faces with ratio between projected and actual area < 0.8
    Minimum ratio (minimum flatness, maximum warpage) = 0.11079
  <<Writing 85 warped faces to set warpedFaces
    Cell determinant (wellposedness) : minimum: 0.0342967 average: 10.9159
    Cell determinant check OK.
 ***Concave cells (using face planes) found, number of cells: 574493
  <<Writing 574493 concave cells to set concaveCells
    Face interpolation weight : minimum: 0.0329211 average: 0.477303
 ***Faces with small interpolation weight (< 0.05) found, number of faces: 2
  <<Writing 2 faces with low interpolation weights to set lowWeightFaces
    Face volume ratio : minimum: -1.3994 average: 0.894135
 ***Faces with small volume ratio (< 0.01) found, number of faces: 16
  <<Writing 16 faces with low volume ratio cells to set lowVolRatioFaces

Failed 9 mesh checks.

End

Apparently several things went better and 1 mesh check has been deleted, but many others remain the same. Unfortunately, my simulation cannot run (I get the famous floating point exception error), and I've checked if it does without the addLayers step and the answer is yes, with decent residuals (~e-07). Do you know what could be wrong or causing the issue apart from the parameter that I changed? I could upload the log file of snappyhexmesh if it helps.

Thanks again!

[Atzori]

Hi Max!

Let me to make a couple of questions to better understand:

1) Are you performing RANS/URANS, right? Are you using wallfunctions as well?

2) In that case, are you sure that so high refinement levels (=> number of cells) are needed? If not, try simply to change wildly nCellsBetweenLevels 3 => nCellsBetweenLevels 2, and consider if the mesh is still good enough or not.

3) If I understand, you obtain a good convergence without layers: so, why do you want to add them?

Cheers!

Marco

[jet_engine]

Originally Posted by Atzori

Hi Max!

Let me to make a couple of questions to better understand:

1) Are you performing RANS/URANS, right? Are you using wallfunctions as well?

2) In that case, are you sure that so high refinement levels (=> number of cells) are needed? If not, try simply to change wildly nCellsBetweenLevels 3 => nCellsBetweenLevels 2, and consider if the mesh is still good enough or not.

3) If I understand, you obtain a good convergence without layers: so, why do you want to add them?

Cheers!

Marco

Dear Marco,

Yes, I am using wall functions for the variables that need them. My simulation is a SpalartAllmaras DDES.

There is no need for a perfect refinement, and therefore I have done the change you mentioned, but the output did not change very much:

Checking geometry...
    Overall domain bounding box (-0.225 -0.35 -0.125) (1.12 0.35 0.84)
    Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
    Mesh has 3 solution (non-empty) directions (1 1 1)
    Boundary openness (-6.15461e-16 -2.68982e-15 -7.85125e-14) OK.
 ***High aspect ratio cells found, Max aspect ratio: 4.6537e+92, number of cells 2
  <<Writing 2 cells with high aspect ratio to set highAspectRatioCells
    Minimum face area = 1.95889e-09. Maximum face area = 0.00355656.  Face area magnitudes OK.
 ***Zero or negative cell volume detected.  Minimum negative volume: -6.44127e-13, Number of negative volume cells: 2
  <<Writing 2 zero volume cells to set zeroVolumeCells
    Mesh non-orthogonality Max: 169.126 average: 8.14453
   *Number of severely non-orthogonal (> 70 degrees) faces: 16.
 ***Number of non-orthogonality errors: 32.
  <<Writing 48 non-orthogonal faces to set nonOrthoFaces
 ***Error in face pyramids: 64 faces are incorrectly oriented.
  <<Writing 40 faces with incorrect orientation to set wrongOrientedFaces
 ***Max skewness = 31.6768, 2 highly skew faces detected which may impair the quality of the results
  <<Writing 2 skew faces to set skewFaces
    Coupled point location match (average 0) OK.
 ***Error in face tets: 240 faces with low quality or negative volume decomposition tets.
  <<Writing 106 faces with low quality or negative volume decomposition tets to set lowQualityTetFaces
    Min/max edge length = 2.99844e-05 0.0606271 OK.
   *There are 59376 faces with concave angles between consecutive edges. Max concave angle = 89.8234 degrees.
  <<Writing 59376 faces with concave angles to set concaveFaces
    Face flatness (1 = flat, 0 = butterfly) : min = 0.11079  average = 0.999975
   *There are 85 faces with ratio between projected and actual area < 0.8
    Minimum ratio (minimum flatness, maximum warpage) = 0.11079
  <<Writing 85 warped faces to set warpedFaces
    Cell determinant (wellposedness) : minimum: 0.0342967 average: 10.9159
    Cell determinant check OK.
 ***Concave cells (using face planes) found, number of cells: 574493
  <<Writing 574493 concave cells to set concaveCells
    Face interpolation weight : minimum: 0.0329211 average: 0.477303
 ***Faces with small interpolation weight (< 0.05) found, number of faces: 2
  <<Writing 2 faces with low interpolation weights to set lowWeightFaces
    Face volume ratio : minimum: -1.3994 average: 0.894135
 ***Faces with small volume ratio (< 0.01) found, number of faces: 16
  <<Writing 16 faces with low volume ratio cells to set lowVolRatioFaces

Failed 9 mesh checks.

And therefore I am kinda lost now... The reason I was adding layers was because I thought that the near-wall treatment would be more precise, but if I am obtaining this kind of errors I do not know what do do.

Thank you again for your response.

Best regards,

Max.

[Atzori]

Hi Max!
The world is small I’m using now a IDDES

About the refinement:
In my personal experience the refinement could have an influence, depending also on the Re and on what you are more interested in, and in my case layers are needed… I think that first of all is better to have a clue about the importance of the layers in this case.

Please, could you check what is your yPlus, without layers?

BTW, do you have a reference case that can be used to check your results? Sometime “good residuals” do not mean “good results” …

About the dictionary:
I’m surprise a bit that the change in nCellsBetweenLevels has apparently no effects on the mesh: what is the total number of cells now? It is reduced has it should?

Cheers!!
Marco Atzori

[Hemmati]

Hi. l run snappyHexMeshDict -overwrite then checkMesh, but similar error happend.
please help me.
Best wishes

[Hemmati]

Hi Max.
Is stl files at the origin is drawn?

[Atzori]

Dear Hemmati,

Could you please give us more info about your case?

(In my experience the .stl is rarely the main source of the problems, more often they are caused by some parameters in the mesher dictionary but without precise info is impossible to give precise suggestions )

Best!

Marco

[Hemmati]

Hi Marco.
I modelled the steeped spillway. I draw the basic geometry in SolidWorks. But my mistake was that I did not draw in the origin. If the initial geometry drawing in the origin, most likely, CheckMesh after snappyHexMesh will be ok.

Best...