[chaos3]

A few days ago when I first created the snappyHexMesh of my geometry, it was fine and there were no floating cells above and below the mesh and I decided to run my simulation in parallel and after that, the mesh just seemed to have this error of having floating cells. I am not sure if it was because of running in parallel because I have no idea. I tried to use a smaller resolve angle, lower the toleration, load the castellated mesh first, then snap, etc., but it wouldn't solve itself. I don't honestly know if this is an issue because my simulation runs fine but I find it odd and ideally would like to fix it. To make this clear, the odd floating cell artifacts only appear on the internal mesh.

[Yann]

Hello Joseph,

Having screenshots showing the issue would help to understand what kind of artifacts you are talking about.

Regards,
Yann

[chaos3]

Hello Yann, thanks for reaching out

https://imgur.com/a/IiuaKs3

The circled floating cells are the artifacts I am talking about and appear when I select the internal mesh in paraview.

I will also send the log file for my snappyHexMesh.

https://files.catbox.moe/gzvswt.gz

[Yann]

Hell Joseph,

Not sure what is going on.
How do you obtain this view ? Wireframe mode in paraview ? On the internal mesh, on the boundaries, or on a slice ?
And where are the cells located? Close to a surface ?

Yann

[chaos3]

Hey Yann,

This is a wireframe view of the internal mesh on the upper surface of my internal mesh and then the disconnected cells are floating right above the top of my mesh, yes it is close to the surface of my internal mesh. This is not a slice of it, a slice of the mesh wouldn't show the artifacts.

[Yann]

Well, I'm not sure what is going on.
Would it be possible to share your mesh here?

[AtoHM]

What Yann is too polite to say: its hard to tell anything from this image . Look into other threads here to see what kind of images people provide and try to replicate. Maybe add an overview and 2 or three sections with slices using crincle option in paraview. If it still shows there, we might get a hang what it is.
Also, please attach files to the forum post directly, especially images. You can paste dictionary as plain text too.


Some random examples how to provide helpful images:
How to generate Layers at the corner
No layer in corners and high skewness
snappyhexmesh with some mesh cells missing

[chaos3]

Hello AtoHM,

I thought the circles might be enough, but I guess not haha, I can provide some more zoomed out images, if that helps.

Here, I have a 3d blockMesh domain surrounding the airfoil that I am simulating and you (I hope) can see the these floating cells in this wireframe view and the last surface view, and again, this only occurs in the internalMesh option selected. Slices do not show this artifact. The second image is of the surface of the airfoil, not the internalMesh, and it is clear that there are no artifacts on this one.

[AtoHM]

Hello Joseph,

way better! These spots are mysterious. If you cannot spot them on slices, I assume they are not actually there. Maybe try another Paraview version and see if they still exist? Can you grab the location of these edges in Paraview and check your simulation results at these locations for anything weird?

[Yann]

Originally Posted by AtoHM

What Yann is too polite to say: its hard to tell anything from this image .

My dear AtoHM, it sounds like you are overestimating my politeness!

I'm still not sure what is going on with the additional screenshot.
I'd be tempted to think this is mostly a visualization issue rather than an actual problem with the mesh.

Joseph, in addition to AtoHM's recommendations, could you post a checkMesh log?
Also, since you mentioned this happens only when you run in parallel: are you loading the decomposed or reconstructed mesh in ParaView? Do you still see the artifacts when doing it the other way around?

[chaos3]

/*---------------------------------------------------------------------------*\
========= |
\\ / F ield | OpenFOAM: The Open Source CFD Toolbox
\\ / O peration | Website: https://openfoam.org
\\ / A nd | Version: 9
\\/ M anipulation |
\*---------------------------------------------------------------------------*/
Build : 9-8dd6acb16f4b
Exec : checkMesh
Date : Aug 08 2023
Time : 04:57:27
Host : "localhost.localdomain"
PID : 2516
I/O : uncollated
Case : /home/cfd-user/OpenFOAM/OpenFOAM-9/CFD_1_revised
nProcs : 1
sigFpe : Enabling floating point exception trapping (FOAM_SIGFPE).
fileModificationChecking : Monitoring run-time modified files using timeStampMaster (fileModificationSkew 10)
allowSystemOperations : Allowing user-supplied system call operations

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

Create polyMesh for time = 0

Time = 0

Mesh stats
points: 9261
faces: 25200
internal faces: 22800
cells: 8000
faces per cell: 6
boundary patches: 6
point zones: 0
face zones: 0
cell zones: 0

Overall number of cells of each type:
hexahedra: 8000
prisms: 0
wedges: 0
pyramids: 0
tet wedges: 0
tetrahedra: 0
polyhedra: 0

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
side2 400 441 ok (non-closed singly connected)
inlet 400 441 ok (non-closed singly connected)
side1 400 441 ok (non-closed singly connected)
side3 400 441 ok (non-closed singly connected)
outlet 400 441 ok (non-closed singly connected)
side4 400 441 ok (non-closed singly connected)

Checking geometry...
Overall domain bounding box (-1.5 -1 -0.7) (1.2 0.5 0.8)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (-2.7236e-18 1.26543e-17 -1.00564e-17) OK.
Max cell openness = 1.71331e-16 OK.
Max aspect ratio = 1.8 OK.
Minimum face area = 0.005625. Maximum face area = 0.010125. Face area magnitudes OK.
Min volume = 0.000759375. Max volume = 0.000759375. Total volume = 6.075. Cell volumes OK.
Mesh non-orthogonality Max: 0 average: 0
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 1.11893e-14 OK.
Coupled point location match (average 0) OK.

Mesh OK.

--> FOAM Warning :
From function virtual Foam:olyMesh::readUpdateState Foam:olyMesh::readUpdate()
in file meshes/polyMesh/polyMeshIO.C at line 209
boundary changed, proceed with care.
Time = 5e-15

Mesh stats
points: 146740
faces: 360177
internal faces: 331344
cells: 107257
faces per cell: 6.44733
boundary patches: 7
point zones: 0
face zones: 0
cell zones: 0

Overall number of cells of each type:
hexahedra: 88966
prisms: 0
wedges: 0
pyramids: 0
tet wedges: 0
tetrahedra: 0
polyhedra: 18291
Breakdown of polyhedra by number of faces:
faces number of cells
6 6514
8 269
9 7772
10 41
11 144
12 2958
13 3
14 8
15 524
17 4
18 54

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
side2 400 441 ok (non-closed singly connected)
inlet 400 441 ok (non-closed singly connected)
side1 400 441 ok (non-closed singly connected)
side3 400 441 ok (non-closed singly connected)
outlet 508 564 ok (non-closed singly connected)
side4 400 441 ok (non-closed singly connected)
NACA 26325 27313 ok (closed singly connected)

Checking geometry...
Overall domain bounding box (-1.5 -1 -0.7) (1.2 0.5 0.8)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (8.97067e-17 1.11815e-17 -1.44111e-16) OK.
Max cell openness = 2.31296e-16 OK.
Max aspect ratio = 1.80107 OK.
Minimum face area = 2.1968e-05. Maximum face area = 0.010125. Face area magnitudes OK.
Min volume = 1.8521e-07. Max volume = 0.000759375. Total volume = 5.91383. Cell volumes OK.
Mesh non-orthogonality Max: 34.4898 average: 13.6718
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 1.00021 OK.
Coupled point location match (average 0) OK.

Mesh OK.

Time = 1e-14

Mesh stats
points: 142298
faces: 355355
internal faces: 331344
cells: 107257
faces per cell: 6.40237
boundary patches: 7
point zones: 0
face zones: 0
cell zones: 0

Overall number of cells of each type:
hexahedra: 84905
prisms: 3315
wedges: 0
pyramids: 0
tet wedges: 144
tetrahedra: 0
polyhedra: 18893
Breakdown of polyhedra by number of faces:
faces number of cells
4 38
5 598
6 6537
7 207
8 142
9 7692
10 2
11 136
12 2950
13 1
14 8
15 524
17 4
18 54

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
side2 400 441 ok (non-closed singly connected)
inlet 400 441 ok (non-closed singly connected)
side1 400 441 ok (non-closed singly connected)
side3 400 441 ok (non-closed singly connected)
outlet 508 564 ok (non-closed singly connected)
side4 400 441 ok (non-closed singly connected)
NACA 21503 22871 ok (closed singly connected)

Checking geometry...
Overall domain bounding box (-1.5 -1 -0.7) (1.2 0.5 0.8)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (4.93046e-18 3.43733e-17 1.35376e-16) OK.
Max cell openness = 3.50631e-16 OK.
Max aspect ratio = 8.16114 OK.
Minimum face area = 7.35658e-06. Maximum face area = 0.0102876. Face area magnitudes OK.
Min volume = 3.32478e-08. Max volume = 0.000772639. Total volume = 5.91293. Cell volumes OK.
Mesh non-orthogonality Max: 55.9184 average: 14.609
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 2.80216 OK.
Coupled point location match (average 0) OK.

Mesh OK.

Time = 1.5e-14

Mesh stats
points: 265907
faces: 698464
internal faces: 674556
cells: 217008
faces per cell: 6.32705
boundary patches: 7
point zones: 0
face zones: 0
cell zones: 0

Overall number of cells of each type:
hexahedra: 171407
prisms: 3400
wedges: 0
pyramids: 0
tet wedges: 146
tetrahedra: 0
polyhedra: 42055
Breakdown of polyhedra by number of faces:
faces number of cells
4 230
5 515
6 7598
7 17914
8 3071
9 8664
10 286
11 181
12 2982
13 22
14 6
15 524
16 4
17 4
18 54

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
side2 400 441 ok (non-closed singly connected)
inlet 400 441 ok (non-closed singly connected)
side1 400 441 ok (non-closed singly connected)
side3 400 441 ok (non-closed singly connected)
outlet 508 564 ok (non-closed singly connected)
side4 400 441 ok (non-closed singly connected)
NACA 21400 22724 ok (closed singly connected)

Checking geometry...
Overall domain bounding box (-1.5 -1 -0.7) (1.2 0.5 0.8)
Mesh has 3 geometric (non-empty/wedge) directions (1 1 1)
Mesh has 3 solution (non-empty) directions (1 1 1)
Boundary openness (-1.39448e-17 9.17593e-17 6.0023e-17) OK.
Max cell openness = 5.95927e-16 OK.
Max aspect ratio = 34.2641 OK.
Minimum face area = 1.17145e-06. Maximum face area = 0.0108978. Face area magnitudes OK.
Min volume = 5.25522e-09. Max volume = 0.000802761. Total volume = 5.91293. Cell volumes OK.
Mesh non-orthogonality Max: 73.2163 average: 14.8943
*Number of severely non-orthogonal (> 70 degrees) faces: 134.
Non-orthogonality check OK.
<<Writing 134 non-orthogonal faces to set nonOrthoFaces
Face pyramids OK.
Max skewness = 2.80216 OK.
Coupled point location match (average 0) OK.

Mesh OK.


I have posted my checkMesh here, by the way, this occurs in parallel but also not in parallel as well, it doesn't matter.

[Yann]

OK so let me guess:
You run snappyHexMesh without the overwrite option so snappy writes a new timestep for each phase of the meshing process.

The last timestep corresponds to to final mesh (after adding layers) and you have non orthogonality errors.

I'm not sure it is related to those artifacts you see in the mesh, but you can run this (or its parallel version if you ran in parallel):

checkMesh -latestTime -writeSets vtk

It will write a vtk file containing your non-orthogonal cells in the postProcessing directory. You can open it in ParaView in addition to your mesh, and see where these cells are located

[chaos3]

Ok, I will check it out, thanks.

Edit: I have looked and it and it contains the following, I didnt include the actual points since its meaningless here I imagine-- I also provided some images of this vtk file in paraview.

# vtk DataFile Version 2.0
sampleSurface
ASCII
DATASET POLYDATA
POINTS 364 float