CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ParaView (https://www.cfd-online.com/Forums/paraview/)
-   -   [OpenFOAM] Difference in the number of cells between ParaView and checkMesh (https://www.cfd-online.com/Forums/paraview/156916-difference-number-cells-between-paraview-checkmesh.html)

MartinBlx July 15, 2015 07:39
Difference in the number of cells between ParaView and checkMesh
 
Hello everyone,

I'm comparing the efficiency of different snappyHexMesh meshes and to do so I'm paying attention to the number of cells of the mesh. And I made the weird discovery that checkMesh and ParaView don't agree on the number of cells of the mesh!

For example, for the same mesh, the Statistics inspector of ParaView gives 3.995.096 cells:

https://drive.google.com/file/d/0B_T...ew?usp=sharing

While the checkMesh result gives 1.105.136 cells:

Code:
Create time

Create polyMesh for time = 0

Time = 0

Mesh stats
    points:          1593493
    faces:            3707462
    internal faces:  3262532
    cells:            1105136
    faces per cell:  6.3069106
    boundary patches: 3
    point zones:      0
    face zones:      0
    cell zones:      0
Which one should I trust?

And any explanations for this huge difference? I have compared several meshes and the ratio is typically around 4 in favor of ParaView everytime.

Martin

user10600 January 18, 2016 04:46
Hello everyone !

I have noticed exactly the same thing a few days ago! Does someone has an idea why there is such a difference ?

I would tend to trust more the data informations given by "checkMesh", but should I ?


Thanks in advance.

Antimony January 18, 2016 04:54
Hi,

If I remember correctly (unfortunately I can't remember where I read it though), Paraview and the rest can't handle polyhedral cells, and instead decompose it to tets and the like.

Consequently the number reported by Paraview would be higher, because of this decomposition.

If I can find the link to the post on this, I will put it here.

Hope this helps.

Cheers,
Antimony

user10600 January 18, 2016 05:01
Wow it makes sense!

Thanks a lot!

himanshu28 January 29, 2017 15:31
Quote:
Originally Posted by Antimony (Post 581355)
Hi,

If I remember correctly (unfortunately I can't remember where I read it though), Paraview and the rest can't handle polyhedral cells, and instead decompose it to tets and the like.

Consequently the number reported by Paraview would be higher, because of this decomposition.

If I can find the link to the post on this, I will put it here.

Hope this helps.

Cheers,
Antimony
Hi Antimony,

Did you find the link.

Thank you.

Traction January 30, 2017 01:57
Quote:
Originally Posted by himanshu28 (Post 635084)
Hi Antimony,

Did you find the link.

Thank you.
Hey,
this topic has been analyzed here (with a few examples as well): https://www.cfd-online.com/Forums/en...gion-case.html


All times are GMT -4. The time now is 03:21.