CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Volume Mesh - Skewness? (

David Banks August 6, 2007 06:30

Volume Mesh - Skewness?
Hi everyone

I have successfully imported an STL file of a seabed bathymetric shape,made up from several faces stitched together and wrapped with 5 other flat faces to create a volume.

I already have face meshes and need to mesh the volume now but keep getting the same message;

"Initialisation failed; perturb boundary nodes and try again:

Error: TG_MESH_DOMAIN failed with error code 1

Error: Tetrahedral meshing has failed for volume v_volume.1 This is usually caused by a problem in the face meshes. Check the skewness of your face meshes and make sure the face mesh sizes are not too large in areas of small gaps"

Does this mean the mesh is no fine enough in certain faces? Can this be easily fixed?



nino August 6, 2007 07:21

Re: Volume Mesh - Skewness?
no it means that you have high-skewed cells. Check if you have small edges or small faces

David Banks August 6, 2007 08:12

Re: Volume Mesh - Skewness?
I have some relatively small faces. Do I have to reduce the amount of elements in the smaller cells?

nino August 6, 2007 09:31

Re: Volume Mesh - Skewness?
no. With this small faces, you enforce Gambit to mesh them, even if your cell size isn't appropriate. I would either refine your mesh on those faces (and use a Size Function for controling your mesh), or try to merge them (if it is possible)

cfdproject April 4, 2009 19:53

I am trying to mesh the continuum between the fins. and I am getting the same error.
I tried to assigning a size function I have a error of Equiskew Angle>0.97
Can anybosy help me

panda April 5, 2009 19:54

you may try to use Tgrid to mesh the volume, it would be easier to generate a good quality mesh.

-mAx- April 6, 2009 00:43

The question is why do you get cell's equiskew angle over 0.97.
It is possible that you have short edges, small angle etc...
Have you tried with other parameters in size function?
Have you tried without size function?

cfdproject April 6, 2009 19:45

My question is how would I be able to deal with a error message saying that the Equisize angle>0.97. and therefore the geometry could not be meshed.
I did try the geometry with varying size function and also without size function but I end up getting the same message..
Please help
Thank you

-mAx- April 7, 2009 00:28

Go and examine your mesh with the icon at the bottom right.
Select skewness and set the lower value with 0.9 (cells with higher value, may cause problems with the convergence)-
Gambit will show the cells with skewness over 0.9.
It is a good way for knowing where is the problem

cfdproject April 9, 2009 00:37

Thanks Max.
it works

onionmon February 17, 2011 09:33

Hi, I am having the same problem here.

I am meshing a 3D volume using the Tet/Hybrid on a TGrid. I have cleaned my geometry to remove small edges and small areas and small angles. I have implemented size functions where applicable to mesh small areas. My face mesh check shows no elements with an equisize skew > 0.9. A check using geometry command button>check vertices; geometry command button>check edges; geometry command button>check faces; geometry command button>check volumes show 0 failures in the validity check for all different geometries.

What else am I doing wrong for this not to mesh??

Thanks ahead for your help!

onionmon February 19, 2011 05:20

OK, I solved my problem. Here is my follow up.

I discovered a vertex from one surface that was very close to another surface so that a 3D tet element was not able to be created during 3D meshing. The vertex closeness to a surface is shown in a rough diagram below.


|-----|    |-----|
|      |_._|      |

I solved the problem by zooming out so I can see my entire system. I noticed a white vertex (multi-edged vertices are purple) sitting by itself in the middle of a face. I created a geometry and boolean-subtracted it from my base geometry so that the vertex was removed. Then I was able to mesh.

All times are GMT -4. The time now is 10:24.