[Gmsh] Elements neighbors - physical surfaces question

Vitor Vasconcelos vitors.vasconcelos at gmail.com
Sun Jul 3 18:17:23 CEST 2016


Dear members,

I'm generating a mesh to be used by a finite volumes formulation of a
nuclear reactor code.

I'm experiencing errors when the software makes its checkings on the mesh.

The mesh is generated in gmsh 2.9.3 and is attached.

I'm stucked in the following situation:

The actual geo file, when meshed and loaded in the nuclear code gives
me the error:

-------    ------      -----       --       -
diffusion :: finite volumes 3D
building...
error: mesh inconsistency, element 1260 has less neighbors (5) than faces (6)

To solve this error, a added a Physical Surface (uncommented line 113)
Physical Surface(300) = {100}

After doing that, I get a new error:

-------    ------      -----       --       -
diffusion :: finite volumes 3D
building...
error: mesh inconsistency, element 1011 has at least one more neighbor
than faces (6)

I'm still learning how to use gmsh and could not managed to visualize
in GUI the surface 300.
However, I could check the elements and they are far appart in the mesh.

In the .msh file they're:

1011 5 2 3 1 146 438 430 145 247 857 849 248
1260 5 2 2 6 96 116 117 97 366 325 326 365

I'm not sure what to ask, but I'll try:

There is any form of check neighbor consistence of the generated mesh from gmsh?

I'll try to elaborate more my question next time.

If someone had similar problems before, would be great to hear from
you any suggestion.
And please, if I'm doing something wrong in my geo file, feel free to
let me know.

Thanks,

Vitor
-------------- next part --------------
A non-text attachment was scrubbed...
Name: subcanal.geo
Type: application/vnd.dynageo
Size: 3769 bytes
Desc: not available
URL: <http://onelab.info/pipermail/gmsh/attachments/20160703/328ea395/attachment.geo>


More information about the gmsh mailing list