[Gmsh] Fails to mesh 3d region with hole.

Christophe Geuzaine cgeuzaine at ulg.ac.be
Tue Apr 8 19:25:02 CEST 2008


Martin Aunskjær wrote:
> Hi,
> 
> I'm having trouble generating a 3D mesh of a bounding
> box with a hole (another closed volume) embedded
> inside of it. Gmsh does generate a mesh, but it meshes
> the entire bounding box, including the inside of the
> hole. I believe this will happen if the volume
> enclosing the hole is not closed. Thus, I tried
> dumping the 2D surface mesh, converted it to a .poly
> file for TetGen and sure enough TetGen detects an open
> face in the interior volume and no mesh is generated
> as all tets are eaten away. I would appreciate some
> assistance in understanding the following:
> 
> 1) Is there something foolish about the way I generate
> the geometry (see included nose_in_box.geo)? The
> surface bounding the interior hole seems to me to be
> closed. Or is it gmsh that produces an erroneous
> surface mesh that will make the 3D meshing fail ?


No, you have duplicate points in your geometry.

Gmsh can take care of that for you if you add

Coherence;

at the end of your .geo file. (But it's probably wiser to change the way 
you generate your input file so that you don't have duplicate entities 
to begin with.)








> 
> 2) I do not understand the remark in the user guide
> about consistent orientation of surfaces when creating
> a surface loop. Supposedly, this means that the
> normals of all elementary surfaces of a surface loop
> should all either point into or out of the volume it
> encloses. Is it so that the surface normals should
> point into the volume being meshed; i.e. for an
> exterior bounding box they should point inward, and
> for an inside hole they should point outward? And what
> determines the direction of a surface normal compared
> to the way its underlying line loop is defined ?
> 
> Thanks in advance,
> Martin Aunskjaer
> 
> 
>       Trænger du til at se det store billede? Kelkoo giver dig gode tilbud på LCD TV! Se her http://dk.yahoo.com/r/pat/lcd
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> gmsh mailing list
> gmsh at geuz.org
> http://www.geuz.org/mailman/listinfo/gmsh


-- 
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine