[Gmsh] Malformed cells at boundary between surfaces

Christophe Geuzaine cgeuzaine at ulg.ac.be
Sun Mar 17 23:05:18 CET 2013


On 17 Mar 2013, at 21:10, Johannes Reinhardt <johannes.reinhardt at unifr.ch> wrote:

> Hi everybody,
> 
> I try to generate a 2D quad mesh for a geometry that consists of
> several surfaces. However sometimes I trigger a bug that results
> in a mesh with malformed cells at a boundary between surfaces.
> 
> I attached one of the geometries where I run into this problem. The
> artifacts appear at the top left and bottom right inner
> boundary of the dumbbell shaped inner surface. See also the
> attached screenshots. Slightly changing some of the mesh element sizes
> for some points sometimes helps. It also depends on the meshing
> algorithm: For this geometry MeshAdapt, DelQuad and BAMG have
> artifacts, Frontal and Delaunay not.
> 
> I am running gmsh 2.7.0 Linux 64bit precompiled binary on Ubuntu 11.10
> 64bit. I mesh the geometry using the commandline
> 
> gmsh -2 -format msh1 artifact.geo
> 

Hi Johannes,

Indeed, the all-quad algo is still very much a work-in-progress. I would recommend using the Frontal algorithm instead of MeshAdapt.

You can also give the experimental DelQuad triangulation algorithm a try (Mesh.Algorithm=8;): it is specifically designed to generate triangulations for later recombination into quads. 

Christophe

> If I mesh from the gui, the artifacts are not visible. I need to use the
> msh1 format, because the solver that I use only reads msh1.
> 
> 
> Regards
> 
> Johannes Reinhardt
> 
> 
> 
> -- 
> Johannes Reinhardt <johannes.reinhardt at unifr.ch>
> <Artifact_close_up.png><Artifact_overview.png><artifact.geo><artifact.msh>_______________________________________________
> 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