[Gmsh] Possible bug when extruding relatively small surfaces

Christophe Geuzaine cgeuzaine at ulg.ac.be
Wed Sep 21 20:45:14 CEST 2016

Dear Kate,

This is "normal" : by default the tolerance to distinguish separate points/vertices is set to 1e-6 times the size of the model bounding box. Set e.g. 

Geometry.Tolerance = 1e-12;

at the beginning of your script.


> On 21 Sep 2016, at 14:26, look <mlook at me.com> wrote:
> Hello all,
> I think I found a bug related to the extrusion (and recombination) of relatively small surfaces.
> The problem occurs when you try to extrude very small 2-D cells into the third dimension while having a huge domain, like it is often necessary for a resolved boundary layer mesh for a 2-D airfoil simulation.
> Gmsh then fails to recombine the extruded mesh into quadrangles. 
> You can test this with the attached minimal example. It creates the mesh well until you uncomment the last line. You can then see in the GUI statistics-panel, that gmsh created triangles, which is obviously not intended.
> Please see also the related discussion at cfd-online.com:
> http://www.cfd-online.com/Forums/openfoam-solving/164600-gmshtofoam-fatal-error.html
> Best regards,
> Kate
> <main.geo.tar.gz>_______________________________________________
> gmsh mailing list
> gmsh at onelab.info
> http://onelab.info/mailman/listinfo/gmsh

Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science 

Free software: http://gmsh.info | http://getdp.info | http://onelab.info

More information about the gmsh mailing list