[Gmsh] Unable to mesh layered sphere with netgen algorithm

[Christophe Geuzaine]

Julian Swartz wrote:
> Hi,
> 
> I would like to use the NETGEN algorithm to mesh a layered sphere (.geo 
> file attached). It is stated that all surface normals need to be 
> outwardly directed in order for the NETGEN algorithm to work. The 
> layered spheres are built by extruding semicircles in such a way that 
> both the sphere and the outer layer consist of three volumes. I can 
> ensure that the normals of the sphere outer surfaces are outward 
> pointing by extruding in the direction of -theta. How should I treat the 
> surfaces bounding the sides of the three segments making up the sphere? 
> The concept of an outwardly directed normal seems to lose meaning here 
> and thusfar all attempts to mesh the domain with NETGEN has resulted in 
> gmsh crashing.
> 
> Any ideas?
> 

To change the orientations you just need to reference the surfaces with 
the correct sign in the surface loops. Here is a modified version of 
your geometry, with all the volumes defined in terms of surfaces having 
outward normals.

(Obviously, we need to make this orientation step automatic. If you have 
some spare time it would be great if you could help us out with that...)


> Julian
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> gmsh mailing list
> gmsh at geuz.org
> http://www.geuz.org/mailman/listinfo/gmsh


-- 
Christophe Geuzaine
Assistant Professor, Case Western Reserve University, Mathematics
http://www.case.edu/artsci/math/geuzaine
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: sphere_layer_netgen.geo
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20060130/3df98181/attachment.geo>