[Gmsh] Possible Bug with Orientations

Christophe Geuzaine geuzaine at acm.caltech.edu
Thu May 15 23:11:58 CEST 2003


Archi Forshop wrote:
> Christophe,
> 
> 
>>In the meantime,
>>you can simply change the definition of the Physical
>>Surface to control
>>the orientation, e.g.,
>>
>>Physical Surface(surfPhys) = {s1,-s2,s3,s4,-s5,s6};
>>
> 
> 
> I've tried that too. Didn't work. 

Yes it did :-)

The sign changes take place when you *save* the file. Try to reload the
.msh file from disk: you'll see that the orientation changes have indeed
been taken into account...

> I can see however
> from the code that it should react to the sign of {s}.
> I'll need to go through the code to check where it's
> been lost.
> 
> Christophe, would you please provide some references
> regarding the meshing (3D) algorithm that is used
> within gmsh, if available?
> 
> ...Roman
> 
> P.S. 1) Orientations are needed because my FEM is just
> a piece of the overall computation; 2) I just wish
> there were some technical docs on Gmsh.
> 

Agreed, but this takes time... Meanwhile, you can find a good reference
for the 3D algorithm in "George, P. L. and Borouchaki, H., Delaunay
Triangulation and Meshing. Editions Hermes, Paris, 1998. ISBN
2-86601-692-0".

Christophe

-- 
Christophe Geuzaine

Tel: (626) 395-4552    http://www.geuz.org
Fax: (626) 578-0124    mailto:geuzaine at acm.caltech.edu