[Gmsh] Mesh by physical

Christophe Geuzaine cag32 at case.edu
Tue Dec 5 23:13:54 CET 2006


Duane Malcolm wrote:
> Hi Christophe, hi Jean-François,
> 
> We are using Gmsh to a tree structure. But it takes too long to mesh the 
> entire tree at once.
> 
> So we have written a perl script that breaks the tree into individual 
> branches, generates a .geo file for each branch, run Gmsh on the .geo 
> file, and then rejoin all the branches by matching the face nodes.
> 

Hi Duane - If you add up the time it takes to mesh the separate parts, 
you get less than the time it takes to mesh the whole model?

(If this is the case, could you send me or Jean-Francois a sample model? 
This is definitely not normal behavior.)

Cheers,

Christophe



> We are thinking there must be a better way to do this in Gmsh.
> 
> Is there a better way you know of?
> 
> If not, I was thinking of meshing a geometry according to physical 
> volumes. Each branch would be a different physical volume and Gmsh would 
> loop through each physical volume and mesh it. Gmsh will maintain common 
> nodes at the face between branches. What do you think of this approach?
> 
> If you like it, can I implement it in the Gmsh code and can you give me 
> some pointers on where to start?
> 
> Thanks for your time.
> 
> Kind regards, Duane Malcolm
> 
> _______________________________________________
> 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