[Gmsh] Hybrid mesh and combine triangles in quadrilaterals

[Christophe Geuzaine]

Durufle Marc wrote:
> Hi,
> I have two questions :
> 
> 1- How do you create hybrid meshes in 2-D (quad-dominant with a few 
> triangles) ?
> Until now, I am using the Recombine option, but often I get non-convex 
> quadrilaterals.
> In the picture attached to this message, you can see that two elements 
> (in black) are non-convex
> quads and that's a problem ... So, often I change the characterisc 
> length in the hope that I will have only
> fine quadrilaterals. Is it possible to require that all quadrilaterals 
> are convex ?
> 

Yes. In your file you use

Recombine Surface {1,s0[0]} = 100;

which means that you allow quads with angles between 90-100 and 90+100 
degrees (which does not make sense).

Use

Recombine Surface {1,s0[0]} = 45;

instead, which will allow angles between 90-45=45 and 90+45=135 degrees. 
(45 is the default if you don't specify anything, i.e., with just 
"Recombine Surface {1,s0[0]}".)

We are working on an automatic full quad algorithm, but it's not ready 
yet. In the meantime for simple geometries you could also use 
Transfinite meshes to generate a full quad mesh (see attached file).


> 2- In 3-D, when I try "Recombine Volume", it doesn't work. So, how is it 
> possible with gmsh
> to create 3-D hexahedral-dominant meshes (with some tets, pyramids and 
> wedges) ?

Non. Your only choices for generating hex/prism/pyramid meshes are 
extruded (see "Extrude Surface") or transfinite (see "Transfinite 
Volume") grids.



> 
> Thank you for your help.
> 
> PS : I am using gmsh 2.24
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> 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
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: ellipse_pml_2.geo
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20081022/40634e80/attachment.geo>