[Gmsh] coplanar points problem

Christophe Geuzaine geuzaine at acm.caltech.edu
Wed Oct 6 07:40:05 CEST 2004


Krishna Mohan Gundu wrote:
> Dear Christophe Geuzaine,
> 
> I have attached pcf1.geo file. It basically has six "holes" in a
> planar slab. Note that, I realize these are not holes by gmsh
> definition. I generated this file using a program. First I create a
> slab, then subtract the holes one by one from it.
> 
> When I try to mesh it, gmsh gives coplanar points warnings, some other
> error and does not generate a mesh. Along the suggestion given in this
> forum I reduced the char. length for the hole nodes from 0.10 to 0.05
> (pcf2.geo). It again gives me coplanar points warnings and errors like
> 
> Info    : 1/6: Edge 3082->1942 => 0 division(s)
> Error   : Missing edge without any intersection (0.573162,0.190714,0)
> (0.530627,0.193289,0)
> I
> It generates a mesh this time. What do the above error mean? Can I
> reliably use the mesh despite the above errors? Also the resulting
> mesh size is huge. Is there any intelligible way of controlling this?

Krishna - I'll look into the problem when I have some spare time.
Meanwhile, you can always use the Netgen 3D algorithm (beware that all 
the surfaces that bound a volume have to be defined with exterior 
pointing normals in this case).

Best,

Christophe

-- 
Christophe Geuzaine
Applied and Computational Mathematics, Caltech
geuzaine at acm.caltech.edu - http://geuz.org