[Gmsh] Re : Problems for creating ruled surfaces belonging to a cylinder

[Durufle marc]

On these pictures, you see the surface of the object after meshing, and it is clearly not a cylinder ...
 Otherwise I checked coordinates of points of the lines obtained by extruding a point (with combined rotation 
and translation), they are not on the cylinder too, so probably a first step would be to specify directly spline with more than
 5 points in order to have curves on the cylinder, and probably split the surfaces into small patches of ruled surfaces that would
 generate a better approximation of the cylinder. Well, I will try Salome, I hope that Gmsh correctly exploits the CAD informations in order to generate the curved mesh.
Thanks for your comments.




________________________________
De : David Colignon <David.Colignon at ulg.ac.be>
À : Durufle marc <marc_durufle at yahoo.fr>
Cc : gmsh at geuz.org
Envoyé le : Dim 27 Décembre 2009, 10 h 26 min 58 s
Objet : Re: [Gmsh] Problems for creating ruled surfaces belonging to a cylinder


On 26/12/09 20:58, Durufle marc wrote:
> Hello,
> your help has been very useful for the problem of the cavity, thanks
> very much. I have another problem with
> the geo file attached. I want that "ruled surfaces" 36, 38, 40, 46, 48
> and 50 are part of a cylinder, so I tried to
> construct nurbs which are belonging the cylinder, but the generated
> ruled surfaces are not part of the cylinder :(

Hi Marc,

the mesh seems fine here. How do you have verified that these ruled surfaces are not part of the cylinder ?


> Do you know a way to generate those surfaces correctly ?

We have an option to force the mesh on the surface of a sphere, but not on a cylinder...

http://geuz.org/gmsh/doc/texinfo/gmsh.html#Surfaces

Ruled Surface ( expression ) = { expression-list } < In Sphere { expression } >;
    Creates a ruled surface, i.e., a surface that can be interpolated using transfinite interpolation. The expression inside the parentheses is the ruled surface's identification number; the first expression-list on the right hand side should the identification number of a line loop composed of either three or four elementary lines. The optional In Sphere argument forces the surface to be a spherical patch (the extra parameter gives the identification number of the center of the sphere).


Is it possible
> (or soon possible) to specify directly
> a nurb surface in gmsh (not only a nurb curve), so that I could
> approximate the cylinder with nurb patches ?


I don't think we will add nurbs surfaces support natively in Gmsh ( Christophe, Jean-François, a comment ? )

You could/should generate these critical nurbs surfaces in a real CAD software/Solid Modeler
( Salome http://www.salome-platform.org/ is free and based on the same CAD engine as Gmsh, OpenCascade),
export it in the STEP  format, import in Gmsh, finalize the geometry and mesh ...

Regards,

Dave

--
David Colignon, Ph.D.
Collaborateur Logistique du F.R.S.-FNRS
CÉCI - Consortium des Équipements de Calcul Intensif
ACE - Applied & Computational Electromagnetics
Institut Montefiore B28
Université de Liège
4000 Liège - BELGIQUE
Tél: +32 (0)4 366 37 32
Fax: +32 (0)4 366 29 10
WWW:    http://hpc.montefiore.ulg.ac.be/
Agenda: http://www.google.com/calendar/embed?src=david.colignon%40gmail.com


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20091227/fea5017a/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: genec1.jpg
Type: image/jpeg
Size: 16224 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20091227/fea5017a/attachment.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: genec2.jpg
Type: image/jpeg
Size: 91038 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20091227/fea5017a/attachment-0001.jpg>