[Gmsh] Two overlapping spheres

Mon Apr 8 10:43:28 CEST 2013

Hi Christophe,

I have asked for the availability of the new CAD module because, as I posted earlier on this list, I am working on a solution for the "duplicated faces/lines" problem since quite some time - unfortunately I have very little time to actually work on it because of other projects.  I have now a script with the desired functionality implemented (see below) which still needs some debugging. It is a scilab script - I chose scilab because of easier implementation of vector/matrix-related algorithms, and because I do not know anything about the inner workings  and structure of the gmsh code. My approach is to attack the problem on the mesh level, i.e. clean up a surface mesh. The 2D mesh created by gmsh is loaded, the duplicated lines and surfaces detected and merged. Then the cleaned mesh is saved for gmsh to perform the 3D meshing. Unfortunately the script is awfully slow, probably due to scilab being fast in vector/matrix operations, but rather slow for the rest (e.g. explicit loops). My plan was so far, time permitting,  to finish debugging the script and then implement the functionality either in a stand-alone program in C or Delphi, or directly inside gmsh. Especially the re-implementation would represent quite some work, for which I have little time...
Since apparently you are planning to implement new functionality which could do all that on CAD level, I am asking myself if I just finish debugging my script, live with its slowliness for a while and wait for the CAD solution in gmsh to be available instead of doing the re-implementation in C. To fully address the problem, gmsh  would have to be able to

* detect and merge identical lines,
* detect and merge partially overlapping lines,
* detect and merge identical surfaces,
* detect and merge partially overlapping surfaces,
* do the above for surfaces and lines of arbitrary shape.

I can send you a test geometry in STEP format containing all of the cases above.

It would be of interest for me to know if the projected new CAD module in gmsh will be able to do all that, and if it will be available, say, in the course of this year.

In addition, do you think the scilab script which cleans a 2D mesh would be of interest for gmsh once I consider it sufficiently stable? If so, I will be happy to contribute it.

Best regards,

Matthias

> -----Ursprüngliche Nachricht-----
> Von: Christophe Geuzaine [mailto:cgeuzaine at ulg.ac.be]
> Gesendet: Freitag, 5. April 2013 14:07
> An: Zenker, Dr. Matthias
> Cc: gmsh at geuz.org list
> Betreff: Re: [Gmsh] Two overlapping spheres
>
>
> On 05 Apr 2013, at 12:46, "Zenker, Dr. Matthias" <Matthias.Zenker at erbe-
> med.com> wrote:
>
> > Sounds promising. Do you have an approximate idea when a beta will be
> > available for testing...? ;)
> >
>
> not yet...
>
> >> -----Ursprüngliche Nachricht-----
> >> Von: Christophe Geuzaine [mailto:cgeuzaine at ulg.ac.be]
> >> Gesendet: Freitag, 5. April 2013 12:33
> >> An: Zenker, Dr. Matthias
> >> Cc: gmsh at geuz.org list
> >> Betreff: Re: [Gmsh] Two overlapping spheres
> >>
> >>
> >> On 05 Apr 2013, at 12:21, "Zenker, Dr. Matthias"
> >> <Matthias.Zenker at erbe- med.com> wrote:
> >>
> >>> Hi Christophe,
> >>>
> >>> since you are working on "real" CAD routines: Do you also plan to
> >> implement a possibility to detect and merge duplicated surfaces
> >> (including partial overlaps) on CAD level? The "cut and merge faces"
> >> does not work in quite some cases...
> >>>
> >>
> >> Indeed. The current idea for the OpenCASCADE port is to rewrite the
> >> model factory using the Salome/FreeCAD routines.
> >>
> >>
> >>> Matthias
> >>>
> >>>
> >>>> -----Ursprüngliche Nachricht-----
> >>>> Von: Christophe Geuzaine [mailto:cgeuzaine at ulg.ac.be]
> >>>> Gesendet: Donnerstag, 4. April 2013 13:15
> >>>> An: Songting Luo
> >>>> Cc: gmsh at geuz.org
> >>>> Betreff: Re: [Gmsh] Two overlapping spheres
> >>>>
> >>>>
> >>>> On 27 Mar 2013, at 00:40, Songting Luo <luosongting at gmail.com>
> wrote:
> >>>>
> >>>>> Dear All,
> >>>>>
> >>>>> I want to generate a 3D mesh for two overlapping spheres.  How can
> >>>>> I combine these two spheres as one single object (as a union), and
> >>>>> generate
> >>>> the mesh?  Please let me know, thank you very much.
> >>>>>
> >>>>
> >>>> Gmsh's internal small CAS engine does not compute boolean
> >>>> operations; you'll need to either create the geometry directly
> >>>> (computing the intersections yourself), or use a CAD modeler (e.g.
> >>>> Freecad) to create the geometry and export e.g. in STEP.
> >>>>
> >>>> We are working toward integrating "real" CAD creation routines in
> >>>> Gmsh, but it's not ready yet for general use.
> >>>>
> >>>>>
> >>>>> best,
> >>>>>
> >>>>> Luo
> >>>>> _______________________________________________
> >>>>> 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
> >>>>
> >>>>
> >>>>
> >>>>
> >>>
> >>>
> >>>
> >>>
> >>
> __________________________________________________________
> >> ___________
> >>> ERBE Elektromedizin GmbH
> >>> Firmensitz: 72072 Tuebingen
> >>> Geschaeftsfuehrer: Christian O. Erbe, Reiner Thede
> >>> Registergericht: Stuttgart HRB 380137
> >>>
> >>
> >> --
> >> Prof. Christophe Geuzaine
> >> University of Liege, Electrical Engineering and Computer Science
> >> http://www.montefiore.ulg.ac.be/~geuzaine
> >>
> >>
> >
> >
> >
> >
> __________________________________________________________
> ___________
> > ERBE Elektromedizin GmbH
> > Firmensitz: 72072 Tuebingen
> > Geschaeftsfuehrer: Christian O. Erbe, Reiner Thede
> > Registergericht: Stuttgart HRB 380137
> >
>
> --
> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science
> http://www.montefiore.ulg.ac.be/~geuzaine
>
>

_____________________________________________________________________
ERBE Elektromedizin GmbH
Firmensitz: 72072 Tuebingen
Geschaeftsfuehrer: Christian O. Erbe, Reiner Thede
Registergericht: Stuttgart HRB 380137