[Gmsh] Gmsh-4.0.4 Volume mesh issue

Christophe Geuzaine cgeuzaine at uliege.be
Tue Nov 13 11:31:32 CET 2018



> On 13 Nov 2018, at 10:43, Simone GREMMO <Simone.GREMMO at umons.ac.be> wrote:
> 
> Dear Christophe,
> thank you very much for your reply. I downloaded the latest snapshot and now I can get the mesh as expected. I will continue my testing.
> 
> One more question: I have seen that with Gmsh 4 you have introduced many modifications to the mesh generation algorithms. Could you tell me the section where the pyramids are added and "inflated" inside the domain?

The (flat) pyramids are created in splitQuadRecovery::buildPyramids() in Mesh/meshGRegion.cpp ; they are "inflated" by RelocateVertices() (in the same file). An easy optimisation would be to limit the relocation to only a few layers of elements touching the pyramids.

Christophe


> 
> In a previous version (I think 2.x) it was in "MeshDelaunayVolumeTetgen" in meshGRegion.cpp, but now, I cannot find the relevant section in the new implementation.
> In the old code I introduced a modification (after a discussion with prof. Remacle) to this part and would like to try to replicate it also in the latest version of Gmsh.
> 
> Thanks in advance,
> Simone G.
> From: Christophe Geuzaine <cgeuzaine at uliege.be>
> Sent: Monday, November 12, 2018 3:33:25 PM
> To: Simone GREMMO
> Cc: gmsh at onelab.info
> Subject: Re: [Gmsh] Gmsh-4.0.4 Volume mesh issue
>  
> 
> Dear Simone,
> 
> Indeed, Gmsh 4.0.0 -> 4.0.4 do not have the ability to automatically create hybrid tetrahedral+pyramidal meshes (your surface contains quadrangles).
> 
> The latest snapshot (future Gmsh 4.0.5 or Gmsh 4.1) has a new algorithm to generate such meshes, similar but more general to what we did in Gmsh 3. Give it a try and let us know how it works out for you?
> 
> Christophe
> 
> 
> > On 12 Nov 2018, at 14:38, Simone GREMMO <Simone.GREMMO at umons.ac.be> wrote:
> > 
> > Dear all, 
> > I have just downloaded the last version of Gmsh (4.0.4), but I am facing a problem with the generation of a "simple" volume mesh.
> > 
> > I would like to mesh a volume delimited by a surface mesh in .msh format, but it looks like Gmsh cannot create the volume mesh.
> > Error Message is :
> > .....
> > Info    : Boundary Recovery...
> > Info    : Writing to GRegion...
> > Info    : Reconstruct time : 0.116395 sec
> > Error   : No tetrahedra in region 5
> > Info    : 3D point insertion terminated (3195 points created):
> > Info    :  - 0 Delaunay cavities modified for star shapeness
> > Info    :  - 0 points could not be inserted
> > Info    :  - 0 tetrahedra created in 1.6e-05 sec. (0 tets/sec.)
> > Error   : No tetrahedra in region 5 
> > Info    : Done meshing 3D (0.133228 s)
> > Info    : 4857 vertices 6617 elements
> > Error   : ------------------------------
> > Error   : Mesh generation error summary
> > Error   :     0 warnings
> > Error   :     2 errors
> > Error   : Check the full log for details
> > Error   : ------------------------------
> > 
> > I am attaching the .geo file I have created as well as the surface mesh I used. To launch the mesh generation I run the command:
> > gmsh-4.0.4 createGeom.geo -3 -o volumeMesh.msh
> > 
> > Note that the same test case is successfully meshed using Gmsh version 3.0.6.
> > 
> > Do you have any suggestion on how to solve the problem? 
> > Is it possible that the surface mesh, containing both triangles and quadrangles is at the origin of the problem?
> > 
> > Thanks in advance for your suggestions.
> > Regards,
> > Simone G.
> > 
> > <createGeom.geo><surfaceMesh.msh>_______________________________________________
> > gmsh mailing list
> > gmsh at onelab.info
> > http://onelab.info/mailman/listinfo/gmsh
> gmsh Info Page
> onelab.info
> Gmsh is a free 3D finite element mesh generator with built-in pre- and post-processing facilities. This list is for questions and discussions about Gmsh features, resources, bugs, and so forth. Useful links: Gmsh homepage (for documentation, downloads, etc.); gmsh-announce mailing list (for announcements about new Gmsh releases and major bug fixes - you probably want to subscribe to that, as well)
> 
> 
>> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science 
> http://www.montefiore.ulg.ac.be/~geuzaine
> 
> Free software: http://gmsh.info | http://getdp.info | http://onelab.info
> 

— 
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science 
http://www.montefiore.ulg.ac.be/~geuzaine

Free software: http://gmsh.info | http://getdp.info | http://onelab.info




More information about the gmsh mailing list