[Gmsh] Possible Error in Tetrahedral Meshing

[Christophe Geuzaine]

> On 17 Apr 2019, at 11:27, Jackman, K, Dr [kjackman at sun.ac.za] <kjackman at sun.ac.za> wrote:
> 
> Dear Christophe,
> 
> The advice you gave me in the previous email solved my problem. Thank you.
> 
> However, I have noticed a new problem when creating an "internal" surface within a volume. I am using "Line{…} In Surface{…}" to link the lines of an "internal" surface to the boundary surface of a volume. This seems to work fine for most cases, but under certain circumstances the "internal" surface is not properly connected. More specifically, some of the nodes are connected to the triangles of the "internal" surface, but not the tetrahedral mesh.

You also need to embed these surfaces in the volumes, using "Surface{...} In Volume{...};".

Christophe


> 
> I have attached the geometry and mesh files, including an image, to demonstrate the problem. Nodes 605, 606, 1067 and 1068 are connected to the "internal" surfaces, but not the tetrahedral mesh.
> 
> Hopefully, this is just a simple mistake in my geometry file.
> 
> Best regards,
> Kyle
> From: Christophe Geuzaine <cgeuzaine at uliege.be>
> Sent: Monday, 08 April 2019 6:25 PM
> To: Jackman, K, Dr [kjackman at sun.ac.za]
> Cc: gmsh at onelab.info
> Subject: Re: [Gmsh] Possible Error in Tetrahedral Meshing
>  
> 
> 
> > On 8 Apr 2019, at 17:59, Jackman, K, Dr [kjackman at sun.ac.za] <kjackman at sun.ac.za> wrote:
> > 
> > Dear Christophe,
> > 
> > Thank you for your prompt reply.
> > 
> > I can successfully mesh the geometry file with Gmsh, without any errors, but the tetrahedral mesh does not seem to be correct at some locations. The faces of some tetrahedrons (please see attached images) are not aligned correctly, and therefore, not properly connected. I have attached the mesh file and some images to demonstrate the problem.
> 
> The mesh you sent is correct; but it has some fairly flat tetrahedra ("slivers") that the optimizer could not remove. This is to be expected for a mesh with two very close surfaces - with different meshes on both sides. For such geometries you would get better results by extrusion, making sure that the meshes on the two surfaces "match".
> 
> To visualize the poor-quality elements, you can either use Tools->Options->Mesh->Visibility and enter e.g. [0,1e-4] for the Gamma range ; or use Tools->Statistics and click "3D" for the quality you are interested in.
> 
> Christophe
> 
> > 
> > Best regards,
> > Kyle
> > 
> > 
> > From: Christophe Geuzaine <cgeuzaine at uliege.be>
> > Sent: Monday, 08 April 2019 3:30 PM
> > To: Jackman, K, Dr [kjackman at sun.ac.za]
> > Cc: gmsh at onelab.info
> > Subject: Re: [Gmsh] Possible Error in Tetrahedral Meshing
> >  
> > 
> > 
> > > On 8 Apr 2019, at 12:47, Jackman, K, Dr [kjackman at sun.ac.za] <kjackman at sun.ac.za> wrote:
> > > 
> > > To Whom It May Concern,
> > > 
> > > I have noticed some errors in the tetrahedral meshes, generated by Gmsh (both v3.0.6 and v4.2.2). Under certain circumstances, e.g. volumes with holes, some tetrahedron elements do not align with adjacent tetrahedrons. An example is attached that demonstrates this error, including images that show the location of the edges that do not connect.
> > > 
> > > I have tried several geometric examples, including meshing algorithms (surface and volume), but this does not fix the problem. The attached example seems to work when selecting “Mesh.Algorithm = 2;”, but this is not always the case for other geometries.
> > > 
> > > Hopefully, this is just a simple mistake in my geometry file.
> > 
> > The file you sent meshes fine here will all the algorithms. Can you send the .msh file that you think is wrong?
> > 
> > Christophe
> > 
> > > 
> > > Best regards,
> > > Kyle
> > > 
> > > 
> > > The integrity and confidentiality of this email are governed by these terms. Disclaimer
> > > Die integriteit en vertroulikheid van hierdie e-pos word deur die volgende bepalings bereël. Vrywaringsklousule <Gmsh_error_overview.zip>_______________________________________________
> > > gmsh mailing list
> > > gmsh at onelab.info
> > > http://onelab.info/mailman/listinfo/gmsh
> > 
> > — 
> > Prof. Christophe Geuzaine
> > University of Liege, Electrical Engineering and Computer Science 
> > http://www.montefiore.ulg.ac.be/~geuzaine
> > 
> > 
> > 
> > <mesh_file_and_images.zip>
> 
> — 
> Prof. Christophe Geuzaine
> University of Liege, Electrical Engineering and Computer Science 
> http://www.montefiore.ulg.ac.be/~geuzaine
> 
> 
> 
> <files.zip>

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