[Gmsh] Bug with transfinite interface

[Christophe Geuzaine]

> On 25 May 2018, at 16:26, Quentin Goestchel <quentin.goestchel at ens-cachan.fr> wrote:
> 
> Dear Sir or Madam,
> 
> I would like to thank you for the great work which has been done to develop the open-source Gmsh software. 
> I am currently working on a 3D numerical method and I think I have found a bug in the 3D meshing algorithm. I was trying to generate a transfinite bulk in a mesh (to have a constant orthogonality of the elements), and an unstructured layer around it, as shown in the enclosed "sphereTransfinite.png". 
> 

This is "normal" the (non-recombined) 3D transfinite algorithm does not try to enforce the compatibility between the volume and surface meshes. It's actually impossible without modifying the 2D mesh a posteriori in some configuration (since we also generate the 2D mesh first). You should specify the corners of the transfinite interpolation in the right order to have the match.

PS: Using Extrusion instead of Transfinite, Gmsh will try to perform the modifications automatically.

> Everything was going fine until I noticed a big error on my discretized spherical surface, computed with the sum of the areas of the boundary faces. I found out that my code was unable to find the neighbours of many tetrahedrons around the bulk mesh because many faces were not matching but crossing each other edges, as shown in sphereTransfinite_err.png . 
> I didn't find any topics about such problem so I assume that you didn't know about it, but I might be wrong. 
> Best regards, 
> -- 
> Quentin Goestchel  | Etudiant - Année ARPE 
> Département Génie Civil 
> +33(0)6 83 75 57 53 <tel:+33683755753>  |  quentin.goestchel at ens-paris-saclay.fr <mailto:prenom.nom at ens-paris-saclay.fr> 
> ENS Cachan - 61 avenue du Président Wilson 94235 Cachan CEDEX <geo://?q=61+Avenue+du+Pr%C3%A9sident+Wilson+Cachan+France>
> www.ens-paris-saclay.fr <http://www.ens-paris-saclay.fr/>
> <imblgnlknopffhfn.png>                        <http://www.ens-paris-saclay.fr/><sphereTransfinite.png><sphereTransfinite_err.png><sphereTransfinite.geo>_______________________________________________
> 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

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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://onelab.info/pipermail/gmsh/attachments/20180530/f6600079/attachment-0001.html>