[Gmsh] Bug with periodic entities in extruded meshes with layers?

[Christophe Geuzaine]

> On 1 May 2019, at 14:13, Marco Tiberga <M.Tiberga at tudelft.nl> wrote:
> 
> Dear Gmsh developers,
>  
> I have recently moved from Gmsh 3.0.6 to the most recent 4.3.0, and I have noticed that the .msh files for extruded meshes with layers are slightly different than before, because they contain periodicity relations.
> To me this is puzzling, because I do not have any periodic mesh. Why are the periodicity relations written even on non-periodic meshes? is there a way to turn this off?
>  
> Second, I have noticed that Gmsh is returning an error when reading these .msh files. I have attached two meshes (2D and 3D) that prove this: the .msh files have been generated from the attached .geo ones. For example, when I open square.msh, I get “Error: Cannot find periodic counterpart of vertex 3 on edge 2 nor on 2”.
> Since these are two extremely simple meshes, I was wondering whether there is a bug in the definition of the periodicity relations, or whether some commands I am using need to be changed with the new Gmsh version.
>  

New Gmsh versions indeed output by default periodic node information for extruded meshes (see https://gitlab.onelab.info/gmsh/gmsh/issues/439). The problem you encountered is due to the fact that the MSH2 writer saves the information even if the full mesh is not saved (i.e. only same parts are in physical groups).

The error will be fixed in the next development build. (Quick workaround until then: set "Mesh.IgnorePeriodicity = 1" when reading the mesh, e.g. "gmsh cube.msh -setnumber Mesh.IgnorePeriodicity 1".)

Christophe



> For your information, I have to use the MSH 2.2 format for compatibility with my CFD code.
>  
> Thanks.
>  
> Best regards,
> Marco Tiberga
> PhD candidate
> Delft University of Technology
> Faculty of Applied Sciences
> Radiation Science & Technology Department
> Mekelweg 15, 2629 JB Delft, The Netherlands
> E-Mail: m.tiberga at tudelft.nl
> Website: http://www.nera.rst.tudelft.nl/
>  
>  
> <square.geo><cube.geo><cube.msh><square.msh>_______________________________________________
> 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