[Gmsh] Perodic surface command does not find connections on the opposite surface

[Christophe Geuzaine]

Dear Christoph,

There were two issues:

(1) your script assumed that Sxmin() and Sxmax() contain the corresponding surfaces in the right order; this is not guaranteed, as the master and slave surfaces are generated in different orders, so that the bounding box search will also return them in different order

(2) our curve matching code was using too strict a tolerance in case the match cannot be done purely topologically (when more than one curve has the same end points as the master curve)

Here is your script with a solution for (1). The latest development snapshot of Gmsh contains the fix for (2).

Best,

Christophe


-------------- next part --------------
A non-text attachment was scrubbed...
Name: gmsh_in_test_CG.geo
Type: application/octet-stream
Size: 11697 bytes
Desc: not available
URL: <http://onelab.info/pipermail/gmsh/attachments/20200127/35c3960d/attachment.geo>
-------------- next part --------------


> On 27 Jan 2020, at 09:47, Grunwald, Christoph <Christoph.Grunwald at emi.fraunhofer.de> wrote:
> 
> Dear all,
>  
> I am trying to build a periodic mesh for a cube with some spheres in it and where some of the spheres protrude the surfaces and are cut on the cubes sides. As far as I see, there is the same number of curves and points on each of the geometric surfaces. However, when collecting opposite surfaces and applying ?Periodic Surface? command, I regularly receive the Error ?No corresponding point xx for periodic connection of surface xx to xx? or ?Different number of points (x vs x) for periodic correspondence between surfaces xx and xx?.
> The issue does not appear if the spheres cut only one side of the cube (i.e. not spheres protruding the cubes edges).
>  
> Any help would be highly appreciated!
> Thank you very much,
> --
> Christoph Grunwald
> Fraunhofer Institut f?r Kurzzeitdynamik, Ernst-Mach-Institut, EMI
> Am Klingelberg 1 | 79588 Efringen-Kirchen
> Fon +49 7628 9050 630 | Fax -677
> Christoph.grunwald at emi.fraunhofer.de
>  
> <gmsh_in_test.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