[Gmsh] Help with gmsh

[Christophe Geuzaine]

Hi Ana,

I've looked at your file:

1) There's indeed an incompatibility between occt 7.3.0 (used for our linux builds) and the pre-release version of occt 7.3.1 that we use on Windows and macOS, which leads to different numberings of entities. We will update the build environments so that all future builds are made with the same version.

2) You are using the "Attractor" field with FacesList. As the documentation says: "Warning, this feature is still experimental. It might (read: will probably) give wrong results for complex surfaces". We haven't improved this in a long time, as a completely different strategy for boundary layers will eventually be implemented. In the meantime you should use another method to impose the size field (from points, in a box, using a simpler distance function to e.g. some points/curves in conjunction with a mathematical formula, ...)

Christophe


> On 26 Jul 2019, at 18:43, Ana Gomes <ana.gomes at strath.ac.uk> wrote:
> 
> Sorry to bother you once again, but this problem has been persisting for a long time and if there is anything you can do to help, if you can have access to a Linux machine to reproduce my script, I would greatly appreciate it.
> 
> I have been trying to simplify the meshing definitions to debug this. At this point, I use an attractor field to, in a very simple way, have more refinement near the vehicle, and coarsening as we go away from the vehicle.
> 
> The 2D results work fine (see attractor-y.png), but it seems that the output for the volume mesh is not respecting the parameters defined for the attractor (see attractor-x.png).
> 
> What am I missing here?
> 
> Best regards,
> Ana
> <attractor-x.png><attractor-y.png><attractor.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