[Gmsh] conformal mulitregion meshing from commandline +step file import

[Christophe Geuzaine]

> On 27 Oct 2015, at 16:35, Anan Im <ananim343297 at yandex.com> wrote:
> 
> 
> 
> 27.10.2015, 13:10, "Zenker, Dr. Matthias" <Matthias.Zenker at erbe-med.com>:
>> Hi,
>> 
>> the mesh in the picture in the post http://forum.freecadweb.org/viewtopic.php?f=18&t=10846&sid=2d71e90e5bc6cb707fb96796681a84ca&start=10#p103981 looks conformal, but if I open it in gmsh, it looks loke a single body. In the picture in the post http://forum.freecadweb.org/viewtopic.php?f=18&t=10846&sid=2d71e90e5bc6cb707fb96796681a84ca&start=20#p103993 is not conformal at the interface.
>> By "conformal" I mean that the triangles and nodes on both touching faces are the same. Only then you can have heat or current flow between bodies. 
>> 
>> Matthias
> Yes,  basically agree on your conception of "conformal". Thanks for that clarification. It was expected and helped. Others speak of "conforming"; "conformity" (like gmsh man) one reads. Guess I goes all down to the very same, does not it?
> But then there is extra info such as region, or boundary by. There the question is: If there is a standardized format to be be qualified as "conformal"? This mean basically ready for solving. How would such a mesh preparation state be called? Not at all?
> Well and then there was found a definition, where the claim to have a patent on 
> http://www.sonnetsoftware.com/support/help-current-version/help_topics/what_is_conformal_mesh_.htm
> conformal meshing: "Conformal meshing is a technique which can dramatically reduce the memory and time required for analysis"
> Could you please help to sort those terms in? tia.
> 
> 
>> If the OCC functions for connecting/sewing faces work only for flat faces as suggested in the thread you cite, they are next to useless since curved faces occur in 3D.
> It is rumor...
> 
> It seems that at the moment only Salomé can sew such faces - but AFAIK Salomé uses OCC also, so there must be a way...
> 
> Would be a help, if some gmsh experts could drop a commentary on:
> (As I see there are several ways of representing and  controlling multiregion (zones,bodies), thus importing and exporting signaling. 
> -Given the cited command line where along do the ready-to-solve data flow? And what is the typical problem about it all, if there is?
>> gmsh face_com_step.step -2 -format unv -o compound_of_faces.unv -optimize -string "Geometry.OCCSewFaces=1;" -string "Mesh.CharacteristicLengthFactor = 1.5;"
> -What is about "Names" of such "Groups" for the (external) solver to identify? How are they introduced?
> -And yes, what is about bespoken unflat interboundary faces, can gmsh handle it multiregionally or can it not?
> Thanks in advance

There's no need to construct elaborate command lines to drive the process: just create a script file (say, "script.geo"), that contains the command

  Merge "file.step";

followed by any command you want, exactly in the same way as for "geo" geometries. You can even merge a STEP file interactively and use the GUI to add commands (like physical groups, for the solver). Gmsh will create the file for you, automatically.

See e.g. demos/component8_in_a_box.geo for an example.

Christophe


> 
> 
> 
> _______________________________________________
> gmsh mailing list
> gmsh at geuz.org
> http://www.geuz.org/mailman/listinfo/gmsh

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

Tetrahedron V, July 4-5 2016: http://tetrahedron.montefiore.ulg.ac.be
Free software: http://gmsh.info | http://getdp.info | http://onelab.info