[Gmsh] Extruding a NAS closed surface consisting of 2-3 open shells

[Christophe Geuzaine]

Christophe Geuzaine wrote:
> CEM ALBUKREK wrote:
>> Dear Christophe,
>>
>> I have a quick question on extruding closed NAS surfaces, with multiple 
>> part ids (PIDS). What I mean by this kind of mesh is that a given closed 
>> NAS mesh consists of unclosed shell partitions. I have been able to 
>> build nice volume meshes around such surface mesh data successfully 
>> using GMSH. So somehow GMSH must be interpreting the vertices at the 
>> seams of the PIDS correctly when volume meshing. In fact, there is no 
>> duplicate vertex at the seam in the NAS file definition.
>>
>> Unfortunately, I do not see consistent behaviour with surface extrusion 
>> from the same mesh, receiving "could not find extruded vertex" error at 
>> the seams of PIDS. I think the problem is that GMSH interprets each PID 
>> (shell) independently of its neighbors when applying extrusion. Hence 
>> the vertex normals can not be computed correctly at the seams. Am I 
>> correct with this? Do you see a way around, or could you point me to the 
>> section of the code which might require some improvement for this? I 
>> think vertex normal calculation at shared edges needs the tweak...
>>
>> Attached is a sample file pair (.geo and .nas) for you to observe the 
>> problem. At the header of the .geo file there is a switch to turn off 
>> the extrusion (boundary layer mesh) to verify the volume meshing works 
>> well. Please let me know how I can tackle the issue. Note that I have 
>> tested the same case for which the NAS file has one single PID - one 
>> closed shell - and extrusion works...
>>
> 
> We'll have a look.
> 

There is unfortunately no easy fix. The problem comes from the fact that
we don't have the topology information (telling us how the two surfaces
are connected). The vertices on the common edge are thus categorized on
one of the two surfaces, and cannot be easily recovered when extruding
the other...


> 
> 
>> Thank you.
>>
>> Cem
>>
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> 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