[Gmsh] Gmsh may mistakenly treats an interface surface between two regions as a boundary surface

cgeuzaine at ulg.ac.be cgeuzaine at ulg.ac.be
Mon Aug 29 23:13:01 CEST 2016



> On 29 Aug 2016, at 22:57, Junhui Liu <jhliu at lcp.nrl.navy.mil> wrote:
> 
> Hi Christophe,
> 
> Thanks a lot. But I got the same error when I used the geo file you modified. Still complains about the "Boundary face construction failed with: xxx faces remaining”. 

Not sure what this means... The mesh seems ok. Does the definition of your physical surfaces influence this? (You have 2 physicals for the same parts)


> 
> Jun
>> On Aug 29, 2016, at 4:38 PM, Christophe Geuzaine <cgeuzaine at ulg.ac.be> wrote:
>> 
>> 
>> Hi Jun,
>> 
>> It seems that your last volume definition is wrong. Attached is a file where the volume is defined interactively - does it work? 
>> 
>> Christophe
>> 
>> <pipe2.geo>
>> 
>>> On 29 Aug 2016, at 21:22, Junhui Liu <jhliu at lcp.nrl.navy.mil> wrote:
>>> 
>>> Hi All, 
>>> 
>>> I am generating structured tetra meshes (without the final recombine)  inside and outside a pipe. Please see the attache plot. Using structured tetra meshes outside the pipe is to control the time step size.  There are three regions generated. Two are structured tetra meshes and another one is the regular unstructured tetra mesh.  Somehow our internal code complains that "Boundary face construction failed with: xxx faces remaining”. if I use a version that does not care about boundary construction, It appears that the results of the cylindrical surface pointed by blue arrows behave very strangely and look like those of a boundary face. If I take region III away, the problem disappears. I have also attached the pipe.geo file. 
>>> 
>>> I have to say GMSH has lots of functionalities to control mesh and I really like it. I hope this problem can be resolved soon. Thanks a lot for your help.
>>> 
>>> Jun
>>> <pipe.geo>
>>> 
>>> <mesh_topology.pdf>
>>> 
>>> _______________________________________________
>>> 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
>> 
>> Free software: http://gmsh.info | http://getdp.info | http://onelab.info
> 


More information about the gmsh mailing list