[Gmsh] half of structure disappears

Juan E. Sanchez juan.e.sanchez at gmail.com
Mon Mar 5 16:02:53 CET 2018


Thank you,

That fixed the problem!

Regards,

Juan

On 3/5/18 1:29 AM, Christophe Geuzaine wrote:
> 
> 
>> On 4 Mar 2018, at 18:46, Juan E. Sanchez <juan.e.sanchez at gmail.com 
>> <mailto:juan.e.sanchez at gmail.com>> wrote:
>>
>> Hi Christophe,
>>
>> In my actual structure, it is unknown whether or not it will be split. 
>> There will also be many more volumes with different tags.
>>
>> I also do not want "air" to contain any of "gate".
>>
>> How do I get all volumes created from the BooleanDifference into the 
>> same Physical Volume?
>>
> 
> Use the list (note the "()") returned by the boolean command :
> 
> air()=BooleanDifference{Volume{air}; Delete;}{ Volume{gate};};
> 
> Recursive Delete{Volume{unit_box};}
> BooleanFragments{ Volume{air()}; Volume{gate}; Delete;}{}
> 
> Physical Volume ("gate") = {gate};
> Physical Volume ("air") = {air()};
> 
> 
> 
>> Regards,
>>
>> Juan
>>
>>
>>
>> On 3/4/18 11:29 AM, Christophe Geuzaine wrote:
>>>> On 4 Mar 2018, at 17:46, Juan E. Sanchez <juan.e.sanchez at gmail.com 
>>>> <mailto:juan.e.sanchez at gmail.com> <mailto:juan.e.sanchez at gmail.com>> 
>>>> wrote:
>>>>
>>>> Using the latest Mac OS snapshot, it is the same behavior.
>>>>
>>>> 3 sections appear (air, gate, air)
>>>>
>>>> I go into the visibility dialog, visualize both "gate" and "air", 
>>>> and the last "air" disappears.
>>>>
>>> I think you are misunderstanding the definition of physical groups. 
>>> Your boolean operations create 3 elementary volumes : left, center 
>>> and right. You put the left one in the physical group "air" and the 
>>> middle one in "gate". (Note that your boolean operations are 
>>> unnecessarily complicated - you could just fragment the 2 boxes 
>>> instead of doing a difference followed by a fragment.)
>>> So if you select the 2 physical groups, indeed you hide the right 
>>> volume...
>>>> If I mesh before opening the visibility dialog, all 3 sections get 
>>>> meshed.
>>>>
>>> Normal again: by default with the MSH2 format Gmsh only saves 
>>> elements that belong to physical groupe.
>>>> When I save the mesh, and then open the mesh file, one of the "air" 
>>>> sections is missing.
>>>>
>>>>
>>>> Starting over:
>>>>
>>>> If I open the visibility dialog, and select both physical groups, 
>>>> one of the "air" sections disappear.
>>>>
>>>> Looking in the Elementary entities, it appears that there are 3 
>>>> volumes.  It looks like one of the "air" sections did not get 
>>>> assigned to the Physical Group.
>>>>
>>>> How do I ensure that both "air" sections get assigned to the same 
>>>> Physical Group?
>>>>
>>> Physical Group ("air") = {3,4};
>>> Christophe
>>>> Regards,
>>>>
>>>> Juan
>>>>
>>>>
>>>>
>>>>
>>>> On 3/4/18 1:21 AM, Christophe Geuzaine wrote:
>>>>> Dear Juan,
>>>>> Not sure what you observe - your file seems to behave normally with 
>>>>> the latest version. Can you try with the latest automated snapshot ?
>>>>> Christophe
>>>>>> On 3 Mar 2018, at 20:40, Juan E. Sanchez <juan.e.sanchez at gmail.com 
>>>>>> <mailto:juan.e.sanchez at gmail.com> 
>>>>>> <mailto:juan.e.sanchez at gmail.com> 
>>>>>> <mailto:juan.e.sanchez at gmail.com>> wrote:
>>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> I am using the OpenCASCADE factory to create a fairly complicated 
>>>>>> structure.  In this simplified test case, I run into the issue 
>>>>>> where part of the structure disappears when the center of it is 
>>>>>> removed.
>>>>>>
>>>>>> When GMSH opens the file, it looks correct.  Then when I start 
>>>>>> manipulating it in the Visibility dialog, half of the "air" volume 
>>>>>> disappears.
>>>>>>
>>>>>> When 3d meshing half of the "air" volume is also missing.
>>>>>>
>>>>>> When I reopen the saved mesh, half the "air" volume is also missing.
>>>>>>
>>>>>> Could someone please take a look?
>>>>>>
>>>>>>
>>>>>> Regards,
>>>>>>
>>>>>> Juan Sanchez
>>>>>> <test.geo>_______________________________________________
>>>>>> gmsh mailing list
>>>>>> gmsh at onelab.info <mailto:gmsh at onelab.info> 
>>>>>> <mailto:gmsh at onelab.info> <mailto: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
>>>>
>>>>>> 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
>>
> 
>> 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