[Gmsh] Fragment + addSurfaceLoop

[Christophe Geuzaine]

> On 20 Oct 2020, at 15:40, Bruno <bruno.agostini at gmail.com> wrote:
> 
> Dear Gmsh Community:
> 
> In the code below I import a 3D cooling fin CAD in Brep format. I want to modify the model to insert a heating area smaller than the base area and have a conformal meshing. I then tag the heating area for further use to apply boundary conditions. To this end I used the python code below:
> 
> ```python
> 
> # New heating area
> 
> qa.append((2, factory.addRectangle(0, 0, 0, dz, -dy)))
> factory.rotate([qa], 0, 0, 0, 0, 1, 0, pi/2)
> tx = -Hd/2 - HEV
> ty = ( dy - ( Wd + CH0 ) / 2 ) / 2
> tz = Ld/2 + zq
> factory.translate([qa], tx, ty, tz)
> factory.synchronize()
> 
> s = model.getEntitiesInBoundingBox(-Hd/2-HEV-tol, -Wd/2-CH0/2-tol, -Ld/2-tol, -Hd/2-HEV+tol, Wd/2+CH0/2+tol, Ld/2+tol, dim = 2) # Get fin bottom surface
> sc = list(set(s) - set(qa)) # Remove heating area (duplicates)
> sf = factory.fragment(sc, qa, removeObject = True, removeTool = False)[0]
> factory.synchronize()
> s = model.getEntities(2)
> s = list(set(s) - set(sc)) # Remove original base area (duplicates)
> 
> sl = factory.addSurfaceLoop([s[i][1] for i in range(len(s))], sewing = True)
> factory.synchronize()
> 
> vol = factory.addVolume([sl])
> factory.synchronize()
> 
> p = model.addPhysicalGroup(2, [qa[i][1] for i in range(len(qa))])
> model.setPhysicalName(2, p, "base")
> 
> p = model.addPhysicalGroup(3, [vol])
> model.setPhysicalName(3, p, "fin")
> 
> model.mesh.generate(3)
> 
> ```
> 
> If I use `sewing=True` then I get the results in the attached picture (superimposed_mesh.png): the heating area and the volume seem to have superimposed meshes even though I used fragment to cut-out the heating area.
> 

I would fragment the whole model, which will ensure that everything is conforming; then extract the parts needed for boundary conditions.

Christophe



> If I use `sewing=False` then I get facet intersection errors (fin_NOK.png).
> 
> Any idea what I am doing wrong here?
> 
> 
> Best regards
> 
> 
> Bruno
> 
> 
> 
> 
> <superimposed_mesh.png><fin_NOK.png>_______________________________________________
> 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://people.montefiore.ulg.ac.be/geuzaine