Hello Jose, I also use Gmsh as a mesh generator for the 3D BEM method. I use Gmsh as: (1) surface modeling tool and (2) 3D surface mesh generator.
I my software I also use discontinues elements in two ways: (1) Geometrically BEM node is the same as mesh node, but functionally they are the different nodes. I use this approach when I treat the normals direction in the corner. (2) BEM node geometrically shifted into the element. But I don't use Gmsh for such nodes generation. I working only with resulting 3D mesh and use local coordinates bases. For example, when I use the shifted nodes for each element I set the shifted BEM nodes local coordinates as (0.75, 1.0, 1.0), (1.0, 0.75, 1.0), (1.0, 1.0, 0.75). Sincerely, Alexander Kalinin On Sat, Sep 22, 2012 at 11:09 PM, José Jeferson Rêgo Silva < [email protected]> wrote: > Dear Sirs. > > We are trying to use *Gmsh* as a mesh generator for a 3D boundary element > software (surface modeling). > Do you have any experience in doing so? > > The purpose of our work is academic, for teaching and research, helping to > use our BE software. > We understand that for others purposes it will be necessary a different > type of license. > > It seems that modeling surfaces with *Gmsh* could help. > Our major difficulty is when dealing with adjacent elements that do not > share the same functional nodes. > The geometric nodes, however, are kept the same, naturally. > > For these cases, we are used to employ the so-called "discontinuous > element" (also known as "non-conforming" element). > An alternative is to use "double nodes", that is not our case. > > The idea of discontinuous element is to place the functional nodes inside > the element, instead of on its boundary (edge). > On the contrary of FEM, this do not affect BEM formulations. > > So, we would like to know if *Gmsh *has this capability of creating new > nodes inside the elements, when necessary, accordingly to different types > of connectivity between adjacent elements. Sometimes an surface element can > share functional nodes with one (or more) of its adjacent elements, and do > not share functional nodes with the others. > > It is not necessary *Gmsh* generate the coordinates for these new nodes. > Our software can handle it. > It is necessary, however, to number these new nodes and associate them to > each element (keeping the same element orientation - ordering of their > nodes), as it is already done for the others standard (geometric) nodes. > > Thanks in advance. > Greetings. > > J. Jeferson Rêgo Silva > Civil Engineering Department - UFPE > > _______________________________________________ > gmsh mailing list > [email protected] > http://www.geuz.org/mailman/listinfo/gmsh > >
_______________________________________________ gmsh mailing list [email protected] http://www.geuz.org/mailman/listinfo/gmsh
