Hi Jerome, Yes indeed - all our meshing algorithms currently generate 1st order geometrical elements, starting from 1st order geometrical elements on the boundary. The meshes are then "curved" later on. I will add the ability to re-use a given high-order discrete surface input when we curve to our TODO list. But it's not trivial.
Christophe > On 1 Jun 2018, at 14:56, Jerome Robert <[email protected]> wrote: > > Hi all, > > I am trying to fill an existing order 2 3D triangular mesh with order 2 > tetrahedron. In the result mesh, the order 2 vertices of my input mesh are > all moved so they were linearly interpolated. So everything is as if my input > mesh was used by gmsh as an order 1 surface. > > Here is my .geo: > > // Various options I tryed with no luck > //Mesh.SecondOrderExperimental=1; > //Mesh.ElementOrder = 2; > //Mesh.SecondOrderLinear = 0; > //Mesh.Algorithm3D = 1; > > // surf.unv contains only order 2 triangles > Merge "surf.unv"; > Surface Loop(2) = {1}; > Volume(3) = {2}; > Characteristic Length {3} = 1000; > > Which I run with gmsh 3.0.6: > > $ gmsh -3 my.geo -o my.msh > > My understanding is that my input mesh is converted to order 1 by this > SetOrder1 call: > > https://gitlab.onelab.info/gmsh/gmsh/blob/a4306e50a3916dd52f72edce56c05f36954f3f0a/Mesh/Generator.cpp#L979 > > Interpolation on GEntity::DiscreteSurface or GEntity::DiscreteCurve is also > always linear. > > So there seems to be no way to use an order 2 mesh as a gmsh GModel. > > Is that correct ? If not what would be the way to do it ? > > Regards, > > Jerome > > _______________________________________________ > gmsh mailing list > [email protected] > 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
_______________________________________________ gmsh mailing list [email protected] http://onelab.info/mailman/listinfo/gmsh
