Thanks Chris! If paraview doesn’t handle them correctly then I guess there’s no point in it.
If you have it lying around could you send me the xdmf file with quadratic elements? If I have time I might see if someone at kitware knows what the problem is. Dave On Feb 7, 2014, at 12:23 PM, Chris Richardson <[email protected]> wrote: > Hi Dave > > I don't think dolfin can do this at present. > > In my experience, the visualisation software (e.g. paraview) doesn't draw > quadratic elements properly anyway, but approximates as piecewise linear. > I did test this out experimentally for xdmf. > > I guess you could do a global refine of your mesh and interpolate to P1 to > get a similar result. > > Chris > > On 7 Feb 2014 19:27, David Bernstein <[email protected]> wrote: >> >> Hi Everyone, When I write a pvd file from a simulation which uses second >> order triangles the VTK element type that gets written is #5, i.e., a 3-node >> triangle. >> >> Is there a way to get fenics to realize that if the function getting written >> is in a space of second order Lagrange elements to write a pvd file using >> the VTK_QUADRATIC_TRIANGLE (#22) type? >> >> I would like to see what effect this has on how paraview displays the >> results. >> >> Regards, >> Dave >> >> >> VTK file format reference: http://www.vtk.org/VTK/img/file-formats.pdf >> _______________________________________________ >> fenics mailing list >> [email protected] >> http://fenicsproject.org/mailman/listinfo/fenics _______________________________________________ fenics mailing list [email protected] http://fenicsproject.org/mailman/listinfo/fenics
