[Gmsh] high order finite elements

Christophe Geuzaine cgeuzaine at ulg.ac.be
Mon Mar 19 23:43:26 CET 2007


Hi Axel - All the high-order stuff from the paper is implemented, but we
still need to "wrap it" nicely in the interface, and extend the file
formats to take advantage of the new features. This should be ready for
the next major release (probably in a few months).



Axel Gerstenberger wrote:
> Dear developers and users,
> 
> some time ago I asked for a way to visualize bi-linear elements with 
> real bilinear interpolation. The provided solution sent to me can be 
> seen in the Gmsh archive under:
> http://www.geuz.org/pipermail/gmsh/2006/001816.html
> 
> cite:
> #############
> There is a new functionality in gmsh for displaying high order FEM fields.
> 
> add
> 
> INTERPOLATION_SCHEME
> {
>    {.25,-.25,.25,-.25},
>    {.25,.25,-.25,-.25},
>    {.25,.25,.25,.25},
>    {.25,-.25,-.25,.25}}
> { {0,0,0},
>    {1,0,0},
>    {0,1,0},
>    {1,1,0}
> };
> 
> after the last scalar quad of your map. Then, load your map and go into
> the options -> Visibility
> 
> Maximum recursion level = 3 and target error = 0.
> 
> ##################################
> 
> It works beautifully, however, I am not able to figure out how to apply 
> this to second order elements (say quad9). I guess it must be trivial, 
> but somehow I don't see it. Could you give me a hint, please? I think I 
> tried hard to find an advice within the documentation, so I hope I did 
> not miss it there...
> 
> An additional feature would be to show second order triangles and quads 
> in one plot, but so far I can only provide one map at the end of the 
> post-processing file. Is there a way to show second order triangles and 
> quads in one plot?
> 
> Thanks in advance for your help and your software.
> 
> Best regards,
> 
> Axel Gerstenberger
> 
> 
> 
> BTW, I read one of the developers recent publications
> "Efficient visualization of high-order finite elements"
> J.-F. Remacle, N. Chevaugeon, É. Marchandise, Ch. Geuzaine
> http://www3.interscience.wiley.com/cgi-bin/abstract/112659093/ABSTRACT
> 
> Will these high order visualization techniques be implemented into 
> Gmsh2? That would be really a great feature.
> 
> 
> _______________________________________________
> gmsh mailing list
> gmsh at geuz.org
> http://www.geuz.org/mailman/listinfo/gmsh


-- 
Prof. Christophe Geuzaine
University of Liege, Electrical Engineering and Computer Science
http://www.montefiore.ulg.ac.be/~geuzaine