[Gmsh] Is about a bug in Von-Mises vs time representation

Jean-Luc FLEJOU jean-luc.flejou at edf.fr
Mon Feb 23 09:59:30 CET 2004



Hello,

I have just met an error (GMSH 1.50) in the representation of the tensors,
when there is a time list.
Here a file to highlight the bug

(See attached file: tenseur_cub8.pos)
It seems to me that the calculation of Von-Misès do not take into account
the time list.

In file "Graphics/PostElement.cpp" procedure "void Draw_TensorElement"

There is
  // View->TensorType == DRAW_POST_VONMISES
  double V_VonMises[8];
  for(int i = 0; i < nbnod; i++){
    V_VonMises[i] = ComputeVonMises(V + 9*i);
  }

There should be something of that
  // View->TensorType == DRAW_POST_VONMISES
  int ts = View->TimeStep;
  View->TimeStep = 0;

  double V_VonMises[8];
  for(int i = 0; i < nbnod; i++){
     V_VonMises[i] = ComputeVonMises(V + 9*(i + nbnod * ts));
  }

and right before leaving this procedure

  View->TimeStep = ts;

Regards,
Jean-Luc Fléjou

-------------- next part --------------
A non-text attachment was scrubbed...
Name: tenseur_cub8.pos
Type: application/octet-stream
Size: 3035 bytes
Desc: not available
URL: <http://www.geuz.org/pipermail/gmsh/attachments/20040223/d0ad43ad/attachment.pos>