×

Notice

The forum is in read only mode.

Matrice non factorisable

  • Fergus Rhodes
  • Topic Author
  • Offline
  • New Member
  • New Member
More
17 years 9 months ago #1375 by Fergus Rhodes
Matrice non factorisable was created by Fergus Rhodes
Hi, I'm having some trouble getting Aster to solve a mesh generated in Salome from a Step model of a piping component. The model is reasonably complex with curved surfaces and tricky radii.

I am getting the error \"Matrice non factorisable\". The model is fully constrained, so I do not think that lack of boundary conditions is the issue.

The initial alarms I am getting are:
1) presence of flattened meshs!
2) la maille MW porte un element fini de bord. mais elle ne borde aucun element ayant une \"rigidite\".
i.e. the mesh MW carries a finite element of edge. but it does not border any element having a “rigidity”.

Can anyone out there explain these alarms to me? Is some problem with the mesh making it unsolvable? Any ideas on how to avoid this problem?

Any help would be much appreciated.

Thanks
More
17 years 9 months ago #1378 by johannes ackva
Replied by johannes ackva on topic Re:Matrice non factorisable
Hi, The reason for \"Matrice non factorisable\" is Your alarm 2). I suppose how it happened. Look to this diskussion:

www.caelinux.com/CMS/index.php?option=co...;id=397&catid=14

Does it help to solve Your problem? Good luck

Johannes
  • Fergus Rhodes
  • Topic Author
  • Offline
  • New Member
  • New Member
More
17 years 9 months ago #1382 by Fergus Rhodes
Replied by Fergus Rhodes on topic Re:Matrice non factorisable
Thank you for help. This has given me an insight into why this error message is generated.

I can now get my model to mesh and solve OK without the troublesome radius in place. However with this radius in the step model, I still get the \"Matrice non factorisable\" fatal error.

This radius is a variable fillet radius from R20 to R0 on an internal curved surface in my piping component. There is a point on the surface where four curved surfaces meet at a single node, where the radius decays to a point.

The start of the problem (in Aster) seems to be at the verification of the mesh stage in the Aster run.

====== VERIFICATION DU MAILLAGE ======
LA MAILLE POSSEDE DES NOEUDS CONFONDUS GEOMETRIQUEMENT
MAILLE:MIK DM/DP= 0. TYPE:SEG2

i.e. \"The mesh has geometrically confused notes\". Any idea how to locate these nodes in Salome? My guess is that there must be nodes occupying the same co-ordinates in space, or joined in the same position to the same elements. Is there a better approach I could be using to meshing in Salome? I'm tried using sub-meshes to refine the mesh around the radius, however this doesn't seem to help.

Thanks
Moderators: catux
Time to create page: 0.153 seconds
Powered by Kunena Forum