While I did not find a way to solve my problem using code aster I solved it within ParaVis.

I explained my solution in the following post:

code-aster.org/forum2/viewtopic.php?id=23835

I wasn't sure whether to put this in the Code Aster or Salome Meca part but as I am hoping to solve this using code aster I chose to put it here. Sorry if this was wrong.

I am looking for an easy way to extract data (stresses SIXX, SIYY and SIXY) along the path of an analytically defined curve.

I have a curve defined by

x=(5+3*cos(theta))*cos(theta)

y=(5+3*cos(theta))*sin(theta)

z=0

for theta from 0 to pi.

This curve is placed randomly over the mesh so does not necessarily coincide with any nodes.

I know I can extract data from a group of nodes using POST_RELEVE_T but as nodes are not placed on the curve this does not work.

I was also able to force nodes to be on this curve by adding an analytical curve into my structure in the geometry module of salome meca. This however leads to meshing being very difficult and forcing me to have a weird mesh. This would not be a dealbreaker except for that I want to alter the dimensions of my structure automatically and having this curve in my mesh creates al lot of errors when generating the mesh.

In the ParaVis module I found "Probe location" which can get the data from any location (even in between nodes) by interpolation. However, as doing this by hand for very many points would take forever this is not practical.

TLDR: So in short. Is there a way to extract stresses from either an analytically defined curve (or a list of coordinates lying on this curve)?

]]>