Interference-fit analysis

theCapitain wrote:

I'm looking forward to read your considerations.

I am still trying to run your job. Since I am using CodeAster version 10.2+ I had to make changes and the job started this morning is still running. Hopefully it will converge. A few past attempts have been un-successful. I am trying and have not forgotten.

Regards,
JMB

deleted dual posting...<br /><br />Post edited by: JMB, at: 2010/08/25 20:36

Peter Halverson wrote:

1. [...] Code Aster has some alternatives but I forget what they are. You can either consult the documentation or try them all.
2. Quadratic elements are not always better than linear elements when contact is concerned different methods will treat quadratic elements differently. Sometimes linear elements will do better.

Hi Peter,
I've read the documentation and I'm quite sure of the commands … well, cause my bad french I could have missed some important content. Anyway, following your advice I tried to run the problem with linear elements ad I obtained a better result (and time saving) as you can see in the picture. So aren't quadratic elements suitable for contact problems? In my initial post some “esclave” node penetrate the “maitre” surface.

I still have a trouble in the result plot.
The image is the “deformed shape and scalar map” plot, field EQUI_NOEU_SIGM, scale factor 0,005, scalar mode [1] VMIS-. The gap generated by DIST_MAIT is distinctly visible (but it shouldn't, it is 0,02 while the external radius of the model is 30); I wander what scalar factor means!
Moreover I imposed symmetry conditions on XZ and YZ faces but the plot displays incongruous displacements. Note: is only a visualization problem because the plot of DX or DY from the DEPL field shows that symmetry faces are on symmetry plane. I have to set the scale factor to 70 in order to obtain a visualization of the distance between the contact faces comparable with that from EQUI_NOEU_SIGM.
Where am I mistaken?

Regards.

Hello,

I am trying to run your previous (quad) problem on Ver 10.2+ and getting some bizzare results. Can you post your latest (linear mesh) files (tar.gz) please, so I can try to reproduce your findings?

Regards,
JMB

Of course I can.
Using the same mesh I've modified the command file as in the attachment. The .med file I posted was exported from Salome-Meca … I don't know if this could be the cause of your bizarre results.
Thanks and regards.