3 Replies Latest reply on Oct 2, 2006 7:45 AM by

    Wrong behaviour with incompatible mesh

      Hi Francesco, the stress errors that one might see with an incompatible mesh will be localized to the bonded area, not the model in general as the internal "glue" (constraint equations) tries to tie the closest nodes on each part together. Depending on the coarseness of the mesh on the bonded faces, those nodes could be far from the interface so you would expect to see some ugly, spotty stresses where these equations are attached. Could you post a couple of JPGs in the interim illustrating your problem? Without looking at the database, an option to consider when the incompatible bonded option is selected is to apply mesh control to those surfaces so that the bond doesn't have to look too far from the interface for attachment nodes.

      -- Vince
        • Wrong behaviour with incompatible mesh
          Sorry, I have troubles in finding an FTP resource to post my 2.6 MB zip file.
          Maybe in a few hours.
          FR
          • Wrong behaviour with incompatible mesh
            I often use the "Bonded+Incompatible mesh" in assemblies, as it solves many meshing issues.
            I know it yelds less accurate stress values, but so far it has never failed to give correct overall responses.

            Now I am dealing with this assembly (SWX2007sp1.0) coming from the Scorpion Demo.
            If you try the two different studies, you will find a big difference.

            I am wondering if this it to be considered a normal quality limitation of incompatible mesh, or rather a bug.

            NOTES:
            1) Mesh with global size of *15mm* (default fails)
            2) I will try to post the assembly zip on a subsequent message.
            • Wrong behaviour with incompatible mesh
              Enfin! You can download my model from:

              http://www.rustichelli.com/Materiale/CongruenzaMesh.zip

              Again:
              1) Mesh with global size of *15mm* (default fails)
              2) top assy is "FrontSuspension.SLDASM"

              There are 2 studies:
              CompatibleMesh
              IncompatibleMesh

              I know that mesh size is coarse to an obscene extent,
              but I am doing some refinings on a copy of it, and it does not seem to improve...