8 Replies Latest reply on Aug 11, 2011 12:14 PM by Mark Smith

    Bonded Bodies Overlap During Drop Test

    Mark Smith

      After running a drop test simulation on a an assembly, surfaces that have Bonded contact sets intersect and thus show significant differences in resulting displacement.

       

      More detail:

      I designed a plastic part with a stainless steel bracket (spine) insert molded into the part to add strength during drop.  In terms of modelling, the SS bracket was inserted into the part in an assembly, and then I used the Cavity feature to shell out the plastic part in the geometry of the bracket.  As shown below:

       

      Pre-Drop Test Model.JPG

       

      All the surfaces are contacting each other between the bracket and the plastic housing surrounding it, so I added Bonded contact sets to each of the surfaces.

      After running the drop test, the deformed model of the view above looks like this:

      Overlapping Bodies.JPG

      Why are the bodies overlapping?  The Von Mises stress is well below the yield points for both the plastic and SS.  Maybe I'm missing something simple.

       

      *Note that I also ran the exact same drop test changing the Bonded contact sets to No Penetration.  This alleviated the overlapping body issue, but the resulting Von Mises stresses were literally 15 times greater than the same test with Bonded contacts.

        • Re: Bonded Bodies Overlap During Drop Test
          Michael Feeney

          Have you looked at the mesh between these components?

          • Re: Bonded Bodies Overlap During Drop Test
            Michael Feeney

            Yep. That'll do it. : ] You're running solids for your main part and a shell for you thin SS bracket?

            • Re: Bonded Bodies Overlap During Drop Test
              Michael Feeney

              If the simulation is taking too long with a proper density of all solid elements, I believe you can do either one of the following:

              1. Use solids for your main plastic part and use shells for the thin walled SS bracket. You might have to do a bit of digging through the forums to make sure that the shell elements of the SS part properly link with the solid elements of the plastic part. I don't beleive this is very difficult and is, for the most part automated      or..........

               

                 2.  Still use solid elements but have a localized high density of elements in your SS part and around the plastic housing of the plastic part.

               

              In your picture it is diffcult to see what's going on...but it sure looks like your mesh is too coarse and relatively non-uniform.

                • Re: Bonded Bodies Overlap During Drop Test
                  Mark Smith

                  Awesome, thanks for the info.

                   

                  I wanted to update with some new information from our CAD resellers.

                   

                  SolidWorks 2011 changed the default mesh mode to Curvature Based from Standard.  When I sent the assembly referenced in this post to them for evaluation, SW2011 showed the same results with overlapping bodies when the default Curvature Based mesh was used.  In other words, there seems to be an inherent problem with the Curvature Based mesh when it comes to coincident faces with bonded contact sets because the meshes aren't guaranteed to be compatible between the faces.  No Penetration contact sets between coincident faces are not an issue, but that's because the meshes between those faces don't have to be compatible to make the calculations.

                   

                  The issue is being reported up the chain, and hopefully there will be a resolution with a subsequent service pack.  Until then, if you have mating surfaces that need to be bonded for a simulation, you should mesh your model with the standard mesh, not curvature based.