2 Replies Latest reply on Feb 8, 2012 4:08 PM by Samuel Haworth

    Having Trouble With Contact Sets in Assembly Simulation

    Dean Baragar

      I have a hitch stub that is welded to a large heavy blade and no matter what I define - component contact or contact set I have several items that have no stress while the parts they are welded to are experiencing half their yield strength in stress. I am relatively new to Simulation at the assembly level but I have had success with most of my contact sets by using the "bonded" type.


      You'll notice that one 'wing' as we call them is under quite significant stress while the other has absolutely no stress. This wing has two fixed surfaces on the far top right of the image - the same as the other wing, which essentially puts one wing under compression, and one under tension as the simulation was to pull the hitch at 45 degrees.


      Can anyone speculate why this isnt working as expected?

        • Re: Having Trouble With Contact Sets in Assembly Simulation
          Ryan Werner

          Hi Dean,


          It would certainly appear that something is not touching or no contact set exists somewhere.  However, without seeing your actual model it is hard to offer much advice.  Your model should have a globally bonded contact set as default so that any touching faces will act as though they are welded together.  However, it is important to note that any small amount of interference will cause the pieces to no longer be bonded together.  If there is no interference and the faces are touching than you may need to add a separate bonded contact set between the faces you want to have bonded.  Hope this helps.  If you want to upload a model I would be happy to take a look.


          Ryan W.

          • Re: Having Trouble With Contact Sets in Assembly Simulation
            Samuel Haworth

            Is the bonded contact incompatible mesh? If so, in your study properties, select "Simplified" for the incompatible mesh bonded contact setting. The "more accurate" bonding method is full of bugs, and will cause the kind of random connectivity that you are showing in your image.


            If you are using a global bonded contact setting, make sure you have "remesh failed components with incompatible mesh" turned off. Or at least use the simplified contact setting already mentioned. Also note, compatible mesh will produce more accurate stress results in the proximity of the contact, but will not produce accurate free body forces anywhere. Global bonding screws up the free body forces calculation.


            Frankly, this is all ridiculous. Bonded contact worked fine in SW 2006. And the problem with free body force calculations was first reported in 2008. It is now 4 years later and bonded contact and free body forces are a mess. Add to this the fact that study results and meshes can be corrupted by simply rebuilding the model or switching to an exploded view (neither of which were problems in 2006) and this software appears to be regressing with regards to some fundamental functionality.