7 Replies Latest reply on Mar 25, 2014 2:39 PM by Shaun Densberger

    Box with rubber seal

    Marc Holbach
    Marc Holbach Mar 22, 2014 9:49 PM

      I am trying to simulate what stresses there would be on a steel box with a steel frame, when a 15psi force is acting on it. The three long bars on the top are attached to the lid from the outside, and the lid is supported by the 5 top pieces of the frame (the two long pieces and the three shorter ones)

      Untitled1.png

      The simulation ran perfect, but as soon as I add a rubber gasket between the frame and the lid, either the following happens, or the simulation ignores the center support completely and deforms as if it weren't there, even though the universal contact is set to bonded.

      Untitled.png

      I have tried various conacts, but dont know all that much about simulation, and have absolutely no idea what the issue could be. Any help or tips would be highly appreciated.

        • Re: Box with rubber seal
          Shaun Densberger
          Shaun Densberger Mar 24, 2014 3:00 PM (in response to Marc Holbach)

          It looks like you have an issue with how your parts are connected together (based off of the second image). Do you have soft spring enabled? Why do you think you need the rubber gasket? What effect do you expect it to have on your stress results? Keep in mind that rubber is a hyper-elastic material, which will require a non-linear analysis.

            • Re: Box with rubber seal
              Marc Holbach
              Marc Holbach Mar 24, 2014 6:23 PM (in response to Shaun Densberger)

              Well thats the issue I'm having, that the parts are all mated a few times over, to the point where nothing should be moving out of place. The rubber gasket will serve as a seal for the enclosure and I would like to know what kind of stresses there are in the seal, and I'm hoping it will reduce some of the stresses in the lid. I will try the changes you suggested and post the results.

            • Re: Box with rubber seal
              Jared Conway
              Jared Conway Mar 24, 2014 4:38 PM (in response to Marc Holbach)

              if all of your parts are touching (interference detection > treat coincidence as interference) then it should solve as bonded

              if your parts aren't touching, you're going to have trouble like what you're seeing