4 Replies Latest reply on Mar 6, 2014 10:32 AM by Jared Conway

    Fixture assisstance

    Robert Smith

      Hi All,

       

      please see attached screenshot of skid in query.

       

      I have a skid im trying to simulate with 2 loads of 8T and 12T evenly over the four points to see if it would pass a load test.

       

      I currently just have it fixed on the under side of the bottom rails. I was just wondering if any1 could suggest a better method of restraining them model as i dont get any results through the bottom rails as its fixed.(if im right in sayin tht, please correct me if not)

       

      The skid doesnt get bolted down or anything, just sits on the surface then the load is applied to the 4 points evenly.

       

      If i add a virtual wall to the underside of the skid, would this be an efficent enough restraint.

       

      Thanks for any assisstance

        • Re: Fixture assisstance
          Aaron Gradeen

          Hey Robert,

            It sounds like changing your boundary restraints in this case might not have a big effect on your results. Are you looking for stresses in those bottom beams? If the load is purely vertical I would expect it to mostly travel directly down the columns and into the floor. Creating a virtual wall and setting up a no pen contact between the bottom beams would be another option but your solution would take much longer. This looks like a good example for using symmetry planes to cut down on simulation time.

            • Re: Fixture assisstance
              Robert Smith

              Hi Aaron,

               

              Thanks for the reply.

               

              Well just the stress all round really, but am i correct in saying that with the bottom rails being fixed, it wont show the streses in those members?

               

              I'm still quite new to the simulation side of things, when applying a symmetry, do i need to cut the model in half then apply the symmetry to those cut faces? or is there a way to do it using planes?

               

              I'll go run it again and i'll put it up for anyone to have alook.

               

              Thanks

                • Re: Fixture assisstance
                  Aaron Gradeen

                  Hey Robert - Since those beams are not truly fixed to the floor you will just not see the correct stress. From a simulation perspective, using a fixed face cuts down on simulation time and gives you a good initial result of where your stress hot spots might be. If those beams weren't fixed in place and no friction was applied to the contact between the beam and the floor, the smallest lateral load could result in the entire structure flying off into space (you would see a large displacement warning in the middle of the analysis). In reality it is the friction force which would resist any potential lateral loads. You could set a coefficient of friction in the no penetration contact but I don't think you need to. Those bottom beams really aren't doing much in your loading case.

                  As for symmetry, it looks like you could cut the model into a 1/4 and apply symmetry to each cut face. I would suggest making a new config and applying the cut at the assembly level. Let me know how it goes.

                  -A

              • Re: Fixture assisstance
                Jared Conway

                The symmetry suggestion is a great one.

                 

                Fixed is a good first approximation.

                 

                From there I'd go with 1 fixed and the rest Fixed in the normal. This allows the rest to move in the plane. With symmetry, you can just use normal because the sym bc will keep things in place.

                 

                After that you are taking about contact. I don't think you need that level.