10 Replies Latest reply on Aug 6, 2013 7:08 PM by Jerry Steiger

    What am I doing wrong in this Simulation?

    Eric Henderson

      I'm pretty new to Simulation and I'm trying to do a simple analysis on an arm that is bolted together with 4 socket-cap screws, fixed (for all intents and purposes) with no rotation through a hole in the bottom, and a small load on the opposite end. This should be pretty easy but for some reason SW is taking HOURS to run this before giving me an error. Can someone help? Pack & Go files attached in zip folder.

      ARM TEST.JPG

        • Re: What am I doing wrong in this Simulation?
          Jared Conway

          study 1 or study 2?

           

          i noticed you have gravity in there. does it solve with gravity only? what about just the preload from the bolts only? that will tell you a lot about whats going on.

           

          also, any specific reason for using ffe plus?

           

          any reason that symmetry isn't being leveraged? half the problem = half the solution time.

           

          the mesh is cranked way down so i'm surprised thati t took a long time but haven't had a chance try it myself. can do that tomorrow once i know which one to take a look at.

          • Re: What am I doing wrong in this Simulation?
            Chatura Elkaduwa

            HI Eric,

             

            1. You may set a local no penetration contact set with "node to surface" option, between the faces indicated below.
              This will override the default "surface to surface" contact condition and will simplify the analysis process
              Untitled.png
            2. Use curvature based mesh at default settings; you may keep the pre defined the mesh control settings.

            3. Use the direct sparse solver; it is more stable, when it comes to studies with contact conditions.

             

            It solved in my workstation within 8mins (2.6Ghz xeon processor and 8GB RAM)

             

            If you need to check if the solver manages to complete without giving any errors, you may use a coarse mesh with draft quality elements (which solves in a few seconds)

            • Re: What am I doing wrong in this Simulation?
              Eric Henderson

              Thanks for the responses.

               

              Jared,

              For some reason, FFE plus was selected as the default solver. I changed it to automatic later and was able to get a solution but it didn't seem reasonable. The assembly isn't really symmetrical.....would symmetry still apply in this situation? Do my fixtures and connections seem correct? I was having trouble determining which fixture to use.

               

              Chatura,

              1. I was getting a warning about not having "contact set" with the bolt connections. Is that what you are referring to here?

               

               

              I WAS finally able to get a solution but it didn't seem reasonable. According to the simulation, the metal yielded with just the 10lb force (and gravity). I know from "real" testing that it can take way more than that. Any suggestions as to why that would be?

                • Re: What am I doing wrong in this Simulation?
                  Jared Conway

                  only apply symmetry if everything is symmetric but it looked pretty close. FE is an approximation so if even if there is a small assymetry and you apply it, it may not be the biggest assumption you make. regarding your loading condition, we don't have enough information about what you're trying to study to make that determination. the setup looks reasonable.

                   

                  regarding the contact set error, you may get that because you're using the global contact. i think that is a known issue. but whether you use a contact set or global, that is still valid for a bolt and you can ignore the message. but just check that the plates separate and you'll be ok.

                   

                  just out of curiosity, what did you change to get it to run and how long did it take? something i didn't mention is checking where your results are going. lots of people are using dropbox these days. that can slow things down or cause you trouble when it otherwise wouldn't.

                   

                  can you post a picture of your results. make sure that you're not just reading the results in an area where high stress is numerical. do some research on singularities for more info on this.

                    • Re: What am I doing wrong in this Simulation?
                      Eric Henderson

                      Here's the stress plot I was able to get. For what it is, it seems reasonable.

                      STUDY2 STRESS.jpg

                       

                      The Displacement plot seems ridiculous however.

                      STUDY2 DISPL.JPG

                       

                      to get a result, i believe i changed the contact set and maybe the fixture type (i've ran it so many times now, it all runs together). it only took a few minutes to run. now, in an effort to make the model a little more accurate, i've added some steel washers around the "mounting" hole. i've gotten some results that seem reasonable but when i change the value of the load, the results dont change. this is giving me a headache. 

                       

                      to give you an idea of what i'm trying to do, here's a shot of the full model.

                      FD171ARM.JPG