9 Replies Latest reply on May 28, 2013 2:27 PM by Bill McEachern

    Stress in Butress Threads

    William Turk

      I have an assembly that consists of a collared metal cylinder captured by two plastic pieces joined via buttress style threads.  I can make the simulation work with no friction, but I know that’s not realistic.  When I tried to run the simulation, SolidWorks keeps telling me to reduce the coefficient of friction.  I’ve run it down to 0.05, and it still stalls.  I would greatly appreciate it if someone could take a look at the attached assembly and tell me if there is anything I could do to make it work.

        • Re: Stress in Butress Threads
          Jared Conway

          What is the exact error message that you get?

            • Re: Stress in Butress Threads
              William Turk

              "Slow convergence to equilibrium.  Reduce coefficient of friction and try again."

              Torque Error 2.bmp

                • Re: Stress in Butress Threads
                  Jared Conway

                  how much do you think this will rotate?

                   

                  when you solved it without friction, how long did it take and did it solve without errors?

                   

                  this is going to be a pretty tough simulation. probably better as a nonlinear analysis.

                   

                  in the KB, there is one suggestion which is to shut down solidworks and delete the CWR and start over which has worked for me on specific problems where I've received this error but in other times, it is because of the setup of the problem. you may have to try different contact conditions and try simplfying your problem to get it to solve with linear.

                    • Re: Stress in Butress Threads
                      William Turk

                      In the initial state, I have the green part rotated so there is 0.0005" clearance between the threads.  Any less than that and SolidWorks starts picking up small interferences.  With that in mind, it shouldn’t rotate much before the threads come into contact.  Given that the torque is relatively small, I don’t think it’ll rotate much after contact either.  Studies on the individual parts indicate that they should only stretch a few thousandths of an inch.

                       

                      I don’t recall exactly how long it took the no-friction model to run, but it was less than thirty minutes.

                        • Re: Stress in Butress Threads
                          Jared Conway

                          and you purposely have the middle component as rigid?

                            • Re: Stress in Butress Threads
                              Jared Conway

                              Hi William, ran the no friction model in 2013 sp3.0. As expected, there is a large displacement error. Meaning the results are basically no good if you choose the option "no". And I suspect that it will fail to solve if you enable the large displacement option.

                               

                              Is it possible to run this analysis in solidworks simulation? Probably, but it will need a bit of work and i'd suggest nonlinear. We can certainly help you out with this at Hawk Ridge if you'd like. My guess is you're going to have to go to all parts being deformable, use the friction that is appropriate, try different contact options and start the analysis with everything touching.

                               

                              If you can get away with running the analysis on individual parts, that is probably your best bet to start.

                                • Re: Stress in Butress Threads
                                  William Turk

                                  Jared,

                                   

                                  Thank you for all your help.  If nothing else, I have a greater appreciation for why SolidWorks Simulation has Bolt Connections!

                                   

                                  I don’t think I can run a non-linear analysis.  I have SolidWorks Premium, but my only option seems to be basic static solutions.  I’ll check with my IT guy to tee if that’s something we can load.  In the meantime, I’ll evaluate the parts individually.  I think the best way is to determine the theoretical load by Torque = (Coefficient of Friction)*(Major Diameter)*(Force) and apply it directly to the threads, unless you can suggest a better solution.

                                   

                                  William

                                • Re: Stress in Butress Threads
                                  William Turk

                                  (This is a reply to the question about leaving the central part rigid)

                                   

                                  Yes, I did that to try and simplify the problem.  That part is steel, while the other pieces are nylon, so I wasn't particularly concerned with the stresses there.

                                   

                                  Message was edited by: William Turk

                                    • Re: Stress in Butress Threads
                                      Bill McEachern

                                      You could try using a shrink fit to avoid the interference issues. The program will automatically do a sequenced analysis with the first load step as the shrink fit solution and then apply the torque. Depends what you are looking for as to whether it makes any sense to do this. It will run in the software you have though.