9 Replies Latest reply on Jun 21, 2015 8:39 PM by John Willett

    Preload a Joint (as though a bolt)?

    John Willett

      Another ignorant question:

       

      I have a round rod with a flat and a smaller round projection on one end (like a bolt).  At the moment I'm inserting it into an equal-diameter hole in a flat surface and using bonded contact between the projection an the hole to simulate a threaded joint (see sketch).  But since it's really screwed in and tightened, I want a preload between the flat surfaces.  Is there a simple way to achieve this?

      Simulated Bolt.png

      SolidWorks Help says, "You can also define a bolt by selecting entities of the same component," but it doesn't say how.

       

      It also says, "You can assign mate properties for use in SOLIDWORKS Motion and SOLIDWORKS Simulation analysis."  Under Mate Property Manager/Analysis Tab/Bushing/Translational/Force, "Enter the preload applied."  I've tried this with a coincident mate between the two flats, but it doesn't seem to result in a preloaded joint in a simulation.

       

      How should be done? -- John Willett

        • Re: Preload a Joint (as though a bolt)?
          Shivani Patel

          I think the simplest way would be to apply a torque to the faces instead.

           

          "You can also define a bolt by selecting entities of the same component,"

          This would still define a bolt connector in the study. It would be a blue bolt created between two circular edges on a single part. You wouldn't use this in your model unless you got rid of the rod.

           

          "You can assign mate properties for use in SOLIDWORKS Motion and SOLIDWORKS Simulation analysis."

          Yes, SOLIDWORKS Motion can use mates. Once a Motion Analysis has been run, results can be exported for a particular time step into a Linear Static analysis in SOLIDWORKS Simulation. I'm sure you've noticed that the exported results are a bit odd and it's hard to tell whether the correct preload was translated into Simulation.

            • Re: Preload a Joint (as though a bolt)?
              John Willett

              >>I think the simplest way would be to apply a torque to the faces instead.<<

               

               

              Shivani -- Not sure I understand.  Do you mean to apply external forces normal to the opposing faces, like a couple but with no moment arm -- and a no-penetration contact condition, I presume -- to press them together with the desired preload?  If so, that sounds very straightforward.  I wish I had thought of it! -- John Willett

              • Re: Preload a Joint (as though a bolt)?
                John Willett

                >>What you have here is a shrink fit...<<

                 

                 

                Andrei -- I've heard the suggestion to use a shrink fit instead of bonded contact before in connection with a simple threaded rod into a mating threaded hole -- no flat.  I've tried that, but I never understood the point.  Can you explain?

                 

                But based on one of your references (Mate Property Manager/Analysis Tab/Option/Treat interference as a shrink/press fit -- "Treats mates that force interference as a shrink fit in SOLIDWORKS Simulation."), maybe you mean to make the two opposing flats interfere slightly and then apply a shrink-fit connection between them (or maybe shrink fit is automatically applied to any interferences caused by the mate in question when that option is selected?)?

                 

                If so, it would seem that an extremely small interference would be required to get a reasonable preload (with steel at least).  SW Help cautions about very small interferences in another of your references.  ("For accurate results, the overlap should be large enough to overcome approximations introduced by meshing. For example, the interference of cylindrical or spherical faces should be larger than 0.1% of the larger diameter at the interface for accurate results.")  Any tricks or pitfalls I need to know about?  Any particular solver or other study options required?

                 

                (Please note that I'm running SW Premium, hence do not have a non-linear solver available, just the Large-Displacement option if needed). -- John Willett

                  • Re: Preload a Joint (as though a bolt)?
                    Andrei Popov

                    Hm maybe I misunderstood your question, I was thinking you have a shrink fit between the two cylindrical faces of the bolt and hole, but you want a pressure load between the two flat faces... but what are you trying to simulate? motion analysis or stress analysis? are you trying to simulate the friction when the bolt rotates or the stress in the bolt neck? from your reply I think you want to calculate stress, in this case you can load the bolt by setting a non-penetration condition between the two flat faces and apply a displacement or a force on the bolt tip surface. but with static study this is pretty much you can do, if you want to pre-load and then apply other loads in different places with a different time schedule then you need the non-linear package which by the way it is available in the Premium but not available in the Professional package.

                      • Re: Preload a Joint (as though a bolt)?
                        John Willett

                        >>from your reply I think you want to calculate stress, in this case you can load the bolt by setting a non-penetration condition between the two flat faces and apply a displacement or a force on the bolt tip surface.<<

                         

                         

                        Andrei -- Yes, that might be made to would work, as also suggested by James.  The problem is to avoid affecting the joint at the other end of the rod that hasn't been shown in my sketch, not to mention the threaded (smaller diameter) section that penetrates the plate.  I guess I was chasing the apparently built-in bolt preload idea and didn't see the forest for the trees.  Maybe Shivani was trying to tell me the same thing, but I was thrown off by the word, torque.

                         

                        You raise a good point about the time factor.  I want the "bolt" to be preloaded, but otherwise I'm really not interested in the order in which the forces are applied.  (Maybe a face-to-face "shrink fit" would help with that?  No non-linear package available to me.)

                         

                        In any case, if you have any light to shed on the shrink-fit idea (either for simulating a threaded joint, as you mentioned above, or for this flat-to-flat purpose), I'd appreciated an explanation.  As I said, I've been told this before in connection with screw threads but haven't understood why.  Maybe the idea is that bonded contact effectively converts the two parts into one, with the resulting "re-entrant corners" over-concentrating the stresses?

                          • Re: Preload a Joint (as though a bolt)?
                            Andrei Popov

                            Shrink fit applies only to cylindrical surfaces, not flat to flat.

                            Bonded contact makes the two faces to share same nodes in the mesh.

                            I would do something like this or something even simpler, all you need is a circle around the neck of the bolt that shows where the thread starts and impose a preset displacement to that entity.

                             

                            bolt preload.png

                              • Re: Preload a Joint (as though a bolt)?
                                John Willett

                                Andrei -- Thanks very much.  You've given me a great clue.

                                 

                                >>Shrink fit applies only to cylindrical surfaces, not flat to flat.<<

                                 

                                FYI this appears not to be correct.  SW Help says under Shrink Fitting, "The contacting faces need not be cylindrical."  I tried mating the bolt so that its flat interferes slightly with the plate and setting a shrink-fit contact set between the opposing flats.  It does appear to work and to give reasonable answers.  Nevertheless, your way may be better:

                                 

                                >>...all you need is a circle around the neck of the bolt that shows where the thread starts and impose a preset displacement to that entity.<<

                                 

                                Instead of setting a displacement on a circle, I set equal and opposite distributed forces on facing segments of the bolt and the inside of the hole (to simulate loading over the first three threads or so).  This appears to work well and, because it's balanced, does not necessarily create any reaction forces elsewhere.  While this approach might be more realistic, it has at least one drawback:  It will be necessary to adjust the forces to hold the opposing faces together under other external loads without overloading the "threads"...

                                 

                                Perhaps this is why people like to simulate a threaded joint with a shrink fit, as the contact forces are automatically calculated.  With appropriate friction it appears to allow some sliding to distribute the forces over several threads' length.  Preload may be accomplished with a flat-to-flat shrink fit, as mentioned above.  This combination appears to give reasonable results and to satisfy my need.  The main problem seems to be interpreting the deformed results (as image below).  The initial interferences mean that a gap in these plots does not necessarily mean a real gap.

                                Double Shrink Fit.png

                                If anyone wants to see the model, I can attach a P&G.  Am I missing anything here? -- John Willett

                      • Re: Preload a Joint (as though a bolt)?
                        James Riddell

                        John Willett, with something this simple it seems (to me at least) that if you apply a force on the end of your 'bolt' equal to the clamping force of the preload that you should get similar results.  Have you tried that?