7 Replies Latest reply on Dec 3, 2013 4:41 PM by Jared Conway

    Bolt Stress & Pre-Load

    Amit Gupta

      Hi All,

       

      I have an assembly where I am trying to see the stress profile developed through various bolts within my assembly.  The issue is that I can't figure out how to accurately model the pre-load.  If I use the bolted connector, then I get total reaction forces on the bolt but can't see how the stresses are distributed within the bolt.  If I model the bolt & nut then I am unsure how to accurately apply the pre-load.  If I know what my pre-load clamping force should be, is it fair to apply that as an axial force on the end of the bolt?  My thought process is that any force acting against the bolt would decrease the clamping force while maintaining the total load on the bolt - thereby mimicking a pre-load.  Does this seem fair? The pitfall is that when I do this, I get vastly different results from using a bolted connector.  Any thoughts would be greatly appreciated.  Thanks!

       

      -Amit

        • Re: Bolt Stress & Pre-Load
          Jared Conway

          post a sample file that shows what you've done and all the different configs. the methods you describe are the ways to handle it. i suspect a setup issue here is the problem.

           

          another option to consider for bolt preload, temperature contraction in the axial direction only onl a modeled bolt.

            • Re: Bolt Stress & Pre-Load
              Amit Gupta

              Hi Jared,

               

              Please find a sample of the setup attached.  I have three simulation scenarios set up:

               

              1. With the bolt modeled and no pre-load where the bolt is bonded to the threaded connection
              2. With the bolt connector in place of the bolt/washer.  A pre-load torque is given to the connector.
              3. With the bolt modeled and a force on the end of the bolt to simulate the pre-load.

               

              The issue is that the forces shown by the "list bolt/pin forces" result in average stresses much higher than I see with the modeled bolt.

               

              Any help you can provide would be appreciated!  Thanks!

               

              -Amit

                • Re: Bolt Stress & Pre-Load
                  Jared Conway

                  Hi Amit, I have not been able to look at your model. Could you provide a summary of your outputs. How different are they? How did you validate that the loads are equivalent in all tests? IE are the displacements and overall forces the same?

                    • Re: Bolt Stress & Pre-Load
                      Amit Gupta

                      Hi Jared,

                       

                      In all three of the setups the fixtures and external loads are the same (with the exception of the bolt pre-load).  The displacements do vary significantly (factor of 20 between the simulations) with the bolt connector (scenario 2) showing the lowest displacement.  The run with the external force applied to the bolt (Scenario 3) has the largest deformation and I was attributing the difference in deformation to the joint stiffness being ignored.  For reference, I attached a screen shot of the displacement results for all three scenarios. 

                       

                       

                      Clearly all three scenarios are not consistent with each other.  I don't expect the simulation with no pre-load to be accurate (Scenario 1) however I am surprised at the difference between Scenario 2 & 3.  Right now I'm not sure how to model my joint to create good results.  Nor am I sure that I should be trusting the bolt connector results or using those results to validate my other models.

                       

                      Any help you could provide would be appreciated!

                       

                      Best,

                      Amit

                        • Re: Bolt Stress & Pre-Load
                          Jared Conway

                          how did you validate they were equivalent loads other than looking at the loads? (see a recent forum post where someone was trying to use torque and equate it to axial load and the results were different because of the internal calculation for torque vs axial)

                           

                          i don't know what i'm looking at in the screenshot. what are you comparing. this is why i was asking for a summary of the setups vs results and ideally, vs what your expectations/calculations are. this is the golden answer.

                  • Re: Bolt Stress & Pre-Load
                    Peter Biggert

                    If you want to model the bolt use the "shrink fit" contact on one end of the bolt to simulate the bolt preload. You have to create your model with the appropriate interference.  You can run your model without external loads and see how much stress occurs in your bolts. Then add the external loads and you can see how the stress changes.

                     

                    Adding a force to the end of the bolt does not model the interaction between the bolt stiffness and the joint stiffness. A good place to learn about this is at www.boltscience.com.

                      • Re: Bolt Stress & Pre-Load
                        Amit Gupta

                        Hi Peter,

                         

                        Thank you for the reference and information.  I read everything at Bolt Science however I still don't understand why my approach is wrong.  You are correct that by adding a force to the end of the bolt, I am negating all effects of joint stiffness.  However, if I am trying to simulate the worst case condition, shouldn't I assume that my bolt stiffness is much greater base material stiffness and that my applied load goes completely into the bolt?  I am trying to be conservative and believe that this is the conservative way to model it (at least in relation to the bolt).  If that is not correct, can you please help me understand why? 

                         

                        Your tip about modelling the pre-load as a shrink fit is intriguing and I haven't tried that yet.  I will be doing that next to see if the the stresses develop as I would expect (and how they compare to other methods).

                         

                        Thanks for all the help!

                         

                        Best,

                        Amit