6 Replies Latest reply on Mar 30, 2010 2:06 PM by Matt Medford

    Interference Fits

    Matt Medford

      I'm interested in simulating a bolted joint.  In Ansys, I can use a bolt whose geometry is line-on-line at the bolt and nut interfaces with the mating flanges.  I can then set the initial interference at either location through options on the contact elements and it will put the bolt in tension for me.  (Alternatively, I can model the initial interference into the geomety and tell the contact element to use the geometry to determine the initial interference, which seem to be the approach that Simulation uses.)  I would like to be able to set the initial interference through an option in the contact (rather than using the actual part geometry) because I need to iterate on the values to get the correct preload in the bolt (it's not known precisely a priori) and that's more easily done through multiple studies in a single assembly than through creating multiple bolts with different lengths.  Is this something I can do in Simulation?  If it is, it is not obvious.

       

      Thank you.

        • Re: Interference Fits

          Hello Matt,

           

          You can use Solidworks Simulation "Bolt connectors" instead of an actual bolt part.  The bolt connector comes with a property manager that allows you to specify bolt preload (torque or axial).  You could create several studies with different bolt preloads and see if the displacement results turn out the way you want.  I realize that this is coming at the problem from the opposite direction compared to the way you would like, but it does satisfy your need for not changing part geometry and having a property manager to tweak interference settings.

           

          There could be other options to set the initial displacement for a bolt connector.  You can read more in the simulation help files under "bolt connector."

           

          I hope that helps!

          David

            • Re: Interference Fits
              Matt Medford

              Yes, I've considered using the "bolt connectors" but I don't believe it will work in all cases I'm interested in.  Specifically, I'm looking at a clamp system which uses an array of jack screws to push a plate down onto another surface.  So it's bolts in compression rather than tension and there are no bolt heads or nut heads.  At any rate, I don't like that SW dumbs down the analysis of bolted joints.  Try telling P&W or GE that they should just use the "bolt connectors" feature to analyze a bolted joint in their engines.

               

              Thanks for the response.  If nothing else it confirms that there's no option to change a shrink fit value on the fly.

                • Re: Interference Fits
                  Bill McEachern
                  SWX provides the connector and you decide whether it is appropriate to use them for your application. If you want  a full fidelity solid fastener representation knock yourself out. - You can load it anyway you want (tension & compression) by applying temperature as a load. Though you may have to use an NL study to get the sequencing right.
                    • Re: Interference Fits
                      Matt Medford
                      Actually, temperature is the problem.  Analyzing the clamp at room temp is fairly straight-forward because I can apply equal and opposite pressures where the jackscrews contact the pressure plate and it's a reasonable assumption.  But then I can't apply temperatures because no matter what the temperature field is, I'm always going to get the load in the jack screws that I specified via the pressures.  In my case, there are different materials in the assembly so the CTE's are all quite different.  The loads in the screws (and bolts, which are also in the model) need to be able to relax or increase when the assembly is taken to temp.  All of this would be much simpler if SWX simply provided the ability to change shrink fit and/or gap values w/o having to update the geometry - like every other analytical program on the planet does.
                • Re: Interference Fits

                  Hello Matt,

                   

                   

                   

                  I was just re-reading your post and I agree, that would be really cool to have a PM for interference without having to change the model geometry.  Have you submitted an Enhancement Request about that yet?

                   

                  Staying within the realm of Solidworks at present, I thought of something.  Have you considered using Design Scenarios?  You can set up Simulation to change the length of the bolt for you (after you set up the shrink fit in the study) so that would change the preload automatically, iteratively.  I think that's what you want.  Design scenarios are fairly easy to set up, and you can have any number of parameters that vary so there's infinite possibilities of studies you can run.  All you would probably need is the dimension name of your bolt length.  Set that as a parameter and pick a range of values that you want to test.  The design scenario basically re-meshes and runs with the new parameters and spits out results when done.

                   

                  Getting the results you want out of design scenarios is another thing that I'm not too familiar with, since I didn't actually get that far.  It didn't work on the study I was doing b'c SW kept re-setting the mesh to the highest value on the slider (my setting that worked was a bit higher than that) which caused a mesh failure.  I didn't have time to fix it, so I just did the design scenario manually.  Setting up, however, was quite easy.

                    • Re: Interference Fits
                      Matt Medford

                      Hi David,

                       

                      Yes, I've submitted it as an Enhancement Request.  Lord only knows if it'll get done, though.

                       

                      Thanks for your thoughts on Design Scenarios.  I briefly looked into that as an option but threw it out because it wasn't as intuitive as I'd like.  I'll take another look at it.