4 Replies Latest reply on Nov 16, 2017 7:55 PM by Swme Eng

    Foundation Bolt for Random Vibration

    Swme Eng

      I've been messing around with connectors and foundation bolts in static simulation, and was wondering if there was another way to do something similar in random vibration? I know that obviously connectors aren't supported in linear dynamic, but I was wondering if you could just create "landing pads" for the bolt based on the pressure cone of the bolt (seen on page 19 here: http://prod.sandia.gov/techlib/access-control.cgi/2008/080371.pdf ). In doing so, you would end up creating a split face on the bottom of the part and fixing there (diameter is determined using 30deg pressure angle), then creating a similar split face on top and applying clamping force there. While this seems to help solve the issue of artificial stiffness when fixing an entire bottom face of a part, it doesn't constrain motion into the "virtual wall" plane. Just curious if anyone had explored this topic more, and what others do to more accurately simulate a part that is bolted down to a much more rigid baseplate.

        • Re: Foundation Bolt for Random Vibration
          Bill McEachern

          A random analysis is a linear dynamic procedure so by definition a non linearity is not allowed. The item is bolted to a rigid surface by the sounds of it. What subtlety are your trying to approximate? The stiffness of the connection between the item and the "rigid" plate? Springs are usually the way its done in my experience but maybe I don't really grasp what is being sought.

            • Re: Foundation Bolt for Random Vibration
              Swme Eng

              With limitations on what I can share, were are basically looking to increase the fidelity of our simulations that look specifically at displacements as a result of random vibration.

                • Re: Foundation Bolt for Random Vibration
                  Bill McEachern

                  The difference between bolted with a pre load and fixed, assuming the vibration regime in not extreme relative to the bolt pre loads, is probably pretty subtle. In ABAQUS for example you can use the state of the analysis after the pre loads are applied as the base state-for the frequency extraction and use it in the random. No similar capability exists in the SWX Sim products as far as I know, though you could do a time based direct integration analysis of say a sine in a non-linear dynamic procedure and see how it compares to modal sine analysis and see how they compare. At least it will give you some idea where you stand.