12 Replies Latest reply on Aug 20, 2014 10:34 PM by Jared Conway

    Frequency Analysis on Assembly

    Nathan Cassou

      Happy Wednesday,

       

      I need to know the resonant frequency of the default (6) modes on a frequency analysis on Simulaion Pro. I simply have an aluminum plate that have several small circuit boards (FR4 laminate material) fastened to it via 0-80 bolts/nuts. What I have done is instead of modeling the bolts and going through and applying contacts and loads on them, I just made them all Bolt Connections and applied a pound of axial load to each (just to start out). The problem is that I always get a warning message stating that there are no contact pairs on the bolt connections. I run the simulation anyway and my parts seem to just fly off the plate at only 2 Hz. I basically need to know what the frequencies are on the whole assembly. Frequency analysis does not allow for No Penetration on parts.

       

      Thanks

        • Re: Frequency Analysis on Assembly
          Jared Conway

          What is your default contact condition?

           

          You'll need it to be bonded but recognize that will change the way your bolts behave.

          • Re: Frequency Analysis on Assembly
            Mike Pogue

            A good way to get frequency is to delete the bolts and use pin connections. Allow rotation but not translation (pin + snap ring).

            • Re: Frequency Analysis on Assembly
              Nathan Cassou

              The default connection is a bonded condition and is really my only option besides allowing penetration which would not be accurate.

               

              I have been playing around with pinning the connections but to be accurate there is an axial load on the bolts and screws. I have tried both the Bolt Connection feature and also mating very simplified bolts to the assembly and applying loads/bonded connections.

               

              Both have not been successful. When I use the Bolt Connections feature it always says that "Contact Pairs" are not specified.

                • Re: Frequency Analysis on Assembly
                  Mike Pogue

                  Allowing penetration is much more conservative than bonded contact.

                   

                  I believe the error introduced by the pinned connectors is dwarfed by the error introduced by the other simplifications in the model, for instance, ignoring structural damping. The goal of the frequency studies I've done has not been to calculate the natural frequencies to the 3rd decimal place. That's not really possible for a design engineer using SolidWorks Simulation. The goal is usually to get an idea where the frequencies are so you can drive them up out of the range of frequencies likely to be seen in service.

                   

                  No penetration is a non-linearity that I don't think a normal modal analysis solver can handle. You'd have to write your own specialized code to handle that situation.

                    • Re: Frequency Analysis on Assembly
                      Nathan Cassou

                      Mike,

                       

                      I am still working on the frequency analysis for the PCBs mounted to the aluminum frame. You mentioned that allowing for penetration on the parts would be more accurate than using bonded contact? I am also using pins as connections.

                       

                      Thanks

                       

                      Nate

                      • Re: Frequency Analysis on Assembly
                        Davide Lauria

                        Hello guys,

                         

                        I'm experiencing the same problem in a recent job. In case of 2 isotropic plates bolted together with M24 bolts (so 1 plate with passing through holes and the other with threaded holes), how would you simulate correctly the contact condition in a modal analisys? If the contact is completely bonded, then to simulate the bolts is not very meaningful I suppose. Alternatively, simulating the bolts with a penetration contact would be very conservative...

                         

                        Thanks

                        Bye

                          • Re: Frequency Analysis on Assembly
                            Jared Conway

                            there is no contact in frequency, so there is no way to handle the bolts exactly as they are in the physical world.

                             

                            i would recommend running both simulations, bonded and allow pen (with some type of connector) and recognize thta the behavior is somewhere in between. one is more stiff one is less stiff. (basically what you describe)

                        • Re: Frequency Analysis on Assembly
                          Jerry Steiger

                          Nathan,

                           

                          I really doubt that the axial loads on the bolts are going to have any effect on your frequencies. The resonant frequency of the bolt in bending is so high above the resonant frequencies of your system that you can very safely ignore it, no matter what the tension is. The same thing goes for the details of what happens at the interface between the nuts and the board or the board and the standoffs. Any minor differences between the clamped friction and a bonded connection is trivial compared to the other simplifications.

                           

                          Jerry S.

                        • Re: Frequency Analysis on Assembly
                          Nathan Cassou

                          Okay Mike and Jared,

                           

                          I think I have a good simulation after using Pin Connections and using Bonding on edges to faces. The assembly is resonating in the modes that were expected. Basically what this will determine is the optimized geometry of the circuit board mounting plate so that it creates a higher frequency in the system. The PCBs have sensitive gyros and accelerometers printed on them and any weird resonant mode shapes will have an effect on the chip sensitivity.