9 Replies Latest reply on Mar 27, 2013 8:42 AM by Matt Taylor

    Reducing vibration by changing mass properties, Iterative method

    Matt Taylor

      Hi everyone,

       

      I am working on a project fo my dissertation to reduce vibration. A method I am considering to do this is to change the weight of the part vibrating to see I can minimise it that way. I could simulate this by changing the mass properties of the part randomly and using a trial and error method to get the best result but I haven't got the time (or patience) to do that. Does anyone know of an iterative method that can be used on solidworks to achieve this? I have seen on the tutorials a way to minimise the volume of a shelf bracket ("shape optimisation") using an iterative mehod, but I have not yet managed to find a way of applying a simular method to the weight of a part to recuce vibration (increasing fatigue life would also be an acceptable optimisation goal)

      Any help would b greatly appreciated.

      Thanks

        • Re: Reducing vibration by changing mass properties, Iterative method
          Roland Schwarz

          Interesting.  I suppose it depends on what you mean by "reduce vibration".  I suppose if you are trying to minimize fatigue, then reducing total strain per cycle or keeping strain below threshold would be a goal.

           

          There will always be vibration.  There will always be modal frequencies.  Loads and constraints can change.  Results depend not just on mass but on distribution and location of mass.

           

          I'm sure you know all the above already.  Hard to picture reducing this problem to a few discreet inputs and outputs for an optimization routine.

            • Re: Reducing vibration by changing mass properties, Iterative method
              Matt Taylor

              Hi Roland,

               

              Thanks for your response.

               

              To give you the full picture I am trying to reduce the vibration of a mirror assembly at the end of a supporting arm. I have attached a picture to try to give you a better idea.

               

              My main aim is to reduce the vibration by as much as I can, as in the current configuration at certain times the mirror vibrates too much to use effectively. An increase in fatigue life would be beneficial but is a secondary aim.

               

              I am aware that the distribution of mass would also effect the modal frequencies of vibration but as I am time constrained I was planning on choosing the most suitable location to add mass (the middle (largest) mirror) and seeing what an increase in mass does there. I have found through physical testing that there is a definite change but it would take far too long to find the optimum mass like that. 

               

              This is not the only method that I am considering. Any input on if an iterative method is possible and if it how to do it would be really appreciated. Also on a side topic if anyone has any other ideas on how to reduce the vibration they would be appreciated as well.

               

              Thanks


              Mirror bracket attached to telehandler.jpg

                • Re: Reducing vibration by changing mass properties, Iterative method
                  Bill McEachern

                  What is casuing the mirror to vibrate? Engine speed? Wheel speed? Vortex shedding? actuation of the lift? Once you know that then you can figure out what frequency you need to avoid and possibly what other frequencies you need to avoid to get the response you want out of the mirror. going stiffer is the usual thing to do if don't know where you are heading. Better to know though as it is far less taxing. Then you can just modify the stiffness or the mass of mirror so that it does not have a resonance at those frequencies you know to avoid (and their harmonics). Then say you could use the optimization to minimize the mass and maintain the required frequency.

                    • Re: Reducing vibration by changing mass properties, Iterative method
                      Roland Schwarz

                      It would be an interesting exercise to try to get the mirrors attached at node points so they do not move as much.  I'm thinking maybe that horizontal bar at the bottom could be jockeyed around so that the vertical portion is close to a node.

                       

                      Disclaimer:  I'm not an expert.  I am a mechanical engineer with a great deal of interest in sound and music, but I don't do NVH for a living.

                      • Re: Reducing vibration by changing mass properties, Iterative method
                        Matt Taylor

                        Hi Bill,

                         

                        There are a few different causes of the vibration, which makes the task a bit harder, the engine speed itself will cause vibration, I have done a full sweep of tests from idle to max revs to find the worst frequencies of this, however the mirrors also suffer from vibration when roading, I believe this to be a combination of rough surface and the tyre treads, it is hard to know for sure as I didn't have a completely smooth test track to check to see if here was the vibration pattern was consistent with the tread cycle rate.

                         

                        I have done a frequency test on the mirror assembly in solidworks, will changing he natural frequencies of the assembly away from what the machine is likely to produce make a big difference? I am working on the assumption that it will but as the vibration in a forced vibration and not natural I am not certain.

                         

                        Cheer

                          • Re: Reducing vibration by changing mass properties, Iterative method
                            Bill McEachern

                            It seems tome you need to figure out the possibilites. In my view they are one or some combination of the following:

                             

                            1) The mirror vibrates because it get excited by some external source at some resonant mode.

                            2) The mirror self excites - vortex shedding or some similar phenomena.

                            3) It is not resonating and the support structure is moving and the morror is moving in response by not vibrating excessively.

                             

                            So for situation 1 - change the natural frequencies so the excitations do not cause a resonnace.

                            Situation 2 - change the design to avoid self excitation issues - break up the energy of coherent vortices as an example.

                            Situation 3 - put in a structural filter so that the motions seen by the mirror are reduced in amplitude - could be a high pass, low pass or band pass filter. Examples of this are elastomeric engine mounts - they are going the other way but the idea is the same.

                    • Re: Reducing vibration by changing mass properties, Iterative method
                      Jared Conway

                      I agree with Bill here. The problem needs to be defined better.

                       

                      Once you know the frequencies, then you can analyze.

                       

                      Freq will only tell you what the natural freqs are. Optimize the assy to avoid them.

                       

                      If you do this, you should improve fatigue life.

                       

                      But it you want to know the displacements and stresses caused by your forcing functions, you need a dynamic analysis. Once you have that, you can investigate improvements in the geometry or if you have to go with another solution that is non geometry related.

                       

                      I don't know if I understand how linear stress or fatigue as implemented in simulation help solve your problem other than there are some optimization tools you can leverage. But they don't seem like the right types of analyses for your loading condition. Freq could be optimized but you don't know which freq needs to be avoided the most without dynamic. 1st mode is probably worst but then what?