8 Replies Latest reply on Jun 10, 2010 4:56 AM by Stefan Wozniak

    Shaft Vibration / Deflection

    Corey Cooper

      I have a problem I am trying to analyze using Simulation.  I need to analyze a rotating shaft to determine harmonic vibrations and deflections at certain RPMs. I have set up the simulation to use bearing fixtures at two split portions at the end of the shaft.  The details are (I'm also attaching the solidworks file of the shaft):


      RPM: 3200

      Shaft dia: 3.001"

      Shaft face length: 92.125"

      Bearing face length: 2.125"

      Material: Hot Rolled Steel 1018


      I've tried using the Frequency analysis solver but get extreme deformations that I know aren't accurate through observation.


      Anyone have any suggestions?

        • Re: Shaft Vibration / Deflection
          Poovaiah Devenira


          I am not sure if the deflection is accurate, however for a 3.00 in dia  shaft, turning at 3500 RPM the bearings would not be spaced more than 58.00 in apart

          • Re: Shaft Vibration / Deflection
            Jason Nicholson

            When you run a frequency analysis, you are just getting the mode shapes.  You are not getting actual deformation.  You are only getting the shape in which it would vibrate.  If you remember matrix algebra and differential equations and want a great tangent, look up eigenvectors.


            My hunch is ‘Simulation Advanced Professional” will allow you do a dynamic analysis that could give you what you need.  I can not confirm this.


            Be careful that your model is not to "stiff."  This happens when your mesh is to course.  Big tetrahedral elements are very "stiff" (like spring stiffness) which makes the displacements smaller than they actually should be.  You can check for proper stiffness by running with smaller and smaller elements until the largest displacement stabilizes and doesn’t change more than a couple percent.  This can be done manually or using "h adaptive" method (h stands for the size of the element).  Check the help file and ask more questions on the forum.


            There are many pitfalls doing a simulation like this, but you can do it.  Stick with it and learn a lot!

              • Re: Shaft Vibration / Deflection
                Poovaiah Devenira


                I agree with Jason, the resonance frequency analysis looks correct. However, the deflection does not look accurate.  I am not sure if you could depend on the deflection, the program calculates.

                  • Re: Shaft Vibration / Deflection
                    Jason Nicholson

                    Do you remember Differential equations?


                    differential equation = forcing function

                    a(t)*y(t)+ b(t)*y'(t)+y''(t)=f(t)


                    The solution was

                    solution = transient response + steady state response

                    transient response=C1*y1(t)+C2*y(t)

                    steady state response = yp(t)



                    The mode shapes are like y1(t) and y2(t).  You multiple by constants C1 and C2 to get the real displacement which is dependent on the intial conditions.  The frequency response in Solidworks is like solving for y1(t) and y2(t).  To get real displacements, you have to run a dynamic analysis with intial conditions; this is like solving for C1 and C2 in differential equations.

                • Re: Shaft Vibration / Deflection
                  Stefan Wozniak

                  Frequency study geaves you only mode shapes and natural frequencies with mas participation factor. You shouldn't use displacement or any other thata from this study.

                  To get correct displacement you should use dynamic study. Perhaps dynamic harmonic study shoul be correct.