2 Replies Latest reply on Oct 30, 2015 4:01 PM by Davide Tamaddon

    simulation in a spline shaft

    Davide Tamaddon

      Hi,

      I have a problem with the calculation of the torsional rigidity of a spline shaft.

      Using the static analysis, I selected a fixed surface, then I want to apply the torque on the spline (in order to evaluate the angular deformation and the torsional stiffnes as ratio between Torque and theta).

      I can I apply the torque? Do I have to select torsion, and then select each lateral surface of every thooth? If I do so the value of the torque will be split on every thooth?

      Please help a poor student

        • Re: simulation in a spline shaft
          Janko Stellaard

          There are multiple ways to do it. What have you tried, shouldn't be that complicated.

           

          You can select a torque and apply it to all the faces. Then select a axis or circular face. There is an option to apply the input torque value 'per item' or 'total'. So that means your value divided by the amount of faces selected (total) or your value applied to every single selected face (per item).

           

            • Re: simulation in a spline shaft
              Davide Tamaddon

              Hi Janko,

              Thank you very much for your answer.

              The most important problem that I have with this simulation, is  that I don't know what is the best way to simulate the fixed geometry. What I mean is: since the torque is applied only in the middle of the lenght of the shaft, can I cut and delete the ends of the shaft (which are not stressed)??

              In this way I would have I comfortable way to select a section (the end one) and select the fixed geometry (which is the reaction to the torque applied).

              I have this idea because in the real shaft the torque (trasmitted by the spline) arrives from a coupling (Periflex) and so on my shaft there is the keyway. But if i select the lateral surface of this keyway in my simulation I would have even a bending moment ( caused by the eccentricity between action and reaction). Having bending moment, the circonferential displacement y ( SW doesen't give me directly the angular displacement) wouldn't be only caused by the torsion. In this way when calculating tehta=y/radius I wouldn't get the theta I was looking for...

              Don't know if I can explain properly my problem.... I am very confused too

              If posting photo of my problem would help, please let me know...

              Thank you for the time you're wasting with me

              Waiting for your answer.....