6 Replies Latest reply on Mar 28, 2010 7:42 AM by Derek Bishop

    Shaft constraints



      Does anybody know what is the best way to constrain that shaft at the bearings journals? The shaft is loaded radially by force and a torque is applied also?

      If I use bearing support (rigid, self aligning) the torque does not pass through them. If I use fixed hinges, the torque passes through them, but now the supports do not bend. See below the images.





      Shaft with bearing supports (rigid, slef-aligned)


      Shaft with fixed hindges at the bearing journals


        • Re: Shaft constraints

          Bearing supports only input a torque if there's friction in them.  (usually bearings are fairly close to frictionLESS.)  You should use rigid, self-aligning bearing supports.  You need to define some other restraint that will oppose the input of a torque.  For example, a zero-rotation about a cylindrical axis constraint.  This can be found under "advanced fixtures."


          Good luck!


          David Fletcher

            • Re: Shaft constraints

              It does not work. On the pic below you see that the toque can not pass through the bearing support. The animation shows that there is no twist in the bearings supports  no matter how I restrain the opposing torque. The shaft should twist from the application of the torque until the torque restraint but it does not. Bearings do not work like that. They would allow the shaft to rotate and generally could have some axial and radila movement.

              I am confused about your idea, I did not get it.




              Message was edited by: Vanyo Kirov

            • Re: Shaft constraints
              Have you tried using reference geometry instead of a hinge restraint?  You can limit movement in the axial and radial directions and not in the circumferential direction.
              • Re: Shaft constraints
                Pete Yodis

                Have you tried modeling the housing structure and then using either bonded or no penetration contact between the bearing and housing and bearing and shaft?  I am working on a similar project as we speak.  I would imagine that modeling the bearing support areas as infinitely stiff will change your deflection in your shaft compared to modeling the bearing supports as having some stiffness.  The stiffer the bearing support areas are made, the greater the deflection in your shaft.  Also, is your application of forces and torque area a gear mesh?  What would happen if you applied the torque to the area that is now fixed and modeled the mesh contact and restrained that other gear mesh shaft from rotating?  The gear mesh applies a load in a way to output shaft, but there is also some stiffening effect on the output shaft from the gear interaction with the other gear shaft in mesh.  I don't think your setup accounts for this, and I don't know how much of a difference it would make.  Like I said I am working on a similar problem myself.  I am more interested in the bearing reaction loads for consideration of bearing life.  Traditonal statically determinate methods of applying a gear separating force and a torque and overhung load to the output shaft, while considering the bearing area to be under no deflection - will yield reaction forces at the bearing locations.  I would like to know if those hand calc methods are overly conservative b/c they consider the bearing areas to be infinitely stiff while also not taking into consideration the load sharing that might occur from the gear mesh also acting like a bearing in some regards.  It would be good to continue this dialogue.  I would like learn more about what is really happening rather than assume traditional hand calc methods capture reality to a great enough degree.

                • Re: Shaft constraints
                  Peter Biggert
                  Try using Remote Load Displacement constraints with cylindrical coordinate system. (Create a Reference axis colinear with your shaft axis.) You can then only restrain the radial deflection from a point defined on the shaft axis allowing deflections in the axial and circumferential directions. Your axial and circumferential restraints can be put elsewhere on the shaft.
                  • Re: Shaft constraints
                    Derek Bishop

                    How do you know the torque does not pass through the bearing using a self aligning bearing? I did an example and found that it worked fine. It may just be that the bending stesses are overshadowing the torque stresses so they are not easily detected in the stress plot.



                    Shaft 1.JPG




                    Shaft 2.JPG