Page 154 - Cam Design Handbook
P. 154

THB5  8/15/03  1:53 PM  Page 142

          142                      CAM DESIGN HANDBOOK

                 10








               Acceleration of output motion  .5

















                              Spline (k = 10)
                              Optimized Polydyne
                 –10
                    0                         .5                        1
                                        Normalized time
              FIGURE 5.26.  Normalized acceleration of output motions for spline (k = 10) and optimized
              polydyne in Example 8.




          Evaluation of the Follower Response The physical parameters of the cam-follower
          system used are: M = 0.75lb, K s = 200lb/in, K f = 10000lb/in, w d = 1000rpm, b = p/3rad.
          h = 0.5in. and F p = 70lb. A damping ratio of 0.5(C s + C f)/M/w n = 0.2 is used (Rothbart,
          1958) and the damping coefficients C s and C f are set equal.
             The performance of the spline-based motion as compared to the optimized polydyne
          motion is shown in the following figures. Table 5.7 shows the peak values of vibrations
          and contact forces for these two output motions. Figure 5.29 compares the amplitudes of
          primary and residual vibrations of the two motions with the damping values given above.

          Finding the Cam Motion To solve differential Eq. (5.18) for the cam displacement and
          its  derivatives,  the  spline  collocation  method  is  applied  again.  Here  the  seven
          boundary conditions cited earlier must be satisfied. Also the left side of Eq. (5.18) at the
          64 collocation points must be satisfied. Application of these constraints yields a total of
          71 conditions requiring 71 B-splines. Splines of order k = 11 are used, requiring the knots
          at t = 0 and 1 to be repeated eleven times. Again, at mesh points of t = 1/16, 2/16,... ,
          and 15/16 the knots are repeated four times each. As before a system of linear equations
          is assembled and solved for the unknown coefficients in Eq. (5.6). Once this is done, the
   149   150   151   152   153   154   155   156   157   158   159