Page 197 - Biomedical Engineering and Design Handbook Volume 1, Fundamentals
P. 197

174  BIOMECHANICS OF THE HUMAN BODY

                                                    FFT of the EMG Data
                            70


                            60


                            50



                            40
                           Power

                            30


                            20


                            10



                             0
                              0     50   100   150   200   250   300   350   400   450   500
                                                       Frequency (Hz)
                         FIGURE 7.19  The fast Fourier transform (FFT) of an EMG sample. Most of the signal power is in the 20- to
                         200-Hz range.

                       Modeling Activation Dynamics. As noted in the subsection “Neural Excitation of Muscle” muscle
                       cannot activate or relax instantaneously. The delay between excitation and activation (or the development
                       of muscle force) is due mainly to the time taken for calcium pumped out of the sarcoplasmic reticulum
                       to travel down the T-tubule system and bind to troponin (Ebashi and Endo, 1968). This delay is often
                       modeled as a first-order process (Zajac and Gordon, 1989; Pandy et al., 1992):


                                     ⎛ 1  ⎞        ⎛ 1  ⎞
                                                m
                                            2
                                  m
                                                             m
                                                                            m
                                                                                m
                                    a =   ( u − ua )  +  ( u a )  u u =  u t()  a =  a ()  (7.6)
                                                          −
                                                                                 t
                                     ⎜ ⎝ τ rise ⎟ ⎠  ⎜ ⎝ τ fall  ⎟ ⎠
                       where  u  represents the net neural drive to muscle and  a  m  the activation level. Other forms of this
                       relation are also possible; for example, an equation that is linear in the control  u  has been used to
                       simulate jumping (Pandy et al., 1990) as well as pedaling (Raasch et al., 1997). Implicit in the for-
                       mulation of Eq. (7.6) is the assumption that muscle activation depends only on a single variable  . u
                       Other models assume that  a  depends on two inputs,  u 1  and  u 2  say, which represent the separate
                       effects of recruitment and stimulation frequency (Hatze, 1978). In simulations of multijoint move-
                       ment, whether or not both recruitment and stimulation frequency are incorporated in a model of
                       excitation-contraction (activation) dynamics is probably not as important as the values assumed for
                       the time constants,  τ rise  and  τ fall  . Values of these constants range from 12 to 20 ms for rise time,
                       τ  rise ,  and from 24 to 200 ms for relaxation time,  τ fall  (Pandy, 2001). Changes in the values of  τ rise
                       and  τ fall  within the ranges indicated can have a significant effect on predictions of movement coor-
                       dination (Anderson and Pandy, unpublished results).
   192   193   194   195   196   197   198   199   200   201   202