| The overall goal of the four studies presented herein is to identify physiological factors (i.e. form) that ultimately govern mechanical behavior at the muscle, joint, and limb level (function), as well as the mechanical and energetic effects of modifying form through the use of passive elastic exoskeletons. This was done using computational and experimental models of vertical hopping.;Computational model of unassisted hopping consisted of a Hill-type muscle model of the triceps surae-Achilles tendon complex working against a gravitational load during cyclic contractions. In chapter 1, we used this model to sweep a 2D parameter space of frequency and magnitude of muscle stimulation centered on the passive resonant frequency (o0) of the modeled biological-inertial system. Results from this study indicate that frequency of stimulation plays a primary role in regulating whole muscle-tendon unit (MTU) dynamics, including phase of stimulation and peak force, average positive power, and apparent efficiency. Stimulation amplitude was primarily responsible for regulating peak active forces, contractile element (i.e. active muscle, CE) and series elastic element (i.e. tendon, aponeurosis, SEE) interaction, and overall metabolic demand. Peak 'tuning' of CE-SEE interaction in this study was observed at driving frequencies just above o0 of the passive MTU system.;Due to known inaccuracies of Hill-type muscle models, chapter 2 replaced the modeled MTU with a biological one from the American bullfrog Rana Catesbeiana, and simulated inertial environments similar to those in chapter 1 on a feedback controlled ergometer. We characterized o 0 by allowing the MTU to oscillate passively against simulated gravitational loads, and drove the muscle contraction via direct nerve stimulation across a range of frequencies centered on o0. We found that a driving frequency of o0 maximized force output, minimized the ratio of CE to MTU work, minimized estimated metabolic demand, and maximized MTU apparent efficiency due to inherent frequency-phase coupling of system dynamics. This study concludes that high level active control are not required to 'tune' muscle tendon interactions if driving frequency matches o0 of the passive biomechanical system.;Computational models of Exo assisted hopping in chapter 3 were developed using the same base model as chapter 1, with the addition of a linear spring in parallel to modeled biological components. By sweeping a 2D parameter space of stimulation amplitude and Exo stiffness at a fixed frequency, we were able to identify trends in observed behavior that mimic human response to Exo assisted hopping. This included constant MTU+Exo stiffness and CE positive power output, reductions in MTU force, SEE energy cycling, MTU apparent efficiency, and metabolic demand, and increases in MTU+Exo efficiency. This model also provided insight into mechanisms underlying metabolic cost minimization and enhanced performance in Exo assisted hopping, and concludes that these outcomes cannot be optimized simultaneously, i.e. one must come at the detriment of the other.;Our final experimental study used a preparation similar to that form Chapter 2, with modified environment controllers that simulated an exoskeleton in parallel with the biological MTU. We used a 'pulsed' rate coding approach to control relative levels of stimulation amplitude, and selected stimulation amplitude-Exo stiffness combinations from chapter 3 that mimicked human response to exoskeleton assistance. In all conditions, the biological MTU was driven at its estimated o0. This study indicates that increased due to an artificially stiffened MTU+Exo system, along with invariant stimulation frequencies, may be a critical factor limiting beneficial response to springy assistance. |
