A Dynamic Spring-Charging Mechanism for Energetically Enhanced Locomotion

This dissertation focuses on the design and development of a dynamic spring-charging mechanism suitable for energetically enhanced locomotion. The mechanism employs high-speed mechanical logic, triggering a high-powered mechanical reflex that energizes the locomotion system with an undersized motor while out-of-contact with the environment. Before the actual linkage design of the mechanism is instantiated, the essential mechanism characteristics for dynamic spring-charging were investigated using a simulation framework where all physical mechanism parameters were excluded, concentrating instead on the dynamic utility of the linkage to be designed. This mechanismless simulation framework allowed us to determine the essential kinematic requirements and parameters of the to-be-designed linkage, and these kinematic requirements and parameters were used to synthesize the physical mechanism for dynamic spring-charging locomotion. Once synthesized and constructed, the mechanism was bolted to a benchtop to conduct inverted kicking experiments, affirming the outcomes from the mechanismless simulation study. Subsequently, the mechanism was mounted onto a pitch-constrained planarizing boom to perform hopping tasks while the motor receives constant torque signals, demonstrating its ability to not only offload repetitive subtasks of locomotion from conventional computer-actuator-sensor setups to mechanical processes, but enable a new capability that the standard setups alone cannot achieve: out-of-contact elastic energy accumulation & its automatic release upon contact, outputting 1.6 - 2 times the kinetic energy compared to the conventional points of comparison while maintaining a more consistent trajectory. Finally, using a purpose-built rigid body dynamic simulation framework, the effects of length scaling on the dynamic spring-charging mechanism are explored, showing that compared to conventional counterparts, the dynamic spring-charging mechanism demonstrates its superior power output regardless of scale, and the best normalized power output is found at 0.75 times the built size. By raising the robot's energetic profile without upsizing the motor, this work shows how the dynamic spring-charging mechanism for mechanical processing offers new capabilities beyond alleviating control tasks.