72.3 Saturday, Jan. 5 The role of natural frequency in a jumping robot AGUILAR, JA*; LESOV, A; WIESENFELD, K; GOLDMAN, DI; Georgia Tech; Georgia Tech; Georgia Tech; Georgia Tech email@example.com
Many animals and robots jump to reach higher ground, to escape from predators, and even as primary mode of locomotion. At a basic level, jumping involves transient bursts of actuation of a mass coupled with internal elastic elements to generate movement. A hypothesis, then, is that this system’s natural frequency, f0, (from mass and elasticity) should play a crucial role in maximizing jump performance. While there have been many models created to simulate jumping, these often have many parameters and multi-link legs, making it a challenge to analyze the dynamics of such systems. To probe in detail how natural frequency affects jumping performance, we study a simple robot comprising a periodically actuated mass-spring arrangement. The actuator frequency and phase are systematically varied to find optimal performance. If forced for N=2 or more cycles, robot lift-off is achieved optimally at resonance. However, for the fastest lift-off, (N=1), maximal jump heights surprisingly occur above and below (but not at) f0. A simple model reveals how jumping, which occurs at transient time scales, is optimized less by resonant build up and more by proper timing and phasing. Two distinct jumping modes emerge: a simple jump, which is optimal above f0, is achievable with a squat maneuver, and a “stutter jump”, which is optimal below f0, is generated with a counter-movement. The stutter jump is slow but uses less power, while the single jump has a fast time to takeoff but requires higher power input. We propose that animal musculoskeletal systems can target these different jumping templates to make situation-appropriate tradeoffs between time-to-takeoff and internal power.