Parallel pulsed jets for precise underwater propulsion

Abstract

A significant limitation for underwater robots is their ability to maneuver in tight spaces or for complex tracking tasks. Next generation vehicles require thrusters that can overcome this problem and efficiently provide high maneuverability at low speeds. Recently, thruster design has begun to draw inspiration from nature's swimmers, applying the principles of pulsed jet propulsion to robotic thrusters. Although most developments have focused on single jet actuators, nature provides some indications that multi-jet systems can provide propulsive benefits -- marine invertebrates called sales connect into chains of individual animals that each eject short jets to collaboratively move the entire chain efficiently around the ocean. However, despite the promise of multi-jet propulsion, there are no existing models or empirical data that explain the physics of multi-jet propulsion. As a result, there are no physically motivated rules to guide the design of man-made multi-jet thrusters. In this thesis, I experimentally investigate how interactions between neighboring jets in a multi-jet thruster will affect the system's propulsive performance. I use high-speed fluorescence imaging to investigate the mutual influence of two pulsed jets under conditions relevant to low-speed maneuvering in a vehicle (Re ~ 350). Using a new force estimation technique developed in this thesis, I analyze the video data to evaluate how thrust and efficiency are affected by the jet spacing. This analysis reveals that, compared to non-interacting jets, the efficiency and thrust generated by the pair of interacting jets can fall by nearly 10% as the jets are brought into close proximity. Based on this data, I develop a model of vortex interactions to explain the thrust and efficiency drop. The data and model described in this thesis contribute new insights to understand vortex formation in pulsed jets, and these results can be used to guide the design of multi-jet underwater propulsion syste