Dissertation/Thesis Abstract

A hierarchy of models for the control of fish -like locomotion
by Melli, Juan B., Ph.D., Princeton University, 2008, 236; 3324314
Abstract (Summary)

Inspired by the advanced capabilities of fish and other underwater swimmers, we seek a greater understanding of fish-like propulsion and control.

Our study begins by modeling two extremes of fish-like locomotion: a potential flow model, which ignores the effects of viscosity, and a Stokes flow model in which inertial forces are negligible relative to viscous forces. We emphasize the importance of the local form of a mathematical object called the connection, which appears in the equations of motion and relates internal shape changes to body velocities. We demonstrate how the process of designing large-amplitude gaits for systems characterized by Abelian connections can be facilitated by visualizing the curvature of the connection over the shape space. These results are partially extended for a class of non-Abelian connections where the group is the semidirect product of an Abelian group and a vector space.

A third model accounts for the effects of both inertia and viscosity. Although still in potential flow, the effects of viscosity are partially modeled through the shedding of vorticity from sharp trailing edges. Our focus is on the interaction of the swimmer with its own vortex wake. We take a heuristic approach and perform a series of numerical experiments to identify a strategy for producing near-optimal thrust-producing gaits. We implement a phase-locked loop controller to achieve the control objective and demonstrate its effectiveness at generating high thrust-producing gaits.

Indexing (document details)
Advisor: Rowley, C.W.
School: Princeton University
School Location: United States -- New Jersey
Source: DAI-B 69/08, Dissertation Abstracts International
Subjects: Mechanical engineering
Keywords: Fish-like locomotion, Locomotion, Swimming, Viscosity
Publication Number: 3324314
ISBN: 978-0-549-76578-3
Copyright © 2020 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy