Achieving Mechanical Versatility in Robots and Structures Through Laminar Jamming

Abstract

There are two major physical paradigms in robotics---soft robots and traditional rigid robots. Soft robots are made of compliant materials and have excellent adaptivity, robustness, and safety, whereas traditional rigid robots are made of stiff materials and have outstanding resolution, precision, speed, and load capacity. Building a single system that can selectively behave like either a soft or traditional rigid robot has been a grand challenge of the field. In this thesis, we rigorously investigate a promising mechanism that can help unite these paradigms. The mechanism is laminar jamming, in which a stack of flexible layers can exhibit dramatic changes in its mechanical properties (e.g., stiffness) when a pressure gradient is applied. When laminar jamming structures are integrated into soft robots, they can begin to exhibit the form, and consequently the function of traditional rigid systems. The mechanism was first reported in the robotics literature in 2000. However, a surprising number of fundamental questions have been left unexplored: For instance, what is the physical mechanism behind the phenomenon? How can the deformation of laminar jamming structures be predicted for both small and large loads, as well as during dynamic motions? Beyond stiffness, what other mechanical properties can the phenomenon change? In this thesis, we demonstrate how the laminar jamming phenomenon works; how laminar jamming structures can transform the stiffness, damping, kinematics, and dynamic response of robotic systems; how designers can relate design parameters to performance metrics; and how the performance of laminar jamming structures can be pushed well past the state-of-the-art. In doing so, we aim to foster robots and structures that cannot simply be classified as "soft" or "rigid," but instead exhibit highly versatile mechanical behavior.