McDonald, Gillian2021-06-292021-06-292021-04https://hdl.handle.net/11299/220587University of Minnesota Ph.D. dissertation. April 2021. Major: Mechanical Engineering. Advisors: Timothy Kowalewski, James Van de Ven. 1 computer file (PDF); xi, 247 pages.The advancement of soft robotics and the inherent ability of soft robots to interact safely with delicate environments has created a host of opportunities for innovation in a wide range of disciplines, from pipe inspection to muscle rehabilitation. The compliance of soft robots has potential to be particularly valuable in medicine where robots are becoming increasingly present in clinical settings. However, developing medically relevant soft robots at millimeter size scales and accurately predicting how they will interact with their environments is a challenge that has yet to be overcome. This work investigates how soft robot behavior is affected as the size of the robot is reduced using both novel experimental prototypes and efficient modeling methods. One core contribution of this work is a soft robot design that is capable of locomoting through tube-like environments, such as arteries or the intestinal tract. The overall robot is modeled using components of fluid power systems to enable the robot, comprised of multiple individual sections referred to as actuators, to move in sequence using just one control input. The experimental prototype was developed using custom fabrication methods to allow new designs and material combinations to be efficiently explored. A second key contribution is an interaction model that predicts the actuator shapes and forces that develop as a result of soft robots interacting with environmental constraints. The model utilizes a combination of Hencky bar-chain and linear complementarity methods to create a simple, efficient contact model that does not require computationally expensive finite element modeling and estimates shapes with errors of 1.06\% and forces on the order of grams-force. A third major contribution is the determination of the factors controlling the underlying dependence of soft actuator bending stiffness on actuation pressure, which ultimately plays a role in how robots behave. The presented work introduces the free-fold test to soft robotics to empirically estimate the bending stiffness of soft actuators, whether composite or homogeneous. This work concludes by tying together the proposed models and corresponding empirical studies to provide a design tool and overall understanding of how soft robot behavior is affected by size reduction. The work identifies fundamental challenges and performance limitations of producing increasingly smaller soft robots at the millimeter scale and provides a foundation on which to build in order to advance the viability of soft robots in medicine.enHencky bar-chainLinear Complementarity Problem (LCP)medical roboticssmall-scalesoft robotsThesis or Dissertation