Submersibles

Recent successes in Autonomous Underwater Vehicle (AUV) technology have generated demand for broader, more general use of these vehicles, along with demand for longer, more complicated missions. However, AUVs are becoming more complex, and hence more difficult to program, test and maintain. A discrete event system would provide conditional execution for mission management, failure detection, and resource allocation techniques. The goal of this thesis is to provide a convenient formalism for discrete event systems that reduces the apparent complexity of the system while maintaining its robust capabilities. The chosen formalism allows the convenient representation of hierarchies of concurrent hierarchical state machines. The formal semantics of the model are discussed, along with complete data structures and algorithms. As an example, a complete design of the navigator state machine is provided.

Designing a propeller for optimum performance on a human powered underwater vehicle presents a significant engineering challenge. The propeller must be highly efficient to utilize the inherently low power output of a human. Also, the propeller must be correctly matched to the maximum sustainable torque of the propulsor. This thesis experimentally investigates a minimum induced loss propeller design program and its application to a human powered underwater vehicle. The design program is based on the vortex theory of propellers. The work includes experimental measurements of the velocity and rotational rate of three propellers designed with the minimum induced loss propeller design program. This positively verifies the output of the design algorithm. Also, the research, through the use of an underwater ergometer, determines the maximum power and torque sustainable by a human pedaling underwater. Final results of the research show that the design algorithm overestimates the blade section angles by 25% because the design program neglects the influence of the wake of the vehicle.