Design, Operation, and Control of a Robotic Manta Ray

Abstract

The manta ray is able to maneuver very effectively in tight spaces with minimal noise output. In order to reap the benefits of the ray’s approach to locomotion, Princeton University has sponsored the design of numerous batoid-inspired underwater vehicles. The potential applications for clandestine surveillance, cave mapping, and tracking different animals without disturbing the local ecosystem are just some of the possible advantages of an oscillation-driven underwater vehicle. There is a strong incentive to support ray-inspired designs given their comparative advantages over the industrystandard, Myring hull. Myring hull designs have limited maneuverability, consume energy typical of motorized propulsion, and produce an easily detectable sound profile due to the heavy turbulence created by the propeller. Inheriting Mantabot 3.0, a significant portion of its modification was devoted to the preparation of the autonomous underwater vehicle (AUV) for consistent and prolonged underwater operation. The most significant modifications to the design of the modular AUV were the implementation of a multi-celled tensegrity actuation system, the switch to softer silicone fins, the switch to a submersible servo for the elevator, the design and manufacture of aluminum heat sinks, and experimentation with different fin sections. This project validated the tensegrity structure as a viable and effective means for achieving high range-of-motion, oscillatory fin flapping. Through tens of hours of swim testing, the Mantabot was made capable of executing a number of autonomous, open-loop swimming tasks in calm water conditions. These include the ability to perform a U-turn, obstacle avoidance turn, loiter, and dive and surface routine. However, external disturbances (e.g. waves and currents) significantly altered the desired swim trajectory. This problem can be solved with more powerful actuating motors and the incorporation of feedback control. Work was also begun on creating an aerodynamic model of the Mantabot’s dynamics for flight and control simulations. Such a model will complement physical testing of the Mantabot in the water, and inform future design decisions and parameters. Using a 3D panel method CFD simulator of unsteady swimmers, data was gathered and post-processed to compute the non-dimensional, aerodynamic coefficients of the Mantabot. The current model is limited to longitudinal motion and shows behavior confirmed in reality. Future progress on this dynamic model will involve gathering CFD data on lateral-directional coefficients to incorporate the effects of sideslip, roll, and yaw. With the establishment of a modeling framework for simulation and the achievement of open-loop maneuverability and control, the Mantabot project is on a promising path to realizing autonomous, closed-loop swimming functionality.