Development of Hybrid Air-Water Rotor Transition Thrust Prediction and Control

Abstract

Hybrid vehicles are able to function in some combination of aerial, underwater, and terrestrial environments, which greatly expands the scope of missions a vehicle can perform. Hybrid aerial-water (HAW) vehicles are a promising subcategory that are designed to operate in two vastly different fluid mediums. Multirotor HAW vehicles configurations have advantages in maneuverability, but pose a challenge in the water entry or water exit transitions. The interaction of a powered rotor with the air-water interface and its performance in a mixed air-water medium are poorly understood. Previous HAW vehicle strategies avoid a powered rotor with additional propulsion and buoyancy systems, constraining the design space. A custom test stand was constructed to better understand rotor performance during the air-water transition. By recording powered rotor performance during controlled water entries and exits in a large tank, several novel observations were made. Previously unrecorded phenomenon such as the gradual height and RPM dependent transition and the underwater ceiling effect are determined. These observations inform the development of the Transition Index TI, a novel metric that indicates the transition state of the rotor, without the need for specialized sensors or computationally intensive modeling. TI is applied to experimental data to make further observations, and is also used in a novel thrust prediction formulation. The first known low-order prediction of thrust through the transition is validated against experimental data, and allows for the development of a TI based controller. A preliminary controller implementation shows promising results in maintaining constant thrust through the air-water transition. Finally, a HAW vehicle to apply this controller is built. Careful consideration to the waterproofing and motor choice is shown and preliminary flight tests are demonstrated. Future expansion on the application of the novel TI and thrust prediction has great potential to advance the capabilities of hybrid aerial-water vehicles.