Designing a vertical / horizontal AUV for deep ocean sampling

Document Type

Conference Proceeding

Date of Original Version



Faculty and students at the URI Dept. of Ocean Engineering have conceived of a new class of AUV for deep sea profiling and sampling called a Mini Ocean Elevator (or MiniOE). The MiniOE is intended to operate as both a vertical and a horizontal AUV on the same mission, beginning its mission as a vertical profiler and then converting to a horizontal AUV for sampling prior to homing on an acoustic beacon at the surface. Such a vehicle has tremendous potential for ocean research while presenting unique challenges for stability and control. Horizontal AUVs typically achieve stability against roll and pitch through significant mass below the center of buoyancy along with tail fins. For a vertical vehicle this mass-ballast displacement must be in an orthogonal direction along the longitudinal axis to achieve stability in pitch and yaw, forsaking roll stability. Designing a cigar shaped AUV capable of both horizontal and vertical stability represents an unusual challenge. This paper describes the design work and testing involved in developing a stable vertical AUV that will later be equipped to convert itself to a stable horizontal AUV. This design effort used a balanced combination of physical experimentation and computer simulation to provide detailed information about how different configurations of an autonomous underwater vehicle (AUV) might affect its stability, and thus the controllability, while using conventional hardware and software resources. Passive vertical ocean projectiles like XBTs typically use maximum displacement between mass and buoyancy centers along with rotation and / or tail drag to achieve stability. But for a vertical AUV to guide its descent and maintain sensor orientation means rotation is undesirable, and thus must be stabilized or controlled. Of course vertical and horizontal vehicles need this stability in different directions. As a result converting from one to the other is particularly challenging. Descent rate control and orientation stability can be influenced by using tail drag to balance vehicle weight. For the purposes of designing a stable vertical descent profile for the Mini O.E., a computer model was constructed to test parameters that affect control and stability of the vehicle through hydrodynamics. The model was verified by measurements on a physical test-bed of the vehicle in shallow depths and with varying initial conditions. The test-bed vehicle included a Tattletale microcontroller/logger with rapid data acquisition from 3-axis magnetometer, 3-axis accelerometer, rotation and precision pressure sensors. The vehicle was dropped vertically first into a 5 meter deep tank, then into 20 meter deep water in Narragansett Bay. Experimental results served as "ground truth" for the vehicle's dynamic characteristics, showing the stabilizing tendencies in descent inherent with different configurations. The test-bed observations were reconciled with a custom Matlab simulation model which provided a cost effective method of demonstrating the influence of changing physical parameters. The model was able to show how the vehicle would act with active controls and under a wide variety of configurations and conditions, which would otherwise be costly to physically build and test. ©2007 MTS.

Publication Title

Oceans Conference Record (IEEE)