In the Spring of 2020, I worked in a team of seventeen on RAPID (Figure 1) – Rapid Autonomous Pods for In-Situ Data. Autonomous underwater vehicles are a critical tool for sampling oceanographic data, however, current ocean exploration technologies are prohibitively expensive and slow. RAPID is an Autonomous Underwater Vehicle (AUV) platform designed to address these challenges by utilizing autonomy and simple mechanical design to efficiently and adaptively collect volumetric samples of ocean data.
RAPID utilizes a piston-based buoyancy engine to vertically descend and ascend down to 1000m below sea level. One AUV is equipped with two actively controlled fins in order to move laterally as well as vertically with each dive. Using these fins and an on-board processing suite of satellite transceivers and multiple sensors, multiple RAPID AUVs can swarm to areas of interest and characterize a volumetric region of the ocean. While each RAPID AUV can measure water temperature, depth, and salinity, RAPID is designed to handle a multitude of possible data collection needs with its modular sensor interface.
While working on RAPID, we divided into four subteams: buoyancy, structures, electrical engineering and autonomy, and power. I was part of the structures subteam which had a broad range of responsibilities, from the pressure vessel that houses the guts of RAPID to the fins and hydrodynamic properties of the vehicle.
Figure 2: The structures subsystem includes the fins, the hull, the nose and tail end caps, the pressure release valve, and the linear rail.
While determining the selection and design of the structures subsystem components (Figure 2), we considered the following requirements:
Operating Depth of 1000m
Lateral Mobility during ascent and descent
Streamlined Dynamics for stability and speed
Total Vehicle Weight of less than 46kg
Manufacturable from stock materials
Robust mechanical design
Safe (able to release built up pressure)
Though we all discussed each of these components, I primarily worked on the fin assemblies and the nose fairing within the structures subteam.
Fin Assembly
Figure 3: One side of the fin assembly attached to the hull.
The primary component of horizontal pod movement stems from the ability to translate up and down at a controllable angle. Additionally, RAPID must be able to swivel clockwise and counterclockwise to adjust the direction of horizontal translation. In order to achieve this horizontal motion, fins were added to RAPID.
After considering RAPID’s desired direction and velocity as well as the conditions under which it would operate, a two-fin design was chosen to allow the pod to swivel clockwise and counterclockwise by pitching the two fins in opposing directions. Once we simulated variations of our design, we arrived at a magnetic torque coupler, controlled by a servo, to move each fin using a set of alternating, circularly arranged magnets (Figure 3). To see more details about this as well as the several designs we considered and more details on our final design, please see sections 5.1.1-5.1.2 of the white paper linked at the top of this page. A visual walkthrough of some of the custom designed components is also shown below (Figures 4-11).
Figure 4: A cross-section of one of the fin assemblies attached to the hull. The assemblies are made of a combination of off the shelf parts and custom designed parts that we outsourced to be manufactured. The numbered parts are those we designed and are described in greater detail in section 5.1.2 of the white paper and in Fig. 5-11.
Figure 5: The back and front respectively of the Outer Hull Mount that has a lip to easily line up the part onto the hull for the weld connection. The Outer Hull Mount is the only component within the fin assembly that connects with the hull. (#1 in Fig. 4)
Figure 6: The front and back respectively of the Fin Housing component, which connects the fin to the Outer Hull Mount (Fig. 5) via screws. (#11 in Fig. 4)
Figure 7: The mounting component for the servo, made from removing as little material from an aluminum block as possible, connects the servo to the fin assembly. The connection aspect of this part lines up with the rectangular pattern of holes on the back of the Outer Hull Mount (Fig. 5). (#4 in Figure 4)
Figure 8: This image shows the Inner Mag Coupler (an off the shelf part) including the holes drilled to allow for the spring pin to be included to secure the rod (Fig. 9) – which holds the two bodies that make up the fin. The rectangle on the image depicts which holes in the image are meant for the spring pin.(#12 in Figure 4)
Figure 9: The 1/4" Diameter Rod would be ordered as a stock part then cut to length and with 3 holes drilled into it to allow for connection with the Fin Bodies (Fig. 11) and Inner Mag Coupler (Fig. 8). To the left and right are magnified views of each end of the full rod (middle), with the hole for connecting the Inner Mag Coupler (Fig. 8) highlighted with a red circle and the holes for connecting the Fin Body highlighted with red rectangles.
Figure 10: An isotropic view of a Fin Endcap, which provides structural support for the Fin Bodies (Fig. 11). (#15 in Fig. 4)
Figure 11: The Fin Bodies with a hole for the rod (Fig. 9) to connect the fin to the Fin Housing (Fig. 6) and an insert for the Fin Endcap (Fig. 10). (#16 and #17 in Fig. 4)
Figure 12: Drag coefficient based on shape.
Nose Fairing
The primary consideration in designing the nose fairing was the compromise between drag coefficient and fair feasibility since one of the basic ideas of our strategy is to have the ability of rapid sink and rise (Figure 12). As you can see from the chart, the shape with the lowest drag while still being feasible given the shape of the rest of the pod was the half-sphere.
We decided to use the hemisphere design (Figure 13) for our nose fairing. We decided to manufacture this part out of ABS by 3D printing to get a cheap and fast prototype for our specific design.
Figure 13: Nose fairing for the prototype with space to insert electronics.
Prototype Construction
With our design completed remotely after leaving campus due to COVID-19, we worked with manufacturers to get parts 3D printed and machined and then shipped to campus. We have worked on assembly plans remotely working with 3D printed parts shipped to our homes and have all parts for the final prototype waiting for us on campus or on their way to campus. Once we are able, our team plans to meet back on campus for the full assembly and testing of RAPID.