Capstone Senior Design Project:
AIAA 2001 Design-Build-Fly UAV “Shamu"
A capstone senior design project completes a formal undergraduate engineering program. Thus, much like the architectural capstone, the engineering capstone course proves, finalizes and protects the underlying structure. My aerospace engineering senior design course at Oklahoma State University followed this tradition by being a semester long, high pressure, team oriented, and successful UAV (Unmanned Aerial Vehicle). Starting from nothing, we designed, built, and flew a competition winning UAV in 4 months.
 |
| On runway just before a testflight |
Vehicle, Mission, and Competition
The AIAA Design-Build-Fly (DBF) contest provides a yearly international competitive environment for university aerospace engineering students when given a specified mission profile and physical constraints. For 2001, the DBF contest tasked teams with flying two payloads, low density tennis balls and high density steel, around a given flight pattern with a time limit. Score depended on ball volume, steel weight, flight laps, and written project proposal and test-flight reports. The contest was held at NAS Patuxent River, Maryland in April 2001.
Our final vehicle was a low wing, conventional configuration aircraft with a 10 ft wingspan, 1500 Watt motor, and a 16 pound maximum payload. Takeoff at the 35 pound gross weight required 200 ft. The primary structure consisted of carbon fiber skin with a foam core. The untapered, unswept, polyhedral wing consisted of a carbon fiber skin with a foam core; the spar was built-up of both a carbon fiber shear web and a wood spar cap. Configuration details and photos are available at
http://bit.ly/Shamu2001. A large horizontal was required for short-field takeoff rotation, so we named the aircraft after a famous killer whale, Shamu. With this aircraft, our team won 1st place among 27 teams. An official AIAA write-up is available at
http://bit.ly/Shamu2001-AIAA.
My Role
My role in the Aerodynamics Group was critical for the aircraft's performance and the team's success. My primary responsibility consisted of aircraft performance and contest score optimization. I converted the DBF contest's written “request for proposal” (i.e. rules) into a computational aircraft flight simulation in MathCad for tracking the aircraft's energy budget throughout the missions. A global optimization routine allowed for optimizing and selecting the plane, motor, and batteries simultaneously. Coordination of detailed performance data and interactions between the Structures and Propulsion groups became increasingly critical as aerodynamics performance, structural weight, and propulsion models developed from conceptual to final design tolerances. As the design process wound down, my responsibilities shifted to construction and then to flight testing. I constructed the wing spars and the carbon-fiber wing skins; the low Re airfoil necessitated smooth skins. Flight test data allowed for performance and simulation tuning, which allowed further score optimization.
Flight test clips of the prototype and final aircraft are at
http://bit.ly/ShamuFlightTest.
Challenges
Challenges always exist in a complex project.
Weather uncertainties were amplified by the 200 foot takeoff constraint. Headwind determined the gross weight. This challenge forced a statistical optimization based on the plane and historical weather. Better yet, solving the challenge allowed real-time performance envelope and mission loading data for the flight crew.
During the prototype's flight tests, the cruise speed was significantly less than predicted. Working with experimental propeller, and motor curves, we identified that the motor winding was customized and differed from the motor's specification. A correctly wound motor fixed the cruise speed problem.
Airfoil selection was strongly constrained to low Re and high CL. Most airfoils meeting those criteria had strong maximum surface curvature with delicate trailing edges. After browsing experimental wind tunnel data and airfoil profiles, I selected a set of candidates, parameterized their CL, CD, and polar curves, and then simulated the competition score. The Eppler 423 was selected.
I enjoyed my senior design project and the challenges it presented.
