AKKA – HALE UAV

AKKA is a surveillance system that consists of a HALE UAV. The UAV becomes airborne by a large cargo plane, which delivers it to its target area. Upon arrival, the UAV deploys its wings and ascends to the desired height  and conducts an 80 hour surveillance mission. Additionally, the drone should be capable of recognizing vehicles from a distance of 250 kilometers. At the end of the mission, the UAV autonomously returns to its operations base and lands by itself.

This report follows the complete process that was performed by our team, beginning with system requirements, followed by theoretical analysis of the aerodynamics, flight control system, flight mechanics, propulsion, structure, communication systems, surveillance and radar systems and concluding with the assembly and wind tunnels tests on our models.

First we chose the various systems of the UAV according to the tactical requirements. We selected SAR Radar by SELEX, Day vision -Star Safire 380-H by Flir, where the Star Safire acts as a complementary system to the SAR radar. The optical system is redundant and was chosen for its low weight and power requirements. The communication system consists of a Planner Phase Antenna, which broadcasts to a GEO – synchronous satellite. The antenna is 30 cm long and requires 200 watts of direct current. The cargo plane’s cargo bay was a constraint on the UAV’s maximum size. The chosen cargo plane is An-124 by the Antonov Company. It was chosen for its flight envelope suited for our goals, market availability and large cargo bay. It has a length of 43.7 m, width of 6.4 m, and height of 4.4 m.

The main and unique requirement for flight is the long endurance mission of 80 hours, which demands an economical weight and fuel design, and requires the flight method to generate minimum drag. This time requirement was not met and a more realistic maximum mission duration was re-determined to be 60 hours. The UAV’s size, weight, and flight conditions had to be taken into consideration when choosing a propulsion method (engine), and therefore a jet engine was chosen. The commercial engine we chose is AE-3007 by Rolls-Royce – SFC: 36 kg/(kN*hr) , with a maximal thrust of 42-26.1 KN

This report also presents the aerodynamic and propulsion models and the analysis of wind tunnel testing that led to the final flight conditions: Height: 60,000 feet, Velocity: 180-220 m/s.

The final flight control system was chosen from several options: Standard Proportional control, and two “revolutionary” methods: “Bang Bang” and PWM for statically unstable aircraft. The various methods were compared both by their degree of over-shoot following a steering command, and by the ability of their flight control system to maintain the desired height. The modeling and analysis was performed using Simulink. The proportional method was chosen because it allows for safe landing, straight and level flight and “smoother” ascending and descending compared to the other methods. The other methods are of greater academic challenge, but are unsuitable for this project.

The many specific flight requirements mentioned above disqualified the use of any conventional plane designs suited for sub-sonic flights. Therefore, our group proposed three distinct configurations.

The various models parameters are presented in the following table:

Model is 1:100

Our group designed and created Models B and C, which were then studied in the sub-sonic wind tunnel at the Aerospace Engineering Department at the Technion.

For each model, the lift to drag ratio was studied in reliance to the flight velocity and angle of attack. The models’ dimensions and maximum flight velocity were derived from the tunnel’s limitations. Despite structural additions to our models that were added to achieve a more reasonable Reynolds number, their flight characteristics were measured in a different Reynolds number than the one that characterizes the desired flight conditions. Also, the models’ weight was significantly lower than desired, which forced us to use force gauges in order to measure the very low forces on the model. This is one explanation of the high error percentage in the results.
The deviation in the Reynolds numbers and the high error percentage did not enable us to obtain an accurate estimation of the UAVs’ behavior during real life flight, but we were able to conduct a comparison between the three models. The comparison shows that model C is preferable according to all the flight parameters: lift, drag, aerodynamic efficiency, and moment about the aerodynamic center. Therefore, we chose the design of model C for AKKA.