A plethora of studies examining the aerodynamic design of UAVs already exist. For example, Panagiotou et al.
conducted a study on Medium-Altitude-Long-Endurance (MALE) UAVs, emphasizing the conceptual and preliminary
design phases, while also running computer simulations as a part of their optimization technique [2]. While their
study was not specialized for Hunter UAVs (which are more commonly used for close-range operations compared
to the MALE UAVs which are more commonly used for longer distance operations), the methodology used in their
optimization study was rigorous and can be applied to the study of Hunter UAVs.
In Panagiotou et al. study, the first stage was to create a conceptual design of the MALE UAV [2]. Eight engineers
were divided into four groups and they all came up with optimal UAV designs with different fuselages, tails, and wings
using CAD software. Many limitations were taken into account, such as the total weight of the aircraft which included
the payload weight, fuel weight, and empty weight, temperature, maximum speed, and much more. These limitations
reflected the terrain They also used the Breguet equation to estimate the required fuel for the flight of the UAV. With
each of the four optimal designs that were created, the thrust-to-weight ratio (T/W) and the wing loading (W/S) were
calculated. The findings were that the team 2 had the best overall L/D ratio, with the design of a boom-mounted tail
configuration featuring raked-type winglets and an airfoil-shaped fuselage.
Another study performed by Kontogiannis and Ekaterianris focused on small-sized UAVs, again utilizing a general
design that could possibly be implemented for different UAVs used for various tasks[5]. Again, they optimized wing
geometry after picking an optimal existing fuselage, that being the PARSONS-F2-49. Optimized wing features included:
wing twist, airfoil angle of attack, and type of winglet design. The study presented here, in contrast, focuses solely on
the Hunter UAV, which is used for military operations.
In the proceeding sections, the procedural stages and methodology are thoroughly discussed, the results of the simulation
runs are discussed and the most aerodynamic UAV is chosen, and the wing and airfoil are optimized on the best UAV
model, and finally the conclusion and final UAV design concept are presented.
2 Methodology
In this experiment, the GW-19 airfoil and its angle of attack (AoA) is incrementally changed to find the optimal AoA.
In aerodynamics, it is generally agreed upon that the best metric for aerodynamic efficiency is the lift-to-drag (L/D)
ratio [6]. This study, therefore, seeks to maximize the L/D ratio in order to determine the most aerodynamic AoA.
The program of choice was Flowsquare 4.0, which solves high-order differential equations to simulate two-dimensional,
incompressible flows [7]. The chosen input airfoil was the GW-19 airfoil as it is the root airfoil of the MQ-9 Reaper.
Every angle was created in MATLAB and made into an input boundary condition file for Flowsquare to read and use
in the simulation as shown in Figures 1-3 [8]. For the Flowsquare parameter file, which was labeled grid.txt, certain
conditions were used to simulate the real environment that the MQ-9 Reaper UAV flies in, assuming clear weather.
UAVs such as the MQ-9 Reaper operate at altitudes below 9,100 m and have a broad variety of altitudes to cruise in
based on the mission [9]. The altitude of 5,000 m above sea level was chosen for simplicity. The average pressure at an
altitude of around 5,000 m is 16.22 in Hg (mercury), which is approximately 54,930 Pa of pressure [10]. At this altitude,
air has a density of 0.746 kg/m-3 [11] and, according to the US Standard Atmosphere 1976 model, has a dynamic
viscosity of around 3.4 * 10-7 lb*s/ft2, which in standard units is 1.63 * 10-5 kg/(m*s). The last parameter needed
for the simulation is the relative velocity between the aircraft and the air, which will stay relatively constant for the
majority of the flight. Because drag and lift are associated with the velocity squared, the maximum speed is chosen to
clearly show the difference between the different AoAs in their respective L/D ratios. The maximum speed of the MQ-9
Reaper is approximately 240 knots, or 123.467 m/s [12].
Other Flowsquare parameters were chosen purely based on simulation time constraints and preference, and every
simulation run kept the same parameters. A high-order scheme with a 4th order difference and 3rd order time integral
was used for the numerical scheme in the simulation. 5000 was the last time step used and each time step lasted around
6.1138 x 10-5 seconds. The reason for this choice was because when preparing the simulation by running pre-trials,
after around 3000 time steps, there seemed to be no changes at all to the x and y velocity vector fields. A simulation
was run for each AoA, 0 through 20 degrees in increments of 1 degree. Examples of the boundary condition with the
airfoil at various angles are provided in Figures 3-5. It is generally advised to not go above around 17 degrees before
airflow above the upper surface becomes detached, a phenomenon known as flow separation [13].
The cruise velocity can be helpful to calculate the aircraft range to find the fuel efficiency of the aircraft. Using the
lift-drag analysis provided by Flowsquare, the instantaneous air density and velocities were read from the collected
simulation data, and then converted to the ratios of drag and lift forces to drag (CD) and lift coefficients (CL), respectively.
Next, the drag and lift coefficients were approximated using NASA FoilSim JS, whose values are experimentally derived
[14]. The parameters used for FoilSim were a 2.2% maximum camber at 44.2% chord and a maximum thickness of
2