Task 4

Design, Construction, and Flight

Shane Hills, Matthew Berk, Nur Harell, Jae Hwan Choi

Note: all figures and photos can be found at the following link for this task:

Flickr Album with Figures

Design Concept

This section focuses on the design, manufacturing, and testing process involved in the creation of an initial UAV concept. The main challenges associated with aircraft design stems from the fundamentally multidisciplinary nature of the problem: it is necessary to integrate a wide range of fields when designing an effective drone. Aerodynamic, control, structural, mission, and manufacturing considerations must be made in order to ensure flight efficiency and mission effectiveness.

From the outset the goal was to design an aircraft that was conceptually atypical, but aerodynamically stable and efficient. The typical aircraft employs a fuselage that provides low drag but no lift, contributing to only a reduction in overall aerodynamic efficiency. Inspiration was taken from the competitive RC sailplane community, where a common design concept pioneered by Mark Drela involves a high wing configuration with suspended fuselage pod. As mentioned however, this type of fuselage only contributes to drag. To this end, a high wing configuration with a lifting fuselage design was conceptualized in order to make use of the fuselage as both a vessel for containing the electronics, and a component that contributes lift to the aircraft. Thus, by integrating the electronics and other components into a spar and rib design it is possible to use the fuselage as a wing that creates lift for the aircraft. This type of design has an advantage over a blended wing-type design in that if the aircraft is designed modularly such that each main component can be easily swapped in or out, it is easy to tweak the CG location with respect to the neutral point. A mock-up of the design is shown in Figure 1.

Design Considerations

Airfoil Choices

Since the intent was to create a fuselage that has lifting capabilities, it was necessary to choose appropriate airfoils for both the main wings, tail, and fuselage. The main considerations involved with choosing airfoils for this application were the lift and drag characteristics, thickness, and Reynolds number suitability. For obvious reasons airfoils that have high lift and low drag characteristics were chosen, but in particular it was necessary to optimize these parameters based on the appropriate flight conditions. It was determined by the mission planning team that the optimal air speed was approximately 12 m/s (determined by a simple optimizing calculation involving the picture-taking frequency), and given the kinematic viscosity and density of air at sea level for early morning temperatures, the appropriate Reynolds number for this application was in the range 150 000 to 200 000. As a result, it was necessary to choose airfoils that were aerodynamically efficient in low Reynolds number scenarios. A variety of airfoils were considered for the main wing, including the E214, E387, AG35, AG40, and SD7037. These airfoils each have different thicknesses, with the AG40 having the minimum thickness at 8%, and the E214 having the maximum at 11.1%.

XFLR was used as a preliminary tool to model the full span aerodynamic characteristics of these airfoils from the perspective of lift and drag coefficients and their relation at varying angles of attack. Specifically, XFLR was used as a quick and dirty comparison tool to choose the appropriate airfoil for this application. Figure 2 shows plots of C_L vs. C_D, C_L vs. AoA, C_L/C_D vs. speed, and C_L/C_D vs. C_L. Considering the thin nature of AG35 (weight savings), broad C_L/C_D vs. speed in the area of our desired flight speed, and superior lift coefficient, it was decided to use this airfoil for the main wing.

Having chosen an airfoil using XFLR, AVL was used to accurately model the aerodynamic characteristics. From Task 2 it was determined that a more accurate lift coefficient of 0.33 was appropriate, which gave the ability to size the wing accordingly.

L = Fg

C_L*rho*v²*S/2 = mg

Given a conservative weight estimate of approximately 750 grams, the planform area of the main wing was determined to be approximately 0.2 m². The main wings were given a dihedral angle of 5 degrees for spiral stability and a slight taper or more efficient lift distribution.

The airfoil shape for the fuselage involved very different design objectives. As before it was important to achieve good aerodynamic properties, but in this case a large thickness was required to provide sufficient room to house all the electronic components necessary for flight, sensing, and data capture. Furthermore, since the intent was to construct the fuselage using a somewhat standard spar and rib design, ease of construction was also a driving factor. The Clark Y airfoil fit all of the prescribed design criteria, particularly considering that it has a flat bottom edge that allows for easy skin attachment and electronic housing. The fuselage airfoil was sized by simply arranging the internal components in a compact manner, taking rough measurements of area and height, and using the Clark Y thickness (11.7%) to size the ribs appropriately.

Finally, the tail considerations had similarly different design objectives. Since the main wings were sized to provide sufficient lift for the aircraft, in the spirit of quickly and effectively developing a prototype aircraft, the tail was not given a large priority from an aerodynamic standpoint, but rather a stability perspective. As a result, the tail was designed using a flat plate in both the horizontal and vertical directions. From Figure 1 it can be seen that the initial design was to use a thin carbon boom to connect the fore components of the aircraft to the aft. Given the length of the boom as a moment arm, the tail was sized appropriately using AVL to provide stability and damping.

Component Materials

An important consideration in designing the aircraft was one of construction materials. The main objective here was to utilize materials that provide optimal strength to weight ratios. Once again however the importance of quickly developing and manufacturing the aircraft prototype played a large role in the selection of materials. As a result a combination of balsa wood, basswood, carbon, and foam were used to construct the aircraft (with appropriate fasteners).

Given the time constraint that was imposed, foam provided the ideal material to construct the main wings from, particularly considering the ease with which it is cut and high potential for reinforcement. It was decided to use a combination of basswood, balsa, and plastic skin for the fuselage in order to provide space for the electronics. The boom was specially ordered in carbon fibre, and the tail was made from solid balsa once again for it's weight and ease of construction.

Design Modelling

The vast majority of the design work was done using SolidWorks. By modelling all of the component geometries and weights in CAD, it was possible to easily arrange and modify the positions of all of the electronic components in the fuselage to achieve good weight distribution from a CG and moment perspective. Figures 3 and 4 show various components and assemblies of the aircraft in SolidWorks.

Manufacturing

Pictures of the manufacturing process can be seen in Figures 5 and 6.

Main Wing

The main wing was perhaps the easiest component to manufacture considering it was done using the foam cutter. However, a slight complication arose when it was determined that the initial desired aspect ratio of 10 was not possible without making the wing in multiple sections given the maximum allowable span dictated by the cutter and the desired wing area. As a result, rather than make the wing in multiple sections (reducing structural integrity) the aspect ratio was reduced to approximately 7.3 to keep the required wing area and reduce the span to the point where each wing could be made in one cut. Obviously this comes with an aerodynamic efficiency loss, but it was determined that the ease of manufacturing and quicker timeline afforded by this approach was a higher priority for the first aircraft prototype. To increase bending rigidity carbon rods were inserted span-wise in the wings.

Fuselage

The fuselage was constructed using 4 ribs cut from balsa and a basswood plate for attaching the electronics. While the use of basswood increases the overall weight of the aircraft, the strength afforded by the stiffer wood was deemed a priority, especially considering that it provides the main structural member of the fuselage and the electronics are directly attached to it. Once the 4 ribs were glued to the basswood plate the electronics were arranged appropriately and affixed using velcro. With the electronics in the correct positions, supporting stiffeners were added to improve the span-wise strength in the fuselage. Finally, the majority of the fuselage was covered using book laminating film to create the airfoil. The top of the fuselage where the electronics could be accessed was covered using a simple piece of paper that taped down, allowing easy accessibility.

Tail

The tail construction was incredibly simple for the reasons mentioned above. Two pieces of balsa were simply glued together perpendicularly to create the tail, and small supports were glued in place to ensure rigidity. One of the main challenges with the tail was attaching it to the boom. An important lesson learned here was the need to consider the grain direction in the balsa. Originally the tail was attached to the boom using small balsa fittings, but unfortunately the grain direction was not considered and ended up being parallel to the horizontal tail. Upon initial glide tests the tail attachment immediately sheared along the grain direction. This problem was fixed by redirecting the grain.

Construction

The main difficulty with manufacturing at this stage of development was a lack of planning in terms of how each of the components were going to fit together properly, as hinted at by the difficulty in attaching the tail. The front of the boom was attached between the 2 centre ribs of the fuselage using glue, which provided sufficient adhering strength. The main wings were connected to the fuselage using a pylon structure made from balsa (keeping in mind grain direction) that attached directly to the boom. This ensured good rigidity and robustness. Finally, to ensure that the wings had maximum bending stiffness, vertical carbon rods were attached to the outside edges of the fuselage and inserted into the wings. The final constructed weight of the aircraft was 601 grams, including all necessary components but not water weight required for the mission.

Flight Testing

The maiden flight of Spaero 1 occurred on April 28th, 2015 in the evening, and was highly successful. The first flight lasted for approximately 8 minutes and ran without much threat of failure. A dramatized short video of the flight can be seen here:

Spaero 1 Maiden Flight

The aircraft was found to be reasonably stable and controllable by the RC pilot, indicating a good initial prototype design. The pilot reported some slight pitch instability, which was counteracted by adding 27 grams of weight to the nose. The second test flight of Spaero 1 included this additional 27 grams, bringing the total to 628 grams, and was reported to improve stability greatly and lasted for approximately 6 minutes. Pictures of flight testing can be found in Figures 7, 8, and 9.

Both of the initial flights were sufficient to collect valuable data from the prototype aircraft in terms of some basic flight parameters. Using data from these 2 flights, L/D was found experimentally found to be 6.32, and C_Lmax was found to be 1.315. Using the heavier second flight data, which is perhaps the more conservative approach, the power consumption was between 25 and 35 watts at cruising speeds between 8 and 12 m/s. The figures below show altitude and distance traveled for gliding conditions (L/D), stall tests (C_Lmax), and power consumption as functions of time.