Conceptual Design of a Model Solar-Powered Unmanned Aerial Vehicle

50th AIAA Aerospace Sciences Meeting including the New Horizons Forum and Aerospace Exposition 09-12 January 2012, Nashville, Tennessee AIAA 2012-0134 Conceptual Design of a Model Solar-Powered Unmanned Aerial Vehicle Christopher J. Hartney 1 San Jose State University, San Jose, California, 95192 In recent news, there has been more conversation pertaining to solar-powered aircraft due to the flight of the Solar Impulse on July 8, 2010. Because of this, more technology and studies are needed to fully understand how solar-powered aircraft operate. Over the last fifty years, there have been more studies on solar-powered aircraft because of the negative environmental impact that commercial aircraft have. This paper will study a conceptual design of a model solar-powered unmanned aerial vehicle (UAV) that will have a wingspan no more than 7 meters and have a total mass of 10 kg, 2.27 kg of which will be devoted to payload. The UAV will be used to study wildfires specifically in California and will have the necessary equipment needed for such a mission, including a Global Positioning System (GPS) and cameras. AR A sc α TD b c c HT c VT C D C L d D p e η bec η ctrl η grb η mot η plr η sc η wthr F L g I p L HT L VT m n n p P av P bl P electot P lev Nomenclature = aspect ratio = solar cell area = tail dihedral angle = wingspan = chord length = horizontal tail volume coefficient = vertical tail volume coefficient = wing drag coefficient = wing lift coefficient = propeller diameter = propeller diameter = Oswald efficiency factor = BEC efficiency = motor controller efficiency = gearbox efficiency = motor efficiency = propeller efficiency = solar cell efficiency = weather efficiency = fuselage length = acceleration due to gravity = power index variable = horizontal tail arm = vertical tail arm = aircraft total mass = engine rotational rate = number of propeller blades = power required for avionics = power loading per blade = total electrical power available = power required for steady level flight 1 Graduate Student, Mechanical and Aerospace Engineering Department, San Jose State University 1 Copyright 2012 by the, Inc. All rights reserved.

P max = maximum power per engine P pld = power required for payload RC = rate of climb RCP = rate of climb parameter ρ = density of air S HT = horizontal tail area S VT = vertical tail area σ = air density ratio T day = total day time T night = total night time V = aircraft velocity (V tip ) helical = helical propeller tip velocity (V tip ) static = static propeller tip velocity I. Introduction HE main motivation behind this paper comes from the environmental challenges that the world has been facing T over the last decade. From global warming to a lack of natural resources, engineers have been challenged to fix these issues over the last fifty years. Commercial aircraft currently use several thousand pounds of fuel to complete their flight, which has a negative impact on the atmosphere because of the carbon emissions that are released. One way to solve this environmental crisis is to eliminate the use of jet fuel and find an alternative source of energy. Many companies are looking into bio-fuels for commercial aircraft, but these will also become scarce over time. This is why the idea of a solar-powered aircraft not only can be a successful one, but it can be a solution to current environmental problems and become the future of aviation. A. Objectives The primary objective for this paper is to present a design of a solar-powered UAV that will be hand-launched and has a total mass of 10 kg. Another objective is to provide an initial selection of a configuration for the solarpowered UAV. A comparative study of similar solar-powered UAVs will be discussed and will be used for a preliminary configuration. There will be a discussion of the propulsion system and a selection of the initial system, as well as how the propulsion system will be integrated into the aircraft. There will be an overall configuration that will be chosen, as well as a wing configuration and empennage configuration. A proposal for an initial configuration will be shown and discussed. Since power is one of the most critical parameters for solar-powered UAVs, we will look at the fundamental power equations required for steady level flight, and discuss their important and how the variables will affect the overall design of the aircraft. A weight and balance analysis, including an estimate and the choice of components, will also be shown. B. Literature Review To understand the history of solar-powered flight, a discussion of solar cell history is required. Humans started using the sun s rays to benefit themselves as early as 7 th Century B.C., where they studied how a magnifying glass can be used to make fires. There were not many more advances in solar technology up until 1767 when Horace de Saussure built the world s first solar collector, later used for cooking. In 1839, Edmund Becquerel was experimenting with electrolytic cells and found that when these cells are exposed to light, they create electricity. This was the first ever photovoltaic cell, which is the basis of all solar cell technology. For the next fifty years, electrolytic cells were shown to power devices, including steam engines and water heaters. Also, selenium was shown to conduct electricity when exposed to the sun during this time. There were some advances in the early 1900s, but the key advancement in solar cell technology was in 1954, when photovoltaic technology was born in the United States. Daryl Chapin, Calvin Fuller, and Gerald Pearson created the first ever photovoltaic (PV) cell capable of using the sun s rays to power everyday electrical equipment 1. Thus, PV cells were built. It will not be for another 15 years until the usage of solar cells will be incorporated into aircraft. On November 4 th 1974, at Camp Irwin, California, the first solar-powered flight took place. The aircraft, named Sunrise I, flew for 20 minutes at an altitude of 100 meters during the flight 2. Designed by Astro Flight Inc., they were later able to get flights upwards of three to four hours, but the aircraft was damaged after a sand storm and was unable to fly. The company later built Sunrise II, which weighed 2.04 kg less than the Sunrise I and was able to generate a power output of 600 watts, which was a 150 watt increase in comparison to the Sunrise I. 2

After the Sunrise I and II flights, there was a greater passion for solar-powered flight and many model aircraft enthusiasts began to use the solar cell technology. One example was Dave Beck from Wisconsin, who built Solar Solitude in 1996. This aircraft set model airplane records for distance flown in a straight line (38.84 km) and altitude (1283 m). During the same time in the 1990 s, Wolfgang Schaeper from Germany built his Solar Excel aircraft, which was also a model aircraft. The Solar Excel broke all the model aircraft records from 1990 to 1999, including duration, distance in a straight line, gain in altitude, speed, distance in a closed circuit, and speed in a closed circuit. This aircraft still holds all FAI F5-SOL model aircraft records. Since then, there have not been many advances in solarpowered model aircraft because there has been an emphasis Figure 1. SoLong Aircraft on manned aerial vehicles (MAVs) and the idea of a high altitude long endurance (HALE) UAV. One example of a long endurance UAV is the SoLong aircraft, shown in Fig. 1. Designed by AC Propulsion, the SoLong aircraft has a wingspan of less than 5 meters and is able to fly continuously using only solar and thermal energy. On April 22 nd 2005, the SoLong aircraft flew for 24 hours and 11 minutes continuously, which was the longest flight for solarpowered UAVs. Two months later, the aircraft was able to fly for over 48 hours, thus opening the door for the possibility of eternal fuel-less flight. An example of a HALE aircraft is the Zephyr, which was first flown in December 2005. A British company named QinetiQ first tested the Zephyr aircraft in New Mexico and was able to fly for six hours and reached a peak altitude of 7,925 meters. The crowning achievement of the Zephyr aircraft was on September 10 th 2007, when the aircraft flew continuously for 54 hours, a new world record for unmanned flight. During the flight, it was able to reach a peak altitude of 17,786 meters. In the future, QinetiQ hopes to have the Zephyr aircraft do a test flight of several months at an altitude over 15,240 meters. The goal of solar-powered manned flight was achieved with the success of the Solar Impulse 3. In Switzerland, a project by Bertrand Piccard was started in 2003 to develop a solar-powered MAV that can circumnavigate the globe using only solar power. The group of engineers in Switzerland went through the aircraft design methodology and was able to build a fully-functional prototype by 2009. In more recent news, the Solar Impulse completed a 26 hour flight on July 8, 2010 using only solar energy, the first of its kind. They soon plan to complete a full circumnavigation flight in 2012. There also have been solar-powered UAVs that were primarily from student creation. One example is the SunSailor UAV, which was done in Israel by undergraduate students 4. The main goal of the SunSailor, shown in Fig. 2, was to break the world record for distance flight at 139 km. They have done several test flights in 2006, but when they tried for the record flight on August 12 th of that year, the aircraft entered a dive and was damaged and could not be flown again. To this day, they are still trying to break the world record. II. Mission Specification Figure 2. SunSailor Aircraft and Team A. Mission Requirements The mission requirements for the model solar-powered UAV are listed below: 1) Total mass of the aircraft is not to exceed 10 kg, 2.27 kg devoted to payload. 2) Requires four crew members. At least two members will be needed to do the hand-launch of the aircraft, one member to be responsible for flight tracking and data analysis, and one member will check the status of the aircraft. 3) Maximum cruise speed: 13.5 m/s (35 knots). 4) Maximum cruise altitude: 2 km. 5) Endurance: at least ten hours at the cruising altitude. 6) Take-off: hand-launched. 3

7) Landing: skid land softly in a designated landing area. B. Mission Profile The mission profile for the solar-powered UAV is shown in Fig. 3. Details of each phase of the Figure 3. Mission Profile mission are as follows: Phase 1: UAV will be hand-launched ideally at 4 A.M. and climb continuously for 4 hours to peak cruise altitude if needed. The UAV is not required to be launched at 4 A.M; however, to achieve maximum endurance during flight, it will need to be launched at this time. Phase 2: At 8 A.M., the UAV will cruise for at least 10 hours at the peak cruise altitude for wildfire analysis. Phase 3: At 6 P.M., the UAV will begin the descent back to ground and land in a designated landing area at 4 A.M. the next day. C. Comparative Study The solar-powered UAV will be compared to other solar-powered UAV that have been built previously. The important design parameters that will be analyzed are the mass, endurance, wingspan, and aspect ratio. The results are shown in Table 1. Weight (in kg) Endurance Wingspan (in m) Aspect Ratio Table 1. Important Design Parameters for Similar Aircraft Sunrise I Sunrise II Solar Solar Excel SoLong Zephyr Sunsailor Solitude 1/2 12.25 10.21 2 0.72 12.6 50 3.6 Maximum flight time was 4 hours Unknown Unknown 11 hours, 34 minutes, 18 seconds 48 hours, 16 minutes 336 hours, 21 minutes (Over 14 days) 9.75 9.75 2.7 2.1 4.75 22.5 4.2 Unknown 11.4 11.4 13.3 12.8 15 11.6 13.15 III. Initial Configuration An initial configuration will be based on similar solar-powered aircraft that are out there, specifically, the Sky- Sailor and SoLong aircraft. The selection of the propulsion system type will be electrical with photovoltaic cells powering the motors and payload. There will be a single engine propeller located at the tip of the fuselage, ahead of the wing. Many model solar-powered aircraft choose a single engine propeller because the aircraft is light. The thrust-to-weight ratio is a critical parameter for the selection of engines, and since smaller solar-powered aircraft are very light, they need a minimal amount of thrust to takeoff. The propeller sizing depends on the fuselage diameter. The larger propeller diameter would then make the propeller the most efficient. However, this adds weight. One of the limitations that Raymer 5 discusses is the propeller tip speed, which is the sum of the rotational speed and the aircraft s forward speed, as shown in Eqs. (1) and (2). ( V ) tip static πnd 60 2 2 ( V ) V V = (1) tip = helical tip + (2) One can also use Roskam 6 to determine the ideal propeller diameter based on the maximum power per engine, number of propeller blades, and the power loading per blade, shown in Eq. (3). 1/ 2 4 Pmax D p = (3) πn ppbl 4

For an initial configuration, a propeller with an 11 inch diameter with a 6 to 8 inch pitch and a rotational speed of 1000 to 2000 rpm will be selected based on similar aircraft. The overall configuration of the aircraft will be a land based conventional configuration from Roskam. The wing configuration will be a high wing installation with no sweep and slight dihedral located at the tips of the wing. The airfoil that will be used is a Selig 1223, Figure 4. Selig 1223 Airfoil with Unit Chord Length shown in Fig. 4. This airfoil has a 12.14% maximum thickness-to-chord ratio at roughly 20% from the leading edge, and is ideal for low Reynolds number flow. Several aerodynamic characteristics of this airfoil were determined using XFLR5, which is software that uses the vortex panel method to simulate the flow over the airfoil. From XFLR5, the maximum lift coefficient is approximately 2.25 at an angle of attack of 12.5 degrees. Also, this airfoil has a low drag coefficient, so the lift-to-drag ratio ranges from 80 to 120 depending on the angle of attack. The empennage configuration will be a v-tail because of weight savings and stability. This aircraft will have no landing gear as was previously discussed. A sample initial configuration is shown in Fig. 5. IV. Mission Weight Estimates The weight estimates of a solar-powered aircraft are unlike other commercial aircraft because of how sensitive it is to weight. Noth 7 provides a method of estimating the different groups of the aircraft, and will be used to estimate the weights for our solar-powered UAV. Before the weight estimate equations are presented, a discussion of how to determine the available power is required. For this, we will assume steady level flight, which means the total Figure 5. Solar-Powered UAV Proposed Configuration lift of the aircraft will equal to the total weight, and the total drag will equal to the thrust. From this, and knowing that the power available is equal to the thrust available multiplied by the velocity, we then get Eq. (4) below. 3 3/ 2 CD 2ARg m Plev = 3/ 2 C ρ b L (4) We also can calculate the daily required electrical energy that the aircraft must use for steady level flight using Eq. (5). 1 1 P electot = Plev + ηctrlη motη grbη plr ηbec ( Pav + Ppld ) (5) Noth provides a MATLAB file that optimizes these equations to come up with a result using initial parameters that the user identifies. We now need to identify all the mass groups so that the total mass will then be used as a variable in Eq. (4) to optimize the design. The details of the mass groups will be omitted in this paper; however, interested readers should consult Noth s text for more information. Using the equations, along with the MATLAB code, we then used the weight restriction of 10 kg and the payload weight restriction of 2.268 kg to create what is shown in Fig. 6. Table 2 summarizes these results. Thus we have now estimated the weight of the aircraft as accurately as we can for now. Figure 6. Proposed Mass and Wingspan Mass Distribution 5

Table 2. Breakdown of Mass Available for Groups Components Mass (in kg) Payload 2.27 Avionics 1 Airframe 1.96 Batteries 3.06 Solar Panels 0.84 Maximum Power Point Tracker (MPPT) 0.1 Propulsion Group 0.42 TOTAL 9.65 V. Performance Constraint Analysis Using the parameters defined in the Mission Specification, we will now analyze the performance constraints to see if this aircraft can complete its mission. To do this, we will follow Roskam 8 methods. One critical requirement that needs to be met is the sizing to cruise speed requirement. From Roskam, we can define a Power Index formula shown in Eq. (6). W S I p = (6) W σ P From Roskam, for a velocity of 30 mph, we get a power index of 0.1714. We then can use Eq. (6) to solve for a relationship between the wing loading and power loading. Another critical requirement is the sizing to FAR 23 rate of climb requirements. We will assume that the rate of climb requirement for our aircraft is 0.3 m/s. From Roskam 9, the rate of climb and rate of climb parameter are defined by Eqs. (7) and (8). dh 1/ 3 ( RCP) RC = = 33, 000 (7) dt W η plr S RCP = (8) W 3/ 2 C 1/ 2 19 L P σ CD For this analysis, we will assume a propeller efficiency of 50%, an air density ratio at 2,000 m of 0.82168 slug/ft 3, a lift coefficient of 1.5 from the Selig 1223 airfoil, and a drag coefficient from Noth of 0.0967. Using these parameters, we then can solve for another relationship between wing loading and power loading to meet our mission requirements. We now can graph both of the sizing to cruise speed and rate of climb requirements on the same graph, which is shown in Fig. 7. The takeoff speed requirement from Roskam is defined as 70% of the cruise speed requirement graph. The intersection point that we will choose is where the takeoff curve and rate of climb curve intersect. Therefore, the wing loading is 0.9715 and power loading is 0.2204. From these values, knowing the weight, the power required ends up being 100.06 W and the wing area is 2.11 m 2. Knowing that our aspect ratio is 13, we can then solve for the new wingspan, which ends up being 5.24 m. Knowing the wing area and wingspan, we can solve for the chord length, which is 0.403 m. Next, we will look at a power analysis to see if we can meet the power required for steady level flight. Including Power Loading (lbs/w) 1/ 2 0.35 0.25 0.15 0.05 0 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Wing Loading (psf) 6 Figure 7. Performance Sizing Graph 0.5 0.45 0.4 0.3 0.2 0.1 Cruise Takeoff Rate of Climb

efficiencies such as the propeller and motor, the power required then becomes 235.4 W. To analyze how much power can be outputted by the solar cells, we need to know the exact solar cell that will be used. For this paper, the solar cell that will be used is AzurSpace s S 32 solar cell because of low weight and high efficiency. From Noth, it was shown that using a configuration of 36 solar cells connected twice in series, the total power output was about 30 W. Thus, the power output per solar cell using this configuration becomes 2.4 watts per solar cell. The dimensions for this solar cell are 31.9 mm by 74 mm. Using these dimensions along with our chord length and wingspan, we can show that the maximum amount of solar cells that can be installed onto the wing is 12 by 70 or 840 solar cells. Since we are using the same Noth configuration however, the maximum amount of solar panels we can install onto the wing are 11. Therefore, the total cells will be 792. Solving for the total power gives us 1.9 kw. From this analysis, it is clear that this aircraft will have more than enough available solar cell power to complete its mission. Note that this does not take into account the high camber and any weather issues we may have, which will reduce the power slightly. VI. Preliminary Design A. Fuselage Design The fuselage design for a solar-powered aircraft is relatively simple since there is no fuel needed or passengers to load onto the aircraft. The primary components that will go into the fuselage will be the payload, autopilot computer, batteries, and a motor for the propeller. To size the fuselage, a comparison of similar aircraft was used. For this analysis, the SunSailor and Sky-Sailor designs were most applicable. A comparison of the fuselage length versus wingspan was done for these two aircraft and a power relationship was found and is shown in Eq. (9): 0.5289 F L = b (9) Using the calculated wingspan, we then find that the fuselage length for our aircraft is 2.4 m. There will be two primary sections of the fuselage design. One section will be where the payload and any other necessary electronic or propulsion equipment will go. This section will be cylindrical in shape primarily for simplicity of building. This section will also be from the tip of the fuselage to the midpoint of the fuselage length, which is 1.2 m. For the diameter of this section, it was chosen from Raymer 10 to have a maximum diameter of 0.3 m to minimize subsonic drag. The back section of the fuselage will be a smaller diameter aluminum tube which the empennage will connect to. This again was done for simplicity and weight savings. For the diameter, it was arbitrarily chosen to have a maximum diameter of 1 inch since that is very common and easy to get off-the-shelf. B. Wing and Lateral Controls Design The airfoil, wingspan, wing area, and chord length for the wing were defined in the initial design stage. Now, we will concentrate on taper ratio, dihedral angle, incidence angle, high lift devices, and control surfaces in this section. The taper ratio is defined as the ratio of the wing tip chord to the wing root chord. From Raymer 11, for a rectangular unswept wing, this value is close to 0.45 in order to have an elliptical lift distribution throughout on the wing. For this aircraft, the taper ratio will be 1, meaning that there will be no change in the wing chord from tip to root. The main reason why this was chosen is because there needs to be enough area for the solar panels so that there is enough power available to fly. Dihedral angle on the wing is primarily used for increasing the dihedral effect of the wing. The dihedral effect is a rolling moment that results from the fact that the aircraft has a non-zero sideslip angle. Thus, the dihedral is primarily used for stability of the aircraft. For this aircraft, a dihedral will be added, but only in towards the root 0.5 m on each side from the tip of the wing. Also, the majority of the wing will be parallel to the ground, thus generating the most lift. For the exact angle, 6 will be chosen based on similar aircraft as well as the minimum used on average by sailplanes. Incidence angle of the aircraft is the angle between the chord line of the wing and the longitudinal axis of the fuselage. This value is fixed because it depends on how the wing is mounted onto the fuselage. From similar aircraft, most either use zero or negligible angle of incidence on the aircraft, and because of this, there will be no incidence angle applied for this design. High lift devices are primarily used for takeoff and landing to control the aircraft and provide extra lift and drag when necessary. For this aircraft, there will be no high lift devices installed because they are unnecessary on a model aircraft. Also, they add to the overall weight of the aircraft, which we would like to decrease as much as possible. The control surfaces that will be installed on the wing will be ailerons, which are devices on the trailing edge that help maneuver and control the aircraft. From Raymer, it is suggested to use ailerons that extend from 50% to 90% on each side of the wing. Therefore, the ratio between the total aileron span and the wing span will be 0.4. Using 7

Raymer 12, it is also suggested that for a span ratio of 0.4, the chord ratio should be 0.2 to 0.24. Therefore, we will use 22% of the total wing chord for our aileron chord. These values, as well as the total aileron surface area, are calculated and are shown in Table 3. Table 3. Aileron Dimensions Aileron Span 2.096 m Aileron Chord 0.089 m Total Aileron Surface Area 0.186 m 2 C. Empennage, Longitudinal, and Directional Controls Design As was discussed during the initial configuration, this aircraft will incorporate a v-tail empennage. This section will specify the dimensions of the empennage, as well as the control surfaces, in more detail. The location of the empennage will be calculated using Raymer s methods for a front-mounted propeller engine. From Raymer 13, it is suggested that the tail arm be 60% of the fuselage length. The tail arm is defined as the distance from the tail quarter-chord to the wing quarter-chord. Know that our fuselage length is 2.4 m, our tail arm then becomes 1.44 m. The sizing of the empennage was based on Raymer s methods for sailplanes. To calculate the tail area, we will first calculate the total horizontal and vertical area needed from Eqs. (10) and (11). cvtbs S VT = (10) L 8 c VT cs HT S HT = (11) LHT From Raymer, the values for the horizontal and vertical tail volume coefficients will be taken for sailplanes, which are 0.5 and 0.02 respectfully. Using these values, as well as the values that were calculated in the previous sections for the aircraft, we can solve for the vertical and horizontal tail area, which is 0.15356 m 2 and 0.2953 m 2 respectfully. Since our aircraft is a v-tail configuration, we also need to solve for the tail dihedral angle 14 that will agree with the vertical and horizontal areas that were calculated. Using Eq. (12) below, we then can calculate the tail dihedral angle, which is 35.8. S VT α TD = arctan (12) SHT Using the tail dihedral angle, as well as the horizontal area calculated above, the total area for one side of the v- tail can be solved for using Eq. (13). S / 2 S HT T = (13) cos( αtd ) Using similar aircraft from Noth, it is safe to assume that the tail chord length is 60% of the wing chord length. From this assumption, we find that our tail chord length is 0.2418 m. Knowing the v-tail chord length, as well as the area of one side of the v-tail, we can calculate the wingspan of half of the v-tail, which is 0.7527 m. The total wing becomes 1.505 m, and the aspect ratio for the v-tail is 12.45. The airfoil for the empennage will be a NACA 0008, shown in Fig. 8. The taper ratio will be 1 and no sweep angle will be added since there is a possibility of installing solar panels onto the empennage. Also, there will be no incidence angle from studies done on similar aircraft. The longitudinal and directional control surfaces are unique for a v-tail. Because of the shape, there will need to be a combination rudder-elevator device, more commonly known as a ruddervator. installed on each side of the v- tail. From Raymer, a common ruddervator spans from the fuselage root to about 90% of the tail span. For the solar-powered aircraft, it is better to extend the ruddervator all the way out to the tip for better control. For the chord length of the ruddervators, Raymer suggests that 25-50% of the tail chord is sufficient. Therefore, 35% of the tail chord was chosen. A summary of the aileron sizing is shown in Table 4. Figure 8. NACA 0008 Airfoil with Unit Chord Length

Table 4. Ruddervator Sizing (For One Ruddervator Only) Ruddervator Length 0.753 m Ruddervator Chord 0.085 m Total Ruddervator Surface Area 0.064 m 2 VII. Weight and Balance A. Component Weight Breakdown Before an explanation of the component weight is shown, a discussion on the list of components is required. One of the most vital components for a solar-powered aircraft is the batteries. In the case of this solar-powered aircraft, the decision was made to go with lithium-ion batteries because they provide more power than other types of batteries, and can be installed to motors fairly easily. The configuration is also important because of sizing purposes and the amount of energy needed for the aircraft. It was chosen to have 8 batteries in series, and then those batteries would be connected in parallel to 5 other sets of 8 batteries. In more common terms, it is written as 8S6P. Table 5 shows several lithium-ion batteries that were chosen for this aircraft. It is important to note that the battery with the highest energy density is preferable because with solar-powered aircrafts, weight is more important than power, so a study was done to find out which batteries gave the highest energy density. From Table 5, it is clear that the Panasonic NCR-18650A gives the highest energy density, and will therefore be used for our aircraft. Table 5. Battery Specifications Panasonic Batteries NCR 18500 NCR 18650 NCR 18650 A Nominal Voltage (V) 3.6 3.6 3.6 Nominal Capacity (Typical) (Ah) 2.0 2.9 3.1 Cell Configuration 8S6P 8S6P 8S6P Cell Voltage (V) 28.8 28.8 28.8 Cell Capacity (Ah) 12.0 17.4 18.6 Cell Energy (Wh) 345.6 501.1 535.7 Cell Mass (kg) 1.61 2.16 2.18 Energy density (Wh/kg) 214.66 232.0 245.7 As was previously mentioned, the solar cell that will be used is AzurSpace s S 32 cell because of its low weight and relatively high efficiency. The configuration for the separate solar panels will be 36 solar cells connected twice in series, totaling 72 solar cells for each solar panel. The maximum amount of solar panels that can be done is 11 since the maximum allowable area on the wing can only fit a total of 840 solar cells. With 11 solar panels, there will be 5 on each side of the wing, and then 1 located at the center of the wing. To get the most energy from the solar panels to the batteries, a maximum power point tracker, or MPPT, must be used. For business and home use, this device can be readily purchased anywhere solar panels can be purchased. However for a solar-powered aircraft, these MPPTs are not ideal because of weight. Since weight is critical for the MPPTs on this aircraft, it was decided that the MPPTs will be built using a microcontroller and a processor unit. For this paper, we will assume that Noth s model for a self-built MPPT will be used, and the weight will be taken into account in this analysis. We now must decide on the components that need power in the aircraft, such as motors, motor controllers, GPS, etc. We will start with the motor controller which is a device that is connected to the motor to supply power to it. From other solar-powered aircraft, it was suggested to look into the JETI JES 045 Plus DC speed controller. It allows connections to lithium-ion batteries, and it has its own battery eliminator circuit, or BEC, which controls the amount of voltage that gets supplied to the motor. The motor controller is then connected to the motor for the aircraft. From previous solar-powered aircraft, it is suggested to use brushless motors because they provide more power than other model aircraft motors. It was decided 9

to use an E-flite Power 90 Brushless Outrunner Motor 325Kv because it recommends a proper range of propellers for our aircraft, and it can be installed easily with the amount of battery cells we are using. After the motor was chosen, the next component is the propeller. The propeller s diameter was chosen to be a maximum of 16 inches, or 0.4064 meters, because of our fuselage size. There are two types of propellers that are currently on the market: beechwood and APC. Beechwood is made of high-quality wood that is quite efficient. APC propellers are made from lightweight molded nylon, as well as other composites, to minimize drag and increase efficiency. For this aircraft, I will try both propellers and see which one provides the aircraft with more thrust. The pitch for the aircraft was chosen to be 8 inches because this is readily available. Therefore, we will try a PJN Electric Beechwood Propeller and an APC 16 x 8 Competition Propeller, available from Hobby People. Connected to the batteries will be a series of microcontrollers and processing boards to supply power and direction to the servos and autopilot. These boards will primarily consist of dspic33 16-bit microcontrollers and digital signal controllers, as well as 8-bit PIC microcontrollers available from Microchip. From Noth, you can build your own servo board, autopilot, and energy board with these microcontrollers, and the Noth model will be used for the dimensioning of these boards and autopilot. Once the boards are installed and connected correctly, the servo motors are then chosen. For the ailerons installed on the wing, it was chosen to use Futaba BLS154 Brushless Low Profile Servos primarily because these can be installed into the wings with ease. For the ruddervators installed on the empennage, Futaba BLS151 Brushless Standard Air Servos will be used since we can fit these within the empennage and supply power to them through the fuselage. All other components, such as the GPS, RC Receiver, BECs, and the 3DOF Orientation Tracker, were based on Noth s model during his analysis. Other components that can be neglected due to weight include an altimeter and an airspeed sensor. For both of these, we will use the How Fast Airspeed MPH instrument and the How High Model Aircraft Altimeter respectfully. Since both of these together weigh less than 5 grams, they can be ignored for our weight estimates. Table 6 shows the component weights in detail. The values obtained from Table 6 will allow up to 3.91 kg for the airframe for the fuselage, wing, and empennage. Table 6. Component Weights Component Weight (in kg) Batteries 2.18 Solar Panels 0.6 MPPT 0.1 Motor Controller 0.026 Motor 0.45 Propeller 0.036 Servos 0.19 Microcontroller Boards (Autopilot, Energy Board, 0.04 Servo Board) Other (GPS, RC Receiver, IMU, Wiring, etc.) 0.2 TOTAL 3.822 B. Center of Gravity Calculations and Discussion For the center of gravity calculations, we will assume that 1.955 kg is for the large fuselage structure, 0.6517 kg is for the small fuselage, 0.9775 kg is for the wing structure, and 0.3258 kg is for the empennage. Figures 9 and 10 show the aircraft with the components inside, as well as the overall dimensions. Using these dimensions, we then can determine the center of gravity for the aircraft. For the calculations, we assumed that since every component is either rectangular or square, the center of gravity will be located at the center point relative to the x-direction. Also, the point that the moments were taken was at the nose of the large fuselage. Using this point and the dimensions and weights for each component, we then find the aircraft center of gravity with and without payload to be 66.04 and 61.59 cm respectfully. If we assume that the center of lift is acting at the quarter-chord point of the wing, which is 66.0564, then we see that for either scenario (with and without payload), the center of gravity is ahead of the center of lift, which is desired for Figure 9. Aircraft with Components 10

static stability. Also, the largest static of margin is 11.1%, which is comparable with other model R/C aircraft. Therefore, this aircraft would not need any augmentation system. VIII. Conclusion The solar-powered design discussed in this report can be applied to many applications and not only wildfire analysis. Based on the analysis discussed in this report, the idea of a solar-powered aircraft, whether it is model or commercial, can be implemented depending on the solar technology and Figure 10. Aircraft with Dimensions in Centimeters structural restraints. Also, with the current necessity for a greener society, an alternative source of energy for all aircraft is needed. There are many alternative energy solutions that are promising, including bio-fuel and hydrogen fuel cells, but nothing is as limitless as solar energy. Solar-powered aircraft can be a vital part of the future of aviation, and can be a solution to a greener society. Acknowledgments The author would like to thank Nikos Mourtos for his technical expertise and support during the writing of this report, Periklis Papadopoulos for his helpful suggestions, and James Mokri for his solar knowledge and experience. Without their guidance, none of this would have been possible. References 1 U.S. Department of Energy, The History of Solar, U.S. Department of Energy, Energy Efficiency and Renewable Energy, URL: http://www1.eere.energy.gov/solar/pdfs/solar_timeline.pdf [cited 21 September 2010] 2 Noth, A., History of Solar Flight, Autonomous Systems Lab, 2008, ETH Zürich, Switzerland 3 Wikipedia, Solar Impulse Project, Wikipedia, URL: http://en.wikipedia.org/wiki/solar_impulse_project [cited 12 July 2010] 4 Weider, A., Levy, H., Regev, I., Ankri, L., Goldenberg, T., Ehrlich, Y., Vadimirsky, A., Cohen, M., SunSailor: Solar Powered UAV, Undergraduate Report, Technion IIT, Haifa Israel, 2006 5 Raymer, D.P., Aircraft Design: A Conceptual Approach, 2 nd ed., AIAA Education Series, AIAA, Washington D.C., 1992, pp. 220-221 6 Roskam, J., Airplane Design Part II: Preliminary Configuration Design and Integration of the Propulsion System, Roskam Aviation and Engineering Corporation, Ottawa, Kansas, 1988, pp. 127-128 7 Noth, A., Design of Solar Powered Airplanes for Continuous Flight, Ph.D. Dissertation, ETH Zürich, 2008 8 Roskam, J., Airplane Design Part I: Preliminary Sizing of Airplanes, Roskam Aviation and Engineering Corporation, Ottawa, Kansas, 1988, pp. 162-166 9 Roskam, J., Airplane Design Part I: Preliminary Sizing of Airplanes, Roskam Aviation and Engineering Corporation, Ottawa, Kansas, 1988, pp. 131-132 10 Raymer, D.P., Aircraft Design: A Conceptual Approach, 2 nd ed., AIAA Education Series, AIAA, Washington D.C., 1992, pp. 109-110 11 Raymer, D.P., Aircraft Design: A Conceptual Approach, 2 nd ed., AIAA Education Series, AIAA, Washington D.C., 1992, pp. 55-57 12 Raymer, D.P., Aircraft Design: A Conceptual Approach, 2 nd ed., AIAA Education Series, AIAA, Washington D.C., 1992, pp. 113-116 13 Raymer, D.P., Aircraft Design: A Conceptual Approach, 2 nd ed., AIAA Education Series, AIAA, Washington D.C., 1992, pp. 110-113 14 Raymer, D.P., Aircraft Design: A Conceptual Approach, 2 nd ed., AIAA Education Series, AIAA, Washington D.C., 1992, p. 112 11