PHOTOVOLTAIC POWER FOR LONG ENDURANCE UNMANNED AERIAL VEHICLES

Transcription

1 PHOTOVOLTAIC POWER FOR LONG ENDURANCE UNMANNED AERIAL VEHICLES Thomas R. Lamp Aero Propulsion and Power Directorate Wright Laboratory Wright Patterson AFB, OH , Fax: Kitt C. Reinhardt Space Technology Directorate Phillips Laboratory Kirkland AFB, NM , FAX: Anthony J. Colozza Aerospace Technology Department NYMA Corporation 2001 Aerospace Parkway Brookpark, OH , Fax: ABSTRACT Computer simulation trade-studies were conducted within the Aerospace Power Division at Wright (1 of 15) [ :21:06]

2 Laboratory to determine how photovoltaic (PV) cell efficiency and mass influence the performance of high altitude long endurance (HALE) unmanned aerial vehicles (UAV). These simulations employed existing UAV experimental data and associated weight and power algorithms, state-of-the-art propulsion and energy storage conversion efficiencies, and a range of PV cell efficiencies and array mass-densities. It was found that for a given payload weight and set of flight conditions, both the weight and size (wing-span) of the UAV are not only dependent on PV cell performance, but to an even greater extent, on the performance of energy storage systems. The primary intent of conducting the simulations was to provide insight into the remaining power system deficiencies that currently limit solar-powered UAVs. It was concluded that an energy storage specific energy (total system) of Whr/kg is required to enable most useful missions, and that PV cells with efficiencies greater than ~12% are suitable for use in lower latitudes. Improve PV cell efficiencies can work to extend the operating regime to more northerly climes. This paper summarizes the computer simulations and present photovoltaic development efforts that are aimed at the UAV application. INTRODUCTION The Duke of Wellington once remarked that "...all the business of war...is guessing what is on the other side of the hill". Throughout history, military commanders have sought the advantage of knowing the intentions and strengths of their enemies. In modern times, both success and failure in this quest is demonstrated by actions in the Gulf War. Coalition forces were reasonably successful in determining the disposition of the Iraqi order of battle. However, the lack of a continuous on-station reconnaissance system foiled attempts to locate and destroy scud missile launchers in a timely fashion. The subject of this paper, solar power UAVs, holds the promise of just such a long duration reconnaissance system. Due to the long endurance and high altitudes at which these aircraft will be required to fly (20 km or higher), the method of propulsion is the major design factor in constructing the aircraft. An attractive method of supplying power for this type of aircraft is using solar photovoltaic modules coupled with an energy storage system. The principal advantage of employing this all-electric approach over others, such as open cycle combustion engines or air breathing fuel cells, is that it eliminates the need to carry fuel and to extract and compress air at the oxygen-lean high altitudes. Further, for use in military applications, electrically powered aircraft will have a very small thermal signature which is useful in avoiding detection. In the 1980s, analyses showed that the performance of state-of-the-art strong and lightweight airframe, energy storage, and photovoltaic cell technologies could not satisfy the demands of a solar-powered UAV. However, with the recent advent of advanced composite materials for airframe construction, high efficiency motors and propellers for propulsion, improvements in energy storage, and lightweight photovoltaic cells for power generation, the feasibility of a long endurance solar-powered UAV is approaching reality. (2 of 15) [ :21:06]

3 SUMMARY OF COMPUTER SIMULATIONS Method The method used for our computer simulations is based on the work performed by NASA Langley Research center and Lockheed Missiles and Space Company on the high altitude powered platform (HAPP), and NASA Lewis Research Center (Hall et al., 1983; Youngblood and Taley, 1984; Colozza, 1994). In order for a solar powered aircraft to be capable of continuous flight, enough energy must be collected by the photovoltaic array to charge the energy storage system for propulsion at night, while providing sufficient power for propulsion during the day. Figure 1 illustrates the solar power (Psc) produced by the photovoltaic cells per unit area (S), Psc/S, as a function of time (Colozza, 1994) and was used to determine the amount of solar energy available to the aircraft. The Psc/S curve is a function of PV cell efficiency, altitude, location, and time of year. The area under the Psc/S curve represents the total power collected over a 24 hr period used for propulsion during day and night hours. The area under the Psc/S curve and above the Ptot/S value, denoted AN, is the energy generated during daylight hours that FIGURE 1 (3 of 15) [ :21:06]

4 Energy Balance Diagram. can be used to power the aircraft during the night. The energy required for aircraft propulsion during the night is shown as A'N, resulting in the energy balance relation ANhes = 2A'N, where hes is the energy storage system efficiency. A value for Ptot/S is obtained from the diagram using geometric arguments, and represents the output power per unit wing area (Sw) available to fly the aircraft and run any necessary equipment. Aircraft power requirements based on aircraft mass and flight altitude are then determined considering the power conservation equation, Ptot/Sw = [Preq + Pres + Ppay] Sw, (1) where Preq is the power required for level flight obtained using the velocity for minimum power or maximum endurance [Colozza, 1994], Pres is the power required for aircraft climbing and maneuvering, and Ppay is the payload power. Expressions for Preq and Pres can be obtained for a particular airframe type in terms of aircraft total mass (mtot), wingspan (B), and aspect ratio (AR). Hence, the value of Ptot/S obtained from the energy balance diagram can be expressed as a function of mtot, AR, and B. In order to eliminate one of these variable, an expression for mtot is determined in terms of AR and B; where mtot consists of mass contributions from the airframe, engine and propeller, photovoltaic cells, energy storage, and power conditioning and control electronics. The airframe considered in this analysis is illustrated be in Figure 2, and was previously studied under the HAPP program in the mid-1980s. This airframe design was selected because it offers enhanced enhanced structural stability for the very large wingspans that the HALE-UAV requires. The propulsion and power systems consisted of an electric motor, gear box, and propeller for propulsion; PV modules for power generation; and fuel cells for energy storage. In these simulations it is assumed that the PV modules are mounted only on the main wing of the aircraft and that the packing factor is 100 percent. Input Parameters Photovoltaic input parameters used in the PV-powered UAV simulations are contained in Table 1, while the parameters concerning energy storage, power management and distribution (PMAD), mission characteristics and electric propulsion motors are shown in Table 2. (4 of 15) [ :21:06]

5 FIGURE 2 Dual-Boom Airframe. Photovoltaic Modules Air Mass Zero Efficiency (%) Specific Mass (kg/m^2) Thin Film CuInGaSe Si (110 µm) GaAs (105 µm) GaAs (70 µm) GaInP / GaAs (70 µm) TABLE 1 Photovoltaic Performance Parameters used to Conduct Simulations. Values used for module specific mass were based on the flexible photovoltaic module fabrication process employed on the Pathfinder solar-plane (Carey et al., 1994); using 90 g/m2 for interconnects and diodes, 178 g/m2 for module laminant and adhesive, 12 g/m2 for module/wing adhesive, and a PV cell packing factor of Values for photovoltaic cell thickness and mass were based on projections resulting from on-going Air Force photovoltaic cell programs (Reinhardt, 1996) and the thin 110µm Si photovoltaic cells flown on the Pathfinder UAV. (5 of 15) [ :21:06]

6 Fuel Cell / Electrolyzer System: Efficiency 67% Specific Enegry: Whr/kg Power Conditioning & Control Electronics: Efficiency: 90-99% Mission Characteristics: Flight Altitude: km Flight Latitude: 0-40 N Flight Date: 3/22, 6/22, 9/22, 12/22 Payload Mass: kg Payload Power: W TABLE 2 Parameters used to Conduct Performance Simulations of Solar-Powered UAVs. Results The output for our simulations can be separated into two categories: the first considers the result of varying the power system components, the second the results of varying mission requirements. Solutions to the maximum endurance calculations were plotted as wingspan versus aspect ratio and wingspan versus fuel cell specific energy curves. Aircraft design points below the plotted curves represent an aircraft system configuration not capable of collecting enough energy during the day to fly continuously for 24 hours, whereas any design point on or above the curve can provide a continuous flight capability. How PV Module Mass & Efficiency Influence UAV Size. The impact of PV module efficiency and specific mass on UAV wingspan versus aspect ratio is shown in Figure 3. Values used for power system performance and mission scenario are as follows: fuel cell specific mass is 400 Watt-hour per kilogram, payload mass and power are 50 kilograms and 50 Watts; and the date is June 22. The plots indicate that for fixed aspect ratio only a modest reduction in wingspan results with increased module efficiency. For example, an increase in module efficiency from 10 to 24 percent and 18 to 24 percent reduces the wingspan by only about 20 feet and 2 feet, respectively. The plots also show that the specific mass of the PV module has a greater impact on wing size than module (6 of 15) [ :21:06]

7 efficiency does. The lighter (0.286 kilogram per square meter)10 percent thin film and heavier (0.485 kilogram per square meter)18 percent silicon modules yielded nearly equal wingspans, and the lighter (0.295 kilogram per square meter) 18 percent module yielded a wingspan 13 feet shorter than the heavier (0.485 kilogram per square meter) 18 percent module. How Energy Storage Influences UAV Size. Next we analyzed the impact of PV module performance on wingspan versus fuel cell specific energy for a fixed wing aspect ratio of 25. A value of 25 was selected for practical design reasons, and was used for the remainder of analysis; a larger aspect ratio would have compromised the structural integrity of the airframe. Figure 4 shows the dramatic effect of the fuel cell's specific energy on wingspan and, by comparison, the relatively marginal impact of PV module performance. In the 18 percent, kilogram per square meter silicon module, there is a very strong dependence of wingspan on fuel cell specific energy. A specific energy of at least 280 Watt-hour per kilogram is necessary to maintain a wingspan of less than 250 feet. The maximum practical wingspan for most projected UAV missions is under 250 feet, and a value closer to 150 to 200 feet is highly desirable for both aircraft structural integrity and cost reasons. Figure 4 also shows that a specific energy of about 400 Watt-hours per kilogram is needed to yield a wingspan of less than 200 feet. Existing energy storage systems are limited to specific energies of less than about 150 Watt-hours per kilogram. Hence, even in the case of the lightweight 24 percent PV module, the resulting wingspan would be greater than 400 feet. There are two reasons for this dependence. First, the required wing size is proportional to the amount of lift needed to balance the aircraft's total mass for level flight, where a considerable part of the total mass is due to energy storage (typically about 30 percent). Consequently, in Figure 4 the wingspan is shown to increase almost exponentially with decreasing fuel cell specific energy, i.e., increasing fuel cell mass. Second, although contrary to PV module efficiencies greater than about 15 to 18 percent produce only a minimal reduction in wingspan results, as shown in Figure 5. However, for power-intensive missions, higher PV cell efficiencies (as well as lower mass densities) will provide enhanced mission performance in terms of higher payload carrying capability. Also, for low PV module efficiencies (less than 10 about percent), the percentage increase in power corresponding to a change in efficiency, say 5 to 10 percent, is large (100 percent increase) and produces a significant reduction in aircraft size. However, at higher PV module efficiencies, the increase in efficiency from, for example, 18 to 24 percent, represents an increase of only 33 percent, and consequently the reduction in aircraft size is much smaller. (7 of 15) [ :21:06]

8 FIGURE 3 Wingspan vs Aspect Ratio for Various Photovoltaic Module Designs. How Altitude, Season & Latitude Influence UAV Size. The flight altitude range that encompasses the majority of reconnaissance, scientific, and environmental missions of interest is between 20 and 30 kilometers. Our simulations showed a significant increase in wingspan with increasing altitude as illustrated in Figure 6. This is caused by the decrease in air density that occurs with higher altitude, having the effect of reducing the wing lift for a given wing velocity. Because wing lift is proportional to aircraft wing area and the square of velocity and because there is little design latitude for increasing aircraft velocity due to minimum power (maximum endurance) and propulsion system mass arguments, the wing area must increase to compensate for the reduced air density. To maintain an acceptable wingspan at high altitudes, significantly higher values of fuel cell specific energy are FIGURE 4 Wingspan vs Fuel Cell Specific Energy (8 of 15) [ :21:06]

9 required, and operation above 28 kilometers may be impossible. Season and latitude at which the UAV operates determine the charge/discharge time available for solar energy storage and the total amount of solar power available for propulsion. In general, the lower the available solar energy per wing area, the larger the wing area must be to generate sufficient power to propel the airframe and energy storage mass (airframe and energy storage comprise about 70 percent of the total mass). On the summer solstice, June 22, the longest period of sunshine and shortest period of eclipse occur, resulting in the lowest demand for energy storage. Consequently, the small energy storage mass drives the wingspan to a minimum. Conversely, on December 22, the winter solstice, the shortest sunshine and longest eclipse periods occur, which substantially increase the mass of the energy storage system, resulting in the largest wingspan. December 22 represents the worst-case scenario, and any aircraft power system/mission configuration feasible on this day will be capable of operation throughout the year. The dependence of UAV size on season and latitude is shown in Figure 7. Although operation may be feasible in the lower latitudes in December, it is not possible at latitudes greater than about 20 N. A fairly significant reduction in aircraft size can be obtained by restricting the time of year that the aircraft is required to fly at the different latitudes. A fuel cell energy storage specific energy between 250 to 400 Watt-hours per kilogram is required to enable most missions between mid-february and mid-october. How Payload Mass & Power Influence UAV Size. As expected the impact of payload mass on aircraft size is significant; greater values for fuel cell specific energy are required to support the larger payloads. Therefore, any reduction in payload mass, such as through the use of lightweight, low-power, miniature satellite instruments and equipment, will be very important. Also, our analysis showed that moderate increases in payload power have a very small effect on aircraft size. This is reasonable since payload power, which is on the order of hundreds of watts, is significantly less than the power required to fly the aircraft, which is typically 15 to 20 kilowatts. FIGURE 5 UAV Wingspan vs PV Module Efficiency (9 of 15) [ :21:06]

10 FIGURE 6 Impact of Altitude on UAV Wingspan For the sake of completeness, the effect of electronics efficiency was also examined. Electronics are necessary for PV power conditioning, charging/discharging the energy storage system, and switching power to the electric motors used for propulsion. In our analyses, silicon-based electronics were assumed with an efficiency of 93 percent. Depending on the operating conditions, state-of-the-art silicon power devices (diodes, MOSFETs, IGBTs, etc.) and power conditioning and control circuits (converter/inverter circuits) can have efficiencies in the range of 90 to 93 percent. FIGURE 7 Impact of Flight Latitude and Season on UAV Wingspan Next-generation wide-bandgap electronics using semiconductors such as silicon carbide show excellent (10 of 15) [ :21:06]

11 potential for yielding power devices with efficiencies approaching 99 percent. These electronics have enormous potential in military and commercial aircraft, ground vehicle, and utility applications. In our analysis an increase in electronics efficiency from 90 to 99 percent was found to reduce the wingspan by only 18 feet for a specific energy of 300 Watt-hours kilogram. However, increased PMAD (Power Management and Distribution) efficiency will reduce the demands on energy storage thereby enabling sufficiently smaller wingspans. PV CELL DEVELOPMENT EFFORTS The initial effort to investigate PV cells for UAV applications consisted of a phase I SBIR contract that was concluded in 1995 with Spire Corporation (Wojtczuk, 1995). This work resulted in the development of a thin (5µm), 16-18% (AM0), GaInP2/GaAs PV cell bonded to a m Pilkington GMC coverglass. The performance demonstrated with these Spire cells was 1040 W/Kg and 0.21 Kg/m2. The GaInP2/GaAs cell could boost conversion efficiency to the range of 25%. The theoretical AM0 cell efficiency for GaInP2/GaAs is about 28%. When the cost associated with different PV module technologies was considered (see Table 3), it was apparent that the cost of the high-efficiency (low mass) technologies would be prohibitive in the short term. The cost-per-watt values shown are for PV cells only and do not include the cost of module fabrication. The apparent expense associated with these more advanced PV cells prompted the Aerospace Power Division to look at lower cost alternatives. Work is currently underway in a phase II project that focuses on developing new, lightweight PV modules that are based on Si-PV cell technology. Hardware from this project will be installed on a NASA-Dryden Research Center UAV for flight testing. The modules measure slightly less than 0.5x1.0m and consist of modified low-cost terrestrial Si-based PV cells. Prototype panels are being installed on wing sections to determine how robust these modules are during handling and UAV operations. Figure 8 shows one such prototype panel being applied to a UAV wing. Typical performance characteristics are illustrated in Table 4. The principal objective of the present effort is low cost combined with moderate efficiency and robust handling characteristics. The present work will be followed up with more advanced PV cells as they become available. Solar Module Thin Film 10%, KG/M^2 Thin Si "Pathfinder" 15%, KG/M2 Thin GaAs or GaAs/Ge 18%, KG/M2 GaInP2/GaAs 24%, KG/M2 Cost per Watt UAV Power Required Wing Span Total Cost $3.50-$ kw 92 ft $100 k $ kw 200 ft $2.2 M $ kw 175 ft $ 9.0 M $500 - $1000 (use $500) 16 kw 173 ft $ 8.0 M (11 of 15) [ :21:06]

12 TABLE 3 Different PV Cell Technologies vs Cost This seventeen foot wingspan UAV featured in Figure 9 was built by NASA Lewis Research Center for experiments in solar powered aircraft. The aircraft is shown during a test flight in which electric power was supplied by battery packs. The UAV is currently being modified by the installation of advanced photovoltaic cells that were supplied through a Manufacturing Technology program sponsored by the USAF Wright Laboratory. Parameter Module A Module B Cell Efficiency (%) Cell Thickness (µm) Module Power (W) Module Mass (g) Module Specific Power (W/kg) Module Area (m^2) Module Power Density (W/m^2) TABLE 4 Key Characteristics for Lightweight PV Modules. Power is Rated at 1 kw/m2, AM1.5, 25 C. (Nowlan, 1997) CONCLUSIONS Results of this study showed the impact of power system component performance and mission conditions on UAV aircraft size. The most significant reduction in aircraft size was found to occur by increasing the energy storage (fuel cells in this case) specific energy, whereas the effects of PV module and power electronics efficiency and mass play a marginal role in comparison. Flight altitude, flight latitude, time of year, and payload mass also play significant roles in determining aircraft size. It was found that an energy storage specific energy of Whr/kg is required to enable most useful missions, and that operation in the winter at northern latitudes may not be possible. Also, it is concluded that low-cost thin-film PV cells, at least 10% efficient, are excellent candidates for use on solar-powered UAVs. Although this study focused on the technical issues of sustained flight of a solar-powered UAV, thermal management of UAV platforms will also play an important role in the exploitation of not only solar UAVs but conventionally powered UAVs as well. Unmanned combat aerial vehicles (UCAVs) equipped (12 of 15) [ :21:06]

13 with directed energy weapons (DEWs) are already under consideration. The reconnaissance community is seeking unmanned reconnaissance aerial vehicles (URAVs) capable of working HALE missions at altitudes of 20 to 25 kilometers. Due to the extreme levels of heat flux and waste heat, the DEW and HALE missions produce special concerns for thermal management. The rarefied atmosphere at extreme altitude provides little excess power for thermal management assets, and virtually no means of rejecting waste heat. The message is...thermal management issues also must be considered early in development, and heroic efforts might be needed to keep UAVs reliable throughout the intended mission profiles. FIGURE 8 PV Module Being Mounted on a UAV Wing. These manufacturing rejects were used only to evaluate robustness of installation and handling methods. FIGURE 9 Electric-Powered UAV During Flight Testing at NASA Lewis Research Center. (13 of 15) [ :21:06]

14 REFERENCES Carey, P.G., Aceves, R.C., Colella, N.J., Sinton, R.A., and Glenn, G.S., 1994, "A Solar Module Fabrication Process for HALE Solar Electric Vehicles," Proceedings of 1st World Conference on Photovoltaics and Energy Conversion, p.1963, Hawaii. Colozza, A.J., 1994, "Effect of Power System Technology and Mission Requirements on High Altitude Long Duration Aircraft," NASA Contractor Report Hall, D.W., Fortenbach, C.D., Dimiceli, E.V. and Parks, R.W., 1983, "A Preliminary Study of Solar Powered Aircraft and Associated Power trains," Lockheed Missiles and Space Company Inc., NASA Contractor Report Nowlan, M.J., Maglitta, J.C., Darkazalli, G., 1997, "Ultralight Photovoltaic Modules for Unmanned Aerial Vehicles", IEEE Photovoltaic Specialists Conference, Anaheim, CA, September Reinhardt, K.C., 1996, "Feasibility of Solar-Powered Unmanned Aerial Vehicles (UAVs): Impact of Photovoltaics, Energy Storage and PMAD," Aerospace Power Division (WL/POOC-2), Air Force Wright Laboratory, internal study and report. Youngblood, J.W. and Taley, T.A., 1984, "Design of Long-Endurance Unmanned Airplanes Incorporating Solar and Fuel Cell Propulsion," AIAA For information on ordering hard copies of GRC reports, look at the GRC Technical Reports Server FAQ Other Home Pages Back to ERAST, PTD Publications For more information or to provide feedback on these subjects, contact or Tony Colozza who provided the information on this page. Questions about this server: (14 of 15) [ :21:07]

15 This page mantained by: Last Updated: 13:58 Tues, Feb. 23, (15 of 15) [ :21:07]