Electric UAV Using Regenerative Soaring and Solar Power (project proposal)

Transcription

1 Electric UAV Using Regenerative Soaring and Solar Power (project proposal) Abstract: Autonomous Electric Aircraft using no Fuel (Unmanned Aerial Vehicle UAV) Propeller powered electric UAV takes off on batteries and actively searches for updrafts. After encountering an updraft the UAV switches of the propulsion electric motor and soars. Air passing through the propeller during soaring revolves it and the movement is transmitted to the electric motor. Electro motor works as a generator in this mode. The produced energy recharges batteries and powers the electric equipment of the UAV. Energy gain is improved using solar power. The proposed UAV can stay aloft for long (indefinite) periods of time and can be used in reconnaissance and other applications. The control system of the UAV is responsible for autonomous behavior (searching for updrafts, optimization of flight trajectory with regard to the mission objective and power management, solving critical situations, etc.) and for implementation of the human issued commands. Table of Contents 1 Introduction Outline of the Idea Regenerative Soaring Electric Aircrafts Features, Equipment and Instruments of the Proposed UAV Functions of the Control System Feasibility Analysis The Aircraft Energy Balance Self Launch and Climb to 3m Cruise for 2 minutes in search for updrafts Recharge batteries to full capacity Cruise till 5% of the battery capacity remains Recharge batteries to full capacity and land Conclusion, energy balance AI and Control System Estimated Project Impact...16

2 1 Introduction 1.1 Outline of the Idea We propose an electric unmanned aerial vehicle (UAV) capable to take-off and fly using electric motor and to land with fully charged batteries. Batteries of the UAV are recharged by regenerative soaring and solar power. The UAV is expected to stay aloft for a long time (hours and possibly days). During the time aloft the UAV searches autonomously for updrafts, optimizes flight trajectory with regard to the mission objectives and power management, solves critical situations and responds to human issued instructions. The instructions are expected to be defined as partial and general objectives instead of detailed commands. The advantages of the UAV are the following: quiet and clean fossil fuel independent very low operational cost can potentially remain aloft indefinitely "parked" in air when not in use has interesting application potential study for manned regenerative soaring aircraft The possible applications include: Long time surveillance of road traffic, traffic jams avoidance Bird eye view during rescue operations or catastrophic events Search for missing persons at sea (flock of UAV's to cover large areas) Transponding radio signals in mountainous areas Following migrant birds, marine mammals etc. Weather forecast and study Many of the ongoing projects of electric aircrafts are aimed on high altitude solar powered UAV's. These try to avoid weather in order to get maximum exposure to sun and to protect their fragile lightweight construction. The UAV proposed here should operate in low altitudes in visual contact with the surface of Earth and take advantage of the vertical atmospheric motions.

3 1.2 Regenerative Soaring Illustration 1: Regenerative soaring, aircraft design (Credit [4]). Traditional sailplanes utilize updrafts to stay aloft and travel. In the typical scenario the sailplane searches for an updraft after initial climb (tow, self launch ), gains altitude by circling in an updraft and then glides in direction of the intended destination. Two important parameters the sailplane's performance are minimum sink rate and glide ratio. Minimum sink rate determines how fast will the sailplane gain altitude and the glide ratio (usually expressed as X :1, meaning the sailplane will glide to the distance of X kilometers if starting 1km above ground) determines how good will the sailplane utilize the gained altitude. Most of the updrafts used by sailplanes are either thermal columns or upwind slope lifts. Illustration 2: Soaring in thermals

4 Thermal columns (thermals) are basically bubbles of rising warm air which was warmed over sun irradiated surfaces. Upwind slope lift arises when air is forced to flow over an obstacle. It is also possible to apply dynamic soaring technique used by many migrant birds and remote control pilots. This technique uses the change of wind speed in the wind profile close to surface to gain energy. Albatrosses travel almost effortlessly thousands of kilometers in any direction using dynamic soaring. However, this technique requires to make sharp, high speed maneuvers close to the ground. Regenerative soaring feature is easily added to most of the Illustration 3: Upwind slope lift self launching sailplanes. In this case the propeller (or propellers) serves as a wind turbine while flying in the updraft. Because of the increased drag during regen the aircraft can not climb as fast as a clean sailplane but the energy generated by the turbine can be stored for future use. The most common means to store the energy are batteries, pinwheels, springs or twisted rubber bands. This energy can be used for free or emergency cruising, for new start and climb or to power devices on board of the aircraft. In the optimal case the aircraft will fly entirely without fuel or recharging on the ground. 1.3 Electric Aircrafts Electric motors are widely used in remote controlled (RC) airplanes and in several manned airplanes. The limiting factor in application of electric motors in aviation is the energy storage. Batteries or ultracapacitors provide low energy to weight ratio which means that electric airplanes have small range. Solar and atmospheric energy can be used to increase the range and the time in air. Helios prototype was a UAV of NASA, ultralightweight flying wing aircraft with a wingspan of 75.3m, powered by solar cells, batteries and Illustration 5: Sunseeker II Illustration 4: Helios prototype hydrogen-air fuel cell. Sunseeker II was as of Dec, 28 the only manned solar powered airplane in flying condition. In 29 it became the first solarpowered aircraft to cross the Alps. Its solar array charges Li-Polymer battery powering a 6kW electro motor. Max speed on solar power is 64kph. Sunseeker II takes advantage of thermals if possible. Several self launching electric sailplanes are available on the market, e.g. Antares 2E of Lange Aviation. Antares 2E is equipped with 42kW electric motor. It climbs to 3 meters in app. 13 minutes when the batteries are depleted. Electric aircrafts slowly start to appear also as commercial products. Electric Aircraft Corporation [1] produces

5 two types of electric aircrafts: rigid wing Electraflyer-C and Electraflyer trike (motor hang glider with Stratus wing [3]). Their lithium-polymer battery pack with capacity 5.6kWh lasts for hours flying. The electric motor used is 13.5kW brushless motor with 9% efficiency. Further examples of electric aircrafts are described in Chapter 2 Feasibility Analysis. Illustration 6: Electraflyer trike, Electric Aircraft Corporation 1.4 Features, Equipment and Instruments of the Proposed UAV Propulsion system components: propeller large (more efficient) with symmetrical blades cross section possibly mounted as a ducted fan brushless electric motor (dual role as turbine) battery pack (Li-polymer, Li-Ion... ) battery heating system to ensure optimal performance of the batteries solar cell array motor-generator controller charger Flight controls (fly-by-wire): primary controls: 3 (roll, pitch, yaw) secondary controls: elevator trim, wing flaps, airbrakes etc. (optional) Other equipment: retractable landing gear (optional) rocket parachute deployed in critical situations

6 Avionics and Instruments (digital): GPS altimeter airspeed indicator rate-of-climb indicator (variometer) attitude indicator (gyro horizon) turn coordinator indicators (battery, g-force, fire, stall, landing aids etc.) Special devices: Force feedback (aerodynamic load on the steering surfaces) Diagnostic system Radio receiver: for data and commands Cameras and optional sensing devices Video and data transmission system Control computer 1.5 Functions of the Control System The proposed control system intends to use ground based and on-board AI. The ground based system will analyze available meteorological data and 3D model of the surface of Earth. This system will provide hints to the UAV regarding areas with high likelihood to produce updrafts. The ground based AI system would be vital for night operation when thermal updrafts and solar array do not provide energy and UAV depends mostly on upwind slopes lift. The onboard AI will be responsible for the following: Automated take-off and landing Reaction in critical situations: stall and spin recovery, parachute deployment etc. Optimization of flight trajectory with respect to the mission objective and to the atmospheric conditions (power management) Localization and utilization of updrafts (variometer, vision and ground AI data) "Parking" in air when not in use aircraft is commanded to stay within certain space and fly autonomously until new mission objective is uploaded.

7 2 Feasibility Analysis 2.1 The Aircraft We have chosen SWIFT rigid wing hang glider produced by Belgium based firm Aériane as an candidate aircraft [2]. It was designed by Bright Star Gliders in collaboration with engineers at Stanford University The SWIFT is a high performance sailplane, designed to combine some of the convenience of hang gliders with the soaring performance of sailplanes. It takes off and lands like a hang glider, yet maintains exceptional performance at high speeds, achieving a lift-to-drag ratio of about 25:1. Although it is a fully-cantilevered rigid wing with aerodynamic controls and flaps, it weighs only 48 kg and is easily transported on the top of a car. It is sold in many countries including Japan. Aériane produces also engine Illustration 7: SWIFT rigid wing hang glider kit (with combustion engine) which is easy to adapt to a standard SWIFT frame (with steerable front wheel, disk brake, wheeled tiplets for taxing). The optional equipment of SWIFT includes rocket parachute and car roof transportation container. Instead of human pilot the payload of the aircraft will consist of lithiumpolymer battery pack (5.6kWh [1]), brushless electric motor (13.5kW [1]), solar array, electric equipment (control computer, servos, radio transmission system etc.) and other equipment used for adaptation of the aircraft. Table 1 summarizes the specifications of the SWIFT with the proposed adaptations. The estimated payload weights approximately as much as an average pilot. The propulsion system is taken from Electraflyer trike [1] which weights 112kg empty and approximately 2kg with a pilot. This ensures that the UAV will be sufficiently powered. In fact, human pilot can be on board of the aircraft during initial Illustration 8: SWIFT with engine kit experiments, observing the behavior of the control system, making measurements and ensuring safe operation. The motor and the propeller will enable to climb at estimated rate of 1ms -1. The propulsion system of Electraflyer trike is sold also adapted for regenerative soaring [2]. It will be necessary to develop a new propeller optimized

8 for regenerative soaring. The commercially available propellers are optimized to provide maximum trust. The propeller for our purpose should be optimized to work also as a turbine with high efficiency which among other requires symmetrical blades sections [4]. Optional features of the propeller include adjustable pitch and collapsibility. Glide ratio (best, at 75km/h) [2] 24:1 Minimum sink rate (at 45km/h) [2].65ms-1 Never exceed speed (VNE) [2] 12 km/h Climb rate (estimated) 1ms-1 Wing area [2] 12.5m2 Maximum load [2] +5.3g/-2.65g tested +7.95g/-3.98g Weight empty [2] 48kg Payload total: 74kg battery pack (5.6kWh [1]) el. motor (13.5kW [1]) solar array 1 electric equipment other Gross weight 35kg 12kg 1kg 7kg 1kg 122kg 2 Estimated costs : 7.2Mil Yen (8 USD) SWIFT LIGHT with pod + closed fairing [2] Air brakes [2] Rocket parachute [2] lithium-polymer pack 5.6kwh [1] Electraflyer propulsion kit 3 [1] Solar array (1m2, with installation costs) Carbon fiber propeller 4 Other (modifications of SWIFT) 5 Transportation container [2] 2182EURO 2EURO 185EURO 85USD 42USD 8USD 1USD 5USD 355EURO Table 1: Proposed UAV specifications. 1 With installation aids and accessories 2 Does not include electric equipment developed during the project 3 Includes: motor, electronic controller, power dial and switch, fuse, connectors, ammeter and shunt, voltmeter, custom machined propeller hub, and digital motor temperature display with probe 4 Development, design, manufacture costs - estimated 5 Development, design, manufacture costs - estimated

9 2.2 Energy Balance The purpose of this section is to show that it is possible to build an electric aircraft capable of self launch, long daytime flight and landing with a full battery. The proposed aircraft is described in the previous section. The scenario investigated consists of 5 phases: 1. Self launch and climb to 3m, 2. 2 minutes cruise in search for updrafts, 3. Recharge batteries to full capacity 4. Cruise till 5% of the battery capacity remains. 5. Recharge batteries to full capacity and land The following calculation requires to estimate certain values. The estimated values were carefully calculated, consulted (see Acknowledgments) and adjusted to slightly under or over estimate the actual values against the benefit of the UAV. Also, only the energy balance considering the electric energy is shown here. It is ignored that the plane can use the potential energy gained by climbing in the updrafts during regeneration for travel by gliding. But please note that traditional sailplanes utilize only this form of energy to fly over large distances (Free out-and-return distance record: km, [6]) and to stay aloft for long periods (~15 hours). The proposed battery pack lasts for hours of Electraflyer trike's powered flight. By takeoff the proposed UAV weights 78kg less than Electraflyer trike. Therefore is is assumed that the battery pack will last for at least 1.5 hours in the UAV. The proposed propeller works with 85% estimated efficiency ([4], efficiency of airborne turbine is defined differently than efficiency of a ground based turbine and Betz limit does not apply here). The brushless electro motor used in this example is 9% efficient as a generator [2].When the propeller works in a turbine mode it creates drag which leads to increased sink rate. If we assume the sailplane to be in level flight in steady air it is loosing potential energy at certain rate corresponding to the sink rate at given speed. If additionally electric energy is to be generated the sailplane has to loose potential energy at higher rate. Assuming the potential energy is converted into kinetic energy and then to electric energy with the overall efficiency corresponding to the efficiency of the propeller-turbine times the efficiency of the generator the increase of the sink rate of the aircraft can be calculated if the gross weight of the aircraft is known. We assume that increasing the sink rate by.75ms-1 will not seriously impair the flight performance of the UAV. Therefore we estimate to obtain 687W on the output of the generator during regenerative soaring. The battery charger works with 8% estimated efficiency (charging efficiency of the batteries is 99.9%). The charger works only when the power balance is positive. We estimate that 8% of the wing area can be used to install solar cells. The efficiency of solar cells suitable for airplanes varies between ~6% to ~2% [5]. Let us assume installing solar cells with 15% efficiency providing maximum output 15W per m 2. The solar cells are estimated to provide on average 7% of their maximum output during cruising (the UAV flies sometimes under the shadows of clouds and the wing is not always in the optimal position relative to the sun). During regeneration are the solar cells estimated to provide on average 4% of their maximum output (the UAV is moving in banked turns and the wing is only part of the time exposed to the sun). The electric equipment of the UAV (controll computer, servos, radio transmission system etc.) is estimated to consume on average 15W. Table 2. summarizes the values and estimates used in this example.

10 Battery pack [1] 5.6kWh (2.2MJ) Average powered flight time [1] 1.5 hour Electric motor [1] 13.5kW, 9% efficient Average energy consumption rate motor 3.7kW Equipment input 15W Propeller efficiency (turbine mode) 85% Electric motor efficiency (generator mode) 9% Charger efficiency 8% Generator output in regen [4] 687W Solar power per m2 max [5] 15W Solar cells surface 1m2 Solar array output - max 15W Solar array output - cruising 15W Solar array output - regen 6W Sink rate total in regen [2] 1.4ms-1 Table 2: Values and estimates used for calculation.

11 2.2.1 Self Launch and Climb to 3m 3m is an altitude suitable for search for thermal updrafts [4]. The electro motor is running at full power during the start only [2]. The climb is done at reduced power (high discharge rate more rapidly depletes the battery capacity). During climbing electric energy is used to sustain flight (compensate the drag) and also converted to potential energy of the aircraft The only source of energy now is the solar array. Table 3 summarizes the energy balance for this phase. The values are rounded. Altitude gained = 3m Balance: Time spent = 5 minutes Power (Watt) Solar array Energy (Joule) Motor (sustain flight) Motor (to potential energy) Total Turbine (Regen) Equipment Battery capacity spent: 6% Battery capacity spent total: 6% Table 3: Self launch and climb to 3m, energy balance Cruise for 2 minutes in search for updrafts It is likely that the aircraft will encounter updraft in 2 minutes after start. As it maintains the same altitude the motor spends energy only to sustain flight level flight. Table 4 summarizes the energy balance for this phase. The values are rounded. Altitude gained = m Balance: Time spent = 2 minutes Power (Watt) Solar array Turbine (Regen) Equipment Motor (sustain flight) Motor (to potential energy) Total Energy (Joule) Battery capacity spent: 17% Battery capacity spent total: 23% Table 4: Cruise for 2 minutes in search for updrafts, energy balance.

12 2.2.3 Recharge batteries to full capacity The aircraft has found an updraft (thermal, upwind slope lift) and it is regenerating. The propeller works as a turbine. The aircraft will continue to recharge until the batteries are full. The efficiency of the charger has to be taken into account in this phase. If the updraft is strong enough (1.4ms-1 is the minimum sink rate in this phase) the aircraft gains altitude. It can then glide in the direction of the mission objective and continue recharging. Altitude gained = Ignored Balance: Time spent = 85 minutes Power (Watt) Energy (Joule) Solar array Turbine (Regen) NA NA Total Recharge (8% eff.) Battery capacity spent: -23% Equipment Motor (sustain flight) Motor (to potential energy) Battery capacity spent total: % Table 5: Recharge batteries to full capacity, energy balance Cruise till 5% of the battery capacity remains The aircraft is cruising freely and accomplishing the mission objectives. Altitude gained = m Balance: Time spent = 59 minutes Power (Watt) Solar array Turbine (Regen) Equipment Motor (sustain flight) Motor (to potential energy) Total Energy (Joule) Battery capacity spent: 5% Battery capacity spent total: 5% Table 6: Cruise till 5% of the battery capacity remains, energy balance.

13 2.2.5 Recharge batteries to full capacity and land The aircraft is recharging and continues to accomplish the mission objectives. The turbine generated power is higher during the final descent what can shorten the recharge time before landing but the turbine power is considered to be constant during the entire final phase of the investigated scenario for simplicity. Altitude gained = NA Balance: Time spent = 185 minutes Power (Watt) Energy (Joule) Solar array Turbine (Regen) NA NA Total Recharge (8% eff.) Battery capacity spent: -5% Equipment Motor (sustain flight) Motor (to potential energy) Battery capacity spent total: 1%

14 2.2.6 Conclusion, energy balance Following this analysis it has been shown that it is possible to build an unmanned aircraft powered with solar power and regenerative soaring from an existing aircraft so that it would be capable to self launch, fly and land with full battery. The total flight time in the investigated scenario was 5 hours and 53 minutes. In this configuration the ratio between free cruising and regen is 1:3.1 which means that the aircraft would need to regenerate for 3.1 minutes for every minute of powered cruising to have the batteries fully charged by landing. It should be considered that the regeneration time may be used to accomplish mission objectives (e.g. the UAV can perform surveillance and regen at the same time). The analysis shows that using regenerative soaring shortens the recharge time significantly during the day. Regenerative soaring is the only source of energy for the aircraft during the night. To maximize the efficiency of the UAV the following should be considered: solar cells should cover maximum of the suitable surfaces and have the highest possible efficiency power input of the equipment should be minimized drag of the aircraft should be minimized large and slow (more efficient) propellers should be used (possibly installed as ducted fans) the control system should efficiently search for strong updrafts enabling higher recharge rates. A grand challenge for the aircraft would be to stay aloft during the nighttime when two important sources of energy vanish: sun and thermal updrafts. UAV could still use upwind slopes and dynamic soaring. Using upwind slopes for lift would require reliable ground analysis of atmospheric conditions (information on wind speed and direction used to model air motion in a 3D map of the operation area). Dynamic soaring requires precise flight control in sharp low altitude maneuvers (2-15m over ground) and a reasonably small and sturdy aircraft (dynamic soaring is often used by RC pilots).

15 2.3 AI and Control System Autonomous soaring for UAV's was first proposed in 1998 [7]. Recursive learning was used to center updrafts and neural networks were used to identify updraft positions. Algorithms were too intensive for real-time use at that time. Since then several other works were focused on soaring UAV's (e.g. [8,9]). The SoLong unmanned aerial vehicle from AC Propulsion flew on June 13, 25 over 48 hr nonstop fueled only by solar energy. The plane sports a wingspan of 4.75 m and weighs 12.6 kg. NASA Dryden Flight Research Center supported a project of an autonomous soaring UAV [8] (CloudSwift, Span: 4.26m, Weight: 6.58kg, Stall speed: 33kt, Mission speed: 46kph). Illustration 9: SoLong UAV, AC Propulsion. The experiments were first performed in simulation. Simulation results showed that a small UAV can benefit significantly by exploiting updrafts and simulation study assumed that a small UAV could autonomously detect and center updrafts. CloudSwift UAV was used for real world experiments. Updraft detection sensors were not used. Updrafts were only detected after the UAV had physically encountered them. Archimedes spiral pattern was chosen for the UAV to fly while searching for updrafts. During 17 test flights CloudSwift found 23 updrafts, climbed maximum 844m in a single updraft and gained 172 meters in an updraft in average. These results indicate that it is possible for an UAV to autonomously search and utilize updrafts. We propose to improve the performance of the UAV by using a vision system for recognition of typical signs of updrafts (cumulus clouds) and ground wind and weather analysis for Illustration 1: CloudSwift, NASA Dryden identification of areas producing updrafts (upwind slopes and thermals). We also propose application of artificial neural network learning to fly by observing actions of a human pilot. This can be done either off-line using recorded flight data or on-line because the proposed aircraft enables presence of human pilot on board during experiments. The trained artificial neural network can be used as a subsystem of the control system. The core of the control system is proposed to be rule based.

16 3 Estimated Project Impact Results of this project would contribute to several scientific fields: aeronautical engineering, electric engineering, robotics and artificial intelligence. The proposed UAV is intended to fly with or without a human pilot. The possibilities of this design are important because it enables straightforward step to manned flight without fuel. In this setup artificial neural network can be used to learn how to fly from a human pilot. Comparison of performance of manned and unmanned aircraft would be possible. Regenerative soaring has not been practically tested on an aircraft yet. The measurements taken could be helpful for future applied research. Important are also the results obtained in the field of artificial intelligence and robotics. These results could be important for development of completely autonomous UAV's for extra-terrestrial research e.g. for Mars exploration.

17 Acknowledgments I would like to thank to the following people who contributed to the project proposal by providing advice and consultation: Phil Barnes is Principal Engineer at Northrop Grumman Corporation. He has a Master s Degree in Aerospace Engineering from Cal Poly Pomona and a Bachelor s Degree in Mechanical Engineering from the University of Arizona. He has 25-years of experience in the performance analysis and computer modeling of aerospace vehicles and subsystems at Northrop Grumman. Phil has authored technical papers on aerodynamics, gears, and flight mechanics. Phil Barnes is the author of the paper Flight Without Fuel - Regenerative Soaring Feasibility Study [4] Randall Fishman is the president of Electric Aircraft Corporation. He has won numerous awards and accolades for his work on electric flight and already has built an electric-powered ultralight and a single-seat motorglider. In April 27 the Electric Aircraft Corporation began offering complete electric ultralights and engine kits under the ElectraFlyer brand name, to convert existing ultralight aircraft to electric power, in what is the first commercial offering of an electric aircraft. Jukka Tervamaki graduated from the Helsinki University of Technology in 1963 specialized in Aeronautical Engineering. Experimenting, creating, designing and building has been everyday work as well as a hobby for him for four decades. He designed and build several rotary wing aircrafts (autogyros), a motor glider and cooperated on development of a fixed wing tow plane. He has logged total 22 flight hours of which 15 hours in autogyros. He is aviation writer for several Finnish and foreign aviation magazines.

18 Sources: [1] Electric Aircraft Corporation, correspondence with Randall Fishman, president of Electric Aircraft Corporation [2] Aériane SWIFT rigid wing hang glider [3] Northwing, Stratus, [4] Philip Barnes - Pelican Aero Group, Flight Without Fuel - Regenerative Soaring Feasibility Study, presented at General Aviation Technology Conference & Exhibition, August 26, Wichita, KS, USA, Session: Propulsion Dynamics and Advanced Engine Concepts, correspondence with Philip Barnes. [5] NASA Dryden Fact Sheet - Pathfinder Solar-Powered Aircraft news/factsheets/fs-34-dfrc.html; [6] The Worlds Air Sports Federation as of Oct [7] J. Wharington, I. Herszberg (1998), 'Control of High Endurance unmanned air vehicle', Proc. 21st Congress of the International Council of the Aeronautical Sciences, Rodney S. Thomson & Murray L. Scott, eds., AIAA Electronic Publication - CD ROM, ISBN: , (ICAS '98, 1318 Sept. '98 - RMIT Fishermen's Bend, Melb. Vic.) [8] Allen, Michael J. (25) Autonomous Soaring for Improved Endurance of a Small Uninhabited Air Vehicle. Meeting Presentation AIAA , Research Engineering, NASA Dryden Flight Research Center. [9] Guidance and control for an autonomous soaring UAV, United States Patent