Fuel Cell Powered Unmanned Aerial Vehicle

MAE 435 Midterm Report

Fuel Cell Powered Unmanned Aerial Vehicle

March 1, 2016

Team Members:

Maryam Amer

Romeo Bacus

Mitch Deighan

Ryan Dudley

Bradlee Garant

Gaston Gatete

Corey Guilbault

Colby Henson

Mark Mintzer

Duong Nguyen

George Michael Portillo

Krista Saunders

John Shaw

Christopher Venning

Advisors:

Ayodeji Demuren

Xiaoyu Zhang

Table of Contents

List of Figures3

Abstract4

Introduction5

Methods7

Preliminary Results11

Discussion 18

Appendices19

References35

Table of Figures

Figure 113

Figure 219

Figure 320

Figure 420

Figure 521

Figure 621

Abstract

Unmanned Aerial Vehicles (UAVs) play a huge role in today’s society and continue to rise in popularity. However, the current life expectancy of a battery is a major limiting factor on flight time. Fuel cells are of particular interest because of their potential to improve the long-range flight performance of small-scale electric unmanned aerial vehicles With the help of a fuel cell, battery life could be increased by only using the battery for peak power demand while the fuel cell is used for cruising. The aerospace community takes special interest in high endurance fuel cell UAVs because of their great potential in telecommunications, reconnaissance, and remote sensing missions. To address the scarcity of relevant data, and to serve as a proof of concept, a commercially available airframe will be retrofitted with a fuel cell based propulsion system. Thus far selections have been made regarding the airframe, the fuel cell system, and the fuel storage system. Integration and testing of the subsystems is underway and the final stage of the project will consist of integrating, assembling, and testing the fuel cell UAV through a flight demonstration.

Introduction

Unmanned Aerial Vehicles (UAVs) are used by both the military and civilian sectors. UAVs are used for surveillance and reconnaissance purposes, traffic monitoring, pollution control, meteorological data collection, pipeline survey, early fire detection, wildlife population tracking, data collection for precision farming, etc. [1].

In recent years, climate change has been one of the most debated topics across the board. A large percentage of the harmful chemicals currently emitted in the atmosphere corresponds to the transport sector. In particular, aviation-generated carbon dioxide is projected to grow approximately 6% by 2050, due to the increase in global travel demand [2]. Fuel cells can be a clean alternative for their use in the aeronautic sector. Due to the depletion of petroleum energy, many studies have looked at alternative energy sources. Fuel cell systems have become popular and attractive because they have a great potential for high efficiency [3]. Moreover, hydrogen used as fuel to power human made mechanical devices is a cheap, sustainable, and clean source of energy.

As used in the military, fuel cell UAVs and associated technologies provide advanced capabilities from air, ground, and logistics equipment [4]. Using a single fuel cell can extend UAV flight times, provide continuous power, and reduce weight of the UAV [5]. Fuel cell power offers a great number of benefits over other more conventional power supplies.

When it comes to operating a UAV powered by a fuel cell, the advantages are quite apparent. Compared to the more common UAV power source batteries, fuel cells offer much longer operating times allowing for record breaking flight durations [6]. In order to increase the operational time of a fuel cell powered UAV, one only needs to increase the amount of fuel and not increase the capacity of the unit as you would with a battery powered UAV [7]. Unlike ground vehicle applications, aviation applications have seen much less application and testing, and comes with more challenges associated with them, such as increased weight and weight distribution [8]. High endurance unmanned aerial vehicles are needed in a variety of different areas and can offer better performance in many of them [9].

Although it’s not a new concept, fuel cell integration aboard autonomous aircrafts presents various design challenges. The power demand of UAV propulsion systems fluctuates similar to a traditional flight vehicle. Hybridizing the propulsion system with a small conventional battery allows for a smaller, less expensive fuel cell. The fuel cell and battery hybrid system allows the UAV to address power fluctuations during peak demands experienced at takeoff and landing while alternatively having the opportunity to recharge the battery when under low demand times such as cruising [10, 11]. In a flight vehicle, size and weight are critical because they affect lift and maneuverability. The small size and weight of the UAV makes these factors even more critical. The weight of a fuel cell can be a big design factor when considering their use in UAVs as it requires more lift to get off the ground [12]. For fuel cells to be successfully integrated into UAVs, a launching system is necessary to produce enough speed needed for takeoff while conserving power.

Fuel cells can provide the constant, long-term power needed by long distance UAVs and can be incorporated into a UAV’s subsystems. The lack of designs and experimental data involving the implementation of fuel cells with aircrafts calls for more research. Therefore, the purpose of this project was to prove the concept of a converted battery powered UAV to a fuel cell powered UAV, mitigate the weight of the fuel cell system, perform ground tests, improve the power delivery system to leverage the power demand between takeoff and cruising, and increase the flight time of the electric powered UAV.

Methods

Nomenclature

α Angle of attack (degrees)

CDDrag coefficient (non-dimensional)

CLαLift coefficient vs. α slope (deg-1)

CmMoment coefficient (non-dimensional)

CoGCenter of Gravity

CoPCenter of Pressure

D Drag (N)

δfFlap deflection (degrees)

Iyy Inertia about the pitch axis (kg-m2)

m Mass(kg)

t Time(seconds)

T Thrust (N)

ρ Air density (kg/m3)

V0Launch Speed (m/s)

XcpCenter of lift chord wise location (mm)

Completed Methods

Our team has split up into four sub-teams that work collaboratively due to the complexity of the system. These sub-teams are flight management, fuel cell, airframe, and catapult. Three of these sub-teams are interdisciplinary, containing mechanical and electrical engineering students.

The fuel cell sub-team was in charge of the power management of the hybrid system, along with the theory, basic testing and wiring, and refueling the hydrogen supply in the Hydrostiks. One of the most notable attributes of fuel cell usage was the ability to cleanly convert energy while operating on pure hydrogen. There were many different types of fuel cells to choose from, but our team used the Polymer Electrolyte/Proton Exchange Membrane Fuel Cell (PEM). The PEM operated by using a plastic electrode to move protons from the anode to the cathode. The power to weight ratio was a high contributing factor to maximize the performance of the fuel cell UAV. This concept was brought into the design due to studies that fuel cell powered UAVs were characterized by highly efficient airframes and propulsion systems, low weight structures, and low power payloads. It was more effective to use a hybrid system in high power consumption regions since the operating voltage of fuel cells is variable and decreases as the load increases. The system can be modified and reduced based on the required power output from the battery needed to operate the fuel cell controller. The number of cells and the cell area could resize the fuel cell system since one stack outputs one volt of power. The creation of an integrated fuel cell hybrid system to increase flight time of a battery powered drone was needed in order for UAVs to maintain relevance in today’s society.

The typical solution is a high aspect ratio (AR) wing, however achieving an efficient and reliable aerodynamics platform with desirable AR is difficult. A carbon-balsa-carbon composite can support high ARs in a sailplane type platform, yielding cruising capability at 80W and 5kg of payload. The AR of this UAV was 16.27 with a 5.5m span, for performance of a smaller, more maneuverable and transportable UAV with a similar flight envelope, further investigation is required. For this project, an off-shelf Blended Wing Body (BWB) flying-wing airframe with reflex airfoil stabilizers and swept wings was considered. Using XFOIL software and verified with wind tunnel tests, reflex airfoils compared to cambered in related application can significantly reduce pitching moment (Cm0) in flying wings, however, CL max also decreased. The local chord-wise distribution of Cl is elliptical for BWB[6], which increases Cm0 on the lever arm of a swept wing. Further research into the combined effect of BWB and reflex airfoils on static aerodynamic stability is necessary. In order to achieve this, the airframe sub-group obtained aerodynamic data using XFLR5 and CFD software to optimize stability and performance of our selected airframe.

Fuel cells underperform by the metric W/kg[1] compared to conventional batteries leading to poor performance during takeoff or climbing. A fuel cell weighing around 1300 grams that delivers 100 watts of power would be able to support a delta wing UAV for significantly longer. However, it would only be possible if a launching system was designed for reliable takeoff. Over the last decade, many industries have been researching the use of drones for various applications. Many of these applications would require a consistent and uniform launching system that could launch drones consistently without human intervention. The catapult subgroup created a launching system that was portable, reliable, and able to provide a consistent increased launch speed to reduce the hydrogen power needed at takeoff.

The flight management subgroup is responsible of piloting Skywalker X8 delta wing airframe, an essential responsibility due to a limited budget. To do this, each member of flight management is going through various training to become familiar with flying a UAV. Flight management started their training with the Parrot AR Drone 2.0. The Parrot AR Drone 2.0 is a quadcopter that is controlled via smartphone or tablet with a Wi-Fi signal. The quadcopter has an automated landing feature that had been used multiple times during flight practices. With the additional weight of 50 grams, the quadcopter miscalculated the distance from the ground and crashed, rendering it non-operational. Under the advisement of Dr. Drew Landman, the sub-group continues its training on the flight simulator in the Phoenix Lab located Kaufman Hall. The flight simulator has various models of drones to fly including a drone similar to the Skywalker. The Skywalker X8 delta wing airframe is a much more difficult drone to fly than the previous drones. Flight management has endure a significant amount their pilot training over Old Dominion University’s winter break to account for the difficulty.

While each team is responsible for ensuring the completion and success of each required task, collaboration will be required to achieve the optimal integration and performance of each subsystem as well as a successful flight demonstration.

Proposed Methods

For the testing phase of the project, it is planned to first fly the manufacturer recommended battery powered configuration of the UAV. This will provide baseline experimental data for what the exact power requirements are for sustained flight with the initial loading. Data on how much power the initial system produces under normal conditions will also be collected during this test. The initial setup will then be loaded with additional weight to simulate the weight of the fuel cell components that are planned to be in the UAV. This will allow for experimental data to be collected regarding the effects of weight on flight endurance.

The fuel cell will be tested for operation upon receipt in the fuel cell lab using a horizon fuel cell trainer kit. This test will involve making a basic circuit with a variable resistor and using a multimeter to read the voltages and current of the output. Then the fuel cell will be integrated with the electric motor and the electric motor will be tested for operation with the fuel cell power. This test will involve running the electric motor powered by the fuel cell and verifying that the motor operates within the limits of the manufacturer.

Once it is verified that the fuel cell and electric motor operate properly, they will be integrated with the selected aircraft. A ground test will be performed to verify proper operation of all the flight controls and electric motor. This test will involve running the fuel cell to operate the aircraft on the ground and checking all the systems. Once this ground test is complete, the aircraft will make its first test flight.

Take-off Thrust

At this point in the project it is unclear if a more powerful motor is required. In the wind tunnel (Kaufman Hall, Old Dominion University), thrust experiments will be conducted. The current setup is a 14X10E Propeller and 770kv E-Flite motor (Horizon Hobby, Inc., Fieldstone Road, Champaign, Illinois, USA). This and two other set ups will be tested, namely with the 13X10E and 12X10E propeller variations. Thrust and power data will be collected for each propeller at full throttle and air speeds varying from launch speed to steady flight speed, approximately 10 – 40 mph. At steadily flight speed the throttle will then be incrementally reduced to zero percent. From this data, the propeller motor combination can be optimized, efficiency estimates will be update, and a decision regarding motor replacement can be made. Whether or not the motor can produce sufficient thrust will be determined from the following force relationship.

The value of CLα will be determined using a software model for a maximum flap deflection of -20 degrees. This simulates the pilot’s attempt to climb after launch. The drag coefficient will be determined simultaneously. By solving the nonlinear ordinary differential equation (existence of a solution is yet to be determined), then using the acquired thrust data it can be determined what launch speed or thrust increase (i.e. motor size) is necessary to reach steady flight speed.

Wing Reinforcement and Transient Stability

Overall few predictions regarding the UAVs transient stability have been attempted, as of yet no undesirable transient flight characteristics have been observed. However, at moderate to high flight speeds it does demonstrate a degree of aeroelastic instability, characterized by violent wing oscillations or “flutter”. The speeds at which this phenomenon occurs need to be determined experimentally, if they are within the range need to generate sufficient lift, the wings many need to be reinforced in order to prevent the dislodging of critical components.

Preliminary Results

XFLR5

Figure 1

XFLR5 is an extension of the widely use XFOIL software, which provides a 3D analysis capability. A wing can be defined using cross-sections, on which an analysis using either the Lifting-Line Theory (LLT) or the Vortex Lattice Method (VLM) can be conducted. For higher accuracy the manual recommends VLM. However, considering the limitation such as absence of flap and winglet and the underlying VLM small angle assumption, result are interpreted with caution. For our purposes XFLR5’s VLM capability provides a qualitative look at the aerodynamic trends of the Skywalker X8 airframe.

Modeling and CFD

Continuing an aerodynamic assessment of the Skywalker X8 airframe, certain results from the XFLR5 software were reevaluated for accuracy. The purpose of this reevaluation was to strengthen the previous findings that this airframe is suitable for H-100 series fuel cell along with its control and hydrogen supply components, in particular. Primarily there were two sets of data that were the focus. 1. The location of the airplane’s average center of lift (xcp) as a function of α. This data could not be accurately collected from XFLR5 because of limitations with modeling the wing geometry. 2. The airframes total drag as a function of α. This is information is critical when designing any aircraft where power is the basic limitation. Also it is particularly difficult to estimate for the blended body, reflex airfoil combination. The unique overall design of the Skywalker X8 has the unforeseen drawback of being incomparable with much of the published wing aerodynamic data and empirical methods for drag estimation.

The method chosen for obtaining the desired airplane data was modeling and CFD analysis using SolidWorks (Dassault Systèmes SolidWorks Corp., Waltham, Massachusetts, USA), which seemed to offer the most user friendly interface while still maintaining a high level of accuracy. The airframe was first modeled in three parts, the wings, the fuselage top, and bottom. The complex, stream-line surfaces of the blended wing and body as well as the internal structure were captured from approximately 30 cross-sections at varying half-span locations using Rhinoceros (Robert McNeel & Associates, Woodland Park, Seattle, Washington) and imported into Solidworks. Once complete, the CFD mesh capturer and solver were used to query results for the UAV at different values of α. Results pertaining to lift and torque were then processed in a spreadsheet in order to extract the desired information.

Internal Layout and Static Stability

Compared to the XFLR5, the CFD data is considered to give a more accurate prediction of the basic drag-lift relationship for the airframe. With this information and a fully defined model of the airframe, it was helpful to now model the UAVs fuel cell power plant and other internal components. This simplified the internal layout optimization process. Not only could the center of gravity and inertia values for different layouts could be calculated in a matter of seconds, but visualization also help to avoid overlooking any critical design issues.