Abstract Two years ago, a team of students at Parks College began the design and construction of the schools first Unmanned Aerial Vehicle (UAV). Since that time, the project has gone through many phases of design: including conceptual design, preliminary wind tunnel testing, construction of scale prototype, wind tunnel flying quality verification testing, and finally into the flight test phase. The project began with a goal of proving the concept for a long range tactical unmanned bomber. Since that time, the 1/8 th scales prototype of the conceptual design has even drawn interest as its own micro UAV. There is currently a very lucrative market for small unmanned vehicles, both air and ground, that can perform military tasks such as surveillance and ordinance delivery. The Association for Unmanned Vehicles Systems International (AUVSI) Student Competition poses a great opportunity to gain experience in an area of engineering that will thrive for years to come. The Parks UAV is a very unique design that employs cutting edge technology to achieve a high speed surveillance flight platform. By combining large aircraft technology with cutting edge radio controlled aircraft components, it was possible to create a vehicle that is small in scale but advanced in ability. The aircraft is equipped with twin turbine engines, pneumatic retractable gears and proportional disc brakes, an autonomous flight control system, and utilizes a completely composite airframe. This report documents the design of the aircraft, construction details, an analysis of the test data, and a description of the flight control systems. Figure 1 Computer Aided Drawing of Parks UAV 2
Table of Contents Introduction 4 Objective 4 DESIGN Design Concept 5 Wind Tunnel Testing 5-7 Performance Analysis 7 Stability Analysis 7-8 CONSTRUCTION Composite Material Construction Process 8-11 Systems Integration 11-13 METHOD OF AUTONOMY Autonomous Control System 13 Control Analysis 14 CONCLUSION 14 APPENDIX 14-2 3
Introduction The AUVSI Student UAV Competition poses college students with the challenge of designing an unmanned aircraft that can autonomously navigate several waypoints and record visual images of the targeted areas. Aside from the main objective of the competition, points are also awarded for autonomous takeoff and landing capability, ability to update waypoint navigation while the aircraft is in flight, and overall aircraft performance. Parks College s entry to this year s competition is a design that began as a proof of concept prototype design project two years ago. The UAV is a non-conventional configuration that was originally designed for high speed, high altitude, long range surveillance and ordinance delivery. Since the AUVSI Competition requires a aircraft that can perform autonomous surveillance, the design is well suited as an entry into the competition. This report details the design, construction, analysis, and testing that was done on the UAV. The first section outlines the design criteria created by the design team as well as the methods of design that were used to achieve the goals. Much of the design work that was done is outside the scope of this document; however it is important to outline the performance and stability characteristics of the aircraft, which is also covered in the design section. The second section of this report details the construction of the Parks College UAV. The airframe was construction using primarily composite materials. Carbon fiber is the primary material, chosen for its high strength to weight ratio and ease of creating complex geometry. Much of the construction time was dedicated to system integration. The UAV has many complex systems that must work together to successfully fly the aircraft. Beginning with the propulsion system, the aircraft is powered by twin turbine engines, each with its own engine control system. Other features are the pneumatic retractable gears and pneumatic proportional main wheel disc brake, data acquisition systems, and of course the autonomous flight control system. Finally, as the objective of the competition revolves around successful autonomous flight, the last section of this report is dedicated to detailing the autonomous flight control package, analysis of the response of the UAV to totally autonomous control, and an analytical stability analysis performed to determine proper aircraft controller parameters. Matlab and Simulink were used extensively to analyze the aircraft control requirements. The key to the success of this project has been careful attention to detail. Numerous tests have been completed to ensure the aircraft will respond as designed when flown. Since it is not only a new design, it is also a very unconventional design, all of the analysis that was done was adapted as best as possible for an unconventional configuration, and then compared to the test data for confirmation. Objective The objectives of this report are to outline the goals set forth by the AUVSI Student UAV Competition and outline the design of the Parks College UAV. It should become clear why some of the unconventional features of the aircraft were utilized and why the Parks College entry will be successful in achieving the goals of the competition. 4
Design_- Design Concepts \ Park s UAV design concept was based on the premonition that surveillance and ordinance delivery could be accomplished in one mission by one unmanned aircraft. Current military tactics include long duration of surveillance by and aircraft either manned or unmanned followed by a second aircraft being called upon to deliver the ordinance. This means two aircraft over enemy territory, two aircraft burning fuel, and two pilots in harms way. The main objective of the design was to create an airframe that was capable of carrying a large bomb payload, travel to an area of interest at a high speeds, loiter over that area while acquiring necessary information about the territory that could be transmitted live back to a base, and finally when a target was located the aircraft can deliver the final assault before returning back to base. allowing it to be a valid design concept for the AUVSI UAV Competition. With high speed capabilities, effective maneuverability, and versatile payload options the Parks UAV is capable of performing the objectives outlined by the competition in a quick and efficient manor. As discussed earlier, much of the design work done is outside the scope of this document because it related to the full size conceptual design and is very detailed. However, a great deal of design work was also done at the prototype scale as well. Again, much of this work is outside the scope of this document, but the next few sections will outline much of the analysis that was done. The competition calls for an effective means of performing surveillance analysis over a given airspace. Therefore, the main area of design interest is the performance of the aircraft to demonstrate the speed and effectiveness with which the UAV will cover the course, and the stability of the aircraft giving it the ability to record accurate surveillance data. The rest of the design section of this report will detail the testing as well as analysis related to the discussed design criteria. Design - Wind Tunnel Testing Figure 2 Concept of missions performed by the Parks UAV. Upon completion of the conceptual design, all the parameters of the aircraft were scaled down to create a proof of concept prototype. The prototype is a 1/8 th scale to the conceptual design and models its geometry, thrust to weight ratio, and mission effectiveness. Since the prototype modeled the mission objective of the conceptual design, it is essentially a miniature UAV with a very stable surveillance platform, Before the performance and stability analysis can be covered, it is important to understand the wind tunnel testing that was done to acquire the true aircraft parameters used. Three main wind tunnel tests were performed to develop a basis for analysis that would return accurate results. Since the aircraft is non-conventionally configured i.e. blended lifting body, swept forward wings, tailless, combined with the low Reynolds number on the prototype; many standard means of analytical analysis proved to be inaccurate due to a lack of information about the true behavior of the aircraft. The solution to the issue was to develop a set of 5
tests that would provide the necessary information about the aircraft so the analysis could be done accurately. The three tests that were performed were first a small scale (1/25 th of the original conceptual design) model tested in the low speed wind tunnel at Parks College, second the engines that are used in the prototype were tested in the low speed wind tunnel at Parks College, and finally the prototype itself was tested at the Wichita State University NIAR Aerodynamic Lab Walter H. Beech Memorial low speed wind tunnel. The first test of the wind tunnel model was designed to determine the relationship of the wings to the blended wing body. Since the model was small and the wind tunnel was low speed (-14 mph) precision data about the aircraft, such as lift coefficient and drag coefficient would be very inaccurate without extensive Reynolds number corrections. Therefore, the objective, mostly out of curiosity, was the contribution of lift by the lifting body fuselage. The body itself covers much of the wing planform because of its blended nature, thus it must be designed as a lifting body to compensate for the lost wing area. This is where the airfoil type shape of the fuselage is derived from. Figure 3 1 st Low Speed Wind Tunnel Test Alpha sweeps at zero degrees yaw were performed with the wings attached and wings removed. As can be seen in Figure 3, trip strips were also used to simulate higher Reynolds numbers to make the data as accurate as possible. Although in terms of numerical results the data wasn t accurate, the relationship of lift with and without wings was accurate. This relationship is plotted in Figure A-1 of the Appendix. It can be seen that the fuselage produces about 2% of the overall lift of the aircraft. These results verified the fuselage design was effective and the blends performed as expected. The second wind tunnel test that was performed was designed to determine the exact performance of the engines used on the prototype. The twin turbines used on the Parks UAV are commonly used by scale radio control airplane enthusiasts. Jet engine performance is well understood on a large scale, however at a small scale dealing with such low Reynolds numbers, the performance characteristics are not well documented. In order to ignite, fuel must be under a certain amount of compression, thus the first stage of turbines engines is generally and axial compressor. The diameter of the compressor on the small turbines is usually about 4 inches, a fraction of the size of a full size jet engine. Since the fuel for these engines must be under the same amount of compression as it does for the large engines, the compressor must spin at a very high rpm to generate the same type of compression. For example, the engines on the Parks UAV idle at 5, rpm, higher than some jet engine run full throttle. The issue that arises is the inefficiency of the compressor at such low Reynolds number and such high speed. The small turbines 6
also only have a single stage compressor. Usually, the thrust vs. airspeed profile for a jet engine is a constant, which factors heavily into aircraft performance calculations. Knowing the inefficiencies that existed with the small engines, the thrust profile needed to be plotted to make the performance analysis on the Parks UAV accurate. the Reynolds number of the test matches the actual flight Reynolds number, making the results very accurate as pertaining to the prototype performance. Figure 5 Prototype in WSU Wind Tunnel Figure 4 T-5 Engine in Parks Wind Tunnel The engine was run at 15, rpm (full throttle) and held constant throughout the test. Thrust was measured at 1 mph increments from static to 14 mph. Standard conditions static thrust was determined to be 12.5 lbs per engine. However, because of the described inefficiencies, this thrust constantly declined with airspeed to about 6 lbs of thrust at 14 mph. The results of this test can be seen in Figure A-2 in the Appendix. Since the tunnel was run within the flight speed envelope of the prototype, the results are very accurate. Finally, the last wind tunnel testing that was done was completed after the prototype construction was complete. The aircraft was retrofitted with a mounting bracket that could be fixed to the Walter H. Beech wind tunnel at Wichita State University. The tunnel has a 1 ft cross section, and again, the aircraft could be run within its actual flight speed envelope, thus In this testing, alpha sweeps were performed at zero and four degrees of yaw with varying control configuration; clean, full aileron, full elevator, full split rudder, and full flaps. The array of precision data acquired during this testing was used to verify the actual aircraft stability and control and derive necessary parameters such as lift and drag for performance analysis. All the tests were run at constant dynamic pressure to aid in the ease of analysis. Plotted lift and drag can be seen in Figures A-3 and A-4 in the Appendix. Also, tabularized stability results for one angle of attack can be seen in Table A-1 in the Appendix. It was determined that the aircraft possessed desirable handling qualities and demonstrated stability on all axes. The first test was performed by simply running the tunnel up in 5 mph increments until the aircraft produced enough lift to take off (approx 4 lbs). All of the subsequent tests were run at the dynamic pressure (15 psf) that related to this takeoff speed of 8 mph. Further analysis of this data is seen in the performance and stability analysis sections. 7
Design Performance Analysis With wind tunnel testing completed, an accurate platform for doing performance analysis was available. Through testing the engine, the available thrust at all airspeeds was known, and all through the large scale wind tunnel testing, lift and drag was known at takeoff speed. All the performance calculations done in this section are done at 8 mph, the approximate liftoff speed of the aircraft and assuming standard sea level conditions as Webster Airfield in Maryland (location of the competition) is only 21 ft above sea level and the competition is in the summer. In the Appendix, Figure A-3 is a plot of the wind tunnel data for lift and drag coefficients. From this plot, thrust required at takeoff can be determined because is: T R := WC D C L Eqn. 1 Available thrust is taken from the engine wind tunnel test, Figure A-2 in the Appendix. By estimating airspeed as: HP A HP R RC := 55 6 W Eqn. 3 The performance characteristics of the Parks UAV gives it the ability to take off and climb to altitude very quickly and immediately enter the flight profile, cruising at a high velocity. Since speed is one of the requirements, as the course must be completed in under 4 min and points are awarded for quickness, the Parks UAV has a favorable performance profile. Design Stability Evaluation Although alternate control methods were utilized, the prototype wind tunnel testing and prior analysis has shown the aircraft to be stable in all flight regimes. Pitch is controlled by an all moving stabilator/thrust vectoring system. Air is ducted in through inlets on each side of the fuselage then blown over the hidden tail with an electric ducted fan to create a blown tail. The system can be seen below in Figure 6. Eqn. 2 Since the values were all taken at zero angle of attack and 8 mph, thrust required and thrust available can be plotted against airspeed, seen in Figure A-5. From the figure, it can be seen that the aircraft maximum speed is around 15 mph. Figure A-6 in the Appendix shows the rate of climb with and without flaps. The Parks UAV best climb is approximately 275 fpm at around 9 mph. The rate of climb was calculated using the following equation: Figure 6 Tail Fan System Roll control is achieved using a standard aileron configuration, however split drag rudders are used for directional control. The control scheme was chosen because of 8
the swept forward wing design and a desire to eliminate the vertical radar cross section of the conceptual aircraft. There is some vertical surface at the back of the fuselage where the ducted fan exhausts, and inherent lateral stability was designed into the fuselage. Figure A-7 in the Appendix is the Cm vs. Angle of Attack plot for the aircraft. This plot includes effects of the fuselage. It can be seen from the plot that the aircraft will trim at approximately 2.5 degrees AOA which requires approximately 2 degrees of elevator to trim. The neutral point was calculated to exist at.436 of the mean aerodynamic chord using Equation 4. This gives a static margin of.15 for the current center of gravity location. Eqn. 4 Numerous calculations and simulations were set up to determine the exact stability and control properties of the aircraft. Damping and time decay for all stability modes were set up and then verified using the Opinicus Engineering Flight Simulator at Parks College. The simulator is fully reprogrammable and the flight characteristics of the aircraft were modeled in it. Figure A-8-11 are the plant matrices set up for longitudinal and lateral control along with the time rate of decay graphs derived from these matrices. All of the dynamic stability data was verified using the simulation. It can be seen from the time decay plots that all modes of dynamic stability are stable except spiral, which has an inherent necessity for instability. In the final wind tunnel test, it was impossible to test the dynamic stability because the aircraft position is fixed, however it was possible to verify all the static stability characteristics. All static stability was verified in the wind tunnel and the results can be seen in Table A-1 in the Appendix. The results show effective control authority and static stability at zero degrees angle of attack. This data was analyzed, with similar results, at all angles of attack. Further evaluation of the aircraft stability will be discussed in the autonomous control section where the aircraft control loops are modeled in Matlab. This allows for the further verification of control, and does agree with the data discussed in this section of the report. Construction Construction Process Fabrication of the aircraft was done using primarily composite materials with honeycomb composite substructure. The lay-up methods used were developed last year and slightly modified to suit the complex construction of the aircraft. Comparable to industry plug/mold procedures in many ways, the major difference with the procedure is absence of a heat curing stage. Instead, the single stage wet lay-ups are cured at room temperature in a vacuum bag at approximately 1 psi. The use of a wet lay-up procedure was necessary due mostly in part to the absence of a largescale autoclave. Using Rhinoceros CAD software, all the components of the aircraft were modeled and laid out, as they would be during construction. From these drawings, full size templates for the exterior skins of the aircraft were created and used to cut out foam representations of the components. The foam model is inspected and cleared of surface imperfections and then laid-up using bi-directional S-type fiberglass. Using composite fillers, these lay- ups eventually become the male plugs that are used to create the female molds. The plugs are sanded and painted repeatedly and then 9
waxed to create the desired skin texture, in this case as smooth as possible. Once the plug is prepped, it is coated with mold release and using a combination of deck cloth, S-type glass, and glass matting a rigid female mold of the component is created. Both the plug and the mold are cured at room temperature and 1 psi. The primary material used for final component construction is 5.2 oz bidirectional carbon fiber. Carbon fiber was chosen for its lightweight and highly structural material properties. Though each lay-up was tailored differently, all aircraft skins were created from a fiberglass mold, which allows for repeatability in case of a lay-up error or future need for repair. Fuselage Details The fuselage was created using 3 separate molds. First, fuselage halves were made which overlapped an imaginary centerline drawn from nose to tail. Once the molds were created, the plug was sliced into sections in order to create exact templates for bulkhead formers. The formers were traced onto foam, cut out, and laid-up using a 45 x 45 ply setup. Bulkheads were then bonded to one of the fuselage halves. Figure 7 Fuselage Half Once complete, the other half was bonded to the assembly and the center joint was cut, bonded, and filled. The removable bottom panels were created using the third mold. All fuselage skins were manufactured using a x 45 x 9 carbon fiber ply arrangement and reinforced with fiberglass or Kevlar in areas where metal fasteners would be attached. Figure 8 Assembled Fuselage Wing Details Advanced structural tailoring was used in designing the structure for the wing in order to withstand specific forces associated with swept forward wings. The rib pattern is not a standard setup in which the ribs are aligned with the air flow; instead they cross at 9 angles with the joints aligned on the elastic axis of the wing. This is the axis of rotation at which any type of torsional divergence would take place. By arranging the ribs in this manor, the wing is most ridged at the elastic axis, which is favorable to withstand divergence. The main bending structure in the wing is the ¼ chord spar. The spar is an H- boom design created using 4 layers of carbon fiber and two layers of Kevlar. The carbon plies are arranged in a x 45 x 45 x 9 pattern with the two layers of Kevlar between the 45 layers of carbon. The carbon fiber provides both the desired bending and torsional stiffness values required for the swept forward design. Necessary bending and torsional stiffness values were attained using a method 1
developed by Dr. Marty Ferman of Parks College. The method outputs stiffness values based on wing geometry, atmospheric conditions, and flight speed envelope. The addition of Kevlar to the spar provides the spar with the ability to flex slightly without failing. At the root of the wing, the spar is laminated to an aluminum C-channel with a 1/8th inch wall thickness. This was done as a means of fastening the wing to the torque box using standard bolts so that the wings would be easily removable for aircraft transportation. Wing skins were manufactured using two molds, top and bottom. The bottom skin extends around the leading edge and joins the top skin at the ¼ chord spar. Each wing skin has two layers of carbon fiber tailored such that they are arranged in a x 9 relative to the airflow over the wing. The combination of the ply tailoring in the wing skins, the arrangement of the ribs, and the rigid spar, the wing is able to withstand the complete flight envelope of the aircrafts capabilities. UAV are the propulsion system, flight control system, video system, and the pneumatic systems. The propulsion system on the aircraft is comprised of twin T-5 gas turbine engines and their independent engine control system. Each engine has its own Engine Control Unit that controls the fuel mixture and rpm setting during flight. During the start phase, a Start Box Computer is attached to the engine to monitor the start sequence and adjust the engine run parameter. The engines start on propane gas and ramp up until there is sufficient compression to ignite Jet A fuel. The main fuel solenoid is engaged and the engines ramp to the final idle of 5, rpm. At this time the engines are ready for operation. A diagram of the engine system is seen in Figure 1. Figure 1 Propulsion System Figure 9 Aircraft Skins Prior to Assembly Construction System Integration Once the airframe was assembled, integration of the various systems began. During system integration, safety of the aircraft and effective functionality were the main goals. The main systems in the Parks The flight control system of the aircraft is made up of Futaba High Torque Metal Gear Servos, a Futaba 9U Computerized Radio System, and a Micro Pilot MP228g Autonomous Flight Computer. The servos used were chosen based on the necessary force to move the control surfaces. The tail and flaps each have two servos for additional reliability. All control surface linkages and clevises are metal with safety locking springs. No control surface 11
force will ever exceed the amount of available torque dedicated to that surface. All servos were tested in the wind tunnel to verify that they were capable of moving the control surface. The radio system itself provides the extra channels needed to run the auxiliary systems such as camera movement. The MP228g autopilot was chosen because of its ability to easily integrate with the split rudder and independent flaps control scheme of the Parks UAV. Detailed discussion of the system configuration will be discussed in detail in the Autonomous Control Section of this report. The flight data is transmitted through a radio modem at 915 mhz. Finally, the last major system on the aircraft is the pneumatics. There are two independently charged pneumatic systems on board the aircraft. The landing gears are retractable and the main landing gears are equipped with proportional disc brakes. Each system has its own reservoir tank that must be charged before each flight. The landing gear is actuated with a servo that is controlled by the pilot upon takeoff and landing. Figure 11 MP228g Autopilot The camera used on the Parks UAV is a radio control aircraft flight camera. It is fixed to the nose gear, which is retractable. This was done so that on takeoff, the camera view would be forward to give a pilot view from the aircraft and when in flight the camera view is down to perform surveillance. Also, when the autopilot is engaged, the nose gear is independently controlled from the radio, therefore by moving the nose gear the camera can be panned across the ground to gain a panoramic view of a given territory. The camera also sends a live audio feed. The transmission frequency for the video and audio is 2.4 GHz, so as to not interfere with the flight computer transmissions. Figure 13 Retract Pneumatic System The brakes are controlled by an electronic solenoid that is tapped into the elevator channel and landing gear channel. Figure 12 Parks UAV Camera Figure 14 Disc Brakes and Wheels The splice into the landing gear channel assures that the brakes are not used when the gears are up, thus saving air 12
power. The brakes are actuated through the elevator, upon landing, full up elevator (nose down) locks the brakes, and they are proportional with varied up elevator input. Figure 14 Brake Electronic Solenoid Autonomy Autonomous System As discussed in the system integration portion of this report, the main component of the Parks UAV autonomous flight control system is a Micro Pilot MP228g autonomous flight computer. The computer, servo attachment circuit board, and the radio controlled receiver are mounted in a custom made fan cooled carbon fiber case on the aerodynamic center of the aircraft. The autopilot transmits flight information back to a ground control station via a 915 MHz data link. The ground control system is a PC laptop, equipped with the Micro Pilot custom Horizon software, which is used to set up flight patterns, update flight patterns while in flight, and display flight data such as altitude and airspeed. The flight computer software is a system of PID based controllers that take aircraft position and orientation information from a GPS, three axis position gyro, y-axis rate gyro, and a pito-static system. All of the PID gains were calculated using analytical methods with data from the final wind tunnel test. This analysis will be discussed in detail in the next section. The autopilot program is configured to the flight control system of the aircraft. That is, it controls yaw using drag rudders, pitch using elevator, and roll using predominantly ailerons. For added safety, a redundant roll control system was added, that being differential flaps. This was added because when the drag rudder is fully deployed, the corresponding aileron is totally negated. Even though the opposite aileron will still provide roll control, the autopilot will still have positive roll control at all times with the flaps. This also provides a redundant system in case of aileron failure. The engine ECU s are also integrated with the autopilot. They utilize the same pulse signals that a typical servo understands, thus the autopilot sends a standard servo position signal based on desired airspeed, and the ECU transforms that signal into fuel flow rate and desired engine rpm. Nose gear steering, ducted fan speed, and landing gear retracts are not controlled by the autopilot. Although developing a plug-in to control these functions would not be difficult, the aircraft landing and takeoff speeds are 8 mph and faster, and the autopilot is not fast enough to make up for precision errors that need to be corrected rapidly when landing. Since landing and takeoffs will not be autonomous, there was no need to integrate these systems into the autopilot. The following section will detail the analysis of the controls that was done to ensure safe operation of the aircraft. It is important to have precise gains for the PID controllers, understand how the aircraft will respond to autonomous control, and be able to transmit updated commands in case of variable atmospheric conditions. 13
Autonomy System Control Analysis It is necessary to have a detail understanding of the PID controllers set up in the flight computer. If the gains are not set properly, the aircraft will either respond too slow or over respond to position errors. Either of these cases could lead to a crash. Although a complete analysis off the control system was done, the main focus was given to roll, pitch, and yaw control. These systems were modeled as control loops in Simulink, as seen in Figure 15. Figure 15 Pitch, Yaw, and Roll Controls Based on aircraft information attained in the wind tunnel testing and prior analytical calculations, the PID controller gains were calculated using the Ziegler- Nichols method. The PID gains can be seen in Table A-2 in the Appendix. After the PID gains were calculated, the aircraft transfer functions were modeled and overall aircraft response to a step input was analyzed to ensure the PID gains were accurate. Although the autopilot doesn t respond exactly in this manner because it doesn t know the exact aircraft transfer functions, this did give a starting place for setting the PID gains, which will be fine tuned in flight. Conclusion The Parks UAV is capable of completing the objectives of the AUVSI Student UAV Competition. High speed efficient performance with a stable platform is the key that gives the UAV an advantage in the competition. The autonomous control system is adequate for navigation of a GPS waypoint course, and a quick means of reducing visual data from the aircraft is in place. This document has outlined the design of the Parks UAV including the initial concepts for the design, a brief final performance analysis, verification of a stable surveillance platform, and the testing that was done to prove these criteria. Also, a detailed explanation of the aircraft construction procedure and system integration process was covered. Finally, the autonomous flight control package, ground control station, and control analysis was discussed. A careful and detailed approach to the design and system integration was used. The testing and simulations that were done verified all of the design data that was calculated throughout the initial design phase. When scaling data, it was also very important to realize the change in Reynolds Number and therefore change in aircraft response and behavior. The prototype is a representation of the initial concept, but is also its own design. Park s first UAV could fill many demands, both Military and civilian. As the design continues to draw interest, the roll of this versatile design is ever being modified. 14
Appendix Corrected Lift Coefficient vs. Corrected Angle of Attack for the Aircraft vs. Fuselage.6.5 y =.31x +.27.4 Corrected Lift Coefficient.3.2.1 y =.48x +.82. -6. -4. -2.. 2. 4. 6. 8. 1. 12. 14. 16. 18. -.1 -.2 Corrected Angle of Attack (deg) Aircraft Fuselage Linear (Aircraft) Linear (Fuselage) Figure A-1 Results from first wind tunnel test, fuselage lift and aircraft lift plotted Thrust Profile 14 12 1 Thrust (lbs) 8 6 4 2 2 4 6 8 1 12 14 16 Velocity (mph) Poly. (Thrust Profile) Figure A-2 Thrust Profile for a T-5 Turbine Engine, 15, constant rpm 15
Lift and Drag (8 mph) 1.6 1.4 1.2 CL and CD 1.8.6.4.2-5 5 1 15 2 Alpha (deg) Poly. (CL (flap)) Poly. (CL ) Poly. (CD) Poly. (CD (flap)) Figure A-3 Lift and Drag coefficient from wind tunnel test performed on prototype Lift to Drag Ratio (8 mph) L/D 1 9 8 7 6 5 4 3 2 1-5 5 1 15 2 Alpha (deg) Poly. (Flaps) Poly. (No Flaps) Figure A-4 Lift to Drag Ratio from wind tunnel test performed on prototype 16
Thrust Available vs Thrust Required 25. 2. Thrust (lbs) 15. 1. 5.. 2 4 6 8 1 12 14 16 Velocity (mph) Poly. (Thrust Available) Poly. (Thrust Required) Poly. (Thrust Required w ith Flaps) Figure A-5 Thrust Ratio for Prototype, derived from wind tunnel data 3 Rate of Climb Rate of Climb (feet per min.) 25 2 15 1 5 2 4 6 8 1 12 14 16 Velocity (mph) Poly. (Rate of Climb) Poly. (Rate of Climb w ith Flaps) Figure A-6 Rate of Climb for Prototype, derived from wind tunnel data 17
( ) :=.12 +.547 C m α α.2.1 C m ( α) ( α).1.2 1 2 3 4 5 α Figure A-7 Cm vs Alpha plot for prototype, including fuselage effects Figure A-8 Plant Matrix for Longitudinal Dyanamic Stability Calculations Exact Method for Dutch Roll, Spiral, and Roll Mode A := Y β u o L β N β Y p u o L p N p 1 Y r 1 u o L r N r 32.2 cos θ o u o ( ) Figure A-9 Plant Matrix for Lateral Dyanamic Stability Calcuations 18
Phugoid () t := 2 e.177 t Exact() t := 2 e.278 t 2 Phugoid Time Decay 1.5 Phugoid() t Exact() t 1.5 2 4 6 8 1 SP() t 2 e λ real := t Exact() t := 2 e.786 t t seconds 2 Short Period Time Decay 1.5 SP() t Exact() t 1.5 2 4 6 8 1 t Seconds Figure A-1 Phugoid and Short Period Time Decay Plots for Stability Figure A-11 Dutch Roll, Roll, and Spiral Time Decay Plots for Stability 19
Crm (roll moment) Cpm (pitching moment) Cym (yawing moment) Csf (side force) deg of Yaw Clean.1557.191738 -.145594473 -.2671297 Full Spoiler -.13248376 -.365547 -.1412979 -.1219722 Full Aileron.1454212.37723124 -.16655333 -.2812418 4 deg of Yaw Clean.2349965.99433 -.1429249 -.497416 Full Spoiler.415787 -.5727574 -.1532782 -.12446411 Full Aileron.27525187.292642 -.124625776 -.224717 Table A-1 Stability Data from Final Wind Tunnel Test (AOA = ) Roll Pitch Yaw Kp.1229 1.596.7488 Ki.166.185.7286 Kd.23 3.433 1.924 Table A-2 PID Gains for the Parks UAV References 1. Hoerner, Sighard F. Fluid-Dynamic Lift. Mrs Liselotte A. Hoerner. 1975 2. Lan, Dr. Chuan-Tua Edward; Roskam, Dr. Jan. Airplane Aerodynamics and Performance. DARcorporation. 1997 3. Nelson, Robert C. Flight Stability and Automatic Control. McGraw Hill. 1998 4. Raymer, Daniel P. Aircraft Design: A Conceptual Approach. AIAA Inc. 1999 5. Roskam, Dr. Jan. Airplane Design: Part I VII. The University of Kansas. 199 6. Schmidt, Louis V. Intorduction to Aircraft Flight Dynamics. AIAA Inc. 1998 7. Stinton, Darrol. The Design of the Airplane. Pg. 164-18. AIAA Inc. 21 8. Von Mises, Richard. Theory of Flight. Dover Publications Inc. 1959 2