CASI Free Flight Glider Competition Aircraft Design Report By the University of Toronto Free Flight Glider Team Aircraft Design Report The University of Toronto Free Flight Glider Team Team Members: Terri Chu Jeff Kwan Nicolas Lee Jay Loftus Eric Ng Leo Ng Sarah Razzaqi Rav Singh Andrew Sun Danny Yen i EXECUTIVE SUMMARY The University of Toronto's entry for this year's CASI free flight competition builds upon last year's successes. The improved Zephyr II employs carbon fibre/foam wing, carbon fibre fuselage, and balsa wood empennage, while it retains a conventional configuration from the original Zephyr. The limiting factors in the success of the original Zephyr were the lift characteristics, and the robustness of the aircraft. To improve the aerodynamic characteristics, the airfoil of Zephyr II was changed to a modified SA7038 and the wingspan was increased by almost a metre. The component integration methods were also improved to make Zephyr II more robust, and thus it would give more consistent performance. In addition, more refined techniques of analysis in aerodynamics, flight dynamics and structural mechanics facilitated the design optimization process. Overall, the performance of Zephyr II has increased without sacrificing robustness. With improved design and construction methods, our team expects Zephyr II will not only outperform its predecessor, but its competitors as well. ii Table of Contents 1 Introduction ................................................................................................................. 1 1.1 Design Philosophy .............................................................................................. 1 1.1.1 Problem Definition...................................................................................... 1 1.2 Design Factors .................................................................................................... 2 1.2.1 Competition Criteria ................................................................................... 2 1.2.1.1 Self-imposed Constraints ........................................................................ 2 1.2.2 Design Approach ........................................................................................ 3 2 Glider Design .............................................................................................................. 3 2.1 Aerodynamics and Stability ................................................................................ 3 2.1.1 Aerodynamic Surfaces Sizing ..................................................................... 4 2.1.2 Airfoil modification .................................................................................... 4 2.1.3 Finite Wing Correction of XFoil Analysis.................................................. 6 2.1.4 Longitudinal Stability ................................................................................. 7 2.1.5 Longitudinal Motion Simulation................................................................. 8 2.2 Glider Structure and Construction ...................................................................... 9 2.2.1 Material Selection ..................................................................................... 10 2.2.1.1 Wing Material ....................................................................................... 10 2.2.1.2 Fuselage Material: ................................................................................. 11 2.2.1.3 Empennage Material: ............................................................................ 11 2.2.1.4 Cargo Bay Material: .............................................................................. 11 2.2.1.5 Nose Cone Material: ............................................................................. 11 2.2.2 Structural Design ...................................................................................... 12 2.2.2.1 Wing...................................................................................................... 12 2.2.2.2 Fuselage ................................................................................................ 12 2.2.2.3 Empennage ............................................................................................ 12 2.2.2.4 Cargo Bay ............................................................................................. 12 2.2.2.5 Nose Cone ............................................................................................. 12 2.2.3 Wing Structural Analysis .......................................................................... 13 2.2.3.1 Experimental ......................................................................................... 13 2.2.3.2 ANSYS Analysis .................................................................................. 13 2.2.4 Construction Method ................................................................................ 14 2.3 Systems Integration ........................................................................................... 15 2.3.1 Wing Mounting ......................................................................................... 15 2.3.2 Cargo Carriage .......................................................................................... 15 2.3.3 Tow Hook ................................................................................................. 15 2.3.4 Empennage Mounting/Dethermalizer ....................................................... 16 2.3.5 Nose Cone ................................................................................................. 16 3 Design Alternatives ................................................................................................... 17 3.1 Canard Configuration........................................................................................ 17 3.2 Flying Wing Configuration ............................................................................... 17 3.3 Biplane Configuration ....................................................................................... 17 4 Future Considerations ............................................................................................... 18 4.1 Mechanically Assisted Wire Cutter .................................................................. 19 4.2 Computer or Microcontroller Controlled Motorized Wire Cutter .................... 19 4.3 3D Motion Simulator ........................................................................................ 20 5 References ................................................................................................................. 21 iii 6 Appendix A: Load Calculations................................................................................ 22 6.1 Wing Loads ....................................................................................................... 22 6.1.1 Flight Loads: ............................................................................................. 22 6.1.2 Launch Loads: ........................................................................................... 22 6.1.3 Impact Loads:............................................................................................ 22 6.2 Fuselage Loads.................................................................................................. 22 6.3 Empennage Loads ............................................................................................. 23 6.4 Cargo Bay Loads............................................................................................... 23 6.5 Nose Cone Loads .............................................................................................. 23 7 Appendix B: Drawings ............................................................................................. 24 iv Nomenclature Symbol Definition Angle of attack Induced drag factor Downwash Fraction of dynamic pressure at the tail relative to the wing t Glide angle Air density Correction factor for C L calculation Unit rad or degree rad or degree rad or degree kg/m3 AR CD C D ,i Aspect ratio Drag coefficient Induced drag coefficient CL C L Cl _ 2 D Lift Coefficient of a finite wing Curve slope of lift coefficient of a finite wing Lift coefficient of an infinite wing Cm C mo C.G. L.E. M MAC MGC N .P. Re S Coefficient of moment about wing’s ¼ chord Coefficient of moment about wing’s ¼ chord at 0 Centre of gravity Leading edge Mass Mean aerodynamic chord Mean geometric chord Natural point Reynolds number Area kilogram metre metre a ao Curve slope of lift coefficient of a finite wing Curve slope of lift coefficient of an infinite wing Free fall acceleration due to gravity Altitude Length of “x” or Location of “x” from tip of aircraft nose rad-1 or degree-1 rad-1 or degree-1 m/s2 metre metre Time second g h lx t w t Variable associates with the wing / wing-body Variable associates with the tail v rad-1 or degree-1 m2 1 INTRODUCTION The goal of the CASI Free Flight competition is to give students a chance to apply their book-learned knowledge to the designing, building and flying of a model aircraft. Free Flight encourages students to be problem solvers and innovators by imposing constraints on the design that the team must work within in order to be successful. The University of Toronto Free Flight Glider Team is comprised of ten undergraduate students drawn from various engineering disciplines including aerospace, mechanical and electrical. Nearly half of the team has returned from previous years’ glider teams to take up core positions within the group. This year’s entry (Zephyr II) builds upon the performance of last year’s 2004 entry. It was the intention of the group to increase the payload capacity and flight time of Zephyr II without sacrificing robustness. To do this, the hot-wire foam cutting process for wing construction had to be refined. More advanced analysis and modeling techniques, including the use of XFoil to optimize the airfoil, have also been added to our design tools. Our longitudinal motion simulator was custom designed and programmed in house. Backed by these rigorous analysis methods, improved manufacturing techniques, and most importantly, the dedication of all of the members of the flight team, we are confident that Zephyr II will perform to our expectations with consistent competitive flights. 1.1 DESIGN PHILOSOPHY 1.1.1 Problem Definition The team must design and build a heavier than air aircraft, which is tow-launched and which is capable of sustaining flight for the maximum amount of time while carrying maximum payload. Each flight of the glider will be scored according to the formula: FlightScore P t 2 *10 4 The formula suggests that there will have to be a tradeoff made between flight time and payload carried. Ideally, the team would optimize flight time to 75 seconds, and no more, while carrying as much payload as possible. In reality, the team must design the 1 airfoil and structure such that the plane is capable of at least a flight time of 75 seconds, while carrying a significant amount of payload. In the past, designs which have carried close to 1kg and which achieved flight times of 60 seconds or more have been quite successful. This is not to be used as a design objective, merely a benchmark. 1.2 DESIGN FACTORS 1.2.1 Competition Criteria The fundamental, and perhaps most limiting, constraint on the design of the aircraft is that the entire aircraft must fit inside a 2x2x1m box. This constraint most significantly affects the design of the wing, in that the size, planform and dihedral are limited. Also, the stability of the aircraft is impacted by this constraint, as the longitudinal moment arm is restricted. The new requirement to include a nose with a radius of curvature of at least 2 inches significantly affects the aerodynamics of the glider. The mandatory inclusion of a dethermalizing mechanism has the potential to add considerable weight to the aircraft, along with increasing the complexity of the systems integration. 1.2.1.1 Self-imposed Constraints 1.2.1.1.1 Funding A major obstacle faced each year by the UofT team is the uncertainty surrounding the funding of the glider project. For the past two years, the team has been able to count on the generous support of the UofT Engineering Society, but the exact amount of sponsorship is generally not known until mid-March to April. As such, the team set a modest budget of approximately $600 (based on experience), and designed within this constraint. 1.2.1.1.2 Weight Another significant self-imposed constraint was to minimize the weight of the glider. Thus material selection, structural design and construction were all carried out 2 with this in mind. As a benchmark, the team aimed to have a glider with a weight comparable to last years (approx 900g), even though this year’s glider will be 15% larger. However, along with the constraint to minimize weight comes the complimentary constraint to ensure that the glider can survive hard landings and maintain the repeatability of the configuration. That is to say factors such as wing alignment, empennage integrity, angle of attack etc. should remain constant from one flight to the next. The lack of repeatability between flights was a major challenge of last years design, and could well have been the reason the team did not fare better in the flight competition. 1.2.1.1.3 Stability Finally, static and dynamic stability of the aircraft is vitally important in an uncontrolled glider such as we have here. The configuration of the glider, which affected both the location of the centre of gravity and the neutral point, was motivated by the desire to design a stable plane. 1.2.2 Design Approach Over the last two years the UofT Free Flight team has gained a considerable amount of knowledge and expertise in the designing and building of model gliders. The team has chosen to build on the work of previous years, while also striving to be innovative in solving the problems with earlier designs and the constraints imposed by new competition rules. The team goal was to strike a balance between using tried and true techniques and new innovating. 2 2.1 GLIDER DESIGN AERODYNAMICS AND STABILITY In this section of the report, the methods of aerodynamics and stability analysis of Zephyr II are summarized. Aerodynamic surfaces were designed based on a trade-off between aerodynamic, structural and fabrication considerations, while under the constraints imposed by the rules. Airfoil section used in the original Zephyr was redesigned and the associate aerodynamic characteristics were determined. The results 3 were then used for stability calculation, and were inputted into the longitudinal motion simulation code for more detailed study. 2.1.1 Aerodynamic Surfaces Sizing Since the dimensions of glider are limited by the rule’s size constraint, the length of the wing span is set at 2.90 m (sum of all wing panels’ lengths). To ensure roll stability, the dihedral angle is specified to be 15º from previous experience. The wing’s planform is determined based on a trade-off between aerodynamic performance, structural strength and fabrication considerations. The mean aerodynamic chord (MAC) and its location are approximated by those of the mean geometric chord (MGC). The method followed was given in Ref 1 for a multi-panel wing planform with each section having distinct tapering ratio and dihedral angle. The geometry of the wing is summarized in Table 1a. The sizing of the empennage was based on recommended tail volume ratios found in Ref 2. The geometry of the tail was determined with stability and ease of fabrication in mind after studying numerous glider designs found in Ref 3. The design of the empennage was modified slightly after performing various aerodynamic and structural analyses. The geometries of the empennage are provided in Table 1b and Table 1c. Wing geometry Airfoil Section Customized SA7038 Area 0.5205 m2 Aspect Ratio 15.9 Dihedral 15º Root Chord 0.20 m Tip Chord 0.13 m MAC 0.18 m Span 2.90 m Sweep (L.E.) None Tapering Ratio 0.66 Horizontal tail geometry Airfoil Section Flat plate Table 1a Table 1b 0.095 m2 4.27 None 0.15 m 0.15 m 0.15 m 0.64 m None 1.0 Area Aspect Ratio Dihedral Root Chord Tip Chord MAC Span Sweep (L.E.) Tapering Ratio Vertical tail geometry Airfoil Section Flat plate Area Aspect Ratio Dihedral Root Chord Tip Chord MAC Span Sweep (L.E.) Tapering Ratio 0.027 m2 1.51 None 0.15 m 0.12 m 0.13 m 0.20 m 25.4º 0.77 Table 1c 2.1.2 Airfoil modification Instead of choosing a completely new airfoil section, our team decided to modify the airfoil section used for previous year’s entry, SA 7038, for Zephyr II. The goal was to 4 improve the glider performance by manipulating the camber and thickness. Progress is being made in extending our analysis to include fully inverse design. Different combinations of camber and thickness were analyzed using XFoil (Ref 4). The cross section of the improved airfoil, SA 7038 t9 c5 (Note: t9 denotes 9% thickness, and c5 denotes 5% camber), is shown in Fig 1. In Fig 2-4, a comparison of the 2D airfoil aerodynamic characteristics of the original SA 7038 and the improved SA 7038 t9 c5 is given. Fig 2 shows that SA7038 t9 c5 generates more lift than SA 7038 throughout the entire test range. Fig 3 shows that there was no significant increase in drag due to our Fig 1 modification. Fig 4, however, indicates that due to the increased camber, the nose down pitching moment has increased. Thus, a larger downward force is required from the tail to provide the trimming moment. This has a negative effect on the glider’s performance since it decreases the overall lift and increases induced drag from the tail. Nevertheless, in later section of this report, the stability calculations and our self-developed longitudinal motion simulation code showed that the increase in trim drag was small. In terms of maximizing points, the gain in lift due to our modification more than offset the increase in trim drag. XFoil 2D Analysis: Cl Comparsion, Re=100,000 XFoil 2D Analysis: Cd Comparsion, Re=100,000 Xfoil 2D Analysis: Cm Comparsion, Re=100,000 0.1400 0.0000 -10 1.4000 0.1200 0.1000 0.0600 0.4000 0 5 10 15 -0.0400 0.0800 Cm SA 7038 SA 7038 t9 c5 Cd Cl 0.9000 -5 -0.0200 SA 7038 SA 7038 t9 c5 -0.0600 -0.0800 0.0400 -10 -5 -0.1000 -0.1000 0 5 10 15 0.0200 20 -0.1200 0.0000 -0.6000 -10 Angle of Attack (deg) Fig 2 -5 0 5 10 Angle of Attack (deg) Fig 3 5 15 20 -0.1400 SA 7038 SA 7038 t9 c5 Angle of Attack (deg) Fig 4 20 2.1.3 Finite Wing Correction of XFoil Analysis Before the aerodynamic data from XFoil can be used in the longitudinal motion simulation code, correction for finite wing effects had to be made. The effects to be accounted for include: 1. 2. dC L a. d Induced drag caused by wing tip vortices Decrease in lift curve slope, C L The expression for 3-D lift coefficient, CL, was calculated using equation (2.2.2.1), which can be found from Ref 5 and Ref 7 : a0 dC L (2.2.2.1) a d 1 (a0 AR)(1 ) where a0 was the lift versus α slope in 2-D. The parameter τ for a uniformly tapered wing was obtained from data provided in Ref 7. However, no data of τ for a multi-panel wing, like Zephyr II’s, could be found in our technical literature survey. Therefore, our team proposed the following original approximation method. Fig 5 Suppose there exists an equivalent uniform taper wing such that its finite wing aerodynamic coefficients are the same as the multi-panel wing in question. Our estimation method assumes that such wing will have the same planform area, MAC and same span as the original multi-panel wing. Thus, the aspect ratio is also guaranteed to be the same. By equating these values and solving the associated quadratic equation, the root chord and tip chord, and thus tapering ratio, of the equivalent uniform taper wing can be obtained. Using this newly found tapering ratio, τ in equation (2.2.2.1) can be estimated from Ref 7. To correct for the finite wing effect, equation (2.2.2.2) is applied. a Cl _ 2 D (2.2.2.2) a0 The results of the finite wing correction on CL for SA7038, SA7038t9c5 and the CL horizontal tail using (2.2.2.2) are given in Fig 6-8 respectively. 6 Zephyr II's SA7038 3D Aero. Coeff. at Re=100,000 Zephyr II's SA7038t9c5 3D Aero. Coeff. at Re=100,000 1.4000 1.4000 1.2000 1.2000 1.4000 1.2000 1.0000 1.0000 Cl Cd Cm Fitting line 0.8000 0.6000 0.4000 -10 Zephyr II's Horiz. Tail 3D Aero. Coeff. at Re=90,000 0.6000 0.4000 0.2000 0.2000 0.0000 -5 -0.2000 0 0.0000 5 10 15 20 -10 -5 0.8000 0.6000 0.4000 0.2000 0.0000 -0.2000 0 5 10 15 20 -15 -10 -0.4000 -0.4000 -0.6000 -0.6000 Angle of Attack (deg) Fig 6 Cl Cd Cm Fitting line 1.0000 Cl Cd Cm Fitting line 0.8000 -5 -0.2000 0 5 10 15 -0.4000 Angle of Attack (deg) Fig 7 -0.6000 Angle of Attack (deg) Fig 8 To properly represent the drag acting on the wing, equation (2.2.2.3) was used to calculate the induced drag (Ref 5): C L2 (2.2.2.3) 1 AR In equation (2.2.2.3), the induced drag factor, , was estimated from data found in Ref 5 C D ,i and Ref 6 using the equivalent uniform taper wing mentioned above. Due to Zephyr II’s large aspect ratio, extrapolation was required to obtain an estimate of . As a reminder, the CL used in (2.2.2.3) was obtained from equation (2.2.2.2). Results of the drag coefficients with the appropriate C D ,i has been added and plotted along with CL and C m in Fig 6-8. Currently, our team is researching the necessity for C m correction. If such correction is required, an analysis from a vortex-lattice code may be needed. 2.1.4 Longitudinal Stability To achieve longitudinal stability, two conditions must be satisfied (Ref. 8): 1. Positive static margin 2. Positive pitching moment at zero angle of attack ( C mo 0 ). Positive pitching moment can be satisfied by adjusting the angle of the horizontal tail. To guarantee that there is a positive static margin, an estimation of the location of the centre of gravity (C.G.) and the neutral point (N.P.) is required. From Table 2, the C.G. is found to be located at 0.55 m from the tip of the nose. Using the definition of N.P., one can easily derive (2.2.3.1) to find the N.P.’s location. l np 1 C L C L w l w C L t 1 lt S t S t 7 (2.2.3.1) where C L C L w C L t 1 S t S t and is to be estimated with USAF DATCOM method outlined in Ref 8’s appendices. The location of the N.P. was calculated to be at 0.64 m behind the tip of the nose. Thus, yielding a static margin of 0.50 and the first longitudinal stability condition is satisfied. Following the method outlined in Ref. 8, in order to satisfy the second condition ( C mo 0 ), only a minimum of 1º downward horizontal tail angle is required. Thus, Zephyr II can easily be set to be statically stable in longitudinal motions. Dist of components' CM to tip of nose (m) Component's mass (kg) Wing 0.55 0.4 Vertical Tail 1.45 0.03 Horizontal Tail 1.4 0.03 Fuselage 0.7 0.1 Nose 0.06 0.07 Ballast 0.1 0.085 De-Thermalizer 0.7 0.03 Payload 0.55 0.2 Total weight 0.945 CG location from tip of nose 0.55 kg m Table 2 2.1.5 Longitudinal Motion Simulation The development of the longitudinal motion simulation code using C++ began before last year’s competition and it was finally tested and verified in Jan 2005 before the construction of Zephyr II. The purpose of the simulation code is to serve as an economic mean to try out different glider configurations and airfoil sections. The theory of the simulation is based on Ref. 8. The operation of the simulation is outlined in Fig.9 Fig. 9 8 During initialization, the necessary data such as aerodynamic coefficients, glider’s configurations and initial conditions are entered. Then, using a time marching method, the forces, velocities and position of the glider are calculated through successive integrations. As a demonstration of Zephyr II’s longitudinal stability, a simulation was run for the stress case in which Zephyr II pre-maturely detached from the tow hook at 40m altitude at a remarkably low air speed of 5m/s and 3º pitch up angle (none are steady conditions). Fig 10 shows an initial height loss due to low airspeed, but the oscillation in the glider’s altitude eventually flattens out and the flight path stabilizes. Fig 11 illustrates the fluctuation in the glider’s air speed settles as the oscillation in motions dampens out. Finally, Fig 12 shows the pitching motions as a function of time and demonstrates Zephyr II’s natural stability in its attitude. Air speed vs. Time Airspeed (m/s) Height (m) Height vs. Time 60 40 20 0 20 40 60 10 5 0 0 80 20 40 Time (sec) Time (sec) Fig 10 Fig 11 Angle (deg) Ang Vel.(deg/s) 0 15 20 10 0 -10 0 -20 -30 60 80 Euler Pitch Angle and Ang Velocity 20 40 60 80 Euler Pitch Ang. Ang. Velo Time (sec) Fig 12 2.2 GLIDER STRUCTURE AND CONSTRUCTION In this section we will determine the glider structure and composition, based on our design objectives. In terms of determining structure and composition the most important criteria are the following: strength, weight, cost and experience. For strength, our glider must withstand loads associated with launch, flight and landing. For weight, 9 our glider should be as light as possible and definitely less than 900g. For cost, our total glider cost should be less than $600. Finally, our team must be experienced, or at least confident, with the construction method and materials that we use. Objectives Main Wing Fuselage Empennage Cargo Bay Nose Cone Miscellaneous Max Load† 15 Nm 680 MPa n/a 9N 45 MPa n/a Max Weight 500g 100g 100g 100g 50g 50g Max Cost $350 $50 $50 $50 $50 $100 Experience* 4 2 3 1 1 n/a † See ‘Appendix A: Load Calculations’ for the derivations *A rating from 1-5 of how important experience with the materials and construction is, for a given part of the glider (5 being most important). Table 3 In addition, during construction we will attempt to minimize excess weight and maximize the accuracy of construction. In this way our glider will perform as close to our theoretical predictions as possible. 2.2.1 Material Selection In this section we want to choose materials for each part of the glider in order to meet the previously stated objectives. Material Properties Material Cost [$/kg] Density [kg/m3] Aluminum Steel Carbon Fibre Balsa Spruce 2 0.6 240 2700 7750 1570 Elastic Mod. [GPa] 73 207 190 1 0.7 200 400 3.4 10.8 Tensile Strength [MPa] 90 380 810 Yield Strength [MPa] 48 207 200 Specific Yield Strength 0.018 0.027 0.13 n/a 76 22 40 0.11 0.10 Note: The highlighted entries are the ‘best’ for that column. Table 4 2.2.1.1 Wing Material From the material properties we can clearly see how carbon fibre has the best strength to weight ratio (specific yield strength). A carbon fibre wing would be the 10 lightest possible for a given strength. And although the cost of carbon fibre is much more than the cost of the other materials we plan on using at most 400g, which will cost only $95. Furthermore, our team is experienced with carbon fibre construction, since our glider wings last year were of the same construction. 2.2.1.2 Fuselage Material: For the lightest fuselage for a given strength we will again use carbon fibre. In fact, discarded carbon fibre golf club shafts can be obtained for free and are suitable for our fuselage. 2.2.1.3 Empennage Material: Since the empennage will experience very small stresses and moments, strength is a minor issue, so we will choose the lightest material, balsa wood, for its construction. In addition, members of our team are familiar with balsa wood construction. Keeping the empennage light is also important in maintaining a positive static margin. 2.2.1.4 Cargo Bay Material: For the cargo bay we will use an aluminum can since they are light, cheap, and strong enough for a tow hook. We will also use foam to create an aerodynamic shape around the cargo bay. 2.2.1.5 Nose Cone Material: We will use foam for the nose cone, in order to meet the nose cone requirements. An aluminum sheet embedded in the nose cone will prevent the fuselage from puncturing the nose cone when the glider hits the ground. Final Material Selection Main Wing Fuselage Empennage Cargo Bay Nose Cone Carbon Fibre Carbon Fibre Balsa Wood Aluminum with Foam Foam with Aluminum Table 5 11 2.2.2 Structural Design This section contains the finalized structural design of the wing, fuselage, and empennage. 2.2.2.1 Wing The wing will be composed of a foam core, with a balsa wood trailing edge, with a carbon-fibre/epoxy outer shell weighing 5.8oz/1800in2 (0.142kg/m2). Fig 13 – Wing Cross Section 2.2.2.2 Fuselage The fuselage will consist of a pre-made carbon fibre shaft with a 1cm2 crosssection. 2.2.2.3 Empennage The empennage and horizontal stabilizer will be made of a frame of balsa wood, coated with heat-shrinking plastic wrap like Monokote. Fig 14 - Empennage 2.2.2.4 Cargo Bay The cargo bay will be constructed from an aluminum cylinder to house payload, with foam for aerodynamic shaping. 2.2.2.5 Nose Cone The nose cone will be made of foam, with an aluminum sheet to prevent the fuselage from puncturing it, upon impact. 12 Fig 15 - Nose Cone 2.2.3 Wing Structural Analysis 2.2.3.1 Experimental Based on our moment loading on the wing (Appendix A: Load Calculations), the most likely location for wing failure is mid-span. Thus, we will focus our analysis at that location. We used the glider wing from last year to perform the structural analysis, since this year’s glider wing is of similar construction. We applied a measured force 28 cm from the center line, until the breaking point. The wing began to buckle at 32 lbs (140N) of applied force, or a moment of 39Nm. Clearly, our wing from last year was able to easily withstand the 15Nm from launch. This year we plan on reducing the weight of the wing by eliminating the spruce spars, and carbon fibre trailing edge of last year’s design. 2.2.3.2 ANSYS Analysis A finite element structural analysis was performed on the wing using ANSYS. The model was constructed using 1774 4-node shell elements representing half of the wing cantilevered at the root. It was loaded with a uniform upward pressure on all surfaces, as a conservative approximation of the actual lift the wing will experience. The total lift load was set at 21.582 N, to represent a 1 kg plane under a 4.4g pull-up, as specified in (Ref 9). Under this load case, the maximum von Mises stress was determined to be slightly less than 50 MPa, which is lower than the carbon fibre wing’s yield stress of 200 MPa. The stress distribution on the upper and lower wing surfaces is shown in Figure 16 13 Figure 16 2.2.4 Construction Method The main glider wing is manually cut out of foam using a manual hotwire. Several foam wing sections are cut using airfoil templates as guides for the hotwire. Each section was then glued together by epoxy and coated with carbon fibre using the vacuum bag method. The cargo bay of the glider is made out of an aluminium cylinder cut to size. Epoxy and balsa wood fillets are used to fix the cylinder to the fuselage. The rest of the cargo bay, including the planform for the wing, is cut out of foam using a manual hotwire. The empennage was constructed using a balsa wood truss which is coated with heat shrinking plastic wrap. The empennage is connected to the cabin with the graphite shaft fuselage. The nose cone is made of a hemisphere of foam, with a tapered trailing edge, and a small aluminum insertion. Only after careful consideration of the objectives of our glider design, we determined the most suitable materials, and structural design to use. The carbon fibre wing and fuselage will have the largest strength to weight ratio. The aluminum and foam cargo bay will be cheap, strong and light and can support the tow hook. The balsa empennage will be light and the foam nose cone will conform to requirements. 14 2.3 SYSTEMS INTEGRATION 2.3.1 Wing Mounting In the mounting of the wing, the team has decided to continue to use elastic bands to secure the wing to the fuselage, allowing the wing to pop off on impact, which prevents significant damage. This technique has worked well for the past two years, thus there was no need to change it. However, the team did suffer some problems with wing alignment in previous designs. To solve this problem a small piece of balsa wood was mounted in front of the wing in order to align it with the empennage. This is a simple and light solution, which successfully addresses the repeatability constraint identified earlier. 2.3.2 Cargo Carriage The cargo carriage is located under the fuselage, exactly at the centre of gravity of the glider. The carriage is simply an aluminum can, with a foam cap for safety, and foam fairing for aerodynamic purposes. The payload itself is a small cylindrical disk of lead shot and epoxy. This payload comes in a range of weights and sizes appropriate for this particular glider. Cylindrical foam disks are used as plugs of sort to ensure that the payload is secure in the container during flight and upon impact. 2.3.3 Tow Hook The dynamics of the glider during tow are actually quite difficult to analyze, constraining a degree of freedom of the aircraft makes the analysis more complicated. However, in depth theory of tethered aircraft is not really necessary for our purposes, instead we identified two possible cases for tow hook placement. In all cases the tow hook must be placed ahead of the cg for stability considerations. First, for the tow hook mounted closer to the nose the tow will be slower but the glider should come off the hook at a level attitude. However, this method runs the risk of placing the tow hook too far ahead such that the nose is pulled down during tow. The second option is to mount the took just barely in front of the cg, which results in a much faster climb rate. There are two potential risks with this method. One, the cg of the aircraft could change over the 15 course of the competition and the tow hook might end up aft of the cg. Two, with such a large nose up angle during tow there is the potential for the glider to stall on tow, thus losing a lot altitude during the initial stages of the flight. We decided to go with the second option of placing the tow hook just in front of the cg, and have analyzed the aerodynamics sufficiently to ensure that the glider won’t stall as it comes off the hook. 2.3.4 Empennage Mounting/Dethermalizer The empennage mounting is very similar to the wing mounting, but there exists the added complexity of the dethermalizer mechanism. It was imperative to allow the tail to pop off on impact, in a similar manner to the wing. Thus the empennage is mounted onto the fuselage via elastic bands as well. There is a small hinged board attached to the fuselage on which the empennage sits. On impact, the elastics come off and the empennage pops off the board. The dethermalizer is a small mechanical timer; a radio control activated device was deemed to be too heavy. There is a pin on the timer that is connected via a string to the board on which the empennage is mounted. When 75 seconds of flight time has elapsed, the pin pops out creating slack in the string, which allows elastics connected to the empennage to pitch the tail down. There is a stop in place, which restricts this pitch down to 45 degrees. This is a simple, effective and innovative solution to the new mandatory dethermalizing constraint. 2.3.5 Nose Cone In order to ensure the safety of all participants, it was necessary to include a round foam nose cone in the design of the glider. The presence of a bulbous nose at the front of the glider will wreak havoc with aerodynamics, thus it was important to try and minimize its impact. A 2-inch radius semi-circle was attached to the front of the fuselage, then a foam fairing of sorts was mounted behind the cylinder on the fuselage to reduce the potential for flow separation around the nose. A small metal plate was placed at the tip of the golf shaft fuselage before the foam semi-circle nose in order to distribute the load on impact and prevent the fuselage from piercing the foam nose. The foam fairing is a simple technique that will greatly improve the glider aerodynamics. In addition the metal plate to distribute load is a clever way to avoid the problems of the fuselage piercing the nose cone as encountered by other teams in last year’s competition. 16 3 3.1 DESIGN ALTERNATIVES CANARD CONFIGURATION The advantage of a canard configuration is that the trim surface produces positive lift to the aircraft, instead of negative in the case of a conventional configuration. Nevertheless, this does not guarantee an overall lift increase in the aircraft. The downwash from the canard reduces the local angle of attack of the wing, and therefore, the lift as well. To study the effect of the wing being “unloaded” by the canard’s downwash, the following method can be used. The downwash can be estimated by USAF DATCOM method. If the downwash is assumed to be of the same width as the span of the canard span, Schrenk’s method (Ref 7, Ref 9) can be used to calculate the lift deficiency. However, our team’s previous experience with canard has found that such configuration is prone to failure because performance of the aircraft is quite sensitive to the sizing of the aerodynamic surfaces. Idea of a canard configuration was abandoned during the early stage of our design process. 3.2 FLYING WING CONFIGURATION Recent research by NASA and Boeing in the blended wing body (BWB) concept spurs general interest in the flying wing configuration. Combining the wing with fuselage has the potential to reduce weight while maintaining the aircraft’s cargo storage space. However, cargo storage space is not a critical constraint to our glider. Thus, the merit of a flying wing is limited for our purposes. In addition, in order to generate a nose down moment to trim the aircraft, sweep back and twisting of the wing and/or a reflex airfoil section are required. None of these wing features are aerodynamically favorable, and they are difficult to construct. 3.3 BIPLANE CONFIGURATION Despite years of neglect, our team’s recent interest in biplanes prompted us to investigate such a configuration. With the span of the glider limited by the rules, biplane appears to be an attractive alternative in reducing induced drag from the wing. The method of our analysis followed that given in Ref 9 and Ref 10. Better theoretical lift 17 coefficient can be obtained following Ref 11 and Ref 12. Given the lift coefficient, induced drag coefficient can be calculated from Ref 13. Three cases were tested in our analysis: 1. Basic monoplane based on last year’s Zephyr configuration 2. Biplane with 2 full size wings, both wings same as the one used in case 1 3. Biplane with 2 wings having chord length half of the one used in case 1 To simplify calculation, we assumed 1. The wing is the aerodynamically dominant part on the glider (i.e. trim force assumed to be small) 2. Glider is under steady condition throughout the flight (i.e. net force is zero, constant velocity, etc) Following these assumptions, expression of flight time can be easily derived from force balance and basic kinematics equation. 1/ 2 S C 2 C 2 L D 2 gM where tan 1 D L and h is the initial altitude of the glider. h t sin (4.3.1) Using the lift and drag coefficients obtained from the method outlined in Ref 10, it was found that monoplane (case 1) and biplane with 2 full size wings (case 2) had similar flight durations (147sec and 145sec respectively under the assumed ideal conditions). However, biplane with 2 half size wings (case 3) yielded a substantially longer flight time than the other 2 cases (337sec under the assumed ideal conditions). Thus, case 3 has the potential to carry more cargo than the other two cases. This "halfsize" configuration will be the focus of our future research. Although biplane is aerodynamically favorable, reduction in individual wing size increases the structural load. A finite element analysis, similar to the one given earlier, is required to determine the feasibility of biplane for future competitions. 4 FUTURE CONSIDERATIONS In the course of constructing the glider, the most important tool was the wire cutter. The wire cutter was simply a long wooden handle with wire running across it, connecting to a power transformer. This vital tool used to cut through foam would 18 lengthen and lose tension when heated. The lost in tension causes the wire to drag along the surface of the foam, creating undesirable grooves. The process of cutting required two people, one person on either side to hold the wire cutter steady. Much time and material was spent for practicing with the wire cutter. The construction of the glider can be drastically improved in aspect of quality and speed if we had better method of cutting the foam. Two options are worth considering for the future: mechanically assisted and computer or microcontroller controlled motorized wire cutter. 4.1 MECHANICALLY ASSISTED WIRE CUTTER Mechanically assisted wire cutter is basically two highly flexible mechanical arms mounted on a platform, with the wire held between them. In essence, a mechanically assisted wire cutter is similar to current crude wire cutter; it will still require two people to operate most of time. However, there are four important differences. 1. Adjustable distance between the two ends that holds the wire, to ensure constant tautness even when heated. 2. Mechanical arms assistance minimizing tremors in user’s hands allow for smoother cutting of foam. 3. Allow for more precise cutting when following a template because the wire cutter mechanical arms are mounted on platform. 4. Allow for single person operations by locking the two mechanical arms to move in synchronization when there is an obvious symmetry in shaping the foam. With this device, the foam cutting time and learning curve would be dramatically shortened. The greatest challenge of building such construction is determining the amount of flexibility of the mechanical arms. Not only the joints cannot be too loose or tight, but the number of joints is also important factor in creating an ideal arm. Such task will require much time in research and development. 4.2 COMPUTER OR MICROCONTROLLER CONTROLLED MOTORIZED WIRE CUTTER The principle of a computer or microcontroller controlled motorized wire cutter is similar to the current foam cutting method except the hot-wire cutter will be guided by stepper motors. The dimension of the airfoil section will be entered into the control software, and the path of the stepper motors will be calculated automatically. This type of wire cutter has all the advantages of the mechanically assisted wire cutter plus three additional points: 19 1. Minimized human errors by minimizing human involvement. 2. Very precise foam cutting. 3. Automatic adjustment for tautness of wire. This is much more ambitious task than the previous type of wire cutter. Constructing such contraption will likely require more than a year of development. Aside from the previously mentioned mechanical challenge, there will be electrical and software challenges. Controlling this device will require a microcontroller or a computer, depending on the interface of the user and the device. Both types of wire cutter will be a definite improvement over the current tool. Both will decrease time to learn and build the glider, as well as minimizing wastage of materials. However, the amount of work that required setting up either one is not feasible at this moment. Nevertheless, when our team’s techniques in aerodynamic analysis have matured, these ideas will be worth considering. 4.3 3D MOTION SIMULATOR The U of T Free Flight Glider Team is currently implementing a 3D motion simulator in MATLAB Simulink. However, calculating the aerodynamic coefficients and stability derivatives prove to be quite difficult using USAF DATCOM method. We are currently attempting in estimating stability derivatives using computational fluid dynamics (Ref 14). 20 5 REFERENCES [1] Jones, B., “Elements of Practical Aerodynamics”, 4th ed., John Wiley & Sons, Inc. New York, 1950 [2] Raymer, D.P., “Aircraft Design: A Conceptual Approach”, 3rd ed., AIAA Education Series, 1999 [3] Coates, A., “Jane’s World Sailplanes and Motor Gliders”, New Edition, Jane’s Pub. Co., London, 1980 [4] Drela, M., “Xfoil Manual”, Aerospace Computation Design Laboratory, Dept. of Aeronautics & Astronautics, M.I.T., 2001 [5] Anderson, J.D., “Fundamentals of Aerodynamics”, McGraw-Hill, 2001 [6] McCormick, B.W.., “Aerodynamics, Aeronautics, and Flight Mechanics”, John Wiley & Sons, New York, 1979 [7] Schrenk, O., “A Simple Approximation Method for Obtaining the Spanwise Lift Distribution”, N.A.C.A. Technical Memorandum No. 948, 1940 [8] Etkin, B., Reid, L.D., “Dynamics of Flight: Stability and Control”, 3rd ed., John Wiley & Sons, Inc., 1996 [9] Hiscocks, R.D., “Design of Light Aircraft”, R.D. Hiscocks, 1995 [10] Munk, M.M., “General Biplane Theory”, Technical Report No. 151, N.A.C.A., 1922 [11] Diehl, W.S., “Relative Loading on Biplane Wings”, Technical Report No. 258, N.A.C.A. , 1922 [12] Diehl, W.S., “Relative Loading on Biplane Wings of Unequal Chord”, Technical Report No. 501, N.A.C.A. , 1934 [13] Prandtl, L, “Induced Drag of Biplanes”, Technical Note No. 182, N.A.C.A. , 1924 [14] Park, M.A., Green, L.L., Montgomery, R.C., Raney, D.L., “Determination of Stability and Control Derivatives Using Computational Fluid Dynamics and Automatic Differentiation”, 17th AIAA Applied Aerodynamics Conference, Norfolk, Virginia, AIAA 99-3136, 1999 21 6 6.1 APPENDIX A: LOAD CALCULATIONS WING LOADS 6.1.1 Flight Loads: We assume level flight Lift = Weight. A glider of 1 kg and a span of 3m has a wing loading of 3.27N/m. Then the moment along the wing can be described as follows: M 3.7 4.9 x 1.6 x 2 Moment Load [Nm] Moment Load Along Wing During Flight 4 3 2 1 0 0.0 0.5 1.0 1.5 Distance From Fuselage [m] 6.1.2 Launch Loads: During launch the glider is at an inclined configuration for climbing. This increases the drag and lift of the wing substantially. We will assume that the moment on the wing during launch is between 2 to 4 times greater than the moment on the wing during normal flight. A maximum of moment 14.8 Nm may be experience. 6.1.3 Impact Loads: The impact loads are highly unpredictable. Typically a wing-tip will impact the ground upon landing, resulting in a maximum moment near the fuselage and a large shearing force near the wing-tip. These factors will be qualitatively taken into account. Thus we will design our wing to withstand a maximum moment in the wing of 3.8 Nm during flight and 15 Nm during launch. 6.2 FUSELAGE LOADS The maximum loading for the fuselage will occur during landing. The glider travels at approximately 5.5m/s (22km/hr) and upon impact could stop within 20cm. This corresponds to a deceleration of 76 m/s2. Assuming the mass of the glider is 900g and the 22 cross-section of the fuselage is 1cm2, the corresponding load is 680 MPa. This is within the limits of the carbon fibre properties. Assuming constant accel. Vav = 5.5/2 = 2.75 m/s Δt = Δd/2.75 = 0.2m/2.75m/s = 0.0727s a = Δv/Δt = 5.5m/s/0.0727s = 75.6 m/s2 stress = ma/Area = 0.9kg*75.6m/s2/0.0001m2= 681 MPa 6.3 EMPENNAGE LOADS The maximum loading on the empennage will occur during landing. The empennage is attached with rubber bands to absorb any impact forces. We will adjust the rubber band configuration to ensure that the yield strength of 22MPa of the balsa wood is not surpassed. 6.4 CARGO BAY LOADS The maximum loading for the cargo bay will occur during launch, through the tow hook. From experience we can estimate the tension in the rope during launch at approximately 3 times the glider weight. Thus the maximum load on the cargo bay tow hook is 27N. Force = (3)mg = 3*0.9kg*9.81m/s2= 26.5N 6.5 NOSE CONE LOADS The maximum loading on the nose cone will occur during landing. The calculation is similar to that of the fuselage, except the cross-sectional area is now the contact surface (~15cm2). Thus the maximum stress is 45MPa. 23 7 APPENDIX B: DRAWINGS 24