RocketSat Senior Design Project

Transcription

1 RocketSat Senior Design Project Jacob Bailey 1, Travis Brady 1, Alexander Grammer 1, David Gronstal 1, Brendan Mangan 1, Eric McVay 1, Noelle Ridlehuber 1, Kellen Schroeter 1, and Dylan Stapp 1 The University of Alabama, Tuscaloosa, Alabama, The ability to launch a payload and collect data is of great significance. A rapid deployment glider is specifically beneficial for surveillance and data collecting. In the case of locations without adequate space for a runway, rapid deployment gliders are able to deploy where orthodox drones are not. Potentially, this system can be used as an important tool in humanitarian and military missions. This paper discusses the yearlong design project of the RocketSat team, who examined and executed the process of how to launch a rocket, deploy a glider in flight, capture high-definition video of the launch, and log atmospheric data of the descent. Although the team was met with significant financial challenges and a difficult degree of coordination and integration, this paper demonstrates the logical design process, technical analysis, and budgetary decisions of the project. Nomenclature b 1/2 = half-span length = center of gravity = fin coefficient (Barrowman Eqns) = nose cone coefficient (Barrowman Eqns) = resultant coefficient (Barrowman Eqns) = center of pressure = root chord of fin (Barrowman Eqns) = tip chord of fin (Barrowman Eqns) c 1 = section 1 chord length c 2 = section 2 chord length D = rocket body diameter = diameter at aft section of transition (Barrowman Eqns) = diameter at fore section of transition (Barrowman Eqns) = diameter of base of nose (Barrowman Eqns) EI = beam stiffness f EI = non-dimensional stiffness value f m = non-dimensional mass value K = non-dimensional stiffness matrix LE = leading edge length = length of fin at center chord line (Barrowman Eqns) = nose length (Barrowman Eqns) = transition length (Barrowman Eqns) M = non-dimensional mass matrix M = bending moment m = mass of beam = number of fins (Barrowman Eqns) q = distributed load = base radius of rocket (Barrowman Eqns) = semispan of fin (Barrowman Eqns) TE = trailing edge length V = shear force 1 Student, Department of Aerospace Engineering and Mechanics, AIAA Student Member 1

2 = distance from nose tip to root chord LE of fin = fin center of pressure (Barrowman Eqns) = nose cone center of pressure (Barrowman Eqns) = distance from nose tip to front of transition (Barrowman Eqns) = distance between fin LE at root and at tip (Barrowman Eqns) x = spanwise beam location βl = roots of beam characteristic equation ξ = non-dimensional x coordinate η = static margin Λ aft = wing sweep (aft section) Λ fore = wing sweep (forward section) λ = vector of non-dimensional frequencies squared (eigenvalues) v = beam displacement = eigenfunction (or mode shape) Ψ = vector of modal amplitudes (eigenvector) ω = vector of non-dimensional vibrational frequencies I. Introduction ROM the earliest beginnings of the project, the RocketSat design team met many obstacles in its path to F completion. Given major budgetary constraints, the team still completed the project detailed below with great ingenuity and resourcefulness. A. Concept of the Operations During mission operation, five primary events (mission stages) will occur: Launch, Deployment 1, Deployment 2, Descent, and Landing and recovery. These stages and associated mission operations are discussed in more detail below. Stage 1: Launch The rocket is propelled to an apogee of 1000ft, taking video and atmospheric data during ascent and storing the data on-board. During this stage, the wings are held in a collapsed position and the wing is tucked inside the glider body tube. Stage 2: Deployment 1 At apogee, a black powder charge detonates, breaking shear pins and separating the glider from the booster section. At this point the wings of the glider open and the glider stabilizes. Stage 3: Deployment 2 As the glider begins its stable, gliding descent towards the ground, a second black powder charge is detonated at 600 feet in the booster section. This charge deploys the booster parachute. The chute unfurls and the booster begins to decelerate. Stage 4: Descent Having successfully activated all deployables. The glider and payload maintain a slow, constant, controlled descent to the ground. The kinetic energy of the glider should not exceed that of the booster section during descent. GPS transmission, video recording, and atmospheric data recording continue throughout descent. Stage 5: Landing and recovery At the end of their descents, the glider and booster sections make contact with the ground. Conspicuously colored parachutes and fins make the two pieces of the glider easy to spot and recover. A GPS unit mounted in the nose of the glider also aids in recovery. 2

3 B. Original Project Goals and Requirements In Table 1 below, the initial project goals are given assuming a full budget from the Alabama Space Grant Consortium. Table 1. Initial Project Goals Rqmt Num Requirement Comply / No Comply / Partial Comments or Notes 1 Build and launch rocket with a budget under $6,000 Comply -- 2 The rocket shall safely deliver the glider to an altitude of 5000ft (+/- 100ft) No Comply Change of requirement 3 After deployment, the glider will right itself before glide Comply -- 4 The glide ratio of the glider will be >1 No Comply The glider electronics will be removable without excessive effort The glider electronics may be activated in launch configuration GoPro video will be recorded in-flight. The data will be stored onboard the glider and transmitted to a ground station in real time. Atmospheric data will be recorded from the glider in-flight. The data will be stored on-board and transmitted to a ground station in real time at 1Hz After separation, the wings of the glider will deploy to 45 degrees The glider will land at a kinetic energy that is less than that of the booster at landing Comply -- Comply -- Partial Partial No real time transmission to GS No real time transmission to GS Comply -- No Comply For all intents and purposes, the glider hit the ground too hard 11 The rocket shall have a stable ascent No Comply Nose dive After glider separation, the booster will deploy a parachute and assume a stable descent to the ground A GPS device will be used to ensure recovery of the glider after launch The color scheme of the rocket booster, glider, and parachutes will be chosen to necessitate the easy recovery of the system Upon recovery of the system, it shall be in a condition ready for immediate re-launch Other than the on-board batteries, no materials will be chosen that present a hazard to the environment No Comply No deployment Comply -- Comply -- Comply -- Comply -- 3

4 C. Budget Funding and Constraints RocketSat was met with a significant budget constraint in January 2015 when the team was informed that a grant from the Alabama Space Grant Consortium would not be received. This led the team to ask for money from the Student Government Association FCA committee. Due to budget constraints the team was only given $ Table 2below shows the distribution of funds with $25.94 remaining in the account. The team s finances were supplemented by a t-shirt fundraiser which earned a profit of $160. The money raised was used to buy the materials used to build a vibration mount and to pay for gas to and from the launch site in Manchester, TN. Additionally, Dr. Zeiler bought the team two respirators to be used during all sanding. The respirators were bought under the condition that they were to remain in use by senior design teams only. Table 2. RocketSat FAC Budget RocketSat Budget Price Paid Shipping Paid ITEM VENDOR QTY 2.6" inch G12 Thin Wall Airframe, 3' Chris' Rocket Supplies 1 $53.00 $ " G12 coupler, 6" Chris' Rocket Supplies 1 $12.99 $ count armorall car cleaning wipes Home Depot 1 mechanics gloves (large/blue/nitrile) Home Depot 1 $18.16 $ mm 2 grain motor casing Chris' Rocket Supplies 1 $ mm motor Chris' Rocket Supplies 1 $32.95 $0.00 Spring (1" x 1") Grainger (Spring 44U867) 3 $13.96 $0.00 MicroSD Card Breakout Board+ Adafruit (Product ID 254) 1 $14.95 $9.18 Spring Housing/Cage McMaster-Carr (1878K19) 2 Rigid Carbon Fiber Rod McMaster-Carr ( 2153T56) 1 Rigid Carbon Fiber Tube McMaster-Carr ( 2153T41) 2 Clevises 2pk Horizon Hobby 4 Pushrods Horizon Hobby 1 Servo Arm Horizon Hobby 1 Servo Gears Horizon Hobby 1 $68.15 $8.29 $11.98 $2.99 Payload Amazon $21.00 $0.00 Totals: $ $27.47 Grand Total: $ D. Revised Project Goals and Mission Requirements Upon the realization that funding would not be available to the team from the Alabama Space Grant Consortium, new projected goals were outlined. They are listed in Table 3. 4

5 Table 3. Revised Project Goals Rqmt Num Requirement Comply / No Comply / Partial Comments or Notes 1 Build and launch rocket with a budget under $500 Comply The rocket shall safely deliver the glider to an altitude of 1000ft (+/-100ft) GoPro video will be recorded in-flight and stored onboard the glider Atmospheric data will be recorded from the glider in-flight and stored onboard at a packet rate of 0.5 Hz Comply -- Comply -- Comply -- E. Educational Engagement and Outreach The educational engagements completed by the team included The University of Alabama E-Day, Tuscaloosa County's Journey2Jobs, and Woodland Forrest Elementary School. A general exhibition table was set up at both E- Day and Journey2Jobs where students and professionals could come to the table and ask questions about the different rockets. At Woodland Forrest Elementary School a pre-determined presentation was given to each class level, and then a small Estes rocket launch was performed for each grade. The number of students engaged at E-Day is unknown, however approximately 800 8th graders were met with at Journey2Jobs and 400 students were met with at Woodland Forrest Elementary. II. Procedure, Design, and Manufacturing In the section below, the steps to selecting the materials and building the structure of the rocket and glider are outlined. A. Recovery Electronics The avionics bay was housed inside a coupler joining the booster and payload. A SolidWorks drawing of the avionics sled and components is shown below in Fig. (1). The sled was fitted with an on/off switch, PerfectFlite StratoLogger SL100 altimeter, and 9v battery. The altimeter is shown below in Fig. (2). The final design of the avionics sled was selected with the battery on the underside-beneath the altimeter to conserve space inside the bay. Figure 1. A SolidWorks drawing of of the avionics sled 5

6 Figure 2. The StratoLogger altimeter by PerfectFlite To track the rocket, Yellowhammer Rocketry s tagg PetTracker GPS unit was used. To activate the unit, a monthly subscription to PetTracker was purchased for $9.99. The GPS unit was secured in the nosecone of the rocket with duct tape. The unit was mounted such that it would provide additional stability by adding mass to the underside of the glider. A tagg PetTracker GPS unit is shown below: Figure 3. The tagg PetTracker GPS unit B. Booster The booster section was designed to be 24 inches long so that it would be large enough to house the 9 inch motor mount, a parachute and parachute protector with shock cord, and 2.5 inches of coupler from the avionics bay. The diameter was 2.6" fiberglass because of its small size and low cost compared to larger fiberglass body tubes. The construction of the booster was completed in three steps. The first step was building the motor mount which was made from a motor tube, a motor retainer, and two centering rings. The motor tube was fiberglass and 38 mm in diameter to allow housing of a Cesasroni 2 grain 38 mm motor. Next, an AeroPack 38 mm motor retainer was epoxied to the bottom of the tube. This retainer was made of two parts, as seen in Fig. 4, where the right piece formed a lip off of the motor tube in order to catch the motor casing and has a matching thread so that the left part can be screwed in place. Then the two centering rings that center the motor tube inside the body tube needed to be epoxied in place. The aft centering ring was glued as close to the base of the motor retainer as possible to allow a greater surface area for the motor retainer to be glued to the motor tube. The forward centering ring had an eye-bolt with a shock cord knotted through it so that the shock cord had an anchoring point during recovery. The forward centering ring was glued above the fin slots and high enough so that the shock cord knot did not interfere with the motor casing being put into the motor mount. For this design that meant that the forward centering ring was placed a 1/2 in below the forward end of the motor tube. Step two was to epoxy the motor mount inside the body tube. This was done so that the aft centering ring was 1/2 in from the aft end of the body tube to maintain clearance of the fin slots. The final step was to glue the fins in place by putting epoxy on the each edge of the fin that would interface with the motor body tube, and then slid the fin into the fin slot. Each fin was epoxied individually to ensure that it was straight before the epoxy was set. Once all fins were dry, a fillet of epoxy was built up in where the body tube meets the fin in order to increase the strength of the fins attachment to the body tube. 6

7 Figure 4. AeroPack Motor Retainer C. Fin Design The purpose of the fins on a rocket are to ensure stability of the rocket in flight. It is crucial that the rocket maintain its orientation. Therefore, adequate fin designs for the given size and mass of the rocket were needed in order to fulfill mission requirements. Due to the unique mission objectives of the rocket, it was decided that two full sets of fins be incorporated into the design. This decision was a result of multiple factors. With the fore section of the rocket deploying the rogallo wings to become a glider, fins at the aft of the glider (mid-section of the rocket) were required to act as a tail for the glider and to help provide glider stability. Furthermore, the external glider features on the rocket provided some instabilities in the rocket; rather than have abnormally large rear fins to compensate for these instabilities, it was found that two sets of fins could have the same results. RockSim was a vital tool in determining the size and shape of the fins required to ensure stability of the rocket. A useful measurement of stability that RockSim provides is the static margin of the rocket. The static margin is a dimensionless value that provides a means by which to easily quantify the stability of a rocket; it is given by Eqn (1). (1) RockSim uses the Barrowman Equations to calculate the center of pressure. These equations essentially split the rocket body into several pieces and calculate the center of pressure for each section separately. They then use the center of pressure for each of the sections to find the center of pressure for the entire rocket. An abbreviated explanation of the Barrowman Equations for a rocket with an ogive nose cone and no conical transitions (i.e. a smooth, constant diameter body) is given by Eqns (2) through (7) and Fig 4. It should be noted that Fig 4 is a general diagram for the Barrowman Equations; the RocketSat rocket did not have any conical transitions. Furthermore, the wing mount block on the side of the rocket was considered to be a series of closely spaced fins in the shape of the block for the RockSim Barrowman calculations. (2) (3) (4) 7

8 (5) (6) (7) Figure 4: Barrowman Equation Variables The key factor in designing the fins and determining the fin shape was the static margin. One of the team captains has a contact with NASA that highly recommended a static margin of over 2.0 to ensure rocket stability. Therefore, the fins were designed to help achieve that value. The rear fins were attached via a fin-can assembly. A fin-can assembly is a stronger way to attach the fins that involves epoxying the fins to an internal rocket tube and sliding the tube into the main rocket body, ensuring that slits are cut in the main rocket body for the fins to slide into. Due to a combination of the necessity to quickly build (due to the lack of funding and need for a last minute redesign) and miscommunication between the rocket and glider sub-teams, the wing mount block ended up being much larger than the rocket sub-team previously thought which drastically changed the static margin with the current fin design. To counteract this loss of stability the fins on the rocket needed to be increased in size. However, the rear fins were already epoxied in the rocket body by the time that the rocket sub-team realized that the wing mount block would cause these changes. Therefore, a solution was needed to fix the stability of the rocket. The solution was to 3-D print larger sleeve fins that would be epoxied over the existing fins. However, the sleeve fins were designed to not be too much larger than the existing small fins. This design decision was made because the 3-D material had significantly less strength than the original fiberglass fins. Since the 3-D fins were designed to be sleeves, part of the 3-D fins had strong fiberglass backbones. However, making the fins too much larger than the existing fiberglass fins would result in large sections of the rear fins with no fiberglass backbone and this result led to concerns about the 3-D fins breaking in flight. Thus, the 3-D fins were designed to increase the size of the rear fins while simultaneously ensuring that a majority of the sleeve fins had a strong fiberglass backbone. The sleeve fins were printed with the Objet 3-D printer and epoxied over the existing smaller fins. The rear fin design is shown in Fig. 5. 8

9 Figure 5: Rear Fin Design When considering the design for the mid-section fins, both rocket and glider stability needed to be considered because these fins were integral parts of both vehicles. There were two options to attain a static margin of 2.0 for the rocket. Option one was to make extremely large mid-section fins needed to be used to move the center of pressure farther aft, increasing the distance between the center of pressure and center of gravity, and increasing the static margin. Option two was to add a significant amount of mass to the nose cone which would move the center of gravity forward and, have the same effect as the extremely large mid-section fins. Adding weight to the glider was undesirable, and there were some relatively large fins already cut that were readily available for use, so option one was chosen. After completing a static margin calculation on the rocket, it was found that using these fins in accompaniment with an ounce of added mass to the nose cone would put the static margin of the rocket at the desired 2.0. Using these fins would eliminate the need for a significant amount of added mass to the nose cone, a highly desired outcome for the glider design. Therefore, it was decided to use these existing fins on the rocket and to add an ounce of mass to the nose cone of the rocket. The mid-section fin design is shown in Fig. 6. Figure 6: Mid-Section Fin Design D. Electronics Sled A rapid deployment glider would be beneficial for surveillance and data collecting. A flying mechanism of this style could be launched quickly, even in locations with no conventional runway or sufficient space to deploy more orthodox reconnaissance drones. This system could be seen as an important tool in disaster area assessment and relief or combat zone aid. To cater to these applications the team decided it would be critical to carry a payload that could capture high-definition video of the launch and surrounding area and log atmospheric data of the launch and descent. For the video subsystem of the payload the team explored many different options. The three options considered can be seen in Fig. 7. This narrowed field included the light weight, customizable Raspberry Pi Camera Module, the reliable and readily available GoPro Hero 3+ and the technically impressive, yet expensive Sky Drone FPV system. The Sky Drone FPV system was quickly identified as the ideal solution. It offered full 1080p HD video at 30 frames per second and operated a first person view system over standard cellular networks. This unique FPV system would allow users to record and watch real time flight video easily over any cell phone. This transmission system, as opposed to standard 5.8 GHz transponders, would allow for long range FPV without the hassle of complicated antenna systems which may be difficult to set up at a remote launch location. However, due to unforeseen design and budget constraints a switch to a 2.5 in diameter glider meant the Sky Drone system, with a maximum dimension 9

10 of 3.3 in, would be unable to fit in the designated glider bay. The GoPro Hero 3+ was subsequently chosen over the Raspberry Pi Camera. Both systems were already owned by a member of the team so experience was not a short coming. However, the Pi Camera only offered 1080p 30fps and was controlled by a computer that could not fit the minimum diameter. Meanwhile, the GoPro Hero 3+ could record 1080p video at twice the frame rate and packages the camera, processing and battery into and astounding 2.3x1.6x0.8 in. While it is clear that aid workers or troops on the ground would benefit immensely from a real time FPV video, the team realized that a more expensive and sophisticated system could accomplish this task and that the record-and-recover strategy settled upon would serve as a more than sufficient proof of concept. Figure 7. Payload Camera Options. From left to right: Raspberry Pi Camera Module, GoPro Hero 3+, Sky Drone FPV Recording flight data is essential to prospective customers. This aspect of the payload could monitor disaster information such as chemical composition of the air, radiation levels or structure of a lava plume. It could also be used as a way to intercept radio signals in a time of war. The team decided to take a catch-all approach to the remainder of the bay and collect basic atmospheric data from a simulated research perspective. This proof of concept application would measure and record data using the following packet format: <Packet Count, Mission Time, Altitude, Temperature, Pressure>. The altitude, temperature and pressure readings were carried out using a Bosch BMP180 barometric pressure and temperature sensor. This device is compact and low power - measuring 1.8x1.8x0.2 in and operating between 3.3 and 5V. It also has a resolution of up to.03hpa which would plenty accurate for the team s application. Additionally it communicates via i2c protocol which means the physical layer allows for easy transfer with any microcontroller containing SCL and SDA lines. A real time data transfer system using 2.4 GHz series 2 Xbee radios and Matlab was developed. However, it was ultimately scrapped due to prohibitively expensive long range antennas. For this system a back up on board SD card storage system was created. Again using i2c protocol, a text file was written containing each packet of flight data from each glider flight. This on board storage system now became the primary flight logging strategy and a back up system was now necessary. The back up storage need was the critical driver for the microprocessor selection. Initially an ATmega328 based Arduino Pro Mini was used for prototyping. This processor is extremely small, measuring only 0.7x1.3x.0.1 in. But it contains only 1 kb of on board EEPROM memory. At this storage capacity and with a packet rate of 1 Hz only a few minutes of data could be logged on the back up memory system. Subsequently, an ATmega2560 based Arduino Mega was selected. While significantly larger this controller was still small enough to fit in the payload bay and boasted a much more acceptable 4 kb of on board EEPROM. A back up storage protocol was then written were an incoming packet was disassembled from a string and each character was converted to its conventionally corresponding hex value and written byte by byte into an EEPROM address. Data values were separated by an ASCII comma (,) and packets were delimitated with a pound sign (#). At the end of each packet an at symbol (@) was written. This was over written by the start of a new incoming packet at each loop. But in the instance of an abrupt termination the last, would serve to locate the end of relevant data. This backup system only starts when the sensor reads a change in altitude so the limited backup storage space isn t filled with baseline launchpad data. The various data acquisition components used can be seen in Fig

11 Figure 8. Payload Electronics Components. From left to right: Bosch BMP180, Arduino Mega 2560, Adafruit Micro SD Card Breakout Board Once the system was devised and each component thoroughly bench tested the final electronics sled assembly could begin. The sled was allotted half of the diameter of the glider tube to leave plenty of room for the retracted wing material. This dimension was verified to house all required electronics through SolidWorks modeling. The CAD constructed model was then milled out onto a piece of steel sheet metal. Steel was selected because unlike aluminum, steel will not yield to a 90-degree bend. Aluminum requires various bend radii to achieve a 90-degree turn and the tools available in the student machine shop were not sufficient to meet this requirement. The milled sheet was then bent into the correct 3D formation using a break. Semicircle bulkheads to secure the sled to the body tube were constructed out of wood with a 2.5in hole saw. A custom atmospheric data shield was created to fit into the female headers on the Arduino. This model is common among hobbyists and allows a clutter free, secure connection into the processor. The backbone of this shield is a standard.1 in grid, fiberglass perfboard. A connection plan was carefully laid out to ensure a simple set up that correctly mated all components to their necessary processor pins. The SD card logger and BMP 180 are the main components, but an LED was also outfitted to visually confirm to a user when a packet had been logged along with a screw terminal for battery replacement and a system on/off switch. This shield was then carefully soldered together and fit onto the Arduino. This data logging system was then connected to the electronics sled using industry standard nylon hex standoffs to eliminate the possibility of a short circuit and provide a rigid flight-worthy attachment. A 9V battery powers this system. The GoPro camera was a simple stand-alone system and was rigidly secured into place. The sled was designed to fit at any point along the body tube and could be used by the mechanical team as ballast for center of gravity control. Once the appropriate location was determined an access port was cut into the body tube to control all electronics and provide a window for the camera. This port was covered with clear packing tape to maintain aerodynamic soundness. Fig. 9 shows the complete electronics sled. Figure 9. Payload Electronics Sled. The final product, complete with Arduino and Altitude Data Logging Shield, GoPro Hero 3+ and the physical sled housing. 11

12 E. Rogallo Wing History, Selection, and Design The Rogallo wing was originally designed by Francis M. Rogallo of NASA Langley Research Center as a reentry recovery system for the Mercury, Gemini, and Apollo capsules, as well as the Saturn rocket stages 1. The Rogallo flexible wing is essentially a controllable paraglider. This type of recovery system was developed out of necessity for a semi-controlled capsule descent and landing though that initial interest would be replaced by the time-constrained desire to achieve a lunar landing before the end of the 1960s 1. While being similar to today s hang gliders in appearance, the Rogallo wing s lift-to-drag ratio averaged between where as hang gliders average much higher around 15. Aerojet s study of the Fabrication of Inflatable Reentry Structures for Test (FIRST) used an inflatable paraglider design with a relatively high lift-to-drag ratio of 8.0. For this study, engineers estimated a landing speed of 55 kph (~34 mph); to mitigate this speed to a reasonable level, they included a flare before touchdown that reduced the horizontal velocity to 9 kph (~5.6 mph) 3. Each proposed Rogallo wing configuration underwent a series of static wind-tunnel tests in the Langley 300 mph tunnel. Naeseth and Fournier outline the results gained from each test, as well as how modifying different planform variables affected the aerodynamics of the system 2. This research was used to begin the RocketSat design, though it should be noted that the initial Rogallo wing configurations were highly experimental and each design required substantial wind tunnel and in-flight testing. Of particular concern was that most Rogallo wings use an extended high wing design (i.e. extended high above the fuselage), and our design was a standard high wing design (i.e. attached to the fuselage). The standard Rogallo wing also operates using a shifting center of gravity, CG, and a shifting center of pressure, CP. At steady level flight conditions, the CG and the CP are collocated in the 2D horizontal plane. A shift of the CG in any direction causes the wing to flare in the same direction. The CP then automatically adjusts back to stable equilibrium, with thecp set directly above the CG once again 4 this shifting equilibrium allows the hang glider to maintain both static and dynamic stability. Fig. 10 depicts a Rogallo wing with inflatable tubes carrying a Mercury capsule 5. Many of the aforementioned configurations were also entirely flexible wings, meaning they had no rigid structure or support elements similar to today s paraglider designs. Figure 10. Rogallo Wing Carrying Mercury Capsule 5 The team decided to take a highly experimental approach to this project and develop a fixed Rogallo wing glider. Theyimplemented a leading-edge-supported flexible wing model with a partial keel batten (the keel length was prevented from vertical movement by a clothes hanger wire). The wing fabric was rip-stop nylon with carbon fiber fixed leading edges; the trailing edges were hemmed to prevent fraying, but were otherwise free. The design was established from a parawing canopy planform geometry described in earlier NASA research 6. This geometry uses a fore sweep angle of 45 degrees and an aft sweep angle of 22.5 degrees. The leading edge length was constrained to be inches because of the length of the carbon fiber rods that the team had available. The wing geometry and dimensional values for the deployed wing planform area are displayed in Fig. 11 and Table 4, respectively. 12

13 Figure 11. Deployed Wing Planform Area Geometry Table 4. Dimensional Values for Deployed Wing Planform Area Parameter Value [Unit] Λ fore 45.0 [deg] Λ aft 22.5 [deg] b 1/ [in] LE [in] TE [in] c [in] c [in] A series of geometric relations and calculations were then applied to obtain the wing curvature that is characteristic of Rogallo wings. First, the desired radius of curvature and arc length for the wing trailing edge were determined by mapping out multiple radii and the corresponding arcs, then determining which one would be used for the design this approach, much like that of the initial Rogallo engineers, is highly experimental and would have benefitted greatly from wind tunnel and in-flight testing. The half-span length, b 1/2, was then set to be equal to the resultant arc length. With the half-span and leading edge lengths known, the corresponding sweep angle for the desired curvature was able to be determined. Additionally, the adjusted trailing edge length was determined using the new half-span length and the original section 2 chord length. Table 5 outlines the dimensional values that were used to cut the wing fabric, accounting for the desired wing curvature. Table 5. Dimensional Values for Wing Fabric Cut Parameter Value [Unit] Λ fore 38.0 [deg] Λ aft 20.4 [deg] b 1/ [in] LE [in] TE [in] c [in] c [in] As stated earlier, the team took a highly experimental approach in our design of the fixed Rogallo wing glider; particularly because this design had never been done before or at least not published. It was immediately evident that the glider, with the fixed Rogallo wing, would have no means of maintaining static or dynamic stability; the CG was fixed (all of the glider s subcomponents were fixed in place), but there were no control surfaces to adjust for the changes in the CP location. To mitigate this, the team tried to align all the components so that the CG was directly under the wing area. The top-view of the glider modeled in RockSim is shown in Fig

14 Figure 12. Glider Top View The last-minute stability margin requirement forced the CG to be moved forward, no longer directly beneath the CP. Had the glider deployed and begun in steady flight, the relative CP would have caused a pitch down moment, thus causing the glider to nosedive. While there was a significant amount of research completed by NASA on the Rogallo wing configurations, most of the literature is purely experimental, only discussing testing methods and results for the Rogallo and Rogallovariant wings. Of particular interest to them were wind tunnel testing, wing deployment, and scaled dimensional similarity analyses 7. Due to budget and time constraints, the team was only able to test for one of these sections - wing deployment. If the lack of funding were predicted, the team would have built subscale models and completed wind tunnel and in-flight testing, similar to the NASA engineers, before completing the final design. Future research and development should be completed for fixed Rogallo wing gliders by examining aerodynamic effects from variances in the curvature of the trailing edge, sweep angle, planform area (wing configurations), fuselage size, wing deployment mechanism. F. Wing Deployment Mechanism The wing deployment mechanism is driven by a compressed spring housed inside the rocket body. The spring pushes a rack which in turn spins a gear. This gear is attached via a rod to a servo arm sitting on the outside of the rocket. The servo arm spins with the gear, pushing or pulling the pieces that hold the wing rods. These pieces rotate about a pivot point, thus deploying the wing to its gliding state. Initially, the mechanism was laid out as shown in Fig

15 Figure 13. Deployment mechanism geometric relations The equations shown there were made using geometric relations to determine the dimensions of the various parts in the system. There are four equations and nine unknowns. Five of the variables would have to be set to determine the other four. However, the orientation shown in this figure did not allow for a reasonable width such that it could fit on the rocket. The servo arm pushed/pulled outward on the pivots. To fix this problem, the servo arm was switched such that it could sit forward of the wing pivots. This would have the servo arm push/pull the wing pivots along the axis of the rocket, allowing for a thinner profile. See Fig. 14 for the final mechanism. Fig. 18 illustrates the internal gear system. Figure 14. Glider Wing Mechanism. Wings folded on left; wings extened on right. 15

16 When the glider section is attached to the booster section, the wing rods are held in their closed position (compressing the spring) by inserting the ends into two tubes. These two tubes are attached to the booster section but extend to the glider section so that this is possible. When the booster section is ejected during flight, the wing rods come loose from the booster section tubes, and the compressed spring deploys the wings. See Fig. 15 below. Figure 15. Glider Wing in its Folded Position. The wing material is stowed away during ascent by folding it into the rocket body through a slit running down the length of the glider section. To keep the wing from sliding off of the wing rods, fishing line is threaded through the wing and tied off at a screw in the wooden block holding the wing deployment mechanism. This can be seen in Fig. 14. To keep the wing centered along the body a small section of a rod is sewn into the center of the trailing edge of the wing. This section acts as a stopper as it is bigger than the slit in the body, preventing the whole wing from coming out of the slit. To keep the rest of the wing centerline anchored near the body during flight, a coat hanger wire is mounted over the slit in the body. The wire holds down the wing at the centerline giving the two arches desired in the wing. The wire and slit can be seen in Fig. 16. Figure 16. Slit and Centering Wire on Glider Section 16

17 G. Wing Deployment System Manufacturing For the manufacture of the wing deployment mechanism, there were several important considerations: 1) The materials needed to be low-cost and preferably already available in the lab 2) The materials needed to be workable so that manufacturing could be done in-house and necessary modifications could be made easily and quickly 3) The materials needed to be strong and lightweight to support the aerodynamic load on the wings without adding too much additional weight to the glider 4) Components needed to be compact to avoid taking up too much space inside the body tube and presenting a large drag profile on the outside of the tube A fully functioning prototype of the wing deployment mechanism was constructed before the final model to test the functionality of the system and make sure that the mechanism would fit within the restricted envelope of the rocket body tube. The prototype is shown below in Fig. 17. The part of the deployment mechanism to be mounted to the exterior of the body tube is shown on the right side while the interior portion is shown on the left. Figure 17. Prototype deployment mechanism interior (left) and exterior (right) of rocket Wood was the material of choice for the mounting bases and primary structure of the wing deployment mechanism due to its availability, workability, and its relatively lightweight nature. The other option considered for the wooden pieces was 3D printed ABS. This was decided against to save time and cost. In order to 3D print the bases with good fidelity, additional funding would have been required to use the Objet 3D printer. Additionally, the lines to use the Objet 3D printer are usually lengthy and a lot of time is wasted in the queue waiting for printing. It may have been possible to use a lower fidelity printer, but this would require the time to model each piece in CAD software before waiting most of the day to actually finish printing the piece. The parts used to make the actual deployment mechanism came almost entirely from model aviation. The red, aluminum, servo arm (left image in Fig. 17) is intended to operate the control surfaces of a RC airplane. Brass clevises were used as the push rods to attach pivoting pieces, mounting the wings to the wooden base, to the red servo arm. The red servo arm was then attached to a threaded steel axel using two nuts (one on either side) counterturned to fix the servo arm in place. The rack and pinion of the mechanism (right image in Fig. 17) were selected 17

18 from an assorted pack of various plastic racks and gears meant for remote controlled cars. The gear with the greatest radius that would still fit within the diameter of the body tube was chosen to afford the spring mechanism the largest moment arm and, therefore, highest mechanical advantage possible within the space constraint. By increasing the mechanical advantage of the spring system, less stress is place on the rack, pinion, and other components of the deployment mechanism. A spring with a spring constant of 15 lb/in, uncompressed length of 1.5, and diameter of 0.6 was chosen. Different spring constants were tested (5 lb/in and 10 lb/in) to determine the least spring force that could be used to adequately deploy the wings and resist aerodynamic loading that would attempt to fold the wings during flight. It was determined through testing with the prototype that the stronger spring (15 lb/in) was the minimum acceptable spring constant to deploy the wings with adequate force and prevent wing snagging during deployment. Unfortunately, the team was unable to put the glider through wind tunnel testing to determine exactly how much force would be required to hold the wings open during flight. A decision was made based on the apparent rigidity of the wing leading edge rods when fully deployed and the ease by which the wing overcame snags during deployment. A 3D printed ABS piece (blue-green in Fig. 18) was used to mate the rack with the spring assembly. Figure 18. Actual pieces inside (left) and outside (right) of the deployment mechanism Fig. 18 above shows the pieces of the deployment mechanism used on the actual glider on launch day. Most obviously, all of the extra wood from the prototype has been excluded. The interior mounting piece has been divided into two separate pieces to make it possible to fully assemble the mechanism inside the glider tube. The team found it quite challenging to properly seat the rack and pinion once the axel and top assembly had been mounted to the body tube. However, it was necessary to mount the top section and gear first due to space restrictions and difficulty in fastening the internal components. The team had to make custom tools (attach wrenches to long rods) in order to fasten internal parts from the mouth of the body tube. Fig. 19 below provides further insight into how the various portions of the deployment mechanism were assembled and mounted to the body tube. Several notable modifications were made to these pieces after the photos shown in Fig. 18 were taken. First, it may be noticed that no guide wall exists on one side of the rack where there were previously fiberglass pieces used as guides (see the prototype in Fig. 17). Originally, it was felt that the precaution of this extra guide was unnecessary since the rack was already tightly held on 3 sides. However, upon testing of the deployment mechanism after 18

19 integration with the body tube, it was discovered that the rack was slipping off of the gear. To remedy this problem, a small plate of fiberglass was again positioned to hold the rack along its track. Another change that was enacted on the final deployment system was the method by which the gear was attached to the axel. Originally, the gear was to be attached to the threaded axel solely by epoxy. However, upon testing, it was discovered that the epoxy would not bond to the plastic gear. The gear slipped and broke free from the epoxy easily. After some research, it was found that many other glues/epoxies that were readily available would not bond well to most plastics. To securely attach the gear, part of the gear was grinded down and two large washers were used along with two oppositely turned nuts and thread locker to secure the gear. Once these measures had been taken, no slipping was observed in the gear and the deployment system functioned without further problems in testing. The final modification made to the parts shown in Fig. 18 was to further sand material off of the wooden piece mounted to the exterior of the tube to reduce unnecessary weight and decrease the drag profile. The clevises attaching the wing leading edges to the mount were also reinforced with epoxy. This reinforcement had the duel purpose of making the wing attachment more rigid during flight. Looking at Fig. 19 below, on the right side may be observed the full deployment mechanism without the body tube and leading edge rods. On the left the pieces may be scene as they are mounted inside the body tube. Figure 19. Deployment components integrated (left for demo only) It was decided that the wing should be hidden inside the body of the rocket during launch to both protect the wing and prevent it from destabilizing the rocket during launch due to the creation of extra drag. The wing was stowed by cutting a slit in the glider body 23 in length and 0.2 in width. This slit was cut with the assistance of the machine shop and an effort was made to make the slit as thin as possible without causing the wing to jam to avoid weakening the body tube. Sharp edges were removed from the slit using sandpaper and a Dremel tool. To hold the centerline of the wing down during flight, wire from a clothes hanger was secured to the body of the rocket using drilled holes and epoxy. Care was taken when folding the wing into the body to lessen the probability of snags. This was done by folding the wing in equally spaced pleats (like an accordion). To arm the deployment system, the carbon fiber leading edge rods of the wing were folded and inserted into two, larger diameter, fiberglass rods attached to the booster section of the rocket. This procedure had to be done during attachment of the full glider section to the booster section. The fiberglass rods served to hold the leading edge rods collapsed in launch configuration against the force of the spring. Using this system, the leading edge rods of the wings would be under maximum loading during launch (folded) configuration and, upon separation of the glider from the booster, the leading edge rods would withdraw from the fiber glass rods and spring open. Fig. 19 below shows the glider in both launch (left) and glide configuration (right). The major downside of this method is that the least amount of force 19

20 holding the wings open is exerted when they are fully deployed, meaning that it is more likely that the wings will collapse under the force of drag during flight. To counteract this, the 3D printed part used to attach the rack to the spring was modified to ensure that the spring was still under some compression when the wings were in the deployed configuration. The opening of the wings is halted when the 3D printed piece wrests against the wooden mounting piece at the mouth of the spring collar. Figure 20. Deployed and launch configurations of the glider wing H. Wing Material Manufacturing The wing itself was made from ripstop nylon a material used in most parachutes. The wing shape was traced onto a length of this fabric with some extra on the edges to allow for hemming. The shape was then cut and sewn. All stitches were zigzag stiches to cover more surface area and increase strength. The trailing edge was sewn with a standard hem as was the center of the leading edge. The rest of the leading edge was sewn as a loop to allow the wing rods to be threaded into the wing material. Duct tape was used on the center of the leading edge to reduce the likelihood of the fishing line previously mentioned from ripping through the material. A section of a carbon fiber rod was used as the stopper in the center of the trailing edge. This was sewn into place with more fabric acting as a sealed pocket. III. Testing In the section below, the computational design and vibrational testing of the rocket are outlined. A. Computational Design and Analysis The initial conceptual design of the rocket was completed with the use of RockSim. RockSim is a rocket simulation tool that allows users to design and test fly a rocket. It provides valuable information such as maximum height that the rocket will reach, the maximum speed of the rocket, the weight of the design, and the static margin of the design to determine whether or not the rocket will have a stable flight. For the RocketSat team, RockSim was primarily used as a means to visualize the rocket and to ensure that the design was stable. Fig. 21 displays the rocket design in RockSim. 20

21 Figure 21: RocketSat RockSim Design RockSim was also utilized to determine certain launch parameters. A separate open source rocket analysis program called RASAero was used to validate the RockSim results. The launch data that was obtained from RockSim is displayed in Table 6 and the corresponding RASAero calculations are given in Table 7. It is crucial to note that there will be differences in the RockSim and RASAero calculations. RASAero has some severe design limitations that prevented the RocketSat team from analyzing their design in RASAero. For instance, in RASAero only one set of fins per rocket is allowed, external objects (i.e. the wing mount box) cannot be modeled, and the closest engine in RASAero to the one that the RocketSat team used produced more thrust than the RocketSat rocket engine. Therefore, it was not expected that the RASAero results and the RockSim results would agree perfectly; however, it was expected that the results should be somewhat similar. The stability of the rocket was expected be relatively unaffected because the second set of fins (the mid-section fins) primary purpose were to counteract the extreme loss in stability created by the wing mount box. Table 6: RockSim Launch Data Table 7: RASAero Data Comparing the RockSim and RASAero results, it was shown that both simulations produced similar results despite the differences between them. As expected, the RASAero simulation produced a higher amount of thrust and, correspondingly, had a higher acceleration. The time to apogee, projected altitude, and maximum velocity for both simulations was very similar for both simulations. Another analysis tool considered in the aerodynamic testing of the Rocket was Missile DATCOM. Missile DATCOM is a program that analyzes missile aerodynamics. Due to time restrictions and complications with the code, a complete rocket/glider analysis could not be completed prior to the launch; however, some sample test cases were run at varying altitudes and Mach numbers. Fig. 22 displays the rocket input file that was created to be used in Missile DATCOM and Fig. 23 displays the corresponding sample output file. 21

22 Figure 22: Missile DATCOM Input File Figure 23: Sample Missile DATCOM Output File In the Missile DATCOM sample input file, the simulation was done for a variety of Mach numbers at an altitude of 800 feet. Through varying the Mach number, altitude, and angle of attack, the aerodynamic coefficients shown in Fig. 23 can be determined at each stage of flight. The untrimmed flight condition was considered for this simulation. The wing mount block was modeled on the rocket as a protuberance in the flow around the rocket. Due to a lack of time and manpower coupled with some significant difficulties encountered with formatting the input file, a complete aerodynamic analysis of the rocket was not able to be completed. 22

23 The final rocket analysis that was completed was a rocket vibration study. To simplify the vibrational analysis, the rocket was approximated to be a non-uniform beam under free-free boundary conditions. When attempting to mathematically describe the vibration of a continuous beam, the team began by considering the equation of motion for a continuous beam. Coupled with the equation of motion there was, as always, the free body diagram of a segment of a beam with variable mass,, variable stiffness,, and distributed load,. The free body diagram is given in Fig. 24 and the subsequent dynamic summation of forces (where the net force is equated to the mass of the beam segment multiplied by its acceleration) is given by Eqn. (8). It should be noted that, since derivations of beam vibration equations are not the focus of this paper, only brief a derivation is provided. A more thorough derivation can be found in Structural Vibration 8 by C.Y. Wang Figure 24: Free Body Diagram of a Continuous Beam Segment (8) Taking the limit of Eqn. (8) as approaches zero yields Eqn. (9). (9) In conjunction with Eqn. (9), Euler-Bernoulli beam theory was also considered, as shown by Eqns. (10) and (11). Eqn. (11) is merely the derivative of Eqn. (10), the classical Euler-Bernoulli beam theory equation. (10) (11) Taking the derivative of Eqn. (11) and substituting the result into Eqn. (9) yielded Eqn. (12), the differential equation of motion of the beam in question. It is important to note that Eqn. (12) has both spatial and temporal components. Allowing for the deflection of the beam to vary with both time and location allowed for any vibration within the beam to be accurately described Using the uniform beam case allowed for a great simplification of Eqn. (12). By definition, a uniform beam is uniform and, thus, mass and stiffness are not functions of spanwise location. Furthermore, it was assumed that the distributed loading was negligible; thus, Eqn. (12) becomes Eqn. (13). (13) 23

24 Following this simplification, a separation of variables was done that divided the beam deflection into spatial and temporal components. Then each of the two differential equations (spatial and temporal) was solved simultaneously. The solution process (including finding the required constants through the use of pertinent boundary conditions) is lengthy and can be found in Structural Vibration 8. A characteristic equation resulted from the solution of these equations. This equation is given by Eqn. (14). The boundary conditions used in the aforementioned solution process were free-free boundary conditions to imitate a rocket in flight. (14) Once the roots of Eqn. (14) were found, they were used in Eqn. (15), the eigenfunction. Like before, this derivation is lengthy and Eqn. (15) was taken as a given for this study; however, like Eqn. (14), this derivation can be found in Structural Vibration 8. The eigenfunction is a spatial function describing the location of the beam. By substituting the various roots of Eqn. (14) into Eqn. (15), the plots for each of the mode shapes were obtained. (15) To obtain the desired vibrational frequencies, Eqn. (16) was considered. Eqn. (16) is a multi-degree-of-freedom (MDOF) matrix equation of motion containing the required mass and stiffness matrices. (16) The mass and stiffness matrices in Eqn. (16) are defined via Eqn. (17) and Eqn. (18), respectively. (17) (18) It should be noted that for the uniform beam case if i is not equal to j. Therefore, for a uniform beam, M and K are diagonal matrices. However, for the non-uniform beam case, this does not hold. The non-uniform M and K matrices are symmetric, but they are not diagonal. To approximate the non-uniform beam vibration, multiple uniform beam modes were used. The idea is analogous to dividing a non-uniform beam into multiple uniform beams. As more uniform beam modes were used, the non-uniform beam vibrational frequencies began to converge. The rate of convergence was entirely dependent upon the properties and variations of these properties in the nonuniform beam. Once M and K were determined, Eqn. (19) was used to find the vibrational frequencies. The eigenvalues of Eqn. (19) are functions of the non-dimensional frequencies. (19) Once the eigenvalues were determined from Eqn. (19), Eqn. (20) was used to determine the non-dimensional natural vibrational frequencies. (20) 24

25 Thus, the natural vibrational frequencies corresponding to each mode shape were found for both a uniform and non-uniform beam with free-free boundary conditions. The RocketSat rocket was modeled as both a uniform beam and a non-uniform beam. It is crucial to note that both of these simulations are approximations. Within both the uniform and non-uniform cases, the mass and stiffness variations were approximated. The idea behind performing this vibration study was merely to obtain an idea of the vibrational frequencies that the rocket would see in flight. Since the mode shapes (given by Eqn. (14) and Eqn. (15)) are not functions of mass and stiffness, the uniform and non-uniform approximations both had the same mode shapes. Fig. 25 displays the approximate mode shapes for the RocketSat rocket. Figure 25: The Free-Free Mode Shapes for the RocketSat Rocket Table 8 gives the natural frequencies for the non-uniform approximation the uniform approximation. Table 8: Approximate Rocket Natural Vibrational Frequencies An online search found that the typical vibrational frequencies of a rocket in flight were anywhere from 8 Hertz to 20 Hertz. Therefore, assuming that the RocketSat rocket will have a typical vibrational frequency, it likely will not vibrate at a natural frequency. However, a vibration test of the actual rocket on the shaker table (0 Hertz to 30 Hertz) did not show that a natural frequency was obtained. Therefore, it can safely be assumed that the rocket will not vibrate at a natural frequency. 25

26 B. Vibration Test Mount The vibration mount was designed and built by a team led by Brendan Mangan. The dimensions of the rocket and the threaded rod on the shake table were taken and a design was built around them. The mount was connected to the shake table via a coupling nut that had been attached to a square piece of wood. On the top end of this wooden block a PVC cross was attached using PVC clamps. PVC elbows were then attached to the to this cross allowing for four vertical PVC pipe pieces to hold the rocket. Figure 26. Vibration mount. Rocket loaded in vibration mount that is attached to shake table. A half inch diameter pipe was used to ensure that the structure would remain stable during testing, but this created the problem of the rocket diameter being smaller than the distance between the vertical pipes across from one another. To solve this problem, rubber matting was cut into strips and wrapped around each of the vertical pipes. To ensure that the rocket remained snugly attached to the mount during testing, tape was wrapped around the vertical pipes with the rocket inside. The materials were purchased at Home Depot and the total cost of the materials was approximately $25. C. Vibration Test Procedure and Results Before launch, the rocket was mounted to the University of Alabama shaker table to ensure that components would withstand typical vibrations to be experienced during launch. A variety of frequencies were tested to search for observable vibrational modes. The range of frequencies tested was from 0-30Hz, which is typical in rocket launches. Figure 27 below shows the glider mounted to the rocket test stand. It may be observed that the entire rocket was not mounted to the shaker stand in one piece, as it was feared that the weight and height of the full vehicle could cause damaging loads to be placed on the shaker table. Instead, the glider and booster sections were mounted separately. While this arrangement definitely skewed testing for vibrational modes, it was still valuable for testing whether the system was structurally sound under vibration. The vehicle was secured to the test stand at the lowest point possible to best simulate vibrations from the motor section. However, the method of restraining the base of the rocket during the vibration testing was definitely not a perfect representation of actual conditions experienced during flight due to the large area of contact between the vehicle and the test mount. In the air, the rocket would be unrestrained. A test stand with firm but minimal contact with the base of the rocket would be ideal. 26

27 Figure 27. Glider section during vibration testing During vibration testing there were no observable modes found. No structural failures were observed during the test or inspection after the test. Each test lasted 1 minute and was conducted twice. After testing was completed, the full glider-rocket system was again integrated and tested for full functionality. No malfunction was observed. Having found no observable modes in the range of operation, structural damage, or loss of system functionality, it was determined that the vehicle passed the vibration test. IV. Launch Day and Results The team arrived at the launch site at approximately 10 am on April 4th, Music City Missile Club (MC2) had begun setting up the range in Manchester, TN and announcements for the day were made shortly after. These announcements included wind speeds and general safety guidelines. It was during the meeting that team members notified the MC2 officers running the launch that the RocketSat launch vehicle would be deemed a heads-up flight. This was done because it was unknown how far the glider would go and how well it would work. By notifying the officers of MC2, the club was able to tell all spectators to keep an eye on the launch and ensure that everyone stayed safe. After the meeting team members reviewed safety rules and on-site launch rules to ensure that everyone was compliant that day. 27

28 Figure 28. E-match wires taped to the exterior of the rocket prior to the ground test Next, the rocket was prepared for a ground deployment test. This test was conducted to ensure that the black powder charges used during descent were large enough to separate the rocket. The rocket was prepared as if it were to be launched so that the weight distribution during the test was the same as during launch. The black powder charges were placed on the designated bulkheads but instead of the electric matches running to the altimeter, they were pulled through a breather hole in the avionics body tube as can be seen in Fig. 28 above. Ten feet of wire then ran from the e-matches to a team member. With the range notified and everyone standing back, each charge was tested by connecting the e-match to a 9V battery. Figure 29. Rocket in launch configuration prior to ground charge test The first test was unsuccessful because the charges were too small and were unable to break the shear pins. After speaking with the Vice-President of MC2, Chris Dondanville, it was decided to triple both charges such that the separation charge would be 1g of black powder and the parachute charge would be 1.5g of black powder. Fortunately, the team was able to quickly replace the black powder charges with new ones and re-run the test sequence as can be seen in Fig. 29. Just prior to the test the spring in the wing deployment mechanism was compressed such that the wing rods would deploy at separation. This was done to test the wing deployment mechanism and to see if the rods that had been glued to the booster body tube to hold the wing rods in place during ascent would affect separation. The second test was successful with a full separation of the payload from the avionics bay and the booster from the avionics bay. The wing deployment mechanism also worked correctly with the wing rods deployed at the correct preset angles. 28

29 Figure 30. The wing was folded inside the body tube during launch preparation At this time the team prepared the rocket for launch. The payload team put the wing on the glider and folded the fabric inside of the body tube as shown in the left and right images of Fig. 30. One team member prepared the motor and recovery electronics, while others prepared the parachute. Figure 31. Launch preparation Once all preparation was complete on the sub-components, the rocket was put together with either rivets and shear pins at each interface. The left image in Fig. 31 shows the team putting in shear pins. It was at this time the team decided to name the rocket Lawn Dart, and each team member present signed a fin. The right image of Fig. 31 shows the name written on the side of the rocket. 29

30 Figure 32. Placing the rocket on the launch rail Once the rocket was checked by Chris Dondanville at the Range Safety Officer station the rocket was taken out to the launch pad. Due to the settings on the payload electronics there was a limited window for the data to be captured. This meant that the team had to be the last on the pad, and the first one to launch in the group. MC2 was able to accommodate this range and had the rocket launched within minutes of the payload electronics being turned on. Fig. 32 above shows the team placing the rocket on the launch stand just moments after turning on the payload electronics. Figure 33. Recovery configuration of the rocket 30

31 The ascent of the rocket was smooth with little wobble, it began to arc over and the altimeter worked perfectly in separating the glider from the booster. The booster continued to arc over and unfortunately the parachute did not deploy. This was determined to be that the black powder charge was not large enough to separate. The glider initially stabilized and then nosedived to the ground. The problem was determined to be too much weight in the nosecone. The recovery configuration of both the booster and glider can be seen in Fig. 33. Figure 34. Recovery of the booster and glider Recovery of the glider was not difficult as only the nosecone and a few inches of body tube had punctured the ground, as seen in the right image of Fig. 34. The glider was pulled from the wet ground easily and it was found that there was no exterior damage. The booster underwent severe damage with over a foot of the body tube being underground. After retrieving the booster the team discovered that the entire avionics bay had been pushed inside the booster body tube by approximately 2 inches which resulted in the fiberglass zippering 10 inches. The right image of Fig. 34 shows the full extent of the damage the booster underwent. Figure 35. The team after the launch 31

32 Although damage did occur, all parts of the rocket were recovered and the team was able to return to the university that night. Before the team returned to the university the team took one last team photo to commemorate their hard work as seen in Fig. 35. A. In-Flight Data The altimeter reports that the rocket reached an apogee of 740 feet, where the separation charge was ignited. The altimeter data states that the second charge was blown at 598 feet, which is the first data point below 600 feet for the altimeter. The altimeter data can be seen in Fig. 36 below. The conclusion is that the altimeter was properly working during descent and the charge was not large enough to separate the booster section from the avionics bay. This is disconcerting as the charge was able to separate with the same amount of black powder during a ground test an hour before. The approximate maximum velocity achieved during flight was 245 ft/s, as can be seen from the graph in Fig. 37 below. Figure 36. Altimeter Data- Altitude v. Time Figure 37. Altimeter Data- Velocity vs. Time 32