AUVSI Student Competition 2014 Journal Paper

Transcription

1 AUVSI Student Competition 2014 Journal Paper Mechanical Engineering Team: Javier Lichtscheidl, Adam Nguyen, Eder Avila, Robert Taylor, Travis Morgan, Adam Dominguez, Mark Nguyen, Jerome Moscoso, Urian Borda, Rene Correa, Timothy Gulliver, Alex Bautista, and Chris St. Dennis. Advisors: Dr. Nina Robson and Jake Bailey Abstract This paper describes the design, manufacturing, and testing approach taken by the Titan UAV team in order to be able to compete in the annual Student Unmanned Aerial System (SUAS) competition. The design approach for this year s team was based on two main criteria: Improvement in structural and stabilizing design from the plane designed the previous year, and the requirements needed for the SUAS competition. The way this was completed was through a vast range of material analysis, cost effective design, and creating a new plane design to fulfill the requirements for the competition. Since the team is entirely made up of mechanical engineers, the plane was not off-the-shelf, but rather designed and manufactured at CSU- Fullerton. Most of the material on the plane was made out of carbon-fiber in order to create a high strength-to-weight ratio that would improve on rigidity and structural stability. The payload consists of all the electrical components that will enable the UAV to fly autonomously, follow the corresponding waypoint navigation system, and gather imagery that will be streamed to ground control. This includes the Raspberry Pi and Raspberry Pi camera, the Ardupilot board for autonomous flight controls, and LiPo batteries needed to power the electronic components as well as the 450 hp motor. This composition makes it crucial to provide stable flight, intelligence, surveillance, and reconnaissance to ground control during flight missions.

2 Table of Contents 1. Systems Engineering Approach Mission requirements analysis 1.2 Design rationale 1.3 Expected performance 1.4 Programmatic risks and mitigation methods 2. UAS Design Design Description Design of air vehicle Method of autonomy Data link Payload Ground Control Station Data processing Mission planning 2.2 Mission Tasks being attempted 3. Test evaluation results Mission task performance 3.2 Payload system performance 3.3 Guidance system performance 3.4 Evaluation results supporting evidence of likely mission accomplishment 4. Safety considerations/approach Specific safety criteria for both operations and design 4.2 Safety risks and mitigation methods 5. Conclusion.19 2

3 1. Systems Engineering Approach 1.1 Mission requirements analysis The goal of the SUS mission is to gather accurate ISR autonomously and in real time in order to aid firefighters in identifying fire activity. Based on these goals and the SUAS competition requirements the mission will entail the following requirements: Perform autonomous flight Identify Targets (Alpha-Numerical) Locate Targets (GPS stamp) Complete mission in required time (less than 40 minutes) Weigh less than 55 lbs. Be able to carry a payload Fly waypoints with maximum 100 feet tolerance Provide real time target information 1.2 Design Rationale The air vehicle design was derived from the competition requirements listed in 1.1 and also improvements from last year s design. Last year s design had some structural flaws that added instability within the vehicle during flight. This year s team chose to add structural integrity, improve flight stability, and to have a maximum vehicle weight 30 ± 5 lbs. The autonomous flight board system was chosen to remain the same since it worked last year, but to improve the image recognition system as well as the amplitude of the signal being sent from the air vehicle to ground control. Besides these main design needs and changes the team based the rest of the design on the mission requirements. The air vehicle detailed designed for both structural and payload components will be described in detail in section 2.1 of the report. 1.3 Expected Performance Having a stall speed of about 37 mph, the motor will need to run at 76% of its power in order to generate lift, which then will reduce the amount of power to about 40% at steady state in order to cruise at its calculated cruise speed. The autonomous system used last year is being used again and therefore autonomous performance is expected to work efficiently and effectively. Structural rigidity and design of the air vehicle will have a better stability than that of last year s design as well as high strength-to-weight ratio in order to carry the necessary payload. 1.4 Programmatic risks and mitigation methods Section 4 correlates with the risks plausible and the mitigation methods involved. As required in the SUAS competition rules, the Ardupilot board is capable of implementing the air vehicle to react in the following cases: loss of signal of about 30 seconds, loss of signal for 3 minutes and termination, and safety kill-switches used to prevent the air vehicle from causing any injuries, damages, or any other hazard to any team member, spectator, or any other person whom might be within close proximity or flight range of the air vehicle. 3

4 2. UAS Design 2.1 Design Description The Titan UAV was designed to assimilate the structural design of Israel s Heron UAV (shown below) and consists of a twin-tail design that is attached to the wings via booms and is carried via a pusher propeller. Figure 1: Israel s Heron UAV. The air vehicle will be able to carry a payload and have a maximum weight of 30±5 lbs. depending on necessary payload components. The structure of vehicle consists of three materials: carbon fiber, expanded polypropylene (EPP) foam, and aluminum. The fuselage consists of a three-ply inner and outer prepreg carbon fiber cloth of about inches thick and honeycomb layer in between (Nomex) of about inches. This resulted in a fuselage thickness of about 0.2 inches, but an increase in strength, stiffness, and overall structural stability. The wing profile chosen was E-214 from Airfoil Tools and was selected for its high lift to drag ratio. The profile specified was also chosen for having lift at a negative angle of attack in order to minimize the chance of having any tail-strike. Both the rudders and elevator were chosen to have the airfoil profile NACA-0012 from Airfoil Tools. The air profiles chosen were for stability purposes and to be able to handle the backwash generated from the pusher propeller. The EPP foam for both the wing and empennage were covered in carbon fiber cloth using a wetlayup technique. For structural rigidity and support for the empennage, the wings have two carbon fiber spars that are bonded on the inside of the wing at the chord length height. The overall structural composition of the air vehicle is to be able to carry its payload with added strength and minimizing the amount of weight added to the entire system at the same time. The rest of the structural components will be addressed in more detail in section of the report. The payload consists of the following: the 7.4V LiPo batteries which will power the entire electronic system, the gimbal for the camera system, the autonomous flight control board, servo receivers, motor, electronic speed controller, Raspberry pi, Raspberry pi camera, Airmax Bullet, Pitot tube, 3DR Telemetry, Receiver. These components were used in last year s design and will be used again due to the simplicity of the power setup and communication between the air vehicle and ground control. The rest of the payload components will be described in section of the report. 4

5 2.1.1 Design Description of air vehicle: When approaching an in-house designed airplane for a competition such as the one described above, it s obviously desirable to design towards ensuring mission success. It was decided that the best way to ensure mission success was to incorporate redundancy and safety into the design. What his specifically meant for the structure team was to follow the design of a proven system and apply it to this particular application. The other constraint that was a driving factor for many of the decisions made was that of cost. To lower the cost, while at the same time design for reliability the team decided to explore various layouts of airplanes and weigh their benefits, putting an emphasis on low stall speed, high payload capacity, and ease of fabrication as well as modification. This year s design can be split into two portions: Structural and Payload designs. The structural design included the following parts: Fuselage, Wings, Empennage, and Landing Gear. Payload consisted of two branches: On-board communication and Flight Controls. To assist with this task, a design process involving various morphological charts, pugh charts, and basic literature surveys led us to make the basic decisions outlined below. Table 1: Structure Decisions Table. Fuselage Topic Construction Type Wing Placement Wing Shape Tail Style Motor Arrangement Landing Gear Style Decision Twin boom tail pusher plane configuration High Wing Rectangular (constant chord) U-Tail Single Motor, Aft Fuselage mounted Tricycle The Fuselage, as shown below in Figure 2, consists of three sections: the nose, the middle, and aft sections of the fuselage. These sections are connected at two locations with bonded and bolted bulkheads. The two joint locations are between the nose and mid fuselage and the second is between the back portion of the mid fuselage and aft section. The design of these bulkheads is mainly to be able to support shearing stresses from both static and dynamic loadings. The body of the fuselage is made Figure 2: Fuselage Top Assembly. out of carbon fiber and an inner layer of Nomex (honeycomb layer). This composition allows for a light weight body, while maintaining strength to support the weight of the entire plane and its corresponding static and dynamic force loadings. The shape chosen has relatively low effects on flight in terms of lift to drag ratio, but each radius created at the corners decreases the amount of vortices generated in flight. The main purpose of the fuselage is to support be able to carry a payload of about 15 lbs. and support a gross weight of approximately 30 lbs. 5

6 Wings The wing assembly, as shown in Figure 3, consists of a 7 ft. wing span and a chord length of 1 ft. The airfoil chosen was the E-214 foil for its high lift to drag ratio and low Reynolds number applications. Since the plane will be locating targets, the flight speed will be around 55 mi/h and therefore it requires a low Reynolds number airfoil to be able to fly at a laminar state and reduce form drag. Another reason why the E-214 foil was chosen was to generate lift at a negative angle of attack. Since the UAV will consist of a pusher motor configuration and a twin tail via boom configuration, the high lift at a negative angle of attack will allow for the plane to lift before either the tail or propeller can strike the ground during takeoff. The materials chosen for the wing consisted of expanded polypropylene foam for its impact energy absorption, chemical resistance for the use of resin coating, and its high strength to weight ratio. The wings will have an outer layer consisting of Figure 3: Wing Assembly. carbon fiber and a PT-2050 resin. The coated carbon fiber will add strength to the wings and a smoother surface finish to reduce parasitic drag from skin friction. Since the wings have to carry large force loads during takeoff and surveillance flight, two carbon fiber spars were added to support the wing from bending moments. Empennage The empennage, Figure 4, is responsible for keeping the aircraft directionally and longitudinally stable throughout all aircraft weight loads, feasible flight speeds, altitudes, and allowable center of gravity. All these scenarios can be summarized between the extreme conditions during takeoff and landing. Our design consists of booms attached to both the vertical and horizontal stabilizers; the booms are made of composite carbon fiber contributing strength and reduced weight. The booms thread into the boom tail interface that is machined from Aluminum In order to be able to sustain the load from the empennage, the wings will have two inner spars running through the wing span centered with the chord length of the airfoil. The Horizontal and Vertical Stabilizers are the NACA 0012 airfoil made of expanded polypropylene foam, which allows the profile to generate an optimum lift, drag and control. The airfoils also consist of the carbon fiber-pt2050 layup for high strength to weight ratio and reduced skin friction. Arriving at this design was through a down selection process of reverse engineering from the previous years Figure 4: Emepennage Top Assembly. 6

7 Unmanned Aerial Vehicle in Last year s team mentioned how the aircraft was not as stable as they preferred forcing the pilot to rely heavily on the controls, in order to improve upon this issue a primary objective of ours is to increase the overall stability. We initially narrowed the empennage design to two choices being an inverted V-tail or a U-tail. During the preliminary design review we discovered the inverted V-tail had a novel look but it would have been difficult and expensive to manufacture. The inverted V-tail also combined the functions of the elevator and rudder into one control surface meaning reduced control of the aircraft which is not what we desired. The U-tail was chosen for the feasibility of manufacture, more stability control, and strength to support the backwash created from the propeller onto the tail. The improvement of this UAV design is the critical benefit in increasing the horizontal tail volume coefficient (HTVC) to have a value that upgraded the overall stability. With the increase in the HTVC the aircraft isn t as maneuverable as a fighter jet, but more like a transport jet which is ideal for our flight regime being reconnaissance and surveillance. By raising the HTVC, there is a larger area and that means a larger force equipping greater leverage when restoring the overall longitudinal trim once the UAV encounters any disturbances through its operating rigorous envelope. Landing Gear The final structural component consists of the main landing gear and the steering gear Figure 5. The landing gear is made mostly from aluminum 6061 in order to be able to carry the plane s weight at static loadings and sustain impacts generated from landing. The tires are pneumatic and made of rubber material to dampen the impact forces it receives. The main landing gear consists of an aluminum bracketing that connects to the fuselage and extends at an angle of 45. This configuration will allow the main landing gear to distribute the impact force equally and to reduce the bending moments generated from landing. The specific height chosen for the main landing gear was designed to reduce the chances of having a tail or propeller strike during the takeoff and landing phases of the plane. The main design for the front landing gear was to have a steering mechanism that can taxi the air vehicle and also keep it straight during takeoff. Figure 5: Steering and Main Gear Design. 7

8 2.1.2 Method of Autonomy: As expected of the APM 2.5, autonomous flight for the air vehicle is capable in conjunction with software called Mission Planner. Mission planner is a ground control station used for the dynamic control of an autonomous vehicle. Using the proper telemetry, Mission Planner can monitor the vehicle s status while in operation and aid in the overall stability of the air vehicle while in flight through its given waypoint coordinates. The flight controller board used this year was also used the previous year and accomplished the given autonomous flight mission requirements. Mission Planner is interfaced using windows, which will be controlled from ground control and will relay the information back to the air vehicle via the Airmax Bullet. As shown below, the waypoint coordinates can be created generated through Mission Planner to communicate to the APM 2.5 in order to be able to have autonomous flight. Figure 6: Mission Planner interfaces with the APM 2.5 for Autonomous Flight Data Link: Data relating to image recognition will be done based on the figure below: Figure 7: On-board to Ground Control Communication. 8

9 2.1.4 Payload: The payload system consists of two branches: Flight controls and on-board communication. The Flight control system consists of the Ardupilot APM 2.5, which consists of an open source development, 3-axis gyro, accelerometer, and barometer. The APM contains GPIO ports for servos and sensors that will be used for the plane to be able to fly autonomously. The software involved is Mission Planner. Mission planner interfaces google and will use the coordinate system specified by the SUAS competition rules. With the Mission Planner, the APM will be able to set up the flight control system to follow a waypoint system and be able to configure the vehicle s controls to fly within the discrepancy value of less than 100 feet. Based on criteria needed to fulfill the competition rules, the APM 2.5 was chosen as the primary source for flight controls (Table 2). Table 2: Flight Controller Matrix. The servos chosen were the A402 for cost, durability, volume, and compatibility with the APM 2.5. The motor chosen was a Scorpion motor that will run on two 25.9V in series and a single 7.4V LiPo battery in parallel to run a total of 33.3V. This single motor has the capability of producing about 450 hp and will need the appropriate amount of voltage in order to reach takeoff speed at about 76% throttle. The air vehicle control frequencies that the air vehicle will be operating under are 2.4 GHzand 915 mhz. The second branch of the payload system consists of the on-board communication system and image recognition. The system will consists of a 7.4V LiPo battery that will power that sector of the payload system, the Raspberry pi, Raspberry pi camera, the camera gymbal system, and the Airmax Bullet. The Raspberry pi camera will feed a live stream to the Raspberry pi, which in turn will be sent to ground control via the Airmax Bullet. More will be explained in section

10 in relation to ground control and data link that was described in section The gimbal system designed below will enable for the camera chosen to be rotate 360 degrees and also reaching tilt of about 90 degrees. The gimbal system was 3-D printed out of ABS and can be seen on the right: Figure 8: Camera Gimbal System. The air vehicle block diagram containing the entire payload can be seen below: Figure 9: UAV Block Diagram Consisting of payload in corresponding locations on air vehicle. 10

11 2.1.5 Ground Control: The ground control station objectives are to process an autopilot system, image processing, and navigation along with manual flight. For the autopilot system, the Mission Planner software will be utilized to meet the competition rules. The Mission Planner Software entails navigational controls where wave points can be set through GPS coordinates that will allow the UAV to fly autonomously. The Mission Planner software will also display the elevation, velocity of the UAV, and the GPS coordinates. The 3DR telemetry, transmitting 915 Mhz, will be used for the navigation and autonomous flight. One of the main objectives of the ground control station is to maintain a strong communication link with the UAV. The Airmax Bullet M5 was chosen to be part of the UAV because it s capabilities of transferring data up to 100Mbps+ at longer ranges. The Airmax Bullet will be connected to the Airgrid Antenna with an antenna gain of 27 db. The performances of these two components were chosen based on the analysis on the component s specifications. The analysis is composed of three equations below: Table 3: Airmax Bullet signal analysis. 54 Mbps With Airmax Bullet Distance (miles) FPSL (dbm) EIRP (dbm) SOM (dbm) The least system of operating margin that a 5 GHz antenna can have is 10 db before the connection depreciates. From the figure shown above, at a distance of 4 miles the system of operating margin is 20.4 db. The ground control station will still have a strong link with the UAV between 1 to 4 miles which meets the requirements of the UAV transmitting data to the ground control station. The Airmax Bullet and Airgrid will be interfaced with the image processing computer system. The image processing system will composed of a computer and connects with the UAV via I.P. addresses. This will allow us to control the onboard system and camera from the ground control station. The image processing computer will be connected to a gigabit switch which routes the IP addresses connected to the navigation Computer shown in the block diagram. The gigabit switch is used to access any data from the navigation computer such as the GPS coordinates Data Processing: Data transfer will be made using the Airmax Bullet and Receiver Antenna as shown in Figures 10, 11, and 12 below: 11

12 Figure 10: Receiver Antenna. Figure 11: Transmitter antenna. The data processed will be made using the Raspberry Pi in the following configuration shown in Figure 11: Figure 12: Data Processing Block Diagram. When the target is recognized through the Raspberry Pi camera, the image will process the video feed through the Raspberry Pi and transmit the live feed to the Receiver antenna, which in turn will be visible by Ground Station. The goal is to be able to capture images of targets while also live streaming video. This will require the Raspberry Pi to compress the images and send them to ground Control. The other goal is to be able to attach and retract an infrared filter to the system. This would enable the UAV to capture the I.R. targets during the mission, but it is still undergoing construction and design planning. 12

13 2.1.7 Mission Planning: Pre-Flight Assignments: Before the flight it is very important that every member of the flight crew knows what the mission requirements are and how to fully achieve all the mission goals. This briefing must always be achieved at least 12 hours before the deployment assignment. During this pre-flight assignment, the captain of the team announces the limits, the essential objectives, the supplemental objectives and safety concerns depending on the environment, time of day, region, and category of flight regime. The weather forecasts need to be evaluated at this period to establish precipitation, temperature, and wind with respect the restrictions of the systems. After all these checks have been secure each member has to confirm they re ready for the next deployment assignment. Deployment Assignments: Before approaching the flight demonstration, all the essential parts for certain subsystems are put into separate containers for organization. Once the team arrives onto the field each recognized group will open their respective container and start to assemble the specific system they are in charge of. The containers are categorized into an aircraft preparation consisting of payload, fuselage, wings, empennage, and landing gear. Also, payload support hardware such as ground control and imaging stations will have their own containers. Once the team has performed all appropriate tasks the head commander must update all flight crew of all changes to the mission. Checks are then completed to the operation criteria while adhering to the competition radio restrictions. In order to cut down potential issues connecting and fastening the modules, it s crucial to make certain the standalone functionalities of each component. Then the final inspection is finished by the head commander to bring to completion the operation criteria. Takeoff Technique: The takeoff stage of the mission is the most dangerous segment of the flight regime. All the necessary flight precautions need to be acknowledged so that the aircraft can fulfill a smooth operation. The safety pilots are trained for many hours on virtual simulations and various aircrafts. The position on the airstrip will be dependent on which direction the wind is facing. Once the plane is ready for takeoff we will have all members of the flight crew that aren t responsible for in flight objectives be advised to leave the area for safety reasons. Once the members are evacuated the communications will be tested and the engine will throttle to 100%. Once the takeoff speed is reached the aircraft will begin to lift and will climb at a constant rate. At this point the aircraft will hold the desired course in an autonomous mode but at any time throughout the flight the safety pilot may take over control of the aircraft to prevent any catastrophic crashes. Mission Arrangement: Once takeoff has successfully been achieved the flight crew personnel that is in charge of the ground station will keep a close eye on the autopilot overseeing that the speed and altitude is calibrated correctly. Now the mission will begin with the predetermined routes which will be inputted as the waypoints in Mission Planner within the command center. After the waypoints have been initiated the chosen search pattern can be performed. At 13

14 any time the mission can be modified to achieve all possible objectives. The next task is to recognize all the targets using the digital camera within the payload. Once the targets are identified they will be tracked using their position and characteristics confirmed using the live video feed. In case of a failure to the autopilot the live video feed can be used to navigate the aircraft. Landing Technique: After the mission is finished it is time to achieve a safe landing. The first step is for the landing strip to be cleared for a safety. The autonomous landing is broken down into two parts being an approach and compass lock. The approach of autonomous landing will align the airplane for the precise altitude and the compass lock will orient the plane in the correct position ensuring a safe approach to the landing strip. The final approach is directed by an established gliding rate until the aircraft touches down to perform its slow landing. Post Flight Assignments: Immediately after the plane lands the flight data is saved on the computer to interpret by the right personnel. At the end of the mission each flight crew personnel disassemble the subsystem components, place them in the correct container, and leave the flight zone. Once the data is firmly analyzed we will look for potential failures during the flight and make the appropriate changes in the software and hardware to prevent them from happening again. Mission Tasks Being Attempted: Our flight control system is responsible for navigating through the waypoints without direct manual inputs. In Addition, the flight control system is required to have a set of safety aspects, such as manual override. In order to operate the mission sufficiently the system also requires a ground station to contribute critical flight information to the Autopilot Operator, as well as enable in-flight re-tasking. The autopilot system that was chosen is the ArduPilot APM 2.5 Autopilot. It consists of the ArduPilot APM 2.5 Autopilot, a ground station, and the Mission Planner command center running on a standard laptop. The Adrupilot APM 2.5 Autopilot and ground station each connect to a GPS antenna which determines geographic position. The Ardupilot APM 2.5 Autopilot employs gyroscopes to determine the aircraft orientation and is connected to a pitot tube to calculate airspeeds. Through the Mission Planner command center, the Autopilot operator can program flight plans, adjust control parameters, and ultimately command the aircraft. The Mission Planner command center also provides the ability to perform inflight re-tasking. Furthermore, the Mission Planer command center supports a manual control mode, which is toggled by a switch on the hand held transmitter, and allows the safety pilot to take complete control of the airplane. During each of the flights, telemetry data is automatically saved and can be interpreted at any time. 14

15 2.2 Mission Tasks Being Attempted: Section Task Probability of Attempting Primary 7.1 Autonomous Flight Task 100% Primary 7.2 Search Area Task 100% Secondary 7.5 Off-Axis Target Task 80% Secondary 7.6 Emergent Target Task 80% The full assembly model of UAV is shown below: Figure 13: TITAN UAV Top Assembly. Due to complications within the manufacturing process, a full assembly of the actual air vehicle has not being met yet, but will be ready before proof of flight. Sub-assembly Pictures will be shown below in order to give the judges a visual of how the UAV is turning out after the design process: 15

16 Figure 14: Fuselage after Prepreg Carbon Fiber Layup curing Process. The Fuselage will be ready to assemble after the proper clearance holes have been made on the plane. Below is a picture of the curing process for the carbon fiber booms, boom holders, and middle fuselage-wing foam. The center part of the wing will go through a wet layup process in order to be able to reduce skin friction and assemble to the top of the fuselage. The empennage is complete and is being used as a gauge to where the boom holders will be allocated as shown below: Figure 15: Wing-Empennage Assembly and curing process of interfaces. 16

17 3. Test Evaluation Results Mission Task Performance: Due to the long manufacturing process, the TITAN UAV has not been completed yet in order to be able to be tested for all the mission task performances. Since the payload configuration is the same as last year s plane, the autonomous system has proved to work for the flight mission. Flight tests have been made with test planes in which the same payload configuration has been placed. The PixHawk flight controller has also been tested as a backup source. The interface between Mission Planner and both flight controller boards have been tested to communicate properly multiple times. 3.2 Payload Performance: The components associated with the payload have been tested for connectivity, continuity, and power required for system to be at steady state during operation. More tests will be conducted once the full assembly of the Titan UAV is completed. As far as using the payload components with test planes, it has been successful and working properly for the given operations. Also, the configuration for this year s design is the same as last year s design and has been successfully used to fly its mission. The camera has been also tested to provide video feedback using the Airmax Bullet. The interface between the Raspberri pi camera to the Raspberri pi and Airmax Bullet transmitter have been tested for communication compatibility. The Airmax Bullet transmitter and receiver have been tested within a mile range to test the signal strength. Although there were obstructions such as buildings, trees, and other objects, the signal strength was about 25 dbm for the test duration. The Airmax Bullet systems have been rated for 2 miles with a given signal strength of about 28 dbm with a ±2 dbm tolerance under no disturbances. Since the competition site will not have obstructions such as large structures or buildings, the operable distance should increase. 3.3 Guidance System Performance: The Guidance System has been tested using the flight controller and interfacing software Mission Planner using a test plane. The autonomous flight capabilities will enable for the TITAN UAV to be able to complete the primary mission requirements. More flight tests will be completed using the completed manufactured plane. Since it has worked for the test planes and the previous year s plane design, the system will be able to operate for the newly constructed plane. 3.4 Evaluation Results: Task Autonomous Flight Camera integration with Ground Control Airmax Bullet Signal Strength Motor Endurance Test Flight Stability Test Under disturbances Autonomous Recognition Test Status To be completed To be completed To be completed 17

18 4. Safety Considerations Approach 4.1 Specific safety criteria for both operation and design: Titan UAV is designed to meet the AUVSI competition regulations as well as system level requirements. The aircraft was designed to not exceed 100 KIAS (115 mph) or the 55 lb weight limit. Landing gear was designed to provide enough support for the aircraft when taxing, but specifically the impact force when landing. The landing gear also plays a huge role in ensuring that the propeller and empennage do not hit the ground when taking off or landing. Titan UAV also has a fail-safe system in case of loss of signal or in need of manual override through the flight controller board. In the manufacturing process the team operated machine tools and worked with flammable and corrosive material. While working with machine tools and corrosive materials the team wore personal protective equipment (PPE). A safe-house was built to anticipate any battery failure in order to provide adequate safety for the team member or any other personnel working in the affected perimeter. 4.2 Safety Risk and Mitigation methods: Safety was a very important factor in the design process of Titan UAV. This year s team has learned from past experiences and designed the aircraft as a pusher propeller configuration. This configuration was designed in order for the motor to be mounted in between the boom holders so that if a member or anyone is near the aircraft they will not be in immediate danger from the propeller. The system is composed of switches so that certain components of the aircraft will turn on when specified as shown below. This will prevent the motor from turning on unexpectedly. The APM system has a built in fail-safe mode in which it will activate if there is loss of signal for more than 30 seconds in which it will return to base. If there is signal loss for more than 3 minutes, the APM will initiate death spiral to ground. Figure 16: Kill-switch orientation for Battery configuration. 18

19 5. Conclusion The purpose of this journal paper was to show the rigorous design process and the engineering approach to complete the autopilot, airframe, and payload systems. The improvements made from last year s team were the structural integrity from the carbon fiber and honeycomb Nomex composition for the fuselage. This also gave a higher strength to weight ratio for the air vehicle in order to maximize flight stability from bending moments created from the tail and motor loads. The structural design improved from last year s design, but increased the amount of time devoted to the production of the vehicle and decreasing the testing time. This in turn caused some delay within testing the autonomous system on the complete system. The autonomous system and payload was tested on test planes and had more hours used from last year s team. Using the same flight controller board that successfully worked for the previous year made it simpler to focus the design on the structural aspect of the UAV. The goal of this year s team is to solidify a structurally sound UAV structure that can be used continuously in order for next year s team to focus on the autonomous target recognition and secondary tasks. This competition has given us the most learning experience and practice of our engineering skills. Therefore, we are very thankful that AUVSI has provided us with this opportunity. 19