Design of a Miniature Aircraft Deployment System

Transcription

1 Project Customer Prof. Eric Frew Project Advisors Prof. Bill Emery Prof. Kurt Maute Design of a Miniature Aircraft Deployment System Leah Crumbaker Jason Farmer Michael Gordon Matt Lenda Jeffrey Mullen Scott Tatum Travis Schafhausen Kristina Wang

2 Briefing Overview & Content Purpose To present the Design of a Miniature Aircraft Deployment System AIAA Paper. System Architecture Subsystem Design Deployment Electronics & Software Analysis System Development 2

3 System Architecture Deployment Subsystem Design Deployment Electronics and Software Deployment Analysis System Development

4 Project Objectives and Purpose System Architecture Objective Integrate a mechanism into an existing UAS that can carry and deploy four small subvehicles on demand during flight Purpose Provide a test platform for cooperative control protocols Provide a proof of concept that in-flight deployment of air vehicles and the overall dynamics associated with such an action is predictable and reliable 4

5 Concept of Operations System Architecture 5

6 System Design System Architecture 1. Primary Vehicle (PV) On-board electronics controls deployment through wireless commands from a ground station Flown by an RC Pilot 2. Deployment Mechanism (DM) Consists of mounting point for the SV and linear actuator for pin-movement release Attached to the PV with bracketing system 3. Sub-Vehicle (SV) CUPIC Autopilot commands the control surfaces and motor settings autonomously from mission commands given from the ground station 6

7 System Architecture Deployment Subsystem Design Deployment Electronics and Software Deployment Analysis System Development 7

8 Deployment Subsystem Design Deployment Subsystem Design Attachment of Sub-Vehicles to Primary Vehicles Sub-Vehicle selection Attachment configuration and reinforcements Finite element analysis models Verification of models with structural testing Deployment Mechanism Design Robustness Functionality Test functionality in a vibration environment 8

9 Sub-Vehicle Configuration Deployment Subsystem Design Configuration Vehicle is mounted in flight ready configuration Reasons Removes complexity of designing foldable SV Minimizes possible failure points COTS options available Deployment dynamics easily predicted Chosen Vehicle Super Fly 9

10 Attachment Configuration Deployment Subsystem Design Cantilever Beam Mounted Psuedo-Stacked Configuration Reason Has largest available control margins in worst case load No wing modifications Fuselage easy to reinforce Reduces structural complexity Attachment Design Choice Cantilever Beam Location Selection Numerical drop model CFD aircraft stability model 10

11 Attachment Reinforcement Model Deployment Subsystem Design FEA Model Estimated worst-case aerodynamic forces and moments applied from each Sub-Vehicle Near-impossible flight condition with a 1.5 safety factor Used ANSYS to design and determine estimated stresses in body wall Iterative design process to maximize reinforcement strength and minimize mass Boundary Conditions Fixed beam ends Loads applied to body Results External aluminum plate Internal ½ thick aluminum block Doubles as retention device with use of set screws Withstands estimated worst-case loads 11

12 Deflection [mm] Non-Destructive Structural Testing Purpose: Verify FEA model Matched boundary conditions in FEA model Applied load (to the fuselage of the aircraft) Displacement from the bottom of the fuselage measured Comparison to FEA model was used to verify the results Deployment Subsystem Design Deflection of Plane Fuselage for Varying Applied Loads* Experimental Data Theoretical (Immovable BCs) Applied Load [kg] FEA model is validated. The structure will support all 4 Sub-Vehicles in flight with estimated in-flight maximum loads (with no gusts) * Measured with LVDT on a 12-bit ADC. Accurate to within 1.831x10-6 mm 12

13 Deployment Mechanism Method Pin-sleeve mechanism Manufactured out of aluminum Stainless steel pin for strength and decreased friction Action Linear actuator Benefits Semi-circle surface only allows pitching motion. Flat sides inhibit roll, yaw, and lateral motion Actuator Mounting Bracket Deployment Subsystem Design Assembled Deployment Mechanism Sub-Vehicle Mounting Point 13

14 Deployment Mechanism Deployment Subsystem Design Linear Actuator Mount Linear Actuator Deployment Mechanism 14

15 Force [N] Vibration Test Testing Results Vibrations from 0Hz to 150Hz (max expected forced frequency from engine running at 9000rpm) Deployed 121/124 times. 95% confidence for non-aerodynamic loads in a vibration environment Failed at 5Hz Vibration mode of the beam predicted by ANSYS modal analysis Deployment Subsystem Design 15 Maximum Force to Pull the Pin Experimental Data Best Fit Line Frequency [rad/s] 15

16 System Architecture Deployment Subsystem Design Deployment Electronics and Software Deployment Analysis System Development 16

17 Deployment Electronics & Software Deployment Electronics & Software Wireless Communication Communication protocol is IEEE (ZigBee) utilized due to pre-existing technology on the autopilot Primary vehicle Sub-vehicle Ground Station Deployment Process Ground station sends deployment command to the primary vehicle Primary vehicle initiates sub-vehicle deployment sequence once all failsafe conditions have been met 17

18 Deployment Electronics Deployment Electronics & Software Linear Actuator Operates at +/- 5V to extend and retract (unlike PWM controlled servos) Internal potentiometer relays pin position data Linear Actuator Control Board Designed custom control boards 5-MOSFET for current control to retract/extend actuator Externally-mounted Manual Override System on the primary vehicle for ease of mounting Sub-Vehicles Image From: Sub-Vehicle Outfitted with in-house CUPIC autopilot Primary Vehicle Outfitted with CUPIC hardware with the software stripped 18

19 Electrical/Software Flow Deployment Electronics & Software Primary Vehicle XBee Radio Ground Station Ground Station GUI (MATLAB) Sub-Vehicle XBee Radio CUPIC Actuators Potentiometers Realterm& ActiveX XBee Radio CUPIC Servos/Throttle GPS GPS Flow Meter Legend Wireless Communication Wired Communication Roll Gyro Altimeter 19

20 Primary Vehicle Design Deployment Electronics & Software The Primary vehicle utilizes an 8-bit PIC microcontroller Equipped with ZigBee communication, GPS receiver, and pressure transducer Controls the deployment of the Sub-Vehicles through the linear actuator control boards Retracts the pin and monitors pin position data Able to command the Sub-Vehicle to begin its deployment sequence Telemetry is streamed to the ground station 20

21 Sub-Vehicle Design Deployment Electronics & Software Sub-Vehicle is outfitted with the CUPIC Autopilot, developed by CU PhD Student Bill Pisano Develop autopilot with minimal sensors PIC18F Microcontroller integrates GPS, roll gyro, pressure sensor, and XBee radio Interrupt-based controller: 100Hz Low Priority, Asynchronous High Priority Heading rate controlled by roll angle Altitude controlled by throttle Utilizes Lyapunov vector field to track predetermined loiter circle Includes autonomous landing and deployment sequences Deployment sequence combines roll and pitch stabilization of the autopilot to prevent collision with the Primary Vehicle Images Courtesy: Bill Pisano 21

22 Error [m] Latitude [deg] Sub-Vehicle Autopilot Flight Deployment Electronics & Software Autopilot loiter circle tracking shows convergence from deviations Wind causes repeated tracking error SV CUPIC GPS Position for 50m Loiter Track CLC, = +/- 1.96m WP Loiter Longitude [deg] 20 SV CUPIC Tracking Error in 50m Loiter Track The minimal autopilot can control the Sub- Vehicle Time [s] 22

23 Ground Station Deployment Electronics & Software Development Methods Pre-existing in-house ground station software heavily modified Ground station written in MATLAB interfacing to ZigBee with RealTerm Telemetry GPS positions of up to 5 vehicles Sub-vehicle data Autopilot status Primary vehicle data Deployment mechanism status 23

24 System Architecture Deployment Subsystem Design Deployment Electronics and Software Deployment Analysis System Development 24

25 Deployment Analysis Deployment Analysis Numerical Deployment Model Deployment dynamics need to be modeled Avoid primary vehicle strikes Full System Stability Analysis Analyze unusual geometry of system Predict aerodynamic capability Ensure stability such that it can be flown by an RC pilot 25

26 Vertical Motion (m) Deployment Model Deployment Analysis Ensure no collisions with the primary vehicle Modeled non-linear longitudinal dynamics using Newton s and Euler s equations of motion Tail Relative Motion of SV to SIG Wing Upper SV Wing Strut ma ma I yy x y q F F x y M q q y,,, T,,,, T, q, q,, e e e Path Start End Restricted Zone Wheel Strut Wheel Casing Horizontal Motion (m) Equations integrated with explicit Euler method Each time step, all aerodynamic variables updated Numerous cases run with varying elevon deflection and initial conditions PV strike is not an issue 26

27 Distance [m] Deployment Test Model was compared with in-flight deployment data Aircraft would drop for 2 seconds to ensure separation from the PV Switch to autopilot control Deployment Analysis SV Distance from SIG during Autonomous Deployment * Measured Data Predicted Data The drop model accurately predicts deployment dynamics Time [sec] * Error bars from ±1σ statistical distribution from static altitude data 27

28 System Stability Model Deployment Analysis Conventional stability analysis tools utilize vortex lattice solvers Solvers fail to converge for complex and unusual geometries Current stability modeling requires wind tunnel with forced oscillation balance Can Computational Fluid Dynamics be used to estimate stability derivatives? Navier Stokes CFD popular in aerodynamic design and research New focus on Lattice Boltzmann Method (LBM)* Due to nature of equations, good for complicated geometries LBM is more numerically stable than Navier Stokes CFD Fast meshing/problem set-up for complicated geometry due to Immersed Boundary Techniques * Master s Thesis: Herrmann, Dominik. Study of the Suitability of PowerFLOW as an Educational Engineering Design Tool for Undergraduate Students. 28

29 Coefficient System Stability Model Deployment Analysis Solution Create Custom Stability Tool Using LBM Utilized the off-the-shelf LBM solver, PowerFLOW, to analyze stability Forced oscillation simulations created in PowerFLOW Cross-coupled and linear stability derivatives estimated (N Θ, X r, etc.) at various airspeeds Each flight case takes ~2 weeks to run Due to time constraints, only the full-system stability could be simulated Each simulation provides forces and moments in each of the three body axes Complete drag polar can be estimated System Aerodynamic Data at Re = Lift Drag Moment Stall Angle of Attack [deg] System Drag Polar at Re = C L C D 29

30 Imaginary Axis Imaginary Axis System Stability Model Deployment Analysis All the aircraft parameters are known (span, mass, moments of inertia) and the flow conditions are known (velocity, pressure, density) Linear stability derivatives can be calculated Poles of Longitudinal Modes Short-Period Phugoid Roll Subsidence Real Axis Poles of Lateral Modes Spiral Divergence Dutch Roll Results show the natural modes of a traditional aircraft Slightly unstable in the longitudinal phugoid mode No concerns because the primary vehicle is controlled by an RC pilot Real Axis 30

31 System Architecture Deployment Subsystem Design Deployment Electronics and Software Deployment Analysis System Development 31

32 Accomplishments System Development System Analysis and ground-testing was performed on nearly every aspect of the project before flight test Several successful flight tests and deployments 14/18 successful deployments 4 failed deployments Due to damaged linear actuators 29-sensor, data collection package is flown at every flight Verifies theoretical models and allows for proper analysis if unexpected events occur Sub-Vehicle Autonomous flight has been demonstrated from 3 sensors: GPS, roll rate, and altitude 32

33 Future Applications System Development Design SVs with foldable wings and a more advanced and robust PV Longer data collection times Larger areas to cover and characterize Dangerous/difficult situations that require data-collection Severe weather Wildfire observation Military applications 33

34 Conclusion System Development The design of a miniature aircraft deployment system and the overall dynamics associated with such an action can be predictable and reliable! 34

35 Acknowledgements Prof. Bill Emery and Prof. Kurt Maute Our faculty advisors have given useful feedback throughout the design and testing phases. Prof. Eric Frew Our customer has offered much support and guidance in helping us to define and achieve our project goal and requirements. Arvada Associated Modelers The whole crew at AAM has provided an abundance of useful information regarding RC planes, and has greatly helped in the testing of the system by offering their field, assistance, and piloting skills. Trudy Schwartz Trudy has been a great help in all of our electronic endeavors. Matt Rhode Matt has been essential in manufacturing all mechanical components of the system. Eric Dickey Has been a great help with the CU DataLoggers and many other electronics issues. Bill Pisano Bill has helped a great deal in aiding our understanding of the CUPIC. PAB The members of the PAB have provided constructive criticism, feedback, and guidance throughout project definition, development, and design. 35

36 Thank You Questions?