UAV Plane Project

i
University Of Adelaide
School of Mechanical Engineering
2008

Honours Project 637:
Design and Build of a Pulsejet UAV

Ryan Anderson 1132309
Nicholas Lukacs 1133184
Mitchell OCallaghan 1131620
Karn Schumacher 1133398
Michael Sipols 1133364
Terry Walladge 1133113

ii
iii
Sir Ross and Sir Keith Smith Fund Acknowledgement and
Disclaimer
Research undertaken for this report has been assisted with a grant from the Sir Ross
and Sir Keith Smith Fund (Smith Fund) (www.smithfund.org.au). The support is
acknowledged and greatly appreciated.
The Smith Fund by providing funding for this project does not verify the accuracy of
any findings or any representations contained in it. Any reliance on the findings in any
written report or information provided to you should be based solely on your own
assessment and conclusions.
The Smith fund does not accept any responsibility or liability from any person,
company or entity that may have relied on any written report or representations
contained in this report if that person, company or entity suffers any loss (financial or
otherwise) as a result.

iv
v
Executive Summary
The pulsejet powered Unmanned Aerial Vehicle (UAV) was designed and
manufactured by a group of six undergraduate engineering students from the School
of Mechanical Engineering at the University of Adelaide. The students, studying a mix
of mechanical and aerospace engineering, aimed to design and build a UAV powered
by a valveless pulsejet engine, which was also developed throughout the year. The use
of pulsejets in aviation history has been almost non-existent since the end of World
War II. However, interest in pulsejet technology has increased in recent years, as they
offer a cheap and viable alternative from turbojet and ducted fan engines. The design
of the aircraft was based around the pulsejet engine and is ultimately intended for use
as a high speed target drone or decoy aircraft.

The development of the valveless pulsejet engines followed of from work completed
by Coombes et al in 2007, with the aim to produce an engine and fuel system capable
for use in flight. A wide range of development was undertaken on three different
engines throughout the year, with over 100 static tests performed by the students.
Significant improvements were achieved in the areas of engine thrust, thrust specific
fuel consumption, engine weight and engine fuelling; most notably achieving
successful operation using liquid fuels.

The allowance for pulsejet engine installation meant that a conventional airframe
design was not suitable. A classical approach was taken to determine the performance
and stability of the airframe. This design incorporated low swept wings, dual vertical
stabilizers and an elevated swept tail, to produce an airframe that is capable of
pulsejet powered flight. The airframe was manufactured by the students under the
supervision and assistance of the Mechanical Engineering Workshop staff, and was
constructed primarily from composite materials.

Successful flight of the aircraft was achieved on a ducted fan as it was seen as a more
conventional power source, which has similar operational characteristics to the

vi
pulsejet engines. The flight tests showed that the airframe was stable, controllable and
maneuverable. A cruise speed of 150km/hr was achieved during a four minute flight.
The aircraft performed all handling requirements during the test flight.

The project goals set by the students at the beginning of the project reflected the
ambitious nature of the project. The extension goals were particularly ambitious and
related primarily towards the performance of the aircraft and engine. While some
goals were not completely achieved, most were well within the performance
capabilities of the aircraft.
vii
Acknowledgements
The group would like to acknowledge the many people who have helped make this
project possible. We would especially like to thank our supervisor Dr. Maziar
Arjomandi, whose guidance, support and technical knowledge has been invaluable in
ensuring the projects success.

The group would like to acknowledge the Sir Ross and Keith Smith Fund, whose
generous contribution was vital for the success of the project. Without the funds
passion for the development of Aerospace design and technology in South Australia,
the project would not be possible.

The group would also like to thank the School Of Mechanical Engineering, ASC and
Australian Aerospace for their generous contributions to the project.

The authors would also like to thank and acknowledge all of the individuals who have
spent countless hours with the group throughout the year. In particular, a special
thanks to Bill Finch, from the Mechanical Engineering Workshop, whose technical
knowledge and dedication were invaluable. The personal contribution of James Irvine,
from Irvine Aeropulse, for his in-kind sponsorship, guidance and assistance in the
development and operation of pulsejet engines was greatly appreciated. Finally, we
would like to thank John Modistach, for both his time and effort spent assisting us with
aircraft manufacture, as well as for passing on a wealth of knowledge, which assisted
us in the manufacture of the airframe.

viii
ix
Disclaimer
This statement confirms that the work presented in entirely our own, unless
identified otherwise. The work presented was completed as part of the
requirements for the Degree of Bachelor of Engineering (Aerospace and
Mechanical respectively) at the University of Adelaide. This document describes
the work carried out by the students, as recorded in individual project
workbooks throughout 2008. The students acknowledge the penalties for
plagiarism, fabrication and unacknowledged syndication and declare that the
work presented is free of these.

Ryan Anderson Nicholas Lukacs

------------------------------------------- ----------------------------------------
Date: Date:

Mitchell OCallaghan Karn Schumacher

------------------------------------------- ----------------------------------------
Date: Date:

Michael Sipols Terry Walladge

------------------------------------------- ----------------------------------------
Date: Date:

x
xi
Contents
Sir Ross and Sir Keith Smith Fund Acknowledgement and Disclaimer ............................ iii
Executive Summary........................................................................................................... v
Acknowledgements .........................................................................................................vii
Disclaimer .........................................................................................................................ix
Contents............................................................................................................................xi
List of Figures .................................................................................................................xvii
List of Tables ................................................................................................................. xxiii
1 Introduction.............................................................................................................. 1
1.1 Project definition................................................................................................ 2
1.2 Project Aims ....................................................................................................... 2
1.2.1 Pulsejet Development................................................................................. 2
1.2.2 Airframe Development ............................................................................... 3
1.3 Project Goals ...................................................................................................... 3
1.4 Extension Goals.................................................................................................. 3
1.5 Scope.................................................................................................................. 4
2 Feasibility Study ........................................................................................................ 5
2.1 What is a Pulsejet............................................................................................... 5
2.2 Advantages + Disadvantages.............................................................................. 6
2.3 Pulsejet Engines in Aviation History................................................................... 7
2.4 Market Research and Benchmarking................................................................. 7
2.4.1 V-1............................................................................................................... 8
2.4.2 ENICS Drones .............................................................................................. 9
2.4.3 AMT Pulsejet Hobby Aircraft .................................................................... 10
2.4.4 Comparison to turbine engine UAVs or Target Drones............................ 10
2.5 Mission Profile Specifications .......................................................................... 11
2.5.1 Mission Profile .......................................................................................... 11
2.5.2 System Requirements............................................................................... 11
2.5.3 Takeoff methods....................................................................................... 14
2.5.4 Landing Options ........................................................................................ 15

xii
2.6 Power plant Design...........................................................................................15
2.6.1 Valveless Pulsejet - Thermodynamic Cycle ...............................................15
2.6.2 Review of Previous Work ..........................................................................17
2.6.3 Alternative Engine Designs........................................................................21
2.6.4 Exhaust Pipe Development .......................................................................24
2.6.5 Liquid Fuelling ...........................................................................................27
2.6.6 Thrust Augmentation ................................................................................29
2.7 Feasibility Study Summary................................................................................30
3 Conceptual Design...................................................................................................31
3.1 Aircraft Conceptual Design Introduction..........................................................31
3.2 Selecting Preliminary Aircraft Concept.............................................................31
3.2.1 General Configuration...............................................................................31
3.2.2 Fuselage Configuration..............................................................................32
3.2.3 Engine Configuration.................................................................................32
3.2.4 Wing Configuration ...................................................................................32
3.2.5 Empennage Configuration ........................................................................32
3.2.6 Landing Gear Configuration ......................................................................33
3.2.7 Basic Wing Parameters..............................................................................33
3.3 Developing concept for selected configuration...............................................34
3.3.1 Concept Sketches ......................................................................................35
3.3.2 Statistical Calculations...............................................................................35
3.4 Designing technical parameters for concept....................................................38
3.4.1 Weight Estimation.....................................................................................38
3.4.2 Matching Diagram.....................................................................................39
3.4.3 Aerofoil Selection......................................................................................42
3.5 Finalisation of Preliminary Aircraft Concept ....................................................45
3.5.1 Variation of Pulsejet Position in Concept Development...........................45
3.5.2 Empennage Design....................................................................................47
3.6 Finalization of Preliminary Aircraft Concept ....................................................52
3.6.1 Preliminary Conceptual Fuselage Design..................................................52
3.7 Practical Modifications to Final Concept..........................................................53
xiii
3.8 Engine Design................................................................................................... 54
3.8.1 Exhaust Design Two Stroke Exhaust Similarities ................................... 54
3.8.2 Steady State Diffuser Design..................................................................... 57
3.9 FWE Bellmouth Development.......................................................................... 63
3.9.1 Starting Vortices........................................................................................ 63
3.9.2 Bellmouth Design...................................................................................... 64
3.9.3 Final Design............................................................................................... 67
3.9.4 Flight considerations................................................................................. 68
3.10 Flight Engine Development .......................................................................... 71
3.11 Liquid Fuel System Design............................................................................ 76
3.11.1 Fuel Choice............................................................................................ 77
3.12 Fuel Injector Design...................................................................................... 78
3.13 Conceptual Design Summary........................................................................ 83
4 Detailed Design....................................................................................................... 85
4.1 Fuselage Structure Design................................................................................ 85
4.1.1 Fuselage Structural Layout ....................................................................... 85
4.1.2 Fuselage Structure Selection .................................................................... 88
4.2 Wing Design ..................................................................................................... 89
4.3 Wing Structural Design..................................................................................... 89
4.3.1 Lifting force profile.................................................................................... 89
4.3.2 Spar Design ............................................................................................... 95
4.3.3 Torsion ...................................................................................................... 97
4.4 Wing Connection............................................................................................ 101
4.5 Control Surface Sizing .................................................................................... 103
4.5.1 Aileron Sizing........................................................................................... 103
4.5.2 Elevator Sizing......................................................................................... 104
4.5.3 Servo Motor Sizing.................................................................................. 105
4.6 Pulsejet Engine Mount ................................................................................... 106
4.6.1 Mounting Locations ................................................................................ 106
4.6.2 Thermal Isolation.................................................................................... 107
4.6.3 Vibration Isolation .................................................................................. 108
4.6.4 Vibration Isolation Method: ................................................................... 110

xiv
4.6.5 Engine Mount Materials..........................................................................112
4.6.6 Final Design .............................................................................................113
4.6.7 Engine Modal Analysis.............................................................................117
4.7 Pulsejet Launch System..................................................................................118
4.7.1 Launch process........................................................................................118
4.7.2 Launch Stand Components .....................................................................119
4.8 Electrical and Electronic Components............................................................121
4.8.1 Pump and related components...............................................................121
4.8.2 Radio Controller ......................................................................................122
4.9 Ducted Fan......................................................................................................123
4.9.1 Purpose of fan.........................................................................................123
4.9.2 Selection of fan system...........................................................................123
4.9.3 Modifications to the airframe for Ducted Fan Testing ...........................127
4.10 Final Stability Analysis.................................................................................129
4.10.1 Longitudinal Moment Analysis ............................................................130
4.10.2 Roll Stability Analysis ...........................................................................134
4.10.3 Ground Performance...........................................................................135
5 Airframe Manufacture ..........................................................................................137
5.1 Available Manufacturing Methods.................................................................137
5.2 Wing Construction..........................................................................................138
5.3 Empennage Construction...............................................................................140
5.4 Fuselage Construction ....................................................................................141
5.5 Internal Fuselage Construction ......................................................................144
5.6 Internal Access................................................................................................145
5.7 Propulsion System..........................................................................................146
5.7.1 Ducted fan...............................................................................................146
5.7.2 Pulsejet ....................................................................................................147
5.8 Landing Gears and Wheels .............................................................................148
5.9 Control System Installation ............................................................................149
6 Testing...................................................................................................................151
6.1 Engine Testing.................................................................................................151
xv
6.1.1 Phase One Testing................................................................................... 152
6.1.2 Phase Two Testing .................................................................................. 155
6.1.3 Phase Three Testing................................................................................ 160
6.2 Aircraft Testing............................................................................................... 168
6.2.1 Wing Structural Testing .......................................................................... 168
6.2.2 Electrical Component Testing................................................................. 169
6.3 Aircraft Pre-flight Tests .................................................................................. 170
6.3.1 C.G. Test .................................................................................................. 170
6.3.2 Other pre-flight checks ........................................................................... 171
6.3.3 Location for flying ................................................................................... 172
6.3.4 Pilot ......................................................................................................... 173
6.3.5 Engine and Flight Tests ........................................................................... 173
6.4 Pulsejet Flight Test ......................................................................................... 179
6.5 Discussion of experimental results ................................................................ 180
7 Management......................................................................................................... 181
7.1 Time Management ......................................................................................... 182
7.2 Financial Management................................................................................... 184
7.3 Risk Management........................................................................................... 185
8 Conclusion and Future Work ................................................................................ 187
8.1 Review of project goals.................................................................................. 187
8.1.1 Extension Goals....................................................................................... 189
8.2 Project Concerns ............................................................................................ 190
8.3 Future Developments and Recommendations .............................................. 191
References .................................................................................................................... 195
Appendix A - Configuration Selection........................................................................... 199
Appendix B- Weight Calculation Method..................................................................... 211
Appendix C Matching Diagram.................................................................................. 221
Drag polar estimation........................................................................................... 221
Initial estimate of drag polar ................................................................................ 222
Climb requirements .............................................................................................. 223
Stall Requirement ................................................................................................. 223
Takeoff Field Length Requirement ....................................................................... 224

xvi
Cruise Requirement...............................................................................................224
Adjusting to take-off values ..................................................................................225
Appendix D Sensitivity Analysis..................................................................................227
Appendix E Engine Mounting Calculations.................................................................233
Appendix F Liquid Fuels..............................................................................................235
Appendix G Component Weight Breakdown.............................................................239
Appendix H Test Log Books ........................................................................................241
Appendix I Fuselage Stress Analysis...........................................................................279
Appendix J- Gantt Charts...............................................................................................281
Appendix K- Risk Register..............................................................................................285
Appendix L- Meeting Minutes.......................................................................................287
Appendix M- Drawings..................................................................................................367

xvii
List of Figures
Figure 1 - Valved and Valveless Pulsejet Designs ............................................................. 5
Figure 2 - Comparison of Engine Costs ............................................................................. 6
Figure 3-View of the V-1................................................................................................... 8
Figure 4-E-95 Ramp Launch.............................................................................................. 9
Figure 5- Flight Profile..................................................................................................... 11
Figure 6 - Statistical Trends of Target Drone UAVs ........................................................ 13
Figure 7 - Ideal Lenoir Cycle............................................................................................ 16
Figure 8-Pobezhimov modified Lenoir cycle................................................................... 17
Figure 9 - Valveless Pulsejet Engine (Carolina State University) .................................... 18
Figure 10 - Focus Wave Energy (FWE) Pulsejet Engine .................................................. 20
Figure 11-Chinese Valveless Pulsejet Engine.................................................................. 21
Figure 12-Lockwood Prototype ...................................................................................... 22
Figure 13-Escopette Valveless Engine ............................................................................ 23
Figure 14 - Interaction of Escopette Pressure Waves .................................................... 24
Figure 15 A Focused Wave (FWE) Pulsejet engine ...................................................... 25
Figure 16 A Lockwood-Hiller style Pulsejet engine,..................................................... 26
Figure 17 A Focused Wave engine variation, the FWE VIII - Lady Anne Boleyn. ..... 27
Figure 18- optimised thrust augmenter as used on a valved pulsejet ........................... 29
Figure 19-Configuration Concept Sketches .................................................................... 35
Figure 20: Statistical Thrust Loadings of Jet UAVs.......................................................... 36
Figure 21: Wing Loading Versus Weight of Jet UAVs ..................................................... 37
Figure 22: Statistical Concept ......................................................................................... 38
Figure 23: Matching Diagram......................................................................................... 40
Figure 24: Example of Early Design................................................................................. 46
Figure 25: Second Phase Design Example ...................................................................... 46
Figure 26: Final Engine Position...................................................................................... 47
Figure 27: Centre of Gravity Excursion Diagram............................................................. 49
Figure 28: Longitudinal X-Plot......................................................................................... 50
Figure 29: Lateral Stability X-Plot.................................................................................... 52
Figure 30 -Conceptual Fuselage Design.......................................................................... 53

xviii
Figure 31: Modifications to final aircraft concept...........................................................53
Figure 32- Advancements in Two Stroke Exhaust Design...............................................55
Figure 33 The effect of expansion angle on wave behaviour. .....................................56
Figure 34: Loss Coefficient for a Conical Diffuser ...........................................................58
Figure 35 The UFLOW1D model used to investigate expansion angles.......................59
Figure 36 Combustion chamber pressure extremes for different expansion angles. .60
Figure 37 - Statistical data showing exhaust expansion angles from similar engine
designs.............................................................................................................................61
Figure 38 - statistical data showing a trend between combustion chamber diameter
and expansion diameter..................................................................................................61
Figure 39 - The final expansion design............................................................................63
Figure 40 - PIV images of vortex interaction...................................................................64
Figure 41 - Bellmouth designs considered (Blair, Cahoon 2006) ....................................65
Figure 42 - Performance of bellmouth designs...............................................................65
Figure 43 - The data obtained in 2007 using UFLOW1D (blue) and textbook
recommendations (red) ..................................................................................................66
Figure 44 - The adjustable bellmouth design..................................................................68
Figure 45- Three intake geometries................................................................................69
Figure 46- Domain Layout ...............................................................................................69
Figure 47- Effect of intake geometry on mass flow rate.................................................70
Figure 48 -Static Pressure Contours of Aerodynamic Flare at 80m/s.............................70
Figure 49-Static Pressure Contours on Standard Flare at 80m/s....................................71
Figure 50- Statistical trend of Chinese and FWE engines ...............................................72
Figure 51-Variation of material properties of 310 stainless steel with temperature.....73
Figure 52-Operating pressure of the Escopette pulsejet................................................74
Figure 53- full engine mesh.............................................................................................75
Figure 54-pressure loading input for flexible dynamic solver.........................................75
Figure 55-Stress Results on Combustion Chamber End Cap...........................................76
Figure 56- Final Results of the Axi Symmetric Model .....................................................76
Figure 57- 12 hole swirl injector......................................................................................79
Figure 58- 6 hole opposed spray injector........................................................................79
xix
Figure 59-BETE PJ Cone Spray Injector ........................................................................... 80
Figure 60-5mm stainless steel injectors ......................................................................... 81
Figure 61: Conceptual Design Three View...................................................................... 83
Figure 62: Overview of Fuselage Structural Layout....................................................... 85
Figure 63: Fuselage Internal Reinforcing Structure........................................................ 86
Figure 64 - Schrenk's Approximation.............................................................................. 90
Figure 65 -Lifting Force Distribution............................................................................... 91
Figure 66- wing shear distribution.................................................................................. 92
Figure 67 - wing bending force distribution ................................................................... 92
Figure 68 - Corrected Cl Distribution.............................................................................. 93
Figure 69 - Lift Distribution at 88m/s.............................................................................. 94
Figure 70 - Maximum spar thickness from root to tip of the wing ................................ 96
Figure 71 - Position of Centre of Pressure with AOA...................................................... 98
Figure 72 - Wing Torque at Takeoff (70km/hr)............................................................... 99
Figure 73 - Wing Torque at Climb Speed (150km/hr)..................................................... 99
Figure 74 - Wing Torque at Cruise Speed (300km/hr).................................................. 100
Figure 75- Wing Connection System............................................................................. 101
Figure 77 - The mounting extension on the front of the Chinese engine .................... 107
Figure 78 - Force transmissibility as a function of frequency ratio and damping ratio109
Figure 79 - Yield stress relative to room temperature as a function of temperature for
301,302,304,321,347 annealed stainless steels ........................................................... 113
Figure 80 - The final engine mount design. The modification made to the front of the
engine is shown in green. ............................................................................................. 114
Figure 81 - Thermal analysis results of the engine mount. .......................................... 114
Figure 82 - Stress distribution within the initial design under an 80N load................. 115
Figure 83 - Stress distribution within the design under dynamic loading of 40N +- 20N
...................................................................................................................................... 116
Figure 84 - 208Hz vibration mode of the engine, mounted at ends ............................ 117
Figure 85 - Release tab attached to the intake of the engine, and release tab on the
launch stand.................................................................................................................. 119
Figure 86 - Launch stand for pulsejet flight .................................................................. 120
Figure 87: Flight Works Fuel Pump.............................................................................. 121

xx
Figure 88 - Schubeler Ducted Fan .................................................................................123
Figure 89 - Lehner electric motors ................................................................................125
Figure 90 - ZIPPY-R battery pack ...................................................................................126
Figure 91 - Ducted fan mounted in the airframe..........................................................127
Figure 92 - Ducted Fan Mounting Tabs.........................................................................128
Figure 93 - The cover design for the ducted fan...........................................................129
Figure 94 - The cover installed on the plane.................................................................129
Figure 95: Centre of Gravity and Aerodynamic Centre Excursion Diagram..................131
Figure 96:Cm-Cl Graph (Power On)...............................................................................132
Figure 97: Cm-Cl Graph (Power On)..............................................................................132
Figure 98: Cm-Cl Graph (Pulsejet) .................................................................................133
Figure 99: Roll Stability Contributions...........................................................................135
Figure 100 - Rib Installation in Wings............................................................................139
Figure 101 - Wing structure schematic .........................................................................140
Figure 102 - Servo Installation.......................................................................................140
Figure 103 - a) Horizontal tail joined as a single piece, b) horizontatal tail after glassing,
c) installation of vertical tail onto fuselage...................................................................141
Figure 104 - Fuselage plug.............................................................................................142
Figure 105 - Gel coat being applied to plugs in preparation for creating the moulds..143
Figure 106 - Fuselage.....................................................................................................144
Figure 107 - Location of bulkheads (blue) and longerons (red)....................................145
Figure 108 - The aircraft showing the both access panels a) removed and b) attached
.......................................................................................................................................146
Figure 109 Schubeler ducted fan (Schubeler Jets, 2008) ...........................................147
Figure 110 - Front pulsejet engine mount ....................................................................148
Figure 111 - Front landing gear steering system...........................................................149
Figure 112 - Test System Layout ...................................................................................151
Figure 113 Reducing thrust during extended operation............................................153
Figure 114 Effect of fuel injection position on engine performance.........................153
Figure 115 Effect of exhaust and intake length on engine performance ..................154
xxi
Figure 116 - The adjustable FWE engine with expanding tail section and 100mm
extension....................................................................................................................... 156
Figure 117 - Affect of injector position on engine thrust ............................................. 157
Figure 118 - Thrust Results ........................................................................................... 158
Figure 119 - Visible damage to ceramic coating........................................................... 159
Figure 120 - liquid fuel injectors placed mid way along the intake tube ..................... 161
Figure 121 - Opposed injector configuration................................................................ 163
Figure 122 - Performance of the Chinese engine with different injector placements. 163
Figure 123 - engine performance on various fuels....................................................... 164
Figure 124 - Performance of the Chinese engine for various lengths.......................... 166
Figure 125 - Aircraft testing flow chart......................................................................... 168
Figure 126 Load zones for wing structural testing .................................................... 168
Figure 127 - Experimental Wing Deflection.................................................................. 169
Figure 128 - C.G. Test Setup.......................................................................................... 170
Figure 129 - C.G. Test Photo ......................................................................................... 171
Figure 130 - Ground Roll Test at Gawler Airfield.......................................................... 174
Figure 131 - Plotted flight path from GPS logger.......................................................... 176
Figure 132-Compact Gantt Chart.................................................................................. 183
Figure 133-Cost Breakdown.......................................................................................... 185
Figure 134: Mock graphic of selected configuration................................................... 209
Figure 135- Graph of WE/WO Vs WO........................................................................... 212
Figure 136- Graph of WE/WO Vs WO........................................................................... 217
Figure 137- Graph of WE/WO Vs WO for Consistent Data........................................... 219
Figure 138: First Estimate of Drag Polar ...................................................................... 222
Figure 139: Sensitivity to fuel consumption ................................................................ 227
Figure 140: Sensitivity to Engine Weight ..................................................................... 228
Figure 141: Sensitivity of Cruise Speed to W/S ............................................................ 228
Figure 142: Sensitivity of Takeoff Distance to W/S ...................................................... 229
Figure 143: Sensitivity of Climb Rate to W/S................................................................ 229
Figure 144 : Sensitivity of Stall Speed to W/S............................................................... 230
Figure 145: Sensitivity of Cruise Speed to T/W............................................................ 230
Figure 146 : Sensitivity of Takeoff Distance to T/W..................................................... 231

xxii
Figure 147 : Sensitivity of Climb Rate to T/W...............................................................231
Figure 148- Engine during Test......................................................................................245
Figure 149- Thrust Vs Time for Test 6 ...........................................................................247
Figure 153- Thrust Vs Time for Test 10 .........................................................................250
Figure 157- Thrust Vs Time with injector 32mm from intake mouth...........................259
Figure 158- Time Vs Thrust for FWE with expanding Exhaust ......................................268
Figure 159 - Engine performance on liquid fuels..........................................................278

xxiii
List of Tables
Table 1- Specifications of the V-1 ..................................................................................... 8
Table 2- Specifications of ENICS E95 Target Decoy .......................................................... 9
Table 3- Specifications of AMT Pulsejet aircraft............................................................. 10
Table 4- Lockwood Performance Data ........................................................................... 23
Table 5 : Requirements and Input Data of Matching Diagram....................................... 40
Table 6 Characteristics of Suitable Aircraft.................................................................. 41
Table 7: Characteristics of Possible Aircraft ................................................................... 41
Table 8- Initial Aerofoil Analysis ..................................................................................... 43
Table 9: NACA 4 Digit Aerofoil Analysis .......................................................................... 43
Table 10- Suitable Tail Aerofoils ..................................................................................... 44
Table 11........................................................................................................................... 82
Table 12: Fuselage Stress Analysis Results ..................................................................... 88
Table 13 - Aileron Dimensions...................................................................................... 104
Table 14- Servo Requirements...................................................................................... 105
Table 15 - Spring stiffness and deflection under a 40N thrust load, for various frequecy
ratios. ............................................................................................................................ 110
Table 16: Material Selection for Engine Mount............................................................ 112
Table 17 - Ducted Fan Parameters ............................................................................... 124
Table 18 - Parameters of Lehner 1950 Electric Motor ................................................. 125
Table 19 - Expanding Exhaust Test Results................................................................... 156
Table 20 - Maximum control surface/servo motor deflection..................................... 170
Table 21 - Flight data from GPS logger ......................................................................... 176
Table 22- Hours Worked By Group Members .............................................................. 184
Table 23: General Configuration Decision Matrix ........................................................ 201
Table 24: Fuselage Configuration Decision Matrix....................................................... 202
Table 25: Engine Configuration Decision Matrix .......................................................... 203
Table 26: Wing Configuration Decision Matrix............................................................. 204
Table 27: Wing Height Decision Matrix ........................................................................ 205
Table 28: Wing Sweep Decision Matrix ........................................................................ 206
Table 29: Empennage Configuration Decision Matrix.................................................. 207

xxiv
Table 30: Landing Gear Type Decision Matrix...............................................................208
Table 31: Landing Gear Arrangement Decision Matrix.................................................208
Table 32- Weight Data for Piston UAVs ........................................................................211
Table 33- Table of First Iterations .................................................................................216
Table 34- UAV Data .......................................................................................................216
Table 35- Consistent UAV Weight Data ........................................................................218
Table 36- Iteration Results ............................................................................................220
Table 37- Fuel Flash Point Data.....................................................................................235
Table 38: Fuel Energy Density Data...............................................................................236
Table 39: Fuel Optimal AFR Data...................................................................................236
Table 40- Fuel Flammability Limit Data.........................................................................237
Table 41- Latent Heat of Vaporisation Data..................................................................238
Table 42 : Component Weight Breakdown...................................................................239

1
1 Introduction
The purpose of this project was to design and manufacture a Valveless Pulsejet
Powered Unmanned Aerial Vehicle (UAV), suitable for use as a Target Drone or Decoy
UAV. The project aimed to develop an understanding of valveless pulsejet engines, as
well as developing a prototype engine, with the aim of showing that they are a cheap
and viable alternative form of propulsion.

A feasibility study was initially conducted in order to develop project goals and define a
realistic scope. This stage included an extensive study of all valveless pulsejet engines,
developed by academics and enthusiasts, in order understand the working
characteristics of the engines, and to better understand how to optimise and improve
the operating characteristics of these engines. A study of target drones and decoy
aircraft, powered by both pulsejet and turbo jet engines was undertaken to help
develop the fundamentals of the aircraft, as well as to identify some of the key issues
which needed to be addressed in the following design stages.

The engine design was initially a continuation and modification of a Focused Wave
Energy (FWE) valveless pulsejet engine developed by Coombes et al in 2007 at the
University of Adelaide.

The aircraft design was produced initially using a combination of statistical and
numerical analysis, in consultation with aircraft design literature. The aircraft design
was then progressively refined in an iterative manner.

This project has involved a significant testing section, with over 100 static engine tests,
conducted, and two successful aircraft flights. The data obtained from this project has
helped to better develop the understandings of valveless pulsejet operational
characteristics, particularly with liquid fuels.

This report shows the development steps which were utilised to ensure the project
was completed on time; on budget and that all goals were achieved. This project has
Chapter 1 Introduction
2
shown through proof of concept that valveless pulsejet engines are a viable form of
propulsion for short range target drones and decoy UAVs.
1.1 Project definition
The project continues on from the work conducted by Coombes et.al in 2007. This
project expands from that work to concern the development of a valveless pulsejet
powered UAV.

The preceding data and research from the detailed feasibility study and bench-marking
has been synthesised to produce this project definition, which outlines the aims and
objectives of the project. This project definition is categorised into the Pulsejet and
Airframe development areas.
1.2 Project Aims
The project aims to show that valveless pulsejets are a viable alternative engine for
short range and low cost UAV aircraft. To ensure the project is completed the
following must be achieved.
1.2.1 Pulsejet Development
Continued development of a valveless pulsejet, with the aim of increasing the
overall thrust of the engine and a reduction in engine weight based on 2007
results.
Research and development of a liquid fuel delivery system for pulsejet engines.
Develop a fuel system that is of suitably low weight for flight.
Successfully test and record key performance criteria of the pulsejet on both
gas and liquid fuel mixes.
Development of alternative engine designs which may be suitable for future
development.
Completion of all tasks within the allocated budget.
Section 1.3 Project Goals
3
1.2.2 Airframe Development
Successful design of a low weight air-frame, based on the parameters of an
estimated 3kg of thrust and of plausible 300km/hr max speed.
Successful manufacture of the airframe using composite materials within the
specified weight.
Successful flight of the UAV.
Completion of all tasks within the allocated budget.
1.3 Project Goals
The project success was based on the completion of the following goals:
To modify, build and manufacture a valveless pulsejet, with the aim of
producing 3kg of thrust, with an engine weight of 1.5kg or less. This goal will be
quantified by the output received from the thrust measurement stand
constructed during the 2007 Project.
Develop a liquid fuelled system for a pulsejet engine and integrate a flight
weight version into the UAV design.
Based on the desired pulsejet specifications; design, develop and build a
lightweight UAV capable of sustaining flight for 10 minutes with thrust supplied
by a valveless pulsejet engine.
Achieve a cruise speed of over 200km/h. As measured by onboard GPS or a
similar system.
Achieve a flight time of 10 minutes
Gain a better engineering perspective on the workings of pulsejets, with the
aim of developing different engine design alternatives.
1.4 Extension Goals
Completion of extension goals will show above expected outcomes from the project:
Achieve 3.5kg of thrust from a valveless pulsejet engine.
Chapter 1 Introduction
4
Achieve a cruise speed of 250 km/h or above.
Increase flight time of the proposed liquid fuelled pulsejet UAV to over 15
minutes.
Manufacture an alternative engine design for future development.
1.5 Scope
The project scope is limited to the successful completion of the goals specified above.
The project aimed to design an airframe capable of supporting a pulsejet engine. While
this involved some optimisation of the airframe structure, the project only intended to
develop a proof of concept aircraft. The project is aimed to develop an aircraft capable
of flight, the scope of the project is limited to the initial development and
manufacturing stages, further alteration and optimisation after successful flight was
minimal.

The engine and fuel delivery systems were designed to be capable of producing 3kg of
static thrust for an estimated 10 minute flight time. The scope of this project was
therefore limited to the development of these systems to a level which will allow the
aircraft to sustain flight for the desired time period. Continued development of the
systems to optimise flight of the aircraft is not anticipated unless project goals are not
fulfilled.

5
2 Feasibility Study
The aim of this feasibility study was to construct realistic project goals, understand the
challenges and risks involved with the project and formulate a logical and progressive
development plan for the remainder of the project. This was completed through a
literature review and market survey of both the airframe and pulsejet components of
the project.
2.1 What is a Pulsejet
A pulsejet engine is a form of combustion engine with few to no moving parts. The
engine comes in two forms, valved or valveless (Figure 1). Both engines have similar
layouts, consisting of an intake, combustion chamber and exhaust.

Figure 1 - Valved and Valveless Pulsejet Designs
(Pulse-jet.com 2008)

The main difference between the two engines is in the use of a valve to direct the flow
out of the exhaust tube. This valve was the main source of problems in early pulsejets.

As can be seen in Figure 1, the valve in the valved engine is positioned inside the
combustion chamber. The combination of extreme heat and violent closing movement
of the valve meant that valved pulsejets often experienced lifetimes only lasting
several minutes before the vales fatigued.

Chapter 2 Feasibility Study
6
The valveless pulsejet uses an aerodynamic valve, created by the differences in length
between the intake and exhaust, in order to sustain operation. This means the engine
has no internal parts, and thus is significantly more reliable once and effective engine
layout has been created. It is for this reason that valveless pulsejet engines have been
investigated in this project.
2.2 Advantages + Disadvantages
The main advantage of valveless pulsejet engines is in their extreme low cost, as
shown in Figure 2. This is due to the engines simple design, and use of low cost and
readily available materials and manufacturing methods. This makes them an excellent
power plant for low cost target drones and decoy UAV (Tao 2006).

Figure 2 - Comparison of Engine Costs
Pulsejet engines however suffer mostly from their poor thermodynamic efficiency
(outlined in section 2.6.1), which means the specific fuel consumption of the engines is
significantly greater than that of common turbojet or turbofan aircraft.

The other disadvantage of pulsejet engines is the extreme levels of noise and vibration
they emit. This factor rules out the use of pulsejets in anything other than military
applications.

However the most interesting and exciting area of pulsejet engines is in the
combustion mechanism. Pulsating combustion is self compressing, so that the air fuel
Section 2.3 Pulsejet Engines in Aviation History
7
mixture does not burn steadily, but in bursts. This makes pulsejet engines an excellent
research engine, as many of the fundamental theories have been investigated on
pulsejets, before the construction of a large scale Pulse Detonation Engines (Wilson,
Dougherty 2002).
2.3 Pulsejet Engines in Aviation History
The pulsejet engine first found application in aircraft in 1891. Pulsejet engines have
been used throughout aviation history in several applications, including unmanned
military vehicles, early missile development, and vertical takeoff and landing (VTOL)
research, however much of the recent research has been undertaken by model aircraft
enthusiasts.

Sometime after the invention of the Pulsejet the Pulsejet powered German V-1 Missile
was produced. This missile is the pulsejet powered aerial vehicle produced in the
largest quantities with approximately 30,000 units manufactured. The V-1 missile
utilised a valved pulsejet engine and during tests of the V-1 significant failures
occurred, even though the aircraft only flew for less than 20 minutes (Goeble 2003).

In modern times much development in pulsejet engines has come from model aircraft
hobbyists, due to its low cost and comparable ease of manufacture.

The Pulsejet Engine has been of interest to commercial manufacturers throughout
several brief periods in history. Pulsejets have been used commercially and for the
military as propulsion devices for target drones.
2.4 Market Research and Benchmarking
The aim of this section was to gain an understanding into the capabilities and aircraft
configuration styles of pulsejet powered aircraft. Due to the lack of such aircraft, the
study was extended to both hobby aircraft and jet powered target drones and decoy
aircracft.
8
2.4.1 V-1
The V-1 was the first pulsejet powered aircraft, used by the German Air force during
World War II as a low cost and high quantity missile. It was the first mass-produced
guided missile and first jet powered aircraft.
Specifications of the V-1 can be seen in Table 1. The design of the V-1 is shown in
Figure 3.

Figure 3-View of the V-1
(Naughton 2001)
Table 1- Specifications of the V-1
(Combined from: Werrel 1985, Goebel 2003, Naughton 2001)
Engine Argus valved pulsejet 109-014
Thrust (kg) 272
Take-off weight (kg) 2150
Speed (kph) 630
Span (m) 5.3

While the size and weight of the aircraft is significantly larger than the anticipated UAV
weight, it useful for analysis as it is one of few aircraft which has been powered by a
pulsejet engine.
Section 2.4 Market Research and Benchmarking
9
2.4.2 ENICS Drones
ENICS is a Russian company which provides pulsejet powered decoy aircraft for
military training. ENICS produces fully manufactured drones and engines in three
different configurations. Full details of their E95 target decoy can be seen in Table 2.
Table 2- Specifications of ENICS E95 Target Decoy
(Enics 2006a)

Engine Enics M44D pulsejet
900 mm length, 75mm diameter
Engine weight (kg) 0.9
Thrust (kg) 20
SFC (kg/kg/hr) 6.61
Take-off weight (kg) 70
Span (m) 2.4
Speed (kph) 400
Range (km) 70
Endurance (min) 30
Launch Ramp, pneumatic

Figure 4-E-95 Ramp Launch
(Enics 2006b)
The aircraft is larger than the estimated project design, however its use as a target
drone and use of a valveless pulsejet engine make it an excellent aircraft for analysis.
10
2.4.3 AMT Pulsejet Hobby Aircraft
Pulsejets are moderately popular as propulsion systems for jet model aircraft. Pulsejets
are attractive to many pilots as they are low cost and offer good thrust to weight
ratios. In most cases commercially available valved engines are used.
This AMT Pulsejet is a custom built delta wing aircraft with a modified valved pulsejet
producing 8.7kg of thrust. The specifications of this aircraft can be seen in Table 3. This
aircraft is useful for analysis as it is close to the expected weight of the aircraft and its
use of a pulsejet allows analysis of expected fuel consumption during flight.
Table 3- Specifications of AMT Pulsejet aircraft
(AMT 1998)
Engine Custom valved pulsejet
Engine length (mm) 880
Engine diameter (mm) 90
Thrust (kg) 8.7
Take-off weight (kg) 7.5
Empty weight (kg) 5.9
Speed (kph) 390
Span (m) 1.12
Fuel Consumption 500 mL/min

[50% Kerosene, 40% Gasoline, 10% Propylene
Oxide]
2.4.4 Comparison to turbine engine UAVs or Target Drones
Turbine engines UAVs similar in size to the project aircraft have a large advantage in
terms of thrust to weight comparison to pulsejet engines. This is due to the
compactness of the engines, as well as comparably lower fuel consumption figures.
However the main disadvantage of these types of engines is the cost of the engine for
the similar amount of thrust, as shown in Section 2.2. For the statistical design of the
aircraft, turbine powered UAVs will be included in the analysis, due to the low number
of pulsejet powered aircraft, specifically of comparable size to the anticipated design
size of the aircraft.
Section 2.5 Mission Profile Specifications
11
2.5 Mission Profile Specifications
2.5.1 Mission Profile
Based on the analysis of the aircraft in Section 2.4 , the mission profile of the aircraft
was developed. As the aircraft was aimed to be developed as a proof of concept
aircraft, it was decided that the mission profile would be kept simple. The profile can
be seen in Figure 5.

Figure 5- Flight Profile
Further details are specified for some sections of this profile:
Start up and warm up - with pulsejets this is especially critical, as the engines
must be stable before launch. As a result, the fuel consumption during this
period will be significantly higher than for other engine types.
Loiter The goal of the flight is top remain airborne for 10-15 minutes with no
set range goal therefore the flight will take place within the line of sight of the
pilot.
2.5.2 System Requirements
The system requirements define the abilities of the aircraft and its components. These
values are determined from know requirements, calculations and the market research
performed on similar aircraft.
12
Cruise Speed Requirements
In section 2.4, different aircraft that utilized pulsejet engines or jet engines for power
were analysed. A suitable requirement for the cruise speed can be decided based on
that data and other calculations. A realistic cruise speed requirement was determined
based on numerous things.
Direct Bench Marking
The direct bench marking here refers to other pulsejet aircraft of similar size. A small
pulsejet aircraft presented earlier that was similar was the AMT. This had a top speed
of 390 kph, but also had a thrust to weight ratio greater than what we are aiming for.
Collated statistics of other Jet UAVs
Of the aircraft identified in the research and benchmarking section it can be seen that
most jet powered target drone aircraft have a cruise speed of approximately
400km/hr. However these aircraft have high thrust loadings and also high wing
loadings which reduce wing area and thus drag/weight. These characteristics are
allowed by the use of rocket-assisted launch and/or ramp launches and multiple or
more powerful engines. The following graph shows the thrust loading for a variety of
aircraft. It can be seen that the mean thrust loading is approximately 0.3. This was the
basis for all further aircraft development.

Section 2.5 Mission Profile Specifications
13

Figure 6 - Statistical Trends of Target Drone UAVs
Estimated available thrust
The amount of thrust currently available from the engine that is to be used for this
aircraft is 2.3kg (Coombes et.al 2007). This is described in more detail in Section 2.6.2
Estimated possible speed
Using calculations that estimate the drag based on the estimated drag in conjunction
with thrust and weight, it was possible to estimate the possible top. For thrust around
2.3 kg, and with a thrust loading of 0.3, from the mean of the collated data, top speeds
of 200-250 kph are possible.
Plausible Cruise Speed Requirement
Based on the above considerations, a realistic cruise speed requirement was deemed
to be 200 kph.
Control and Electronic Requirements
There were two separate control functions for the aircraft, control of the flight and
control of the engine.
Control of Flight
14
In selecting a control mechanism for the flight of the UAV it was determined that
simplicity was of importance, due to this a common remote controlled system was
deemed appropriate over other systems.
Control of the Engine
The thrust produced by the engine could be controlled by varying the fuel flow rate
supplied to the engine by the fuel pump.
2.5.3 Takeoff methods
There are numerous ways that a UAV system can be launched including trolley
launched and fixed gear. These methods have significantly different characteristics and
will be discussed.
Trolley Launched
The idea of a UAV being launched from a trolley or with a detachable landing gear is
that once the aircraft leaves the ground the trolley or gears detach from the aircraft.
This can be done using the propulsive power of the aircrafts own propulsion system
and or with a supplementary propulsion system such as rockets or sling shot. This take-
off method requires that the aircraft has an alternate landing method other than via
landing gear. The advantage of this launch method is that it has no need for a landing
gear which would decrease drag during flight. However the main disadvantage is that
it requires an alternate landing method such as a parachute.
Fixed Gear
For an aircraft taking off from an attached landing gear both fixed and retractable
types of landing gear designs can be considered. The main advantage of a fixed gear is
that the system is reliable and simple however it has a disadvantage of increased drag
during flight.
Section 2.6 Power plant Design
15
2.5.4 Landing Options
There are several alternative landing methods for UAVs without a conventional landing
system. These options have been considered to determine the overall risk and
feasibility of designing an aircraft without a conventional style landing gear. The four
options considered are parachute, belly landing, net catch and air cushioned landing.
Parachute
There are numerous advantages to landing an aircraft with a parachute. Recovery
parachutes are commercially available at a relatively low cost and they produce
minimal extra drag in comparison to a fixed landing gear system. The main
disadvantages of a parachute recovery are the complexity and weight of the system
and the high loads experienced when the parachute is first deployed.
Belly Landing
The use of a belly landing for an aircraft has numerous benefits, primarily the minimal
effect on drag, the slight effect on the weight of the aircraft and the low complexity of
the system. For a belly landing the underside of the aircraft is reinforced to withstand
the forces created by the impact of the aircraft with the ground, which is the main
disadvantage of this system.
2.6 Power plant Design
The power plant for the engine was defined by the initial project outline. This section
outlines the initial research that was conducted by the group into the workings,
research and challenges that exist in designing a valveless pulsejet engine.
2.6.1 Valveless Pulsejet - Thermodynamic Cycle
Pulsating combustion is the main area of confusion for researchers attempting to
successfully understand the operation of pulsejet engines. In the research conducted,
it has been noticed that different authors associated the behaviour of the engines to
16
different phenomenon. The self sustaining, periodical nature of the combustion is
generally associated with either wavy, acoustic or vortex nature (Pobezhimov 2006),
however models using these analysis generally can only describe parts of the
combustion process accurately. A thermodynamic approach can be used to explain the
operating process of a pulsejet engine, and show the advantages that exist in pulsating
combustion. The operating cycle of a pulsejet engine can be described by
modifications to the Lenoir Cycle which can be seen in Figure 7.

Figure 7 - Ideal Lenoir Cycle
(Pobezhimov 2006)
The operating cycle is described in thee steps:
1-2 Constant volume (isochoric) heat addition
2-3 isentropic expansion.
3-1 Constant pressure (isobaric) heat rejection - compression to the volume
at the start of the cycle.

The main difference between the Lenoir cycle and a pulsejet cycle is that during heat
addition the process is neither isochoric or isobaric, as there is a combination of
pressure release, and heat release (McCalley 2006). This is because the engine
operates from wave compression which is relatively weak; therefore combustion is not
confined to the combustion chamber, but occurs down the length of the engine. A
more realistic diagram can be seen in Figure 8.
17

Figure 8-Pobezhimov modified Lenoir cycle
(Pobezhimov 2006)
2.6.2 Review of Previous Work
From the research conducted, it was found that the development of pulsejet engines
has been the recent study of several universities. The two of interest to this project
were studies conducted by North Carolina State University and The University of
Adelaide. The work conducted by these two bodies allowed for a better understanding
on the fundamentals of pulsejet engine operation and optimisation.
North Carolina State University
Within the past decade, numerous investigations have been conducted by North
Carolina State University Masters students, under the direction of Dr. William L.
Roberts into various areas of pulsejet engine development. Studies have included:

Experimental Investigations into Pulsejet Engines
Experimental Investigations Into The Operational Parameters of a 50
Centimetre Class Pulsejet Engine
Experimental Investigations in 15 Centimetre Class Pulsejet Engines
Experimental Investigations of 8 Centimetre Class Pulsejet Engines
Experimental Investigations of Liquid Fuelled Pulsejet Engines
Numerical Simulations of Pulsejet Engines

18
These investigations have aimed to better understand the operating characteristics of
valved and valveless pulsejet engines, as well as attempting to develop small scale
engines for use with small size UAV and MAV aircraft, as the efficiency of commonly
used turbojets becomes lower as the size of the engine decreases (Tao 2006).

This section covers some of the key research conducted by these projects, with focus
on fundamental operating theories and engine performance. Work into the
development of liquid fuelled pulsejet engines, as conducted by McCalley in 2006, is
covered in section 27.
Valveless Engine Studies
Valveless engines studies conducted at Carolina State University have revolved around
the analysis of straight exhaust valveless engines, known as Schubert jets, (Figure 9),
with varying lengths and diameters of the intake and exhaust pipes. Schubert jets are
known for their ease of manufacture, but low thrust and high specific fuel
consumption.

Figure 9 - Valveless Pulsejet Engine (Carolina State University)
Experimental data was taken from the engines via a number of different mechanisms,
including instantaneous pressure sensors, manometers, thermocouples and SPL
meters.

19
Experiments in varying the length of the valveless pulsejet engine showed a direct
correlation between operating frequency and length. This frequency can be linked
directly to the Helmholtz frequency for the intake pipe (Equation 1) and a 1/4 wave
tube frequency for the exhaust (Equation 2).

Equation 1

Equation 2
L
C
f
4
=
It was found that both these frequencies act together to give the engine operating
characteristics which are similar to that of a 1/6 wave tube. This was compared to
tested data and was found to be accurate to within 5%. The equation is temperature
dependant, which suggests that changes in area in the engine can cause a change in
the operation frequency. Also, changes in fuel will alter the burn temperature and thus
affect the engines operating characteristics. However it was found that if the intake
and exhaust frequency are within 10% of each other, the engine will successfully
operate.

Studies by Ordon in 2006 showed that this frequency characteristic is altered
significantly by changes in geometry, as these cause reflections in the waves, which
effect how the jet operates. It was found later by Kiker that the operating frequency of
the pulsejet scales as the inverse of the inlet length and reducing the exhaust diameter
of the pulsejet has very little effect on its operating frequency. With respect to
combustion chamber pressures, Kiker found that pressure scaled inversely with exit
diameter and directly to fuel flow rates. He also investigated the use of platinum
coating in a 5cm pulsejet to act as a catalyst and increase chemical reaction time.

20
2007 Study University of Adelaide
In 2007, a study by Coombes et.al Al 2007, was conducted at Adelaide University into
the devolvement and testing of a valveless pulsejet engine and thrust measurement
stand. The work aimed to create an engine capable of 3kg of thrust, with an engine
weight of under 2kg, a stand capable of accurately measuring the engines thrust during
tests and a software package to be used to predict pulsejet performance and allow the
optimization of engines
The groups work focused on the development of a Focus Wave Energy (FWE) Valveless
pulsejet engine, as shown in Figure 10, which was originally developed by notable
pulsejet engine developer, Larry Contril.

Figure 10 - Focus Wave Energy (FWE) Pulsejet Engine
Two engines were developed, the first based on statistical design, with adjustable
lengths. This engine aimed to investigate the effect of the intake and exhaust lengths
on engine performance, and a fixed length engine, developed based on findings from
the engine prediction program.
The work completed produced an optimum engine configuration which produced
2.392kg of static thrust with a total length of 1035mm. The second engine developed
was not successful in achieving sustained combustion.
21
The notable areas of interest are in the relationships which lead to the design of their
statistical based engine, the testing procedure they utilized, the theory behind the
development of the engine design software, and finally the problems and risks they
encountered throughout the project.
2.6.3 Alternative Engine Designs
Numerous different valveless engine designs have been developed, with the aim of
improving the performance of the engines. In selecting a valveless engine for use on a
UAV, thrust output, fuel consumption and aerodynamic performance must be
considered. This section outlines some of the most successful pulsejet engines which
have been developed, with the aim of identifying the most suitable engine for a flight
weight aircraft.
Chinese Pulsejet Engine
The Chinese Pulsejet engine was developed in the 1960s by CS manufacturing, a 2-
stroke motor designer from Shanghai. The engine is characterised by its expanding tail
exhaust and cylindrical combustion chamber (Figure 11). CS manufactured two
commercially available engines, which were designed to run on regular gasoline. In
1993 the designs for the engine became public, and it has since been developed by
enthusiasts for use with propane fuel systems.

Figure 11-Chinese Valveless Pulsejet Engine
(Beck 2008)
22

The engine is streamline in design, with rearward facing exhaust and intake to ensure
all thrust created acts in the same direction. No analytical research has been
conducted into this specific design, however specific fuel consumptions of between
3kg/kg/hr and 6kg/kg/hr have been noted from enthusiasts. Thrust to weight ratios of
between three and five have been achieved.
Lockwood Valveless Engine
The Lockwood valveless engine has been the most successful valveless pulsejet
developed in recorded history. The engine was investigated between during the 1960s
as a form of propulsion for vertical takeoff and landing (VTOL) aircraft. The engine is a
U-shape, with the exhaust bent around 180 degrees to direct both the intake and
exhaust thrusts in the same direction. A table of the final engine performance claims
can be seen in Table 4, however it should be noted that these values have never been
achieved using the patented design, specific fuel consumptions closer to 5kg/kg/hr
have been seen, with thrust results approximately 25% less than claimed. The
aerodynamic performance of the engine is also poor, in comparison to the Chinese and
FWE designs shown earlier.

Figure 12-Lockwood Prototype
(Lockwood 1957)
23
Table 4- Lockwood Performance Data
(Lockwood 1957)
Model HH 5.25-7
Valveless Engine
Military max thrust (lbs) 300
Maximum continuous (lbs) 280
Minimum idle (lbs) 30
Idle to mil. max time (secs) 0.1
Fuel/thrust (lb/lb/hr) 0.85
Dry weight (lbs) 30
Escopette
The Escopette was developed by the French research agency SNECMA (Societe
Nationale d'Etude et de Construction de Moteurs d'Aviation) in 1950. The engine was
the first developed with a rearward facing intake, and with expanding sections in the
exhaust.

Figure 13-Escopette Valveless Engine
The engines operating characteristics are different to a normal pulsejet, due to the
unique exhaust design and separation between the curved intake and the main engine.

The split intake allows the engine to behave as if its length were variable long during
the exhaust phase of the cycle and short during the intake phase. During expansion, it
treats the curved intake as a part of the effective length of the engine and uses it to
turn the escaping gas around and increase thrust (Ogorelec 2004). During the intake
cycle the curved section is not used. This reduces the effective length of the intake and
lets the engine inhale more easily.

The tailpipe is a series of steps of increasing section. Each transition from a straight
24
section into a diffusing section represents a point from which the pressure waves
travelling up and down the tube will reflect. Each of these waves reflects in an area of
varying temperature, and therefore they all travel at different speeds. The interaction
and timing of these waves are critical to the engines operation (Figure 13).

Figure 14 - Interaction of Escopette Pressure Waves
(Belfast University 1983)

The unique design of the engine means that it inhales twice for each expansion cycle,
with the aim of increasing the amount of cool air drawn into the exhaust section. This
increases the mass of the air in the exhaust and thus allows energy from the
combustion process to be converted more efficiently into thrust.

The original engine produced 108N of thrust, with a fuel consumption of 19.8kg/hr.
The engine however was extremely long at over 2.6m.
2.6.4 Exhaust Pipe Development
From the analysis of the pulsejet engines in Section 2.6.3, it can be seen that the
performance of a pulsejet engine is reliant on the behavior of the dominant waves in
the engines exhaust. Modifications of the engine exhaust characteristics can have a
dramatic effect on the engine performance. Studies by Artt and Balair in 1983 found
that altering the exhaust of a valved pulsejet engine could improve its performance by
up to 25%. The following section investigates the various exhaust designs,
25
characteristics and theories, in order to provide a knowledge base from which
modifications to the existing engines can be made.
Straight Pipe
Straight pipe exhausts are generally found on basic engines designed for first time
builders. The most common engine design to use a straight pipe is the Focused Wave
Engine, shown in Figure 15.

Figure 15 A Focused Wave (FWE) Pulsejet engine
(Beck, 2008)
The advantage of this type of exhaust is primarily ease of manufacture and cost
reduction, as commercially available pipe can be used, without the hassle of forming
conical sections. The section only operates on a single refraction wave returning from
the end of the exhaust, significantly reducing the engines throttle range, and thrust
output (Artt 1983).
Expansion Pipe
This type is the most common exhaust found on designs that produce reasonable to
high levels of thrust. Popular designs include the Lockwood-Hiller engine (Figure 16), as
well as the Chinese engine.
26

Figure 16 A Lockwood-Hiller style Pulsejet engine,
(Kontou 2007)
The most common justification of the expansion pipe is related to the drawing of cool
air into the exhaust section after the combustion pulse. The larger entrance volume
draws a larger volume of cool air into the exhaust. This dense air then acts as a cold air
piston which is accelerated by the combustion wave leading to more thrust.

The expansion may also be justified through energy and momentum laws. The thrust a
jet engine creates is related to the momentum of the expanded gasses leaving the
engine, and hence proportional to the product of mass and velocity. The energy of a
particle however is proportional to the product of mass and the square of velocity.
Hence, it can be deduced that for a given energy input, the thrust of the engine can be
maximized by reducing the velocity of the gasses leaving the engine. By expanding the
exhaust pipe, the flow velocity will reduce, resulting in a larger momentum of the
gasses leaving the engine, and hence a thrust increase.
Advanced Pipes
Some advanced pulsejet engines feature unique exhaust designs that help to improve
the efficiency of the engine. An example of this is the SNECMA Escopette engine,
27
which was explained in the previous section. Another unique exhaust design is found
on the FWE VIII - Lady Anne Boleyn engine as shown in Figure 17. The exhaust on this
engine has a diverging then converging section between the combustion chamber and
the final expansion.

Figure 17 A Focused Wave engine variation, the FWE VIII - Lady Anne Boleyn.
(Cottrill 2006)
2.6.5 Liquid Fuelling
There are a number of critical design aspects when it comes to using liquid fuel to
power a valveless pulsejet engine. These include the choice between direct injection
and carburetion for supplying fuel to the combustion zone, as well as the advantages
and disadvantages of different fuels depending on the desired use of the engine. These
aspects of liquid fuelling are discussed in more detail below.
Carburetion or Direct Injector
An atomisation, or carburetion system, is potentially the simplest system available for
liquid fuelling. The principals are similar to those for carburettor in an internal
combustion engine. By Incorporating a venturi into the intake the pressure within the
pipe at the injector point is reduced, causing fuel to be drawn from the fuel tank.

While this system is simple in design it is very sensitive to pressure head, and therefore
fuel tank placement. This is compounded in flight by the aircraft performing
manoeuvres. The other disadvantage to a carburetion system for pulsejet applications
is that the engine cannot be throttled. In a normal carburetion system the engine is
28
throttled by increasing and decreasing the air flow rate, and thus the draw of the
venturi, in a pulsejet, the intake flow cannot be varied, as it will affect the operating
characteristics of the engine.

A direct injection fuel system has the added complication of a fuel pump, which is used
to provide a constant pressurised fuel supply to the engine. A pump requires its own
power supply and requires the operator to control the fuel flow rate into the engine.
Due to the constant fuel supply from the pump, slightly more fuel is wasted during the
combustion cycle of a direct injection system when the exhaust gasses are pushed out
of the intake pipe. The system however is not affected by fuel tank placement or
aircraft angle and has been shown to be able to be throttled (McCalley 2004).
Fuels
There are a wide range of fuel properties that must be considered when determining
the type of fuel to be used to power a valveless pulsejet engine, such as flash point,
energy density, air-fuel ratio and latent heat of evaporation. Unlike many engines, a
pulsejet engine can run on a wide range of liquid fuels, without the need for major
redesign (Simpson, 2006). However, as valveless pulsejet engines are heavily
dependent on the acoustic lengths of the engine to obtain resonance, temperature
changes within the engine will directly impact the operation of the engine. Therefore,
different fuels, with different combustion temperatures, will directly impact the
engines operation. Successful operation of a valveless Schubert jet on kerosene and
unleaded petrol was achieved by McCalley in 2006.
Injector Design
The studies by McCalley in 2006 were directed towards the operation of a valveless
pulsejet engine on heavy fuels, such as kerosene. The investigation suggested that
injector design and fuel atomisation were vital in ensuring the engines operation. It
was also noted that the engine must be brought up to operating temperature using
propane first, as the compression of the pulsejet was not enough to cause combustion
at room temperature.
29

The study developed numerous pinhole injector designs, varying from single hole
injectors, to 6 hole swirl patterns. Results suggested that a high pressure injector was
required, as low restriction injectors (multiple hole injectors) were not capable of
producing atomisation. These injectors were also found to lead to boiling of the fuel
within the fuel injector head, starving the engine of fuel. Successful operation of the
jet on liquid fuel was achieved, however it was noted that variations in the injector
position were vital in achieving self sustained combustion.
2.6.6 Thrust Augmentation
Thrust augmenters or ejectors, are a simple addition which can be made to pulsejet
engines. The augmenter is positioned behind the exhaust section of the pulsejet and is
used to capture the starting vortices which form with each cycle of the engine (Paxton
2006). The augmenter uses these starting vortices to draw in and mix cold air with the
hot exhaust gasses adding extra mass to the exhaust flow. Studies have shown that
exhaust thrust augmentation can increase engine thrust by over 100%, with designs
which are much simpler and more compact that conventional steady state ejectors
(Kailasanathan 2007). However, as it can be seen in Figure 18, the addition of an
optimised ejector to a pulsejet engine would increase the drag of the engine
significantly in flight.

Figure 18- optimised thrust augmenter as used on a valved pulsejet
30
2.7 Feasibility Study Summary
The above data from the detailed feasibility study and bench-marking has been
synthesised to produce the project definition, which outlines the aims and objectives
of the project, as well as develop the direction for further work.

31
3 Conceptual Design
The conceptual design stage aimed to develop a feasible and realistic concept aircraft
and engine which met all the criteria outlined within the feasibility study and project
definition. The aircraft design process entailed an initial statistical overview of existing
aircraft and configuration options before sizing of the aircraft to meet the
requirements outlined in the project definition. The engine development involved an
initial testing of the engines developed by Coombs et al. (2007) and the design of
modifications to the second prototype engine developed by the group. In order to
manage the risks associated with this project a second Chinese valveless engine was
investigated and designed from statistical data and professional opinion.
3.1 Aircraft Conceptual Design Introduction
The first step was to determine a configuration then using concept sketches and
statistical considerations to evolve towards a concept based on the configuration.
3.2 Selecting Preliminary Aircraft Concept
To develop a concept, a general configuration first had to be selected. The initial
configuration of the aircraft was determined through use of a decision matrix method.
A summary is presented here while the complete breakdown is presented in Appendix
A and the review of pulsejet and target drone aircraft conducted in the feasibility study
has been taken into consideration when determining the rankings
3.2.1 General Configuration
The general configuration is the overall configuration of the aircraft, entailing options
such as monoplanes, biplanes. In this area, the available layouts were rated in the
areas of pulsejet suitability, complexity, weight, stability and drag. The highest rated
option was a conventional monoplane due to high ratings in all areas, particularly
pulsejet suitability.
Chapter 3 Conceptual Design
32
3.2.2 Fuselage Configuration
The fuselage configuration entails the design of the fuselage, with several variations on
a conventional tubular fuselage available. The fuselage options were rated in the areas
of cost, complexity and weight, considering that no payload would be carried. After
consideration, the highest rated fuselage was a conventional fuselage due to its
strength in all areas.
3.2.3 Engine Configuration
The engine configuration includes both the number of engines and the mounting
method used as a pulsejet was specified in the project brief. These options were rated
in the areas of cost, reliability, weight, heat and performance. A single engine was
optimal for all areas and so was highest rated. However, in this area final design thrust
required may be the determining factor if two engines are required.
3.2.4 Wing Configuration
The wing configuration included shape, height and sweep areas with common options
considered for all. In the area of a linearly tapered wing was preferable due to a
balance of all factors. The best wing height was a low wing due primarily to minimal
landing gear weight. A non-swept wing was determined as the preferred wing but aft
sweep may be required for stability purposes depending on future calculations.
3.2.5 Empennage Configuration
The empennage configuration options included a conventional tail, T-tail, H-tail and V-
tail. These options were rated in the areas of suitability to a pulsejet, complexity,
weight and size. The H-tail was the preferred option based on these factors, primarily
due to the suitability of this design to a pulsejet.
Section 3.2 Selecting Preliminary Aircraft Concept
33
3.2.6 Landing Gear Configuration
Landing gear configuration encompasses both the ability to be retracted and wheel
arrangement. In the area of the ability to be retracted, fixed, retractable and
removable landing gears were considered as well as a design with no landing gear
requiring a launcher. These options were rated based on drag, complexity, weight and
reliability with a fixed landing gear the best option for this aircraft. Wheel
arrangements considered were tail-dragger, bicycle and tricycle designs. The areas of
consideration for these designs were drag, stability and weight with a tricycle landing
gear being the selected design.
3.2.7 Basic Wing Parameters
For later conceptual calculations, some basic wing parameters were required to be
determined.
Aspect Ratio
Aspect ratio is usually determined by a trade study between structural weight and
aerodynamic efficiency, however the scope of this project does not allow for such a
complex optimisation therefore the aspect ratio of this aircraft was be selected based
on statistics. A medium aspect ratio of 6.5 will be selected here, because it is
statistically consistent with other similar aircraft and provides a balance between drag
and weight.
Taper Ratio
The taper ratio of an aircraft can increase its aerodynamic performance by more
closely approximating an ideal elliptical wing, reducing induced drag. For the
determination of taper ratio, Raymer (1992) determines that a taper ratio of 0.45 is
ideal for a non-swept wing and as the wing used for this aircraft is slightly swept, a
slightly smaller taper ratio such as 0.4 should be used.
34
Wing Sweep
The wing sweep of the aircraft is normally zero for an aircraft which does not approach
a sonic Mach number. However, sweep is also used to adjust the aerodynamic centre
of the aircraft and allow aircraft. As this aircraft will have a heavy engine located at
the rear, stability will be a concern and so the wing will be swept so that the trailing
edge is straight with the taper ratio and the front edge swept.
Wing Twist
Wing twist is used to prevent the onset of stall at the wingtips and enable the pilot to
retain control of the aircraft during stall. The non-use of twist would also simplify
manufacture, although this is not the overriding requirement.
Wing Incidence Angle
The wing incidence angle is set to balance between trim drag in flight and drag at
takeoff. However, generally the wing incidence angle is set to that which will minimise
cruise drag. This will be decided once an aerofoil has been selected for this aircraft.
Dihedral Angle
A dihedral angle can be applied to the wings to provide additional roll stability.
However, the use of an H-tail to provide two surfaces and rearward sweep on the
wings reduces the requirement for dihedral angle. As a result, a dihedral angle is not
required on this aircraft.
3.3 Developing concept for selected configuration
Basic concept based on the selected configuration were developed using sketches and
statistics.
Section 3.3 Developing concept for selected configuration
35
3.3.1 Concept Sketches
Some preliminary concept sketches were prepared, shown in Figure 19. These
outlined various options for positioning the pulsejet engine on the aircraft. The top
concept has the pulsejet mounted underneath and an aerofoil shaped fuselage. While
the third concept is a sketch whose main feature is an open fuselage allowing suitable
cooling airflow past the pulsejet.

Figure 19-Configuration Concept Sketches
3.3.2 Statistical Calculations
An extensive collection of all the available pulsejet and jet UAV statistical data was
compiled from which trends were noted in some important statistics.
Statistical Estimates of Thrust and Wing Loadings
Thrust loading is the thrust to weight ratio of the aircraft. The majority of thrust
loadings of the aircraft surveyed were between 0.3 and 0.6, providing a reasonable
minimum thrust for flight. All aircraft before this loading used an alternate launch
36
method such as release from aircraft or rocket assisted. The median thrust loading of
jet UAVs is approximately 0.4 and this is approximately the required thrust loading for
the project vehicle. A graphical representation of statistical thrust loadings is below in
Figure 20.
Thrust Loading of UAVs
0
0.2
0.4
0.6
0.8
1
1.2
1.4
T
/
W

Figure 20: Statistical Thrust Loadings of Jet UAVs
The range of wing loadings of UAVs was also analyzed, showing correlation with weight
of the aircraft. Below in Figure 21 is a graph of the wing loading versus the weight of
statistical UAVs, showing this relationship. This is as reasonably expected as a larger
aircraft would allow a longer runway length for takeoff and landing, hence allowing a
greater wing loading
Section 3.3 Developing concept for selected configuration
37
Wing Loading - Weight
W/S = 83.872Ln(W) - 182.62
0
100
200
300
400
500
600
700
0 100 200 300 400 500 600 700 800
Weight (kg)
W
i
n
g

L
o
a
d
i
n
g

Figure 21: Wing Loading Versus Weight of Jet UAVs
Statistical Estimates of Length and Span
From statistics of weight and length, for the estimated size of the aircraft of
approximately 8 kg, it was possible to estimate the main spatial parameters length and
span from trend-lines found in plots of length and span versus length. Via this method,
the length was statistically estimated at approximately 1.4 m while span was
statistically estimated at approximately 1.6 m.
Statistical Concept
Based on the statistical review, a statistical concept was produced this is shown in
Figure 22. It has a span of 1.6m, length of 1.4 m, and conventional layout with H-tail.
38

Figure 22: Statistical Concept
3.4 Designing technical parameters for concept
Following the development of a basic concept, the technical parameters of the
concept to meet the performance requirements in the project goals were determined.
The weight is being determined, and the size of the aircraft to meet climb, cruise and
requirements was determined.
3.4.1 Weight Estimation
The weight estimations for the aircraft were completed based on statistical
information obtained from Munson (2001) for similar sized piston powered UAVs.
Piston powered UAVs were used as a basis due to expected size similarity to the
pulsejet UAV and ease of data acquisition. The additional weight of a pulsejet and
additional fuel was taken into account as a payload for a piston powered aircraft.
Weight estimation calculations aimed to determine:
W
e
empty weight of the aircraft
W
f
weight of the fuel the aircraft can carry
W
o
the takeoff weight of the aircraft.
Section 3.4 Designing technical parameters for concept
39
The calculations are based on equating Equation 3 and Equation 4 below to compare
values required by the mission and statistical values from past UAVs.
Equation 3:
O W
W
O W
W
O
W W W
O
E
O
F
+ =

Equation 4:
S
C
O
O
E
K AW
W
W
=
The values for A, c and K
s
were determined from a statistical review of piston powered
UAVs. An iterative procedure was then utilised in the Microsoft Excel spreadsheet
program to determine W
o
. From the converged procedure weight values were
obtained. The empty weight of the aircraft was 6.1kg, fuel weight was 1.9kg and
takeoff weight was 8kg. A full outline of the weight calculation estimation procedure
can be seen in Appendix B.
3.4.2 Matching Diagram
A matching diagram is a plot of take-off thrust loading versus take-off wing loading. It
is used to find two or three possible limit case combinations of wing loading and
thrust loading to evaluate and determine an optimal aircraft. For this aircraft the
matching diagram contains a stall requirement, a cruise requirement, a single climb
requirement and a takeoff length requirement. Details of the matching diagram
calculation are contained in Appendix C Matching Diagram
Summary of requirements and assumptions
The requirements used to produce the matching diagram are summarized in Table 5.
40
Table 5 : Requirements and Input Data of Matching Diagram
Parameter Requirement
Stall Speed 60km/hr
Cruise Speed 260km/hr
Climb rate 8.33% at 78km/hr
Takeoff distance 100m
Input Data
Takeoff Weight 8kg
Fuel Weight 2kg
Specific Fuel Consumption 7.2kg/kg/hr
Cl
max
1.2
Aspect Ratio 6.5
Matching Diagram Plot
All of the sizing requirements that were found were plotted onto the matching
diagram seen in Figure 23.

Figure 23: Matching Diagram
41
The matching diagram combining all of the requirements (climb, cruise, take-off and
stall) is plotted above. Only one intersection point (designated A) fits both cruise and
stall requirements. However, if the stall speed was extended another point (designated
B) could be considered at the intersection of the cruise and takeoff requirements. The
data for the intersection points can be seen in Table 6.
Table 6 Characteristics of Suitable Aircraft
Point T/W (W/S)
TO
[kg/m2]
A 0.375 20.81
B 0.323 29.31

Discussion of Suitable Points: A & B
Two possible aircraft can be found from the above diagram. Table 7 below summarises
the performance of Aircraft A and Aircraft B.
Table 7: Characteristics of Possible Aircraft

W/S
(kg/m
2
)
T/W
(kg/kg)
V
s
(km/hr)
V
cr
(km/hr)
s
to

(m)
S
ref

(m
2
)
T
(kg)
Aircraft A 20.8 0.38 60 260 57.8 0.38 3.04
Aircraft B 29.3 0.32 71.2 260 100 0.27 2.56

From the above data it can be seen that Aircraft A has superior performance in both
stall speed and takeoff distance. Aircraft B would have superior efficiency in cruise due
to a lower range of operational speeds. The result of this improved efficiency is a lower
thrust loading requirement. Aircraft B also requires a smaller wing area which would
reduce the susceptibility of the aircraft to turbulence.
After a consideration of the above characteristics, Aircraft A has been selected due to
the lower risk of failure in the design. The lower stall speed is significant to reduce the
chance of failure on landing. A shorter runway length will also allow more freedom in
possible launch locations. For a proof of concept project such as this a safe design is
42
more critical than optimal efficiency as efficiency can easily improved in future designs.
A sensitivity analysis was performed for Aircraft A, with the results available in
Appendix D Sensitivity Analysis
3.4.3 Aerofoil Selection
To enable the tail to be sized correctly, aerofoil characteristics for both wing and tail
were required. Different aerofoils were required for the wings and tail due to varying
requirements.

An aerofoil for the wing was required to meet various parameters to be suitable for
use on the aircraft and provide performance as expected. The aerofoil was also
required to perform adequately across the range of Reynolds numbers which would be
encountered during operation. The selected wing aerofoil was the NACA 4412.

The tail aerofoils were selected to provide suitable symmetric performance in the
variable conditions at which they will be operating. The NACA 0012 aerofoil was
selected for the horizontal and vertical tails via a similar method to the wing aerofoil.
Wing Aerofoil Selection
An aerofoil for the wing was required to meet various parameters to be suitable for
use on the aircraft and provide performance as expected in the initial sizing. The
aerofoil needed a three dimensional lift coefficient equal to or greater than 1.2,
minimal drag across the expected range of operation (from a C
l
of 1.2 to 0.1 at
maximum speed level cruise).

Initially, various aerofoils with characteristics indicative of their families were tested in
Javafoil analysis to determine which were worthy of closer examination. The Clark,
SD, NACA 4, 5 and 6 digit series aerofoils were analysed with the results of this testing
presented in Table 8. This initial analysis also operated as a test of the accuracy
provided by Javafoil by comparison with two dimensional wind tunnel test data from
Abbott, von Doenhoff and Stivers 1945 report.
43
Table 8- Initial Aerofoil Analysis
Profile C
lmax
(C
lmax
) C
m
(C
lmax
) (L/D)
max
C
Do
(C
Do
)
NACA 4412 1.22 12 -0.12 23.6 0.011 -4
Clark Y 1.12 11 -0.10 22.9 0.012 -2
SD7034 1.12 12 -0.08 23.8 0.011 -3
NACA 23015 1.13 15 -0.04 20.4 0.010 -1
NACA 63-615 1.21 11 -0.15 22.0 0.011 -4
Experimental Data
NACA 4412 1.28 11 -0.10 N/A 0.010 -4
NACA 23012 1.24 10 -0.02 N/A 0.010 -2
NACA 63-615 1.38 10 -0.12 N/A 0.011 -4

The three dimensional approximations used by Javafoil produced data that
approximates that determined from NACA wind tunnel testing and so were suitable for
use. The test reports by Abbott, von Doenhoff and Stivers (1945) give lift coefficients of
between 5% and 14% more than those from Javafoil and drag coefficients within 8%
as shown in Table 8.

The initial testing shows that the NACA 4 digit series is the best for this application. As
a result, the family chosen for closer analysis is the NACA 4 digit series which has
superior aerodynamic efficiency to other families tested.

Various NACA four digit aerofoils were analysed in Javafoil three dimensional
analyses to determine the optimal camber and thickness. These tests were performed
at the Reynolds numbers of 2.9 x 10
5
and 1.3 x 10
6
as before. The worst value for each
variable parameter between the tested Reynolds numbers is recorded in Table 9.
Table 9: NACA 4 Digit Aerofoil Analysis
Profile C
lmax
(C
lmax
) C
m
(C
lmax
) (L/D)
max
C
Do
(C
Do
)
NACA 2412 1.07 12 -0.06 23.2 0.009 -2
NACA 4409 1.04 9 -0.11 25.1 0.009 -4
NACA 4412 1.22 12 -0.12 23.6 0.011 -4
NACA 4415 1.37 15 -0.13 21.9 0.012 -4
NACA 4418 1.5 15 -0.14 20.6 0.013 -4
NACA 6409 1.24 10 -0.16 22.6 0.014 -3
44
From this data it can be seen that the NACA 4412 is the best available aerofoil. A
smaller thickness such as in the NACA 4409 aerofoil will reduce the C
lmax
below 1.2
while a higher thickness will reduce the available (L/D)
max
value. Reducing the camber
reduces the maximum lift coefficient and the maximum L/D so a less cambered wing is
inferior in both key areas despite decreasing the moment coefficient generated. An
excessively cambered wing would result in poor cruise performance for an aircraft with
such a speed range as this aircraft and exceptional wing moment. As a result, the NACA
4412 aerofoil profile was chosen as the best option to meet the requirements outlined
above.
Empennage Aerofoil Selection
The tail aerofoils were selected to provide suitable symmetric performance in the
variable conditions at which they will be operating. The requirements for the
horizontal and vertical tails are similar, with a reasonable aerodynamic performance
and slow stalling key requirements.
For these aerofoils, as symmetry was required and performance requirements were
not particularly rigorous only the NACA 4 digit aerofoil series was considered as these
aerofoils are in common use on many aircraft tails due to their gradual stall
characteristics. The results of Javafoil testing of three reasonable thickness aerofoils
are presented in Table 10. This shows the worst value from the two tested Reynolds
numbers and the difference between the upper and lower values.
Table 10- Suitable Tail Aerofoils
Profile C
lmax
(C
lmax
) C
m
(C
lmax
) (L/D)
max
C
Do
(C
Do
)
NACA 0009 0.54 9 -0.01 17.1 0.007 0
NACA 0012 0.67 12 -0.01 17.5 0.009 0
NACA 0015 0.83 15 -0.02 16.2 0.010 0

This initial testing shows that the NACA 0012 is the best for this application. This
aerofoil provides the best aerodynamic efficiency and an angle of attack versus
coefficient of lift graph shows that the stall of this aerofoil is suitably gradual for use.
Section 3.5 Finalisation of Preliminary Aircraft Concept
45
The angle of attack of stall is around that of the wing and the delayed stall will enable
this tail to stall after the wing. The NACA 0012 has better aerodynamic efficiency and
gradual stall than other options and so has been selected. Although this testing is for
the aspect ratio of the horizontal tail, it is expected that the aerofoil properties would
not be too dissimilar for the vertical aerofoil and so the vertical tail will also be fitted
with a NACA 0012 aerofoil.
3.5 Finalisation of Preliminary Aircraft Concept
With the configuration and basic parameters determined, the aircraft concept was
able to be finalised.
3.5.1 Variation of Pulsejet Position in Concept Development
The main variable in sketches was the engine position, and various engine positions
were tried in a preliminary CAD model to work out a final concept considering
longitudinal stability, thrust line, practicality and overall plane length. This analysis can
be characterised into three phases with differing pulsejet positions, attempting to
solve the inherent challenges of pulsejet engine airframe design. The first phase
utilised an engine mounted external to the airframe, directly above the fuselage. The
second phase used a twin boom layout, enabling the engine to be mounted inline with
the airframe. The engine was mounted inside a fuselage cavity in the final phase of
development.

The initial designs focussed on a usable airframe, with pulsejet mounting considered a
less important aspect of design. As such, the pulsejet was initially mounted above the
fuselage, enabling sufficient air flow for cooling but with additional drag and an
uncertain mounting method. Eventually, when these designs were analysed the drag
provided by the airframe was exceptionally high due to the pulsejet mounting
location.. An example of an early design with engine mounted above the airframe is
shown below in Figure 24.
46

Figure 24: Example of Early Design
The inline engine phase was characterised by mounting the pulsejet inline with the
fuselage, reducing profile drag and moment produced by the engine thrust. Mounting
the engine inline with the fuselage allowed more efficient flight but the exceptional
length of the pulsejet engine lead to stability and size issues. An example of an aircraft
with this engine position is shown below in Figure 25.

Figure 25: Second Phase Design Example
The final phase of design combined the previous two by mounting the engine out of
the oncoming airflow in a fuselage cut out to reduce drag without compromising
stability. This phase developed from designs with engine mounted above a cut down
fuselage section in the second phase, by moving the engine forward so that it would be
almost over the centre of gravity of the aircraft.

However, this engine position was not without drawbacks. Heat generated by the
pulsejet was a critical issue and without cooling air passing over most of the pulsejet
surface the mountings and connections required additional heat resistance through
the use of coatings. The direct mounting of the engine to the fuselage meant that extra
vibration isolation was required in the design of the pulsejet mountings. Considering all
areas however, designs such as shown in Figure 26 were superior to those that came
before them and have been continued into detailed design.
47

Figure 26: Final Engine Position
3.5.2 Empennage Design
To determine the required horizontal and vertical tails, longitudinal and lateral stability
of the aircraft must be considered. This will enable the exact size of these tails to be
determined for a reasonable stability margin. Stability will be considered in more detail
once other details of the aircraft have been determined.
Longitudinal Stability
Longitudinal stability is based on the margin between aerodynamic centre and centre
of gravity. When the centre of gravity is ahead of the aerodynamic centre, an aircraft
which pitches upwards will have a natural moment downwards as the lift increases
due to an increased angle of attack, returning it to the original attitude. An unstable
aircraft would continue to pitch upwards until stall. However, in all aircraft and
particularly this aircraft with a large fuel fraction the centre of gravity will move
depending on the load configuration. As a result, the movement of the centre of
gravity must first be determined to find the critical most aft centre of gravity. The
aerodynamic centre of the aircraft can then be computed for varying horizontal tail
size to find the tail size for a defined stability margin.
48
Centre of Gravity Envelope
An envelope of potential centre of gravity positions was determined using a
spreadsheet computation. The centre of gravity is determined through Equation 5 by
taking into account the centre of gravity of all (n) components of the aircraft and their
relative masses.
Equation 5:
cg n
i
i
n
i
i i
x
M
M x
=
=
=
1
1

With no payload or crew for this UAV, four loading options were considered, based on
variable depletion of the fuel bags used. As the weight of various structural
components was still uncertain, initially statistical assumptions were made with the
use of equations. This method used empirical equations based on many aircraft
properties to estimate the weight of aircraft components. Once the components had
been manufactured, the measured weights and centres of gravity of these were used.
A component weight breakdown is presented in Appendix G Component Weight
Breakdown A together with applied lever arms from the front of the aircraft used to
determine the position of the centre of gravity. When the above variable configuration
options are considered the centre of gravity excursion diagram presented in Figure 27
can be determined. This diagram shows how the location of the centre of gravity varies
with the loading of the aircraft as well as the locations of the wing and whole aircraft
stick fixed aerodynamic centres. The aerodynamic centre of the aircraft (also known as
the neutral point) was determined from Equation 6 (Raymer 1992) and shows the
stability margin which is determined by later tail sizing.
49
Equation 6:
w
pa
h
h l
w
h
h l
p
p
w
p
ach
h
h l
w
h
h fus m acw l
np
qS
F
C
S
S
C
X
qS
F
X C
S
S
C X C
X
+
+
=

Figure 27: Centre of Gravity Excursion Diagram
Horizontal Tail Sizing
The horizontal tail size required was determined through use of an X-Plot which is a
graph of neutral point and centre of gravity versus horizontal tail size. A required
minimum stability margin of approximately ten percent of the mean aerodynamic
chord is required for an aircraft with inherent static stability (Roskam 1985b). As at this
stage the location of the stick-free aerodynamic centre was unknown, a slightly greater
margin was used to allow for the stick-free point being further forward than the stick-
fixed position calculated. The most aft centre of gravity (with no tail) was determined
from the above excursion diagram. The neutral point of the aircraft was determined by
use of Equation 6 (Raymer 1992) for each tail size tested. The X-Plot determined could
then be used to determine the required horizontal tail size by applying the required
stability margin as shown in Figure 27. During iterations of this procedure, it was found
50
that the downwash from the wings was having a detrimental effect on the
performance of the tail. This was then corrected by utilising the modified T-tail layout
outlined earlier. Sweep was incorporated on both the vertical and horizontal tails to
maximise the moment arm between the wing and the tail by moving the aerodynamic
centre of the tail rearward. This allowed a significantly smaller horizontal tail area and
hence a more efficient aircraft. The figure shown below is for the final centre of gravity
position, showing a greater stability margin than required as the theoretical centre of
gravity was significantly further backwards than the This method gave a horizontal tail
size of 0.145m
2
for the aircraft.
Longitudinal X-plot
0.0%
20.0%
40.0%
60.0%
80.0%
100.0%
120.0%
0 0.05 0.1 0.15 0.2 0.25 0.3
S(m
2
)
L
o
c
a
t
i
o
n

(
%

M
A
C
)
Centre of Gravity
Static Margin
Neutral Point

Figure 28: Longitudinal X-Plot
Lateral Stability
The required vertical tail size was determined through use of a directional X-Plot of
vertical tail size versus lateral stability margin. To enable this sizing to be completed,
the coefficient of directional stability (C
n
) for the aircraft was required. This was
influenced by the required stability margin and three area specific coefficients C
nf
, C
ni

and C
np
which were considered separately. C
np
is only relevant to aircraft with
propeller engines and so was zero for this design. For a low winged aircraft, C
ni
is
0.024 (Torenbeek, 1982) while C
nf
is dependent on fuselage geometry and can be
51
calculated from Equation 7 (Torenbeek 1982) (correcting for specific geometry using
Equation 8).
Equation 7
3
1
1
2
2
1
=
f
f
f
f f fs
f n
b
b
h
h
Sb
l S
C

Equation 8
105 . 0 75 . 0 3 . 0
max
+ =
f
f
f
cg
l
h
l
l

Equation 9
2
) / (
) (
V V C
C C C C
Sb
l S
v yv
i n p n f n n
v v

+ +
=
Equation 9 (Torenbeek 1982) can then be used to determine the stability margin for a
given vertical tail size. This can be repeated for varying vertical tail sizes to determine
the required size for a certain stability margin. The stability margin required can be
estimated via consultation with various sources, Roskam (1985b) recommends 0.001
deg
-1
while Torenbeek (1982) recommends a range from 0.04 to 0.1 rad
-1
so a required
margin of 0.06 rad
-1
has been chosen to fit within these values. The lateral X-plot can
then be drawn as in Figure 29 and used to determine required vertical tail area. From
this a vertical tail size of 0.05 m
2
was determined. However, as the vertical tail size also
contributes to the effectiveness of the horizontal tail by raising it upwards from the
downwash of the main wing the vertical tail size was somewhat larger than required to
achieve the stability margin set. The uncertain effects of locating the pulsejet between
the vertical tails altering the flow profile also required a greater margin of error. As a
result, the combined area of both vertical tails was set at 0.11m
2
.
52
Lateral X-Plot
-0.1
-0.05
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0 0.02 0.04 0.06 0.08 0.1 0.12
S
vertical tail
(m
2
)
S
t
a
b
i
l
i
t
y

m
a
r
g
i
n

(
r
a
d
-
1
)

Figure 29: Lateral Stability X-Plot
3.6 Finalization of Preliminary Aircraft Concept
The engine data development (such as selection of Chinese engine and engine length)
was now able to be incorporated into the aircraft concept.
3.6.1 Preliminary Conceptual Fuselage Design
With the main conceptual parameters determined such as engine length and based on
some of the sketches developed earlier, a conceptual design for the fuselage was able
to be produced.

The overall shape of the fuselage is roughly cylindrical, with a cut out section for the
pulsejet and an aerofoil shaped nose. This circular shape enables a simpler
manufacturing process for both the fuselage and bulkheads and easier, more accurate
structural analysis. This circular shape also allows the use of a rotated aerofoil cone as
the nose of the fuselage, providing minimal drag and hence better aircraft
performance.

Section 3.7 Practical Modifications to Final Concept
53
The conceptual fuselage diameter selection is a balance between providing minimal
centre of gravity movement and minimizing drag. A smaller diameter fuselage would
reduce the cross sectional area and as a result, the drag of the fuselage. The Chinese
pulsejet used required a cavity of 80mm to be left in the top of the fuselage to provide
space and clearance for the engines exhaust (diameter 74mm) without allowing too
much of the pulsejet to protrude from the fuselage. This protrusion would significantly
increase the drag of the aircraft as the front of the engine is simply a cylinder with no
streamlining. The final design can be seen in Figure 30.

Figure 30 -Conceptual Fuselage Design
3.7 Practical Modifications to Final Concept

Figure 31: Modifications to final aircraft concept
With all the conceptual parameters of the aircraft the first concept showing all the
elements of the design is shown in Figure 31 (a). However some modifications had to
be made to this concept and this is shown in Figure 31 (b). A fillet has been added
around the root of the wing to reduce interference drag and to reduce the number of
(a) (b)
54
sharp edges needed to be cut using CNC milling. The linearly tapered engine cut-away
has been changed to a flat engine cutaway as the tapered cutaway would be difficult
and expensive to manufacture due to complex moulds being required. The tail has
been changed to a modified T-tail design to, this allows for elevated positioning of the
horizontal tail to limit impacts of main wing downwash on the tail. Sweep has been
incorporated to give the tail later stall characteristics than the wing, while the swept
vertical allows for further rearward position of the horizontal tail to improve
longitudinal stability. Finally, the diameter of the fuselage has been increased slightly
from 150 mm to 160 mm.
3.8 Engine Design
Initial research performed in the feasibility study section suggested the use of an
expanding tail section would improve engine performance. The following section
investigates the design aspects related to the expansion design. The similarity between
pulsejet expansions and the exhaust systems used on two-stroke internal combustion
engines was noted and used as a partial basis of the expansion theory. This theory was
combined with steady state diffuser theory, fluid dynamics software and statistical
research.

The effect of intake and exhaust flares was investigated based on the research
conducted by Coombes et al in 2007, flares for both the intake and exhaust were
developed for subsequent testing based on a literature review and 1-D fluid dynamics
software.
3.8.1 Exhaust Design Two Stroke Exhaust Similarities
The evolution of pulsejet engine exhausts appears to carry some similarities to the
evolution of 2-stroke engine exhaust design. The use of this observation as a guide
into the workings of the advanced exhaust designs may provide a means of
understanding the complex behaviour that is present in the systems, as well as provide
the possibility to produce an exhaust that dramatically improves the performance of
the current engine.
Section 3.8 Engine Design
55
2-Stroke exhaust design evolution:
Two stroke engines typically have one of three exhaust designs (Figure 32). Early
engines used straight pipes, which evolved into megaphone expansions, and finally
the expansion chamber style pipes that are seen currently. (Jennings, 1987)

Figure 32- Advancements in Two Stroke Exhaust Design
The Initial straight pipe exhaust systems on 2-stroke engines could be tuned by simply
changing the length of the pipe based on the desired speed range. (Jennings, 1987) Thi
allowed for significant performance gains.

It was then discovered that a Megaphone produced a stronger low pressure region
behind the pressure wave. This helps the exhaust extract all the burnt gasses from the
combustion chamber, allowing more fresh air and fuel into the cylinder. It proved
possible to build an expansion that worked so well that it would draw unburnt fuel into
the start of the exhaust pipe. (Jennings, 1987)

The addition of a convergent section to the end of the expansion results in a pressure
wave being sent back towards the exhaust port. It is found that if this wave was timed
correctly, raw mixture drawn into the start of the exhaust could be pushed back into
the combustion chamber just prior to the port closing. This effectively supercharges
56
the engine, forcing more air into the combustion chamber than it could normally hold.
The extent of this effect is very significant, often producing volumetric efficiencies of
around 140% on modern engines. (Jennings, 1987)
Effects of chamber characteristic lengths:
The diffuser angle is responsible for the intensity and duration of the returning low
pressure wave. A higher angle results in a stronger wave with a shorter duration. This
relationship is seen in Figure 33.

Figure 33 The effect of expansion angle on wave behaviour.
(Jennings, 1987)
On a 2-stroke engine, it is claimed that an 8 degree included angle provides the best
energy recovery. The use of multi-cone sections is common, as it reduces the angle
required between cone steps, and hence allows greater angles before flow separation
occurs. This optimal angle is a compromise between peak power and the power range,
and hence the optimal angle for a pulsejet, which has a constant operating frequency,
will possibly be larger than commonly found on 2-stroke engines. (Jennings, 1987)

As with the diffuser angle, the convergent cone angle determines the intensity and
duration of the returning wave. This angle is typically about twice the diffuser angle.
This angle tends to effect the tuning location of an engine dramatically, with subtle
57
changes in location producing noticeable translation in the tuned frequency range. The
small exit pipe is the main dimension that determines the pressure inside the exhaust
chamber. A higher pressure in the chamber results in stronger pressure waves, and as
such can be used to improve the performance of the engine. (Jennings, 1987)

A study using 2-stroke exhaust theory with valved pulsejet engines was conducted in
1983 by Arrt and Blair through Belfast University. The experiment showed that the
megaphone design had the effect of improving the specific fuel consumption of the
engine by 44%, whereas the expansion chamber style engines increased the engine
fuel consumption and reduced thrust.
3.8.2 Steady State Diffuser Design
Background research suggested that the use of an expanding exhaust section results in
an increase thrust output from the engine, for a number of reasons. Firstly, the
expanded section results in a more efficient use of available energy within the engine.
A valveless pulsejet engine does not lack energy, as can be deduced by the high
temperature exhaust gasses that exit the engine (Coombes et al, 2007). Instead, the
major challenge for a valveless pulsejet engine is being able to convert the available
heat energy into momentum and thrust (Equation 10 and Equation 11)
Equation 10
2
2
1
mv KE =
Equation 11
mv p =

The precise sizing of the design of an expanding section is, as with most pulsejet
related design, not documented. However, approximate sizings can be estimated from
a range of different sources including statistical reviews, information provided by
enthusiasts and some fluid dynamics theory.
58

The design of the expanding tail pipe is effectively a conical diffuser during the
combustion cycle and a contractor during the intake cycle. It is therefore expected to
behave as such, which will make the expansion angle critical. Figure 34 below shows
results for the behaviour of a typical diffuser. Due to the complex nature of flow
through a diffuser, the figure can only be use as a guideline, with Individual flows
varying with characteristics of the design such as area ratio and Reynolds number
(Munson, Young, Okiishi, 2006).

Figure 34: Loss Coefficient for a Conical Diffuser
(Munson, Young, Okiishi, 2006)
Figure 34 shows that for small expansion angles, the diffuser is excessively long, with
large losses a result of wall shear stress. For moderate to large angles, losses are a
result of flow separation, which effectively reduces the diameter of the expanded
section. Commonly, an optimum diffusion angle is about 8 with angles similar to this
(+/- 1) used widely in pulsejets, including the Messerschmidt pulse-ram jet, the
Escopette and the Laird Chinese. Half angles of 4 or 5 are very common, with
optimum performance angles for individual engines generally requiring extensive
testing.

Figure 34 shows that it is very difficult to efficiently decelerate a fluid. Acceleration of a
fluid in a contractor on the other hand is more efficient and less complicated, with loss
59
coefficients only ranging from 0.02 at 30 to 0.07 at 60 (Munson, Young, Okiishi,
2006). Therefore, as the expanding section acts as both a diffuser and a contractor, it
was decided that the section should be designed to minimise losses for expansion,
which is during the combustion stage of the engines cycle.
3.8.2.1 Exhaust Analysis
In order to produce the optimal expansion section, a 1 dimensional analysis of the
engine with various characteristic lengths was performed using the UFLOW1D
software package. Uflow is a one dimensional compressible flow program for analysing
flow in pipes.

The basic geometry that was used for the investigation was based on the engine
developed by Coombes et al in 2007. These characteristics include the combustion
chamber, as well as the diameter of the middle sections. The lengths of the remaining
sections, as well as the expansion angle, were then investigated using the software.

The performance of particular engine geometries was gauged by the ability of the
engine to produce a large pressure swing in the combustion chamber. This criterion
was chosen as it was considered to be a key factor in determining the power the
engine can produce. The basic engine was therefore modelled using varying expansion
angle sections, in order to find the optimal divergence angle. The model used in the
process is shown in Figure 35. The results obtained from the investigation are shown in
Figure 36.

Figure 35 The UFLOW1D model used to investigate expansion angles.
60

Figure 36 Combustion chamber pressure extremes for different expansion angles.
The UFLOW1D analysis shows that the combustion chamber pressure swings are
significantly increased by the presence of the divergent section. Results suggest that
while the best cone angle is around 10 degrees, most of the expansion effect is seen by
around 6 degrees.

A statistical review of some similar engines is presented in Figure 37, which suggests
that the expansion angle should be around 8 degrees. The reasoning for the use of 8
degrees was not known, but was assumed to be as a result of the onset of flow
separation. The engine with the greatest expansion angle in the statistical data was,
however, the most efficient engine that was considered, which suggests that the use of
a 10 degree expansion is reasonable, and perhaps optimal in some circumstances.
61

Figure 37 - Statistical data showing exhaust expansion angles from similar engine designs
The design expansion angle was 10 degrees, chosen mainly due to the length
constraints on the exhaust pipe discussed in the expansion location section.
3.8.2.2 Expansion Diameter
From statistical research, it was found that the expansion typically expands to
approximatly the diameter of the combusition chamber, as shown in Figure 38. It was
decided to comply with the trend, as this would remove a variable that would be
diffucult to alter. Hence, the expansion diameter was chosen to be the same as the
combustion chamber.

Figure 38 - statistical data showing a trend between combustion chamber diameter and expansion
diameter.
62
3.8.2.3 Expansion location
To investigate the effect of moving the expansion closer towards the combustion
chamber, the UFLOW1D model was altered. It was found that the location of the
expantion along the exhaust length did not make an appreciable change to the
pressure swing in the combustion chamber. However, this finding does not
compliment the general trend of engines seen. A common trend seen in engines is that
the the exhaust sections are approximatly equal lengths, resulting in the expansion
being located in the middle of the exhaust section.

In order to achive the required expansion angle and diameter, it was required that the
expansion section be 285mm long. Using this length, and assuming equal length
sections, the complete exhaust would be 855mm long. This was seen as a potential
problem, as the optimal length of the exhaust with a straight pipe was found to be
approximately 600mm, significantly shorter than the new design. The length of the
pipe is also a dominant reason for the use of a 10 degree expansion angle, as it
reduced the length of the expansion section. It was decided that, in order to shorten
the length of the pipe, the end section should be shortened.

As a reuslt of this, the expansion was designed with a 150mm extension at its end,
rather than the 285mm extension that was originally intended.
3.8.2.4 Adjustability:
In order allow for investigation into the behaviour of the engine, it was required that
the exhaust section be made adjustable, in a similar fashion to the 2007 development.
In order to aid manufacture, and to make testing of the full range of lengths possible, it
was decided that only one characteristic length would be adjustable. The most suitable
length to adjust was the centre section, as it will modify the length of the engine easily,
and sections of centre pipe were already available. The shortest exhaust length
sections from straight pipe testing can be used for this, however some shorter lengths
still required to be manufactured. The final design is shown in Figure 39.
Section 3.9 FWE Bellmouth Development
63

Figure 39 - The final expansion design
3.9 FWE Bellmouth Development
Based on the research conducted by Coombes et al, it was noted that the omission of
flares from the engine design could possibly effect the ability of the engine to achieve
a sustained thrust. To determine the effect of the flares on engine performance,
suitable intake and exhaust flares were developed for a FWE engine.

It was proposed that the engine did not sustain due to the lack of an intake bellmouth,
resulting in a reduced ability for the engine to draw air into the combustion chamber
(Tao 2007). Based on the bellmouth research performed in 2007, it was also proposed
that a flare on the exhaust of the engine would strengthen the intensity of the
returning waves, and hence improve the chances of reaching sustained operation. This
hypothesis can be confirmed through research conducted by Tao in 2007 into the
effect of starting vortices on unsteady combustion devices.
3.9.1 Starting Vortices
Starting vortices have been shown to be a significant contributor into the operation of
pulsejet engines, due to their unsteady operational characteristics. Studies of vortex
ring formation date back to 1900s and this phenomenon has been extensively
investigated theoretically, numerically, and experimentally. The link between to
pulsejet engines was made by Gharib et al. (1998), and was shown to improve low
pressure swing, allowing the engine to draw more fuel and air during the intake cycle.
64
The vortex structure, captured using particle image velocimetry (PIV), shows the
interaction of vortices during the interchange between intake and exhaust cycles
(Kailasanathan 2007).

Figure 40 - PIV images of vortex interaction
3.9.2 Bellmouth Design
Bellmouth shape
In order to optimize the effect of the bellmouth on the engine, the shape the design
was considered. A shape that produced the good results and was relatively simple to
manufacture was desired for the project. Due to the future use of the engine on a
UAV, the overall size of the bellmouth was also considered, as increased diameter
would result in a larger drag on the engine, reversing any positive effects seen by
addition.
The article Best bell (Blair, G and Cahoon, M 2006) was used as a basis of the
investigation of the performance of various bellmouth designs. Figure 41 shows the
bellmouth designs that were considered for the second 2007 engine.
65

Figure 41 - Bellmouth designs considered (Blair, Cahoon 2006)
The performance of the bellmouth sections were laboratory tested in comparison to a
standard radius bellmouth, producing the data in Figure 42. It is estimated that the
minimum pressure in the combustion chamber will be around 0.7 bar, resulting in a
pressure ratio of up to 1.4 at the intake during the intake stage of operation.

Figure 42 - Performance of bellmouth designs
The article shows that in low pressure ratio conditions, the best results are seen with
the elliptical and aerofoil shaped bellmouths. However, the improvement seen by the
66
use of a more complex shape is minimal, in the realm of 2% for the pressure ratio
expected for the engine. In the interests of ease of manufacturing, it was decided that
a simple radius bellmouth would be best suited to the intake of the engine.
Radius
The 2007 project group conducted an investigation into the effect of bellmouth radius
on the intensity of a returning pressure wave. The data was obtained using the one
dimensional fluid dynamics software UFLOW1D. The results from the investigation are
shown in Figure 43. These results obtained were confirmed with the statement the
minimum radius of the bellmouth lip is specified as being best between 0.25 and 0.3
times the diameter of the tube. (R. Alan Wallis, 1983).

Figure 43 - The data obtained in 2007 using UFLOW1D (blue) and textbook recommendations (red)
Adjustability
Engine 2 was designed by the 2007 project group using scaling of a significantly smaller
engine. As a result the engines state of tune is questionable, and the ability to adjust
the intake and exhaust lengths is desirable. Due to this, it was decided that an
adjustable bell mouth design would be beneficial to the engine development. It was
also noted that a continuously variably pipe length would provide better resolution
67
data than discrete length sections, and hence allow better understanding of the engine
behaviour.
Manufacturing
The ease of manufacturing of the bellmouth was considered in order to ensure that
the final product was at the desired standard, and could be produced at a reasonable
cost. Initially, the bellmouth was only be used for testing purposes, and hence was not
required to be optimised for flight purposes.
Discussion with workshop staff resulted in the conclusion that manufacturing a
stainless steel bellmouth to the optimal radius would be very difficult, as the metal
would not allow the amount of deformation required to be deformed into the desired
shape. As such, it was decided that a machined curve would be much easier to
produce, and would produce the best solution.
3.9.3 Final Design
The final design was manufactured from a combination of pipe and machined billet.
The machined curve was welded to the pipe section, which was sized to slide over
engine pipe. This solution was particularly easy to manufacture, and provided a
smooth curve to improve air flow. The model can be seen in Figure 44.

The greatest concern with this solution was the presence of a lip where the engine
pipe ends, when the flow transfers to the inner surface of the bellmouth. While this
situation was not ideal for flow, it was not of a major concern at the time of
manufacture. This was justified by the presence of steps and uneven joints in the
engine, due to the presence of welds and metal irregularities. It was decided that if the
inner lip of the pipe were ground to smooth the lip, the transition would produce less
interference to the flow than the joints that existed in the pipe, and hence it was not a
significant concern for the performance of the engine.
68

Figure 44 - The adjustable bellmouth design.
3.9.4 Flight considerations
While the bellmouth manufactured for the second engine was not designed to be
optimised for a flight condition, future engines will require consideration in this area.
Concerns have been identified regarding the bellmouth and its impact on the plane
aerodynamics. The greatest concern is that the frontal area of the bellmouth will
create a significant amount of drag, which will not only reduce the performance of the
UAV, but will create a low pressure region in the area that the intake draws from,
potentially limiting the volume of air that the engine can intake. This would be
detrimental to the performance of the engine, as thrust would dramatically reduce as
the UAV velocity increases, limiting the aircrafts top speed.

In order to improve engine performance at high air speeds, a more streamlined
bellmouth was designed, such that drag is reduced, and the pressure in the intake
maximised.

To perform this analysis a Axi-Symmetric model of the intake was created using
computational fluid dynamics (CFD) programs Gambit and Fluent. Three geometries
were modelled as shown in Figure 45.
69

Figure 45- Three intake geometries
The domain was setup with flow from left to right, over the intake, with a low pressure
zone in the intake to simulate the intake cycle of the engine (Figure 46). Compressible
flow was modelled as flow speeds reached over Mach 0.4 inside the intake. The height
of the domain was set to 20 times the radius of the intake, whilst the length was 30
times longer than the intake pipe, to ensure the flow had stabilised before reaching
the far field pressure outlet.

Figure 46- Domain Layout
Results were obtained at three different flow speeds, 10m/s, 40m/s and 80m/s. The
results confirmed that the addition of a flare increases the mass flow rate into the
70
engine, however little difference in mass flow was shown wit the addition of a flare.
The results can be seen in Figure 47.
Mass Flow Rate Comparisons
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0 10 20 30 40 50 60 70 80 90
External Flow Velocity (m/s)
M
a
s
s

F
l
o
w

R
a
t
e

a
t

i
n
l
e
t

(
k
g
/
s
)
No Flare
Standard flare
Aerodynamic Flare

Figure 47- Effect of intake geometry on mass flow rate
The major gain found from the aerodynamic flare was a reduction in the drag
coefficient by up to 50% at 80m/s. A comparison of the pressure field plots for the
standard and aerodynamic flare at 80m/s can be seen in Figure 48 and Figure 49.

Figure 48 -Static Pressure Contours of Aerodynamic Flare at 80m/s
Section 3.10 Flight Engine Development
71

Figure 49-Static Pressure Contours on Standard Flare at 80m/s
3.10 Flight Engine Development
From the analysis conducted in Section 2.6.3 it was determined that the Chinese
Pulsejet Engine was the most suitable engine to develop for flight, most notably due to
its documented success on liquid fuels. An extensive search produced numerous
models developed with thrust outputs ranging from 2.6lbs to 50lbs. Overall, seven fully
dimensioned engine models were discovered. This allowed a simple statistical analysis
to be performed in order to determine the approximate size required for a 3kg engine.

The statistical analysis produced extremely accurate results with R
2
values no lower
than 0.97. This allowed trend lines to be used to predict all dimensions of the engine.
This also provided information on the critical dimensions that must be considered
when attempting to scale a pulsejet.

72
Comparison of Thrust vs Engine Length
0
10
20
30
40
50
60
0 20 40 60 80 100
Length (inches)
T
h
r
u
s
t

(
L
b
s
)
Chinese Engine
FWE

Figure 50- Statistical trend of Chinese and FWE engines
From the statistical analysis it was determined that an engine of similar length, but
significantly smaller combustion chamber size, when compared to the current FWE
design could be produced to achieve the required 3kg. This can be seen from the data
collected in Figure 50.

From contact with Irvine Aeropulse, a small Brisbane based company, it was
determined that the exhaust section of the statistical Chinese model could be better
optimized to reduce engine length and improve performance. As mentioned earlier,
the expansion for the exhaust can be optimized at around 5 degrees. The statistical
engine had an expansion section with half angle of approximately 3.4 degrees,
therefore by expanding this angle and maintaining the same area of the sections, in
accordance with studies by Mcalley (2006). the final engine design reduced the length
by 200mm and the phone diameter increased by 10mm, giving a final engine length of
850mm.
Material Selection and Stress Analysis
To ensure the weight of the engine was kept to a minimum a finite element analysis
(FEA) was conducted. Varying pressure and temperature loads were applied to a full
engine model. Based on the results a separate axi-symmetric model of the combustion
73
chamber end cap was created, as this was found to be the area of highest stress. The
material chosen was 310 Stainless Steel, as it was the only stainless steel found which
was capable of withstanding cyclic loading at 1150 degrees. Figure 51 shows the
material properties of 310 stainless steel at high temperatures.

310 and 310s Material Properties
0
100
200
300
400
500
600
700
0 200 400 600 800 1000 1200
Temperature (C)
S
t
r
e
n
g
t
h

(
M
P
a
)
310 Yield
310 Tensile
310s Yield
310s Tensile

Figure 51-Variation of material properties of 310 stainless steel with temperature
The analysis required coupled thermal and pressure loads, with load data based on the
performance of the SNECMA Escopette (SNECMA 1951). From
Figure 52 it can be seen that the pressure waves fluctuate in a sinusoidal manner
between 50kPa to 150kPa (absolute) during each cycle. It can also be seen in
Figure 52 that the pressure waves are constant in magnitude along the length of the
engine. For this modelling case the engines operational frequency was assumed to be
200Hz.
74

Figure 52-Operating pressure of the Escopette pulsejet
The model was created in pro engineer, and imported into ANSYS workbench where it
was thinned and simplified for modelling. The analysis was conducted using shell
elements with the aim of reducing the computational time of the coupled solver. A
wall thickness of 0.5mm was initially investigated, as this was found to be a common
value used in hobby pulsejet manufacture.

The meshing of the engine was created using element sizing and manually setting the
divisions along the length. This ensures the mesh remains aligned and reduces element
deformation. Elemental deformation leads to errors in both the temperature and
stress gradients, causing inaccuracies in the model. The final mesh of the model is
shown in Figure 53.

75

Figure 53- full engine mesh

The thermal and pressure loads were input into the flexible dynamic solver using
Equation 12. The solver used, completes the thermal analysis first and determines
thermal stresses on the model, before the pressure model is solved.
Equation 12
time * 180/pi) * 200 * pi * sin((2 * 50000 P = )

Figure 54-pressure loading input for flexible dynamic solver
The results of the full body model showed that the highest stresses existed around the
combustion chamber end cap, as shown in Figure 54. Based on this analysis an axi-
symmetric model of the combustion chamber end cap was created to determine the
strength of the cap with the inclusion of fillet welds.
76

Figure 55-Stress Results on Combustion Chamber End Cap
The axi-symmetric model showed high stress regions around the base of the welds.
The stress of 109MPa represents the yield stress of the material at 950 degress, which
was determined to the operational temperature of the combustion chamber.
Therefore the thickness of this section was increased to 1mm for manufacture, and the
thickness of the TIG welds was also increased.

Figure 56- Final Results of the Axi Symmetric Model
3.11 Liquid Fuel System Design
This section contains comparisons between a number of potential liquid fuels and the
fuel characteristics that were identified as important for use in a valveless pulsejet
engine.

Section 3.11 Liquid Fuel System Design
77
3.11.1 Fuel Choice
The choice of fuel was based on several key chemical properties. These include the
energy density, flash point, latent heat of vaporisation and flammability limits. Based
on these properties it was determined that, petrol, kerosene and methanol were the
most suitable fuels for testing. A detailed analysis of the fuels that were considered
can be found in Appendix F Liquid Fuels.

The chemical properties of petrol suggest that it was the most suitable fuel to use to
power a valveless pulsejet engine. The high energy density means a relatively low fuel
flow rate is required to produce a given thrust, thereby reducing the mass of fuel that
would need to be carried onboard the aircraft. The low flash point of petrol (-46C)
means that it will ignite easily which should allow the engine to be started relatively
easily. The main disadvantage of petrol is its small flammability range (1.3% 6%),
potentially limiting the throttleable range of the engine. Petrol also has a low latent
heat of vaporisation, which will cause the engine to run hot. Too much heat may
eventually affect the structural integrity of the engine, especially under static
operating conditions.

A fuel such as methanols main advantages over a fuel such as petrol is its latent heat
of evaporation and larger flammability range. As methanol expands, it absorbs a large
amount of heat from its surroundings, resulting in an overall cooler burn temperature.
The air being mixed with the methanol will therefore also be cooled, producing a
higher density mixture in the combustion chamber. Therefore, a larger mass is able to
be ejected from the chamber, thereby producing more thrust compared to lower
density, higher temperature combustion mixes. The large flammability range would
likely help to increase the throttle range of the engine. The trade-off for the cooler
burn and large flammability range is a much lower energy density, being approximately
half that of petrol. Therefore an engine running on methanol will require
approximately twice the fuel compared to an engine running on petrol, to produce the
same thrust. As a result, the weight of fuel needing to be carried to produce the same
amount of thrust for the same time period will also increase by around a factor of 2.
78
For flight purposes, the fuel mass required will likely be the limiting factor for
methanol.
Outcome
The successful operation of a valveless pulsejet engine is highly based on acoustic
resonance within the engine. Therefore, the combustion temperature of a given fuel
will impact on the acoustic properties and hence the performance of the engine. To
test the effect of the fuel properties on the engine a range of different fuel mixes were
tested. From the information gathered, petrol will be used as the main fuel, due to its
chemical properties and because it can be easily obtained. Petrol-methanol mixes
were also tested, with as based on the fuel properties it reduced the fuel consumption
of the engine. Fuels such as shellite (a.k.a naphtha) can also be mixed with petrol as
based on its chemical properties it increases the frequency of the engine by raising the
combustion temperature.

3.12 Fuel Injector Design
As stated in Section 2.6.5 studies were conducted by Mcalley in 2006 into the
development of a liquid fuel system for pulsejet engines, with much work conducted
on valveless engines. While his testing showed limited success, mostly due to poor
injector placement, he developed numerous injector styles which showed produced
useful results. The initial injector studies revolved around the design of multi-holed pin
injectors, created from 1/8 stainless steel. Flow visualisation and engine tests were
performed on varying size holes and hole positions. The initial tests produced 6 and 12
hole injectors with 1.7mm holes. Flow visulisation of the injectors can be seen in Figure
57 and Figure 58. It can be seen that the flow pattern from the 12 hole swirl injector
aims to promote better mixing in the engine, however it was found that this injector
design was unable to sustain thrust within the engine. The 6 hole opposed spray
injector was successful in sustaining thrust, but only with forced air. It can be seen in
the flow visualisation that flow is only exiting through the bottom injector. This
Section 3.12 Fuel Injector Design
79
concluded that the injector hole size was too large, as the pressure drop was not great
enough to force fuel out of the other injector holes.

Figure 57- 12 hole swirl injector

Figure 58- 6 hole opposed spray injector
Continued development with decreasing hole sizes progressively showed that a single
or double hole injector, with hole size 0.51mm was the most effective injector design,
as they provided high pressure injection, which promoted better air/fuel mixing.

After looking at the work undertaken by Mcalley, we undertook an investigation into
the performance of a simple pin hole injector, to allow estimates to be made regarding
required injector sizing and performance. Initial calculations based of work from
Williams (1990), suggested that an orifice diameter of 0.1mm would be large enough
to supply 300ml/min of petrol for a supply pressure of 3bar. We were unable to test
these calculations as such small orifice sizes were not able to be manufactured. Also,
80
the calculations did not account for the dynamic behaviour of the engine, or the high
operating temperatures. The high temperatures during operation would reduce the
diameter of the orifice, causing a significant reduction in flow rate.

The operating characteristics of swirl injectors were also analysed, but in depth
numerical calculations were not undertaken. Swirl injectors produce greater spray
angles than pin hole injectors, due to the added tangential component of velocity.
However, the performance of these injectors varies dramatically depending on the
internal structure. Therefore, as we were not designing our own injectors, this exercise
was done to obtain knowledge regarding general injector performance, to allow
decisions to be made during engine testing, regarding the performance of the system.

Following this, a review of commercially available injectors was conducted. It was
found that very few injectors met the specific requirements of both size and fuel flow
rate, whilst being able to handle the required temperatures. Two main injectors were
found, the first manufactured by BETE, who are a custom nozzle manufacture for a
wide range of industries in the United States.

The BETE PJ series offered flow rates between 0.043 to 5.34 L/min from a supply
pressure of between 3 and 10Bar. The injector offered was a cone spray type injector,
as shown in Figure 59.

Figure 59-BETE PJ Cone Spray Injector
Cone spray injectors produce highly atomised fuel spray, in a 90 degree arc. This
increases mixing compared with a single pinhole injector, producing droplet sizes of
Section 3.12 Fuel Injector Design
81
under 50 microns (BETE 2008). From the data provided a maximum flow rate of
0.34l/min could be achieved at 5 Bar through a 0.51mm orifice. However despite this
the size of the injector was still larger than desired, as the intake of the pulsejet only
has a diameter of 29mm.

The other injector which was found was a similar design spray injector available
through a local Australian dealer. The designs were constructed from 5mm stainless
steel tube, with the injector nozzle fitting inside the tube diameter, as seen in Figure
60. The manufacturer also quoted droplet sizes of 25 microns.

Figure 60-5mm stainless steel injectors

From the data available it was seen that a seen that a single injector could only supply
a maximum of 170m/min though a 1mm orifice. However due to the size of the
injectors it was determined that 3 injectors could be placed down the length of the
intake to achieve the required fuel flow rates. Due to the availability, small droplet size
and significantly smaller physical size, these injectors were purchased.

Six injectors were purchased as the fuel flow rate was unknown. Two 1mm, two
0.8mm and two 0.6mm injectors as the flow rate was expected to be between 200 and
350ml/min. The different sizes would ensure that once the required fuel flow rate was
82
found, that the optimum injector configuration could be determined. A table of fuel
flow rates for each size injector can be found in Table 11.

Table 11
Orifice size
Fuel Flow rate
(5 bar)
0.6mm 0.101
0.8mm 0.138
1mm 0.178

Section 3.13 Conceptual Design Summary
83
3.13 Conceptual Design Summary
From the previous analyses, a three viewed drawing of the aircraft can be developed.
This is shown below in Figure 61.

Figure 61: Conceptual Design Three View

Chapter 0
84
.

85
4 Detailed Design
4.1 Fuselage Structure Design
In this section, the detail structural layout and analysis of the fuselage is described.
Under structural layout, the detailed components of the fuselage are described, while
under structural analysis the fibreglass skin and longerons used are checked for
strength.
4.1.1 Fuselage Structural Layout
The detailed layout of the fuselage is illustrated above in Figure 62. It has an external
fibreglass skin and internal support structure, comprising of bulkheads and longerons.
A removable heat shield is located between the fuselage and the engine however this
is not a structural component.

Figure 62: Overview of Fuselage Structural Layout
Fuselage Skin
For the main surface of the fuselage, fibreglass material was selected. It provides a
light weight yet stiff structure and is commonly used in similar UAVs.

The basis of the skin structure was 320 gsm, 0/90 aircraft grade glass cloth, with a
50:50 volume resin mix. Three layers of the cloth are used, as determined in section
Fibreglass skin
Internal reinforcing
Removable heat
protection
section
Landing gear notch
Chapter 4 Detailed Design
86
4.1.2. The resin selected for use with the fibreglass was a high temperature version, in
order to resist heat from the engine. The chosen resin is a vinyl ester type with a 177
C glass transition temperature.

Radio waves can propagate through fibreglass and thus the antenna for the radio
controller is able to be placed internally inside the fuselage. This is not possible with a
carbon fibre construction without a high risk of the signal being broken.
Removable aluminium section
The removable aluminium section is of sufficient size that when removed, provides
access to the majority of the fuselage. This is required to provide access to the fuel
storage. The aluminium section also functions as a heat shield, reducing heat transfer
to the fuselage.

Alternatively, smaller access holes could have been placed along the bottom of the
fuselage. This would however make manufacture more difficult, provide inferior access
to the area and not provide heat shielding.
Internal Reinforcing Structure
As the design for an aluminium section provides a discontinuity in the fibreglass
structure, reinforcement was designed to strengthen this area. Figure 63 shows how
the bulkheads and longerons will be used to reinforced the fibreglass skin.

Figure 63: Fuselage Internal Reinforcing Structure
Bulkhead
Longeron
Neutral
Moment of Inertia
Contribution of Longerons
Section 4.1 Fuselage Structure Design
87
The longerons have been added to improve the bending strength of the fuselage.
Unidirectional carbon fibre tubing is used, as the fibers are orientated along the length
of the tube, resulting in a very stiff longeron-bulkhead structure, and hence providing
strong resistance to bending. Structural analysis of the longerons is conducted in the
Fuselage Structural Analysis section.

Three longerons are to be used, distributed at the top corners and lowest point of the
rear bulkheads. The bulkheads perform the function of connecting the longerons to
the fuselage skin. Bulkheads also providing stiffening to help the fuselage hold its
cross sectional shape and prevent bulking. The specific structural contribution of the
bulkheads has not been analysed, however it is expected they will significantly stiffen
the fuselage.

Six bulkheads are to be used, including five main bulkheads and one nose bulkhead.
The three rear bulkheads have a half-circular shape. The rearmost bulkhead assists
with the tail fuselage join, while the remaining two provide strength around the wing-
fuselage interface. The bulkheads also provide segmentation between the fuel bags.
Two full circle bulkheads sit in the back half the nose section. The main engine mount
attaches to the bulkhead at the rear, while the forward bulkhead provides attachment
for the front landing gear and electronics tray. The front most bulkhead in the aircraft
sits toward the front of the nose and has a smaller radius, providing the second
attachment point for the electronics tray.
Landing gear notch
A landing gear notch has been incorporated into the fuselage shape at the required
location of the main landing gear. This is required as the central part of the main
landing gear is flat and a suitable contact surface is necessary. A landing gear tray was
avoided, as the extra weight and complexity was considered unnecessary.
88
4.1.2 Fuselage Structure Selection
To determine required specification of the fuselage skin and longerons, a stress
analysis was performed. Component weights and estimated aerodynamic forces were
used to estimate the force distribution on the fuselage and thus the stresses were
determined. The details of the analysis can be found iAppendix I Fuselage Stress
Analysis. Table 12, shows the maximum stresses found at the most critical locations in
the fuselage.
Table 12: Fuselage Stress Analysis Results

For shear, the fuselage has been analysed for one, two or three layers of fibreglass on
the basis of .25 mm per ply. While for bending, longerons of 7mm outer diameter and
5.5 mm inner diameter have been assumed
Fibreglass thickness selection
While calculations suggest 1-ply of 80 gsm fibre-glass would be satisfactory, the
calculations are based on idealised flight conditions, and do not consider bulking
failure. In order to ensure that the plane is able to withstand a reasonable crash, 3
layers of 320gsm glass was selected. In addition to this, a layer of Kevlar was selected
to be used in the nose area
Longeron selection
Since the longerons do not solely carry the load of the aircraft, the fuselage carries
some of the load, the size of longerons previously mentioned are sufficient for use in
the aircraft. Due to availability, the final longerons selected had an outer diameter of 8
Fibreglass Longerons
Max Shear Stress
3-ply (MPa)
Max Shear Stress
2-ply (MPa)
Max Shear Stress
1-ply (MPa)
Max Bending Stress
(MPa)
13.88 20.83 41.65 70.91
Section 4.2 Wing Design
89
mm and an inner diameter of 6 mm, making them stronger due to the larger diameter
and extra thickness.
4.2 Wing Design
The aircraft wing was partially specified by the decisions made in conceptual design;
however, more detailed specifications were required to enable construction of a final
wing. The airfoil selected was a NACA 4412 to enable high lift and low drag at a varying
range of angles of attack. The ailerons were sized to provide adequate performance in
roll.
4.3 Wing Structural Design
The wing structure was designed to withstand all operational loads anticipated within
the aircrafts flight regime, with considerations made to ensure the structure is easy to
assemble. All flight loads were determined from the United States Federal Aviation
Regulations, section 23, (FAR-23) which governs the airworthiness standards for
Normal, Utility, Acrobatic and Commuter aircraft. Material safety factors were
obtained from the Australian Civil Aviation Safety Regulations (CASR), Section 101,
which governs the use of Unmanned Aerial Vehicles and Rockets. The results are
summarised below and have been taken into account in all calculations:
Maximum wing load factor of 4.4 in accordance with FAR-23
A safety factor of 2.25 in accordance with CASR 101
A safety factor of 1.5 on composite materials in accordance with CASR 101
4.3.1 Lifting force profile
The lifting force generated by the wing was initially determined so that the stresses in
the wing could be designed for accurately. Two methods were initially utilised, to
ensure that all conditions of flight could be analysed. In the analysis, the following
assumptions were made.
All lifting forces are due to the wing, the fuselage generates no lift.
90
The aerodynamic lifting force is the most significant force on the wings, drag
forces are not considered.
Both methods are approximations; they do not take into account wing tip
vortices.
4.3.1.1 Shrenks Approximation
The lifting profile was initially generated using Schrenks approximation. The method
states that any untwisted planar wings spanwise load distribution shape can be
approximated by the average of its actual planform shape and an elliptic wing shape of
the same span and area (Schrenk, 1940), as illustrated in Figure 64.

Figure 64 - Schrenk's Approximation
(ISOAR 2007)
The wing planform and elliptical planform are determined using Equation 13 and
Equation 14 respectively.
Equation 13

Equation 14

Section 4.3 Wing Structural Design
91
The calculation of C(y) gives the overall area of the wing. In order to determine the
lifting force, an approximation of the load to be supported by the wing is made. In this
case each wing is assumed to support half the weight of the aircraft, multiplied by the
load factor determined using FAR-23 calculations for V-n diagrams. Using the load
estimation and the surface area a force factor Equation 15 can be calculated to
determine the load distribution.
Equation 15

The load created using this method was corrected for the weight of wing, giving the
load profile as shown in Figure 65. It should be noted however that the wing is
supported from the edge of the fuselage, such that all further calculations will only
involve the area from 0.05m to 0.79m.
Schrenk Approximated Lift Force
0
50
100
150
200
250
300
350
400
0 0.2 0.4 0.6 0.8 1
Spanwise Distance (m)
L
o
a
d

(
N
)

Figure 65 -Lifting Force Distribution
Based on the load distribution calculated shear force diagram could be created from
the integral of the load distribution between 0.05 and 0.79, including appropriate
boundary conditions. Similarly the bending force diagram can be determined. The
calculations were completed with the assistance of Matlab 7.4.
92
0.1 0.2 0.3 0.4 0.5 0.6 0.7
-200
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
distance from root (m)
Shear Forces
F
o
r
c
e

(
N
)

Figure 66- wing shear distribution
0.1 0.2 0.3 0.4 0.5 0.6 0.7
0
10
20
30
40
50
60
distance from root (m)
Bending Force
F
o
r
c
e

(
N
m
)

Figure 67 - wing bending force distribution
4.3.1.2 Alternative Distribution Calculations
Whist Schreks method provides a good estimate of the wing load distribution it has no
theoretical reason for its accuracy. To better understand the aerodynamic forces
applied to the wing, a more theoretical method was investigated. The method chosen
is outlined in Theory of Wing Sections (Abbot 1959). The method considers the
spanwise distribution to consist of two parts. The first part, known as the basic
distribution is based on the twist of the wing, which for in this case is zero. The second
part is known as the additional distribution and is the lift due to the wing angle of
attack. The method, uses tables to easily select parameters for the wing shape, and
93
uses data from the aerofoil at different speeds to determine an accurate lift profile.
The data required to find the spanwise lift distribution is given in table form, and can
be interpolated based on the aspect ratio and taper ratio of the wing. From the data
provided, and data gained from the Javafoil aerofoil simulator, and approximation fro
the wing lifting for can be found using Equation 16.
Equation 16

Where CL is the lift coefficient of the wing, as determined from Javafoil at a certain
condition, and
al
cl is determined from Equation 17.
Equation 17

Based on these calculations the following lift distribution was determined for a cruise
speed of 88m/s at zero angle of attack (Figure 68).

Corrected Cl Distribution (88m/s)
0
0.02
0.04
0.06
0.08
0.1
0.12
0 0.2 0.4 0.6 0.8 1
Spanwise Postion from Root (m)
C
o
r
r
e
c
t
e
d

C
l

Figure 68 - Corrected Cl Distribution

94
By estimating the area of the wing area at span wise stations along the wing, the lift
distribution was determined. The distribution is then corrected for weight, by
estimating the shape of the wing as a trapezoid (Beer 2006).

Lift Distribution at 88m/s (Including Load Factor)
0
2
4
6
8
10
12
0 0.2 0.4 0.6 0.8 1
Spanwise Distance from root (m)
L
i
f
t

f
o
r
c
e

(
N
/
m
)
Weight Corrected Lift
Distribution

Figure 69 - Lift Distribution at 88m/s
This method offers the opportunity to analyse the loads on the wing at different points
in the aircrafts flight profile. It also allows more design flexibility, as the effect of such
factors as sweep and twist can be introduced to the analysis.
95
4.3.2 Spar Design
Spars are utilised as the primary method for handling the bending and shear forces
due to the aerodynamic lifting forces generated during flight. Spars were located at the
20% and 80% wing location (Avalakki et al, 2007). It was determined that two spars on
the top and bottom surfaces of the wing would be the best option for transferring
loads along the wing into the fuselage. Calculations for the bending stresses in the
spars were undertaken by determining the moment of inertia for the four spar system.
To determine the moment of inertia first the centroid of the spar system was
calculated using Equation 18 and Equation 19.
Equation 18

Equation 19

From the geometry of the aerofoil the moment of inertia along the wing (x direction)
could be determined using Equation 20, assuming that only the distance between the
spar and centroid caused inertia in the system.
Equation 20

The analysis of the bending stresses in the wing at different positions was used to
determine the cross sectional area required for each spar using Equation 21.
Equation 21

96
By analysing the bending stresses at the root of the wing, it was determined that an
overall spar area of 1.25x10-
5
m was required. This was equivalent to four 1.25mm
thick by 11mm wide spars manufactured from uni-directional carbon fibre with a
tensile stress of 580MPa and a compounded safety factor of 3.75, based on the CASR
regulations outlined in Section 4.3. The bending stresses were also analysed at
spanwise sections along the wing, as shown in Figure 70

Optimum Spar Thickness
0
0.002
0.004
0.006
0.008
0.01
0.012
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Distance from root (m)
S
p
a
r

T
h
i
c
k
n
e
s
s

(
m
)

Figure 70 - Maximum spar thickness from root to tip of the wing
This initial analysis assumed that the bending force was taken equally by each spar,
however further analysis was undertaken due to the non-symmetric load conditions
which exist on the wing. In general the centre of pressure for a wing is at
approximately 25% of the chord. Therefore the percentage of the load carried by the
front and rear spars was calculated using Equation 22 and E
quation 23.
97
Equation 22

E
quation 23

Analysis utilising this method concluded that four 1.25mm thick, and 12mm wide spars
would be capable of resisting the bending stressed induced by the wing lifting forces.

The shear stress was also analysed at the root of the wing. Assuming that all the shear
is taken by a single spar, a shear stress of 24Mpa was calculated using Equation 24.
Equation 24

The shear strength of unidirectional carbon fibre is 70Mpa, therefore it was deemed
unnecessary to analyse the shear at other positions along the wing.
4.3.3 Torsion
Torsion in the wing structure was estimated by assuming that the lift force of the wing
acts through the centre of pressure of the aerofoil, creating a moment around the
static neutral point of the wing. The static neutral point is calculated by analysing the
spars and the foam between the spars as a single structure. Using a method defined by
Beer et al (2006), the structural neutral point for a 2-D body consisting of multiple
materials can be approximated using a ratio of the materials Modulus of Elasticity, to
create a scaled moment of inertia.

98
The structural neutral point of the wing was determined at the root chord of the wing.
The torsion was analysed in both takeoff and cruise scenarios between -5 and 15
degrees angle of attack.
The centre of pressure from the aerofoil was determined using Designfoil software.
This software re-creates the pressure profile over the aerofoil at various angles of
attack. From this data the centre of pressure is determined using Equation 25.
Equation 25

The centre of pressure is not affected by changes in Reynolds number. Figure 71 shows
the change in position of centre of pressure with angle of attack.

Change in Centre of Pressure (Cp) of NACA 4412 with Angle of
Attack (AOA)
0
10
20
30
40
50
60
70
80
90
100
-5 0 5 10 15 20
AOA (degrees)
C
p

(
x
/
c
)

Figure 71 - Position of Centre of Pressure with AOA

Based on the structural neutral point and the centre of pressure, a level arm could be
determined. The root chord was analysed at takeoff (70km/hr), climb (150km/hr) and
cruise speed (300km/hr), at various angles of attack, as shown in the following figures.
99
Torsion vs Angle of Attack (Take-off)
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 5 10 15 20
AOA (degrees)
T
o
r
s
i
o
n

N
m

Figure 72 - Wing Torque at Takeoff (70km/hr)
Torsion vs AOA (Climb Speed)
0
5
10
15
20
25
0 5 10 15 20
AOA (degrees)
T
o
r
s
i
o
n

(
N
m
)

Figure 73 - Wing Torque at Climb Speed (150km/hr)
100
Torsion vs AOA (Cruise Speed)
0
20
40
60
80
100
0 5 10 15 20
AOA (degrees)
T
o
r
q
u
e

N
/
m

Figure 74 - Wing Torque at Cruise Speed (300km/hr)
The maximum torque of 77Nm occurs at 13 degrees angle of attack at 300km/hr.
Although it is unlikely that these conditions will be experienced by the aircraft, it has
been designed for to ensure the wing structure is capable of handling all conditions
within the speed profile of the aircraft.

The wing torque was analysed using Equation 21, assuming that the fibreglass skin
could take the torque in the wing.

The moment of inertia of the structure was calculated initially assuming a 0.5mm
thickness. This equates to approximately 3 layers of 84gsm fibreglass. Stresses at the
root of the wing were calculated resulting in a maximum stress of 23.2Mpa. This allows
a 12.9 times safety factor, based on a tensile strength of 300MPa for fibreglass (FGI,
2007).
Section 4.4 Wing Connection
101
4.4 Wing Connection
The wing connection was designed with three aims in mind.
Ability of the structure to withstand normal flight loads
Be easy to assemble and disassemble
Easy to replace in the event of a crash
Protect the wing structure from damage during the event of a crash

For this reason it was determined that a twin beam system, as shown in was the most
suitable method for this application. The system was designed to cope with ten times
load factors, in accordance with a heavy landing where the wing comes into contact
with the ground.

Figure 75- Wing Connection System
The wing system was designed so that it would fail in three stages during a crash. The
aim of this was to progressively protect the more expensive and difficult to replace
components. The first phase of failure was through a tensioning system. This protected
the main load bearing beams from the initial Impact loads. Once failure of the tension
system failed the loads were transferred to the wing connection beams, which were
manufactured from uni-directional rolled carbon fibre. While these beams were
relatively low in cost, it was found that obtaining the required diameter was difficult,
102
with up to two months lead time quoted by some companies. Finally, in the event of a
serious crash the beams would fail and the wings would detach from the main body.

The main beam section was designed to take the loads during standard flight
operations. The problem can be analysed as a simple cantilever beam in bending.
Force loads of 10x the normal flight loads were used to take into account landing, and
to ensure deflection of the wing was kept to a minimum during normal 4G flight
maneuvers. Two carbon fibre rods were used to support the wing. They were situated
at 20% and 80% of the root chord, and extended to the start of the ailerons, where it
was mounted into a 4mm ply rib. The ply rib was analyzed to take shear stresses
during landing, based on a 10x load factor. Based on a shear stress of 50MPa for
plywood, there was a 3.5x reserve factor, suggesting that plywood was the most
suitable choice.
From the stress analysis conducted it was determined from that two beams of 12mm
and 10mm 1mm thick uni-directional carbon fibre were able to take the aircraft
loadings, however further analysis showed that a deflection of 60mm in the wingtip
would occur under maximum load conditions. This was deemed excessive and
therefore the thickness of the main beam was increased to 2mm. this reduced the
deflection by 2/3, which was seen as suitable for flight.
Section 4.5 Control Surface Sizing
103
4.5 Control Surface Sizing
In this section the sizing of the control surfaces and servo motors to actuate them is
presented. The aircraft had a rudder less design as sufficient control for utility flight
can be achieved using the elevators and ailerons. Standard radio control servo motors
were selected for use, with compact motors preferred.
4.5.1 Aileron Sizing
The ailerons will provide roll control. Considerations relevant to their design are:
It is preferable for ailerons to be located as far out-ward as possible - where
they will provide maximum moment. The same actuating effort will be
required regardless of position, as the effort depends only on the size of the
surface being deflected.
Guidelines state that suitable percent chord for the ailerons to occupy is
between 35-20% (Raymer 2004).
The ailerons should occupy approximately 7% of the total wing area, according
to Eger (1983).

Since the trailing edge of the wing is straight and the rear spar is straight, then for
simplicity the aileron can be rectangular.

The required area for each aileron is:
2
aileron
2
aileron
2
m 0133 . S
m .0266 S 07 . 2S
m .38 S
=
= =
=

The required span for each aileron is:
b
aileron
= .0133 m
2
/ .0694 m = 191.6 mm

104
Thus the required dimensions of each aileron are approximately 70 mm by 200 mm,
the parameters of the ailerons are tabulated in Table 13 and the ailerons are
illustrated in Figure 76.
Table 13 - Aileron Dimensions
S
aileron
(m
2
per aileron) .0133
C
aileron
(mm) 70
b
aileron
(mm) 200

Functions in the radio controller allow for the ailerons to double as flaperons, where
the motion of the surfaces is together instead of opposed, for additional lift at take-off.
This will be integrated into the control system to provide added lift on takeoff and
landing if it is considered necessary during testing.
4.5.2 Elevator Sizing
The elevators were sized with a chord of approximately 20% of the horizontal tail. The
elevators were designed to span as much of the horizontal tail as possible for
maximum control authority. The elevators run from either side of the central section
of the T-tail from the vertical tail outward.

The positioning of another elevator surface in the central part of the tail between the
vertical tails was also considered. The extra complexity of manufacture and extra
20% of root
chord
(~69.4 mm)
20% of
root
chord
40 mm 200 mm
S
aileron

Figure 76 - Aileron Dimensions
Section 4.5 Control Surface Sizing
105
servo required was thought to be too complex for the small increase in elevator
power.
4.5.3 Servo Motor Sizing
The servo motors to operate the control surfaces were sized based on the required
torque to actuate the surfaces, which was found based on the dynamic pressure and
the size of the control surfaces. The torques were calculated for the extended goal
design speed of 250 km/hr, with greater velocity requiring more actuating torque and
thus providing for additional reserve factor at lower flight speeds. The servo
requirements can be seen in Table 14
Table 14- Servo Requirements
Surface Deflection
(degrees)
Required torque
for maximum
deflection (kg.m)
Selected
servo
Torque of
selected
servo
(kg.m)
Reserve
Factor
Aileron 20 2.0 Eagle E711 4.2 2.1
Elevator 15 1.2 JR NES 331 3.25 2.7

The servo motors supplied with the radio transmitter were sufficiently powerful to
operate the control surfaces. However, their physical size was too large for the
intended wing based. Therefore, smaller servos were sought.

The servo selected for the ailerons was the Eagle E711. It had the same width and
depth as the servos supplied with the transmitter but was a low profile servo with a
much reduced height that enabled it to fit inside a wing mounting. They have
dimensions measured to be 44 x 22 x 23 mm.

The servo selected for the elevators was the JR NES 71. It was a micro servo, with
measured dimensions of 30 x 13 x 28
106
For the nose wheel steering, the servo supplied with the radio transmitter was
deemed suitable as mounting space was not an issue and the servo torque was
sufficient.
4.6 Pulsejet Engine Mount
The engine mount system is required to take the loads of the engine in all operating
conditions, as well as provide vibration and thermal isolation to the fuselage. Initial
engine testing showed that the temperature of the engine could reach as an estimated
1000C if left for an extended period without any forced air cooling, requiring that
thermal conditions be considered in the design of the mount.

Initial testing on the flight engine produced a maximum sustained thrust of 3.5kg on
propane. As the engine performance at speed was not known, the peak thrust
expected from the flight engine was safely defined at 4kg.

A safety factor of 2 was incorporated into the design to allow for any increased load
due to the cyclic vibrations seen in the engine. It was also decided that the front
engine mount would be designed to take the full thrust load of the engine, in the case
that the rear mount was not aligned correctly.
4.6.1 Mounting Locations
In order mount the engine securely, it was decided that the engine needed at least 2
rigid mounts. Initial concepts involved tabs welded to the engine; however it concern
was raised in regards to the loading on the mounting locations. While the 2007 engines
did not suffer any form of damage from the mounting tabs, the transition to 0.5mm
sheet, from 1.2mm sheet significantly reduced the ability of the engine walls to take
the bending and moment stresses. For this reason, it was decided that the spark plug
nut was a better mounting location, as it is the strongest component of the engine,
and hence least susceptible to warping or failure due to the engine mount loads.

Section 4.6 Pulsejet Engine Mount
107
Following initial engine tests using the spark plug nut as the primary engine mount, a
crack was discovered at the weld interface of the nut. It is believed that this occurred
due to increased stresses as a result of the engine oscillating about the spark plug nut,
and hence it was decided that another mounting point should be used.

It was decided that the addition of a mount forward of the combustion chamber would
permit the mounting of the engine, while transferring the thrust and vibration loads to
the combustion chamber. This extension is shown in Figure 77. Mounting the
extension above the sparkplugs allows the engine to have a longer thermal path to the
bulkhead, improving the thermal performance of the engine mount.

Figure 77 - The mounting extension on the front of the Chinese engine
4.6.2 Thermal Isolation
The expected temperatures of the engine near the combustion chamber approach
1000C, which, if this temperature was not effectively isolated, would cause severe
problems to the fibreglass and timber materials used in the aircraft. The maximum
temperature for the plywood bulkhead is approximately 100C, after which the material
begins to degrade. In order to ensure that the materials are not exposed to excessive
temperatures, a thermal analysis was required. The model used considered both the
effects of convection and radiation, in order to produce a realistic representation of
the actual situation.
108

The emissivity of stainless steel varies dramatically with surface conditions, which, at
the temperatures expected is not well known. Because of this, the value of emissivity
was conservatively estimated at 0.5.

The convection coefficient also had to be estimated, due to the uncertainty of the
conditions that will surround the mount. The value of 15 W/m
2
K was estimated from
common values given for forced and natural flows. It was assumed that the flow
around the mount would be greater than natural convection, due to the movement of
the engine and airflow, and as such, the high value in the common range of 3 to 20
W/m
2
K was used (Mills, 1998).
4.6.3 Vibration Isolation
Due to the operation of the pulse jet engine, a large amount of vibration is produced,
which could lead to problems if not correctly isolated. No information was available in
regards to the vibration tolerances of the electronic systems that will be present in the
plane, so it was decided that the best vibration isolation that is reasonable obtainable
should be used.

Engine tests on propane produced an operating frequency of around 200Hz, and hence
this value was considered the dominant frequency to isolate.

As can be seen in Figure 78, the best isolation of a vibration force is achieved when the
frequency ratio is as high as possible, and damping ratio is as low as possible. As such,
it is desired to produce an engine mount and engine system that has the lowest
natural frequency that is obtainable while maintaining enough structural strength to
resist failure.
109

Figure 78 - Force transmissibility as a function of frequency ratio and damping ratio
(Chan 2007)
The natural frequency of a spring-mass system is given by Equation 26 below:
Equation 26
m
k
w =

As such, to reduce the natural frequency, the spring stiffness must be as low as
possible, as the mass of the engine is not changeable. Figure 78 shows that in order to
achieve a reasonable level of vibration isolation of 50%, it is necessary to have a
frequency ratio of around 2 for low damping systems, or around 3.5 for systems with
high damping levels.
To achieve a frequency ratio of 2, the engine system must have a natural frequency of
around 100Hz, which, for a mass of 0.8kg, gives a spring stiffness of 315827N/m. For a
spring stiffness of 315827N/m, the deflection that the spring would have under a
thrust load of 40N is 0.13mm. Table 15 shows the expected spring stiffness and
deflection over a range of frequency ratios.
110
Table 15 - Spring stiffness and deflection under a 40N thrust load, for various frequecy ratios.
The shaded reigion of the table will not provide any vibration isolation.
Frequency
Ratio
Stiffness
(N/M) Deflection (mm)
0.5 5053237 0.01
1 1263309 0.03
1.5 561471 0.07
2 315827 0.13
2.5 202129 0.20
3 140368 0.28
4 78957 0.51
5 50532 0.79
6 35092 1.14
8 19739 2.03
10 12633 3.17
The actual vibration levels produced by the engine were not available, as during testing
the vibration intensity is affected by the significant mass (around 14kg) of the
mounting trolley. Recoded data suggests that the engine and trolley has a variation in
thrust of around 0.3kg about the mean thrust. In order to obtain more accurate
vibration quantification, the test stand was approximated to be a spring-mass system
with an input force equal to the thrust variation, approximated as a sinusoidal force.
Data obtained from the manufacturer of the load cell and an ANSYS simulation of the
trolley was used to estimate a spring coefficient for the system. From this an estimate
of the natural frequency of the trolley-engine system was found to be around 940Hz.
This gives a frequency ratio of around 8. Using Figure 78, and the assumption that the
trolley system was almost critically damped due to friction, it was determined that the
vibrations transmission is in the order of 15%. Thus, the actual force variation
generated by the engine is estimated to be in the order of +-2kg. The calculations used
to determine this can be found in Appendix E Engine Mounting Calculations
4.6.4 Vibration Isolation Method:
In order to develop the optimal design, it was necessary to consider various methods
which would allow for vibration isolation.
111
Buckling beam
The use of a buckling beam style vibration isolation system was considered to provide
the spring support. This approach exploits the non-linear spring co-efficient of a
buckling beam, and would hence be able to provide a high level of vibration isolation,
without excessive deflection. The problem that limits the use of this design is that
under part thrust conditions, where the average loading is less than that at peak
thrust, the spring stiffness will be much higher than is required to isolate vibrations. As
a result, this approach would only provide vibration isolation under full thrust
conditions, and is hence not suitable for the application.
Vibration damping
A soft foam or rubber vibration bed was considered in order to provide or supplement
the isolation. Initial research produced no materials that could provide stiffness as low
as is required, but still operate in the thermal conditions near the engine. The use of
the rubber as a supplement to the mount deflection is however reasonable. In order to
investigate this approach, a sample of the most suitable rubber obtainable from a
retail outlet was obtained. As the outlet could not provide material data for the
product, the youngs modulus of the material was estimated from a basic deflection
test to be in the order of 5.6 MPa. A problem associated with the use of a rubber is the
increased damping of the system will increase force transmissibility, and hence its use
is not ideal if a spring system is possible. The material was also flame tested to ensure
that it did not pose a significant fire hazard, which it did not.
Spring and Linkage System
The use of a linkage system and a conventional coil spring was also considered for this
task. This approach could easily allow the required deflection, and would not have any
structural issues due to cantilever bending. The major concern that was associated
with this design is the complexity of the system, which requires 4 separate parts and 2
pivot axis. The primary concern was related to the thermal conditions at the top pivot
112
point. It was thought that a pivot point that is tight enough to stop vibration within the
joint when cold may bind when the joint gets hot due to thermal expansion.
Cantilever beam
The use of a cantilever beam spring system is the simplest and most conventional
method considered. This approach would provide the required stiffness for vibration
isolation, while also providing thermal resistance. The potential downside to this
approach is that the ideal vibration isolation may not be possible whilst retaining
thermal isolation and a low mass. This approach was the most promising of the options
considered, and hence was investigated further.
4.6.5 Engine Mount Materials
The material used for the engine mount is required to be able to support the loads of
the engine at high temperatures for a reasonable period of time. Initial investigations
showed that few materials were available that were able to provide the required
structural strength at the expected operating temperatures, as well as provide good
thermal isolation. The best materials for thermal isolation, predominantly ceramics,
would be very difficult to manufacture, and would be prone to brittle failure, which is
undesirable, as it would most likely result in a crash of the model, and a significant fire
hazard. Aluminium alloys were not considered due to melting points commonly in the
order of 600C. The materials reviewed for selection were restricted to mild and
stainless steels on the grounds of manufacturing and durability requirements.
Table 16: Material Selection for Engine Mount
Matweb 2008
Material Thermal
conductivity
Yield strength
(MPa)
Max Service
Temperature
Density
1018 Mild Steel ~50 275 (not given) 7.87
310 Stainless Steel ~20 310 1150 8.00
304 Stainless Steel ~20 215 925 8.00
302 Stainless Steel ~20 275 925 7.86
113

The material properties of stainless steel at elevated temperatures can be estimated
from Figure 79.

Figure 79 - Yield stress relative to room temperature as a function of temperature for
301,302,304,321,347 annealed stainless steels
(USADOD, 1998)
The increased thermal conductivity of the mild steel will result in significantly reduced
thermal isolation, and as such is not suitable for the application. Of the stainless steels
considered, 310 stainless steel provides the best performance characteristics. 310
stainless steel is also readily available, and hence is the material chosen for the
manufacture of the engine mount.
4.6.6 Final Design
This design incorporates the engine extension to reduce stresses on the spark plug nut.
As a result of the extension, it also produced much better thermal analysis results than
the initial design. The final design is shown in Figure 80.

114

Figure 80 - The final engine mount design. The modification made to the front of the engine is shown
in green.
Thermal Analysis
Analysis of the final engine mount design showed that the temperatures at the
mounting points were at quite safe levels. The addition of the extension to the
combustion chamber provides the majority of the thermal isolation, reducing the
temperature at the interface with the engine mount to around 175C. If further thermal
isolation were required, a layer of ceramic material could be placed between the two
parts at the interface, and would significantly reduce the temperature at the bulkhead.

Figure 81 - Thermal analysis results of the engine mount.
115
The results of the thermal analysis suggest that there will not be thermal damage the
bulkhead, and hence the design is considered to be satisfactory.
Vibration Analysis
Using a preliminary ANSYS structural model, the spring co-efficient of the mount was
derived from the deflection due to the loading. This value was then used to calculate
the natural frequency of the engine system. The natural frequency was found to be 31
Hz, resulting in a frequency ratio of 6.3. This value will reduce vibration to around 10%
of the initial value.

This result was considered to be acceptable for the application, and hence no further
design changes were required on the basis of vibration isolation.
Structural Analysis
Structural analysis of the engine mount showed that stresses in the critical locations
are below the yield stress of the material under the 80N design load. The peak
reported stresses are singularities at the constraint points, and not of concern.

Figure 82 - Stress distribution within the initial design under an 80N load
116

In order to determine if the stress behaviour of the engine changes under dynamic
conditions, a flexural dynamic analysis was performed. In this analysis, a load of 40N
was applied, with a variation of 20N at 200 Hz about this mean value. The stress
distribution is shown in Figure 83, showing that all critical locations are significantly
below the material capabilities. As with the static analysis, the peak reported stresses
are singularities at the constraint points, and not of concern.

Figure 83 - Stress distribution within the design under dynamic loading of 40N +- 20N
Verification
In order to verify the thermal and vibration analysis results, the engine mount was
used throughout the testing stage of the Chinese engine. During this testing, it was
found that the temperatures during operation were never high enough to cause
damage to the plywood mounting plate, even on extended runs. Due to technical
issues with the load cell, vibration data was not obtained, and hence the vibration
analysis results could not be verified.
117
4.6.7 Engine Modal Analysis
A modal analysis of the engine was performed using ANSYS software to determine if
specific mounting systems would cause excessive vibration transmission to the
fuselage of the plane.
It was determined that mounting the engine at the spark plug nut and the end of the
tail resulted in oscillation modes at 197 and 208 Hz, with a shape similar to the first
mode of a simply supported beam (Figure 84). This was reason for concern, as the
operation of the engine may excite this mode, resulting in excessive vibration of the
engine.

Figure 84 - 208Hz vibration mode of the engine, mounted at ends
Due to the presence of this vibration mode, the rear engine mount was moved to
inwards from the tip of the tail, to the joint between the exhaust cone and the final
pipe. An analysis of this mounting method did not produce any modal frequencies in
the operation range, and as such it was considered the best solution.

118
4.7 Pulsejet Launch System
In order to fly the aircraft powered by a valveless pulsejet engine, it is necessary to
develop a method of safely starting the engine while mounted to the fuselage. This
requires that the aircraft is supplied with propane, compressed air, and spark during
the start up process. Due to safety concerns, the operators are required to be
positioned several metres away from the aircraft, to ensure safety even in the
occurrence of a fire. As such, the aircraft will be started and released from a distance
of at least 5 metres.

As propane, air and spark are only required for the initial start up, these systems are
integrated into the launch system, and not attached directly to the plane. After the
engine is started and the plane is released, the propane and compressed air injectors
will be left behind.

4.7.1 Launch process
The following is the launch process required to start the pulsejet engine and release
the plane on the runway.

Start the engine on propane: Propane and compressed air are supplied via the supply
lines. Spark is supplied from the generator circuit, located near to, but not in the way
of the plane.
Transition to onboard liquid fuel: Throttle on the RC unit is increased as the propane
flow is decreased.
Completely stop Propane and air: turn off both propane and air. This must be done
before release to prevent flames coming in contact with the tail of the plane.
Remove spark: The spark plug is hooked up to the spark generation box with a quick
release coupling. Once the engine has settled, the spark circuit is turned off. The wires
can then be pulled off from a remote location, releasing the coupling from the aircraft.
Section 4.7 Pulsejet Launch System
119
Release: Once the engine is running on liquid fuel, the plane is released from the stand
immediately. This is done by releasing the pin that joins the plane to the launch stand.
This leaves the air and propane injectors behind. If complete automation was required,
this could be done with an electric solenoid.
4.7.2 Launch Stand Components
The launch stand consists of four main components which are required to allow each
step of the launch process. These components are discussed in the following section.
Release mechanism
The launch stand is required to restrain the plane until the engine is started and
settled. The supply lines will provide the restraint force required to resist engine thrust
during start up. A tab welded on the supply lines and tabs welded on the engine intake
are connected with a 4mm rod. The rod is pulled releasing the coupling and allowing
the plane to move under thrust.

Figure 85 - Release tab attached to the intake of the engine, and release tab on the launch stand
Spark Plug wires
The ignition circuit requires both an active wire, and a ground wire. The active wire is
attached to the head of the spark plug using an alligator clip with external insulation.
This is required to prevent sparking against the engine mount. The ground wire is
attached to the engine mount with another alligator clip.
The use of alligator clips allows the wires to be disconnected remotely, by simply
pulling on the wires. Care must be taken to ensure that the spark generator has been
120
turned off before pulling the wires, to ensure that the connectors do not touch and
short the discharge coil.
Supply lines
The supply lines are required to provide the engine with propane fuel and compressed
air. Additionally, they act as injectors and the restraint system. These lines are 5mm ID
steel tubing, with fittings on the inlet end to allow connection to the compressed air
and propane gas supplies.
Stand
The stand must be able to support the thrust force of the plane during the start up
process. Due to the likelihood of starting the plane on a sealed runway, it is necessary
to use weights to hold the stand down, and prevent it from moving. The stand
structure is shown in Figure 86 below. Bricks or other non-flammable heavy objects are
to be placed on the three legs of the structure, preventing it from moving.

Figure 86 - Launch stand for pulsejet flight

Section 4.8 Electrical and Electronic Components
121
4.8 Electrical and Electronic Components
The electronic and electrical components installed inside the fuselage include the
pump and its related components, and radio equipment.
4.8.1 Pump and related components
A Flightworks 200 C fuel pump was used to supply the engine with high pressure liquid
fuel. The pump is compact, lightweight and designed for use on an aircraft. The
dimensions of the pump are shown in Figure 87.

Figure 87: Flight Works Fuel Pump
(FlightWorks Inc. 2008)
Pump battery
A separate battery will be used for the pump to ensure that the pump operates reliably
without interference during the flightA FlightPower 7.4 V (3 cell lithium polymer)
battery was selected with a capacity of 3750 mAh according to the packaging. This
allowed for multiple flights on a single charge, and ample reserve capacity to avoid
excessive voltage drop during discharge of the cells.
Throttle
To regulate the voltage of the pump and thereby throttle the fuel supply, an Electronic
Speed Controller (ESC) was required. The ESC provides the pump with a Pulse Width
122
Modulated (PWM) signal, allowing variation in pump power. ESCs are lightweight and
will form the only link from the radio control circuit to the engine control circuit.

As the pump is DC and requires only a small amount of power thus the smallest
available brushed ESC was selected for use
4.8.2 Radio Controller
The radio controller selected was based on the number of channels required and ease
of programming. A minimum of five channels were required for the following
functions:
Throttle
Aileron (Left)
Aileron (Right)
Elevator
Nose wheel steering

Two ailerons channels were required as the ailerons were to have separate servos and
be programmable for use in unison as flaperons to assist take-off.
The controller selected was the Spektrum DX7. Although its 7 channels exceeded the
requirements, it was selected for its more powerful programming capabilities than
most 4-5 channel radios, and its 2.4 GHz radio frequency which did not require the use
of frequency pins. The radio system was supplied with a receiver and additional
antenna that formed part of a loom including a small 4 x AA style NiCad battery
package to power the servo motors. This loom was to be installed inside the nose of
the aircraft, with extension cables used to connect the servo motors. (Model Flight
2007)
Section 4.9 Ducted Fan
123
4.9 Ducted Fan
4.9.1 Purpose of fan
For the first flight tests of the aircraft it was decided that an alternate propulsion
system was to be used in order to ensure the reliability of the propulsion. A ducted fan
was chosen for this first flight as it had similar propulsive capabilities and it was
estimated that the system would have the same weight as the pulsejet engine and fuel
system.
4.9.2 Selection of fan system
The fan system that was to be selected needed to have the capacity to produce the
same thrust as the pulsejet engine, it also was required to be able to have enough
power for 10 minutes of flight whilst having a system weight similar to that of the
pulsejet engine system. The fan also was required to be capable of being mounted to
the aircraft.
Ducted Fan
The fan selected was the compact Schubeler DS-51-DIA HDT electric ducted fan, which
has a 90mm inner diameter and 51 cm
2
swept area. The fan is manufactured of
carbon fibre wrapped around aluminium, and the duct is carbon-fibre sandwiching a
honeycomb core. The outer diameter of the fan is sufficiently small that it would fit
inside the rounded cut away inside the bulkheads for the pulsejet engine. (Schuebeler-
Jets 2008) The fan can be seen below in Figure 88

Figure 88 - Schubeler Ducted Fan
(Schuebeler-Jets 2008)

124
Other reasons as to the selection of this ducted fan are that it offers the most static
thrust output for the fan size compared to others and that the fan is supplied
balanced, which was thought to reduce the set-up time required.

The thrust range of the fan is 10-33+ N, closely approximating the design thrust range
expected from the valveless pulsejet engine. The fan can produce up to 38 N static
thrust with a suitably low-revving, high torque motor (Schuebeler-Jets 2008),
approximating the extended thrust goals for the valveless pulsejet engine.

As the fan is supplied balanced there shouldnt have be any balancing issues associated
with the high angular velocity, however the fan will be tested statically before
attempted use in flight to ensure the fan is balanced and to verify the thrust output.
The parameters of the ducted fan can be seen below in Table 17
Table 17 - Ducted Fan Parameters
Inner Diameter (mm) 90
Outer Diameter (mm) 94.5
Thrust Range 10-33+ N
Speed Range (rev/min) 35,000 45,000+
Maximum motor diameter (mm) 38
Fits motor shaft size (mm) 5
Mass of Fan (kg) .058
(Schuebeler-Jets 2008)
Electric Motor
The motor selected for use was the Lehner 1950/11 D, a brushless electric motor
manufactured by Lehner Motoren Technik. Similar motors are shown in Figure 89.

125

Figure 89 - Lehner electric motors
(Ductedfans.com 2008)

This motor was selected on the basis that it would fit inside the motor cavity provided
in the fan, it would produce the revolutions per minute required by the fan in order to
produce the required thrust and the motor was available.

The parameters of the motor can be seen in Table 18
Table 18 - Parameters of Lehner 1950 Electric Motor
(Fine Design RC 2008)

Mass (kg) 0.355
Maximum Angular Velocity
(rpm)
50000
Maximum power input (kW) 3.0
Maximum current input (A) 100

Batteries
The batteries selected for use were 2 sets of 5-cell Lithium-Polymer batteries in series
for a total of 4 batteries. The battery packs that were selected can be seen below in
Figure 90
126

Figure 90 - ZIPPY-R battery pack
(Hobbycity.com 2008)

These battery packs in such a configuration were selected based on; the required
voltage to operate the motor, the required capacity to fly for 10 minutes, cost &
availability and capability for the required current draw.

Voltage equivalent to 10 cells (or 37 volts) were required to operate the fan. However
10-cell battery packs were found to be extremely bulky and difficult to package they
consisted of 2 5-cell packs joined together lengthways and therefore would be difficult
to package between the bulkheads in the airframe, subsequently it was thought that
two 5-cell battery packs connected in series would be easier to package inside the
airframe.

The amount of capacity required for 10 minutes of flight was conservatively estimated
based on full throttle settings which indicated a requirement for around 10 Ah. Thus
each battery packs in parallel needed to have a capacity of around 5 Ah.

Based on the first three considerations, ZIPPY-R 4800 mAh 25C 5S1P battery packs
were selected. The mass and dimensions of these packs are 636 grams and 170 x 43x
45 mm respectively. Each battery pack was capable of a claimed maximum current
draw of 125 A, so these battery packs are capable of the required current draw.
(Hobbycity 2008)

127
4.9.3 Modifications to the airframe for Ducted Fan Testing

A ducted fan requires relatively consistent airflow into the duct. This required the
addition of cover to reduce the turbulence induced by the sudden drop from the nose
of the aircraft to the pulsejet compartment. A mounting for the ducted fan past this
cover was also designed to provide a suitable mounting point without excessive
additional drag. The ducted fan mounted in the airframe is shown in Figure 91.

Figure 91 - Ducted fan mounted in the airframe
Ducted Fan Mounting
A fan mount was developed based on three considerations:
Maximum length for a smoothly ramped intake,
Minimise the pitching moment contribution of the fan, and
Minimise the amount of weight added towards the rear of the airframe.

To ensure the maximum length for a smoothly ramped intake, the fan was mounted
towards close to the rear of the airframe. It was positioned just forward, and central
of, the two vertical tails. This allowed for an intake with a length of 280 mm, described
in the next section.

The pitching moment contribution of the fan was calculated, and it was found that if
the fan was positioned above the fuselage then the nose down contribution of the fan
due to the high thrust line was such that the elevators did not have enough power to
pitch the aircraft up during take-off. Therefore the fan has been positioned as low in
the airframe as possible to minimise the pitching the moment contribution.
128

Several designs for the actual mount were prepared and considered from plywood. All
the designs took into consideration the provided fan mounts, which were supplied
with the fan itself. To minimise the amount of weight added at the rear of the
airframe, the final mount consisted of two wood tabs attached with epoxy to sides of
the fuselage with the fan was attached using 4 screws to the suppled mountings. The
fan mount can be seen below in Figure 92.

Figure 92 - Ducted Fan Mounting Tabs
Ducted Fan Cover
In order to achieve optimal performance from the ducted fan, it is required that the
incident airflow is as undisturbed as possible. Ideally, this would be achieved by placing
the fan in free stream air. However, due to constraints on the height of the thrust line,
this is not a possible option. As such, it was necessary to create a duct that would
provide the fan with the best possible quality of incident air.
To direct flow into the fan, it was necessary to have a downward sloping cutout, due to
the fan being mounted much lower than the top of the fuselage. The lowest angle that
could be achieved, utilizing the entire length of the cover, was approximately 2
degrees. This angle was chosen as it was thought that the smallest angle possible
would help prevent separation at non-smooth points on the surface, produced in the
manufacturing process. The final design is shown in Figure 93.
Section 4.10 Final Stability Analysis
129

Figure 93 - The cover design for the ducted fan
The cover is made from 0.5mm aluminum sheet in order to obtain the lowest weight
possible. Initial plans were to cut the cover from foam. However the complexity of the
cutting due to the presence of the bulkheads and longerons made the use of metal
sheet more suitable. The cover is secured with 6 screws along the lower side edges, as
well as sliding in under the fan duct. The cover installed on the airframe is shown in
Figure 94.

Figure 94 - The cover installed on the plane

4.10 Final Stability Analysis
A final analysis of the aircraft stability was conducted to ensure the flight quality of the
aircraft on the ducted fan.
130
4.10.1 Longitudinal Moment Analysis
To ensure the aircraft can be trimmed and to take into account the effects of engine
power on aircraft stability, the longitudinal moment derivatives of the aircraft were
determined using approximations developed by Roskam (1985f). The wing component
of this was determined by applying corrections for the aspect ratio, sweep and twist to
the aerofoil zero lift pitching moment coefficient. The fuselage component was found
by applying Hoak (1978)s empirical equation and approximating the fuselage as a
constant area (this gives an overestimate of the magnitude of this moment which is
conservative in determining whether the aircraft is able to be trimmed). The
contribution of the tail to this value was neglected as the tail produced no lift at zero
aircraft angle of attack (without any control surface deflection). The pulsejet and
ducted fan configurations provide a small difference in moment in the power-on
configuration due to differing thrust lines and maximum thrust levels. The base
moment coefficient at zero angle of attack was hence determined as between -0.109
(pulsejet power on) and -0.107 (power off).

The effective overall aerodynamic centre of the aircraft was then determined for both
stick free and stick fixed conditions. The stick fixed, power off value was determined
earlier in Section 3.5.2. To determine the stick free position, the effectiveness of the
elevator in changing the lift of the tail was first determined. This entailed
determination of the effects of changing the aircraft angle of attack on the hinge
moment of the elevator, changing the deflection angle on the hinge moment, and
changing the elevator deflection ion the angle of attack. These were determined
through empirical and statistical methods developed by Roskam (1985f). The effects of
changing the angle of attack of the aircraft on the thrust produced by the ducted fan
were then approximated by using Roskam (1985f)s empirical approach, with a ducted
fan approximating a propeller in this respect. The effects of the normal force produced
by the ducted fan on the aerodynamic centre was also calculated but was so small that
it could have been neglected. The thrust produced by the pulsejet is relatively
independent of angle of attack and so it has a negligible effect on the C
m
term being
131
considered. The aerodynamic centres and centres of gravity under different flight
conditions can be seen below in Figure 95.
Centre of Gravity and Aerodynamic Centre Excursion Diagram
0
1
2
3
4
5
6
7
8
0.00% 20.00% 40.00% 60.00% 80.00% 100.00% 120.00%
%MAC
W
(
k
g
)
Stick Fixed,No Power
Stick Free, No Power
Stick Fixed, Ducted Fan Power
Stick Free,Ducted Fan Power
Centre of Gravity (Ducted Fan)
Centre of Gravity (Pulsejet)

Figure 95: Centre of Gravity and Aerodynamic Centre Excursion Diagram

With the data on the effectiveness of the elevators calculated for the stick-free
aerodynamic centre and the moment at zero angle of attack calculated earlier, trim
ability diagrams for the aircraft were drawn for three key configurations. The trim
ability diagram for the ducted fan configuration with power off can be seen below in
Figure 96.
132
Cm-Cl graph (Power Off)
-1
-0.5
0
0.5
1
1.5
2
-1.2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
Cm(ducted fan CG)
C
l
Engine Off
Engine Off, Elevator -10 degs
Engine Off, Elevator +10 degs
Alpha=0
Alpha=12 (Stall)

Figure 96:Cm-Cl Graph (Power On)

This shows that with power off, elevator deflection downwards of 7 degrees is
required to trim the aircraft at maximum lift coefficient. In cruise flight, a smaller
elevator deflection is required, in the vicinity of 4 degrees. With ducted fan power on,
the trim ability diagram can be seen below in Figure 97.
Cm-Cl graph (Power On)
-1
-0.5
0
0.5
1
1.5
2
-2 -1.5 -1 -0.5 0 0.5 1
Cm(ducted fan CG)
C
l
Engine On
Engine On, Elevator -10 degs
Engine On, Elevator +10 degs
Alpha=0
Alpha=12 (Stall)

Figure 97: Cm-Cl Graph (Power On)
133
With power on, the ducted fan based aircraft requires around 14 degrees of elevator
deflection downwards to trim at maximum lift coefficient. At cruise, the required
deflection is 7 degrees with power on. Below in Figure 98 the trim ability diagram for
pulsejet flight with the centre of gravity at its most forward point can be seen. This is
sufficient as all other centre of gravity positions will produce a trim ability diagram
between this and the diagram produced for power off, ducted fan flight (with a small
offset to the left in the case of pulsejet power on).
Cm-Cl graph (Pulsejet)
-1
-0.5
0
0.5
1
1.5
2
-2 -1.5 -1 -0.5 0 0.5 1
Cm(ducted fan CG)
C
l
Pulsejet
Pulsejet, Elevator -10 degs
Pulsejet, Elevator +10 degs
Alpha=0
Alpha=12 (Stall)

Figure 98: Cm-Cl Graph (Pulsejet)
This shows similar results to under ducted fan, power on conditions as the distance
between the centre of gravity and the aerodynamic centre is within several percent of
the mean aerodynamic chord. The elevator deflection to trim the aircraft for maximum
lift is 15 degrees while the trim deflection at approximately level flight is 7 degrees.

The overall result of this trim ability analysis shows that if manufacturing allows an
elevator deflection of at least 15 degrees downwards and 4 degrees upwards (seen
from the lower end of the diagrams) then the aircraft will be trimmable and capable of
flying at any constant angle of attack the aerofoil is capable of. It also shows that a
more efficient aircraft could be obtained by inclining the horizontal tail at an incidence
angle equal to around 5 degrees elevator deflection downwards. This has been noted
134
for future designs, however as manufacturing of the tail had commenced by this point,
this alteration could not be performed.
4.10.2 Roll Stability Analysis
Before final manufacturing, the aircrafts roll stability derivative was determined.
Although roll stability is not a necessity for any aircraft, if the aircraft is too unstable in
roll it will lead to a continually tightening turn until the aircraft is flying at 90 degrees
to level. If this occurs too quickly the aircraft will be significantly harder for the pilot to
control. The roll stability contributions from the wing-fuselage, horizontal tail and
vertical tail were calculated using empirical methods outlined by Roskam (1985f).

The wing and fuselage contribution took into account the wing geometry and the
location and size of the wing relative to the fuselage. The sweep and aspect ratio
effects produced a stable contribution which slightly outweighed the unstable
contribution from the low wing configuration.

The effect of the horizontal tail of the aircraft produced a significant stabilising effect
on the roll stability of the aircraft as a whole. This combined with a location above the
fuselage with significant sweep and low aspect ratio to produce a stabilising effect
around four times larger than the wing. The vertical tails also produced a small
stabilising effect due to the coupling of roll with yaw. The contributions of each
component can be seen below in Figure 99.
135
Roll Stability Contributions
-0.16
-0.14
-0.12
-0.1
-0.08
-0.06
-0.04
-0.02
0
C
lB
Vertical Tails
Horizontal Tail
Wing and Fuselage

Figure 99: Roll Stability Contributions
As the overall roll stability derivative of the aircraft is negative, the aircraft is stable in
roll, providing for a good test platform. For a future design to better fill a target drone
role, an aircraft that is slightly unstable in roll would provide better turning
performance but as a proof of concept design, stability in roll is suitable.
4.10.3 Ground Performance
The first consideration for suitable ground performance is that the front landing gear
must take a reasonable fraction of the aircrafts weight when on the ground so that
steering is effective. The fraction of weight applied to the front and rear landing gears
was calculated through a simple force balance between the weight of the aircraft and
two reactions at the landing gear locations. For the locations of front and rear landing
gears outlined earlier and the centre of gravity point for the ducted fan, the front
landing gear takes 22% of the weight of the aircraft. For the pulsejet and full fuel
configuration (as it will be on the runway), the front landing gear takes 28% of the
aircraft weight. Both these values were within a reasonable range for the landing gear
configuration and would allow suitable steering effectiveness.

For the aircraft to be able to rotate for takeoff climb, a suitable pitch angular
acceleration must be provided by the elevators at the point where lift equals weight of
the aircraft. To calculate this angular acceleration, Equation 27 seen below outlined by
Roskam (1985g) was used.
136

Equation 27:Pitch angular acceleration
For this analysis the angle of attack of the aircraft was measured as 3 degrees (from
the CAD model) and the elevator deflection allowed was only up to 10 degrees
(providing a margin of safety). The moment of inertia was estimated from the CAD
model available. This enabled the angular acceleration of the aircraft in ducted fan and
pulsejet configurations to be determined at the instant of takeoff. In ducted fan
configuration, the angular acceleration was 4.3 deg/s
2
while pulsejet configuration
only provided 1.5 deg/s
2
.This was a worrying result as these values are significantly
lower than those recommended by Roskam (1985g). However, as these are still
positive (allowing a rotation upwards) the aircraft will still takeoff, just at a slower
climb than ideal. For the pulsejet configuration, the engine may require throttling at
the point of takeoff to improve rotational performance (if the thrust is dropped to 1kg
then the rotational acceleration is increased to 2.2 deg/s
2
). Ideally, the location of the
rear landing gear would be brought further forward to improve rotation while still
allowing suitable front landing gear loads.

The position of ideal balance between requirements would been around 900mm
(50mm further forward of the position previously considered) from the nose, providing
17% weight on the front landing gear and 6.3 deg/s
2
of rotation in the ducted fan
configuration. This would also allow 23% weight on the front landing gear and 3.5
deg/s
2
of angular acceleration in the pulsejet configuration, allowing a better takeoff
and margin for calculation error. However, as at the stage these calculations were
performed the fuselage plans were already finalised and sent away for manufacture
with the landing gear location set these modifications could not be made.

137
5 Airframe Manufacture
The aircraft design weight was required to be kept to a minimum to allow greater
performance from the aircraft. A conscious effort was made throughout the entire
design and build process to minimise weight, while still ensuring the structural
integrity of the final product. The use of composite materials was chosen, to produce a
high strength airframe with minimal weight. All composite work was completed by the
group, with the guidance, supervision and assistance mechanical engineering
workshop staff as well as aircraft modelling enthusiasts. This working environment was
critical to the success of the project on the whole, as all members of the group had
little to no manufacturing experience. The airframe was manufactured in a number of
sections including the wings, empennage, fuselage and internal structure. On
completion of all sections, the airframe was completed through assembly.
5.1 Available Manufacturing Methods
The available manufacturing methods were a key driver throughout the entire design
process. It was important for the group to know, and be familiar with, the
manufacturing techniques that were available to them. It was also important that all
designs were able to be manufactured correctly and under budget.

The primary tool available for precisely manufacturing a wide range of parts is a CNC
machine. The machine allows CAD images to be imported and cut to shape. A CNC
machine was available through the mechanical engineering workshop, and others
were sourced external to the university.

The fuselage can be manufactured in two main ways, using either male or female
moulds. In either case, the manufacture of a plug is required, to form the desired
fuselage shape. The plug will commonly be made from either wood or foam depending
on the application. A male mould system allows the final product to be formed straight
from the plug. This is a relatively quick process, but results in a poor external surface
finish. A female mould system uses the plug to produce a female mould, which is then
used to create the final product. This is a longer process, but can produce a final
Chapter 5 Airframe Manufacture
138
product with a high quality surface finish. Female moulds also allow parts to be re-
made quickly.

A number of different techniques were considered for the manufacture the body of
the wings. Two techniques commonly used are CNC and hot wire cutting. A primary
advantage of hot wire cutting over CNC cutting is the time flexibility available. This
would increase the flexibility of the manufacturing process (no machine needed) and
would reduce production costs. The disadvantage of wire cutting over CNC cutting is a
reduction in cutting accuracy.

The most common technique used for the production of composite structures similar
to those being considered in this project, is hand lay-up production. This technique
was suitable for applying a fibreglass skin to the surface of the wing core, as well as for
constructing the outer structure of the fuselage. Hand lay up is relatively simple,
requiring manual addition of a reinforcing fibres and resin. The main reasons for using
this technique are that it can be completed in a home workshop, without requiring a
high skills base.

A range of other manufacturing approaches and techniques were suggested to the
group from a range of sources including modelling enthusiasts and business experts.
The principal approaches suggested did not differ greatly from those previously
considered. However, numerous tricks of the trade designs and techniques were
suggested for individual components of the design. These suggestions were considered
by the group.
5.2 Wing Construction
The wing structure comprised of a foam core, fibreglass skin with carbon fibre and
plywood reinforcement. The wings are connected to the fuselage using two
unidirectional carbon fibre rods as can be seen in Figure 101.

The core of the wings, made from 24kg/m
3
density polystyrene foam, was shaped
using a hot wire cutting method. Foam blocks were cut to the appropriate size, with
Section 5.2 Wing Construction
139
wing and tip profiles used to achieve the NACA 4412 aerofoil shape. The end profiles
were cut from a kitchen laminate, using a CNC machine. The profiles were used to
guide the hot wire to achieve the desired wing dimensions, including the taper and
sweep.

Due to the nature of hot wire cutting, imperfections were present in the final cut
product. The addition of taper made cutting more difficult as it required different
cutting speeds along the length of the wing. Two sets of wings were cut, with the best
set used for the aircraft and the second set used for fibreglassing practice. A
lightweight, mouldable plaster paste was used to achieve the final desired aerofoil
shape and surface finish.

Three vertical plywood ribs were installed into each wing. One at the root, one at the
tip and one just inboard of the aileron. The ribs were made from 4mm plywood and
were cut using a CNC machine. The root and centre rib had holes pre drilled, to allow
the ribs to hold and support the wing to fuselage connection carbon rods. The foam
structure of the wing was cut into two sections, to allow the centre rib to be installed.
Two holes were drilled down the length of the wing between the root and centre rib,
to accommodate the wing-fuselage carbon rods. A third hole was also drilled to allow
servo wiring to pass through the centre of the wing. Epoxy resin was used to re-join
the wing structure into a single piece. The sectioned wing is shown in Figure 100.

Figure 100 - Rib Installation in Wings
Reinforcing spars were added to support the basic structure of the wing at 20% and
80% of the root chord. Slots were cut out of the foam core to accommodate the spars
140
and ensure they were assembled flush to the surface. Epoxy resin was used to bond
the spars to the foam core. The full structure of the wings is shown in Figure 101.

Figure 101 - Wing structure schematic
The prepared wings were covered with two layers of 0/90 85gsm fibreglass cloth. The
middle layer of glass was added at 45 to increase the torsional strength of the wings.
A 50% volume of resin to fibreglass was maintained to minimise the total volume used.
The wings were vacuum bagged during drying to ensure a well bonded and smooth
finish, as well as to remove any excess resin. Finally, the wings were heated at 40
degrees for two hours to allow the resin to cure and reach maximum strength.

Figure 102 - Servo Installation
Prior to glassing, balsawood reinforcement was added either side of the aileron hinge
line. The ailerons themselves were cut from the main structure of the wing after
fibreglassing. Hinges were installed inside the balsa reinforcement, and the required
servo, pushrod and servo horn were installed on the bottom side of the wing as
illustrated in Figure 102.
5.3 Empennage Construction
The entire empennage section, comprising two vertical and one swept horizontal
surface, was produced in four separate sections. The cores for each section were hot
Section 5.4 Fuselage Construction
141
wire cut from polystyrene blocks, using the same technique as for the wings. However
blue foam, of higher density was used, as it was found to be more stable under
hotwire cutting. The two horizontal surfaces were attached using epoxy resin to create
the horizontal tail section (Figure 103a). The addition of control surfaces and glassing
of the remaining three sections was completed in the same way as the wings. The
glassed horizontal tail is shown in Figure 103b.

A half aerofoil shape was cut into the ends of the vertical tail sections to allow the
horizontal tail to allow a smooth transition from vertical to horizontal sections. The
two verticals surfaces were first attached directly to the fuselage (Figure 103c) using an
epoxy/micro-balloon mix and the horizontal tail was then attached to the verticals. As
the empennage section was permanently fixed to the fuselage, much care was taken
to ensure all surfaces were aligned correctly before attachment.

Figure 103 - a) Horizontal tail joined as a single piece, b) horizontatal tail after glassing, c) installation
of vertical tail onto fuselage
5.4 Fuselage Construction
The final shape of the fuselage was based around the design of the pulsejet engine. It
allowed for streamlining of the installed engine, sufficient fuel storage for 10 minutes
of flight, as well as room for all other auxiliary systems. The final design was carefully
planned to ensure that manufacturing could be completed with minimal
complications. Wing flanges were important for both aerodynamic performance and
for ease of manufacturing. The smooth transition from fuselage to wings was made to
ensure that mould release was able to be achieved without damaging the final
a b c
142
product. A recess was also designed in the under-belly of the fuselage, which allowed
the rear landing gear to be mounted flush with the body.

The information in the CAD package was imported into the CNC milling machine,
allowing the aircrafts plugs to be shaped accurately. Two plugs were made, one for the
port and starboard side of the aircraft fuselage. The plugs were made from Jelutong,
which is a timber commonly used for modelling due to its consistent texture, allowing
a smooth surface finish to be achieved in minimal time. After machining, each plug was
painted and sanded multiple times, until the surface was determined to be sufficiently
smooth for producing the moulds. Figure 104 shows the port plug straight after
machining.

Figure 104 - Fuselage plug
The first step in creating the moulds from the plugs is to ensure that the moulds will be
able to be removed from the plugs after the glass and resin has been applied. This was
achieved by applying multiple coats of wax to the plug surface, with a final coat of PVA
release agent. Following this, a layer of gel coat (Figure 105) was added over the
surface of the plugs, allowing the same smooth surface finish of the plugs to be
achieved in the moulds. This coat formed the inside surface of the moulds, which
inturn forms the outer surface of the fuselage. Once the gel coat had set, 6 layers of
300gsm chop strand fibreglass mat were added, thereby forming the structure of the
mould. The combination of the chop strand mat and vinyl ester resin softened the
fibreglass, making it significantly easier to mould to the desired shape. The vinyl ester
resin was also used to reduce setting times, compared to those required for most
epoxy resins. Once set, the fibreglass moulds were released from the plugs and heated
Section 5.4 Fuselage Construction
143
to allow the resin to cure, thereby significantly increasing its strength. Wet and dry
sandpaper, as well as surface polish were again used on the mould to obtain the
smoothest surface finish possible. Further layers of wax and PVA release agent were
then added to the mould surface in preparation for laying up the final fuselage.

Figure 105 - Gel coat being applied to plugs in preparation for creating the moulds
A thin coat of an epoxy resin/micro-balloon mix was chosen for the outer surface of
the fuselage for a number of reasons. It provided the smoothest surface finish, helped
to fill tight corners to allow the correct shape to be produced and also made the
sanding and cleaning of the final surface easier. Onto this coat, one layer of 0/90
85gsm and then three layers of 0/90 320gsm aircraft grade fibreglass cloth were
added. The middle layer of 320gsm cloth was added at 45 to increase the structures
torsional stiffness. A high temperature vinyl ester resin, with glass transition
temperature of 177 C was used in the fuselage to help protect it from the heat
produced by the pulsejet engine. Fibreglass rovings were added to the corners of the
flanges in the engine cut-away section for increased stiffness.
144

Figure 106 - Fuselage
After being released from the moulds, the two halves of the fuselage were joined using
a layer of 0/90 85gsm fibreglass on the outside, and a layer of 0/90 320gsm Kevlar
cloth on the inside. A layer of Kevlar was also applied to the nose of the fuselage for
added impact resistance. The resultant fuselage is shown in Figure 106. Although it
was expected that significant weight saving could have been achieved through the use
of carbon fibre in the fuselage structure, this was avoided to help prevent radio
interference between the R/C controller and receiver. Depending on the degree of
interference, the radio signal can be lost mid flight. This was an unacceptable risk and
hence was avoided.
5.5 Internal Fuselage Construction
The primary aim of the fuselage internal structure was to stiffen the structure, while
also allowing for a range of component mountings. The three aft most bulkheads were
carefully shaped carefully to allow for the installation of the pulsejet and
corresponding heat shielding. A total of six bulkheads were added and are shown in
Figure 107. Each bulkhead was shaped using a CNC machine from 4mm plywood,
covered in two layers of 0/90 85gsm fibreglass.

The bulkheads were joined to the fuselage structure using an epoxy resin/micro-
balloon mix. The shape and layout of the fuselage was designed to allow fuel bags and
Section 5.6 Internal Access
145
batteries to be attached directly to the fuselage itself, removing the need to have
additional mounting platforms. The details of each bulkhead are outlined below:
- Bulkheads One and Two were installed to allow for the addition of the
electronics tray and to attach the front landing gear. The electronics tray was
used as the mounting point for the receiver and power supply for the remote
control equipment. The use of the tray allows for vibration isolation to be
added later, as is required for pulsejet flight.
- Bulkhead Three was used as the mounting point for the pulsejet vibration
isolation mount, as well as to provide significant structural rigidity to the
fuselage, which was weakened due to the engine cut away section.
- Bulkheads Four and Five were mounted inline with the leading and trailing
edge of the wings respectively, adding structural support, while still allowing
space for the installation of the pulsejet fuel system components, namely the
fuel bags.
- Bulkhead Six, positioned at the aft end of the aircraft, was a primary point of
reinforcement for the two vertical tails.
Holes were drilled in each bulkhead and unidirectional carbon fibre longerons were
inserted the length of the fuselage as shown in Figure 107. The longerons provided
additional bending strength to the fuse and were secured using epoxy resin.

Figure 107 - Location of bulkheads (blue) and longerons (red)

5.6 Internal Access
Due to the design and layout of the pulsejet fuel system, internal access was required
for a large portion of the aircraft fuselage. This was achieved with two access areas.
146
The largest of these two access areas ran over half the length of the fuselage and was
required for installation and maintenance of the fuel bags. The fuselage was designed
with a cut-away to allow access, requiring a separate cover to make the design as
streamlined as possible. The cover was made from 0.5mm aluminium sheeting and was
moulded by the mechanical engineering workshop staff. The cover was fixed to the
fuselage using a total of 6 screws, three on the port and starboard sides respectively.

The nose hatch was required for access to the electronics tray, GPS logger and front
landing gear. A removable panel was therefore cut from the nose of the aircraft to
achieve this access. The front two bulkheads were then used as securing points, to
allow the hatch to be easily added and removed, with only two screws. Figure 108
shows the aircraft with and without the two access panels installed.

Figure 108 - The aircraft showing the both access panels a) removed and b) attached
5.7 Propulsion System
Two propulsion systems were installed at different times on the aircraft. The ducted
fan was installed for use during initial flight test, to ensure the flight readiness of the
airframe, while the pulsejet system was installed in preparation for final pulsejet
flights.
5.7.1 Ducted fan
The assembly of the ducted fan motor was critical in achieving successful operation.
The rotor-blade-housing system, shown in Figure 109, is manufactured with extremely
fine tolerances, with only a fraction of a millimetre separating the blade, spinning at
45,000rpm, from its carbon-fibre housing. The external mounting bracket also required
a) b)
Section 5.7 Propulsion System
147
careful assembly, as even the smallest deviations caused enough deformation of the
housing to produce interference with the blades.

The ducted fan was mounted to the rear of the fuselage, as close as possible to the
centreline of the aircraft to minimise the pitching moment produced. This position was
also chosen in order to ensure the aircraft behaviour is similar to when the pulsejet is
used, as the thrust lines are at similar heights. The fan itself was fixed to two mounting
blocks, which were inturn fixed to the fuselage. The mounting blocks were held in
place with epoxy resin, and the fan was secured to the mounting with four screws, as
was allowed for in the design.

Finalising the mounting angle and position of the fan was important to ensure the
thrust line of the fan was along the centre line of the aircraft. Therefore, a number of
minor adjustments were made to the surface of the mounting blocks during
installation so as to achieve the optimal mounting conditions.

Figure 109 Schubeler ducted fan (Schubeler Jets, 2008)
5.7.2 Pulsejet
Installation of a pulsejet engine is achieved through two mounting points, located at
the front and rear of the engines. The front mount (Figure 110), located on the end cap
of the engine, next to the spark plug, attaches to the third fuselage bulkhead using
three screws. This mount provides vibration and heat isolation to the fuselage which is
critical for pulsejet operation. The rear engine mount is a tab welded to the exhaust of
148
the engine, in line with the rear bulkhead. The tab extends to the bottom of the
bulkhead, where it is fixed in place with a rubber mount, to damp vibration.

Figure 110 - Front pulsejet engine mount
Mounting of the pulsejet engine also requires the installation of the engines fuel
system. This comprises two fuel bags, a pump, ESC, battery and fuel injectors. The fuel
bags are soft mounted to the bottom of the fuselage in between bulkheads 3, 4 and 5,
6. The pump, ESC and battery were all installed in between bulkheads 4 and 5, and
were again soft mounted to protect them from the engine vibration. Finally, the
injectors were able to be secured to the sheet metal which covers the majority of the
fuselage.
5.8 Landing Gears and Wheels
The rear landing gear was a carbon-fibre unit, Model F3A large, sourced from Bolly
Propellers. The landing gear was attached to the fuselage using five nylon bolts, which
were designed to shear on heavy impact, to protect the rest of the aircraft, in
particular the tail. The design of the fuselage allowed the landing gear to be mounted
flush with the surface of the under-belly.

The basic structure of the front landing gear was a pre-manufactured item, made from
spring steel. A hole was drilled into the fuselage to allow the landing gear to be
mounted inside the fuselage to the second fuselage bulkhead. A servo, also mounted
to the bulkhead, was connected to the landing gear. Once the servo was programmed
Section 5.9 Control System Installation
149
into the radio control system, it allowed the aircraft to be safely manoeuvred on the
runway. The steering system is shown in Figure 111.

Figure 111 - Front landing gear steering system
5.9 Control System Installation
As previously mentioned, allowances were made during manufacture of both the
wings and empennage for wiring to actuate the control surfaces. A total of five servos
were used, using a total of four channels on the controller. The two wing servos used
two channels, allowing them to act both dependently or independently as either
ailerons or flaps. The two tail servos used a single channel, as did the front landing
gear. All wiring for the servos was connected directly to the radio control receiver,
which was mounted to the electronics tray.

The ESC selected was mounted in the centre of the fuselage, between the wings, and
was connected directly to the ducted fan, allowing the speed of the fan to be
controlled. The ESC and two lithium polymer batteries were all soft mounted to the
fuselage, to protect them during high impacts.
150

151
6 Testing
Significant testing on both the engine and aircraft systems were conducted. The results
can be found in the following sections.
6.1 Engine Testing
Three successful testing stages have been completed over the duration of this project.
All testing was conducted at Bunker 5 leased from the Federal Government, near the
RAAF Edinburgh. Over the testing period, 100 successful tests on three different
pulsejet engines were conducted. Measurements of thrust, and fuel consumption were
made at various stages during testing. Footage was recorded via a digital camera, and
all data was recorded on a laptop computer.

The testing setup used was developed by previous students for their studies into
valveless pulsejets. A schematic of the system layout can be seen Figure 112.

Figure 112 - Test System Layout
(Coombes et al 2007)
Chapter 6 Testing
152
6.1.1 Phase One Testing
Phase one of testing intended to test the engine control systems and establish baseline
performance of the two FWE engines. As part of this, the effect of injector position on
engine performance is investigated for both engines. The fixed length FWE, which was
previously unable to sustain operation, was tested with the pipe flares developed. All
tests were conducted using propane gas for fuel.
Auxiliary system implementation and engine familiarisation
Initial testing quickly showed that all auxiliary systems such as the fuel and ignition had
been implemented correctly. During the two days of testing, the safety systems and
procedures planned and implemented performed as intended, with no safety concerns
encountered. The time spent conducting live tests provided the operators with
valuable experience about the operational characteristics of valveless pulsejet engines,
which could only be learnt during live testing. Initial problems were encountered when
attempting to obtain self-sustaining internal combustion with the engines. A rethink of
the starting procedure with relation to the physics behind the system allowed a
repeatable and reliable starting procedure to be found.

In all tests where self-sustaining operation was achieved, the data logger showed
engine thrust decreasing with time, as shown in Figure 113. This was believed to be as
a result of the gas bottle cooling, and therefore not being capable of providing the
same fuel flow to the engine. This was undesirable and ultimately affected the results.
All tests over approximately 30 seconds in length were dramatically affected by this
phenomenon. This made it almost impossible to accurately test the throttle range of
the engines.
Section 6.1 Engine Testing
153

Figure 113 Reducing thrust during extended operation
Test the importance of injector position
The tests undertaken suggested a relationship between intake position and engine
performance. Figure 114 shows a reduction in thrust as the fuel injector was placed
further along the length of the intake pipe, closer to the combustion chamber.
Effect of Injector Postion
0
0.2
0.4
0.6
0.8
1
1.2
0 20 40 60 80 100 120 140
Injector Postion (mm)
T
h
r
u
s
t

(
k
g
)

Figure 114 Effect of fuel injection position on engine performance
This result is thought to be caused by differences in air-fuel mixing in the engine. With
the injector placed closer to the opening of the intake, the fuel and air had a larger
time to mix than an injector placed closer to the combustion chamber. More testing
was required to confirm this result in later tests.
Chapter 6 Testing
154
Effect of flared pipes on valveless pulsejet engine performance
The impact the addition of flares had on the engine was clear. Without any flares, the
engine was incapable of producing self-sustaining thrust. However, with the addition
of the flares, the engine immediately obtained continual operation and thrust. As the
design of the flares allowed the lengths of the intake and exhaust pipe to be varied, a
range of configurations were able to be tested. With this set-up, the engine obtained
self-sustaining thrust in almost every test undertaken, proving the theory behind the
flare design, and their importance in pulsejet design. The results from the different
configurations however showed very little correlation between the engine
performance and the lengths of the exhaust and intake. The erratic behaviour of the
engine can be seen in Figure 115.
Effect of intake and exhaust length on thrust
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0 20 40 60 80 100 120
Exhaust extension (mm)
T
h
r
u
s
t

(
k
g
)
66 series
68 series
78 series

Figure 115 Effect of exhaust and intake length on engine performance
The range of thrusts produced by engine 2 varied from as low as 0.7kg up to 1.7kg,
which was less than the 1.8kg obtained from engine 1. This was unexpected due to the
larger physical size of engine 2. It is believed that the low thrust is a result of poor
tuning of the exhaust and intake pipes to the combustion chamber. Comparisons to
more successful designs suggests that the large diameter exhaust allows combustion
gasses to escape the combustion chamber too easily, and therefore reduces the
155
intensity of the pressure waves produced. The results from these tests clearly suggest
that there is more to engine performance than physical size.
Conclusion
The systems implemented for the tests all performed as desired. Cooling of the
propane gas bottle was thought to impede the engine performance as the bottle
cooled and lost pressure, and no solution to this problem was able to be found in the
time available. Some trends were found with regards to injector placement within the
intake pipe; however more testing was required to confirm the results. The flares
added to engine 2 allowed self-sustaining thrust to be obtained. The flares were
thought to enhance the engines ability to produce thrust, even though the engine was
not tuned correctly. The poor tuning of engine 2 meant the engine produced less
thrust than engine 1, which was smaller in size. The thrust produced by both engines
tested was too low to be capable of powering a UAV in its present state. As a result,
further engine development was required. A full testing document is supplied in
Appendix J.
6.1.2 Phase Two Testing
Phase two of testing intended to investigate the performance of the engine
modifications, as well as test the performance of the Chinese style engine. The primary
modification tested was the expanding tail pipe for the adjustable FWE engine, which
was tested at various engine lengths to determine best engine performance. The
ceramic coating on the fixed length FWE engine was also tested to determine any
changes in performance.
Effectiveness of an expanding tail pipe on an FWE engine
After the results of the first round of testing, the adjustable FWE engine configuration
failed to produce 2kg of thrust. In order to improve the thrust of this engine an
investigation into the effect of an expanding exhaust section was conducted. Tests
were conducted to compare the performance of the test engine from 2007, to the
Chapter 6 Testing
156
equivalent engine, with an expanding exhaust section. Short lengths of 50mm
diameter pipe were cut to allow for fine tuning of the exhaust section.

Figure 116 - The adjustable FWE engine with expanding tail section and 100mm extension.
Initially the expansion section was added to the original FWE engine, with an
additional 100mm extension section added to the exhaust, as shown in Figure 116. The
engine length was then increased in 50mm and 100mm intervals, to determine the
most suitable operating conditions. Based on the results, the most effective length was
then selected, and 25mm sections were added and removed in order to find the
optimal length of the engine. The results are summarised in Table 19.
Table 19 - Expanding Exhaust Test Results
Section Length
(mm) 100 150 300 325 350 375 400 450 650
Max Thrust (kg) 1 3 3.25 3.3 3.55 3 3 3.2 2

From the tests it was concluded that a total exhaust length of 1080mm was optimum
for this configuration, producing 3.55kg of thrust, achieving both the project goal of
producing an engine with 3kg of thrust and extension goal of achieving 3.5kg of thrust.
The results suggest that the engine is also less sensitive to changes in exhaust length,
which is beneficial from an engine design point of view.

Testing of the expansion section also revealed that the engines were extremely
throttle able, with numerous engine configurations still maintaining self sustained
combustion to as low as 0.5kg of thrust. This is a significant achievement, as a throttle
able engine will give an aircraft more control during flight.

157
The effect of the intake length was planned as an additional investigation, however
during the final testings, a failure of the main fuel cut-off valve occurred, which
jeopardised the safety of the remaining tests, hence the tests were aborted.
Test Chinese Engine for thrust measurement
The Chinese engine, which was manufactured after the mixed success of the first
round of testing, was tested to determine the engines maximum thrust, most effective
injector position, and throttle range.

Ten tests were performed on the Chinese engine, with the main aim of determining a
suitable injector position for producing maximum thrust. The fuel consumption and
throttle range of the engine were also of interest.

Seven different injector positions were tested, with the results shown in Figure 117. It
was noticed during testing that as the injector position was moved into the engine, the
engine became more throttle able, with the engine throttling as low as 0.8kg on
several occasions. The engine was also easier to flame out, which suggested the fuel
consumption of the engine was less, however data could not be collected regarding
this due to issues encountered with the load cell.

Thrust vs Injector Position
0
0.5
1
1.5
2
2.5
3
3.5
4
0 25 50 75 100
Injector Position from Intake Mouth (mm)
M
a
x

T
h
r
u
s
t

(
k
g
)

Figure 117 - Affect of injector position on engine thrust
Chapter 6 Testing
158
Maximum thrust was achieved with the injector positioned at 32mm inside the engine
intake, with a maximum thrust of 3.62kg achieved. The results from this test can be
seen in Figure 118.

Figure 118 - Thrust Results
A fuel consumption test was performed at this intake position, yielding a specific fuel
consumption of 5kg/kg/hr. It is expected that better results can be achieved by moving
the injector position further into the intake.
Effect of Ceramic coating on Engine Performance
In an attempt to reduce the temperature of the engine, the interior of the larger FWE
engine was ceramic coated by Ceramic Coats Australia. The ceramic coating was
intended to reduce heat transfer to the metal, in an attempt to reduce the need for
heat shielding on the aircraft.

Four tests were performed on the ceramic coated engine, with thrust results of 1.3kg
achieved, which was similar to the results from the first stage of testing. From video
footage it could be determined that the engine ran visibly cooler, however it was
noticed that damage occurred to the ceramic coating after several runs (Figure 119). It
is anticipated that the ceramic coating applied was not suitable for the temperatures
159
achieved in the pulsejet. Further investigation is required to determine if ceramic
coating is a feasible option for reducing the temperatures experienced by the engine
structure.

Figure 119 - Visible damage to ceramic coating
Conclusion
The second phase of testing was successful, with two engines achieving the projects
extended goal of 3.5kg of thrust. Specific fuel consumption figures of 5kg/kg/hr were
also promising, as they showed improvement on both results from the studies of
Coombs et al, who recorded 7.1kg/kg/hr at 1.6kg of thrust, and Enics pulsejets, which
advertise figures of 6.6kg/kg/hr for their engines.

The testing program proved that an expanding exhaust is capable of improving the
performance of an engine; however significant work must still be conducted to
determine a method for optimizing the length and expansion of these sections.

Finally the tests showed that ceramic coating is an option for reducing the heat
transfer out of the engine, with similar thrust results achieved for the engine with and
without ceramic coating. However the damage to the coating suggests that more
research into the most suitable coating for the engine must also be conducted before
this is added to the flight weight engine. The added weight of the engine must also be
determined.

Chapter 6 Testing
160
A full report of all tests can be found in Appendix H Test Log Books

6.1.3 Phase Three Testing
Phase three of engine testing was aimed at testing the liquid fuelling system developed
and readying the engine for use on the UAV platform. Various configurations of
injectors, fuels and engines were trialled throughout the testing phase.
Test the performance of the Chinese engine on liquid fuel
Initial testing was performed with the injectors placed down the intake in a similar
layout as with propane testing. While this setup is not ideal due to both excessive
restriction of the intake flow, and poor spray pattern distribution, the layout allowed
the modification of the position without altering the engine. This was done to establish
a reasonable idea of required injector position. This testing indicated that transition
from gas to liquid was not feasible, as the engine would not maintain operation as gas
flow was reduced. It was however, possible to start the engine on liquid fuel only,
providing that the engine had been brought up to operating temperature on propane.

It was later determined that transitioning could occur with the injectors placed
through the side of the intake pipe. It is thought that the inability to do so in initial
testing was due to the overheating of the injectors while the engine warms on
propane. This would dramatically reduce the flow rate of the injectors to a level that is
unable to sustain the engine operation. It was noted from later testing that the fuel
flow rates during operation were significantly less than those achieved during flow
bench testing, which is most likely also due to injector heating.

The most effective method of startup determined from testing is as follows:
Turn on ignition circuit and air supply
Turn on gas supply system and adjust to start engine
Turn off air supply and ignition circuit
161
Allow the engine to warm until the combustion chamber begins to glow (~5
seconds)
Set gas flow to medium throttle, turn on liquid at medium throttle
Simultaneously reduce gas flow and increase liquid flow, ensuring not to over
fuel the engine.
Test the effect of injector position on performance
Testing was performed on the Chinese engine to determine the optimal injector
position on liquid fuels. Flow requirements developed from equating energy quantities
between fuels suggested that 3 injectors were required for operation. The injectors
obtained were of various sizes, denoted 6, 8 and 10, with the number relating to
the size of the orifice in tenths of a millimeter.

Initial tests established that having the larger injectors closest to the combustion
chamber resulted in the best performance, both in thrust and fuel consumption. This is
thought to be due to less wastage of the fuel, as the larger injector is further from the
mouth of the intake. It is thought that having the larger injector furthermost from the
combustion chamber may allow for easier starting, however transitioning to liquid fuel
after operation is established on propane makes this irrelevant. The injector layout is
shown in Figure 120

Figure 120 - liquid fuel injectors placed mid way along the intake tube
Chapter 6 Testing
162
The effect of injector position was investigated by drilling holes along the intake tube
at regular intervals, allowing the injectors to be moved as required. The remaining
holes were sealed by fireproof fabric with a metal backing held in place with hose
clamps.

Best performance (both thrust and fuel consumption) was achieved with the injectors
placed midway along the intake tube. This is unlike the propane tests, where maximum
thrust was obtained with the injector at the mouth of the intake. It is possible that
excessive losses in fuel are the reason for reduced thrust at the outermost position, as
the pump and injector set was unable to over-fuel the engine during testing.

A larger pump was purchased in order to provide a larger fuel flow rate. However, this
did not provide significant improvements in engine performance. It is believed that
both pumps are able to supply the engine with all the fuel it required, but not enough
to make the engine flame out. To determine if fuel supply limitations were affecting
thrust output, a fourth injector was added to increase fuel flow. This did not produce
any more thrust, but did increase fuel usage, and hence it was determined that extra
fuel was not required.

In the interests of improving fuel atomization, the injectors were positioned in an
opposing configuration, such that the sprays would interfere with each other. This is
setup is shown in Figure 121. This also did not produce positive results, and hence it
was determined that the best performance was with the three injectors placed
midway along the intake tube, on the same side. Comparison of performance with the
various injector setups is shown in Figure 122.

163

Figure 121 - Opposed injector configuration

Figure 122 - Performance of the Chinese engine with different injector placements
Test the fuel mixtures the Chinese engine
In order to achieve the best engine performance, testing was done to determine the
best fuel mixture for the engine. Standard unleaded petrol was used as the bas fuel for
most testing, with additions of other fuels tested. A single test on kerosene was
performed, however, it was noted that the engine produced a lot of fuel vapor during
the test. Due to this, and a lack of any performance improvements, kerosene was
considered to be unsuitable for the application.

Methanol was trialed as an additive, as on vaporization, a relatively large cooling effect
is present. It was thought that this may improve engine performance by increasing the
Chapter 6 Testing
164
density and hence mass of air drawn into the engine each cycle. Conversely, the
addition of Shellite should increase the heat produced during combustion, and hence
potentially improve performance by creating larger combustion pressures.

The best performance was obtained from the addition of 10% methanol to standard
unleaded petrol. This addition produced similar thrust, but reduced the fuel
consumption of the engine significantly. The relatively consistent thrust levels suggest
that the engine is not very sensitive to the degree of variation between the fuels
trialed. Engine performance on the trialed fuels is shown in Figure 123.

Figure 123 - engine performance on various fuels
Test the performance of the FWE engine with expanding tail section on
liquid fuel
Testing was performed to determine the ability of the FWE engine to operate on liquid
fuel. The injector setup used was that which provided the best performance with the
Chinese engine, and the engine configuration was that which provided the best
performance on propane fuel (350mm extension on the tail section).

During the first and only test performed, sustained operation on straight unleaded
petrol was achieved, producing a maximum thrust of 1.5kg. During the test a
significant amount of un-burnt fuel vapor was released into the air and was therefore
165
considered a serious safety risk. Hence, no further liquid fuel testing was performed
with the FWE engine.
The excessive vapor release from the FWE engine on liquid fuels was a factor that had
been expected from the background research stage. It is thought that the FWE style
combustion chamber does provide the correct environment to support, which was a
primary reason for investigation into the Chinese engine style.
Engine Geometry changes
Due to the relatively poor performance achieved on liquid fuel, the engine was
modified with the aim of obtaining more thrust. To do this, the engine was cut at both
the intake and exhaust. The pipe was then either extended by wrapping with metal
sheet, or shortened by replacement of the cut section with a smaller piece.

It was found that modifications to the engine generally resulted in a reduction in thrust
from the engine. It is believed that this results from the interference on the flow due
to the irregularities in the engine walls after modification. This is especially evident
when the original configuration was retried and produced less thrust than was
obtainable previously. Based on this testing, it is thought that the intake length should
be increased slightly over the original length, as the extra 5mm produced
approximately 0.5kg of extra thrust compared to the original sized engine after cutting.
This is shown in Figure 124, which compares the original performance of the engine
with performance after it was cut, as well as the thrust variations with various exhaust
lengths. It was also noted that longer exhaust lengths were unable to achieve
sustained operation on propane fuel, but would sustain on liquid. Further testing is
required in order to determine if a greater extension of the intake will produce better
performance, and if the performance degradation post modification can be rectified by
re-welding the joins to the original.

Chapter 6 Testing
166

Figure 124 - Performance of the Chinese engine for various lengths
The following theories have been developed to explain the reduced performance of
the engine on liquid fuels:
The slower burn rate of the propane fuel means that combustion is still
occurring in the expanding section of the exhaust. This combustion creates a
positive pressure gradient, which helps to prevent separation. The expansion
angle in the engine is relatively large, so on liquid fuel, where the combustion
occurs more so in the combustion chamber, the pressure gradient is not
present, and separation occurs, causing large losses in the system. Faster
combustion may also increase the speed of the gasses at the expansion point,
further increasing the likelihood of separation.
Changes in burn rate have affected the mean temperatures in the exhaust, and
hence changed the acoustic length. This therefore means the intake is not
tuned correctly to the exhaust, reducing engine performance. This theory is
derived from the results of the tests after modification of the engine.

Conclusion
Phase three of engine testing produced positive results, with operation achieved on
both the FWE and Chinese engines. Thrust levels obtained while operating on liquid
167
fuel were significantly down on those achieved on propane fuel. The Chinese engine
was found to run well on a variety of liquid fuel mixtures, with the best performance
achieved with the addition of 10% methanol to unleaded petrol. Liquid fuel operation
of the FWE engine was obtained, however excessive fuel misting was produced, posing
a safety risk.

The optimal injector setup was found to be with three injectors placed midway along
the intake, with the larger injectors closest to the combustion chamber. The maximum
thrust achieved was 2.25kg, with a thrust specific fuel consumption of 4.8kg/kg/hr.

Modifications to the Chinese engine suggest that improved performance would be
achieved by increasing the length of the intake. Further testing is required to confirm
this, as a thrust reduction occurred as a result of cutting the engine. Airframe testing
has however, shown that 2.2kg of thrust is enough to successfully power the UAV, and
as such greater engine performance is not specifically required in order to be
implemented for flight. In current form, the engine is capable of fulfilling the goal of
flight, however, it is expected that more thrust would be required in order to achieve
the speed goals.

A full report of all tests can be found in Appendix H Test Log Books.
Chapter 6 Testing
168
6.2 Aircraft Testing

The aircraft testing consisted of three phases: ground tests to test components, ducted
fan tests and pulsejet tests. The phases of testing are summarised in Figure 125.

Figure 125 - Aircraft testing flow chart
6.2.1 Wing Structural Testing
The wings were tested for structural strength to confirm they were capable of carrying
the aircrafts weight up to the maximum load factor, the deflections of the wings
under loading was measured and compared to the calculated values, this test also
determine if there were design problems of the wing fuselage interface.

Figure 126 Load zones for wing structural testing
To ensure the strength of the wings, they were tested with applied loads
approximating those expected on the aircraft. This was completed using sandbags
applied to the wings with a distribution attempting to match the lift distribution that
would be encounter during flight. Loads started at 9kg per wing and increased to 18kg
Section 6.2 Aircraft Testing
169
which correspond to a 4.4g manoeuvre. Results of the wing structural test can be seen
below in Figure 127.

Figure 127 - Experimental Wing Deflection
The results suggested that all bending could be contributed to the wing beams and the
testing rig. Based on calculations including the aircraft stiffness, it was determined that
the mounting mechanism was suitable for flight.
6.2.2 Electrical Component Testing
The servo motors that were installed on the aircraft were tested for smooth and
consistent operation and deflection using a servo tester. The servo tester is a
connection between a battery source and the servo, with a dial that can be used to
turn the position of the servo motor. The servos that were installed worked smoothly
and with consistent deflection..

The radio transmitter and receiver were testing twice before the range test at the
flight location. Firstly the servos were connected to the receiver circuit outside the
airframe, where they operated effectively. The servos and receiver circuit was then
installed inside the airframe, where longer extension and junction cables were during
the tests the transmitter, receiver and servos again operated effectively.
Chapter 6 Testing
170

The servo motor attached to the nose steering wheel was tested. The motors torque
was found to be sufficient to turn the front wheel stationary or at speed. The
maximum amount of deflection was set, based on estimates of the amount of turn
that would provide reasonable but not excessive steering sensitivity. Limits were set
for maximum servo deflection on the controller to match the maximum deflection
possible on the control surfaces without grinding. The maximum deflections possible
on the control surfaces are shown below in Table 20
Table 20 - Maximum control surface/servo motor deflection
Surface Maximum deflection (degrees)
Ailerons 15
Elevators 20

6.3 Aircraft Pre-flight Tests
6.3.1 C.G. Test
To ensure that the aircraft was stable in the longitudinal direction, the centre of gravity
was experimentally determined, and compared to the expected values from the C.G.
Excursion diagram shown previously in Section 3.5.2. This required the aircraft to be
loaded with the systems required in the airframe for flight, before the airframe was
mounted on a pin joint that allowed the centre of gravity to be determined when the
airframe became statically stable, shown in Figure 128.

Figure 128 - C.G. Test Setup
Section 6.3 Aircraft Pre-flight Tests
171
The mounting point could be moved in the longitudinal direction to determine the
point at which no pitching moment would result from the distance between the
weight force and the reaction at the pinning point.

Figure 129 - C.G. Test Photo
The centre of gravity was determined through this method shown in Figure 129, and it
was located at 840 mm from the nose of the aircraft (ducted fan configuration), which
equates to 97% of MAC, giving the aircraft a static margin of 10% MAC. This is slightly
rearward of the expected position of the C.G. in ducted fan configuration by 41.6 mm.

This provides sufficient stability for an average human pilot. Ballast will be added at
the front of the aircraft to provide a margin of error if the aerodynamic centre of the
aircraft was away from the expected value in first flights.

However for the flight testing the batteries that are used to power the ducted fan
were separated and were moved forward, while a GPS system was added in the nose
of the aircraft these changes to the aircraft moved the C.G. forward by approximately
40mm
6.3.2 Other pre-flight checks
Radio Control Range Test
At the flight location, to check that radio control was both free from interference at
the aircraft and free from interference due to other radio signals, range tests were
performed on each flight day.

Chapter 6 Testing
172
The transmitter was positioned at up to 100 m from the aircraft, and the throttle
control was positioned to 100 % for 2 seconds with the aircraft orientated straight on
and at 45 degrees to the left and to the right.

For each orientation the ducted fan ran smoothly, indicating good radio contact
between the transmitter and receiver showing a successful range test.
Weight distribution on the landing gear
The weight distribution of the aircraft was measured, and it was found that 20 % of the
weight of the aircraft was being supported by the nose wheel while 80% was being
supported by the main landing gear. This agreed with expected and acceptable values.
Fasteners and Pre-Flight Assembly
Before each flight all fasteners were checked, and all other connections were also
checked. The servo motors and control surfaces were again checked for operation at
the flight location before each flight.
6.3.3 Location for flying
Ducted Fan Tests
Ducted fan flight tests were arranged to be conducted at the Gawler airfield where the
Adelaide Soaring Club is located. The Soaring Club has provided their airfield for use
for several University of Adelaide projects in the past and this aided in the choice of
this airfield.

The airfield has runways that exceeded requirements with two 1 kilometre long
runways, consisting of a 50 m wide central gravel runway with several long tarmac
runways of approximately 5 m width at the sides within the gravel runways. The
airfield was available for early morning tests before others needed to use the runway.
Pulsejet Tests
Pulsejet flights test were arranged to be conducted at the Adelaide Model Aerosport
club airfield at Monarto. Jet models are regularly flown at this airfield.

173
The airfield has a 400 m tarmac runway which exceeds the design take-off. There is a
large clear area around the runway providing for excellent visibility.

The loud operation of pulsejet engine restricts use in urban built up areas, however at
the Monarto airfield there is no housing nearby, and the adjoining properties include
motorcycle and rally racing facilities where loud noise production is not an issue.
Contact with the club suggested that Sunday would be the best flying day, once the
aircraft was checked for its heavy model certification, which the pilot was able to do.
6.3.4 Pilot
Murray Scott who has over 20 years experience with model aircraft including jet
models kindly offered to fly the aircraft for the tests. The expertise of a
knowledgeable and experienced model aircraft pilot was extremely helpful in
completing flight tests.
6.3.5 Engine and Flight Tests
Static Ducted Fan Test
The ducted fan was statically tested to check effective operation and to determine the
maximum thrust possible. The fan was controlled using the full set-up as intended for
the airframe. The test configuration involved the fan mounted on the airframe, the
aircraft was attached to a digital strain gauge.

The first objective of the static test of the ducted fan was to check the basic
functionality of ducted fan and if the operation was smooth and free from inference.
On the first running, the fan ran, and it ran smoothly and appeared to be balanced,
suggesting correct assembly and no balancing issues.

Further objectives were to determine the maximum thrust possible to check these
against manufacturers data, and if possible account for some installation effects.
With the ducted fan cover in place, the fan produced 2.2 kg of static thrust on
maximum throttle, less than the 3 kg expected. Without the ducted fan cover in place
the fan produced 2.4 kg of thrust, which was thought to be due to the fan being able to
Chapter 6 Testing
174
access greater air in static running, despite this the ducted fan cover would still be
used in flight tests as it would greatly reduce turbulence and interference drag around
the fuselage and into the ducted fan in flight.

To determine the frequency reached by the fan, a spectrum analysis was performed
which found the fan to be running at a maximum of 37,500 rev/min even on full
throttle settings, below the maximum angular velocity of 45,000 rev/min expected.
The amount of thrust the fan produced was consistent with the manufacturers data
for this frequency.

It was determined that the electronic speed controller (ESC) was restricting the
maximum frequency of the ducted fan. The ESC was found to limit the with 11 pole
motor to 37,500 rpm, confirming the results from the spectral analysis.

Calculations suggested that the aircraft would be able to take-off on 2.2 kg of thrust,
and an ESC capable of allowing the motor to reach higher frequency could not be
sourced from Australia.
Ground Roll Test

The ground roll test was the first test performed at the flight test location. The
objective of the ground roll test was to see if the aircraft travelled straight and was
controllable on the runway. The pilot was to accelerate gradually and the behaviour of
the aircraft was to be observed. The pilot would stop accelerating if the aircraft veered
off course. Figure 130 shows this test in progress.

Figure 130 - Ground Roll Test at Gawler Airfield
175
The results of this test suggested that the nose steering was too sensitive, this was
therefore reduced and the test was completed successfully.
Touch and Go Test
The purpose of a touch and go test is to practise a large amount of landings in a short
amount of time. Landings are generally considered the most dangerous part of flying,
and are quite risky for a new aircraft. This test, involving repeated landings and take-
off, allows a large amount of information about an aircrafts dynamics and behaviour
to be determined quickly. The test was to be performed at a very low altitude to allow
the stability and controllability of the aircraft to be tested in a situation where minimal
damage could result in the case of a failure. The touch and go test was attempted on
the same day as the ground roll test.

For this first flight test, in order to ensure extra longitudinal stability 0.5 kg of ballast
was added to the nose of the aircraft.

The results of the touch and go were that the airplane made one successful take off in
the required 70m distance. The plane drifted to the right so power was removed.
Subsequently the plane had a heavy landing in grass to the right of the runway. Thus
multiple landings were not able to be completed.

Inherent longitudinal stability appeared to be satisfactory the aircraft did not pitch
up as soon as it took off, so this indicates there are no major longitudinal stability
issues. The duration of the flight was 7 seconds. It was thought that the plane could
have continued climbing and would not have had the heavy landing, if power had not
removed. The power was only removed as a touch and go test was being attempted
it was decided the next test would involve an immediate flight quality test attempt.

There was a small amount of structural damage from the heavy landing. Nylon bolts
used to attach the main landing gear sheared off during landing. This was a design
feature of the aircraft to ensure the aircraft performed a belly landing, in order to
protect the tail from structural damage.
Chapter 6 Testing
176
Flight Quality Test
The objective of the flight quality test was to verify the stability, controllability and
manoeuvrability of the airframe. The main result of the test was that the airframe flew
successfully, and the airframe was stable, controllable and satisfactorily
manoeuvrable. The duration of the flight was 4 minutes.

The method of the test was to perform a take-off, check manoeuvrability in roll by
performing figure eights, determine whether the inherent power off stick-fixed pitch
response was nose up or nose down, perform some speed runs and then attempt a
powered landing on the first flight, and a power off (or gliding) landing on the second
flight this was to be done to test whether the aircraft could land with no throttle to
determine whether for the pulsejet flight the engine would be able to be switched off
during flight to reduce the heat.

A GPS data logger was purchased for use for this flight and it was positioned in the
bottom of the nose below the electronics tray. A plot of the flight path taken from the
GPS data logger is shown in Figure 131

Figure 131 - Plotted flight path from GPS logger
The data from the flight tests can be seen below in Table 21.
Table 21 - Flight data from GPS logger
Duration 4 minutes
177
Maximum altitude

190 m (240 m absolute)
Total distance covered 5.61 km
Top Speed 147 km/hr

The results of specific tests and other comments are detailed below:
Take-off
The aircraft took off successfully with a take-off distance of 70 m, which was within the
expected range. On review of video from the flight, it was found that the amount of
rotation of airframe required to lift-off was quite substantial and this was thought to
be a function of the rear landing gear. As the position of the landing gear notch was
fixed, in order to reduce the likelihood of the airframe tipping over on the ground, the
distance from the CG to the main landing gear wheels had been increased by reversing
the direction of orientation this gear; this was thought to be the cause of the issue.

The aircraft also veered off to the left from the tarmac part of the runway to the gravel
part, this was thought to be due to over sensitive nose gear steering which was a
common problem on jet powered model aircraft flying off tarmac runways after being
used to grass airfields.
Figure eights & roll
The figure eights were performed successfully, first at half aileron deflection and the
second at maximum aileron deflection. The lack of a rudder did not hinder the yaw
control of the aircraft. The aircraft was at an altitude sufficiently high that it was
difficult to observe the aircraft clearly enough to time the roll motion and record the
roll-rate.
Pitching moment on power-off
Chapter 6 Testing
178
Untrimmed, the aircraft displayed a desirable nose up moment when power was
removed. Once trimmed, the aircraft was neutral and didnt pitch when power was
removed.
Top Speed
A top speed run was not attempted, however from the GPS the recorded top speed
was 150 km/hr.
Landing
The landing was quite hard, and the right wingtip hit the ground which broke the
central beam between the wings. There was no other structural damage. The heavy
landing was not unexpected, as it was only the second landing for the aircraft.

The landing distance was approximately 100 metres and the flaps were not used. It
was thought that flaps could be considered for use in the up position to provide
washout in order to have less tip stall to assist in a softer landing.
Other comments
The glide angle of the aircraft was shallow (which is desirable), and the pilot stated
that for much of the flight, between tasks, the aircraft was gliding on low power.

For level flight some roll trim was required (possibly to counter-act the torque exerted
by the ducted fan) and the aircraft displayed a self pitch up quality where the pilot ran
out of pitch trim on the controller, this may have been wind related.

It was noted that visibility of the airframe was low due to the white and blue
(unpainted) colours that blended into the sky. Red, orange or yellow was suggested as
preferable colours for visibility with different colours on different surfaces to assist
visualisation of the orientation of the aircraft.

Section 6.4 Pulsejet Flight Test
179
As noted for take-off, the nose gear steering may have been too sensitive which is a
common problem on asphalt runways.
6.4 Pulsejet Flight Test
In order to fly the airframe with the valveless pulsejet engine on liquid fuel the
following had been prepared:
Flight engine demonstrated to produce sustained, continuous, operation with
excellent ability to be throttled.
Flight engine capable of 3.5 kg of thrust (extended goal) on propane, and 2.25
kg of thrust on liquid fuels.
Development of an optimal liquid fuel blend and injectors to best atomise and
supply this fuel.
A liquid fuel system with fuel bags, connectors, and pump which has been
demonstrated to produce continual, uninterrupted fuel flow.
A valveless pulsejet and airframe interfacing consisting of an engine mount to
resist heat and vibration being transferred, rear engine mount and installation
for the fuel system and injectors into the airframe.
Airframe demonstrated to be capable of take-off, stable, controllable flight and
landing when propelled by a system that produces 2.2 kg of static thrust.
Larger version of the flight engine had been designed to produce 3.5 kg of
thrust, for extended aircraft goals.
Due to restrictions related to occupational health and safety in being able to start the
engine and the impending fire ban season the expected outcomes of a pulsejet
powered flight test will be described.

Based on the similar thrust of the alternative propulsion system and the valveless
pulsejet, and the continuous operation and ability to be throttled available from the
valveless pulsejet, it is expected that the aircraft would be able to fly successfully and
based on the fuel consumption and the amount of fuel carried onboard the airframe, it
would be capable of sustaining flight for 10 minutes with thrust supplied by this
valveless pulsejet engine. The larger engine capable of producing 3.5 kg of thrust
Chapter 6 Testing
180
would propel the aircraft to over the goal of 200 km/hr, possibly reaching the
extended goal of 250 km/hr. In this way, the pulsejet testing would be split into
separate sprint and endurance tests.

6.5 Discussion of experimental results

In this section the testing program with the airframe has been presented. Preliminary
tests focused on checking components and determining parameters of the airframe
like the centre of gravity and weight distribution across the nose and main landing
gear.

Flight tests of the airframe demonstrated that it was capable of smooth, controllable
and manoeuvrable flight. The developments prepared to fly the airframe with the
valveless pulsejet engine on liquid fuel have been summarised, along with the
expected results.

181
7 Management
A project management role was initiated during the preliminary stages of the project.
The aim of this position was to ensure the project remained on time, on budget and to
ensure that all risks were managed appropriately. The project was managed using
several tools, including Gantt charts, Microsoft Excel for maintaining control of the
project budget, and a risk management strategy, which was utilised to identify, control
and monitor all risks associated with the project.

During the initial stages of the project, a basic outline of deliverables and target dates
were developed to help create a suitable project timeline. These dates were set as
milestones using Microsoft Office Project. As specific goals and tasks were issued, the
project plan was developed. The timeline was developed utilising start no later than
and finish no later than constraints. This ensured that most tasks relied on the
completion of another, to ensure the project continued to progress successfully.

The group was given an allowance of $2,000 per student from the School of
Mechanical Engineering, totalling in $12,000. This was used to cover the
manufacturing costs of the project. To determine an approximate cost of the project, a
projected budget was prepared. Risk factors were included into the budget to include
factors such as rebuilding of the aircraft, express shipping of key components and high
manufacturing costs. Based on the results of this study it was determined that more
sponsorship was required to ensure the project would be completed successfully. An
initial attempt to achieve sponsorship from a number of defence related companies
showed reasonable success, with presentations prepared for BAE Systems, and
successful recruitment of ASC and Australian Aerospace, who offered the group $2,000
and $1,000 respectively. A grant was also prepared for the Sir Ross and Keith Smith
Fund, who generously offered to sponsor the group to the sum of $12,000, this
ensured the projects financial stability.

Chapter 7 Management
182
The development of a prototype pulsejet powered UAV has numerous risks associated
with it. These risks include developmental and manufacturing risks, which affect the
projects critical path line, as well as occupational health and safety risks, which are
associated with the testing of the engines and handling of fuel. As mentioned earlier
risks were also included into the financial plan, in order to determine a suitable project
budget.
7.1 Time Management
The timeline for the project was managed using Microsoft Project to maintain control
of both internal and School set deliverables. Gantt charts were used to determine the
allowable length of tasks, as well as the personnel who were responsible for that task.

Time management became exceptionally crucial during the manufacturing stage of the
project. Significant delays were experienced during the manufacture of the fuselage,
due to the use of an external contractor. This delayed the project by over one month,
before it was taken and manufactured at university by the project team with the
assistance of Bill Finch.

The Gantt charts used throughout the project can be found in Appendix J- Gantt
Charts. An example of the overall project timeline can be seen in Figure 132.
Section 7.1 Time Management
183
ID Task Name Dur ation Start Finish
1 Research and Benchmarking 31 days? Mon 3/12/07 Mon 14/01/08
2 Pulsejet Background 31 days? Mon 3/12/07 Mon 14/01/08
7 Air craft Research 16 days? Mon 3/12/07 Mon 24/12/07
11 Air craft Prelim Design 43 days? Tue 25/12/07 Thu 21/02/08
18 Pulsejet Design 82 days? Mon 21/01/08 Tue 13/05/08
33 Air craft Detailed Design 67 days? Fri 22/02/08 Mon 26/05/08
43 Pulsejet Testi ng 69 days? Mon 21/04/08 Thu 24/07/08
48 Manufacturing 135 days? Tue 1/04/08 Mon 6/10/08
49 Wings 109 days? Tue 1/04/08 Fri 29/08/08
56 Fuselage 81 days? Tue 27/05/08 Tue 16/09/08
67 Tail 26. 5 days? Thu 14/08/08 Fri 19/09/08
73 Air craft 11. 5 days? Fri 19/09/08 Mon 6/10/08
80 Air craft Testing 8 days? Tue 7/10/08 Thu 16/10/08
3 10 17 24 31 7 14 21 28 4 11 18 25 3 10 17 24 31 7 14 21 28 5 12 19 26 2 9 16 23 30 7 14 21 28 4 11 18 25 1 8 15 22 29 6 13 20 27 3
Dec '07 Jan '08 Feb '08 Mar '08 Apr '08 May '08 Jun '08 Jul '08 Aug '08 Sep '08 Oct '08 Nov '08

Figure 132-Compact Gantt Chart
184
7.2 Financial Management
The project required a level of finance in excess of a standard final year project and so
additional sources of income were required. As such, additional sponsors were sought
to fill the budgetary shortfall. Thanks to the generous support of private companies
Australian Submarine Corporation and Australian Aerospace the project received
additional inputs of $2000 and $1000 respectively. Despite this capital the project still
required a major sponsor to fill requirements and The Sir Ross and Sir Keith Smith
Funds generous support provided an additional input of $12000 to the project. This
enabled a cash budget of $16000 for the project with $12000 in kind support from the
University of Adelaide.

As all manufacturing was completed in house much of the cash expense of the project
was reduced, however this increased the hours worked by the group members (Table
22), so this was taken into account when producing the final cost analysis of the
project.
Table 22- Hours Worked By Group Members
Hours Total Cost
Karn 446.5 $29,767
Michael 484 $32,267
Mitchell 565 $37,667
Nick 799 $53,267
Ryan 631 $41,646
Terry 613 $40,867
Total 3538.5 $235,479

The cost breakdown of the aircraft can be seen in Figure 133, This represents the cash
spent by the group on components and external manufacturing. In total XXX was spent
on workshop manufacture, which can be contributed to time spent manufacturing the
fuselage in house.

Including approximate times spent by the students on design manufacture and testing
brings the total expenditure of the project cost to $245 000.
Section 7.3 Risk Management
185
$1,083.57
$2,353.85
$543.66
$1,920.68
$580.30
Engine Manuf acture and Testing Costs
Fuselage Manufacture
Wing and Tail Manufacturing
Ducted Fan
Electronics + Control Systems

Figure 133-Cost Breakdown
7.3 Risk Management
A risk assessment was made to determine key areas of the project which could have an
effect on the final outcome. Based on the analysis a risk management plan was
developed based on Australian Standard AS/NZS 4360:1999. The risk register
developed can be seen in Appendix K- Risk Register

It can be seen from the risk register that the project has a high level of risk associated
with its development. Throughout the project numerous other risks were identified
which effected the timeline of the project. The risk register was invaluable in
identifying future risks and ensuring they were avoided or mitigated appropriately.
186

187
8 Conclusion and Future Work
This project has aimed to design and build a pulsejet UAV. The goals set for the project
were ambitious, and the project has been considered a success. This section contains
an analysis of project goals, concerns encountered during the project as well as
outlining potential future work and recommendations.
8.1 Review of project goals
The various primary and extended goals of the project were the basis of all the work
performed. The goals were completed to varying extents, with most uncompleted
goals within the capability of the UAV, but not obtained due to various reasons.
1. To modify and manufacture a valveless pulsejet, with the aim of producing
3kg of thrust, with an engine weight of 1.5kg or less. This goal will be
quantified by the output received from the thrust measurement stand
constructed during the 2007 Project.
The development of the Chinese style pulsejet engine occurred over the duration
of the project. The resulting engine was capable of producing well over 3kg of
thrust, as tested on the thrust measurement stand, The engine weight was also
well below the requirement, at just 760gm.
2. Develop a liquid fuelled system for a pulsejet engine and integrate a flight
weight version into the UAV design.
A liquid fuelling system for the Chinese pulsejet engine was developed over the
engine testing period. The fuel system was designed to be integrated into the
fuselage, and is capable of providing fuel to pulsejet during all flight stages.
3. Based on the desired pulsejet specifications; design, develop and build a
lightweight UAV capable of sustaining flight for 10 minutes with thrust
supplied by a valveless pulsejet engine.
Chapter 8 Conclusion and Future Work
188
The airframe was designed, developed and built based on the anticipated
performance of the pulsejet engine. The airframe was flight tested with a ducted
fan in place of the pulsejet, which provided similar thrust and weight
characteristics to the pulsejet engine. From this test it was concluded that the
airframe is capable of sustaining 10 minutes of flight on the developed pulsejet
engine.
4. Achieve a cruise speed of over 200km/h. As measured by onboard GPS or a
similar system.
Due to limitations in thrust achievable on the ducted fan, this goal was not
obtainable. A maximum speed of 147km/h was achieved with only 2.2kg of static
thrust, significantly less than the 3.5kg used to produce the performance goals. It
is expected that the non-static thrust performance of the pulsejet will allow for a
cruise speed of significantly higher than was obtained with the ducted fan.
5. Achieve a flight time of 10 minutes
Due to time constraints, an endurance flight was not conducted. The flight test
conducted had a duration of over 4 minutes, and was completed with less than
30% of the battery capacity. This goal was not obtained, but is well within the
capability of the airframe in its current state.
6. Gain a better engineering perspective on the workings of pulsejets, with the
aim of developing different engine design alternatives.
The dynamic behaviour of the pulsejet engine was the focus of extensive
investigation for the duration of the project. At the completion of the project, a
solid understanding of the behavioural aspects of the engine has been developed,
which was used to develop the performance of the Chinese engine.
Section 8.1 Review of project goals
189
8.1.1 Extension Goals
Completion of the ambitious extension goals was also completed to varying degrees
over the project.
1. Achieve 3.5kg of thrust from a valveless pulsejet engine.
Peak thrust of 3.6kg was achieved from the Chinese engine on propane fuel during
testing. 3.5kg of thrust was also achieved from the adjustable FWE engine after the
development of the expanding tail section. The performance of both the developed
engines is sufficient for the completion of this extension goal.
2. Achieve a cruise speed of 250 km/h or above.
Due to limitations in thrust achievable on the ducted fan, this goal was not
obtainable. If a correctly designed duct was implemented on the airframe, and an
upgraded speed controller was purchased, it is likely that the maximum speed of
the UAV would be significantly improved.
3. Increase flight time of the proposed liquid fuelled pulsejet UAV to over 15
minutes.
The thrust specific fuel consumption of the pulsejet engine was in the order of 60%
of the value used for performance calculations. Considering this, in cruise
conditions, it is quite likely that 15 minutes of flight time would be obtainable for
the Chinese engine.
4. Manufacture an alternative engine design for future development.
The investigations into the Chinese style engine for use on the UAV lead to the
manufacture of the alternatively designed engine. In order for improved
performance in a UAV application, the engine requires further development,
particularly in the area of performance on liquid fuels.
190
8.2 Project Concerns
Manufacturing Experience
The manufacture of the PENGUIN was completed primarily with composite materials.
Due to a lack of previous manufacturing experience, the group were required to either
outsource the manufacturing, or obtain guidance from industry professionals. Initially,
outsourcing was attempted, however significant delays were incurred, which
threatened to prevent the overall success of the project. It was therefore decided by
the authors to complete all other manufacturing in-house.
Airframe Weight
When working with composite materials, significant experience and care is required to
ensure all manufacture is completed within the designed weight allowances. The
airframe was designed with a maximum takeoff weight of 8kg. The importance placed
on manufacturing the airframe as light as possible, without affecting the structural
integrity, resulted in a final takeoff weight of 6.5kg; almost 20% below maximum. This
gave the authors flexibility during final installation of componentry, to ensure
characteristics of the aircraft such as the centre of gravity were optimised for flight.
Engine Performance
The design thrust for the aircraft was set at 3kg for an aircraft takeoff weight of 8kg.
The maximum thrust obtained from the flight engine, running on liquid fuel, was
2.25kg; 25% below design. Concerns were initially raised regarding the ability of the
aircraft to take off with such power. However, the lower final weight of the airframe
inturn reduced the thrust required for takeoff. The ducted fan, installed to test the
performance of the airframe, produced a maximum thrust of 2.2kg during static
testing. The successful flight achieved on the ducted fan strongly suggested that
pulsejet flight would also be successful.
Pulsejet Flight
Section 8.3 Future Developments and Recommendations
191
The successful flight of the aircraft using pulsejet propulsion was not achieved during
the duration of this project primarily due of a range of occupational health and safety
reasons. Significant restrictions were placed on the authors, relating to both the high
temperature and noise level produced by the engines, which delayed the final
installation and flight using the pulsejet engine. Further restrictions were also
encountered due to fire concerns at the airfield, which would require the presence of
the Country Fire Service. However, all flight systems were designed in preparation for
flight, to minimise delays once all other safety concerns had been dealt with.
8.3 Future Developments and Recommendations
This project has shown that a pulsejet powered UAV is a feasible option for target
drone applications. There are nonetheless several areas of development for any future
designs following on from the work undertaken in this project. These have been split
into the areas of airframe design, manufacturing and engine development.
Airframe Design
Design of the airframe was significantly limited by the imposition of stall speeds and
takeoff distances suitable for wheeled launch. Although this was required to reduce
the complexity of what was a proof of concept aircraft, significant improvements in
cruise performance could be achieved through a launching system and alternate
landing method such as a parachute. These would allow a lower maximum lift aerofoil
to be chosen and for maximum cruise speed to occur closer to the maximum L/D for
the aircraft and hence significantly reduce drag.
The aircraft had a large longitudinal centre of gravity range, depending on fuel loads.
This could be reduced by pressurising the fuel or allocating more vertical space for fuel
bags. This would enable a smaller tail and hence more aerodynamically efficient
aircraft.
The effect of downwash on the tail design was significant, requiring a modified T-tail to
move the horizontal tail away from the downwash of the wing. The relatively aft
location of the wing due to the rearward engine weight also contributed to this.
Considering the wing location, a canard design should be considered for any future
192
designs with a similar engine and fuel location, improving aerodynamic efficiency and
reducing downwash effects.
The possibility of including a deployable payload such as flares or chaff should be
considered for target drone applications. This would allow a more flexible and
effective target drone and could probably be achieved with only a small reduction in
cruise speed.
An autopilot system could be integrated within the design to improve the aircrafts
usefulness as a target drone. This would allow the PENGUIN to be used without direct
control input, improving the safety of use.
Any future pulsejet engines composite UAVs should be designed with heat shielding of
the airframe a prime consideration in pulsejet mounting. The compromises inherent in
this design lead to the tail mounting close to the pulsejet and suitable heat shielding
being improvised for the situation. A canard or H-tailed design would have reduced
this danger and allowed the aluminium heat shielding to protect the fuselage only.
Stress analysis of composite components was learned by the project group during this
project. It is recommended that more subjects incorporating this matter are included
in the School of Mechanical Engineering curriculum to better prepare students for any
projects including composites and their future employment.
Manufacture
Fuselage manufacturing was outsourced to a subcontractor and was subsequently met
with significant delays. After more than two months, the mould produced by the
contractor were taken back in-house and project members completed manufacturing
of the fuselage. It is recommended that future projects use in-house manufacturing
techniques for any major components to avoid the delays inherent in subcontractor
usage.
Engine Development
Studies undertaken during this project have shown that augmenters can improve
thrust of pulsejet engines by up to 100% without increasing fuel flow required.
However, detailed design was not considered applicable to the airframe designed..
Section 8.3 Future Developments and Recommendations
193
Future projects using pulsejet engines should consider the possibility of using
augmenters during airframe design due to the significant thrust specific fuel
consumption improvements achievable.
Further development of liquid fuelling systems to improve the tuning of the engine
used and hence improve thrust and efficiency should be considered by future projects.
The engines initially tuned for propane used in this project showed a thrust reduction
of X% when run on liquid fuels due to differing burn rates.
To reduce the impact of engine performance during liquid fuelled operation, an
investigation into the operation and installation of a liquid draw propane system
should be undertaken. By using a lightweight pressure cylinder, drawing fuel in its
liquid state, and using the available engine heat the fuel to its gaseous state, could
eliminate the need for a conventional liquid fuel system and the associated reduction
in thrust.
Ceramic coating of the flight engine should be further investigated to reduce the heat
transfer from engine to airframe. This was briefly considered by this project but
neglected due to time constraints.
Further development into a better theoretical understanding of pulsed combustion,
building on the achievements of this project is another area that requires additional
development. This would enable better tuning of liquid engines and probable further
performance increases.
It would be advantageous to develop a system for starting the engine effectively
without direct input. This would improve the safety of the aircraft by reducing the risk
to operators in the case of a pulsejet failure.
Chapter 8
194
References
195
References
Abbott, I, von Doenhoff, A and Stivers, L 1945, Summary of Airfoil Data, National
Advisory Committee for Aeronautics (NACA), Washington
Abbott et al, 1959, Theory of Wing Sections, Dover Publications Inc.
Agency for Toxic Substances and Disease Registry (ATSDR). 1995. Toxicological profile
for fuel oils. Atlanta, GA: U.S. Department of Health and Human Services, Public
Health Service
AMT 1998, Technical info of the KB-70 and the used pulse jet engine, Advanced Micro
Turbines, <http://pulsejet.amtjets.com/index1.html>, last viewed 8/5/08
AMT Netherlands B.V, 2008, AMT Netherlands retail pricelist,
<http://www.amtjets.com/ last viewed on 6/5/08 at 10:32pm>
Avalakki, N.U, Bannister J, Chartier B.J.J, Downie T.M, Gibson B.A.A, Gottwald C.A,
Moncrieff P.I, Williams M.S, 2008 ,ISOAR Search and Resuce UAV, University of
Adelaide
Beck, E., 2005, Dyna-Jet Pictures and Video, <http://www.beck-
technologies.com/enginedynajet.html, last viewed on 5/5/08 at 2:48pm>
Beer, F., Johnston, R., & DeWolf, J, 2006. Mechanics of Materials: 4th Edition. US:
McGraw-Hill.
Blair, G and Cahoon, M 2006, Best bell, Race Engine Technology magazine, September
2006 edition, High Power Media, Somerset
Blast Wave Jet Corporation 2004, Plans For Pulsejet Engine, Blast Wave Jet
Corporation, last viewed 8/5/08, <http://blastwavejet.com/pulsejet.htm>
Caltex Australia Petroleum Pty Ltd, Product Glossary, viewed May 7, 2008.
<http://www.caltex.com.au/products_glo.asp>
Civil Aviation Saftey Regulations, 1998 Civil Aviation Safety Regulations Section 101,
Unmanned Aerial Vehicles and Rocket Operations, Civil Aviation Safety
Authority.
ConocoPhillips Company, 2004, Glossary of Terms, viewed May 8, 2008.
<http://lnglicensing.conocophillips.com/about/glossary/index.htm>
References
196
Coombes, J., Hollands, M., Jones, J., Matthewson, T., and Smith, R., 2007, Level IV
Design Project 2007: Design and Build of a Pulsejet Engine and Thrust
Measurement Stand
Defence Update, 2007, Advanced PulseJet Vertical Lifters, < http://www.defense-
update.com/events/2007/summary/mdm07_vertical.htm, last viewed 5/5/08
at 12:34pm>
Ducted Fans.com 2008, Lehner motors, United States, viewed 28/10/08,
<http://www.ductedfans.com/electric_motors_Lehner.html>.
Eger S 1983, Proektirovanie Samoletov, Moscow.
Enics, 2006a, , (Translation: ENICS PRODUCTS), 2006,
<http://www.enics.ru/product, last viewed on 5/5/08 at 2:32pm>
Enics, 2006b, E95M, (Translation: Ariel Target E95M), 2006,
<http://www.enics.ru/detail?product_id=22>, last viewed on 5/5/08 at 2:39pm
Euromodels, 2005, Jeti 6 Amp Microprocessor Motor Controller,
<http://www.euromodels.com.au/shop/index.php?main_page=product_info&
cPath=8_9&products_id=291, last viewed on 7/5/08 at 12:00pm>
Experience Festival 2008, Pulse jet engine History, Experience Festival, last viewed
8/5/08, <http://www.experiencefestival.com/a/Pulse_jet_engine_-
_History/id/4691790>
Falconet, 2001, <http://www.inetres.com/gp/military/ar/uav/Falconet.html last
viewed on 6/5/08 at 10:40am>
Federal Aviation Authority, 2008, Federal Aviation Regualtions, Section 23 (FAR-23),
Fine Design RC 2008, Lehner motors, United States, viewed 28/10/08,
<http://www.finedesignrc.com/cars-trucks-motors.asp>.
Goebel, G 2003, The V-1 Flying Bomb, Axis History Factbook,
<http://www.axishistory.com/index.php?id=1362>, last viewed 8/5/08
Hoak, D et al, USAF Stability and Control Datcom, Flight Control Division, Air Force
Flight Dynamics Laboratory, 1978
Hobbycity.com 2008, ZIPPY-R 4800 mAh 25C 5S1P, United States, viewed 28/10/08,
<http://www.hobbycity.com/>.
References
197
Irvine, J, Irvine Aeropulse Labs 2007, Pressurised Liquid Fuel Injection in Valveless
Pulsejets
Jennings, G 1987, Two-stroke Tuners Handbook, HP Books, Tucson
Kailasanathan, K, Experimental Investigation of pulsejet engines North Carolina State
University
Mcalley, CT, 2006, Liquid Fuel Development of a Pulsejet Engine, North Carolina State
University
Mills, A 1998, Heat Transfer, Prentice Hall, New Jersey
Model Flight, 2007, Spektrum DX7
<http://www.modelflight.com.au/spektrum/spektrum_dx7.htm> last viewed 29/10/08
Munson, H. & Young, D. & Okiishi, H. 2006, Fundamentals of Fluid Mechanics, 5th
Edition
Munson, K, 2001, Janes All The Worlds UAVs, Janes Information Group
Naughton, R 2001, The Aerial Target and Aerial Torpedo in Germany, Monash
University, last viewed 8/5/08,
<http://www.ctie.monash.edu.au/hargrave/rpav_germany_fiesler.html>
Parsch, A., 2003, Directory of U.S. Military Rockets and Missiles, Appendix 1: Early
Missiles and Drones, KDH, <http://designation-
systems.net/dusrm/app1/kdh.html, last viewed 5/5/08 at 12:48pm>
Press Digital, 2007, eBonTek DL-3200BT Bluetooth GPS Datalogger,
<http://www.pressdigital.com.au/ebontek-dl3200bt-data-logger-51ch-p-
5558.html, last viewed on 18/4/08 at 1:00pm
R. Alan Wallis, 1983, Axial Flow Fans and Ducts, John Wiley and Sons, Hoboken
Raymer, D 1992, Aircraft Design: A Conceptual Approach Second Edition, American
Institute of Aeronautics and Astronautics, Ohio
Raymer, D 2006, Aircraft Design: A Conceptual Approach Fourth Edition, American
Institute of Aeronautics and Astronautics, Reston
Roskam, J 1985a, Airplane Design Part 1: Preliminary Design of Airplanes, Roskam
Aviation and Engineering Corporation, Kansas
References
198
Roskam, J 1985b, Airplane Design Part 2: Preliminary Configuration Design and
Integration of the Propulsion System, Roskam Aviation and Engineering
Corporation, Kansas
Roskam, J 1985c, Airplane Design Part 3: Layout Design of Cockpit Fuselage, Wing and
Empennage: Cutaways and Inboard Profiles, Roskam Aviation and Engineering
Corporation, Kansas
Roskam, J 1985d, Airplane Design Part 4: Layout Design of Landing Gear and Systems,
Roskam Aviation and Engineering Corporation, Kansas
Roskam, J 1985e, Airplane Design Part 5: Component Weight Estimation, Roskam
Aviation and Engineering Corporation, Kansas
Schrenk, O, 1940, A Simple Approximation Method for Obtaining the Spanwise Lift
Distribution, NACA TM-948,
Schuebeler-Jets 2008, Schuebeler DS-51-DIA HDT, Germany, viewed 29/10/08,
<http://www.schuebeler-
jets.com/index.php?option=com_content&task=view&id=13&Itemid=50>.
Southeast Fisheries Science Centre (SFSC), viewed May 7, 2008.
<http://www.sefsc.noaa.gov/MSDS.jsp>
Tao, G. 2007 Numerical Simulations of Pulsejet Engines North Carolina State
University
Tharpe, B 2005, BTE Double Whammy, BTE Models, last viewed 8/5/08,
<http://www.btemodels.com/doublewhammy.html>
Torenbeek, E 1982, Synthesis of Subsonic Airplane Design, Kluwer Academic Publishers,
The Netherlands
Warwick, G 2003, Boeing Seeks Cash To Pursue VTOL Experiment, Flight Global, last
viewed 8/5/08,
<http://www.flightglobal.com/articles/2003/11/04/173384/boeing-seeks-cash-
to-pursue-pulsejet-vtol-experiment.html>
Werrel, K 1985, The Evolution of the Cruise Missile, Air University Press, Alabama

Appendix A - Configuration Selection
199
This section details the selection of the initial configuration of the aircraft. The results
of this section will form the basis for the continued development of the aircraft. The
advantages and disadvantages of each configuration have been considered based on
the experience of the group and consultation with common aeronautical references.
Considered in the configuration selection will be the:
General configuration,
Fuselage configuration,
Engine configuration,
Engine position,
Wing configuration and the
Empennage configuration.
Decision Methodology
The configuration of the aircraft will be split into areas of configuration, with decisions
that impact on other areas most considered first. Various options will be considered
for each area, with sketches and description of major advantages and disadvantages of
each option. The final decision on each area will be provided by use of a decision
matrix, with area specific factors rated on level of importance from the number of
factors down to one, with one being the area of least importance. The options in the
area will then be ranked from best to worst in each factor, with resultant scores from
the number of options to one, one being the worst option. The final score for each
option will then be calculated by multiplying the options rating for each factor by the
factors weighting. The highest scoring option is hence the best overall and so shall be
selected.
200
General Configuration
The general configuration entails the overall wing to fuselage relationship. A
conventional monoplane configuration has wings as primary lifting surfaces either side
of the fuselage and a smaller tail to provide stability. This is simple, proven, effective
and reduces the amount of materials needed compared with many more radical
designs. As a result, this design scored highest for pulsejet suitability and relatively well
for other areas. A biplane configuration is similar to a monoplane but uses two wings,
mounted on top of each other. The biplane design has higher lift but greater drag than
the monoplane and much of the theoretical lift is destroyed by interference between
the wings. This design scores relatively highly for stability but poorly for weight and
drag. A flying winged design uses only a wing, with no fuselage and little or no tail. A
flying winged design is generally more aerodynamically efficient than a conventional
aircraft but due to the entire surface generating lift and the reduction or negation of
tail structures the flying winged design lacks stability. This design scored maximum
ratings for complexity, weight and drag while scoring poorly for stability. A canard
aircraft uses the rearward wing as the primary lifting surface, with the front wing used
to provide both lift and control. A canard design can prevent stall of the main wing and
allow smaller total wing area but will increase the wing area vulnerable to pulsejet
heat and provide instability in pitch movements. The canard design scored reasonable
ratings in all areas, with high ratings for weight and drag.

A three surface design would provide a canard like wing configuration but also include
a tail for stability. This design allows a moment to be produced by surfaces acting in
opposite directions either side of the main wing, allowing greater stability. However,
this design would be heavier and more complex than others mentioned previously in
this section. The three surface design scored maximum points for stability but
reasonable to poor ratings in other areas. A joined wing design has two wings, a front
wing swept backwards and a rear wing swept forwards that connects to the front wing,
enhancing structural strength and reducing weight required compared with a
conventional biplane. This is a complex concept however, and has increased
interference drag while the joined wings would interfere with potential placement of
201
the pulsejet. The joined wing concept scored poorly in most areas, providing good
stability but poor suitability to the application and complexity. The general
configuration will affect the design of the wings, tail and fuselage equally, depending
on which design is chosen. The general configurations have been compared in the
areas of suitability to pulsejet usage (amount of exposed surface near where a pulsejet
may be mounted), complexity (simple designs are preferable) and weight (different
designs tend to heavier or lighter than others). Stability (some designs have inherent
stability issues or are more likely to have stability issues) and drag (both likely induced
drag and profile drag depend on general configuration) are also considered. A
summarised decision matrix is shown in Table 23.
Table 23: General Configuration Decision Matrix
Pulsejet
Suitability Complexity Weight Stability Drag Total
Weighting 3 1 4 5 2 N/A
Conventional
Monoplane 6 5 4 4 5 69
Conventional Biplane 4 4 1 5 1 47
Flying Wing 5 6 6 1 6 62
Canard 3 3 5 3 4 55
Three Surface 2 2 3 6 3 56
Joined Wing 1 1 2 4 2 36
As seen in the above decision matrix, the conventional configuration provides the best
balance between performance and pulsejet suitability.
Fuselage Configuration
The fuselage configuration entails how the fuselage is designed to provide for storage
of required components and fuel. A conventional fuselage uses a roughly tubular
shape to achieve simplicity and low weight. A conventional fuselage scored highly in all
areas rated. A twin fuselage design uses multiple fuselages, using two fuselages that
may be similar or different depending on requirements. This type of design may allow
202
for greater fuel storage on a second fuselage or the carrying of a specific payload but
will nearly double the amount of material and weight required for the fuselage with
little benefit for this project. A twin fuselage design scored poorly in all areas due to its
unsuitability to an UAV design. A twin boom design uses a conventional fuselage up to
a point where it splits into two booms which carry the tail section. These are usually
heavier than a conventional fuselage-tail combination with the major benefit being to
allow a pusher propeller or move the centre of gravity of the aircraft. A twin boom
design scored reasonably in all areas except complexity where it scored poorly. The
fuselage configuration will have a major effect on the mounting of the pulsejet and the
empennage selection. The fuselage options have been rated in the areas of cost
(materials required to manufacture the fuselage are significant financially), complexity
(a simple design is best) and weight (the weight of the fuselage is significant in the
overall aircraft weight) as it can be seen in Table 24.
Table 24: Fuselage Configuration Decision Matrix
Fuselage Configuration Cost Complexity Weight Total
Weighting 2 1 3 N/A
Conventional Fuselage 3 3 3 18
Twin Fuselage 1 2 1 7
Twin Boom 2 1 2 11
The above decision matrix shows that a conventional fuselage is the best option,
primarily for its simplicity and low weight.
Engine Configuration
As stated in the project brief a pulsejet must be used as the propulsive device. A single
pulsejet is the simplest option, allowing attachment on the empennage. A single
engine design scored highly in all areas except performance. Twin pulsejets could be
used instead, with one mounted on each wing or on the empennage. Two pulsejets
would increase the speed achievable by the aircraft. The unsteady combustion and
non-uniformity in construction of twin engines may introduce additional stability
issues while the use of two engines would increase the chance of one engine failure.
203
The twin engine design scored highly in performance but poorly in all other areas. The
engine configuration will affect the empennage, wings and other components which
are varied depending on performance expectations. The engine options have been
rated in the areas of cost (manufacturing engines and mounts as well as likely fuel
costs) and reliability (chance of an engine failing and causing the aircraft to operate in
an unforseen manner). The engine options are also rated on weight (weight of engine
and mounting is a significant part of the aircrafts total weight), heat (the amount of
heat the aircraft produces and the amount of additional heat absorbing material
needed to ensure safe operation of the aircraft) and performance (power and lift
generated by the aircraft with the option used). A summarising decision matrix can be
seen below in Table 25.
Table 25: Engine Configuration Decision Matrix
Engine
Configuration Cost Reliability Weight Heat Performance Total
Weighting 2 3 4 1 5 N/A
Single Engine 2 2 2 2 1 25
Twin Engine 1 1 1 1 2 19
The above decision matrix shows that a single engine design is preferable for the
reasons of reliability, weight and cost and so shall be used.
Engine Position
The top mounted placement is selected as this is consistent with other pulsejet
powered UAVs and ensures low likelihood of the pulsejet being damaged in landing.
The pulsejet will sit partially inside the fuselage to reduce drag.
Wing Configuration
The wing configuration entails all wing parameters, including critically shape, height
and sweep.
204
Wing Shape
The wing shape can be rectangular, linearly tapered, and elliptic or delta winged. A
rectangular shape is simple and cheap but is least efficient. A linearly tapered wing
provides better performance than a rectangular wing at a somewhat greater cost. An
elliptic wing provides the best possible aerodynamic efficiency but is difficult and
costly to manufacture. Delta wings have improved performance in transonic and
supersonic regions. They can allow less wing structural weight than a similar swept
wing and increased internal volume (again, compared with a swept wing). However,
the weight advantages are due to the added structural weight of a swept wing,
meaning it is no lighter than an unswept wing and rather more costly. A delta winged
design would require more length than an aft swept or non-swept design due to the
increased root chord. Wing shape will affect the design of all other wing
characteristics. The wing shapes will be rated in the areas of performance (efficiency at
relatively low Mach number), cost (manufacturing and material) and complexity
(simple designs are best) as shown in Table 26.
Table 26: Wing Configuration Decision Matrix
Wing Configuration
Shape Performance Cost Complexity Total
Weighting 3 2 1 N/A
Rectangular 1 4 4 15
Linearly Tapered 3 3 3 18
Elliptic 4 1 1 15
Delta 2 2 2 12
The above decision matrix shows that a linearly tapered configuration should be used
for this application.
Wing Height
The height of the wing can be low-winged, mid-winged or high-winged. A low winged
design enables shorter landing gear height and so less drag but can compromise
205
stability. A high winged design provides a stable platform but would require the
landing gear to be mounted to the fuselage (if a landing gear is used). A mid-winged
aircraft provides better stability than a low wing but produces structural problems
which add to the complexity of the design. Wing height will primarily affect the landing
gear and other wing parameters. The wing heights have been rated in the areas of
expected landing gear weight (a lower wing will require a shorter and lighter landing
gear), stability(a higher wing is more stable) and complexity (a simpler design is
preferable, without requiring additional structural connections) as it can be seen in
Table 27.
Table 27: Wing Height Decision Matrix
Wing Configuration
Height
Landing Gear
Weight Stability Complexity Total
Weighting 3 2 1 N/A
Low Wing 3 1 2 13
Mid Wing 2 2 1 11
High Wing 1 3 3 12
Wing Sweep
The wing can be swept forward, aft or not swept. Sweep allows aircraft to perform
better in the area of flight approaching and beyond a Mach number of one. A swept
wing improves stability of an aircraft but can lead to pitch up of the wing at low
speeds. A fore swept wing introduces structural divergence issues, wherein the
wingtips may be bent upwards by the lift load and so increase angle of attack at the
tips, continuing until breakage. Wing sweep in either direction does improve lateral
stability but a swept wing will be heavier than an unswept wing as additional structural
weight is required. The wings sweep will affect other wing parameters such as
dihedral angle. The wing sweep options have been rated in the areas of performance
(in the low Mach number region where flight will occur), stability (a more stable
aircraft is preferable) and complexity (simple designs are better than more complex
designs) as it can be seen in Table 28.
206
Table 28: Wing Sweep Decision Matrix
Wing Configuration - Sweep
Low Mach
Performance
Stabilit
y
Complexit
y Total
Weighting 3 2 1 N/A
Fore Sweep 1 1 1 6
No Sweep 3 2 3 16
Aft Sweep 2 3 2 14
A non-swept wing will be chosen at this stage, although an aft swept wing may be
considered depending on the results of stability calculations.
Empennage Configuration
The empennage configuration entails the horizontal and vertical stabilisers and their
mounting on the fuselage. A conventional tail uses horizontal stabilisers either side of
the fuselage with a vertical stabiliser mounted on the fuselage at 90 degrees from the
horizontal stabilisers. This is a simple, well proven, effective design but provides
difficulties in mounting the pulsejet to the aircraft. A T-tail has a vertical stabiliser
mounted on the fuselage and horizontal stabilisers mounted on the vertical stabiliser.
This design adds weight required due to the stronger vertical stabiliser required and
still provides some difficulty mounting the pulsejet. This may be mounted on top of the
horizontal stabilisers, this would further increase the strength of the vertical stabiliser
required. A H-tail uses conventional horizontal stabilisers but mounts the vertical
stabilisers at the tips, allowing space for the pulsejet to be mounted. A H-tail would
need additional weight compared with a conventional tail due to the strengthening
required of the horizontal stabiliser. A V-tail would enable the pulsejet to be mounted
between two stabilisers, acting in both the horizontal and vertical directions. However,
this design requires additional actuation complexity and would require additional heat
shielding weight compared with an H-tail arrangement. The empennage configuration
affects the engine configuration. The empennage options have been considered in the
areas of suitability to pulsejet (expected heat shielding required for a tail mounted
pulsejet), complexity (simple designs are best), weight (a tail can have significant
weight) and size (different tails will require varying wing areas, leading to preferable
207
size characteristics). Table 29 shows a summary of the above discussion in the form of
a decision matrix.
Table 29: Empennage Configuration Decision Matrix
Empennage
Configuration
Pulsejet
Suitability Complexity Weight Size Total
Weighting 4 1 3 2 N/A
Conventional Tail 1 4 4 1 22
T-Tail 3 3 2 2 25
H-Tail 4 2 1 3 27
V-Tail 2 1 3 4 26
The above decision matrix shows that an H-tail is preferable, primarily due to the ease
of pulsejet mounting.
Landing Gear
The landing gear configuration includes the ability to be retracted, of the landing gear
and the arrangement of the wheels in the landing gear. The landing gear may be either
fixed, retractable, removable or a launcher may be used. A fixed landing gear is a
landing gear of any arrangement that remains lowered throughout the flight. A fixed
landing gear has high drag but is simple, effective and reliable. A retractable landing
gear is a landing gear that can be retracted through the use of a system (generally
hydraulic or pneumatic) and allows lower drag but has additional weight and
complexity. A removable landing gear is used only for take-off and is jettisoned after
this stage, allowing lower drag. However, this system requires the weight and drag of a
fixed landing gear (or higher, as a jettisoned landing gear will not be streamlined) at
takeoff and additional weight for the jettison system. A launcher can be used, giving
the aircraft a boost into the air, at which point it will fly under its own power. This
would require a launcher to be designed and tested, leading to significantly increased
complexity. The ability to be retracted, of the landing gear will affect the landing gear
arrangement and wing height. The ability to be retracted, options are rated in the
208
areas of drag, complexity, weight and reliability. The decision matrix used to determine
the best option can be seen in Table 30
Table 30: Landing Gear Type Decision Matrix
Landing Gear Type Drag Complexity Weight Reliability Total
Weighting 3 1 2 4 N/A
Fixed 1 4 2 4 27
Retractable 2 2 1 2 18
Removable 3 3 3 3 24
Launcher 4 1 4 1 24
The decision matrix above shows that a fixed landing gear is appropriate for this
aircraft design.
The arrangement of wheels in the landing gear structure can reasonably be a tail
dragger, tricycle or bicycle arrangement. A tail dragger has two main front wheels and
a small rear wheel, providing less drag and weight than other alternatives. However, a
tail dragger is inherently unstable and requires precise control of the rudder to retain
control. A tricycle has two main rearward wheels and smaller wheels at the front of
the aircraft. This design is stable on the ground but has poorer rough field performance
than a tail dragger configuration. A bicycle landing gear configuration uses four wheels,
front, rear and two outrigger wheels on the wings to prevent tipping of the aircraft.
The bicycle landing gear must takeoff and land at a flat attitude, increasing landing
difficulty and requiring high lift at low angles of attack which will likely increase cruise
drag. A bicycle gear will also provide worse rough field performance than a tail
dragger. The arrangement of the landing gear will affect the wing geometry required
for performance at differing angles of attack. The landing gear wheel orientations have
been rated on drag (during flight), stability (on ground, on runways of varying
roughness and in the air) and weight (the weight of the landing gear wheels). The
decision matrix can be seen in Table 31
Table 31: Landing Gear Arrangement Decision Matrix
Landing Gear Arrangement Drag Stability Weight Total
209
Weighting 1 2 3 N/A
Tail dragger 3 1 3 16
Bicycle 1 3 1 8
Tricycle 2 2 2 12
As shown above, the tail dragger is the best option for the landing gear arrangement.
Selected General Configuration
In summary, the selected general configuration has a conventional fuselage,
conventional wing with linear taper, H-tail, and a tail dragger with a top mounted
engine. The engine will also be partially cut into the fuselage, which is not shown in
Figure 134

Figure 134: Mock graphic of selected configuration

210
Appendix B- Weight Calculation Method
211

O W
W
O W
W
O
E F O
W W W
W W W
O
E
O
F
+ =
+ =

1
?
?
=
=
=
=
S
S
C
O
O
E
K
C
A
K AW
W
W

There is no published data for single piston engines UAV in any book, however the
literature review that was done produced the required data from Janes All the Worlds
UAVs, this data was tabulated in Table 32- Weight Data for Piston ;

Table 32- Weight Data for Piston UAVs
Aircraft Name We Wo We/Wo
Stephansen Dolphine 6 7.2 0.8333
NRIST Yk7 12.5 14 0.8929
Avaruustekniika AT85 12.3 19 0.6474
TAT Aoussou Mk1 15.5 25 0.62
TAI Keklik 7 10 0.7

A graph of W
E
/W
0
Vs W
E
can be plotted and a trend line can be added and the
equation will give the values for A and C, this graph can be seen in Figure 135

212
Chart Title
y = 1.2627x
-0.2088
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 5 10 15 20 25 30
WO
W
E
/
W
O

Figure 135- Graph of WE/WO Vs WO

As you can see the trend line for the points is of the type Y=AX^C

Therefore we can read off the values for A and C to be
A=1.2627
C=-0.2088

213
8949282 . 27
50 557899 . 0
557899 . 0
1 50 2627 . 1
2088 . 0
=
=
=
=
=
=
E
E
O
O
E
E
O
E
O
E
S
C
O
O
E
W
W
W
W
W
W
W
W
W
W
K AW
W
W

995 . 0
985 . 0
97 . 0
1 06 . 1
3
4
1
2
0
1
3
4
2
3
1
2
0
1
0
4
0
4
=
=
=
=
=
=
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
O
F
O
O
F
F

The above values come from (Raymer 2006)

=
D L
EC
W
W
exp
2
3

214
If we set the duration of the flight as being 12 minutes this allows for the aircraft to
still be able to be in the range of sight of the operators
E=0.2 hours
E=0.2 HOURS

9 max
13 9 max
max 866 . 0
=
=
=
D L
D L
D L D L

This is the range of L/D for a fixed landing gear, general propeller aircraft from (Raymer
2006 Fig 3.6)

The lowest value of the range was chosen, this is an assumption based on the
reasoning that choosing the lowest value would give the Maximum of the weight
range.

The specific fuel consumption was taken from last years report and they gave two
values which were found, the lowest being 7.2kg/kg/hour and the highest being
23.2kg/kg/hour
The sustained thrust that was produced by the engine had the specific fuel
consumption of 7.2kg/kg/hour therefore this is the specific fuel consumption that will
be used for calculation purposes.

831306 . 0
866 . 0 9
3600 2 . 7 3600 2 . 0
exp
exp
2
3
2
3
2
3
=

=
W
W
W
W
D L
EC
W
W

215
( )
222282 . 0
7903 . 0 1 06 . 1
1 06 . 1
7903 . 0
831306 . 0 995 . 0 985 . 0 97 . 0
0
0
0
4
0
0
4
0
4
3
4
2
3
1
2
0
1
0
4
=
=
=
=
=
=
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
W
F
F
F

Now that the initial W
0
value is known the fuel weight can be found.

00904 . 39 W
1 11.1141098 8949282 . 27 W
W W W
48501 . 10 W
06 . 1
W
11410981 . 11
50 222282 . 0
222282 . 0
0
0
F E 0
Usable F
Usable F
0
=
+ =
+ =
=
=
=
=
=
F
F
F
F
W
W
W
W
W

We need to iterate until the change in W
0
as a percentage is small, less than 0.5%

The following table, Table 33 shows the first three iterations

216
Table 33- Table of First Iterations
Initial Guess Wo We/Wo We Wf Wf usable Wo diff %
50 0.557899 27.8949282 11.11410981 10.48501 39.00904
39.00903801 0.587577 22.92081831 8.671014637 8.180202 31.59183 23.47824
31.59183295 0.614029 19.39830491 7.022302008 6.624813 26.42061 19.5727

After a sufficient number of iterations the difference between the two successive
iterations is less than 0.3%, this resulted in a take off weight W
O
=10.2kg at this point
the iteration process was stopped as the numbers were considered to be conceived.

The other characteristic weights corresponding to this are;

W
E
=7.932863748Kg

W
F
=2.26810571Kg

W
F USABLE
=2.139722Kg

W
0
=10.20097Kg
To try and increase the accuracy of the calculation more data, this data can be seen in
Table 34 was used to create a new curve to obtain the values for A and C in the
following formula;

S
C
O
O
E
K AW
W
W
=

Table 34- UAV Data
NRIST Yk7 12.5 14 0.8929
217
Avaruustekniika AT97 5 7.3 0.6849
TAI Keklik 7 10 0.7
Straero ATS 0101 27.5 40 0.6875
Meteor Mirach 10 67 82 0.8171
AES SAT 400 Astoe 40 80 0.5
Dragon DA/3 Targetmaser 6.4 10 0.64
Dragon DA/5 Firefly 2.49 4 0.6225
Tasuma SAGT 50 36 48 0.75
Tasuma SAGT 60 36 52 0.6923

As before it can be plotted to determine the equation co-efficient, this graph can be
seen in Figure 136- Graph of WE/WO Vs WO
y = 0.7324x
-0.0189
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 10 20 30 40 50 60 70 80 90
WO
W
E
/
W
O

Figure 136- Graph of WE/WO Vs WO

From this graph the values obtained are not feasible as they do not produce a
converging result to a realistic value.

218
Also it can be seen on the graph that there are some points on the graph which are not
consistent with the other data, by eliminating these points form the curve a realistic
set of coefficients to produce a converging results.

Table 35 shows only the consistent data sets that will be used;

Table 35- Consistent UAV Weight Data
NRIST Yk7 12.5 14 0.8929
Avaruustekniika AT97 5 7.3 0.6849
TAI Keklik 7 10 0.7
Dragon DA/3 Targetmaser 6.4 10 0.64

The graph produced from this data can be seen in Figure 137

219
y = 0.9351x
-0.1105
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 5 10 15 20 25 30
WO
W
E
/
W
O

Figure 137- Graph of WE/WO Vs WO for Consistent Data

220

After the co-efficients were obtained the method outlined previously was followed to
produce the following:

Table 36 shows the first three rows of the table that shows the iterations to find the
parameters we are after.

Table 36- Iteration Results
Initial Guess Wo We/Wo We Wf Wf usable Wo diff %
10 0.725033 7.250334724 2.222821961 2.097002 9.473156686
9.473156686 0.729383 6.909555522 2.105714072 1.986523 9.015269594 5.079017
9.015269594 0.733386 6.611676973 2.003933924 1.890504 8.615610897 4.638774

After a 200 iterations the difference between the two successive iterations was
9.4310
-08
this resulted in a take off weight W
O
=5.3kg at this point the iteration
process was stopped as the numbers were considered to be conceived.
The other characteristic weights corresponding to this are;
W
E
=4.122207805 kg

W
F
=1.178182369 kg

W
F USABLE
=1.111493 kg

W
0
=5.300390174 kg

Appendix C Matching Diagram
221
The details of the matching diagram calculation are presented here. The resulting
matching diagram is located in the conceptual design section.
Drag polar estimation
To graph the matching diagram, an estimation of the drag polar (Equation 28) was
required. For the curve, a standard shape was assumed but estimations of C
D0
, aspect
ratio and Oswald efficiency factor were required.
Equation 28:
2
0 L D D
KC C C + =
To determine the Oswald efficiency factor of the wing (e), the approximation in
Equation 29 was used. This efficiency factor was then used to determine the factor K in
the drag polar using Equation 30.
Equation 29:
e = 1.78(1 0.045A 0.68) 0.64
Equation 30:
A e
K
1
=
Estimating zero lift drag coefficient, C
D0

The zero lift drag coefficient was estimated by Equation 31 below.
Equation 31:
S
f
C
Do
=
222
The equivalent parasite area, f, was found by a statistical relation shown in Equation
32. The coefficients a and b were taken from Roskam (1985a) for Cf=0.006 which was
chosen as a realistic upper bound of possible skin friction factors while S
wet
was
estimated from preliminary drawings of the aircraft.
Equation 32:
wet
S b a f
10 10
log log + =
When the landing gear is in position C
D0
will be increased by 0.20 as recommended by
Roskam (1985a).
Initial estimate of drag polar
Thus an initial estimate of the drag polar was plotted and can be seen in Figure 138
below.
First Estimate of Drag Polar
C
L
versus C
D
-0.5
0
0.5
1
1.5
0.00 0.05 0.10 0.15 0.20
C
D
C
L

Figure 138: First Estimate of Drag Polar
This will be revised when an aerofoil is selected but should approximate expected
results.
223
For use in matching diagram calculations, aerodynamic efficiencies for different
regimes were estimated by dividing the values of C
L
found on the first estimate drag
polar by the values of C
D
.
Climb requirements
The climb gradient requirement was given by the Equation 33, which gives a minimum
wing loading for a certain thrust loading for a climb gradient, G.
Equation 33:

=
Ae q
Ae
C
G
W
T
G
W
T
S
W
D
2
4
0
2

The thrust loading required must meet the requirements of Equation 34 to produce a
real solution.
Equation 34:
Ae
C
G
W
T
D
0
2 +
For required climb gradient, the CASA requirement of 8.33% climb gradient at 1.3Vstall
was followed.
For stall speed, based on a reasonable speed considering the use of a sealed run-way
and the stall speeds of similar sized jet powered aircraft, stall speed was set to 60 kph.
This stall speed would allow a relatively skilled pilot to land the aircraft while
minimising the effects of a low stall speed on reducing aerodynamic efficiency.
Stall Requirement
The stall requirement is given by Equation 35 which describes the maximum wing
loading for a specific stall speed.
224
Equation 35
2 / 1
max
2
=
L
s
C
S
W
V

At this stage, maximum lift coefficient was estimated at 1.2 for a clean wing.
Takeoff Field Length Requirement
Equation 36 calculates T/W required for a given wing loading and ground roll distance
and can be used to determine the takeoff field length requirement.

Equation 36
( )[ ]
( )
( )

+
=
+
1
1
1 . 1
2
1
2 2
0
2
0
S
W
C KC C g S
stall L L D
L L D g
e
S
W
V C KC C
W
T

The starting speed was set to V
i
=0 as the launch method is not likely to be a catapult
launch and the final speed was set to V
f
=V
TO
=1.1V
stall
.
The friction coefficient takes a value from 0.02 to 0.08 depending on the surface, it
will be set to the intermediate value of = 0.05 for hard turf which is likely to be the
surface of operation.
A take-off distance of 100m was specified for this UAV, to ensure operation in a
reasonably short runway, without overly restricting design.
Cruise Requirement
The cruise requirement was given by Equation 37, which gives the minimum thrust
loading for a certain wing loading.
225
Equation 37:
( )
Ae q
S
W
qC
W
T
S
W
D
+ =
0

At the stage of calculation, a cruise speed requirement of 260 kph was the target
based on expected available thrust of no more than 3kg.
Adjusting to take-off values
As the matching diagram is a plot of (W/S)
TO
versus (T/W)
TO
, W/S and T/W must be
adjusted to take-off values. This can be done through the use of Equation 38 shown
below and a similar equation for (T/W).
Equation 38
( )
Ae q
S
W
qC
W
T
S
W
D
+ =
0

If little fuel is burnt early in the flight of the aircraft then (W/S)
TO
equals (W/S).
Assuming thrust does not vary with altitude as the altitude of flight expected is small
(and assuming little variation in weight), then T/W is equivalent to (T/W)
TO
.
226

227
Trade studies were performed using the matching diagram for the aircraft, considering
consequences if engine weight and fuel consumption are different to those expected
for the aircraft. Fuel consumption is an area of specific concern as last years pulsejet
tests showed variability between 7.2 and 23.2kg/kg/hr (Coombes et al. 2007) while
engine weight is still uncertain, depending on final engine design. The graphs of takeoff
weight versus these parameters are shown below in Figure 139 through Figure 147:

Sensitivity to Fuel Consumption
Wto = 4.5181e
0.0786C
R
2
= 0.9995
0
2
4
6
8
10
12
14
2 4 6 8 10 12
C(kg/kg/hr)
W
t
o
(
k
g
)

Figure 139: Sensitivity to fuel consumption
228
Sensitivity to Engine Weight
Wto = 2.6151Wengine + 4.0742
R
2
= 1
0
1
2
3
4
5
6
7
8
9
10
0.75 1 1.25 1.5 1.75 2 2.25
Wengine(kg)
W
t
o
(
k
g
)

Figure 140: Sensitivity to Engine Weight
Trade studies were also performed to determine the impact of variations in design on
the performance of the aircraft to determine if possible improvements to the design
could be made. The main parameters of thrust to weight ratio and wing loading were
varied to determine the impact of these on the performance parameters of the
aircraft.

Cruise Speed Sensitivity to W/S
235
240
245
250
255
260
265
270
275
0 5 10 15 20 25 30
W/S
C
r
u
i
s
e

S
p
e
e
d

(
k
m
/
h
r
)

Figure 141: Sensitivity of Cruise Speed to W/S
229
Takeoff Distance Sensitivity to W/S
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30
W/S
T
a
k
e
o
f
f

D
i
s
t
a
n
c
e
(
m
)

Figure 142: Sensitivity of Takeoff Distance to W/S
Climb Rate Sensitivity to W/S
28.80%
29.00%
29.20%
29.40%
29.60%
29.80%
30.00%
30.20%
30.40%
0 5 10 15 20 25 30
W/S
C
l
i
m
b

R
a
t
e

Figure 143: Sensitivity of Climb Rate to W/S
230
Stall Speed Sensitivity to W/S
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30
W/S
S
t
a
l
l

S
p
e
e
d

(
k
m
/
h
r
)

Figure 144 : Sensitivity of Stall Speed to W/S
Cruise Speed Sensitivity to T/W
0
50
100
150
200
250
300
0 0.1 0.2 0.3 0.4 0.5
T/W
C
r
u
i
s
e

S
p
e
e
d

(
k
m
/
h
r
)

Figure 145: Sensitivity of Cruise Speed to T/W
231
Takeoff Distance Sensitivity to T/W
0
10
20
30
40
50
60
70
80
90
0 0.1 0.2 0.3 0.4 0.5
W/S
T
a
k
e
o
f
f

D
i
s
t
a
n
c
e
(
m
)

Figure 146 : Sensitivity of Takeoff Distance to T/W
Climb Rate Sensitivity to T/W
0.00%
5.00%
10.00%
15.00%
20.00%
25.00%
30.00%
35.00%
40.00%
0 0.1 0.2 0.3 0.4 0.5
W/S
C
l
i
m
b

R
a
t
e

Figure 147 : Sensitivity of Climb Rate to T/W

232
Appendix E Engine Mounting Calculations
233

234

Appendix F Liquid Fuels
235
This liquid fuel data has been tabulated for the purposes of the discussion found in
section 3.11.1.
Flash point
The flash point of a fuel is defined as the temperature at which vapour given off will
ignite when an external flame is applied under specified test conditions (Caltex
Australia Petroleum, 2008). Fuels with higher flash points are more stable and hence
will combust slower than fuels with low flash points. High flash points fuels are safer
for normal storage and handling as they are less volatile. However, when considering a
valveless pulsejet engine (VPJE), quick combustion will allow a higher instantaneous
pressure within the combustion chamber, resulting in greater engine performance.
Therefore a fuel with a low flash point is desirable. Table 37shows a range of
commonly used fuels and their corresponding flash points.
Table 37- Fuel Flash Point Data
(SEFSC, 2008)
Fuel Flash Point (C)
Methanol (CH
3
OH) 16
Nitro Methane (CH
3
NO
2
) 43
Acetone (CH
3
COCH
3
) -18
Ether (C
2
H
5
OC
2
H
5
) -40
Benzole (C
6
H
6
) -11
Petrol (variable) -46
Naphtha (Shellite) (variable) -20
Diesel (variable) ~75
Energy Density
The energy density of a fuel, also known as the higher calorific value, defines the
amount of energy stored in the fuel per unit mass (kJ/kg). It is therefore clear that
higher energy density will in theory allow for a greater thrust output from the engine.
236
Table 38 shows a range of commonly used fuels and their corresponding energy
density.
Table 38: Fuel Energy Density Data
(SEFSC, 2008)
Fuel Energy Density (kJ/kg)
Methanol (CH
3
OH) 23,200
Nitro Methane (CH
3
NO
2
) 11,630
Acetone (CH
3
COCH
3
) 29,075
Ether (C
2
H
5
OC
2
H
5
) 34,900
Benzole (C
6
H
6
) 40,200
Petrol (variable) ~ 47,000
Naphtha (Shellite) (variable)
Diesel (variable) ~ 46,000
Air-Fuel Ratio (AFR)
The effect of a fuels optimal AFR on the operation of a VPJE must be considered
before deciding upon a fuel. The AFR of a fuel will ultimately decide the required fuel
flow rate for the engine. By knowing the frequency at which the engine operates, the
combustion chamber volume and the optimal AFR of the fuel, the required fuel flow
rate can be estimated. Knowledge of the required flow rate will allow for more
efficient operation. Table 39 displays a range of common fuels and their corresponding
optimal AFR (SEFSC, 2008).
Table 39: Fuel Optimal AFR Data
Fuel Optimal AFR
Methanol (CH
3
OH) 4.5-6.5:1
Nitro Methane (CH
3
NO
2
) 0.5-2.5:1
Acetone (CH
3
COCH
3
) 9.5-10.5:1
Ether (C
2
H
5
OC
2
H
5
) 9.8:1
Benzole (C
6
H
6
) 11-11.5:1
Petrol (variable) 12.5-13.5:1
Naphtha (Shellite) (variable)
Diesel (variable)

237
Another aspect of a fuel that must be considered is its flammability limit. The
flammability limit of a fuel is defined as being the concentration of fuel (by volume)
that must be present in air for an ignition to occur when an ignition source is present
(ConocoPhillips, 2004). Therefore, the flammability limit will affect the throttle range
of the engine. Fuel concentrations above or below a fuels flammability limit will cause
the engine to flame out. The lower and upper flammability limits can be seen in Table
40.
Table 40- Fuel Flammability Limit Data
(SEFSC, 2008)
Flammability Limit Fuel
Lower
(%)
Upper (%)
Methanol
(CH
3
OH)
6 36.5
Nitro Methane
(CH
3
NO
2
)
7.3 63
Acetone
(CH
3
COCH
3
)
2.5 12.8
Ether
(C
2
H
5
OC
2
H
5
)
1.9 36
Benzole (C
6
H
6
)
Petrol
(variable)
1.3 6
Naphtha
(Shellite)
(variable)
1.1 5.9
Diesel
(variable)

Latent Heat of Vaporisation
The latent heat of vaporisation of a fuel is the energy required to change the phase of
a fluid from liquid to vapour. As fuels with a high latent heat of vaporisation vaporise,
they will cool down the surrounding air, resulting in a higher density. In the case of
pulsejet engines, this increase in density is thought to potentially improve the
performance of an engine, by allowing a greater mass flow to be produced. The values
for the latent heat of vaporization for different fuels can be seen in Table 41.
238
Table 41- Latent Heat of Vaporisation Data
(SEFSC, 2008)
Fuel Latent Heat of Vaporisation (kJ/kg)
Methanol (CH
3
OH) 1100 (methanol handling guide)
Nitro Methane (CH
3
NO
2
) 600
Acetone (CH
3
COCH
3
) 520
Ether (C
2
H
5
OC
2
H
5
) 350
Benzole (C
6
H
6
) 390
Petrol (variable) 300
Naphtha (Shellite) (variable) N/A
Diesel (variable) N/A
Appendix G Component Weight Breakdown
239
Appendix G Component Weight Breakdown
A table of components and their relative distance from the front of the aircraft was
prepared as shown below in Table 42. This was used to determine the travel of the
aircraft centre of mass in the C.G. Excursion diagram and determine the size of
horizontal tail required. The use of a percentage aircraft mass column allowed the
applicability of mass approximation equations to be determined.

Table 42 : Component Weight Breakdown
Item
Weight
Estimate(kg)
C.O.M. Location
(mm) Number
Total
Mass(kg)
Mass Centre
Contribution
(kg.mm) Weight (%)
Fuselage 0.930 643 1 0.930 598.3 18.12%
Engine 0.700 821 1 0.700 574.7 11.36%
Fuel Full 2.700 636 1 2.700 1717.2 44.46%
Fuel Empty 0.200 636 1 0.200 127.2 3.89%
Fuel 67% Front 1.875 570 1 1.875 1068.8 31.08%
Fuel 67% Rear 1.875 700 1 1.875 1312.5 31.08%
Landing Gear
Front 0.100 500 1 0.100 50.0 1.62%
Landing Gear
Rear 0.050 1140 1 0.050 57.0 0.81%
Wing 0.757 703 1 0.757 532.4 14.75%
Horizontal Tail 0.266 1110 1 0.266 295.0 5.18%
Vertical Tail 0.127 1110 1 0.127 140.5 2.46%
RC Receiver 0.028 130 1 0.028 3.6 0.45%
RC Receiver
Battery 0.094 195 1 0.094 18.3 1.53%
Speed
Controller 0.006 85 1 0.006 0.5 0.10%
GPS 0.100 188 1 0.100 18.8 1.62%
Pump Battery 0.100 97 1 0.100 9.7 1.62%
Pump 0.040 300 1 0.040 12.0 0.65%
Wing Servo 0.041 750 2 0.082 61.5 0.67%
Tail Servo 0.041 1140 2 0.082 93.5 0.67%

240

Appendix H Test Log Books
241
Testing Logbook 26-27
th
February
Introduction
The three primary aims of this round of testing were:
To familiarise and implement all auxiliary systems required for safe and
effective operation of the pulsejet engine
To familiarise ourselves with the operational behaviour of the engine, thereby
helping to increase general understanding of the engines
To test the importance of flared exhaust and intake tubes for the successful
operation of a pulsejet engine

Background
The development of pulsejet engines began at Adelaide University in 2007 with an
honours project devoted to design a working pulsejet engine and thrust measurement
stand. The 2007 group successfully tested one pulsejet engine, mounted on a stand
able to measure the thrust being produced from the project. A second engine was
designed and built in late 2007 that was bigger in size, in an attempt to increase the
thrust output of the engine. The testing of this engine was not deemed a success, as
sustained operation without the use of an external air supply was not achieved.

Equipment Setup
The following equipment was used during testing:
Two pulsejet engines
Exhaust and intake flares for the bigger engine
Thrust measurement stand
Air compressor
Propane BBQ gas bottle
Spark plug
Pulse generation circuit and coil
Laptop
20kg load cell
242
Data acquisition unit USB-1208FS
Video camera
Solenoid shut-off valve
Electronically operated gate valve
12 volt battery
Computer power supply
TV set
Fire extinguisher
First aid kit
Ear protection (ear buds and muffs)
Air injection piping
Fuel injection piping
Fuel and gas hosing
Control box
Lutron SL-4022 sound level meter

Much of the equipment used was implemented to ensure the safety of all people
conducting testing. Possibly the most critical safety equipment used was the ear
protection. The two levels of protection were needed during ALL tests, as levels in
excess of 130dB were measured. Such sound levels would cause immediate hearing
damage if no protection was being worn. The fire extinguisher and first aid kit were
not expected to be needed, but were kept onsite for use in the case of an emergency.
The equipment used in the set-up of the two engines was largely the same, with the
only difference being the variable length flares to be used on the bigger engine. The
engines were placed and secured to the thrust measurement stand.

The air system used incorporated an air compressor, hosing and injection piping. The
air compressor was positioned outside the bunker, with hosing connecting it to the air
injection pipe, which was positioned at the intake to the engine, positioned inside the
bunker.

The fuel system incorporated a propane gas bottle, insulated gas hosing, a solenoid
shut-off valve, an electronically operated gate valve and fuel injection pipe. The
solenoid valve and gate valve were both positioned in series, located between the gas
bottle and fuel injection pipe, which was mounted in the intake of the engine. Both
valves were powered by the computer power supply, with the gate valve running off
243
5V and the solenoid running off 24V. The valves were manually operated using
switches on the control box.

The ignition circuit involved a spark plug, pulse generation circuit and coil as well as a
24V battery. The spark plug was mounted in the end of the combustion chamber, and
was electrically connected to the coil and pulse generation circuit, which provided the
voltage required for the spark. This circuit was in turn connected to the 12V battery.
The timing of the spark was manually operated via the control box.

A video was used to record the video analysis, and was also connected to a TV set
located in the control room. A 20kg load cell was attached to the test stand, and was
connected to the data acquisition unit, which supplied data to the laptop in the control
room. A sound pressure level meter was also positioned inside the bunker.

Procedure
The two days of testing that were undertaken took a lot of forward planning and
organisation to ensure all went to plan. Not only did all the control and delivery
systems have to be prepared, but we needed to transport all the equipment out to the
testing bunker at DSTO.

Once at the testing facility, all equipment was unloaded and set-up in such a way that
all operators were located outside the testing bunker during live testing, making for a
much safer environment. Full control was still maintained over the engine through the
use of emergency shut-off valves and variable gate valves, all of which were operate
via a control box, situated outside the bunker.

The specific procedure required to successfully operate the engines was largely
unknown to us prior to testing, so there was initially a lot of trial and error until we
were able to see some patters.
244
Results
Test 1: Variable engine
Fuel and Air position: See picture 21
Comments: Engine ran for <1min. Produced 1kg of self-sustaining thrust.

Test 2:
Fuel and Air: Pic 22 (in further than test 1)
Comments: Unable to maintain combustion without air supply. Engine ran, but cut out
as soon as the air was cut.

Test 3:
Fuel and Air: pic 24
Comments: Engine sustained for about 10seconds, producing 1.2kg of thrust. However
it flamed out after 10 seconds.

Test 4: gas bottle changed was getting too cold
Air and fuel: pic 25 intake moved to original position
Comments:

Test 5:
A-F: Pic 26
Comments: Self-sustaining operation achieved. Injector was quite a way out. It was
noticed that there appeared to be a lot of fuel being burnt outside the intake, this can
be seen in Figure 148
245

Figure 148- Engine during Test

Test 6:
AF: pic 27 injector moved out further
Comments: Achieved 1.9kg of self-sustaining thrust.

Test 7:
AF: Pic 28 injector moved out further again
Comments: Achieved 1.5kg of thrust. This was less than the previous test. Again it was
noted that there were large flames out of the intake. Initial thoughts are that this is
inefficient operation, as a lot of fuel is being burnt outside the engine, and hence not
helping to produce thrust.

Test 8:
AF: Same injector position as test 6, but injector end opened more.
Comments: 1.85 kg thrust

NEW ENGINE (Engine No. 2)
Test 1:
Exhaust Length: 74.5mm
Intake: 68mm
Fuel: 50mm (into the engine)
Comments: The engine ran first try. It was easier to start than any of the configurations
tried with Engine no. 1. The data logger crashed on start-up, so no thrust
246
measurement was taken. As the engine did not run without the addition of the flares,
the addition of flares to the engine has made a vast improvement.

Test 2:
Exhaust: same as test1
Intake: same as test 1
Fuel: same as test 1
Comments: Achieved 1.1kg of thrust. Concerns over the accuracy of the data logger, as
it was expected this engine would produce more thrust than the first engine due to the
large size.

Test 3: pics 31 - 35
Exhaust: no change
Intake: no change
Fuel: no change
Comments: To check the accuracy of the data logger, a spring gauge was connected to
the rolling section of the stand to measure the thrust. This suggested an engine thrust
of 1.2kg, and therefore backed up the data logger. The engine again ran easily though,
but just didnt produce the expected thrust.

Test 4:
Exhaust: same
Intake: same
Fuel: 70mm into the intake pic 36, 37
Comments: less than 1kg of thrust produced. Again, the engine started easily and ran
continuously with no problems.

Test 5:
Exhaust: 25.5mm
Intake: 69mm
Fuel: 80mm photo 38
247
Comments: again ran easily, but again produced minimal thrust

Test 6: pic 39
Exhaust: 33.5mm
Intake: 69mm
Fuel: same
Comments: easy to start and ran continuously. Again 1kg of thrust, the graph of Thrust
Vs Time can be seen in Figure 149

Figure 149- Thrust Vs Time for Test 6

Test 7:
Exhaust: 54.5
Intake: 69mm
Fuel: same
Comments: started and ran easily. Thrust spiked at 3kg, but on reviewing the video
footage, that was a result of the stand rolling back and hitting the load cell. The thrust
settled again at about 1kg. The graph of Thrust Vs Time can be seen in Figure 150

248


Test 8:
Exhaust: 78mm
Intake: 78mm - extended
Fuel: 90mm
Comments: Little thrust <1kg, the graph of Thrust Vs Time can be seen in Figure 151

249

Test 9:
Exhaust: same
Intake: flare removed
Fuel: same
Comments: engine struggled to light without air, but managed to eventually. The
engine sounded different and did not appear to get as hot as on previous runs. The
graph of Thrust Vs Time can be seen in Figure 152


Test 10:
Exhaust: same
Intake: same
Fuel: a long way in
Comments: It ran continuously again, but did not sound like it was running well at all. It
tended to splutter a bit, and just sounded weak. The graph of Thrust Vs Time can be
seen in Figure 153
250


Test 11:
Exhaust: same
Intake: same
Fuel: moved into the neck of the intake
Comments: again obtained continuous thrust, but only about 0.7kg. It kept working at
really low fuel flow rates, to a point where it was running, but not producing any
measurable thrust. The engine was very quiet for a pulsejet. The graph of Thrust Vs
Time can be seen in Figure 154

251
Test 12: gas bottle changed not enough pressure being supplied to the engine, which
may have adversely affected the last test
Exhaust: same
Intake: same
Fuel: same
Comments: again ran continuously

Test 13:
Exhaust:
Intake:
Fuel: injector in the cc
Comments: Did not work at all as expected. Did not splutter or pop. Tried different air
positioning, but this did not help.

Test 14:
Exhaust: 89mm
Intake: 66mm
Fuel: 66mm
Comments: Again the engine ran continuously. The engine was left running for a
reasonable length of time. Review of the video showed that the engine appeared to
reach thermal equilibrium. It was thought that it could therefore have been run
indefinitely. The graph of Thrust Vs Time can be seen in Figure 155

252


Test 15:
Exhaust: 103mm
Intake: same
Fuel: same
Comments: it was much harder to start, but once running it was fine. The graph of
Thrust Vs Time can be seen in Figure 156

253
Discussion/Recommendations
Engine 1:
The primary aims of the tests undertaken on engine no.1 were to ensure all auxiliary
systems such as the fuel and ignition circuits were operating successfully, and for the
group to become familiar with the operational behaviour of an engine that was known
to work. These tests were successful, with self-sustaining operation of the engine
achieved on numerous occasions, with few problems with the set-up. The group
gained valuable experience about the operational characteristics of the engine that
could only be learnt during live testing. Initial problems were encountered when
attempting to start the engine, as the group struggled to achieve combustion inside
the engine. After much thought about the physics behind the problem, a repeatable
and reliable starting procedure was found. The precise numerical results obtained
from these tests were deemed largely unimportant, as the performance of the engine
had already been thoroughly tested during 2007. However, it was noted that the
thrust results obtained, with a maximum of 1.8kg, were significantly lower than those
previously achieved. The reasons for this were largely unknown. Lower results were
not of great concern, as no development had been done on this engine.

The knowledge gained during these tests was however important in preparation for
testing engine no.2, as this engine had not been tested correctly to date.

Engine 2:
The impact the addition of the flares had on the engine was clear. Without any flares,
the engine would not produce self-sustaining thrust. However, with the addition of the
flares, the engine immediately obtained continual operation and thrust. As the design
of the flares allowed the lengths of the intake and exhaust pipe to be varied, a range of
configurations were able to be tested. With this set-up the engine obtained self-
sustaining thrust in almost every test undertaken, proving the theory behind the flare
design, and their importance in pulsejet design. The results from the different
configurations however showed very little correlation between the length of the
254
exhaust to the intake, with thrust outputs varying little for a number of different
configurations.

The maximum thrust obtained from engine no.2 was 1.5kg (check this). This was less
than that obtained from engine no.1, which was unexpected due to the larger physical
size of engine 2. The result was of concern, as a minimum of 3kg of thrust was
determined to be required to power the aircraft. The sound produced by engine 2 was
distinctly different from that produced by engine 1. The sound was less intense and
less crisp than that produced by engine 1. It is thought that that is a result of incorrect
tuning of the exhaust and intake pipes.

The tests clearly suggest that there is a lot more to an engine than physical size. Engine
no.2 is larger in size than engine no.1. However it produced significantly less thrust. As
was suspected, for a pulsejet engine to run well, the exhaust, intake and combustion
chamber sizes and lengths must be tuned to one another. The results obtained suggest
that the sizing of the diameter of the intake and exhaust tubes is more important than
the lengths

General:
In all tests where self-sustaining operation was achieved, the data logger showed the
engine thrust decreasing with time. This was thought to be as a result of the gas bottle
cooling, and therefore not providing the same fuel flow to the engine. This was
undesirable and ultimately affected the results. Any tests over approximately
30seconds in length were dramatically affected. This made it almost impossible to
accurately test the throttle range of the engines

Injector too far in does not allow enough time or distance for the air and fuel to mix
adequately to produce good combustion. This is backed up by the lack of operation
when the injector was put directly into the combustion chamber. The injectors too far
255
out tended to produce better engine thrust, but appeared to be wasting more fuel in
doing so.
Testing logbook 21-25
th
of April
Date: 21
st
of April
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Nick Lukacs
Terry Walladge

Logbook Prepared by Nick Lukacs

Engine Test Conducted on:
Engine 2 ceramic coat on inside of engine
Aim
1. Determine the effect of ceramic coating on the operating temperature of the engine
2. Determine the effect of exhaust flair on start up, and on operational performance

Tests 1 4
Engine Modifications: Intake flare on, exhaust flare off
Fuel and Air line Position: air near mouth of intake, fuel line 100mm in.
Test Report: Unable to obtain sustained thrust. Engine popped and spluttered, but did
not maintain operation. Small bursts of operation were achieved when fuel line was
moved out, and air line was moved in. best results achieved when they were both next
to one another, about 20mm from mouth of the intake.
Problems: Camera suffered from start-up. Alternative setups could not provide a
working condition.

Test 5
Engine Modifications: Exhaust flare attached.
256
Fuel and Air Line Position: Fuel line at 40mm in, air at 10mm. (measurements taken
from the start of the engine, not including flare)
Test Report: Engine started with no problems at all, and maintained thrust throughout
test. Engine visibly cooler than in previous tests.
Problems: Ceramic coating seemed to suffer under the conditions (photo 109-0938)
Thrust: approx 1.3kg

Engine Test Conducted on:
New Chinese Flight Weight Engine
Aim:
1. Run Chinese Engine
2. Determine most suitable Injector position for max thrust and throttle range
3. Ensure thrust results are as expected
4. Determine Max Fuel Consumption

Test 1

Engine Modifications: None
Fuel and Air line Position: air near mouth of intake, Fuel line 100mm in.
Test Report: Unable to obtain sustained thrust. Small bursts of operation were
achieved. The engine was flaming out during most of test.
Problems: Camera still not working, no thrust measurements could be taken due to
interference with data logger upon start-up.

Test 2

Fuel and Air line Position: Fuel line moved out to approx 80mm, airline not moved.
Test Report: Engine Ran successfully, very throttle-able. Test ran from start-up to
flameout, and then engine was restarted. Engine is significantly harder to start than
FWE engines. Frequency of operation sounded significantly higher than FWE engines.
257
Problems: Data logger started after initial start-up, engine was throttled until it flamed
out. When the engine was re-started the data logger cut out. This suggested
interference from the spark plug. NO RESULTS

Test 3

Fuel and Air line Position: Same as Test 2
Test Report: Manual thrust results determined using strain gauge. Engine ran
successfully, seems to need to be warm before it starts. Engine ran from start-up to
max thrust, then restarted and throttled as low as possible.
Problems: manual measurements only really accurate to 0.2kg. Tests ended when
component in the variable valve broke.
Thrust
Thrust: approx 3.2kg down to1.5kg

Day one ended due to damage to variable valve setup.

258
Date: 22
nd
of April
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Nick Lukacs
Terry Walladge
Visitors:
Michael Sipols
Karn Schumacher
Mitchell OCallaghan

Tests Conducted on:
Chinese Flight Weight Engine

Notes:
o Installation of new control system for variable fuel valve installed.
o Continuing tests on Chinese engine to determine best injector position and
throttling range
o New Gas Bottles Purchased

Test 1

Fuel and Air line Position: Air at mouth of engine, near edge, Fuel 32mm in at
centreline
Test Report: First test conducted, engine could not be flamed out. Data logger not
working again therefore reverted to strain gauge measurements.
Problems: Data logger, plus low pressure from empty gas bottle.
Thrust: Unknown

Test 2
259

Fuel and Air line Position: Same as Test 1
Test Report: After changing the gas bottles over, the pressure from the new bottle had
a noticeable impact on the way the engine operated. The old, cold bottle, although
providing enough flow, provided it at a lower pressure. It is thought that higher
pressures result in better fuel mixing from the injectors. This is one problem that will
be alleviated with the implementation of a liquid fuelling system, with a pump capable
of providing constant fuel flow rates and pressures.
Problems: None. Data logger worked. Engine is difficult to start, however runs very
well once started.
Max thrust: 3.618kg, the graph of Thrust Vs Time can be seen in Figure 157

Figure 157- Thrust Vs Time with injector 32mm from intake mouth

Tests 3-4

Fuel and Air line Position: fuel at 48mm in, Air same as previous tests
Test Report: First test plagued by more issues with data logger. Data logger gets error
messages and resets during runs when engine jumps in thrust. Second test using
260
manual strain gauge, engine ran well, started very easily. Engine had more throttle
range, but lower max thrust.
Problems: None, once data logger was removed
Max thrust: 3.25kg

Test 5-7

Fuel and Air line Position: injector at 57mm in. Air in same position.
Test Report: Again difficult to start. Strain gauge used again. Data logger is not being
used for the rest of the day. Thrust results higher than previous test, and engine could
not be flamed out, therefore new gas bottle was used to determine max thrust. The
new gas bottle somehow produced lower thrust measurements, suggesting that the
strain gauge may have been moved/bumped. Test re-run, thrust of 3.25kg confirmed.
Problems: None
Max thrust: approx 3.25kg (voltage =2V)

Test 8

Fuel and Air line Position: injector at 66mm from intake mouth, air in same position
Test Report: Engine flaming out at lower voltage (flow rate) suggesting that the further
in the injector is the less fuel that is wasted. Thrust results lower than previous.
However engine can be throttled to 1kg before it dies.
Problems: None.
Max thrust: 3.1 kg (Voltage =1.94V)
Min thrust: 1kg

Test 9

261
Test Report: Engine again flaming out at lower voltage (flow rate). Thrust results about
the same as previous test. However engine can be throttled to 0.8kg before it dies.
Engine also needs less fuel to start.
Problems: None.
Min thrust: 0.8kg

Test 10

Test Report: Fuel rates seem to be about the same as the previous tests, slightly lower
start-up voltage. Thrust values are the same as previous few tests.
Problems: None.
Min thrust: 0.8kg

END OF TESTS

Tests Conducted on:
Adjustable FWE engine with expanding exhaust

Aim:
1. Determine if expanding exhaust will improve engine performance
2. Determine the effect of length on engine performance
3. Achieve 3kg of thrust with FWE engine

Test 1

Fuel and Air line Position: injector at 30mm from intake mouth, air in same position as
Chinese tests
Exhaust Length: 100mm section between expansion section and combustion chamber
262
Test Report: engine could not sustain thrust without compressed air. Engine does not
sound healthy.
Problems: engine could not sustain operation.
Max thrust: 1kg
END OF DAY 2 TESTS
263
Date: 23
rd
of April
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Nick Lukacs
Terry Walladge

Notes:
o Continuing tests from day 2 on expanding tail pipe.
o Strain Gauge still being used.

Test 2-3

Chinese tests
Test Report: First test, engine could not achieve self sustaining combustion. Intake
position adjusted to 45mm from mouth. Self sustaining operation then achieved, with
max thrust of 3kg being produced. Engine needed significantly more fuel to start.
Throttle range was very low. Engine is very easy to start.
Problems: none
Max thrust: 3kg
Sustained thrust: 2.75kg (1min)

Test 4-5

Chinese tests
264
Test Report: Gas very low during first test, thrust of 3kg achieved. Gas bottle changed,
to a full bottle, which produced 3.5kg of steady thrust. Throttle range has improved
slightly.
Problems: none
Max thrust: 3.5kg (Steady)

Test 6-8

Chinese tests
Test Report: During the first test the engine started well and reached 3kg of thrust.
The engine then started to change in pitch and the thrust dropped by approx 1.5kg.
The engine was then throttled down to 1kg before being shut off. Once the engine was
cooled, it was noticed that the spark plug had come out of the back of the engine,
leaving a hole in the combustion chamber. It should be noted that this did not produce
any damage to the engine or the surrounding test equipment.
The test was re-run with the spark plug firmly screwed into the engine. During this test
the max thrust achieved was 3kg, before the settling at 2.5kg. It should also be noted
that so far it has not been possible to flame out any of the engines, even with new gas
bottles. This suggests that the injector diameter is too small for these engines. For this
reason a larger injector was installed into the same position. The same results thrust
results were achieved with the new injector and the engine could now be flamed out.
Problems: spark plug came lose during test.
Max thrust: 3kg

Test 9-11

Chinese tests
265
Test Report: First test, gas not turned on at tap.
Second test, thrust appeared to hit 4kg very briefly, before dropping to 3.5kg and then
3kg. This seemed unusual so the thrust setup was checked, and engine re tested
Third test: steady thrust of 3.2kg was achieved.
Problems: Whilst setting up the second test, it was determined that the larger injector
should be changed back to the smaller injector, as results would be more comparable
if the same injector was used through all tests
Max thrust: 3.2kg

Test 12

Chinese tests
Test Report: As thrust results seemed to be quite similar for small increments a long
section was tested to determine if length does affect thrust. Surprisingly the engine
operated even with the long section, however only 2kg of thrust was achieved.
Interestingly this around the length of last years engine, with the expanded section
added on. Also it is interesting that last years thrust results were around 2kg. This
suggests that the expansion section is not actually being used by the engine at this
length.
Problems: none
Max thrust: 2kg

Test 13

Chinese tests
266
Test Report: Tests conducted at smaller increments around the 350mm mark, as this is
where maximum thrust was achieved. The engine achieved 3.25kg of thrust, levelling
off at around 3kg.
Problems: none
Max thrust: 3.25kg

Test 14

Chinese tests
Test Report: engine starts very easily, max thrust of 3.3kg achieved. The engine is also
very throttle-able, with minimum thrust of 0.5kg achieved, and held for about
20seconds.
Problems: none
Max thrust: 3.3kg
Min thrust: 0.5kg

Test 15

Chinese tests
Test Report: Engine throttles very low, 0.25kg recorded as min thrust. 3kg max thrust
achieved.
Problems: none
Max thrust: 3kg

END OF TESTS

Date: 24
th
of April
267
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Nick Lukacs
Terry Walladge

Notes
o Hope to get data logger to work, to confirm thrust values achieved for
350mm section
o See if extending the intake length can affect the results.
o Extensive effort made to isolate the load cell. The bolt holes were lined with
electrical tape, the amplifier was also attached to the wall of the testing
bunker, instead of being mounted to the test stand. Also an attempt was
made to earth ALL electrical connections.

Test 16

Chinese tests
Test Report: Engine was run with measurements being taken from the strain gauge.
Engine could not be flamed out. Max thrust was the same as last time
Problems: none
Max thrust: 3.5kg

Test 17

Chinese tests
Test Report: Re-test with larger injector to try flame out engine. Data logger also used.
Both ran successfully. Frequency measured by data logger was exactly 50Hz, which
suggests interference from mains power. Thrust results were the same as those
268
measured by the strain gauge, which suggests both are a suitable form of
measurement.
Problems: interference from mains power somehow.
Max thrust: 3.554kg, the graph of Time Vs Thrust be seen in Figure 158

Figure 158- Time Vs Thrust for FWE with expanding Exhaust

Test 18

Chinese tests
Exhaust Length: 350mm section between expansion section and combustion chamber,
added 50mm section to intake
Test Report: N/A
Problems: When throttling low, the fuel shutoff valve was not working. This was of
significant safety concern therefore tests were abandoned for the day.
Max thrust: not recorded.

269
Date: 16
th
of July
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Nick Lukacs
Terry Walladge
Chinese engine:

Test No. 1
Testing Purpose: Testing Gas Injector for suitable Position
Injector Position: 2.6mm outside intake
Air Position: 4.5mm inside intake
Test Length: 1 min
Results: Engine ran with air. Flamed out.
Safety Report: Tighten bolts on engine mount

Test No. 2
Testing Purpose: Testing Gas Shutoff Valve
Air Position: 4.5mm inside intake
Test Length: 1 min
Results: Engine ran with air. Flamed out.
Safety Report: Shutoff valve working.

Test No. 3
Testing Purpose: Testing Gas Injector for suitable Position
Air Position: 45mm inside intake
Test Length: 1 min
Results: Engine sustained thrust for 30 seconds
Safety Report: none
270

Test No. 4
Testing Purpose: Testing liquid and gas injectors
Gas Injector Position: 2.6mm outside intake
Liquid Injector: 8 50mm from plate
10 90mm from plate
30mm gap from plate to intake
Test Length: 1 min
Results: ran on gas, started to add petrol and gas, reduce engine reduced in revs, test
stopped after 1 minute.
Safety Report: none

Test No. 5
10 90mm from plate
Test Length: 1 min
Results: ran on gas, started to add petrol and gas, reduce engine reduced in revs, could
not sustain thrust only on petrol
Safety Report: none

Test No. 6
10 90mm from plate
271
Test Length: 1 min
Results: ran on gas, started to add petrol and gas, reduce engine reduced in revs.
Petrol does not seem to be effecting the operation of the engine. I.e. Adding and
taking away petrol does not seem to effect the engines operation.
Safety Report: fine mist stopped test.

Test No. 7
10 90mm from plate
Test Length: 1 min
Results: ran on gas, started to add petrol and reduce gas flow. Turned off gas and tried
to start on petrol. Unsuccessful.
Safety Report: Power supply is not supplying required 7.2 volts. Reaching 6V and 2.5A,
this does not appear to be enough to supply the required fuel flow rate.

Testing Shows current flow rate with 10 and 8 injector = 148ml/min
Adding third injector

Test No. 8
8 73mm from plate
10 85mm from plate
272
Test Length: 1 min

Test No. 9
8 73mm from plate
10 85mm from plate
Test Length: 1 min
Results: spluttered and almost started on petrol after heating the engine on gas.
Safety Report: large plume of fuel seen. Test aborted.

Test No. 10
8 73mm from plate
10 85mm from plate
Test Length: 1 min
273

Test No. 11
8 73mm from plate
10 85mm from plate
Test Length: 1 min
Results: retest of test 10. Same results. Engine does not appear to be getting enough
petrol to start.
Safety Report: none.

Tests aborted for larger power supply.
274
Date: 23-24 July
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Nick Lukacs
Terry Walladge

The following tables are records of the tests done on the Chinese engine. These tests
were done on the engine with the injectors placed in holes along the side of the intake.

Injector positions Fuel comment
Position 1 = close to CC
test injector no 6 8 10
1 1 2 3 petrol
2 3 2 1 petrol
3 3 2
125mm
inline petrol
4 3 2 1 Shellite 20%
5 3 2 1 petrol Injectors at centre of intake.
6 3 2 1 petrol 5mm in
7 3 2 1
kerosene fuel
test
8 4 3 2 petrol
9 5 4 3 petrol
10 5 4 3 10% methanol Results too good (error?)
11 5 4 3 10% methanol retest
12 4 injectors 10% methanol
13 3 opposed 10% methanol

14
F W E
engine 5 4 3 10% methanol

275

Fuel density 0.743
Fuel usage fuel used
time
rant thrust sfc
weight before weight after fuel weight used fuel used (ml) ml/min
Test
1 530 355 175 235.53 235.53 175 60 2 5.25
2 355 160 195 262.45 224.96 195 70 2.2 4.56
3 870 730 140 188.43 173.93 140 65 1.3 5.96
4 330 140 190 255.72 264.54 190 58 2.2 5.36
5 560 360 200 269.18 248.47 200 65 2.1 5.27
6 360 215 145 195.15 243.94 145 48 2 5.44
7 650 545 105 141.32 256.94 105 33 2 5.73
8 365 235 130 174.97 244.14 130 43 2.25 4.84
9 885 685 200 269.18 323.01 200 50 2.2 6.55
10 215 95 120 161.51 161.51 120 60 2.25 3.20
11 905 645 260 349.93 247.01 260 85 2.25 4.89
12 715 525 190 255.72 306.86 190 50 2.1 6.51
13 515 255 260 349.93 299.94 260 70 2.1 6.37

14 625 525 100 134.59 179.45 100 45 1.5 5.33

276
Date: 7 August
Time: 8.30am 5pm
Testing Personnel:
Ryan Anderson
Terry Walladge

The following tables are records of the tests done on the Chinese engine. For these
tests, the engine was cut at both the exhaust and intake to allow for modification of its
geometry. The rough edge resulting from the extension was smoothed on the intake
with the addition exhaust sealant in the gap, resulting in a reasonable surface finish.
Due to the size of the exhaust, this was not done, resulting in a step where the joint
occurs. This is thought to be the main cause of the reduction in thrust.

Thrust (kg)
test no
intake
(mm)
Exhaust
extensio
n length
(mm)
exhaust
(mm)
propan
e
liqui
d comment
1 0 0 100 unable to sustain without air
2 0 270 370
Aluminum extension failed before
thrust measurement could be taken.
A large hole was made in the side of
the exhaust extension
3 0 30 130
new extension - wouldnt sustain
without air, flamed out easily on
liquid, sounded very rough
4 0 50 150
wouldnt sustain without air, still
rough sounding
5 0 70 170
wouldnt sustain without air, very
close to sustained operation however
(did not stop immediately on air shut
off)
6 0 77 177 1.5
Sustained on fuel, flamed out at
7volts
7 0 115 215 2.2 1.5
Comparison test with original section
replaced. Produced much less thrust
than original engine. Unable to flame
out on propane, flamed out at
7.5volts on liquid
8 0 135 235 1.5 1.5 normal run, no issues
9 0 145 245 1.5 1.5 could not flame out on liquid
10 0 165 265 1.5
wouldnt sustain on propane - could
not flame out on liquid even at 8.9
277
volts
11 0 165 265 1.7
Replaced '6' injector with an '8'. Still
unable to flame out at 9 volts
12 0 185 285 1.5 9 volts, unable to flame out
13 5 185 285 1.5
Increased intake length by 5mm.
Wouldnt sustain on propane,
sounded rough. Unable to flame out
on liquid
14 5 20 120 ran on propane, but not on liquid
15 5 45 145 2.2 1.5 normal run, no issues
16 5 115 215 2
flamed out on liquid, best result so
far
benchmar
k 215 3.6 2.2

The addition of the 5mm extension on the intake appears to improve the engine
performance. However the actual thrust obtained was less than the original engine.
The modification would need to be made in the workshop (to a much better surface
finish) and retested in order to determine if the apparent improvements are consistent
with the actual performance of the engine without modification damage. The exhaust
length appears to be at least close to the ideal value, with further testing required to
verify this.

Figure 159 highlights the variation in performance seen in the testing.
278

Figure 159 - Engine performance on liquid fuels
Appendix I Fuselage Stress Analysis
279
The following stress analysis of the fuselage was carried out to determine the required
thicknesses of the fuselage. It was carried out using the load factor of 3.8 and a safety
factor of 2.25. For simplification, it was assumed that the fibre-glass skin took the
shear load while the longerons took the bending load, which gave the shear and
moment distributions.

The shear stresses in the fuselage skin were determined using Jourawskis formula:

Where V is the shear force, Q and I are the first and second moment of area of the
fuselage skin, and t is the fuselage thickness.

Various thicknesses of fibre-glass were assumed, on the basis of .25 mm per ply:
three-ply fibre-glass with an approximate thickness of 0.75 mm, two-ply with a
thickness of 0.5 mm and one-ply with a thickness of 0.25 mm. Assuming thin walls
then the thickness, t, is directly proportional to Q and I as they are directly
proportional to the sectional area (Q is also proportional to the distribution of the
sectional area by displacement y and I is proportional to the square of the distribution
of the sectional area by y
2
these will stay much the same for a thin wall structure).
Since such changes in Q and I will cancel out, then the wall thickness is simply inversely
proportional to the shear stress.

While, as the area of the longerons was small, the second moment of area for the
longerons was estimated using the following simplification:

Where y is the distance of each longeron from the neutral axis of the longeron
arrangement and A is the area of each longeron. Longerons of 7mm diameter and 1.5
mm thickness were assumed actual longerons of larger diameter and thickness will
reduce the stresses.

280
The bending stresses were determined using the flexure formula:

Where y is the distance from the neutral axis, and only the maximum distance from
the neutral axis was considered.

The stresses, analysed at critical locations and evenly spaced positions either side,
were estimated to be the following:

Station
Position
(m)
Shear
(N)
Moment
(Nm)
Shear
Stress
3-ply
(MPa)
Shear
Stress
2-ply
(MPa)
Shear
Stress
1-ply
(MPa)
Max
Bending
Stress
(MPa)
1 0.09 -0.50 -0.01 -0.01 -0.02 -0.04 0.00
2 0.20 -33.38 -2.04 -0.60 -0.89 -1.79 -1.19
3 0.59 -106.77 -30.32 -2.25 -3.38 -6.76 -17.68
4 0.68 -251.88 -41.38 -5.32 -7.98 -15.95 -24.13
5 0.71 482.01 -41.60 10.17 15.26 30.52 -24.25
6 0.76 347.22 -20.23 7.33 10.99 21.99 -11.80
7 1.00 288.50 53.12 6.09 9.13 18.27 30.97
8 1.30 215.11 121.64 13.88 20.83 41.65 70.91
Appendix J- Gantt Charts
281
ID Task Name Dur ation Start Finish Predecessors
1 Res earch and Benchmark ing 31 days? Mon 3/12/07 Mon 14/01/08
2 Pul sejet Background 31 days? Mon 3/12/07 Mon 14/01/08
3 Puls ejet Operational Cycles 15 days? Mon 3/12/ 07 Fri 21/12/ 07
4 Last Year s Research 15 days? Mon 3/12/ 07 Fri 21/12/ 07
5 Types of Pulsejet Engines 14 days? Fri 7/12/07 Wed 26/12/07
6 Find Papers of Pulsejet Research 27 days? Fri 7/12/07 Mon 14/01/08
7 Air craft Research 16 days? Mon 3/12/07 Mon 24/12/07
8 Puls ejet Aircraft 15 days? Mon 3/12/ 07 Fri 21/12/ 07
9 Statistics on Tar get Dr one UAVs 12 days? Fri 7/12/07 Mon 24/12/07
10 Analysis Techniques ( Roskam + Raymer )11 days? Fri 7/12/07 Fri 21/12/ 07
11 Air craft Prelim Desi gn 43 days? Tue 25/12/07 Thu 21/02/08
12 Statistical Design 11 days? Tue 25/12/07 Tue 8/01/ 089
13 Weight Calcs 11 days? Wed 9/01/08 Wed 23/01/0812
14 Aircraf t Concepts 21 days? Wed 9/01/08 Wed 6/02/0812
15 Mat ching Diagram 21 days? Thu 24/01/08 Thu 21/02/0813
16 Finalise Aircraf t Concepts 6 days? Thu 7/02/ 08 Thu 14/02/0814
17 Take off + Landing Methods 6 days? Mon 11/02/08 Mon 18/02/08
18 Pul sejet Design 82 days? Mon 21/01/08 Tue 13/05/082
19 Prepare Engine Development Plan 11 days? Mon 21/01/08 Mon 4/02/ 08
20 Prepare 2007 Engine For Testing 21 days? Tue 5/02/ 08 Tue 4/03/ 0819
21 Test 2007 Engine 2 days? Tue 26/02/08 Wed 27/02/08
22 Development of f lares f or 2007 engine 6 days? Tue 4/03/ 08 Tue 11/03/0821FS+3 days
23 Manf acture of f lares 10 days? Wed 12/03/08 Tue 25/03/0822
24 Expanding exhaust Researc h 20 days? Thu 28/02/08 Wed 26/03/0821
25 Chinese Engine De sign 33 days? Fri 28/03/ 08 Tue 13/05/08
26 Statistical Analysis 7 days? Fri 28/03/ 08 Mon 7/04/ 08
27 FEA 26 days? Tue 8/04/ 08 Tue 13/05/0826
28 Liquid Fuel Re search 46 days? Mon 10/03/08 Mon 12/05/08
29 Sys tem Requirements 11 days? Mon 10/03/08 Mon 24/03/08
30 Fuels 10 days? Tue 25/03/08 Mon 7/04/ 0829
31 Injector Researc h 20 days? Tue 8/04/ 08 Mon 5/05/ 08
mit ch
mit ch
ter ry,ryan,nick
ter ry
workshop
ter ry
nick
nick,ryan,terry
ryan
ryan
ryan,nick
26 3 10 17 24 31 7 14 21 28 4 11 18 25 3 10 17 24 31 7 14 21 28 5 12 19 26 2 9 16 23 30 7 14 21 28 4 11 18 25 1 8 15 22 29 6 13 20 27 3
Dec '07 Jan '08 Feb '08 Mar '08 Apr '08 May '08 Jun '08 Jul '08 Aug '08 Sep '08 Oct '08 Nov '08

282

31 Injector Research 20 days? Tue 8/04/08 Mon 5/05/08
32 Injector Design 6 days? Mon 5/05/08 Mon 12/05/08
33 Air craft Detaile d Design 67 days? Fri 22/02/08 Mon 26/05/08
34 Electronics Requirements 11 days? Fri 22/02/08 Fri 7/03/08
35 Fuse Shape (Finalise) 39 days? Mon 10/03/08 Thu 1/05/08
36 Aer ofoil selections 1 day? Mon 31/03/08 Mon 31/03/08
37 Wing Structual Design 20 days? Tue 1/04/08 Mon 28/04/08
38 Engine position and mounting method 14 days? Fri 2/05/08 Wed 21/05/08
39 Wing Connection 14 days? Fri 2/05/08 Wed 21/05/08
40 Tail Design 14 days? Fri 2/05/08 Wed 21/05/08
41 Fuse Stresses 14 days? Fri 2/05/08 Wed 21/05/08
42 Landing Gears 3 days Thu 22/05/08 Mon 26/05/08
43 Pulsejet Testing 69 days? Mon 21/04/08 Thu 24/07/08
44 Flar es 4 days? Mon 21/04/08 Thu 24/04/08
45 Expanding Exhaust 4 days? Mon 21/04/08 Thu 24/04/08
46 Chinese 4 days? Mon 21/04/08 Thu 24/04/08
47 Liquid Fuels 9 days? Mon 14/07/08 Thu 24/07/08
48 Manufacturing 135 days ? Tue 1/04/08 Mon 6/10/08
49 Wings 109 days ? Tue 1/04/08 Fri 29/08/ 08
50 Wing FoamCut 10 days? Tue 29/04/08 Mon 12/05/08
51 Source Wing Spars 21 days? Thu 22/05/08 Thu 19/06/08
52 Source Fibreglass 5 days Tue 1/04/08 Mon 7/04/08
53 Cut Wing Ribs 21 days? Mon 5/05/08 Mon 2/06/08
54 Assemble Wing Structure 1 day? Mon 21/07/08 Mon 21/07/08
55 Preperation and Glassing 29 days? Tue 22/07/08 Fri 29/08/08
56 Fus elage 81 days? Tue 27/05/08 Tue 16/09/08
57 Source Manufacturer 1 day? Tue 27/05/08 Tue 27/05/08
58 Source Wood For Plug 1 day? Wed 28/05/08 Wed 28/05/08
59 CNC Plugs 57 days? Thu 29/05/08 Fri 15/08/08
60 Sanding and Finishing 10 days? Mon 18/08/08 Fri 29/08/08
61 Make Moulds 2.5 days? Mon 1/09/08 Wed 3/09/08
ryan,nick
ryan,nick
kar n
mik e
mike,ter ry
nick
ryan
nick
mike
kar n
mit ch
nick,ryan,terry
nick,ryan,terry
nick,ryan,terry
nick,ryan,terry
nick,terr y,ryan
nick
nick
nick
John
nick
ryan
ryan,terr y
Mincham
Mincham,nick ,terry,ryan
1 4 7 10 13 16 19 22 25 28 2 5 8 11 14 17 20 23 26 29 1 4 7 10 13 16 19 22 25 28 1 4 7 10 13 16 19 22 25 28 31 3 6 9 12 15 18 21 24 27 30 3 6 9 12 15 18 21 24 27 30 2 5 8 11 14 17 20 23 26 29 1
February 2008 Mar ch 2008 Apr il 2008 May 2008 June 2008 July 2008 August 2008 September 2008

283

58 Source Wood For Plug 1 day? Wed 28/05/08 Wed 28/05/08
59 CNC Plugs 57 days? Thu 29/05/08 Fri 15/08/08
60 Sanding and Finishing 10 days? Mon 18/08/08 Fri 29/08/08
61 Make Moulds 2.5 days? Mon 1/09/ 08 Wed 3/09/08
62 Prepare Moulds f or Fuse Manuf act ure 3 days? Wed 3/09/08 Mon 8/09/ 08
63 Manf actur e Fuse 2.5 days? Mon 8/09/ 08 Wed 10/09/08
64 Sanding and f inishing 2 days? Thu 11/09/08 Fri 12/09/08
65 Joining 1 day? Mon 15/09/08 Mon 15/09/08
66 Finishing 1 day? Tue 16/09/08 Tue 16/09/08
67 Tai l 26.5 days ? Thu 14/08/08 Fri 19/09/ 08
68 Cut Foam 6 days? Thu 14/08/08 Thu 21/08/08
69 Prepare Surf aces 1 day? Fri 22/08/08 Fri 22/08/08
70 Glass Sections 1 day? Mon 25/08/08 Mon 25/08/08
71 Sanding and Finishing 1 day? Tue 26/08/08 Tue 26/08/08
72 Assembly 1.5 days? Thu 18/09/08 Fri 19/09/08
73 Air craft 11.5 days ? Fri 19/09/ 08 Mon 6/10/08
74 Assembly of tail to Fuse 1 day? Fri 19/09/08 Mon 22/09/08
75 Cut ting Wing holes and alignment 4.5 days? Mon 22/09/08 Fri 26/09/08
76 Engine Mount 1 day? Mon 6/10/ 08 Mon 6/10/ 08
77 Engine Duct 6 days? Mon 29/09/08 Mon 6/10/ 08
78 Landing Gear Attachments 1 day? Mon 6/10/ 08 Mon 6/10/ 08
79 Wing Mounts 1 day? Mon 6/10/ 08 Mon 6/10/ 08
80 Air craft Testing 8 days? Tue 7/10/ 08 Thu 16/10/08
81 Fan Thrust Test 1 day? Tue 7/10/08 Tue 7/10/08
82 Assembly Checks 1 day? Tue 7/10/08 Tue 7/10/08
83 Touch and Go Test 1 day? Tue 14/10/08 Tue 14/10/08
84 Final Flight Test 1 day? Thu 16/10/08 Thu 16/10/08
Mincham,nick ,terry,ryan
Bill ,nick,r yan
nick,terr y,ryan
Bill ,nick,t erry,r yan
nick,ryan
nick,Bill,mitch
nick
Bob
Murray
Murray
S M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S M T W T F S S
1 Sep '08 8 Sep '08 15 Sep '08 22 Sep '08 29 Sep '08 6 Oct '08 13 Oct '08
284
Appendix K- Risk Register
285
The Risk

The consequences of an
event happening
Suggested Controls

Level of
Risk
Consequence Likelihood
Aircraft Crash Catastrophic

Possible

-Manufacture Spare Flight
Surfaces.
-Reusable fuselage mould.
Extreme
Unsuccessful
Development of
Liquid Fuel
System
Major Possible -Ensure Backup Propane
System is available
-Commercial Liquid fuel
system purchased
High
Vibrations
interfering with
electronics or
aircraft structure
Major Likely -Design vibration isolation
system for mounting.
-Vibration isolation for all
electronic components
-Extensive Pre-Flight
Testing.
Extreme
Failure to reach
target speeds or
flight time
Moderate Possible -Design for possibility of
belly landing and removable
landing gear
-Research Augmentor
Design
High
Unavailable
flight testing
facility
Major Possible -Investigate the possibility
of flying at Lake Ayre
-Assess possibility of using
Terrys farm in Keith
High
286

High
temperature of
engine effecting
composite
aircraft structure
Moderate Possible -Investigation of heat
shielding ducting to
increase airflow around the
engine
-Use of High Temperature
Resins
-Pre-Flight Testing
High
Slow
Development
Times
Minor Possible -Ensure Continuing use of
Gantt Charts and other
project management
techniques
Medium
Engine Flameout
During Flight
Major Unlikely -Extensive pre flight testing
with induced air flow.
-Voltage Limit on pump to
restrict high fuel flow
-Run engine at maximum
90% throttle
High
Unable to reach
3kg of Thrust
with FWE engine
Major likely -Development of alternative
engine designs.
-Investigation into
augmentor design
-Expanding Tail Pipe design
Medium
Appendix L- Meeting Minutes
287

Pulsejet UAV level IV Project Group

Date: 23/11/08
Time: 1pm-2.30pm
Minutes: Nick Lukacs

Present: Maziar Arjomandi
Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher

Absent: none

Apologies: none

Topics Discussed:

o Use last years engine, test by end of 07, design such that it can be used on the
aircraft if required.
o Need 6-8 people by the end of next week
o Technical Manager and Financial Manager appointed:
TM Nick Lukacs
FM Michael Sipols
o Aeronuatical II books for Karn and Michael
Raymer, D.P aircraft design, a conceptual approach
Roskam, J Airplane design pt 1 + pt 2
lecture handouts
o Begin aircraft concept
look at small aircraft/models etc
288
Adelaide model airclub
prepare 20 aircraft concepts (spreadsheet)
engine type, thrust, weight etc

o Engine (Nick and Ryan)
Research liquid fuel pulsejets/possible delivery systems
Types of liquid fuels (specific energy and air fuel ratios
and safety)
Prepare table (good detail)

o Engine (Terry)
Drawing of engine to make (based on last years design)

o Log Book (everyone)
Big folder to include:
Calculations
Internal meeting records
General note taking

o Project Definition
2 pages
Initial idea
Submitted by TM
Gantt chart (look at software for this, poss. Freeware)
o Planned from Dec 1, monthly. 20ish tasks
o Check MyUni for deliverables

o List of Potential Sponsors
Inc. possible contact people
Prepare presentation (eventually)
Think about what we could gain from each sponsor ie.
Parts, small money, big money(applications)

o By Monday have a time planned for meeting with Maziar

Tasks issued:

289
Technical
Michael and Karn
Research aircraft designs (by next meeting)
Get notes and borrow books
Nick and Ryan:
Research possible delivery systems for liquid fuels and types
of liquid fuels
Terry:
Drawings of old pulsejet for manufacture

Management
All
Organise logbook
List of possible sponsors (based on personal contacts)
Gantt chart freeware
Nick
Start Project definition
Karn
Check times of deliverables
Ryan/Terry
Recruitment of other members
290

Date: 29/11/07
Time: 5pm-6pm
Minutes: Karn Schumacher

Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher

Absent: none

Apologies: Michell OCallaghan

Topics Discussed:

o Review of last weeks meeting
o Sponsorship
Nick has books listing defense companies and access to place
notice that can be viewed by 100 defense companies.
o Review of last weeks technical tasks:
o Fuels
Candidates starting point: methyl oxide & propyl oxide, and
nitromethane.
o Injection
Atomisation
Direct Injection (requires pressurised fuel feed, obstacle with
liquid fuel)
o Pulsejet designs
Chinese
Other bigger design
o Last years engine
1035 mm length, 2.392 kg weight
o Engine
291
Pulsejet lengths and widths methods [statistics, program
(Excel) ]
Last year they made the pulsejet before their understanding
was complete.
Smaller increments in length of pulsejet for better data this
year?
Last year is 12 mm, this year try 20 gauge (8mm) thickness?
Test engine thickness
o temperature concerns only
o reduce weight of engine (refer last years ANSYS)
o Review of last weeks organisational tasks:
o Log books in order.
o Financial reports in order.
o Deliverables schedule is in table for now.
o Organisation
o Last years report in pdf format.
o Information:
Headings, citations for pieces of information through e-mail.
Technical manager and others collect all the pieces of info and
store them under folders like pulsejet, fuel, aircraft etc.
Not just reporting but decision and reason (justifiable).
o Extended Aircraft Report
o Consider by (relevant) aircraft class
o Consider Jet UAVs 1 kg to 20 kg
o All parameters you can find (vel., cruise speed, takeoff weight, range)
o James(?) All the World Aircraft (also All the world engine, all the world UAV)
o Roskam can be purchased from publisher at darcorp.com or has been ordered
into Unibooks in February, (Parts I and II needed for now)
o W
O
W
E
curve
o Long report
Generate curves, sizing etc.
(Specify/decide) Type of aircraft:
Want this type of engine, this type of
wing, this type of tail etc..
Tasks:
Technical
Pulsejet
Proposal (Engine)
Modifications, thickness and drawings
(from second engine of last years if can get it running)

Continuing with understanding last years report and design,
fuels etc.
(Terry, Nick)
292

Specifics on direct injection (Ryan)

Aircraft
Long aircraft report (Michael, Karn, Mitchell)

Management
Draft of Project definition (Ryan)
Nick had some brief ideas to start

List of companies
Listed in table with the name, why, type (in-kind,
cash, contract) and rank (5 rank 1s, 5 rank 2s...).

Powerpoint sponsorship presentation (draft)
10-15 slides
Applications?

Bill of material estimate of money needed
293

Date: 06/12/07
Time: 5pm-6pm
Minutes: Ryan Anderson

Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher

Absent: Mitchell OCallaghan

Apologies: Michell OCallaghan (via email)

Topics Discussed:

o Review of last weeks technical tasks:
o CAD
CAD of 2007 engine nearly done. A few areas still
needing to be cleaned up. (Terry)
o Liquid fuel injection
For liquid fuels to work, the fuel needs to be heated
before entering the combustion chamber problems
when starting a cold engine. Need to look at different
options for heating the engine either using a heating
element or consider starting the engine using gas.
o Aircraft Comparisons
Comparisons were done largely on propeller aircraft
because information on pulsejet information is had to
come by. However, info. On pulsejet aircraft is critical.
Nick offered to join this area of research for the time
being, to help in finding the desired info.

o Review of last weeks organisational tasks:
294

o Log books in order.
o Financial reports in order.
o Drafts for project definition and presentation have been started. More
work needed
o An initial sponsorship list was put together. More work needed
o Using Microsoft project for creation of Gantt chart

Tasks for this week

Organisational Tasks

o Keeping in touch:
Need to put in a lot more effort into our communication
skills throughout the group. All members need to send
more emails, to ensure everyone is well informed.
(all)
o Ensure all members of the group have access to a pdf of last years
report (soft copy given to Terry)
o Continue with project definition (Ryan)
Feedback on the current draft wanted from all members
(all)
o Continue draft of PowerPoint sponsorship presentation (Ryan)
Feedback on the current draft wanted from all members
(all)
o List of companies
Continue preparation of company sponsorship list. More
wanted for the reasons why we would seek funding
from the companies. Also need to finish ranking
companies ie favourite least favourite. (Nick)
o Bill of materials and general costing (Mitch)
o Continued updating and improvement of Gantt chart

Technical Tasks

Pulsejet Engine
Contact Adelaide University workshop supervisor Richard
Pateman. Talk to him about our project and get an idea of the
level of work that can be undertaken in the workshop.
Design flanges to be attached to the inlet and exhaust tubes on
the 2
nd
engine from 2007. Try and organize, with the workshop,
a time when the flanges could be added (Terry)
295
Consider a more in-depth construction P.O.A for the next week
or two. Have this done by Monday and send to Maziar.
Begin initial planning for testing of both engines designed in
2007. This requires every member of the group to carefully read
the sections of last years report regarding testing. (all)
Continue research into liquid fuel systems (Ryan)
Continue general research into pulsejet designs (all)

UAV
Literature review investigate all aircraft powered by pulsejets
and small convention jet engines (DO NOT look at other types of
planes ie propeller/big passenger planes). Need to focus
research on small aircraft ie similar in size to what we are aiming
at designing. Attempt to find pictures, as well as technical
information on as many as possible, to allow comparison and
analysis of different designs. This should help decision making
for our design. (all aero)
Investigate likely wind and thrust loadings that will be present in
the design
Begin sketches of possible designs (Mike)
Look at costing of the UAV materials/labour etc (Mitch)
The following was a task for last week, but was not done
satisfactorily therefore more work is still required in this are:
Extended Aircraft Report (long report)
Consider by (relevant) aircraft class
Consider Jet UAVs 1 kg to 20 kg
All parameters you can find (vel., cruise speed, takeoff
weight, range)
Janes All the World Aircraft (also All the world engine,
all the world UAV)
Roskam can be purchased from publisher at darcorp.com
or has been ordered into Unibooks in February, (Parts I
and II needed for now)
W
O
W
E
curve
Generate curves, sizing etc.
(Specify/decide) Type of aircraft:
Want this type of engine, this type
of wing, this type of tail etc..
296

Date: 13/12/07
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher
Mitchell OCallaghan

Absent: none

Apologies: none

Topics Discussed:

Review of last weeks tasks:
o Nick reviewed all the pulsejet UAVs.
o Terry prepared drawings and has spoken to Bob at the workshop who
preferred spun stainless steel flange.
o Propose sliding adjustable design, which is preferred.
o Flare reduces flow losses.
o Manufacture flange in first week of January.
o Project definition, presentation and manufacturing schedule has been prepared
by Ryan.

o Sponsorship, Mitchell knows someone at BAE.
Some sort of letter for a consistent approach when contacting different
companies.
o For testing:
297
Safety Officer (Issues, procedures, paperwork. Eg. Procedure if
flame etc.)
Test (Operations) Manager
Video
Emergency mop
High voltage.
Strain gauge.
Safety forms, school needs to be open.
Compressor etc.

Tasks:

A3 Overhead view (and side view?) fully dimensioned, all angles & reasoning

Designed using statistics:

o e.g. Angle of sweep WHY?

o MQM, Lockheed Martin professional UAVs are more optimal than
hobbyist ones

o For configuration - 10 kg or 100kg UAV statistics does not matter.

o Janes All the worlds UAVs electronic copy?

Statistics
Designed using
Not textbooks
W/S, T/W
All UAVs
Any jet(? Or any)
10kg 100 kg
weight configuration
pulsejet
298

Date: 20/12/07
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher


Apologies: Mitchell OCallaghan (via email)

Topics Discussed:

Pulsejet

o Irvine? 20 lbs thrust, expansion zones.
o Talked to Richard Pateman.
o When fitted to plane is there a need to direct air into intake, so engine doesnt
stall, as there is a low pressure region behind engine due to the airflow.
o Need to build an air intake?
o Bell mouth:
o Elliptical shape best by 3%.
o Bell mouth shapes: parabola, etc.
o NACA inlet airfoils.

Actions
o Tolerance for inside diameter of new flange, notes for parts that are to be
assembled together.
o Signature of drawer.
299

Sponsorship progress

o Composites manufacturer (supplies UniSA SAE) can supply carbon fibre and self
manufacture.
o BAE senior contact to reply in January.
o Entech electronics machine shop access.
o Next year there is a proposal: $2000 a student, charge $50 an hour for
workshop jobs. Encourage to use workshops, specialists etc. outside the
University when its better.

Actions
o Beginning of January, template preparation of grant application. (Sir Ross &
Keith Smith Fund)
o (End of Jan., decisions from some sponsors, prefer many small sponsors using
tax deductions.)

Testing

o Terry Engine manager
Ryan Test manager
Nick Safety manager

Actions
o Water and power is at site, take a fire extinguisher from school.
o Contact school office for access to S237.

[Safety]
o Video camera & feed needed, not allowed to stand in chamber.
o Last years safety reports/notes Josh.

[Test manager]
o Test procedure.
o Table to collect data.
o Risk assessment what if it doesnt work.

[Engine]
o The engine can be cut and shortened if needed (built into risk assessment).
300

[Test form]
o Fill in
o School.
o Supervisor signs.
o Mobile numbers.

Airframe

o Pulsejet mounting? Al-foil shielding?

Actions
o Try to go into the structure of the aircraft.
o For example, know how to design bolt, bending beam dont know how to
design aircraft.
o E.g. actuator, wires, position of batteries, spars etc.
o Dont put dimensions.
o (Wing plan form needs to be fixed.)
o Try to put as much detail as you can.

Bill of material

o Eg. 125 parts

Drawing
No.
Item Manufacturing Purchasing
(material)
Tooling Testing
Assembly
Subassembly

(e.g. wing
section,
fuselage)
(have
correlation
between
part &
drawing)
(item
name)

Total price for each subassembly
Total price overall
301

e.g. An actuator has a purchase cost and manufacturing cost is zero.

Management

o Sections: wings, fuselage, landing gear, mounting of pulsejet.
o Structure: primary job, related secondary job.
o Everyone takes actions & initials at meetings?

Actions

o Task: Everyone
o Write down something that could go in report (at least dot points) of
what you have done so far including websites & references where you
got stuff from.

Next meeting is 11/1/08
302

Date: 24/1/08
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher
Mitchell OCallaghan

Absent: none

Apologies: none

Topics Discussed:

Sponsorship progress

o Email from DSTO supervisor.
o Contacted Messenger, they took details.
o Contact at Advertiser was on lunch, details were taken.
o Video of pulsejet running.
o FGI.

Actions
o Final Smith Fund application prepared by the end of next week. (1/2/08)
o Contact Adelaidean.

Pulsejet

303
o Workshop done with the flanges.
o Review of 2007 report started.
o Method of characteristics (textbooks, talk to lecturers in that area) OR CFD
approach instead.
o Lockheed empirical spreadsheet for other design off pulsejet forums.
o Goals: - version of code
- combustion

Actions
o Method of characteristics.

Testing
o Test procedure has been started.
o Parts list for testing has been compiled.

- Progress with testing as quickly as possible, in case unforeseen issues
arise.
- Allow two weeks after first visit and mock set-up.
- Attempt to bring equipment for thrust measurements.

Actions
o Contact school office for access to S237.
o Contact Joe Walker safety officer at the school, mobile number.

Management

o In industry and project want to support CFD, FEA with statistics, theory, hand
calculations etc.

Actions
o Need A4 page, project summary. [Goes to school website]

Airframe:

Test procedure

o Test structures (see iSOAR)
304
o Install pulsejet engine or shape of pulsejet , test on car at 100, 110, 150? kph
to check drag.
o Devise a stand to check behaviors of the aircraft.
o Then install actuators, test control surfaces.

o Need propeller piston engine (budget? & need to modify the fuselage or make
two fuselages)
o OR run engine on aircraft on car, see if it works at 110 kph instead (better).

Actions
o Make test procedure.

Gantt Chart

o All should be on the same Gantt chart.
o Increase number of operations.
o Remove long on-going objectives (e.g. R&D pulsejet, aircraft)
o Task for each week, for each month.

Actions
o Revise Gantt chart, ensuring to break tasks of 6 months into months and then
weeks, so they are more manageable, progress more easily tracked etc.


o Explanations for weightings, in matrices.
o All UAVs should have parachutes according to CASA.

Actions
o Redo the 3-plan view with latest statistics.
o According to general configuration text, 20 comments why this wont work, and
a response why the comment is wrong.

Weight Calculations

o WF=.82 kg, W0=10.7 kg
o Replace coefficients (agricultural) with ones for piston engine UAV OR the
collected statistics (jet).
o SFC 4-7 on last years pulsejet
305
o W0 vs L/D method doesnt actually work

Actions
o Redo with a & c values for piston UAV.
o SFC vs. T/O.

306


Date: 31/1/08
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Terry Walladge
Michael Sipols
Karn Schumacher
Mitchell OCallaghan

Absent: none

Apologies: none

Topics Discussed:

Management

WBS Work Breakdown Structure prepared.
Grant Application three quarters.

Actions
1 page project definition (project summary)

Aircraft

General configuration - finished
307
13 kg WTO, 2.8 kg fuel
Piston engine, pulsejet fuel consumption.
Checking?

Statistics review
90% between .3 and .6 thrust loading.
Wing loading: - 10/20 to 700 kg/sq m.
No clear pattern
100 kg/ sq m

Matching diagram
Relationship between wing loading and thrust loading for different
configurations of aircraft [e.g. weight 10 kg, L/D 9, thrust of such and such etc.]
Wing loading to get cruise of 300 kph
Wing loading to get climb of such and such
Boundaries defined by cruise speed, climb, stall etc.
o At the end 2-3, 4-5 intersections
o Find the range that contains the possible aircraft
Refer:
o Roskam Part 1
Unibooks
o Lecture notes MyUni
Contains full explanation of matching diagram

Three (four) plan view
3 view with lofting/sections
For example 5 cuts.

Actions
1. Weight Calculations check?
2. Matching diagram
3. Three (four) view sketch with section views.
4. Composites research different materials and manufacturing methods.

Pulsejet

308
Fuel review conducted:
Energy values Propane highest
Methane, similar power output, more fuel weight, less tank weight as no gas
tank.
High pressure fuel pump may significantly improve efficiency
Gas may have problem with flow rate dropping if pressure drops.
Usually valveless pulsejets are not throttle-able.
Extensive fuel research will be needed.

Actions
Start on liquid fuel theory.

Testing

Mock set-up at 7
th
Februrary [12 OClock]
o S237 can put in access request via e-mail.
Senior technician of electronic workshop
o Strain gauge, data logger
Propose testing during O Week, talk to Vicky.

Sponsorship

BAE meeting being arranged.
DSTO.

309

Date: 14/2/08
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan

Apologies: Terry Walladge

Topics Discussed:

Management

FGI Composites quote obtained.
Manufacture:
o Wing:
Wire cutting - foam cores.
CNC machine.
o Fuselage.
Foam core nose.
Middle has structural support.
Back: Firewall and fuel tank.
Fuselage: Remove a foam core or mold.
How to join wing and fuselage? Depends on configuration. E.g. Aluminium tongue on
iSOAR.
Can have access to read through the Aero II project on a Pulsejet UAV (high
maneuverable) target craft.

Airframe:

3 plan view
310
size of tail is related to fuselage
1500 mm length
190 mm diameter [fits 120 nm dia. Fuel tank and parachute]
o fuel tank placed at centre of mass, parachute placed for clearance needs
low wing
fuel pump ~50g with 200g batteries

Want little difference in stability between full fuel and empty.

A layout to
consider?

Half drag.
Aligns thrust with axis of aircraft.

Or cut pulsejet into fuselage?

Matching Diagram

Climb and
TO curves
Cruise
Stall
A
B
311
Internals

4 servos, battery pack, receivers, 300g-400g
4 channels $300
Probably no more than 5 channels needed.

Configuration & structure

Trolley launch?
Protection for wing in parachute fall?

Airframe Actions

2-3 designs each (hand-drawn) for internal meeting.
o Some dimensions?, scale paper? to fit parts.

Matching diagram critical.

Sketches and internal design.

312

Date: 21/2/08
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Apologies: none

Topics Discussed:

Testing

Testing on Tuesday and Wednesday (26 & 27 February).
Have camera, need a cable to connect to computer or T.V. monitor.

Sponsorship

Waiting for BAE contact
Meeting to clarify how the three UAV projects will appear in Adelaidean.

Airframe

Tables of values: [Aircraft B] W=10.2kg?, T=2.9kg?, S=.4m
2
?, climb angle, C
Lcruise
,
C
D0 cruise
, C
D0 take-off
, L/D
cruise
(not more than 15) [L/D
TO
, stall speed, top speed ?]

313
Conservative lift coefficient and L/D (more C
D
, less C
L
)

Carpet Plot Method, last chapter from Raymer.

(Matching diagram) Comparison of points A and B:

A is more forgiving in build tolerance
A is better in terms of aerodynamics.
B is more likely to crash.
B is lighter.

(That the high wing loading is structurally more complex is not significant for us
as both wing loadings are low.)

[Move the point B by applying a takeoff safety factor e.g. 20%, and also some
safety factor for stall line.]

Parameters like C
DO
, C
L
quite critical.

Parachute, appoint a person for the parachute?

Scale sketches, have a list of approximate internal component dimensions.

Actions
Calculate the table of values.
Parachute.
Fuselage size [based on internals, pumps close to fuel tank, items located
laterally on central axis as much as possible].
Sketches (scale drawings which fit the parts).
(and look at Carpet Plot, Raymer).
314

Date: 28/2/08
Time: 5pm-6pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Apologies: none

Topics Discussed:

Testing

Tested on Tuesday and Wednesday (26 & 27 February).
Both engines ran.
Thrust measured was lower than last years results, maximum sustained thrust
of just under 2 kg from the smaller engine.
With the addition of flares, the second engine ran first go, and ran in every
configuration, even with the intake flares removed. Was louder than first
engine but sounded less crisp.
Had to isolate data logger [max 10 V] from spark plug [20,000 V].

Actions
Roadmap Pulsejet test plan.

Airframe

315
Fixed weights Engine weight, SFC
For engine weight and SFC use the number for manufacturers and
add/reduce 20%?

8 Kg (T.O.), 3 kg Thrust, .375 T/W

3.5 kg for pulsejet and fuel

Actions
5 new sketches, 1 each.
All the components, weights.

Item Weight X
COG

1 Pulsejet
2 .

o Then look at the other configurations and calculate check if COG stays
within 25% to 35% MAC for each of: full and completely empty tank.
o So will have 5-6 sketches with pros and cons for each.

Matching diagram, 5 slide presentation.

Aerofoils, 5 options for main wing (thin aerofoils).

Pros and Cons of Parachute versus Wheels (and basic designs for each).

Calculate sensitivity to fuel weight, fuel consumption and engine weight [e.g. If
engine was heavier then aircraft TO weight would be, wingspan would be
etc.]
Calculate empty weight, W
known
(incl. pulsejet) and W
unknown
(statistics or guess).
List of required R/C channels?
25 to 35% MAC 0
100
1.5 kg
200
3 kg
300
3.4 kg
316
Composite parts.
o How to manufacture?
o How to analyse?
o Maybe Aluminium/wood?
Start calculating wing.

1 Aerofoil.
2 Lift distribution.
3 Break wing into 3 zones A, B, C with description of proposed
structure for each zone.

Testing (Aircraft)

Pilot? The contact in R/C plane building was involved in a glider competition
with the University 2-3 years ago.

Sponsorship

ASC will sponsor.
BAE are difficult for progress.

A B C
317

Date: 7/3/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Apologies: none

Topics Discussed:

Sponsorship

ASC - $2000.
Australian Aerospace $1000.
o Does it clash with BAE?
BAE meeting 3 people
o Updated presentation, pictures, CAD plan, e.g.:
o 2 to present, 1 to audit.
Try to end with clear decision from company.
20 companies contacted.
Spring Valley foods contact.

Exhbition
At a Pavilion at the Showgrounds, advertised in papers, expect 10,000 to attend
(high schools invited etc).
One day for Mechanical, Electrical and Civil (Mechanical is largest display).
2 days to set-up Wednesday afternoon & Thursday with some time on early
Friday morning for expensive items.
318

Actions
Update presentation for BAE, send to supervisor on Monday.

Management

Structure: Technical Manager Nick, Airframe Manager Michael, Engine Manager
Terry.
Gantt chart for project prepared, consisting of two parts:
Pulsejet development
Airframe development.
Try to break down tasks to less than one week, 2-3 days.

Pulsejet

Tail pipe section and liquid fuel to develop.

Wing structure

Structure has been prepared and analysed.
Spar assumed to carry all bending moment (conservative).

Aerofoil

Aerofoil search completed.
6 aerofoils short-listed out of a thousand.
NACA 4412 seems quite stable.
SD aerofoil generates large moment but otherwise performance was better.
Most stalled at C
L
=1.55, generally want C
L
=1.4 as reduction in CL due to 3D effects
of wing.

Actions

Best C
L
/C
D
(tangent?) want to find angle of attack for stall.
Airfoil for nose see preferred concept.

Airframe

319
Empty weight fractions calculated.
o 6 kg empty weight.
o 4.5 kg for everything apart from pulsejet and fuel.

Tapered wing to be used, not more difficult to manufacture using wire cutting
foam.

More lift is needed on wing at take-off due to negative lift on tail to provide nose
up moment.
o e.g. Wing 8 kg lift, Tail 1-2 kg negative lift wing needs to produce more
lift.

Summary of Parameters
Stall Speed<60km/hr
Cruise Speed >260km/hr
Climb rate of 8.33% at 78km/hr
Takeoff distance<100m
Takeoff Weight=8kg
Fuel Weight=2kg
Specific Fuel Consumption=7.2kg/kg/hr
Clmax=1.2
Aspect Ratio=6.5

Sensitivity analysis presented.
Discussions of points A and B presented.

Actions

Sensitivity of performance parameters to W/S and T/W.
Find CL for the take off point (the lower CL at the selected point for the same take-
off length of 100m.)

Parachute v. Landing Gear

Development velocity for the UAV (150 kph) is much larger than a rocket (zero).
50:50 decision for now.

Actions

Parachute that is large enough but wont cause excessive deceleration.

320
Concepts

1. Aerofoil fuselage, pulsejet behind fuselage and empennage mounted on pulsejet
preferred option. Airfoil for
2. Standard fuselage with top mounted pulsejet. (Relatively short total length, to
meet COG requirements parachute must be mounted on the underside and
pulsejet must be brought forwards.)
3. Boom tail or similar frame to hold tail, with pulsejet behind fuselage. (The boom
or frame is thought to be un-necessary, total length is very long much longer
than first sketches - as parachute and fuselage need long lever arm to balance the
engine assuming fuel tank and wing are placed around the COG in the middle.)
4. Pulsejet mounted underneath. (Ruled out, total length when corrected for COG is
far too long, fuselage shape is too complex.)
5. Pulsejet cut into top side of fuselage. (Lower drag than option 2).

Prefer interface at bottom in high pressure zone of aerofoil.
o Use an Augmenter to improve air conditions at intake?
Use pulsejet as structure for tail?
o Control & heat resistance issues, but gives structural benefits.

Actions
List position of components and mass to find contribution to COG, calculate COG
to stay between 20% and 35% MAC for preferred concept.

Testing (airframe)

Actions
Ask if pilot contact can land the aircraft at 70 kph?
321

Date: 14/3/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Apologies: none

Topics Discussed:

Pulsejet

Pulsejet Development Presentation

Comparison of FEW & Chinese designs presented based on extensive research
and consultation with experts.
Last years project only compared FWE and Lockwood-Hiller designs.
FWE:
o FWE has poor combustion chamber shape and poor intake shape:
Only half of chamber works.
Poor air/fuel mixing.
Slow burning.
CFD simulation shows how the air travels to back of chamber
and mixes poorly.
No combustion energy added in the simulation.
Chinese:
o Combustion chamber:
Intake positioned to direct air for even mixing more thrust.
322
o Exit:
Megaphone shape.
Liquid fuel:
o Chinese works and can be throttled.
Fuels:
Petrol (more energy but burns hot implies low density).
Methanol (despite less energy burns cooler implies high
density and better mass flow).
Costing:
o Engine $300 (can roll any diameter pipe)
o Fuel pump $US 139.95 (flightworks)
o Injection nozzles 25-50 (for particular flow rate)
$AU 500 Total

Lots of testing can be done on FWE still.

Actions
Ryan, Irvine, Queensland.

Expansion angle calculations presented

For varying exhaust expansion angle Minimum pressure after out flow and
max. pres after in flow simulation shows maximum in flow pressure benefit
after about ~5.7 degrees.

Plots of pressure versus time presented [assumed 1000 degrees Celcius, but
from earlier comparisons feel the program gives a reasonable description of the
acoustic theory.]
P
R
E
ANGLE
5.7 deg 20 deg
323
o With the exhaust expansion there is a 40% greater peak difference in
pressure compared to without exhaust expansion.
o With exhaust expansion there appears to be less harmonics.

2 stroke exhaust expansion was investigated (wider range frequencies of
operations required in that case though).

Finding: Expansion on exhaust emphasizes pressure changes in combustion
chamber to help the pumping [more air, more fuel, more thrust].

Actions

To further investigate other modifications.
324
Aiframe

Wing load factor n? VLA, FAR23?

Actions

100 drawings complete drawings of aircraft.

Two weeks.

Full design.

o No FEA the calculations for afterwards, as they would take too much
time otherwise.

Manufacturing

Fixed wing (not removable).
CNC fuselage, wrap around some composite materials.
o Wooden (CNC) fuselage mold preferred, plastics will stick to the
composites while resin molds are very expensive.
Try to avoid molds if possible (apart from fuselage):
o Best method is try foam core for everything.
Terry can get CNC milling for free (900mm axis machine) as long as not too
much machine is used.

Actions

Obtain thick foam to cut the wing.

Other

Sponsorship

BAE e-mail sent to them this week to which they didnt respond, set new e-
mail of times that are available.

325
Management

Invoices? Satisfactory to pay in cash and then use a receipt.
326

Date: 28/3/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Apologies: none

Topics Discussed:

Sponsorship

o BAE meeting went well and was well prepared.
o Mechanical technical manager was there but not a financial person.

Engine

o To measure temperature distribution: Thermocouple? IR? Borrow from
combustion researchers?
o Engine is 65 cm long, 1.5 lbs mass and 8.1-8.2 lbs thrust with 0.5 mm stainless.

Actions

o Test plan, and reasoning for the test plan, for each of the engines.

327
Airframe

o Have section where can put payload and camera?
o Carbon tube booms likely to break on landing.

Actions

o Wing (Finalize shape of wing, foam and template).
o Report about manufacturing processes to be used.
o General configuration (landing gear etc.)
328
Airframe continued

Possibilities:

o No need for rudder?
o Vertical tail in form of H-tail without rudders?
o Mount rear wheels at ends of downward pointing vertical stabilizers?
o Boom should be connected to wing?
o Significant interference drag between pulsejet and boom? - Instead of booms
could put whole engine instead tube e.g. with a 1 inch clearance?

Landing?

o Risk management: landing gear has less risk?
o Extra weight would be needed for reinforcements for belly landing?
o Fairings on wheels?
o Can buy retractable landing gear, not needed?

Meetings

o Before decision:
o Meeting:
Look at variables.
Some calculations or understanding.
Try to categorise the variables.

o Build in some room for modifications?
329


Date: 18/4/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Absent: none

Apologies: none

Topics Discussed:

Testing

Signatures with regards to safety procedures and 2 checklists.
Infra-red temperature measuring system borrowed goes up to 1700 degrees.
[Testing next week.]

Report

Outline of preliminary report was presented.
Replace Literature Review with feasibility study: 1
st
part general literature
review or history, 2
nd
part main challenges (descriptive content includes: is it
330
good, is it bad, does it have history, why do you want to put a pulsejet on a
plane etc.)
Extension of project definition (either at the end, or in separate section
afterwards) description of why the parameters are as they are, (e.g. 10-12
minutes endurance is similar to that target drones).
o The project definition explanation and main challenges should be
linked.
Chapters 2 & 3 should be conceptual design of pulsejet and airframe
respectively, both chapters to end with a general 3-view drawing (not detailed).

Actions

Chapters and percent completions:
o Feasibility Study 100% complete
o Conceptual Design 100%
Pulsejet
Aircraft
o Detail Design 30%
Pulsejet
Aircraft
o Manufacturing and Materials Selection [60% Manufacturing, 100%
Materials]
o Experiments
Test Plan
Test Results
(mainly engine, but also aircraft test plan include place
and how others test aircraft).
o Financing and Management
Bill of materials
Sponsorship
Gantt chart
Management Policy
Safety and Risk Analysis
o Conclusion

Airframe

3-view presented
331
Static Margin (excellent):
Add 20% to all structural weights (wing, fuselage, horizontal tail and vertical
tail) and see how it fares additional curve with the worst case structural
weights.

Airframe manufacturing

Actions

Cut foam for wing.
332


Date: 18/4/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Karn Schumacher
Mitchell OCallaghan
Terry Walladge

Absent: none

Apologies: none

Topics Discussed:

Testing

Signatures with regards to safety procedures and 2 checklists.
Infra-red temperature measuring system borrowed goes up to 1700 degrees.
[Testing next week.]

Report

Outline of preliminary report was presented.
Replace Literature Review with feasibility study: 1
st
part general literature
review or history, 2
nd
part main challenges (descriptive content includes: is it
333
good, is it bad, does it have history, why do you want to put a pulsejet on a
plane etc.)
Extension of project definition (either at the end, or in separate section
afterwards) description of why the parameters are as they are, (e.g. 10-12
minutes endurance is similar to that target drones).
o The project definition explanation and main challenges should be
linked.
Chapters 2 & 3 should be conceptual design of pulsejet and airframe
respectively, both chapters to end with a general 3-view drawing (not detailed).

Actions

Chapters and percent completions:
o Feasibility Study 100% complete
o Conceptual Design 100%
Pulsejet
Aircraft
o Detail Design 30%
Pulsejet
Aircraft
o Manufacturing and Materials Selection [60% Manufacturing, 100%
Materials]
o Experiments
Test Plan
Test Results
(mainly engine, but also aircraft test plan include place
and how others test aircraft).
o Financing and Management
Bill of materials
Sponsorship
Gantt chart
Management Policy
Safety and Risk Analysis
o Conclusion

Airframe

3-view presented
334
Static Margin (excellent):
Add 20% to all structural weights (wing, fuselage, horizontal tail and vertical
tail) and see how it fares additional curve with the worst case structural
weights.

Airframe manufacturing

Actions

Cut foam for wing.
335


Date: 2/5/08
Time: 10am-11am
Minutes: Michael Sipols

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge

Absent: Karn Schumacher

Apologies: Karn Schumacher

Topics Discussed:

Sponsorship

Sir Ross and Sir Keith Smith Fund have put up $20000 for the three UAV
projects
We should get around $7000 of that
If any sponsors chase us we will still take their money but we dont need to
chase them

Advertising

When we have articles anywhere we need to mention all of our sponsors:
Smith fund, ASC, Australian Aerospace
Contact Adelaidian and get them to change the article to put our sponsors in
336
Try to contact the Advertiser and any other marketing opportunities: chance
for us to get known in the community and publicise school
Ryan is responsible for articles

Testing

Testing overview presented
Expansion for FWE engine tested, achieved 3.5kg of thrust
Lengths of intake and exhaust tested for FWE, optimal lengths determined
Only got one load cell fuel reading
Camera issues due to vibration interference
Footage from 1
st
test is coherent and can potentially be used for presentations
Analysis of tests is very important- Nick doing test analysis
Need to justify why engine lengths etc are better with theory also
Maziar wants a report on the testing to help convince Smith fund of the success
of our project
Send some movies of testing (if any are coherent) to Maziar

Test Setup

The test setup needs to be improved to allow better measurements
The hybrid rocket/other pulsejet from last year had a setup that may be more
useful for measurement of characteristics
However, the hybrid rocket setup may not be available as they were industry
sponsored so this needs to be checked

Manufacturing

Ryan bought enough foam for 3 wings: $60, 24g/m^3 density foam
FGI should have all we need for composites and the technical guy should be
able to help with manufacturing advice
We have the number of a guy who teaches composite manufacturing: should
be useful
End profiles have run into communications difficulties
The workshop only has Solid Edge V17 which is incompatible with the V19 and
V20 available
The end profiles might be best if they are transferred into .igs format
337
Wings will be cut as 2 pieces- wire cut by enthusiast

Ailerons
A small aileron will be less jumpy but might be able to lose control if wingtips
stall
Could use flaperons: control could be programmed into electronics relatively
easily
Would need better servos and rough clacs to work out efficiency of the flap
Wing-tail distance would definitely need to be increased if we were to use
these

Wing Detail
Could use vertical spars or cut the wing and glue together
May weaken the wing in torsion though
Could wrap UD carbon around aerofoil
Rod method may be the best option for connection to fuselage?

Vibration
Is a major issue
With liquid engined rockets which have similar vibration issues they use an
isolated engine bay
Fatigue failure may be an issue need to do engine-> fuselage mount perfectly
High optimal and high frequency vibrations are hard to damp
Maybe talk to Anthony Zander, Carl Howard or Colin Hanson about vibrations

Report
Everyone get into groups of 2, edit each others stuff before its sent to Nick
Make sure everything is ready for report before sending (refs etc)
Maziar wants a whole report so that it can be checked for structure, doesnt
care about little mistakes, technical stuff and format more important

Liquid Fuel
Pump chosen
Lots of academics have done liquid fuelling tests, might be able to just use one
of their systems
Fred Zackel has experience with diesel injectors- talk to him about them (eng
Nth 2
nd
floor , opposite lift/toilet)
338

Next Week
Plan on how to improve test stand
Sketches of installation of engine in aircraft
Progress of manufacturing and plan-list of issues
Any liquid fuel plans
Draft of report to Maziar
339


Date: 9/5/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent:

Apologies:

Topics Discussed:

Manufacturing - General

Workshop has time free this week and next week
Try to get stuff manufactured by mid year break
Maziars manufacturing plan (attached) discussed

Manufacturing - Fuse

Could use Isoar mould and modify it
Workshop usually CNCs core then makes outside mould from that
External workshop could make mould using one operation
Need to look at making fuselage manufacturable
340
More flexibility in wing position?

Manufacturing - Wing

Geometry finished
Control surfaces need to be finished designing
Need to finish internal structure
Getting help from model aircraft guy on joints and attachment
Joints are nearly done, need to finalise getting what we need

Test Setup

Amateur setup at the moment
Proximity of spark plug to load cell is a possible problem ? Maz seems to think
not biggest issue
Data logger doesnt have enough time*frequency for good testing (40secs at
800Hz)
Program for data logger is a bit poor
Load cell isnt rated for dynamc case possibly?
Vibration of engine may be being amplified in something (load cell?)?
Frequency cant be properly measured with load cell due to dynamic issues
(frequency of load cell is not good enough)
Post Grad Amin has a good datataker 145 datalogger that is independent of
stuff (generally found in S331?(possibly 231))
Could borrow this for next test to determine if the data logger is the problem
but is frequently used by pHd students
Ask guys in electronic workshop about frequency of load cell

Report
Was handed in by 4 oclock on the day
Peer assessment of between 1->3 for everyone for final prelim report
341


Date: 16/5/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge

Absent: Karn Schumacher

Apologies: Karn Schumacher

Topics Discussed:

General

Plane has been put into X-plane by Irvine
Task list given out for next few weeks

Engine Mounting

Terrys initial engine mounting turned out to make vibrations worse
Slab of rubber bought by Terry is not burnable but is too stiff to be useful
Shock mounts are used in aircraft and cars to isolate engine vibrations from
structure
Needs to be very soft
342
Maziar doesnt like the mounting position: transmitting loads through spark
plug on flat plate is bad stresswise
Just using a spring should be better than too much damping according to
vibration resonance graph
Mounting definitely needs to be inline with thrust or other problems are
introduced
Terrys new solution should be capable of solving the problem with a bit more
work
Could use washer at connection to reduce heat transfer
Zander, Howard and Cazzalato are all knowledgeable about vibrations
Tail

Tail isnt pretty enough
Being above the fuselage as it is is good
Need to increase arm of tail somehow to reduce downwash problems and tail
size
Could increase sweep on vertical tails
Can extend sidewalls of fuselage to increase moment arm for tail
Try to reduce size ratio to more like Sh=0.25Sw (more like 0.4 or 0.5 at the
moment)
Maziars idea of
fuselage extension

343
Report

Need to have Smith fund on front page and stuff off their website on the 2
nd

page
Title page is bad: see above
In the draft report, NW means that a section needs work, RW means it needs a
rewrite
The exec summary is too long: needs to be down to 1-1 and 1/2 pages
Needs serious restructuring-engine stuff is based too much on last years
project (use as a reference)
Use proper referencing, not just like The works of Raymer and Roskam)
See Myuni on page bordering
Some more proof reading required- if put references, need them to be decent
references
Table and picture formatting needs to be similar across report
Maximum of 3 steps on heading numbering: 1.2.3 and no more
Need to make sure stuff is linked properly, esp. pictures
When pictures are put in, they cant take up a page, if making them small
makes them illegible then find another picture
Restructure needed: stuff should go from general to detailed
Put references for pictures in the caption instead of in-text
Summaries for entire chapters only, not individual sections
Is a bit long, needs to be reduced somehow
Feasibility study has stuff which needs to be later, is way too long
There is no preliminary design for this project, that type of stuff is conceptual
design
Need to be more concise and certain with decisions
Decision matrix values need to be more justified
Need to start aircraft design conceptual section with the sketches we did way
back
Dont put pages in references, just author and year is needed
Only need safety factor for things that are uncertain in calculations
Put stuff in an order that makes sense
Report should have decisions at the end of each section, not just discussion
Someone should write a structure outline for report, not based on draft coz
draft was bad structure
Karn did some editing also of prelim report draft, on shared drive
Cant afford to spend whole week on report, need to do other stuff too
Conceptual design is everything up to a 3 view
344

Sponsorship
Hungry Jacks is going forwards- send them photos of exhibition
At exhibition there is a sponsors board they will get on and an awards dinner
with 200-300 people where sponsors names will be mentioned and there will
be a sponsors board
BAE are sending us a letter to say they wont be giving us any money
Publicity
We could try to get a joint paper into Engineers Australia for publicity
345


Date: 23/5/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher?

Absent:

Apologies:

Topics Discussed:

General

Report has been finished and handed in to front office

Sponsorship

Can get reimbursed for report printing from sponsors money
Hungry Jacks is looking likely, asking for $10K
Will need to write a report for the smith fund about what weve done with
their money so cant waste any of their money
Fuselage
346

Model hasnt been taking into account joints
Forget about interference drag-5% of total
Design fuselage to have sufficient space and have proper connections
Wing fuselage connection explained
Connecting composite I beam to bulkhead with spars wont work well
Need to try to increase contact area of spars
Need wing central body

Ducting

Near engine need to use metal at fuselage curved bit
Cavity between engine and fuselage is a problem, will create drag
Mitchell designing ducts for airflow from front, need a fairing over gap
All surfaces near engine need to be metal
Could use a lid type structure with aluminium foil skin over insulation over fibre
which is removable for access to fuselage internals
Side walls are bad

Manufacture
Cutting holes in UD carbon fibre is very bad for strength
Need to change fuselage to make more manufacturable
Need to get started into making the mould for fuselage
Minchim aviation is being helpful and can make mould in one action (straight to
wood)
Should do all composite manufacturing in one place
General Again
X-plane model made
X-plane not technical simulation, ant use it to tell how aircraft will fly
Useful for exhibition
Flight Gear is free download- better than X-plane but still not technical
347


Date: 6/6/08
Time: 10am-11am

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent:

Apologies:

Topics Discussed:

General

Report peer assessment results provided by Maziar (if you are not satisfied, talk
to Nick)

Manufacturing

Basically everything is designed to be sent to manufacturing we have about 4
weeks for manufacture
Then from way through July to mid August we need to put the plane together
348
The School wants to do more in-kind stuff because they cant afford to give
every student $2000 and have them spend it elsewhere
Get approval from Richard Pateman for stuff we outsource (to check that the
workshop couldnt do it or it was uneconomical for the workshop to do it)
Meeting with Mincham Aviation at Parafield airport Tuesday 2 oclock
Mincham made a uniSa solar car a few years ago and have lots of experience
with composites
Get a quote from them and then offer sponsorship to cut down price
Check Mincham can do vacuum bagging
Go to FGI afterwards to get wing stuff mention Adelaide uni and get cost
price
Injectors

2 types we can use- flat plate orifice or centrifugal
Trying to figure out hole diameter from known flow rate- finding spray angles,
velocities
Need to make sure fuel vaporises before reaching edge
Fred Zoekel has 20 ears experience designing injectors so send him an email
because hes only a part time academic
Workshop is scared by holes less than 1mm so if we need to manufacture these
ity will need to be outsourced
Cant just use regular injectors due to extra heat- diesel injectors would be
alright with heat but need heaps of pressure
See if we can buy a few different ones from Irvine, test them and work out
which works best
Can use two injectors in one engine- reduces fuel losses
In most jet engines 50x more air is used than fuel due to material properties
not allowing the heat of proper combustion
Kelso/Bassam/Gus/Zoekel are all useful people for this
Need to design setup such that it can be modified with different injectors
Wing Box/Fuselage

Wing box idea outlined
Need some changes to transmit loads dont use foam to transmit loads
See CAD for updated idea ( way too many sketches to put into minutes)
Need to design plates for fuse-wing connection perfectly
Skin should work as langeron to transmit load from tail to centre
Try to remove as little area as possible from skin for access to retain strength
(atleast 40% unaffected)
349
Having 3 openings for fuel would be better structurally than a massive cutout
Give Mincham several sketches for what can be done and several choices
Produce drawings with bulkheads etc for Mincham

Engine
Ideally should test soon (within the next few weeks)
Need a whole new test setup because of liquid fuels
General Again
Need to order electronics stuff (Karn making a list of whats required)
350
Date: 14/7/08
Time: 2pm-3pm
Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent:
Apologies:
Topics Discussed:
Wings
Need to get a time to get together with John to put the bits inside the wings
and get them glassed
Fuselage
Still waiting on Mincham to get back to us regarding manufacture but once they
are ready it should only take a day
Liquid Fuelling
We have injectors, brazing kit and ancillaries needed for test of injectors
For liquid fuels, the combustion chamber needs to be heated before running
2 options for this heating- 1. Start engine on propane and then go to liquid fuel
2. Heat it manually maybe problems with starting behind a wall if heated
manually as it cools down when starting it
However, theres a lot of safety issues with standing next to the engine when its
running, heaps more areas that safety procedures will need to be developed in
May have problems with composites when heating up the engine but these can
be solved using low temperature heating and should be no worse than when
engine is running
May be better to start on propane initially, dont need to have propane injector
in proper location as can just flamethrower the engine to heat it up
Could possibly use an electric heating element on the engine
Forget about standing next to the engine for tests-not going to happen
351

Testing
Testing planned, generally every second day from tomorrow until the end of
holidays
Fuel system will be mounted on a stool to provide proper height
Each test should use less than 1L of fuel
We can put the propane on the other side of the engine, nearer to wall access
so that if there is a failure in system, the gas can be shut off manually
Send Maziar all test documents before test
Fill paper safety procedure document out
Electronics
We have got most of the electronics
Dont have a GPS yet- dont need to order it as we can measure speed other
ways
The cheap servos we have are fine- dont need to buy more expensive ones
Need to be careful with connection between servo and hinge
Make aircraft with what we have- if we need to change anything afterwards it
should be easy
Landing Gear
Theres is one main manufacturer of landing gears in SA buy a front landing
gear and make connections to suit it
Wing Tests
Need to test the wing structure
Design a simple stand- could just be a block of wood
Use small, long sandbags with long direction parallel to chord
Load most of the wing chord but the 25% mostly (Aero centre)
Start with 50% required weight (SF*Aircraft weight)
Need to measure the deflection at the wing tip then check with what is
expected from calcs
If this is within 10% of expected then go up to 75% of weight
Then step up in smaller increments (10% or 5%) towards 100%
Only load these higher (75% and above) loads for a short time (3 seconds or so)
The first sign of composite delamination is the noise of failure so keep it quiet
Presentation
Everyone make their own outline of presentation with times estimate for
sections
Check on Maziars webpage and download past presentations to get the
structure of the presentations
352

Pilot
Need to fix pilot in stone
Isoar pilot is one of the best in Adelaide
He is doing the fuel cell plane (fuel cell are planning to go to Murray Bridge)
Need to have a flight test plan worked out
Could fly on ducted fan first
Open Day
Weve been listed on the open day presentations list but theres no force on us
to do it
Would be good experience but would take time
Presentation
Looking at getting augmentors made for FWE engine to show that augmentors
are good and can improve pulsejet thrust generally
353


Date: 21/7/08
Time: 2pm-3pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Terry Walladge
Karn Schumacher


Apologies: Mitchell OCallaghan

Manufacturing
Mincham, in contact for fuse. quote.
o Purchased hardwood.
o Bill could offer some advice for wings & fuse. temp issues etc.
To assemble & glass wing.
4 layers, to purchase composites from FGI.
Actions: Assem. drawing of wing.

Airframe
Test Stand
Support similar to fuselage.
Prepare sandbags small.
Risk assessment for aircraft.
Test whole system for electrical interference or other issues.
Test Plan
354
e.g. 10 tests: purpose of each test, what to measure etc. (all) equipment to
take.

Engine
Runs on petrol.
Wont transfer from propane to petrol when running.
Looking to recover thrust on petrol:
o Hasnt run flat out on petrol.
o Will test injector positions.
o Pump wont deliver sufficient pres.
Additional current draw from pump.
o Possibly back pressure or expansion of metal that reduces size of orifice.
Waiting for larger pump.

Presentation
Actions: First draft next week, 30-35 mins presentation.

Finances
Actions: Next week financial support.
355


Date: 28/7/08
Time: 2pm-3pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent:

Apologies:

Pulsejet

Injector position found.
Excellent sustained thrust.
Starting:
o Choking intake seems to increase air vel. & make starting easier
Run on propane to heat up, then starts on petrol.
Larger pump borrowed (on order).

Manufacturing

Wood & CAD ready for, and quote ready from Mincham.
o (50 % resin, fibre-glass only).
Wings assembled.
356
o Suggest 3 layers for whole wing, then 4
th
layer to 10 cm beyond rib, and
5
th
layer for 10 cm beyond root chord.

Flight Test

30-31
st
August
Ground test separate from airfield.
Ducted Fan
o 4 kg fan.
o 1 m above ground, talk to pilot, a close place to fly on ducted fan.
Place with long runway (1000 m) near Adelaide.
Heavy vehicle test required or not required? (Less than 7 kg without fuel).

Actions: Find definite place for ground test.
Presentation

White background, slide number etc.
Acknowledgements (first slide Smith Fund, University, School of Mechanical
Engineering)
35 Mins ( 6 x 5 min plus 1.5 min introduction and 1.5 min conclusion, 2 minutes
spare)
Money, financial issues and management in last 5 minutes
Citations for pictures or graphics from elsewhere
Each slide should have separate title and graph or picture.

357


Date: 11/8/08
Time: 3pm 4pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent:

Apologies:

Manufacturing

Fuselage Plugs done, waiting on fuselages
Wings curing required, 40 deg oven elsewhere.

Testing (Wing)

Once wings are cured.

Presentation

Layout & Introduction
358
No need for picture of plane on first slide.
Slide design:
o Smaller project title (footer)
o No need for University logo on every slide.
o Guide to presentation outline and position on every slide.
o Names in order of presenting.
o Photo of final design on project purpose.
o Flow chart layout at start with who will talk, for first presenter who also
presents project definition and introduction.
o Project goals after feasibility study.
o Slide titles in top-left.
o Font size 24 is too large.

Aircraft conceptual design
Summary of literature survey (aircraft) before config selection.
Then technical goals.
Add hand sketches to configuration selection slide, dont put final designs
until design stage.
Put statistics in (slide 10).
Stall speed, cruise speed etc should be on project definition (from slide 11).
Redo graph to readable fonts (slide 12).
Put table result of design on one side of three view (no for dimensions directly
on the three view).
Talk about sensitivity challenging area.
Especially thrust and cruise speed.
Maybe SFC.
Remove slide 14, put large (animated GIF?) isometric with main dimensions in
table.

Pulsejet design
More on Why pulsejet? and other pulsejet aircraft to start.
Redraw outlines on Pulsejet design types valved and valveless.
Mention papers in pulsejet conceptual section what is it, description, result.
Lots more diagrams/pictures.
Thrust Augmentation -> Papers -> Results, (water injection same) at beginning
of pulsejet section for literature review.
Engine section with feasibility study & literature review as first bit and then
design and testing second or however it flows best with the two engines.
Next week talk about engine parts (expand to 15 min 3 speakers).

Detail aircraft design
Aerofoil add pictures (CL v. Alpha).
359
Put loads for calcs.
Slide 32 put drawing and talk.
Flight control stuff electronics (stability).
Manufacturing pictures in Aircraft Detail Design & combine those two sections.
Talk about manufacturing after the detail design of each bit.
Ducted fan covered in testing. Suggested titles: Flight test 1, 2 etc.

Finances
Logos should go at end, titled acknowledgements instead of finances. Smith
Fund text.
360


Date: 25/8/08
Time: 3pm 4pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent:

Apologies:

Manufacturing

Wings testing report by Thursday.
Redesign plug for ducted fan to intake shape, Pro/E drawings of ducted fan
plug after the redesign for end profiles. Glued out of spare pieces of foam.

Ducted fan

Ducted fan order & batteries order.

Presentation

Parts sent on Wed. Parts put together and practice on Thurs.
361

Engine

Slightly larger engine to produce req. thrust.
With Richard Craig have developed OH&S procedure to resume testing.
362


Date: 1/9/08
Time: 3pm 4pm

Nick Lukacs
Ryan Anderson
Mitchell OCallaghan
Terry Walladge
Karn Schumacher

Absent: Michael Sipols

Apologies: Michael Sipols

Manufacturing
Following up fuselage.
(Gawler gliding club? If flying fan and batteries.)
Testing
SOP, Risk Assessment, Table of tests
Extra meeting
Tuesday 4-5 pm next week if possible.
o Supervisor not available on Monday.
Presentation
20-40 seconds per slide.
363
Not too formal or too informal.
Arrange separate time to time speech and practice.
In presentation graphs dont need title repeated from slide title.
3 pieces of information per slide photo, text, graph/animation etc.
Dont need Smith Fund disclaimer.
Notes
o University logo & year on first slide.
o No need for repetition of project title on first slide.
o Slide 1,3,4 add pictures, sketch or animation.
o Previous work at University of Adelaide moved to engine section and
more technical without the picture from the exhibition.
o Integrate the slides on early pulsejet aircraft, with pictures and thrust
produced.
o Slide 8, Russian aircraft put clearer picture or schematic.
o Slide 11 statistical study graphs redrawn to larger scale with design
aircraft placed on them.
o Rewrite as bench-marking in the first section general UAVs with
similar speed, UAVs with pulsejet.
o Conceptual design slide 15 changed to wing area, aspect ratio etc as
listed parameters of final concept.
o Less bright colours on engine introduction.
o Animation is being prepared for slide 19 how a pulsejet works.
o Slide 22, explanation of dual engine plan.
o Slide 35 isnt clear that its liquid.
o Consider slide 37 (fuel data) as a bar chart more comparative.
o Testing introduction of each test & test location at start.
o Video/picture at start.
o Add temperature & vibration to plane detail design.
364
Date: 15/9/08
Time: 3pm-4pm

Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher?
Absent:
Apologies:
Topics Discussed:
Manufacturing
Bulkheads to be installed, assembly, painting etc.
Servos, electronics to be connected/sorted etc.
Testing
Pulsejet test after exhibition.
Aim to assemble with pulsejet for exhibition.
Aim to assemble with fan before seminar.
Book for airfield for Monday (edit Thursday) Gawler gliding club and arrange for
pilot.
Heavy model certification can be done by pilot.
Ring supervisor to tell result 0400 669 063.
Presentation
Extra practice on timing. Rm S238a booked on 4-6pm Thursday.
Prepare for possible questions.
Dont say Thank you for listening.
First speaker to inform final speaker about how much time is short. Remove some
management/balance sheet stuff.

365
Date: 29/9/08
Location: S237A
Time: 3pm 4pm
Nick Lukacs
Ryan Anderson
Michael Sipols
Mitchell OCallaghan
Terry Walladge
Karn Schumacher
Absent:
Apologies:
Manufacturing
Horizontal tail to be fitted.
Ducted fan installation prepared, may have to be remade to fit reconfigured
ducted fan position after ducted fan pitching moment contribution calculation.
Calculations
Stabiltiy and Design Text: Calc torque/yaw (result of roll) contribution of fan,
and offset angles.
o Estimate stability derivatives
o Install fan straight for now
Find pitching moment contribution of fan
o High thrust line airplanes -> Power-up, then remove power to pitch up,
then power again to take-off.
Test
Check-list, prepare aircraft
Talk to pilot, understand dynamics, bring graphs, drop by to discuss with
supervisor on wednesday. Plans for weight distribution changes etc that can be
used to fix various handling dynamics.
Can test fan in room where co-axial tests. Check for max rpm, possibly measure
power if gather equipment, possibly measure thrust if devise suitable set-up.
C.G./fan test Thurs/Fri/Mon?.
Aircraft test Mon/Tues.
366
Engine
Larger engine from Irvine.
Prof. Kelso suggests liquid propane.
o e.g. 50 g aluminium 1 L tanks used in Olympic torch.
Finances
No expenditure after exhibition.
Report
Send detailed plan to supervisor, include all the new testing stuff in preference
to design stuff.
Decision matrices, detailed calculation in appendices.
Meeting
Send Tuesday times to arrange an extra meeting.
Exhibition
No logos in exhibition area
Reimbursement for 1 poster (overall poster for school), already talked to
Dorothy, extra posters optional.
Requirement:
o 1 computer (recommend 1 laptop with files/data)
o (Recommend video, 10-15 minutes with mix of visual and data)
o Products

Appendix M- Drawings
367
Appendix M- Drawings

UAV Plane Project

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

UAV Plane Project

Uploaded by

Copyright:

Available Formats

i

You might also like