CRANFIELD UNIVERSITY

SAMUEL JOSEPH BRADLEY

AE-16 COMPOSITE AILERON DESIGN

SCHOOL OF AEROSPACE, TRANSPORT AND MANUFACTURING

Masters of Science in Aerospace Vehicle Design

MSc(AVD)
Academic Year: 2016 - 2017

Supervisor: Dr Wenli Liu

April 2017
 CRANFIELD UNIVERSITY

SCHOOL OF AEROSPACE, TRANSPORT AND MANUFACTURING

Masters of Science in Aerospace Vehicle Design

MSc(AVD)

Academic Year 2016 - 2017

SAMUEL JOSEPH BRADLEY

AE-16 COMPOSITE AILERON DESIGN

Supervisor: Dr Wenli Liu

April 2017

This thesis is submitted in partial fulfilment of the requirements for

the degree of MSc(AVD)

© Cranfield University 2016. All rights reserved. No part of this

publication may be reproduced without the written permission of the
copyright owner.
This thesis has been assessed as of satisfactory standard for the award of a
Master of Science degree in Aerospace Vehicle Design. This thesis covers the
part of the assessment concerned with the Group Design Project. Readers must
be aware that the work contained is not necessarily 100% correct, and caution
should be exercised if the thesis or the data it contains is being used for future
work. If in doubt, please refer to the supervisor named in the thesis, or the Centre
for Aeronautics.
ABSTRACT
The Ae-16 is a semi-wide body aircraft designed as part of the Group Design
Project (GDP) for the MSc(AVD) course at Cranfield University. This thesis
documents the author’s contributions to the GDP, focussing on the preliminary
and detailed design of the composite ailerons. The Ae-16 aircraft utilises the
latest aerodynamic, systems and materials technology in order to inherit an
environmentally friendly operational design. The primary design goals were to
reduce emission and noise generation, to create a safer, healthier cabin
environment, and to focus on sustainability issues (through choice of materials
and recyclability).

The wing structure of the Ae-16 is comprised of a high aspect ratio, laminar flow
wing. Consequently, this presented a structural challenge in designing the aileron
structure, as high torsional loads needed to be reacted within a relatively small
design space compared to a conventional wing design.

From the initial conceptual design of the Ae-16, the operational envelope of the
aircraft was determined based on CS25[1] regulatory requirements. The loads on
the wing and aileron were then derived by the author to progress through to a
detailed design stage.

An iterative design process was used to optimise the structure and laminate layup
of the ailerons. The structure was analysed using conventional stress analysis
methods, finite element analysis, aeroeleastic analysis, fatigue and damage
tolerance analysis, and weight and balance analysis.

The final aileron design is of a conventional single spar, multi-rib configuration,

utilising advanced composite bonding methods and tailored strength laminates.
The final design weight of 36kg per aileron exhibits a 45% weight saving over the
equivalent metallic design.

Keywords:

High Aspect Ratio, Carbon Fibre, Natural Laminar Airflow Aerofoil, Sandwich
Core, Distributed Flange Loading Distribution

i
ACKNOWLEDGEMENTS
First of all, I would like to acknowledge my fiancée, Zaria, for her immense support
during this project. Without her love, support, and sacrifice, this project would
have been far more challenging to complete. Secondly, I would like to
acknowledge the lecturers, class mates, and in particular the fellow military
officers on course for their support, comradeship, and guidance. Finally, I would
like to thank the Royal New Zealand Air Force, whose sponsorship and support
have allowed me to complete the MSc(AVD) program at Cranfield University.

Per ardua ad astra

ii
TABLE OF CONTENTS
ABSTRACT ......................................................................................................... i
ACKNOWLEDGEMENTS.................................................................................... ii
LIST OF FIGURES ............................................................................................. vi
LIST OF TABLES ............................................................................................... x
NOMENTCLATURE .......................................................................................... xii
1 Introduction...................................................................................................... 1
1.1 Background of AE-16 design .................................................................... 1
1.1.1 AE-16 design Brief ............................................................................. 2
1.2 Component selection and design process ................................................ 2
1.2.1 Aileron Design Considerations for the Ae-16 ..................................... 3
2 Loading............................................................................................................ 5
2.1 Initial Requirements Capture .................................................................... 6
2.1.1 Flight Manoeuvre and Gust Envelope Requirements ......................... 6
2.1.2 Structural Loading from Aileron Deflection Requirements ................ 10
2.1.3 Roll Performance Requirements ...................................................... 10
2.1.4 Gust Loading/Performance Requirements ....................................... 10
2.2 Generation of the nV Diagram ................................................................ 11
2.3 Wing Loads Generated from Aileron Deflection ...................................... 13
2.3.1 Spanwise loading ............................................................................. 13
2.3.2 Chordwise loading ............................................................................ 15
2.4 Aircraft Roll Performance ........................................................................ 16
2.4.1 Roll rate performance ....................................................................... 16
2.4.2 Gust performance ............................................................................ 17
2.5 Shear Force, Bending Moment, and Torque Diagrams........................... 17
2.5.1 Shear Force Diagram ....................................................................... 18
2.5.2 Bending Moment diagram ................................................................ 19
2.5.3 Torque diagram ................................................................................ 19
2.6 Loads Acting on Aileron .......................................................................... 20
3 Initial Structural Layout .................................................................................. 22
3.1 Review of existing composite designs .................................................... 22
3.2 Design Philosophy .................................................................................. 25
3.3 Individual Component Layout ................................................................. 26
3.3.1 Aileron Spar ..................................................................................... 26
3.3.2 Aileron Skin ...................................................................................... 27
3.3.3 Aileron Ribs ...................................................................................... 28
3.3.4 Aileron Hinge Locations ................................................................... 28
3.3.5 Aileron Actuation Position ................................................................ 31
4 Material Selection .......................................................................................... 33
4.1 Composite Material Selection Methodology ............................................ 33
4.2 Sandwich Core Material Selection .......................................................... 35

iii
 4.3 Hinge Material Selection ......................................................................... 36
5 Initial Sizing ................................................................................................... 37
5.1 Laminate Strength and Stiffness for Initial Sizing .................................... 37
5.2 Initial Sizing - Aileron Skin ...................................................................... 37
5.3 Initial Sizing - Aileron Spar ...................................................................... 38
5.3.1 Spar Flanges .................................................................................... 38
5.3.2 Spar Web ......................................................................................... 39
5.4 Initial Sizing - Aileron Ribs ...................................................................... 39
5.5 Initial Sizing – Summary and Commentary ............................................. 40
6 Detailed Design ............................................................................................. 41
6.1 Detailed Design – Iteration Methodology ................................................ 41
6.2 Iteration Step 1: Laminate Layup Selection ............................................ 42
6.3 Iteration Step 2: Laminate Properties from COALA ................................ 46
6.4 Iteration Steps 3 and 4: ........................................................................... 46
6.5 Iteration Step 5: Assess Laminate Failure .............................................. 47
6.6 Iteration Step 6: Optimise Laminate Layup ............................................. 49
6.7 Spanwise Detailed Analysis - Final Iteration ........................................... 50
6.8 Detailed Design: Ribs ............................................................................. 51
6.8.1 Torsion Box Shear Flow ................................................................... 51
6.8.2 Chordwise Shear Force and Bending Moment Diagram .................. 53
6.8.3 Detailed Sizing Analysis: Ribs .......................................................... 55
6.9 Detailed Design: Hinge Analysis ............................................................. 56
7 - Aeroelastic Analysis .................................................................................... 61
8 - Fatigue and Damage Tolerance Analysis.................................................... 62
8.1 Aileron Fatigue Analysis Process ........................................................... 62
8.2 Initial AFGROW Output, Analysis, and Modifications.............................. 63
8.3 Final AFGROW Output and Analysis ...................................................... 64
8.4 Damage Tolerant Inspection Regime...................................................... 65
8.5 Secondary Task – Aircraft Fatigue Spectrum Development ................... 66
9 – Mass and Centre of Gravity Assessment ................................................... 69
10 – Finite Element Analysis ............................................................................ 71
10.1 FEA Model Generation ......................................................................... 71
10.2 Global Model FEA Results – Deflection Analysis .................................. 73
10.3 Global Model FEA Results – Strain Analysis ........................................ 75
10.4 Detailed Model FEA Results ................................................................. 77
11 – Ae-16 Composite Aileron Features .......................................................... 78
11.1 Trailing Edge Design ............................................................................. 78
11.2 Lifting and Balance Points .................................................................... 80
11.3 Electrostatic Discharge Points .............................................................. 80
11.4 Lightening Protection System (LPS) ..................................................... 81
11.5 Integrated Vehicle Health Monitoring (IVHM) ........................................ 82
12 – Manufacture and Assembly ...................................................................... 83

iv
13 – Conclusion................................................................................................ 85
14 – Recommendations ................................................................................... 86
REFERENCES ................................................................................................. 88
APPENDICES ................................................................................................. A-1
Appendix A - Compliance Matrix.................................................................. A-1
Appendix B - Loading Calculations .............................................................. B-1
Appendix C - Tabulated Loading Data and Graphs .................................... C-1
Appendix D – Initial Sizing .......................................................................... D-1
Appendix E – Laminate Selection Iteration Process .................................... E-1
Appendix F – Detailed Stressing Report ...................................................... F-1
Appendix G – Aeroelastic Analysis ............................................................. G-1
Appendix H – Fatigue and Damage Tolerance Analysis ............................. H-1
Appendix I – Finite Element Analysis ............................................................ I-1

v
LIST OF FIGURES
Figure 1 - nV Diagram for the Ae-16 (MTOM, Max Flap Altitude)..................... 12
Figure 2 - nV Diagram for the Ae-16 (OEM, Max Ceiling Altitude) ................... 13
Figure 3 - Aileron Spanwise Load Distribution due to Control Deflection ......... 14
Figure 4 - Unit Distribution of Roll Dampening Lift Force with Median Point .... 15
Figure 5 - Chordwise Pressure Distribution - Aileron Deflection ....................... 16
Figure 6 - Shear Force Diagram (MZFM, Sea Level, Roll Initiation) ................. 18
Figure 7 - Bending Moment Diagram (MZFM, Sea Level, Roll Initiation) ......... 19
Figure 8 - Torque Diagram (MTOM, Sea Level, Roll Initiation)......................... 20
Figure 9 - Aileron Spanwise Aerodynamic Load Distribution ............................ 21
Figure 10 - Aileron Spanwise Torque Distribution ............................................ 22
Figure 11 - Composite Aileron Design - Sandwich Covers [10] ........................ 23
Figure 12 - Thin laminate post-buckled skin aileron design [11] ....................... 24
Figure 13 – Co-Cured RTM Composite Aileron Design [12] ............................. 25
Figure 14 - Strand7 Stress Plot and Loading Profile ........................................ 30
Figure 15 - Spanwise Shear Force Diagram at Critical Aileron Loading Case . 31
Figure 16 - Spanwise Bending Moment Diagram at Critical Aileron Loading Case
.................................................................................................................. 31
Figure 17 - Chordwise Accumulated Torque across Aileron Span ................... 32
Figure 18 - Detailed Design Iteration Process .................................................. 42
Figure 19 - View of INBD End (Looking Down Pivot Axis) ................................ 45
Figure 20 - Combined Torsion Box Shear Flow from Chordwise Loading ........ 52
Figure 21 - Combined Torsion Box Shear Flow from Chordwise Loading ........ 52
Figure 22- Combined Applied and Reactive Chordwise Loading Systems....... 53
Figure 23 - Chordwise Shear Force Diagram - Critical Rib .............................. 54
Figure 24 - Chordwise Bending Moment Diagram - Critical Rib ....................... 54
Figure 26 - Hinge Pivot Axis Alignment ............................................................ 57
Figure 27 - Wing Side Triple Lug Hinge and Linear EHA ................................. 58
Figure 28 - Aileron Side Hinge Components .................................................... 60

vi
Figure 29 - Aileron F&DT Assessment Process ............................................... 62
Figure 30 - Crack Growth History - Aileron Lug AFGROW Output ................... 65
Figure 31 – Ae-16 Acceleration Spectrum Development Process .................... 67
Figure 32- TWIST Exceedance Plot for Ae-16 ................................................. 69
Figure 33 - Final Mass Breakdown ................................................................... 70
Figure 35 - FEA Model Loading Distribution ..................................................... 72
Figure 34 - ABAQUS Ply Stack Plot for Ae-16 Sandwich Skins ....................... 73
Figure 36 - FEA Deflection Result (x10 Scale) ................................................. 74
Figure 37 - Ply Maximum Strain Envelope (Ultimate Loading) ......................... 75
Figure 38 – Ply Maximum Strain Envelope - Ply Location ................................ 76
Figure 39 - Detailed FEA Stress Results – Hinge............................................. 78
Figure 40 - Titanium Trailing Edge Design ....................................................... 79
Figure 41 - Sandwich Panel Adhesion Surface at Trailing Edge (Upper Sandwich
Panel and Trailing Edge Hidden) ............................................................... 79
Figure 42 - Titanium Lifting Point (cut away sandwich view) ............................ 80
Figure 43 - Trailing Edge Static Wicks ............................................................. 81

Figure B-1 - Chord-wise Pressure Distribution due to Deflection of Aileron .. B-11

Figure B-2 - Combined Chord-wise Pressure Distribution due to Deflection of
Aileron and Angle of Attack .................................................................... B-13
Figure B-3 - Critical Case Aerodynamic Load Distribution (N/Aileron Span) . B-14
Figure C-1 - Wing Spanwise Shear Force Diagram – Critical Cases ............. C-3
Figure C-2 - Wing Spanwise Bending Moment Diagram - Critical Cases ....... C-4
Figure C-3 - Wing Spanwise Torque Diagram - Critical Case (MTOM,Sea Level,
Vc) ........................................................................................................... C-5
Figure C-4 - Design Speeds at Mass and Altitude Variations ......................... C-6
Figure C-5 - Gust Accelerations at Mass and Altitude Variations ................... C-6
Figure D-1 - Torsion Box Required Thickness due to Torsion Induced Stress D-4
Figure D-2 – Flat Plate Buckling Co-efficient [25] ........................................... D-6
Figure D-3 - Hinge 2 Section Cut and Load Distribution ............................... D-10

vii
Figure F-1 - Cross Section at Second Hinge Position ..................................... F-5
Figure F-2 - Torsion Box Shear Flow due to Torsion....................................... F-9
Figure F-3 - Torsion Box Integration Points used in Shear Flow Analysis ..... F-10
Figure F-4 - Open Section Shear Flow due to Accumulated Shear ............... F-11
Figure F-5 - Closed Section Shear Flow due to Accumulated Shear ............ F-12
Figure F-6 - Torsion Box Shear Flow due to Accumulated Torsion and Shear
............................................................................................................... F-14
Figure F-7 - ESDU 80023 Panel Dimensions & Load Case [29] ................... F-17
Figure F-8 - Fig. 7.2.2 from Niu [25] used to find and ............................. F-19
Figure F-9 – ESDU 80023 Panel Dimensions & Load Case [29]................... F-25
Figure F-10 - Aerodynamic Load Distribution - Hinge Two ............................ F-30
Figure F-11 - Torsion Box Integration Points used in Shear Flow Analysis ... F-31
Figure F-12 - Torsion Box Shear Flow from Chordwise Aerodynamic Loading
............................................................................................................... F-34
Figure F-13 - Loads at hinge line - Hinge Two .............................................. F-35
Figure F-14 - Torsion Box Integration Points used in Shear Flow Analysis ... F-35
Figure F-15 - Torsion Box Shear Flow from Chordwise Hinge Loading ........ F-38
Figure F-16 - Combined Torsion Box Shear Flow from Chordwise Loading . F-39
Figure F-17 - Combined Torsion Box Shear Flow from Chordwise Loading . F-40
Figure F-18 – Applied and Reactive Loading Systems Combined ................ F-41
Figure F-19 - Chordwise Shear Force Diagram - Critical Rib ........................ F-41
Figure F-20 - Chordwise Bending Moment Diagram - Critical Rib ................. F-42
Figure F-21 – ESDU 80023 Panel Dimensions & Load Case [29]................. F-44
Figure F-22 - Tearout Planes ........................................................................ F-53
Figure F-23 - Tensile Rupture Failure Mode - ESDU 91008 [51]................... F-59
Figure F-24 - Shear Tearout Failure Mode - ESDU 91008 [51] ..................... F-60
Figure F-25 - Bearing Failure Mode - ESDU 91008 [51] ............................... F-62
Figure F-26 - Lug Geometry Nomenclature from ESDU 06021 [56] .............. F-62
Figure F-27 - Lug Uniform Loading Detailing Moment Arm (b) - ESDU 91008 [51]
............................................................................................................... F-67

viii
Figure F-28 – Comparison Between Loading Distributions - ESDU 91008 [51]
............................................................................................................... F-68
Figure F-29 - Tensile Rupture Failure Mode - ESDU 91008 [51]................... F-70
Figure F-30 - Shear Tearout Failure Mode - ESDU 91008 [51] ..................... F-71
Figure F-31 - Bearing Failure Mode - ESDU 91008 [51] ............................... F-72
Figure H-1 - Aileron Fatigue Spectrum Process ............................................. H-1
Figure I-1 - FEA Boundary Conditions.............................................................. I-4

ix
LIST OF TABLES
Table 1 - Design Load Cases [2] ........................................................................ 7
Table 2 - Design Air Speed Requirements from CS25[1] ................................... 8
Table 3 - Limit Manoeuvring Factors from CS 25[1] ........................................... 9
Table 4 - Composite aileron configurations trade-off matrix [10] ...................... 23
Table 5 - Optimum Hinge Locations and Reaction Loads ................................ 30
Table 6 - Composite Material Trade Off Study [17,18] ..................................... 34
Table 7 - HRH-10-4.8-48 Aramid Fibre Honeycomb Properties ....................... 36
Table 8 - Initial Sizing - Sandwich Skin Required Thickness ............................ 38
Table 9 - Initial Sizing Results - Spar Flange Required Thickness ................... 38
Table 10 - Initial Sizing - Spar Web Required Thickness ................................. 39
Table 11 - Initial Sizing - Rib Required Thickness ............................................ 40
Table 12 - Initial Sizing Summary Thicknesses ................................................ 40
Table 13 - Iteration Extract Showing First and Final Iterations ......................... 44
Table 14 - Final Laminate Reserve Factors...................................................... 48
Table 15 - Optimised Spar and Skin Laminates Stacking Sequences.............. 50
Table 16 - Rib Laminate Stacking Sequence Iterations.................................... 55
Table 17 - Reserve Factors for Final Hinge Configuration ............................... 58
Table 18 - Hinge Components.......................................................................... 59
Table 19 - Recommended Inspection Schedules - Aileron Hinges................... 66
Table 20 – Ae-16 Mission Mix and Expected Usage ........................................ 67
Table 21 - Final TWIST Spectrum for Ae-16 .................................................... 68

Table A-1 - Compliance MOC Codes and Compliance Status Levels ............. A-1
Table A-2 - Compliance Matrix - Ae-16 Composite Aileron ............................. A-2
Table B-1 - ESDU Wings 01.01.05 Input and Output Variables ...................... B-8
Table B-2 - ESDU Controls 01.01.03 Input and Output Variables ................... B-9
Table B-3 - ESDU Controls 04.01.02 Input and Output Variables ................... B-9
Table C-1 - Applied Aileron and Roll Dampening Loads ................................ C-1

x
Table C-2 - Maximum Accelerations at Mass and Altitude Variations ............ C-7
Table H-1 - Aileron Usage Assumptions ........................................................ H-1
Table H-2 - Mission Mix Aileron Usage .......................................................... H-2
Table H-3 - Ae-16 Aileron Deflection Spectrum.............................................. H-3
Table H-4 - Single Aileron Pivot Lug Load, Geometry, and Gross Stress ...... H-4
Table H-5 - Ae-16 Aileron Pivot Lug Fatigue Spectrum.................................. H-4
Table I-1 - Mesh Convergence Check .............................................................. I-5

xi
NOMENTCLATURE
Overall zero lift angle of the wing
( ) Local lift curve slope as a function of span
( , ) Local lift curve slope as a function of span and time
( ) Lift-co-efficient slope with incidence
( ) Theoretical lift-co-efficient slope with incidence
( ) Lift-co-efficient slope with control deflection
( ) Theoretical lift-co-efficient slope with control deflection
Lift Curve Slope due to Incidence
Lift Curve Slope due to Control Surface Deflection
Angle of attack
Panel deflection coefficient
( ) Rate of change of hinge moment co-efficient with control deflection
( ) Theoretical rate of change of hinge moment co-efficient with
control deflection
Wing Span
Panel bending stress coefficient
̅ Mean Aerodynamic Chord
Control chord, aft of hinge line
Coefficient of lift
( ) Co-efficient of lift at span position
Aircraft pitching moment coefficient at zero lift
( ) Local chord at span ( )
Diameter
Edge margin
Material stiffness
Product of inertia (x,z axis) due to asymmetry and inclination of body axis in
flight direction
Gust alleviating factor or total force acting on aileron
Allowable Stress
Gravitational acceleration constant
Shear stiffness
Load peaking effect factor
wing aerodynamic centre
Moment of inertia about the rolling axis of the aircraft
Coefficient value
Rolling moment due to rate of roll
Rolling moment due to the rate of yaw
Rolling moment due to side velocity
Wing lift due to incidence

xii
Rolling moment due to lateral motivator deflection
Rolling moment due to roll motivator deflection
mass
Mach speed
Mass Parameter
Normal acceleration factor
Allowable load
Pressure on control surface due to angle of attack
Angular roll rate
Angular acceleration in roll
Roll angle of the aircraft
Angular acceleration in yaw
Density of Air at Local Altitude
Density of Air at Sea Level
Shear Flow
Statical moment of area
Wing planform area
Applied or accumulated shear force
Standard Mean Chord (geometric)
Applied or accumulated torque
Shear stress
Reference gust speed
Applied shear force
Design manoeuvring speed
Design speed for maximum gust intensity
Design cruising speed
Design dive speed
Equivalent Air Speed
Design wing-flap speeds
Stalling speed with wing-flaps retracted
Stalling speed at current condition
Width
Angle of roll motivator deflection
Spanwise median point of discreet roll dampening load
Spanwise median point of discreet wing lift load
Spanwise median point of discreet aileron loading

xiii
 Section modulus
Roll dampening load on wing
Aerodynamic load on wing due to aileron deflection

_ Total aerodynamic load due to angle of attack and aileron deflection acting
on the wing across the span width of the aileron
Total aerodynamic load due to angle of attack and aileron deflection acting
on aileron

AVD Aerospace Vehicle Design

AMC Acceptable Means of Compliance
BCAR British Civil Airworthiness Requirements
BL Butt Line
BM Bending Moment
CATIA Computer Aided Three-Dimensional Interactive Application v5r20
CFRP Carbon Fibre Reinforced Polymer
CoG Centre of Gravity
CNC Computer Numerically Controlled
CoP Centre of Pressure
CS Certification Standard
EAS Equivalent Air Speed
EASA European Aviation Safety Agency
ESDU Engineering Sciences Data Unit
FEA Finite Element Analysis
GDP Group Design Project
GFRP Glass Fibre Reinforced Polymer
HLD High Lift Device
ILSS Inter Laminar Shear Strength
LPOP Laser Ply Outline Projection
MTOM Max Take-off Mass
MZFM Max Zero Fuel Mass
OEM Operating Empty Mass
RAL Rough Air Line
RF Reserve Factor
UD Uni Directional Fibre
UDL Uniform Distributed Load

xiv
1 Introduction
This thesis is a summary of the work completed by the author as part of the
MSc(AVD) Group Design Project (GDP) at Cranfield University. The GDP
requires a group of MSc(AVD) students to progress a conceptual aircraft design
through to the detailed design level. The numerous structural, avionic, and
support components required to form an entire aircraft design were distributed to
individual GDP members, with the author designated the design of the composite
ailerons. As a secondary task, the author also undertook a fatigue spectrum
analysis of the aircraft to assess the fatigue resistance, inspection regimes, and
Life of Type (LOT) for the Ae-16.

1.1 Background of AE-16 design

Boeing’s B737 and Airbus’ A-320 families of aircraft are undoubtedly the most
commercially successful aircraft in the aviation industry. The continued demand
for aircraft in this sector of the market creates opportunities for competing aircraft
to attain high commercial success. Ebraer’s E-190/195 and C-Series aircraft, and
COMAC’s C-919 aircraft are examples where competing aircraft manufacturers
are starting to realise the gains to be found in the narrow body, medium range
market[2].

The air transport density has increased significantly over the last few decades,
and is expected to rise up to 3% per year until 2050 [3]. A major concern with
these statistics, is the increasing levels of environmental damage that is occurring
due to the greenhouse gases emitted from aviation. Further, aspects such as
noise pollution, contrail generation, recyclability, and sustainability continue to
present challenges to today’s aircraft manufacturers.

The project specification of the Ae-16 aims to explore possible solutions to these
issues, while at the same time presenting an economically viable aircraft within
the narrow body market.

1
1.1.1 AE-16 design Brief
The Ae-16 is presented as an advanced semi wide body aircraft, with the capacity
to carry 200 passengers within the short to medium haul range. The aircraft will
be an eco-friendly design that aims to reduce emissions and noise pollution, while
enhancing the levels of comfort and safety for the passengers. Further, the Ae-
16 will incorporate the latest aerodynamic, systems and materials technologies
for optimisation of the reliability, maintainability, service life, and, where
compatible, low cost[2].

The Ae-16 design achieves these design goals through the incorporation of
advanced design concepts such as a semi-elliptical fuselage, battery powered
turbofan engines, natural laminar flow wings with morphing leading edge
technology, and advanced avionics concepts such as single pilot operation
capability.

The conceptual design of the Ae-16 was presented to the GDP team by Professor
Smith within the aircraft’s project specification booklet [2].

1.2 Component selection and design process

This thesis focusses of the design and development of the composite ailerons of
the Ae-16. The author chose this component due to the industry movement
toward composite based structures, and the motivation of the author to enhance
his skills in composite structural analysis. Further, as the aileron hinge structure
is a component highly susceptible to high cycle fatigue damage, the investigation
into mass optimisation while maintaining sufficient fatigue resistance presented
itself as an interesting challenge.

The design process presented in this thesis is as follows:

-Analysis of conceptual aileron design variations to provide the most viable option

-Requirements capture: identifying the regulatory requirements for aileron design

2
-Generation of the design loading cases and the Ae-16’s acceleration velocity
(NV) diagram (the design operational envelope)

-Derivation of the loads imparted on the wing by asymmetric rolling manoeuvres

at each loading case

-Derivation of the rolling performance figures at each loading case.

-Derivation of the maximum loading on the aileron and hinges

-Determination of the initial structural layout and incorporated technologies

-Initial sizing of structural components and material selection

-Optimisation and qualification of the structure through detail stressing and FEA

-Fatigue and damage tolerance assessment of fatigue critical structure

-Assessment of manufacturability, maintainability, and compliance of the design

1.2.1 Aileron Design Considerations for the Ae-16

An aircraft’s ailerons primarily provide roll authority; however, they can also be
utilised for secondary purposes that may be utilised in the Ae-16’s design.

Firstly, a number of Airbus aircraft incorporate a function known as Aileron Droop

Function (ADF) or more commonly known in the industry as ‘Flaperons’.
Essentially the ailerons on each side of the aircraft are deflected in unison
downward in order to increase lift at lower aircraft velocities. This is used to
augment the wing’s lift when the flaps are extended, while also smoothing the
spanwise lift distribution across the wing. Another secondary use of the ailerons
is to provide what’s known as a Load Alleviation Factor (LAF). The LAF system
uses an accelerometer to detect uncommanded accelerations and automatically
deflects the ailerons to oppose them. Generally used to oppose gusts (but can
also be used for dynamic manoeuvres) the LAF system results in an alleviated
load on the wing structure, a smoother ride for passengers, and reduced fatigue
effects on the wing structure.

3
The major difference between the wing structure of conventional narrow body
aircraft and the Ae-16 is the high aspect ratio geometry of the wing. The high
aspect ratio results in the aileron load acting at a greater distance from the roll
axis of the aircraft, leading to an inherently higher coefficient of rolling moment.
Therefore, aileron deflections of the Ae-16 result in greater rolling moments, and
greater loads applied to the wing compared to conventional wing designs.

The high aspect ratio of the Ae-16 wing also means that the height of the wing
box section at the aileron attachment point is relatively small. It is highly likely that
due to restricted space, the aileron actuators will need to be placed eccentrically
toward the inboard end of the aileron. This fact, combined with the required large
span of the ailerons, means torsional loading will be a structural challenge.

A second design consideration which must be taken into account due to the Ae-
16’s high aspect ratio wing is the potential for aeroelastic twisting of the wing
under aileron deflection force. As the wing box at the spanwise location of the
ailerons is relatively small, the torsional forces on the wing may cause significant
twisting deflection due to reduced torsional stiffness. Any twisting during aileron
deflection will significantly reduce the aileron’s effectiveness, as the aerodynamic
effect of the twisting opposes the effect of the aileron. Further, the aerodynamic
effect of wing twisting increases exponentially with speed, and at the extreme
cases can cause a condition known as ‘control reversal’.

In order to overcome the Ae-16’s susceptibility to aeroelastic performance losses,

one solution is to split the conventional single aileron design into separate inboard
and outboard ailerons. The benefit of multiple ailerons per wing is the capability
to ‘lock out’ the outboard ailerons at high speed and rely on the inboard ailerons
only (where the torsional stiffness is higher). For large wide-body aircraft, such
as the A380, a multiple aileron design is essential in order to have a smooth
response curve of input/output for such large control surfaces. The split aileron
design also allows for greater accuracy of attaining roll trim of the aircraft as
compared to a single large aileron. The downsides to employing a multiple aileron
design is the increased complexity of the system, increased weight, higher

4
maintenance burden, and reduced reliability due to increased number of
components.

For the Ae-16, a multiple aileron design would present some additional
challenges in that the actuator(s) for the outboard aileron surface would need to
be placed at a spanwise location where the wingbox height is very small. It is
likely that external blisters on the lower surface of the wing would need to be
designed to house such an actuator system (greatly affecting aerodynamic
smoothness).

With a single aileron design, it is still prudent to reduce the load imparted on the
wing from aileron deflection in order to reduce aero elasticity. This can be
accomplished by limiting the control authority for aileron roll at higher speeds,
and through the use of spoiler roll augmentation. As multiple spoilers are located
towards the inboard span of each wing (at locations of higher torsional stiffness),
their use in rolling manoeuvres will reduce the required deflection, and hence
load, imparted by the ailerons at the outboard spans of the wing.

Taking the above considerations into account, the loading calculations will be
completed assuming a single aileron design and complete roll authority from the
ailerons as an initial design case. At a later design stage, these considerations
will be investigated in more detail.

2 Loading
Prior to conducting structural layout assessment and sizing analyses, the loads
on each aircraft component needed to be ascertained. A significant proportion of
the aircraft’s structure experiences loading generated from asymmetric rolling
manoeuvres. The loads applied to a deflected aileron are transmitted directly
through the actuators and hinges to the outer wing structure. This in turn
generates shear, bending, and torque forces across the wing, to the wing root,
and into the fuselage. The angular acceleration from an asymmetric rolling
manoeuvre also generates inertial forces on all structure located at a distance
from the rolling axis (significant for the engine pylon structure). Finally, the roll

5
acceleration and roll rate would generate gyroscopic induced forces on the
engine pylon structure due to rotation of the engine rotation plane (the engine
rotation axis is not aligned with the rolling axis of the aircraft). This section details
how the asymmetrical loads were determined, from initial requirements capture,
to final loading results.

2.1 Initial Requirements Capture

The European Aviation Safety Agency (EASA) was chosen as the regulatory
body for which the AE-16 would be certified. As the AE-16 is classified as a large,
turbine powered aircraft, the design falls within the Certification Specifications
(CS) and Acceptable Means of Compliance (AMC) Part 25[1]. The standards
required by CS25, pertinent to the asymmetric manoeuvres, are detailed below.

2.1.1 Flight Manoeuvre and Gust Envelope Requirements

In order to generate the acceleration/velocity design envelope (known as the nV
diagram) for the Ae-16, the author referred to a number of EASA certification
standards.

Firstly, the GPD project team required a set of load cases for which they would
analyse the loads and confirm compliance with CS25[1]. It was important to
ensure the worst case loading cases were accounted for, whilst minimising the
number of cases to be analysed.

CS25.321 [1] requires that load cases are selected:

- At each critical altitude within the range of altitudes selected
- At each weight from the design minimum weight to the design maximum
weight, appropriate to each particular flight condition.

Further, for each required altitude and weight, the load cases must be selected
for any practicable distribution of disposable load CS25.321(b),3 [1]

The author, in consultation with the wing team, selected the following critical load
cases for which the GDP project team would conduct loading analyses:

6
 Table 1 - Design Load Cases [2]

Max Take-off Mass (MTOM), 87,663kg

Aircraft Mass Cases: Max Zero Fuel Mass (MZFM), 87,087 kg
Operating Empty Mass (OEM), 60,133 kg
Sea Level (0 ft)
Critical Operating Max Flap (20,000 ft)
Altitudes: Cruise (35,000 ft)
Max Ceiling (41,000 ft)

The design airspeeds to be used to construct the nV Diagram are described as

follows:

Design airspeeds:
-  Design manoeuvring speed
-  Design speed for maximum gust intensity
-  Design cruising speed
-  Design dive speed
-  Design wing-flap speeds
-  Stalling speed with clean wing configuration
-  Stalling speed at current condition

The standards in determining the design airspeeds were extracted from multiple
certification specification sources detailed in Table 2 below. It must be noted that
all standards relating to gust accelerations, including derivation of the rough
airspeed velocity ( ), have been drawn from EASA CS23 standards (Normal,
Utility, Aerobatic and Commuter Aeroplanes) [4]. The reason behind this is for
simplification purposes; the CS23 requires gust accelerations to be derived using
a sharp edge alleviated gust assessment. Comparatively, CS25 requires a
continuous turbulence assessment, which was agreed to be too complex to
assess within the GDP project timelines1.

The regulatory requirements used to define the design airspeeds are detailed in
Table 2 below. All design airspeeds are stated in Equivalent Air Speeds (EAS):

1 Ae-16 Decision Log – 10 NOV 16.

7
 Table 2 - Design Air Speed Requirements from CS25[1]

VC – Design cruising speed

CS 25.335 (a)(1) The minimum value of must be sufficiently greater than V
CS 25.335 (a)(2) may not be less than + 1.32* , but need not exceed the
maximum speed in level flight at maximum continuous power for the
corresponding altitude.
CS 25.335 (a)(3) At altitudes where is limited by a Mach, then may be limited to a
Mach number too.
VD – Design dive speed
CS 25.335 (b)(2) In any case, the margin [between and ] may not be reduced to
less than 0.05M.
VA – Design manoeuvring speed
CS 25.335 (c)(1) may not be less than VS1·n0.5.
CS 25.335 (c)(3) need not be more than or the speed at which the positive CNmax
curve intersects the positive manoeuvre load factor line, whichever is
less.
VB – Design speed for maximum gust intensity
CS 23.333 (c)(1) The aeroplane is assumed to be subjected to symmetrical vertical
gusts in level flight. The resulting limit load factors must correspond to
the conditions determined as follows:
(i) Positive (up) and negative (down) gusts of 50 fps at VC
must be considered at altitudes between sea level and
6096 m (20 000 ft). The gust velocity may be reduced
linearly from 50 fps at 6096 m (20 000 ft) to 25 fps at 15240
m (50 000 ft); and
(ii) (ii) Positive and negative gusts of 25 fps at VD must be
considered at altitudes between sea level and 6096 m (20
000 ft). The gust velocity may be reduced linearly from 25
fps at 6096 m (20 000 ft) to 12·5 fps at 15240 m (50 000
ft).
(iii) (iii) In addition, for commuter category aeroplanes, positive
(up) and negative (down) rough air gusts of 66 fps at VB
must be considered at altitudes between sea level and
6096 m (20 000 ft). The gust velocity may be reduced
linearly from 66 fps at 6096 m (20 000 ft) to 38 fps at 15240
m (50 000 ft).
CS 23.335 (d)(1) may not be less than the speed determined by the intersection of
the line representing the maximum positive lift CN MAX and the line
representing the rough air gust velocity on the gust V-n diagram, or
√ , whichever is less, where –

8
 (i) ng the positive aeroplane gust load factor due to gust, at
speed (in accordance with CS 23.341), and at the
particular weight under consideration; and
(ii) (ii) VS1 is the stalling speed with the flaps retracted at the
particular weight under consideration.
CS 23.335 (d)(2) need not be greater than .
VF
CS 25.335 (e)(3) may not be less than:
(i) 1.6*VS1 with flaps in take-off position at MTOW
(ii) 1.8*VS1 with flaps in approach position at MLW
(iii) 1.8*VS0 with flaps in landing position at MLW

CS25 standards were also used to determine the limit manoeuvring factors,
detailed in Table 3 below. These factors represent the maximum acceleration that
the aircraft structure must be designed to withstand.

Table 3 - Limit Manoeuvring Factors from CS 25[1]

POSITIVE LIMIT
CS 25.337 (b) The positive limit manoeuvring load factor ‘n’ for any speed up to
may not be less than:
24000
2.1 +
+ 10000
except that ‘n’ may not be less than 2·5 and need not be greater than
3·8 – where ‘W’ is the design maximum take-off weight (lb).
NEGATIVE LIMIT
CS 2.5337 (c) The negative limit manoeuvring load factor:
- May not be less than -1.0 at speeds up to .
- Must vary linearly with speed from the value at to zero at
.
HIGH LIFT DEVICES
CS 25.345 (a)(1) Manoeuvring to a positive limit load factor of 2.0.

These regulatory certification requirements were used to generate the flight

manoeuvre envelope (the nV Diagram) as detailed in Section 2.2.

9
2.1.2 Structural Loading from Aileron Deflection Requirements
CS25.349(a)(1) [1] requires that loading is assessed at the maximum angular
acceleration and steady rolling velocity cases. The standard requires the
following specified conditions are assessed:

-At , a sudden deflection of the aileron is assumed.

-At , the aileron deflection must be that required to produce the roll rate as that
obtained for the first case ( , max aileron deflection).

-At , the aileron deflection must be that required to produce the roll rate of one-
third that of the first case.

As per CS25.349(a)[1], these asymmetric loads must be considered in

combination with an aeroplane load factor of two-thirds of positive manoeuvre
factor used in design (used later to derive shear, bending moment, and torque
diagrams).

2.1.3 Roll Performance Requirements

The lateral control requirements prescribed in CS25.147(d) state that with the
critical engine inoperative, the roll response must be sufficient for normal
manoeuvres, and the roll rate performance prescribed in AMC25.147(d) must be
met (a 60 bank angle achieved in 11s). A second case prescribed in CS25.147(f)
and AMC25.147(f) requires higher rolling performance, yet with all engines
operating normally (a 60 bank angle achieved in 7s) [1].

2.1.4 Gust Loading/Performance Requirements

CS25.349(b) requires that the design gust accelerations are assumed to act at
an 80/100 imbalance between the wings, hence resulting in asymmetric rolling
loads acting on the wing. The requirements of CS25.147(f) also state the roll
response of the aircraft must be able to recover from these asymmetrical gusts
(essentially the aileron control authority must be sufficient to maintain steady level
flight) [1].

10
2.2 Generation of the nV Diagram
In order to generate the nV diagrams for all loading cases, the author constructed
an algorithm in the Matlab programming language. The benefits of using Matlab,
was the flexibility to run loops of code while changing the various input variables.
This allowed the author’s algorithm to cycle through all loading case variables
and to generate the nV diagram for each case.

It must be noted that the generation of the nV diagram was not within the authors
original work scope, however due to a willingness to learn the Matlab
programming language, the author generated an independent algorithm to
compare the results with team dedicated to forming the nV diagrams. This proved
invaluable to the team, as the authors independent checks found several errors
and incorrect assumptions in the main nV team’s calculations.

In generation of the nV diagram, several assumptions were used as per below:

-The gust requirements were sourced from CS23 [4] as discussed in section 2.1.4

-The requirement set by CS 25.335 (a)(2), that may not be less than +
1.32( ), is not complied with, as the use of CS23 for gust cases makes
compliance impossible for certain loading cases.

-The dynamic coefficient of lift under gust loading is: = 1.25 ×

-The High lift Device (HLD) setting is constant for both approach and landing

-The maximum negative and positive coefficients of lift are equivalent

It must be noted that the gust envelope for the loading case of an Operating
Empty Mass (OEM) aircraft exceeded the manoeuvre envelope at certain
altitudes. The author’s algorithm determined that the maximum gust envelope for
the OEM case occurs at 20,100ft. As several teams had already performed
calculations at max flap altitude (20,000ft), the maximum gust accelerations
found by the author at 20,100ft were published as occurring at 20,000ft. This
avoided requiring the GDP team to undertake calculations at another altitude,
whilst also maintaining conservatism.

11
Finally, the team dedicated to forming the nV diagrams had used an assumption
that the Rough Air Line (RAL) is a straight line within the nV Diagram. The RAL
is used to determine the rough air speed ( ), by finding the intersection of the
RAL and the stall line. In reality, the RAL is a curve as it varies not only with
speed, but with the lift curve slope ( ). The authors Matlab algorithm forms the
curve shape and finds the true value of . The results presented in Appendix
C.6 use the simplified, straight line (RAL) approximation method for consistency.

Two examples of the nV diagram generated by the author’s algorithm can be

seen in Figure 1 and Figure 2 below.

Figure 1 - nV Diagram for the Ae-16 (MTOM, Max Flap Altitude)

Figure 2 below shows the versatility of the authors algorithm at a different loading
case, where at higher altitudes the positive manoeuvre limit is reduced below
2.5g to ensure is not greater than (required by CS 25.335(c)(1)). Further,
the rough air speed is limited by √ rather than the intersection of the RAL
with the stall curve (required by CS 23.335 (d)(1)). Finally, the effect of dynamic
stall limits can be seen with the gust acceleration at existing above the stall
curve. Note the high lift device curves do not apply to higher altitudes:

12
 Figure 2 - nV Diagram for the Ae-16 (OEM, Max Ceiling Altitude)

2.3 Wing Loads Generated from Aileron Deflection

2.3.1 Spanwise loading

Spanwise loading on the aileron and wing are derived from the unitised spanwise
loading distribution provided in the project specification [2]. Integration of the
loading distribution allows a spanwise median point of loading to be determined.
Details of how this integration is performed is detailed in Appendix B.1.2. The
median point of aileron lift load for the Ae-16 was found to be at 15.533m from
the aircraft centreline (B.L. 0) as shown in Figure 3 below.

The derivation of the spanwise median point for aileron lift force allows the
determination of aileron load from generalised rolling equations as detailed in
Appendix B.1.1. All loading cases were analysed to find that the maximum aileron
lift load for the Ae-16 was 178.85kN occurring at the load case of MTOM, Sea
Level, and speed (tabulated load results can be seen in Appendix 14B.1.9

13
 Figure 3 - Aileron Spanwise Load Distribution due to Control Deflection

The roll dampening aerodynamic load was derived from the use of the ‘Schrenk
Method’ to find an approximation of the spanwise lift distribution. The Schrenk
method uses the assumption that the lift distribution per unit span length is the
mean value of the actual wing chord distribution and an elliptical wing chord
distribution that has the same area and span [5]. The derivation of the distributed
roll dampening load using this method is detailed in appendix B.1.4. Again,
integration of the roll dampening distribution can yield a median acting point for
roll dampening lift, and, when used with the generalised rolling equation, allows
determination of the roll dampening load. The Schrenk distribution and median
point of roll dampening lift load for the Ae-16 is shown in Figure 4 below.

All loading cases were analysed to find that the maximum roll dampening lift load
for the AE-16 was 246.74kN occurring at the load case of MTOM, Sea Level, and
speed (tabulated load results can be seen in Appendix C.1)

14
 Figure 4 - Unit Distribution of Roll Dampening Lift Force with Median Point

2.3.2 Chordwise loading

The chordwise pressure distribution acting on the wing and aileron during aileron
deflection was determined through the use of several ESDU sheets [6–8] and
equations sourced from Howe [9]. The overall chordwise pressure distribution
was found by the summation of the wing incidence pressure distribution and the
aileron deflection pressure distribution as shown in Figure 5 below:

The aileron deflection pressure distribution was integrated in isolation to find the
median acting point of the deflection lift. The chordwise distance between this
median point and the shear centre of the wing box was used to determine the
applied torque on the wing produced from aileron deflection (Derivation can be
viewed in Appendix 14B.1.7) (see Section 2.5.3 for Wing Torque Diagram).

Integration of the combined chordwise pressure distribution allows the calculation

of the percentage of load that is reacted forward and aft of the aileron hinge line.
For the Ae-16 chordwise profile, the aileron reacts 21.6% of the total chordwise
load (derivation can be viewed in Appendix 14B.1.8)

15
 Unit Chord-wise Pressure Distribution Due to Aileron
Deflection
3
2.5
2
1.5
1
0.5
0
-0.5 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

-1 Chordwise Position, LE to TE (%)

Aileron Deflection Distribution Angle of Attack Distribution

Combined Distribution Hinge Line

Figure 5 - Chordwise Pressure Distribution - Aileron Deflection

2.4 Aircraft Roll Performance

2.4.1 Roll rate performance

The roll performance requirements described in chapter 2.1.3 were met at all
loading cases, without the necessity for spoiler roll augmentation. The critical
engine failure condition of CS25.147(d) however, has not been quantitatively
proven. This is acceptable for preliminary performance calculation, as the
engines of the Ae-16 are mounted to the fuselage. This entails that they share a
close proximity to the aircraft yaw axis and thus, would induce a relatively small
moment under asymmetrical thrust.

Conservative calculations were used by assuming that the approach speed was
equivalent to the stall speed of the aircraft in landing configuration. In reality, the
approach speed will be greater which would result in improved roll performance.

Further conservatism was used by assuming that the performance of the ailerons
was only two-thirds effective due to aero elastic twisting of the wing under load.
This is especially pertinent for the high aspect ratio wing of the Ae-16, which
incorporates relatively small wing box sections to react the aileron loads.

16
2.4.2 Gust performance
The gust correction performance requirements (see Section 2.1.4) for the Ae-16
were met at all loading cases. The Ae-16 therefore has the control authority to
maintain level flight under worst case asymmetrical gust loading. Again,
conservatism was used by assuming that the performance of the ailerons was
only two-thirds effective due to aero elastic twisting of the wing under load. Details
of the calculations in proving compliance can be seen in Appendix B.1.6.

2.5 Shear Force, Bending Moment, and Torque Diagrams

The wing spanwise shear force, bending moment, and torque diagrams for
asymmetrical loading were produced by the author. As per CS25.349(a) [1], the
asymmetric loads were considered in combination with an aeroplane load factor
of two-thirds of positive manoeuvre factor used in design (representing a rolling-
pull up manoeuvre). As the positive manoeuver factor reduces at the high altitude
load cases, the author’s Matlab algorithm was used to find the actual positive
manoeuver factor to be used. Details of the derivation of spanwise envelopes can
be seen in Appendix B.1.7.

The envelope of the worst case spanwise loads for each loading type were
passed to the wing team to compare against symmetrical loads at the positive
manoeuvre factor. By superimposing the author’s envelope (which already
included factored symmetrical loading), the overall maximum envelopes could be
produced. The envelopes of asymmetrical loading produced by the author can be
viewed in Appendix C.

Note that the sign convention is as per industry standard, with positive shear
tending to rotate an element in the clockwise direction, the positive bending
moment producing a ‘u’ shape, and the right hand rule applying for torque.

17
2.5.1 Shear Force Diagram
All load cases were assessed to generate the maximum shear load envelope for
the wing. The maximum envelope was a combination of cases due to variations
such as fuel weight, which affects the inertial dampening, rolling inertia and the
roll acceleration.

Below in figure Figure 6 is an example of the shear force diagram created by the
author for the loading case of MTOM, Sea Level, . As can be seen, the author
has assumed all shear loads are reacted at wing root, where the inner wing would
connect to the centre wing (B.L 10% of wing span).

Figure 6 - Shear Force Diagram (MZFM, Sea Level, Roll Initiation)

18
2.5.2 Bending Moment diagram
As with the shear force diagram, all load cases were assessed to form a
maximum bending moment envelope (as can be viewed in Appendix C.3). An
example of a bending moment diagram can be seen below (MTOM, Sea Level,
). As the shear force is reacted at the wing root, the bending moment in the
centre wing box is constant.

Figure 7 - Bending Moment Diagram (MZFM, Sea Level, Roll Initiation)

2.5.3 Torque diagram

All load cases were assessed to determine the maximum spanwise torque
loading on the wing. The spanwise torque was discovered to be dominated by
mass/fuel inertial loads, rather than the applied torque from the ailerons. As such,
the critical torque case was discovered to be the fully fuelled, MTOM case, even
though it has lower roll acceleration as compared to an equivalent MZFM case:

19
 Figure 8 - Torque Diagram (MTOM, Sea Level, Roll Initiation)

2.6 Loads Acting on Aileron

The loads acting on the aileron are ascertained by isolating spanwise distributions
for the critical case between the span limits of the aileron (between WS13.41m
and WS19.05m). To find the actual aerodynamic load reacted by the aileron, the
wing load distributions are multiplied by the percentage reacted by the aileron
(21.6% as detailed in Section 2.3.2). The resulting spanwise load distribution
acting on the aileron is shown in Figure 9 below.

Integration of this load distribution yields the total aerodynamic load acting on the
aileron of 65,094N at the critical case (derivation of the distribution and load can
be seen in Appendix 14B.1.9).

20
 Spanwise Aerodynamic Load Distribution
(N/Aileron Span)
100,000
90,000
80,000
Load Distribution N/m

70,000
60,000
50,000
40,000
30,000
20,000
10,000
0
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Aileron Span %

Figure 9 - Aileron Spanwise Aerodynamic Load Distribution

The spanwise torque distribution for the aileron can be derived by multiplying the
local spanwise load distribution by the local chordwise centre of pressure aft of
the hinge line. As the chordwise pressure distribution aft of the hinge line is
triangular in shape (see Figure 5) the centre of pressure is 1/3rd of the aileron
chord at each spanwise location. The resulting torque distribution is detailed in
Figure 10 below.

Integration of this torque distribution yields the total hinge moment of 11,257Nm
at the critical case (derivation of the distribution and load can be seen in Appendix
14B.1.9). This total limit torque value is to be distributed and reacted by the
aileron actuators.

21
 Torque Distribution (Nm/Aileron Span)
20000

18000
Torque Distribution (Nm/AIleron Span)

16000

14000

12000

10000

8000

6000

4000

2000

0
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Aileron Span %

Figure 10 - Aileron Spanwise Torque Distribution

3 Initial Structural Layout

3.1 Review of existing composite designs
A literature review provided great insight into current and emerging technology
used in the design of composite aileron structure.

James & Vaughn [10] conducted a trade-off comparison between common

composite aileron configurations. Comparisons across a variety of attributes were
made between sandwich composite panel, stiffened laminate, and honeycomb
core designs, as can be seen in Table 4 below.

As can be seen, the composite sandwich panel design has numerous advantages
over the other design alternatives, particularly in weight savings.

22
 Table 4 - Composite aileron configurations trade-off matrix [10]

James & Vaughn [10] also designed and tested an optimal composite sandwich
panel aileron design as can be seen in Figure 11 below:

Figure 11 - Composite Aileron Design - Sandwich Covers [10]

This design inherits a mass-boom configuration, with a composite front spar,

sandwich panel covers, composite ribs reacting hinge/actuator loads, and a rear
aluminium spar separating the main torque box from a honeycomb trailing edge.
Mass-boom loading distributions are advantageous in lightly loaded structures as
they are simple to analyse and allow large cutouts to exist in the skins [9].
However, when the load intensity is moderate to high it becomes practical to
design the skins to aid in reacting the sectional bending loads (a distributed flange

23
loading distribution). Whilst a distributed flange design is more complex to
analyse, it’s high structural efficiencies and damage tolerance characteristics are
a significant advantage [9].

Scott et.al.[11] researched a composite aileron design that avoids the use of
sandwich panels due to their disadvantages (poor impact resistance, water
ingression, and skin delamination). Their panel-skin design requires an increased
number of ribs or stiffeners in order to prevent unwanted deformation or buckling
of the skins (and therefore a trade off in weight occurs). Scott et.al.[11] used
stiffened ‘blades’ and designed the skin to inherit a post buckled load capacity in
order to reduce this weight penalty as can be seen in Figure 12 below:

Figure 12 - Thin laminate post-buckled skin aileron design [11]

Romano et.al.[12] researched a design that significantly reduces the number of

individual parts by using co-curing and Resin Transfer Moulding (RTM)
processes. The skins and ribs are co-cured and then bonded to the main spar
assembly. Figure 13 below details how the plies are split at the trailing edge to
form a quasi-rear spar. The tailorability of composites is also demonstrated with
the use of symmetrical, but unbalanced layups used to maximise stiffness
efficiency.

24
 Figure 13 – Co-Cured RTM Composite Aileron Design [12]

The downside to the use of an RTM process is the high tooling cost and high
energy usage inherent with autoclave curing. As the primary design objective of
the Ae-16 was to reduce environmental impact, the high energy demands of
implementing an RTM manufacturing process were not ideal.

3.2 Design Philosophy

The design philosophy taken by the author is to prioritise eco-friendly and
sustainable design considerations, thereby aligning with the primary design
objectives of the Ae-16 project. After a review of prior designs and assessing their
qualities with regard to the design objectives, a sandwich panel skin design,
utilising a distributed flange design philosophy, with a single front spar and
minimal ribs has been selected with the following benefits:

-A distributed flange design will utilise the skin to react spanwise and chordwise
shear and bending loads. Regardless of the design philosophy, the skin will need
to be relatively thick to resist the high torsional forces of the Ae-16 aileron. This
means that efficiencies will be observed if the spar and ribs are sized based on a
sharing of bending load with the stiff skins (thereby reducing the required spar
and rib thicknesses). A distributed flange design also has the advantages of
attaining a higher level of damage tolerance (not reliant on critical booms) [9]

-The stiff sandwich panels will reduce the required amount of ribs (thereby
reducing weight and therefore fuel/emission penalties).

-The reduced amount of ribs also reduces the amount of carbon fibre material
used (which reduces manufacturing energy cost and wastage)

25
-Hand layup, out-of-autoclave manufacturing will be possible with a sandwich
panel design, which will reduce manufacturing energy costs. Secondly, this
method will avoid tooling energy costs and wastage currently present with RTM.

-Bonding will be used where possible to reduce the amount of fasteners used to
reduce weight/fuel/emission penalties

-Material selection will be focussed on sustainability and recyclability.

3.3 Individual Component Layout

The following sections describe the methodology used to define the layout and
placement of the primary aileron structural components. The design philosophy
of attaining a minimal weight design is adhered to for all component layout
decisions.

3.3.1 Aileron Spar

The aileron spar’s primary structural purpose is to react spanwise bending and
shear loads, primarily through the spar flanges and web respectively. The spar
also serves as an attachment point for the aileron hinges, and as such, must
transmit concentrated hinge and actuation forces through to the surrounding
aileron skin and rib structure.

The orientation of the spar can be chosen to have the flanges facing aft or
forward. The benefits of an aft facing spar is the increase in size of the enclosed
torsion box area, which has benefits in reducing the required skin and spar
thicknesses. A forward facing spar has the benefit of greater surface area with
which to attach the hinges (as the hinges can be connected directly to the spar
flanges). As both the shear and torque forces acting on the Ae-16’s aileron are
relatively large (due to the relatively long length and high aspect ratio) the author
deemed it critical that the hinges have ample room for attachment, and in the
ideal orientations. Therefore, a forward facing spar has been selected.

26
There are several factors involved in the chordwise placement of the front spar.
A spar placed as far forward as possible has the advantages of minimising the
moment arm between the outer wing spar and the aileron spar, thereby reducing
the bending forces within the hinges. Further, as the spar is moved forward, the
enclosed area of the aileron torsion box is increased (thereby increasing torsional
stiffness). Conversely, a spar that is placed further aft will reduce the chordwise
bending loads reacted by the aileron ribs, and also allow greater room forward of
the spar web with which to attach the hinges with mechanical fasteners.

As it is likely there will be high accumulated torsional loads on the Ae-16 aileron
(due to the high aspect ratio and limitations on actuator placement) the torsional
criteria for spar placement is considered critical. As such, a minimal distance of
50mm between the hinge line and spar web has been chosen as an optimal
balance of the above requirements. The spar will be aligned with the hinge line
(which is at 70% wing chord) in order to have consistent spar flange surface areas
for all hinges (regardless of spanwise location).

3.3.2 Aileron Skin

The aileron skin serves the primary purpose of forming part of the aileron torsion
box, reacting torsional loads in conjunction with the spar web[13]. As the torsion
box enclosed area is relatively small (due to the high aspect ratio geometry of the
Ae-16 aileron) the skin thickness will need to be relatively large to react the
torsional loads. By designing for a distributed flange loading distribution, these
relatively thick skins will also contribute to reacting the shear and bending forces
(i.e. the skins are considered fully effective). This will reduce the loads reacted
by the spar and ribs, and lead to a weight optimised and efficient solution.

A sandwich panel design enables the bending resistance of the panel to be

increased or decreased by tailoring the core thickness. This in turn dictates the
panel’s resistance to panel buckling and out-of-plane deformation. Further, by
providing a sufficient core thickness, there is no need for intermediary supporting
ribs between the hinge locations, which would increase complexity and weight.

27
3.3.3 Aileron Ribs
The primary purpose of the aileron ribs is to maintain the aerodynamic profile, to
reduce the buckling and compression forces on the skin, and to transmit
chordwise shear and bending forces through to the spar/hinges[13]. As the skin
will aid in reacting the chordwise bending forces, the required thickness of the
ribs will be less than for an equivalent mass boom design. With composite
manufacturing techniques in mind, the ribs will essentially be a C-section design,
with the flanges providing bending resistance and to act as a connection surface
to the sandwich panel skins. By incorporating large sandwich panels in the aileron
design, the ribs are only required at locations of high stressed areas (at each
hinge point) and at the end cap locations.

The ribs will be aligned to the airflow direction as is standard for unswept wings
[9]. The alignment of the ribs with the airflow direction will give greater torsional
stiffness [9], which is deemed a critical factor for the Ae-16 ailerons. Further, as
the outer wing design has ribs aligned with the airflow direction, the load path will
be linear across the entire chord of the wing (straight load paths exhibit high
structural efficiencies compared to kinked designs).

3.3.4 Aileron Hinge Locations

The number of hinges and the spanwise placement of these hinges is a critical
design aspect as it affects the loads distribution within the aileron structure. In
deciding the number and placement of the Ae-16’s aileron hinges, the author
followed the following principles as described in Howe [9] to optimise the aileron
design:

-The shear and bending stresses along the span of the aileron spar should be
equalised. As such, the larger geometry of the inboard end of the aileron will react
higher loads than the tapered outboard end

-The spar should not deflect more than 2% of the local spar depth

-The number of hinges should be minimised to reduce the complexity and

installation burden.

28
The author used Strand7 FEA software to model the Ae-16 aileron spar, and
hence tailor the number of hinges and placement for the optimum design. The
spar length was divided into 200 beam elements, and a polynomial equation for
the spanwise aerodynamic load distribution was applied across the span
(identical to that of Figure 9). Hinge supports were modelled as node constraints
at various nodal positions, with various constraint types to reflect the fixed inboard
actuation hinges and floating outboard hinges. The beam element profile used
was a standard C-section beam with tailorable thicknesses. The model was
simplified in that it was constrained to 2D beam analysis (i.e. nil section twisting
was allowed). An isotropic aluminium material was selected as the modelling
material (note the intent of the Stand7 modelling was to find optimal hinge
placement, which is not dependant on the material choice).

The iteration process used by the author was as follows:

3 - Optimise spar
2 - Distribute hinges
thickness so that
1 - Select number of to equalise shear
stress and deflection
hinges and bending
requirements are
stresses
met

The final step in the iteration process ensured that the bending and shear
stresses within the beam did not exceed the yield limit of the aluminium material,
nor allow the beam to exceed the 2% local spar depth requirement.

Configurations consisting of four, five, and six hinge designs were analysed by
the author. Within each configuration, there were several iterations to meet the
requirements of steps two and three of the iteration process above. The five hinge
design was chosen as the optimal amount of hinges for the Ae-16 aileron due to:

-A four hinge design required a 31% increase to the spar thickness in order to
meet deflection requirements

-A six hinge design allowed a reduction in spar thickness of only 6% from the five
hinge design in order to meet deflection requirements.

29
As such, a five hinge design was the optimum design to meet the design
principles described in Howe [9]. The optimum placement of the five hinges was
determined through many iterations to be as per Table 5 below:

Table 5 - Optimum Hinge Locations and Reaction Loads

Maximum Positive Vertical

Hinge Aileron Span % Wing Station (mm) Reaction, Ultimate Load (N)
1 0.1 13974 25190
2 0.31 15158 24940
3 0.51 16286 20306
4 0.71 17414 16236
5 0.9 18486 8496

The final Strand7 model for five hinges can be seen in Figure 14 below. Note how
the stresses have been equalised across the hinge locations. The exact stress
values are not presented in tabular form, as they are not representative of what
the stresses will be in the final design.

Figure 14 - Strand7 Stress Plot and Loading Profile

Validation of the Strand7 model was accomplished by determining the flexural

rigidity of the spar using methodology discussed in Appendix D.3.4. Using the
same isotropic material as in the model, calculated deflection values were
assessed to be within 7% of the Strand7 model (deemed acceptable by the
author).

With a five hinge design, the shear force and bending moment diagrams can be
seen in Figure 15 and Figure 16 respectively.

30
 Shear Force Diagram at Critical Aileron Loading Case
15,000

10,000

Limit Load Shear

5,000
Force
Shear Force (N)

0 Ultimate Load
0% 20% 40% 60% 80% 100% Shear Force
-5,000

-10,000

-15,000
Aileron Spar Span %

Figure 15 - Spanwise Shear Force Diagram at Critical Aileron Loading Case

Bending Moment Diagram at Critical Aileron Loading

Case
3,500
3,000
2,500
Bending Moment (Nm)

2,000
Limit Load Bending
1,500 Moment
1,000 Ultimate Load
500 Bending Moment

0
0% 20% 40% 60% 80% 100%
-500
-1,000
-1,500
Aileron Spar Span %

Figure 16 - Spanwise Bending Moment Diagram at Critical Aileron Loading Case

3.3.5 Aileron Actuation Position

When the total hinge moment results for the Ae-16 aileron were passed to the
actuation team it was quickly realised that there would be sizing issues with the

31
aileron actuators. Due to the high aspect ratio and laminar flow characteristics of
the Ae-16 wing, the wing box is much smaller than for a conventional wing design.
Further, due to the small wing chord, the space between the wing rear spar and
aileron hinge line is minimal. In coordination with the actuation designer, the
author determined that the only solution was to fit the actuators asymmetrically
toward the inboard end of the aileron. Whilst this allows the required room to fit
the actuators, it does have drawbacks in that it will lead to high accumulated
torque within the outboard end. This is one notable drawback to the high aspect
ratio design of the wing, in that the torsion box structure is required to be relatively
thick compared to a conventional design.

Two actuators will be used for each aileron in order to meet redundancy
requirements. Figure 17 below details the accumulated torque transferred
through the aileron structure (primarily through the torsion box) and the actuation
reaction forces at the first and second hinge positions. The accumulated torque
was determined by integrating the torque distribution detailed in Section 2.6,
Figure 10.

Chordwise Accumulated Torque across Aileron Span

4000

2000
Accumulated Torque (Nm)

0
0% 20% 40% 60% 80% 100%
-2000
Limit Load
-4000 Torque

Ultimate Load
-6000
Torque

-8000

-10000
Aileron Span (%)

Figure 17 - Chordwise Accumulated Torque across Aileron Span

As can be seen, there is a significant accumulated torsion value across the

outboard span of the aileron, peaking at -9,141Nm ultimate load at the second

32
actuation hinge. This is the ultimate load that must be reacted by the torsion box
structure when sizing the components.

The total limit torque across the whole aileron of 11,257Nm is reacted equally
between the two actuators. Thus, each actuator reacts 5,629Nm limit (and each
actuation support structure must be designed to 8,444Nm ultimate).

4 Material Selection
The author conducted a full trade off study between various composite material
options. Further, the connecting metallic components (hinges, fasteners, trailing
edge metallic lip) required an optimal yet compatible material to be selected.

In selecting the optimal materials, the author focussed on prioritising the overall
design objective of the Ae-16, to “focus on sustainability issues through choice of
materials and recyclability” [2]. Further to this primary goal, the secondary factors
of structural performance, manufacturing process, weight, cost, and
maintainability were also considered in the material selection phase.

4.1 Composite Material Selection Methodology

In order to simplify the design, assembly, and maintainability of the aileron, a
single composite material was desired (with the addition of the core material).

In terms of a current industry accepted composite material choice for structural

applications, Carbon Fibre Reinforced Plastic (CRFP) is the most commonly used
material for structural performance. Carbon fibres inherit high compressive
strength, are relatively easy to manufacture, and display superior structural
properties compared to other fibre types [13]. An epoxy matrix is the most widely
used matrix type for low heat structural applications, inheriting high resistance to
delamination, and a relatively high toughness. Further, an epoxy matrix is
relatively cheap to manufacture (low material and energy costs) and is easy to
form into pre-impregnated (pre-preg) laminae [13].

Following the primary design objective and to reduce manufacturing energy

costs, the author selected pre-preg carbon fibre epoxy material which has the

33
option to enable a hand-layup procedure. While also significantly reducing the
manufacturing energy and environmental impact, a hand layup procedure would
result in a reduced manufacturing unit cost (based on the expected initial
production run of the Ae-16). Pre-preg material is also less susceptible to
manufacturing variances, and can lead to a highly efficient fibre volume fraction
compared to matrix injection/addition layup methods [14].

Due to the large variance in required laminate directional properties for the aileron
(i.e. high shear resistance required in skins and webs, high axial strength required
in flanges) the author has selected uni-directional fibre laminae over a bi-
directional weave. Uni-directional laminae offer much greater flexibility in terms
of being able to tailor more fibres aligned in certain directions, and so not suffer
the loss in properties due to fibre-crimp which occurs with weaved laminae [13].

In order to select the optimal CFRP material for the aileron, the Hexcel Pre-Preg
Aerospace Selector Guide was used [15]. Hexel is a composite manufacturer that
caters to the aerospace industry, and “designs, constructs, and operates its
facilities in a manner that minimizes the impact of its operations on the
environment” [16] thereby aligning with the primary design objective. The author
examined many material options, however selected the following four material
options to conduct a comparative trade-off study:

Table 6 - Composite Material Trade Off Study [17,18]

Cured
ply Density Xt E1 Xc ILSS
Material Description
thickness
(mm) (g/cm3) (Mpa) (Gpa) (Mpa) (Mpa)

UD, CFRP Prepreg, Out of

M56 AS7 0.134 1.53 2330 141 1308 115
autoclave manufacture

UD, CFRP Prepreg, Out of

M56 IM7 0.253 1.53 2730 182 1550 81
autoclave manufacture

UD, CFRP Prepreg,

8552 AS4 Autoclave only 0.13 1.58 2207 141 1531 114
manufacture
UD, CFRP Prepreg,
8552 IM7 Autoclave only 0.131 1.57 2724 164 1690 120
manufacture

34
As can be seen, the author compared pre-pregs which could be cured by an out-
of-autoclave manufacturing process, to those which required autoclave curing.
Further, for each curing method two fibre types were assessed based on variation
in the fibre stiffness (AS vs. IM).

Autoclave cured laminates inherit a lower variance in cured mechanical

properties compared to oven cured laminates, mainly due to the relatively low
presence of matrix voids, which primarily affects the compressive performance
[14]. Further, autoclave cured laminates are less susceptible to variances in
processing variables (de-baulking, curing regime etc.) compared to oven cured
laminates [14]. The downside of autoclave curing is the high manufacturing cost
and the high energy demand of the manufacturing process (pertinent to the Ae-
16). As the primary design objective is to reduce environmental impact, the low
energy manufacture of an out-of-autoclave method is the ideal choice for the Ae-
16 aileron.

As can be seen in the trade-off table, the autoclave cured samples have
significantly higher compressive strengths for both fibre types. The high
performance IM fibres appear to be more sensitive to the curing method
compared to the AS fibres, which see similar tensile strength, stiffness, and shear
strength between the two curing methods. As the IM fibres are around four times
the cost of the AS fibres [19], the AS fibres are the ideal choice for the aileron
application.

A full table of MS56-AS7 properties extracted from [17] can be viewed in

Appendix F.2.1

4.2 Sandwich Core Material Selection

Selection of the sandwich core material was accomplished using Hexcel’s
Honeycomb Selection Guide [20]. Honeycomb core material is widely used in for
aerospace panel applications due to its high strength and stiffness to weight ratio
and minimal moisture pickup properties [21]. Further, the aramid fibre honeycomb
materials manufactured by Hexel are highly compatible with the M56 facing
epoxy, and inherit high bonding strengths with nil corrosive susceptibility. The

35
author chose HRH10-4.8-48, Phenolic Resin Hexagonal Honeycomb due to its
ideal stiffness properties and environmental resistance. HRH10-4.8-48 properties
can be seen in Table 7 below:

Table 7 - HRH-10-4.8-48 Aramid Fibre Honeycomb Properties

Compressive Compressive Shear Shear

Density
Material Description Strength Modulus Strength Modulus

(kg/m3) (Mpa) (Mpa) (Mpa) (Mpa)

Aramid Fibre Phenolic
HRH-10-
Resin Hexagonal 48 1.86 138 0.59 23
4.8-48
Honeycomb

4.3 Hinge Material Selection

The primary selection criteria for the hinge material were high static strength,
fatigue durability characteristics, and compatibility with CFRP. Composite
materials are not ideal for highly concentrated loading in multiple directions (as
exists in the aileron hinges) as such, a high strength metallic alternative is
essential.

The use of aluminium alloys is not recommended for use with CFRP composites
due to the galvanic corrosive potential between the two materials [13]. Carbon is
cathodic, so the best compatible materials should also have cathodic potentials.
Titanium or passive stainless steel are suitable, opposed to aluminium which is
highly anodic. Whilst there are novel solutions to this issue (such as using a
GFRP interface between aluminium), the industry accepted practice is to use
titanium for composite connection points.

The author has selected the titanium alloy Ti-6Al-4V due to its excellent structural
performance qualities, and its prevalent use within the aerospace industry.

36
5 Initial Sizing
The purpose of the design stage was to find a rough approximation of the required
structural size. This was accomplished by using simplified analysis techniques
combined with multiple simplifying assumptions. The initial sizing enables a
starting point for iterations and optimisations to occur in the detailed design stage.

5.1 Laminate Strength and Stiffness for Initial Sizing

Conducting initial sizing analysis using composite materials inherits an added
level of complexity due to the anisotropic nature of the laminate plies. The number
of plies in each direction, and their stacking sequence within the laminate,
significantly affects the laminate directional strength and stiffness. In order to
simplify the initial sizing analysis, and to achieve an acceptable representation of
the laminate properties, the author analysed a quasi-isotropic layup consisting of
the following percentage of ply directions (0° = 25%, ±45° = 50%, 90° = 20%)
(Recommended by Howe [9]). Further, in order to approximate the laminate
directional stiffness and strength, the “10% rule” recommended by Hart-Smith
[22] was utilised.

Detailed derivation of the quasi-isotropic laminate properties for initial sizing can
be seen in Appendix D.1.

5.2 Initial Sizing - Aileron Skin

Whilst the final design will be analysed as a distributed flange load distribution,
for simplification purposes the initial sizing of the skin was assessed assuming a
mass-boom load distribution. Essentially this means that the skin was considered
ineffective in reacting the shear and bending forces. While this method is
unconservative in relation to skin sizing, it was expected that the skin size would
be driven by the requirement to react the high torsional forces (see Figure 17),
coupled with the small enclosed area of the high aspect ratio torsion box.

Under a mass-boom distributed loading analysis, the skin was assessed for
torsional stiffness, shear stress failure from applied torsion, bending stress from
out-of-plane pressure, and out of plane deflection failure criterion. Detailed

37
analysis for all calculations can be seen in Appendix D. The resulting skin
thickness requirements can be seen in Table 8.

Table 8 - Initial Sizing - Sandwich Skin Required Thickness

Skin Component Assessment Criteria Required Thickness

Sandwich Facing Torsional Stiffness 7.6mm (2 x 3.8mm)

Sandwich Facing Shear Stress (Torsion Induced) 0.8mm (2 x 0.4mm)
Sandwich Core Bending Stress (Out-of-plane Pressure Induced) 1.918mm
Sandwich Core Out-of-plane Deflection 4.22mm

5.3 Initial Sizing - Aileron Spar

The initial sizing of the spar was conducted assuming a mass-boom load
distribution. Essentially this means that the spar is assumed to react 100% of the
spanwise accumulated shear (primarily through the spar web) and 100% of the
spanwise bending moment (primarily through the spar flanges). This is a
conservative simplification, as in reality the skin will also react a portion of each
loading type. It is expected that due to this simplification, and the use of the quasi-
isotropic layup, that the initial thickness of the spar components will be larger than
the final design.

5.3.1 Spar Flanges

Analysed as a mass-boom configuration, the spar flanges are assumed to react
100% of the spanwise bending moment through membrane axial reactive forces.
As such, they have been assessed for compressive stress failure and
compressive instability failure (under ultimate loading), and flexural rigidity (under
limit loading). The resulting spar flange thickness can be seen in Table 9 below:

Table 9 - Initial Sizing Results - Spar Flange Required Thickness

Spar Component Assessment Criteria Required Thickness

Flange Compressive Stress Failure 1.18mm

Flange Compressive Instability Failure 4.01mm
Flange Flexural Rigidity Assessment 2.23mm

As can be seen, the critical failure criterion is with local buckling. It was expected
that this initial thickness would be reduced in detailed sizing by tailoring the

38
laminate ply percentages to increase the buckling resistance (i.e. by including
higher percentages of ±45° plies compared to the quasi-isotropic layup).

5.3.2 Spar Web

Analysed under a mass-boom loading distribution, the spar web was assumed to
react spanwise vertical shear forces and torsional shear forces (as it forms a
component of the aileron torsion box). By undertaking a maximum stress
assessment at the critical spanwise location (see Appendix D.4) the maximum
required web thickness was identified and detailed in Table 10 below.

Table 10 - Initial Sizing - Spar Web Required Thickness

Spar Component Assessment Criteria Required Thickness

Web Shear Stress Failure 2.35mm

5.4 Initial Sizing - Aileron Ribs

Initial sizing of the ribs assessed the critically loaded rib only. For the Ae-16
aileron, the critical rib location is at the second hinge position due to the relatively
high air loads (24,940N ultimate, see Table 5), the actuation reaction loads, and
the smaller geometrical dimensions compared to the inboard ribs.

The aileron ribs transfer the chordwise shear loads (primarily through the rib web)
through to the spar web, and react the chordwise bending moments through the
rib flanges.

Assessed as a mass-boom load distribution (for simplicity) the skins were

assumed to be ineffective in reacting the chordwise shear and bending forces. In
reality, the skin will share a portion of these loads, which would be accounted for
in future detailed sizing calculations (using a distributed flange load distribution
assessment). As such, the initial thicknesses detailed in Table 11 below were
expected to be reduced in future iterations. Please see Appendix D.4 for detailed
derivation:

39
 Table 11 - Initial Sizing - Rib Required Thickness

Rib Component Assessment Criteria Required Thickness

Rib Web Shear Stress Failure 2.06mm

Rib Flange Shear Stress Failure 5.8mm

5.5 Initial Sizing – Summary and Commentary

The required thicknesses of the Ae-16 aileron components determined from initial
sizing analysis is presented in Table 12 below. The cured ply thickness of M56-
AS7 material is 0.134mm [17], and was used to determine the required number
of plies to meet the initial sizing requirements:

Table 12 - Initial Sizing Summary Thicknesses

Component Required Thickness Driving Failure Number of M56-AS7 Plies

Sandwich Facing 7.6mm (2 x 3.8mm) Torsional Stiffness 58 (2 x 29)
Sandwich Core 4.22mm Panel Deflection N/A
Spar Flange 4.01mm Flange Buckling 30
Spar Web 2.35mm Shear Stress 18
Rib Web 2.06m Shear Stress 16
Rib Flange 5.8m Shear Stress 44

As discussed previously, the author expected the majority of these thicknesses

to be reduced during detailed analysis, primarily due to the assumptions of
reduced skin effectiveness in reacting shear and bending forces, but secondly
due to the calculations using a quasi-isotropic laminate. The main advantage of
using composites is the ability to tailor the layup sequence to match the loading
type on the individual component, which was not done for initial sizing. For
example, a quasi-isotropic laminate is considerably weak in shear (the laminate
used in initial sizing had an in-plane shear modulus of only 5.6GPa). Laminate
tailoring should enable a four-fold increase in this shear stiffness value for shear
critical components such as the sandwich skin facings, and spar web [19].

These thicknesses, whilst based off simplistic analysis techniques, provided an

acceptable base line for which to assess and optimise in future detailed sizing
steps.

40
6 Detailed Design
The purpose of the detailed design stage was to refine and optimise the
component sizing based on complex and interactive analysis techniques. Further,
the detailed design stage forms the transition point between a basic initial
structure, to one that inherits finer details such as connection types, surface
coatings, secondary structure, and non-structural additions (i.e. static wicks).

The detailed design stage provides final component sizing which can then be
used to produce final engineering drawings, and enable the assessment of the
structure using Finite Element Analysis (FEA) and Fatigue and Damage
Tolerance (F&DT) analysis.

6.1 Detailed Design – Iteration Methodology

Contrary to the analysis methods used in initial sizing, the detailed sizing analysis
will assess the aileron structure under a distributed flange load distribution. This
method of analysis assesses the skin as fully effective in resisting shear and
bending forces, as opposed to a mass-boom distribution where the skins only
resist torsion through shear [9]. As discussed previously, the skin is required to
be relatively thick due to the combined factors of the high accumulated torsional
load and the small enclosed area of the aileron torsion box (a negative
consequence of the high aspect ratio design). Therefore, by also assessing the
skins capacity to react shear and bending forces, the load will be distributed away
from the other primary structural components (spar, ribs etc.)

While the distributed flange loading distribution analysis will lead to a highly
efficient and optimised design, it does make the design iteration and optimisation
process markedly more complex to undertake. This is due to the load sharing
interrelationship between the components, where a reduction in weight in one
component, may lead to an excessive load in another component (for example,
reducing the skin thickness will mean the spar and rib flanges will need to react
a higher percentage of the bending moment). This complexity is exacerbated by
the use of composites, as there are the extra variables of fibre orientation
percentages and stacking sequence within each component. While the majority

41
of prior GDP students have assessed their designs as simplified mass-boom
loading distributions, the author deemed these solutions far too overweight, and
do not reflect the optimal design that could be achieved with the additional effort
inherent in a distributed flange analysis.

The detailed design iteration process used to optimise the skin and spar
thicknesses (spanwise load reaction components) is detailed in Figure 18 below:

1. Select laminate
layup for all
components

6. Propose iterations for 2. Calculate membrane

each laminate layup to stiffness properties for all
optimise for failure/weight components using COALA

5. Assess for failure using 3. Calculate effective

COALA, ESDU, and other bending stiffness Ixx' for
analysis techniques the aggregate cross section

4. Calculate membrane
loads Nx, Ny, Nxy within
each component

Figure 18 - Detailed Design Iteration Process

As can be seen, each iteration presented a significant amount of work. An

equivalent iteration process for a metallic design would require far less effort per
iteration (as there is no ply tailoring process, no ply failure analysis, and less
variables per iteration step). The methodology within each step are described in
detail below.

6.2 Iteration Step 1: Laminate Layup Selection

Starting from the initial sizing thicknesses, the number of plies within each
component were selected to meet these initial sizes. Contrary to the initial sizing
layup, the ply orientations could be tailored to meet the loading requirements of
the individual components (as opposed to a quasi-isotropic layup). For example,

42
components under high shear loads could be assigned a higher percentage of
±45° plies, and components under axial membrane loads could share a higher
percentage of 0° plies.

Each laminate stacking sequence iteration was based on guidelines from ESDU
82013 [23], Hart’s 10% rule [22], as well as the general guidelines from [24], [25],
and [26], detailed below:

1. Ensure a balanced stacking sequence to prevent shear-extension

coupling
2. Ensure a symmetric stacking sequence to prevent extension-bending
coupling
3. Stack no more than four plies of the same orientation together to limit edge
splitting and transverse shear stress
4. Place at least one 45° ply at the surface of the laminate to improve the
damage tolerance qualities of the panel
5. Positon a 90° ply close to the surface (ideally just after the 45° ply) in order
to maximise transverse bending stiffness
6. Assign a higher percentage of ±45° plies to laminates under high shear
loads or those that are buckling critical

Further, the author followed the guidance of ESDU 82013 [23] in selecting a
stacking sequence where the flexural stiffness sub-matrix ( ) is orthotropic, with
the bending and twisting stiffnesses uncoupled so that and are zero.
These specially orthotropic stacking sequences are accomplished by a specific
ordering of the laminate orientations within the stacking sequence, and result in
maximised laminate stability and reduction in distortion during manufacture.
Specially orthotropic laminate layups are critical for thin laminates (less than 21
plies), however less so for larger laminates [23].

As can be seen from the initial sizing summary (Table 12), the initial laminate
thicknesses were considered large by EDSU 82013 standards (larger than 21
plies). Therefore, the author modified the specially orthotropic stacking
sequences presented in ESDU 82013 in order to attain flexural stiffness sub-
matrix component values and as close to zero as possible. Throughout

43
the iteration process, the laminate thicknesses were reduced to within ESDU
82013 thresholds, and fully specially orthotropic stacking sequences could be
used (and selected based on the other iteration guidelines detailed above).

After a complete detailed stress analysis of an iteration, the laminate layups were
adjusted based on the returned reserve factors. During the first iteration (using
initial thicknesses) it was readily apparent that the laminates were far too large
(with high reserve factors for all failure criterion). This was expected, due to the
departure from the inefficient quasi-isotropic layup sequence that was used
during initial sizing.

Subsequent iterations were aimed at reducing the thickness of the component

laminates, with full detailed stressing performed on each iteration to ensure
reserve factors remained positive. The first and last iteration layups are detailed
in Table 13 below:

Table 13 - Iteration Extract Showing First and Final Iterations

Iteration t
Number Component Layup Plies (mm)
Spar Flange S11 MOD [(+45/90/-45/-45/0/0/0/+45)s]2 32 4.288
Spar Web S11 MOD [(+45/90/-45/-45/0/0/0/+45)s]2 32 4.288
S31 MOD (+45/90/90/-45/0/0/-45/-
1 Sandwich Facing
45/+45/+45/-45/+45/-45/+45)s
28 3.752

S31 MOD [(+45/90/90/-45/0/0/-45/-

Skin at Flange 56 7.504
45/+45/+45/-45/+45/-45/+45)s]2
Sandwich Core Aramid Honeycomb 1 4.22

Iteration t
Number Component Layup Plies (mm)
Spar Flange S1 (+45/-45/-45/90/+45/0)s 12 1.608
Spar Web S1 ([+45/-45/-45/90/+45/0]s)s 24 3.216

5 Sandwich Facing
Skin at Flange
S1 (+45/-45/-45/90/+45/0)s
S1 (+45/-45/-45/90/+45/0)s
12
12
1.608
1.608
Sandwich Core Aramid Honeycomb 1 4

The full iteration table, detailing the layup iterations and methodology for each
iteration selection can be seen in Appendix E. The ‘skin at flange’ component is

44
where the sandwich skin facing panels come together at the spar web location,
and extend forward over the spar flange. This skin is adhered to the spar flange,
and contributes to the spanwise bending resistance of the cross section (see
Figure 19 below):

Figure 19 - View of INBD End (Looking Down Pivot Axis)

Local instability failure tended to be the driving failure criterion for the majority of
components. As such, final iterations incorporated a far higher percentage of ±45°
plies than one would expect. Even for components primarily under axial loading
(such as the spar flanges), the buckling failure criterions were far more critical.
Regardless, the final thickness of the laminates were deemed to be at the
minimum acceptable values, without leading to extremely thin sections that would
not inherit sufficient damage tolerance (from impacts or lightning etc.)

The author also performed the laminate stacking iterations with manufacturing
criteria in mind. As such, the final laminate stacking sequences are duplicated
over several components, thereby significantly simplifying the manufacturing
process. Further, transition between the spar web and spar flange is simple, in
that the outer six plies on either side of the spar web are continuous as they
transition to for the spar flange (they share a continuous stacking sequence). This
also applies to the transition between the sandwich facing panels and the skin
above the flange.

45
The final aspect of the stacking sequence that the author wishes to discuss is the
balanced interaction between the spar flanges and the skin adhered to these
flanges. As these components are adhered to one another, they act essentially
as one single laminate. Therefore, by ensuring that the stacking sequence of both
the spar flange and adhered skin are identical, there will be no aggregate shear-
extension coupling due to asymmetric balancing (the components acting in
unison will behave as a specially orthotropic laminate).

6.3 Iteration Step 2: Laminate Properties from COALA

COALA is a composite analysis software program distributed by Cranfield
University to enable easy determination of laminate properties and ply failure
mechanisms. By entering various input parameters such as ply stiffness/strength,
stacking sequence, ply thickness, temperature and moisture factors into COALA,
the equivalent laminate structural properties can be easily obtained.

As part of each iteration step, each component laminate layup sequence was
entered into COALA to ascertain the equivalent laminate membrane stiffness
values and the flexural stiffness matrix [D]. The laminate stiffness values were
subsequently used to ascertain the equivalent bending stiffness of the aggregate
aileron cross section, and the flexural stiffness values were used to determine
buckling resistance of the laminates.

A 1% moisture content, and a 155K temperature difference were used within the
COALA analysis, based of cure data for MS56 obtained from Hexel [17].

6.4 Iteration Steps 3 and 4:

The methodology used to determine the effective bending stiffness ( ′) and the
membrane loading on each laminate ( , , ) is covered in detail in Appendix
F.2, and so will not be repeated here.

The main assumption used in deriving the loading distribution was that the spar
and skin flanges forward of the spar web do not react shear or torsional loads
(i.e. only the torsion box components are considered effective in reacting these
loading types). This assumption is valid due to the sub-optimal orientation of the

46
flanges (they are perpendicular with respect to the vertical applied shear force),
and the open celled configuration leading to poor torsional stiffness
characteristics.

6.5 Iteration Step 5: Assess Laminate Failure

There are several failure modes possible for each component of the composite
aileron structure, which are dependent on loading and support conditions. The
following assumptions were followed throughout the analysis, based on guidance
from [24] and [27].

- All deformations are small

- Linear elasticity and Hooke’s Law are applicable in ply analysis
- Plane stress is applicable in ply analysis
- Moduli values are equal in tension and compression
- All plies are macroscopically homogeneous

Firstly, all laminates were assessed in COALA for individual ply failures using the
maximum load intensities determined in the previous step (Nx, Ny, Nxy). Stress
failure of individual plies were determined using an interactive failure criterion
(Tsai-Hill) and a non-interactive failure criterion (Maximum stress) in order to
identify the failure mode (recommended by [24]). The Tsai-Hill failure analysis
better reflects the actual loading state within the laminate, as it accounts for the
stress relieving loads acting on each ply in the transverse direction.

The maximum strain failure criterion was also assessed using COALA, based on
a maximum allowable strain of 3500 . This limit is driven by damage tolerance
allowances and applicable knockdown factors as recommended in the AVD
Damage Tolerance and Fatigue notes [28].

Shear and compression buckling failure modes were assessed for all
components using ESDU sheets 80023 [29], 03013 [30], and AVD stability lecture
notes [27]. All panels were conservatively assumed as simply supported. Further,
the panels were assumed as flat to achieve consistency with the ESDU
methodology.

47
The sandwich panels were assessed for maximum bending stresses within the
facing panels using methodology in Niu [25]. These out of plane bending loads
are a result of the aerodynamic pressurisation load acting on the panel surface.
Sandwich panel deflection was also assessed using limits and methodology as
recommended in Niu [25] (maximum panel deflection limited to 0.5 inch).

Overall sectional deflection was assessed at the critical free ends of the aileron
(the sections inboard and outboard of the first and fifth hinges respectively). The
bending stiffness of the aileron cross section ( ) was used with methodology
specified in [31] to find tip vertical deflection of the inboard and outboard ends
(see Appendix F.2.18 for detailed methodology). This tip deflection was checked
to ensure it remained within the 2% spar height limit, as recommended by Howe
[9]. The deflection analysis presented in Appendix F.2.18 also served as a
validation check for the FEA analysis (discussed in Section 10.2 of this report).

Finally, as the Ae-16 aileron reacts large torsional loads within a small enclosed
torsion box, the author deemed it important to check the torsional stiffness
requirements were met at each successive iteration. As such, a torsional stiffness
check was undertaken using the methodology detailed in Appendix G.1.

The critical failure modes and associated reserve factors for each iteration were
recorded within the table shown in Appendix E. These reserve factors were then
used to identify areas for optimisation (discussed in the following section). An
extract detailing the final iteration reserve factors is shown in Table 14 below:

Table 14 - Final Laminate Reserve Factors

Iteration Strength Buckling Deflection Torsional

Number Component Layup RF RF RF Stiffness
Spar Flange S1 (+45/-45/-45/90/+45/0)s 1.427 1.202
Spar Web S1 ([+45/-45/-45/90/+45/0]s)s 3.311 1.67

5 Sandwich Facing S1 (+45/-45/-45/90/+45/0)s 1.508 1.252

3.205
2.006

Skin at Flange S1 (+45/-45/-45/90/+45/0)s 1.425 1.167

Sandwich Core Aramid Honeycomb N/A N/A 1.024

48
6.6 Iteration Step 6: Optimise Laminate Layup
If the difference between the returned reserve factor was significantly above the
ideal value of 1.0, the stacking sequence would be amended for the proceeding
iteration. This would usually be accomplished by selecting a specially orthotropic
laminate (using ESDU 82013 [23]) that was comprised of fewer plies.

If the returned reserve factor from failure analysis was below 1.0, this indicated
the laminate failed prior to achieving the required loading. For the several
iterations conducted by the author, the consistent critical failure mode was local
instability (buckling) of the laminate panels. As detailed in the buckling
methodology used in ESDU sheets 80023 [29] and 03013 [30], the bending
stiffness [ ] matrix is the driving laminate property that defines buckling
resistance (with higher [ ] matrix values leading to larger allowable buckling
loads). The bending stiffness [ ] matrix is determined by multiplying the
individual ply stiffness by a cubic function of its distance from the neutral axis of
the laminate [24]. This means that a small increase in laminate thickness will tend
to increase the overall laminate bending stiffness in an exponential manner.
Further, as recommended by [24], [25], and [26], the addition of a higher
percentage of ±45° plies can be used to increase the buckling resistance of the
laminate (which effectively increases the bending stiffness values).

The laminates analysed by the author through the successive iterations showed
high sensitivities to small changes to laminate thickness and composition. For
example, between iterations four and five, the spar flange laminates were
amended to incorporate a higher percentage of ±45° plies and to increase the
overall thickness by two plies only. The reserve factor against buckling jumped
from 0.436 to 1.202 based on this small change. Thus, while optimisation is
possible, there is a limitation to how close the reserve factors can be brought
close to the ideal value of 1.0 (see Appendix E for further details).

49
6.7 Spanwise Detailed Analysis - Final Iteration
The above iteration process was completed several times by the author in order
to identify highly optimised laminates for the spar web, spar flanges, and
sandwich skin panels. The final laminate stacking sequence for each of these
components is detailed in Table 15 below:

Table 15 - Optimised Spar and Skin Laminates Stacking Sequences

Component Layup Plies t (mm)

Spar Flange S1 (+45/-45/-45/90/+45/0)s 12 1.608
Spar Web S1 ([+45/-45/-45/90/+45/0]s)s 24 3.216
Sandwich Facing S1 (+45/-45/-45/90/+45/0)s 12 1.608
Skin at Flange S1 (+45/-45/-45/90/+45/0)s 12 1.608
Sandwich Core Aramid Honeycomb 1 4

Each optimised component is comprised a specially orthotropic laminate

(balanced, symmetric, with , = 0). As discussed previously, a commonalty
in stacking sequence is also used across all laminates in order to significantly
simplify the manufacturing process (the impact on failure indices for having this
commonality is negligible as can be seen in the full iteration table within Appendix
E). Further, as the spar web to spar flange transition is simple (due to the same
stacking sequence) the manufacturing process is further simplified. The full
detailed stressing method and process used for each iteration can be seen in
Appendix F.2.

It must be noted that the iteration process was a lengthy and demanding stage of
the design work, particularly due to the authors choice to analyse the structure
as a distributed loading distribution (where changes to each component affect the
overall loading distribution). The author is confident that the additional effort put
into this analysis has led to a highly optimised design that exhibits minimal weight.

50
6.8 Detailed Design: Ribs
The ribs transmit the chordwise shear and bending forces, resulting from
aerodynamic pressure, forward through to the spar/hinges [13]. As the skin is
considered effective in transmitting bending loads (distributed flange loading
distribution) it will aid in reacting the chordwise forces, and must be accounted
for in the analysis of the ribs (they will share the chordwise bending loads with
the rib flanges). The iteration process is less complicated than the spanwise
reactive components (spar web, flange, and skin) discussed above, as there are
only two interactive components, the skin and the rib.

In order to determine the loads passed through the rib laminates, the chordwise
shear force and bending moment diagrams were created for the critical loading
case. In order to produce these diagrams, the shear flow within the torsion box
due to chordwise loading was first ascertained. The shear flow is made up of two
contributions:

-Shear flow due to applied local aerodynamic loading. This applies a shear
and torsional force which must be reacted by the torsion box.

-Shear flow due to applied hinge loads. This also applies a shear and
torsional force which must be reacted by the torsion box.

For all analysis steps, the critically loaded rib at Hinge Position 2 was analysed.
This rib location is critical due to the high shear reaction loads (24,940N, see
Table 5), the actuation loads, and the smaller enclosed area of the torsion box
compared to the more inboard ribs. Analysis of this critical rib location will lead to
conservative thicknesses for the remaining ribs.

6.8.1 Torsion Box Shear Flow

After a lengthy process (detailed in Appendix F.2.20) the shear flow within the
torsion box due to aerodynamic loading and hinge/actuation loading was
ascertained, and is shown in Figure 20 below:

51
 Figure 20 - Combined Torsion Box Shear Flow from Chordwise Loading

Where, the red shear flow depicts the contribution from hinge/actuation loading,
and the blue shear flow result from aerodynamic loading. Summing these two
shear flows together gives the resulting torsion box shear flow detailed in Figure
21 below:

Figure 21 - Combined Torsion Box Shear Flow from Chordwise Loading

As one would expect, as the applied aerodynamic load (24,940N) is equal to the
vertical hinge reaction (24,940N), the shear flow effects from vertical shear cancel
each other out. Essentially this results in the overall shear flow to equal the
difference in torsional reactions between the two loading cases (which means
shear flow is constant throughout the torsion box, opposing the combined torsion
load).

52
6.8.2 Chordwise Shear Force and Bending Moment Diagram
To create the chordwise shear force and bending moment diagrams for the critical
rib, the applied loading system (aerodynamic and hinge loads) was combined
with the reactive loading system (torsion box shear flow) as detailed below:

Figure 22- Combined Applied and Reactive Chordwise Loading Systems

The resulting shear flow and bending moment diagrams in the chordwise
direction are detailed in Figure 23 and Figure 24 below:

As can be seen in the diagrams, the highest shear force in the rib is -28,304.9N
at the spar web to rib interface. This maximum shear force must be reacted by
the rib web. The largest bending moment is also at the same chordwise location,
with a value of -7,306Nm. The maximum bending moment must be reacted by
the rib flanges and skins (due to the effectiveness of the skins in reacting bending
loads).

53
 Chordwise Shear Force - Critical Rib
0% 20% 40% 60% 80% 100%
0

-5000
Shear Force (N)

-10000

-15000

-20000

-25000
Spar Web Location
-30000

-35000

Chord (%)

Figure 23 - Chordwise Shear Force Diagram - Critical Rib

Chordwise Bending Moment- Critical Rib

0% 20% 40% 60% 80% 100%
0
-1000
Bending Moment (Nm)

-2000
-3000
-4000
-5000
-6000
-7000 Spar Web Location

-8000
-9000

Chord (%)

Figure 24 - Chordwise Bending Moment Diagram - Critical Rib

54
6.8.3 Detailed Sizing Analysis: Ribs
The maximum chordwise shear force and bending moments were used to
perform detailed sizing analysis for the ribs. The iteration process was near to
identical as the process used for the spanwise reactive components (see Figure
18). The only variation was that the direction of effective bending stiffness was
instead assessed in the chordwise direction. Please see Appendix F.2.20 for the
detailed methodology used for rib analysis.

Initial sizing thickness requirements were used to select an appropriate specially

orthotropic laminate using ESDU 82013 [23]. Detailed stressing was then
conducted for all applicable failure modes for each rib component. The reserve
factors for the first and second iterations are detailed in Table 16 below:

Table 16 - Rib Laminate Stacking Sequence Iterations

Iteration
Number Component Layup Plies t (mm) Strain Buckling
S6 (+45/90/-45/-45/-
Rib Web 45/+45/0/+45)s 16 2.144 1.05 0.46
1 S6 ([+45/90/-45/-45/-
Rib Flange 45/+45/0/+45]s)s 32 4.288 3.659 110.2

Iteration
Number Component Layup Plies t (mm) Strain Buckling
S1 ([+45/-45/-
Rib Web 45/90/+45/0]s)s 24 3.216 1.174 1.582
2 S1 ([+45/-45/-
Rib Flange 45/90/+45/0]s)s 24 3.216 3.671 51.155

For the web, the initial iteration failed in buckling, and was very near to failing
under maximum strain allowable. As such, for the second iteration the thickness
was increased and the percentage of ±45° plies increased to improve buckling
resistance.

For the flanges, the reserve factors were very high for the initial iteration. This
was because the thickness from initial sizing did not take into account the
effectiveness of the skin to react chordwise bending. Subsequently, the thickness
was reduced for the second iteration to optimise on weight.

55
The reserve factors from the second iteration were near to ideal, with the reserve
factor against web shear buckling at = 1.174. Further iterations performed by
the author could not improve on this second iteration. The final layup for the rib
is thus what is shown in Iteration 2, Table 16. This layup is a specially orthotropic
layup (balanced, symmetric, with , = 0) which leads to maximum stability
and minimal distortion from the curing process.

The author decided to keep the stacking sequence of the rib flange identical to
that of the rib web, as this would significantly simplify the manufacturing process.
This would only come at a minimal penalty in weight due to the small geometry
of the rib flanges. The final rib flange width was 20mm, and while this value could
be smaller based on strength, there needed to be sufficient surface area to
adhere the rib and skin together (Analysed in Appendix F.3.1).

It must be noted at this stage that the author had considered the inclusion of
lightening holes within the rib webs, which is a standard design feature for metallic
ribs. The author chose not to implement this design feature due to:

-Minimal weight advantage possible due to the small geometrical area of

the rib webs. This small rib web size is an inherent characteristic of a high
aspect ratio aileron design

-Complexity and manufacturing demands from using cutouts within

composite panels. Adding lightning holes would increase the interlaminate
stress and reduce rib load bearing capacity, which would require
significant reinforcement [13].

6.9 Detailed Design: Hinge Analysis

The aileron structure is connected to the outer wing by way of five hinge
attachment points, with the two inboard hinges also transferring the actuation
forces by way of two linear Electronic Hydrostatic Actuators (EHA). The most
critical hinge is the second most inboard hinge, as the vertical reaction forces are
high (24,940N, see Table 5) and the size of the hinge is minimal due to tapering
of the aileron spar. The required actuation force is highly dependent on the
distance between the pivoting lug and the actuation lug. This distance is

56
essentially a moment arm, and by maximising this distance, the required
actuation force is minimised (smaller reaction force acting at a larger moment
arm). Thus, the overall goal of detailed hinge sizing was to maximise this
distance, while attaining overall positive reserve factor values.

Another factor which influenced where the connection lugs could be positioned
was the need to have a single rotation axis along all five pivoting hinges. Due to
the taper and twist of the aileron along its spanwise length, this requirement
reduced the allowable placement of the lug positions at the inboard locations
(which essentially reduced the distance possible between the pivoting lugs and
actuation lugs). The vertical location of the pivot rotation axis was checked and
adjusted throughout the iteration process to ensure a valid pivot axis was
maintained (see Figure 25 below):

Figure 25 - Hinge Pivot Axis Alignment

The author conducted multiple iterations to find the optimal pivot and lug
placement, as well as the size of the bolts, bearings, and lugs. This process took
a significant amount of time to complete, as all factors were interrelated. For
example, a smaller moment arm between the pivot and actuation points meant a
higher actuation load, which required larger bolts/bearing/lugs, which in turn
reduced the net tensile areas of the lugs.

57
After a lengthy iteration process the optimal hinge configuration was determined
by the author (with all RF values close to the value of 1.0). Unfortunately, upon
subsequent fatigue analysis of the lugs, the optimal static configuration of the
lugs resulted unacceptably short fatigue lifetimes (around 1/10th of the aircraft
lifetime). As a result, the entire hinge configuration had to be redesigned based
on a damage tolerant basis. Further, the original configuration of a single wing
side lug mating with double aileron side lugs had to be modified to enable three
wing side lugs to mate with double aileron lugs (see Figure 26 below)

Figure 26 - Wing Side Triple Lug Hinge and Linear EHA

The entire detailed stressing analysis was performed again with the new
configuration, and the fatigue analysis (discussed in Section 8.3) now showed
acceptable fatigue lives. The results of the detailed stressing analysis and critical
reserve factors can be seen in Table 17 below.

Table 17 - Reserve Factors for Final Hinge Configuration

Reserve
Hinge Component Critical Failure Mode
Factor
Wing Side Triple Lugs Shear-Bearing Ultimate Oblique Rupture 2.069
Bolt Bolt Bending Failure 1.808
Aileron Side Upper Pivot Lugs Shear-Bearing Ultimate Axial Rupture 3.043
Aileron Side Lower Actuation Lugs Shear-Bearing Ultimate Axial Rupture 3.917
Actuator Rod End Bearing Ultimate Load Failure 1.138

58
The final hinge design is damage tolerant (with acceptable inspection limits
detailed in Section 8.4) and fail safe, in that no single lug failure could lead to loss
of the aileron (three lugs on one side of the hinge, and two on the other).

For the lower actuation connection point, a single central rod end bearing extends
from the linear EHA, and mates with two aileron side lugs. Unlike the upper pivot
lugs, these two lugs do not require bearings (due to the central rod end bearing).
The instead have bushings to enable easy maintenance should excessive wear
occur.

As can be seen in Figure 26 above, the linear Electro Hydrostatic Actuator (EHA)
was of a size that could not fit between the rear spar of the wing and the spar
web of the aileron. A solution was chosen after numerous collaborations with the
actuation team and the outer wing designers to place the actuator within the outer
wing. This solution requires a reinforced cutout within the rear spar web to allow
for the actuation arm. The placement of the actuator within the wing does not
impact on fuel capacity, as the outer wing at this location holds no fuel (due to
the location existing outboard of the wing fold).

The final configuration of the hinge components is shown in Table 18 and Figure
27 below. All components were selected to meet required loading and fatigue
requirements.

Table 18 - Hinge Components

Actuator Hinge Pivot Hinge

Component Details Component Details
Configuration 1 to 2 (wing to aileron) Configuration 3 to 2 (wing to aileron)
Bearing 1 x P/N M81935/4-10 Bearing 2 x P/N MS14102-10
Bushings 2mm Flanged MIL-B-81934/2 Bushings 2mm Flanged MIL-B-81934/2
Bolt S99 5/8" Bolt S99 5/8"
Washer S99 Washer S99
Nut 5/8" UNF Castellated Nut S99 Nut 5/8" UNF Castellated Nut S99
Split Pin 1/16" Split pin MS24665 Split Pin 1/16" Split pin MS24665

59
 Figure 27 - Aileron Side Hinge Components

As can be seen within Figure 27 above, the actuation hinges incorporate a

number of pockets in order to reduce the overall weight of each hinge. The
required thickness of the lugs to meet the fatigue requirements, and to match the
required bearing surface, is 14.40 . The initial optimisation completed by the
author proved that lugs that were 5 thick provided positive reserve factors for
static loading. As such the ‘web’ of each hinge is 5 with thicker areas around
the lugs and bracing. The pockets are on the outer faces of each lug, as this
enables a one-sided milling manufacturing process to be used. The pockets have
been analysed in FEA to identify the optimal placement and to validate the
detailed stressing results (see FEA Section 10.4).

The attachment of the hinges to the aileron spar is a critical connection point as
the transfer of all aerodynamic, actuation, and hinge loading occurs here. As
such, the author chose to use mechanical fastening (in lieu of adhesion) between
the hinges and the spar. Mechanical fasteners inherit a higher reliability over
adhesion, as their properties are less affected by environmental factors (heat,
moisture) or adhesion variables (surface finish, glue application etc.). Adhesion
connections also suffer from exceptionally weak properties in interlaminar tension
and peel strengths [13], which would be pertinent to the dynamic oblique loading

60
experienced by the hinges. Finally, mechanical fasteners better meet the damage
tolerance requirements of CS25.571, where there would be no chance of a single
delamination failure leading to loss of the hinge [1].

The fasteners selected are Hi-Lok® HL11-6, titanium flush shear-head pins. The
titanium material is required to preclude galvanic corrosion between the fastener
and composite material. There are 10 fasteners connecting the hinge to the spar
web, and 8 fasteners through each upper and lower flange. Full analysis of these
connections can be seen in Appendix F.3.2.

7 - Aeroelastic Analysis
A detailed analysis of the torsional stiffness of the aileron was conducted by
meeting the aileron stiffness requirements specified within the British Civil
Airworthiness Requirements (BCAR), subsection K3-9 [32]. The analysis method
was similar to that performed in initial sizing, however includes added criteria
such as enhanced aileron geometry, design speeds, and the use of actual
laminate shear stiffness values.

The calculated torsional stiffness of the composite aileron was far higher than
that required by the BCAR standards, indicating that the optimisation process
conducted by the author was highly successful, and detrimental twisting of the
aileron under load would not be an issue. See Appendix G.1 for detailed analysis.

A rudimentary divergence speed calculation was also conducted to ensure that

aileron divergence does not occur within regulatory set design speeds. Control
surface divergence must be prevented as per CS-25.629 [1], which requires no
divergence occurs at speeds up to 1.2 × , which for the Ae-16 is 1.2 ×
192.15 = 230.58 (see Table C-2 for max value reference).

Divergence analysis methodology as specified in AVD-0441 [33] and Babister

[34] was followed to ascertain the divergence speed of the aileron. The calculated
divergence speed for the composite aileron was 545.6 , which is higher than
that required by CS25.629 (230.58 ) [1]. See Appendix G.2 for detailed analysis.

61
8 - Fatigue and Damage Tolerance Analysis
In order to satisfy CS25.571 requirements, detailed design of the aileron structure
must show that catastrophic failure due to fatigue or corrosion can be avoided for
the duration of the aircraft’s specified operating life [1]. While the industry
approach to designing a damage tolerant structure would involve multiple
complex analyses, including static and cyclic load testing of components, this
approach is beyond the scope of this thesis.

The author chose to perform a simplified fatigue analysis of the critically loaded
actuation hinge at the second hinge position. The upper pivot lugs of this hinge
are under critical loading, as they react both shear and actuation reaction loads.
A fatigue and damage tolerance analysis of the pivot lugs is appropriate due to
these relatively large dynamic and oblique loading cases.

Note, that the composite structure of the aileron has been designed damage
tolerant through the implementation of a 3500 strain failure criterion (a significant
knock down from the published failure strains in order to account for a damage
tolerant considerations such as BVID limits [28]).

8.1 Aileron Fatigue Analysis Process

The process used by the author to conduct the fatigue and damage tolerance
assessment is as per Figure 28 below. For detailed discussion of the
methodology and calculations, see Appendix H.

Determine Probability
Determine Aileron Determine Critical
of Reaching Aileron
Usage Frequency per Aileron Lug Gross Stress
Deflection Angles for
Flight Hour at Limit Loading
Ae-16

Determine Occurance
Assess Fatigue
of Each Stress Level Per Calulate Lug Stress At
Spectrum using
Flight Hour (Spectrum Each Deflection Angle
AFGROW
Creation)

Figure 28 - Aileron F&DT Assessment Process

62
The primary assumptions used by the author in the above analysis process are
as per below:

1- The ailerons are activated once every ten seconds during the climb and
decent stages of a flight, and once every two minutes during cruise
2- The aileron deflection probability matches that presented by A.R.B White
Note No. 96 [35] for a commercial transport aircraft
3- The lug stress at each deflection level have a linear relationship with the
limit stress/deflection level (i.e. a 10% aileron deflection results in a stress
level that is 10% of the limit stress level)
4- One aileron cycle represents positive and negative stress states in the lug
(representing an upwards then downwards deflection of the aileron)
5- The initial flaw is a single corner crack at lug hole of 0.05”
6- Loads act axially on the lug

Full validation and discussion of the above assumptions can be viewed in

Appendix H.

The AFGROW input spectrum file was entered as a cycle-by-cycle spectrum, as

opposed to a block spectrum analysis. This was deemed important by the author
in order to better approximate the actual aircraft loading spectrum, where stress
occurrences are random (and are not best reflected by the linear structure of a
block spectrum). A randomisation function within Excel was used to place each
occurrence in a random position relative to other occurrences. While this method
of spectrum creation is more complex, it leads to a smoother crack growth curve,
and a higher accuracy life prediction (as final unstable crack growth is more likely
to occur at critical loading cycle, rather than in the middle of a block loading
analysis). Full details of the AFGROW input file and parameters can be viewed
in Appendix H.

8.2 Initial AFGROW Output, Analysis, and Modifications

Initially the fatigue analysis was conducted on the optimised pivot lug geometry
that was determined from static stress analysis iterations discussed in Section
6.9. This geometry was sized through numerous iterations so that the static stress

63
reserve factors were close to the ideal value of 1.0. When this optimised geometry
was assessed for fatigue using AFGROW, the fatigue life and possible inspection
frequencies were unacceptably short (a fatigue life of only 1/10th that of the
expected life of the aircraft). As a result, a full redesign the hinges and a
significant increase the lug thickness was required.

Further, as the fatigue analysis showed that fatigue was undoubtedly the critical
factor for the Ae-16 hinges, the author chose to modify the configuration of the
hinges to attain a fail safe design. As such, instead of a 1-to-2 configuration of
wing lug to aileron lug design, the author redesigned a 3-to-2 layout (see Figure
26). This ensured that no single lug failure could lead to a loss of the hinge. This
is critical for the actuation hinges, where the loss of one actuation hinge would
lead to overstress and failure of the other actuation hinge.

This redesign required considerable time and effort, as the entire detailed stress
analysis had to be repeated based off the new geometry. The result however,
gave acceptable crack growth rates that could be inspected at reasonable
intervals as discussed in Section 8.4.

8.3 Final AFGROW Output and Analysis

The crack growth history for the aileron pivot lug is presented in Figure 29 below
(AFGOW output converted from imperial to metric units, and aileron deflection
cycles converted to flight hours). The C-Crack is the size of the crack in the
transverse direction (from the lug hole to free edge), the A-Crack direction refers
to the through thickness crack size of the same crack.

The critical crack size in the (width) direction was determined to be 9.3
based on maximum spectrum stress. This equates to a crack which extends 92%
of the width of one arm of the lug. The crack history in Figure 29 shows failure at
a larger crack length of 10.7 due to the random nature of when the high
loading cycle occurred. The analysis predicted failure at 1,731,812 aileron cycles,
equating to 20,545 flight hours. The planned service lifetime of the Ae-16 is
70,000 hours (45,000 GAG cycles).

64
 Aileron Lug Crack Growth - 0.05" Initial Single Edge Crack
14

12

10
Crack Length (mm)

C-Crack
8

6 A-Crack

4
2.0mm Crack
2 Length

0
0 5000 10000 15000 20000
Flight Hours

Figure 29 - Crack Growth History - Aileron Lug AFGROW Output

While the predicted lifetime of the aileron lug does not meet the required crack-
free lifetime requirement of the Ae-16, the analysis is based on the assumption
that there is an initial crack flaw that is 0.050”. This is the largest crack size that
does not meet the 90% probability of detection and 95% confidence of detection
limits as stated by ESDU 91207 [36]. As the majority of fatigue life is spent in the
crack initiation stages, any small reduction in the initial flaw size below 0.050” will
significantly increase the fatigue life. Regardless, the crack growth rate of the
redesigned hinges is low enough to enable a reasonable inspection regime to be
adopted (see Section 8.4 below).

8.4 Damage Tolerant Inspection Regime

The author followed industry accepted practice, and guidance from [37] and
ESDU 91207 [36], to design the aileron lugs to allow at least three opportunities
to detect a crack using the selected inspection technique. The author has
proposed two different inspection regimes for the Ae-16 aileron hinges using
Detailed Visual Inspection (DVIN) and a Bolt Hole Eddy Current (BHEC)
inspection techniques.

65
 Table 19 - Recommended Inspection Schedules - Aileron Hinges

DVIN BHEC
Threshold Inspection Thickness [36] 2mm 1.2mm
Equivalent Flight Hour Threshold 8,000 FH 0 FH
Recommended Inspection Frequency 4,000 FH 6,000 FH
Number of Intermediate Inspections Possible 4 3
prior to Critical Crack Length

For visual reference, the DVIN crack threshold limit of 2mm has been plotted on
Figure 29 (with possible inspections at 8,000FH, 12,000FH, 16,000FH, and
20,000FH before critical crack length is reached).

The inspection type and frequency can be tailored by the customer airline or
maintenance organisation to match their chosen maintenance frequencies. For
example, the organisation may choose to perform more DVIN inspections at
shorter intervals (every 4,000FH, or Intermediate Level Maintenance (ILM)
opportunity) or choose BHEC inspections at larger intervals (every 6,000FH, or
Depot Level Maintenance (DLM) opportunity). The trade-off factor for the airline
is that for BHEC inspections, the removal of the aileron and bushings is required.

8.5 Secondary Task – Aircraft Fatigue Spectrum Development

As part of the Group Design Project, the author undertook the task of generating
the Ae-16’s vertical acceleration spectrum for use by other GDP students to
conduct fatigue analyses. The analysis tasks were split between five GDP team
members, which enabled validation and cross checking of spectrum results. The
author focussed on development of the airborne fatigue spectrum for the Ae-16,
which will be discussed briefly here.

The process used by the author to determine the Ae-16 airborne fatigue spectrum
is detailed in Figure 30 below:

66
 Using ESDU 69023
Convert to vertical
Obtain Mission Mix and methodology, find
acceleration occurances
Operating Stages of cumulative occurances
to per 4000FH (TWIST
Selected Flight Types of accelerations per
spectrum Base)
flight

Use spectrum to create Adjust the acceleration

acceleration exceedance level magnitudes so that
plots and AGROW input occurance rates match
files. the TWIST spectrum

Figure 30 – Ae-16 Acceleration Spectrum Development Process

The mission mix and flight profiles was provided by the Aircraft Performance
Team. There were five mission profiles consisting of various flight profiles,
distances flown, and weights. The pertinent data is displayed below:

Table 20 – Ae-16 Mission Mix and Expected Usage

% of use No of hours in this No of flights per

Sortie Distance Length of sortie
assumed sortie configuration 70,000 life cycle
(nm) Hrs Hrs
1 500 25% 17500 1.5 11560
2 1000 40% 28000 2.8 10161
3 2000 10% 7000 5.2 1336
4 1500 20% 14000 4.0 3500
5 2500 5% 3500 6.5 540
Total 100% 70000 20.0 27097

As can be seen, the yellow column is assumed usage of each mission type. The
total number of hours relates to the 70,000 expected Life-Of-Type (LOT) of the
Ae-16 as prescribed by the project specification document [2].

TWIST (Transport WIng STandard) standard fatigue loading sequence has been
selected for simulating the magnitude and frequency of the applied service
loadings for the Ae-16. The TWIST loading sequence represents the stresses on
the lower wing skin at the wing root of a transport aircraft [38]. A TWIST spectrum
is comprised of 10 different acceleration amplitude levels and 10 different fight

67
types (each flight type equates to a certain severity of gust/manoeuvres
encountered). As the TWIST spectrum derivation is explained in detail within
ESDU 97018, it will not be repeated here.

In order to generate a TWIST spectrum, the author needed to determine the

acceleration amplitude levels specific to the Ae-16 which match the TWIST
occurrence frequencies.

ESDU 69023 [39] was the primary resource used by the author to derive the
vertical acceleration data for the Ae-16. This document compiles empirical
vertical acceleration data from several types of subsonic transport aircraft, and
provides a method to determine the vertical acceleration occurrences for any
aircraft (using geometric, weight, and aerodynamic variables as inputs). As the
methodology and derivation is detailed within ESDU 69023, it will not be repeated
in this report due to brevity requirements.

Using ESDU 69023, the author determined the occurrences (per flight) of various
amplitude accelerations between +2.5g and -1.0g. The amplitude levels were
then adjusted so that each acceleration level occurred at the same frequency as
that prescribed by the TWIST spectrum (for example, the highest amplitude
TWIST acceleration occurs once per 4000 flights. The Ae-16 acceleration
amplitude levels were adjusted so that exceedance to the highest amplitude level
also only occurred once per 4000 flights).

The final TWIST spectrum for the Ae-16, detailing all 10 amplitude acceleration
levels, and the occurrences per 4000 flights is shown in Table 21 below. The
TWIST exceedance plot can be seen also in Figure 31:

Table 21 - Final TWIST Spectrum for Ae-16

Amplitude Occurrence Per Upwards Downwards

Level 4000 Flights Acceleration Acceleration
i 1 1.66 0.35
ii 2 1.64 0.37
iii 5 1.63 0.38
iv 18 1.61 0.39
v 52 1.60 0.44
vi 152 1.55 0.49

68
 vii 800 1.48 0.57
viii 4170 1.37 0.66
ix 34800 1.28 0.76
x 358665 1.17 0.86

Figure 31- TWIST Exceedance Plot for Ae-16

The TWIST fatigue spectrum was converted to an AFGROW block file for
analysis of aircraft components where vertical acceleration is applicable (wing,
tail, engine pylon etc.). If a GDP student wished to analyse the TWIST spectrum
as a cycle-by-cycle analysis in AFGROW (higher accuracy, yet computationally
expensive), they could modify the spectrum similar to how the author conducted
his analysis in Section 8.1.

9 – Mass and Centre of Gravity Assessment

The final aileron mass and centre of gravity was assessed through both CATIA
modelling and validated through hand calculations. The final mass of the Ae-16
composite aileron is 36.65kg, with a longitudinal CoG at 25780.5mm aft of aircraft
nose reference plane. The composition of this mass can be seen in Figure 32:

69
 Figure 32 - Final Mass Breakdown

Compared to the mass calculated during initial sizing (using the same
methodology), the mass has reduced significantly, from 77.4kg down to 36.65kg.
This primarily highlights the significant efficiencies that can be gained by
assessing the structure under a distributed flange loading distribution. The upper
and lower skins are by far the highest contributors to overall mass, and their mass
is driven by the high torsional loading inherent with the high aspect ratio design.
By analysing the skin’s effectiveness and contribution in reacting bending and
shear loads, the mass of the remaining aileron structure has been significantly
reduced.

The mass of composite design compares favourably to the equivalent metallic

aileron design for the Ae-16 (assessed at 66.52kgs [40]). This is primarily due to
the tailoring advantages of composite materials (i.e. aligning ±45° plies in high
shear loaded components), and secondly due to the high bending stiffness
inherent with sandwich panels, allowing far fewer supporting webs to be required
compared to the metallic design.

70
10 – Finite Element Analysis
A Finite Element Analysis (FEA) was conducted on both a global model of the
aileron and a detailed model of a titanium hinge.

The primary purpose of the global model FEA was to assess the deflection and
torsional stiffness characteristics of the aileron, as this is a critical design factor
for control surfaces (which can only be assessed using simplified hand analysis
otherwise). An FEA model allows complex pressurisation loading to be applied to
the actual geometry of the aileron (including factors such as twist, taper, and
camber) which cannot be analysed with hand calculations. Further, the
interactions between different structural components can be assessed to an
accuracy not possible with hand calculations (for example, non-linear stress
distribution within panels supported by stiffening structure).

The detailed model of the titanium hinge serves as a validation of the stressing
calculations performed in Appendix F.4. Further, the FEA model was used to
tailor and optimise the weight saving pockets discussed in Section 6.9

While this section discusses the main methods and results of the FEA modelling,
the detailed methodology, model creation, and validation checks conducted by
the author can be viewed in Appendix I.

10.1 FEA Model Generation

The author conducted FEA modelling using ABAQUS software due to the relative
ease, tools, and capacity to accurately model composite layups. Further, as both
ABAQUS and CATIA are produced by Dassault Systèmes, the two programs
interact well together when transferring models. The author learned to use
ABAQUS software over the course of the GDP period, primarily learning from the
approved ABAQUS documentation and example analysis files [41].

The global FEA model created by the author was highly accurate in terms of
representing the actual aileron geometry and loading. The model geometry was
created in CATIA using the actual aileron surface profile (including taper, twist,
and camber). Aerodynamic loading was applied to the global model using a

71
complex polynomial function that accurately modelled both the span wise and
chordwise pressure distributions acting aft of the hinge line. A computationally
efficient model was achieved by representing the free edge flanges as beam
elements attached to the closed form surface model (free edges require overly
refined meshing to accurately model in-plane bending stresses).

Figure 33 - FEA Model Loading Distribution

Notice in Figure 33 above that the loading varies in both spanwise and chordwise
directions, as well as acting on the upper and lower surfaces. The yellow arrows
represent the aerodynamic loading forward of the spar web, applied directly to
the beam elements which model of the skin and spar flanges forward of that point.

Accurate laminae properties and stacking sequences (as defined in Section 6.7)
were applied to each individual component. The surface elements were modelled
as thick shell elements, whilst all flanges were modelled as beam elements using
the equivalent membrane stiffness properties calculated using COALA. An
example of the composite stacking plot used for the sandwich skins is shown in
Figure 34 below.

72
 Figure 34 - ABAQUS Ply Stack Plot for Ae-16 Sandwich Skins

The detailed FEA model of the critical hinge was also created in CATIA by
assessing only one half of the hinge (allowable due to symmetry), and omitting
the hinge flange structure, which is not a critically loaded component within the
structure. Contrary to the global model, the detailed hinge model incorporated
continuum elements instead of shell elements, and therefore required 3D
meshing controls to be used within the model.

Full detailed methodology, model creation, and validation checks conducted by

the author can be viewed in Appendix I.

10.2 Global Model FEA Results – Deflection Analysis

The results of the ABAQUS deflection modelling can be seen in Figure 35 below.
The image is a view looking upwards and outboard at the underside of the aileron.
Limit loading has been applied for the deflection analysis.

Several points of interest can be discussed pertaining to the deflection results

detailed in Figure 35. Firstly, the maximum deflection value of 7.38mm occurs
within the central portion of the largest sandwich panel. This was expected, as

73
 Figure 35 - FEA Deflection Result (x10 Scale)

the large panel size combined with the relatively high aerodynamic pressure at
this location (see Figure 33) leads to a larger deflection. This value compares to
the hand calculation value of 12.39mm (see Appendix F.2.10). The difference is
primarily due to the simplified uniform pressure distribution used within the hand
calculations (a requirement from using the methodology specified in Niu [25]).
The actual distribution used within the FEA model is roughly triangular in shape
in the chordwise direction, therefore the loading at the central portion of the panel
would be relatively lower. Regardless, the panel deflection in both forms of
analysis meets the deflection requirements prescribed by Niu of 0.5” (12.7mm)
[25].

It can be seen that the spar deflects very little in the vertical direction, with a value
of 0.702mm at the inboard most end. This correlates very closely to the hand
calculated value of 0.811mm derived in detailed stressing. This high accuracy
was expected due to the comprehensiveness of the detailed stressing method; it
includes the bending stiffness contributions from all relative components (spar
web and flange, skin flange, and sandwich skins).

74
The deflection of the trailing edge gets progressively larger as it nears the
outboard end of the aileron. This is expected, as only the inboard two hinges
reaction torsion (due to the actuators), while the outer structure is allowed to twist
under torsional loads. The outboard aft tip total deflection is 5.1mm which
equates to a twist at the outboard end of 0.78°. This is lower than the twist of
1.699° calculated during the aeroelastic analysis (see Appendix G.1). The reason
behind this difference is due to the simplifications inherent within the hand
calculations (such as using average sectional shear moduli, or the omission of
the additional torsional stiffness provided by the spar flanges). Further, Bredt and
Batho’s theorem forms the basis for the torsional stiffness hand calculations,
which in itself is a simplified approximation of the actual torsional stiffness.

10.3 Global Model FEA Results – Strain Analysis

Strain analysis was conducted using the same model as for the deflection
analysis, however the loading magnitudes were modified to that representing
ultimate loading. The author used the numerous tools available within ABAQUS
to generate the maximum strain envelope for the laminates shown in Figure 36
below.

Figure 36 - Ply Maximum Strain Envelope (Ultimate Loading)

75
To generate Figure 36 above, ABAQUS determines the maximum in plane strain
within each individual ply, and displays the maximum strain value, regardless of
which ply it occurs in. This enables a visual check to ensure the strain at any ply,
at any location within the structure, does not exceed a prescribed limit (3500 to
meet fatigue and damage tolerance requirements). ABAQUS can also identify
which individual ply that the maximum strain occurs in, detailed in Figure 37:

Figure 37 – Ply Maximum Strain Envelope - Ply Location

As can be seen from Figure 36 that the maximum ply strain occurs at the second
actuation hinge location, and at this location Figure 37 shows it is the outermost
ply PLY-1 which experiences this strain (outer +45° ply).

The location of maximum strain at the second actuation hinge location is

expected, as this spanwise location reacts the largest accumulated torque
(9142Nm Ultimate, see Figure 17) and a large vertical shear reaction force
(24,940N Ultimate, see Table 5).

Through detailed stressing and COALA analysis, the maximum strain is indeed
in the outermost +45° ply, however a value of 2293 strain was expected

76
compared to the 1630 returned in FEA (Figure 36). This FEA maximum strain
result is thus 28.9% smaller in magnitude than detailed stressing maximum strain.
The primary reason behind this is the conservative assumptions used in the
detailed stressing methodology. The contributions from the spar and skin flanges
forward for the spar web were omitted from torsional stiffness and shear reaction
shear flow calculations, which in turn increase the calculated loading within the
sandwich skin panels. Further, the hand calculations conservatively assumed the
entire panel was subject to the critical in-plane load, compared to the FEA model,
which allows a realistic distribution of the loading. By applying the critical loading
to the entire panel, there was no account for load shedding from areas of high
loading to areas of low loading. As the FEA model allows for this type of loading
distribution, the loads at the critical skin/web interface are lower. See Section 6
for a full list of assumptions that apply to the detailed stressing analysis that would
be fully accounted for in the FEA model.

One final observation from Figure 37 is that it is apparent that the maximum strain
values occur in either of the two outermost plies (+45° plies) within any section of
the structure (excluding the in-plane strain of the core material, which inherently
has very low in plane stiffness). This is expected, as the outermost plies
experience the highest fibre stresses in bending, combined with the in plane axial
and shearing forces.

10.4 Detailed Model FEA Results

The results plot of Mises stress values is depicted in Figure 38 below. The
maximum stress from FEA analysis of 344.7MPa compares to the stress value of
294.57MPa calculated in detailed stressing (Shear bearing rupture stress, see
Appendix F.4.6). The stress value calculated in detailed analysis is lower as it
does not take into account the twisting or loading offset resulting from split lugs.
Further, the hand calculations do not account for load distribution peaking near
to the edges of the lug. These effects are prevalent in the FEA model (see the
maximum stress occurs at the lug edge in Figure 38). The average stress across
the FEA upper lug position closely matches the 294.57MPa stress calculated
during detailed stressing. Regardless, the ultimate stress in both forms of analysis

77
is still far lower than the ultimate stress allowable values for the titanium material
( = 896 [42])

Figure 38 - Detailed FEA Stress Results – Hinge

11 – Ae-16 Composite Aileron Features

11.1 Trailing Edge Design
The trailing edge of an aileron is susceptible to both accidental impact damage
during maintenance or ground operations, and to damage from lightening strikes.
In order to increase the damage tolerance of the trailing edge structure, a titanium
sheet (2mm thick) encapsulates the composite edges of the sandwich panels as
they come together (see Figure 39 below). A phenolic/epoxy filler is used to
smooth the transition between the composite panels and the titanium edge in
order to maintain smooth airflow. The titanium trailing edge, combined with the
attached static wicks, aid in attaining lightning protection and static discharge
requirements of CS25.581 and CS25.899 respectively.

78
 Figure 39 - Titanium Trailing Edge Design

The upper and lower sandwich panels come together at the trailing edge as
depicted in Figure 40 below. As can be seen, the outermost laminate facing of
the panel follows the aerodynamic shape of the aileron, whilst the inner laminate
facing follows the contours of the core material. The sandwich panels come
together at the trailing edge at which point they are adhered between their
common surface profiles, and encapsulated by the trailing edge structure.

Figure 40 - Sandwich Panel Adhesion Surface at Trailing Edge (Upper Sandwich

Panel and Trailing Edge Hidden)

79
11.2 Lifting and Balance Points
To enable easier installation of the ailerons there are three titanium lifting points
spread across the upper surface of the Ae-16 composite aileron. These points
are aligned with the chordwise centre of mass axis to also enable control surface
balancing procedures to be accomplished using the same fixing points. The
points are adhered to the sandwich panels as depicted in Figure 41 below:

Figure 41 - Titanium Lifting Point (cut away sandwich view)

11.3 Electrostatic Discharge Points

CS25.899[1] requires that electrical bonding and protection against static
electricity must be accounted for in order to minimise accumulation of
electrostatic discharge. As such, the composite aileron design incorporates three
trailing edge static wicks attached directly to the titanium trailing edge. These
design features, combined with the composite conductive surface treatment
(discussed in Section 11.4) ensure a continuous conductive path from the
grounded hinges through to the static wicks (see Figure 42 below):

80
 Figure 42 - Trailing Edge Static Wicks

11.4 Lightening Protection System (LPS)

Carbon fibre composites are highly susceptible to lightening damage and a
lightening strike to an unprotected laminate can cause severe damage.

Traditionally, composites have been protected from lightening by co-curing with

conductive aluminium or copper materials. Aluminium is a potential LPS material
choice for the Ae-16 aileron, as it has inherently low weight, but the risk of
galvanic corrosion in contact with the carbon fibre laminates is a concern, and an
isolation ply of fiberglass adds weight. Copper eliminates the galvanic reaction
risk, but weighs at least twice as much as aluminium [43].

Configurations of LPSs have traditionally taken the form of flame sprays,

aluminium/copper foil, or aluminium/copper mesh. Foil protection systems ensure
a uniform surface conductivity is achieved, but this comes at the expense of high
weight and poor reparability characteristics [13]. Boeing, Embraer, Airbus, and
Bombardier currently use metallic mesh LPSs over the majority of the aircraft’s
composite structure [43]. The main benefits of a metallic mesh LPS application is
the high conductivity and high heat of vaporization necessary to handle massive
lightning strike current levels [43].

A recent advancement in LPS technology is the use of conductive surface

treatments. Conductive surface treatments have the benefits of a complete
uniform distribution of protection (similar to foils) however inherit significant

81
weight and compatibility advantages over current LPS types. As the Ae-16 is
designed for a 2035 operational date, this recent LPS technology is an ideal
choice.

The author has chosen the LORD® UltraConductive LPS Surface Coating for
application with the Ae-16 aileron [44]. This coating provides a 55% direct weight
savings as compared to copper expanded foils, and provides a shielding just as
effective as a solid aluminium LPS [43]. The coating application is suitable for
current manufacturing processes, and presents good reparability characteristics.
Currently going through the certification process with several OEMs, the system
is likely to be certified prior to commencement of Ae-16 manufacture.

The conductive surface treatment provides a continuous conductive path from

the trailing edge titanium protection forward through to the titanium fasteners and
hinge grounding pathways, thereby fully meeting the requirements of CS25.581.
Further, the continuity across the surface allows for static charge to flow
effectively to the static wick dissipaters, meeting CS25.899 requirements.

11.5 Integrated Vehicle Health Monitoring (IVHM)

The Ae-16 will incorporate advanced, real time, integrated vehicle health
monitoring systems in order to aid in the detection of damage or degradation
between maintenance checks. The author worked extensively with the IVHM
team to find the ideal IVHM solutions to apply to the composite aileron.

Firstly, Comparative Vacuum Monitoring (CVM) sensors will be applied to the

outer surface of the critical pivot lug locations. CVM sensors can detect the
presence of an underlying crack by assessing air pressure differentials between
fine air galleries within the sensor. Any formation of a crack will alert the
maintenance team upon post-flight checks, and replacement of the lug can be
scheduled at the next opportune maintenance period.

A total of three Acoustic Emission (AE) sensors will also be attached along the
length of the composite spar web. This passive NDT method monitors for the

82
natural elastic wave properties of a structure, and detects if unusual deformation
or response of the structure is occurring [45]. The main concern for the composite
aileron, is delamination of the adhesive surfaces between the spar flanges, skin,
and rib flange structures. The AE IVHM method will allow for early detection of
this form of failure mechanism and allow for maintenance to occur prior to
significant structural degradation occurring.

12 – Manufacture and Assembly

As discussed previously, the design philosophy for the composite aileron design
was to prioritise environmental considerations and sustainability in all facets of
design. This flows through to the choice of method of manufacture and assembly,
where low energy manufacturing processes and low waste methods are selected.

Firstly, a Laser Ply Outline Projection (LPOP) tool is proposed for the cutting and
placement of all CFRP plies. This method projects a laser outline of the required
laminae shape onto the raw pre-preg material. The technician then cuts along the
laser profile to obtain a perfect shaped ply for the application. LPOP methodology
ensures minimal wastage of raw material, and significantly reduces the layup time
for complex parts [19]. Further, as the aileron laminate profiles are relatively
complex (due to the taper, twist, and honeycomb inserts) the method will
eliminate many sources of errors. The LPOP method can also be used to identify
the exact placement locations of the honeycomb cores.

Hand-layup methodology is proposed due to the reduced manufacturing energy

cost, and due to the global homogeneity of stacking sequences (the sequence
does not change between inboard and outboard ends). The hand-layup method
also significantly reduces cost compared to automatic tape laying processes.

The M56 epoxy material was chosen as it can be manufactured using an out-of-
autoclave curing process. Autoclave curing, while able to produce a consistent
laminate structure with minimal inclusions, is a high energy manufacturing
process due to the pressurisation requirements. By using an oven cure process,
the energy input per cured laminate will be significantly lower. Secondly, there is

83
generally fewer restrictions on the size/number of components that can fit within
an oven (compared to an autoclave); therefore more samples could be cured
within the same cycle (further reducing energy input per unit cured). Standard
oven curing methodology will be used (vacuum bag, bleed schedule, OEM
recommended cure cycle etc.). The moulding tool will be made from cured
composite material, matching the outer surface profiles of the Ae-16 aileron. A
composite mould is preferable, as the co-efficient of thermal expansion will match
that of the child laminate during curing (as recommended by [19]).

It is proposed that each aileron is comprised of eleven individually cured sections

(spar, nose round, 2 x skins, 5 x hinge ribs, 2 x end cap ribs). These structural
components will be adhered to each other post-cure using Hexcel Redux® 810
Adhesive. This adherent has been chosen due to its excellent compatibility with
the M56 epoxy, and its resistance to environmental factors [48]. A detailed
strength analysis of the adhered surfaces is conducted in Appendix F.3.1. The
critical adhesion interface between the spar and skin is segmented by the
mechanical fastening present at the hinge locations, ensuring no single adhesion
failure can extend the entire length of the aileron (meeting CS25.571 [1]
requirements).

As for the hinges, their profiles will be milled from solid block titanium. Computer
Numerically Controlled (CNC) milling will be utilised to attain the complex profiles
of the hinges, and to ensure a consistency is maintained between parts. As
discussed in Section 6.9, the weight saving pockets in each hinge are designed
to enable milling from the outer surface of each hinge.

84
13 – Conclusion
This thesis summarises the work conducted to design the composite aileron for
the Ae-16. The primary design objective for the aircraft was to ensure that
environmental awareness and sustainability factors were prioritised in all aspects
of design. The composite aileron designed by the author has fully met this
objective, incorporating low weight materials that are able to be manufactured
using low energy methods.

Airworthiness requirements were first assessed in order to generate the

acceleration velocity envelope for the Ae-16. This enabled loads acting on the
wing and aileron structure to the ascertained.

A brief literature review enabled the identification of existing aileron designs and
optimal structural layouts . The structural layout deemed to best meet the design
objectives was determined to be a single spar, sandwich panel design that
incorporated minimal ribs. The sandwich panels would be designed under a
distributed flange loading distribution model and would be fully effective in
reacting bending loads.

Initial sizing was conducted using simplified analysis techniques in order to form
a starting point for the iteration process. During detailed analysis, a thorough and
complex iteration process was used by the author to determine the optimum
laminate layup for all main structural components. Further, the optimum
configuration and sizing of the hinge components was also ascertained.

A fatigue and damage tolerance assessment was conducted on the critical

titanium hinge, leading to a full redesign of the hinge to meet fatigue life and
inspection frequency requirements.

A mass and centre of gravity assessment was conducted on the final design to
highlight the success of the iteration/optimisation process conducted by the
author. The mass is significantly lower than the equivalent metallic design due to
the fully effective sandwich skin panels and the minimisation of supporting ribs.

85
Finite Element Analysis was conducted on a global surface model and detailed
hinge model of the aileron. Highly accurate aerodynamic loading profiles were
applied to the actual aileron geometry to result in an FEA model that best
approximates the real loading and structural response. Numerous validation
checks were performed to ensure model accuracy. The results from FEA varied
little from the values expected from hand calculation methodologies.

Overall, the composite aileron designed by the author fully meets the primary
design objectives. The design is highly optimised and exhibits minimal weight,
which leads to a significant reduction in fuel penalties and wastage. The choice
of out-of-autoclave pre-preg laminae enables a low energy manufacturing
method to be used, thereby reducing the environmental impact. The numerous
detailed features incorporated by the author will enable an easy transition to the
final detailed design stages, should the design be progressed through to
manufacture.

14 – Recommendations

The analysis conducted within this thesis is sufficient for an initial pass of the
detailed design stage for the composite aileron design, however, additional
work is required prior to final acceptance of the Ae-16 composite aileron. The
author recommends the following analyses are performed:

1. Conduct a loading assessment utilising spoiler augmentation as a means

to assist with aircraft roll. This will significantly reduce the loads imparted
on the aileron structure, and will allow aileron component size to be
reduced

2. While the Ae-16 aileron deflection limits were specified in the project
specification [2], the resulting roll performance far exceeded the
regulatory requirements. Reduced maximum deflection limits should be
investigated to reduce loading on the aileron/wing structure.

86
3. Conduct Computational Fluid Dynamics (CFD) analysis of aileron
deflection performance to validate the calculations conducted during
preliminary design.

4. Carry out aileron flutter analysis using CFD methodology.

5. Perform additional FEA modelling of the hinge to spar mechanical

fasteners to investigate the load sharing distribution between the pins.
Further, this will enable accurate assessment of the bearing stress
characteristics of the spar and skin laminates.

6. Conduct detailed stress analysis on sections of the aileron that have

been repaired (e.g. scarf or bolted repairs to composite structure). This
will confirm that the aileron can be repaired after minor damage while
maintaining acceptable strength and durability properties.

87
REFERENCES
1. EASA. ‘Certification Specifications and Acceptable Means of Compliance for Large
Aeroplanes - CS-25’. 2012; (Amdt. 18).

2. Smith H. ‘Hybrid-Electric Airliner Ae-16 Project Specification’. DES 1650. (unpublished);

2016.

3. Transport D for. UK Aviation Forecasts. 2013; (January). Available at:

https://www.gov.uk/

4. EASA. ‘Certification Specifications and Acceptable Means of Compliance for Normal,

Utility, Aerobatic, and Commuter Category Aeroplanes - CS 23’. 2015; (Amdt. 13).

5. Smith H. ‘Loading Actions, Asymmetric Flight – Derivation of Loads’, Part of the

Cranfield University MSc(AVD) course lectures on Loading Actions. DAeT 9586. 2016.

6. IHS ESDU. Controls 04.01.02: ‘Rate of change of hinge-moment coefficient with control
deflection for a plain control in incompressible two-dimensional flow’. Amdt. B. 2003.
Available at: www.esdu.com.

7. IHS ESDU. Controls 01.01.03 ‘Rate of change of lift coefficient with control deflection in
incompressible two-dimensional flow’. Amdt. B. 1978. Available at: www.esdu.com.

8. IHS ESDU. Wings 01.01.05 ’Slope of lift curve for two-dimensional flow". Amdt. E. 2007.
Available at: www.esdu.com.

9. Howe D. Aircraft Loading and Structural Layout. United Kingdom: Professional

Engineering Publishing; 2004.

10. James AM., Vaughn RL. Design of an advanced composites aileron for commercial
aircraft. Composites. 1976; 7(2). Available at: DOI:10.1016/0010-4361(76)90016-1

11. Scott ML., Raju JAS., Cheung AKH. Design and manufacture of a post-buckling co-
cured composite aileron. Composites Science and Technology. 1998; 58(2). Available
at: DOI:10.1016/S0266-3538(97)00116-4

12. Romano F., Fiori J., Mercurio U. Structural design and test capability of a CFRP aileron.
Composite Structures. 2009; 88(3). Available at: DOI:10.1016/j.compstruct.2008.04.010

13. Niu MC-Y. Airframe Structural Design. Burbank, California: Conmillit Press Ltd.; 1988.

14. Bradley S. Processing Variables of Carbon Fibre Epoxy Pre-Preg Laminates. University
of Auckland; 2009.

15. Hexcel. Hexcel Hexply Selector Guide - Aeropace. 2017. Available at:

88
 http://www.hexcel.com/user_area/content_media/raw/Aerospace_SelectorGuide(1).pdf

16. Hexcel - Sustainability Review. Available at:

http://www.hexcel.com/About/Sustainability/Environmental-Commitment/ (Accessed: 5
January 2017)

17. Hexcel. Hexcel Product Data Sheet - Hexply M56. 2017. Available at:
http://www.hexcel.com/Products/Prepregs-and-Resins/HexPly-Prepregs

18. Hexcel. Hexcel Product Data Sheet - Hexply 8552. Available at:
http://www.hexcel.com/Products/Prepregs-and-Resins/HexPly-Prepregs

19. Mills A. Materials and Manufacturing Technology for Composite Aircraft Components.
AVD Lecture - Cranfield University; 2017.

20. Hexcel. Hexcel Hexweb Selector Guide - Aerospace. 2017. Available at:
http://www.hexcel.com/Products/Honeycomb/HexWeb-Honeycomb

21. Hexcel. Hexcel Hexweb HRH-10 Data Sheet. 2017. Available at:
http://www.hexcel.com/Resources/DataSheets/Honeycomb

22. Hart-Smith IJ. The Ten Percent Rule. Aerospace Materials. Vol. 5(No. 2).

23. IHS ESDU. ESDU 82013 - Laminate stacking sequences for special orthotropy.

24. Guo S. AVD 0428 - MACROMECHANICS OF FRP SECTIONS. AVD Lecture - Cranfield
University;

25. Niu MC-Y. Airframe Stress Analysis and Sizing. Second Ed. Hong Kong: Conmillit Press
Ltd.; 1999.

26. Phillips BJ. Composite Wing Structure Design with Practical Constraints. Visiting Lecture
- Cranfield University; 2017.

27. AVD Lecture - Buckling of Composite Sections. AVD Lecture - Cranfield University;
2017.

28. Liu W. Fatigue and Damage Tolerance AVD Notes. AVD Lecture - Cranfield University;
2017.

29. IHS ESDU. ESDU 80023 - BUCKLING OF RECTANGULAR SPECIALLY

ORTHOTROPIC PLATES. 2017. Available at: https://www.esdu.com

30. IHS ESDU. ESDU 03013 Thickness selection for the flanges and web of a composite I-
section beam subjected to bending and shear. 2017. Available at: https://www.esdu.com/

89
31. Giannopoulos I. Major component design & structural layout Synthesis Procedure –
Initial sizing of members. AVD Lecture - Cranfield University; 2017.

32. BCAR. British Civil Airworthiness Requirements (BCAR), Subsection K3-9. London:
BCAR;

33. Guo S. AVD 0441 - Aeroelastcity Lecture Notes. AVD Lecture - Cranfield University;
2017.

34. A.W.Babister. Aircraft Stability and Control. London: Pergamon Press; 1961.

35. Control Surface Usage. A.R.B White Note. No.96(Issue 1).

36. IHS ESDU. ESDU 91027 - NON-DESTRUCTIVE EXAMINATION – CHOICE OF

METHODS. 2017. Available at: https://www.esdu.com/

37. Irving P. AVD Lecture - Inspection Thresholds and Inspection Considerations. AVD
Lecture - Cranfield University; 2017.

38. IHS ESDU. ESDU 97018 - Standard fatigue loading sequences. 2003. Available at:
https://www.esdu.com/

39. IHS ESDU. ESDU 69023 - Average gust frequencies Subsonic transport aircraft. 1989.
Available at: https://www.esdu.com/

40. Joyce C. AVD Structures Submeeting - 16 March 2017 - Metallic Aileron Update. 2017.

41. Abaqus 6.14 Documentation Collection. Dassault Systèmes; 2017.

42. Metallic Material Properties and Development Standards - MMPDS-11.

43. BLACK S. Lightning strike protection strategies for composite aircraft. Composites
World. Available at: http://www.compositesworld.com/articles/lightning-strike-protection-
strategies-for-composite-aircraft

44. LORD. LORD® UltraConductive Film and Coatings for Lightning Strike Protection. 2017.
Available at:
http://www.lord.com/sites/default/files/PB6071_UltraConductiveFilmandCoatings.pdf

45. Harris HLDDO. Continuous monitoring of fatigue-crack growth by acoustic-emission

techniques. Experimental Mechanics. 1974; 14(2): pp 71–81.

46. DESIGN OF COMPOSITE MATERIAL STRUCTURES FOR BUCKLING - TECHNICAL

REPORT AFWAL-TR-81-3102. LOCKHEED MISSILES & SPACE COMPANY, INC.;
1980. Available at: http://www.dtic.mil/dtic/tr/fulltext/u2/a112218.pdf

90
47. Giannopoulos I. AVD Detailed Stressing Lecture 10.2 - Simple Rib Analysis Notes. AVD
Lecture - Cranfield University; 2017.

48. Hexcel. Redux 810 Data Sheet. 2017; Available at:

http://www.hexcel.com/Products/Adhesives/

49. High Sear Corporation. HI-LOK Pin - HL11 Data Sheet. 2017;

50. MIL-HDBK-5J, DEPARTMENT OF DEFENSE HANDBOOK: METALLIC MATERIALS

AND ELEMENTS FOR AEROSPACE VEHICLE STRUCTURES (31 JAN 2003). 2017.

51. IHS ESDU. ESDU 91008 - Strength of lugs under axial load. 2011; Available at:
https://www.esdu.com/

52. RBC Aerospace Bearings. M81935/4 SELF-LUBRICATED ROD END BEARINGS DATA
SHEET. 2017; Available at: http://www.rbcbearings.com/

53. Cranfield University AVD 9632 - Cranfield University Stressing Data Sheets.

54. MS14102 SELF-LUBRICATED SPHERICAL BEARINGS DATA SHEET. RBC

Aerospace Bearings; 2017. Available at: http://www.rbcbearings.com/

55. IHS ESDU. ESDU 81006 - Stress concentration factors axially loaded lugs with
clearance-fit pins. 1982; Available at: https://www.esdu.com/

56. IHS ESDU. ESDU 06021 - Strength of lugs under transverse load. 2006; Available at:
https://www.esdu.com/

57. IHS ESDU. ESDU 08007 - Strength of lugs under oblique load. 2009; Available at:
https://www.esdu.com/

58. IHS ESDU. ESDU STRUCT 01.06.01 - Form factors for circular sections under
combined bending and axial load. Available at: https://www.esdu.com/

91
APPENDICES

Appendix A - Compliance Matrix

This compliance matrix is primarily focussed on those regulatory requirements that apply directly to the design of the composite
ailerons. Regulatory requirements that apply to the aircraft as a whole have been included in the main body of this thesis (for example
the requirements used to derive the operating envelope is detailed in Section 2.1, Table 2). Table A-1 below details the Means of
Compliance (MOC) coding and compliance status for reference with the full compliance matrix.

Table A-1 - Compliance MOC Codes and Compliance Status Levels

Means of Compliance
MOC Code Compliance MOC Code Compliance
0 Definition 5 Aircraft Ground Testing
1 Report/Drawing 6 Aircraft Flight Test
2 Analysis/Calculations 7 Inspection/Survey
3 Safety Assessment 8 Simulation
4 Lab or Rig Testing 9 Equipment Qualification and Procurement
Compliance Status
C Compliant PC Partially Compliant NC Not Compliant

A-1
 Table A-2 - Compliance Matrix - Ae-16 Composite Aileron

Compliance
Compliance Compliance Compliance
Description Method Means of Compliance
Reference Status Reference
Code
Flight - CS25 Regulatory Requirements
CS25.25 Maximum weight and centre of gravity limits must
Weight and CoG remains within
CS25.27 be established and remain within aircraft C 1,2 Section 9
specification limits
CS25.29 specification limits
Analytical assessment confirms rolling
Roll performance must meet the various
performance requirements are met.
CS25.147 requirements specified within CS25.147 and PC 2,4,6,8 Appendix B.1.5
Further simulation and flight tests
AMC25.147
required.
Torsional stiffness has been proved to be
Aircraft must demonstrate in flight to be free from
within required limits. Further analysis Section 7
CS25.251 undue vibration and buffeting that would adversely PC 2,4,5,6,8
and testing against other instabilities Appendix G
impact the flight
must conducted
Design loads from aileron deflection must be
Loading has been based on CS25.349
CS25.349 assessed at various critical aileron deflections at C 2 Appendix B.1.9
requirements.
various design speeds
Structure - CS25 Regulatory Requirements
A factor of safety of 1.5 must be applied to external
CS25.303 C 2 Factor of safety of 1.5 applied in analysis Appendix F
limit loads
The structure must support limit loads without
permanent or detrimental deformation. The Structure has been analysed and designed
CS25.305 C 2 Appendix F
structure must withstand at least 3 seconds of against limit and ultimate loading.
ultimate loading without failure.
Partially met through stressing and
Strength and deformation requirements must be deformation analysis. There is an
CS25.307 PC 2,4 Appendix F
met at each critical loading condition additional requirement for static/dynamic
testing for complete verification.

A-2
 Control surface hinges must be designed to Hinges have been designed to react loads
CS25.393 C 2 Appendix F.4
withstand loads parallel to the hinge line parallel to the hinge line.
Flight control systems must been designed to
Further analysis and ground testing is
CS25.415 withstand limit loading from horizontal gusts while PC 4,5 N/A
required
parked or during taxi.
AFGROW crack growth analysis
The structure must be proven to preclude
conducted on critical hinge. Composite
catastrophic failure due to fatigue, accidental Section 8
CS25.571 PC 2,4,8 structure analysed using a knock-down
damage, or corrosion throughout the operational life Appendix H
failure strain. Dynamic testing required
of the aircraft.
for full compliance
Lightening protection applied within
The aircraft must be protected from the catastrophic
CS25.581 PC 1,2,4 composite structure. Validation testing Section 11.4
effects from lightening (see also CS25.899)
required for full compliance
Design and Construction - CS25 Regulatory Requirements
Materials used in construction must conform to
All materials and fabrication methods
CS25.603 approved specifications. Fabrication must occur
conform to approved specifications and
CS25.605 under an approved process specification. Removable C 1,2,3 Section 6
processes. Fastener locking criteria has
CS25.607 fasteners must incorporate two separate locking
been met for removable fasteners.
devices.
Structure must be suitable protected against
CS25.609 C 1 Drainage system incorporated into design Section 11
degradation caused by environmental factors
The normally replaceable components of
the Ae-16 aileron are the hinges. These
Accessibility must be allowed for inspection,
are readily accessible and easily removed
CS25.611 replacement of parts normally requiring C 1,2,9 Section 6.9
due to mechanical fastening system used.
replacement, adjustment and lubrication.
Hinges are easily inspected using a visual
technique
Material strength properties must be based on
Further testing is required to verify the
CS25.613 adequate testing to establish design values on a PC 2,4,5,6 Section 4
strength of chosen laminated materials
statistical basis
An adequate bearing factor must be applied to
Bearing factor applied in detailed
CS25.623 clearance fit parts that are subject to vibration or C 2 Appendix F.4
stressing
pounding.

A-3
 For each fitting whose strength is not proven by limit
Fitting factor of 1.15 applied in detailed
CS25.625 and ultimate load test, a fitting factor of 1.15 must C 2 Appendix F
stressing
be applied.
Torsional stiffness and divergence has
The aircraft must be designed to be free from
been proved to be within allowable limits. Section 7
CS25.629 aeroelastic instability for all configurations and PC 2,4,8
Further analysis and testing against flutter Appendix G
design conditions
and other instabilities must conducted
Structure has been analysed and designed
Limit load testing of the control surface must be
CS25.651 PC 4 against limit and ultimate loading. Proof Appendix F
conducted
testing must be conducted
Control surface hinge components (including Bearings selected with higher allowable
CS25.657 C 2 Appendix F.4
bearings) must not exceed the approved ratings. proof loads than required
The control system must have stops that limit the
range of motion of the aerodynamic surface. The The linear actuators include stops and
CS25.675
control system must prevent damage from control C 1,2,5 static stiffness characteristics in order to N/A
CS25.679
surface movement due to gusts while the aircraft is meet this requirement
on the ground.
Static wicks have been incorporated into
Electrical bonding and protection against static
the design. Continuous electrical path
CS25.899 electricity must be designed to minimise PC 1,4,6 Section 11.3
through hinges, fasteners, metallic mesh,
accumulation of electrostatic discharge
and static wicks

A-4
Appendix B - Loading Calculations
B.1 Aileron Loads
This appendix details the aileron/wing loading and rolling performance analysis
conducted by the author.

B.1.1 Derivation of rolling equation

The derivation of the rolling equation of motion is documented in Smith [5] and
Howe [9] and summarised here:

The general rolling moment equation is as follows:

+ + + + − + =0 eqn.(A-1)

By expressing these terms in non-dimensional form, and through rearranging:

2 2 eqn.(A-2)
. + + − + = −( + )

One simplifying assumption is that the product of inertia ( ) is small when the
aircraft is at low angles of attack; thus, the effect on the aircraft can be negated.
Further, by assuming that the aircraft is kept in straight flight by rudder
application, the sideslip velocity and rate of yaw ( and ) can also be removed
from the equation. Finally, as the effect of the rolling moment due to lateral
motivator deflection ( ) is relatively small, it can be removed to give the simplified
roll equation:

2 eqn.(A-3)
− −=

+ = eqn.(A-4)

Where:

=− and, =

B-1
B.1.2 Load due to aileron deflection
The maximum distributed loading due to aileron deflections was determined for
all load cases. The first step, was to establish the spanwise acting point of the
aileron force on the wing. This point can be represented as the spanwise point at
which the lift force to either side is equal. By integrating the unitised spanwise
load distribution (provided in [2]) across the wingspan, and equated half of this
value to the integral between the root and the aileron acting point ( ), the point
can be found algebraically as follows:

1 ( ) ( ) ( ) ( ) eqn.(A-5)
=
2

For the Ae-16, this median point of aileron loading exists as = 15.53

The dimensional rolling moment due to aileron deflection can be expressed as:

1 eqn.(A-6)
=
2

By dividing this moment by the spanwise acting point ( ), the load due to aileron
deflection ( ) can be ascertained:

1 eqn.(A-7)
=2

These aileron loads are tabulated for all loading cases in Appendix C.1.

B.1.3 Roll rate

The maximum angular roll velocity was determined for all load cases. For the
assumed case of instantaneous aileron deflection, eqn.(A-4) becomes:

eqn.(A-8), [5]
= = (1 − )

will occur when a steady roll rate has been achieved as → ∞, also:

As → ∞, →0 therefore:

B-2
 eqn.(A-9)
=

These maximum roll rates are tabulated for all loading cases in Appendix C.1.

B.1.4 Roll Dampening Load

The distribution of the aerodynamic roll dampening load can be approximated by
use of the simplified Shrenk equation as expressed in Smith [5]

eqn.(A-10)
( ) 2 2
( ) ( )= + 1−
2

Unitising the Schrenk equation gives the unit distribution equation for the roll
dampening force:

eqn.(A-11)
( ) ( ) ( ) 2 2
= + 1−
2

The spanwise acting point of the roll dampening force on the wing can be
represented as the point at which the lift force to either side is equal. By
integrating the unitised spanwise load distribution across the wingspan, and
equated half of this value to the integral between the root and the roll dampening
acting point ( ), the point can be found algebraically as follows:

1 ( ) ( ) ( ) ( ) eqn.(A-12)
=
2

For the Ae-16, this median point of roll dampening load equates to = 11.26

The dimensional rolling moment due to roll dampening can be expressed as:

1 eqn.(A-13), [5]
=
2

By dividing this moment by the spanwise acting point ( ), the load due to roll
dampening ( ) can be ascertained:

B-3
 1 eqn.(A-14)
=2

These roll dampening loads are tabulated for all loading cases in Appendix C.1.

B.1.5 Roll performance

The performance of the ailerons to meet roll requirements as required by CS25
[1] was also proven. To calculate the roll angle ( ) the aircraft will complete within
time (t), the following integration of eqn.(A-9) can be performed

eqn.(A-15)
= (1 − ) = +

1 eqn.(A-16)
= + −

This can also be rearranged to find the required aileron deflection to meet the
CS25 performance targets:

eqn.(A-17)
=
( + − 1)

B.1.6 Asymmetrical gust loads and performance

The aircraft acceleration due to gusts was determined as part of the generation
of the nV diagram through the author’s Matlab algorithm. The sharp edged,
alleviated gust methodology as required by CS23 [4] was used as per below:

eqn.(A-18)
=1±
2 /

. /
Where: = .
= ( )

This acceleration (n) was then multiplied by half the weight of the aircraft to derive
the force on each wing. This force acts at the spanwise centre of lift, which is

B-4
found by integration of the unitised spanwise load distribution for wing lift and
algebraically solving for :

1 ( ) ( ) ( ) ( ) eqn.(A-19)
=
2

For the AE-16, this median point of wing lift acts at = 7.42

By assuming an 20% strength differential of gust force between the wings (as
required by CS25[1]) the differential of force generates a rolling moment ( ) and
roll acceleration ( ):

∆ eqn.(A-20)
= 0.2 × ×
2

= / eqn.(A-21)

The aileron deflection required to counteract this rolling moment is calculated

using equation A1-15:

1 eqn.(A-22)
=2

Providing that the required aileron deflection is less than the maximum design
deflection (20 for the Ae-16), then steady level flight can be maintained.

B.1.7 Shear Force, Bending Moment, and Torque Diagrams

In order to generate the SF, BM, and T diagrams for the wing, the asymmetrical
loads were combined with symmetrical loads as required by CS25[1].

The symmetrical contributions to the loading diagrams consisted of: the wing lift
from incidence, and the normal inertial relief forces:

The wing lift due to incidence per unit span was derived from the unitised lift
distribution provided in the project specification [2] as per the following equation
(essentially isolating the local lift force):

B-5
 eqn.(A-23)
( ) ( ) 1
= × ×( )
Δ 1 2
2

Where is the positive manoeuvring factor multiplied by two thirds as required

by CS25[1].

The normal inertia load per unit span was found by multiplying the mass per unit
span (derived from the project specification [2]) by :

eqn.(A-24)
=
Δ Δ

As for asymmetrical contributions, these consisted of the aileron lift, the rolling
inertial relief forces, and the aerodynamic roll dampening forces:

The aileron lift per unit span was calculated by multiplying the unitised spanwise
load distribution of aileron deflection (provided in the project specification [2]) by
the distributed aileron load derived in eqn.(A-7):

eqn.(A-25)
( ) ( )
= ×
Δ 1
2

The roll inertia load per unit span was found by multiplying the mass per unit span
(derived from the project specification[2]) by the spanwise position and the
angular acceleration :

eqn.(A-26)
= × ×
∆ Δ

The roll dampening load per unit span was calculated by multiplying the unitised
spanwise load distribution of roll dampening load (calculated previously in section
eqn.(A-11)eqn.(A-14)) by the distributed roll dampening load derived in
eqn.(A-14):

eqn.(A-27)
( ) ( )
= ×
Δ 1
2

B-6
The spanwise lift an inertial forces, from both symmetric and asymmetric
accelerations, were applied across the span of the wing to generate the spanwise
shear force and bending moment diagrams presented in Appendix C.

The torque diagram was generated by deriving the chordwise loading due to
aileron deflection and roll acceleration, and calculating the distance at which
these forces act from the shear centre of the wing box.

The shear centre of the wing box was calculated as per Howe [9], eq15.1, pg 427:

eqn.(A-28)
ℎ = = 0.4069
ℎ
1+
ℎ

Where from aerofoil geometry provided in the project specification [2]:

= 0.5c, ℎ =0.081718c, ℎ = 0.086379

The chordwise centre of gravity of the structure and fuel was assumed to act at
the centroid of the aerofoil section. This was calculated by the following equation:

∑ eqn.(A-29)
= = 0.457
∑ ×

The chordwise centre of pressure due to aileron deflection was found by

integration of the pressure distribution for aileron deflection (in isolation) as
detailed in eqn.(A-38)&eqn.(A-39). Integration allows for the median point of lift
distribution to be determined, and represents the chordwise CoP for aileron
deflection (assumed constant across the spanwise region influenced by the
aileron).

1 eqn.(A-30)
+ + = + +
2 ̅

Where the , , coeffecients are detailed later in

eqn.(A-35)eqn.(A-36)eqn.(A-37).

Integration yielded the chordwise CoP for aileron deflection as ̅ = 0377

B-7
Spanwise torque per unit span was calculated by multiplying the force/inertia by
the distance that it acts from the shear centre (at the local chord):

eqn.(A-31)
= + ( ( ) − ℎ ( ))
Δ ∆ Δ

This torque was applied across the span of the wing to generate the spanwise
torque diagram from asymmetrical manoeuvres, as can be seen in Ch. 2.5.3.

B.1.8 Chordwise Aerodynamic Pressure Distribution

The chordwise pressure distribution due to aileron deflection and angle of wing
incidence must be determined in order to determine the percentage of chordwise
load that acts aft of the hinge line and on the aileron.

Firstly, the chordwise pressure distribution due to aileron deflection is affected by

the ratio of the hinge moment coefficient due to deflection of the aileron  b20 and
the lift curve slope due to deflection of the aileron  a20 . In order to find these
two co-efficients, several ESDU sheets ([6–8]) were used. The inputs and output
parameters used within the ESDU sheets are detailed below:

Table B-1 - ESDU Wings 01.01.05 Input and Output Variables

ESDU Wings 01.01.05 [8]

Parameter: Value/Units: Notes:
175 / EAS
1.225 /
Median point of aileron lift
y 0.8108η
load (discreet load point)
c(y) = c(0.8108η ) 1.773 Chord at y
μ 1.789×10−5 · ( . / ) Dynamic viscosity at Sea Level
= ( ℎ )/( ) 21.246 × 10 Local Reynolds Number
Log10Re 7.327
= (0.0245 + 0.0036) 0.028147 Thickness of aerofoil at 90%(c)
= (0.0074 + 0.0004) 0.00776 Thickness of aerofoil at 99%(c)
= tan (( − )/0.09) 0.222762842 rad Trailing edge angle
tan(τa/2) 0.118
Figure 1 gives for leading-edge
 a10/a10T 0.882
transition

B-8
 Figure 2 gives for mid-chord
 a10/a10T 0.896
transition
= ((0.5-0.3)/0.5)*(0.896-
 a10/a10T Linear interpolation
0.882)+0.882 = 0.8876
From Figure 3, with
a10T 6.77 = 0.22276 = 12.76
and t/c = 0.1

Slope of lift coefficient curve

 a10 6.01
with incidence for 2D flow

Table B-2 - ESDU Controls 01.01.03 Input and Output Variables

ESDU Controls 01.01.03 [7]

Parameter: Value/Units: Notes:
 a10/a10T 0.8876 From ESDU Wings 01.01.05
From Figure 1 with t/c = 0.1
a20T 4.45
and c f/c = 0.3
From Figure 2 with
 a20/a20T 0.84  a10/a10T = 0.8876 and
c f/c = 0.3
Lift curve slope due to
 a20 3.738
deflection of the aileron

Table B-3 - ESDU Controls 04.01.02 Input and Output Variables

ESDU Controls 04.01.02 [6]

Parameter: Value/Units: Notes:
From Figure 1 with t/c = 0.1
b20T 0.885
and c f/c = 0.3
From Figure 2 with
 b20/b20T 0.86  a20/a20T = 0.84 and
c f/c = 0.3
Hinge moment coefficient due
 b20 -0.7611
to deflection of the aileron

The ratio of =− can now be used to determine the pressure distribution due

to aileron deflection by using the equations presented in Howe [9], ch 9.4.4.3, pp

296-297. The inputs to the distribution equation are as follows:

0.7611 eqn.(A-32)
= − = = 0.2036
3.738

B-9
 eqn.(A-33)
= = 0.3

From Howe[9] p298,table 9.1 = Centre of pressure forward of hinge line:

0.405 − 0.72 eqn.(A-34)

= = 0.31638
1 − 0.9

From Howe [9] p297 eq 9.20b:

1− eqn.(A-35)
2(1 − 3 ) −
=6 − 2 = 1.894
(1 − )

2 1− eqn.(A-36)
2(1 − 3 )
36 3 − 4 −
(1 − )
= = −4.312
1−

(1 − 3 )(3 − (1 − )) eqn.(A-37)
12 3 −
(1 − )
= = 4.788
(1 − )

The pressure distribution forward of the hinge line can be represented by the
following equation From Howe [9], ch 9.4.4.3, pp 296-297:

_ = + + = 1.894 − 4.312 + 4.788 eqn.(A-38)

As the pressure distribution aft of the hinge line is assumed as a triangular

distribution, and calculating that the value of at the hinge line is 6R, the
following polynomial can represent the distribution aft of the hinge line:

_ = −4.072 + 4.072 eqn.(A-39)

The chordwise pressure distribution due to aileron defection is plotted as per

Figure B-1 below. Note that the centre of pressure of this distribution has been
calculated by integration of the distribution as described in eqn.(A-30):

B-10
 Figure B-1 - Chord-wise Pressure Distribution due to Deflection of Aileron

Next, the chordwise pressure distribution due to the influence of the angle of
attack must be determined. Firstly, the chord-wise centre of pressure position, ̅
must be determined by calculating several contributing co-effecients below:

The aircraft pitching moment coefficient at zero lift as per the specification[2] is:

= −0.12 eqn.(A-40)

The local chord at the spanwise median point of aileron lift load (discreet load
point at 0.8108η span) = 1.773m, determined from wing geometry.

The wing aerodynamic centre ( ) at the spanwise median point of aileron lift
load was obtained from figure 8 of the project specification[2] (presented as a
fraction of local chord):

(0.8108η) = 0.223 eqn.(A-41)

The mean aerodynamic chord of the wing, provided by the specification[2] is:

̅ = 3.887 eqn.(A-42)

The at the critical aileron loading case (MTOM, 175EAS, Sea Level) is:

B-11
 eqn.(A-43)
= = 0.375
1
2

The unitised wing spanwise load due acting at the spanwise median point of
aileron lift load can be drawn from figure 14 of the project specification[2]:

( ) ( ) eqn.(A-44)
= 0.69
̅

The local wing coefficient of lift acting at the aileron loading spanwise point can
be calculated as per the equation below:

0.69( ̅ ) 0.69 × 3.887 × 0.375 eqn.(A-45)

( )= = = 0.567
( ) 1.773

The chord-wise centre of pressure position, ̅ , can now be determined using

localised coefficients for the aerodynamic centre and lift:

−0.12 eqn.(A-46)
̅= − = 0.223 − = 0.543
0.375

The chord-wise centre of pressure can now be used in the equations presented
in Howe [9] for the chordwise non dimensional pressure distribution due to the
angle of attack:

= 6[1 − 2 ̅ + (8 ̅ − 3) + 2(1 − 3 ̅ ) ] eqn.(A-47)

= −7.548 + 8.064 − 0.516 eqn.(A-48)

The combined pressure distribution with the contributions from angle of attack
and from aileron deflection can be plotted as per Figure B-2 below:

B-12
 Figure B-2 - Combined Chord-wise Pressure Distribution due to Deflection of
Aileron and Angle of Attack

The equations for chordwise pressure distribution due to aileron deflection and
angle of attack can be combined to find the equation for the overall chordwise
pressure distribution.

/ = / + eqn.(A-49)

= −2.76 + 3.752 + 1.378 eqn.(A-50)

= −7.548 + 3.992 + 3.556 eqn.(A-51)

In order to determine the percentage of force that acts on the aileron, the above
equations must be integrated with respect to the chord (integral of the pressure
distribution is equivalent to applied force)

. eqn.(A-52)
= (−2.76 + 3.752 + 1.378).

+ (−7.548 + 3.992 + 3.556).

.

B-13
 . eqn.(A-53)
−2.76 3.752
= + + 1.378
3 2
−7.548 3.992
+ + + 3.556
3 2 .

= 1.568 + 0.432 = 2 eqn.(A-54)

Which is correct, as both distributions are unitary. Therefore the percentage of

aerodynamic pressure acting aft on the hinge line and on the aileron is:

0.432 eqn.(A-55)
% = = 21.6%
2

B.1.9 Aileron Loading

As part of the loading derivations to determine the shear force, bending moment,
and torque diagrams (Section B-5B.1.7 above), the spanwise aerodynamic load
distribution due to angle of incidence and aileron defection was plotted by the
author. The plot in Figure B-3 below shows the aerodynamic load distribution
acting on the wing across the aileron span (WS13.41m – WS19.05m), plotted as
a load per percentage of aileron span:

Figure B-3 - Critical Case Aerodynamic Load Distribution (N/Aileron Span)

B-14
The plot in Figure B-3 is plotted as (N/Aileron Span) rather than (N/m) so that the
distribution can be applied to a span of the aileron not aligned with the aircraft
axis. For example, as the spar of the aileron will not be aligned with the aircraft
spanwise axis (due to sweep) the spar will be slightly longer than 5.64m (19.05m-
13.41m). The distribution shown in Figure B-3 can be applied to the actual length
of the spar, and still maintain the same distribution shape.

The incremental load at each point will be the load distribution ( / )

multiplied by the incremental change in aileron span (∆ ). Note that
Figure B-3 also plots several polynomial curves with various degrees of order.
While the 6th order polynomial best fits the actual curve, the 4th order polynomial
equation attains a high enough accuracy to be used in future calculations (with
an error percentage calculated at 0.25%). Integration of the polynomial equation
between the aileron span limits yields the total spanwise load on the wing (for the
critical case of MTOM, , Sea Level):

_ = 301,359 eqn.(A-56)

As this represents the total load on the wing, the portion of this load that acts on
the aileron is found by multiplying the total wing load by the percentage calculated
in eqn.(A-55:

= _ × 21.6% = 65,094 eqn.(A-57)

The resulting distribution is plotted within the main body of this report in Figure 9.

The above aileron loading value represents the limit loading on the structure
which will be used for actuator sizing and deflection calculations. Structural sizing
calculations will incorporate an ultimate loading factor of 1.5 as required by
CS25.303[1].

B-15
Appendix C - Tabulated Loading Data and Graphs
C.1 Applied Aileron and Roll Dampening Loads
Table C-1 - Applied Aileron and Roll Dampening Loads

Roll
Roll Rate Acceleration Aileron Force Roll Dampening Force
Mass Altitude Speed (kN) (kN)
MTOM Sea Level VA -0.885 -4.025 -137.34 189.47
MTOM Sea Level VC -0.885 -5.242 -178.85 246.74
MTOM Sea Level VD -0.295 -1.947 -66.43 91.64
MTOM Cruise VA -0.979 -1.510 -51.53 71.08
MTOM Cruise VC -0.979 -1.510 -51.53 71.08
MTOM Cruise VD -0.326 -0.541 -18.45 25.46
MTOM Ceiling VA -0.832 -0.830 -28.33 39.09
MTOM Ceiling VC -0.832 -0.830 -28.33 39.09
MTOM Ceiling VD -0.277 -0.297 -10.12 13.97

Mass Altitude Speed (kN) (kN)

MZFM Sea Level VA -0.882 -4.764 -136.34 188.08
MZFM Sea Level VC -0.882 -6.227 -178.18 245.82
MZFM Sea Level VD -0.294 -2.313 -66.18 91.30
MZFM Cruise VA -0.979 -1.801 -51.53 71.08
MZFM Cruise VC -0.979 -1.801 -51.53 71.08
MZFM Cruise VD -0.326 -0.645 -18.45 25.46
MZFM Ceiling VA -0.832 -0.990 -28.33 39.09
MZFM Ceiling VC -0.832 -0.990 -28.33 39.09
MZFM Ceiling VD -0.277 -0.354 -10.12 13.97

Mass Altitude Speed (kN) (kN)

OEM Sea Level VA -0.717 -3.205 -90.89 125.39
OEM Sea Level VC -0.717 -5.111 -144.93 199.95
OEM Sea Level VD -0.239 -1.898 -53.83 74.26
OEM Cruise VA -0.853 -1.386 -39.30 54.21
OEM Cruise VC -0.853 -1.582 -44.85 61.87
OEM Cruise VD -0.284 -0.566 -16.06 22.16
OEM Ceiling VA -0.832 -0.999 -28.33 39.09
OEM Ceiling VC -0.832 -0.999 -28.33 39.09
OEM Ceiling VD -0.277 -0.357 -10.12 13.97

Max Max Max (kN) Max (kN)

0.979485108 6.226987067 178.85 246.74

C-1
C.2 Asymmetric Gust Loads and Required Aileron Correction Deflections

Table - Asymmetric Gust Loads and Required Aileron Correction Deflections

Mass Altitude Speed ΔnG (kN) (kNm) (rad/s2) ξ required [º] Check < 20º
MTOM Sea Level VB 0.945 406.280 603.14 0.57 3.53 OK
MTOM Sea Level VC 1.106 475.434 705.80 0.67 1.93 OK
MTOM Sea Level VD 0.630 270.909 402.17 0.38 0.88 OK
MTOM Cruise VB 0.941 404.740 600.85 0.57 9.37 OK
MTOM Cruise VC 0.790 339.731 504.34 0.48 6.30 OK
MTOM Cruise VD 0.444 191.030 283.59 0.27 3.06 OK
MTOM Ceiling VB 0.841 361.508 536.67 0.51 12.43 OK
MTOM Ceiling VC 0.597 256.889 381.36 0.36 8.66 OK
MTOM Ceiling VD 0.336 144.511 214.53 0.20 4.23 OK

Mass Altitude Speed ΔnG (kN) (kNm) (rad/s2) ξ required [º] Check < 20º
MZFM Sea Level VB 0.948 404.961 601.18 0.68 3.54 OK
MZFM Sea Level VC 1.112 475.188 705.43 0.79 1.92 OK
MZFM Sea Level VD 0.634 270.764 401.96 0.45 0.88 OK
MZFM Cruise VB 0.945 403.515 599.03 0.67 9.38 OK
MZFM Cruise VC 0.795 339.661 504.24 0.57 6.30 OK
MZFM Cruise VD 0.447 190.988 283.53 0.32 3.06 OK
MZFM Ceiling VB 0.843 360.176 534.69 0.60 12.44 OK
MZFM Ceiling VC 0.601 256.849 381.30 0.43 8.66 OK
MZFM Ceiling VD 0.338 144.487 214.50 0.24 4.23 OK

Mass Altitude Speed ΔnG (kN) (kNm) (rad/s2) ξ required [º] Check < 20º
OEM Sea Level VB 1.146 338.043 501.84 0.57 3.91 OK
OEM Sea Level VC 1.556 458.999 681.40 0.77 1.86 OK
OEM Sea Level VD 0.886 261.195 387.75 0.44 0.85 OK
OEM Cruise VB 1.168 344.525 511.46 0.58 10.01 OK
OEM Cruise VC 1.136 334.920 497.20 0.56 6.21 OK
OEM Cruise VD 0.638 188.173 279.35 0.32 3.01 OK
OEM Ceiling VB 1.018 300.354 445.89 0.51 13.26 OK
OEM Ceiling VC 0.862 254.104 377.23 0.43 8.57 OK
OEM Ceiling VD 0.484 142.855 212.07 0.24 4.18 OK

C-2
C.3 Wing Spanwise Shear Force Diagram – Critical Cases

Figure C-1 - Wing Spanwise Shear Force Diagram – Critical Cases

C-3
C.4 Wing Spanwise Bending Moment Diagram – Critical Cases

Figure C-2 - Wing Spanwise Bending Moment Diagram - Critical Cases

C-4
C.5 Wing Spanwise Torque Diagram – Critical Case

Figure C-3 - Wing Spanwise Torque Diagram - Critical Case (MTOM,Sea Level, Vc)

C-5
C.6 NV Diagram Tabulated Results

h Sea Level (SL) Max Flap (MF) Cruise (CR) Ceiling (CE)
M MTOM MZFM OEM MTOM MZFM OEM MTOM MZFM OEM MTOM MZFM OEM
Va 138.297 137.842 114.541 - - 114.541 128.755 128.755 114.541 111.481 111.481 111.481
Vb 125.012 124.720 109.524 - - 121.744 117.025 116.805 105.862 110.549 110.318 98.838
Vc 175.000 175.000 175.000 - - 178.832 128.755 128.755 128.755 111.481 111.481 111.481
Vd 192.015 192.015 192.015 - - 190.295 137.009 137.009 137.009 118.627 118.627 118.627
Vf 118.276 122.287 101.615 118.276 122.287 101.615 - - - - - -
Vs1 87.467 87.179 72.442 - - 72.442 - - - - - -

Figure C-4 - Design Speeds at Mass and Altitude Variations

ΔnG ΔnG Sea Level ΔnG Max Flap ΔnG Cruise ΔnG Ceiling
M MTOM MZFM OEM MTOM MZFM OEM MTOM MZFM OEM MTOM MZFM OEM
Vb 0.945 0.948 1.146 1.233 1.238 1.521 0.941 0.945 1.168 0.841 0.843 1.018
Vc 1.106 1.112 1.556 1.375 1.383 2.028 0.790 0.795 1.136 0.597 0.601 0.862
Vd 0.630 0.634 0.886 0.798 0.803 1.138 0.444 0.447 0.638 0.336 0.338 0.484

Figure C-5 - Gust Accelerations at Mass and Altitude Variations

C-6
 Table C-2 - Maximum Accelerations at Mass and Altitude Variations

h Sea Level (SL) Max Flap (MF)

M MTOM MZFM OEM MTOM MZFM OEM
n Positive Negative Positive Negative Positive Negative Positive Negative Positive Negative Positive Negative
n1 n3 n1 n3 n1 n3 ngust n3
Va - - -
2.50 -1.00 2.50 -1.00 2.500000 -1.00 - 2.584413 -1.00
nstall nstall nstall nstall
Vb - - - - - - - -
2.042780 2.046679 2.285802 2.521057
n1 n3 n1 n3 n_gust+ n3 ngust n3
Vc - - - -
2.50 -1.00 2.50 -1.00 2.556180 -1.00 3.028316 -1.00
n1 n0 n1 n0 n1 n0 n1 n3
Vd - - - -
2.50 0.00 2.50 0.00 2.50 0.00 2.50 -0.137641
nflaps nflaps nflaps nflaps nflaps nflaps
Vf - - - - - -
2.00 2.00 2.00 2.00 2.00 2.00
nto nto nto nto nto nto
Vto - - - - - -
1.69 1.69 1.69 - - 1.69
h Cruise (CR) Ceiling (CE)
M MTOM MZFM OEM MTOM MZFM OEM
n Positive Negative Positive Negative Positive Negative Positive Negative Positive Negative Positive Negative
n1 n3 n1 n3 n1 n3 n1 n3 n1 n3 n1 n3
Va
2.166934 -1.00 2.181266 -1.00 2.50 -1.00 1.624486 -1.00 1.635231 -1 2.368210 -1.00
ngust ngust ngust ngust ngust ndyn
Vb - - - - - -
1.941285 1.944642 2.168069 1.840742 1.843183 2.018312
n1 n3 n1 n3 n1 n3 n1 n3 n1 n3 n1 n3
Vc
2.166934 -1.00 2.181266 -1.00 2.50 -1.00 1.624486 -1.00 1.635231 -1.00 2.368210 -1.00
n1 n0 n1 n0 n1 n0 n1 n3 n1 n0 n1 n0
Vd
2.166934 0.00 2.181266 0.00 2.50 0.00 1.624486 0.00 1.635231 0.00 2.368210 0.00

C-7
Appendix D – Initial Sizing
D.1 – Laminate Strength and Stiffness
Hart-Smith [22] presents a method for approximating the directional strength and
stiffness properties of a laminate for initial sizing purposes, named the “10% rule”.
The predicted strength is based on an assumption that relative to the 0° direction,
each ±45° and 90° ply contributes 10% of its strength or stiffness. Thus, using
the strength and stiffness properties for M56-AS7 ply material as detailed in the
previous section:

= 2330 = 1308 1 = 141

The following approximate laminate properties for a quasi-isotropic laminate

(0° = 25%, ±45° = 50%, 90° = 20%) are as follows (see Howe [9] for derivation):

= 0.325 × = 757.25

= 0.325 × = 425.1

= 0.325 × 1 = 45.825

= 0.122 × = 92.38

= 0.122 × = 5.6

To determine 0.2% proof stress, Howe [9] recommends allowable strains for a
quasi isotropic layup are 0.0055 and 0.004 in tension and compression
respectively. Therefore:

= 1 × 0.325 × 0.0055 = 252.04

= 1 × 0.325 × 0.004 = 183.3

It is acknowledged by the author that in reality, the actual failure strengths will be
dictated by individual ply failures and the overall laminate properties will depend
on the stacking sequence used. Further, actual failure strains will likely be lower
to take account of damage tolerant knock down factors. However, for initial sizing,
the laminate properties detailed above were deemed sufficient.

D-1
D.2 – Initial Sizing – Aileron Skin

D.2.1 Initial Skin Sizing – Torsional Stiffness Requirements

Torsional stiffness for the aileron is based off a recommendation on Howe [9] to
keep the twist at the control surface inboard and outboard ends to within 10% of
the overall deflection range. For the Ae-16 ailerons, the maximum angular
deflection is 20° (range 40°), giving a maximum allowable spanwise twist at the
tips of:

= 2 × 20° × ( ) × 10% = 0.07

180°

Bredt-Batho theory for twisting of thin walled sections was rearranged in order to
assess required torsion box thickness:

( ) ( ) eqn.(A-58)
= 2
4 ( ) ( )

( ) ( ) eqn.(A-59)
∴ =
4 ( ) ( )

( ) ( ) eqn.(A-60)
∴ ( ) =
4 ( )

Where ( ) is the local section perimeter of the structural box, ( ) is the local
section area of the enclosed structural box, G is the laminate shear modulus,
( ) is the accumulated torque at the spanwise location. It is important to note
that the torque value is at the limit load value, as torsional deflection limits only
apply to maximum expected operational loads.

The critical span of the aileron is between the second actuation hinge and the
outboard end of the aileron due to the maximum accumulated torque (see figure
Figure 17) and the largest spanwise distance ( ). Performing the integration
. ( ) −2 .
across this aileron span, the value of .
= 1956.451 Thus, the
( )

required torsion box thickness along this length of span in order to meet torsional
stiffness requirements is:

D-2
 6094 −2 eqn.(A-61)
( ) = × 1956.451 = 7.6
4 × 5.6 × 0.07

The author acknowledges this is a very thick value for the torsion box thickness.
This is a consequence of the asymmetrical placement of the actuators, leading
( )
to high accumulated torque. Further, the very high geometric integral 2 is
( )

a consequence of the high aspect ratio design of the aileron (small torsion box
area). In the detailed design stage, the layup sequence would be optimised to
reduce the required thickness of the torsion box.

D.2.2 Initial Skin Sizing – Torsional Stress Failure Assessment

The accumulated torsion at any given cross section results in a torsional stress
within the torsion box (skins and spar web). The minimum thickness of these
torsion box components was derived by assessing against the maximum
allowable shear stress of the composite material.

The shear flow within the torsion box at a given span was represented by
implementing the Bredt-Batho formula, applicable to thin walled sections:

( ) eqn.(A-62)
=
( ) 2 ( )

Where ( ) represents the local enclosed area of the torsion box, and ( ) is the
accumulated ultimate torque load.

The mean material thickness needed to react the torsion moment is then:

( ) = ( ) /2 ( ) eqn.(A-63)

As the accumulated torque and torsion box enclosed area vary along the span of
the aileron, the required thickness of the torsion box to resist torsion stresses is
detailed in Figure D-1 below:

D-3
 Torsion Box Required Thickness due to
Torsion Induced Stress (mm)
0.90

SKin Thickness Required (mm) 0.80

0.70

0.60

0.50

0.40

0.30

0.20

0.10

0.00
0% 20% 40% 60% 80% 100%
Aileron Span (%)

Figure D-1 - Torsion Box Required Thickness due to Torsion Induced Stress

The maximum value of 0.8mm is significantly lower than the 7.6mm thickness
required to for torsional rigidity. Thus, torsional stress is not a driving criterion.

D.2.3 Initial Skin Sizing – Sandwich Core Sizing

The initial depth of the sandwich material was calculated by taking into account
the initial sizing thickness of the sandwich facing (7.6mm (2 x 3.8mm) as required
to meet torsional stiffness requirements). The benefit of a sandwich panel
configuration, is the ability to tailor the core thickness to optimise flexural rigidity.
The required core thickness was calculated using the methodology within Niu,
Airframe Stress Analysis and Sizing, pg 216 [25].

Note, that as the sizing methodology is an industry accepted method using

empirical data, the method is repeated in the detailed stressing analysis using
final sandwich facing thicknesses. As such, in order to avoid duplication, the
derivation and methodology can be seen in detail in Appendix F.2.10.

D-4
D.3 – Initial Sizing – Aileron Spar

D.3.1 Spar Flange Sizing – Spar Flange Loading

Assuming that the flanges react 100% of the bending moment, the axial force ( )

acting on each flange at any spanwise location ( ) is as follows:

( ) eqn.(A-64)
( ) =
ℎ( )

Where ℎ( ) represents the local distance between the flanges, and ( ) the local
moment. With equal thickness flanges the critical flange will be on the
compression side due to a lower compressive allowable stress, and the
susceptibility to compressive instability failure (buckling/crippling). Note, that the
flange width is equal to the distance between the hinge line and the spar web
(50mm).

D.3.2 Spar Flange Sizing – Ultimate Compressive Stress Failure

The thickness required to preclude compressive stress failure was found by
rearranging the standard engineer’s stress formula below:

3088.10 eqn.(A-65)
= = = = 1.18
ℎ 0.123 × 0.05 × 425.1

Where, M represents the largest bending moment at the critical design case,
which is 3088.10Nm (ultimate) occurring at the first hinge (see Figure 16). The
local spar height at this location is 123mm (obtained from CATIA model), and with
flange width is constant along the span at 0.05m, the required thickness of the
quasi-isotropic laminate is 1.18mm to preclude compressive overstress failure

D.3.3 Spar Flange Sizing – Compressive Instability Failure

Buckling calculations for composite laminates inherit a higher level of complexity
compared to isotropic materials, as the stacking order severely affects the flexural
stiffness of the laminate panel [24]. There are no simplified equations available
to determine buckling characteristics of laminate beams/panels without
prescribing the layup and resulting flexural stiffness [D] matrix. To gain a rough
estimate for a flange thickness requirement, the author used a methodology for

D-5
panel buckling of isotropic thin walled panels as detailed in Niu [25]. The equation
below gives the allowable buckling stress for an isotropic flat plate under an in-
plane compression load:

2 eqn.(A-66)
_ =

Where is the flat plate buckling co-efficient for a panel simply supported on
three sides (which best approximates the flange supporting conditions). can
be derived from Niu, Fig. 11.3.1, pg 458[25]. The length of flange compared to
the width means the value approaches the asymptote value of 0.426 for a
material with a poisons ratio of 0.3 as can be seen in Figure D-2 below:

Figure D-2 – Flat Plate Buckling Co-efficient [25]

By dividing eqn.(A-64) by the flange area (to obtain applied stress) and combining
this with equation eqn.(A-66, the expression for required thickness ( ) to avoid
flange instability can be represented as:

D-6
 2 eqn.(A-67)
3 3
3088.1 × 0.052
= = = 4.01
ℎ 0.123 × 0.05 × 0.426 × 45.825

Where, under this simplified bucking analysis, is value of the membrane

compressive stiffness value of 45.825GPa as derived in Section D.1. The critical
bending moment at the second hinge positon was used (3088.1Nm with 123mm
spar height). Entering this thickness back into the rearranged form of eqn.(A-64,
the resulting stress under load is 125.53MPa. As this is significantly lower than
the compressive yield allowable ( = 183.3 ), the critical buckling load is
unlikely to be affected by premature ply failures.

The author acknowledges the isotropic analysis for buckling may differ
significantly from the actual composite buckling response. The actual layup
sequence for the laminate will affect the flexural stiffness, and hence the buckling
response. The isotropic analysis has also been derived from first order shell
theory, which neglects the effects of transverse shear in the material (this effect
can significantly affect the buckling response in composite laminates [46].
Regardless, the methodology used provided a good starting point for detailed
analysis. Final flange sizing would be based on laminate plate theory, and take
actual flexural rigidity and buckling reserve factors into account.

D.3.4 Spar Flange Sizing – Flexural Rigidity Requirement

The flexural rigidity ( ) of the spar was assessed at the initial sizing stage to
ensure the deflections remain with acceptable limits, and to provide a validation
check against Strand7 FEA results. Deflection at the critical location of the
inboard spar end was checked using the equation in AVD Initial Sizing Notes [31]:

eqn.(A-68)
=
8

Where, is the uniform distributed loading = 13,000 / (actual loading

simplified to a uniform distribution for validation purposes)

is the length between fist hinge and the inboard end of the spar = 0.56

D-7
 is the moment of intertia of the two flanges representative of the end span of
the spar, found through parallel axis theorem to be = 1.44 × 10

13,000 / × (0.56 ) eqn.(A-69)

= = = 2.23
8 8 × 45.825GPa × 1.44 × 10

This is within 2% of the local spar depth at the end of the spar (2% × 130 =
2.6 ), which is the recommended maximum allowable deflection stated by
Howe [9]). Therefore, the initial thickness of the spar flange was deemed
sufficient for deflection requirements.

D.3.5 Spar Web Sizing

Under the mass boom analysis method, the spar web is assumed to react all
vertical shear loads. These loads would need to be combined with the shear flow
in the web resulting from torsion calculated earlier:

The shear flow induced by torsional load was found by implementing the Bredt-
Batho formula for thin walled closed torsion sections:

( ) eqn.(A-70)
=
( ) 2 ( )

The shear flow component due to the applied shear loads can be represented as:

( ) eqn.(A-71)
=
( ) ℎ ( )

As per Howe, equation 15.13 [9], the net shear flow can thus be represented by
the following equation:

( ) = ( ) +2 ( ) × / eqn.(A-72)

Where, / is the chordwise ratio of where the concentrated shear force acts
relative to the torsion box. As the pressure distribution aft of the hinge line is
triangular in shape, the value of the / ratio is 1/3 (CoP of triangular distribution)

The required spar web thickness to resist torsional shear flow and ultimate shear
force was determined by the equation below:

D-8
 = / eqn.(A-73)

As the spar height, applied shear force, torsion box area, and local torque all vary
along the span of the aileron, the required thickness also varies along the span.
The critical spanwise location is at the second hinge position, where the shear
flow due to torsion is highest (due to the high accumulated torsion). The local
shear flow at the second hinge position is as per below:

( ) ( ) 1 eqn.(A-74)
= +2 ×
( ) ℎ ( ) 2 ( ) 3
12,578 9141 1
= +2× ×
0.108 2 × 0.03028 2 3
= 217,090 /

Thus, as per equation eqn.(A-73, the required thickness of the web to resist this
shear flow is as follows:

217,090 eqn.(A-75)
= = = 2.35
92.38

D.4 – Initial Sizing – Aileron Ribs

As discussed in the main body of this thesis, the critical rib at the second hinge
location was analysed (returning conservative sizing for the remaining ribs). The
aerodynamic load reacted at Hinge 2 is 24,940N ultimate, see Table 5.

D.4.1 Rib Sizing – Rib Web

The aerodynamic pressure distribution aft of the hinge line is triangular in shape.
To find the vertical shear force transferred by the rib web to the spar, the
percentage of aerodynamic load aft of the hinge line must be established. The
cross sectional geometry at the critical rib location (hinge two) and its chordwise
aerodynamic distribution is shown in Figure D-3 below:

D-9
 Figure D-3 - Hinge 2 Section Cut and Load Distribution

Through calculation of triangular areas, the proportion of aerodynamic load acting

forward of the spar web is 17.4%, and aft of the spar web is 82.6%. Thus, the total
aerodynamic load aft of the spar web is 24,940 × 82.6% = 20,600 . Due to
the triangular distribution, this acts at 1/3rd of the distance between the spar web
and the trailing edge, or 168 aft of the spar web.

The vertical shear force applied to the aft section is reacted at the spar web/rib
web interface. The standard average shear stress equation, applied to this critical
location, can be rearranged to calculate required rib web thickness:

ℎ 20,600 eqn.(A-76)
= = = 2.06
ℎ 92.38 × 108

D.4.2 Rib Sizing – Rib Flange

The rib flanges primarily transfer the bending loads forward to the main spar.
Under initial sizing, it is conservatively assumed that the skins do not transfer
these chordwise bending loads. The largest bending moment will occur at the
spar/rib flange interface due to its maximal distance from the centre of pressure.
The axial load reacted by each flange due to bending at this location is:

D-10
 20,600 × 168 eqn.(A-77)
= = = 32,044
ℎ 108

32,044 eqn.(A-78)
= = = 5.8
× 30 × 183.3

Where the flange width has been set at 30 . While this thickness seems large,
this conservative calculation assumes all chordwise bending moment is reacted
by the rib flange. In detailed sizing, the skin would be included as contributing to
the chordwise bending moment reaction, therefore reducing the required flange
area.

D-11
Appendix E – Laminate Selection Iteration Process
This appendix includes the details of each detailed analysis iteration, stacking selection, thickness, reserve factors, and progressive
changes. Each stacking sequence variant is noted by a code (S11, S6 etc.) which correlates to the specially orthotropic stacking
sequences detailed in [23]. Thick laminates that are modified from the layups specified in [23] are noted with ‘MOD’:

E-1
E-2
Appendix F – Detailed Stressing Report
Table of Contents

F.1 Introduction............................................................................................................ F-3

F.2 Composite Analysis ............................................................................................... F-3

F.2.1 Laminate Properties (Starting Variables) ........................................................ F-3

F.2.2 In-Plane Laminate Distributed Loads due to Spanwise Bending .................... F-4

F.2.3 Torsion Box Shear Flow.................................................................................. F-8

F.2.4 Torsion Box Shear Flow due to Accumulated Torsion .................................... F-8

F.2.5 Torsion Box Shear Flow due to Accumulated Shear ...................................... F-9

F.2.6 Torsion Box Shear Flow – Combined Loads................................................. F-13

F.2.7 In-Plane Laminate Distributed Loads Summary............................................ F-15

F.2.8 Sandwich Skin – Strength and Strain Analysis ............................................. F-15

F.2.9 Sandwich Skin – Compression and Shear Buckling Analysis ....................... F-17

F.2.10 Sandwich Skin – Core Deflection Analysis ................................................. F-18

F.2.11 Sandwich Skin – Facing Laminate Bending Strength Analysis ................... F-20

F.2.12 Skin Adhered to Spar Flange–Strength Analysis ........................................ F-22

F.2.13 Skin Adhered to Spar Flange– Local Compression Buckling Analysis ....... F-23

F.2.14 Spar Web - Strength and Strain Analysis.................................................... F-24

F.2.15 Spar Web - Shear Buckling Analysis .......................................................... F-25

F.2.16 Spar Flange - Strength and Strain Analysis ................................................ F-26

F.2.17 Spar Flange - Local Compression Buckling Analysis .................................. F-27

F.2.18 Spanwise Flexural Rigidity Analysis............................................................ F-27

F.2.19 Reserve Factor Summary for Spanwise Loaded Laminates ....................... F-28

F.2.20 Rib Analysis ................................................................................................ F-29

F-1
 F.2.21 Rib Analysis - Aerodynamic Loading Contribution ...................................... F-30

F.2.22 Rib Analysis – Actuation Hinge Loading Contribution ................................. F-34

F.2.23 Rib Analysis – Combined Rib Shear Flows ................................................. F-39

F.2.24 Rib Analysis – Chordwise Shear Force and Bending Moment Plots........... F-40

F.2.25 Rib Analysis – Rib Web .............................................................................. F-42

F.2.26 Rib Analysis – Rib Flange ........................................................................... F-45

F.2.27 Rib Analysis – Reserve Factor Summary ................................................... F-47

F.3 Connection Analysis ............................................................................................ F-48

F.3.1 Adhesion Analysis ........................................................................................ F-48

F.3.2 Mechanical Fastener Analysis – Hinge to Spar Connection ......................... F-49

F.4 Lug Analysis ........................................................................................................ F-54

F.4.1 Lug Material Properties ................................................................................. F-54

F.4.2 Lug Configuration and Geometry .................................................................. F-55

F.4.3 Lug Loading .................................................................................................. F-58

F.4.4 Wing Side Triple Lugs - Upper Pivot Location .............................................. F-59

F.4.5 Bolt Analysis ................................................................................................. F-65

F.4.6 Upper Pivot Connection Point - Aileron Side (two lugs) ................................ F-69

F.4.7 Lower Actuation Connection Point - Aileron Side (two lugs) ......................... F-73

F.4.8 Lower Actuation Connection Point - Actuator Single Lug Analysis ............... F-76

F.4.9 Hinge Reserve Factor Summary................................................................... F-76

F-2
F.1 Introduction
The detailed stressing was undertaken using PTC MathCad® mathematical equation
software. This software takes inputs as variables, which can then be used by any equation
throughout the analysis. Using this software ensures that there are no calculation or
rounding carry-through errors within the analysis. Further, as variables such as material
properties and geometries are set at the start of an analysis, the user can easily change the
variable values, which in turn affects the equations return results.

Note that throughout the detailed stressing analysis, the variables will not be repeated, as
they would have been stated earlier in the report.

F.2 Composite Analysis

This section will include all detailed stressing analysis steps used to determine the optimum
laminate stacking sequences described in Section 6 of the main body of this report. The
analysis contained within this section was repeated several times, as each iteration was
affected by changes to laminate thicknesses, stacking sequences, and hence overall
stiffness properties.

The analysis detailed in the section below is for the final laminates selected (see Section
6.7 for results discussion of the final laminate selections).

F.2.1 Laminate Properties (Starting Variables)

The thickness of all laminates detailed below are based off the number of plies used (each
ply is 0.134mm thick [17]). The flexural stiffness matrix [ ] values and equivalent membrane
stiffnesses are determined using COALA for each individual laminate:

Sandwich Skin Properties:

F-3
Skin Adhered to Flange Properties:

Spar Properties:

Rib Properties:

F.2.2 In-Plane Laminate Distributed Loads due to Spanwise Bending

The in plane laminate loads (Nx) from spanwise bending need to be calculated for each
component reacting spanwise loading (spar web, spar flange, sandwich panel, & skin flange
adhered to spar flange).

Note that compared to a mass boom loading distribution, the distributed flange loading
distribution analysed by the author considered the skin as fully effective in reacting the

F-4
bending and shear loads. While this leads to a higher level of complexity in the analysis
steps, it leads to a highly efficient and optimised design.

The analysis will occur at the second actuation hinge position. The laminates at this location
react the highest accumulated torque (9141.94Nm, see Figure 17) and a high vertical shear
reaction value (12,578.4N). The geometry at this location is simplified to a triangular
enclosed area are shown in Figure F-1 below:

Figure F-1 - Cross Section at Second Hinge Position

Width of spar flange (and skin adhered to spar flange)

Height of web spar at second hinge location

Chordwise length of torsion box aft of the spar web

Angle that each sandwich panel makes with the horizontal

The length of the sandwich panel aft of the spar web, and thickness are as per below:

F-5
In order to ascertain the bending loads reacted by each component, their second moment
of area properties must first be determined:

As each laminated component has a different equivalent membrane stiffness, each

component’s contribution to the overall sectional bending stiffness will depend both on their
second moment of area, and their relative stiffness values. The bending stiffness of the cross
section is as per the equation below:

The load reacted within each component from the spanwise bending force can now be
determined using a variation of the standard engineers bending theory (including the
influence of equivalent membrane stiffness ).

The ultimate spanwise moment at the second hinge position is as per below:

(see Figure 16 in main body of this report)

F-6
The in-plane bending load distribution reacted by the spar flange is:

The in-plane bending load distribution reacted by the skin adhered to the spar flange is:

The in-plane bending load distribution reacted by one sandwich panel is a maximum at the
forward most end of the sandwich panel (at the spar web), as this location is furthest from
the neutral axis:

The bending resistance from the spar web is not included in the bending analysis, as its
contribution is relatively insignificant compared to the other components detailed above.

F-7
F.2.3 Torsion Box Shear Flow
The torsion box of the aileron is comprised of the sandwich panels aft of the spar web, and
the spar web (which together form an enclosed area). The shear flow ( ) due to loading
within these torsion box components is comprised of two elements, shear flow due to
accumulated torque, and shear flow due to the accumulated spanwise shear load.

The accumulated ultimate torque at hinge position 2 is 9141.94Nm (see Figure 17).

The accumulated ultimate shear load at hinge 2 is 12578.4N (half the hinge reaction load).

For the shear flow calculations, it is assumed the spar flanges and skin adhered to the spar
flange do not contribute to reacting the vertical shear or torsional loading. This is due to their
perpendicular orientation to the vertical shear direction, and their discontinuous open cell
configuration (poor at torsional resistance). Further, the curved aerodynamic nose laminate
(forward of the spar) is not considered effective in torsional resistance due to thin thickness
and small enclosed area. These assumptions are conservative.

F.2.4 Torsion Box Shear Flow due to Accumulated Torsion

The critical accumulated torsion has been determined by integrating the polynomial equation
for distributed spanwise torque between the critical hinge location and the outboard end of
the aileron (see Figure 17). See Section 2.6 within the main body of this report for an
explanation of the derivation.

Note that a positive torsion value indicates the torque acting in the anticlockwise direction
when viewing the aileron from the left hand side.

Ultimate torque to be reacted by the torsion box

Area enclosed by the torsion box

The shear flow within the torsion box due to accumulated torsion can be determined by
implementing the Bredt-Batho formula, applicable to thin walled sections. A negative
value indicates a shear flow in the clockwise direction (shear flow always opposes the
applied torque, in this case it opposes the anticlockwise accumulated torque):

Torsion box shear flow due to torsion

F-8
 Figure F-2 - Torsion Box Shear Flow due to Torsion

F.2.5 Torsion Box Shear Flow due to Accumulated Shear

The second loading type that is reacted by the torsion box components is the accumulated
vertical shear. As discussed above, only the spar web and sandwich skins are considered
effective in reacting the vertical shear force. In reacting this vertical shear, a shear flow is
created within the spar web and sandwich skin panels (which will be combined with the
shear flow due to accumulated torsion, calculated in the previous section).

The bending stiffness of the torsion box is as per below (only web and sandwich skin
considered effective in reacting torsion and vertical shear):

The ultimate accumulated shear flow at the second hinge location is as per below. This is
from a downwards aileron deflection (critical case). Therefore it acts in the positive direction.

As shear flow due to torsion has already be calculated in Section F.2.4, it is important that
the shear flow is assumed to act at a chordwise location that does not lead to additional
torsion (as this has already been accounted for in torsional shear flow analysis). An applied
shear load will produce zero twist when applied at the shear centre of the torsion box. The
following analysis will determine the shear flow due to the shear load applied at the shear
centre of the torsion box and in doing so, identify the shear centre location.

F-9
The first step is to cut the cross section at an arbitrary point and determine the open shear
flow of members around the torsion box. The arbitrary cut has been chosen at the trailing
edge (point 1 detailed below). The directional notation and integration direction is also shown
below:

Figure F-3 - Torsion Box Integration Points used in Shear Flow Analysis

With the cut existing at the trailing edge, the open shear flow for the torsion box members is
integrated in a clockwise direction, starting at point 1 (zero open shear flow):

F-10
These positive shear flow values indicate an anticlockwise shear flow. The final equation
above is a check to ensure the integration around the torsion box was correct, ensuring that
the open shear flow returns to zero.

Figure F-4 - Open Section Shear Flow due to Accumulated Shear

By integrating the shear flow equations for each torsion box member along its length, the
open shear load reacted by each member can be determined:

The next step is to determine the closed shear flow magnitude that ensures zero twisting of
the torsion box. As discussed above, the shear load is required to pass through the shear
centre to ensure the overall section will not twist. The equation below is a variant of the
Breth-Batho formula for sectional twist, adjusted for shear flows. By setting this to equation
equal to zero, the closed shear flow required to keep the section from twisting can be
ascertained:

1
∅=
2

F-11
 1
0= +
2
1
+ ( + )
2
1
+ ( + )
2

Rearranging to determine the closed section shear flow ( ):

The negative sign of the closed shear flow indicates a clockwise direction for the flow.

Figure F-5 - Closed Section Shear Flow due to Accumulated Shear

The vertical shear load reacted by each torsion box member can be determined by
integrating the shear flow equations again, this time including the closed shear flow:

F-12
A simple check can now be performed to ensure that the shear force in the vertical direction
within the torsion box members equates to the accumulated shear load of 12,578.4N:

Further, another check can be made to ensure that there is zero moment at the shear centre.
To determine the chordwise location of the shear centre of the torsion box, moments are
taken about the trailing edge (point 1) which will equate to zero. By taking moments about
the trailing edge, the distance between the shear centre and trailing edge is found as per
below:

The moment arm between the shear centre point and each sandwich panel is:

And checking that the moment about the shear centre point is zero:

With both the vertical and moment checks returning the correct results, the shear flows that
have been calculated are confirmed to be the correct values.

F.2.6 Torsion Box Shear Flow – Combined Loads

The total shear flow in the torsion box members can now be assessed by combining the
shear flow from torsional loading with the shear flow from shear loading. The shear flow at
the vertex points of the torsion box are detailed below, and the overall shear flow is shown
in Figure F-6:

F-13
A negative shear flow indicates shear flow in the clockwise direction, opposing the positive
torque acting in the anticlockwise direction.

Figure F-6 - Torsion Box Shear Flow due to Accumulated Torsion and Shear

The maximum shear distribution in each member equates to the largest shear flow within
each torsion box member as follows:

F-14
F.2.7 In-Plane Laminate Distributed Loads Summary
The in-plane loading distribution for the laminates which react spanwise loads are
summarised below. These loading distributions will be used in the proceeding analysis
sections.

Sandwich Panel Distributed Loading:

Skin Adhered to Spar Flange Distributed Loading:

Spar Distributed Loading:

F.2.8 Sandwich Skin – Strength and Strain Analysis

The laminate strength and strain analysis is conducted with COALA, using the maximum
load intensities as calculated above (Nx, Nxy). Stress failure criteria are presented as both
an interactive criteria (Tsai-Hill) and one non-interactive failure criteria (max stress/strain).
The interactive failure criteria assessment better reflects the actual stress/strain conditions
in the material due to the interactions between variable direction stresses. The non-
interactive criteria allows the identification of the failure mode.

F-15
The maximum strain failure criterion below is based on a maximum allowable strain of
3500 strain. This limit is based off reccomended damage tolerance allowances as
specified in the AVD Fatigue and Damage Tolerance notes [28].

Entering the load intensity magnitudes to COALA, the following failure indices are calculated
and converted to reserve factors:

(Mode of Failure: Transverse Direction Compressive Failure in +45 Plies)

The minimum reserve factor for the sandwich skin is as per below:

F-16
F.2.9 Sandwich Skin – Compression and Shear Buckling Analysis
Sandwich skin buckling analysis has been carried out using ESDU 80023 [29]. The
sandwich panel is conservatively assumed as simply supported panels.

Figure F-7 - ESDU 80023 Panel Dimensions & Load Case [29]

The critical panel dimensions are conservatively assumed as square, with the following edge
lengths:

Ratio used in ESDU 80023 [29]

The following are input equations for ESDU 80023 [29] to determine buckling load:

F-17
Using ESDU 80023 Figure 7 [29], the allowable shear load :

Using ESDU 80023 Figure 7 [29], the allowable compressive load :

F.2.10 Sandwich Skin – Core Deflection Analysis

Methodology from Niu, Airframe Stress Analysis and Sizing [25] is used to assess the
sandwich skin panels for out of plane deflection. The requirement from Niu is that the panel
deflection should not be more than 0.5inch (1.27cm).

The maximum deflection at the centre of a homogenous plate can be ascertained by using
equation Eq. 7.2.2 within [25] and finding from Fig. 7.2.2 within [25].

F-18
 Figure F-8 - Fig. 7.2.2 from Niu [25] used to find and

From Figure F-8, the alpha value for panel deflection is:

The total surface area of the aileron has been used to determine the equivalent uniform
distributed loading pressure applied to the plate. The total surface area of the lower (critical)
skin stated below is taken directly from the CATIA surface model of the Ae-16 aileron:

The pressure distribution between the upper and lower skins of the aileron is (1/3 upper)
and (2/3 lower) as discussed earlier in this report. The UDL for the critical lower skin is
derived below:

F-19
It must be noted that the actual chordwise pressure distribution is triangular in shape, not
uniformly distributed. The UDL is used for simplicity, and is conservative as it will result in
larger than actual panel deflection and bending stresses.

Methodology from Niu, CH 7, pp216-217 [25], is used to determine panel deflection for a
non-homogenous sandwich panel. Derivation of the equation below is not stated here as it
is detailed extensively within the reference text.

Niu recommends that the deflection of the sandwich panel does not exceed 0.5”, therefore
the reserve factor against panel deflection requirements is:

F.2.11 Sandwich Skin – Facing Laminate Bending Strength Analysis

As a result of the applied pressure load, there is a bending stress that is created within the
sandwich panels. Again, the methodology from Niu, CH 7, pp216-217 [25], is used to
determine the bending stress within the sandwich panel face plates.

Firstly, the bending stress coefficient needs to be ascertained from Niu, Fig. 7.2.2 [25].
(See Figure F-8 above).

The following equation to determine sandwich panel bending strength is derived in detail
within Niu, CH 7, pp216-217 [25], and so will not be derived within this report:

F-20
This stress represents the stress within the sandwich facing as a result of bending. The
stress is in the critical direction with respect to the plate boundary lengths (i.e. the maximum
bending moment occurs perpendicular to the longest boundary).

In order to find the reserve factors for the individual plies within the sandwich facing layup,
the stress will be converted to direct in-plane loading as per below:

It must be noted that the loading on the sandwich facing is not combined with the in-plane
loading from global bending of the aileron structure. This has been omitted, as combining
the two has been deemed to produce an overly conservative design. The in-plane loading
from global bending is maximum in the sandwich skin at the rib locations, near to the spar.
The maximum bending stress in the panel however, is in the centre of the panel (far from
the boundary). Further, as the pressure is assumed UDL, the bending stress equations are
conservative. By ensuring that both reserve factors for the global bending, and local bending
due to pressure are positive, the combined reserve factor is also very likely to be positive.

The local in-plane loading due to applied pressure on the sandwich skin has been inputted
to COALA to derive the following reserve factors:

F-21
Mode of Failure: Transverse Direction Compressive Failure in +45 Plies

The minimum reserve factor for the sandwich skin facing laminates is as per below:

F.2.12 Skin Adhered to Spar Flange–Strength Analysis

The two sandwich faces for each sandwich panel come together at the forward and aft edges
of each sandwich panel, effectively making a macroscopically homogenous plate. The skin
adhered to the spar flange has considerable strength and will react the bending loads
applied to the global aileron structure. The skin adhered to the flange must be assessed for
strength/strain, and also be assessed for local buckling due to its free edge.

The in-plane compressive load has been derived in the first section of this detailed stressing
report. This has been inputted into COALA to determine the interactive and non-interactive
failure criteria:

F-22
Mode of Failure: Transverse Direction Compressive Failure in +45 Plies

The minimum reserve factor for the skin flange laminate is as per below:

F.2.13 Skin Adhered to Spar Flange– Local Compression Buckling Analysis

Local skin flange buckling analysis has been carried out using ESDU 03013 [30]. This
analysis method is applicable to flanges behaving as long flat plates with one side free and
the other simply-supported, subjected to a uniform end-load; the effect of the shear load
distribution is neglected. For this buckling analysis, the simply supported edge is assumed
as the spar web location, where the sandwich faces come together to create the plate
laminate. The width of the nose skin has been included in the effective width of the flange
as both layups are identical, and this achieves conservatism assuming they act in unison.
The derivation of the buckling equation is detailed in ESDU 03013, and so will not be
replicated here:

Note, that for simplicity and conservatism, the skin flange has been analysed in isolation
from the spar flange. In reality the two laminates adhered to one another will lead to much
higher buckling resistance (the combined flexural stiffness would be much greater). Both
F-23
the spar flange and the skin adhered to the spar flange are comprised of the same
stacking sequence, so the adhered structure is overall balanced under load.

F.2.14 Spar Web - Strength and Strain Analysis

The maximum shear load intensity (Nxy) that was calculated earlier in this report is entered
into COALA to determine failure indices against strength and strain criteria. Stress failure
criteria are presented as both an interactive criteria (Tsai-Hill) and one non-interactive failure
criteria (max stress/strain). The interactive failure criteria assessment better reflects the
actual stress/strain conditions in the material due to the interactions between variable
direction stresses. The non-interactive criteria allows the identification of the failure mode.

The maximum strain failure criterion below is based on a maximum allowable strain of 3500μ
strain. This limit is based off recommended damage tolerance allowances as specified in
the AVD Fatigue and Damage Tolerance notes [28].

Entering the load intensity magnitudes to COALA, the following failure indices are calculated
and converted to reserve factors:

Mode of Failure: Transverse Direction Compressive Failure in +45 Plies

F-24
The minimum reserve factor for the spar web laminate is as per below:

F.2.15 Spar Web - Shear Buckling Analysis

Spar web buckling analysis has been carried out using ESDU 80023 [29]. The web panel is
assumed as simply supported (conservative)

Figure F-9 – ESDU 80023 Panel Dimensions & Load Case [29]

The panel dimensions used in the ESDU analysis are as per below:

The input equations for ESDU 80023, Figure 5 [29], are detailed below:

F-25
Using ESDU 80023 Figure 5 [29], the allowable shear load is:

F.2.16 Spar Flange - Strength and Strain Analysis

The spar flange primarily reacts bending loads. It has been assumed the shear flow within
the spar flange is negligible due to the free edge condition as discussed earlier in this
detailed stressing report.

Similar to previous analyses of strength and strain within this report, the interactive and non-
interactive failure criteria have been assessed in COALA and are presented below:

Mode of Failure: Transverse Direction Compressive Failure in +45 Plies

F-26
The minimum reserve factor for the spar flange laminate is as per below:

F.2.17 Spar Flange - Local Compression Buckling Analysis

Local spar flange buckling analysis carried out using ESDU 03013 [30]. This analysis
method is applicable to flanges behaving as long flat plates with one side free and the other
simply-supported, subjected to a uniform end-load; the effect of the shear load distribution
is neglected. The derivation of the buckling equation is detailed in ESDU 03013, and so will
not be replicated here:

F.2.18 Spanwise Flexural Rigidity Analysis

The vertical deflection at any point along the aileron is not to exceed 2% of the local spar
height as recommended by Howe [9]. Deflection is assessed using the standard beam
flexural formula below (under uniform distributed load).

F-27
The critical location is the inboard most free end of the aileron (inboard of the first hinge
position) as this free end section is under the largest load. This section is 0.56m in length.
By taking the average loading on this section as 13,000N/m (see Figure 9 in the main body
of this thesis), the deflection at the tip is:

As the local spar height at the inboard end is 130mm.The reserve factor is based on the 2%
magnitude of this depth as per below:

F.2.19 Reserve Factor Summary for Spanwise Loaded Laminates

F-28
F.2.20 Rib Analysis
This section will assess the laminates which react the chordwise loading of the aileron,
namely the rib web, rib flanges, and the sandwich skin. As the skin is considered fully
effective in the distributed loading distribution analysis, the required thickness of the rib
components will be lower than an equivalent mass-boom configuration.

Chordwise shear force and bending moment diagrams need to be produced at the critical
rib position in order to identify the maximum forces reacted by the rib components. The
maximum shear force within the rib will drive the rib web thickness, and the maximum
bending moment will drive the rib flange dimensions.

The critical rib location is at the second hinge location due to the high reacted air loads
(24,940N ultimate, see Table 5), the actuation reaction loads, and the smaller geometrical
dimensions compared to the inboard ribs. All ribs will be sized off the analysis assessed at
the critical rib location, thereby obtaining simplicity and conservatism.

Similar to earlier in the report, the methodology used to derive the shear force and bending
moment diagrams will follow the methodology of the AVD Detailed Stressing Lecture Notes
[47]. However, due to the author choosing to optimise a distributed flange design, the
calculations inherit additional complexity.

In order to derive the shear force and bending moment diagrams in the chordwise direction
for the rib, the shear flow in the torsion box needs to be ascertained. This shear flow is made
up of two contributions:

-Shear flow due to applied local aerodynamic loading. This applies a shear and torsional
force which must be reacted by the torsion box.

-Shear flow due to applied hinge loads. This also applies a shear and torsional force which
must be reacted by the torsion box.

F-29
F.2.21 Rib Analysis - Aerodynamic Loading Contribution
It must be noted that the analysis is similar to that conducted earlier in the detailed stressing
report. The second moment of area for members effective in shear and torsion at the critical
rib location has already been determined (torsion box components).

The portion of aerodynamic load acting on the hinge/rib is equivalent to the hinge reaction
load ascertained in Section 2.6 of this thesis. This applied shear force is 24,940N, detailed
in Table 5. As the pressure distribution aft of the hinge line is triangular in shape, the centre
of pressure acts at a distance of 1/3rd of the aileron chord (552/3) = 184mm.

Figure F-10 - Aerodynamic Load Distribution - Hinge Two

The flexural stiffness of the torsion box in the spanwise direction is detailed below. The spar
and skin flanges are not considered effective in reacting vertical shear or torsional forces as
explained previously in this stress report:

The first step is to cut the cross section at an arbitrary point and determine the open shear
flow of members around the torsion box. The arbitrary cut has been chosen at the trailing
edge (point 1 detailed below). The directional notation and integration direction is also shown
in Figure F-11 below:

F-30
 Figure F-11 - Torsion Box Integration Points used in Shear Flow Analysis

With the cut existing at the trailing edge, the open shear flow for the torsion box members is
integrated in a clockwise direction, starting at point 1 (zero open shear flow):

These positive shear flow values indicate an anticlockwise shear flow. The final equation
above is a check to ensure the integration around the torsion box was correct, ensuring that
the open shear flow returns to zero.

By integrating the shear flow equations relating to each torsion box member, the open shear
load at each member can be determined:

F-31
In order to determine the closed section shear flow, it must be recognised that the moments
at each point within the torsion box must equate to zero. Therefore, taking moments about
the trailing edge (vertex point 1):

Note above, that the vertical shear force acts at a distance of (2/3rd) the total chord.

Using the enclosed area of the torsion box, the closed shear flow is calculated as:

The shear load within each member when reacting the total shear load can be determined
by integrating the shear flow equations again, this time including the closed shear flow:

A simple check can now be performed to ensure that the shear force in the vertical direction
within the torsion box members equates to the accumulated shear load of 24,940N:

Further, another check can be made to ensure that moments equate to zero at all sections
within the torsion box. Therefore, taking moments about the trailing edge (vertex 1):

F-32
With both these checks returning the correct results, the shear flows that have been
calculated are confirmed to be the correct values.

The shear flow due to aerodynamic loads at each vertex of the torsion box is presented
below. These will be added to the shear flow due to hinge flows in the next section, where
a final shear flow diagram will be produced:

A positive shear flow indicates shear flow in the anticlockwise direction. As expected, the
shear flow in the web is acting in the downwards direction (opposing aerodynamic upwards
shear).

F-33
 Figure F-12 - Torsion Box Shear Flow from Chordwise Aerodynamic Loading

F.2.22 Rib Analysis – Actuation Hinge Loading Contribution

Additional to the shear flow produced by aerodynamic loads, the loading from the actuation
and hinge reaction loads must be reacted by the critical rib. The analysis is similar to the
aerodynamic shear flow calculations above.

The hinge loads at the critical rib position, at the second inboard hinge are:

-24,940N reaction in the downwards direction (opposing aerodynamic load, see Table 5)

-8443.5Nm torque reaction from actuation coupling (see Figure 17)

These loads act at the hinge line chord, which is 50mm forward of the spar web.

F-34
 Figure F-13 - Loads at hinge line - Hinge Two

The flexural stiffness of the torsion box is detailed below. The spar and skin flanges are not
considered effective in reacting vertical shear or torsional forces as explained previously in
this stress report:

The first step is to cut the cross section at an arbitrary point and determine the open shear
flow of members around the torsion box. The arbitrary cut has been chosen at the trailing
edge (point 1 detailed below). The directional notation and integration direction is also shown
in Figure F-11 below:

Figure F-14 - Torsion Box Integration Points used in Shear Flow Analysis

F-35
With the cut existing at the trailing edge, the open shear flow for the torsion box members is
integrated in a clockwise direction, starting at point 1 (zero open shear flow):

These negative shear flow values indicate a clockwise shear flow. The final equation above
is a check to ensure the integration around the torsion box was correct, ensuring that the
open shear flow returns to zero.

By integrating the equation relating to each torsion box member, the open shear load at
each member can be determined:

With negative values indicating shear in a clockwise direction around the torsion box.

In order to determine the closed section shear flow, it must be recognised that the moments
at each point within the torsion box must equate to zero. Therefore, taking moments about
the trailing edge (vertex point 1):

F-36
Using the enclosed area of the torsion box, the closed shear flow is calculated as:

The shear load within each member when reacting the total shear load can be determined
by integrating the shear flow equations again, this time including the closed shear flow:

A simple check can now be performed to ensure that the shear force in the vertical direction
within the torsion box members equates to the vertical hinge load of 24,940N:

Further, another check can be made to ensure that moments equate to zero at all sections
within the torsion box. Therefore, taking moments about the trailing edge (vertex 1):

With both these checks returning the correct results, the shear flows that have been
calculated are confirmed to be the correct values.

The shear flow in the rib due to hinge loading at each vertex of the torsion box is presented
below. These will be added to the shear flow due to aerodynamic flows in the next section,
where a final shear flow diagram will be produced.

F-37
A negative shear flow indicates shear flow in the clockwise direction. As expected, the shear
flow in the web (vertex points 2,3,4) is acting in the upwards direction (opposing hinge
downwards load).

Figure F-15 - Torsion Box Shear Flow from Chordwise Hinge Loading

F-38
F.2.23 Rib Analysis – Combined Rib Shear Flows
The shear flow from aerodynamic loading can now be combined with the shear flow from
hinge/actuation loading, depicted in Figure F-16 below:

Figure F-16 - Combined Torsion Box Shear Flow from Chordwise Loading

This gives the overall shear flow acting on the critical rib, which will enable the generation
of the shear force and bending moment diagrams. The resulting shear flow at each vertex
of the torsion box is presented below.

A positive shear flow indicates shear flow in the anticlockwise direction.

F-39
 Figure F-17 - Combined Torsion Box Shear Flow from Chordwise Loading

As one would expect, as the applied aerodynamic load (24,940N) is equal to the vertical
hinge reaction (24,940N), the shear flow effects from vertical shear cancel each other out.
Essentially this results in the overall shear flow to equal the difference in torsional reactions
between the two loading cases (which means shear flow is constant throughout the torsion
box, opposing the combined torsion load). The overall shear flow in the torsion box is simply
the combination of the two closed shear flows:

F.2.24 Rib Analysis – Chordwise Shear Force and Bending Moment Plots
To create the chordwise shear force and bending moment diagrams for the critical rib, the
applied loading system (aerodynamic and hinge loads) is combined with the reactive loading
system (torsion box shear flow) as detailed below:

F-40
 Figure F-18 – Applied and Reactive Loading Systems Combined

The resulting shear flow and bending moment diagrams in the chordwise direction are
detailed in Figure F-19and Figure F-20 below:

Chordwise Shear Force - Critical Rib

0% 20% 40% 60% 80% 100%
0

-5000
Shear Force (N)

-10000

-15000

-20000

-25000
Spar Web Location
-30000

-35000

Chord (%)

Figure F-19 - Chordwise Shear Force Diagram - Critical Rib

F-41
 Chordwise Bending Moment- Critical Rib
0% 20% 40% 60% 80% 100%
0
-1000
Bending Moment (Nm)
-2000
-3000
-4000
-5000
-6000
-7000 Spar Web Location

-8000
-9000

Chord (%)

Figure F-20 - Chordwise Bending Moment Diagram - Critical Rib

As can be seen in the diagrams above, the highest shear force is -28,304.9N at the spar
web to rib interface. This maximum shear force must be reacted by the rib web. The largest
bending moment is also at the same chordwise location, with a value of -7,306Nm. The
maximum bending moment must be reacted by the rib flanges and skins (due to the
effectiveness of the skins in reacting bending loads). Thus:

F.2.25 Rib Analysis – Rib Web

Using these maximum shear value determined in Section F.2.24 above, the shear load
distribution on the rib web ( ) can be ascertained. For the following calculations it is
assumed that the shear force is completely reacted by the rib web (conservative).

The second moment of area for the rib web at the forward spar location (critical section
under highest applied shear load) is as per below:

F-42
Maximum shear flow within the rib web can be calculated using the standard engineers
shear formula:

Using this maximum shear load distribution, the laminate strength and strain analysis can
be conducted using COALA. Stress failure criteria are presented as both an interactive
criteria (Tsai-Hill) and one non-interactive failure criteria (max stress/strain). The interactive
failure criteria assessment better reflects the actual stress/strain conditions in the material
due to the interactions between variable direction stresses. The non-interactive criteria
allows the identification of the failure mode.

The maximum strain failure criterion below is based on a maximum allowable strain of
3500 strain. This limit is based off reccomended damage tolerance allowances as
specified in the AVD Fatigue and Damage Tolerance notes [28].

Entering the load intensity magnitudes to COALA, the following failure indices are calculated
and converted to reserve factors:

F-43
Mode of Failure: Transverse Direction Compressive Failure in +45 Plies

The minimum reserve factor for the rib web is as per below:

Shear Buckling of Rib Web:

Web buckling analysis carried out using ESDU 80023 [29] assuming a simply supported
panel. The panel dimensions are taken as the average rib chordwise rib height and the
length equal to the chord length of the torsion box.

Figure F-21 – ESDU 80023 Panel Dimensions & Load Case [29]

The following equations act as inputs for ESDU 80023, Figure 5 [29]:
F-44
Using ESDU 80023 Figure 5 [29], the allowable shear load is:

F.2.26 Rib Analysis – Rib Flange

The rib flange reacts the chordwise bending moment within the torsion box region. As the
sandwich skin has been proven to be fully effective under critical load, it will not buckle.
Therefore the skin also aids in reacting the chordwise bending, along with the rib flange.
This is one of the benefits of a distributed flange loading distribution, as the rib flange will be
smaller than a comparable mass-boom rib flange design.

The second moment of areas about the spanwise axis are calculated below. As the cross
section is symmetrical about this axis, the neutral axis coincides with the geometrical axis:

F-45
As the ribs and sandwich skin inherit differing equivalent membrane stiffness properties, the
flexural stiffness of the combined section is calculated as per below:

The maximum axial load distribution in the rib flange is calculated using the modified
engineers bending theory:

Strength and strain analysis is conducted using COALA (method as per previous examples):

Mode of Failure: Transverse Direction Compressive Failure in +45 Plies

F-46
The minimum reserve factor for the rib flanges is as per below:

Local Compression Buckling of the Rib Flange:

Local rib flange buckling analysis carried out using ESDU 03013 [30]. This analysis method
is applicable to flanges behaving as long flat plates with one side free and the other simply-
supported, subjected to a uniform end-load. For this buckling analysis, the simply supported
edge is assumed as the rib web location. The derivation of the buckling equation is detailed
in ESDU 03013, and so will not be replicated here:

F.2.27 Rib Analysis – Reserve Factor Summary

F-47
F.3 Connection Analysis
This section will include all detailed stressing analysis used to determine the connection
strength between the various Ae-16 composite aileron structural components. There are two
forms of connection types, adhesion and mechanical fastening:

F.3.1 Adhesion Analysis

The Ae-16 aileron design utilises composite adhesion connection techniques in two main
areas: the connection of the spar flanges to the sandwich panel skins, and the connection
of the rib flanges to the sandwich panel skins and spar web.

The bonding material used is Redux 810 [48], which has the following cured shear strength
under the most conservative environmental conditions (test environment of 40° C and a
humidity condition of 60°C-95%RH):

The shear stress at the adhesion boundary between the spar flange and the sandwich skin
panel is calculated below using standard engineers shear theory. A number of the
geometrical values have been calculated earlier in this report (see Section F.2):

Shear stress = =

The reserve factor against static adhesion failure between the spar flange in shear is thus:

The shear stress at the adhesion boundary between the critical rib flange and the sandwich
skin panel is calculated below. A number of the geometrical values have been calculated
earlier in this report (see Section F.2.20). The loads acting on the rib flange are in the
chordwise direction, hence:

F-48
The reserve factor against static adhesion failure between the rib flange in shear is thus:

As these reserve factors for adhesion are large (even with most conservative/harshest
environmental condition chosen) no further durability analysis of the adhesion interface is
deemed necessary.

F.3.2 Mechanical Fastener Analysis – Hinge to Spar Connection

An iterative process has been used to find the ideal sized fasteners to fix the hinge flanges
to the composite spar. Mechanical fastening was selected over adhesion for this interface
as the connection between the hinge and aileron structure is critical to the safe operation of
the control surface (and better aligns with the damage tolerance requirements of CS25.571)

The fasteners selected are Hi-Lok HL11-6, titanium flush shear-head pins. The titanium
material is required to preclude galvanic corrosion between the fastener and composite
material. The shear strength of the fastener is as per below (REF [49] states shear strength
in double shear, therefore the strength of 5381lbf is divided by two):

There are 10 fasteners between the hinge aft flange and the spar web (1 row of 5 on each
hinge flange). These fasteners are spaced at distances of 4D as an industry accepted
spacing to avoid interfastener shearout failure. These fasteners react the accumulated shear
at the critical hinge location (this value has been calculated in previously in Section F.2.5):

F-49
Thus, the shear load applied to each fastener is:

Web fastener shear failure:

The reserve factor for web fastener shear is calculated below. A fitting factor of 1.15 has
been applied as required by CS25.625. Further, MIL-HDBK-5 [50] was investigated to see
if a shear strength correction factor was required for the pin-laminate thickness combination.
For a 3/16" lock bolt within the 3.216mm thick laminate, no shear strength correction factor
was deemed necessary (MIL-HDBK-5 [50], Table 8.1.5(a)).

Web Fastener Bearing Failure:

Bearing failure of the 5mm thick hinge flange material (Ti-4Al-4V) will be assessed as per
below:

(Dry pin ultimate bearing stress, e/D 2.0 - MMPDS Ti-6Al-4V, 0.1875"-2.000", AMS 4911,
A-Basis Values)

The edge margin has been designed to allow the replacement of the fasteners with the
next size up (1/4") fasteners, should a repair be necessary. Hence the edge margin is 2D
of a 1/4" size hole:

The resulting edge margin for 6/32” pins is as per below:

F-50
The bearing stress within the titanium material is as per below:

The reserve factor for bearing failure is as per below. Note, that the ultimate bearing
strength has been factored by 90% to account for a realistic wet pin value, as
recommended by ESDU 91008 [51] Section 1.3.2.2:

Flange Fasteners:

The fasteners which connect the critical hinge to the upper and lower flanges of the spar
must be assessed for joint failure conditions. There are 16 fasteners in total (eight
fasteners on each of the upper and lower flanges). These fasteners react the shear
caused by the bending moment acting on the aileron structure, which must be passed
through the critical actuation hinge. Each actuation hinge reacts half of the hinge moment
(11,300Nm). As such the shear force reacted by the flange fasteners is calculated as per
below:

(on each flange fastener)

Flange fastener shear failure:

The reserve factor for flange fastener shear is calculated below. A fitting factor of 1.15 has
been applied as required by CS25.625. Further, MIL-HDBK-5 [50] was investigated to see
if a shear strength correction factor was required for the pin-laminate thickness

F-51
combination. For a 3/16" lock bolt within the 3.216mm thick laminate, no shear strength
correction factor was deemed necessary (MIL-HDBK-5 [50], Table 8.1.5(a)).

Flange Fastener Bearing Failure:

Bearing failure of the 5mm thick hinge flange material (Ti-4Al-4V) will be assessed as per
below:

(Dry pin ultimate bearing stress, e/D 2.0 - MMPDS Ti-6Al-4V, 0.1875"-2.000", AMS 4911,
A-Basis Values)

Similar to the hinge web flanges, the edge margin has been designed to allow the
replacement of the fasteners with the next size up (1/4") fasteners should a repair be
necessary. Hence the edge margin is 2D of a 1/4" size hole:

(edge margin for 6/32” fasteners)

The bearing stress within the titanium material is as per below:

The reserve factor for bearing failure is as per below. Note, that the ultimate bearing
strength has been factored by 90% to account for a realistic wet pin value, as
recommended by ESDU 91008 [51], Section 1.3.2.2:

F-52
Flange Fastener Tearout Failure:

As the connection at the upper and lower flanges of the hinge have free edges in line with
the shear loading direction, a tearout analysis must be conducted. For this analysis the
shear planes within the hinge flanges is measured from the free edge to the 45 degree
location of the fastener hole (standard industry practice):

(Ultimate shear stress - MMPDS Ti-6Al-4V, 0.1875"-2.000", AMS 4911, A-Basis Values)

The area of two tearout planes is calculated as per below (45 degree location starting point
on fastener hole (sin(45°)=√2/2)):

Figure F-22 - Tearout Planes

F-53
F.4 Lug Analysis
This section will include all detailed stressing analysis used by the author in the hinge sizing
iteration process. Several iterations were performed in order to size the hinges to the
optimum size/strength relationship. The following analysis results relate to the final hinge/lug
sizing iteration, which was sized based on the fatigue and damage tolerance requirements.

It must be noted that the geometry and layout presented in this detailed stress report is for
the final iteration only. The entire detailed stress analysis had to be re-done multiple times
in order to find the optimum solution. Further, at a late stage of the thesis, the fatigue
characteristics of the lugs were found to be the driving factor for lug sizing. As such, the
sizes of the lugs had to be significantly increased to meet acceptable crack growth rates.

F.4.1 Lug Material Properties

The lugs are milled from Ti-6Al-4V titanium. The following material properties for this
material are stated below:

MMPDS [42] Ti-6Al-4V, 0.1875"-2.000", AMS 4911, A-Basis Values:

Ultimate tensile stress

Ultimate tensile stress in transverse direction

Yield tensile stress

Yield tensile stress in transverse direction

Dry pin ultimate stress, e/D 2.0

Dry pin yield stress, e/D 2.0

Ultimate shear stress

F-54
 Strain at ultimate failure

F.4.2 Lug Configuration and Geometry

The critical hinge exists at the second inboard location (second actuation hinge). This hinge
is critical as the local spar height is smaller than that for the inboard hinge, and so the
reaction loads to the actuation force are higher.

The upper connection point is a pivot point made up of three wing side lugs containing
bushings, and two aileron side lugs containing bearings (see Figure 26).

The lower connection point is the actuation connection point, made up of two aileron side
lugs containing bushings, and one central actuation rod end bearing (see Figure 26).

General Size Parameters:

Local spar height at second hinge location

The distance between the upper pivot lug and the lower actuation lug is a key dimension as
it represents the moment arm for the actuation force. This distance also affects the lug
geometry and required bolt/bearing sizing. The distance/moment arm detailed below was
optimise after many iteration steps:

Upper Aileron Lugs at Pivot Connection (Two lugs containing bearings):

Distance from lug centre to lug edge

Lug total width

Lug thickness

Upper Wing Lugs at Pivot Connection (Three lugs containing static bushings):

As the aileron moves relative to the fixed wing side lugs, there must be an allowance gap in
the vertical direction between the two structures. This enables the aileron to move relative
to the wing side lugs without contact under full deflection (detailed below).

F-55
The distance from the lug centre to the lug edge will thus be the same as the aileron side
lug, less the interference allowance gap:

Lower Aileron Lugs at Actuation Connection (Teo lugs containing static bushings):

The thickness is allowed to be lower than the upper pivot lugs,

as the actuation load is lower than the upper reaction/actuation loading.

Central Lower Actuation Rod End Bearing:

The rod end bearing selected for the actuation connection an M81935/4-10 self lubricated
rod end bearing manufactured by RBC aerospace bearings [52]. This rod end bearing has
in internal bolt diameter of 5/8", and the following characteristics:

Ultimate radial load of the bearing

Bolt for Upper and Lower Connections:

S99 bolt, 5/8" D (15.875mm), 154kN single shear strength [53]

This bolt has been chosen through an iterative process for both rotation axis (upper) and
actuation connection point (lower). The bolt diameter is the smallest allowable while
maintaining positive bolt reserve factors within the detailed stress report:

F-56
Bush for Bolt/Lug Interface:

The bush chosen is a 2mm thick, flanged bush meeting MIL-B-81934/2 specifications.

Bearing for upper pivot connection (2 in aileron side lugs):

The bearing has been selected in order to fit the required bolt size (5/8" Diameter), which
itself has been selected by an iterative process to find positive RF values in the detailed
stress report. A self lubricating spherical wide bearing from RBC [54] has been selected
based on the ideal properties suited to aileron application:

Bearing: AS81820 standard (formerly MIL-B-81820) self-lubricated spherical wide bearing

P/N MS14102-10

128500N radial limit load rating

Internal Bolt Diameter= 5/8" (15.875mm),

The final geometry to define is the gap between the inner and outer lugs. Within this gap is
the extended sphere of the bearing and the flange of the mating lug bush. It is important that
there is no gap between the surfaces, as this would lead to lug bending when the bolt is
tightened. As such, the gap is as follows:

F-57
F.4.3 Lug Loading
The loads reacted at the actuation lug are driven from the analysis performed earlier in the
thesis, and re-presented here:

Each actuation hinge reacts half of the total limit hinge moment. The total limit hinge moment
is 11,300Nm (detailed earlier in this report):

The linear actuation force is determined by the moment arm between the pivot lug and
actuation lug (detailed above). This is the linear force applied to the lower lug of the hinge:

The maximum vertical shear force reacted is 25,190N (see Table 5). This is reacted by the
upper pivot lug of the hinge.

The upper pivot lug reacts both the vertical shear load, and the reaction to the linear
actuation force (to attain static equilibrium). As such, an oblique load acts on the upper lug
with the following magnitude and angle is as follows:

F-58
F.4.4 Wing Side Triple Lugs - Upper Pivot Location
For the wing side triple lugs, the lug hole size will be the bushing diameter. The following
ratios are used within the ESDU analysis sheets:

Tensile Rupture Failure - Wing Side Triple Lugs - Upper Pivot Location

Figure F-23 - Tensile Rupture Failure Mode - ESDU 91008 [51]

The first step is to determine elastic stress concentration factor (based off lug geometry) for
axially loaded lug as per ESDU 81006, Figure 2 [55]:

Note that a conservative clearance tolerance of 0.2% has been selected for determination
of the above elastic stress concentration factor.

Next, a value is obtained for the axial tensile rupture factor using ESDU 91008 Figure 1a
[51]:

F-59
The load to cause tensile rupture failure is determined as per ESDU 91008 eq.1.1 [51]
methodology and presented below (factor x allowable x net tensile area)

Shear-Bearing Rupture Failure - Wing Side Triple Lugs - Upper Pivot Location

This failure mode is affected by the combination of shear plane strength and bearing
deformation susceptibility.

Figure F-24 - Shear Tearout Failure Mode - ESDU 91008 [51]

The first step is to determine the shear-bearing reduction factor for the lug from ESDU 91008
figure 2b [51]. The shear bearing reduction factor is dependent on the ratio of the ultimate
bearing strength to the ultimate tensile strength of the material in the weak grain direction
(transverse direction):

Note, that the ultimate bearing strength has been factored by 90% to account for a realistic
wet pin value, as recommended by ESDU 91008 1.3.2.2 [51].

The axial shear bearing rupture factor can now be determined from ESDU 91008 figure 2a
[51] using the shear-bearing reduction factor and the following ratio:

F-60
The load to cause shear-bearing rupture is given by ESDU 91008 eq. 1.2 [51] and detailed
below. (shear-bearing rupture factor x ultimate tensile stress in weak(transverse) grain
direction x bearing area)

Lug permanent deformation failure mode - Wing Side Triple Lugs - Upper Pivot Location

First, the axial permanent deformation factor ( ) needs to be ascertained from ESDU
91008, Figure 3 [51], using the following ratios:

= Minimum load before first failure

The load required to cause an overall lug permanent deformation failure is a follows. The
assessment of the overall lug permanent deformation is related to the minimum of the tensile
rupture mode, and the shear-bearing rupture mode:

F-61
Pure Bearing Permanent Deformation - Wing Side Triple Lugs - Upper Pivot Location

Figure F-25 - Bearing Failure Mode - ESDU 91008 [51]

The load to cause unacceptable bearing permanent deformation of the lug at the lug-to-pin
interface is given by:

Note, that the limit bearing strength above has been factored by 90% to account for a
realistic wet pin value, as recommended by ESDU 91008 1.3.2.2 [51]

Transverse Loading Analysis - Wing Side Triple Lugs - Upper Pivot Location

Transverse loading analysis was carried out using the methodology in ESDU 06021 [56]

Geometrical Factors for analysis: To conduct the analysis in accordance with ESDU 06021
[56], the factors relating to the shape and thickness of the lug first need to be ascertained.
For detailed descriptions of the factors, please refer to ESDU 06021 [56]:

Figure F-26 - Lug Geometry Nomenclature from ESDU 06021 [56]

F-62
The following ratios are used within the curves from ESDU 06021 [56]:

Total strain at 0.2 percent proof stress as defined in

ESDU 06021, Section 1.1 [56]

Ultimate strain divided by strain at 0.2 percent proof stress

From ESDU 06021, Figure 1 [56]:

Then, using ESDU 06021, Figure 2 [56]:

F-63
This is transverse rupture factor for the lug, which will be used to determine maximum
allowable lug load to avoid transverse rupture:

Transverse Loading - Lug Rupture - Wing Side Triple Lugs - Upper Pivot Location

This is the maximum allowable transverse load on the lug to avoid ultimate lug rupture. The
applied transverse load on the wing side pivot lug is the vertical shear load reacted at the
hinge point:

Transverse Loading - Permanent Deformation - Wing Side Triple Lugs - Upper Pivot
Location

This is the allowable limit transverse load on the lug to avoid permanent deformation.

Oblique Loading - Lug Rupture - Wing Side Triple Lugs - Upper Pivot Location

Oblique loading analysis was carried out using the methodology in ESDU 08007 [57]

F-64
Where, is the minimum failure load as calculated eariler in this report (in accordance
with ESDU 91008). is the maximum allowable oblique load before lug rupture
occurs:

Oblique Loading - Permanent Deformation - Wing Side Triple Lugs - Upper Pivot Location

Oblique loading permanent deformation analysis was carried out using the methodology in
ESDU 08007 [57].

The maximum allowable limit load for an oblique loading is similar to the analysis above,
however, the ultimate loads have been substituted by the proof loads (U to P): has
been substituted by , and substituted by :

Where, and have been calculated earlier in this report

F.4.5 Bolt Analysis

Bolt Shear - Upper Pivot Location

The bolt shear strength stated earlier in this report is the bolt’s single shear strength. The
upper pivot configuration includes three wing side lugs, and two aileron side lugs; thus, there
are a total of four shear planes. To compare with the single shear strength value of the bolt,
the ultimate loading in the equation below is divided by four (shear is shared by the four
shear planes). Further, a fitting factor of 1.15 has been applied as required by CS25.625 [1].
MIL-HDBK-5 [50] was used to confirm that a shear strength correction was not required for
this combination of bolt and lug thickness.

F-65
Bolt Bending - Upper Pivot Location

ESDU Struct 01.06.01 [58] is used to define the maximum allowable bending moment.

First, the form factor is ascertained from ESDU Struct 01.06.01, which, for a circular section
the value is 1.535:

Next, the section modulus (Z) is calculated for a circular section:

The maximum moment can now be determined using the following formula obtained from
ESDU Struct 01.06.01. The maximum moment is essentially the point at which yielding will
occur at the outer fibres of the bolt:

F-66
Bolt Bending Assuming Uniform Loading Across Thickness of Lug - Upper Pivot Location

Figure F-27 - Lug Uniform Loading Detailing Moment Arm (b) - ESDU 91008 [51]

Assuming a uniform loading across this thickness of the lugs, the maximum moment arm of
the bolt can be represented by the following equation:

Note that the maximum moment arm exists between the outer wing lug and the aileron lug
(rather than between the middle wing lug and the aileron lug).

The applied moment acting on the bolt can be represented by ESDU 91008 eq. 1.5 [51]. As
there are four shear planes, the ultimate force is divided by four:

While this reserve factor meets the failure requirements (RF > 1), the author wished to
perform further analysis on the bolt bending failure condition in order to gain a better
reflection of the actual strength of the bolt. As such, the analysis in the next section

F-67
includes the effect of the load peaking nearer to the shear faces (the analysis above
assumed uniform load distribution across the lug width):

Pin bending analysis assuming load peaking between pin and lug - Upper Pivot Location

To increase the accuracy of the reserve factor for bolt bending the following analysis will
take into account the tendency of the bearing stresses due to pin loading to peak near the
shear faces of the inner lug thus reducing the moment on the pin. This analysis more
accurately reflects the actual loading condition of lug/bolt interface.

Figure F-28 – Comparison Between Loading Distributions - ESDU 91008 [51]

First the load peaking effect factor ( ) is determined from ESDU 91008, figure 4 [51], and
the following ratio:

F-68
The equivalent moment arm using the peak loading distribution is represented as the
following (following ESDU 91008, eq. 1.7 [51])

The applied moment acting on the bolt, using a peak loading distribution, can be represented
by the following equation. Essentially the ultimate load is divided by four (as there are four
shear planes) then multiplied by the peak loading distribution moment arm:

As expected, this more accurate analysis has yielded a higher RF, which better reflects the
actual loading conditions.

F.4.6 Upper Pivot Connection Point - Aileron Side (two lugs)

The layout of the upper pivot connection point is comprised of two lugs on the aileron side
containing bearings, surrounded by the three wing side lugs (which are bushed). The
important geometrical ratios to be used in the lug analysis are stated below:

F-69
Tensile Rupture Failure - Upper Pivot Connection Point - Aileron Side (two lugs)

Figure F-29 - Tensile Rupture Failure Mode - ESDU 91008 [51]

The first step is to determine elastic stress concentration factor (based off lug geometry) for
axially loaded lug as per ESDU 81006, Figure 2 [55]:

Note that a conservative clearance tolerance of 0.2% has been selected for determination
of the above elastic stress concentration factor

Load to cause tensile rupture failure is as per ESDU 91008 eq.1.1 [51] and presented below
(factor x allowable x net tensile area)

Thus, the reserve factor for tensile rupture in one of the two aileron lugs at the actuation
connection point is as follows. Note, that the allowable is doubled as there are two aileron
side lugs sharing the ultimate load.

F-70
Shear-Bearing Rupture Failure - Upper Pivot Connection Point - Aileron Side (two lugs)

This failure mode is affected by the combination of shear plane strength and bearing
deformation susceptibility.

Figure F-30 - Shear Tearout Failure Mode - ESDU 91008 [51]

The first step is to determine the shear-bearing reduction factor for the lug from ESDU 91008
figure 2b [51]. The shear bearing reduction factor is dependent on the ratio of the ultimate
bearing strength to the ultimate tensile strength of the material in the weak grain direction
(transverse direction):

Note, that the ultimate bearing strength has been factored by 90% to account for a realistic
wet pin value, as recommended by ESDU 91008 1.3.2.2 [51].

The axial shear bearing rupture factor can now be determined from ESDU 91008 figure 2a
[51] using the shear-bearing reduction factor and the following ratio:

The load to cause shear-bearing rupture is calculated using ESDU 91008 eq. 1.2 [51] and
detailed below. (shear-bearing rupture factor x ultimate tensile stress in weak(transverse)
grain direction x bearing area):

F-71
Lug permanent deformation failure mode - Upper Pivot Connection Point - Aileron Side
(two lugs)

First, the axial permanent deformation factor ( ) needs to be ascertained from ESDU
91008, Figure 3 [51], using the following ratios:

= Minimum load before first failure

The load required to cause an overall lug permanent deformation failure is a follows. The
assessment of the overall lug permanent deformation is related to the minimum of the tensile
rupture mode, and the shear-bearing rupture mode.

Pure Bearing Permanent Deformation - Upper Pivot Connection Point - Aileron Side (two
lugs)

Figure F-31 - Bearing Failure Mode - ESDU 91008 [51]

The load to cause unacceptable bearing permanent deformation of the lug at the lug-to-
bearing interface is given by:

F-72
Note, that the limit bearing strength above has been factored by 90% to account for a
realistic wet pin value, as recommended by ESDU 91008 1.3.2.2 [51]

F.4.7 Lower Actuation Connection Point - Aileron Side (two lugs)

The layout of the lower actuation connection point consists of two outer aileron lugs, and
one central rod end bearing from the aileron actuator. The aileron side lugs are static (with
an installed bush in each). The important geometrical ratios to be used in the lug analysis
are stated below:

Tensile Rupture Failure - Lower Actuation Connection Point - Aileron Side (two lugs)

The first step is to determine elastic stress concentration factor (based off lug geometry) for
axially loaded lug as per ESDU 81006, Figure 2 [55]:

Note that a conservative clearance tolerance of 0.2% has been selected for determination
of the above elastic stress concentration factor.

Obtain a value for the axial tensile rupture factor using ESDU 91008 figure 1a [51]:

Load to cause tensile rupture failure is as per ESDU 91008 eq.1.1 [51] and presented below
(factor x allowable x net tensile area)

F-73
Thus, the reserve factor for tensile rupture in one of the two aileron lugs at the actuation
connection point is as follows. Note, that the allowable is doubled as there are two aileron
side lugs sharing the ultimate load.

Shear-Bearing Rupture Failure - Lower Actuation Connection Point - Aileron Side (two lugs)

This failure mode is affected by the combination of shear plane strength and bearing
deformation susceptibility.

The first step is to determine the shear-bearing reduction factor for the lug from ESDU 91008
figure 2b [51]. The shear bearing reduction factor is dependent on the ratio of the ultimate
bearing strength to the ultimate tensile strength of the material in the weak grain direction
(transverse direction).

Note, that the ultimate bearing strength has been factored by 90% to account for a realistic
wet pin value, as recommended by ESDU 91008 1.3.2.2 [51]

The axial shear bearing rupture factor can now be determined from ESDU 91008 figure 2a
[51]using the shear-bearing reduction factor and the following ratio:

The load to cause shear-bearing rupture is given by ESDU 91008 eq. 1.2 [51] and detailed
below. (shear-bearing rupture factor x ultimate tensile stress in weak(transverse) grain
direction x bearing area)

F-74
Lug permanent deformation failure mode - Lower Actuation Connection Point - Aileron
Side (two lugs)

First, the axial permanent deformation factor ( ) needs to be ascertained from ESDU
91008, figure 3 [51], using the following ratios:

= Minimum load before first failure

The load required to cause an overall lug permanent deformation failure is a follows. The
assessment of the overall lug permanent deformation is related to the minimum of the tensile
rupture mode, and the shear-bearing rupture mode.

Pure Bearing Permanent Deformation - Lower Actuation Connection Point - Aileron Side
(two lugs)

The load to cause unacceptable bearing permanent deformation of the lug at the lug-to-pin
interface is given by:

F-75
Note, that the limit bearing strength above has been factored by 90% to account for a
realistic wet pin value, as recommended by ESDU 91008 1.3.2.2 [51]

F.4.8 Lower Actuation Connection Point - Actuator Single Lug Analysis

For the actuator central rod end bearing, the ultimate reserve factor is simply found by
comparing the rated ultimate load (from M81935/4 Data Sheet [52]) with the applied
ultimate loading. Note that a reduction factor of 0.8 has been applied as required by
CS25.623 [1] due to the clearance fit of the bolt and presence of vibrations.

F.4.9 Hinge Reserve Factor Summary

Tensile Ultimate Axial Rupture - Wing Lug

Shear-Bearing Ultimate Axial Rupture - Wing Lug

Proof Bearing Failure - Wing Lug

Proof Axial Deformation Failure - Wing Lug

Shear-Bearing Ultimate Transverse Rupture- Wing Lug

Shear-Bearing Ultimate Oblique Rupture - Wing Lug

Proof Oblique Bearing Failure - Wing Lug

F-76
 Proof Bearing Failure Transverse - Wing Lug

Bolt Shear Failure

Bolt Bending Failure

Tensile Ultimate Axial Rupture - Aileron Lug

Shear-Bearing Ultimate Axial Rupture - Aileron Lug

Proof Permanent Deformation Axial - Aileron Lug

Proof Bearing Failure - Aileron Lug

Tensile Ultimate Axial Rupture - Aileron Lug

Shear-Bearing Ultimate Axial Rupture - Aileron Lug

Proof Permanent Deformation Axial - Aileron Lug

Proof Bearing Failure - Aileron Lug

Actuator Rod End Bearing - Ultimate Load

----------------------------------------End of Detailed Stressing Report-------------------------------------

F-77
Appendix G – Aeroelastic Analysis
G.1 – Torsional Stiffness
A detailed analysis of the torsional stiffness of the aileron was accomplished by
meeting the requirements of the British Civil Airworthiness Requirements
(BCAR), subsection K3-9 [32]. The following analysis is similar to that conducted
for during the preliminary sizing, however includes added criteria such as
enhanced aileron geometry, design speeds, and will use the actual laminate
shear stiffness values.

The BCAR subsection K, calculates the required torsional stiffness for an aileron
using the following variables and equation as follows:

Note that as the BCAR utilises imperial units, the torsional criteria

( ) specified within the BCAR as 0.016 , has been converted to metric

units (0.064 ).

This torsional stiffness requirement for the Ae-16 ailerons can now be compared
to the actual torsional stiffness of the aileron torsion box.

The first step is to determine the actual torsional stiffness of the aileron torsion
box using the Bredt-Batho theory for twisting of thin walled sections (shown
below):

G-1
 ( ) ( ) eqn.(G-79)
=
4 ( ) ( )

( ) ( ) eqn.(G-80)
∴ =
4 ( ) ( )

Where ( ) is the local section perimeter of the structural box, ( ) is the local
section area of the enclosed structural box, is the laminate shear modulus,
( ) is the accumulated torque at the spanwise location. This can be re-
arranged to isolate the torsional stiffness of the structure (Torsional stiffness is
the torsion load required to twist the section by one radian):

( ) 4 ( ) eqn.(G-81)
= =
( )

( )

As the BCAR measures the torsional stiffness of the aileron between 10% and
90% span, the integration of the geometrical values above is conducted within
with these limits. Integration of this between 10% and 90% span yields:

. ( )
.
= 1647.067 .
( )

As the torsion box is comprised of the sandwich panels and the spar web, both
components have differing shear modulus ( ) and thicknesses. As such, the
average of the ( ) multiple is found as per below:

Where the thicknesses and shear moduli are as per those figures detailed in
Table 15 and Appendix F.2.1.

Thus, the torsional stiffness for the aileron is:

eqn.(G-82)

G-2
As the actual torsional stiffness is far higher than the required torsional stiffness
(2.055E5 >> 223), the Ae-16 aileron design easily meets the torsional stiffness
requirements.

Under limit aerodynamic loading, the maximum accumulated torque is 6094Nm

acting on the structure outboard of hinge number two (see Figure 10). By
rearranging eqn.(G-81), the twist angle of the outboard end of the aileron under
this accumulated torsion can be determined:

eqn.(G-83)

G.2 – Aileron Divergence Speed

A rudimentary divergence speed calculation has been performed to ensure that
aileron divergence does not occur within regulatory set design speeds. Control
surface divergence must be prevented as per CS-25.629, which requires no
divergence occurs at speeds up to 1.2 × , which for the Ae-16 is 1.2 ×
192.15 = 230.58

The divergence analysis methodology as specified in AVD-0441 [33] and

Babister [34] is followed as per below:

eqn.(G-84)

Where, the hinge moment co-efficient ( ) is as per that calculated in Appendix

B.1.8. The torsional stiffness used is that of the structural torsion stiffness
calculated in the preceding section:

As the Ae-16 aileron divergence speed is higher than that required by CS25
(545.6 230.6 ), divergence is not an issue for the Ae-16 aileron design.

G-3
Appendix H – Fatigue and Damage Tolerance Analysis
The process used to produce the aileron fatigue spectrum is as per Figure H-1:

Determine Probability
Determine Aileron Determine Critical
of Reaching Aileron
Usage Frequency per Aileron Lug Gross
Deflection Angles for
Flight Hour Stress at Limit Loading
Ae-16

Determine Occurance
Assess Fatigue
of Each Stress Level Calulate Lug Stress At
Spectrum using
Per Flight Hour Each Deflection Angle
AFGROW
(Spectrum Creation)

Figure H-1 - Aileron Fatigue Spectrum Process

Aileron usage per flight hour assumes that the aileron is activated once every
ten seconds during climb and decent phases of flight, and once every two
minutes during cruise (detailed in Table H-1 below).

Table H-1 - Aileron Usage Assumptions

Flight Stage Aileron Use (assumed) Aileron Input per hour

Climb Once every 10 seconds 360
Cruise Once per 2 minutes 30
Decent Once every 10 seconds 360

These assumptions are sufficient for the preliminary fatigue assessment, but will
require validation during the flight testing stage.

The flight profile and mission mix obtained from the Aircraft Performance Team
includes the amount of hours spent within each stage of flight, and the length of
each type of flight as per Table H-2 below. The total hours spent within each
flight stage is also provided which allows the calculation of aileron usage for
each flight type over the 70,000 lifetime of the aircraft:

H-1
 Table H-2 - Mission Mix Aileron Usage

% %
% time No of hours in this
time time Total Aileron
% of Length spent sortie
Sortie spent spent Cycles per
use of sortie in configuration for
in in 70,000 FH
Decent 70,000hr Life Limit
Climb Cruise
% Hrs % % % Hrs Cycles
1 25% 1.5 23 27 50 17500 4,720,738
2 40% 2.8 13 60 27 28000 4,555,427
3 20% 5.2 7 78 14 14000 1,417,928
4 10% 4.0 9 72 19 7000 859,223
5 5% 6.5 6 83 12 3500 306,776
100% 70,000 11,860,092

For example, to calculate lifetime aileron cycles due to Sortie 1 :

Aggregating all of the aileron usage values across all flight types leads to a total
of 11,860,092 cycles over the lifetime of the aircraft.

A.R.B White Note 96 [35] was used to determine the probability of the aileron
reaching a certain percentage of its maximum deflection angle at any one
deflection occurrence. The probability data extracted from A.R.B White Note 96
[35] is detailed in Table H-3 below:

H-2
 Table H-3 - Ae-16 Aileron Deflection Spectrum

% Aileron Probability of Exceedances Occurrences Cycles per Cycles per

Deflection Exceedance per 70,000FH per 70,000FH 70,000FH 2,500 FH
0 1 11860092 59300 - -
0.1 0.995 11800792 59301 29651 1059
0.2 0.99 11741491 1067408 533704 19061
0.3 0.9 10674083 4269633 2134817 76243
0.4 0.54 6404450 3558028 1779014 63536
0.5 0.24 2846422 1719713 859857 30709
0.6 0.095 1126709 593005 296503 10589
0.7 0.045 533704 296502 148251 5295
0.8 0.02 237202 183832 91916 3283
0.9 0.0045 53370 53311 26656 952
1 0.000005 59 59 30 1

The exceedances per 70,000FH were calculated by multiplying the total aileron
usage value (11,860,092 activations per 70,000FH) by the probability of
exceedance value. For example, the aileron exceeds the 10% deflection point
11,860,092 x 0.995 = 11,800,792 times within the aircraft’s life.

The occurrences per 70,000FH were calculated by subtracting all exceedances

past the next magnitude of aileron deflection. For example, the occurrences of a
deflection between 10% and 20% deflection is 11,800,792-11,741,491 = 59,301

As each aileron is deflected upwards and downwards by the same number of

occurrences, one cycle can represent both directions of travel. Therefore cycles
per 70,000 flight hours is simply the occurrence value divided by two.

The cycles were divided by the lowest common denominator of all occurrences
(30) in order to reduce the amount of occurrences to be randomised. This gave
cycles per 2,500FH. The spectrum is unaffected by this division, and AFGROW
will repeat the spectrum (regardless of the number occurrences) until failure.

The limit loads within each aileron lug under maximum aileron deflection were
extracted from Appendix F.4.3. There are two pivot lugs on the aileron side of
the hinge, and so each shares half of the limit load. The limit loading on the
pivot lugs is 81,882N, comprised of both vertical shear and actuation reaction
components. While this loading is oblique in direction, for the fatigue

H-3
assessment it is assumed to act axially (conservative as now the loading is
perpendicular to the critical crack growth direction). As there are two aileron
side lugs, each lug reacts 40,941N. The loading, geometry, and calculated
gross stress is shown in Table H-4 below:

Table H-4 - Single Aileron Pivot Lug Load, Geometry, and Gross Stress

Limit Load 40,941 N

Width Aileron Lug 50.30 mm
Lug Hole 30.16 mm
Thickness Aileron Lug 14.40 mm
Gross Stress 56.52 Mpa

The gross stress value of 56.52MPa equates to a maximum aileron deflection in

the critical downwards direction. For simplicity this stress is also assumed to
apply for an upwards aileron deflection (conservative as an upwards aileron
deflection relieves some of the aerodynamic pressure from wing incidence).

As discussed in the main body of this report, it is assumed that the gross stress
at any aileron deflection angle inherits a linear relationship to the maximum
deflection/stress values (i.e. a 10% aileron deflection results in 10% of the
maximum gross stress value). By calculating the gross stress at each deflection
angle, the fatigue spectrum is derived and detailed in Table H-5 below.

Table H-5 - Ae-16 Aileron Pivot Lug Fatigue Spectrum

Stress (Mpa) Stress (Ksi)

% deflection Max Min Max Min Cycles per 2,500 FH
0% - - - - -
10% 5.7 -5.7 0.8 -0.8 1059
20% 11.3 -11.3 1.6 -1.6 19061
30% 17.0 -17.0 2.5 -2.5 76243
40% 22.6 -22.6 3.3 -3.3 63536
50% 28.3 -28.3 4.1 -4.1 30709
60% 33.9 -33.9 4.9 -4.9 10589
70% 39.6 -39.6 5.7 -5.7 5295
80% 45.2 -45.2 6.6 -6.6 3283
90% 50.9 -50.9 7.4 -7.4 952
100% 56.5 -56.5 8.2 -8.2 1

There are 210,728 total aileron cycles per 2,500FH. A randomisation function
within excel was used to place each of these cycles randomly within a 2,500FH

H-4
block. This allows the formation of an AFGROW input file that was in a cycle-by-
cycle format, made up of 210,728 individual cycles at their respective stress
levels. AFGROW repeats these 210,728 cycles until final failure occurs. As
discussed in the main body of this report, the author used a cycle-by-cycle
fatigue analysis as this better reflects the actual loading the aircraft
experiences, and produces a smoother, more realistic crack growth curve.

The AFGROW model inputs are as per below, results are presented within the
main body of this report.

Crack Model: 1030 - Single Corner Crack at Hole -

Standard Solution
English Units [ Length(in), Stress(Ksi)]
Initial crack depth (a) : 0.0500
Initial surface crack length (c): 0.0500
Thickness : 0.570
Width : 1.980
Hole Diameter: 1.187
Retardation: WILLENBORG
Shut-off ratio : 2.800
Adjust Yield Zone Size for Compressive Cycles = Yes
Harter T-Method crack growth rate approach used
Material: Ti-6-4 AMS 4911G ANNEALED
Plane strain fracture toughness: 60
Yield stress: 130
Cycle by cycle beta and spectrum calculation

H-5
Appendix I – Finite Element Analysis
I.1 – Global Model Parameters
This section will detail how the model was created, assessed, and validated within
ABAQUS. The purpose of this section is to provide confidence that the author
has generated a sufficiently accurate model, and has thought about the numerous
variables that can adversely affect FEA results.

I.1.1 Global FEA Model - General Description

The primary goal of the global FEA model was to assess torsional
stiffness/deflection, and to validate maximum ply strain values calculated during
detailed stressing. The surface model was imported directly from CATIA with the
accurate aileron twist, camber, and taper. Flanges were omitted from the
importation process in order to avoid free edges within the FEA model. Free
edges under in plane bending require a high mesh density to provide accurate
results (hence computationally expensive). The author modelled the flanges
within ABAQUS using beam/stiffener elements attached to the web/skin
extremities, and ensured accurate membrane stiffness properties and flange
profile orientations.

I.1.2 Global FEA Model – Analysis Type

ABAQUS/Standard analysis has been utilised due to the linearity of the geometric
and material properties of the Ae-16 aileron model. As the composite material
exhibits linear stiffness until ultimate failure, there is no advantage to using the
incremental procedure provided by an explicit analysis procedure
(ABAQUS/Standard is more efficient for solving smooth nonlinear problems [41]).
Further, a dynamic loading analysis of the aileron is not required due to the
loading and stressing analysis assuming an instantaneous aileron deflection
(required by CS25.349 [1]).

I-1
I.1.3 Global FEA Model – Element Type
Thick conventional shell elements were used to model the skin and webs of the
aileron structure. Thick shell modelling was used to account for the very compliant
interior layers of the sandwich composite skins (which exhibit very low transverse
shear stiffness) and to model the influence of shear flexibility in laminated
composite shell models [41]. The linear, finite-membrane-strain, fully integrated,
quadrilateral shell element (S4) was used, as in plane bending is expected in
some sections of the model (recommended by [41])

First-order, shear-deformable beam elements (B31) was chosen to model the

spar and rib flanges.

I.1.4 Global FEA Model – Material/Section Properties

The layup sequences were accurately modelled for all composite shell structure,
using the laminae material properties from material data sheets. For the beam
elements, the orthotropic properties of the laminates were entered directly (found
from COALA analysis). Section integration was selected to be conducted prior to
the finite analysis process in order to increase computational performance (this
is allowable due to the linearity of the composite material until failure). Three
through thickness integration/section points within each laminae were used to
enable accurate analysis of individual plies under load.

I.1.5 Global FEA Model – Interaction Properties

Modelling of the web to skin attachments was accomplished through merging of
parts to ensure shared edge nodes with common nodal DOF. The flanges were
modelled as stiffeners with shared nodes with the shell edges (both for spar and
rib flanges, and for the skin flange which is adhered to the spar flanges)

I.1.6 . Global FEA Model – Mesh

An optimised quadratic dominated mesh with a global size of 32.5mm was used
to provide accurate, yet efficient, modelling results. The spar web and skin
surfaces inherit 100% quad shaped mesh, with similar sizing and shape across
the entire structure. See validation Section I.1.9 below for mesh check process.

I-2
I.1.7 . Global FEA Model – Loads
Both spanwise and chordwise aerodynamic pressure distributions were modelled
using a complex polynomial field function. The pressure distribution thus
accurately matches that calculated determined in Section 2.6 (see Figure 9 and
Figure 5 within the main body of this report). The pressure was modelled as
2/3rds acting on the lower surface, and 1/3rd acting on the lower surface as
recommended by Howe [9].

I.1.8 . Global FEA Model – Boundary Conditions

Reference points were placed at the pivot location for each of the five hinge
positions. All reference points were constrained in the chordwise plane directions.
The two actuation hinges were further constrained in the spanwise direction to
reflect the fixed nature of the spherical bearings (required to ensure actuation
reference points to not shift axially). The actuation reaction torque was modelled
by a combination of a constraining rotation in the chordwise plane at one of the
actuation hinges, and a reaction torque applied to the other actuation hinge
equating to exactly half of the reaction torque. This method ensures that each
actuation hinge reacts exactly half of the hinge torque.

The nodes of the spar web and flange that attach to the hinges were fixed relative
to each other using a rigid body tie constraint. These rigid body nodes were then
tied to the reference points where the boundary conditions were specified. This
allows the outboard sections to rotate about the pivoting reference point. See
Figure I-1 below:

I-3
 Figure I-1 - FEA Boundary Conditions

I.1.9 . Global FEA Model – Validation Checks

It was important that several validation checks were performed to ensure that the
FEA model returned accurate results (and to ensure it was not affected by poor
model setup or incorrect assumptions).

The mesh was checked for size and shape appropriateness using relative size
and skew filters within ABAQUS. It is important particularly for quad shaped mesh
elements that the shape is not overly extended, as this greatly distorts the
integration accuracy and returned results. Mesh size appropriateness was
checked by comparing the output stresses at the integration points to the stresses
reported at the nodal points (accomplished by generating a field report with both
outputs). The stress output at the nodes was extrapolated from the element
integration points and averaged over all elements containing a given node. If the
difference between the nodal stresses and stresses at the integration points was
large enough to be of concern, this indicated that the mesh may have been too
coarse. For the final mesh size chosen by the author (global seed of 32mm) the
difference between the two stress values for all points was acceptably low.

A mesh convergence check was accomplished for the model to ensure that the
maximum strain values were not influenced by singularities, and to ensure the
mesh was sufficiently sized to return accurate strain results. The results of the
converge check is detailed in Table I-1 below:

I-4
 Table I-1 - Mesh Convergence Check

Global Seed Size Maximum Strain Maximum Deflection

64mm 810μ 7.35mm
32mm 1087μ 7.329mm
16mm 1190μ 7.321mm

As can be seen, the maximum strain values are reducing in step magnitude as
the mesh is refined. This indicates an acceptable approach towards an asymptote
and provides confidence that the results were not affected by singularity errors.

The reaction forces and reaction moments at the hinge positons were checked
by requesting a field output report results file, tailored to include unique nodal
outputs for the hinge reference points. The total vertical reaction force across all
hinges was confirmed to match that of limit and ultimate loading for each version
of analysis (65,094N and 97,641N respectively). The moment reaction at the
actuation hinge that was fixed in chord wise rotation was confirmed to react
exactly half of the hinge moment (5,629Nm limit and 8,444Nm ultimate). The
loading on the other actuation hinge was also set to these torque values so that
total reaction torque matched the total hinge moment value.

By monitoring the submitted model during ABAQUS processing, the author

confirmed that zero warnings or errors were produced from the model (which
provides confidence that the model was prepared correctly).

The final validation check for the FEA model was to compare the results to those
determined during detailed stressing and torsional stiffness analyses. These
validations and discussion have been included in the main body of the report, and
so will not be repeated here.

I-5
I.1.10 . Detailed FEA Model – Hinge
Detailed FEA modelling of the critical lug was accomplished by using similar
methodology to that used for the global FEA modelling. The primary differences
were:

-Continuum elements were used in the form of second order tetrahedral

elements. The mesh was intentionally restricted to Tet based controls due to the
complex contours of the hinge (tetrahedral elements are less sensitive to element
distortion [41]). Second-order Tet elements were used to avoid the excessive
stiffness characteristics inherent with first order tetrahedral elements. The main
disadvantage of utilising a tetrahedral dominated mesh is the computational cost
compared to brick elements.

-The model was simplified to one side of the critical lug (as the hinge is roughly
symmetrical, this leads to increased computational efficiencies.

-Boundary constraints were applied to reference points at the centre of each lug.
The constraints were tailored to reflect the linear restraint of the actuation lug,
and the chordwise plane constraints of the pivot lug. The nodes around the inner
surface of each lug were tied to the movement of the reference points.

-A reference point was created at 184mm aft of the hinge chord line, and the
aerodynamic load was applied to this point (this point lies at the centre of pressure
of the aerodynamic load (see Figure F-10). The nodes of the hinge/spar interface
were tied to the movement of this loading reference point.

All validation steps conducted for the global model were repeated for the detailed
hinge model in order to ensure model accuracy (mesh, convergence, RF, and
detailed stress checks).

I-6