Structural Analysis of A Jet Trainer Cockpit

STRUCTURAL ANALYSIS OF A JET TRAINER COCKPIT
A THESIS SUBMITTED TO
THE GRADUATE SCHOOL OF NATURAL AND APPLIED SCIENCES
OF
MIDDLE EAST TECHNICAL UNIVERSITY
BY
MUHİTTİN NAMİ ALTUĞ
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

FOR
THE DEGREE OF MASTER OF SCIENCE
IN
AEROSPACE ENGINEERING
FEBRUARY 2012
Approval of the thesis:
submitted by MUHİTTİN NAMİ ALTUĞ in partial fulfillment of the requirements

for the degree of Master of Science in Aerospace Engineering Department,
Middle East Technical University by,
Prof. Dr. Canan Özgen _____________________

Dean, Graduate School of Natural and Applied Sciences
Prof. Dr. Ozan Tekinalp _____________________

Head of Department, Aerospace Engineering
Assist. Prof. Dr. Melin Şahin _____________________

Supervisor, Aerospace Engineering Dept., METU
Examining Committee Members:
Prof. Dr. Yavuz YAMAN _____________________

Aerospace Engineering Dept., METU
Assist. Prof. Dr. Melin ŞAHİN _____________________

Assist. Prof. Dr. Demirkan ÇÖKER _____________________

Assist. Prof. Dr. Ercan GÜRSES _____________________

Dr. Muvaffak HASAN _____________________

Chief Engineer, TAI
Date: 10.02.2012
I hereby declare that all information in this document has been obtained and
presented in accordance with academic rules and ethical conduct. I also declare
that, as required by these rules and conduct, I have fully cited and referenced
all material and results that are not original to this work.
Name, Last Name : Muhittin Nami Altuğ
Signature :
iii
ABSTRACT
Altuğ, Muhittin Nami

M.Sc., Department of Aerospace Engineering
Supervisor: Assist. Prof. Dr. Melin Şahin
February 2012, Pages 128
This thesis presents structural analysis of a cockpit of a jet trainer type aircraft and
the correlation studies performed by using ground pressurisation test results. For this
purpose, first the response of the complex integrated fuselage structure is
investigated under the complex type cabin pressure load. Then, cockpit part of the
fuselage structure is modelled using commercial finite element software
MSC/PATRAN® and MSC/NASTRAN®. The finite element model (FEM) of the
cockpit structure is improved by the examination of the ground pressurisation test
data and is finalised after achieving a good correlation between the finite element
analysis (FEA) and the test results. This final form of the FEM of the cockpit
structure serving as a benchmark is proved to be reliable for any future
modifications.
Keywords: Structural Analysis, Jet Trainer Aircraft, Cockpit, Finite Element

Modelling and Analysis, Experimental Correlation
iv
ÖZ
BİR JET EĞİTİM UÇAĞI KOKPİTİNİN YAPISAL ANALİZLERİ
Altuğ, Muhittin Nami

Yüksek Lisans, Havacılık ve Uzay Mühendisliği Bölümü
Tez Yöneticisi : Yrd. Doç. Dr. Melin Şahin
Şubat 2012, 128 sayfa
Bu tez, bir jet eğitim uçağı kokpitinin yapısal analizlerini ve yer basınçlandırma test
sonuçları ile korelasyonu çalışmalarını sunmaktadır. Ulaşılmak istenen nihai amaca
yönelik, ilk olarak bütünleşik gövde yapısının karmaşık bir yük türü olan kabin
basıncı yükü altında davranışı incelenmiştir. Sonra, gövdenin kokpite ait bölgesi
MSC/PATRAN® ve MSC/NASTRAN® ticari sonlu elemanlar yazılımları
kullanılarak modellenmiştir. Kokpite ait bu sonlu elemanlar modeli yer
basınçlandırma test verileri incelenerek geliştirilmiş ve sonlu elemanlar analizleri ile
test sonuçları arasında iyi bir korelasyon sağlanmasıyla da son halini almıştır. Bu
çalışma ile ayrıca kokpit modelinin elde edilmiliş bu son halinin ileriki çalışmalara
güvenilir bir referans teşkil edeceği de ispatlanmıştır.
Anahtar Kelimeler: Yapısal Analiz, Jet Eğitim Uçağı, Kokpit, Sonlu Elemanlar
Model ve Analizi, Deneysel Korelasyon
v
to my Mami
vi
ACKNOWLEDGMENTS
I would like to express my gratitude to my supervisor Assist Prof. Dr. Melin

Şahin for his guidance and patience throughout this study.
I must also express my appreciation to my superiors at TAI, Dr. Gürsel

Erarslanoğlu and Dr. Muvaffak Hasan for providing me the necessary data and their
guidance.
I want to thank my colleagues Ali Baki Uygur, Ömer Faruk Türkmen and
Selçuk Topçu for their help and perfect performance during the ground pressurisation
test.
I would also like to thank my colleagues Enver Özakkaş, Bilgin Çelik,

Muhsin Öcal, Zuhal Gökbulut, Abdulkadir Çekiç, Emre Ünay, Engin Kahraman,
Fatih Mutlu Karadal, Derya Gürak, Ahmet Ufuk Yavuz, Mehmet Efruz Yalçın and
Erdoğan Tolga İnsuyu for giving me the strength to complete this work.
Last, but not least, I would like to express my deepest thanks to my sister
Rihal and her husband Martin for their endless support and Eda for her support at
every stage of this study, understanding and love.
vii
TABLE OF CONTENTS
ABSTRACT ................................................................................................................ iv
ÖZ ................................................................................................................................ v
ACKNOWLEDGMENTS ......................................................................................... vii
TABLE OF CONTENTS .......................................................................................... viii
LIST OF TABLES ....................................................................................................... x
LIST OF FIGURES .................................................................................................... xi
CHAPTERS
1. INTRODUCTION.................................................................................................... 1
1.1 Motivations of the Study .............................................................................. 1

1.2 Objectives of the Study ................................................................................ 2
1.3 Literature Survey.......................................................................................... 2
1.4 Limitations of the Study ............................................................................. 15
1.5 Contents of the Study ................................................................................. 17
2.FINITE ELEMENT MODELING AND ANALYSIS OF COCKPIT STRUCTURE

.................................................................................................................................... 18
2.1 Introduction ................................................................................................ 18

2.2 Structural Model of the Cockpit ................................................................. 18
2.3 Finite Element Model of the Cockpit ......................................................... 24
2.4 Checks for Finite Element Model .............................................................. 44
2.5 Finite Element Analysis ............................................................................. 50
2.6 Summary .................................................................................................... 54
viii
3. GROUND PRESSURISATION TEST .................................................................. 55
3.1 Introduction ................................................................................................ 55

3.2 Test Configuration ..................................................................................... 55
3.3 Sensor Types .............................................................................................. 56
3.4 Criteria for the Selection of the Sensors .................................................... 58
3.5 Installations of the Sensors ........................................................................ 59
3.6 Data Acquisition System ............................................................................ 67
3.7 Test Steps ................................................................................................... 68
3.8 Test Results ................................................................................................ 69
3.9 Summary .................................................................................................... 81
4. CORRELATION STUDIES BETWEEN FINITE ELEMENT ANALYSIS AND

TEST RESULTS ........................................................................................................ 82
4.1 Introduction ................................................................................................ 82

4.2 Deformation Checks .................................................................................. 83
4.3 Methodology to Interpret the Results......................................................... 86
4.4 Checks for FEA and Test Results Correlation ........................................... 91
4.5 Actions taken to Update the FEM ............................................................ 103
4.6 Updated Results ....................................................................................... 110
4.7 Summary .................................................................................................. 122
5. CONCLUSION .................................................................................................... 123
5.1 General Conclusions ................................................................................ 123

5.2 Recommendations for the Future Work ................................................... 125
REFERENCES......................................................................................................... 126
ix
LIST OF TABLES
TABLES
Table 2.2.1: The number and location of the structures in the cockpit. ..................... 23
Table 2.3.1: The number of the elements and grid point used in the model .............. 27
Table 2.3.2: Mechanical properties of the materials .................................................. 43
Table 2.3.3: Web thicknesses used in the model ....................................................... 43
Table 2.4.1: Element Quality criteria ......................................................................... 45
Table 2.4.2: Natural modes analysis results ............................................................... 46
Table 2.5.1: Canopy Hook Loads-Left Side .............................................................. 52
Table 3.3.1: Strain Gauge Specifications [29], [30] .................................................. 57
Table 3.5.1: The number and the locations of the strain gauges ................................ 66
Table 3.9.1: The maximum strain and stress values for the strain gauges ................. 80
Table 4.4.1: The comparison between test and FEA results ...................................... 92
Table 4.4.2: The percentage of difference between test and FEM results ................. 93
Table 4.6.1: The comparison between test and updated FEM results...................... 110
Table 4.6.2: The percentage of difference between test and updated FEM results . 111
Table 4.6.3: The comparison between updated FEM, original FEM and test results
.................................................................................................................................. 112
x
LIST OF FIGURES
FIGURES
Figure 1.3.1: NASA Dryden’s T-38 trainer aircraft in flight over Cuddeback Dry
Lake in Southern California [2] ................................................................................... 4
Figure 1.3.2: Semi-monocoque construction [5].......................................................... 5
Figure 1.3.3: Alpha Jet-E, Canopy and Windshield structures [10] ............................ 7
Figure 1.3.4: Aermacchi M-346 cutaway drawing [18]............................................. 10
Figure 1.3.5: Integrated FEM for Boeing 767-400ER [17] ....................................... 11
Figure 1.3.6: Strain gauge types [25] ......................................................................... 13
Figure 2.2.1: General view of cockpit structure ......................................................... 19
Figure 2.2.2: Descriptions of the cockpit structures .................................................. 20
Figure 2.2.3: Bulkhead & Frame numbering ............................................................. 22
Figure 2.3.1: General view of cockpit FEM .............................................................. 26
Figure 2.3.2: Detailed FEM of the upper longeron .................................................... 28
Figure 2.3.3: Detailed FEM of the lower longeron .................................................... 29
Figure 2.3.4: Detailed FEM of the frames and intercostals ....................................... 30
Figure 2.3.5: Detailed FEM of the frame ................................................................... 31
Figure 2.3.6: General view of bulkhead FEM............................................................ 32
Figure 2.3.7: Detailed FEM of the bulkheads ............................................................ 33
Figure 2.3.8: Detailed FEM of the skin ..................................................................... 34
Figure 2.3.9: Detailed FEM of the front and rear floors ............................................ 35
Figure 2.3.10: Detailed FEM of the upper and lower deck ........................................ 36
Figure 2.3.11: Detailed FEM of the windshield......................................................... 37
Figure 2.3.12: Detailed FEM of the horse shoe ......................................................... 38
Figure 2.3.13: Detailed FEM of the sheet support ..................................................... 39
Figure 2.3.14: Detailed FEM of the front tie bar ....................................................... 40
Figure 2.3.15: Detailed FEM of the rear tie bar ......................................................... 41
xi
Figure 2.3.16: Assembly of the structures ................................................................. 42
Figure 2.4.1: Quadrangle element quality check parameters [16] ............................. 45
Figure 2.4.2: Mode shape-Tx ..................................................................................... 47
Figure 2.4.3: Mode shape-Ty ..................................................................................... 47
Figure 2.4.4: Mode shape-Tz ..................................................................................... 48
Figure 2.4.5: Mode shape-Rx ..................................................................................... 48
Figure 2.4.6: Mode shape-Ry ..................................................................................... 49
Figure 2.4.7: Mode shape-Rz ..................................................................................... 49
Figure 2.5.1: Pressure surfaces in the model.............................................................. 51
Figure 2.5.2: Shell element normals .......................................................................... 51
Figure 2.5.3: Schematic representation of canopy hook loads................................... 52
Figure 2.5.4: Displacement boundary conditions ...................................................... 53
Figure 3.2.1: General Test Configuration .................................................................. 56
Figure 3.3.1: Linear and Rosette Gauges with their channel numbers [29], [30] ...... 57
Figure 3.3.2: Pressure transducer used in the study [31] ........................................... 58
Figure 3.5.1: Linear strain gauges on longeron inner and outer flanges .................... 60
Figure 3.5.2: Linear strain gauge on longeron inner flange at the near of its cut-out 61
Figure 3.5.3: Linear strain gauge on frame inner cap ................................................ 61
Figure 3.5.4: Linear strain gauges on longeron web .................................................. 62
Figure 3.5.5: Rosette strain gauge on skin panel ....................................................... 62
Figure 3.5.6: Strain gauges on frames and skin panel................................................ 63
Figure 3.5.7: Linear strain gauge on the tie-bar ......................................................... 63
Figure 3.5.8: Linear strain gauge on the sheet support .............................................. 64
Figure 3.5.9: Linear strain gauge on the bulkhead upper cap .................................... 64
Figure 3.5.10: General view of the strain gauge installation on the cockpit .............. 65
Figure 3.9.1: Longeron web strain-pressure plot ....................................................... 70
Figure 3.9.2: Longeron inner flange strain-pressure plot........................................... 71
Figure 3.9.3: Longeron inner flange non-linear strain behaviours............................. 71
Figure 3.9.4: Longeron outer flange strain-pressure plot........................................... 72
Figure 3.9.5: Skin strain-pressure plot ....................................................................... 72
Figure 3.9.6: Frame inner cap strain-pressure plot .................................................... 73
Figure 3.9.7: Frame inner cap non-linear strain behaviours ...................................... 74
xii
Figure 3.9.8: Bulkhead upper cap strain-pressure plot .............................................. 74
Figure 3.9.9: Tie bars strain-pressure plot ................................................................. 75
Figure 3.9.10: Sheet support strain-pressure plot ...................................................... 75
Figure 3.9.11: Orientation of the rectangular rosette ................................................. 77
Figure 3.9.12: Mohr’s circle for the rectangular rosette ............................................ 77
Figure 4.2.1: Deformation of the upper longerons under cabin pressure-Top view .. 83
Figure 4.2.2: Deformation of the frames under cabin pressure.................................. 84
Figure 4.2.3: Deformation of the frames under cabin pressure (close-up view)........ 85
Figure 4.3.1: Element coordinate systems on 2-D elements ...................................... 87
Figure 4.3.2: Stress plot that has no element to element averaging at the nodes. ...... 88
Figure 4.3.3: The layer selection (Z1/Z2) for 2-D skin elements .............................. 89
Figure 4.3.4: The stress recovery points on beam cross section ................................ 90
Figure 4.3.5: SG installation on beam element .......................................................... 90
Figure 4.4.1: Longeron stress variation at F18 station-Top view .............................. 95
Figure 4.4.2: Longeron inner flange stress distribution with test results ................... 96
Figure 4.4.3: Longeron outer flange stress distribution with test results ................... 97
Figure 4.4.4: F2-Frame inner cap stress distribution with test results ....................... 98
Figure 4.4.5: F-3 Frame inner cap stress distribution with test results ...................... 98
Figure 4.4.8: F6-Frame inner cap stress distribution with test results ..................... 100
Figure 4.4.9: F12-Frame inner cap stress distribution with test results ................... 100
Figure 4.4.10: F13-Frame inner cap stress distribution with test results ................. 101
Figure 4.5.1: Detailed FEM of the original sheet support-a .................................... 104
Figure 4.5.2: Detailed FEM of the updated sheet support-b .................................... 104
Figure 4.5.3: Detailed FEM of the front tie bar-a .................................................... 106
Figure 4.5.4: Detailed FEM of the updated front tie bar-b ...................................... 106
Figure 4.5.5: Detailed FEM of the rear tie bar-a ...................................................... 107
Figure 4.5.6: Detailed FEM of the updated rear tie bar-b ........................................ 107
Figure 4.5.7: Detailed FEM of the canopy support fittings and the drive shaft....... 109
xiii
Figure 4.5.8: Connections of the canopy support fittings and the drive shaft ......... 109
Figure 4.6.1: Updated longeron inner flange stress distribution .............................. 114
Figure 4.6.2: Updated longeron outer flange stress distribution .............................. 115
Figure 4.6.3: Updated F2-Frame inner cap stress distribution ................................. 116
Figure 4.6.8: Updated F12-Frame inner cap stress distribution ............................... 120
Figure 4.6.9: Updated F13-Frame inner cap stress distribution ............................... 120
Figure 4.6.10: Updated F14-Frame inner cap stress distribution ............................. 121
Figure 4.6.11: Updated F17-Frame inner cap stress distribution ............................. 121
xiv
CHAPTER 1
INTRODUCTION
1.1 Motivations of the Study
Because of very high procurement and operating costs for the new-generation jet
trainer aircrafts, modernization of the older aircrafts has become inevitable for many
countries. Also, aircrafts are forced to be modernised for efficiency and safety
reasons according to the constantly updated international aviation rules.
Modernization programs for the jet trainer aircrafts are generally carried out in two
main branches: modernization of avionic systems and structural modifications.
Avionics in the aircrafts are installed especially in the cockpit region in order to
improve the human-machine interface for the pilots. Installations of the avionic
equipments in the cockpit are performed with support structures and sometimes these
supports may change the main load path of the fuselage structure. In addition, for the
older aircrafts, aging problem of the metal fuselage structure has become a dangerous
situation over the years. Structural modification of the fuselage is generally the most
economical solution to extend the life of the aircraft. For the non-manufacturer
countries that have purchased jet trainer aircraft from another country, load path
information on the aircraft is very critical at the beginning of the structural
modification. Under the operational static loads, each sub-structure of the integrated
fuselage structure shares the loads according to their design. Especially under
complex loads, for example cabin pressure, predicting these load distributions among
1
the structures is a very hard task. Because under cabin pressure, the responses of the
sub-structures of the fuselage not only depend on their own elastic properties, but
also strictly depend on boundary conditions, i.e. their form of integration. Under the
light of these requirements, fuselage structure modernization in a jet trainer aircraft is
a must for the non-manufacturer countries, hence, there is an immense need to
construct an experimentally validated high-fidelity finite element models.
1.2 Objectives of the Study
The objectives of this study can be listed as follows:
• Investigating the load carrying/transfer mechanism of the cockpit structure

• Having a detailed FEM of the cockpit structure
• Investigating the complex type cabin pressure load
• Having the response of the complex integrated fuselage structure
• Obtaining a methodology for strain gauge (SG) testing
• Performing a correlation study
• Validating FEM via experimental test results
1.3 Literature Survey
The importance of air dominance in the defence of countries is increasing. This is

only possible with air force equipped with the most modern aircrafts and highly-
qualified personnel. A combat pilot training is a laborious as well as a costly job.
Technological developments and major advances in warfare aircrafts also bring up
the necessity of the development of jet trainer aircrafts and the systems for the pilot
training.
Today, in many countries, training aircrafts are about to complete their lives or do
not meet the requirements of the day. The difficulties in operation and maintenance
2
of the current training aircrafts are effective in forcing the need for innovation and
development in this field.
Especially together with advances in the field of electronics, digital displays and
easy-use control panels are integrated to the last generation jet aircrafts. Thus, the
possibility of collection of more sensitive data has emerged. Using this information
during a high speed manoeuvre gives rise to the search for a more ergonomic
environment for the pilot. Avionic modernizations for the jet trainer aircrafts,
therefore, become inevitable.
Today in the world, there are different kinds of jet trainer aircrafts used by many
countries. The famous company Northrop Grumman produced 1,187 T-38s between
1959 and 1972, the year in which the production program ended. Today, from those
aircrafts, nearly half of them are still in service by U.S. Air Force, U.S. Navy, NASA
(Figure 1.3.1) and air forces around the world. Since 1961, more than 70,000 Air
Force pilots have been trained in the T-38. The average T-38 has flown 15,000 hours,
and the high-time aircraft has flown 19,000 hours. T-38 is maintained by the
production of replacement wings and new structural components to make the service
life of the platform longer. T-38s are currently going through structural and avionics
programs (T-38C) to extend their service life to 2020 [1].
3
Figure 1.3.1: NASA Dryden’s T-38 trainer aircraft in flight over Cuddeback Dry
Lake in Southern California [2]
Another jet trainer, namely Dassault/Dornier Alpha Jet-E has been sold to a number
of countries in Africa and in the Middle East and this jet trainer aircraft modernised
to Alpha Jet-3 (Lancier). In this aircraft, multi-functional controls and a glass cockpit
that will train pilots in the use of navigation and attack systems of the latest and
future generation fighter aircraft are fitted [3].
BAE Hawk, another jet trainer, is used in a wide range of air forces for training
purposes. Hawk is modernised lots of times and there are lots of variants of it. The
Advanced Jet Trainer, the latest version of Hawk trainers, is equipped with glass
cockpit, inertial navigation, and other improvements [4].
Also, Northrop Grumman F-5B, Aermacchi MB-339, L-39 Albatross, G-4 Super
Galeb are some of the examples for jet trainers, having ongoing modernization
programs today.
4
Fuselage Structure
The fuselage is the main structure or body of the aircraft. The function of an aircraft
fuselage is to provide support for the following structures: structure for wings and
tail, structure that contains the cockpit for the pilot and structure that allows aircraft
to carry cargo, passengers, and other equipments. In single-engine aircraft, it also
hoses the power plant. One type of fuselage structure is the monocoque fuselage that
uses formers, frame assemblies and bulkheads to give shape to the fuselage, and it
relies on the skin to carry the primary stresses. A very crucial problem related to
monocoque construction is sustaining enough strength while keeping the weight
within allowable limits. To overcome this problem, a modification, namely semi-
monocoque construction, was developed. In addition to formers, frame assemblies,
and bulkheads, the semi-monocoque construction has the skin reinforced by
longitudinal members. Primary bending loads are taken by the longerons. They are
supplemented by other longitudinal members, called stringers. Bulkheads, frames
and formers are the vertical structural members of the fuselage. These vertical
members are located at intervals to carry concentrated loads and at points where
fittings are used to attach other units, such as the wing, power plant, stabilizers, etc.
[5]. The members of the semi-monocoque construction are shown in Figure 1.3.2.
Figure 1.3.2: Semi-monocoque construction [5]
5
Fuselage structures are subject to various types of loads. There are differences in the
generation of these loads. Aerodynamic forces on the fuselage skin are relatively
low; on the other hand, the fuselage supports large concentrated loads resulted from
the attachments such as wing and landing gear and it carries payloads, which may
cause large inertia forces. Furthermore, for aircrafts designed for high altitude flight,
the fuselage structure must resist internal pressure [6].
The jet fuselage geometry is composed of three parts: a tapered nose section, forward
fuselage that involves cockpit and rear fuselage [7].
Canopy and Windshield
The need for visibility for the pilot of an aircraft and for protecting himself against
the wind stream has placed emphasis on the design of canopy and windshield. The
visibility is provided by the transparent materials which are recently made up of
composite materials. The shape of the canopy is based on streamlining requirements
regarding aerodynamic efficiency. The development of aircrafts with pressurised
cabins introduced complications. The need for high mechanical strength without
adding excessive weight to withstand the cabin pressures resulted in decreased
window areas and in curved surfaces to obtain strength without excessive weight [9].
Canopy and windshield on the aircraft is shown in Figure 1.3.3.
6
Canopies
Windshield
Figure 1.3.3: Alpha Jet-E, Canopy and Windshield structures [10]
Cabin Pressurisation
Inside a pressurised cabin, people can be transported comfortably and safely for long
duration, especially if the cabin altitude is maintained at 8.000 [ft], or below, where
the use of oxygen equipment is not required. Pressurised air is pumped into fuselage
by cabin superchargers which release a relatively constant volume of air at all
altitudes up to a designed maximum. From the fuselage the device called outflow
valve releases the air. Since the superchargers provide a constant inflow of air to the
pressurised area, the outflow valve, by regulating the air exit, is the major controlling
element in the pressurisation system. The degree of pressurisation and, therefore, the
operating altitude of the aircraft are limited by several critical design factors.
Primarily the fuselage is designed to withstand a particular maximum cabin
differential pressure. Cabin differential pressure is the ratio between inside and
outside air pressures and is a measure of the internal stress on the fuselage skin. If the
differential pressure becomes too high, structural damage to the fuselage may occur
[5].
7
Finite Element Theory
Classical analytical methods consider a differential element and develop the

governing equations, usually in the form of partial differential equations. When
applied to real-life problems, it is often difficult to obtain an exact solution to these
equations considering complex geometry and boundary conditions. The finite
element method (FEM) can be defined simply as a method of finding approximate
solutions for partial differential equations [11].
FEM requires a problem defined in geometrical domain to be subdivided into a finite

number of smaller regions (a mesh). These regions are connected at points called
nodes. Element behaviour is approximated in terms of nodal variables called degrees
of freedom. Elements are assembled considering loading and boundary conditions.
The governing equations in the FEM are integrated over each finite element and the
solution assembled over the entire problem domain. Consequently, a set of finite
linear equations in terms of a set of unknown parameters is obtained over each
element. Solution of these equations is performed using linear algebra techniques
[12], [13].
In engineering problems, unknowns are infinite in a continuum. The finite element

procedure reduces such unknowns to a finite number by expressing the unknown
field variables in terms of assumed approximating functions (interpolating
functions/shape functions) within each element. The approximating functions are
defined in terms of field variables of nodes. Thus in the FEM, the unknowns are the
field variables of the nodal points. Once these are found, the field variables at any
point can be found by using interpolation functions [14].
8
Steps in finite element analysis can be summarised as follows:
• Selecting suitable field variables and the elements.

• Discretising the continuum
• Selecting interpolation functions
• Finding the element properties/stiffness
• Assembling element properties/global stiffness to get global properties.
• Imposing the boundary conditions and loading conditions.
• Solving the system equations to get the nodal unknowns.
• Making the additional calculations to get the required values [14], [15].
FEM Construction
For many structural Finite Element Analysis (FEA) applications, there are always
decisions made by the design or analytical engineer on just how to simplify a real
structure into a simulation model.
Successful simplification depends on;

• Understanding the physics of the problem
• Understanding the behaviour of the elements
• Selecting the correct element, the number of elements and their distribution
• Critically evaluating the results and making modification in the conceptual
model to improve the accuracy [16].
While simplifying a structure into a model, one can use coarse, fine and very fine
meshes. Each type of these meshes serves for different analysis purposes. The
ultimate decision of combining these different types is complicated and subjective
[17]. There are many sub-components of aircraft. Since they are interacting with
each other, they form a complex environment causing boundaries and loading
conditions on the aircraft to be complex too. Guessing load paths that result from
9
these conditions can be a hard task. Figure 1.3.4 shows the complexity of airframe
structure.
Figure 1.3.4: Aermacchi M-346 cutaway drawing [18]
In making an analysis of large structural components such as a wing, fuselage etc.,

modelling with simple (low order) elements is most desirable. These simple models
can provide reasonably accurate information about the overall load paths, and the
simplicity of the elements allows easier interpretation of the results. They are also
ideal for parametric studies in preliminary design and optimisation. The use of higher
order elements is appropriate while making a detailed analysis of local areas, such as
a plate with cut-out or a crack or local buckling of a panel etc. [17].
Modelling of the complex structural components with low order elements as

mentioned above is the essence of Global FEM. The main objective of the Global
FEM is to reproduce the global stiffness and overall load paths. Global FEM contains
enough detail to accurately describe the structural behaviour. It includes the major
10
structural elements. The Global FEM does not use detailed models for components
and it is a collection of several individual models (Figure 1.3.5).
Figure 1.3.5: Integrated FEM for Boeing 767-400ER [17]
Advantages:
• Easier to find errors.
• Use simple elements.
• Use simple modelling concepts.
• Keep the model size small [19].
Since the 1990s, all primary structures of commercial airplanes like the B777 and
A340 are certified using such FEM analysis [20]
A given structure which is too complex for classical analysis can be solved by FEM
in a powerful way for determining stresses and deflections. The method appears
complex due to the fact that thousands of elements or members of an airframe
11
structure have each its own set of equations. Because of the very large number of
equations and corresponding data involved, finite element method is only possible
when performed by computer [21].
Many commercial programs use finite element analysis methods, to name a few,
ANSYS® [22], ABAQUS® [23], MSC.Patran® /MSC.Nastran® [24]. In this thesis
MSC.Patran®/ MSC.Nastran® are used since it is well appreciated in aerospace
industry and can be well-suited for aerospace applications.
Structural Testing
Despite encouraging results from simulation and experimental modelling, structural

testing is still a valuable tool in the industrial development of product and process.
Through testing, the response of the structure under applied loads (force, pressure,
temperature, shock, vibration and other loading conditions) is determined. Its success
depends on careful choice of testing method, instrumentation, data acquisition, and
allocation of resources [25].
Structural Testing Activities
Structural testing has three major steps. The first one is the planning of the test. At
this phase, the requirements and the type of test are specified. Examples of different
types can be full scale, coupon tests and quality assurance tests. The second step is
the preparation. This step involves specifying load type, magnitude of loading and
making the loading equipments and subsystems ready for the test. Also instrument
and subsystem calibrations are made at this stage. The final step is the execution.
This refers to obtaining data from transducers and processing data which involves
data transmission, signal conditioning etc. [25]
12
Strain Gauges as Measuring Devices
The purpose of a strain gauge is to convert the physical changes occurring in a

structure due to applied loading into an electrical change, usually by altering voltage
or current [25]. Three types of strain gauges are used on airframe fatigue and static
tests for measuring strain. These are axial, shear, and rosette gauges, as shown in
Figure 1.3.6.
Figure 1.3.6: Strain gauge types [25]
In some cases a single strain gauge will provide sufficient information. In many
cases, however, two or more strain gauges will be necessary to supply the
information necessary to calculate the stresses. Two-dimensional stress state can
often be used as an adequate model for actual stress distributions. In this case the
positions of the strain gauges must be carefully chosen, taking into account the
stresses which are of interest. If the principal directions are known, strains of interest
can be measured by a 00/900 rosette. If the principal directions are not known, three
armed rosettes will be required [27].
13
Temperature Compensation
Some errors may occur during the measurement. One of the sources of errors can be
environmental factors, for example. These factors may be temperature, humidity,
corrosive atmospheres, electromagnetic noise, etc. [26].
If there is no temperature compensation, during a test under temperature effects,

output signals from strain gauges are not related only to strains that are expected to
be generated by the applied load. One should also consider the temperature effect in
the test results. To suppress temperature effects, a dummy gauge should be placed on
a non-deforming area on the structure to be tested. As it is normally impossible to
find a non-deforming area, the gauge is usually bonded on a small sheet of a material
corresponding to that of the component. This sheet is attached to the structure as
close as possible to the active gauge. It is recommended to use active and dummy
gauges from the same batch in order to obtain the same tolerances for the gauge
factor and to keep temperature effects as low as possible [27].
Correlation
The certification rule for the validation of the FEM that is to be used for the
structural analysis is stated as follows below;
“Analyses including finite element models used in place of tests must be

demonstrated to be reliable for the structure under evaluation and the load levels
that have to be covered. This would normally be provided by correlation with
experimental results on the same structure or through comparison with other known
and accepted methods and results or through a combination of both” [28]
14
1.4 Limitations of the Study
In this study, only the cockpit part of the jet trainer fuselage is investigated.
In the static analysis, cabin pressure load is applied to the model. The model is
solved by the algorithm of linear static analysis. Assumptions of the analysis are the
following: displacements are small, stiffness matrix and boundary conditions do not
change and displacements are directly proportional to the loads.
In the detailed FEM of the cockpit structure, canopy was not modelled due to the
lack of information about the canopy hook mechanism, which is responsible for
locking the canopy to the longerons. Under internal pressure, at the cockpit region,
modelling of the canopy becomes unnecessary when the hook mechanism
information is missing. Because the level and the type of the load transferred from
canopy to the longerons depend on the stiffness and behaviour of the mechanism.
Instead of modelling, canopy-hook loads were gathered from the manufacturer
company.
Opposed to the canopy, windshield structure is modelled except its transparent part.
The transparent structure is not modelled because a material property of that part is
unknown. In this situation, the moments on the longerons exerted by the windshield
pressure loads has to be neglected.
Another assumption is concerned with a plastic sealant which is located between the
canopy and the longerons. The working principle of these plastic sealants is to stick
the canopy to the fuselage when air pressure, namely sealant pressure, is given inside
the sealants. So under the cabin pressure, this simple sticking mechanism prevents
the pressure leakage from the cabin. In detailed FEM, it is not concerned due to
unknown contact loads occurring between canopy and the longerons.
15
In the real structure of the model, there is a canopy drive mechanism which is
responsible for opening and closing the canopy. In the mechanism there is a drive
shaft, which lays between the bushings of the canopy support fittings. These fittings
are located either side of the aircraft. In the study, canopy support fittings and drive
shaft are not modelled. It is thought that modelling these structures does not have a
very large impact on the load path of the cockpit under cabin pressure. Also
modelling the fittings brings an additional run time to the FEM.
In addition, in the FEM, the primary structures are modelled according to Global
FEM approach. Fasteners and fittings are not modelled for assembling the primary
structures. Because there are lots of different types of fasteners in the aircraft,
modelling the fasteners with assigning each of their stiffness is time-consuming.
However, at some local points where necessary, rigid-type elements are used for
connection. Also cut-outs, except longeron, are not modelled based on the same
Global FEM approach.
Strain gauge installations are bounded on the aircraft by the limitations of the
minimum space required for the installation. The reason is that structure to be
modified is already assembled and the environmental condition for the labour is not
the same with the condition during the production phase. For the manufacturer
company, it is easier to install the strain gauges on the desired parts before the
assembly.
In static test, ground pressurisation is applied into the cockpit. Also, there is no
temperature compensation used during the test because the test is performed
approximately at room temperature.
16
1.5 Contents of the Study
This study is composed of five chapters. The first chapter is the introduction which
involves literature survey that gives background information for the main topics of
the study. Limitations of the study are also included in this chapter. The second
chapter is about the modelling and the analysis parts of the study. In this particular
chapter, finite element modelling techniques are presented in detail. The third chapter
is concerned with ground pressurisation test. The procedure of the test is mainly
described and the results are presented and discussed. The fourth chapter is about the
correlation between the finite element analysis and the test results. The final chapter
presents the concluding remarks and recommendations for future work.
17
CHAPTER 2
FINITE ELEMENT MODELING AND ANALYSIS OF

COCKPIT STRUCTURE
2.1 Introduction
In this section, finite element model of a jet trainer type cockpit structure is
constructed and static analysis is performed within the framework of finite element
analysis techniques. Firstly, cockpit sub-structures and their structural functions are
described. Secondly, finite element model construction and model verification
methods are presented. Then, finite element analysis is performed by applying the
loads and boundary conditions specified to the model.
2.2 Structural Model of the Cockpit
The cockpit structure consists of a typical semi-monocoque structure. The primary

structures of the cockpit such as longerons, frames, bulkheads and skins are
assembled for resisting the cockpit to the manoeuvre, aerodynamic and internal
pressure loads. As a general feature of a jet trainer aircraft, there are two cockpits in
the structural model which are; a student (front) cockpit and an instructor (rear)
cockpit. In the model, the right and the left sides of the cockpit regions are
18
structurally and geometrically symmetric. General view of the cockpit structure is
shown in Figure 2.2.1.
Figure 2.2.1: General view of cockpit structure
The modelled structure is composed of upper and lower longerons locating through
the cockpits, side skins extending between upper and lower longerons and frames
which are symmetrically placed on the two sides of the cockpit. There are two types
of frames in the model that are called continuous and discontinuous frames.
Continuous frames are formed with side frames and bottom frames lying between the
lower longerons. Side and bottom frames are attached to the lower longerons.
Discontinuous frames consist of only the side frames. Between some of the frames
there are also intercostals. Also there are bulkheads, floors and decks to complete the
primary structures of the model. Bulkheads have stiffeners with side, upper and
lower caps to reinforce their main webs. In addition to the stiffeners and caps, seating
rails are modelled on the bulkheads which are located at the back of the pilot’s
ejection seat. Floor webs are reinforced with the help of floor beams and webs of the
19
decks are also reinforced with stiffeners. Because structural analysis will be done
under specified load condition, potentially critical structures that can significantly
affect the validation of the study are added to the model. These are the tie bars, sheet
support, windshield and horse shoe which are connected between two-sides of the
cockpit. Normally, it is expected that canopies which are connected to the upper
longerons and canopy support fittings are comprised in the complete cockpit model.
However, canopies and canopy support fittings are not modelled in this study. Figure
2.2.2 shows the description and the placements of the structures in the cockpit.
Figure 2.2.2: Descriptions of the cockpit structures
In the assembly, each primary structure has specific functions. The upper longeron
mainly carries the axial loads from primary fuselage bending. In addition to this, it is
subjected to lateral bending induced by cockpit and canopy pressures and
aerodynamic forces. The supporting structure for the upper longeron under the lateral
loads is provided by the bulkheads and tension ties. The upper longeron also
transfers the vertical components of the windshield and canopy lock loads to the
20
adjacent frames. The lower longeron carries only axial load due to primary fuselage
bending. The side skin panels carry fuselage shear flows and transfer the differential
air pressures to the frames and bulkheads. The floor is subjected to cockpit pressure
and shear flows due to fuselage net side loads. The frames react the external and
internal pressure loads and distribute the net loads to the fuselage side skin and floor.
The decks support the aft bulkhead of the cockpit and it carries a portion of the net
side shear load of the fuselage. Horse shoe is one of the supporting member of the
canopies and sheet support which lays between the upper longerons share the side
loads of the longeron.
The functionalities of the modelled structures are provided with proper and sufficient
attachments between the structures. Attachments are made with using fasteners,
fittings or splices. In this study, the real structures of these kinds of attachments are
not modelled. Instead of this, the functions of the attachments are simulated with
some techniques, described in detail in this chapter.
The horizontal locations of the any structure in the cockpit are defined by taking the
bulkhead and frame locations as the reference. The numbering of the bulkheads and
frames is made sequentially from front through the back of the cockpit. Figure 2.2.3
shows the bulkhead and frame numbering in the cockpit. Here, B and F stand for
bulkhead and frame respectively.
21
Figure 2.2.3: Bulkhead & Frame numbering
Side frames, intercostals and longerons are located symmetrically and the numbers of
these structures are equal on both sides of the cockpit. The number and the locations
of the structures involved in the model are given in Table 2.2.1 below.
22
Table 2.2.1: The number and location of the structures in the cockpit.
Part Location Number Part Location Number

Frame (side) F1 2 Bulkhead Web B1 1
Frame (side) F5 2 Bulkhead Caps B1 4
Frame (side) F9 2 Bulkhead Stiffeners B1 9
Frame (side) F13 2 Skin B1-B4 1
Frame (side) F14 2 Upper Longeron B1-B4 2
Frame (side) F15 2 Lower Longeron B1-B4 2
Frame (side) F16 2 Front Floor Web B1-B2 1
Frame (side) F17 2 Front Floor Beams B1-B2 2
Frame (side) F18 2 Rear Floor Web B2-B3 1
Frame (side) F19 2 Rear Floor Beams B2-B3 15
Frame (side) F20 2 Upper Deck Web B3-B4 1
Frame (bottom) F1 1 Upper Deck Stiffeners B3-B4 3
Frame (bottom) F2 1 Lower Deck Web B3-B4 1
Frame (bottom) F4 1 Lower Deck Stiffeners B3-B4 7
Frame (bottom) F6 1 Windshield B1-F5 1
Frame (bottom) F7 1 Horse Shoe F12-F13 1
Frame (bottom) F8 1 Intercostal B1-F1 2
Frame (bottom) F10 1 Intercostal F2-F4 2
Frame (bottom) F11 1 Intercostal F10-F11 2
Frame (bottom) F12 1 Seating Rail B2 1
Frame (bottom) F13 1 Seating Rail B3 1
Frame (bottom) F16 1 Sheet Support F1 1
Frame (bottom) F19 1 Tie-Bar F3 1
Frame (bottom) F20 1 Tie-Bar F12 1
TOTAL NUMBER OF THE STRUCTURES = 146
23
2.3 Finite Element Model of the Cockpit
For the structural analysis purposes, finite element modelling (FEM) techniques are
utilized in order to obtain mathematical model of the real structures. However, true
simulation firstly depends on having a strong knowledge on the physical behaviour
of the problem. Then, simplification begins with selecting the correct elements for
the structures. As the complexity of the model increases, decision for combining
these simplified elements becomes a difficult task.
In this study, the cockpit structure is modelled by using Global FEM approach.
Global FEM contains enough detail to accurately describe the structural behaviour of
the large models, in this case, a cockpit. The main objective of this type of modelling
is to obtain global stiffness and overall load paths by using low order elements. The
cockpit structure has lots of sub-structures in it. However, only the primary structures
are modelled. Each primary component is simplified according to their structural
missions in the cockpit. Based on this approach, fasteners, fittings and splices are not
modelled at the connection regions. However, rigid type elements are used at some
local points where necessary. Also cut-outs on the primary structures, except
longeron cut-outs, are not modelled for the simplification reasons.
In model, canopy is not modelled due to the lack of information about the canopy
hook mechanism. Cabin pressure reactions of the canopy are transmitted by the
canopy hooks through the upper longerons of the cockpit. Because the hook
mechanism information is missing, modelling of the canopy becomes unnecessary.
Also, plastic sealants between the canopy and the upper longerons are not modelled
due to the hardly predictable contact loads. In addition, canopy support fittings and
drive shaft are not modelled due to the reasons that modelling these structures does
not affect the load path significantly and it brings an additional run time to the FEM.
24
Selection of the Elements
In order to satisfy the designated strength to weight ratio, aircraft structures are
generally formed with combination of thin panels with their longitudinal and
transverse stiffeners. These stiffeners provide out of plane resistance for the panels.
For the frames, bulkheads, floor and deck, in this study, the idealization is made such
that webs of the structures are modelled with 2-D CQUAD4 (quadrangle) and
CTRIA3 (triangle) elements and stiffeners are modelled by using CBEAM (1-D)
elements. In this manner, web is modelled to resist mainly the shear load while beam
elements resist mainly the bending and also shear loads with webs. In stiffened
panels, to obtain a correct bending load distribution, offsets are given to the bar
element neutral axis according to the thicknesses of the bars and the panels. During
the design phases, conservative approach could be made with assigning CROD
element properties to the frame, floor and deck stiffeners. In that way, initial sizing
of a shear resistance web could be made. Also, during the simplification of the
stiffeners, CROD element could be sufficient instead of using CBEAM element
according to the load path. However in this study, because of the reason that the
investigation is made for the existing structure, CBEAM elements are preferred to
CROD elements. CTRIA3 elements are used only regions where CQUAD4 element
generation is painful. Because it has higher stiffness property than CQUAD4
element, CTRIA3 element is not preferred for the simulation of the panel structures.
The structures that have major thickness variation, such as machined parts, are
modelled by using CTETRA (3-D) solid elements. The reason is that the chamfers or
flanges on the machined parts provide an additional stiffness to the main body of the
part. In order to include this effect in the model, CTETRA element is used due to its
better topological properties.
Also in the model, multi point constraint elements (MPC’s) are used to connect the
parts and distribute the point loads to the structures. There are two types of MPC
elements used in the model. One of them is RBE2 element (Rigid Body Element-
25
Type 2) which is used for connection purposes. RBE2 element simulates the
fasteners with its infinite stiffness. Deformations of the parts that connected to each
other are directly transmitted with these elements. However, fasteners have finite
stiffness values and deformations are not transmitted directly. Second MPC type
used in the model is RBE3 element (Rigid Body Element-Type 3) which is for load
distribution purposes. It distributes concentrated loads from a specified location to
the other points that are connected to it. Distribution is made by considering the
relative distances of the other points to the specified point.
The general view of the FEM of the cockpit structure is shown in Figure 2.3.1. In
the figure, bar elements are displayed in their 1-D form.
Figure 2.3.1: General view of cockpit FEM
The number of the elements and grid points (nodes) used in the model is given in
Table 2.3.1.
26
Table 2.3.1: The number of the elements and grid point used in the model
Element Type Number

CBEAM 4781
CQUAD4 18313
CTETRA 43971
CTRIA3 133
RBE2 96
RBE3 10
GRID POINTS 108627
27
Upper Longerons
Upper longeron is designed mainly for carrying the primary fuselage bending loads.
Also it supports the windshield and the canopies. Because of these missions, it has
form of a U-shape beam. Thus, there are three sections on the longerons called; web
which supports the windshield and the canopies, outer flange which is connected
with the skin and inner flange which supports the lateral structures of the cockpit.
While web is modelled with CQUAD4 elements, outer and inner flange of the upper
longeron is modelled by using CBEAM elements. Also at the canopy hook locations,
webs are stiffened locally with hook flanges. Hook flanges are also modelled with
CBEAM elements. For the distribution of the canopy loads to the upper longerons,
RBE3 elements are used at the location of canopy hook cut-outs. There are totally
five hook locations on one side of the cockpit. Two of them are at the front cockpit
and three of them are placed at the rear cockpit. The detailed FEM of the upper
longeron is shown in Figure 2.3.2.
Figure 2.3.2: Detailed FEM of the upper longeron
28
Lower Longerons
Another fuselage bending carrying member of the cockpit is the lower longeron. Its
form consists of a tapered T- shape beam. Lower longeron is connected to the skin
from its base. Both sides of the lower longeron support the side and the bottom
frames. At front cockpit, floor is connected to the outer side of the longeron. Lower
longeron is modelled with CBEAM elements. The detailed FEM of the lower
longeron is shown in Figure 2.3.3.
Figure 2.3.3: Detailed FEM of the lower longeron
29
Frames
Frame structure consists of three parts which are inner cap, web and the outer cap.
The cross section is in Z and U-shape. Inner and outer cap are reinforcement
members and prevent the frame web from buckling. For the continuous frames, side
and the bottom frames are connected to the both sides of the lower longerons. Also
there are intercostals between some frames. Intercostals carry a portion of
longitudinal compression loads of the fuselage to prevent the skin from buckling.
These structures, like frames, formed with web and flanges. For the frames and
intercostals, the caps and flanges are modelled with CBEAM elements. While only
CQUAD4 elements are used for modelling the intercostals webs, in addition to
CQUAD4 elements, CTRIA3 elements are also used to model the frame webs.
CTRIA3 elements are generated at the connection points of the lateral structures such
as tie-bars. The detailed FEM of the frames and the intercostals is shown in Figure
2.3.4 and Figure 2.3.5 respectively.
Figure 2.3.4: Detailed FEM of the frames and intercostals
30
Figure 2.3.5: Detailed FEM of the frame
31
Bulkheads
Bulkhead is the main supporting member of the cockpit. It supports longerons, skins,
floors and decks. Bulkhead also resists the compression loads which results from
internal pressure. It consists of caps, web, and horizontal and vertical stiffeners. Caps
are connected to the all side of the web. Side caps distribute the side shear loads from
the skin. Lower caps support the floors and the skin, and upper caps carry a portion
of the internal pressure loads of the cockpit. In the model, also seating rails are
supported from the bulkheads which are located behind the pilots. In the real
structure, bulkhead webs have cut-outs especially for the harness installations.
However in the model, web cut-outs are not modelled for the simplification reasons.
Stiffeners, caps and seating rails are modelled with CBEAM elements. CQUAD4 and
CTRIA3 elements are used for the webs. CTRIA3 elements are generated only at
mesh transition regions where especially at the connection points between parts. The
detailed FEM of the bulkheads is shown in Figure 2.3.6 and Figure 2.3.7
respectively.
Figure 2.3.6: General view of bulkhead FEM
32
Figure 2.3.7: Detailed FEM of the bulkheads
33
Skin
The skin carries fuselage shear flows and transfers the differential air pressures to the
frames and bulkheads. It is supported by longerons, bulkheads and frames. In the
model, skin lies between the upper and the lower longerons. Because of its sheet
form, skin is modelled mainly with CQUAD4 elements. Cut-outs are not modelled
for the simplification reasons. At some transition regions, CTRIA3 elements are used
during the meshing. The detailed FEM of the skin is shown in Figure 2.3.8.
Figure 2.3.8: Detailed FEM of the skin
34
Front & Rear Floor
The front floor is placed between the lower longerons and supported also by the
bottom frames. The rear floor is located between side frames and supported also by
floor posts which are the stiff vertical beams under the floor. Floors are loaded
mainly with cockpit pressure and consist of floor web with its reinforcement
members, longitudinal and transverse beams. Beams of the floor are modelled with
CBEAM elements while webs are modelled with CQUAD4 elements. The detailed
FEM of the front and a rear floor is shown in Figure 2.3.9.
Figure 2.3.9: Detailed FEM of the front and rear floors
35
Upper & Lower Deck
The upper deck lies between the upper longerons whereas the lower deck is placed
between the cockpit panels. Deck structure supports the seat rail at the rear cockpit. It
also provides a mounting platform for the equipments. Deck consists of web with
longitudinal and transverse stiffeners. Stiffeners are modelled by CBEAM elements
and CQUAD4 elements are used in the modelling of the webs. The detailed FEM of
the front and a rear floor is shown in Figure 2.3.10.
Figure 2.3.10: Detailed FEM of the upper and lower deck
36
Windshield
Windshield is modelled with CTETRA (3-D) solid elements. During the modelling,
the volumetric parameters are considered in order to reflect the stiffness of the
structure better. Windshield connection to the upper longerons is performed with the
fasteners, piano type hinges and the fittings. Instead of modelling these attachment
members, RBE2 elements are used at the connection points. While creating RBE2
elements, all of 3 translational and 3 rotational degrees of freedoms (DOF) are fixed.
So, at the connection points, force and moments are directly transmitted between the
windshield and the longerons. The detailed FEM of the windshield is shown in
Figure 2.3.11.
Figure 2.3.11: Detailed FEM of the windshield
37
Horse Shoe
Horse Shoe is modelled in order to add lateral stiffness to the cockpit, especially
places where it connects to the upper longeron. As in the case of the windshield, it is
modelled with CTETRA elements and attachment of horse shoe to the longeron is
performed with RBE2 elements. RBE2 elements are used such that 6-DOF is
constrained. In other words connections are in rigid-fixed forms. The detailed FEM
of the horse shoe is shown in Figure 2.3.12.
Figure 2.3.12: Detailed FEM of the horse shoe
38
Sheet support
The sheet support which mainly supports the secondary structures in the cockpit is
modelled for adding the lateral stiffness to the model. Although there are cut outs on
the structure, they are not modelled. Sheet support is modelled with CQUAD4 and
CTRIA3 elements. RBE2 elements are used to connect the support to the upper
longerons. Three translational DOF is assigned at the connection points. The detailed
FEM of the sheet support is shown in Figure 2.3.13.
Figure 2.3.13: Detailed FEM of the sheet support
39
Tie-Bars
There are two tie bars in the model which are located at front and rear cockpit. These
bars are connected to the frames and support the upper longerons against the lateral
loads caused by the cabin pressure. The sides of the tie bar are in plate form and
these side sections provide connection area on the frames. Main bar section is
modelled with CBEAM elements. Although the side sections are in plate form, due
to the simplification reasons, CBEAM elements are also used for these sections. Also
the connections of the tie-bars to the frames are not modelled with rigid elements as
in the case of the connection of the primary structures between each other. The
detailed FEM of the tie bars are shown in Figure 2.3.14 and Figure 2.3.15
respectively.
Figure 2.3.14: Detailed FEM of the front tie bar
40
Figure 2.3.15: Detailed FEM of the rear tie bar
41
Assembly of the structures
The assembly of the parts is shown in Figure 2.3.16. For better visualization, half of
the cockpit is shown. The other part is the symmetric of this part in X-Z plane. At the
intersection locations, mesh density on the parts is arranged in such a way that the
connectivity requirement of the global model is satisfied.
Figure 2.3.16: Assembly of the structures
Material Properties
In this study, structures are built up with isotropic materials. In the model,
aluminium, steel and magnesium alloys are used to assign the mechanical properties
of the structures. Mechanical properties of the materials according to their structural
forms are given in Table 2.3.2. In Table 2.3.2, E, Ftu, Fty, Fsu, ρ and ν stand for elastic
modulus, ultimate tensile strength, yield tensile strength, ultimate shear strength, density and
Poisson’s ratio.
42
Table 2.3.2: Mechanical properties of the materials
E Ftu Fty Fsu ρ

Part Name Material Form ν
(ksi) (ksi) (ksi) (ksi) (lb/in3)
Frame, Skin, Floor,
Deck, Bulkhead, Al 7075
Sheet 10300 78 70 47 0.101 0.33
Intercostals, Sheet T6
Support
Longeron, Stiffeners,
Al 7075
Bulkhead caps, Seat Extrusion 10400 78 70 41 0.101 0.33
T6
Rail
Al 2024
Tie Bar Plate 10700 64 48 38 0.100 0.33
T351
Al 356 Aluminium
Horse Shoe 10300 22 15 14 0.097 0.33
T6 Casting
AZ91C Magnesium
Windshield 6500 17 12 - 0.065 0.35
T6 Casting
Thickness Properties
In the model, web thickness is defined as a constant over the CQUAD4 and CTRIA3
elements. By assigning the thickness to these 2-D elements, real 3-D web form is
obtained. The web thicknesses used in the model is given in Table 2.3.3.
Table 2.3.3: Web thicknesses used in the model
Thickness
Webs
(inches)
Frame (Side) 0.050
Frame (Bottom) 0.040
Bulkhead 0.040
Skin 0.050
Front/Rear Floor 0.040
Upper/Lower Deck 0.063
Intercostal 0.050
Sheet Support 0.050
43
2.4 Checks for Finite Element Model
Before running the finite element analysis (FEA), in order to have clear load path,
performance of the elements should be in desirable levels as the shape of the
elements significantly affects the stress values calculated in FEA. In order to
understand whether the model has mechanism or not, in other words, whether it has
rigid body motion in it or not, also rigid body check should be done prior to the static
analysis.
Element Quality Check
Due to the difficulties in structural geometries, inevitably, during the meshing

process, some of the elements exceed the default element geometric limits of the
NASTRAN®. These elements are generally located in transition areas. Although
these elements exceed the limits, it does not prevent the solution of the other
elements and NASTRAN® can reach a solution. However, in order to get clear load
path for the future applications of the study, such as comparison of the test results
with the FE analysis results, the element quality check become crucial especially for
the testing points. As a result of these checks, mesh refinement is done for the
regions where clear load path is necessary.
For quadrangle elements, aspect ratio, skew, taper and warping is the geometrical
parameters which are to be verified with respect to NASTRAN® limits. For triangle
elements only two of them which are namely aspect ratio and skew are considered.
The descriptions of the element quality check parameters for the quadrangle elements
are shown in Figure 2.4.1.
44
Figure 2.4.1: Quadrangle element quality check parameters [16]
The aspect ratio parameter is the width over the length ratio. Skew is an angle used to
verify that the CQUAD4 elements are not too trapezoidal and the CTRIA3 elements
not too slender. Taper is a surface area ratio which reflects whether CQUAD4
elements are tapered or not. Warping is defined as the angle which shows the
deviation of the element from being planar. The element quality check parameters
with their NASTRAN® limits are given in Table 2.4.1.
Table 2.4.1: Element Quality criteria
Quality Parameters NASTRAN® Limits

Aspect Ratio 5
Max. Warping Factor 0.05
Max. Taper Ratio 0.5
Min. Quad Skew 30 deg
Max. Quad Angle 30 deg
Min. Quad Angle 150 deg
Max. Tri Angle 160 deg
Min. Tri Skew 10 deg
45
Rigid Body Check
In static analysis, in order to reach a solution, there should not be any mechanism or
rigid body motion in the FEM. The complete structure can have at most six rigid
body modes which are in three translational and three rotational directions. These six
rigid body mode frequencies should be zero for the check purposes. Rigid body
modes are calculated in NASTRAN® (SOL103) solution sequence. Normal modes
analyses are done by assigning free/free boundary condition to the model. For
medium and large scale models, the most effective eigenvalue extraction method
which is Lanchoz Method [24] is used during the control. The mass distribution is
obtained by assigning the density parameters to the materials involved in the model.
Natural modes analysis results for the first six modes are given in Table 2.4.2.
Table 2.4.2: Natural modes analysis results
Rigid body motion

Mode Frequency [Hz]
(Directions)
Tx 1 0.000005
Ty 2 0.000009
Tz 3 0.000010
Rx 4 0.145680
Ry 5 0.252420
Rz 6 0.431200
Here, Tx, Ty, Tz and Rx, Ry, Rz represents for translational and rotational motions
in x, y and z directions respectively.
Rigid body mode shapes for first six modes are shown from Figure 2.4.3 to Figure
2.4.5 respectively.
46
Figure 2.4.2: Mode shape-Tx
Figure 2.4.3: Mode shape-Ty
47
Figure 2.4.4: Mode shape-Tz
Figure 2.4.5: Mode shape-Rx
48
Figure 2.4.6: Mode shape-Ry
Figure 2.4.7: Mode shape-Rz
49
2.5 Finite Element Analysis
In this study, NASTRAN® (SOL101) solution sequence is used for the linear static
analysis. A set of linear equations generated by the finite elements represents the
differential equations and the solution of these equations depends on the load and
displacement boundary conditions. Assumptions of the analysis are the following:
displacements are small, stiffness matrix and boundary conditions do not change and
displacements are directly proportional to the loads. Also, the material used in the
study is homogenous and isotropic.
Applied Loads
In the study, cockpit is loaded with 5 [Psi] internal pressure loads, which is the limit
load for the cockpit pressurisation system. Load is applied to the surfaces which are
exposed to the internal pressure in the cockpit. At the front side; bulkhead-1, at the
rear side; bulkhead-3 and upper deck, at the bottom side; front and rear floors, and at
the right and the left side skin create the pressure surfaces of the cockpit. At the
upper side, between bulkhead-1 and upper deck, there are canopies and windshield to
complete the close shape of the cockpit. However, these structures are not taken into
account as pressure surfaces for the reasons mentioned in section 1.4. Under 5 [Psi]
cabin pressure, the loads exerted by the canopy on the longerons through the canopy
hooks are gathered from the manufacturer company. On the other hand, the loads on
the longerons exerted by the windshield pressure loads are neglected as mentioned in
section 1.4. The shell elements which are located on the pressure surfaces are
oriented in such a way that their normal axes are facing to the outdoors. Thus, loss of
the pressure load is prevented by assigning the pressure force directions as the same.
The pressure surfaces in the model and the shell element normals are shown in
Figure 2.5.1 and Figure 2.5.2 respectively. Canopy hook loads for the left side of the
cockpit are given in Table 2.5.1. Here, Py, Pz and Mx stand for force in lateral, force
50
in vertical and moment about forward direction respectively. Also loads are shown
schematically in Figure 2.5.3.
Figure 2.5.1: Pressure surfaces in the model
Figure 2.5.2: Shell element normals
51
Table 2.5.1: Canopy Hook Loads-Left Side
Location Py (lb) Pz (lb) Mx (lb.in)

F6 635 1530 2140
F9 710 1650 1688
F12 433 1170 287
F15 550 1900 0
F19 235 2460 -559
Figure 2.5.3: Schematic representation of canopy hook loads
Displacement Boundary Conditions
Under the static pressure load, for the solution of the FEM, model is only fixed at its
bulkhead-4 (B4) as a simply supported way. On the real fuselage structure, actually,
there is a nose section which is located at the front side of the cockpit. The nose
section is cantilevered at bulkhead-1 (B1). At the ground condition, nose section
does not support the cockpit effectively. Therefore, at B1, model is released as free.
At the rear side, behind the cockpit, there is a rear fuselage which supports wing,
52
engine and empennage. The rear fuselage also supports the cockpit, especially when
cockpit is loaded under cabin pressure. In the model, the displacement boundary
conditions for the cockpit are assigned by considering these real life end conditions.
The translational DOF’s at the ends of the upper and lower longerons are constraint
in vertical (Z-axis) and in longitudinal (X-axis) directions. However, the rotational
DOF’s at the constraint points are released as free in order to reduce the boundary
effects at the connection points. The lateral (Y-axis) translations are constraint at the
longerons due to this particular reason. On the other hand, in order to create static
solution matrices in the mathematical model, translation is constraint in lateral (Y-
axis) direction at the lower deck-bulkhead web intersection points (see Figure
2.3.16). These points have no impact on the results evaluated for the correlation
study. The final configuration of the displacement boundary conditions assigned to
the model is shown in Figure 2.5.4. In this figure, constraint translational DOF’s in
X, Y and Z directions are shown respectively with numbers 1, 2 and 3.
Figure 2.5.4: Displacement boundary conditions
53
2.6 Summary
In this section, finite element model (FEM) of the cockpit structure is constructed by
considering the functions of the structures involved in the model. Assumptions made
during the modelling are also described in detail. The model is checked in order to
determine whether it is suitable for the static analysis or not. After these checks,
cabin pressure load is applied to the model by assigning displacement boundary
conditions to the specific locations. The finite element model (FEM) is solved by
NASTRAN® (SOL101) solution sequence. The results are not evaluated in this
section and they will be made according to the static test procedure which considers
the locations where strain data is gathered from.
54
CHAPTER 3
GROUND PRESSURISATION TEST
3.1 Introduction
In order to have a reliable FEM, model must be validated with experimental testing
methods. The types of tests and methods which are selected to perform these tests
mainly depend on the type of load applied to the model. In this study, cockpit FEM is
validated only under cabin pressure load. Cabin pressurisation is applied to the real
cockpit structures in a static manner. In this section, ground pressurisation test
procedure and the test results for the real cockpit structure are introduced. First, test
configuration is described. Secondly, type, criteria for the selection, instrumentation
and location of the sensors used in the test are explained. And after having mentioned
about test steps, finally, the obtained test results are also interpreted.
3.2 Test Configuration
The ground pressurisation test is performed on the aircraft itself. The cockpit of the
aircraft is pressurised on the ground by pressure supplier unit. Pressure supplier unit
provides air to the cockpit through a hose pipe which is mounted on the aircraft cabin
pressure system. The obtained signals from the sensors which are installed in the
cockpit are recorded by data acquisition system. Collected data in the data
55
acquisition system is then transferred to the mobile PC. The general configuration of
the ground pressurisation test is shown in Figure 3.2.1 schematically.
Figure 3.2.1: General Test Configuration
3.3 Sensor Types
Under the cabin pressure, in order to examine the strain changes occurring on the
cockpit structures, two types of sensors, namely strain gauges and the pressure
transducers, are installed inside the cockpit in order for measuring strain and cabin
pressure respectively.
In this study, two types of strain gauges, VISHAY® CEA-13-250UN-350 [29] linear
type and VISHAY® CEA-13-125UR-350 [30] rectangular 450 single-plane rosette
type strain gauges are used. The strain gauge specifications are given in Table 3.3.1.
56
Table 3.3.1: Strain Gauge Specifications [29], [30]
Strain
Resistance Dimensions Temperature
Gauge Designation Type Range
(Ohms) (inch) Range (oC )
(%)
CEA-13-250UN-350 Linear 350 ± % 0.3 0.52 × 0.22 ±5 -75o to+175
CEA-13-125UR-350 Rosette 350 ± % 0.4 0.42 × 0.62 ±5 -75o to+175
While linear type strain gauge has a single data output channel, rosette type strain
gauge has three. The strain gauges used in the study are shown in Figure 3.3.1 with
their corresponding output channel numbers.
Figure 3.3.1: Linear and Rosette Gauges with their channel numbers [29], [30]
ENDEVCO® Model 8540 [31] Piezoresistive pressure transducer is chosen for its
high sensitivity, high stability during temperature transients and having broad
measurement ranges from 5 [Psi] to 500 [Psi] pressure. The transducer has a 0.15 [in]
face diameter. Pressure transducer used in the study is shown in Figure 3.3.2.
57
Figure 3.3.2: Pressure transducer used in the study [31]
3.4 Criteria for the Selection of the Sensors
Linear type strain gauges are used in the places where stain changes are mainly
varying in one direction. These types of gauges are generally selected on the caps
and flanges. If the strain values are expected to change on the structure more than in
one direction, rosette type strain gauges are preferred to be installed in those places.
Therefore, these types of strain gauges are installed mainly on the webs and panels.
In other words, if the structure mainly resists the axial loads and the principal strain
directions are known, linear type strain gauges, if the structure mainly resists the
shear loads and the principal strain directions are not known, rosette type strain
gauges are used. The 450 single-plane rosette type gauge has three gauge grids (i.e.
arms). Three grids, with the second and the third grids angularly displaced from the
first grid by 450 and 900, respectively. The principal strains are calculated easier in
450 single-plane rosette and, in this study, it is selected for its computational
advantage. Strain gauges are manufactured from different combinations of grid alloy
in order to meet various application requirements. In this study, gauges are selected
to be suitable for the aluminium alloy materials they are attached. They are also
58
offered in a number of different lengths. The sizes of the gauges are selected by
considering the installation space limitations.
3.5 Installations of the Sensors
Strain gauge installations are bounded on the aircraft by the limitations of the
minimum space required for the installation. The reason is that, structure to be
modified is already assembled and the environmental condition for the labour is not
the same with the condition during the production phase. For the manufacturer
company, it is easier to install the strain gauges on the desired parts before the
assembly. Hence, in the study, sensors are installed on the accessible and/or suitable
areas found in the cockpit.
Strain Gauge Installation
Linear strain gauges are installed on the upper longeron inner and outer flanges,
frame inner caps, bulkhead upper caps, upper longeron web, sheet support and on the
tie bars. On these structures, principal strain exists mainly in one direction. Rosette
strain gauges are installed on the skin panels. On the skin, principal strain directions
are unknown and exist in more than one direction.
Steps in strain gauge installation can be summarised as follows:
• Dirty and greasy surface is cleaned with isopropyl alcohol.

• Marking is done for positioning the strain gauge.
• Paint on the surface is removed with sandpaper.
• Wet sanding is done to the surface by using a proper conditioner.
• Surface is cleaned chemically with neutraliser.
• Surface is dried from centre to the edges.
• Strain gauge is pasted to the surface with the help of bond adhesives.
59
• Soldering process is carried out to the gauge from its solder tab area.
• Wiring process is completed with marking the cables of the strain gauges.
Rosette arms are named as A, B and C. The arms B and C angularly displaced from
the arm A by 450 and 900, respectively. Arm A is attempted to lie in parallel with
aircraft forward direction while arm C is attempted to lie in parallel with vertical axis
of the aircraft. Strain gauge installations on some of the cockpit structures are shown
from Figure 3.5.1 to Figure 3.5.10.
Figure 3.5.1: Linear strain gauges on longeron inner and outer flanges
60
Figure 3.5.2: Linear strain gauge on longeron inner flange at the near of its cut-out
Figure 3.5.3: Linear strain gauge on frame inner cap
61
Figure 3.5.4: Linear strain gauges on longeron web
Figure 3.5.5: Rosette strain gauge on skin panel
62
Figure 3.5.6: Strain gauges on frames and skin panel
Figure 3.5.7: Linear strain gauge on the tie-bar
63
Figure 3.5.8: Linear strain gauge on the sheet support
Figure 3.5.9: Linear strain gauge on the bulkhead upper cap
64
Figure 3.5.10: General view of the strain gauge installation on the cockpit
The number and the locations of the strain gauges installed in to the cockpit are given
in Table 3.5.1.
65
Table 3.5.1: The number and the locations of the strain gauges
Longeron Web Longeron Inner Flange Longeron Outer Flange Skin

(Linear) (Linear) (Linear) (Rosette)
Location Number Location Number Location Number Location Number
B1 1 F4 1 F4 1 F4 1
F1 1 F6 1 F9 1 F13 1
F2 1 F14 1 F12 1 F16 1
F20 1 F15 1 F13 1
F18 1
Frame Inner Cap Bulkhead Upper Cap Tie Bars Sheet Support
(Linear) (Linear) (Linear) (Linear)
Location Number Location Number Location Number Location Number
F2 2 B2 1 F3 1 F1 1
F3 2 B3 1 F12 1
F4 2
F5 2
F6 2
F12 2
F13 2 Total Linear Strain Gauges = 34
F14 1 Total Rosette Strain Gauges = 3
F17 1
As there is a structural symmetry in the cockpit, except bulkhead upper caps, strain
gauges are installed only on one side of the cockpit. In Table 3.5.1, location
information on the longeron, skin and frame is given for only their left hand side
parts. The number of the gauges on these structures indicates the number on left hand
side longeron, skin and frame. For example, at F2 station, there are actually two
frames which are placed at each side of the cockpit symmetrically. However, at F2,
two of the strain gauges are installed only on the left hand side frame. Because tie
bars and the sheet support lie between the each side of the cockpit, strain gauges are
installed on their regions close to the left side. On the bulkhead upper caps, strain
gauges are installed on the locations where the installation space is possible.
66
The numbers of the strain gauges on the structures are determined according to the
expected strain distributions on the area of interest. On the upper longeron, strain
gauges are distributed by considering the longeron size. It lies over the whole cockpit
and because of the supporting structures that are attached to it; there are strain
discontinuities over its whole length. Although the skin is located on the whole of the
cockpit as the longeron, there are few numbers of strain gauge installed on it. The
reason is from the fact that the discontinuities in the strain values are not expected
over the skin. Under the cabin pressure, frame deformation highly depends on the
deformation of the longerons and the skin. While the canopy hook loads pull the
frames upward, lateral cabin pressure forces bend the frames in the radial direction.
In order to have a decision for the frame response under cabin pressure, the strain
gradients on the frames are investigated by using two strain gauges. The load path on
the tie bars, bulkhead upper caps and the sheet support exist mainly in one direction
which is actually in a lateral direction. The strain gradient does not vary much over
their length and therefore a single strain gauge provides accurate measurement
results for the strain levels of these structures.
Pressure Transducer Installation
There are two pressure transducers in the cockpit which are located in front and in
rear cockpit. Transducers are attached to the cockpit structures by using tapes. The
purpose of locating two pressure transducers is to determine whether there is a
pressure gradient or not through the cockpit.
3.6 Data Acquisition System
In this study, data acquisition system consists of NI® CompactRIO-9022 [32] real-
time controller and NI® 9205 [33] industrial I/O modules. NI® CompactRIO is
®
programmed with NI LabVIEW [34] graphical programming software which
collects and analyze the signals and convert them into the physical measurements.
67
Sensors in the cockpit are connected to the module channels with cables. There are
two kinds of modules in the system which are for measuring strain and pressure.
Data acquisition system is placed into the cockpit floor during the test. The controller
has USB data storage unit and the measured data is stored in the flash memory stick
during the test.
3.7 Test Steps
In the study, test steps are carried out as follows:
1) Sensors are connected to the data acquisition system.

2) Data acquisition system is operated and the initial data is started to be taken.
3) The canopies are closed.
4) Pressure supplier unit is operated.
5) The pressure inside the cockpit is increased more than 5 [Psi] step by step
during a period of 30 seconds. Additional pressure is given to the cockpit due
to the difficulty of controlling the manual control unit located on the supplier
and not to miss the data around 5 [Psi].
6) When pressure reach its maximum level, waited at least 10 seconds for the
stabilization of the data.
7) Waited minimum 3 seconds for gathering the data.
8) Cabin is depressurised step by step over a period of 30 seconds.
9) When the pressure drops closer to zero, in order to prove the test
repeatability, test steps are repeated once more following the steps starting
from (5). In other words, two pressure cycles are applied to the cockpit during
the test.
10) The canopies are opened when the pressure drops to zero.
11) Data acquisition unit is closed and the data is transferred to a computer by
USB memory stick.
68
3.8 Test Results
In this section, at first, the strain gauge results are given as strain-pressure plots.
Secondly, the maximum strain values at 5 [Psi] pressure load are tabulated with their
corresponding calculated stress values. For the rosette type strain gauges, the
maximum principal stress values are also calculated. Throughout the study, strain
and pressure values are given in [με] and in [Psi] units respectively.
Under linearly increasing cabin pressure, all of the strain gauges on the longeron web
show linear behaviours. Between the pressurisation and the depressurisation, some of
the strain data follows a different path. This situation is known as ‘hysteresis’. If a
strain gauge is loaded to a high value of strain, resistance value of the gauge has
acquired some settled value and during the unloading, all resistance values will have
higher values than that of in the loading. Thus, between the loading and unloading,
hysteresis loop occurs [35]. For the strain gauges located at the longeron web, these
loops are so narrow that do not affect the reliability of the results. Also at maximum
pressure, strain gauges read the same strain values in two pressure cycles. The stain-
pressure plot shown in Figure 3.8.1 is for the longeron web.
69
Longeron Web
350.0
300.0
250.0
Microstrain (με)
B1
200.0
F1
150.0 F2
F20
100.0
50.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Pressure (Psi)
Figure 3.8.1: Longeron web strain-pressure plot
The strain gauges on the longeron inner flanges show linear and non-linear
behaviours. At first glance, it seemed that the strain gauges which have non-linear
behaviour also have some bonding problems with their bonding surfaces. However,
when load is applied by hand in the vicinity of the gauge, strain gauges show linear
behaviour. Thus, it is decided that there is not any problem for the installations. The
common point for the gauges which show non-linear behaviour is that, their strain
values are too low when compared to the gauges which have linear responses. It is
thought that, the strain gauges could not response properly for the low strain values
where the strains are below the resolution of the gauge. The detailed discussions
about the non-linearity of these gauges are made in Chapter 4. For the longeron inner
flange, the strain-pressure plot is shown in Figure 3.8.2 and non-linear gauges are
70
Longeron Inner Flange
500.0
300.0
F4
Microstrain (με)
100.0 F6
F14
-100.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0
F15
F18
-300.0
-500.0
-700.0
Pressure (Psi)
Figure 3.8.2: Longeron inner flange strain-pressure plot
Longeron Inner Flange‐Non Linear

35.0
25.0
Microstrain (με)
15.0 F4
5.0
F14
-5.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0
-15.0
-25.0
Pressure (Psi)
Figure 3.8.3: Longeron inner flange non-linear strain behaviours
The strain gauges on the outer flange of the longeron and the skin show linearity
under cabin pressure. The hysteresis loops are also involved and their affects on the
results are considered as negligible. While for the longeron outer flange, the strain-
71
pressure plot is shown in Figure 3.8.4, for the skin, the strain-pressure plot of the
three armed rosettes is shown in Figure 3.8.5.
Longeron Outer Flange

500.0
400.0
300.0
Microstrain (με)
F4
200.0
F9
100.0 F12
F13
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
-100.0
-200.0
Pressure (Psi)
Figure 3.8.4: Longeron outer flange strain-pressure plot
Skin
500.0
400.0 F4A
300.0 F4B
Microstrain (με)
F4C
200.0
F13A
100.0 F13B
F13C
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 F16A
-100.0 F16B
F16C
-200.0
-300.0
Pressure (Psi)
Figure 3.8.5: Skin strain-pressure plot
72
The strain gauges on the frame inner caps show linear and non-linear behaviours as
gauges located at longeron inner flange. The installation checks are made and
understood that there is not any problems for the installation. As longeron inner
flange, the common point for the gauges which show non-linear behaviour is that,
their strain values are too low when compared to the gauges which have linear
responses. At the frames where these gauges are located, there are also other gauges
and they show linear behaviour under cabin pressure. It is again thought that, the
strain gauges could not response properly for the low strain values. The detailed
discussions about the non-linearity of these gauges are also made in Chapter 4. For
the frame inner cap, the strain-pressure plot is shown in Figure 3.8.6 and non-linear
gauges are shown in Figure 3.8.7.
Frame Inner Cap

F2-a
700
F2-b
F3-a
500
F3-b
300 F4-a
Microstrain (με)
F4-b
100
F5-a
-100 0.0 1.0 2.0 3.0 4.0 5.0 6.0 F5-b
F6-a
-300 F6-b
F12-a
-500
F12-b
-700 F13-a
F13-b
-900
F14
F17
Pressure (Psi)
Figure 3.8.6: Frame inner cap strain-pressure plot
73
Frame Inner Cap‐Non Linear
30.0
20.0
10.0
Microstrain (με)
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
-10.0
F4-a
-20.0
F14
-30.0
-40.0
-50.0
-60.0
Pressure (Psi)
Figure 3.8.7: Frame inner cap non-linear strain behaviours
The linear strain variations under cabin pressure for the strain gauges located at the
bulkhead upper cap, ties bars and at the sheet support are shown in the plots from
Figure 3.8.8 to Figure 3.8.10.
Bulkhead Upper Cap

400.0
350.0
300.0
Microstrain (με)
250.0
200.0 B2
B3
150.0
100.0
50.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Pressure (Psi)
Figure 3.8.8: Bulkhead upper cap strain-pressure plot
74
Tie Bars
1000.0
900.0
800.0
700.0
Microstrain (με)
600.0
F3
500.0
F12
400.0
300.0
200.0
100.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Pressure (Psi)
Figure 3.8.9: Tie bars strain-pressure plot
Sheet Support
1200.0
1000.0
Microstrain (με)
800.0
F1
600.0
400.0
200.0
0.0
0.0 1.0 2.0 3.0 4.0 5.0 6.0
Pressure (Psi)
Figure 3.8.10: Sheet support strain-pressure plot
75
The maximum strain values obtained at 5 [Psi] cabin pressure for the gauges which
show linear behaviour under linearly increasing cabin pressure are converted to the
stress values by using the Hooke’s Law. If the test material is homogeneous and
isotropic and if the stress/strain relationship is linear, then the uni-axial and biaxial
forms of Hooke's law can be used to convert the principal strains into principal
stresses for both linear and rosette types strain gauges.
The Hooke’s Law in uni-axial form which is used for calculating the axial stress
values from the strain measurements of linear type strain gauges is given in Equation
3.1.
σ = E×ε (Eqn. 3.1)
where, σ , ε and E are the axial stress, axial strain and modulus of elasticity of the
material respectively.
The principal strain calculations for the rosette type strain gauges are derived from
the strain transformation graph, which is known as Mohr’s circle. According to the
Mohr’s circle, the normal strain at any angle θ from the principal axis is calculated
by Equation 3.2.
εP + εQ εP − εQ
εθ = + cos(2θ) (Eqn. 3.2)
2 2
In Equation 3.2, εP and εQ stand for maximum and minimum principal strains
respectively. A rectangular 45o single-plane rosette type strain gauge which is
oriented at θ degrees from the maximum principal direction is shown in Figure
3.8.11. The reflection of this orientation to the Mohr’s circle is shown in Figure
3.8.12.
76
Figure 3.8.11: Orientation of the rectangular rosette
Figure 3.8.12: Mohr’s circle for the rectangular rosette
77
The arms of the rosette gauge with their orientation angles from the maximum
principal direction read the strain values according to the equations derived from the
Equation 3.2.
εA = + cos(2θ) (Eqn. 3.3)
2 2
εB = + cos 2(θ + 45 ο ) (Eqn. 3.4)
2 2
εC = + cos 2(θ + 90 ο ) (Eqn. 3.5)
2 2
Rosette type strain gauge measures the left hand side parameters in the above
equations. The unknowns at the right hand sides are found by solving these three
equations simultaneously. Thus, principal strain values for the 45o rosette type strain
gauge are calculated according to the Equation 3.6 given below.
εA + εC 1
ε P ,Q = ± (ε A − ε B ) 2 + (ε B − ε C ) 2 (Eqn. 3.6)
2 2
The Hooke’s Law in biaxial form is given in Equation 3.7 and Equation 3.8.
E
σP = (ε P + νε Q ) (Eqn. 3.7)
1- ν2
E
σQ = (ε Q + νε P ) (Eqn. 3.8)
1- ν2
where ν is the Poisson's ratio of the material.
78
The maximum (σP) and the minimum (σQ) principal stresses are calculated by
substituting Equation 3.6 into Equation 3.7 and Equation 3.8. The final form of the
equation is given in Equation 3.9 below.
E ⎡εA + εC 2 ⎤
σ P ,Q = ⎢ ± (ε A − ε B ) 2 + (ε B − ε C ) 2 ⎥ (Eqn. 3.9)
2 ⎣ 1- ν 1+ ν ⎦
The maximum strain values at 5 [Psi] pressure load are tabulated with their
calculated stress values in Table 3.8.1. In the calculations, elastic modulus (E) and
Poisson’s ratios (ν) are taken from Table 2.3.2. The gauges which are below the
gauge resolution are coloured in yellow. Although Hooke’s law is not valid for these
gauges, the stress levels are calculated by using this law only to interpret the physical
meaning of the data. In this study, these data are evaluated only qualitatively not
quantitatively.
79
Table 3.8.1: The maximum strain and stress values for the strain gauges
Longeron Strain Stress Longeron Strain Stress Longeron Strain Stress

Web (με) (ksi) inner flange (με) (ksi) outer flange (με) (ksi)
Location Location Location
B1 269 2.8 F4 8 0.1 F4 84 0.9
F1 288 3.0 F6 -448 -4.7 F9 128 1.3
F2 253 2.6 F14 -2 0.0 F12 -81 -0.8
F20 170 1.8 F15 -330 -3.4 F13 279 2.9
F18 -130 -1.4
Strain Stress
Frame Strain Stress Max/Min. Max. Bulkhead Strain Stress
Skin
Inner Cap (με) (ksi) Principal Principal Upper Cap (με) (ksi)
(με) (ksi)
Location Location Location
F2-a 199 2.0 F4 314 / -134 3.1 B2 320 3.3
F2-b 74 0.8 F13 385 / -198 3.7 B3 154 1.6
F3-a 255 2.6 F16 386 / 215 5.4
F3-b 254 2.6
F4-a -27 -0.3 Tie Bars Strain (με) Stress (ksi) Sheet Support Strain (με) Stress (ksi)
F4-b 380 3.9 Location Location
F5-a 403 4.2 F3 863 9.2 F1 952 9.8
F5-b 618 6.4 F12 725 7.8
F6-a 531 5.5
F6-b 321 3.3
F12-a 55 0.6
F12-b -251 -2.6
F13-a -182 -1.9
F13-b 239 2.5
F14 -23 -0.2
F17 -834 -8.6
At it can be seen from the Table 3.8.1 that the maximum stress value calculated for
the strain gauge locations is 9.8 [ksi] at the sheet support. This level of stress is too
low when compared with the material yield stress limit, which is 70 [ksi] as shown in
Table 2.3.2.
80
3.9 Summary
In this section, ground pressurisation test procedure, sensor types, criteria for
selection and installations of the sensors, test steps and the test results are presented.
The test results for the real cockpit structures are shown with graphs and maximum
values are tabulated in tables. The majority of the strain gauges show linear
behaviour under linearly increasing cabin pressure. However, some of the gauges
show non-linear behaviours. The reasons for this situation are also mentioned briefly
in this section. The more detailed discussion about the physical definition of the non-
linear strain gauges is also made in the forthcoming chapter. The strain data obtained
from the strain gauges at 5 [Psi] cabin pressure is also converted to the stress values
which is used for the correlation study in Chapter 4.
81
CHAPTER 4
CORRELATION STUDIES BETWEEN FINITE ELEMENT

ANALYSIS AND TEST RESULTS
4.1 Introduction
In this part of the study, FEA results for the measured points at ground pressurisation
test is compared with the test results and FEM is improved by the examination of the
ground pressurisation test data. FEM is also finalised after achieving a good
correlation between the FEA and the test results. Firstly, the response of the
structures under cabin pressure is checked. The satisfactory FEA is examined by the
evaluation of the deformed shapes of the structures. Secondly, for the specific strain
gauge (SG) locations, methodologies so as to gather the stress results from the FEM
are explained. Thirdly, FEA results and test results are then compared and checks for
the correlation are made in detail. Following those, in order to get a better
correlation, some actions for updating the FEM are performed and updated results
are also presented and discussed. Finally, experimentally validated FEM is obtained
and arguments about the validation are concluded.
82
4.2 Deformation Checks
In FE environment, the satisfactory FEA not only depends on the software but
mainly depends on the engineer who is responsible for interpreting the results.
Firstly, the boundary conditions of the real structure and structural deformations
should reflect to the FEM correctly. In other words, the expectations for the
deformed body under specific load have to be meaningful at the first glance. At this
stage, generally, interpretation of the results is made by the engineer qualitatively.
And if one has some predictions about the response of the structure calculated before
using the analytical solution techniques, quantitative examination could also be made
by the engineer during the interpretation of the results. In this study, however, the
prediction of the response of the complex integrated cockpit structure under complex
type cabin pressure load is extremely challenging subject and for this reason, the
deformed shapes of the structure are examined only qualitatively. The deformation
result for the upper longerons under 5 [Psi] cabin pressure load is shown in Figure
4.2.1 and they are exaggerated for the illustration purposes.
Figure 4.2.1: Deformation of the upper longerons under cabin pressure-Top view
83
Under cabin pressure, the upper longerons deform symmetrically as expected. The
boundary conditions assigned at the rear side prevent the translational motions while
rotational motions are free in the model. At the front side, both the sheet support and
front tie bar support the longerons laterally and at bulkhead-2, longerons do not
deform much. At this location, cross tie bar and bulkhead-2 support the longerons as
expected. The maximum obtained deflection is 0.131 [in].
Under cabin pressure, generally, frames carry the tangential stresses from the skin
and normal stresses from the cabin pressure. In this case however, they also carry
canopy hook and floor loads. Floor pressure loads push the bottom frames down and
canopy hook loads bend and pull the side frames to the upward direction. This
combined loading cause frames to stretch, bend and twist. The deformation result of
the frames under 5 [Psi] cabin pressure load is shown in Figure 4.2.2. The
deformations are exaggerated again for illustration purposes and the maximum
deflection is 0.122 [in].
Figure 4.2.2: Deformation of the frames under cabin pressure
84
Figure 4.2.3 shows combined deformations on some of the frames in a close-up
view. On the frames, there are inflection points which reflect the regions where
compression-tension stress transition occurs suddenly.
Figure 4.2.3: Deformation of the frames under cabin pressure (close-up view)
The deformations of the longerons and frames are taken as reference to check
whether the FEA is providing meaningful results or not. Because the performance of
the secondary structures which support longerons and frames, namely; sheet support,
tie-bars, horse-shoe and windshield can be decided by examining the deformations of
the longerons and the frames. Skin response is also expected to be reliable when
longeron and frame response are satisfactory. The reason is that skin covers all of the
longerons and the frames in the cockpit. By using these deformation results, the
boundary conditions can also be checked whether they reflect the real physical
constraints or not. At the first glance, longeron model seems to be validated. For the
unpredictable frame deformations, decisions for the validation of frame structures
will be made after comparing the FEA results with the test results.
85
4.3 Methodology to Interpret the Results
After the construction of the FEM in PATRAN®, the interpolating functions/shape

functions are automatically selected by the software. These shape functions for each
element represent the field variable in terms of degrees of freedom (DOF). In the
structural analysis, these DOF’s, in other words, nodal variables are the node
displacements solved by the NASTRAN® in its system of equations. When these
nodal displacements are found, by defining stress/strain relations and strain
displacement relations, strains, stresses and forces can be calculated easily. By using
these shape functions, the known nodal displacements can be interpolated through
the assemblage of the elements and at any points in the domain, field variable can be
found. Therefore, the displacement at any point inside the element is the function of
the displacements at the corners of that particular element.
In this study, the shape functions defined within the CQUAD4 and CBEAM
elements are linear functions. The strain gauges are placed on these types of elements
in FE environment. According to the SG dimensions, element refinements are made
on the CBEAM elements to match the element size with the SG size. The mesh
refinements on 1-D elements are made easily by breaking the coarse ones into
desired finer dimensions. Therefore, FEA result for the region where strain gauge is
installed in CBEAM elements is gathered directly from these refined 1-D elements.
Because the model is created by using coarse mesh methodology, for the strain
gauges where installed on CQUAD4 elements on the FEM, the dimension of the SG
does not suit always with the 2-D element size. In those situations, the refinement or
breaking the CQUAD4 element affects its neighbouring elements too much. The
reason for this is that nodal connectivity should be provided for the new elements
after each breaking operation. Therefore, FEA results in 2-D element for the region
where strain gauge is installed is gathered by using shape functions.
86
2-D Element Results
In PATRAN®, there is a default averaging domain to average all elements nodal

results at each node. However in this study, because the SG measurements in the test
are done for specific points, the results for these points in FEA have to be gathered
from the specific 2-D element results. Therefore, averaging is not performed during
the post processing of the 2D element results. As it is also mentioned before, shape
function is used for the SG results in FEA. In PATRAN®, the surfaces of the plate
are named as the Z1/Z2 layers which locate at the distances from the mid-surface of
the plate as half of the thickness symmetrically. Because SG is placed on one of the
faces of the actual 3-D structure in the test, the FEA results are gathered according to
that surface and Z1 and Z2 results are examined according to this requirement.
The stress results of the elements are gathered according to the element local
coordinate system. In order to compare with the test results, maximum principal
stress values from the FEA are obtained by considering these element coordinate
systems. Element coordinate system for 2-D elements is shown in Figure 4.3.1.
Figure 4.3.1: Element coordinate systems on 2-D elements
87
In Figure 4.3.2, the discontinuities on stress variations between the groups of
CQUAD4 elements modelled for the skin can be seen by none-averaging the nodal
values. Here, the stress variation on the elements is obtained under 5 [Psi] cabin
pressure. The layer selection (Z1/Z2) for the other groups of CQUAD4 elements on
the skin is shown in Figure 4.3.3. Here, the stress results are obtained from the Z1
(inner) side of the elements.
Figure 4.3.2: Stress plot that has no element to element averaging at the nodes.
88
Figure 4.3.3: The layer selection (Z1/Z2) for 2-D skin elements
1-D Element Results
As mentioned before, FEA result for the region where SG is installed in 1-D element
is gathered directly from the refined elements which suit with the dimension of the
SG. During the modelling of 1-D element that has standard cross-section type, the
stress recovery points are defined automatically by the NASTRAN®. With these
predefined points, cross section of the beam element is also defined and by
considering the element coordinate system, section properties and moments of inertia
values are then calculated. In beam elements, the stress recovery points (C, D, E and
F) in the element coordinate system are located at each end of the beam element and
specified relative to the shear centre of the beam cross-section. The stress recovery
points created by the NASTRAN® on a standard rectangular beam cross section are
89
Figure 4.3.4: The stress recovery points on beam cross section
The SG installation face considered for the stress result in FEA is shown on beam
element in Figure 4.3.5. Here, both SG dimension and beam element size are nearly
the same between two nodes.
Figure 4.3.5: SG installation on beam element
90
The stress values on the stress recovery points are calculated for each cross section.
Therefore, C, D, E and F values across a beam can be obtained. After this stage, the
decision should be made for the face selection in order to get results from true SG
location. In Figure 4.3.5, the SG face of the beam element is D1D2E2E1 plane. In
order to get the stress results for this SG, D and E stress values are considered across
an element. The gathered stress values are a type of combination of axial and
bending stresses acting on each of these stress recovery points. At the final stage, the
combined stress variations on two edges of the SG plane are averaged to gather the
final beam stress value for the SG.
4.4 Checks for FEA and Test Results Correlation
In this part of the study, the stress results gathered from ground pressurisation test are
compared with FEA stress results obtained in 5 [Psi] cockpit pressure load and
checks are made for FEA and Test results correlation. The comparison of the stress
results are tabulated in Table 4.4.1.
91
Table 4.4.1: The comparison between test and FEA results
Longeron Stress Stress Longeron Stress Stress Longeron Stress Stress

Web (ksi) (ksi) inner flange (ksi) (ksi) outer flange (ksi) (ksi)
Location TEST FEA Location TEST FEA Location TEST FEA
B1 2.8 2.7 F4 0.1 -2.3 F4 0.9 1.8
F1 3.0 2.6 F6 -4.7 -6.4 F9 1.3 -0.5
F2 2.6 2.8 F14 0.0 -1.8 F12 -0.8 2.3
F20 1.8 1.5 F15 -3.4 -3.7 F13 2.9 4.9
F18 -1.4 0.6
Stress Stress
Frame Stress Stress Max. Max. Bulkhead Stress Stress
Skin
Inner Cap (ksi) (ksi) Principal Principal Upper Cap (ksi) (ksi)
(ksi) (ksi)
Location TEST FEA Location TEST FEA Location TEST FEA
F2-a 2.0 2.9 F4 3.1 3.4 B2 3.3 2.8
F2-b 0.8 1.5 F13 3.7 4.1 B3 1.6 1.4
F3-a 2.6 3.4 F16 5.4 5.0
F3-b 2.6 2.3
F4-a -0.3 4.7 Tie Bars Stress (ksi) Stress (ksi) Sheet Support Stress (ksi) Stress (ksi)
F4-b 3.9 3.9 Location TEST FEA Location TEST FEA
F5-a 4.2 4.7 F3 9.2 9.6 F1 9.8 10.2
F5-b 6.4 6.0 F12 7.8 8.4
F6-a 5.5 5.4
F6-b 3.3 3.0
F12-a 0.6 0.3
F12-b -2.6 -1.7
F13-a -1.9 -1.2
F13-b 2.5 1.5
F14 -0.2 0.2
F17 -8.6 -8.5
When magnitudes of the stress levels are inspected in Table 4.4.1, at first glance,
FEA stress results have approximately same order of magnitude with test results for
the most of the SG locations. For the gauges, which has strains that are below the
gauge resolution in the test (coloured in yellow), results do not match in magnitude
and also in sign. Also there are sign and magnitude differences in FEA and test stress
results for longeron outer flange locations at F9 and F12, and longeron inner flange
location at F18. The percentages of difference between test and updated FEM and
results are given in Table 4.4.2.
92
Table 4.4.2: The percentage of difference between test and FEM results
Longeron Difference Longeron Difference Longeron Difference

Web (%) inner flange (%) outer flange (%)
Location TEST-FEA Location TEST-FEA Location TEST-FEA
B1 -3 F4 -2864 F4 106
F1 -13 F6 37 F9 -138
F2 6 F14 8554 F12 -373
F20 -15 F15 8 F13 69
F18 -144
Frame Difference Difference Bulkhead Difference

Skin
Inner Cap (%) (%) Upper Cap (%)
Location TEST-FEA Location TEST-FEA Location TEST-FEA
F2-a 41 F4 9 B2 -16
F2-b 97 F13 11 B3 -13
F3-a 29 F16 -8
F3-b -12
F4-a -1667 Tie Bars Difference (%) Sheet Support Difference (%)
F4-b 0 Location TEST-FEA Location TEST-FEA
F5-a 13 F3 4 F1 4
F5-b -6 F12 8
F6-a -1
F6-b -9
F12-a -47
F12-b -34
F13-a -36
F13-b -39
F14 -184
F17 -1
In the literature, there is not any defined percentage of difference to check the
correlation between FEA and test results for the static analyses. In some applications
while 20% is acceptable, in others even 1% is unacceptable. However, limitation
mainly depends on the order of magnitude of the inspected stress levels. As a general
practice, if the stress levels are too high, even in small percentage of difference, the
magnitude of the errors will be also too high and these high errors lead engineer to
get misleading results during his checks for the structural failure modes such as
material failure. Also, such high stress errors affect the fatigue life and crack
initiation calculations of the structures. In this study however, maximum stress value
calculated from the test is 9.8 [ksi]. This level of stress is too low when compared
93
with the material yield stress limit, which is 70 [ksi] (Table 2.3.2). The percentage
errors between the results given in Table 4.4.2 are unrealistic due to these low stress
levels. In the study, test and FEA correlation is checked by considering the stress
differences with their order of magnitude not the percentage errors.
As it is seen from the Table 4.4.1, for the longeron web, a considerable correlation is
satisfied between the test and the FEA stress results. The difference in stress values
at the F20 location is due to the simply supported boundary condition assigned at B4.
The real elastic effects are not included at the end of the cockpit FEM due to the
assumed fix condition in that region. Although simulating the real deformation slope
of the upper longerons at this location in the FEM is very difficult, the displacement
boundary condition provides a sufficient approximation for the longeron web
regarding the obtained result. At the front side of the cockpit, the differences between
results are due to the elastic end effect at B1 and sheet support connection at F1.
However, the stress values are at most 0.4 [ksi] lower in FEA than stress values
measured in test. This value is acceptable when considering the material yield stress
limit of the longeron which is 70 [ksi] as shown in Table 2.3.2. This means that, such
low order error does not give rise to big deviations during the structural analyses
which are to be re-performed on the FEM.
When longeron inner flange results are inspected, satisfactory correlation can be seen
on the F15 location. At the rear side of the cockpit, the results for the longeron inner
flange are affected as the longeron web from the fix condition assigned at this
particular region. However this time, the stress value between test results and FEA at
F18 station differs not only in magnitude but also in its sign. The stress is measured
as compressive in the test while gathered as tension in FEA. However, the
compression-tension band is so narrow and very close to the zero stress line. The
stress variation of the longeron at F18 station is shown in Figure 4.4.1. The curvature
from fix point to the F18 station determines the sign of the stress.
94
Figure 4.4.1: Longeron stress variation at F18 station-Top view
The test results for the strain gauges located at F4 and F14 stations on longeron inner
flange is already given in Chapter 3. These gauges have strains that are below the
gauge resolution and their strain values are too low when compared to the gauges
which have linear responses. After the checks and confirmations for the installations,
the reasons of such behaviour are discussed in this section. The FEA results for these
gauges are not as low as expected. Upper longerons deform in lateral and vertical
directions under the cabin pressure. The combination of canopy hook loads and side
pressure loads deform the longerons and final deformed body takes its shape
according to the support points of the longerons such as frames, bulkheads, sheet
support, tie bars and horse shoe. The translational and rotational deformation cause
stress transition points occurring locally on the structure. The sign of the stresses
begin to change through these points and at the very vicinity of these points, stresses
become nearly zero as in the case of neutral points on the structures. Thus, support
conditions highly affect the stress values of the longerons, especially at its
connection regions with the supports.
95
At the F3 and F12 stations, there are front and rear tie bars located respectively. Also
there is a horse shoe located between the F12 and F13. These lateral support
structures for the upper longerons affect the stresses that exist on the inner flanges of
the longerons at F4 and F14 stations. Because of the combined loading at these
locations, stress transition points do exist. The results for the F4 and F14 stations
have to be refined by considering the lateral supports in detail. Also for station F6,
longeron inner flange stress result has to be refined as well. It can be concluded that,
longeron deforms more in FEA than they behave in the tests. The FEA longeron
inner flange stress distribution through the longitudinal direction (X-direction) of the
cockpit is shown with longeron inner flange test results in Figure 4.4.2.
B1 F1 F2 F3 F4 F5 F6 F7 F8 F9 B2 F10 F11 F12 F13 F14 F15 F16 F17 F18 B3 F19 F20 B4
Figure 4.4.2: Longeron inner flange stress distribution with test results
As the inner flange, the longeron outer flange stresses are also affected from the
support conditions. The only parameter which is different in the deformation
variation for outer flange is its extra support; namely the skin. The U-shape form of
96
the upper longeron has three sections which are for outer flange, web and inner
flange. The inner flange is free except its lateral supports. However, in addition to the
lateral supports, outer flange is continuously supported by the skin. Therefore outer
flange is less sensitive to deformations than the inner one. This situation is clearly
seen from the station F4 results of the inner and the outer flanges. It can be seen from
Table 4.4.1 that the test and the FEA results for the outer flange are not compatible
with each other. The differences in results are due to the lateral supports as in the
case of inner flange. The FEA longeron outer flange stress distribution through the
longitudinal direction of the cockpit is shown with longeron outer flange test results
in Figure 4.4.3.
Figure 4.4.3: Longeron outer flange stress distribution with test results
As it is mentioned already, the combination of the side cabin pressure, canopy hook
and floor loads cause frames to stretch, bend and twist. As a result of these
deformations, stress distributions on the frames show extreme variations. Frame
97
inner cap stress variations through the height of the frames in FEA are plotted with
frame inner cap test results from Figure 4.4.4 to Figure 4.4.12.
Bottom Top
Figure 4.4.4: F2-Frame inner cap stress distribution with test results
Bottom Top
Figure 4.4.5: F-3 Frame inner cap stress distribution with test results
98
Bottom Top
Bottom Top
99
Bottom Top
Bottom Top
100
Bottom Top
Bottom Top
101
Bottom Top
When frame inner cap stress results are inspected, at the first glance, satisfactory
correlation between FEA and test results can be observed for the F2, F3, F5, F6, F12,
F13 and F17 frame stations. For the frame F4, gauge at station F4-a is below the
gauge resolution level in the test and gauge at station F4-b result shows linear
variation under increasing cabin pressure. In FEA, the stress level for the gauge at
station F4-a is higher than that of in the tests as in the case of the longeron inner
flange gauges that have strains below the gauge resolution. On the other hand, station
F4-b result shows a good correlation with test result on the same frame. As gauge at
station F4-b does, the gauge on the F14 frame also has strains that are below the
gauge resolution in the test. The FEA result for this gauge shows very low stress
level as expected and mentioned in Chapter 3. From Figure 4.4.4 to Figure 4.4.12, it
can be concluded that all of the tested points located at the frame inner caps capture
the stress trends obtained from the FEA except gauge located at station F4-a.
102
As it is observed from the Table 4.4.1, the correlation is satisfied between the test
and the FEA stress results of the SG points at the skin, bulkhead upper caps, tie bars
and the sheet support.
4.5 Actions taken to Update the FEM
According to the checks for the correlation, it is concluded that, the differences
between the FEA and test results are mainly due to the lateral support structures. The
deformations of the longerons and the frames are highly affected from these
supports. When the FEA results for the longerons are inspected, it can be said that,
longerons deform more in FEA more than they deform during the tests. In order to
achieve a better correlation, actions are taken to update the lateral support structures
of the cockpit.
Updating of the Sheet Support
The sheet support is modelled without its cut-outs in the FEM due to the
simplification reasons. The update for the structure is done by modelling these cut-
outs as well. The reason to take this action is such that the amount of material
reduced from the support is at serious levels when compared to the existing size of
the structure. Also, the stress level on the sheet support is higher in FEA than the
ones in test results and it may possibly affect the results of the F2 frame. Therefore,
when lateral stiffness is taken into account, modelling the cut-outs becomes worthy.
The detailed FEM of the original and updated sheet supports are shown in Figure
4.5.2 and Figure 4.5.2 respectively.
103
Figure 4.5.1: Detailed FEM of the original sheet support-a
Figure 4.5.2: Detailed FEM of the updated sheet support-b
104
Updating of the Tie Bars
Tie bars are fully modelled with CBEAM elements and at the connection locations to
the frames and therefore rigid type elements are not used. Although the correlation
checks show that this type of simplification gives reasonable results, side sections of
the tie bars are updated due to its connection region with frames. The actual plate
form of the side sections is simulated with CQUAD4 elements and connection to the
frames are satisfied with RBE2 elements by setting three rotational DOF as free. The
aim is to expand the area of the connection and to get better stress results from the
frames F3 and F12. It is thought that, expanding the connection area and assigning
free rotational DOF for the connections, results are provided better in bending
deformation variation for the frames. At the intersection point of the main bar section
with the side section, RBE2 type element is used. All the six translational and
rotational DOF’s are constrained at the connection points. In real structure, at the
transition region, there is a radius and this radius transmits bending loads between
the main and the side sections. In FEM of the tie bar, simulation of this effect with
lonely common grid point of the bar and the plate element is impossible. In order to
provide the rigidity of the model for the static analysis, RBE2 type elements are used
in these regions. The detailed FEM of the original front and rear tie bars and their
updated versions are shown from Figure 4.5.3 to Figure 4.5.6.
105
Figure 4.5.3: Detailed FEM of the front tie bar-a
Figure 4.5.4: Detailed FEM of the updated front tie bar-b
106
Figure 4.5.5: Detailed FEM of the rear tie bar-a
Figure 4.5.6: Detailed FEM of the updated rear tie bar-b
107
Canopy Support Fittings & Canopy Drive Shaft
Canopy support fittings and canopy drive shaft are not modelled at the beginning of
the study. It was thought that modelling these structures does not have a very large
impact on the load path of the cockpit under cabin pressure. The modelling of the
fittings also brings an additional run time to the FEM during the analysis stage.
However, during the checks performed for the correlation, it is understood that,
deformations of the longerons are not in expected levels. In order to get less longeron
deformation results from the FEA, these structures are also modelled.
Canopy support fitting is a complex-shaped structure which supports mainly the

drive shaft mechanism of the canopy. It also carries portion of upper longeron side
loads under cabin pressure. It locates between frames F9 and F10 in the cockpit. This
fitting is connected to the upper longeron inner flange, bulkhead web and bulkhead
upper cap. Thus, it works all in lateral, longitudinal and vertical directions. Such a
complex geometry has lots of fillets, flanges and radiuses. Therefore, canopy support
fittings are modelled with CTETRA type solid elements. At the connection points,
only three translational DOF is constrained with the help of RBE2 type elements.
The rotational constraints are released for all of the connections, especially for
bulkhead upper cap connection. At this location there is only one fastener and in real
life, the fastener can rotate around its axis. Between the canopy support fittings, there
is a canopy drive shaft. Shaft is connected to the bushings of the fittings. Between the
drive shaft and the bushings, there is a relative motion, and this connection should be
modelled with contact type elements for the static analysis. However, contact-type
element having non-linear feature is a computationally expensive one. Therefore,
instead of using contact type elements, shaft-fitting connection is provided by using
RBE2 type elements. The rotational and translational DOF’s of the shaft in its axis
are assigned as free. Thus, shaft is allowed to rotate around its axis and lateral
stiffness of the shaft is not added into the model. The detailed FEM of the canopy
support fittings and the canopy drive shaft are shown in Figure 4.5.7 and Figure 4.5.8
respectively.
108
Figure 4.5.7: Detailed FEM of the canopy support fittings and the drive shaft
Figure 4.5.8: Connections of the canopy support fittings and the drive shaft
109
4.6 Updated Results
The comparison between the updated FEM (U_FEM) stress results with test results
are tabulated in Table 4.6.3.
Table 4.6.1: The comparison between test and updated FEM results
Longeron Stress Stress Longeron Stress Stress Longeron Stress Stress

Web (ksi) (ksi) inner flange (ksi) (ksi) outer flange (ksi) (ksi)
Location TEST U_FEM Location TEST U_FEM Location TEST U_FEM
B1 2.8 2.6 F4 0.1 -0.2 F4 0.9 0.5
F1 3.0 2.5 F6 -4.7 -5.1 F9 1.3 0.9
F2 2.6 2.7 F14 0.0 -0.6 F12 -0.8 -0.1
F20 1.8 1.5 F15 -3.4 -3.5 F13 2.9 3.0
F18 -1.4 0.7
Stress Stress
Frame Stress Stress Max. Max. Bulkhead Stress Stress
Skin
Inner Cap (ksi) (ksi) Principal Principal Upper Cap (ksi) (ksi)
(ksi) (ksi)
Location TEST U_FEM Location TEST U_FEM Location TEST U_FEM
F2-a 2.0 2.5 F4 3.1 3.3 B2 3.3 3.0
F2-b 0.8 1.1 F13 3.7 4.0 B3 1.6 1.4
F3-a 2.6 3.0 F16 5.4 5.0
F3-b 2.6 2.4
F4-a -0.3 0.5 Tie Bars Stress (ksi) Stress (ksi) Sheet Support Stress (ksi) Stress (ksi)
F4-b 3.9 4.0 Location TEST U_FEM Location TEST U_FEM
F5-a 4.2 4.5 F3 9.2 9.1 F1 9.8 9.6
F5-b 6.4 6.0 F12 7.8 7.9
F6-a 5.5 5.6
F6-b 3.3 2.9
F12-a 0.6 0.4
F12-b -2.6 -2.2
F13-a -1.9 -1.4
F13-b 2.5 1.9
F14 -0.2 0.2 U_FEM : Updated FEM
F17 -8.6 -8.5
As it is observed from Table 4.6.1 that, the gauges which are below the gauge
resolution in the test (coloured in yellow), updated FEA results are closer to the test
results than the results obtained from original FEA. Although still there are sign
differences for these gauges, stress levels are very close to the zero stress level. Also
updated results for longeron outer flange locations at F9 and F12 are better in sign
110
and in magnitude when compared with the results in original results. It is also
observed that, the stress level for longeron inner flange location at F18 does not
change much after updates. The percentages of difference between test and updated
FEM results are given in Table 4.6.2.
Table 4.6.2: The percentage of difference between test and updated FEM results
Longeron Difference Longeron Difference Longeron Difference

Web (%) inner flange (%) outer flange (%)
Location TEST-U_FEM Location TEST-U_FEM Location TEST-U_FEM
B1 -6 F4 -319 F4 -43
F1 -18 F6 9 F9 -32
F2 4 F14 2831 F12 -88
F20 -13 F15 3 F13 3
F18 -152
Frame Difference Difference Bulkhead Difference

Skin
Inner Cap (%) (%) Upper Cap (%)
Location TEST-U_FEM Location TEST-U_FEM Location TEST-U_FEM
F2-a 22 F4 6 B2 -10
F2-b 48 F13 9 B3 -13
F3-a 14 F16 -8
F3-b -8
F4-a -267 Tie Bars Difference (%) Sheet Support Difference (%)
F4-b 1 Location TEST-U_FEM Location TEST-U_FEM
F5-a 8 F3 -1 F1 -2
F5-b -6 F12 2
F6-a 2
F6-b -13
F12-a -29
F12-b -15
F13-a -25
F13-b -23
F14 -184
F17 -1
As stated before in section 4.4, the percentage errors between the results given in
Table 4.6.2 are also unrealistic due to the low stress levels. The comparison of the
updated FEM (U_FEM) stress results with original FEM (O_FEM) and test results
are tabulated in Table 4.6.3.
111
Table 4.6.3: The comparison between updated FEM, original FEM and test results
112
When Table 4.6.3 is inspected, it is seen that longeron web results are not affected
too much from the updates. The only expected effect that may come from the sheet
support update does not even affect the results much. It is interpreted that sheet
support stiffness does not change load distribution of the longeron web, which is
actually supported by longeron inner and outer flanges. It can be concluded from
these results obtained for the longeron web that a considerable correlation is satisfied
between the test and the updated FEM stress results.
The updated FEM results for the longeron inner flange show a better correlation with
test when compared with the original FEM results. For the stations F4 and F14 which
has lower strain values than the gauge resolution during the tests, stress results are
too low as expected but it is not the case in original FEM. Additionally, stress result
of station F6 shows a better correlation. The results at the stations F15 and F18 on
the longeron inner flange are not affected from the updates. It is because of the
update locations which are mainly at the stations F1, F3, F9, F10 and F12 stations in
the cockpit. After station F15, results do not change much due to this particular
reason. The updated longeron inner flange stress distribution through the longitudinal
direction of the cockpit (X-direction) is shown with the test and original FEM results
in Figure 4.6.1.
113
Figure 4.6.1: Updated longeron inner flange stress distribution
It can easily be seen from the Table 4.6.3 that the updated FEM results for the
longeron outer flange is much better than results obtained from the original FEM.
The stress results at the locations of F4, F9, F12 and F13 on the longeron outer
flange are affected in considerable levels from the updates at the locations of F3, F9,
F10 and F12. It can be concluded for the longeron outer flange that a considerable
correlation is satisfied between the test and the updated FEM stress results. The
updated longeron outer flange stress distribution through the longitudinal direction of
the cockpit (X-direction) is shown with the test and original FEM results in Figure
4.6.2.
114
Figure 4.6.2: Updated longeron outer flange stress distribution
When Table 4.6.3 is inspected for the frame inner cap updated results, it can be said
that updated sheet support provides better correlation for the frame F2. Updated
frame F-2 inner cap stress variation through the height of the frame is also plotted in
Figure 4.6.3 with the test and original FEM results for comparison purposes.
115
Bottom Top
Figure 4.6.3: Updated F2-Frame inner cap stress distribution
As it is observed in Table 4.6.3, updated front tie bar located at frame F3 provides
better correlation for the frame F3. Updating the front tie bar also changes the
bending deformation at the top of the frame F3 where it is located near the upper
longeron. Updated frame F-3 inner cap stress variation through the height of the
frame is also plotted with the test and original FEM results in Figure 4.6.4.
116
Bottom Top
The stress result for the point at station F4-a, which shows lower strain value than the
gauge resolution during the tests, turns out to be too low after the updates as
expected. It can be seen from Table 4.6.3 regarding the frames F5 and F6 that
updates do not affect the original results much. Updated frames F4, F5 and F6 inner
cap stress variations through the height of the frames are plotted with the test and
original FEM results from Figure 4.6.5 to Figure 4.6.7 respectively.
117
Bottom Top
Bottom Top
118
Bottom Top
The updated rear tie bar located at F12 provides better correlation for F12 frame.
Updating the rear tie bar also changes the bending deformation at the top of the
frame F12 as it is the case for the front tie bar. For frame F13, updated results show
better correlation with the test results than the original FEM ones. A huge differences
in stress variations are observed at the top of the frame F13. For the F14 and F17
frames, updates do not affect the original FEM results. Updated F-12, F13, F14 an
F17 frames inner cap stress variations through the height of the frames is plotted with
the test and original FEM results from Figure 4.6.8 and Figure 4.6.11 respectively for
comparison purposes.
119
Bottom Top
Bottom Top
120
Bottom Top
Bottom Top
As it can be seen from the Table 4.6.3, while updates do not affect the results for the
skin and bulkhead upper caps much, better correlation is obtained on the tie bars and
the sheet support results after the updates performed for these structures.
121
4.7 Summary
In this section, after confirmations made for the reliability of the FEA, the
methodologies defined in order to gather the results are presented. Then, the
comparisons between FEA and test results are made and correlations are checked. In
order to get better correlation, some actions for updating the FEM are performed and
updated results are re-checked to observe the improvements in the correlation. For
the longeron webs, a good correlation is satisfied with the original FEM and thus
updates do not affect the results much. However, for the longeron inner and outer
flanges, updates affect the results in considerable levels and better correlation is
achieved in comparison with the test results. Because of the update locations, results
do not change in considerable levels from frame F15 through the rear end of the
cockpit. All of the tested points located at the frame inner caps capture the stress
trends obtained from the original FEM except at F4-a location. After performing
further updates, FEA results show better correlation with the test results and expected
results for F4-a location are also obtained. For the frames F3, F12 and F13, stress
variation through the height of the frames change in significant levels after the
updates performed for the tie bars. After updating, the results for skin and bulkhead
upper cap do not change much and for the tie bars and sheet support, better
correlation is obtained.
122
CHAPTER 5
CONCLUSION
5.1 General Conclusions
The main aim of this study is to validate a detailed FEM of the jet trainer cockpit
structure with ground pressurisation test results. For this purpose, first the load
carrying/transfer mechanism of the cockpit structure is investigated. During this
process, assumptions for the FE modelling are made such that to ensure the
necessary and sufficient conditions for the correlation study. The assumptions reflect
mainly the Global FEM logic. According to these assumptions, the FEM of the
cockpit structure is constructed by using commercial finite element software
MSC/PATRAN® and MSC/NASTRAN®. After the checks performed for the
obtained FEM (i.e. the original FEM), loads and boundary conditions are assigned to
the model and linear static FEA is performed.
In order to validate the FEM, cockpit is pressurised on the ground. In the scope of the
test procedure, the criteria for the selection of the sensors used for the test, their
installations and numbering procedures are presented. After performing the tests, the
obtained strain results are then interpreted on the basis of their linear or non-linear
responses under the linearly increasing cabin pressure. The installation checks are
also made for the strain gauges which show non-linear behaviour. After the
confirmation for the installations, it is concluded that their strain values are too low
123
in comparison to the gauges having linear responses. It is thought that the strain
gauges may not response to the low strain values properly and then, in order to make
a physical comparison with FEA, the obtained strain results are converted to their
corresponding stress values.
For the correlation study performed between FEM and the test results, first, the
response of the structures under cabin pressure is checked. Having obtained
reasonable deformation results, methodologies to gather the stresses from the FEA
results are explained. This part of the study is as important as the modelling
techniques applied for the FEM as the commercial finite element software,
NASTRAN®, extracts and offers different types of stress values after the analysis
performed and therefore, in order to make true comparison with the test results, the
methodology chosen by the engineer in gathering the FEA results is vital.
The results obtained from the FEA are compared with the test results and checks are
made for the correlation. At the first run of the model, the obtained stress results are
in reasonable levels for most of the points measured during the test. Moreover, the
differences in stress levels are too low when compared with the material yield stress
limit. However, especially for the longeron inner and outer flanges and for some of
the frames, a need arises for the refinement of the stress results to achieve a better
correlation. For this reason, some actions are taken to update the original FEM. The
updates are performed for some of the lateral structures in the cockpit. After this
second run of the model, the results obtained from the FEA are re-checked for the
correlation purposes. The desirable stress values are then obtained for all of the
points measured in the test and FEM is finalised after achieving a good correlation
between the FEA and the test results. This final form of the FEM (i.e. updated FEM)
of the cockpit structure is now ready to serve as a benchmark for any future
modification and/or correlation studies by also proving itself a very reliable one.
124
5.2 Recommendations for the Future Work
This structural analysis study can also be performed by considering the damage
tolerance issues regarding the aging of the aircraft.
The fasteners can be modelled with their elastic properties at the desired primary
structure connections. This application can be performed in order to get knowledge
of the refined stresses for further fatigue analyses and/or for the repairing purposes to
be performed at specific locations.
The obtained updated model can be correlated with flight tests under some specific
load case; such as pull-up manoeuvre combined with cabin pressure as in the case of
flight above 8000 [ft]. For this purpose, canopies and the windshield with its close
shape should be modelled to transfer the aerodynamic loads to the cockpit. It may
also require modelling of the centre fuselage in order to simulate boundary condition
effects at the end of the rear cockpit in a better way. In addition to the model updates,
also temperature compensation should be used during the flight tests.
The experiments can be performed by changing the locations of the strain gauges and
locating them over the structure where they provide higher strains.
125
REFERENCES
[1] Northrop Grumman Corporation,

http://www.as.northropgrumman.com/products/t38talon/index.html, last
visited on November 18th, 2011.
[2] NASA Dryden Flight Research Center Photo Gallery,

http://www.dfrc.nasa.gov/Gallery/Photo/T-38/Large/ED06-0072-8.jpg, last
[3] Wikipedia The Free Encyclopedia,

http://en.wikipedia.org/wiki/Dassault/Dornier_Alpha_Jet, last visited on
November 18th, 2011.
[4] BAE Systems Company,

http://www.baesystems.com/ProductsServices/bae_prod_air_hawk.html, last
[5] U.S. Department of Transportation Federal Aviation Administration,

Airframe and Powerplant Mechanics Airframe Handbook, 1972.
[6] Yusuf, K., Nukman, Y., Dawal, S. Z., Chandra D., Sofia, N. (2010, April).
Conceptual Design of Fuselage Structure of Very Light Jet Aircraft, Latest
Trends on Theoretical and Applied Mechanics, Fluid Mechanics and Heat
& Mass Transfer.
[7] Megson, T.H.G. (2007). Aircraft Structures. UK: Elsevier Aerospace

Engineering Series.
[8] Aircraft Design: Synthesis and Analysis Textbook, Stanford University,

http://adg.stanford.edu/aa241/AircraftDesign.html, last visited on November
24th, 2011.
[9] Chun, M., Niu, Y. (1988). Airframe Structural Design. California: Conmilit
Press Ltd.
[10] Defence Industry Daily, http://www.defenseindustrydaily.com/france-

modernizing-alpha-jet-trainer-avionics-04661/, last visited on November
24th, 2011.
[11] Seshu, P. (2006, June). Finite Element Analysis. Prentice Hall.
126
[12] Pepper, D. W., Heinrich J. C. (1992). The Finite Element Method: Basic
Concepts and Applications. US: Hemisphere Publishing Corporation.
[13] Chandrupatla, T. R. (2004). Finite Element Analysis for Engineering and

Technology. India: Universities Press.
[14] Bhavikatti, S.S. (2005). Finite Element Analysis. New Delhi: New Age
Publishers.
[15] Liu, G.R., Quek, S.S. (2003). The Finite Element Method: A Practical
Course. UK: Elsevier.
[16] Schaeffer, H.G. (2001). MSC.Nastran Primer for Linear Analysis. US:
MSC.Software Corporation.
[17] Practical Application of Finite Element Analysis to Aircraft Structural

Design (1986). AGARD Lecture Series No.147. US.
[18] Flight Global,

http://www.flightglobal.com/airspace/media/militaryaviation1946-
2006cutaways/images/82812/aermacchi-m346-cutaway-drawing.jpg, last
[19] Montana State University, College of Engineering, Introduction to

Aerospace Lecture Notes,
http://www.coe.montana.edu/me/faculty/cairns/Introduction%20to%20Aero
space-Web/Internal%20Loads.pdf, last visited on November 28th, 2011.
[20] An Assessment of the State-of-the-art in Multidisciplinary Aeromechanical

Analyses (2008, January). AMES Research Center Report. US.
[21] Chun, M., Niu, Y. (1999). Airframe Stress Analysis and Sizing. Hong Kong:
Conmilit Press Ltd.
[22] ANSYS®, User’s Manual for Revision 5.1, ANSYS® Inc., 1994
[23] ABAQUS®, User’s Manual Version 6.4, ABAQUS® Inc., 2003
[24] MSC.NASTRAN®, User’s Guide, MSC.Software©, 2001
[25] Sharpe, Jr., William N. (2008). Springer Handbook of Experimental Solid

Mechanics. New York: Springer.
[26] Window, A.L. (1992, November). Strain Gauge Technology. Springer.
127
[27] Strain Gauge Measurements on Aircraft (1976, April). AGARDograph
No.160 Vol.7. London: Technical Editing and Reproduction Ltd.
[28] Certification Specifications for Normal, Utility, Aerobatic, and Commuter

Category Aero planes, CS-23-AMC 23.307(2010, September). European
Aviation Safety Agency.
[29] Vishay® Inc,

http://www.vishaypg.com/docs/11298/250un.pdf, last visited on February
2nd, 2012.
[30] Vishay® Inc,

http://www.vishaypg.com/docs/11225/125ur.pdf, last visited on February
2nd, 2012.
[31] Endevco® Inc,

http://www.endevco.com/product/prodpdf/8540-500.pdf, last visited on
February 2nd, 2012.
[32] National Instruments® ,

http://www.ni.com/pdf/manuals/375102d.pdf, last visited on February 2nd,
2012.
[33] National Instruments® ,

http://sine.ni.com/nips/cds/view/p/lang/en/nid/208800,last visited on
February 2nd, 2012.
[34] LabVIEW User Manual, National Instruments® Inc., 2011
[35] University of Glasgow,

http://www.mech.gla.ac.uk/~sharpj/strain_extra.pdf, last visited on February
3rd ,2012.
128

Structural Analysis of A Jet Trainer Cockpit

Uploaded by

Copyright:

Available Formats

Structural Analysis of A Jet Trainer Cockpit

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Structural Analysis of A Jet Trainer Cockpit

Uploaded by

Copyright:

Available Formats

STRUCTURAL ANALYSIS OF A JET TRAINER COCKPIT

MUHİTTİN NAMİ ALTUĞ

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS

STRUCTURAL ANALYSIS OF A JET TRAINER COCKPIT

submitted by MUHİTTİN NAMİ ALTUĞ in partial fulfillment of the requirements

Prof. Dr. Canan Özgen _____________________

Prof. Dr. Ozan Tekinalp _____________________

Assist. Prof. Dr. Melin Şahin _____________________

Examining Committee Members:

Prof. Dr. Yavuz YAMAN _____________________

Assist. Prof. Dr. Melin ŞAHİN _____________________

Assist. Prof. Dr. Demirkan ÇÖKER _____________________

Assist. Prof. Dr. Ercan GÜRSES _____________________

Dr. Muvaffak HASAN _____________________

Name, Last Name : Muhittin Nami Altuğ

STRUCTURAL ANALYSIS OF A JET TRAINER COCKPIT

Altuğ, Muhittin Nami

February 2012, Pages 128

Keywords: Structural Analysis, Jet Trainer Aircraft, Cockpit, Finite Element

BİR JET EĞİTİM UÇAĞI KOKPİTİNİN YAPISAL ANALİZLERİ

Altuğ, Muhittin Nami

Şubat 2012, 128 sayfa

I would like to express my gratitude to my supervisor Assist Prof. Dr. Melin

I must also express my appreciation to my superiors at TAI, Dr. Gürsel

I would also like to thank my colleagues Enver Özakkaş, Bilgin Çelik,

ACKNOWLEDGMENTS ......................................................................................... vii

TABLE OF CONTENTS .......................................................................................... viii

LIST OF TABLES ....................................................................................................... x

LIST OF FIGURES .................................................................................................... xi

1.1 Motivations of the Study .............................................................................. 1

2.FINITE ELEMENT MODELING AND ANALYSIS OF COCKPIT STRUCTURE

2.1 Introduction ................................................................................................ 18

3.1 Introduction ................................................................................................ 55

4. CORRELATION STUDIES BETWEEN FINITE ELEMENT ANALYSIS AND

4.1 Introduction ................................................................................................ 82

5. CONCLUSION .................................................................................................... 123

5.1 General Conclusions ................................................................................ 123

1.1 Motivations of the Study

1.2 Objectives of the Study

The objectives of this study can be listed as follows:

• Investigating the load carrying/transfer mechanism of the cockpit structure

1.3 Literature Survey

The importance of air dominance in the defence of countries is increasing. This is

Figure 1.3.2: Semi-monocoque construction [5]

Canopy and Windshield

Figure 1.3.3: Alpha Jet-E, Canopy and Windshield structures [10]

Classical analytical methods consider a differential element and develop the

FEM requires a problem defined in geometrical domain to be subdivided into a finite

In engineering problems, unknowns are infinite in a continuum. The finite element

• Selecting suitable field variables and the elements.

Successful simplification depends on;

Figure 1.3.4: Aermacchi M-346 cutaway drawing [18]

In making an analysis of large structural components such as a wing, fuselage etc.,

Modelling of the complex structural components with low order elements as

Figure 1.3.5: Integrated FEM for Boeing 767-400ER [17]

Despite encouraging results from simulation and experimental modelling, structural

Structural Testing Activities

The purpose of a strain gauge is to convert the physical changes occurring in a

Figure 1.3.6: Strain gauge types [25]