simulation of fretting fatigue, cracking in axial disc-blade … · 2013-07-18 · 1.4.1...
TRANSCRIPT
i
Simulation of Fretting Fatigue, Cracking in
Axial Disc-Blade Dovetail Joint in Aero Engine
Compressors
Zaheer-ul-Hassan 07F-UET/PhD-ME-39
Department of Mechanical Engineering
Faculty of Mechanical & Aeronautical Engineering
University of Engineering & Technology
Taxila - Pakistan
July 2013
ii
Simulation of Fretting Fatigue, Cracking in
Axial Disc–Blade Dovetail Joint in Aero
Engine Compressors
Author
Zaheer-ul-Hassan 07F-UET/PhD-ME-39
A dissertation submitted in partial fulfillment of the requirement for
the degree of
Doctor of Philosophy in Mechanical Engineering
Thesis Supervisor
Prof. Dr. Shahab Khushnood
Thesis Supervisor’s Signature:-____________________________ ___________________________________ ____________________________________
External Examiner’s Signature External Examiner’s Signature
Department of Mechanical Engineering
University of Engineering & Technology-Taxila,
Pakistan
July 2013
iii
Supervisor Prof. Dr. Shahab Khushnood
Members of Research Monitoring Committee Prof. Dr. Zubair
Prof. Dr. Khalid Akhter
Associate Prof. Dr. Mirza Jahanzab
Foreign Research Evaluation Experts
iv
ACKNOWLEDGEMENT
All thanks and pray to my creator ALLAH who gave me strength, ability,
prospect and determination to complete this work.
I am deeply indebted and grateful to Pakistan Ordnance Factories (POFs)
particularly its Chairman Lieutenant General Muhammad Ahsan Mahmood HI
(M) and Member production board Mr. Muhammad Asif without which the
realization of the research would have not been possible.
I am greatly thankful to my supervisor Professor Dr. Shahab Khushnood and
Professor Dr. M. M. I Hammouda for their timely and invaluable guidance,
encouragement and helpful suggestions which was no doubt a challenging
task but due to their missionary zeal and persistence efforts beyond the
normal call of duty to make possible the designing, development and
manufacturing of experimental setup. Special thanks to Professor Dr. Khalid
Akhtar, Associate Professor Dr. Mirza Jahanzaib and Dr. Asif Hussain Malik,
who provided the valuable guidelines regarding the designing, development
and manufacturing different parts of the experimental setup and made this
research possible owing to their hard work, pain taking efforts and
commitment. I would like to express deep gratitude to my officers specially
Mr Aftab Iqbal, Mr Qaiser Nadeem and Mr Muhammad Abid for their
supportive attitude.
I must acknowledge my parents and parents-in law who always prayed for
my success and looked after my personal interests during execution of this
research work. Last but not the least the warmest thanks to my wife Naheed
Zaheer Associate Professor in Physics and children Muhammad Bilal Zaheer
and Amna Zaheer for their patience, support and encouragement. I will
remain indebted in future due to their moral support during the study and
preparation of this research work.
Zaheer-Ul-Hassan
v
DEDICATED TO
All those who are working sincerely with dedication, devotion and have
commendable professional ability displayed by setting an example of
excellence in the performance of their job, beyond the normal call of their
duty for Pakistan.
vi
Table of Content Acknowledgement iv
Research work publication v
Table of contents vii
List of figure ix
List of tables xiii
Nomenclature xiv
Abstract xvi
Chapter I: Introduction1 1
1.1 Introduction 21.2 Fretting fatigue history 4 1.3 Research objectives 6 1.4 Organization and management of thesis 7 1.4.1 Methodology 8 1.4.2 Problem formulation 8 1.4.3 Theoretical study 8 1.4.4 Designing and development of suitable experimental test rig 8
1.4.5 Dimensions of analyzed dove tail joint including disc and blades 8
1.4.6 Simulated loading cycle 9 1.5 Thesis structure 9 Chapter 2: Literature survey 11
2.1 Introduction 12 2.2 Dovetail joint analysis 13 2.3 Fretting fatigue- applications 15 2.4 Fretting fatigue- steps of study 16 2.5 Fretting fatigue- performance affecting variables 172.6 Fretting fatigue - phases 212.7 Fretting fatigue - numerical approaches 24 2.8 Predication of fretting fatigue performance 25 2.9 Fretting fatigue - experimental test rigs 26 Chapter 3: Mathematical formulation 27
3.1 Numerical approaches to fretting fatigue 28 3.2 Prediction of fretting fatigue performance 29
3.2.1 Special parameters for pure fretting 29 3.2.2 Critical plane approaches 30
3.3 Fracture mechanics based approaches 35 Chapter 4: Experimental system and data 39
4.1 Introduction 40
vii
4.2 Experimental fretting fatigue test rigs 40 4.2.1 Bridge type fretting pads 40 4.2.2 Cylindrical type fretting pads 41 4.2.3 Related to Special geometry 41 4.3 Designing and development of experimental test rig 48 4.4 Preparation of component drawing of experimental test rig 534.5 Mechanical parts 53 4.6 Detail of various electrical parts 53 4.7 Working of newly developed test rig 54 4.8 Validation of experimental testing system 55 4.8.1 Running trend of Experimental test rig – without specimen 55 4.8.2 Running trend of Experimental test rig – with specimen 56 4.9 Experimental work 56 4.9.1 Introduction 56 4.9.2 Test rig 57 4.9.3 Materials 57 4.9.4 Test specimen 58 4.9.5 Simulated loading cycle 59 4.9.6 Test program 60 Chapter 5: Numerical analysis
65
5.1 Introduction 66 5.2 Numerical work 66 5.3 Numerical simulation of fretting fatigue of high structural steel 69 5.4 Physical properties of high strength structural steel 69 5.5 Present idealization 71 Chapter 6: Result and discussion 79
6.1 General discussion 80 6.2 Result and discussion 81 Chapter 7: Conclusions and recommendations for future work 86
7.1 Conclusion 87 7.2 Recommendations 87 References 89
Appendix –“A” 116
A-1 Manufacturing detail of mechanical components 117 A-1.1 Fixture of experimental test rig along with specimen 117
A-1.2 Upper holding plat, disc, blades and lower holding plate sub assembly 117
A-1.3 Main motor and motor holding subassembly 118
A-1.4 Main motor, motor holding plate and safety guard sub assembly 118
A-1.5 Main motor, motor holding plate, safety guard ,specimen sub assembly and safety cover plate 119
A-1.6 Stand for specimen holding fixture sub assembly 119
viii
A-1.7 Main motor holding plate 120 A-1.8 Safety guard 120 A-1.9 Lower holding plate 121 A-1.10 Upper holding plate 121 A-1.11 Safety Cover Plate 122 A-1.12 Tightening Knob 122 A-1.13 Gripping piece 123 A-1.14 Hinge pins 123 A-1.15 Pillars 124 A-1.16 Base plate 124 A-1.17 Ratchet sub assembly 125 A-1.18 Base 125A-2 Detail of various electrical parts 126 A-2.1 Main motor drive 126 A-2.2 Variable auto transformer 126 A-2.3 Control motor 127 A-2.4 Timer (main on time control) 127 A-2.5 Timer (on time control) 128 A-2.6 Timer (off time control) 128 A-2.7 Limit switches (S.2 and S.3) 128 A-2.8 Relay (1) 128 A-2.9 Relay (2) 128 A-2.10 Tachometer 129 Appendix – “B” 130
B-1 Running trend of Experimental test rig – without Specimen 131B-2 Running trend of experimental test rig – with Specimen 137
ix
List of Figures Chapter I: Introduction
Figure 1.1 Types of surface degradation on a fretting situation [10~11] 2
Figure 1.2 Case study- formed automotive suspension component: a. a formed component: b. detail of the fretting fatigue damaged location [63]
4
Figure 1.3 A block diagram providing general information of the thesis 7
Chapter 2: Literature survey Figure 2.1 Damaged disc and blades due to fretting fatigue failure
phenomenon 12
Chapter 4: Experimental system and data
Figure 4.1 Exploded view of fretting fatigue test rig [235] 42
Figure 4.2 The specimen geometry [235] 42
Figure 4.3 The fretting fatigue apparatus assembled in the servo hydraulic test machine [236] 43
Figure 4.4 Drawing of the elevated temperature dovetail fretting fatigue fixture [54] 43
Figure 4.5 Drawing of the specimen used in elevated temperature dovetail fretting fatigue fixture 44
Figure 4.6 The most commonly used fretting fatigue test rig [256,257 and 258] 47
Figure 4.7 The other most commonly used fretting fatigue test rig [259] 47
Figure 4.8 Schematic of the commonly used type of fretting fatigue test rig [260] 48
Figure 4.9 Experimental testing rig showing. a) The assembly fixture of testing Rig. b) Speed control system of the rig 49
Figure 4.10 Assembly fixture of fretting fatigue test rig 50 Figure 4.10A The exploded view of fixture of fretting fatigue test Rig 51
Figure 4.11 Detail of mechanical parts of fixture fretting fatigue test Riga. 52
Figure 4.12 Detail of electrical and electronic parts used in the test Rig 52
Figure 4.13 Assembly of two sectors of disc and blades in dovetail joints 1. Blade 2. Disc
58
Figure 4.14 Drawing and dimensions of the disc 59Figure 4.15 Drawing and dimensions of the Blade 59
Figure 4.16 Movement of blades at acceleration and deceleration modes 60
x
Figure 4.17 a) Contact surface of the disc and b) blade after a 100 cycle fretting test stopped by reducing the displacement amplitude to zero
61
Figure 4.18 Sliding surface disc with oxides/debris [261] 62 Figure 4.19 Broken disc made of high strength structural steel
material 62
Figure 4.20 Broken disc made of Bakelite material 63 Figure 4.21 Broken disc made of Mild steel material 63 Figure 4.22 Broken disc made of Cost iron material 64 Chapter 5: Numerical analysisFigure 5.1 Dovetail joint disc and blades assembly configuration
Blades 2. disc 67
Figure 5.2 a. Disc and blades model assembled in dove tail joint. b. (A. frictional less support, B. rotational velocity and C. acceleration)
69
Figure 5.3 Alternating stress MPa 70Figure 5.4 Strain life parameters 70Figure 5.5 Three dimensional model of disc and blade assembly fixed
in a dovetail joint 71
Figure 5.6 Meshing adopted in the present analysis i.e. disc and blades assembly 71
Figure 5.7 The model of disc after meshing. b. The enlarged view of model of disc after meshing in critical region 72
Figure 5.8 The stress probe 1.2,3,4 and 5 on the fillet surface of the disc 73
Figure 5.9 Stress intensity of different probe with respect to time 75 Figure 5.10 Maximum principal stress at different probe with respect
to time 76
Figure 5.11 Equivalent stress intensity of different probe with respect to time 77
Figure 5.12 Shear stress XY at different probe with respect to time 78
Chapter 6: Result and discussion Figure 6.1 Maximum shear stress at stress probe 4 82Figure 6.2 Trend of shear stress variation at stress probes 1 to 7 83 Figure 6.3 Maximum principal shear stress at stress probe 4 83 Figure 6.4 Trend of principal shear stress variation at stress probes
1to 7 84
Figure 6.5 Maximum principal stress at stress probe 4 85
Appendix – A 116Figure A-1.1 Fixture of experimental test rig along with specimen 117Figure A-1.2 Upper holding plat, disc, blades and lower holding plate
sub assembly 117
Figure A-1.3 Main motor and motor holding sub assembly 118
xi
Figure A-1.4 Main motor, motor holding plate and safety guard sub assembly
118
Figure A-1.5 Main motor, motor holding plate, safety guard ,specimen subassembly and safety cover plate 119
Figure A-1.6 Stand for specimen holding fixture subassembly 119Figure A-1.7 Main motor holding plate 120Figure A-1.8 Safety Guard 120Figure A-1.9 Lower holding plate 121Figure A-1.10 Upper holding plate 121Figure A-1.11 Safety cover plate 122Figure A-1.12 Tightening knob 122Figure A-1.13 Gripping piece 123Figure A-1.14 Hinge pins 123Figure A-1.15 Pillars 124Figure A-1.16 Base plate 124Figure A-1.17 Ratchet sub assembly 125Figure A-1.18 Ratchet sub assembly 125
Appendix – B 130 Figure B-1.1 Running trend of experimental test rig without specimen 132Figure B-1.2 Running trend of experimental test rig without specimen 133Figure B-1.3 Running trend of experimental test rig without specimen 134Figure B-1.4 Running trend of experimental test rig without specimen 135Figure B-1.5 Running trend of experimental test rig without specimen 136Figure B-1.6 Running trend of experimental test rig without specimen 137Figure B-2.1 Running trend of experimental test rig with specimen 138Figure B-2.2 Running trend of experimental test rig with specimen 139Figure B-2.3 Running trend of experimental test rig with specimen 140Figure B-2.4 Running trend of experimental test rig with specimen 142
xii
LIST OF TABLES
Chapter 1: Introduction Table 1.1 Detail of accident carried out during the year 1919~2005 3 Chapter 4: Experimental system and data Table 4.1 Material specification of the specimen 57
Chapter 5: Numerical AnalysisTable 5.1 Stress intensity of different probe with respect to time 74 Table 5.2 Maximum principal stress at different probe with respect to
time 75
Table 5.3 Equivalent stress intensity of different probe with respect to time 76
Table 5.4 Shear stress XY at different probe with respect to time 77 Chapter 6: Result and discussion Table 6.1 Stress variation at stress Probes 1~7 82 Table 6.2 Principal Stress variation at stress Probes 1~7 84
Appendix –“B” Table B-1.1 Running trend of experimental test rig without specimen 131Table B-1.2 Running trend of experimental test rig without specimen 132Table B-1.3 Running trend of experimental test rig without specimen 133Table B-1.4 Running trend of experimental test rig without specimen 134Table B-1.5 Running trend of experimental test rig without specimen 135Table B-1.6 Running trend of experimental test rig without specimen 136Table B-2.1 Running trend of experimental test rig with specimen 138Table B-2.2 Running trend of experimental test rig with specimen 139Table B-2.3 Running trend of experimental test rig with specimen 140Table B-2.4 Running trend of experimental test rig with specimen 141
xiii
Nomenclature
Total fatigue strain amplitude.
∆2
Elastic strain amplitude.
∆2
Plastic strain amplitude.
´ Fatigue strength coefficient.
Young´ s modulus
Number of cycles to crack limitation.
Fatigue strength exponent. ´ Fatigue ductility coefficient.
Fatigue ductility exponent.
Maximum tensile stress normal to the crack plain during
loading cycle (SWT in case of plain fatigue case).
Half of contact width.
Half of the stick zone width.
∆ Shear stress amplitude
Maximum shear stress on critical plain.
∆ , Shear stress along with the effect of mean axial shear
stress on fretting fatigue.
Shear stress ratio
Fitting parameter
, , , Constants obtained experimentally.
∆2 Cyclic yield strength.
Influence factor. ´ Shear fatigue coefficient.
´ Shear fatigue ductility exponent.
Shear modulus.
FP Findley Parameter
xiv
FSSSR Fatemi and Socie Shear Stress Range
FS Fatemi – Socei parameter
MSSR Modified Shear Stress Range
SWT Smith-Watson Topper
xv
Abstract
Various fretting fatigue test rigs were used to test the specimens to study the
fretting fatigue crack initiation phenomenon of different components and
assemblies. Such rigs were categorized according to the specimens used.
There was no standard test rig available for the researchers. This research
work presents the development of a unique, safe and flexible fretting fatigue
test rig for the analysis of fretting fatigue crack initiation in dovetail joint of
aero-engine compressor. The test rig is capable of revolving the specimen i.e.
disc and blades from 0 to 20000 rpm in small increments and back to 0 rpm
in the same manner. In this test rig the unique feature is its revolution along
with its specimen by simulating the aero-engine compressor. In addition to
the test the rigs designed by the past researches this conceptual design test
rig is helpful to test different joints and materials in the fretting fatigue testing
field. The test rig is equipped with high speed motor with controller. It is
flexible to adjust from 0ο to ±90ο with the help of hinge pins whereas the
system as a whole could be rotated from 0ο to 360ο with the help of ratchet.
The test rig could be calibrated with or without loading the specimen.
The aero-planes, vehicles and robotic machineries are used for the
transportation of various equipments. The structural integrity of their main
components must be ensured, inspected and analyzed to avoid any damage
to human life as well as the loaded equipments. The dovetail joints are
commonly used to assemble the blades and the disc in the assembly as well
as low pressure stage of rotating compressors in turbines used for aircraft
propulsion. The fretting fatigue is a serious threat for such joints in the
mechanical components and engineering structures including air and space-
craft components, automobile sub-assemblies, various electrical and
manufacturing equipments. The attachments of structural components of disc
and blades are damaged by the fretting fatigue. Various experimental and
numerical approaches were developed to avoid such fatigue failures.
xvi
In this research work the experiments at room temperature are carried out on
the test rig by revolving the specimens of dovetail joints of aero-engine
compressor, presented and discussed their fretting fatigue behavior. In these
experiments the specimens of mild steel, bakelite, cast iron and structural
steel are used and the fretting fatigue failure of the disc took place at the
edge of the common surface near the dovetail notch base. The modeling and
simulation are carried out using the commercially available software ANSYS
11.0. It is found in each iteration analyses that irrespective of thickness and
type of material of the disc, fillet is the weakest portion of the specimen.
1
Chapter 1
Introduction
This Chapter gives background of dovetail joints of disc and blades of aero engine
compressors, fretting fatigue phenomenon and also the research motivation of
this research, research objectives and organization and management of thesis.
2
1.1: Introduction
Fretting fatigue phenomenon has been investigated since long and the
developments in this field are still in progress. Dovetail joints are used to fix the
blades on the rotating disc for the fan assembly as well as in low pressure stage
of rotating compressors in turbine, used for aircraft propulsion. Fretting fatigue
mechanism deals with the failure analysis that develops multi-axial non-
proportional elastic plastic cyclic stress strain field. Fretting fatigue is defined as
the damage resulting from small amplitude; few micrometers to few hundreds of
micrometers, induced at interface between two contacting bodies where at least
one or both bodies are subjected to fatigue loading. The researchers [1~4]
found that mechanical components are being damaged directly or indirectly due
to fretting fatigue phenomenon caused by many variables that could persuade
fretting failure. The investigators [5~6] found the most important parameters
like temperature, material combination, cyclic frequency, environment condition,
slip amplitude, type of matting surface, frictional force, hardness and contact
pressure that might influence on the fretting fatigue phenomenon, while
Dobromirski et al. [7] stated that all these parameters are interconnected with
each other. Madge et al. [8] gave the best example of coefficient of fraction that
affects the slip distribution depending upon the normal loading and it changes
with the change in the numbers of fatigue loading cycles.
Vingsbo and Soderberg [9] investigated that fatigue, corrosion and wear are the
specific factors involved in degrading the contact faces of the joints. The
researchers [10~11] concluded that fretting could be summarized as partial slip
(fretting fatigue) and gross slip (fretting wear) as shown in Figure 1.1.
Figure 1.1: Types of surface degradation on a fretting situation [10~11]
Fretting Fretting wearFretting fatigue
Corrosion
3
Zhu et al. [12] found that these forms of failure are usually coexisting in the
same contact. Fretting fatigue is linked with partial slip differentiated by early
crack nucleation causing ultimate failure. Fretting fatigue is fretting in the
existence of bulk fatigue load producing crack. Hager et al. [13] stated that
fretting wear is an addition of the damage occurred at interfaces subjected to
high contact stresses attached with low amplitude oscillation. Powell [14] found
that corrosion phenomenon is established in the fretting field resulting in the
fretting wear and fretting fatigue. The researchers [15~16] found that as an
alternative of relatively expensive full scale engine testing, the experimental and
numerical simulation of loaded disc and blades assembly should be encouraged.
In 1998 an air craft was significantly damaged when it was taking-off from Lake
Evers, Florida. During metallurgical investigation of failed fittings, it is found that
failure was due to fretting wear. Moreover, the crack connected with fretting
wear was found on the right side of the fitting attached part, whereas fretting
wear was observed on the left and right side top beam. Few examples of failure
data downloaded from internet [17] are tabulated in Table 1.1.
Table 1.1: Accident detail during the year 1919~2005
Sr. No. Year Airline/Train Reason of failure
1 1968 Los Angeles Flight 417
Lost one of its main rotor blades due to fatigue failure.
2 1985 Japan Airline123
Aircraft lost its vertical stabilizer due to faulty repairs on the rear bulkhead and crashed.
3 1988 Aloha Airline Flight 243
Suffered an explosive decompression due to fatigue failure.
4 1989 United Airline 232
Lost its tail engine due to fatigue failure.
5 1992 E1-A1,Flight 1862
Lost both engines on its right-wing
6 2002 China Airlines Flight 611
Disintegrated in-flight due to fatigue failure.
7 2005 Chalk’s Ocean Airways
Lost its right wing due to fatigue failure.
B
w
F
1
F
fo
1
E
su
n
a
sc
m
th
th
[2
a
H
fa
fr
o
Buciumeanu
was occurre
igure 1.2a
.2b gives th
Figure 1.2
ormed com
1.2: Frettin
Eden and R
urfaces of
oted the fo
reas conta
cientific stu
motion betw
heir interes
hat fretting
27] was the
nd consequ
Hoppner [28
atigue to e
retting life
n fretting f
u et al. [18]
ed due to f
shows the
he detail of
: Case st
ponent, (b)
ng fatigue
Rose [19] g
the specim
ormation of
cting the g
udy of the p
ween the c
sts in the fr
is purely a
e first rese
uence of a t
8] stated t
engineering
reduction f
fatigue con
] stated tha
fretting fati
e formation
f fretting fat
tudy- form
). detail of t
e – history
gave first id
men when o
f iron oxide
grips of a
phenomeno
contacting s
retting in th
a mechanic
earcher who
temperatur
the real ap
g practice
factor. The
ntinued in 1
4
at the failur
igue pheno
of suspen
tigue dama
ed automo
the fretting
y
dentificatio
one of them
e on the su
test mach
on and con
surfaces. T
heir work.
cal process
o discussed
re rise in th
pplications
commence
researcher
1970s and
re in the au
omenon as
nsion of co
age area.
otive suspe
g fatigue da
n of frettin
m or both a
urfaces of
ine. Tomiln
cluded that
The researc
Tomilnson
that involv
d on the po
e contact z
of scientifi
ed around
rs [29~30]
80s for fu
utomotive s
shown in
mponents
ension com
amaged loca
ng fatigue
are in fatig
a fatigue s
nson [20]
t it is relate
chers [21~
and Thorp
ved microch
ossibility of
zone.
c investiga
1965 and
emerged in
rther unde
suspension
the Figure
whereas Fi
mponent: (a
ation [18]
in two ma
gue loading
specimen in
performed
ed to tange
25] mainta
pe [26] stre
hip. Waterh
the occurr
ation of fre
developed
n 1968-69 w
rstanding o
part
1.2.
igure
a). a
atting
and
n the
first
ential
ained
essed
ouse
rence
etting
d the
work
of its
5
mechanisms, and dozens of papers dedicated to it were produced. Hurricks [31]
identified three stages present in fretting mechanisms which are initial adhesion
and transfer of metal, production of debris in oxidized state and steady state
wear conditions. Notable among many important developments later in the same
era was the use of fracture mechanics to assess the crack development under
fretting fatigue conditions. Edwards [32] marked the beginning of the
applications of this tool to the problem. The Investigators [33~34] explored the
issue of fretting fatigue in orthopedic.
Buciumeanu [35] stated that if fretting fatigue exists in any system, damage
effect should be there and it is a sum of axial, normal, tangential loading, effect
of wear and synergetic factor between fatigue and wear. Mathematically it can
be written as;
D F FF W S , , (1.1)
In equation 1.1, D is the degree of degradation, F is the effect of fatigue on
its life due to axial load, FF is the effect of fretting on life due to normal and
tangential loads involved in process of contact, W is the effect due to wear and
S , , is the factor of synergetic between fatigue and wear.
Eden and Rose [36] identified fretting fatigue and observed the production of
iron oxide on the contact surface of a fatigue specimen in the testing machine.
Tomilnson [37] carried out scientific study on fretting fatigue phenomenon and
found that it is due to the tangential motion between surfaces in contact with
each other. The researchers [38~39] participated in this field and helped to
create interest in fretting fatigue with their work. Nowell [40] worked on fretting
crack initiation and studied its effects on fatigue life and for this purpose material
selected was aluminum 14% Cu alloy and became a prominent investigator in
this field. He used cylindrical fretting pads for the experimental component of his
research.
6
The researchers [41~45] found fretting fatigue crack initiation in dove tail joints
of aero engine compressors through experimentally, numerically and analytically
whereas researchers [46~49] stated that there are many types of fretting fatigue
testing rigs starting from simplest to more complex systems. In most of the
system servo hydraulic testing machine was utilized. Hoppner [50] carried out
different case studies to observe fretting fatigue phenomenon. In this field, a lot
of progress was carried out in prevention, design for fretting fatigue,
understanding and control.
This experimental work performed on disc and blades fixed in dovetail joint
predict the site of fretting fatigue crack initiation. The disc and blade geometry
corresponding to aero-engine compressors is used and finite element analysis is
carried out by using commercially available software ANYSIS 11.0 to simulate
incremental loading condition during normal engine cycle.
1.3: Research objectives
In this research work the experiments are performed with the help of newly
developed, unique, safe and flexible fretting fatigue test rig whereas numerical
analysis is carried out to study the fretting fatigue crack initiation in the disc and
blades assembled in dovetail joint of aero-engine compressor at room
temperature. The numerical analysis is performed by using the commercially
available software ANSYS 11.0 based on the finite element method. In this
research work the fretting fatigue phenomenon is investigated which is the main
cause of failure in different engineering components, sub-assemblies especially in
disc and blades assembly of aero-engine compressors. The main aims and
objectives of this research work are given below;
a. Fabrication of a new experimental test rig.
b. Validation of newly developed experimental test rig.
7
c. Testing of the disc and blades assembly carried out in dovetail joint of an
aero-engine compressor to evaluate fretting fatigue phenomenon occurred.
d. Validation of results obtained through modeling and simulation by using the
software based on the finite element method to investigate the fretting
fatigue phenomenon.
This research work provides a new direction and a sound plate-form for further
research in this field. It provides update knowledge to engineers in the
maintenance, design of mechanical components and various sub-assemblies
passing through various amplitudes in sliding contact surfaces.
1.4: Organization and management of thesis
A block diagram given in Figure 1.3 provides general information of the thesis.
Thesis consists of seven chapters along with references and three appendices.
Figure 1.3: A block diagram providing general information of the thesis
Numerical Analysis
Analysis of Disc & Blades, Dovetail
Joints of Aero Engine
Compressors
Literature Survey
Fretting Fatigue Pheonomenon, Materials & Cyclic Loading.
Designing and development of test Rig Validation of test rig with and without specimen Testing of rig Dimensions of test rig Experiment step by step Result and Discussion
8
1.4.1: Methodology
In order to conduct the research work problem is formulated, studied theoretical
background, designed and developed suitable apparatus to carry out the
experiments, allocated test piece dimensions, simulated loading cycles, used
physical properties of the material for test pieces and results achieved are
analyzed.
1.4.2: Problem formulation
Disc and blade assembly is mounted with the rotor of the motor and attached
with the axial dovetail joint in a typical aero-engine compressor. Disc is divided
into identical sectors each accommodates a blade. Two such sectors will be
utilized for the analysis.
1.4.3: Theoretical study
Theoretical study of the previous research work is carried out in this field and the
gap found in the research work is addressed in the present research work. Detail
of this study is given in Chapter 2 under the title “Literature Review”.
1.4.4: Designing and development of suitable experimental test rig
It is required to design, develop and manufacture suitable experimental test rig
to perform the experiments.
1.4.5: Dimensions of analyzed dovetail joint including disc and blades
The dimensions are to be allocated to the test pieces manufactured from a
structural steel, mild steel, cast iron and bakelite material to carry out the
experiments on the newly developed test rig.
9
1.4.6: Simulated loading cycle
In this study the dovetail joint is to be analyzed by revolving the specimen which
is the take off phase of an aero-engine whereas deceleration simulates the
landing and stopping of the engine.
1.5: Thesis structure
This thesis consists of seven chapters and two complementary sections i.e.
Annexes and references.
Chapter 1: Introduction
This Chapter gives background of dovetail joints of disc and blades of aero-
engine compressors, fretting fatigue phenomenon and its effects on various
structural mechanical components, fretting crack initiation and its effects on
fatigue life, aims and objectives, organization and management of thesis along
with the flow diagram that explains sequences of work.
Chapter 2: Literature Survey
This chapter overviews the literature available on analysis of dovetail joint and
introduces some important aspects of fretting fatigue currently under research by
a number of investigators in the world. The factors playing pivotal role in the
failure process and the failure data are discussed under the fretting fatigue
fundamentals. The dovetail analyses, applications, different phases, experimental
setup, numerical approaches used predication of fretting fatigue performance in
the literature are briefly summarized.
Chapter 3: Mathematical formulation
In this chapter mathematical formulation is discussed in detail including the
numerical approach to fretting fatigue, prediction of fretting fatigue performance,
special parameter for pure fretting, critical plane approaches and fracture
mechanics based approaches.
10
Chapter 4: Experimental system and data
In this chapter designing and manufacturing of experimental test rig is discussed
in detailed including the verification of experiments test rig with and without
specimen, detail of various tests performed on newly developed experimental
test rig with and without specimen and detail of experiments carried out.
Chapter 5: Numerical analysis
In this chapter numerical analysis is carried out with the help of commercially
available software ANSYS 11.0 and its results are discussed in detail.
Chapter 6: Results and discussion
This chapter consists of the results and discussion obtained from the
experiments carried out on the test rig and the numerical analysis. Fretting
fatigue phenomenon in the disc and blades assembled in dovetail joints of an
aero-engine compressor is also presented.
Chapter 7: Conclusions and future recommendations
In this chapter the conclusions and future recommendations are given showing
the actual findings from the research work performed experimentally and
computed numerically along with the future scope of the research work.
11
Chapter 2
Literature survey
In this chapter, an extended state of the art in fretting fatigue is presented.
Analysis of dovetail joints, its applications and performance effecting variables,
different phases and numerical efforts carried out, Predication of fretting fatigue
performance and different types of fretting fatigue test rigs developed by various
researchers are also part of this chapter.
12
2.1: Introduction
Certification of an aero-engine disc is strongly dependent on being able to
predict the modes of failure. The contact interface between the disc and its
blades is the area of major concern. This interface often takes the form of
dovetail joint. The researchers [51~53] found fretting fatigue phenomenon a
serious issue in these joints. Naboulsi and Calcaterra [54] gave a classical
example of damaged dovetail roots used to fix the blades on the disc of aero-
engine compressor as shown in Figure 2.1. Certainly this damage might be due
to any other reason but the fretting fatigue factor as a major cause of failure
could not be ruled out.
Figure 2.1 Damaged disc and blades due to fretting fatigue failure
phenomenon [54]
The researchers [11,55~63] concluded fretting fatigue as a commonly occurring
phenomenon in mechanical system and structural engineering including different
mechanical and electrical parts, human body, manufacturing equipment,
different components and sub-assemblies used in automobile, air and space craft
subjected to vibration on parts connection. The researchers [64~67] concluded
aerospace as one of the most critical area in which this phenomenon has major
impact with special attention to save human life as well as itself.
13
2.2: Dovetail Joint Analysis
Ruiz and Nowell [68] stated that dovetail joint was investigated experimentally,
analytically and numerically by various researchers. However, loading conditions
applied in these analyses don’t match the actual loading conditions of typical
aero-engine compressor. The efforts were devoted to analyzed dovetail joint.
Durelli et al. [69] carried out the study of disc and blade assembly of turbine and
found that the radial centrifugal force due to loading of the blade and blade
bending because of pressure of gas are the major significant forms of loading.
The researchers [70~71] conducted two- as well as three-dimensional photo
elastic analyses of the disc and blades fir-tree region of a turbine. Centrifugal
loads were applied to the two-dimensional model whereas circumferential,
centrifugal and plane bending loads were applied at the cancroids of the blade
for the three-dimensional mode.
In a series of papers the researchers [72~75] analyzed the dovetail joint
experimentally by employing a biaxial model. Boddington et al. [74] introduced
numerically a fretting damage parameter, found out the crack ignition area and
observed that the maximum friction force is produced to join two plane surfaces
with each other of dovetail and the root radius.
Kenny et al. [76] carried out experimental study using photo-elasticity and finite
element analysis to determine the contact stress distribution. Nurse and Peterson
[77] conducted experiments for stage two fatigue crack growth paths in fir-tree
fixtures. No allowances were provided in their models for contact elements and
an approximate contact pressure between the two matting surfaces of disc and
blades was assumed. The investigators [78~79] found the effect of frictional
force on crack propagation of disc and blades in the dovetail joints and other
geometries. Experimental, numerical and analytical techniques were employed to
study contact problems under normal and tangential loads.
14
Kong et al. [80] investigated dovetail joint of three-dimensional reinforced
carbon-carbon composites material. Dovetail joint of an aero-engine, compressor
disc and blade were studied experimentally as well as numerically by fixing the
same in a rigid box. They concluded that fracture phenomenon could be
controlled by relaxing rounded edges of the disc and blades and composite
materials could be considered suitable for use in aero-engine compressor.
Ciavarella and Demelio [81] studied fretting fatigue and devised different
analytical results relating slip contact problems. They analyzed dovetail joint,
studied different damage parameters and considered the distribution of the
pressure forces and shear tractions acting at the contact area of a pad rounded
at its corners and a semi-infinite half-space as an approximation of a dovetail
joint. For that approximation, a normal force acted on the half-space with equal
and opposite force acting on the pad with a tangential load. The researchers
[82~85] investigated and gave a similar concept by adding moment acting on
the half-space with an equal and opposite moment on the pad.
Papanikos and Meguid [86] carried out experimental and numerical study of
fretting initiated fatigue failures whereas Meguid et al. [87] studied three-
dimensional numerical analysis showing two-dimensional model stress induced in
dovetail joint. Sinclair et al. [88] carried out study by changing geometric
features of dovetail joint by using different coefficients of friction by
incrementally applying constant angular velocity. The researchers [88~89]
performed two- and three-dimensional elastic stress analysis of dovetail joint
using finite element analysis under constant angular velocity. They used sub-
modeling technique to obtain stress distribution accurately near the edge of
contact and applied to both two- and three-dimensional analyses.
The investigators [53, 67] conducted experiments on micro-damage in dovetail
along with the other geometries for Ti-6A1-4V using scanning electron
microscopy. They found cracks to initiate on the both edges of the contact
identified as the sites of cyclic plasticity. They observed multiple cracks to initiate
15
simultaneously and analyzed that most fretting fatigue initiated cracks did not
propagate; only one or two of them were proposed to dominate in propagation.
Conner et al. [53] proposed fracture mechanics based life prediction model.
Nowell [90] conducted a biaxial fatigue experiment on dovetail joint using
dovetail biaxial experimental test rig at Oxford University, which was capable of
getting the effect of blade vibration, expansion and centrifugal loading. This is
the modified form of the test rig used earlier by Ruiz et al. Finite element
analysis with sub-modeling technique was used to calculate stresses and
displacements. This provided surface tractions and sub-surface stress fields using
coefficient of friction of 0.1. Rajasekaran and Nowell [71] developed a semi
analytic, by considering the equivalent, flat and round geometry and employing
half-plane geometry to analyze dovetail joints due to high computational cost of
finite element analysis.
Golden and Calcaterra [91] conducted tests by using TI-6A1-4V specimen
configured with the angles of 35ο, 45ο, 55ο dovetail fixture at Air Force Institute
of Technology. In dovetail joints, a methodology of fracture mechanics was
applied to access the life. Finite element method was used to calculate contact
stresses and compute stress intensity factor for cracks with different sizes and
orientations. Crack nucleation life was predicted using multi-axial stress-life
model and propagation life prediction was made through fracture machines
principals. They concluded that initial choice of crack depth has little effect on life
predictions due to the high initial crack growth rates.
2.3: Fretting fatigue- applications
Fretting fatigue has diverse engineering applications. Presently it is applied in
aircraft industry to cope with many important features subjected to fretting
fatigue of riveted and dovetail joints. Giannakopoulous et al. [92] stated its
applications in cables and ropes, structural joints, bio-implant devices, steam or
gas turbine engine, steam or gas turbine engine disc on the drive shafts in land
16
based turbines, the dovetail aircraft jet engines, the flange joint between the
beveled gear and the drive shaft in gas turbine, helicopter power transmission
system, cable car ropes, and marine hawsers and moorings, rivets in wings
aircraft structures, spline and keyways in shafts and couplings, shrink fitted
components, bio-implant devices, fracture on underside of screw-head and
countersink hole in base plates attached to bone and femoral stem in total hip
replacements where lesser trochometer bone is wired around femoral stem
component. Hoppner [93] discussed some important case studies of fretting
fatigue in engineering components. One could overcome fretting fatigue through
control practice and sufficient deterrence in the design and the mechanism used.
Nowell et al. [94] stated that the ultimate goal of research in fretting fatigue is
the prediction of fretting fatigue life under a given set of contact conditions from
some independently measured material data. The stress gradients under fretting
fatigue in the component is almost much large than in the plain specimen, unless
a very sharp notch is used. The surface damage occurred due to relative motion
of the surface might accelerate crack initiation. Szolwinski [95] stated that some
success was achieved in life prediction for relatively large contacts and less
severe stress gradients. Researchers [96~99] suggested that the life prediction
for smaller contacts depends on solving the stress gradient problem and much
work was concentrated in this area. Hoppner [28] listed the goals of research in
the fretting fatigue as the physical and chemical understanding of the
phenomenon and appropriate models. Successful development of all aviations
and prevention schemes, determination of appropriate maintenance, inspection
and replacement intervals, development of models of the various stages of
fretting fatigue.
2.4: Fretting fatigue- steps of study
Like plain fatigue fretting fatigue can be treated by either conducting tests to
represent service conditions and the total life recorded using S-N approach, or
17
the fatigue life to split up into initiation and propagation phases and fracture
mechanics principles used. Additional variables in fretting fatigue describe the
severity of contact conditions and significantly control the nucleation process.
The fretting fatigue problems are analyzed by solving the partial slip contact
problem, predicting the crack initiation conditions as well as the accelerated
growth occurred in short crack.
2.5: Fretting fatigue- performance affecting variables
Performance under fretting is affected by many factors including microstructure,
the local mechanistic parameters and environmental variables. Puglia and Pratesi
[99] listed thirty-one parameters mentioned in their fretting tests whereas Beard
[100] listed more than fifty parameters of concern in fretting. However most of
the variables affecting fretting fatigue are interdependent and much smaller list
of critical variables is desirable.
The researchers [101~107] focused on relationship between global boundary
conditions, i.e. contact geometry, normal load, and shear load, normal bulk
stress and fretting fatigue life. Iyer [108] identified degradation process in
fretting fatigue and gave certain local mechanistic parameters like contact semi-
width, coefficient of friction, maximum shear stress range, slip amplitude at the
interface, force occurred during meeting and local cyclic tangential shear stress
at the interface. The researchers [109~110] found the factors that might be a
source of significant variation in fretting fatigue activities including normal and
tangential load, applied bulk load, relative displacement amplitude, temperature,
used contact geometry, hardness, frequency, coefficient of friction and
conditions of surface etc. Actual characteristics are the parameters influencing
the fatigue life. However, the exact characteristics of all these influencing
parameters and fatigue life effect due to these individual comparative effects are
not known. By using the testing methods the difference of fretting processes
were not yet standardized.
18
Factors affecting the fretting fatigue life are normal and tangential loads and
relative displacement amplitude but there are many other factors like surface
roughness, contact geometry, materials, environmental conditions’ frequency and
temperature also play vital role. Nakazawa et al. [111] found that by increasing
the normal force fretting fatigue life decreases linearly. In case of austenitic
stainless steel no further degradation was observed above the critical contact
pressure and at a pressure of 15~45 MPa, life was found unchanged where as it
drastically decreases beyond the contact pressure of 60MPa. Ramalho et al. [112]
carried out experiment on EN H3 M steel and observed that there is no
significant effect on fretting fatigue life. Researchers [113~114] concluded
through experiments that normal load made very low effect on the nucleation
conditions under partial slip conditions. Several researchers concluded that effect
on the fretting fatigue life, depends upon the application of load frequency. Lyer
and Mall [115] found that there is an effect of 1 Hz and observed no effect at
200 Hz where as Jin and Mall [116] concluded that there is no effect on fretting
fatigue life at 2 Hz and found that there are other factors affecting more fretting
fatigue life than that normal load. The researchers [117~118] studied the
relationship between the normal load and relative amplitude displacement and
concluded that slip zone size is dependent on the magnitude of the normal load.
The tangential load is produced due to the contact. Friction between pads and
specimens affects the fretting fatigue life. During testing contradictory results
were found in relation to tangential load evaluation. Jin and Mall [5] found that
in same situation tangential load increases or after increasing it suddenly
dropped and reason for sudden drop is decreased in applied normal contact load,
resulting from applied material removal of specimen as well as pads. Jin and Mall
[116] conducted that in the partial and gross slip tangential load is quickly
stabilized and remained constant throughout the test and monotonically
increased and then decreased respectively. Wittkowsky [117] concluded that in
the beginning maximum tangential load decreased after stabilizing when the
19
cycles are increased. Zhou [118] concluded that with the increase in tangential
load slip region increased and stick zone is decreased till full slip occurred within
the contact.
The Researchers [119~120] concluded that relative displacement amplitude is
one of the most important factor in fretting fatigue behavior to control the crack
imitation and propagation process. Huang [121] concluded the significant
difference between relative displacement amplitude on coefficient of friction. In
the results there was a significant difference between alloys AZ91D and AM60B.
In case of AZ91D alloy at low amplitude the increase in the coefficient of friction
with the relative displacement amplitude was much higher than that of alloy
AM60B. Madge et al. [122] found that propagation is less sensitive to relative
displacement amplitude than the initiation time. Pape and Nue [123] carried out
investigation on steel 4340 and concluded that with the increase in relative
displacement amplitude between two contacting bodies there is a slight increase
in the surface roughness. Other factors affecting the fretting fatigue life are
environment, materials, coefficient friction, surface roughness, temperature,
frequency contact geometry etc.
Generally tests are performed in normal atmospheric condition but few tests are
carried out under various conditions. Elliot and Hoeppner [124] carried out tests
on aluminum material 7075-T5 and concluded that the material has 10 to 20
time more life when tests are carried out in vacuums rather than in air.
Temperature affects the fretting fatigue life. Changes in the corrosion and
oxidation along with mechanical properties are due to temperature difference. It
has been observed that fretting fatigue life is reduced at high temperature. Lee
and Mall [125] conducted tests on the specimen at shot penned condition of Ti -
6A1-4v alloy and concluded that with the increase in the temperature the fretting
fatigue life is decreased and surface roughness increased. Jin et al. [126] found
no change in the titanium material at 2600C and no reduction in fretting fatigue
life was observed at room as well as elevated temperatures. The researchers
20
[127~128] concluded that fretting fatigue life reduced with the increase in the
number of cycles.
Various tests were carried out with various materials including combination of
similar as well as dissimilar materials. The investigators [129~130] carried out
tests of similar materials where as Elleuch et al. [131] performed tests using
dissimilar materials for different applications like automobile, disc and blade
assemblies along with various parts of aero engine compressors and jet engine.
The researches [132~133] concluded that titanium and aluminums alloy are the
commonly used materials in the aeronautical applications. They studied the
fretting fatigue behavior between dissimilar materials. Navarro and Domínguez
[133] carried out study on TI 6A -4V and A120, Lee and Mall [134] studied the
materials i.e. TI 6A -4V and 718 whereas Kubiak et al. [135] carried out his study
on materials i.e. 30 Ni Cr Mo and 52100 etc. All these materials under fretting
fatigue conditions have different behavior because of their mechanical and cyclic
properties. In view of this Gaspar et al. [136] investigated the effect of
Aluminum and steel on the fretting behavior and for this purpose specimen was
made of stainless steel i.e. AISI 310 whereas material used for pad was steel to
specification AISI 52100 and Alumina 99.7 % AL2 O3. The steel pad formed a
layer of iron oxide in the sliding area of fretting contacts which had a protective
effect against fretting wear. They observed that wear volume in case of steel
was less as compared to aluminum.
Higher surface roughness in two contacting surfaces affects fretting fatigue
mechanism. The researchers [137~138] concluded that increase in surface finish
affects the fretting fatigue damage. Waterhouse and Trowsdale [138] found that
by increasing the surface finish surface degradation decreases due to incipient
cracks prevented from propagation by compressive strength imposed in the
fretting pads. The researches [137~139] concluded that with the increase in
surface roughness damage in fretting fatigue increases due to increase in co-
efficient of friction and it is not beneficial for fatigue life. Researchers [6,139]
21
suggested that in two contacting bodes the fretting fatigue life alters. Volchok et
al. [140] concluded that when the dynamic co-efficient of friction is higher than
the static co-efficient of friction the stick slip situation occurs.
Researches [141~142] concluded that by increasing the normal load coefficient
of friction decreases. Fouvry et al. [141] concluded that at normal load of 300N
coefficient of friction was 0.97 while at 500N normal load it was decreased to
0.85. The researchers [143~144] observed that in early stage of fretting fatigue
test coefficient of friction increased and after that in couple of hundred of cycles
it reached at a constant value.
Practically, stress concentration is the cause of fatigue crack which could be
avoided by proper designing the content geometry. Araújo and Nowell [145]
showed that there is a great influence of contact geometry on fretting fatigue life
and identified critical contact width at which drastic changes in the fretting
fatigue life was found. Navarro [146] carried out tests by using different types of
geometries and observed that in case of spherical contact the stress level was at
high level than that of cylindrical contact geometry. He concluded that initiation
phase is more important in cylindrical instead of spherical contact.
Load frequency interacts with different parameters i.e. temperature and normal
load and affects on fretting fatigue phenomenon. Bryggman and Soderberg
[147] found that by increasing the load frequency loads to a high temperature
rises in a narrow fretting zone causing more damage.
2.6: Fretting fatigue - phases
Fretting fatigue process can be divided into the three distinct phases of crack
initiation, short and long crack propagation. The investigators [148~149] stated
that the long crack propagation life in fretting fatigue is practically a relatively
small proportion of the total component life whereas the short crack phase
received relatively little attention unit recently. Nowell and Dini [150] found that
22
quantitative accurate models of these phenomena were not yet fully developed
and the difficulty was reported to employ plain fatigue data to predict fretting
fatigue performance. The researchers [48,151] concluded experimentally that
initiation and propagation of cracks responded to different factors. Crack
initiation was found dependent mainly on the contact stresses and occurred in
the vicinity of the high localized stress concentrations caused by frictional forces
between the contact surfaces.
Fatigue crack initiation (FCI) is a consequence of local events. Researchers [52,
151] stated the experimental observations and indicated that local cyclic
plasticity resulted in the nucleation of fatigue cracks. Irreversible flow at the
surface and within depths of only several grains leads to intrusions, extrusions
and finally to cracks. Proudhona et al. [46] stated that the fatigue crack initiation
is a process controlled by local cyclic plastic shear deformation. Researchers
[46,151] concluded that nearly simultaneous multiple contact surface crack
initiation at shallow and Dubourg and Lampacq [152] concluded that steep
entrance angles to the surface is common in fretting fatigue tests throughout
different contact geometries and fixtures. The researchers [45,47,153~154]
found that this could be a result of (a) the very sharp gradient of the multi axial
and possibly non-proportional stress field imposed at the critical contact area and
(b) the difference in size and orientation of the surface grains. Terheci [151]
concluded that early life of the initiated cracks is consumed in stage I mode II
growth. The researchers [45,47,153~154] found that retardation in growth of
some of those cracks is most likely due to local stresses just below the contact
region in terms of (a) sharply decreasing stress fields along the contact surface
and below contact area and (b) tri-axial compressive stresses associated with
friction between crack flanks. Ding et al. [155] concluded that material removal
due to fretting wear can result in residual compressive stresses within the fretted
material and thus, lead to redistribution of contact stresses, stress decrease and
consequently, retardation in growth of initiated cracks. Conner et al. [52] found
23
that Various initiated cracks could neither develop nor reach the opening stage
II model I fatigue crack growth which is mainly controlled by tensile stress
range. Neighboring cracks might merge forming a surface notch. The researchers
[53,151,156] found experimentally that existence of relatively large broken
material particles filling the mouth of that notch nearly with common boundaries.
In the worst case, one or two cracks are able to be dominant and propagate to
cause catastrophic failure.
The investigators [5,45~48,51] investigated fretting FCI experimentally,
numerically and analytically. The researchers [47~48] developed experimental
methodologies to define crack nucleation boundary under different laboratory
conditions.
The researchers [157~158] defined crack nucleation in terms of observed cracks
longer than some small detectable size ranging between 0.1 and 0.5 mm
dependent on the capabilities of available experimental setup. The investigators
[159~162] stated that elastic stress analyses of different contact geometries,
materials and loading conditions are mostly associated with a multi axial
formulation. Smith et al. [163] found that multi-axial was applied to analyze
experimental phenomena. Jin and Mall [164] predicted the life, researchers [21,
22, 24] the location and Proudhona et al. [43] the entrance angle of the initiated
crack. The researchers [45,116,153] concluded the necessity to consider a crack
size effect induced by the very sharp stress gradient imposed below the surface.
Fouvry et al. [153] stated that average stress analysis over a fitting critical
material volume was associated with the SWT parameter to predict successfully
the fretting fatigue crack imitation sites. Proudhona et al. [43] found that SWT
parameter failed to predict the entrance angles of experimentally initiated cracks.
However, argued that the SWT parameter, being mainly dependent on the
tensile stress-strain state, was unable to describe the shear mechanisms which
they found controlling the crack nucleation process. They concluded that the
24
physical meaning of the SWT parameter to predict fretting fatigue crack initiation
appeared disputable for their experiment.
2.7: Fretting fatigue - numerical approaches
Numerical efforts found in the literature addressing fretting fatigue are mostly
within the context of specific experimental setup designed for some real
applications. Finite element analysis are usually carried out to solve contact
stresses and displacements or find stress intensity factor of cracks.
Mcveigh and Farris [165] evaluated the influence of the bulk stress on the
contact pressure for fretting fatigue contact problems. Namjoshi et al. [166~167]
used four node quadrilateral plane strain element to estimate fretting fatigue life.
Lee et al. [130] incorporated elastic-plastic isotropic hardening model with a von
Mises yield criterion in order to obtain the evolution of stress and strain during a
fretting fatigue test, Szolwinski and Farris [168] suggested an accurate
characterization of the near surface contact stress field with a multi-axial fatigue
life model to predict fretting fatigue behavior. Cormier et al. [169] developed
aggressive sub-modeling technique to predict accurate stress concentrations
near the edge of contact region. Szolwinski et al. [170] addressed the effects of
finite width fretting fatigue tests using three-dimensional finite element model
and infrared imaging system.
Tur et al [171] showed the influence of finite dimensions of the specimen in
contact with a spherical pad and subjected to fretting using a three-dimensional
finite element model and an h-adaptive mesh refinement process. Hartle et al.
[172] developed a hybrid method incorporating a coarse three-dimensional finite
element model with two-dimensional singular integral equation based approach,
which could reduce computational time with sufficient solution accuracy of stress
concentration at the edge of contact. Kim and Mall [173] conducted three-
dimensional finite element analysis to investigate the effect of finite contact
25
width on fretting fatigue and a critical plane based fatigue model was used to
characterize the fretting fatigue crack initiation behavior.
Rajasekaran and Nowell [71] used finite element analysis in solving the contact
conditions for fretting fatigue tests and employed two-dimensional plain strain
analysis. Golden and Calcaterra [91] also used finite element analysis to compute
stress intensity factors for crack propagation phase.
2.8: Predication of fretting fatigue performance
Fretting fatigue life is generally considered in crack nucleation/initiation and
crack propagation. It is the time period in which crack is developed to damage
the fretting contact. The time period from starting crack initiation to the
propagation depends upon different theories i.e. crack closure, capability of crack
evaluation technique, growth orientation etc. Some researchers have the opinion
that crack size is of 0.4 to 0.5 mm or even 1 mm from crack initiation, while
others believe that crack initiation corresponds to 10 µm. Lykins et al. [174]
stated that crack nucleation corresponds to 80~90% of the total life whereas
investigators [161,175] stated that 90% is being spent on crack initiation. The
researchers [176~177] considered that crack propagation corresponds to almost
the whole fatigue life and it is worth mentioning here that propagation life
started with a crack length of 10 µm. Numerous attempts have been made to
predict the performance of components under fretting fatigue loadings. These
attempts predict the site to produce crack initiation site, direction and life
prediction under given set of condition. The parameters defined by various
researchers for fretting fatigue performance can be mainly divided in special
parameter for pure fretting; critical plane approaches and fracture mechanics
based approaches. Some of this parameter are used for crack initiation whereas
some of the parameters are being used for both crack initiation and life
prediction.
26
2.9: Fretting fatigue - experimental test rigs
Different experimental setups were used to study fretting fatigue phenomenon
including with simple geometries, cylindrical fretting pads, bridge type fretting
pads and special geometries. Fenner and Field [178] developed first type of
setup by using bridge type fretting pads and remained popular until the early
1990’s. The pad consisted of two flat surfaces that contacted two distinct areas
on a single side of a fretting specimen. Nishioka et al. [29] developed second
type of arrangement i.e., using cylindrical. They used a quite different contact
configuration of cylindrical pads clamped against a flat specimen with number of
advantages. This geometry was also adopted by various researchers including
[179~180] and many others.
Experimental setups are produced for special geometries or actual geometries
like dovetail joints along with the setups used by the researchers [53, 72 and
142]. Ruiz’s apparatus is still in use at Oxford. In this implementation, blade
loads, representing centrifugal force in the engine are applied to two opposing
blade specimens and are fixed on a central disk specimen, which is also subject
to load, simulating disk expansion under centrifugal loading. This allows accurate
representation of relative slip in the engine, which may be important for wear of
coating. High cycle fatigue loads representing blade vibration are applied by
mechanical shaker units clamped to each blade.
27
Chapter 3
MATHEMATICAL FORMULATION
In this chapter numerical approach to fretting fatigue is presented. The prediction of fretting fatigue performance mainly aimed to predict the crack imitation. Categories of fretting fatigue performance i.e Special parameters for pure fretting, Critical plane approaches and Fracture mechanics based approaches are also a part of this chapter.
28
3.1: Numerical Approaches to Fretting Fatigue
Numerical efforts found in the literature addressing fretting fatigue are mostly
within the context of specific experimental setup designed in order to simply
some real applications. Finite element analysis are usually carried out to solve for
contact stresses and displacements or to find stress intensity factor of cracks.
McVeigh and Farris [165] evaluated the influence of the bulk stress on the
contact pressure for fretting fatigue contact problems. Namjoshi et al [166~167]
used four node quadrilateral plane strain element to estimate fretting fatigue life.
Tsai and Mall [181] incorporated elastic-plastic isotropic hardening model with a
von Mises yield criterion in order to obtain the evolution of stress and strain
during a fretting fatigue test. Szlowinski and Frarris [168] suggested an accurate
characterization of the near surface contact stress field with a multiaxial fatigue
life model to predict fretting fatigue behavior. Cormier et al. [169] developed
aggressive submodelling technique to predict accurate stress concentrations near
the edge of contact region. Szolwinski and Farris [170] addressed these and
effects of finite width fretting fatigue tests using three dimensional finite element
model and infrared imaging system.
Tur et al. [171] showed the influence of finite dimensions of the specimen in
contact with a spherical pad and subjected to fretting using a three dimensional
finite element del and an h-adaptive mesh refinement process. Hartle et al. [172]
developed a hybrid method incorporating a coarse three dimensional finite
element model with two dimensional singular integral equation based approach,
which can reduce computational time with sufficient solution accuracy of stress
concentration at the edge of contact. Kim and Mall [173] conducted three
dimensional finite element analyses to investigate the effect of finite contact
width on fretting fatigue and a critical plane based fatigue model was used to
characterize the fretting fatigue crack initiation behavior.
29
Rajasekaran and Nowell [71] used Finite element analysis like no of other
investigators in solving contact conditions for fretting fatigue tests. Most of the
investigators employed two dimensional plain strain analyses. Golden and
Calcaterra [91] applied Finite element analysis applied to compute stress
intensity factors for crack propagation phase.
3.2: Prediction of Fretting Fatigue Performance
Numerous attempts have been made to predict the performance of components
under fretting fatigue loadings. These attempts are mainly aimed to predict the
crack initiation site and direction and life prediction under certain set of
conditions. The parameters defined by numerous researchers for fretting fatigue
performance can be mainly divided into following categories.
Special parameters for pure fretting
Critical plane approaches
Fracture mechanics based approaches
Some of these parameters are used to for crack initiation while some of the
parameters are used for both crack initiation and life prediction.
3.2.1: Special Parameters for Pure Fretting
Early attempts at the prediction of fretting fatigue performance frequently
employed special empirical parameters’ formulated purely for the fretting case.
These were perhaps though necessary because of the features of fretting fatigue
which caused difficulties in applying standard fatigue parameters. Two poplar
fretting parameters were proposed by Ruiz et al. [72~74] which are given as
under.
30
Ruiz [72~74] has proposed two type of fretting parameters which are very
popular and there is Ruiz Damage parameter and Ruiz expensed. Damage
parameter, 1st one parameter is energy based and can be calculated by
multiplying maximum shear traction with local slip amplitude ( . . τ and δ is
local value of shear stress within in the contact region and relative displacement
amplitude respectively. This parameter has successful predicted the crack
location in dovetail joints 2nd type of parameter in which it is applicable for other
shapes including dovetail and in this regard Ruiz and Chen has proposed a 2nd
enhanced parameter. It is the multiplication of 1st parameter ie .τ δ multiplied by
σ i.e. ( . . which is maximum local stress components paralleled to the
contact surfaces. The parameter has given a better location of fretting crack
nucleation on application of sphere on flat shape.
3.2.2: Critical Plane Approaches
Critical planes were developed form a physical interpretation of the fatigue
process whereby cracks were observed to initiate and grow on certain
preferential material planes. In such an approach, stress and strains during the
loading cycle are determined for various planes at the same spatial position in
the component. And empirical combination of these is used to predict the most
severely loaded plane or “critical plane” where cracks are expected to nucleate.
Besides the location of crack initiation, these empirical parameters also provide
the direction of early crack growth of the crack, and a measure of the multi axial
fatigue damage that can be correlated with simple uniaxial fatigue data to
estimate initiation life. Researchers [159~160] concluded that there are at least
two distinct modes of crack growth, depending on strain amplitude, material type
and state of stress, material from either shear cracks of tensile cracks.
Following are some of the commonly used critical plane parameters:
a. Modified Smith-Watson-Topper (SWT) Parameters.
b. Fatemi and Socie Parameter
31
c. Shear Stress Rang (SSR) Parameter
d. Effective Shear Stress Parameter
e. Modified shear Stress Range (MSSR) parameter
f. Findley parameter
a) Modified Smith-Watson-Topper (SWT) Parameter
Szolwinski and Farris [182] modified that Smith-Watson-Topper parameter for
application to fretting fatigue crack initiation.
The SWT parameter in this case of multi axial loading on the basis of aε which is
total fatigue strain amplitude and maxσ maximum tensile stress normal to the
crack plane during loading cycle and it can be written in mathematical form:
S . (3.1) The modified parameter is the product of the normal strain amplitude and the
maximum normal stress. For this parameter, the critical plane is defined as the
plane in which the modified SWT parameter is a maximum. At each location, all
possible planes must be examined in order to find the critical one. Therefore the
critical plane approach, using this parameter, gives both the location and
orientation angle of fretting fatigue crack initiation.
It is possible to estimate the component life by reference to a fully reversed
uniaxial test. For such a test, the stress-life and strain-life curves can often be
modeled satisfactorily using the Basquin and Coffin-Manson laws as given life
∆ / 2 b (3.2)
and total fatigue strain amplitude is total sum of elastic and plastic strain.
Equation can be written if we use the experimental relationship to define the
total no of cycles to fatigue failure of metallic material which is given as under:
32
∆ ∆/
2 b+ / 2 c
(3.3)
In the above equation , ∆ , ∆ , /, , , is the total fatigue strain
amplitude, elastic strain amplitude, plastic strain amplitude, fatigue strength
coefficient, young’s modulus, no of cycles to crack limitation, fatigue strength
exponent, fatigue ductility coefficient and fatigue ductility exponent respectively.
These above two equations i.e. 3.2 and 3.3 can be used in conjunction with the
equation with the equation for SWT value to correlate the SWT fatigue
parameter with life, giving:
./
2 + / / 2
(3.4)
b) Fatemi and Socie (FS) Parameter
For cracks that grow in planes of high shear strain, fatemi and socie proposed
following fatigue parameter.
∆ 1 (3.5)
Where ∆ is the difference between maximum and minimum values of shear
strain experienced during the cycle, smax is the maximum value of the stress
normal to the chosen plane, is the yield strength, and is a constant which
approaches unity at long lives and is reduced a shorter lives. The critical plane is
that having the critical location and plane.
Similarly, an empirical equation fitted to the results of simply reversed tests (this
time under pure shear) can be used to correlate the FS fatigue parameter with
life, giving:
33
∆ 1/
2 + / 2 (3.6)
Where G is the shear modulus and / , /, is constants. It is worth
pointing out that some arbitrary definition of the size of and initiated crack, must
be made in order to estimate initiation life with the help of SWT of FS criterion.
To some extent, the chosen size will depend on the detection methods used in
the calibration experience methods used in the calibration experiments which are
employed to find constants in the equations.
c) Shear Stress Range (SSR) Parameter
The second parameter is known as the shear stress range critical plane
parameter. The shear stress range.
∆ (3.7)
Here in this case shear stress range is calculated on all the planes at all the
points and the plane with maximum value of shear stress range is identified as
critical plane.
d) Effective Shear Stress Parameter
Walker [183] slightly modified this parameter in order to include the effect of the
mean axial/shear stress on the fretting fatigue.
∆ . 1 (3.8)
Where means the maximum shear stress on the critical plane, refers to
the shear stress ratio on the critical plane and n is a fitting parameter.
34
This accounts for the mean shear stress ratio effect on the critical plane. The
following are two parameters are based on both shear and normal stresses and
the critical plane.
e) Findley Parameter
Findley [184] created this parameter for plain fatigue analysis. It involves both
the shear stress amplitude and the maximum stress normal to the orientation for
the maximum shear plane multiplied by an influence factor, k, such that:
(3.9)
The critical plane was such that crack initiation was assumed to occur on the
plane with the maximum findley parameter value. It was shown by Namjoshi
[167] that this parameter could not discern between plain and fretting fatigue
when determining fatigue life, which is obviously in error. So it is probably not
the best choice of a predictive fretting fatigue parameter.
f) Modified Shear Stress Range (MSSR) Parameter
Namjoshi et al. [167] created this parameter and is considered by some to be
the premier fretting fatigue predictive. It is a modified version of the shear stress
range critical plane parameters, MSSR, which combined the better features of
the other critical plane parameters. It is thought that this parameter is the best
for determining the effects of fretting fatigue for several reasons. It is based on
both normal and shear stresses, so therefore it eliminates the effect of pad
geometry. Also it includes aspects of the shear stress range parameter, which
was the only parameter mentioned thus far shown by Namjoshi to be satisfactory
for determination of both crack location and orientation.
35
∆ (3.10)
In the above equation A, B, C and D are constants obtained.
3.3: Fracture mechanics based approaches
Facture mechanics approaches used for fretting fatigue performance prediction
can be divided mainly into following three categories;
Crack analogy and asymptotic curves
Notch analogy
Short crack arrest method
a) Crack Analogy and Asymptotic Curves
Crack analogy was first suggested by Giannakopoulos et al. [185] in 1998. This
approach utilizes the similarity between contact mechanics and fracture
mechanics to investigate fretting fatigue life. They showed that there exists there
are some parallels between the stress field close to the edge of a flat rigid punch
and that at the tip of an elastic crack.
Ciavarella [186] has extended the crack analogy approach in formulating the
crack like notch analoguel. It improves and combines features of previous crack
and notch analogue models developed at MIT. This model considers only to
possible behaviors: either ‘crack-like’ or ‘large blunt notch’. In a general fretting
fatigue situation, the former condition is treated with a single contact problem
corresponding to Crack Analogue model; the latter, with a simple peak stress
condition (as in previous Notch Analogy Models), simply stating that below the
fatigue limit, infinite life is predicted for any size of contact. In the typical
situation of conditions can be readily stated. This it can be stated that the sizes
36
effect means that a fretting contact can effectively be classified into three distinct
regions, in a similar manner to that proposed for notches by Atzori and Lazzarin
[187] (i) cases where the stress concentration affects such a small area that it
may be ignored, (ii) cases of intermediate size where the concentration behaves
in some respects like a notch, and (iii) cases where a large volume of material is
affected and a simple stress concentration, or KT approach may be employed.
Naboulsi [188] modified and is an extension of crack analogy to extend CA
capabilities and improve prediction of crack initiation. This includes various
indenter substrate geometries as well as modifying its crack initiation parameter
to include the effect oif build stress in the substrate. AK parameter-life curves
similar to the stress life S-N are established which show similar trends to plain
fatigue with lower damage tolerance as expected. This model shows potentials in
life prediction such that it can be used as a tool in the design of components
under fretting fatigue.
The asymptotic approaches offer a means of correlating one contact with
another, under different geometric and loading conditions. Thus, it should be
possible to characterize a contact in a component and to carry out fretting
fatigue experiments under an identical stress state. This offers an extremely
useful means of reducing the complexity of the experiments required and
experimentally characterizing initiation lives.
b) Notch Analogy
The original results of Giannakopoulos et al. [185] are rather restrictive in terms
of geometry and the application to practical fretting fatigue situations is
therefore not entirely straightforward.
37
However the original idea has been extended to the case of a flat punch with
radiuses corners (the so-called flat and rounded contact) [189]. In this case, an
analytical model for fretting fatigue at a rounded corner punch contacting a
substrate and an analogy is made with fatigue crack initiation at a notch tip. This
methodology provided a direct connection between the round cornered flat
punch fretting fatigue and the plain fatigue crack initiation of a smooth specimen
of the same material. The result is analogous to that of Barsom and McNicol
where the notch fatigue endurance stress was correlated with stress intensity
factor and the square root of the notch-tip radius.
In fact, stress under a typical frictional contact is highly multiaxial and undergoes
non-proportional loading. This is true for the care of a general point but it is
frequently the case that the most highly loaded point is at or near the edge of
the contact. Johnson [190] concluded that If the most highly stressed is at the
contact edge and the contact is incomplete the normal and shear tractions will
have fallen to zero at this point and hence the only non-zero stress component
at the surface will be that parallel to the surface. Hence the stress state tat the
point of initiation is likely to be uniaxial (or very close to it). This means that the
complication of using multiaxial parameters might to draw an analogy between
the stress state at the contact edge and that in a suitably shaped notch. This is a
rather loose analogy, since it is restricted to matching stresses along a line in
each geometry ( the notch bisector and the surface normal at the edge of
contact). However good results can be achieved by varying the notch size root
possible to apply traditional notch fatigue life prediction approaches, such as the
point, line and area method to the interpretation of fretting fatigue. The
approach certainly has its uses, notably in reading across from experiments at
different contact conditions in the same experimental series, but am degree of
expiricism is required in choosing the critical distance. The high stress gradients
present, when compared to notches introduced as design features. Means that it
is unlikely that standard values can be used.
38
c) Short Crack Arrest Method
For a crack to initiate and grow, the stress field needs to be high and sustained
over a reasonable distance. This observation can be formalized by appealing to
the concepts of short crack arrest. Nowell and Araujo [191] first time suggested
general approach and independently by Chan et al. [192~193].
The method is based on the Kitagawa-Takahshi diagram [194] expressing the
crack propagation threshold as a function of crack length for small cracks. The
short crack approach is in principal more attractive since it relies only on
standard material parameters obtained from plain fatigue experiments. Nowell
and Araujo [191] successfully used this approach explain size effect observed
experimentally.
39
Chapter 4
Experimental System and Data
This chapter gives the detail of experimental test rigs already used by the
researcher. The particular considerations is given in this chapter to the newly
developed fretting fatigue test rig including different type of mechanical and
electrical / electronic parts used, its working and validation, detail of
experimental work, techniques and procedures. The material used and geometry
of the specimen is also a part of this chapter.
40
4.1: Introduction
Experimental analysis is performed to investigate fretting fatigue phenomenon in
the disc by revolving the specimen in sinusoidal wave. Experimental test rig, its
description, electrical circuit diagram, detail of different components used,
working of test rig, experimental program and numerical models, detail of
experimental test rigs for freeing fatigue tests, working of experimental test rig,
validation of newly developed test rig with and without specimen mentioned in
detail.
4.2: Experimental fretting fatigue test rigs
Fretting fatigue of structural components damages the whole assemblies and
such damage is very complex as it is involves surface and subsurface multi axial
non proportional elastic plastic cyclic stress-strain field. Various researchers
investigated dovetail joint experimentally, numerically and analytically. Ruiz and
Nowell [195] concluded that in these analyses loading conditions applied do not
match the actual loading conditions of typical aero engine compressor.
The researchers [49,196~198] concluded that literature is available on fretting
fatigue test rig having various forms, starting from simplest to more complex
system. In most of the systems servo hydraulic testing machines are being used.
Different experimental test rigs are in use to carry out study of fretting fatigue
phenomenon in disc and blades assembled in dovetail joint of aero engine
compressor including experimental test rigs with simple geometries for which
analytic solution for contact stress distribution like bridge, cylindrical and
structurally dependent geometries.
4.2.1: Bridge type fretting pads
Bridge type fretting pads was developed by Fenner and Field [199] and these
pads consisted of two flat surfaces that contact two dissimilar areas on a single
side of a fretting fatigue sample. The test rig remained under use with popularity
41
up to early 1990’s. The chief virtue is simplicity in that a normal fatigue specimen
may be used, either in bending or cyclic tension. In this test rig bridges are only
fixed with the sides of the sample which causes relative motion between the two
bridges of feet and the sample have number of difficulties with this simple
arrangement where as contact conditions are also difficult characterize,
particularly if there is bending in the bridge itself. Further conditions at each foot
will not be absolutely identical and it is likely that one foot will slip before the
other, even under normally symmetric conditions. This means that the slip
regime during this experiment is often unknown.
4.2.2: Cylindrical type fretting pads
Nishioka and Hirakawa [200] introduced this type of arrangement they used a
quite different contact configuration of cylindrical pads clamped against a flat
specimen. In this geometry pad alignment is less critical and stresses are
predicted by classical contact mechanics. The researchers [201~203] also used
this geometry normally fixed. In this type of test, the normal load is, whereas the
tangential load is cycled and applied using springs or a separate actuator.
4.2.3: Related to Special geometry
Some special test rigs have been developed which are called special geometry
experimental test rigs. These equipments are in such design that actual
geometry can be utilized. These test rigs have been used by the researchers
[204~205]. The apparatus developed by Ruiz’s is still in use at Oxford. In this
system, blade loads, representing centrifugal force in the engine are applied in
two opposing blade specimens and are mounted in a central disk specimen,
which is also subject to load, simulating disk expansion under centrifugal loading.
Buciumeanu et al. [206] developed fretting fatigue apparatus as exploded view is
shown in Figure 4.1, whereas specimen geometry is shown in Figure 4.2. This
42
apparatus was utilized with servo hydraulic testing machine as shown in
Figure 4.3
Figure 4.1: Exploded view of fretting fatigue test rig [206]
Figure 4.2: The specimen geometry [206]
F
P
m
4
re
F
Figure 4.3
Patrick and
materials as
.5. Basical
esearchers
Figure 4.4:
: The frettin
Golden [20
s shown in
ly it is a
[208~209]
: Drawing o
ng fatigue a
07] develo
Figure 4.4
improvem
].
of the eleva
43
apparatus a
machine [5
ped fretting
where as s
ent in the
ated temper
[207]
.
assembled
53]
g fatigue t
specimen g
e already d
rature dove
in the servo
test rig for
eometry is
developed
etail fretting
o hydraulic
turbine en
given in Fi
fixture by
g fatigue fix
test
ngine
igure
y the
xture
H
a
F
p
si
te
re
a
co
te
o
m
a
d
d
e
ri
Figure 4
Hills and Ne
ppended be
irst type of
roblem”. T
imulates fre
est rig is
esistance t
ccelerated
onditions. I
est rig is us
f tests un
mechanism
bove ment
ifferent fro
ifferent ty
stablished.
igs are avai
.5: Drawing
ewell [180]
elow:
f fretting fa
his type of
etting conta
used for
to fretting
test may
In this test
sed for mak
nder well
of fretting
tioned rigs
om those u
pes of tes
In the tec
ilable rangi
g of the spe
frett
gave thre
atigue testin
f testing rig
act conditio
making ”m
general wa
be carried
t rig tests
king “idealiz
defined s
fatigue pr
its results
sed in the
st rigs are
chnical lite
ng from ve
44
ecimen use
ting fatigue
ee general
ng rig is “s
g is more u
on itself qu
material ra
ay. Geom
out and li
may be ca
zed fretting
simple con
rocess. It is
are applied
tests them
e used. Ye
rature diffe
ery simple d
ed in elevate
e fixture
types of fr
imulation o
seful than
uite precisel
nking tests
metry chose
ttle attenti
rried out v
fatigue tes
ntact geom
s quite diff
d to a wide
mselves. For
et no stan
erent types
device to m
ed tempera
retting fatig
of real engi
the other o
ly. Second
s”. It asse
en is very
on is paid
very rapidly
st rig”. It ca
metry to u
ferent from
e range of
r different
ndard met
of fretting
more comple
ature dovet
gue test rig
neering fre
ones as it r
type of fre
esses mate
simple so
to the con
y. Third typ
arries out s
understand
m the other
geometry q
research w
thod has
fatigue te
ex systems.
tail
gs as
etting
really
etting
erials
that
ntact
pe of
series
the
r two
quite
works
been
sting
. The
45
researchers [210~211] concluded that all the devices for fretting fatigue are
using the servo hydraulic testing machines. The geometry of pads can be
spherical, cylindrical or flat for most fretting fatigue tests. Pauw et al. [212]
reviewed and classified fretting fatigue testing rigs. They concluded that till date
no standard and generally acceptable fretting fatigue testing rig has been
designed and developed. They used existing available concept and tried to build
new testing rigs according to their specific requirement. Too many test rigs may
be disadvantageous due to mismatching in their results. They reviewed the
available literature and classified on the basis of property of the testing rigs into
two categories on the basis of the geometry of specimen and these are full scale
and coupon scale testing.
Full scale rigs are used to find out the life time estimation of the component and
these are mostly used industrial research. The test specimen geometry is
related to the real application. These testing rigs are more expensive as
compared to the coupon scale test rigs, for example to predict life time of the
assembly. Technique in aero space industry where several type of connection
technique is uses which may fail due to fretting fatigue phenomenon. These
joints are very sensitive to fretting fatigue. Wagle and Kato [213] concluded that
body of aircraft is made of aluminum sheet with different nut bolts or through
riveted together and these joints are very sensitive to fretting fatigue. The
researchers [214~216] investigated the dovetail joint disc and blades of aero
engine compressor and splined to connection to join the parts of the engine may
fail due to fretting fatigue phenomenon. Bertini et al [217] found that these full
scales testing rigs may also be used in the oil and gas industries to mine and to
transport oil and gas pipelines are used. These pipe are joint through welding or
threading technique and fails due fretting fatigue process. To save the expensive
down time and costly maintenance full scale testing rigs are used. The
researchers [218~220] concluded that full scale testing rigs are also used in the
press fitted connection and are used to fix the shaft with bearing, flywheel and
46
different type of gears. Aggarwal [221] found that full scale rigs also essentially
required in the leaf spring packages which are used in auto mobile industry
which may fail due to fretting fatigue. The modern engine are base on the turbo
machine and flowed energy is converted into mech. Energy trough blades so
energy is transferred by the blades to rotor which are mounted on the rotor
through dovetail joint. The researchers [222~223] concluded that due to wedge
shape of dovetail and inter active forces on the blades; a slip arises between
rotor and blades. There is another example of when full scale testing rig is
essentially required and it is overhead conductor which are used for the
transportation of electricity. These are also used in the hoisting mechanism; ski
lift and suspension bridge the fretting fatigue testing rigs. In this fixture 52m
rope has been used and it subject to dynamic loading by the shaker. The small
wire in the wire rope will slip against each other and will ultimately due to
fretting fatigue phenomenon will fail rope.
Coupon scale test rigs are, in which no similarity with the real geometry of an
application. The type of test rig a common practice for material characterization
and same is used academics to perform research. In the coupon scale test rig
designer has the facility to choose all the dimensions of the specimen’s arbitrary,
manageable specimen for laboratory testing and has the convenience to perform
a parametric study on the dimension of the specimen. In the coupon scale
testing rig has one more benefits that is the visibility of the contact whereas it is
not possible in case of full scale test rig to see the contact area, nor the crack.
The most commonly test rig used by the researchers [111,130 and 224] is shown
in the Figure 4.6. In this test rig two fretting pads which are pressed against
surface of the specimen, a constant normal load is being applied in with the help
of lateral spring. Whereas applied force is to these springs is kept constant and is
possible to apply a cyclic axial stress on the specimen. With this test rig only
axial load can be measured by a load cell which is the part of the servo hydraulic
test rig.
47
One another most commonly used fretting fatigue testing rig is shown in the
Figure 4.7 developed by Wittkowsky et al. [117]. This test rig is also used
conjunction with the help of servo hydraulic test system like others. This test rig
a capability of accurate measurement of the normal force, relative displacement
between two contact surfaces and the tangential contact force. In this test rig
two load cells have been used to measure the normal force and four load cells to
measure tangential force carried out on the fretting pads by the specimen.
Pape and Neu [123] used this bridge type test rig first time and is shown in
Figure 4.8. In this test rig specimen is being used with two physically totally
separate contacts connected in the form of a bridge. Almost invariable the
Figure 4.6: The most commonly used fretting fatigue test rig [111,130 and 224]
Figure 4.7: The other most commonly used fretting fatigue test rig [117]
48
bridges have contacting pads which have flat faces and therefore form complete
contacts.
All the test rigs which have been developed till date are different types with
different results and there is no standard this acceptable testing rig for fretting
fatigue experiments. In view of fill their specific researchers used to adopt an
existing concept or try to build a new test rig which full or generally requirement.
The test rig is unique in nature will be help full for other researcher for future
development.
4.3: Designing and development of experimental test rig
Fretting fatigue test rig must be carefully designed in order to meet the objective
of the test to be carried out. Fretting fatigue constitutes a series of problem in a
wide range of mechanical parts and this phenomenon is a dangerous which
arises in a lot of structural as well as in many machines like aircraft, electrical
equipment, spacecraft, human body and manufacturing equipment etc. The
efforts made by various researchers to predict the crack initiation with the help
of variety of available apparatus lacks in real conditions. This fact has lead to
develop the one reliable test rig with actual simulation. Rigorous efforts have
lead to this system from conceptual design to the development of real system
Figure 4.8: Schematic of the commonly used type of fretting fatigue test rig [123]
49
capable of simulating incipient fretting fatigue failure with following
consideration.
a) Revolving the specimen, disc and blades in a sine wave pattern or mark
which is the actual simulation i.e. acceleration and deceleration of aero
engine compressors.
b) The testing system must be capable to revolve continuously from 0 to
20000 and reverse back from 20000 to 0 rpm.
c) Easy fixing of the specimen, disc and blades on the shaft of main motor.
d) The extreme safety is desirable for revolving the specimen, disc and
blades up to 20000 rpm which is very high speed.
The photo graph, assembly drawing and exploded view of testing rig is given in
Figures 4.9, 4.10 and 4.10a respectively. Different mechanical parts used in the
testing fixture and the system to control the speed of the testing rig is given in
Figures 4.11 and 4.12 respectively.
Figure 4.9: Experimental testing rig showing. a) The assembly fixture of testing Rig. b) Speed control system of the rig
50
Figure 4.10: Assembly fixture of fretting fatigue test rig
1. Main motor, 2. Main motor holding plate, 3. Safety guard, 4. Lower holdingplate, 5. Disc, 6. Blade, 7. Upper holding plate, 8. Front safety plate,9. Tightening knob, 10. Knob holding plate, 11. Gripping piece, 12. RevolvingPin, 13. Pillar, 14. Base plate, 15. Ratchet, 16. Base, 17. Nut, 18 to 22 Std. Screw
51
Figure 4.10 A: The exploded view of fixture of fretting fatigue test Rig
1. Main motor, 2. Main motor holding plate, 3. Safety guard, 4. Lower holding plate, 5. Disc, 6. Blade, 7. Upper holding plate, 8. Front safety plate,9. Tightening knob, 10. Knob holding plate, 11. Gripping piece, 12. Revolving Pin, 13. Pillar, 14. Base plate, 15. Ratchet, 16. Base, 17. Nut, 18 to 22 Std. Screw
52
Figure 4.11: Detail of mechanical parts of fixture fretting fatigue test Riga
1. Main motor, 2. Main motor holding plate, 3. Safety guard, 4. Lower holding plate, 5. Disc, 6. Blades, 7. Upper holding plate, 8. Front safety plate, 9.Tightening knobs, 10. Knob holding plate, 11. Gripping piece, 12. Revolving Pins, 13. Pillar, 14. Base plate, 15. Ratchet, 16. Base
Figure 4.12: Detail of electrical and electronic parts used in the test Rig
1. Voltmeter, 2. Counter, 3. Timer, 4. Ampere meter, 5. Tachometer, 6. D.Cmotor, 7. Limit switches, 8. Gear train, 9. Variable auto transformer
53
4.4: Preparation of component drawing of experimental test rig
Before to start the manufacturing of the newly required experimental test rig to
carry out test/trials for various mechanical structural components apparatus
available in this field have been studied in detail. It is also added that in this field
a lot of work has been carried out by the researchers but no standard / generally
acceptable test rig is available. To design a new test rig to full fill the actual
requirement of fretting fatigue test of an aero engine components, Various
mechanical components and materials. For this purpose detail drawing were
prepared for mechanical, Electronic and electrical parts of the test rig. Following
were also taken into consideration which is the basic requirements.
• Easy fixing of the specimen.
• Safety required revolving the specimen at very high speed i.e. up to 20000 rpm.
• Revolving the specimen in a sinusoidal wave which is the actual simulation i.e. acceleration and deceleration of the aero engine compressors.
4.5: Mechanical parts
Mechanical parts required for this purpose has been manufactured. The
assembly consists of twenty one components which includes all the sub
assemblies and individual parts which have been used in the manufacturing of
newly developed experimental test rig to make final assembly. Mechanical parts
required for this purpose has been designed and developed and detail of the
same is given under heading A-1.1 to A-1.18 which are shown in Figure A-1.1 to
A-1.18 in Appendix A.
4.6: Detail of various electrical parts
Electrical and electronic parts required for this experimental test rig has been
chosen and detail of the same is given under heading A-2.1 to A-2.10 in
Appendix A.
54
4.7: Working of newly developed test rig
The newly designed, developed and manufactured experimental test rig is unique
and capable to revolve in acceleration and deceleration mode to the disc and
blades sub assembly held with dovetail joint. Blades are fixed diametrically
opposite on disc. The specimen is fixed on the shaft of the main motor directly
with the help of upper and lower holding plates. The main motor is fixed on the
motor holding plate which is further attached with the safety guard and stand
with the help of the pin hinges. Due to these pin hinges; specimen, disc and
blades can be swiveled at any desired angle. This unit is further mounted on the
base through ratchet which provides the angular movement to fixing unit at
required direction. The safety cover plate is fixed on the safety guard with the
help of three nuts and bolts. The speed of main motor has been controlled
through D.C motor mounted on the voltage regulator with the help of a timer.
The motor can be revolved from 0 to 20,000 rpm and same is reversed back in
same manner. Speed of main motor can be monitored through tachometer. The
acceleration and deceleration mode is the actual simulation of the experimental
test rig of aero engine compressors. To conduct the experiments, the following
steps were carried out.
Two limit switches have been used to revolve the main motor from 0 to 20,000
rpm directly with the help of timer introduced in the circuit which is adjustable
from 0 second to 120 hrs. This time is called as “segment start up time”.
Another timer has been introduced and its function is to revolve the main motor
for required time before going to the next step. Each step can be adjusted from
0 second to 120 hrs. Due to this timer each step of main motor will remain at the
same speed till such time it is not shifted to the next step. This time is called as
“segment stay time”.
55
Finally the third timer has been used in the system to ease the long attendance
to monitor results. The time is also adjustable from 0 second to 120 hrs. This
time is known as “system stop time”.
The specimen’s disc and blades can be very dangerous as the speed of main
motor starts 0 and goes up to 20,000 rpm; as such safety first approach is
advisable to monitor the operation of the testing system. In this regard the
system has been operated through remote control so that the experimental
testing system can start/stop from a safe distance. The detail of assembly of test
rig is given in Figure 4.9 showing complete testing system, Figure 4.10 gives the
exploded view, Figure 4.11 shows the detail of mechanical parts and Figure 4.12
gives the detail of different electrical parts of the system respectively.
Manufacturing of newly developed experimental test rig is based on the
visualization of an idea, conversion of the same on papers i.e. in the form of
component drawing, manufacturing of component according to the prepared
component drawing, sub assemblies, assemblies & final evaluation/testing of the
apparatus electrically as well as mechanically. Experimental test rig capable of
simulating incipient fretting fatigue failure was designed.
The assembly consists of twenty one components which includes all the
sub-assemblies and individual parts which have been used in the manufacturing
of newly developed experimental test rig to make final assembly.
4.8: Validation of experimental testing system
4.8.1: Running trend of Experimental test rig – without specimen
After the successful completion of the experimental test rig i.e. revolving the test
rig from 0 to 20,000 rpm and reverse back from 20,000 to 0 rpm with checking
the speed of each step which is the actual requirement for which three timers
have been used. To evaluate the performance of experimental test rig with and
without specimen, same was revolved from 0 to 20,000 rpm with the help of
56
three types of timers i.e. to control “start up time”, “stay time” and “apparatus
test rig time” which has already been explained above under heading working of
testing system. Experimental test rig working is ensured and speed at all the
steps is recorded using tachometer. As the speed of the rig is very high and
safety first has been ensured. Various tests were carried out. Running trend of
experimental test rig without specimen is given in Table B-1.1 t0 B-1.6 whereas
trend depicted graphically in Figures B-1.1 to B-1.6 respectively in Appendix B.
4.8.2: Running trend of experimental test rig- with specimen
After the successful validation of i.e. revolving the test rig without specimen, the
testing was carried out with specimen made out of mild steel is utilized.
Manufacturing of disc, blades and assembly was carried out according to the
component drawing which was fixed on the spindle of the main motor just to run
the apparatus to check whether it bears the weight of the specimen or
otherwise. The experimental test rig was started from 0 rpm with the help of
three types of timers as already mentioned above. Experimental test rig
functioned /operated successfully for certain time and speed at all the steps was
checked through. Data collected tabulated in Table B-1.7 to B-1.10 whereas
trend depicted graphically in Figures B-2.1 to B-2.4 respectively in Appendix B.
4.9: Experimental work
4.9.1: Introduction
A lot of literature is available on the fretting fatigue experimental test rig to till
date and no standard /generally accepted experimental test rig is available.
Researchers are using existing concept with certain modifications or build a new
experimental test rig according to their specific requirement. It has been
observed that each experimental test rig has its own results and hardly compare
able with each other, These are state of the art experimental test rig having
more functionalities but still imperfection and requirement of new experimental
57
test rig is still exist to evaluate the testing of disc and blades fixed in a dove tail
joints of an aero engine compressors over the previously used dovetail fretting
fatigue fixture.
4.9.2: Test rig
Detail of newly developed test rig has already been discussed including
designing, manufacturing and validation chapter 4. This testing rig has been
utilized to perform the tests.
4.9.3: Materials
Throughout the study the specimen were manufactured from four different
materials which are structural steel, cast iron, mild steel and Bakelite. The
material was used in the form of plate and is tested at room temperature.
Chemical and mechanical properties are given in Table 4.1.
Table 4.1: Material specification of the specimen
Material Chemical composition Mechanical properties
High strength structural steel
C: 0.20-0.25%, Si: 0.25-0.35%, Mn: 1.25-1.35%, P: 0.02%
S: 0.035%.
UTS=70000 PSI Min Yield point=5000PSI Min
Elongation=18 % Min Hardness=160-200 BHN
Grey cast iron C: 3.5-3.80%, Si: 2.40-2.60%, Mn: 0.5-0.7%, P: 0.20-0.80%
S: 0.08-0.13%.
UTS=22000-26000 Psi Hardness=160-200 BHN
Mild steel C: 0.15-0.25%, Si: 0.5 % Max,
Mn: 0.5% Max P: 0.05% Max S: 0.05% Max.
UTS=40000-50000 Psi Yield point=25000-35000 Psi
Elongation=30-50 % Hardness=35-40 HRB
Bakelite(thermo setting plastic
Composed of laminated structural of canvas, line nor Kraft paper impregnated with
30 % or more of thermo setting phenolic resin.
UTS=12500Psi Compressive
strength=35000 Psi
4
S
a
e
a
m
co
o
ca
a
p
4.9.4: Test
Specimen w
nd blades
xperimenta
bout its ce
mm. Disc an
ontact surf
pposite to
arried out
re moving
lates.
Figure 4
t specimen
was used to
fixed in d
ation are giv
entral axis.
nd blades n
faces cons
each other
in laborato
freely in t
4.13: Assem
1. Blade,
n
o investigat
dovetail joi
ven in Figu
The dista
otch base r
stitute angl
have been
ory and at
the dovetai
mbly of two
2. Disc
58
te the frett
nt. The di
res 4.13 to
nce betwee
radius is 5
le of 20o
n selected to
room temp
il joints be
o sectors of
ting fatigue
mensions
4.15. The
en disc an
mm and 4
with centr
o carry out
perature. It
etween the
f disc and b
e phenome
of specime
dovetail joi
d blade as
mm respec
al axis. Th
testing. Al
t was ensu
upper and
blades in do
non in the
en used in
int is symm
ssembly is
ctively. Both
he two sec
l the tests w
red that bl
d lower ho
ovetail joint
disc
n the
metric
1.35
h the
ctors
were
ades
lding
ts
4
A
p
a
4.9.5: Simu
A typical loa
resent stud
nalyzed du
Figure
ulated loa
ading cycle
dy, the disc
ring the no
Figure
4.15: Draw
ding cycle
consists o
c and blade
ormal loadin
e 4.14: Dr
59
wing and di
e
f accelerati
es sub asse
ng cycle of a
awing and
mensions o
ion, and de
embly carr
a typical ae
dimensions
of the Blade
eceleration
ied out in
ero engine c
s of the disc
e
loading. In
dovetail joi
compressor
c
n the
int is
r.
T
th
m
d
e
T
D
b
p
4
si
d
4
D
a
b
Figure 4
1. B
The first pha
he aero e
maximum sp
irection, th
xerts a forc
The second
During dece
y gradually
hases simu
.16. In the
imulated. H
esired.
4.9.6: Test
Disc and bl
ssumed to
een used fo
Accel
Decele
4.16: Move
Blades 2. Di
ase corresp
ngine is a
peed. Due
e blade ten
ce on the d
phase is t
eleration ph
y decreasin
ulate the ta
present an
However, t
t program
ade of an
be studied
or which de
eration
eration
ement of bla
isc 3. Uppe
ponds to th
assumed to
to increase
nds to move
isc which is
the deceler
hase, the ae
ng its spee
akeoff and
nalysis, one
the simulat
aero engin
d. To study
etail is give
60
ades at acc
er holding p
e starting o
o be stati
e in speed w
e radials ou
s symmetric
ration whic
ero engine
ed. Thus ac
landing of
e acceleratio
tion could
ne compre
y the same
n in Table 4
celeration a
plate 4. Lo
of an aero e
onary and
while the d
utwards sho
c on both c
ch simulate
is at rest a
cceleration
f an aero
on and dece
be repeat
ssor assem
the differe
4.1.
and decelera
ower holdin
engine. Du
d the engi
isc is rotati
own in Figu
contact surf
s the stop
and the en
and decel
plane as s
eleration cy
ted for mo
mbled in do
ent type of
Ac
De
ation mode
ng plate
ring this ph
ne reaches
ing in clock
ure 4.16. Th
faces.
ping of eng
gine is sto
leration loa
hown in Fi
ycles have
ore cycles
ove tail joi
materials
cceleration d
celeration
es
hase,
s its
kwise
hus it
gine.
pped
ading
igure
been
if so
nt is
have
61
All out efforts were taken before to run experimental test rig and the specimen
was manufactured on wire cut machine from high strength structural steel
material. The specimen was placed between the upper and lower holding plate
as shown in Figure 4.16 and same was mounted on the main spindle of the
motor. After placing on the main motor of shaft and nut was tighten firmly with
the surety that blades are moving freely in the dove tail of the disc. Due to this
movement blades will move out and inward with centrifugal and centripetal force
receptively. The test rig was started to revolve from 0 rpm with small
increments up to 20,000 rpm and reveres back in the same manner from 20,000
rpm to 0 rpm. The specimen remained revolve in sinusoidal wave. The test rig
was stopped after 100 cycles and the specimen was removed from the spindle
carefully. The specimen was examined critically and found black powder on the
sliding surface of the disc and blades known as debris which is one of the causes
to start crack initiation sliding portion of the disc and blade is shown in the
Figures 4.17a and b respectively. This phenomenon has already been
investigated by many researchers like Dicka et al. [225] as shown in Figure 4.18.
Figure 4.17: a) Contact surface of the disc and b) blade after a 100 cycle fretting test
stopped by reducing the displacement amplitude to zero
62
Figure 4.18: Sliding surface disc with oxides/debris [225]
To carry out further investigation of the dovetail joint the thickness of the
specimen was further reduced without changing the other dimensions the disc
with the same material. The specimen was again mounted on the main spindle of
the motor and was starting revolving from 0 rpm and when it reached at 12,000
rpm disc was broken. The specimen was critically examined and observed that
breakage was from the fillet portion of the dove tail. The broken specimen is
shown in Figure 4.19.
A new specimen was prepared on CNC milling machine with same dimensions
and material used for disc is Bakelite to see the behavior and location of
breakage. Specimen was mounted on the spindle of main motor with the help of
upper and lower holding plate. Before to start revolving the specimen it was
ensured that blades are moving freely in dovetail joints. The specimen was start
Figure 4.19: Broken piece of disc made of high strength structural steel material
63
revolving and completed 50 cycles. Specimen was removed and examined
critically and no crack or damage was observed on the surface of the disc and
blades.
Now blades were replaced with high strength structural steel and start
experimental test rig to revolve from 0 rpm and when it reached at 4000 rpm
disc was broken. The broken disc critically examined and found that breaking
area of disc is again from the fillet portion which is shown in Figure 4.20.
Figure 4.20: Broken disc made of Bakelite material
A new sample was prepared on wire cut machine from mild steel and the same
was again mounted on the main spindle of the test rig and started to revolve
from 0 rpm and when speed of specimen reached at 14,000 rpm specimen i.e.
disc was broken and test rig remained safe. The specimen was removed from
the spindle was critically examined and observed that broken portion is again
from the fillet portion and is shown in Figure 4.21.
Figure 4.21: Broken disc made of Mild steel material
64
One more specimen was prepared to further ensure the weakest area of the
fillet. Now material used was cost iron steel. The specimen of disc was
manufactured on wire cutting machine. The entire dimensions were maintained
accordingly. Assembly of disc and blades was carried out and same was mounted
on spindle of main motor with the help of upper and lower holding plate. It was
ensured that blades are moving freely in dove tail of disc between the other
holding plates. The testing rig started to revolve from 0 rpm and when it reached
at 9,000 rpm disc was broken and after dismantling the specimen same was
critically examined and observed that specimen was broken from the fillet portion
and is shown in Figure 4.22.
Figure 4.22: Broken disc made of Cost iron material
Newly developed testing rig has been used to carry out the actual testing of
specimen made of different materials. Before to run the apparatus more safety
measures were taken as the specimen revolves at very high speed. Dark band
having black powder found on the sliding surfaces causes the crack initiation
development. It is observed that material fracture takes place and the working
life changes with the change of rpm. In three-dimensional finite elements
analysis, it was concluded that fillet was the weakest portion of the dovetail joint
disc for both the cases irrespective of the thickness and type of material.
65
Chapter 5
Numerical Analysis
This chapter contains the analysis detail and results of fretting fatigue of the dovetail models by using the ANSYS 11.0.
66
5.1: Introduction
This Chapter contains the analysis details and results of fretting fatigue of the
dovetail model by using ANSYS 11. The dovetail fretting model which was used
in experimental investigation has been designed in Auto Cad and the same has
been imported in ANSYS 11 for further analysis.
A large number of services failures were found due to fretting fatigue
phenomenon and wide range of applications where degradation in the properties
of the materials is found due to repeated loading and involves the gradual
development and growth of a crack. The dovetail joint of disc and blades of an
aero engine compressor is the best example in which occurrence of fretting
fatigue phenomenon is very common. Due to the centrifugal loading and other
forces fretting fatigue is produced in the disc and blade assembly carried out in
dovetail joints. To calculate the fretting fatigue life, stresses in the neighborhood
of the contact region of disc and blades root is a big challenge for the designers
and researchers due to complexity of the geometry, non linear behavior and
loading.
In the present study numerical analysis of fretting fatigue phenomenon is carried
out by using the commercially available ANSYS 11.0 to access the exact
estimation of stress in the area of contact between the disc surface and blade
root. The present problem of single disk and blades sector of aero engine
compressors is analyzed. In this study with incremental loading, with an elastic
plastic three dimensional plane strain finite element is carried out in the study.
5.2: Numerical work
Fretting fatigue is applied to change the properties of metallic materials during
stress-strain cycles and led to cracking or failure. It describes the situation where
micro slip occurs between two contact surfaces and reduces the fatigue life when
it is compared with plan component. Fretting fatigue is responsible for the large
n
sa
p
e
u
a
co
sh
W
b
co
fa
le
umber of s
afety rule is
ower gene
ngine com
sually occu
ero engine
onsidered a
hown in Fig
When the d
lades is pu
ontact with
atigue phen
ead to crack
Figur
service failu
s implemen
eration etc.
pressors is
urs. A typica
e compress
and disc is
gure 5.1. T
isc revolves
lled as a re
h the two f
nomenon m
k imitation,
re 5.1: Dov
ures and it
nted like ae
. Disc and
s one of th
al disc and
sor is assu
divided into
The disc is
s in clock w
esult of cen
flats contac
may occur
propagatio
vetail joint
1. B
67
t is the mo
erospace, ga
d blades as
he case in
blades asse
umed. Thre
o identical s
assumed to
wise directio
tripetal acc
ct surface o
in the edge
on and ultim
disc and bl
Blades 2.
ost importan
as pipe line
ssembled i
which fret
embly carri
ee dimensi
sectors whi
o be rotatin
on on acce
celeration w
of disc. On
es of the c
mately fatig
ades assem
Disc
nt in the in
es, automob
n dovetail
tting fatigu
ed out in d
onal analy
ich accomm
ng in clock
eleration mo
where as it
n variation
contact reg
gue failure.
mbly configu
Contact fOf disc a
ndustries w
bile and nu
joints of
ue phenome
dovetail join
yses have
modates a b
k wise direc
ode the bas
is restraine
of load fre
gions which
uration
flat surfaces nd blades
where
clear
aero
enon
nts of
been
blade
ction.
se of
ed by
etting
h can
68
Hammouda et al. [226] carried out two dimensional finite element analysis to
predict expected sites of fretting fatigue cracking in axial dovetail joint of disc
and blades of aero engine compressors. It consists of the interface frictional
behavior, expected area of fretting fatigue crack imitation and plasticity induced
in dovetail joints. It was concluded that state of stress and strain was found
multi axial and non proportional everywhere within the model. In the previous
study a developed Hammouda et al. [227] adopted fine element code written in
FORTRAN is adopted. This program was also used previously to;
a. Simulate mechanics of flat contact pad fretting fatigue tests Hammouda et
al. [227].
b. Simulate mixed mode I and II cyclic deformation at the tip of a short
kinked inclined crack with frictional surfaces Hammouda et al. [228].
c. Compute stress intensity factors of a short kinked slant central crack with
frictional surface in uni-axially Hammouda et al. [229] loaded plates.
d. Compute stress intensity factors of a short kinked slant central crack with
frictional surface in bi-axially Hammouda et al. [230] loaded plates. For
central slant cracks with frictional surfaces in uni-axially compressed
plates [231].
e. Analyze friction effects on sliding crack surfaces in full or partial contact
Hammouda et al [232].
Papanikos et al. [233] carried out three-dimensional non linear finite element
analysis of dovetail joint of disc and blades of aero engine compressor by using
the contact element and focused on length of flank angle, fillet and skew angle
upon the resulting stress field. Lucjan [234] carried out failure analysis and
utilized non linear finite element analysis to determine the stress site of disc and
blades segment of an aero engine compressors under operating conditions. He
concluded that high stress zone at the region of lower fire-tree slot. Beisheim
69
and Sinslair [235] carried out three-dimensional finite element analysis of
dovetail joint of aero engine compressors with and without crowning. They
concluded that by adding the crowning the maximum contact stress is reduced
by more than one third.
5.3: Numerical simulation of fretting fatigue of high structural steel
Three-dimensional model of disc and blade assembly fixed in a dovetail joint of
an aero engine compressors was carried out by using the CATIA-V5 and is given
in Figure 5.2.
5.4: Physical properties of high strength structural steel
In the present modeling high strength structural steel is selected as it is the most
commonly used material in the structural components. The young’s modulus
2.e+005MPa, Poison ratio 0.3, density 7.85e-006 kg/mm³ , thermal expansion
1.2e-005 1/°C , tensile yield strength 250MPa, compressive yield strength
250MPa, tensile ultimate strength 460MPa and compressive ultimate strength
0MPa.
Figure 5.2: a). Disc and blades model assembled in dovetail joint.
b). (A. frictional less support, B. rotational velocity and
C. acceleration)
The alternative stress and strain life parameter of high strength structural steel is
given in figure 5.3 and 5.4 respectively.
70
Figure 5.3: Alternating stress MPa
Figure 5.4: Strain life parameters
71
Figure 5.6: Meshing adopted in the present analysis i.e.
disc and blades assembly
5.5: Present idealization
Three-dimensional model of disc and blade assembly fixed in a dovetail joint of
an aero engine compressors is given Figure 5.5. Meshing adopted in the present
analysis is given in Figure 5.6. Identical meshing exists for both disc and blade
assembly whereas only one side of dovetail joint is shown. Volume and mass of
the specimen being used in the analysis is 1.4802e-005 m³, 4.7837e-002 kg
respectively, active bodies are 2, number of nodes are 38729 whereas number of
elements are 19451, element size is 2.e-003m, edge behavior is curve /proximity
refinement, refinement is 3, analysis type is static structural, reference temp is
Figure 5.5: Three dimensional model of disc and blade assembly
fixed in a dovetail joint
72
22°C, number of steps are 101, current step number are 50 and step end time is
250 seconds.
Figure 5.7a shows finest mesh area where the high stress gradient exists. The
magnified model of the lower contact surface of the same area is shown in
Figure 5.7b subjected to fretting fatigue where the finest mesh is used as high
stress gradient. Mesh density decreases with the decrease in distance from
contact surface.
Element size of one is used in the analysis as shown in Figure5.7a, because
further reducing the size of element increases the computational time
exponentially. After performing the initial analysis, local meshing was performed
in critical areas as given in figure 5.7a for precise results. The gravitational effect
and cylindrical frictional support is added in the model as boundary conditions.
To calculate stress/strain and estimate the fatigue life of the model, rotational
velocity applied to the model was increased from zero to a maximum value in
steps, each step is of size 100 rpm in five seconds.
After initial analysis, it was found that the fillet spot area is the weakest link in
the model where the failure due to fatigue initiates. The number of nodes in the
Figure 5.7: a.The model of disc after meshing. b.The enlarged view of model of disc after meshing in critical region
73
critical region was different depending on the mesh size as the mesh size were
changed to a finer value at different stages to have more precise results. To
optimize the fatigue life, different materials and geometrical shapes were
analyzed and then results were compared with the experimental results. Five
points were marked as shown in figure 5.8 on the surface of fillet to investigate
the point of interest where the highest stress/strain was developed.
Stress Probe 1 and 4 are on the edges of fillet and stress Probe 2 and 3 are
between the stress Probe 1 and 4 as we proceed from stress Probe 1 to 4, and
stress Probe 5 is away from stress Probe 4 towards the end of disk.
The numerical analysis results graphs show that the value of Stress intensity,
Principal stress, Equivalent stress and share stress are maximum at Probe 4.
Similarly the value of stress parameters at Probe 2 is comparatively higher than
Figure 5.8: The stress probe 1.2,3,4 and 5 on the fillet surface of the disc
74
at Probe 1 and Probe 5 which are the extreme end points along the fillet of test
specimen. This result also confirms that the maximum stress development point
is at Probe 4 which is the weakest point of the fillet and it shows a good
agreement with the experimental result as well. Stress intensity, Principal stress,
Equivalent stress and share stress at probe 1, 2, 4 and 5 is given in Table 5.1 to
5.4 where as it is shown graphically in Figure 5.9 to 5.12 respectively.
Table 5.1: Stress Intensity at Different Probe With Respect to Time
Time (Sec.)
Stress Intensity Probe 1
[Pa]
Stress Intensity Probe 2
[Pa]
Stress Intensity Probe 4
[Pa]
Stress Intensity Probe 5
[Pa] 25 14198 27039 1.10E+06 1.21E+05 50 47460 2.65E+05 6.65E+05 2.26E+05 75 1.02E+05 6.55E+05 9.66E+05 4.07E+05 100 1.78E+05 1.20E+06 2.26E+06 6.64E+05 125 2.78E+05 1.92E+06 4.03E+06 1.00E+06 150 3.98E+05 2.77E+06 6.16E+06 1.41E+06 175 5.42E+05 3.81E+06 8.74E+06 1.90E+06 200 7.05E+05 4.98E+06 1.17E+07 2.45E+06 225 8.90E+05 6.30E+06 1.50E+07 3.08E+06 250 1.10E+06 7.81E+06 1.87E+07 3.80E+06 275 1.33E+06 9.45E+06 2.28E+07 4.58E+06 300 1.58E+06 1.12E+07 2.73E+07 5.43E+06 325 1.86E+06 1.32E+07 3.23E+07 6.37E+06 350 2.15E+06 1.53E+07 3.75E+07 7.37E+06 375 2.47E+06 1.76E+07 4.32E+07 8.45E+06 400 2.81E+06 2.01E+07 4.93E+07 9.62E+06 425 3.17E+06 2.26E+07 5.58E+07 1.08E+07 450 3.55E+06 2.54E+07 6.26E+07 1.21E+07 475 3.96E+06 2.83E+07 6.99E+07 1.35E+07 500 4.39E+06 3.13E+07 7.75E+07 1.50E+07
75
Figure 5.9: Stress Intensity at Different Probe With Respect to Time
Table 5.2: Maximum Principal Stress at Different Probe With Respect to Time
Time
(Sec.)
Maximum Principal
Stress Probe 1 [Pa]
[Maximum Principal
Stress Probe 2 [Pa]
Maximum Principal
Stress Probe 4 [Pa]
Maximum Principal
Stress Probe 5 [Pa]
25 -148.1 27255 85629 54816 50 -980.02 2.66E+05 2.40E+05 1.14E+05 75 -2341.5 6.57E+05 8.90E+05 2.16E+05 100 -4243.2 1.20E+06 2.39E+06 3.61E+05 125 -6737.6 1.92E+06 4.40E+06 5.52E+05 150 -9730.5 2.78E+06 6.82E+06 7.81E+05 175 -13337 3.81E+06 9.74E+06 1.06E+06 200 -17421 4.99E+06 1.31E+07 1.37E+06 225 -22045 6.31E+06 1.68E+07 1.72E+06 250 -27315 7.83E+06 2.11E+07 2.13E+06 275 -33030 9.47E+06 2.57E+07 2.57E+06 300 -39286 1.13E+07 3.08E+07 3.04E+06 325 -46219 1.33E+07 3.64E+07 3.58E+06 350 -53566 1.54E+07 4.24E+07 4.14E+06 375 -61453 1.76E+07 4.88E+07 4.74E+06 400 -70048 2.01E+07 5.57E+07 5.40E+06 425 -79027 2.27E+07 6.30E+07 6.09E+06 450 -88546 2.54E+07 7.07E+07 6.82E+06 475 -98804 2.84E+07 7.91E+07 7.60E+06 500 -1.09E+05 3.14E+07 8.77E+07 8.42E+06
Stress Intensity With Respect to Time
Time (Sec)0 50 100 150 200 250 300 350 400 450 500 550
Stre
ss I
nte
nsi
ty (
Pa)
0
1e+7
2e+7
3e+7
4e+7
5e+7
6e+7
7e+7
8e+7
9e+7
At Probe 1At Probe 2At Probe 4At Probe 5
76
Maximum Principal Stress With Respect To Time
Time (Sec.)0 50 100 150 200 250 300 350 400 450 500 550
Max
imu
m P
rin
cipa
l Str
ess
(P
a)
-1.5e+7
0.0
1.5e+7
3.0e+7
4.5e+7
6.0e+7
7.5e+7
9.0e+7
1.1e+8
1.2e+8
At Probe 1At Probe 2At Probe 4At Probe 5
Figure 5.10: Maximum Principal Stress at Different Probe With Respect to Time
Time
(Sec.)
Equivalent Stress Probe 1
[Pa]
Equivalent Stress Probe 2
[Pa]
Equivalent Stress Probe 4
[Pa]
Equivalent Stress Probe 5
[Pa] 25 14170 26870 9.82E+05 1.05E+05 50 47238 2.65E+05 5.76E+05 1.95E+05 75 1.01E+05 6.55E+05 8.37E+05 3.53E+05 100 1.77E+05 1.20E+06 2.03E+06 5.77E+05 125 2.76E+05 1.91E+06 3.71E+06 8.71E+05 150 3.95E+05 2.77E+06 5.75E+06 1.23E+06 175 5.39E+05 3.81E+06 8.22E+06 1.65E+06 200 7.01E+05 4.98E+06 1.10E+07 2.14E+06 225 8.85E+05 6.30E+06 1.42E+07 2.68E+06 250 1.10E+06 7.81E+06 1.78E+07 3.31E+06 275 1.32E+06 9.45E+06 2.17E+07 3.99E+06 300 1.57E+06 1.12E+07 2.60E+07 4.73E+06 325 1.85E+06 1.32E+07 3.08E+07 5.55E+06 350 2.14E+06 1.53E+07 3.58E+07 6.42E+06 375 2.45E+06 1.76E+07 4.12E+07 7.35E+06 400 2.80E+06 2.01E+07 4.71E+07 8.37E+06 425 3.15E+06 2.26E+07 5.33E+07 9.44E+06 450 3.53E+06 2.54E+07 5.98E+07 1.06E+07 475 3.94E+06 2.83E+07 6.68E+07 1.18E+07 500 4.36E+06 3.13E+07 7.41E+07 1.30E+07
Table 5.3: Equivalent Stress at Different Probe With Respect to Time
77
Figure 5.11: Equivalent Stress at Different Probe With Respect to Time
Table 5.4: Shear Stress XY at Different Probe With Respect to Time
Time
(Sec.)
Shear Stress XY
Probe 1 [Pa]
[Shear Stress XY
Probe 2 [Pa]
Shear Stress XY
Probe 3 [Pa]
Shear Stress XY
Probe 4 [Pa]
25 350.07 -12411 -3.37E+05 48806 50 1328 -1.22E+05 -1.63E+05 1.05E+05 75 2929.2 -3.01E+05 1.22E+05 1.97E+05 100 5166 -5.51E+05 5.21E+05 3.26E+05 125 8099.8 -8.79E+05 1.04E+06 4.96E+05 150 11620 -1.27E+06 1.67E+06 6.98E+05 175 15862 -1.75E+06 2.43E+06 9.43E+05 200 20665 -2.29E+06 3.28E+06 1.22E+06 225 26104 -2.89E+06 4.25E+06 1.53E+06 250 32302 -3.59E+06 5.36E+06 1.89E+06 275 39025 -4.34E+06 6.56E+06 2.28E+06 300 46383 -5.16E+06 7.87E+06 2.70E+06 325 54536 -6.07E+06 9.32E+06 3.17E+06 350 63178 -7.04E+06 1.09E+07 3.67E+06 375 72455 -8.08E+06 1.25E+07 4.21E+06 400 82564 -9.21E+06 1.43E+07 4.79E+06 425 93125 -1.04E+07 1.62E+07 5.40E+06 450 1.04E+05 -1.16E+07 1.82E+07 6.04E+06 475 1.16E+05 -1.30E+07 2.03E+07 6.74E+06 500 1.29E+05 -1.44E+07 2.26E+07 7.46E+06
Equivalent Stress With Respect to Time
Time (Sec.)0 50 100 150 200 250 300 350 400 450 500 550
Equ
ival
ent
Stre
ss (
Pa)
0
1e+7
2e+7
3e+7
4e+7
5e+7
6e+7
7e+7
8e+7At Probe 1At Probe 2At Probe 4At Probe 5
78
Shear Stress XY With Respect to Time
Time (Sec.)
0 50 100 150 200 250 300 350 400 450 500 550
Sh
ear
Stre
ss X
Y (
Pa)
-2e+7
-1e+7
0
1e+7
2e+7
3e+7
At Probe 1At Probe 2At Probe 4At Probe 5
Figure 5.12: Shear Stress XY at Different Probe With Respect to Time
79
Chapter 6
Results and Discussion
This chapter gives the detail of experimental test results obtained and the
discussion of the fretting fatigue.
80
6.1: General discussion
This chapter presents the results obtained from experimental and numerical
analyses carried out with the help of commercially available software ANSYS 11.0
from the present work. Fretting fatigue apparatus developed by the researchers
for the experimentation have already been discussed in chapter 4 under heading
4.2. The entire above mentioned test rigs are being used till date and main
problem is that all the test rigs have their own results and hardly match these
results with each other. Newly developed test rig can revolve the specimen from
0 to 20,000 rpm in the incremental form and subsequently reduces the speed
from 20,000 to 0 rpm in the same manner.
specimen moves from 0~20,000 rpm which predict the take of the aero plane
called acceleration mode where as when it move from 20,000 to 0 rpm, it
simulates the landing and stopping of the engine i.e. called declaration mode. In
this regard various tests were carried out with and without specimen to ensure
the simulation of acceleration and deceleration mode.
Assembly of disc of typical aero engine compressor is assumed. Disc is divided
into identical sectors and each accommodates a blade. Disc rotates in clockwise
direction from 0 to 20,000 rpm and back to 0 rpm. Disc and blades are fixed on
the main motor shaft in such a way that blades could move easily. Upper and
lower holding plates support blades. Disc-blade assembly revolves up to 20,000
rpm and back to 0 rpm. Gradually increase in speed simulates as an acceleration
mode whereas decrease in speed gradually simulates as deceleration
mode/stopping of the engine.
Forces are acting on the blades during revolution. Due to high speed blades
tends to move outward because of centrifugal force and inward because of
centripetal force. Tests are carried out at ambient temperature by considering
frictional force acting between two contacting surfaces. The specimen speed is
increased and decreased incrementally and each step is defined. Newly
81
developed test rig is verified and tests are performed using different materials. In
this thesis research work on the fretting fatigue phenomenon in dovetail joint for
aero engine compressor is performed experimentally and computed numerically
by using newly developed experimental fretting fatigue test rig and three-
dimensional finite elements analysis through commercially available software
ANSYS 11.0 respectively.
6.2: Results and discussion
Different type of fretting fatigue, experimental testing rigs have been developed
by researchers to evaluate fretting fatigue phenomenon in the disc-blades
assembly fixed in dovetail joint of an aero engine compressor. Pauw et al. [21]
concluded that researchers used the existing rigs as such or with certain
modification or develop new testing rigs to fulfill their requirement but still no
standard and generally acceptable testing rig is available for fretting fatigue
experimental. Results of any two type of testing rigs hardly mach with each
other.
A new test rig is designed and developed with additional features revolving the
specimen in a sinusoidal wave. The test rig is verified with & without specimen
as discussed in para 2.5. Experiment is performed at room temperature and
material used is of bakelite and structural steel. In case of structural steel,
specimen remained in revolving condition in the range of 0 to 20,000 rpm and
after 100 cycles no failure was observed. The specimen is removed from the test
rig and tests are carried out critically and found dark band on the contact
surfaces of disc and blade as investigated by past researchers.
Both the specimens made of bakelite & structural steel are rotated in the range
of 0 to 20,000 rpm and disc made of bakelite material is broken at 5,000 rpm
whereas structural steel is broken at 14,000 rpm from the excepted weakest
portion of the fillet. Comparison of stresses at the five points shows stress
variation by moving from inner edge of the fillet toward the end of the disc.
82
Although the equivalent normal stress is greater than the shear stress, but shear
strain is greater than the normal strain, which clearly indicates that the failure is
due to shearing. The shear strain due to the shear stresses is highest as
compared to the other types of stresses. This shows that the failure is based on
maximum shear theory. The maximum shear stress variation at stress probe 4 is
given in Figure 6.1. Shear stress variation at stress probes 1~7 is given in Table
6.1 and the trend is shown in Figure 6.2. The maximum principal stress at stress
probe 4 is given in Figure 6.3 and maximum Principal Stress variation at stress
probe 1~7 is given in Table 6.2, whereas trend is shown in Figure 6.4.
Figure 6.1: Maximum shear stress at stress probe 4
Table 6.1: Stress variation at stress Probes 1~7
Stress points XY Shear YZ Shear XZ Shear 1 0.00 0.00 0.13 2 0.07 0.00 0.02 3 0.12 0.00 0.02 4 0.25 0.07 22.57 5 0.15 0.05 19.0 6 0.05 0.03 13.0 7 0.01 0.02 7.46
83
Stress Variation Along Stress Points
Stress Points0 1 2 3 4 5 6 7 8
Stre
ss (M
Pa)
0
5
10
15
20
25
XY Shear YZ ShearXZ Shear
Figure 6.2: Trend of shear stress variation at stress probes 1~7
Figure 6.3: Maximum principal shear stress at stress probe 4
84
Table 6.2: Principal Stress variation at stress Probes 1~7
Stress points
Equivalent (von-Mises)
Maximum Principal
Middle Principal
Minimum Principal
1 4.63 0.00 0.00 0.00 2 31.34 31.40 0.07 0.06 3 49.66 49.63 0.02 0.00 4 74.10 87.66 17.56 10.12 5 77.09 82.54 15.56 9.19 6 55.14 60.24 5.10 5.1 7 13.04 8.42 0.00 -0.05
Principal Stress variation at stress Probes
Probes Points
0 1 2 3 4 5 6 7 8
Stre
ss (M
Pa)
-20
0
20
40
60
80
100
Equivalen (von-Mises)Maximum PrincipalMiddle PrincipalMinimum Principal
Figure 6.4: Trend of principal shear stress variation at stress probes 1~7
From Table 6.1 and Figure 6.2 it is shown that shear stress at stress point 4
along XZ, YZ and XY plan is 0.25, 0.07 and 22.57 MPa respectively which is
highest as compared to the other stress points 1~7 whereas in Table 6.2 and
Figure 6.4 principal stresses value at stress point 4 is 10.12, 17.56, 87.66 and
74.1MPa which has the maximum values as compared to other stress points 1
to 7.
85
The exact node numbers from where failure/shear initiates, can be found, but
not given here because of mesh refinement at different stages. Points along the
fillet are provided with stresses to find exact location of failure initiation. The
point of maximum stress is about 3 to 4 the distance between inner edges and
the outer edge of fillet, when moving along the inner of the fillet toward the end
of disc. inner
To optimize the design, a number of iterations were performed by changing the
materials and thickness. Irrespective of the material and thickness of the disc,
each iteration analysis shows that fillet is the weakest link in model as shown in
Figure 6.5.
Figure 6.5: Maximum principal stress at stress probe 4 (crack initiation tip).
By changing the material or thickness of the disc failure initiation located do not
change, only the fatigue life is affected that is same as in the prototype testing.
86
Chapter 7
Conclusions & Recommendations
for Future Work
This chapter gives the major concluding remarks and findings of this research
program together with possible future lines of work related with the present
research work.
87
7.1: Conclusion
This research work designed and developed a unique, flexible, safe and
improved version of fretting fatigue experimental testing rig to study the disc and
blades of dovetail joint of an aero engine compressor. Experimental test rig has
the capability to revolve actual geometry in sinusoidal wave to simulate aero
engine compressor. The research describes fretting fatigue phenomenon in
dovetail joints of aero engine compressor. The main conclusion drawn from this
work are following;
1. Dark band having black powder found on the sliding surfaces causes the
crack initiation development.
2. In three-dimensional finite elements analysis it is concluded that fillet is the
weakest portion of the dovetail joint disc for both the cases irrespective of the
thickness and type of material.
3. When the material fracture takes place the working life changes with the
change of rpm.
4. Fatigue cracks develop in the region of tension rather than compression.
5. By changing the disc material fatigue life changes.
6. With the change of disc thickness fatigue life changes.
7.2: Recommendations
It is expected that application of this apparatus may continue to extend beyond
the present broader range being a new method. This research work is useful to
carryout testing of various materials which could probably be used in different
industries as no such work has been carried out. Testing system is a contribution
to develop such testing apparatus / facilities / centre of study in the teaching and
technical institutions for further research on the subject.
88
It is recommended to use this apparatus with the addition of the following
factors considered during the simulation of aero engine compressors and detail
of the same is given as under:
1. Introduction of controls in the apparatus that could vary temperature of the
specimen to analyze the effect on the life of component due to fretting
fatigue at varying temperatures.
2. The experimental work should continue to determine the effect of a
controlled environment (temperature, corrodibility etc) as most investigations
reported cases of fretting in normal atmospheres, but there are also cases
when fretting occurs in different temperature conditions. These effects should
be incorporated in fretting fatigue experimental test rig.
3. To introduce vibration in the experimental test rig to see the effect in the
specimen being tested.
4. To use high speed camera to investigate the crack initiation, propagation and
failure with the help of boroscopy.
5. To optimize the fillet radius to minimize the stresses due rotation of the disk
and blade.
6. To test various material and different type of joints using this apparatus.
7. It will be possible to assess the influence of different variables and give better
understanding of the fretting phenomena. Further tests and calibrations are
needed in order to properly quantify the variables.
89
References
90
[1] Ramakrishna Naidu, Raman, Effect of contact pressure on fretting fatigue
behaviour of Al-Mg-Si alloy AA6061, International journal of Fatigue 27
(2005): pp 283-291.
[2] Hamid, R.K. Yahya, Influence of fretting on the fatigue strength at the
vise clamp-specimen interface, Bull. mater. Sci. 26 (2003): pp 749-754.
[3] A.L. Hutson, L. Chambon, Observations of fretting fatigue micro damage
of Ti-6Al-4V, Wear 255 (2003): pp 259–268.
[4] M.H. Attia, R. B. Waterhouse, editors. Standardization of fretting fatigue
test methods and equipment, ASTM STP 1159, American Society Testing
Materials, Philadelphia, (1992).
[5] O. Jin, S. Mall, Effects of independent pad displacement on fretting fatigue
behavior of Ti–6Al–4V, Wear 253 (2002): pp 585–596.
[6] G.W. Stachowiak, A.W. Batchelor, Engineering Tribology, 2rd ed.,
Butterworth-Heinemann, (2001)
[7] J.M. Dobromirski, Variables of fretting process: are there 50 of them? In:
M. Helmi Attia and R.B. Waterhouse, Editors, Standardization of fretting
fatigue tests and equipment, ASTM STP 1159, American Society for
Testing and Materials, Philadelphia (1992): pp 60–66.
[8] J.J. Madge, S.B. Leen, P.H. Shipway, The critical role of fretting wear in
the analysis of fretting fatigue, Wear 263 (2007): pp 542-551.
[9] O. Vingsbo, S. Soderberg, On fretting maps, Wear 126 (1988): pp 131–
147.
[10] O. Jin a, S. Mall, J. H. Sanders, S.K. Sharma, Durability of Cu–Al coating
on Ti–6Al–4V substrate under fretting fatigue, Surface & Coatings
Technology 201 (2006): pp 1704–1710.
91
[11] Y. Fu, J. Wei, A.W. Batchelor, Some considerations on the mitigation of
fretting damage by the application of surface-modification technologies,
Journal of Materials Processing Technology 99 (2000): pp 231-245.
[12] M.H. Zhu, Z.R. Zhou, Ph. Kapsa, L. Vincent, Radial fretting fatigue damage
of surface coatings, Wear 250 (2001): pp 650–657.
[13] C.H. Hager, J. Sanders, S. Sharma, A. Voevodin, A. Segall, The effect of
temperature on gross slip fretting wear of cold-sprayed nickel coatings
onTi6Al4Vinterfaces, Tribology International 42 (2009): pp 491–502.
[14] G.W. Powell, ASM Handbook, Volume 11: Failure Analysis and Prevention,
ASM International, Materials Park, OH, (1986).
[15] P. papanikos, S.A. Meguid. (1998) Three-dimensional nonlinear finite
element abnalysis of dovetail joints in aerioengine discs. Finite Elemt Anlys
Desgn 29: pp 173.
[16] N.E.Dowling, Mechanical Behavior of Materials: Engineering Methods for
Deformation, Fracture, and Fatigue, 2nd edition, Pretince Hall, (1998).
[17] Few examples of failure data down loaded from internet
[18] M. Buciumeanu, A.S. Miranda, A.C.M. Pinho, F.S. Silva, Design
improvement of an automotive-formed suspension component subjected
to fretting fatigue, Engineering Failure Analysis 14 (2007): 810-821.
[19] E.M, Eden, w.n. Rose, The endurance of metals. Proceeding of the
Institute for Mech. Engineers, 4,(1911): pp 839.
[20] G.A. Tomilson The rusting of steel surfaces in contact. Proceeding of the
Royal Society, London, Series A, 115, (1927): pp 472.
[21] U.R Evans, H.A Miley, Measurements of oxide films on copper and iron,
Nature, 139, (1931): pp 283.
92
[22] J.O Almen, Lubrication and false brinelling of ball and roller bearings.
Mechanical Engineering, 59, (1937): pp 415.
[23] R.E. Peterson, A.M Wahl, Fatigue of shafts at fitted members, with a
related photoelastic analysis. Transactions of ASME, 57, A-1 (1931).
[24] W.E. Cambell, U.B. Thomas, (1938) Films on freshly abraded copper
surfaces, Nature, 142, (1938): pp 253.
[25] W.E. Cambell, Studies in boundary lubrication. Transactions of ASME, 91,
(1939): pp 633.
[26] G.A. Tomilson, P.L. Thrope, An investigation of fretting corrosion of closely
fitting surfaces. Proceeding of the Institute for Mech.1 Engineers, 141,
(1939): pp 223.
[27] R.B. Waterhouse, Influence of local temperature increase on the fretting
corrosion of mild steel. J Iron steel Insit, 197, (1961): pp 301.
[28] D.W. Hoppner, Mechanisms of fretting fatigue ESIS 18. R.B Waterhouse,
T.C Lindley eds. London: Mech Engg Publications, (1994): pp 3.
[29] K. Nishioka S. Nishimura K. Hirakawa, Fundamental investigation of
fretting fatigue-part 1. On the relative slip amplitude of press-fitted axle
assemblies. Bulletin of JSME, 12, (1968): pp 180.
[30] K. Nishioka, K. Hirakawa, Fundamental Investigation of fretting fatigue-
part 2, Fretting fatigue testing machine as some test results. Bulletin of
JSME, 12, (1969): pp 180.
[31] P.L. Hurricks, The mechanisms of fretting-a review. Wear, 15, (1970): PP
389-409.
93
[32] P.R. Edwards, The application of fracture mechanics to predicting fretting
fatigue. In Fretting Fatigue, R.B Waterhouse, ed. London, Elsevier applied
science publishers, (1981): pp, 67.
[33] S.A. Brown, K. Merritt, Fretting corrosion in saline and serum. J Biomedical
Mater Research, 15, (1981): pp, 479.
[34] S.D. Cook, G.J. Gianoli, Fretting corrosion in orthopedic alloys, I biomat:
Med, Dev, Art. Org, 11, (1983): pp 281.
[35] M.Buciumeanu, Predication of fretting fatigue life. Minho university, thesis
(2009).
[36] E.M Eden, W.N Rose, The endurance of metals. Proceeding of the
Institute for Mech. Engineers, 4, (1911): pp 839.
[37] G.A Tomilson, The rusting of steel surfaces in contact. Proceeding of the
Royal Society, London, Series A, 115, (1927): pp 472.
[38] R.E Peterson, A.M Wahl, Fatigue of shafts at fitted members, with a
related photoelastic analysis. Transactions of ASME, 57, A-1,(1935).
[39] W.E Cambell, U.B Thomas, Films on freshly abraded copper surfaces,
Nature, (1938): pp 142.
[40] D. Nowell. An analysis of Fretting fatigue. Doctor of Philosophy Thesis,
Lincoln College, Oxford University(1988).
[41] R.L. Burguete and E.A. Patterson, A photo elastic study of contact
between a cylinder and a half-space. Expt Mech, 37 (3), (1997): pp 314.
[42] J.O Arau’, D. Nowell, The effect of rapidly varying contact stress fields on
Fretting fatigue. Int J Fatigue; 24 (2002): pp 763.
94
[43] H. Proudhona, S. Fouvry and J. Y. Buffy, A fretting crack initiation
prediction taking into account ;the surface roughness and the crack
nucleation process volume. Int J Fatigue 27 (2005): pp 569.
[44] S. Naboulsi and S. Mall, Fretting fatigue crack initiation behaviour using
process approach and finite element analysis. Tribology Int,; 36, (2003):
pp 121.
[45] J. Calcaterra and S. Naboulsi, Design methodology to investigate contact
fatigue damage in turbine engine hardware. Int J Fatigue 27, (2005): pp
1133.
[46] D.A Hills, D. Nowell , Mechanics of Fretting Fatigue, Kluer Academic
Plubishers, Canada, (1994).
[47] S. Mall, S.A Namjoshi,, W.J Porter, Effects of microstructure on Fretting
fatigue crack initiation behaviour of Ti-6Al-4V, Materials Science and
Engineering A383,(2004): PP 334-340.
[48] W. Ren, S. Mall, J.H Sanders, S.K Sharma, Evaluation of coatings on TI-
6Al-4V substrate under Fretting fatigue, Surface & Coatings Technology,
(2005): pp 177-188.
[49] B.U Wittkowsky, P.R Birch, J. Dominguez, S. Surech, An apparatus for
quantitative Fretting fatigue testing, Blackwell Science Ltd. Fatigue Fract
Engng Struct 22, (1999): pp 307-320.
[50] D.W Hoppner, Fretting fatigue case studies of engineering components.
Tribology internation 39 (2006): PP1271-1276.
[51] D.Nowell and D.Hills, Crack initiation in Fretting fatigue, wear; 136,
(1990): pp 329.
[52] B.P. Conner, A.L. Hutson and L. Chambonc Observations of Fretting
fatigue micro-damage of Ti-6A1-4V. Wear 255, (2003): pp 259.
95
[53] B.P. Conner, T.C. Lindley, T. Nicholas and S. Suresh, Application of
fracture mechanics based life prediction method for contact fatigue. Int. J
Fatigue.26, (2004): pp 511.
[54] Sam Naboulsi and Jeff Calcaterra Fretting fatigue investigation of dovetail,
Department of Aeronautics and Astronautics Air Force Institute of
Technology (AFIT/ENY) Air Force Research Laboratory Materials and
Manufacturing Directorate (AFRL/MLLMN).
[55] A. Vadiraj, M. Kamaraj, Characterization of fretting fatigue damage of PVD
TiN coated biomedical titanium alloys, Surface&Coatings Technology 200
(2006): pp 4538-4542.
[56] D. W. Hoeppner, Fretting fatigue case studies of engineering components,
Tribology International 39 (2006): pp 1271-1276.
[57] O. Jin, S. Mall, Effects of slip on fretting behavior: experiments and
analyses, Wear 256 (2004): pp 671-684.
[58] R. Magaziner, O. Jin, S. Mall, Slip regime explanation of observed size
effects in fretting, Wear 257 (2004): pp 190-197.
[59] C. Santus, Fretting fatigue of aluminium alloy in contact with steel in oil
drill pipe connections, modeling to interpret test results, International
Journal of Fatigue 30 (2008): pp 677-688.
[60] K. Iyer, Peak contact pressure, cyclic stress amplitudes, contact semiwidth
and slip amplitude: relative effects on fretting fatigue life, International
Journal of Fatigue 23 (2001): pp 193-206.
[61] W. Ren, S. Mall, J.H. Sanders, S.K. Sharma, Evaluation of coatings on Ti–
6Al–4V substrate under fretting fatigue, Surface & Coatings Technology
192 (2005): pp 177– 188.
96
[62] C. Navarro, M. Garcia, J. Dominguez, A procedure for estimating the total
life in fretting fatigue, Fatigue Fract Engng Mater Struct 26 (2003): pp
459-468.
[63] C. Li, S. Panier, S. Hariri, Application of a simplified method to evaluate
the inelastic state due to fretting fatigue, Technology intenational
39(2006): pp 1092-1099.
[64] O.P. Datsyshyn, V.M. Kadyra, A fracture mechanics approach to prediction
of pitting under fretting fatigue conditions, International 167 Journal of
Fatigue 28 (2006): pp 375–385.
[65] C.L. Neslen, S. Mall, S. Sathish, Nondestructive Characterization of
Fretting Fatigue Damage, Journal of Nondestructive Evaluation 23 (2004):
pp 153-160.
[66] B. Alfredsson, A. Cadario, A study on fretting friction evolution and fretting
fatigue crack initiation for a spherical contact, International Journal of
Fatigue 26 (2004): pp 1037–1052.
[67] S.Shinde and D.W. Hoeppner, Quantitative analysis of fretting wear crack
nucleation in 7075-T6. Wear 259, (2005): pp 271.
[68] C.Ruiz and D. Nowell, Designing against fretting fatigue in aero-engines.
In: Fuentes M. Elices M, Martin-Meizoso A, Martinez-Esnaola JM, editors.
Fracture mechanics, applications and challenges, Invited papers presented
at the 13th European Conference on Fracture, ESIS publication 26.
Amsterdam: Elsevier, (2000): pp 73.
[69] J. Durelli, K.W. Daly, W.F. Riley, Stress and strength studies on turbine
blade attachment, SESA Proc. XVI (1957: pp 171-186.
97
[70] V.J. Parks, R.J. Sanford, Three-dimensional photoelastic stress analysis of
the dovetail region of the TF-30 turbine engine third-stage fan, NRL
Report, NRL (1978): pp 8276.
[71] R.Rajasekaran, D. Nowell, Fretting fatigue in dovetail blade roots:
Experiments and analysis. Tribology Int., Article in press, Corrected proof,
(2006).
[72] C. Ruiz, P.H.B Boddington, K.C. Chen, An investigation of fatigue and
fretting in a dovetail joint, Exp. Mech. 24, (1984): pp 208.
[73] Ruiz, D. Post, R Czarnek, Moire, Interferometric study of dovetail joints,
ASME J. Appl. Mech. 52 (1985): pp 109.
[74] P.H.B. Boddington, K. Chen, C. Ruiz, The numerical analysis of dovetail
joints, Comput. Struct.20, (1985): pp 731.
[75] M.J. He, C. Ruiz, Fatigue life of dovetail joints: verification of a simple
biaxial model, Exp. Mech.29 (1989): pp 126.
[76] B. Kenny, E. Patterson, M. Said, K. Aradhya, Contact stress distributions in
a turbine disc dovetail type joint – A comparison of photoelastic and finite
element results, Strain 27 (1991): pp 21.
[77] A.D. Nurse, E. Patterson, Experimental determination of stress intensity
factors for cracks in turbine discs, Fatigue Fract. Gng. Mater. Struct. 16
(1993): pp 315.
[78] K.J. Nix T.C. Lindley, The initiation and propagation of small defects in
fretting fatigue. Conference life assessment of dynamically loaded
materials and structures, Engineering Materials Advisory Services Limited,
UK, (1985): pp 89.
98
[79] T.C. Lindley, J.K. Nix, The role of fretting in the initiation and early growth
of fatigue cracks in trubo-generator materials. Multiaxial fatigue. ASME,
Philadelphia, USA, (1985): pp 340.
[80] Y. Kong, H. Hatta, M. Toyoda and T. Sugibayashi, Application of three-
dimensionally reinforced carbon-carbon composites to dovetail joint
structures. Composites Science Tech 62, (2002): pp 2143.
[81] M. Ciavarella, G. Demelio, A review of analytical aspects of fretting
fatigue, with extensions to damage parameters, and applications to
dovetail joints. Int. J. Solid. 38, (2001): pp 1791.
[82] T.N. Farris, M.P. Szolwinski, and G. Harish Fretting in Aerospace
Structures and Materials, Fretting Fatigue: Current Technology and
Practices, D. Hoeppner, V. Chandrasekaran, and C, Elliott, eds., American
Society for Testing and Materials, Philadephia, ASTM STP, (2000): PP
1367.
[83] H. Murthy, G. Harish, and T. Farris, Efficient Modeling of Fretting Fatigue
of Blade/Disk Contacts Including Load History Effects, ASME J. Trigbol
(2002).
[84] P.T. Rajeev, and T.N Farris, Two Dimensional Contact of
Dissimilar/Anisotropic Materials, Proc. of 42nd AIAA/ASME/ASCE/AHS/ASC.
Structures, STrudrtural Dynamics and Materials Conference, Seattle, WA,
AIAA, Washington, DC, I, (2001): pp 515.
[85] P. Papanikos, S.A Meguid, Theoretical and experimental studies of
fretting-initiated fatigue failure of aero engine compressor discs, Fatigue
Frat. Eng Mater. STruct. 17 (1994): pp 539.
99
[86] P. papanikos, S.A. Meguid. Three-dimensional nonlinear finite element
abnalysis of dovetail joints in aero engine discs. Finite Elemt Anlys Desgn
29, (1998): pp 173.
[87] S.A. Meguid, P.S. Kanth and A. Azekanski, Finite element analysis of fir-
tree region in turbine discs. Finite Element Anlys Desgn, 35, (2000): pp
305.
[88] G. B. Sinclair, N. G. Cormier, J.H. Griffin and G. Meda, Contact stresses in
dovetail attachments: Finite element modeling. J Eng. Gas Trib Power
124, (2002): pp 182.
[89] J.R. Beisheim and G.B. Sinclair, On the three-dimensional finite element
analysis of dovetail attachments. J Turbomach 125, (2003):pp 372.
[90] D. Nowell, Advances in the understanding of fretting fatigue with
reference to the gas turbine engines. In: Morton J, Paris F, editors,
Progress in structural mechanics. University of Seville, (2000): pp 61.
[91] P.J. Golden, J.R. Calcaterra, A fracture mechanics life prediction
methodology applied to dovetail fretting. Tribology Int, Article in press,
corrected proof (2006).
[92] Giannakopoulous A.E, Suresh S A, Three-Dimensional Analysis of Fretting
Fatigue, Acta materialia et Metalugica, 46, (1998): pp 177.
[93] d.w. Hoppner, Fretting fatigue case studies of engineering components.
Tribology Int, Corrected proof, Article in press (2006).
[94] D. Nowell, D.A. Hills R. Moobola, Length scale considerations in fretting
fatigue. In: Hoeppner DW, Chandrasekaran V, Elliott CB, editors. Fretting
Fatigue; current technology and practices, ASTM STP 1367. Philadelphia,
PA: ASTM (2000): pp 141.
100
[95] M.P Szolwinski. G. Harish, P.A. Mc Veigh, T.N. Farris, Experimental study
of fretting fatigue crack nucleation in aerospace alloys with emphasis on
life prediction. In: Hoeppner DW,Chandrasekaran V, Elliott CB, editors.
Fretting fatigue: current technology and practices, ASTM STP 1367.
Pennsylvania, PA:ASTM, (2000): pp 267.
[96] H. Murthy, T.N. Farris, D.C. Slavik, Fretting fatigue of Ti-6A1-4V subject to
blade/disk contact loading. In: Materials Science Research, International
Special Technical Publication 1, Materials Science for the 21st Century.
Japan Society of Mat Science, (2001): Pp 947.
[97] D. Nowell, J.A Arau’jo, Analysis of pad size effects in fretting fatigue using
short crack arrest methodologies. Int J Fatigue;21 (1999): pp 947.
[98] D. Nowell, J.A Arau’jo, Prediction of fretting fatigue life using volume
averaged multimaxial initiation parameters. In: BeckerAA, editor. Proc. 4th
Int. Conf. on Modern Practice in Stress and Vibration Analysis,
Nottingham; 2000 Sep 5-7. West Midlands; EMAS; (2000): pp 223.
[99] Del Puglia, F. Pratesi, Experimental procedure and parameters involved in
fretting fatigue tests. Fretting Fatigue, ESIS 18. R. B Waterhouse and T.C
Lindley eds., London: Mechanical Engineering Publications, (1994): pp,
219.
[100] J. Beard, An investigation into the mechanics of fretting fatigue. Doctor of
Philosophy Thesis. University of Salford (1982).
[101] D.W. Hoeppner, G.L. Goss, Fretting of aircraft control surfaces, AGARD
Conference Proceedings No. 161, AGARD, Munich (1974).
[102] D.W. Hoeppner, G.L.A. Goss, Fretting fatigue damage threshold concept.
Wear, 27, (1974): (1974): pp 175.
101
[103] D.W. Hoeppner, G.L Goss, Metallographic analysis of fretting fatigue
damage in Ti-6A1-4V MA and 7075-T6 aluminum. Wear 27, (1974): pp
175.
[104] R.B. Waterhouse, Book Wear, 5 (1962): pp 235.
[105] R.B. Waterhouse, T.C Lindley, Fretting Fatigue. ESIS Publication No. 18,
European Structural Integrity Society. Mechanical Engineering Publication
Limited, London (1994).
[106] R.B. Waterhouse, Fretting Corrosion, Oxford: Pergamon Press (1972).
[107] R.B. Waterhouse, Fretting Fatigue, London: Elservier Applied
Science(1981).
[108] K. Iyer, Peak Contact pressure, cyclic stress amplitudes, contact semi-
width and slip amplitude: relative effects on fretting fatigue life. Int. J
Fatigue, 3, (2001): pp 193.
[109] S.E. Kinyon, D.W. Hoeppner, Spectrum load effects on the fretting
behavior of Ti-6Al-4V, “Fretting Fatigue: Current Technology and
Practices,” ASTM STP 1367, D.W. Hoeppner, V. Chandrasekaran, and C.B.
Elliot, Ed., ASTM, West Conshohocken, PA, (2000): pp. 282.
[110] H.A. Fadag, S. Mall, V.K. Jain, A finite element analysis of fretting fatigue
crack growth behavior in Ti–6Al–4V, Engineering Fracture Mechanics 75
(2008): pp 1384–1399.
[111] K. Nakazawa, N. Maruyama, T. Hanawa, Effect of contact pressure on
fretting fatigue of austenitic stainless steel, Tribology Internatinal 36
(2003): pp 79-85.
[112] A. Ramalho, l.M. Correia, J.D. Costa, Fretting fatigue of zinc coated low
carbon steel EN H320, Tribology International 33 (2000): pp 761-768.
102
[113] S. Fouvry, P. Duó, Ph. Perruchaut, A quantitative approach of Ti-6Al-4V
fretting damage: friction, wear and crack nucleation, Wear 257 (2004): pp
916-929.
[114] H. Proudhon, S. Fouvry, G.R. Yantio, Determination and prediction of the
fretting crack initiation: introduction of the (P, Q, N) representation and
definition of a variable process volume, International Journal of Fatigue 28
(2006): pp 707-713.
[115] K. Iyer, S. Mall, Effects of cyclic frequency and contact pressure on
fretting fatigue under two-level block loading, Fatigue Fract. Eng. Mater.
Struct. 23 (2000): pp 335–346.
[116] O. Jin, S. Mall, Influence of contact configuration of fretting fatigue
behavior of Ti6Al4V under independent pad displacement condition,
International Journal of Fatigue 24 (2002): pp 1243-1253.
[117] B. U Wittkowsky, P. R., Birch, J. Dominguez, S. Surech, “An apparatus for
quantitative fretting fatigue testing”, Blackwell Science Ltd. Fatigue Fract
Engng Struct 22 (1999) pp 307-320.
[118] Z.R. Zhou, K. Nakazawa, M.H. Zhua, N. Maruyama, Ph. Kapsa, L. Vincent,
Progress in fretting maps, Tribology International 39 (2006): pp 1068–
1073.
[119] S. Fouvry, P. Kapsa, L. Vincent, Quantification of fretting damage, Wear
200 (1996): pp 186-205.
[120] E. Matikas, P.D. Nicolaou, Development of a Model for the Prediction of
the Fretting Fatigue Regimes, J. Mater. Res. 16 9 (2001): pp 2716–2723.
[121] W. Huang, B. Hou, Y. Pang, Z. Zhou, Fretting wear behaviour of AZ91D
and AM60B magnesium alloys, Wear 260 (2006): pp 1173–1178.
103
[122] J.J. Madge, S.B. Leen, P.H. Shipway, A combined wear and crack
nucleation–propagation methodology for fretting fatigue prediction,
International Journal of Fatigue 30 (2008): pp 1509–1528.
[123] J.A. Pape, R.W. Neu, A comparative study of the fretting fatigue behavior
of 4340 steel and PH 13-8 Mo stainless steel, International Journal of
Fatigue 29 (2007): pp 2219–2229.
[124] C.B.Elliot, D.W. Hoeppner, The importance of wear and corrosion on the
fretting fatigue behavior of two aluminum alloys, Wear 236 (1999): pp
128-133.
[125] H. Lee, S. Mall, Stress relaxation behaviour of shot-peened Ti-6Al-4v
under fretting fatigue at elevated temperature, Materials Science and
Engineering A366 (2004): pp 412-420.
[126] O. Jin, S. Mall, O. Sahan, Fretting fatigue behaviour of Ti-6Al-4V at
elevated temperature, International Journal of Fatigue 27 (2005): pp 395-
401.
[127] B. Podgornink, M. Kalin, J. Vizintin, F. Vodopivec, Microstructural 172
changes and contact temperatures during fretting in steel-steel contact,
ASME 123 (2001): pp 670-675.
[128] S. Onder, Fretting fatigue behavior of a titanium alloy Ti-6Al-4V at
elevated temperature, Thesis.
[129] A.L. Hutson, C. Neslen, T. Nicholas, Characterization of fretting fatigue
crack initiation processes in CR Ti-6Al-4V, Tribology International 36
(2003): pp 133-143.
[130] H. Lee, S. Sathish, S. Mall, Evolution of residual stress under fretting
fatigue, Journal of materials science 39 (2004): pp 7089-7092.
104
[131] K. Elleuch, H. Proudon, C. Meunier, S. Fouvry, Development of a contact
compliance method to detect the crack propagation under fretting,
Tribology International 39 (2006): pp 1262-1270.
[132] J.A. Araujo, D. Nowell, The fretting fatigue limit based on local stress at
the contact edge, Fatigue Fract Engng Mater Struct 24 (2001): pp 791-
801.
[133] C. Navarro, J. Domínguez, Initiation criteria in fretting fatigue with
spherical contact, International Journal of Fatigue 26 (2004): pp 1253–
1262.
[134] H. Lee, S. Mall, Effect of dissimilar mating materials and contact force on
fretting fatigue behavior of Ti–6Al–4V, Tribology International 37 (2004):
pp 35–44.
[135] K. Kubiak, S. Fouvry, A.M. Marechal, A practical methodology to select
fretting palliatives: Application to shot peening, hard chromium and WC-
Co coatings, Wear 259 (2005): pp 367–376.
[136] M.C Gasper, A. Ramalho, A. Cavaleiro. Effects of the counter face material
on the fretting behavior of spotted W-Si-N coatings ,Wear 255(2003): pp
276-286.
[137] D. Nowell, D. Dini, Stress gradient effects in fretting fatigue, Tribology
International 36 (2003): pp 71–78.
[138] R.B. Waterhouse, A. J. Trowsdale, Residual stress and surface roughness
in fretting fatigue, J. Phys. D: Appl. Phys. 25 (1992): pp A236-A239.
[139] A.E. Giannakopoulos, T.C. Lindley, S. Suresh, Aspects of Equivalence
between Contact Mechanics and Fracture Mechanics: Theoretical
Connections and a Life-Prediction Methodology for Fretting-Fatigue, Acta
mater 46 (1998): pp 2955-2968.
105
[140] A. Volchok, G. Halperin, I. Etsion, The effect of surface regular
microtopography on fretting fatigue life, Wear 253 (2002): pp 509-515.
[141] S. Fouvry, Ph. Kapsa, L. Vincent, Analysis of sliding behaviour for fretting
loadings: determination of transition criteria, Wear 185 (1995): pp 3546.
[142] P.R. Arora, M.S.D. Jacob, M.S. Salit, E.M. Ahmed, M. Saleem , P.Edi,
Experimental evaluation of fretting fatigue test apparatus,International
Journal of Fatigue 29 (2007): pp 941–952.
[143] I.G. Goryacheva, P.T. Rajeev, T.N. Farris, Wear in partial slip contact,
Journal of Tribology 123 (2001): pp 848–856.
[144] H. Lee, S. Mall, Some observations on frictional force during fretting
fatigue, Tribology Letters17, (2004): pp 491-499.
[145] J.A. Araújo, D. Nowell, Analysis of pad size effects in fretting fatigue using
short crack arrest methodology, International Journal of Fatigue 21
(1999): pp 947-956.
[146] C. Navarro, S. Muñoz, J. Domínguez, On the use of multiaxial fatigue
criteria for fretting fatigue life assessment, International Journal of Fatigue
30 (2008): pp 32–44.
[147] U. Bryggman, S. Soderberg, Contact conditions and surface degradation
mechanism in low amplitude fretting, Wear 125(1988): pp 39-52.
[148] B. Alfredsson, A.A. Cadario, Study of fretting friction evolution and
fretting fatigue crack initiation for a spherical contact. Int J Fatigue 26,
(2004): pp 1037.
[149] Y. Kondo and M. Bodai, The Fretting fatigue limit based on local stress at
;the contact edge Fatigue Fract Fract Engng Mater Struct 24, (2001): pp
791.
106
[150] D. Nowell, Dini. Stress gradient effects in Fretting fatigue. Tribology Int.
36, (2003): pp 71.
[151] M. Terheci, Microscopic investigation on the origin of wear by surface
fatigue in dry sliding. Materials Characterization 45, (2000): pp1.
[152] M.C. Dubourg and V. Lampacq, stage II crack propagation direction
determination under fretting fatigue loading: a new approach with
experimental observations, in : D.W. Hoeppner, V. Chandrasekaran, C.B.
Elliot (Eds), Fretting Fatigue: Current Technology and Practices, ASTM STP
1367, Am. Soc. Testing Mater., West Conshohocken, PA, USA, (2000):
pp.436.
[153] S. Fouvry, K. Elleuch and G. Simeon, Prediction of crack initiation under
partial slip fretting conditions. J Strain Annual; 6 (2002): pp 549.
[154] H. Ziegler A modification of Prager hardening rule, Q. App. Math., 19
(1959): pp.55.
[155] J. Ding, S.B Leen and I. R. McColl, The effect of slip regime on fretting
wear-induced stress evolution. Int J Fatigue, 26, (2004): pp 521.
[156] N. K Ramakrishna Naidu and Ganesh Sundara Raman, Effect of shot
blasting on plain fatigue and fretting fatigue behaviour of Al-Mg-Si alloy
AA 6061. Int J Fatigue 27, (2005): pp 663.
[157] S. Fouvry, P. Duo and Ph. Perruchaut A quantitative approach of Ti-6A1-
4V fretting damage: friction, wear and crack nucleation. Wear 257,
(2004): pp 916.
[158] S. Naboulsi & S. Mall L. Hutson, C. Neslen and T. Nicholas,
Characterization of fretting fatigue crack initiation processes in Ti-6A1-4V,
Tribology Int 36 (2003).
107
[159] D.F. Socie Critical plane approaches for multiaxial fatigue damage
assessment. In advances in multiaxial fatigue, D. L. Mc Dowell and R.
Ellis, editors, ASTM STP 1191, (1993): pp 7.
[160] D.L. McDowell, Basic issues in the mechanics of high cycle fatigue Int J
Fract; 80 (1996): pp. 103.
[161] A. Fatemi and D.F. Socie, A critical plane approach to multiaxial fatigue
damage including out of phase loading Fatigue Fract, Engng Mater Struct;
11 (1988): pp 149.
[162] Y. Liu and S. Mahadevan, Multiaxial high-cycle fatigue criterion and life
prediction for metals. Int J Fatigue 27 (2005): pp 790.
[163] K. Smith, P. Watson and P. Topper, A stress strain function for a fatigue of
metals J mater; 5: (1970): pp. 767.
[164] O. Jin and S. Mall, Effects of slip on fretting behavior: experiments and
analyses. Wear 256, (2004): pp 671.
[165] P.A. McVeigh, T.N. Farris, Finite element analysis of fretting stresses, J.
Tribol. 119, (1997): pp 797.
[166] S.A. Namjoshi, V.K Jain, S. Mall, Effects of shot-peening on fretting-
fatigue behavior of Ti-6A1-4V, J. Eng. Mater. Technol. 124 (2002): pp
222.
[167] S.A. Namjoshi, S. Mall, V.K Jain, O. Jin, Fretting fatigue crack initation
mechanism in TI-6A1-4V, Fatigue fract. Engg.Mater. 25, (2002): pp955.
[168] M. P Szolwinski, T. N. Farris, Mechanics of fretting fatigue crack formation,
Wear 198 (1996): pp 93.
108
[169] N.G Cormier, B.S. Smallwood, G.B. Sinclair, G. Meda, Aggressive
submodeling of stress concentration, Int. J. Numer. Meth. Eng. 46 (1999):
pp 889.
[170] M. P Szolwinski, C. Tieche, G. Harish, T. N Farris, T. Sakagami, M.
Heinstein, Mechanics of mite width fretting contacts, Proceedings of the
SEM Conference on Theoretical, Experimental and Computational
Mechanics, Cincinnati, (1999):pp 314.
[171] M. Tur, J. Fuenmayor, J.J. Rodenas, E. Giner, 3 D analysis of the influence
of specimen dimensions on fretting stresses, Finite Elem. Anal. Design 39,
(2003): pp 933.
[172] M.S Hartle, T.N Farris, J.F. Matlik, P.T Rajeev, A hybrid approach to 3D
edge of contact stress analysis using singular integral equations,
Proceedings of the sixth National Turbine Engine High Cycle Fatigue
Conference Jacksonville, F (2001).
[173] H.S. Kim S. Mall, Investigation into three dimensional effects of finite
contact width on fretting fatigue, (2005).
[174] C.D. Lykins, S. Mall, V. Jain, A shear stress-based parameter for fretting
fatigue crack initiation, Fatigue Fract Engng Mater Struct24 (2001): pp
461–473.
[175] C.D. Lykins, S. Mall, V.K. Jain, Combined experimental–numerical
investigation of fretting fatigue crack initiation, International Journal of
Fatigue 23 (2001): pp 703–711.
[176] A. Iwabuchi, J.W. Lee, M. Uchidate, Synergetic effect of fretting wear and
sliding wear of Co-alloy and Ti-alloy in Hanks´solution, Wear 263 (2007):
pp 492-500.
109
[177] D. Nowell, D. Dini, Stress gradient effects in fretting fatigue, Tribology
International 36 (2003): pp 71–78.
[178] A.J. Fenner, J.E. Field, A study of the onset of fatigue damage due to
fretting. Proceedings N. E Coast Institute of Engineers and Shipbuilders,
76, (1960): pp 183.
[179] D. Nowell, Ananalysis of fretting fatigue. Doctor of Philosophy Thesis,
Lincoln College, Oxford University (1988).
[180] D.A. Hills, D. Nowell, Mechanics of Fretting Fatigue. Dordrecht, the
Netherlands: Kluwer Academic Publishers (1994).
[181] C.T. Tsai, S. Mall, Elasto-plastic finite element analysis of fretting stresses in pre-
stressed strip in contact with cylindrical pad, Finite Elem. Anal. Design 36,
(2000): pp 171.
[182] Szolwinski M.P, Farris T. N (1998) Observation, analysis and prediction of fretting
fatigue in 2024-T351 Aluminum alloy. Wear 221, pp 24.
[183] K. Walker, The effect of stress ratio during crack propagation and fatigue for
2024-T3 and 7075-T6 Aluminum”, Effects of Environment and Complex Load
History on Fatigue Life. American Society for Testing and Materials, West
Conshohocken, PA. STO 462, (1970): pp 1.
[184] W.N. Findley, A theory for the effect of mean stress on fatigue of metals under
combined torsion and axial load or bending. Trans. ASME. J Eng Ind, B81,
(1959): pp 301-6.
[185] A.E. Giannakopoulous,, T.C. Lindley, S. Suresh, Aspects of equivalence between
contact mechanic and fracture mechanic theoretical connections and life
prediction methodology for fretting fatigue. Acta materialia 46. (1998): pp2955.
[186] M. Ciavarella, Crack like notch-analogue for a safe life fretting fatigue design
mythology, Fatigue fracture Engg.matter struct 26. (2003): pp1159.
110
[187] B. Atzori, P. Lazzarin, Int J Fract; 107, (2001): pp1.
[188] S. Naboulsi, Modified crack-analogy methodology: the equivalent theory. Int. J
Fatigue 27 (2005): pp 439.
[189] A.E. Giannakopoulos, T.C Lindley, Similarities of stress concentration in contact
at round punches and fatigue at notches: implications to fretting fatigue crack
initiation. Fatigue Fract Fract Engng Master Struct, 23, (2000): pp 561.
[190] K.L. Johnson, Contact mechanic .Cambridge: Cambridge University press
(1985).
[191] J.A. Arauj, D. Nowell, Analysis of pad size effects in fretting fatigue using
short crack arrests methodologies .Int.J Fatigue. 21,(1999): pp947.
[192] K.S. Chan, Y.D. Lee D.L. Davidson, Jr. Hudak SJ, A fracture mechanics approach
to high cycle fretting fatigue based on the worst case fret concept: I. Model
development. Int J Fract; 112, (2001): pp 299.
[193] K.S. Chan, D.Y. Lee, D.L. Davidson,Jr S,J Hudak, A fracture mechanics approach
to high cycle fretting fatigue based on the worst case fret concept: II.
Experimental evaluation. Int J Fract; 112, (2001): pp 331.
[194] H. Kitawgawa, S, Takahash, Proceeding of the 2nd International Conference on
Mechanical Behaviour of Materials. Philadelphia, American Society for Metals,
(1976): pp 627.
[195] Ruiz and D. Nowell, Designing against fretting fatigue in aero engines. In:
Fuentes M. Elices M, Martin-Meizoso A, Martinez-Esnaola JM, editors.
Fracture mechanics, applications and challenges, Invited papers presented
at the 13th European Conference on Fracture, ESIS publication 26.
Amsterdam: Elsevier, (2000): pp 73.
[196] Mall, S.A Namjoshi,, W.J Porter, Effects of microstructure on Fretting
fatigue crack initiation behaviour of Ti-6Al-4V, Materials Science and
Engineering A383, (2000): pp 334—340.
111
[197] Ren, S. Mall, J.H Sanders, S.K Sharma, Evaluation of coatings on TI-6Al-
4V substrate under Fretting fatigue, Surface & Coatings Technology, pag.
192 (2000): pp 177-188.
[198] P.R. Wittkowsky, Birch, J. Dominguez, S. Surech, An apparatus for
quantitative Fretting fatigue testing, Blackwell Science Ltd. Fatigue Fract
Engng Struct 22, pp 307-320, 1999.Anales de Mecánica de la Fractura
Vol. I (2006).
[199] Fenner and Field, A study of the onset of fatigue damage due to fretting.
Proceeding N.E, Cost institute of Engineers and ship builders. 76, pp, 183.
[200] Nishioka, Nishimura and Hirakawa, Fundamental investigation of fretting
fatigue fitted axle assemblies. On the relative slip amplitude of press fitted
Bulletin of JSME, 11, pp 437.
[201] Bramhall Studies in fretting fatigue. Doctor of philosophy Thesis, lincoln
College, Oxford University.
[202] Hills and Novel, Mechanics of fretting fatigue. Dordrecht, the Netherlands:
Kluwer Academic publishers.
[203] Farris, Szolwinski and Harish, Fretting in aerospace structures and
materials, Fretting fatigue: current Technology and practices , Hopner,
Chandrasekaran,and Elliott, American society for testing and materials,
Philadelphia,ASTM STP pp 1367.
[204] Papanikos, S.A Meguid, Theoretical and experimental studies of fretting-
initiated fatigue failure of aeroengine compressor discs, Fattigue Fract.
Eng Mater. STruct. 17 (1994): pp 539.
[205] Conner, T.C. Lindley, T. Nicholas and S. Suresh, Application of fracture
mechanics based life prediction method for contact fatigue. Int. J
Fatigue.26, (2004): pp 511.
112
[206] Buciumeanu ,Miranda and Silva, Development of Fretting Fatigue
Apparatus, Anales de Meccanica de la Fractura vol 1. Mechenical
Engineering Department :school of engineering ,Minho University,
Portugal. (2006):Pp 323.l
[207] Patrick and Golden Development of a dovetail fretting fatigue fixture for
turbine engine materials, international journal of fatigue 31 (2009): pp
620-628.
[208] B.P. Conner, T. Nicholas, ASME J. Engg. Mat. Tech,in press.
[209] P.J. Golden, T. Nicholas, Fatigue fract.Engg.M.28, (2005): pp 1169-75.
[210] Lee, Effect of dissimilar mating materials and contact force on fretting
fatigue behavior of Ti-6Al-4V, Tribology International 37 (2004):
pp 35-44.
[211] R.W. Swalla, D.L. Neu, McDowell, Microstructural Characterisation of
Ti-Al6-4V Subjected to Fretting, Journal of Tribology 126 (2004):
pp 809-816.
[212] Pauw, P.De Baets and W.De Waele Review and classification of fretting
fatigue test rigs, Dhent University , Laboratery Soet, Belgium, Sustainable
construction and design (2011): pp 41-52.
[213] Wagle and H. Kato, Ultrasonic wave intensity reflected from fretting
fatigue cracks at bolt joints of aluminum alloy plates, Ndt & E
International, vol. 42, Dec (2009): pp 690-695.
[214] Houghton, P. M. Wavish, E. J. Williams, and S. B. Leen, Multiaxial fretting
fatigue testing and prediction for splined couplings, International Journal
of Fatigue, vol. 31, , Nov-Dec (2009): pp. 1805-1815.
113
[215] Hyde, S. B. Leen, and I. R. McColl, A simplified fretting test methodology
for complex shaft couplings, Fatigue & Fracture of Engineering Materials &
Structures, vol. 28, Nov (2005): pp 1047-1067.
[216] Leen, T. H. Hyde, C. H. H. Ratsimba, E. J. Williams, and I. R. McColl, An
investigation of the fatigue and fretting performance of a representative
aero-engine spline coupling, Journal of Strain Analysis for Engineering
Design, vol. 37, Nov (2002): pp 565-583.
[217] Bertini, M. Beghini, C. Santus, and A. Baryshnikov, Resonant test rigs for
fatigue full scale testing of oil drill string connections, International Journal
of Fatigue, vol. 30, (2008): pp 978-988.
[218] Alfredsson, Fretting fatigue of a shrink-fit pin subjected to rotating
bending: Experiments and simulations, International Journal of Fatigue,
vol. 31, (2009): pp 1559-1570.
[219] Gutkin and B. Alfredsson, Growth of fretting fatigue cracks in a shrink-
fitted joint subjected to rotating bending, Engineering Failure Analysis, vol.
15, (2008): pp 582-596.
[220] Juuma, Torsional fretting fatigue strength of a shrink-fitted shaft with a
grooved hub, Tribology.
[221] R.A. Aggarwal, Khan, and V. P. Agrawal, Effect of surface roughness on
the fretting fatigue behaviour of EN45A spring steel, in Proceedings of the
Institution of Mechanical Engineers -- Part B-- Engineering Manufacture.
vol. 220: Professional Engineering Publishing, (2006): pp 1325-1331.
[222] Golden and T. Nicholas, The effect of angle on dovetail fretting
experiments in Ti-6Al-4V, Fatigue & Fracture of Engineering Materials &
Structures, vol. 28, Dec (2005): pp 1169-1175.
114
[223] Golden and J. R. Calcaterra, A fracture mechanics life prediction
methodology applied to dovetail fretting, Tribology International, vol. 39,
pp. 1172-1180, Oct 2006.nternational, vol. 33, (2000): pp 537-543.
[224] Mall, S.A. Namjoshi, W.J. Porter, Effects of microstructure on fretting
fatigue crack initiation behaviour of Ti-6Al-4V, Materials Science and
Engineering A383 (2004): pp 334—340.
[225] Dicka, Paulinb, Cailletauda, Fouvry. Experimental and numerical analysis
of local and global plastic behaviour in fretting wear LTDS (CNRS UMR
5513), Ecole Centrale de Lyon, 36 Avenue Guy de Collongue. Tribology
International Elsevier Ltd. (2006): pp 1-9.
[226] Hammouda et al Modelling of cracking sites/development in axial Dovetail
Joints of Aero–engine Compressor Disc, International Journal of fatigue ,
Vol 29,(2007): pp 30-48.
[227] Hammouda, El-Batanony and H.E.M. Sallam Finite element simulation of
the mechanics of flat contact pad fretting fatigue tests. Fatigue Fract Fract
Engng Mater Struct, 26, (2003): pp 1.
[228] Hammouda, Fayed and Sallam Simulation of mixed mode I/II cyclic
deformation at the tip of a short kinked inclined crack with frictional
surfaces, Int J Fatigue, 25, (2003): pp 743.
[229] Hammouda, Fayed and Sallam, Stress intensity factors of ;shortly kinked
slant central crack with frictional surfaces in uniaxially loaded plates, Int J
Fatigue, 25, (2003); pp 283.
[230] Hammouda, Fayed, and Sallam, Stress intensity factors of a central slant
crack with frictional surfaces in plates with biaxial loading, Int J Fract, 129
(2004): pp 141.
115
[231] Hammouda, Sallam and Fayed, Mode II stress intensity factors
numerically computed for central slant cracks with rough surfaces in
uniaxially compressed plates. Int J Fatigue, 12, (2002): pp.1213.
[232] Hammouda and El-Sehily A two-dimensional elastic-plastic finite element
analysis of friction effects on sliding crack surfaces in full or partial
contact. Fatigue Fract Fract Engng Mater Struct, 22 (1998): pp.101
[233] Papanikos,l Three Dimentional Nonlinear Finite element analysis of
Dovetail joint in aero engine disc Finite element in analysis and Designe,
Vol 29,(1998): pp173-186.
[234] Lucjan, Failure analysis of Turbine Disc of an aero engine, Engineering
failure analysisVol13, (2006): pp9-17.
[235] Beisheim and Sincliar, Three Dimentional finite Element Analysis of
Dovetail Attachment with and without crowing, Journal of
Turbomachionery, Vol 130,(2008): pp1-8.
116
Appendix-A
In Appendix A provides a detail of mechanical as well as electrical and electronic parts being used in newly developed fretting fatigue test rig .
117
A-1 Manufacturing detail of mechanical components
Components required for the manufacturing experimental test rig is given below.
A-1.1: Fixture of experimental test rig along with specimen
The assembly fixture consists of twenty one components including all the sub
assemblies and individual parts. Individual components have been shown in
Figure A-1.1. Detail of the same has already been given in Figure 4.11(a ~ p).
Figure A-1.1: Fixture of experimental test rig along with specimen
A-1.2: Upper holding plat, disc, blades and lower holding plate sub
assembly
This subassembly consists of upper holding plate, disc, blades and lower holding
plate. Upper and lower holding plate is helping to hold the specimen ie disc and
blades on the spindle of the main motor which is required to revolve at very high
speed ie at 20000 rpm. It is shown in Figure A-1.2. Detail of the same has
already been given in Figure 4.11 (d ~ g).
Figure A-1.2: Upper holding plat, disc, blades and lower holding plate sub assembly
118
A-1.3: Main motor and motor holding subassembly
This subassembly consists on main motor and main motor holding plat. Assembly
is carried out according to the assembly drawing as shown in Figure A-1.3 with
detail in Figure 4.11 (a ~ b).
Figure A-1.3: Main motor and motor holding sub assembly
A-1.4: Main motor, motor holding plate and safety guard sub assembly
This subassembly consists of main motor, main motor holding plat, safety guard
and pin hinges. It is shown in Figure A-1.4. Detail of the same has already been
given in Figure 4.11 (a ~ c).
Figure A-1.4: Main motor, motor holding plate and safety guard sub assembly
A
a
T
su
F
A
T
ti
w
A-1.5: Mai
assembly a
This subasse
ubassembly
igure 4.11
Figure
A-1.6: Stan
The subasse
ghtening k
with detail in
Fig
in motor,
and safety
embly cons
y and safe
(a ~ g).
A-1.5: Ma
nd for spec
embly is ca
knobs , grip
n Figure 4.1
gure A-1.6
motor ho
y cover pla
sists of main
ety cover p
ain motor, msubassemb
cimen hol
arried out f
ping piece a
11 (i~ p).
6: Stand for
119
olding pla
ate
n motor an
plated as s
motor holdibly and safe
lding fixtu
for the pur
and Ellen h
r specimen
ate, safety
nd main mo
shown in fi
ng plate, saety cover pl
ure sub ass
pose consi
head screw
holding fixt
y guard ,s
otor holding
igure A-1.5
afety guardlate
sembly
sts of pilla
as shown
ture subass
specimen
g plat, spec
5 with deta
d ,specimen
ars, base p
in Figure A
sembly
sub
imen
ail in
n
plate,
A-1.6
120
A-1.7: Main motor holding plate
The plate is manufactured by machining, milling, drilling operation. To fix the
main motor a disc was required as shown in Figure A-1.7 with detail in Figure
4.11 (b).
Figure A-1.7: Main motor holding plate
A-1.8: Safety guard
Safety guard resists the reaction of specimen i.e. disc and blades in case of
breakage fixed on spindle of main motor which is moving at very high speed i.e.
20000 rpm. This guard is manufactured from mild steel material as shown in
Figure A-1.8 with detail in Figure 4.11(c).
Figure A-1.8: Safety guard
121
A-1.9: Lower holding plate
This plate holds the specimen assembly so it should be minimum in weight to
avoid unnecessary load on the motor. It is manufactured from aluminum
material as shown in Figure A-1.9 whose detail is given in Figure 4.11(d).
Figure A-1.9: Lower holding plate
A-1.10: Upper holding plate
The plate is manufactured from aluminum material in such a way that it should
retain minimum weight as its weight ultimately affects the main motor. It was
manufactured from aluminum material as shown in Figure A-1.10 whose detail is
given in Figure 4.11(g).
Figure A-1.10: Upper holding plate
122
A-1.11: Safety Cover Plate
Safety cover plate is required that may resist the reaction of specimen in case of
its breakage fixed on spindle of main motor and move at very high speed. This
plate is manufactured from mild steel material as shown in Figure A-1.11 with
detail in Figure 4.11(h).
Figure A-1.11: Safety cover plate
A-1.12: Tightening Knob
The tightening knob is manufactured from mild steel material and two tightening
knobs have been used to fix the safety guard at required angle as shown in
Figure A-1.12 with detail in Figure 4.11 (i).
Figure A-1.12: Tightening knob
123
A-1.13: Griping Piece
The is manufactured from mild steel material and two gripping piece have bee
have been used to fix the safety guard at required angle as shown in Figure A-
1.13 with detail in Figure 4.11(j).
Figure A-1.13: Gripping piece
A-1.14: Hinge pins
Two hinge pins are manufactured from mild steel material to fix between safety
guard and stand for adjusting the angle of specimen holding fixture sub
assembly if required as shown in Figure A-1.14 with detail in Figure 4.11(l).
Figure A-1.14: Hinge pins
124
A-1.15: Pillars
The pillar is manufactured from mild steel material and two Pillars have been
used to fix the base plate and further more it is required to mount the main
motor holding plate and safety guard sub assembly as shown in Figure A-1.15
with detail in Figure 4.11(m).
Figure A-1.15: Pillars
A-1.16: Base plate
The base plate is manufactured from mild steel material one base plate used to
fix the two pillars to mount on the main ratchet sub assembly as shown in Figure
A-1.16 with detail in Figure 4.11(n).
Figure A-1.16: Base plate
125
A-1.17: Ratchet sub assembly
This slander subassembly consists of upper and lower piece, bearing, upper
holding threaded pin and lower holding threaded pin. It is required to rotate the
upper portion of the specimen holding fixture at desired angle as shown in Figure
A-1.17 with detail in Figure 4.11(o).
Figure A-1.17: Ratchet sub assembly
A-1.18: Base
The base is manufactured from mild steel material one base used to fix two
pillars and mount on the main ratchet sub assembly as shown in Figure A-1.18
with detail in Figure 4.11(p).
Figure A-1.18: Base
126
A-2: Detail of various electrical parts
Following electrical parts required for this purpose has been chosen. Detail is
appended below:
A-2.1: Main motor drive
The motor drive rotates the disc under test up to 25,000 rpm. The drive shall
include a speed control circuit to vary the motor speed linearly or in steps. A
variety of motor drives working on different Principles are available in market
having speed of 25,000 rpm.
a) DC drive with DC drive circuit
b) AC Open motor
c) AC motor with (Frequency) Inverter control
d) Universal motor with (Frequency) inverter control or voltage control
Universal motor with voltage control is chosen for the development of apparatus
to easily handle and maintenance is easy and spare parts are available in local
market. This adopted system can be replaced with all other available systems
easily as shown in Figure 4.12.
A-2.2: Variable auto transformer
The transformer supplies controlled variable voltage to motor drive. One winding
part of auto transformer is used for primary winding whereas the other is used
for secondary winding. By controlling the turn ratio between the two by means of
a moving wiper the output voltage is controlled. Following options are available:
a) Bar type auto transformer
b) Rotary type auto transformer
127
Rotary type auto transformer is selected as per ratings coincident with the
selected motor It is easy to handle as compare to other transformers and its
maintenance is easy and spare parts are available in local market. The adopted
system can replace with all other available systems easily as shown in Figure
4.12.
A-2.3: Control motor
To control the position of auto transformer to have subsequent adjustment of
output voltage of auto transformer. Following options are available;
a) AC motor
b) Stepper motor
c) DC motor
DC motor is selected along with the gear box for torque improvement and
reduction of speed of the motor to have subsequent smooth control of output
voltage of Transformer. It is easy to handle as compare to other motors and
maintenance is easy and spare parts are available in local market. The adopted
system can replace with all other available systems easily as shown in Figure
4.12.
A-2.4: Timer (main on time control)
Main on time control timer is required to control on time of the power circuit
after the initialization of the sequence. A variety of timer circuits are available.
a) Basic RC timer with output transistor
b) IC timer / 555 timer circuit
c) Adjustable / Programmable timer module
128
Adjustable multifunction timer module is selected. It is easy to handle as
compare to other time module and maintenance is easy and spare parts are
available in the local market. The adopted system can replace with all other
available systems easily as shown in Figure 4.12.
A-2.5: Timer (on time control)
On time control of the DC motor. Output voltage of the variable transformer
varies to change the speed of the motor drive as shown in Figure 4.12.
A-2.6: Timer (off time control)
Off time control timer controls off time of the D.C motor. Output voltage of the
variable transformer remains constant to sustain the speed of the motor drive as
shown in Figure 4.12.
A-2.7: Limit switches (S.2 and S.3)
The limit switches are used for activation and de-activation of relay (1). The
polarity of supply of D.C motor is adjusted to adjust its direction. It is shown in
figure 4.12.
A-2.8: Relay (1)
The relay is used to latch the electrical supply of the D.C motor to positive
polarity or negative polarity by means of limit switches S.2 and S.3 as shown in
Figure 4.12.
A-2.9: Relay (2)
The relay is used to give electrical supply to the D.C motor to positive polarity or
negative polarity depending on activation / de-activation of relay (1) as shown in
Figure 4.12.
129
A-2.10: Tachometer
Tachometer is used to measure speed of the motor drive as shown in Figure
4.12.
By using the above mentioned components/parts circuit diagram is prepared.
This circuit diagram is tested before to carry out actual test on the newly
designed, developed and manufactured experimental test rig. To revolve the
specimen in controlled manner different electrical and electronic components are
used to meet the actual simulation.
130
Appendix-B
In Appendix B provides a detail of validation of fretting fatigue tes rig with and without specimen.
131
B-1: Running trend of Experimental test rig – without Specimen
To evaluate the performance of experimental test rig with and without specimen,
same was revolved from 0~20000 rpm with the help of three types of timers i.e.
to control “start up time”, “stay time” and “apparatus setup time” which has
already been explained above under heading working of testing system.
Experimental test rig working is ensured and speed at all the steps is recorded
using tachometer. As the speed of the rig is very high and safety first has been
ensured and various tests were carried out step by step. Running trend of
experimental test rig without specimen is given in table B-1.1 to B-1.6 whereas
trend depicted graphically in figures B-1.1 to B-1.6 respectively.
Table B-1.1: Running trend of experimental test rig without specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 2500 4000 6000 7000 8500 9800 11200 11200 14000
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 15000 16000 17000 18000 19200 20200 21700 22600 23400 24300
S# 21 22 23
Step Time (sec.)
0.5 0.5 0.5
Speed (RPM 25300 26300 27000
132
Figure B-1.1: Running trend of experimental test rig without specimen
Table B-1.2: Running trend of experimental test rig without specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 0 2800 5500 7800 9200 10000 11400 12500 14800
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 15500 16800 17800 18400 20200 14860 11300 9900 8300 6800
S# 21 22 23 24 25 26
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 4000 3600 0 0 1348 4110
133
Speed rpm Vs Time Sec
Time Sec
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Spee
d rp
m
0
5000
10000
15000
20000
25000
Figure B-1.2: Running trend of experimental test rig without specimen
Table B-1.3: Running trend of experimental test rig without specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 0 280 2500 5300 7800 9700 11200 12500 14100
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 15400 16800 17800 18000 19800 20800 20800 19355 18000 16596
S# 21 22 23 24 25 26
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 15300 13900 11900 10000 8000 6000 5000 3000 1600 0
134
Figure B-1.3: Running trend of experimental test rig without specimen
Table B-1.4: Running trend of experimental test rig without specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 1 23 2700 4000 6800 9300 11300 13100 14800
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 16400 18100 19700 21000 22500 21200 20000 18000 16500 14900
S# 21 22 23 24 25 26 27 28 29 30
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 13600 12000 10400 8300 6000 2800 2500 100 1 0
135
Figure B-1.4: Running trend of experimental test rig without specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 1 1400 3900 6000 8400 10600 11900 13300 15200
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 16700 18000 19100 20500 21800 21500 21500 21000 19300 17900
S# 21 22 23 24 25 26 27 28 29
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 16900 15800 13800 12400 10000 8300 5500 2000 0
Table B-1.5: Running trend of experimental test rig without specimen
136
Figure B-1.5: Running trend of experimental test rig without specimen
Table B-1.6: Running trend of experimental test rig without specimen
S# 1 2 3 4 5 6 7 8 9 10 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 83 1000 3400 6400 8800 11200 12900 14400 15600
S# 11 12 13 14 15 16 17 18 19 20 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 17000 18600 19900 21300 22000 22000 22000 21600 21600 20300
S# 21 22 23 24 25 26 27 28 29 30 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 19000 17600 16000 14400 13000 12000 11000 10000 9000 6000
S# 31 32 33 34 35 36 37 38 39 40 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 3300 0 3600 6000 8300 10600 12500 14000 15200 16900
S# 41 42 43 44 45 46 47 48 49 50 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 18400 19600 20800 22000 22000 22000 21700 20700 19400 18000
S# 51 52 53 54 55 56 57 58 59 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 16400 14900 13200 11900 9900 4200 3800 2000 0
137
Figure B-1.6: Running trend of experimental test rig without specimen
B-2: Running trend of experimental test rig – with Specimen
After completion of the above mentioned test i.e. revolving the test rig without
specimen as shown above, the testing was carried out with specimen made out
of mild steel is utilized. Manufacturing of disc and blades and assembly was
carried out accordingly which was fixed on the spindle of the main motor just to
run the test rig to check whether it bears the weight of the specimen or
otherwise. The experimental test rig was started from 0 to 20,000 rpm with the
help of three types of timers as already mentioned above. Experimental test rig
functioned /operated successfully for certain time and speed at all the steps was
checked through tachometer. Data collected tabulated in table B-2.1 to B-2.4
whereas trend depicted graphically in figures B-2.1 to B-2.4 respectively.
138
Table B-2.1: Running trend of experimental test rig with specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 2070 2800 5000 6300 7200 8200 9200 10000 10600
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 11100 11800 12400 13000 13500 13900 14500 14900 15380 15800
S# 21 22 23 24 25 26 27 28 29 30
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 16800 16955 17000 17000 17400 18187 18900 19200 19400 19700
139
Figure B-2.1: Running trend of experimental test rig with specimen
Table B-2.2: Running trend of experimental test rig with specimen
S# 1 2 3 4 5 6 7 8 9 10
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 0 1300 3200 4700 5700 6600 7400 8100 9100
S# 11 12 13 14 15 16 17 18 19 20
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 9900 10500 11500 12000 12700 13700 13800 14400 14600 15000
S# 21 22 23 24 25 26
Step Time (sec.)
0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 16200 17600 18900 19900 20900 21800
140
Figure B-2.2: Running trend of experimental test rig with specimen
Table B-2.3: Running trend of experimental test rig with specimen
S# 1 2 3 4 5 6 7 8 9 10 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 2600 4500 5800 6900 7600 9000 9900 11500 12200
S# 11 12 13 14 15 16 17 18 19 20 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 13000 13700 14000 14800 14800 14800 15300 15800 16400 16400
S# 21 22 23 24 25 26 27 28 29 30 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 17000 17400 16900 16400 15900 16000 15300 14700 13900 13400
S# 31 32 33 34 35 36 37 38 39 40 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 13000 12400 12000 11800 11200 10900 10300 9800 9200 8500
S# 41 42 43 44 45 46 47 48 49 50 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 7500 6500 6200 5300 4800 3700 2600 1500 0
141
Speed rpm vs Time sec
Time sec
0 5 10 15 20 25 30 35 40 45 50 55
Spee
d rp
m
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
Figure B-2.3: Running trend of experimental test rig with specimen
Table B-2.4: Running trend of experimental test rig with specimen
S# 1 2 3 4 5 6 7 8 9 10 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 0 500 2000 4200 5600 6600 7600 8400 9300 10000
S# 11 12 13 14 15 16 17 18 19 20 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 10800 11300 12100 12500 13500 14200 14700 15500 15700 16400
S# 21 22 23 24 25 26 27 28 29 30 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 17000 18000 18000 17000 16400 16000 16000 15800 14400 14400
S# 31 32 33 34 35 36 37 38 39 40 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 13800 13500 12800 12000 11200 10300 9400 8400 7400 6200
S# 41 42 43 44 45 46 47 48 49 50 Step Time
(sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 5200 4000 2200 400 0 2600 4500 5800 6900 7600
142
S# 51 52 53 54 55 56 57 58 59 60
Step Time (sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 9000 9900 11500 12200 13000 13700 14000 14800 14800 14800
S# 61 62 63 64 65 66 67 68 69 70
Step Time (sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 15300 15800 16400 16400 17000 17400 16900 16400 15900 16000
S# 71 72 73 74 75 76 77 78 79 80
Step Time (sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 15300 14700 13900 13400 13000 12400 12000 11800 11200 10900
S# 81 82 83 84 85 86 87 88 89 90
Step Time (sec.) 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5
Speed (RPM 10300 9800 9200 8500 7500 6500 6200 5300 4800 3700
S# 91 92 93
Step Time (sec.) 0.5 0.5 0.5
Speed (RPM 2600 1500 0
Figure B-2.4: Running trend of experimental test rig with specimen