teaming and \T Services |

j Collegiate Learning Centre I *Z O

\ Collegiate Crescent UasTipyS |
j Sheffield S102BP I
w,.:mu&ak-Mwif.

101 923 069 X

W
IE
EV
PR

REFERENCE
 ProQuest N um ber: 10701264

All rights reserved

INFORMATION TO ALL USERS

The quality of this reproduction is d e p e n d e n t upon the quality of the copy subm itted.

In the unlikely e v e n t that the a u thor did not send a c o m p le te m anuscript

and there are missing pages, these will be noted. Also, if m aterial had to be rem oved,
a n o te will ind ica te the deletion.

uest
W
IE
EV
ProQuest 10701264

Published by ProQuest LLC(2017). C opyright of the Dissertation is held by the Author.

All rights reserved.

PR

This work is protected against unauthorized copying under Title 17, United States C o d e
M icroform Edition © ProQuest LLC.

ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 4 8 1 0 6 - 1346
Finite Element Model of a Tennis Ball Impact with a Racket

Thomas Bruce Allen

W
IE
A thesis submitted in partial fulfilment of the requirements of
EV
Sheffield Hallam University
for the degree of Doctor of Philosophy
PR

April 2009

Collaborating Organisation: Prince

Abstract
Previous authors have produced analytical models which accurately simulate
tennis impacts. However, currently there are few published studies on the
simulation of tennis impacts using finite-element (FE) technique. The purpose of
this study was to produce accurate FE models of tennis impacts, which will
serve as design tools as well as aid in furthering the understanding of how the
ball, string-bed and racket behave during play.
An FE model of a pressurised tennis ball was produced in Ansys/LS-DYNA 10.0
and validated against experimental data. The ball model was updated to
simulate the extreme playing temperatures of 10 and 40°C and validated
against experimental data, obtained inside a climate chamber. Following
validation of the ball model, an FE model of a head-clamped racket was
produced and validated against experimental data. The validation included a
range of inbound velocities, angles and spin rates, for impacts at a number of

W
nominal locations on the string-bed. Finally, an FE model of a freely suspended
racket was constructed and validated against experimental data. Impacts were
IE
simulated at a number of nominal impact locations on the string-bed, with a
range of ball inbound velocities, angles and spin rates. The impacts were
EV
recorded using two Phantom v4.2 high-speed cameras and analysed in 3D. The
FE models were all in good agreement with the experimental data, for the
individual stages of the validation.
A parametric modelling program was produced to be used in conjunction with
PR

the model. This program enables the user to adjust a variety of parameters,
such as the inbound velocity of the ball, impact location and mass of the racket,
and run simulations without any specialist knowledge of the FE model. This
program was used to analyse the model against ball to racket impact data
obtained during player testing. There was relatively good agreement between
the model and player testing data.
Finally, the model was used to determine the influence of racket structural
stiffness, mass and the position of the balance point, when performing a typical
topspin forehand. It was found that using a head-heavy racket, with high
structural stiffness and mass, will increase the rebound velocity and topspin of
the ball, for a shot of this type at the centre of the string-bed.
Keywords: tennis ball, tennis racket, high speed cinem atography, finite-elem ent m odelling.
Acknowledgements

I would like to thank Dr Simon Goodwill and Professor Steve Haake for their
continual support, guidance and enthusiasm throughout the study. This thanks
extends to all other members of the Sports Engineering Research Group at
Sheffield Hallam University, in particular; Amanda Brothwell and Carole Harris
for providing administrative support, Terry Senior for providing technical
support, John Kelley for continual assistance and Simon Choppin for assisting
with 3D validation techniques and providing player testing data.
I am also grateful to P rince for their sponsorship of the project. The expertise
bought to the project by all of the members of the Prince engineering team, in
particular Mauro Pezzato, has been invaluable.
I am thankful to the International Tennis Federation (ITF) for allowing the use of
their impressive testing facilities.

W
Finally, I would like to thank Robyn Grant for providing continual support
throughout the project and proof reading the final document.
IE
EV
PR
 Contents

ABSTRACT II

ACKNOWLEDGEMENTS III

CONTENTS IV

LIST OF FIGURES V II

LIST OF TABLES XIX

NOMENCLATURE XXI

W
1. INTRODUCTION 1
1.1. Motivation for the Research 1

1.2.

1.3.
Aim and objectives

Thesis structure
IE 2

2
EV
2. LITERATURE REVIEW 4
2.1. Introduction 4

2.2. The ball 5

2.3. The string-bed 14

PR

2.4. The racket 24

2.5. Player testing 34

2.6. Modelling 38

2.7. The influence of technologicaladvances on tennis 49

2.8. Overview of Ansys/LS-DYNA 52

2.9. Discussion 54

2.10. Chapter summary 58

3. TENNIS BALL MODEL 59

3.1. Introduction 59

3.2. Pressurised tennis ball properties 60

3.3. Finite element model of a pressurised tennis ball 64

3.4. Validation of the pressurised tennis ball model 69

IV
 Validation of the tennis ball model for different temperatures 80

Chapter summary 89

Practical applications 89

HEAD-CLAMPED RACKET MODEL 90

Introduction 90

String properties 91

Finite element model of a tennis racket string-bed 94

Validation of the string-bed model 97

Head-clamped racket model 111

Validation of the head-clamped racket model 115

Chapter summary 128

FREELY SUSPENDED RACKET MODEL 129

Introduction 129

W
FE Model of a freely suspended tennis racket 130

Validation of the freely suspended racket model 136

Results and discussion of the freely suspended racket model validation 147

Chapter summary

Practical applications
IE 157

158
EV
PARAMETRIC MODELLING PROGRAM 159
Introduction 159

Description of the parametric modelling program 159

PR

Discussion 166

Chapter summary 167

Practical applications 167

COMPARISON OF THE FE MODEL W ITH SIMULATED PLAY 168

Introduction 168

Method 168

Results 176

Discussion of player testing 178

Summary of player testing analysis 180

Chapter summary 181

APPLICATIONS OF THE MODEL 182

Introduction 182

V
8.2. Method 182

8.3. Results 186

8.4. Explanation of results 202

8.5. Chapter summary 222

9. CONCLUSIONS 223
9.1. Introduction 223

9.2. Summary of research 223

9.3. Conclusions 226

9.4. Future research 227

REFERENCES 231

PERSONAL BIBLIOGRAPHY 241

W
A BALL MODEL VALIDATION 242
A .I. Ball model mesh convergence study 242

A.2. Impact rig validation IE 244

A.3. Frequency analysis 244

EV
B HEAD-CLAMPED RACKET MODEL 247
B .l. Calculating the impact position on the string-bed 247

B.2. Effect of inbound spin 253

B.3. Difference between string-bed and head-clamped racket model 254

PR

C ALTERNATIVE SPIN CALCULATION 258

D MATERIAL TESTING OF TENNIS STRINGS 259

VI
 List of figures
Figure 2.1 Variation in a) COR and b) contact time, between different balls for
perpendicular impacts on a rigid surface (Haake et al., 2003a).............................7
Figure 2.2 Quasistatic material properties of the a) rubber core and b) felt cover
of a tennis ball (Goodwill et al., 2005)....................................................................... 8
Figure 2.3 Ball impact properties for a perpendicular impact on a rigid surface a)
Force plot and b) COM displacement and maximum deformation (Goodwill,
2002 ) 11
Figure 2.4 a) Dynamic string tester and b) Dynamic stiffness and contact
duration results for a selection of strings (Cross et al. 2000)............................... 18
Figure 2.5 Typical material curves which are used for obtaining the dynamic
stiffness of different strings (Jenkins, 2003)........................................................... 19
Figure 2.6 Analysis of an impact of a ball on a string-bed with an inbound
velocity, angle to the racket plane and backspin of 3.27 m s '1, 58.5° and 34.9

W
rad s '1, respectively (Cross, 2003)...........................................................................22
Figure 2.7 Horizontal and vertical coefficient of restitution of balls incident at 39°
IE
on a head-clamped racket (Goodwill and Haake, 2004a).....................................23
Figure 2.8 A selection of tennis rackets a) 1981 Dunlop Maxply, b) 1977 Prince
EV
oversize and c) 1980 Dunlop Max 200G.................................................................25
Figure 2.9 Racket properties from 1870 to 2007 a) frequency and b) mass
(Haake et al., 2007)................................................................................................... 26
Figure 2.10 Typical lay-up for a composite tennis racket (Jenkin, 2003)............28
PR

Figure 2.11 Racket frequency response A & B) low frequency handle clamped
and c) freely suspended (Brody, 1987)................................................................... 29
Figure 2.12 Vibration modes of a free and hand held tennis racket (Cross,
1998)............................................................................................................................31
Figure 2.13 Variation of ACOR with impact location for a perpendicular impact
between a tennis ball and freely suspended racket (Modified from Brody,
1997a)......................................................................................................................... 32
Figure 2.14 Shearing of the felt during an oblique impact on a rigid surface at 15
m s '1 and 30° with no initial spin (Goodwill et al., 2005)........................................41
Figure 2.15 Comparison of a flexible and rigid body model for simulating
impacts between a tennis ball and freely suspended racket (Goodwill and
Haake, 2003).............................................................................................................. 46
VII
Figure 3.1 Hemispherical rubber cores provided for material testing................. 60
Figure 3.2 (a) Tensile test piece (b) Extensometer (c) Positioning of
compression test piece.......................................................................................... 61
Figure 3.3 Example of cyclic loading of rubber taken from cores used in the
construction of tennis balls a) compression, b) tension.......................................62
Figure 3.4 Results of tennis ball rubber material testing a) 5th cycle from each
compression test, b) 5th cycle from each tensile test........................................... 63
Figure 3.5 Combined tension and compression results for the tennis ball rubber
material testing........................................................................................................63
Figure 3.6 Ball model with a quarter section removed.........................................65
Figure 3.7 Pressure-volume curve used to simulate the internal pressure of the
ball........................................................................................................................... 66
Figure 3.8 Pressure in reality and in the FE model.............................................. 66
Figure 3.9 State of maximum strain (red region) in the rubber core for a

W
perpendicular tennis ball impact at 30 m-s'1on a rigid surface........................... 67
Figure 3.10 Modified quasistatic stress-strain curve for the tennis ball rubber..68
IE
Figure 3.11 Estimated stress relaxation curve for tennis ball rubber..................68
Figure 3.12 Tennis ball impact rig a) Complete rig, b) Light gates and force plate
c) Air-cannon...........................................................................................................70
EV
Figure 3.13 a) Set up for impact rig validation and measuring ball/core
deformation and b) Method for calculating maximum ball/core deformation 71
Figure 3.14 Pressure volume curve for the punctured balls and cores.............. 72
PR

Figure 3.15 Rebound velocity against inbound velocity for perpendicular

impacts on a rigid surface a) Punctured core, b) Core, c) Punctured ball, d) Ball
(Experimental results obtained using light gates).................................................73
Figure 3.16 Deformation against inbound velocity for a perpendicular impact on
a rigid surface a) Punctured core, b) Core, c) Punctured ball, d) Ball
(Experimental results obtained a high speed video camera (Figure 3.13, page
71)............................................................................................................................75
Figure 3.17 Contact time against inbound velocity for perpendicular impacts on
a rigid surface a) Punctured core, b) Core, c) Punctured ball, d) Ball
(Experimental results obtained a force plate)....................................................... 76

VIII
Figure 3.18 Force plot of a 5 m-s'1 perpendicular impact for a a) Punctured core,
b) Core, c) Punctured ball, d) Ball (Experimental results obtained using a force
plate)........................................................................................................................77
Figure 3.19 Force plot of a 15 m-s"1 perpendicular impact for a a) Punctured
core, b) Core, c) Punctured ball and d) Ball..........................................................78
Figure 3.20 Force plot of a 25 m-s"1 perpendicular impact for a a) Core, b)
Punctured ball, c) Ball.............................................................................................79
Figure 3.21 Tennis ball internal pressure against relative volume, for
temperatures in the range from 283.15 to 313.15 K............................................ 81
Figure 3.22 a) Static rubber core material properties, a) Dynamic rubber core
material properties................................................................................................. 82
Figure 3.23 a) COR and b) contact time, comparison between the experimental
data from the original validation data (295.15 K) and Downing (2007a) (298.15
K) and the original FE model................................................................................. 83

W
Figure 3.24 COR for adjusted internal pressure at temperatures of a) 283.15 K
and b) 313.15 K...................................................................................................... 84
IE
Figure 3.25 Contact time for adjusted internal pressure at temperatures of a)
283.15 K and b) 313.15 K. Experimental data from Downing (2007a)............... 84
Figure 3.26 Effect of the quasistatic rubber material stiffness on a) COR and b)
EV
contact time. Experimental data from Downing (2007a)...................................... 85
Figure 3.27 COR results for the complete ball model updated to simulate
temperatures of a) 283.15 K and b) 313.15 K. Experimental data from Downing
PR

(2007a).................................................................................................................... 85
Figure 3.28 Contact time results for the complete ball model updated to
simulate temperatures of a) 283.15 K and b) 313.15 K. Experimental data from
Downing (2007a).................................................................................................... 86
Figure 3.29 Force plots for the complete ball models a) 15 m-s'1, b) 20 m-s'1, c)
25 m-s'1, b) 30 m-s'1................................................................................................ 86
Figure 4.1 Setup for tennis string materials testing using the Instron method...92
Figure 4.2 Instron force extension plots for tennis strings................................... 93
Figure 4.3 a) String-bed model and b) Close-up of string-bed model showing the
rigid cylinders on the ends of every string.............................................................94
Figure 4.4 Convergence of string-bed model....................................................... 96
Figure 4.5 Experimental setup for the head-clamped racket testing.................. 98
Figure 4.6 Camera position for recording the head-clamped racket impacts. ...98
Figure 4.7 Obtaining impact position on the string-bed for a nominal inbound
angle of 40°............................................................................................................. 99
Figure 4.8 Rebound velocity against inbound backspin for ball impacts on a
head-clamped racket at a) 20°, 20 m-s'1, b) 20°, 25 m-s'1, c) 40°, 20 m-s'1, d)
40°, 30 m-s'1, e) 60°, 20 m-s'1, f) 60°, 30 m-s'1................................................... 102
Figure 4.9 : Rebound angle against inbound spin for ball impacts on a head-
clamped racket at a) 20°, 20 m-s"1, b) 20°, 25 m-s'1, c) 40°, 20 m-s'1, d) 40°, 30
m-s'1, e) 60 °, 20 m-s'1, f) 60°, 30 m-s'1............................................................... 104
Figure 4.10 : Rebound spin against inbound spin for ball impacts on a head-
clamped racket at a) 20°, 20 m-s'1, b) 20°, 25 m-s'1, c) 40°, 20 m-s'1, d) 40°, 30
m-s'1, e) 60 °, 20 m-s'1, f) 60°, 30 m-s'1............................................................... 106
Figure 4.11 Results obtained from an FE model of a string-bed for an impact
with an inbound velocity of 30 m-s"1, angle of 40° and backspin of 200 m-s'1 a)

W
Vertical and horizontal force, b) Horizontal velocity and c) Spin (E = 7.2 GN-m2
obtained using the Hammer method).................................................................. 107
IE
Figure 4.12 Effect of ball to string friction in the string-bed model for an impact
with an inbound velocity of 30 m-s'1 and an angle of 40° a) velocity, b) angle and
c) spin (E = 7.2 GN-m2obtained using the Hammer method)..........................108
EV
Figure 4.13 Effect of ball to string coefficient of friction for an impact on the FE
model of a string-bed with an inbound velocity of 30 m-s'1, angle of 40° and
backspin of 200 rad-s"1 a) the horizontal force acting between the ball and
PR

string-bed and b) the spin of the ball throughout the impact............................. 109
Figure 4.14 Effect of string to string friction for an impact on the FE model of a
string-bed with an inbound velocity of 30 m-s'1 and angle 40° a) velocity, b)
angle and c) spin.................................................................................................. 110
Figure 4.15 Racket geometry used in the FE model..........................................112
Figure 4.16 Convergence of head-clamped racket model.................................113
Figure 4.17 a) Camera set up for the head-clamped racket model, b) Racket
markers used as a reference for obtaining the impact location on the string-bed.
116
Figure 4.18 Calculating the impact distance from a) the long axis of the string-
bed (view from camera 1) and b) the short axis of the string-bed (view from
camera 2).............................................................................................................. 117
Figure 4.19 Effect of inbound spin on the experimental data for the head-
clamped racket model a) inbound velocity, b) inbound angle and c) impact
distance from the long axis of the racket.............................................................. 118
Figure 4.20 Horizontal and vertical COR for oblique spinning impacts on a
head-clamped racket a) centre, b) off-centre, c) tip and d) throat..................... 120
Figure 4.21 Definition of horizontal COR.............................................................. 120
Figure 4.22 Rebound topspin for oblique spinning impacts on a head-clamped
racket a) centre, b) off-centre, c) tip and d) throat............................................... 121
Figure 4.23 Results for a centre impact on the head-clamped racket model at 21
m-s"1 and 38° with a-c) no spin, d-f) 200 rad-s'1 backspin, g-i) 400 rad-s'1
backspin, j-l) 600 rad-s'1 backspin......................................................................... 122
Figure 4.24 Impact positions on the string-bed.................................................... 123
Figure 4.25 Results obtained from the head-clamped racketmodel for impacts
with an inbound velocity of 20 m-s'1, an angle of 40° and backspin of 200 rad-s'1

W
at a range of locations on the string-bed a) velocity, b) angle, c) spin.............. 125
Figure 4.26 Effect of impact position on the deformation of the string-bed of a
IE
tennis racket............................................................................................................. 126
Figure 5.1 FE model racket geometry with three separate sections................. 131
Figure 5.2 Bifilar Suspension used to obtain the polar moment of inertia of a
EV
tennis racket............................................................................................................. 132
Figure 5.3 The relationship between apparent Young's modulus and natural
frequency for the racket in the FE model............................................................. 134
PR

Figure 5.4 Convergence of the freely suspended racket model.........................136

Figure 5.5 a) Impact rig used for simulating impacts on a freely suspended
tennis racket (Modified from Choppin, 2008) b) Optimum camera positions for
measuring the trajectory of a tennis ball in 3D (Modified from Choppin, 2008).
................................................................................................................................... 137
Figure 5.6 Impact positions on the string-bed for the validation of the freely
suspended racket model for perpendicular impacts............................................ 138
Figure 5.7 Racket positioning for perpendicular and oblique impacts on a freely
suspended racket (View from above).................................................................... 139
Figure 5.8 Racket position showing throat and side markers and axis coordinate
system....................................................................................................................... 140

XI
Figure 5.9 Method used for calculating the top/back spin of a tennis ball, by
calculating the change in 0 over time..................................................................141
Figure 5.10 Spin error in the experimental data as a result of camera positions.
............................................................................................................................... 142
Figure 5.11 Diagram showing the standard deviations in impact locations for the
perpendicular impacts on the freely suspended racket..................................... 143
Figure 5.12 Comparison of spin calculated from the left and rightcamera a)
inbound and b) rebound....................................................................................... 144
Figure 5.13 Ball rebound velocity for perpendicular impacts on a freely
suspended racket a) Centre, b) Off-centre, c) Tip and a) Throat......................148
Figure 5.14 Ball to racket resultant force plots for perpendicular impacts at 40
m-s'1, a) Centre, b) Off-centre, c) Tip and d) Throat.......................................... 150
Figure 5.15 Ball reboundvelocity for oblique impacts on a freely suspended
racket a) 18 m-s'1 and 24° b) 28 m-s'1 and 23°..................................................152

W
Figure 5.16 Ball rebound angle for oblique impacts on a freely suspended
racket a) 18 m-s'1 and 24° b) 28 m-s"1 and 23°..................................................153
IE
Figure 5.17 Ball rebound spin for oblique impacts on a freely suspended racket
a) 18 m-s'1 and 24° b) 28 m-s'1 and 23°.............................................................. 154
Figure 5.18 a) Force b) Horizontal velocity and c) Spin, throughout an impact at
EV
the centre of a freely suspended racket with an inbound velocity of 28 m-s'1,
angle of 23° and with 200 rad-s'1 of backspin (70 G P a/253 Hz)..................... 155
Figure 6.1 a) Start up screen for the parametric modelling program and b) The
PR

four impact types available in the parametric modelling program..................... 160

Figure 6.2 a) Ball parameters in the parametric modelling program and b) The
display in the parametric modelling program when creating multiple files 161
Figure 6.3 a) String-bed parameters in the parametric modelling program and b)
Impact orientation in the parametric modelling program....................................162
Figure 6.4 Racket parameters in the parametric modelling program................163
Figure 6.5 Results programme, showing results for a 40 m-s"1 perpendicular
impact at the GSC of a freely-suspended racket...............................................164
Figure 6.6 Force plot for a 40 m-s'1 perpendicular impact at the GSC of a freely
suspended racket a) Original force plot from ANSYS/LS-Dyna and b) Force plot
updated in the results program............................................................................ 165

XII
Figure 6.7 Ball diameter for a 40 m-s'1 perpendicular impact at the GSC of a
freely suspended racket a) Ball diameter from ANSYS/LS-Dyna and b) Ball
deformation calculated in the results program....................................................166
Figure 7.1 Comparison of the racket in the FE model and the ITF Carbon Fibre
racket with each of the 19 players' rackets, a) length, b) width, c) mass and d)
balance point.........................................................................................................171
Figure 7.2 Comparison of the racket in the FE model and the ITF test racket
with a selection of the players' rackets, a) length, b) width, c) mass and d)
balance point.........................................................................................................173
Figure 7.3 Impact positions a) player testing and b) FE simulations................ 174
Figure 7.4 Comparison of rebound velocity from the player testing and FE
model a) Horizontal, b) Vertical (perpendicular to string-bed) and c) Resultant
(Player data from Choppin etal. (2007a & b)).................................................... 177
Figure 7.5 Comparison of rebound angle from the player testing and FE model

W
(relative to racket normal) (Choppin et al., 2007a & b)...................................... 178
Figure 7.6 Comparison of rebound topspin from the player testing and FE model
IE
(Player data from Choppin etal., 2007a & b)..................................................... 178
Figure 8.1 Impact locations on the string-bed used to determine the effect of
different racket parameters.................................................................................. 183
EV
Figure 8.2 Relationship between the mass of the racket in the FE model and a)
its natural frequency, b) its moment of inertia.....................................................184
Figure 8.3 Relationship between the position of the balance point of the racket
PR

in the FE model and a) natural frequency, b) moment of inertia....................... 186

Figure 8.4 Effect of the structural stiffness of a tennis racket on the rebound
velocity of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1 of
backspin................................................................................................................ 187
Figure 8.5 Effect of the structural stiffness of a tennis racket on the longitudinal
rebound angle of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s"1of
backspin.................................................................................................................188
Figure 8.6 Effect of the structural stiffness of a tennis racket on the horizontal
rebound angle of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s"1of
backspin.................................................................................................................189

XIII
Figure 8.7 Effect of the structural stiffness of atennis racket on the rebound
sidespin of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1 of
backspin................................................................................................................ 190
Figure 8.8 Effect of the structural stiffness of atennis racket on the rebound
topspin of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1 of
backspin................................................................................................................ 191
Figure 8.9 Diagram to illustrate the difference between using a stiff and flexible
racket when performing a forehand shot............................................................192
Figure 8.10 Effect of the mass of a tennis racket on the rebound velocity of the
ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1of backspin..................193
Figure 8.11 Effect of the mass of a tennis racket on the longitudinal rebound
angle of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1of backspin.
............................................................................................................................... 194
Figure 8.12 Effect of the mass of a tennis racket on the horizontal rebound

W
angle of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1of backspin.
............................................................................................................................... 195
IE
Figure 8.13 Effect of the mass of a tennis racket on the rebound sidespin of the
ball, for an impact at 35 m-s'1 and 20° with 300 rad-s"1of backspin..................196
Figure 8.14 Effect of the mass of a tennis racket on the rebound topspin of the
EV
ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1of backspin..................197
Figure 8.15 Effect of the position of the balance point of a tennis racket on the
rebound velocity of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1of
PR

backspin................................................................................................................ 198
Figure 8.16 Effect of the position of the balance point of a tennis racket on the
longitudinal rebound angle of the ball, for an impact at 35 m-s'1 and 20° with 300
rad-s"1of backspin................................................................................................ 199
Figure 8.17 Effect of the position of the balance point of a tennis racket on the
horizontal rebound angle of the ball, for an impact at 35 m-s'1 and 20° with 300
rad-s'1of backspin.................................................................................................200
Figure 8.18 Effect of the position of the balance point of a tennis racket on the
rebound sidespin of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s'1of
backspin................................................................................................................ 201

XIV
Figure 8.19 Effect of the position of the balance point of a tennis racket on the
rebound topspin of the ball, for an impact at 35 m-s'1 and 20° with 300 rad-s"1of
backspin................................................................................................................ 202
Figure 8.20 Definition of vertical and horizontal velocity for an impact between a
tennis ball and freely suspended racket............................................................. 203
Figure 8.21 Effect of racket structural stiffness on the vertical velocityof an
impact at the tip at 35 m-s'1 and 20° with 300 rad-s'1 of backspin a) Vertical
force, b) Vertical ball velocity, c) Vertical racket tip displacement and d) Vertical
racket COM displacement................................................................................... 204
Figure 8.22 The vertical deformation of a stiff and flexible freely suspended
tennis racket for an impact at the tip................................................................... 204
Figure 8.23 Effect of racket structural stiffness on the vertical velocityof an
impact at the throat at 35 m-s"1 and 20° with 300 rad-s'1of backspin a) vertical
force, b) vertical ball velocity, c) racket tip displacement and d) racket COM

W
displacement.........................................................................................................205
Figure 8.24 The vertical deformation of a stiff and flexible freely suspended
IE
tennis racket for an impact at the throat............................................................. 205
Figure 8.25 Effect of racket mass on the vertical velocity of an impact at the tip
at 35 m-s'1 and 20° with 300 rad-s'1of backspin a) vertical force, b) vertical ball
EV
velocity, c) racket tip displacement and d) racket COM displacement 206
Figure 8.26 Effect of racket mass on the vertical velocity of an impact at the
throat at 35 m-s'1 and 20° with 300 rad-s'1 of backspin a) vertical force, b)
PR

vertical ball velocity, c) racket tip displacement and d) racket COM

displacement.........................................................................................................207
Figure 8.27 Effect of balance point on the vertical velocity of an impact at the tip
at 35 m-s'1 and 20° with 300 rad-s'1of backspin a) vertical force, b) vertical ball
velocity, c) racket tip displacement and d) racket COM displacement 208
Figure 8.28 Effect of balance point on the vertical velocity of an impact at the
throat at 35 m-s'1 and 20° with 300 rad-s'1 of backspin a) vertical force, b)
vertical ball velocity, c) racket tip displacement and d) racket COM
displacement.........................................................................................................209
Figure 8.29 Spin generation for an impact close the GSC of a freely suspended
racket, with an inbound velocity of 28 m-s'1, angle of 23° and zero spin 210

XV
Figure 8.30 Effect of racket stiffness on an impact at the tip at 35 m-s'1 and 20°
with 300 rad-s'1 of backspin a) Horizontal force, b) Horizontal ball velocity, c)
Topspin, d) Horizontal tip displacement and e) Horizontal COM displacement.
212
Figure 8.31 The horizontal deformation of a stiff and flexible freely suspended
tennis racket for an oblique impact at the tip......................................................212
Figure 8.32 Effect of racket stiffness on an impact at the throat at 35 m-s'1 and
20° with 300 rad-s'1of backspin a) Horizontal force, b) Horizontal ball velocity, c)
Topspin, d) Horizontal tip displacement and e) Horizontal COM displacement.
...............................................................................................................................213
Figure 8.33 The horizontal deformation of a stiff and flexible freely suspended
tennis racket for an oblique impact at the throat................................................ 214
Figure 8.34 Effect of racket mass on an impact at the tip at 35 m-s'1 and 20°
with 300 rad-s'1 of backspin a) Horizontal force, b) Horizontal ball velocity, c)

W
Topspin, d) Horizontal tip displacement and e) Horizontal COM displacement.
...............................................................................................................................215
IE
Figure 8.35 Effect of racket mass on an impact at the throat at 35 m-s'1 and 20°
with 300 rad-s'1 of backspin a) Horizontal force, b) Horizontal ball velocity, c)
Topspin, d) Horizontal tip displacement and e) Horizontal COM displacement.
EV
216
Figure 8.36 Effect of racket balance point on an impact at the tip at 35 m-s'1 and
20° with 300 rad-s'1of backspin a) Horizontal force, b) Horizontal ball velocity, c)
PR

Topspin, d) Horizontal tip displacement and e) Horizontal COM displacement.

...............................................................................................................................217
Figure 8.37 Effect of racket balance point on an impact at the throat at 35 m-s"1
and 20° with 300 rad-s'1 of backspin a) Horizontal force, b) Horizontal ball
velocity, c) Topspin, d) Horizontal tip displacement and e) Horizontal COM
displacement.........................................................................................................218
Figure 8.38 Diagram showing a centre and off-centre impact on a freely
suspended racket................................................................................................. 219
Figure 8.39 Predicted optimised tennis racket design for all round performance
221
Figure 8.40 Von Mises stress for an impact between a ball and freely
suspended racket with an inbound velocity of 35 m-s'1, an angle of 20° and a

XVI
backspin of 300 rad-s'1. The racket has a mass of 0.348 kg, a natural frequency
of 143 Hz and a balance point 0.396 m from the butt........................................221
Figure 1.1 Number of elements in the ball model against a) maximum
displacement of the ball and b) maximum von Mises stress in the ball 244
Figure 1.2 Frequency results for a 5 m-s'1 impact on a force plate................. 245
Figure 1.3 Frequency results for a 15 m-s'1 impact on a force plate............... 245
Figure 1.4 Frequency results for a 25 m-s'1 impact on a force plate............... 246
Figure 1.5 Horizontal and vertical positions of the centre of the string-bed...247
Figure 1.6 Set-up for 60° impacts, showing the location of the releasepin and
pivot....................................................................................................................... 248
Figure 1.7 Calculating the horizontal location of the centre of the string-bed, for
60° impacts........................................................................................................... 249
Figure 1.8 Calculating the vertical location of the centre of the string-bed, for 60°
impacts.................................................................................................................. 249

W
Figure 1.9 Calculating impact position at 60° inbound angle........................... 250
Figure 1.10 Calculating impact position at 60° inbound angle......................... 250
IE
Figure 1.11 Set-up for 20° impacts, showing the location of the release pin and
pivot....................................................................................................................... 251
Figure 1.12 Obtaining the horizontal position of the centre of the string-bed for
EV
20° impacts........................................................................................................... 251
Figure 1.13 Obtaining the horizontal position of the centre of the string-bed for
20° impacts........................................................................................................... 252
PR

Figure 1.14 Calculating the impact distance from the string-bed along the
horizontal axis for the 20° impacts...................................................................... 252
Figure 1.15 Calculating impact position at 60° inbound angle.......................... 252
Figure 1.16 Effect of inbound spin inbound on a) velocity, b) angle and c) impact
location for the centre impacts............................................................................ 253
Figure 1.17 Rebound a) velocity, b) angle and c) spin. Inbound velocity = 30
m-s'1, inbound angle = 40°................................................................................... 254
Figure 1.18 Horizontal rebound velocity............................................................. 255
Figure 1.19 Vertical rebound velocity..................................................................256
Figure 1.20 Rebound spin................................................................................... 256

XVII
Figure 1.21 a) Comparison of measured spin rates from each camera using the
revolution method and b) comparison of spin between the angle and revolution
method.................................................................................................................. 258
Figure 1.22 Calculating tension (T) and extended length (L) (Cross et al., 2000).
............................................................................................................................... 259
Figure 1.23 Modified impact rig (Hammer head replaced with bolt)...............260
Figure 1.24 Camera set-up................................................................................. 261
Figure 1.25 Force versus time for the original impact velocity.......................... 263
Figure 1.26 Force plot - String 1 impact 5 higher velocity.................................265
Figure 1.27 Force plot for an impact with the hammer headreplaced by a bolt
266

W
IE
EV
PR

XVIII
 List of tables
Table 2.1 Test limits for ITF approved balls (ITF Technical Department, 2009)..6
Table 2.2 Comparison of ball spin rates from different publications (mean ± SD).
................................................................................................................................. 37
Table 3.1 Richimas tracking repeatability for ball and core impacts on a rigid
surface, (valuei/ value2) = SD / SD as a percentage of the mean................... 71
Table 3.2 RMSE between the model and experiment for rebound velocity 73
Table 3.3 Rebound velocity comparison for the pressurised ball models based
on the measured and extended tensile rubber material data.............................. 74
Table 3.4 RMSE between the model and experimental data for maximum
deformation............................................................................................................. 75
Table 3.5 RMSE between the model and experimental data for contact time...76
Table 3.6 Fit to MAT_OGDEN_RUBBER for temperatures of 283.15 and 313.15
K...............................................................................................................................82

W
Table 4.1 Young's modulus values obtained for the tennis string......................93
Table 4.2 Inbound angles, velocities and impact locations relative to the centre
IE
of the string-bed (mean ± SD)............................................................................. 100
Table 4.3 Standard deviations for the manual tracking method for ball impacts
EV
on a head-clamped racket, (value) = SD as a percentage of the mean 100
Table 4.4 Part material properties for the head-clamped racket model 113
Table 4.5 Actual experimental inbound velocities and angles (mean ± SD) ...117
Table 4.6 Standard deviations for the manual tracking method for the head-
PR

clamped racket, (value) = SD as a percentage of the mean............................. 118

Table 4.7 Impact positions for the FE model investigation................................ 123
Table 5.1 Properties of the ITF Carbon Fibre tennis racket (Goodwill, 2002). 131
Table 5.2 Measured polar moment of inertia for the ITF Carbon Fibre tennis
racket.................................................................................................................... 132
Table 5.3 Racket mass distribution in the FE model..........................................133
Table 5.4 Natural frequencies of the two racket models with different Young's
modulus................................................................................................................. 135
Table 5.5 Impact locations for the perpendicular impacts on a freely suspended
racket (mean ± SD)...............................................................................................142
Table 5.6 Inbound velocities, angles and impact locations for the oblique
impacts on a freely suspended racket (mean ± SD).......................................... 144
XIX
Table 5.7 Results of a repeatability test for impacts with low medium and high
inbound spin, (value) = SD as a percentage of the mean................................. 145
Table 5.8 Initial conditions used in the FE model to simulate an impact between
a tennis ball and freely suspended racket...........................................................146
Table 5.9 RMSE between the FE models and experimental data for rebound
velocity, for perpendicular impacts on a freely suspended racket................... 148
Table 7.1 Pre-impact conditions from the player testing results (+ x offset =
towards inbound path of the ball) (+ y offset = towards tip) (Player data from
Choppin etal. (2007a & b))..................................................................................175
Table 8.1 Impact locations on the string-bed used to determine the effect of
different racket parameters.................................................................................. 183
Table 8.2 Two sets of FE simulations used to determine the effect of racket
structural stiffness.................................................................................................183
Table 8.3 Two sets of FE simulations used to determine the effect of racket

W
mass. The mass moment of inertia, the polar moment of inertia and natural
frequency of the racket in the FE model are also displayed..............................184
IE
Table 8.4 Density and mass of the separate parts of the racket in the two FE
models used to determine the effect of racket mass..........................................184
Table 8.5 Two sets of FE simulations used to determine the effect of the
EV
balance point of the racket. The mass moment of inertia, the polar moment of
inertia and natural frequency of the racket in the FE model are also displayed.
............................................................................................................................... 185
PR

Table 8.6 Density and mass of the separate parts of the racket in the two FE
models used to determine the effect of the position of the balance point 185
Table 1.1 Mesh convergence study.................................................................... 243
Table 1.2 Calculated inbound velocity and impact location.............................. 253
Table 1.3 Previous results.................................................................................. 260
Table 1.4 Quasi-static Young's modulus............................................................ 262
Table 1.5 Overall results - original velocity........................................................ 263
Table 1.6 Displacement sensitivity study - String 1 impact 4 originalvelocity.264
Table 1.7 Overall results for the 0.4 kg pendulum at the higher velocity......... 264
Table 1.8 Overall results......................................................................................265
Table 1.9 Estimation of strain rate......................................................................266

XX
 Nomenclature
Abbreviations
ACOR Apparent coefficient of restitution Page 31
COM Centre of mass
COR Coefficient of restitution Page 7
COF Coefficient of friction
DMA Dynamic Modulus Analysis Page 8
FE Finite element
FFT Fast Fourier transform
GSC Geometric string-bed centre
ITF International Tennis Federation
RDC Racket diagnostic centre Page 20
RMSE Root mean squared error
Standard deviation

W
SD
SPR Surface pace rating Page 13
TDT Tennis Design Tool IE Page 96

Roman letters
A Cross sectional area [m2]
EV
E Young's Modulus [N-rrf2]

fn Natural frequency [Hz]

G Shear Modulus [N-rrf2]
PR

1 Moment of inertia [kg-nr2]

K Dynamic modulus of strings [N-rrf1]
k Structural stiffness of racket [N-rrf1]
L Length [m]
m Racket mass [kg]
P Pressure [N-nrf2]
T Period of torsional vibration [s]
V Volume [m3]

Greek letters
V Poisson's ratio
0 Inbound angle relative to racket normal
/j Coefficient of friction

XXI
1. Introduction
The following chapters contain a three year study into the creation and
validation of a finite element (FE) model of an impact between a tennis ball and
racket.

1.1. Motivation for the Research

Over the years, tennis technology has developed, which has had an enormous
impact on the way the game is played. Racket materials have changed from
wood to aluminium, to the oversized, more exotic composite ones used today
(Haake et al., 2007). These advances have allowed players to hit shots faster
and with greater accuracy (Brody, 1997a), effectively increasing the speed of
the game (Brody, 1997b). However, this is also believed to have increased the
dominance of the server and there is growing apprehension that this is resulting

W
in a reduction in spectator appeal (Kotze et al., 2000). The International Tennis
Federation (ITF) is concerned with maintaining public and commercial interest,
IE
in order to prevent the demise of the sport through lack of financial support. To
successfully regulate a sport, such as tennis, the governing body needs a full
understanding of the physical principles and technologies within the game.
EV
Thus, the ITF set up a Technical Department in 1997 in order to monitor and
direct scientific advances in the sport (ITF Technical Department, 2009).
As a scientific subject area, tennis is well publicised with advances in
PR

knowledge and technology coming from within academia and industry.

Researchers, scientists and engineers have simulated the various aspects of
the game through conventional laboratory investigations, which can be both
costly and time consuming. A large number of published studies have been
concerned with creating analytical models. Discrepancies between publications
have arisen due to errors and assumptions in both experimental and modelling
techniques.
The sponsors of this project are Prince whose principal aim is to design and
manufacture tennis rackets, for use by both amateur and professional players.
Prince require a tool which can be used to aid the design of their next
generation of rackets, which must conform to the current rules of the ITF. This
tool should be straightforward and easy to use to enable it to fit seamlessly into
their existing design process. FE models have been successfully applied to