Principles of Tribology

Principles of Tribology
Edited by

J. Halling
Professor of Engineering Tribology
University of Salford

M
© The Contributors 1975, 1978
All rights reserved. No part of this publication
may be reproduced or transmitted, in any form or
by any means, without permission.
First edition /975
Paperback edition /978
Published by
THE MACMILLAN PRESS LTD
London and Basingstoke
Associated companies in Delhi Dublin
Hong Kong Johannesburg Lagos Melhoum~
New York Singapore and Tokyo
ISBN 978-0-333-24686-3 ISBN 978-1-349-04138-1 (eBook)
DOI 10.1007/978-1-349-04138-1

This book is sold subject to the standard conditions of the

Net Book Agreement.
The paperback edition of this book is sold subject to the condition that it
shall not, by way of trade or otherwise, be lent, resold, hired out, or
otherwise circulated without the publisher's prior consent in any form of
binding or cover other than that in which it is published and without a similar
condition including this condition being imposed on the subsequent purchaser.
This book is dedicated
to the Lima Arms whose
efficacious personal
lubrication service
has often eased the
path to its completion
Contents
Foreword XI
Preface XIII

Introduction 1
J. Halling

1.1 Tribology 1
1.2 Historical 4
1.3 Tribology in industry 7
1.4 Economic considerations 10
1.5 Tribological solutions II

2 Surface Properties and Measurement 16

J. Halling and K. A. Nuri
2.1 The nature of metal surfaces 16
2.2 Surface texture assessment 18
2.3 Surface parameters 22
2.4 The statistical properties of surfaces 25
2.5 Measurement of surface parameters 38

3 Contact of Surfaces 40
J. Halling and K. A. Nuri

3.1 Introduction 40
3.2 Stress distribution due to loading 41
3.3 Displacements due to loading 47

VII
3.4 Hertzian contacts 48
3.5 The contact of rough surfaces 61
3.6 Criterion of deformation mode 65
3. 7 Thermal effects 67

4 Friction Theories 72
D. G. Teer and R. D. Arnell
4.1 Introduction 72
4.2 Friction measurement 74
4.3 Possible causes of friction 77
4.4 The adhesion theory of friction 79
4.5 Modified adhesion theory 81
4.6 Plastic interaction of surface asperities 87
4. 7 Ploughing effect 89
4.8 Elastic hysteresis losses 91
4.9 Discussion of the various friction theories 91

5 Wear 94
D. G. Teer and R. D. Arnell
5.1 Introduction 94
5.2 Types of wear 95
5.3 Various factors affecting wear 113
5.4 Experimental aspects 120
5.5 Wear prevention 122
5.6 Application of wear relationships to design 122
5. 7 An example of wear in practice- wear of an i.e. engine 125
5.8 Conclusions 126

6 Tribological Properties of Solid Materials 128

R. D. Arnell and D. G. Teer
6.1 Introduction 128
6.2 Tribological properties of metals 129
6.3 Self-lubricating materials 133
6.4 Types of solid lubricant 134
6.5 Tribological properties of plastics 140

7 Friction Instability 147

L. Eaton
7.1 Introduction 147
7.2 Characteristics of friction vibrations 151
7.3 Review of analytical methods 154
7.4 Frictional force models 158

viii
7.5 Analysis of stick-slip oscillations 161
7.6 Further analysis 170
7.7 Eliminatidn of stick-slip 171

8 Mechanics of Rolling Motion 174

J. Halling

8.1 Introduction 174

8.2 Free rolling 175
8.3 Microslip in rolling 184
8.4 Tyre-road contacts 198

9 Lubricant Properties and Testing 202

R. B. Howarth
9.1 Introduction 202
9.2 Viscosity 202
9.3 Measurement of viscosity 213
9.4 Lubricating oils 224
9.5 Greases 228

10 Hydrodynamic Lubrication 233

T. L. Whomes
10.1 Introduction 233
10.2 Theory 235
10.3 Application of Reynolds equation to sliding bearings 238
10.4 Contacts in the form of non-conforming discs 243
10.5 The journal bearing 247
10.6 Variable viscosity----'the reduced pressure concept 252
10.7 Shear stresses and traction in hydrodynamic films 254
10.8 Finite length bearings 260
10.9 Thermal effects 268
10.10 Gas-lubricated bearings 271
10.11 Hydrodynamic instability 282

11 Elastohydrodynamic Lubrication 288

T. L. Whomes and J. Halling
11.1 Highly loaded contacts 288
11.2 Elastohydrodynamic theory 291
11.3 Comparison of theory and experiment 296
11.4 Traction 300
11.5 Three-dimensional solutions 304
11.6 Fatigue failure 306

IX
12 Hydrostatic Lubrication 308
P. B. Davies and R. B. Howarth
12.1 Externally pressurised bearings 308
12.2 General description of hydrostatic bearings 309
12.3 Viscous flow through rectangular gaps and circular
tubes
310
12.4 Long rectangular thrust bearings in constant flow
system
312
12.5 The need for compensation in multibearing
arrangements 315
12.6 Characteristics of compensated bearings 317
12.7 Comparison of characteristics 325
12.8 Flow, load and power factors for other shapes of
bearings 333
12.9 Sliding effects in thrust bearings 340
12.10 Hydrostatic journal bearings 349
12.11 Other types of hydrostatic bearings 358

13 Selection of tribological solutions 360

J. Halling
13.1 Introduction 360
13.2 Load and speed 361
13.3 The selection of journal bearings 367
13.4 Matching of tribological solutions 367
13.5 Conclusions 368

Appendix A 369

Appendix B 375

Author Index 394

Subject Index 397

X
Foreword
by H. Peter lost
Chairman: Committee on Tribo/ogy, Department of Industry, 1966-74
President: International Tribo/ogy Council

Without tribology, in other words, without 'interacting surfaces in relative

motion', that is, surfaces rolling on each other, surfaces sliding over each
other and surfaces rubbing on each other, life would be impossible.
This truism applies equally to heavy machinery and to a precision mecha-
nism; to a brake and to a rocket; to a mechanical and to a human joint.
Friction and wear, the principal constituents of tribology, have been with
us since time immemorial, so have been our efforts to control the former and
minimise the latter. However, it is only less than nine years ago, that the
modern interdisciplinary concept of tribology was recognised in the United
Kingdom, since when its progress has spread like wildfire throughout the
industrial world.
The main reason behind this development, which is also the principal
reason for the importance of this book, is the recognition of the close inter-
relationship between tribological design principles and practices on the one
hand, and their economic effect on the other. The days of single disciplinary
designs and of design by trial and error are gone forever. Modern products
must, during their design stages, have incorporated all the factors that lead
to a satisfactory control of friction and prevention of wear.
For this to be accomplished, Professor Halling and his co-authors have

xi
provided, in a single volume, nearly all the basic theory needed for a thorough
grounding in the subject. It will therefore be a book most useful for every
engineering student; in addition, I believe that Principles of Tribology will
be found very useful by practising engineers in industrial design and research
departments. For here, the concentration of valuable information in one
volume, will eliminate the consultation of several sources of books and
scientific papers, a saving of valuable time that will be appreciated by those
employed in industry and others working against the clock.
Similarly, this book should be of considerable value to workers in many
other fields, whose occupation brings them face to face-often unpleasantly
so-with the realities of tribological problems and who may find, in this
volume, fundamental answers to at least some of their problems.
The original Report, that bears my name, estimated that by the better
application of tribological principles and practices, industry in the United
Kingdom could save around £515 million per annum (at 1965 values).
During the years since its publication, it has become apparent that this
estimate of savings has been too conservative. Indeed, in a recent Report,
commissioned by the Congress of the United States of America, it was stated
that there was considerable scope for savings of losses through tribological
causes (friction and wear), which were estimated to cost the U.S. economy
around $100 billion per annum, of which $20 billion were in materials.
The savings through the correct application of tribological principles, as
outlined in this book, can be considerable. A recent report, commissioned
by the S.S.R.C., concluded that 'for individual enterprises-even efficient
ones-the rate of return in improvements in the tribological characteristics
of their capital equipment could be very high indeed-far higher than is
customary on ordinary industrial investments'.
Modern machines, mechanisms and equipment must be reliable. I firmly
believe that the application of the ground rules, contained in Principles of
Tribology, will materially contribute towards greater reliability of industrial
products and therefore to the economy of this country.
I congratulate Professor Halling and his co-authors on their work, which
I can warmly recommend not only to those desiring to become engineers,
but also to those professional engineers and others whose work is connected
with the control of friction and the prevention of wear, in other words, in
work where the minimisation of breakdowns, replacements and outages
through tribological causes are of importance.

London H. PETER JOST

xii
Preface
Tribology is a new word, not yet in common usage, but it deals with problems
which man has encountered throughout the whole of his history. The word
was introduced to focus attention on the problem of carrying load across
solid interfaces in relative motion. The ingredients of the subject are there-
fore well-established batches of knowledge which occur in a wide range of
texts on such topics as lubrication, friction, wear, contact mechanics, surface
physics, and chemistry. The subject is truly interdisciplinary since the basic
knowledge from physics, chemistry, mathematics, materials science and
engineering is used to study problems in all branches of engineering, in
medicine, and in almost all aspects of our daily life from the cleaning of our
teeth to the slicing of our golf drive.
The awareness of the social and economic importance of this subject has
resulted in its introduction into several courses at colleges and universities.
This book is an attempt to bring together in a single volume those topics
which are currently scattered throughout the scientific literature. In particular
this volume concentrates on the basic principles of the subject, using practical
examples only to demonstrate the physical manifestations of these principles.
This book should, therefore, prove a useful supplement to other volumes
dealing with such specific practical applications as the design of bearings.
The book is basically aimed at the final-year level of undergraduate courses
in engineering, but the authors hope that it will prove of interest to a wider
readership. Research workers, new to the subject, and designers and develop-
ment engineers seeking background knowledge to the existing literature
should all find the subject matter relevant.
The authors have deliberately excluded illustrations of standard equipment

XIII
and techniques since these should already be familiar to final-year under-
graduates and practising engineers. The notation may also appear to lack
complete consistency but it has been considered desirable to use that notation
which is already well established in the literature concerned with the various
topics considered. This should facilitate the use of standard references, many
of which are included with each chapter. The symbols used are clearly
identified in the text of each chapter.
It may be that the relatively large number of authors has resulted in some
variation in literary style, but since each chapter is more or less self-contained
this should offer no problem to the reader. Since all the authors belong to the
same department we believe that our collaborative efforts have resulted in a
coherent philosophy for the book. In each topic the authors have concen-
trated on the important physical principles and have not thought it desirable
to include rigorous development of the mathematical treatment, although
the most significant mathematical formulations are included. Since the
authors are mainly concerned with the engineering applications oftribology,
it will be noted that this book does not include much information on the
chemical aspects of tribology. This does not imply any denigration of the
importance of such material but rather that such topics tend to be of a rather
specialist interest.
The final chapter has been included as an indication of the relevance of
the remainder of the book. Indeed this chapter could be taken out of sequence
if the reader seeks to appreciate the value of the topics discussed in the
various chapters.
A number of problems together with outline solutions have been included
in an appendix to this book, since the final proof of assimilation of concepts
must be our ability to handle them in given situations.
Finally the authors would like to acknowledge their deep gratitude for the
enthusiastic and competent way in which Mrs L. M. Chadderton has tran-
scribed their often untidy drafts into a coherent manuscript.

Salford J. HALLING

XIV
1
Introduction
1.1 TRIBOLOGY

What does the word mean? It is derived from the Greek word TRIBOS meaning
rubbing, so that a literal translation would be 'the science of rubbing'. The
word is so new as to appear in only the latest editions of dictionaries where
it is there defined as 'the science and technology of interacting surfaces in
relative motion and of related subjects and practices'. This latter definition,
although embracing the literal translation, is of even wider significance
and was created to bring together the interest in friction and wear of chemists,
engineers, metallurgists, physicists and the like. This wide-ranging concern
with tribology immediately illustrates the interdisciplinary nature of the
subject. In a sense it is the name alone which is new because man's interest
in the constituent parts of tribology is older than recorded history. Clearly,
the invention of the wheel illustrates man's concern with reducing friction in
translationary motion, and this invention certainly predates recorded history.
That man should have been so concerned with the tribological problems of
friction and wear is not surprising because our involvement with such
phenomena affects almost every aspect of our lives.
These problems are not confined to the machines which we use, they also
have profound influences on many other aspects of life. The action of animals'
joints is clearly a tribological situation and cures for such diseases as arthritis
already owe much to the tribologists' expertise. We also rely on the control
of friction in our leisure pursuits, whether they be rock climbing or any form
of ball game-such as the spinning of a cricket or tennis ball, the slicing of
our drive in golf or our proficiency on skates or skis. Holding, cutting and

1
brushing are other manifestations of the impact of tribology on our daily
life, while the cleaning of our teeth is clearly a controlled wear-process where
we wish to avoid wear of the enamel while wearing away unwanted films, etc.
Even our ability to walk is dependent on the existence of appropriate
friction, so that tribological effects have clearly had a major effect on the
whole evolutionary process.
We may examine the effect of friction on the evolutionary process by
considering the way in which developments in translation over the earth's
surface have evolved against the timescale of history 1. In figure 1.1 the resist-
ance to translation is represented by a resistance/weight ratio which might

10- 1

:E
.0' .,
;<
Gs M, Sild1ng
Lubricant
M2
'ec 10- 2 M3 Early wheel
M4 Spoked wheel
~
gj Ms Railway
0:: M6 Modern railways
" G, First reptile
'""' G2 Crawlers
G3 Apes
10- 3 G4 Man
Gs Atheletes
M6
c
-~
.0
u -~ E
0
10- 4 ~ u

~ "' "' Q: "'

1 I·
I
I
10
I
10 2
I
103
Cenozoic era
I
104
I
105
I
10 6
I
107
·I I I, 1..
108 109
Years

Figure 1.1 Improvements in land locomotion throughout

history

be considered as an equivalent coefficient offriction A.. This clearly represents

the ease of translation and its reduction with timescale is seen as evolution
has developed from primeval sliding to the movements of a modern athlete,
that is along line G. In this figure it should be noted that the scales are
logarithmic rather than linear so that the shape of the curve is somewhat
deceptive. Modern man starting some 10 000 years ago has used his inventive-

2
ness to achieve a much better performance than is obtained from the physio-
logical developments in animals. Thus the use of lubricants and improve-
ments in the design of wheels have clearly proved advantageous.
It is also interesting to plot the results in figure 1.1 against the speed
achieved by each method of translation, shown in figure 1.2. Again we note

003 03 3 30
Speed m/s

Figure 1.2 The results from figure 1.1 indicating speeds

achieved

the two curves, one representing the evolutionary process and the other due
to man's inventiveness. What is most interesting in this logarithmic plot
is that the evolutionary line is approximately straight and has a negative
slope of 45°. This means that
log A = - 1 x log S + log C (1.1)

where S is the speed of translation and C is some constant. This equation

may thus be expressed
AS= C ( 1.2)
But AS is the resisting force multiplied by speed and divided by the weight,
that is, power divided by the weight carried, thus this straight line suggests
that the evolutionary line is governed by a substantially constant power/
weight ratio. Clearly this is reasonable, since the power/weight value is

3
defined by the same physiological processes in all animals, including man.
The curve due to man's inventiveness does not suffer from such a restriction,
since our inventions have benefited from a more successful application of
scientific principles and the use of other than purely physiological materials.

1.2 HISTORICAL

Man's invention of the wheel as one of the earliest tribological devices has
already been mentioned, but friction affected the development of civilised
man in many other ways 2 • It is known that drills made during the palaeo-
lithic period for drilling holes or producing fire were fitted with bearings
made from antlers or bones, and potters' wheels or stones for grinding ·
cereals, etc., clearly had a requirement for some form of bearing 3 . Records
show the use of wheels from 3500 B.C., and yet it is interesting to note that the
very advanced Inca civilisation of more recent date never did discover the
principle of the wheel. Lubricants were also used from about this period,
and a tomb in Egypt provided evidence of this fact. A chariot in this tomb
still contained some of the original animal-fat lubricant in its wheel bearings
and it is also interesting that this lubricant was contaminated with road
dirt in the form of quartz sand and compounds of aluminium, iron and lime.
In their monumental tasks of building the Egyptians also showed a clear
appreciation of tribological principles 4 . Surviving illustrations in the form of
bas-reliefs show the use. of rollers and sledges to transport their heavy
weights. Figure 1.3 illustrates one example of such transportation; here 172
slaves are being used to drag a large statue weighing about 6 x 10 5 newtons
along a wooden track. Closer examination shows one man (standing on the
sledge supporting the statue) pouring a liquid into the path of motion;
perhaps one of the earliest lubrication engineers. If we assume that each
slave pulls with about 800 newtons we can see that this picture would suggest
a coefficient of friction of
172 X 800
J1. = 6 X 105 ~ 0.23

This is just about the value we would expect for a lubricated wooden slide,
thus we can infer that this picture is a true record of what actually occurred.
The somewhat artificial arrangement of the slaves probably arises from the
artist's inability to draw perspective.
In 1928 fragments of what must have been a ball thrust-bearing were
found in Lake Nimi near Rome, and probably date from about A.D. 40.
The bearing is shown in figure 1.4. It was probably used to support a statue
in a sculptor's workshop thus facilitating rotation during the sculpting
process. There is little further evidence of tribological development until
the time of Leonardo da Vinci (1452-1519), who first postulated a scientific
approach to friction. He recognised that the friction force is proportional to

4
5
I

I
---0---- ~ 1
- fi1
I

Figure 1.4 Detail of fragment and theoretical

reconstruction of Lake Nemi hearing, estimated
diameter nearly 1 metre (reproduced from Ucelli,
Le Navi di Nemi, Libreria Delio Stato, Roma)

load and independent of the nominal area of contact. It was almost two
hundred years however before these two laws were enunciated by Amon ton
in 1699, who, independently of Leonardo da Vinci, postulated these laws and
is usually credited with their discovery. The eighteenth century saw consider-
able tribological development because of the increasing involvement of man
with new machines, and about 1780 the work of Coulomb led to the third
law of friction which suggested that friction was independent of velocity.

6
These three laws are still used today as reasonably true and are to be found
in the teaching of friction in elementary texts in physics and engineering.
Many other developments occurred during this century, particularly in the
use of improved bearing-materials. As early as 1684 Robert Hooke suggested
the combination of steel shafts and bell-metal bushes as preferable to wood
shod with iron for wheel bearings. Further developments were undoubtedly
associated with the growth of industrialisation in the latter part of the
eighteenth century.
With lubricated bearings, although the essential laws of viscous flow
had earlier been postulated by Newton, any scientific understanding of
their operation did not occur until the end of the nineteenth century. Indeed
our understanding of the principles of hydrodynamic lubrication can be
seen to date from the experimental studies of Beauchamp Tower (1883) 6
and the perspicacious theoretical interpretation by Osborne Reynolds
(1886) 7 • Other work by Stokes and Petrof'~ about this time was also very
pertinent to these developments. Subsequent development in hydrodynamic-
bearing theory and practice was extremely rapid in an attempt to meet the
increasing demand for reliable bearings in the new machinery then being
developed, some of which is still operational to this day.
Since the beginning of this century, and spurred on by the industrial
demand for better tribology, our knowledge in all areas of tribology has
expanded enormously. In this context ball-bearings, which were first intro-
duced for industrial applications about 1700, have now reached an unusual
peak of efficiency. They are available in a wide range of sizes and today offer
a very cheap and flexible solution to many tribological design problems.
Other developments in the quality and service characteristics of lubricants
have also added to our toolkit of tribological solutions, but the pace of
modern industrial society is such that increasing demands for higher speeds
and loads, often in hostile environments such as nuclear reactors and space
vehicles, still necessitate further development of the subject. Today we take
as the norm that a car engine should last about 100 000 miles, whereas less
than twenty-five years ago the life expectancy was only a third of this value.
It is also interesting to note that a modern car contains upward of 2000
tribological contacts, so that it is not surprising that this subject is of increas-
ing importance to engineers engaged in a wide variety of engineering
disciplines.

1.3 TRIBOLOGY IN INDUSTRY

Before discussing the prevalence and types of tribological problems in
industry we must first take a somewhat global view of the phenomena. In
essence we are dealing with the interaction of two solid surfaces within a
given environment which results in two outward manifestations.
(a) There is an energy dissipation which is the resistance to motion and is
indicated by the coefficient of friction. This energy dissipation results

7
in a heat release at the contact and a small, but sometimes significant,
amount of noise. It should be emphasised that since two solid surfaces
are always involved, parameters such as the coefficient of friction must
relate to the pair of interacting materials. To talk about the coefficient
of friction of steel without reference to the mating solid is scientifically
incorrect and misleading. It might also be remarked that the idea of
frictionless surfaces is scientifically impossible, and the often-stated
implication that low friction is associated with the smoothness of the
surface is also basically incorrect.
(b) During the sliding process all surfaces are to a greater or lesser extent
changed in their basic characteristics. They may become smoother or
rougher, have physical properties such as their hardness altered,
and some material may be lost in the so-called wear processes. Such
surface changes may be either beneficial, as happens when surfaces
'run in' to produce near ideal operation, or disastrous, as happens when
surface failure occurs, necessitating the replacement of components.
From the foregoing it may be thought that both friction and wear are
always disadvantageous, but this is not the case. In many engineering appli-
cations we are able to employ friction to fulfil required functions. Thus brakes,
clutches, driving wheels on trains, cars, etc. all operate because ofthe existence
of friction, while the commonplace nut and bolt only work because of the
friction between the two. In the same way the wear of machinery is some-
times advantageous. The initial wear resulting in better mating of com-
ponents (running-in), is evidently desirable, while the fact that components
wear-out provides a strong motivation to replace obsolescent machinery.
In its most extreme form this leads to the concept of 'planned obsolescence'
where designers endeavour to use the wear phenomena to provide machinery
with a specified life-span.
A fairly general misconception exists that friction and wear, which must
be related in some way ~ince they both arise from the interaction of surfaces,
are simply related such that high friction means high wear. That this is not
the case is clearly shown in table 1.1 where it is seen that the lowest friction
is not associated with the lowest wear. These results are also interesting in
that they show the very large differences in wear rates for materials whose
friction coefficients vary in a fairly modest manner. The complexity of the
relationship between friction and wear is also demonstrated by the peculiar
results which sometimes occur when, with the same materials, friction may
decrease after a'given running time and this reduction in friction is associated
with an increase in the wear rate.
In almost all industrial situations the wear effects are more important than
the frictional losses because they tend to have the greater economic conse-
quences. Higher friction can often be tolerated, with its slightly higher running
costs, provided there are consequent savings by increased wear life of the
machine. Some of the industrial situations where friction and wear are
important are indicated by the following categories.

8
TABLE 1.1 FRICTION AND WEAR FROM PIN ON RING
TESTS. RINGS ARE HARDENED TOOL STEEL EXCEPT IN
TESTS 1 AND 7

Coefficient of Wear rate

Materials friction cm 3 jcm x 10- 12

I Mild steel on mild steel 0.62 157,000

2 60/40 leaded brass 0.24 24,000
3 PTFE 0.18 2,000
4 Stellite 0.60 320
5 Ferritic stainless steel 0.53 270
6 Polyethylene 0.65 30
7 Tungsten carbide on itself 0.35 2

Load 400 g Speed 180 cmjs

1.3.1 Energy Losses

While friction may not always be a prime consideration, the tribologist has
continuously to be aware of the absolute magnitude of the energy losses
involved and any likely consequences of the ensuing heat release. Thus in a
500 MW generating-set a frictional dissipation of only one tenth of one per
cent represents 0·5 MW or the heat equivalent of 500 single-bar electric
fires. This heat is confined to a fairly localised region of the system and
therefore requires special consideration from the designer. Indeed in many
machines the tribological heat release can result in undesirable thermal
distortions.

1.3.2 Wear
In industrialised societies about one-half of the gross national product is
used to replace the results of wear and similar effects, so that these phenomena
are central to the development of such societies. Savings by improvements
in the life of machinery also have the economic advantage that they are
deflationary, a particularly attractive feature in these days of inflation. It
is of course true that wear may also be desirable as a spur to the replacement
of obsolescent machinery, that is, the concept of 'planned obsolescence'.
Such concepts do not however exclude the need to understand wear processes,
since planned wear or wear control of any kind implies an appreciation of the
basic rules of the wear processes.

1.3.3 Control
There is a requirement in sophisticated machjnery for the precise control of
parts in relative motion. This clearly involves one in proper tribological
design and is a factor which is increasingly important for systems using

9
automatic control methods for the positioning of machine elements. Where
parts are moving at very slow speeds such problems are accentuated due to
instabilities in the frictional behaviour.

1.3.4 Environment
It is not possible to solve all engineering tribological problems by a liberal
supply of oil. In many applications the use of such a lubricant is either
excluded or technically undesirable. Typical examples of such situations arise
either in whole or in part with space vehicles, nuclear reactors, chemical
plants, textile plants, food processing machines and many other forms of
machinery. Indeed the tribologists' problems arise increasingly from the
designers' requirement for higher load-capacity, higher speeds and the
operation in difficult and sometimes 'hostile' environments.

1.3.5 Friction Devices

As already mentioned, in many situations friction is a prior requirement for

the functioning of the device. In such cases the heat release and the rate of
wear become the primary design considerations, as in the design of auto-
mobile brakes and clutches.

1.4 ECONOMIC CONSIDERATIONS

The economic significance of tribology is so obvious as to require little

detailed discussion, but it is still true that the savings in individual cases are
generally small, and we all find difficulty in applying the maxim 'save the
pennies and the pounds will look after themselves'. It is because of the
enormous prevalence of tribological contacts in machinery that the small
savings on each contact can add up to significant sums for the nation as a
whole. Thus it is not surprising to find that about one-half of the world's
energy production is used to overcome friction in some form. In a very real
sense we therefore see that savings by better tribological design could have
considerable significance in the conservation debate which concerns the
whole future of mankind. For these reasons governments of industrial socie-
ties have placed increasing emphasis on the economic aspects oftribology.
In Britain the recent acceleration in tribological activities owes much to
the 1966 D.E.S. report, Lubrication (Tribology), Education and Research. 9
This report (the Jost Report) suggested that this country could save no less
than £515 million per annum by better tribological practices. This enormous
sum arises from the constituent savings shown in figure 1.5, and it must
always be remembered that such costs include the loss of production etc
which is consequent on tribological failures in modern industry. It is also

10
Manpower sav•ngs
Lubricant - " -
Investment - " -

Less frrct1ona1
diSSIPOI10n

Longer life
of machines

Fewer breakdowns

Less mamtenonce
and replacement

Figure 1.5 Sarings indicated hy

the Jost Report 9

apposite to note that these savings do not require new research but only the
application of current knowledge. As research in this area expands, even
greater savings could be anticipated. Experience since 1966 suggests that
these figures were not unduly optimistic and the introduction of the word
tribology and the attendant publicity have already had considerable effect
on entrenched attitudes and in the best interest of the nation's economy.

1.5 TRIBOLOGICAL SOLUTIONS

Perhaps the most significant effect of the introduction of the word tribology
has been to introduce a problem-orientated view of any system. We thus
tend to think: 'What is the best solution to the problem of carrying load
across the interface with acceptable friction and wear?' rather than heretofore

11
where the lubrication engineer naturally tended to have a predisposition to
adopt a lubrication solution. We now therefore categorise the available
solutions to tribological problems in the manner shown in figure 1.6 and
discussed in the following text.

(a) Dry contoCf ( bl Chem1col f1lms

/////

~
\\\\\

(c) Lamellar sol1ds (d) Pressurised lubncont

f1lms

(el Elostometers ( fl Flexible stnps

(g l Roll1ng elements (h) Mognet1c f1elds

Figure 1.6 Methods of solution of tribological problems

Figure 1.6a We may choose the contacting materials because they have
intrinsically low friction and/or wear characteristics, although this may
mean accepting lower load-carrying capacities as for instance when plastic
materials are employed. In many cases it is possible to use materials as
surface layers supported on substrates which fulfil the basic structural
requirements of the particular component. This method is employed in the
well-known bearing s~el used in automotive engines.
Figure 1.6b We may apply chemical films which protect the surfaces

12
and in part reduce the intimate contact of the base materials. In such systems
the thermal stability of these films is important due to the high local tempera-
tures which are created at the points of intimate contact during sliding..
Figure 1.6c Solid surface coatings may be used since they have low
resistance to transverse shear, for example soft metal layers or lamellar
solids such as graphite and molybdenum disulphide. These latter materials
have a layer structure rather like a pack of playing cards with strength to
carry normal load and weakness along planes at right-angles to facilitate
sliding.
Figure 1.6d The surfaces may be separated with a continuous film of
fluid, which may be either a liquid, a vapour or a gas. In such systems the
fluid film must have a built-in pressure to withstand the effects of the applied
normal load. Such pressures may be provided by two distinct mechanisms.
The most obvious is to supply the fluid at a pressure generated by an external
pumping system as used in the externally pressurised bearing, often referred
to as the hydrostatic or aerostatic bearing. Alternatively the pressure may be
generated by the motion of the surfaces themselves as they tend to drag the
fluid into a converging gap. This action of the hydrodynamic bearing is more
dependent on the viscosity properties of the fluid than in the hydrostatic case.
In both these fluid-film applications a wide range of fluids such as water,
oil, air and even liquid metals in nuclear reactors, have been successfully
employed.
Figure 1.6e Where the degree of transverse displacement is of fairly
small amplitude the surfaces may be separated by elastometers bonded to the
two surfaces. This clearly offers an excellent tribological solution and alter-
native designs might incorporate flexible elastic strips as shown in figure
1.6f.
Figure 1.6g A widely employed tribological solution is to interpose
rolling elements such as balls, cylinders and the like between the two surfaces.
The wide range of rolling contact bearings available is evidence of the value
of this particular solution.
Figure 1.6h The carrying of load without mechanical contact is clearly
possible by using magnetic and similar force fields. Such a bearing is to be
found in the domestic electricity-supply meter.
Having this range of solutions in mind the designer may now consider
such factors as the load to be carried, the speed, the nature of the environment
and any limitations on friction and wear, in arriving at the most appropriate
answer to his design problem. It sometimes happens, however, that by a
complete change of design philosophy particular tribology problems may
be either eliminated or at least modified while still achieving the desired
effect from the resulting machine. This concept is illustrated in figure 1. 7
where a power unit is required to drive an aeroplane. Here the tribological
implications of three possible engines are considered, namely, the internal
combustion reciprocating engine, the jet engine and the ram jet. The shaded
areas illustrate where tribological contacts have to be considered between the

13
'-'
~
(f)
<l
u
'-'
z w
z
Vi
<l
u
~
w
w
z
i3
z
w

Propel lor

I C ENGINE

1'7"'7"'7"71 Energy
t:::..L...L.L.J diSSipOfiOnOf Propuls1ve
beonngs and Jel
sl1d1ng surfaces
RAM JET

Figure 1. 7 Energy dissipation in devices for propulsion

various engine components and it is seen how these are reduced in number,
although sometimes becoming more complex, as we move from one design
to the next.
In the foregoing it has been assumed that surfaces are simply smooth
planes defining the boundaries between solids and their environment.
Unfortunately all engineering surfaces are rather more complex than this,
having geometries with hills and valleys and equally complex physical and
chemical properties which are seldom uniform throughout the depth of the
material. It is these features of the surfaces which contribute to the complexity
of the subject of tribology. The remainder of this book is therefore devoted to
considering the nature of such surfaces and to the underlying scientific
principles involved in the various topics discussed above.

REFERENCES

I. J. Halling, The roll (role) of the wheel. Wear, 24, (1973), 53.
2. D. Dowson, Tribology-Inaugura1 Lecture. University of Leeds Press, (1969).
3. C. St. C. Davison, Bearings since the Stone Age. Engineering, (Jan. 1957), 2.

14
4. C. St. C. Davison, Transporting sixty-ton statues in early Assyria and Egypt. Tech-
nology and Culture, 2, (1961), No. I.
5. J. H. Harris, The Lubrication of Rolling Bearings. Sheli-Mex and B.P., London,
(1967).
6. B. Tower, Report on Friction Experiments. Proc. lnstn mech. Engrs, (1884), 632.
7. 0. 0. Reynolds, On the theory of lubrication and its application to Mr. Beauchamp
Tower's experiments. Phil. Trans. R. Soc., London, (1886), p. 117.
8. N. P. Petroff, Friction in machines and the effects of the lubricant. Engng J., (1883),
St. Petersburg.
9. Lubrication (Tribology) Education and Research, D.E.S. Report, (1966), H.M.S.O.

15
2
Surface Properties
and Measurement

2.1 THE NATURE OF METAL SURFACES

In the field oftribology it is usually necessary to widen the simple interpreta-

tion of a surface as being a geometric plane separating two media. A surface
must be recognised as a layer that grows organically out of the solid and has
physical properties of considerable functional significance. The surface
layer of metals is known to consist of several zones having physico-chemical
characteristics peculiar to the bulk material itself.
At the base of the surface layer, (see figure 2.1) there is a zone of work-
hardened material on top of which is a region of amorphous or micro-
crystalline structure. This so-called Bielby layer is produced by the melting
and surface flow during machining of molecular layers which are subsequently
hardened by quenching as they are deposited on the cool underlying material.
This basic structure is usually contaminated by the products of chemical
reaction with the atmosphere and is covered by dust particles and molecular
films deposited from the environment. Finally, the surface contains atoms of
gas, which live with the surface and have properties somewhat dissimilar to
those of the gaseous environment. In addition, the whole texture of the surface
layer has a geometric property characterised by a series of irregularities
having different amplitudes and frequency of occurrence. This particular
property, the surface texture, is of fundamental importance in the study of

16
Bielby
layer

Deformed
layer

Base
molena!

Figure 2.1 Typical surface layers

friction, wear and lubrication and thus a knowledge of, and a definition of,
the topographic features of surfaces is vital.
Some feel for the scale of the various surface properties is given in figure
2.2, and it should be noted that the vertical scale is a log scale in ascending
orders of magnitude. Furthermore such terms as oxide films must be care-
fully defined since they are complex layers which, on steel for example, tend
to have an increasing oxygen/metal ratio as the atmosphere is approached.

u
.
.g
~
;;:::
u
~
;;:::
5
:I
E
0 0
c:,..
"' E
ilQ "0
e 5
loG Q z
10 5
Iii
,.. E
!!
Q
.2 ",..
s::
.2 0 g
E
104 "'0
"' g
0
~ "' w
"'0
10 3
. E
...: Air

·'~
c:

102
8" eu Fe 3 0 4
~
10 FeO
Surface
Metal Fe
10

102
Iii
103 ~
,..
D
104 9i
ID c:
Q
105 g
10 6
i,..
>
r
2

Figure 2.2 Order of magnitude of surface features

17
The geometric texture of ordinary surfaces is controlled by the character-
istics of the finishing process by which they are produced. Close examination
of these surfaces, even after the most careful finishing, shows that they are
still rough on a microscopic scale. The roughness is formed by fluctuations
in the surface, of short wavelengths (microroughness), characterised by hills
(asperities) and valleys of varying amplitudes and spacings and these are
large compared to molecular dimensions. On many surfaces a longer wave-
length roughness called waviness is also observed and is often referred to as
macroroughness (figure 2.3). In addition the surface also contains undulations

Macroroughness
(waviness)

M icroroughness

Ill

Resultant
surface

Figure 2.3 Components of surface geometry

of very long wavelengths caused by the vibrations of the workpiece or tool

during the preparation of the surface. The distribution of the asperities over
the surface can be either directional or homogeneous in all directions,
depending upon the nature of the processing method. Surfaces which have
been submitted to directional methods of processing, such as turning,
milling and planing, exhibit a definite orientation of asperity distribution.
Those which have been submitted to non-directional methods, such as
electropolishing and lapping, show an isotropic or equiprobable distribution
in all directions along the surface.
From the foregoing description it can be appreciated that the surfaces of
metals can be as complicated as the surface of the earth and, obviously, in
order to make a precise assessment of their topographic features, three-
dimensional maps are needed. Unfortunately, simple methods for producing
such maps have not yet been developed, and for the time being we must rely
on a two-dimensional profile of the surface together with observations using
the various microscopic techniques which are discussed later.

2.2 SURF ACE TEXTURE ASSESSMENT

The quantitative assessment of the topographic features of surfaces is of
vital importance for solving a wide variety of problems in tribology. Trib-
ological phenomena such as friction and wear depend primarily on the nature
of the real area of contact between the surfaces which is, in turn, dependent

18
upon the distributions, sizes and shapes of the asperities. Measurements of
these features will, therefore, prove indispensable in the study of all kinds of
surface contact phenomena. Indeed, they offer a valuable analysis of the
conditions for elastic or plastic contact of metal surfaces and information on
the size of the interstices between them. Furthermore, the application of
surface measurement to such problems has produced correlation, both
qualitative and quantitative, between theoretical arguments and experi-
mental evidence.
Many methods are available for the measurement of the micro- or macro-
geometrical features of surfaces. They include optical methods using electron,
interference or reflection microscopy and mechanical methods such as
oblique sectioning and profilometry. The optical methods show certain
advantages in that they can provide a three-dimensional appreciation of the
surface. However, the obtaining of quantitative assessment of surfaces by
these methods is still relatively tedious.
Because of the difficulties in representing every irregularity within the
whole plane of the surface, it is necessary to accept measurements which are
based on a small sample of the surface, always provided that its size is
large enough to be representative, that is, there must be a high degree of
probability that the surface lying beyond the sample is similar to that which
lies within it. Most of the existing methods of surface measurement give
either a high-resolution picture of a small, and often unrepresentative,
sample of the surface or a representative sample with correspondingly lower
resolution, (table 2.1). This difficulty is overcome with profilometers which
provide a representative length of the surface and a high resolution in a plane
normal to it 1 . The higher vertical magnification results in a distortion of the
recorded shape of the profile and thereby sometimes leads to a physical
misconception as to the general character of the actual surface profile. Most
surfaces have asperities with gentle slopes rather than the jagged character-
istics observed on profilometer traces, (figure 2.4). The major disadvantage of
profilometers however, is that they are restricted to a single line sample which
may not be representative of the whole surface if the texture has character-
istics dependent upon the orientation of the record. Nevertheless, for iso-
tropic surfaces they provide useful measurements of the various parameters
which are necessary for their characterisation.
It is also appropriate to record a difference between a peak on a profilo-
meter trace and an actual summit on the real surface. Since the stylus will
register a peak even when it only traverses a shoulder of an actual summit,
it will be clearly recognised that the number of true summits is considerably
less than the number of peaks recorded. This distinction may easily be shown
by taking a series of line profiles in the same direction but with each traverse
slightly displaced from its neighbour. In this way, the information in each of
the line tracings may be consolidated to produce a true three-dimensional
contour map of the surface and thereby clearly identifies the actual summits,
(figure 2.W.

19
TABLE 2.1 SURFACE MEASUREMENT METHODS

Resolution (J.!m)

Method Lateral Vertical Comments

Results depend on the quality

Optical of the optical system and
0.25-0.35 0.18-0.35
microscope hence depth of field of
photomicrographs

Light Using optimum microscopic

0.25 0.25
profile conditions

Sectioning angle ~ tan- 1 0.1

Oblique
0.25 0.025 and optimum microscopic
section
conditions

Interference Requires high specular

0.25 0.025
microscope reflectivity of surface

Multiple Requires high specular

beam 5 0.005 reflectivity of surface and no
interference angular deviations greater than 5

Reflection O.o3 Vertical resolution from

electron 0.03-0.04 profiles.
microscope 0.02-0.008 Vertical resolution from shadows

0.005 Vertical resolution using

Electron stereo device.
0.005 Vertical resolution using shadowed
microscope 0.0025
replica angle tan- 1 0.3

Finite size of the stylus

Profilometers 1.3-2.5 0.005-0.25 is ultimate limit on resolution

The best-known profilometry method for obtaining surface profiles is the

stylus method, in which a fine diamond stylus traverses the surface and its
vertical movements are recorded, usually by electrical systems 3 . The
Talysurf (Rank, Taylor, Hobson) is one of the most popular instruments of
this type. In such systems the vertical measurements must be recorded with
respect to some appropriate datum. The two most common methods for the
establishment of such datums are
(a) The use of datum-generating attachments which ensure accurate
horizontal motion of the stylus-support system.
(b) The use of large-radius skids or flat shoes which rest on and traverse the
surface being measured, thereby generating the general level of the
surface texture.

20
SlyiUSIIP
I
rodn;s
I

~ I I
I
I
,,,.,....·vocr•rnv;
I
Ac1uo1 prof1le I
wuhout dJSTOr"fron
D•slor eo I
~ed by I
prof1lometer 1

Figure 2.4

Adjocenr prof•les

F(qure 2.5 Typical surface map (lighter areas corre-

spond to higher surface regions)

21
Clearly the use of such devices creates some experimental errors, although
(a) is rather more precise than (b) for accurate work. Another major error
in these instruments arises from the size limitation of the measuring stylus.
Its finite size clearly precludes it from complete penetration of all the valleys,
thus their recorded shape will often appear narrower than their actual shape.
In a similar way the peaks are distorted in shape such that they appear wider
than they actually are. This effect, however, is less harmful with smoother
surfaces where the slopes of the asperities are known to be fairly gentle.
In spite of its defects, the stylus instrument, as far as engineering purposes
are concerned, remains the outstanding instrument for studying the nature
of surface geometry and for evaluating its parameters. The main difficulty
does not, in fact, lie in reproducing surface profiles accurately, it lies in using
these profiles for the evaluation of the important features which will determine
the functional behaviour of real surfaces.

2.3 SURFACE PARAMETERS

The choice of surface parameters is necessarily influenced to a certain extent

by the method chosen to reveal the surface features. The stylus method reveals
only a single plane property of the surface topography and, consequently,
its characterisation has to be based on the nature of the ensuing single line
profile. In production engineering one of two parameters is usually used to
define the texture of surfaces. These parameters are the C.L.A. (centre-line
average) roughness value, and the R.M.S. (root mean square) value. The
C. L.A. value is defined as the arithmetic average value of the vertical devia-
tion of the profile from the centre line, and the R.M.S. value as the square
root of the arithmetic mean of the square of this deviation. In mathematical
form they can be written as
• l n
C.L.A. = - L lzil
ni=l

(2.1)

L (zY ]1/2
l n
R.M.S. = [ -
n i=l

where n is the number of points on the centre-line at which the profile devia-
tion zi is measured, (figure 2.6). The centre-line is taken as a line which
divides the profile in such a way that the sums of the enclosed areas above and
below it are equal.
These parameters are seen to be primarily concerned with the relative
departure of the profile in the vertical direction only; they do not provide
any information about the slopes, shapes and sizes of the asperities or about

22
X

Figure 2.6

the frequency and regularity of their occurrence. It is possible, therefore, for

surfaces of widely differing profiles to give the same C. L.A. or R.M.S. values,
(figure 2.7). These single numerical parameters are mainly useful for classi-
fying surfaces of the same type which are produced by the same method.

(a)

(b)

(c)

(d)

(e)

(f)

Figure 2. 7 Various surfaces having the

same C.L.A. value

Lapped surfaces, for example, differing only in the grades of the lapping
compounds to which they have been submitted, will have the same pattern of
roughness and a single parameter will be adequate to characterise them.
For a more rigorous definition of surface profiles, however, the C.L.A. and
R.M.S. parameters are inadequate and more information will be required.
It is important to realise that the choice of parameters for specifying the

23
surface texture will be controlled by the functional requirements of the
surface, that is, by measurements of those features of the surface which are
known to be significant in any given practical situation.
Abbott and Firestone 1 , the founders of profilometry, were concerned with
the wear behaviour of surfaces, and they chose their parameters accordingly.
Using the profile of a surface they constructed its bearing-area curve by
measuring the fraction of the sample length which lies inside the profile
at various positions above the lowest point on the profile, (figure 2.8). By

~- - - - - - ~me
L
~nh- - - - - - i -roo per cen~

Figure 2.8 Method of deriving the bearing area

CUrl'e

dividing the surface into three height-zones, containing the highest 25 per
cent, middle 50 per cent and lowest 25 per cent of the bearing area, they identi-
fied three parameters-the peak, medial and valley occurrences respectively.
It is interesting to find that they were well aware, some thirty-nine years ago,
of the inadequacy of single numerical parameters to define surface textures.
Meyers 4 was concerned with friction between sliding surfaces and suggested
that the slope and curvature of the profile should affect the frictional be-
haviour of surfaces and found strong correlation between slopes and friction.
Recent work on the dry contact of surfaces has shown that this pheno-
menon may be largely explained by the shapes of the peaks and their distri-
bution through the upper decile of the texture. Indeed, in such problems the
properties of the valleys are almost insignificant. For such a situation,
therefore, the required surface parameters must incorporate knowledge of
the number of peaks of any given height level in the upper part of the texture,
a shape factor for the peaks and a statement of the frequency of occurrence of
such peaks in the plane of the surface. This information would, be particularly
important in considering a large range of engineering problems such as
tribological situations, electrical contacts, thermal contacts and the stiffness
of joints created by mating surfaces. Conversely, in the context of stress
concentrations in surfaces under load, it would be anticipated that the shape
and occurrence of the valleys would be the most significant parameters. For
surfaces separated by lubricated films, or those which are to be covered by
layers such as paint, it is apparent that we shall be concerned with the whole
of the texture characteristics.

24
It should be pointed out that the profiles of surfaces can be considered,
in statistical terms, as stationary processes of the random type and therefore
in order to formulate the various parameters discussed earlier, it is necessary
to apply statistical methods in describing the properties of surfaces.

2.4 THE STATISTICAL PROPERTIES OF SURFACES

It was mentioned earlier that surface profiles often reveal both a periodic and
a random component in their geometric variation. Thus the periodicity is
most marked in fine shaped surfaces and in diamond turning, whereas a
considerable degree of randomness is apparent on abraded surfaces such as
those produced by grinding. It has become common practice to break up
the periodic variations into frequency bands such as waviness and roughness.
In a sense it would be desirable not to make such distinctions, but this would
necessitate the evaluation of a sample of the surface which includes all the
frequency variations, and this inevitably leads to a need for the measurement
of the whole of the surface. This latter requirement is unacceptable and the
division into frequency bands is therefore a practical necessity. In what
follows, we shall restrict ourselves to consideration of the higher frequency
spectra, that is, the roughness of the surface.
Before any quantitative statements can be made about surface profiles
a datum line must first be established. For this purpose the centre-line of the
texture offers the most usual choice, although any line parallel to this line
would be satisfactory. Let us first of all consider the properties of surface
profiles in the vertical plane.

2.4.1 The Height Distribution of Surface Textures

Surface textures can be adequately described in terms of the distribution
function of the heights of their profiles. Indeed, Abbott and Firestone's
bearing-area curve is, in statistical terms, the cumulative distribution of the
all-ordinate distribution curve. This can be written as
X

F(z) = J1/J(z) dz
where z refers to the heights of the profile measured from the centre-line and
1/J(z) is the probability density function of the distribution of these heights.
The practical derivation of such distribution curves involves taking
measurements of z 1 , z 2 , etc. at some discrete interval I and summing the
number of ordinates at any given height level, (figure 2.9). In effect this
may be interpreted as converting the continuous analogue signal of the
profile into a discrete digital record taken at intervals I. Self-evidently the

25
1nterval d1s rn but1on
histogram

Figure 2.9 Method of deriving the all-ordinate distribution

distribution curve is then seen to be the best smooth curve drawn through the
histogram produced by such a sampling procedure.
Many surfaces tend to exhibit a normal, gaussian, distribution of texture
heights. Figure 2.10 shows how a gaussian distribution is a reasonable fit

Figure 2.10 Typical distribution curve for a ground

surface

for the histogram obtained for the all-ordinate distribution of a ground

surface. The curve of the gaussian distribution or its density function is given
by
1/J(z) = I/Jo(z)e-z2/2uz

where rr is the standard deviation of the distribution, which is defined in the

same way as the R.M.S. value given by equation 2.1, and rr 2 is the variance.
I/J0(z) can be calculated from the fact that the area of the curve must be equal

26
to the total number of data summed over the chosen scale. The area of the
gaussian curve is

f
ex,

l/lo(z)e-xz/2az = l/lo(z)u(27t)t/2
-oo
Therefore the curve

(2.2)

has unit area and the gaussian distribution curve is usually written in this
standard form. Since this curve encloses unit area, the foregoing procedure
has simply adjusted the vertical scale of the distribution curve to produce a
probability density function, that is the size of any ordinate on this curve
indicates the probability of occurrence of that event in the total population.
It is worth recording that the origin ofthe above curve is located at the centre-
line or the mean ofthe distribution. If we wish to write the curve with reference
to some other point as origin, then
1
1/J(z) = u(27t)t/2

where m is the distance of the mean from the value chosen as origin.
The values of the ordinates 1/J(z) of the gaussian distribution curve and
those of the corresponding areas are found in most books on statistics. The
form of the gaussian distribution necessitates a spread for the events from
- oo to + oo which cannot happen with practical surfaces; in practice the
distribution curve is truncated to finite limits of about ± 3u. Fortunately
about 99.9 per cent of all events lie within this region and consequently the
truncation would lead to negligible error while providing useful simpli-
fication.
It must be emphasised at this stage that although several common surface
preparations produce near-gaussian distributions, many do not. There are
surfaces whose textures show certain departures from the gaussian distri-
bution and it is necessary to define some statistical parameters which are
used for measuring such departures, as well as to consider some other
simple distributions 5 .

Moments, Skewness and Kurtosis

(a) Moments
The nth moment of the distribution curve 1/J(z) about the mean is defined as

f
CJc

Mn = z"l/l(z) dz (2.3)
-oo

27
so that

R.M.S. =a =[_£ '~()

= [2nd moment of t/t(z)]l' 2
d'r
and

C. L.A. =2 Jzt/t(z) dz
0
= twice the I st moment of half t/t(z)
Obviously, the lst moment of the whole of t/t(z) about the mean is zero, and
is in fact the method by which the centre-line of the distribution can be
located.

(i) Gaussian Distribution For a gaussian distribution curve the nth moment
is given by
X

M =
n
1
0"(2rr)l/2
f z"e-z2/2a dz
-oc

If n is odd this vanishes, as it must for any symmetrical curve. If n is even then

M = n'. (Jn (2.4)

n 2"12 (n/2)!
Thus the 2nd moment becomes simply 0" 2 , the variance.

(ii) Rectangular (Uniform) Distribution The rectangular or uniform distribu-

tion implies that at any level there is an equal number of events, (figure 2.11 ).
We can write
t/t(z) = c = constant
so that
'L

Jt/t(z) dz = 1

For practical purposes let us suppose that this distribution is cut off
at the finite limits ±h, and then work out these limits in terms of the standard
deviation O". From the definition of the second moment we have
h

0" 2 = Jz t/t(z) dz
2

-h
h

= J zzc dz = 2c3h3
-h

28
that is

Now, making the area under the curve equal to unity gives
30"2
2h 2h3 = I

or
h = (3)1/20"

and the distribution curve can be written in the standard form

(iii) The Triangular (Linear) Distribution If the number of events varies

linearly with height then the distribution of the events is said to be triangular
or linear, (figure 2.11).

!
I
-t>
l
0 +t>

Gouss,on Uniform Lmear

Figure 2.11 Forms of distribution

Again let us consider the finite limits ±h of such a distribution. For this
we can write
c
t/J(z) = c - - z O<z<h
h
c
= c + -z -h < z < 0
h
such that
h

Jt/l(z) dz = I
-h

29
Now
h

a 2 = J z 1/J(z) dz
2

-h

that is
6a 2
c = /;3

Making the area under the distribution curve equal to unity gives
h x c =1 or h = (6) 1 ' 2 a
The triangular distribution can therefore be written as
1 z
1/J(z) = (6)112a - 6a2 0 < z < (6) 112 a

1 z
= (6)112a + 6a2

(b) Skewness
The skewness is a measure of the departure of a distribution curve from
symmetry. This is defined as
en
J z 1/J(z) dz
3
S = _-_oo.:__---c;-----
(13

3rd moment of 1/J(z)

(J3
(2.5)

For a large class of moderately skewed distribution curves the skewness

can be calculated from an empirical relationship given by

3 (mean - median)
S=-------
(J
(2.6)

where the median represents that value of the variable whose ordinate divides
the area under the distribution curve into two equal parts.
Clearly, symmetrical including gaussian distribution curves have zero
skewness; unsymmetrical curves can have either negative or positive skew-
ness, as shown in figure 2.12.

30
-ve Skew

Figure 2.12
(c) Kurtosis
The kurtosis is a measure of the hump on a distribution curve. This is defined
as
ex,

J z 1/J(z) dz
4

k = ----"(L'------o---
a4
(2.7)
4th moment of 1/J(z)
a4
For a gaussian distribution, using equation 2.4, we find that
I 4x3x2 4
k = a4 X 2 X 2 X 2a = 3

that is a gaussian distribution curve has a kurtosis of 3 which is usually

taken as a standard value for the kurtosis. Curves with values of k less than 3
are called platykurtic and those with k greater than 3 are called lepto-
kurtic, (figure 2.13).

Figure 2.13

31
In plotting distribution curves of events on actual surfaces it has been
shown that this involves drawing smooth curves through the histograms
obtained by summing the frequency of such events at discrete intervals. This
is obviously a fairly tedious procedure and a more useful presentation is to
plot the distributions on probability paper. This has the advantage that the
scales are adjusted such that a truly gaussian distribution degenerates into a
straight line. Figure 2.14 shows the distribution of events on probability
paper for gaussian, rectangular and triangular distribution of events.

Gauss1an d•str•but•on
Tnangular dlstnbut1on
Un1form dtstnbutton

80

01

--o -+
Figure 2.14 Distributions plotted on probability
paper

2.4.2 The Autocorrelation Function of Surface Profiles

The effect of increasing the sampling interval /, although to some extent

affecting the ensuing frequency histogram and distribution curves, can be
most clearly seen by considering the correlation between adjacent ordinates 7 .
This may best be achieved by plotting the variation of the autocorrelation
function, R(/) against the sampling interval /. The autocorrelation function
for a single profile is obtained by delaying the profile relative to itself by some

32
fixed interval, then multiplying the original profile by the delayed one and
averaging the product values over a representative length of the profile.
Thus
R(l) = E[ z(x)z(x + I)] (2.8a)
where E indicates the expected (average) value and z(x) is the height of the
profile at a given coordinate x along the mean line z(x) = 0 and z(x + I) is
the height at an adjacent coordinate (x + I) taken at an interval I from the
previous one. If the value of the ordinates at discrete intervals I is known
this may be interpreted as
J N-1
R(l) = N _ I x~! z(x) x z(x + I) (2.8b)

where N is the total number of ordinates in a sample length L. For a contin-

uous function, equation 2.8b may be more usefully written
L/2

R(l) = lim
L-oc
_!_
L
J z(x) x z(x + I) dx (2.8c)
-L/2

It can easily be seen that when I = 0, R(l) reduces to the variance u 2 or mean
square value of the profile. The autocorrelation function is therefore usually
plotted in its standardised form, r(l), where
r(l) = R(l) = R(l)
R(lo) u2

A typical plot of the autocorrelation function for two different profiles is

shown in figure 2.15. The shape of this function is most useful in revealing

k(~'-­
~;tv¥MA Surface A Surface B

R(()
R(l)

Autocorretotion AutocorrelatiOn
funct1on funct10n

Figure 2.15 Typical surfaces and the resulting autocorrela-

tion functions

33
some of the characteristics of the profile. The general decay of the function
indicates a decrease of correlation as I increases and is an indication of the
random component of the surface profile, while the oscillatory component
of the function indicates any inherent periodicity of the profile. Figure 2.16
shows some typical results for actual machined surfaces where the auto-
correlation functions can be seen to be demonstrating the general features of
the profiles 8 . For this reason it is possible to describe the features of any
surface profile by two characteristics: the height distribution function t/J(z)
and the autocorrelation function r(l).
The main difficulty which may prevent the use of the autocorrelation
function in practical applications lies in the large volume of digital informa-

zoL ~An
so·~UV w~ V~ VV VlJ~
h Shaped surface
C.L A lf;pm

,~
(a)

DI\Af\1\f\ot
V'VV V'VVlJ v

(b)

sL ~ Ground surface

(c l
~o/\ ""LC.LA IOpm

CLA
Superfin1shed
018pm

(d)

Figure 2.16 Examples of engineering surfaces, their

distributions and autocorrelation functions

34
tion which must be obtained from the surface profile for its evaluation.
However, the autocorrelation function may be broken up into two terms
such that the random component of the surface profile may be expressed by an
exponential decay term and the periodic component by a trigonometric
term. For many practical applications, for example the dry hertzian contact
of surfaces, the random component of the surface may be of primary im-
portance, and consequently the problem can be simplified by neglecting the
periodic component of the profile. The surface will then be statistically defined
by an exponential autocorrelation function.
In certain situations the profile may be required to be presented in terms of
its frequency domains. To this end a plot of the power spectra C is useful.
The power spectrum is obtainable from the autocorrelation function using
the expression
c('

C(w) = ~ JR(l) cos wl dl

0

where w = 2rrf, f being the frequency of occurrence. The power spectrum

is in fact the Laplace transform of the autocorrelation function. For the
purposes of this book the use of the power spectra will not be pursued further,
although it clearly has considerable possibilities for the defining of surface
contours.

2.4.3 The Distribution of the Peaks, Valleys, Curvatures and Slopes

Surface profiles might be considered as comprising a certain number of

peaks of varying heights and an equal number of valleys of varying depths.
These features may therefore be assessed and represented by their appro-
priate distribution curves which can be described by the same sort of char-
acteristics as were used previously for the all-ordinate distributions, for
example, aP and av are the standard deviations of the peak and valley
distributions. These distribution curves will obviously involve summing
all the peaks or valleys at any given height level, (figure 2.17).

Figure 2.17 Peak and valley distributions of a surface

35
As with the all-ordinate distributions of surface textures, the peak and
valley distributions often follow the gaussian curve. Figure 2.18 shows how
close to gaussian distributions are the distributions for the peaks and all-
ordinates for a ground surface.

95 ~L
2501-'m

80

50

••

••
• •- ••

0
01 0
0

0 2 4 6 8
He1gh1 above datum (pm)

Figure 2.18 Distributions of an engineering surface

Although the distribution functions of the peaks and valleys provide the
basic information relating to such features of the surface profile, they do not
include any specific statement on the order in which the particular peaks and
valleys occur along the surface; although it will be recognised that any peak
must necessarily have valleys on either side at a lower level than the peak.
Thus the juxtapcsition of the peak and valley distribution implies some limit
on the choice of order for the peaks and valleys. This is readily seen by the
simple example shown in figure 2.19, where the same distribution of peaks
and valleys always occurs but their separation K varies. Where there is no
overlap of these distributions, figure 2.19a, any peak may be associated with

36
. . ---1-
.~
....
~ ~-=
.... __ __
......
-~,
1/1 (v) (a)

Figure 2.19 The effect of the interaction of the peak and

the valley distributions

any two valleys so that the order of occurrence of the peaks along the surface
is in no way defined. Where the distributions overlap, figure 2.19b, it is seen
that in regions A and C the foregoing still applies. However in region B
one cannot associate a valley in the upper part of this region with a peak in
the lower part of the same region, since peaks cannot be lower than their
adjacent valleys. Thus it is seen that the greater the degree of overlap of the
peak and valley distributions the more the order in which the peak and
valley occur along the surface is implied.
Similar distribution curves can be obtained for the curvatures of the peaks
and valleys. This may be done by simple curve-fitting-such as parabaloids
or circular arcs-to the peaks or valleys and assessing the way in which the
curvature varies along the profile. A typical distribution curve for the curva-
ture of the peaks of real surfaces is shown in figure 2.20, together with the
gaussian distribution of the same mean and standard deviation.
The way in which the slope varies along the profile can also be described
in terms of its distribution curve. The derivation of this curve will involve
measuring the slope at each point along the surface profile. The distribution
curve may then be characterised by its standard deviation d- (figure 2.21 ).

37
Radius of asperity peok (mm)

Figure 2.20 Distribution of the asperity peak radii

Meon Mean
ord1note slope

-z -i
All-ordinate distnbut1on All-slope dlslnbut,on

Figure 2.21 Derivation of the slope distribution of

surfaces

2.5 MEASUREMENT OF SURFACE PARAMETERS

Abbott and Firestone measured their bearing-area curve manually using an

optical comparator. For each height considered a line was drawn along the
profile and the fraction of the sample length embraced by the profile at that
height was measured. However, the practical derivation of such curves, in

38
the form of all-ordinate distributions of surface profiles, and indeed of the
distribution curves of all surface parameters, peaks, valleys, curvatures, etc.,
involves the measurement of a sequence of ordinates along the profile.
Such a process can be very tedious since a considerable amount of digital
information is required, but with the aid of modern electronics, the stylus
instrument can be modified to make the operation fairly easy. The usual
practice is to feed the analogue electrical signal from the Talysurf instrument
through a digital voltmeter to a high-speed paper-tape punch to get a per-
manent numerical record of a sequence of profile ordinates at some discrete
sampling interval. The data on the paper tape can then be analysed by a
digital computer for the required information. The height distributions for
the peaks and valleys and their curvatures can be derived by using the
technique of three-point analysis. A peak may be defined when the middle of
three successive ordinates is higher than those on either side. Similarly, the
peak curvatures can be obtained by fitting a curve to the profile using the
three-point analysis.

I. E. J. Abbot and F. A. Firestone. Specifying surface quality. Mech. Engng, 55, (1933),
569.
2. J. B. P. Williamson. The microtopography of surfaces. Proc. lnstn mech. Engrs,
182 (3K), (1967-8), 21.
3. R. H. Reason. The trend of surface measurement. J. lnstn. Prod. Engrs, (May, 1954).
4. N. 0. Meyers. Characterization of surface roughness. Wear, 5, (1962), 182.
5. J. Halling. The specification of surface quality-quo vadis?. Prod. Engr, 51, No. 5,
(1972), p. 171. •
6. J. Halling and M. El-Refaie. A statistical model for engineering surfaces. Tribo/ogy
Convention, Paper No. C60/71, Instn. mech. Engrs., (May, 1971).
7. D. J. Whitehouse and J. F. Archard. The properties of random surfaces of signifi-
cance in their contact. Proc. R. Soc., 316A, (1970), 97.
8. J. Peklenik. New developments in surface characterization and measurements by
means of random process analysis. Proc. lnstn mech. Engrs, 182 (3K), (1967-8), 108.

39
3
Contact of Surfaces
3.1 INTRODUCTION

It is clear that any study of tribology must incorporate a detailed under-

standing of the mechanics of contact of solid bodies. This involves an under-
standing of the nature of the associated deformations and the stresses induced
by any applied loading to bodies of a wide variety of geometric shapes. In
particular we are concerned not only with the deformation and stresses at the
surfaces of solids but also throughout the depth of the surface layers. Any load
inducing a deformation of solids may readily be resolved into a normal and a
tangential component, and it is generally convenient to consider these two
influences separately with respect to the stresses and deformation which
they induce, and then by superimposing the two obtain the total effect. In
such cases the principle of superposition is acceptable since the systems are
essentially statically determinate.
Solid materials subjected to loads deform in either an elastic or plastic
manner; the former deformation being characterised by simple linear rela-
tions between stress and strain and being basically reversible. With plastic
deformation the stress-strain relations are more complex and some deforma-
tion persists even after removal of the load. In most contact situations we
find a mixture of both elastic and plastic deformations. Thus the loads applied
to solids in contact may induce a general elastic behaviour in the bulk of the
solid bodies, but since the actual contact must occur at the tips of the surface
asperities these may be subjected to localised plastic deformation at their tips.
The amount of plastic/elastic deformation must obviously depend on the
value of the applied load and the degree of plastic deformation increases as

40
the load is increased. Thus, in metal-working processes when the nominal
contact pressures are exceedingly large, the amount of plastic deformation of
the surface is increased.
In much of the following we shall be considering the deformation patterns
induced by loads applied to cylinders and spheres. This study is valuable for
two reasons
(a) Many engineering contacts are concerned with the contact of bodies
defined by circular arcs such as wheels on tracks, rolling element type
bearings, gear-teeth contacts, many variable-speed drives and belt
and rope drives.
(b) All solid bodies have surface asperities which may be considered as
very small spherically shaped protuberances. Thus, the contact of
essentially flat bodies reduces to the study of an array of roughly spheri-
cal contacts where we shall concentrate on the deformation of the tips
of such spherical asperities.
Finally, in sliding contacts we are all aware that the work done in over-
coming any friction is finally manifested as a heat release; try sliding down a
rope and this will become painfully obvious. We shall therefore be very
interested in the nature of this heat release, the temperatures which are
induced and the distribution of these temperatures throughout the bulk of
the contacting solids. In sliding down a rope the surface temperature of our
skin is clearly much in excess of the t~mperau of the bulk of our hands.
From the foregoing it is clear that a detailed study of surface contacts will
necessitate a relatively detailed understanding of such topics as elastic and
plastic deformations and the nature of heat conduction due to moving
heat sources. Clearly it is not possible to develop all these arguments in depth
in a book of this type, so that recourse will often be made to physical argu-
ments where the essential conclusions will be stated. In each case the reader
will be able to find extensive verification of such results in standard works on
elasticity, plasticity and heat conduction. It is not the purpose of this book to
develop such arguments, but only to demonstrate the value of such know-
ledge in the particular problems with which we are concerned.

3.2 STRESS DISTRIBUTION DUE TO LOADING

In almost all our studies we shall be concerned with effects within the outer-
most layers of the surface (typically within the outermost millimetre or so of
the surface). The effects at several centimetres below the surface are of only
secondary importance so that we may often treat the surfaces, from a physical
standpoint, as though they represent the surfaces of bodies of essentially
infinite depths, that is, they may be considered as semi-infinite bodies. This
device enables us to concentrate on the details of the surface contact of solids
rather than considering their overall geometric shape and thereby leads to
considerable mathematical simplification.

41
Consider a single normal line load P per unit length in the plane xz and
applied at a point 0' defined by the coordinates (t:, 0) on the surface (z = 0)
of a semi-infinite solid and having the same value for all values of y, see
figure 3.1 a. The elastic stress field in the plane xz is readily obtained. Consider-
ing a unit length in they direction, the radial stress (J, will be given by 1

2P
(J, = - -cos 0, (3.1)
nr
the tangential stress (J 9 and the shearing stress r, 0 being equal to zero.
This represents a state of simple radial compressive stress, the stress
increasing with decreasing radius r and decreasing angle 0. The use of the
two-dimensional Mohr's circle of stress, figure 3.1 b, for these stresses gives

z
(a)

(b)

(b)

Figure 3.1 Stress distribution due to a line load acting on a

semi-infinite body

42
us the resulting cartesian stresses with respect to 0'
ux = i (1 -cos 20) = u, sin 2 0

2
= _ 2P (sin 0 cos 0) = 2P [ zx 2 J
1t r 1t (xz + z2f
u. = i (1 + cos 20) = u, cos 0 2

=- :
2
t<x2 z;xz2)2] (3.2)

or with respect to the original origin 0 as

(3.3)

A similar approach may be applied to obtain the stresses due to a single

tangential line load T acting at 0', (figure 3.2) where
2T
u =--cos O'
r nr
(3.4)

and

(3.5)

If we make T = p.P, where J.l is the appropriate coefficient of friction and

add the stress components due to P and T at any point (x, y) we will clearly
43
T
o·
8' X

Figure 3.2 Stress due

to a tangential line load
acting on a semi-infinite
body
have the stress distribution arising in a simple frictional contact. The solu-
tion, however, suffers from one serious drawback. If we examine equations 3.1
and 3.4 we find that at 0' (r = 0) the stresses are infinite and such a situation
is obviously inadmissible. This arises from our basic assumption that the
load acts at a single point, that is, over zero contact area. In real cases we
must therefore always have some finite area of contact and this clearly
changes our initial problem. Fortunately, we are still able to use our original
solution for the new situation.
Consider a uniformly distributed load giving rise to a contact pressure p
over a region 0 to a on the surface (z = 0) of a semi-infinite solid, figure 3.3.
Taking a length along the y direction equal to unity, it will be recognised
that the total load P is given by
a

P = J pdx = pa
0

If we consider a vanishingly small load p dt: at some point defined by the

coordinates (t:, 0) we may obtain the stress at any point (X, Z) due to this
Pa

0
X

.- --j 1-
d£_3 z

x-.-
X

Figure 3.3

44
load using equations 3.3. In this case P will be replaced by p de. The total stress
at a point X, Z due to the distributed load P is then clearly obtained by the
summation of the effects of all the p de loads acting at different values of e
from 0 to a, or mathematically speaking

(3.6)

rxz = -1[
2p fa [ Z 2 (X - e)
[(X - e)2 + z2]2
J de
0
In a similar way if we consider a tangential load T = JJ.P distributed over
the region 0 to a, then at every point it follows that t dx = Jl.P dx (figure 3.4)
and
a a

T = f
0
t dx = f
0
Jl.P dx = JJ.P

1----- X ----4
z

Figure 3.4

By using equations 3.5 for each elemental tangential load t de acting on

element de (0, e) we can obtain the stresses at any point (X, Z) due to the total
distributed load T, thus

(3.7)

•xz = - n2t fa [ [(X_Z(Xe)2- +zip

e) Jde 2

0
45
For a sliding contact subjected to a normal load P uniformly distributed
over the contact 0 to a, the total stresses are the sum of the stresses given by
equations 3.6 and 3.7 above.
It is clear that the use of the basic normal and tangential point load solu-
tions may be used to obtain the resultant stress distribution for any type of
load distribution over the contact region. All of the preceding solutions have
assumed elastic behaviour of the bodies, but we have already anticipated that
we are also interested in the possible plastic effects in such situations. The
simplest criterion for the onset of plastic deformation assumes that this occurs
when the maximum shear stress reaches the critical shear stress for the
material k, where k = Y /2, Y being the tensile yield stress. For the cases
considered above, where plane strain conditions apply, the maximum shear
stresses always occur in the xz plane. The maximum shear stress in this plane
is simply the radius of the Mohr's circle of stress, (figure 3.1 b), that is

a, p (}
r max =2=- 1tr cos

If we consider a circle of diameter b drawn in the manner shown in figure 3.5a,

we find that r = b cos (} and

that is, the stress remains constant at all points on the circle. It is therefore
useful to plot the stress distribution as isochromatics or lines of constant
rmax• and it is then possible to determine the location at which rmax will reach
its limiting value of k, that is, the location of the onset of plastic deformation.
These plots are useful since they also indicate the pattern of isochromatics
obtained in photoelastic stress analysis. For a point normal load, not strictly
achieved in practice, and a uniformly distributed normal load, calculations of
rmax yield the pattern of isochromatics shown in figures 3.5b and c. It is seen
that in both cases the material will first reach a yield condition at the surface
where increasing load gives rmax = k, the yield shear stress of the material.
p p p

(o) (b) (c)

F~qure 3.5 Patterns of lines of constant maximum shear

stress ( isochromatics)

46
3.3 DISPLACEMENTS DUE TO LOADING
Having obtained the stress distribution we can now obtain the displacements
in a solid using the usual equations which relate the strain e and the cor-
responding displacements. Thus for a single normal load P acting at 0',
figure 3.la, the horizontal and vertical displacements u and w are given by

ou l
- = e = - (u - vu8 ) = - -cos 0
2P
or ' E ' rrr E

u ow l
- + -- = e = - (u
2P
vu) = v - cos(}
0 8 -
r rt!O E ' rrrE

au c'w w l
r o~
c
+ -;;-
ur
- -r = 'Yr8 = -G r,o = 0

To solve these equations we require a knowledge of the boundary conditions.

For this we assume that points on the z-axis, that is, at 0 = 0, have no lateral
displacements and that at a point on the z-axis at a distance b from the origin
there is no vertical displacement. Clearly we are interested in the displace-
ments occurring at the boundary, z = 0, of the solid, thus by putting
0 = ± rr/2 in the solution of the above equations it can be shown that the
horizontal displacement is given by
(l - v)P
(u)z=O = - 2£ (3.8)

This indicates that at all points on the boundary of the solid there is a constant
displacement directed toward the origin. We can also find that the vertical
displacement of a point on the boundary z = 0 at a distance x from the
origin is given by
2P b (! + v)P
(w)z=o = ~E log~- rr£ (3.9)

Here we also find that at the point of load application (x = 0) the vertical
displacement becomes infinitely large. As mentioned earlier, this is due to our
assumption of a point load, but in practice the load is usually distributed over
a finite area. In this case if the load is distributed over the region 0 to a,
figure 3.3, giving rise to a contact pressure p, the vertical displacement at any
point (X, 0) produced by an element of load p de at a distance c from point
0 is known from equation 3.9 by substituting p d<: for P and (X - c) for x
so that the total displacement at point (X, 0) is given by
a a

(w)-= 0 = 2
•
-
rrE
Jp l oX gb- -c - d c(I-rrE+- -v) Jpdc (3.10)
0 0

In the foregoing we have concerned ourselves with problems of a two-

dimensional nature where the deforming solid is considered to be subjected

47
to plane strain conditions. It is clear that the three-dimensional analogues
of these problems necessitate a more complex treatment, however, the results
are of great importance and are often met in tribological situations. For the
purposes of this book it will suffice to state these results, detailed treatments
can be found in standard books on elasticity. Thus, for a point normal load
P acting on a semi-infinite solid, the horizontal and vertical displacements
along the boundary z = 0 at a distance x from the point of load application
are found to be given by
(I - 2v)(l + v)P
(u)z = o = - 2rr£x (3.11)

(I - v2)P
(w)_=o = EX (3.12)
- 1r

For a load distributed over a region of the boundary, giving rise to a pressure
p acting on an element of area dA taken at some distance x from a point,
the vertical displacement at that point is given by
_ (I - v2 )
(w)z=O- - - -
f p dA (3.13)
rr£ x
The integral in the above equation is, in fact, an elliptic integral.

3.4 HERTZ IAN CONTACTS

As mentioned earlier we are particularly interested in the problems of contact

between bodies whose geometry is defined by circular arcs. This problem was
first solved by Hertz for elastic contact and is generally referred to as hertzian
contact.
Consider the contact of two identical cylinders under conditions of plane
strain. From arguments of symmetry we can see that the zone of contact will
be created by compression of the cylinders to generate a straight line, that is,
produce a plane contact zone, figure 3.6a. Although this is no longer strictly
true for a cylinder in contact with a plane, the error is such as to be negligible
and for this situation a plane contact zone may be assumed. One feature of
such contacts becomes immediately apparent, in that as we increase the load
the width of the contact zone will increase and, by noting that the deforma-
tion at the centre of the zone is larger than that at the extremities, we would
expect a contact pressure which is no longer constant. This problem is there-
fore more complicated than the preceding cases and, before considering the
stress distribution, we need to define both the contact pressure distribution
and the actual size of the contact zone for any given applied load. Detailed
treatment of this problem is found in books on elasticity, but for our pur-
poses we shall use a simpler physical argument.
For two identical elastic cylinders in contact under a normal load P

48
p

1--- 2o ------.j (b)

(a)

Figure 3.6 Pressure distribution for the contact of two

cylinders

per unit axial length, let the resulting plane contact zone have a width of 2a,
figure 3.6b. Since the normal deformation at the centre of the contact zone is
greater than at the extremities, it is not surprising that the actual contact
pressure distribution p has the form 1
p = 2P
n:a
(1 - a~)12 2
(3.14)

Without this information it is physically apparent that the stresses in such

a system would be such that

stress oc ( ~)
Considering the deformation, we note that increasing the load would increase
a and thereby increase the strain so that we would expect the non-dimensional
strain to be related by

strain oc ( i)
where R is the radius of the cylinder. From these relations for stress and
strain we have
2 PR
or a oc-
E
The actual solution for this case is in fact 1

2 4PR(l - v2 )
a=----- (3.15)
nE
49
which clearly substantiates the above simplified argument. The solution
defined by equations 3.14 and 3.15 is almost true for other than identical
cylinders, that is, plane contact geometries. Provided the angle subtended
by the contact width at the centre of the cylinder is less than 30", the results
may be reasonably used for other contact geometries, figure 3.7, by using E'
and R' where
1 I - vi 1 - v~
E'=~+
I 2

and

whence
4PR'
az =~-
rtE'

For the case of a cylinder on a plane, the radius of the plane is taken as
infinity, thereby R' becomes the radius of the cylinder only and for concave
curvatures the radius is taken as negative. It is also worthwhile noting that
when E --+ ·X, the solids become rigid leading to a single point contact where
a--+ 0.

G
(a) (b) (c)

Figure 3.7

3.4.1 Stress Distribution in Hertzian Contacts

Earlier in this chapter we have seen that the onset of plastic deformation may
be associated with the maximum shear stress reaching a critical value k. We
therefore proceed to study the distribution of the maximum shear stress for a
body loaded by a pressure distribution given by equation 3.14, acting over the
contact zone -a to +a. Using equations 3.6 for elemental loads p dt: and
integrating for the actual distribution of P will give the cartesian stress distri-

50
bution within the body. Thus

-a

(3.16)

Again, the maximum shear stress for plane strain conditions is given by the
radius of the Mohr stress circle, see figure 3.1 b, that is

rmax = [( -"'~
(J - (J )2 + Jli2
r;z (3.17)

where (JX' (Jz and rxz are defined by equations 3.16. Equation 3.17 therefore
defines the values of rmax at all points. The evaluation of this equation will
enable us to draw the isochromatics and obtain the somewhat surprising
result shown in figure 3.8. We note immediately that the greatest value of
rmax occurs below the surface at a distance of 0.67a. Furthermore, we find
that as the load is increased rmax at this point will reach the value k when the
maximum pressure at the centre of the contact zone Po is 3.1k. This is again a
surprising result since if we had simple compression conditions we would
expect the surface material to yield when Po reaches a value of 2k. This does
not oceur because the surface elements are subjected to compressive stresses
in all three orthogonal directions, allowing Po to be greater than 2k without
producing yield. This is equivalent to saying that the hydrostatic component
of stress at any point cannot contribute to the plastic deformation of the
material at that point. This is. a very important result since it means that
contact pressures in excess of the yield value for the material do not result in
plastic deformation, so that higher loads than might have been expected can
be carried elastically with hertzian type contacts. Moreover, it will be recog-
nised that even when yielding has taken place below the surface, very little
plastic deformation can occur because the plastic zone is constrained by
elastic material on all sides.

51
Lines of
&.-o~ consranl
rmax

Figure 3.8 Actual isochromatics obtained for the

contact of a cylinder and a plane due to norma/load
alone

As the load is increased further, the plastic zone will increase in size and
ultimately spread to the surface of the body. Plastic flow may then occur
fairly readily and the cylinder will indent the surface of the body. This will
happen when the mean contact pressure Pm is about 6k, that is, more than
twice the contact pressure at which initial yield occurred 2 . The mean pressure
under these conditions is essentially the indentation hardness value of the
material, H, and this is why for metals we find that
H::::: 6k::::: 3Y
where Y is the uniaxial tensile yield strength of the material; see figure 3.9.

Fully plaSIIC
behaviour

PlasiiC behav1our
cons1ra1ned by
surroundmg elasiiC malenal

Purely elasiiC
behaviour

Figure 3.9

52
We have as yet only considered normal loads applied to hertzian contacts
and it is apposite to ask what happens in the presence of both normal and
tangential loads. The stress field due to a tangential load !J.P may readily be
obtained by the method discussed earlier, since at all points it is clear that the
tangential traction t = !J.P· Combining the stress distribution due to the
normal and tangential loads and calculating the values of rmax leads to the
isochromatic pattern shown in figure 3.10. It is seen that the location of the

Lines of
constant
Tmox

Figure 3.10 Actual isochromatics obtained for the

contact of a cylinder and a plane due to combined
normal and tangential loads where T = 0·5!J.P

greatest value of the maximum shear stress is now much nearer to the surface,
thus the plastic deformation can take place more readily than in the previous
case. In other words, macroscopic plastic deformation is facilitated by the
presence of such friction tractions.

3.4.2 Conditions When T ~ !J.P

In many practical situations contacting bodies are subjected to tangential
loads less than !J.P so that macroscopic sliding does not occur. This happens in
situations where friction is used as the mechanism for preventing sl;p
between mating components, for example, nuts and bolts, interference
fits and friction drives such as clutches. Most textbooks explain this situation
by allowing the coefficient of friction to increase from zero to a limiting value
at which slip occurs. This is clearly inadmissible since it makes what should
be a physical constant into a variable. In what follows we show that this
assumption is not necessary if we recognise that the contacting bodies are
deformable rather than rigid. The mechanism will be explained by consider-
ing a cylinder pressed against a plane and subjected to a tangential load less
than !J.P.

53
We shall start by assuming the answer to our problem and subsequently
justify its validity. What happens is that within the contact zone there exists
a central area in which no slip occurs, while at the two extremities a small
degree of slip takes place, see figure 3.11. The coexistence of a zone of sticking
and zones of microslip is possible because of the deformable nature of the
materials in contact and the deformation pattern being such as to allow slip
at the extremities of the contact zone. As the value ofT increases, the areas of
slip increase until, when T = pP, they meet at the centre of the contact

(a)

(b) (c)

~)
[ 0:
\ I
I I
®
Contocf
zone
1
1
fP
Ip
Contact
zone
[CD ,
1
~
® 1-1 --1-L-- ®I
1
Comp ~
1e
~
1
Comp Comp
le
~
'
~Tensil
I I I I I I I I

~
1eA' 1
Comp ~Camp Tensil~ ,....__,
:eA'zl Comp

; BOdy I :

~l•p
~ Body 2

Figure 3.11

and macroslip takes place throughout the contact zone. With this model it is
possible for J1 to have a constant value wherever slip occurs, that is, within
the slip regions t = JlP while within the stick region t < pp. Since T is the
integral oft over the contact zone this can be seen from figure 3.11 to satisfy
the requirements of the problem with J1 always having a constant value.
For the case when T = pP, the distribution of normal pressure p and
tangential traction t = JlP are as shown in figure 3.11 b and 3.11 c. Increasing
the normal load induces equal compression strains ex in both bodies so that

54
no slip occurs due to this effect. With the tangential load, however, we see that
since these must be opposite in direction on the two bodies they will cause
the patterns of strain shown in figure 3.11 c, so that slip must occur every-
where throughout the contact zone. For T < JLP we therefore see that the
central stick areas must have zero strain in both bodies and this consequently
defines the type of distribution for the tangential tractions.
The distribution of tangential traction which results in zero strain over the
central stick area is illustrated by the argument shown in figure 3.12. Con-
sider a tangential load T = JLP which results in the traction distribution

Col

Slope
(b)

(c l
I
I
K:±=:k
I I I
I
I
r"

(d)

~
I I I I I
Slope

Cel I I~ I I I I
T=Ti-T"

I I I I I
I I I I

(f)

~ I
Slip
I SliCk I
Slip
I
ex= e; +e;

-a -a a a

Figure 3.12

shown in figure 3.12a. The strain e~ resulting from such a distribution is

shown in Figure 3.l2b. It will be noted that this strain follows a linear law
within the width of the contact zone and that the slope is, from -a to +a,
e~ = J1X/2R. Now consider a similar shape of tangential traction distribution,
T", in the opposite sense applied over the region -a. to +a.. The strains, e~,
resulting from such a distribution are as shown. and it will be seen that the
slope of the strain distribution is exactly opposite in sign, but of the same
magnitude, as that due to T'. Thus by adding the strains e~ and e~ one

55
obtains a zero strain over the region - rx to + rx and a tangential traction in
this region where t < llP· In the remainder of the contact zone one obtains
t = llP and strains which are non-zero and consequently area of slip, figures
3.12e and 3.12f.
The above arguments show that even when no macroscopic motion occurs
some degree of microslip exists when T < 11P and this gives rise to a mech-
anism known as fretting. For more complicated contact geometries these
arguments are still qualitatively correct and microslip will occur at the
extremities of the contact zone.

3.4.3 General Three -Dimensional Contact

For simplicity we have so far concerned ourselves with two-dimensional

hertzian contacts. In many practical situations we must be able to deal with
the more complicated three-dimensional problems and the following out-
lines the main results of such analysis. In general, the patterns of behaviour
are analogous, but some of our preceding formulae must be modified.
If we subject two identical spheres to a normal load N, the area of contact
will clearly be a plane circle of radius a and the pressure distribution is
hemispherical in form, figure 3.13, and is given by 1

p= 3N ( 1 (3.18)
2rra 2
The value of a is given by

a= (3NR)
8£'
n 1
(3.19)

Contact
zone

Figure 3.13 The pressure distribution due to

the contact of spheres

56
We can recognise that this is correct by applying the simple analysis used
previously.
Although the contact of two dissimilar spheres does not result in a plane
contact area, the results of equations 3.18 and 3.19 still hold with substantial
accuracy, but in this case we have

a= (3NR')1/3 (3.20)
4E'
where R' is related to the radii of the spheres R 1 and R 2 by
1 1 1
R'=p:+-R
1 2

For the contact of a sphere and a plane, R' is merely the radius of the sphere.
A more general approach to the problem is to consider the contact of
two bodies 1 and 2 whose contact geometry is defined by the principal radii
of curvature of each body in two orthogonal planes, figure 3.14. The contact

I
I I
I ..L....l.
rz4.2b
-I 2a 1-r
Plane 1 Plane 2

(R12 radous of body 1 in plene 2 ercl

Figure 3.14

geometry is now clearly elliptical in shape and the contact pressure distri-
bution is given by
p= 3N (1 - x2 - z2)1/2 (3.21)
21tab a2 b2
The size of the contact ellipse is defined by the semi-major and semi-minor
axes a and b which are given by
3N ] 1' 3
a = ka[ 4E'(A + B)

h=
r 3N ]1'3
kl4E'(A + B)
(3.22)

57
where ka and kb are constants which depend on the values of the principal
curvatures of the contacting bodies and on the angle </> between the normal
planes which contain these curvatures. If we denote the principal radii of
curvature of body 1 by R 11 and R 12 and those of body 2 by R 21 and R 22 ,
then the constants A and B may be defined by

1 1 )2 I 1 )2
B- A = 21 [( R,, - R,2 + ( R2, - R22

+ 2( - 1 - - 1 )(- 1 - - 1 ) cos 2</> ]1/2 (3.23)

R,, R,z Rz, R22

A+ B = !(_1_ + _1_ + _1_ + _1_)

2 R 11 R 12 R 21 R 22
In these equations any concave curvature is taken as negative. The coeffi-
cients ka and kb in equations 3.22 are numbers which depend on the ratio
(B - A)/(A + B) and these coefficients can be obtained by introducing an
auxiliary angle y defined by
B-A
cosy= A+ B

Thus, using equations 3.23, the value of y can be easily determined. To

evaluate the values of ka and kb corresponding to a certain value of y we
require complicated numerical calculations involving elliptic integrals. The
results of such calculations are given in figure 3.15.

20
\ ~
~

10
"' ~
!'-- ...............
..--- __..
-- _:::- ~

40 50 60 70 80
r•
Figure 3.15

58
When dealing with such complicated geometries it is clear that our assump-
tion of a plane area of contact will no longer be true. As mentioned earlier,
while the pressure distribution and the size of the contact as determined from
hertzian theory are still substantially correct, we sometimes need to know the
actual shapes of such contacts. For materials having the same elastic proper-
ties it is sufficient to assume that the deformed surface, which has some com-
mon radius Rc, is about midway between the two original surfaces, figure
3.16. Thus the value of the common radius of curvature is given by 1

R =. 2RIR2 (3.24)
c Rt - R2

It is clear that for two identical spheres the above equation gives the expected
result of a plane contact area. Where concave curvatures occur, the radius is
taken as negative.
I
I
I
/
., /

Figure 3.16

Subsequently, we shall find that it is necessary to define the normal approach

of a sphere due to the application of normal load and the consequent deforma-
tion. Consider the contact of a sphere and a plane, figure 3.17. It is easily
seen that the separation u of the surfaces at a distance r from the centre of the
contact zone is given by
u = R - (R2 - r2)tt2

If r is small compared to R then

r2
U=- (3.25)
2R

59
----r---···._··...... ~· ....... •

(a)
~a
(b)

Figure 3.17 The elastic contact between a sphere and a plane

The normal approach is defined as the distance which points on the two
bodies remote from the deformation zone move together on application of a
normal load. This arises from the flattening and general displacement of the
surface within the deformation region. If a is the radius of the contact zone
and w is the displacement of the sphere at the boundary of this zone then the
normal approach c5 will be given by

(3.26)

Obviously, at the centre of the contact zone, c5 is given by the degree of deform-
ation and it is therefore reasonable to assume that the normal approach will
be proportional to the flattening of the sphere. In other words
a2
c) oc-
R
From the results given by equation 3.20 we have

aoc - (NR)t/3
E'
so that
c)
N2
oc ( - -
)t/3
E'2R
The exact results show that

or
(3.27)

60
Combining equations 3.20 and 3.27, the area of contact, A, will be given by
A = rra 2 = rrRJ (3.28)
It will be noted that the relation given by equation 3.28 indicates that the
surface outside the contact region is displaced in such a way that the actual
area of contact is only one half of the geometrical area, which is clearly
equal to 7tiX 2 = 2rrRJ.

3.5 THE CONTACT OF ROUGH SURFACES

Although in general all surfaces have roughness, we shall find some simpli-
fication if we consider the contact of a single rough surface with a perfectly
smooth surface. The results from such an argument are then reasonably
indicative of the effects to be expected from real surfaces. Moreover, the
problem will be simplified further by introducing a theoretical model for the
rough surface in which the asperities are considered as spherical segments
so that their elastic deformation characteristics may be defined by the
hertzian theory. We shall also assume that there is no interaction between
separate asperities, that is, the displacement due to a load on one asperity
does not affect the heights of the neighbouring asperities.
Consider a surface of unit nominal area to consist of an array of identical
spherical asperities all of the same height z with respect to some reference
plane X X', figure 3.18. As the smooth surface approaches due to the applica-
tion of load we see that the normal approach will be given by (z - d), where

Figure 3.18 Contact between a smooth plane and a idealised

rough surface

d is the current separation between the smooth surface and the reference
plane. Clearly, each asperity is deformed equally and carries the same load
N i so that for '1 asperities per unit area the total load N will be ·equal to
IJNi. For each asperity, the load Ni and the area of contact Ai are known
from the hertzian theory, see equations 3.27 and 3.28. Thus if f3 is the asperity
radius we have

and
Ai = rrfJ(z - d)

61
and the total load will be given by

N = 1t~E'f32(/JY 12

that is, the load is related to the total real area of contact A = t~A; by

(3.29)

This result indicates that the real area of contact is related to the two-thirds
power of the load, when the deformation is elastic.
If the loads are such that the asperities are deformed plastically under a
constant flow pressure H, which is closely related to the hardness, we assume
that the displaced material moves vertically down and does not spread hori-
zontally so that the area of contact A' will be equal to the geometrical area
2nf3t5. The individual load N; will then be given by

N; = H A; = 2Hnf3(z - d)
Thus
N' = t~N; = t~HA; = HA' = 2HA (3.30)

that is, the real area of contact is linearly related to the load.
It must be pointed out at this stage that the contact of rough surfaces
should be expected to give a linear relationship between real area of contact
and load, a result which is basic to the laws of friction (chapter 4). From our
simple model of rough surface contact we see that while a plastic mode of
asperity deformation gives this linear relationship, the elastic mode does not.
This is due to our simple and hence unrealistic model of the rough surface.
Later we shall find that the proportionality between load and real contact
area can in fact be obtained with an elastic mode of deformation when we
consider a more realistic surface model.
It is well known that on real surfaces the asperities have different heights
indicated by a probability distribution of their peak heights. We must there-
fore modify our previous surface model accordingly and the analysis of its
contact must now include a probability statement as to the number of
asperities in contact 3. If the separation between the smooth surface and
reference plane is d, then there will be contact at any asperity whose height
was originally greater than d, figure 3.19. If cf>(z) is the probability density
of the asperily peak height distribution then the probability that a particular
asperity has a height between z and z + dz above the reference plane will be
c/>(z) dz. Thus, the probability of contact for any asperity of height z is

prob(z > d) = f"'

d
cf>(z) dz

62
z
Smooth surface

x'

Otstnbutton
of peak hetghts ¢ (zl

F(qure 3.19 Contact between a smooth plane and a rough

surface

If we consider a unit nominal area of the surfaces containing 17 asperities,

the number of contacts n will be given by
oc

n = 11 Jcj>(z) dz (3.31)
d

Since the normal approach is (z - d) for any asperity and Ni and Ai are known
from equations 3.27 and 3.28, the total area of contact and the expected load
will be given by
x;

A = 1t17[J J(z - d)cj>(z) dz (3.32)

d

and

f
oc

N = 111Pl!2E' (z- d)3/2cj>(z) dz (3.33)

d

It is convenient and usual to express these equations in terms of standardised

variables by putting h = dja and s = zja, a being the standard deviation
of the peak height distribution of the surface. Thus
n = 17F 0 (h)
A= 1t17{JaF 1(h)
N = 111P 1' 2a 312 E'F 312 (h)
where
"
Fm(h) = f (s - hrcf>*(s) ds
h

cf>*(s) being the probability density standardised by scaling it to give a unit

standard deviation.

63
Using these equations one may evaluate the total real area, load and
number of contact spots for any given height distribution. Experimental
confirmation of the validity of this method has recently been demonstrated 4 .
An interesting case arises where such a distribution is exponential, that is

In this case we have

so that
n = '1e-h
A = 1t1Jf3ue-h
N = n•'z'1f3t12u312E'e-h
These equations give
N = C 1A and N = C2 n
where C 1 and C 2 are constants of the system. We find, therefore, that even
though the asperities are deforming elastically, there is exact linearity
between the load and the real area of contact.
For other distributions of asperity heights such a simple relationship will
not apply, but for distributions approaching an exponential shape it will be
substantially true. For many practical surfaces the distribution of asperity
peak heights is near to a gaussian shape (see chapter 2) and this itself is
near enough to exponential at the outermost decile of the distribution for the
above result to be valid.
Where the asperities obey a plastic deformation law, equations 3.32 and
3.33 are modified to become

f
00

A' = 2rt'1fJ (z - d)¢(z) dz (3.34)

d

N' = 2rt1Jf3H J(z - d)¢(z) dz (3.35)

d

We see immediately that the load is linearly related to the real area of contact
by N' = H A' and this result is totally independent of the height distribution
¢(z), see equation 3.30.
We have so far based our analysis of surface contact on a theoretical model
of the rough surface. An alternative approach to the problem is to apply the
concept of profilometry using the surface bearing-area curve discussed in
chapter 2. In the absence of asperity interaction, the bearing-area curve
provides a direct method for determining the area of contact at any given

64
normal approach. Thus, if the bearing-area curve or the all-ordinate distri-
bution curve is denoted by t/l(z) and the current separation between the
smooth surface and the reference plane is d, then for a unit nominal surface
area the real area of contact will be given by

f
00

A = t/f(z) dz (3.36)
d

so that for an ideal plastic deformation of the surface, the total load will be
given by
00

N = H J t/f(z) dz (3.37)
d

We may summarise the foregoing by saying that the relationship between

the real area of contact and the load will be dependent on both the mode of
deformation and the distribution of the surface profile. When the asperities
deform plastically, the load is linearly related to the real area of contact for
any distribution of asperity heights. When the asperities deform elastically,
the linearity between load and real area of contact occurs only where the
distribution approaches an exponential form and this is very often true for
many practical surfaces. These results will be found to be of considerable
significance when considering such effects as friction and wear.

3.6 CRITERION OF DEFORMATION MODE

It will of course be recognised that in most practical situations the higher
asperities could be plastically deformed while the lower contacting asperities
are still elastic. Thus we obtain a mixed plastic-elastic system in most real
contacts, where the greater the load and hence the normal approach, the
greater the number of plastic contacts. We therefore see that the normal
approach will in a sense be an indicator of the degree of plasticity which
exists. Referring to equations 3.27 and 3.28 we find that for an elastic asperity
contact the mean pressure Prn is given by
4E'b'i2
Prn = 3rt{Jl/2
or
bl/2 = 31t /3 Prn
1/2
(3.38)
4£'

For a spherical contact we know that the transmission from purely elastic to
completely plastic behaviour occurs over a range of loading. Plasticity is
initially sub-surface when the maximum contact pressure is 3.lk or a mean
pressure of about Y, and becomes macroscopic when the mean pressure is

65
about 3 Y, that is, the hardness value of the material (see section 3.6). Thus in
equation 3.38 we see that the transition from elastic to fully plastic behaviour
occurs in a range of values of <5 112 , and the initial deviation from elastic
behaviour occurs when Pm = H/3 where

<)1/2 = 0.78c1;H)

Recognising that the elastic-fully plastic transition IS not instantaneous

we shall assume a transition point where
<)1/2 ~ ~ ({3)1/2
It is convenient to standardise this by dividing both sides by <J 112 so that

<5* 112 = (~) 1/2 = ; (~Y/2

This parameter decreases as the surface roughness as specified by <J increases,
and it is usual to define a function Q, the plasticity index, as the inverse of
<5* 112 so that
Q-_E' - ((J)1/2
- (3.39)
H {3
This index is clearly indicative of the onset of significant plastic deformation,
being large when the contact is basically plastic and small, say less than unity,
when the contact is essentially elastic. The plasticity index is clearly a useful
parameter since it is a non-dimensional grouping relating both the current
physical and geometrical properties of the surface. In a wear-process study
there are obviously advantages in making measurements of Q at discrete
time intervals, since the trend of such a parameter will then indicate whether
the surfaces are approaching an elastic state, that is, a run-in condition, see
figure 3.20.

~ "' 6
u

~
"'

-
4

2
\ r- ~
.............

0
2 4 6 8
T1me (h)

Figure 3.20 The variation of the plasticity index during

a wear process

66
3.7 THERMAL EFFECTS

In a sliding situation most of the work done against friction will be mani-
fested as a heat release with a consequent rise in temperature. Clearly, since
the heat release is basically a continuous process, temperature gradients will
develop in the contacting bodies with the highest temperature occurring at
the point of heat release (heat source), that is, the contact surface. The points
at which heat is released will obviously depend on the overall geometry of the
contacting bodies. If we consider, for instance, our previous surface contact
model of a smooth plane and a rough flat surface then heat will be released
at each contact spot whose size is determined either elastically or plastically.
Thus each contact spot may be treated as an independent heat source and the
individual temperature analysis may then be applied to obtain the tempera-
ture distribution for the general contact region. The same argument is
obviously true for the contact of curved surfaces, however, in this case the
contact spots will be so close together that at high normal loads the whole
contact zone may be regarded as a single heat source. This argument is
justified by the fact that in an analogous electrical example the results show
that the collective resistance of a large number of closely grouped constrictions
is nearly the same as the resistance of one equivalent large constriction. It
will also be recognised that the temperature effects will depend on whether
a body is stationary or moving with respect to the heat source. Thus if we
consider the simple example of two discs in contact we shall obtain a physical
feel for the problem. In such a consideration we must realise that the thermal
effects happen to discrete particles of material and therefore by contemplating
their contact history we can anticipate their overall behaviour which is
clearly related, at any given time, to the integrated effect of their history.
Consider the situation shown in figure 3.2la where two contacting discs
are assumed to be rolling with very small relative slip. It is clear that all

(al (b)

Figure 3.21

67
particles on the surfaces of both discs pass through the contact zone where
heat is generated and afterwards undergo considerable periods of rest. Their
temperature rise will therefore be relatively modest, partly because the
frictional work is small, owing to the small relative sliding, and partly because
the generated heat will be readily dissipated by heat losses during the rest
periods. If one disc is fixed as in figure 3.21b, we will clearly have a state of
pure sliding. In this case, surface particles on disc 2 will be subjected to rela-
tively high temperatures, sometimes called flash temperatures, when passing
through the contact zone and thereafter have a considerable rest period where
cooling takes place. Surface particles on disc l, on the other hand, never
escape from the contact zone and are therefore subjected to a continuous
build-up of temperature towards a steady state determined by the thermal
properties of the whole system. In this example the contact zone, which is
clearly fixed in space, can be regarded as a stationary heat source with respect
to disc I, and a moving heat source with respect to disc 2, and we shall treat
the problem simply by considering these two distinct effects on the two bodies.

3.7.I Stationary Heat Source

Consider a semi-infinite body subjected to a stationary heat source acting

over a small circular surface area of radius a. Clearly, in this case heat will be
supplied to the body through a fixed area and steady state conditions will
exist. This problem is analogous to its electrical counterpart and the heat
flow is considered as a flow of thermal current through a thermal resistance.
If Q is the rate of heat supply then it can be easily shown that the mean
temperature rise, MJ, at the surface is given by

fl.()
Q
= --
4aoc
where oc is the thermal conductivity of the body. If the temperature of other
points on the body at a far distance from the source is taken as zero, then
the above equation will give the mean surface temperature of the body, that is
Q
() = - (3.40)
• 4aoc
The same argument will reasonably apply to a slow-moving heat source
provided that its speed V is so small that at each position of contact there is
sufficient time for the temperature to acquire the same distribution as that
induced by a statK->nary heat source. For a fast-moving source equation 3.40
no longer applies and this occurs at some speed which is related to a certain
parmet~ by 5
(3.41)

where p is the density and c is the spe:::ific heat of the body. For ~ > 5 the
speed is considered high and we must apply moving heat source analysis.

68
3.7.2 Moving Heat Source

For a moving heat source traversing the surface of a semi-infinite body at a

relatively high speed V, we can neglect the effects of the transverse flow of
heat and the problem can be regarded as one of a linear heat flow. In this case,
if the heat is supplied at a constant rate of q per unit area, then the mean
temperature rise of a point on the surface of the body is given by 5
2qt112
!:!f) = ----='------:--;-,--
(mxpc)112

where t is the time during which heat has been supplied. If heat is supplied
through a circular area of radius a, then by putting q = Qjrr.a 2 and by
considering the effective value oft for all points within this area we can obtain
the mean surface temperature. The traverse time for any point (x, y) is
given by

and therefore the mean effective time is

a

1
t-- -
~
J-a
2(
V
2 -y 2)1/2 d y- -an
-
4V
0
Therefore
2Qa112rr.112 0.318Q
f) = ----,;--:--,-,;----;-;-;;- (3.42)
m 2rr.a 2rr. 112 (apcV) 112 a(aapcV) 112

We may summarise the foregoing results by introducing a standardised

surface temperature, 8*, defined by
f)*= pcV 0
rr.q
so that for a stationary heat source, using equation 3.40, we have

o: = a~:v = 0.5~ (3.43)

and for a moving heat source, using. equation 3.42, we have

f)* = 0.318 X (2apcV) 112 = 0.4 38 ~ 12 (3.44)

m (2a)1/2

where ~is as defined by equation 3.41. Examining equations 3.43 and 3.44
we find that for small values of~ or low speeds both stationary and moving
heat sources produce similar temperatures on the surface of a semi-infinite
body, while for larger values of~ a stationary heat source produces consider-
ably higher temperatures.

69
3.7.3 Application to Sliding Bodies

Let us consider the contact model shown in figure 3.22 where a single
spherical asperity of body 1 makes contact, under a normal load P, with the
surface of body 2 which is assumed to be sliding with a constant speed V.
The amount of heat generated over the contact area A will be given by

Q = J1PV (3.45)
J

where J1 is the coefficient of friction and J is the mechanical equivalent of

heat. During sliding, the contact surface of body 1 will be continuously
subjected to part of Q, say A.Q, while the remainder, (1 - A.)Q, will be flowing

Figure3.22

into body 2. For such a contact both bodies can be reasonably considered as
semi-infinite so that their surface temperatures may be obtained by applying
the results derived for stationary and moving heat sources. Thus, using
equations 3.40 and 3.42 the surface temperature of bodies 1 and 2 will be
given by

(3.46)

and
- 0.318(1 - A.) ( v ) 112 (3.47)
02- --
aJ aa.pc
Now, it remains to define the value of A.. Clearly, this will depend on the heat
transfer characteristics of the contacting bodies, but for simplicity we can

70
assume that A. is defined by the ratio of the thermal diffusivities of the two
bodies, that is
A. rx.tfPtCl
1 -A= rx2/P2C2
Therefore, we may easily use equations 3.46 and 3.47 to determine the surface
temperatures for the two bodies in contact.

REFERENCES

1. S. Timoshenko and J. N. Goodier. Theory of Elasticity. McGraw-Hill, New York,

(1951).
2. D. Tabor. The Hardness of Metals. Oxford University Press, (1951).
3. J. A. Greenwood and J. B. P. Williamson. Contact of nominally flat surfaces.
Proc. R. Soc., 195, A, (1966), 300.
4. K. A. Nuri and J. Halling. The normal approach between rough flat surfaces in
contact. Conference on Mechanics of Contact Effects, Kiev, (June, 1973).
5. J. F. Archard. The temperature of rubbing surfaces. Wear, 2, (1958-9), 438.

71
4
Friction Theories
4.1 INTRODUCTION

Friction is the resistance to motion which is experienced whenever one solid

body slides over another. The resistive force, which is parallel to the direction
of motion is called the 'friction force'. If the solid bodies are loaded together
and a tangential force is applied, then the value of the tangential force which
is required to initiate sliding is the 'static friction force'. The tangential force
required to maintain sliding is the 'kinetic (or dynamic) friction force'.
Kinetic friction is generally lower than static friction.

4.1.1 The Laws of Friction

It has been found experimentally that there are two basic 'laws' of friction,
which are obeyed over a wide range of conditions. There are, however,
a number of notable exceptions, examples of which will be included later. It
should be stressed at this point that the two laws of friction are empirical in
nature and, of course, no basic physical principles are violated in those cases
where the laws of friction are not obeyed.
The first law states that the friction is independent of the apparent area of
contact between the contacting bodies, and the second, that the friction force
is proportional to the normal load between the bodies. Thus a brick can be
slid as easily on its side as on its end and if the load between two sliding
bodies is doubled then the friction force is doubled. These laws are often
referred to as 'Amontons laws' after the French engineer Amontons who
presented them in 1699 1 • Coulomb (1785) introduced a third law, that the

72
kinetic friction is nearly independent of the speed of sliding, but this law has
a smaller range of applicability than the first two.

4.1.2 Coefficient of Friction

The second law of friction enables us to define a coefficient of friction. The

law states that the friction force F is proportional to the normal load W.
That is
F ex W
therefore
F = f.LW (4.1)

where ll is a constant known as the 'coefficient of friction'. It must be stressed

that f.L is a constant only for a given pair of sliding materials under a given set
of ambient conditions and varies for different materials and conditions. For
example, a hard steel surface rubbing against a similar surface under normal
atmospheric conditions would typically have a value of ll equal to about 0.6.
The same combination rubbing under very high vacuum conditions would
have a much higher value of fl. A graphite-on-graphite combination in
normal atmosphere has a value of ll equal to about 0.1 but this rises to over
0.5 if the atmosphere is very dry.
It is the object of this chapter to develop theories of friction which can
explain the variations in friction coefficient between different materials and
under different conditions. Now let us examine the first law a little more
closely and attempt to explain why the friction is independent of the apparent
area of contact.
In chapter 2 it has been shown that nearly all surfaces are rough on a micro-
scopic scale and true contact is obtained over a small fraction of the apparent
contact area. Furthermore the real area of contact is independent of the
apparent area of contact. Thus the first law of friction is explained since
friction is related to the real area of contact.

4.1.3 Surface Roughness and Real Area of Contact

When two bodies are rubbed together, some form of interaction takes place
at the contacting surfaces resulting in a resistance to relative motion. Most
friction theories assume that the resistive force per unit area of contact is a
constant. Thus
F = As
where F is the friction force
A is the real area of contact
sis the constant force per unit area, resisting relative motion, that is,
the specific friction force
73
If the assumption that s is a constant can be justified, then we can see the
importance of A. Let us now summarise some of the more important findings
of chapter 3 with respect to the real area of contact.
(a) For a single spherical contact or an array of similar spheres all at the
same height under loading conditions which produce elastic deformation
only

(b) When the contact is wholly plastic

A cc W
(c) For a surface whose asperity height distribution can be represented by
an exponential function
A cc W
whatever the mode of deformation, that is, either elastic or plastic.
(d) For a surface whose asperity height distribution is gaussian, we can
again write with sufficient accuracy
A cc W for all modes of deformation
(e) Many practical surfaces have an asperity height distribution that is
close to gaussian.
Results (a) and (b) above, combined with experimental findings confirming
the second law of friction, led in the past to the conclusion that asperity
contacts must be plastic. However, various physical arguments and experi-
mental results indicated that under certain conditions, for example, well
run-in surfaces, where it was found F cc W, elastic conditions must exist.
This apparent anomaly is now explained by (d) and (e) above. The analysis of
the deformation of model and real surfaces has now made possible the formu-
lation of reasonably realistic friction theories, and it is certain that any future
improvements on current theories must be based on such analysis.

4.2 FRICTION MEASUREMENT

It is not the object of this chapter to present a comprehensive survey of

friction measurement but a brief account of some of the available methods
will now be given.
Any apparatus for measuring friction must be capable of supplying relative
motion between two specimens, of applying a measurable normal load and of
measuring the tangential resistance to motion. There are a large number of
methods available and the final choice will depend largely on the exact
conditions of rubbing contact under investigation. For example, probably the
simplest arrangement is the tilting plane where a specimen is placed on a

74
flat surface which is gradually tilted until sliding starts, figure 4.1. The
coefficient of friction is then tan e. This method is obviously unsuitable in
those cases where a study of the variation of friction with continued rubbing
is required but its simplicity makes it attractive in many cases.

Figure 4.1 Measurement

of friction using tilting
plane, Jl. = tan (}

Where continuous friction measurement is required over a period of time

then an alternative approach must be used. Here one specimen, usually a
disc or a cylinder, is driven continuously, while a second specimen, nominally
stationary, is loaded against it. Commonly used combinations are crossed
cylinders, pin-on-cylinder or -disc, and disc-on-disc. The loading of the
stationary specimen can be by simple deadweight or, if the experimental
conditions demand it, by some more complicated method such as hydro-
static or magnetic loading. The measurement of the friction force is usually
accomplished by mounting the nominally stationary specimen so that a
very small tangential movement, proportional to the frictional force, occurs.
This small movement is measured and recorded. Two typical arrangements, a
crossed cylinder and a pin-on-disc are illustrated in figures 4.2 and 4.3.
In each case the specimen is mounted on leaf springs which allow a small
movement in the direction of the friction force. The movement can be

Load

Rotalong
cylinder

Figure 4.2 Simple crossed cylinder arrangement for the

measurement offriction and wear

75
tload

Rotating
diSC

Figure 4.3a Arrangement

of pin-and-disc machine
Loading moror

Disc To leaf
spring for
friction
force
measurement

Triode 10n pump

Figure 4.3b Schematic diagram of ultra-high vacuum pin-and-

disc machine

calibrated to give the friction force and measured by a capacitance or in-

ductance method and continuously recorded.
The apparatus shown in figure 4.2 is a very simple and convenient method
of measuring friction. That shown in figure 4.3b has been used for ultra high
vacuum and controlled atmosphere friction tests. The motion is provided

76
by a magnetic drive through the walls of the chamber and the load applied
and friction are measured outside the chamber wall, each force being trans-
mitted through a bellows. ·
Just as is the case in wear testing, great care must be taken in ensuring
cleanliness during these tests, since small amounts of contamination can
significantly affect the measured friction, and this is why so much emphasis
is placed on controlled atmosphere tests.

4.3 POSSIBLE CAUSES OF FRICTION

In section 4.1.3, we stated that friction must be due to some interaction

between the opposing surfaces and that this results in resistance to relative
motion. As the surfaces move relative to one another, work is done by
the forces causing the motion, that is, there is an energy loss at ihe contacting
surfaces. In considering the possible causes of friction it is convenient to
consider separately the surface interaction and the mechanism of the energy
loss.

4.3.1 Surface Interactions

When two surfaces are loaded together they can adhere over some part of
the contact and this adhesion is therefore one form of surface interaction
causing friction.
If no adhesion takes place then the only alternative interaction which
results in a resistance to motion is one in which material must be deformed
and displaced to accommodate the relative motion. We need consider only
two interactions of this type. The first is asperity interlocking. Considering the
situation illustrated in figure 4.4, it is obvious that relative motion cannot
Otrect1on of mot1on
A

Figure 4.4 Asperity interlocking-motion cannot

take place without deformation of the asperities

take place between surfaces A and B without displacement of the material of

the asperities.
A second example of the displacement type of interaction is illustrated in
figure 4.5. Here a hard sphere A is loaded against a relatively soft flat surface,
B. In order for relative motion to take place some of the material B must be
displaced. Although the surfaces of both sphere A and flat B will be rough on

77
Figure 4.5 Macro-displacement-a
hard sphere A, loaded against a softer
surface B, causes displacement of
material B during motion

a microscopic scale the material displacement at the individual asperities

will, in this case, be small compared to the 'macro-displacement'. Thus we
have only two types of interaction, Adhesion and Material Displacement,
although we will find it convenient to think of the material displacement as
either, Asperity Interlocking or Macro-Displacement.

4.3.2 Types of Energy Loss

There are only three mechanisms which can cause appreciable loss of energy
at the interacting surfaces. As relative motion takes place, material must be
deformed. The deformation can be either elastic or plastic; additionally the
material may be fractured. Plastic deformation will always be accompanied
by a loss in energy and it is this energy loss which accounts for the major
part of the friction of metals under most practical circumstances. Fracture
must occur when the surface interactions are adhesive and can also take place
due to relative motion of interlocking asperities. The formation of wear
debris is of course evidence that fracture has taken place. However, the energy
losses associated with fracture will, in most cases of sliding metals, be small
in comparison with those due to plastic deformation. One reason for this is
given in chapter 6 where it is shown that a wear particle is not formed at
each asperity contact but that for most metals in normal atmospheric con-
ditions, an asperity makes more than 1000 contacts before the formation of a
wear particle.
Although energy is required to deform a metal elastically, most of this
energy is recoverable and elastic energy losses are negligible compared with
the energy losses associated with plastic deformation. However, some rubbers
exhibit large irreversible energy losses due to their elastic deformation
(elastic hysteresis) and in certain cases this is the major source of friction,
(see section 4.8).
Summarising the above considerations shows that there are two sources
of surface interaction, that is, adhesion and material displacement, and these
can cause energy losses due to both elastic and plastic deformation and to
fracture.
We will now consider various proposed mechanisms of friction in the light
of the above considerations.

78
4.4 THE ADHESION THEORY OF FRICTION

4.4.1 Simple Theory

This theory due to Bowden and Tabor 2 , uses as a starting point the fact that
when metal surfaces are loaded against each other they make contact only
at the tips of the asperities. Because the real contact area is small the pressure
over the contacting asperities is assumed high enough to cause them to
deform plastically. This plastic flow of the contacts causes an increase in the
area of contact until the real area of contact is just sufficient to support
the load, figure 4.6. Under these conditions, for an ideal elastic-plastic
material
Ap0 = W

where A is the real area of contact, p o is the yield pressure of the metal W is
the normal load.

Pressure Po

lllll!!ll
Area A

Figure 4.6 A single asperity

contact-the asperities yield
plastically until the area of
contact has grown sufficiently
to support the load

Over the regions of intimate metal-to-metal contact, Bowden and Tabor

state that strong adhesion takes place, and that the junctions 'cold weld'.
If s is the force per unit area of contact necessary to shear the junctions,
that is, s is the shear stress necessary to cause plastic flow and final fracture
and F is the friction force, then
F =As+ Pe
where Pe is a term introduced by Bowden and Tabor to take account of the
force required to 'plough' hard asperities through a softer surface, that is, a
material displacement interaction causing plastic deformation. Bowden and
Tabor state that for most situations involving unlubricated metals Pe is small
compared with As and may be neglected. (The ploughing term is discussed
in section 4.7.) Ignoring the ploughing term we can write
Ws
F =As=-
Po
79
and
F s
f.1.=-=-
w Po
Thus this simple theory provides an explanation of the two laws of friction,
that is, that the friction is independent of the apparent area of contact and
the friction force is proportional to the load.
In the above analysis we have considered an ideal elastic-plastic material,
and have ignored the effects of work-hardening. Therefore it is reasonable to
takes equal to S0 the critical shear stress, and both Po and S0 must refer to the
softer of the two metals.
Now

(4.2)

This ratio sofPo is fairly constant for most metals and the above analysis
gives an indication of why the friction coefficient of a large range of metals
varies little, while their mechanical properties, for example, hardness, vary
by orders of magnitude. In the case of two hard metals in rubbing contact
Po is high, A is low and S 0 is high. For soft metals, Po is low, as is S 0 , but A is
large.
One way of obtaining lower coefficients of friction is to deposit a thin layer
of soft metal onto a hard metal substrate. Now, the load carrying capacity is
really due to the substrate and Po is the yield pressure for the substrate.
However, shearing takes place in the soft metal layer and the critical shear
stress of the soft metal is the value we must use, and therefore

f.1. ~ s0 (soft) = low

Po (hard)

4.4.2 Discussion of Simple Theory

There is no doubt that junction welding can take place during the rubbing of
metals. For cleaned metal surfaces in high vacuum very high adhesion and
friction coefficients have been recorded. For metals rubbed in normal atmos-
pheric conditions, adhesion and transfer of metal fragments have been
demonstrated using radioactive tracer techniques. However, the simple
adhesion theory has been criticised for a number of reasons and it can be
shown to be inadequate by a comparison of the absolute values of the friction
coefficient predicted by the simple theory and those found exprim~ntaly.
For most metals, S0 is about one-fifth of Po and therefore the simple adhesion
theory predicts that f.1. ~ 0.2.
Many metal combinations in air give a friction coefficient higher than 0.5,
and metals in high vacuum give much higher values of f.l.· This led Bowden

80
and Tabor to re-examine some of the assumptions in the simple theory, and to
present a modified and more realistic description of friction in terms of
adhesion 2 .

4.5 MODIFIED ADHESION THEORY

The fact that very high values of friction are obtained for metals under high
vacuum conditions, where adhesion is unimpeded by oxide films or other
contaminants, indicates that the real contact area must be considerably
larger than is indicated by the simple theory. In the simple theory it was
assumed that A was defined by the yield pressure of the softer metal Po
and the normal load W. This is approximately true for static contact but in the
case of friction, where a tangential force is also applied, yielding must take
place as a result of the combined normal and shear stresses. To illustrate this
let us look at the simplified two-dimensional stress system illustrated in
figure 4.7a and assume that yielding occurs when the maximum shear stress
attains a critical value 3 .
Shear
stress

llllllll
Normal
s stress

(al

(b)

Figure 4.7 The Mohr's circle construction for finding the

maximum shear stress for an idealised two-dimensional
junction under normal and tangential stress

We can find the maximum shear stress in the system by using the Mohr's
circle 4 construction, figure 4. 7b. The maximum shear stress is the radius R
of the circle therefore

When R reaches the critical resolved shear stress, yielding takes place. From
this it is easy to see that yielding is dependent on the action of the combined
stresses and not on p alone.

81
We will now examine how consideration of the combined stresses affects
the value of the real area of contact in an asperity junction.
Consider first a single asperity contact under a normal load W. The area
of contact will be A where W/A =Po· If a tangential force is now gradually
applied up to a value F, further plastic flow will take place. This flow causes
an increase in the contact area, that is, junction growth is brought about by
the superposition of the shear stress on the normal stress. The normal and
shear stresses caused by the normal and shear forces must decrease as the
area over which the forces act increases, and junction growth continues until
the combined stresses obey a relationship of the form previously given for the
two-dimensional system. The exact solution for the three-dimensional case
is not known but we will assume that it is of the form
p2 + rxs2 = k2
where rx and k are constants yet to be determined, that is

(4.3)

where A is the area of contact of the junction. Now, if s is zero, the pressure
over the junction must be Po and therefore
k2 = p;
that is
(4.4)
If F increases to very large values then junction growth continues until
W/A is small in comparison with F/A in which case we can write

In this case s must be approximately equal to S0 , the critical shear stress.

Therefore

or
p;
(1,~2
so
Now Po ~ 5s0 therefore rx ~ 25. However, experiment indicates that rx should
have a value somewhat lower than 25 and Bowden and Tabor assume that
rx = 9. (This also implies that Po= 3s0 .) We will see later that the exact value
of rx does not greatly affect the amount of junction growth taking place in
many practically important cases.
From equations 4.3 and 4.4 we have

82
W !Po is the area of contact derived from the simple theory in which only the
effect of normal load is considered, and the additional term a(F/p 0 ) 2 repre-
sents the increase caused by the shear or friction force.
It can be seen from the above discussion that for clean metal surfaces
(that is, metals in high vacuum), large-scale junction growth is possible,
resulting in very high friction coefficients. This has been confirmed experi-
mentally5. In normal atmospheres metals are covered by an oxide or other
contaminant film. We will now see how the above theory can be applied to
such cases.

4.5.1 Adhesion Theory of Metals with Contaminant Films 2

Consider an asperity junction under normal load W and a gradually increas-

ing shear force F. We assume that at the junction there is a thin contaminating
film with critical shear stress sf. We also assumes f = cs 0 where S 0 is the critical
shear stress for the metal and cis less than unity. While F and A, have values
such that F I A < sf, then junction growth will proceed as described pre-
viously for uncontaminated metals. However, when F I A = sf then the con-
taminating film will shear, junction growth will end, and gross sliding will
occur.
Thus the condition for the start of gross sliding is
p2 + IY.SJ = p;
But it has already been shown that
p; =as;
Therefore

or

Therefore
sf= c
p [a(l _ c2)Jli2
The coefficient of friction 11 = (F /W) = (s fA/pA). Therefore
c
(4.5)
11 = [a(l _ c2)Jli2

As c tends to 1 then 11 tends to infinity in agreement with results obtained for

uncontaminated metals. We can plot 11 against c for various values of 1:1.,
figure 4.8, and we see that the value of 11 drops rapidly as c reduces from unity.

83
0

Figure 4.8 The variation of Jl. with c for different values

of IX. It can be seen that except at large values of c, the
exact value of IX is not of major importance

Thus a small amount of weakening at the interface produces a drastic reduc-

tion in Jl..
When c is small then equation 4.5 can be written
c
Jl. = (1X)l/2

but

Therefore
cs0 s1
)1.=-=-
Po Po
84
that is
critical shear stress of the interface
~ =-~
yield pressure of the bulk metal
which is essentially the same result as obtained by the simple theory (com-
pare discussion of soft metal films, or hard substrates). This is understandable
since when the interface is weak in shear, appreciable junction growth does
not occur and the real contact area depends only on the normal load and the
yield pressure po.
This then is a more realistic theory. Although it is still based on a simple
model and contains a number of assumptions, it is able to explain a remark-
able range of friction phenomena.
However, it is no longer exclusively an adhesion theory. Let us examine
what are the main points of the theory
(a) The real contact area is defined by plastic deformation.
(b) The two rubbing surfaces are separated by a film of shear strength
which can vary from low values up to the bulk shear stress of the sub-
strate material.
(c) The friction force is the force required to shear the separating film.
At one extreme, where metal-to-metal contact and junction welding take
place, we can consider that either the separating film is of zero thickness,
or that it is composed of the softer of the two metals. Under atmospheric
conditions of dry sliding where the metals are covered by oxide films it is
only at those points where the oxide film is broken that metal-to-metal
contact and welding take place. In this case the effective shear strength of the
interface will lie between the shear strength of the softer metal and that of
the oxide film; its exact value depending on the relative amount of metal to
metal and metal to oxide contact.
We will now discuss a number of criticisms of the theory that have been
advanced.
(i) If two hard metals are rubbed together under atmospheric conditions,
in general no adhesive component normal to the surfaces can be
detected when the normal force has been released. Bowden and Tabor
answer this by pointing out that the adhesion under rubbing conditions
occurs while the normal force is exerted. To measure adhesion in the
normal direction the normal load must first be released and elastic
recovery will break many of the bonds during this process.
(ii) The adhesion theory of friction can explain the transfer of material
from one rubbing surface to the other but offers no explanation of the
formation of loose wear debris. Experimental work has shown that
transfer occurs during single traversals but subsequent traversals on
the same track produce loose wear particles. This suggests a possible
change of mechanism during the rubbing process. It may be significant

85
that much of the experimental work carried out by Bowden and Tabor,
giving results consistent with the adhesion theory, consisted of single
rather than multiple pass experiments.
(iii) The adhesion theory is based on plastic deformation of asperities.
On continued rubbing under conditions of low wear, work hardening
will take place and a proportion of asperity contacts will now deform
elasticially. However, although the deformation of the asperities is now
largely elastic, the area of contact is still related to the previous plastic
deformation. This can be understood by considering an individual as-
perity contact under an initial normal load W. The deformation curve
is as shown in figure 4.9. If the load is now released and reapplied up to
a maximum value of W, the deformation is elastic but the total deforma-
tion or strain is exactly the same as in the plastic deformation case.

We will return to these criticisms in section 4.9 which contains a general

discussion of friction theories.

I
I

I
I
I
I
I
I
I
I
I
I
I
I
I 8

Figure 4.9 Typical stress-

strain curve of metals. If the
stress is relaxed at A and then
reapplied the deformation is
elastic along BA, but the resul-
tant strain is equal to that
caused by the initial plastic
deformation

86
4.6 PLASTIC INTERACTION OF SURFACE ASPERITIES

In the Bowden and Tabor theory, the normal and yield stress on a single
asperity were assumed to be representative of the stresses of all asperities.
No consideration was given to the possibility that at a single asperity,
s and p could vary with time over large ranges so that s/p could also vary.
This evolves as a feature of a theory based on the plastic interaction of
asperities. This type of theory was first introduced by Green 6 and has been
extended by Edwards and Halling 7 whose treatment is followed here. The
basis of this analysis is that in the sliding of macroscopically flat surfaces the
motion is parallel to the surface.
Consider two wedge-shaped asperities of equal angle interacting as shown
in figure 4.10. Assuming that motion is in the direction shown and that

Figure 4. 10 The idealised wedge-shaped as-

perities studied in the 'plastic interaction theory'

deformation is plastic it is possible to calculate the instantaneous shearing

force F 1 and normal force W1 over the complete horizontal displacement l
from the first contact until the asperities separate, both for the case where the
asperities are in intimate contact and adhesion is possible and for the case
where the asperities are separated by a film of shear strength cs 0 where c and
s0 have the same meanings as in the Bowden and Tabor theory. The develop-
ment of the expressions for F 1 and W 1 is beyond the scope of this book
but the form of the plot of normal and shear forces against displacement is
shown in figure 4.11 for junction angles of 10° and for various values of c.
The coefficient of friction is the sum of instantaneous shearing forces F 1
for all contacting asperities divided by the sum of instantaneous normal
forces W 1 for all contacting asperities. Alternatively J1 could be regarded as
the mean FjW value for a single pair of interacting asperities, which are
assumed to be typical of all the interacting asperities. Thus the coefficient of
friction can be calculated from figure 4.11 and plots of J1 against c, so obtained
are shown in figure 4.12.
It is interesting to note that if the results are extrapolated the curve cor-
responding to junction angle (}equal to zero, corresponds identically with the

87
W
- .... '
\ c•OB
c•O 2 \

''
'

I
I
I
I
I

Figure 4.11 The variation of the normal force P

and the friction force F throughout the junction life
for a junction of angle 10°

Bowden and Tabor relationship

c

The general relationship given by the Edwards and Halling theory is

p. = [ a(l _c c2)1/2 + <P]/[ 1 - a(l _c c2)1/2 <P J (4.6)

where <P is a function of c and the geometry of the junctions and equal to
zero when() equals zero.
This then indicates that the Bowden and Tabor theory is a special case of
the more general Edwards and Halling theory. The Edwards and Halling
theory assumes plastic deformation and an array of similarly shaped asper-

88
20

15

::t_ 10

Junction
angle

,_c_l
p a(l-c 2 )2

...... ..._ --. oo

10 05 0
c
Figure 4.12 The variation of J.l. with cfor various junction
angles

ities at the same height. However, using the methods developed in chapter 3
for normal load effects on rough surfaces, it is clearly possible to extend this
type of theory to cover surfaces consisting of an array of asperities with a
given height distribution which are interacting elastically and plastically
depending on their height. The effects due to work hardening and dissimilar
materials could also be included.

4.7 PLOUGHING EFFECT

Ploughing is that part of friction, included by Bowden and Tabor in their total
friction force, caused by asperities on a hard metal penetrating into a softer
metal and 'ploughing' out a groove by plastic flow in the softer metal. This is
a major component of friction during abrasion processes and also it is prob-
ably important in cases where the adhesion term is small, for example, for

89
well-lubricated surfaces where the shear strength of the interfacial film is low.
Consider a hard material whose surface is composed of a large number of
similar conical asperities of semi-angle 0 in contact with a softer material
whose surface is comparatively flat, figure 4.13. During rubbing only the

Figure 4.13 The ploughing of a

soft surface by a hard conical
asperity

front surface of each conical asperity is in contact with the opposing material
and the vertically projected area of contact is given by
nr 2
A =n-·
2
where n is the total number of asperities
nr 2
W = Ap0 = n -
2 p0
The friction force F is obtained in a similar manner by considering the total
projected area of material which is being displaced by plastic flow, that is
F = nrhp0
Therefore
F 2h
J.l=-=-
w nr
but
h
-=cot 0
r
therefore
2
11 =-cot() (4.7)
1r
Similar expressions can be obtained for asperities of a different shape.
Using this approach 11 always equals one half the vertical projected area of the
asperity divided by the horizontal projected area of the asperity. The above

90
theory assumes that the yield pressure is the same in the vertical and horizon-
tal direction. Kragelskii 8 introduces an extra multiplying factor which allows
for a difference in work hardening in the two directions.
Bowden and Tabor state that the contribution of ploughing to friction is
small for the following reason. The tangential resistance to sliding for a
single asperity is made up of a shear term and a ploughing term, that is
F = AvS0 + AHPo
where Av and A" are the vertical and horizontal projected areas.
For asperities on most metal surfaces, the angle () is large and A"/Av is
very small so that the ploughing term is negligible in comparison with the
adhesion term. For rough surfaces where() is larger, the ploughing term can
be comparable to the adhesion term.

4.8 ELASTIC HYSTERESIS LOSSES

In chapter 8 it is shown that the major part of rolling friction of elastic (and
visco-elastic) solids is accounted for by hysteresis losses. Also theoretical
expressions for friction arising in this way are derived and shown to agree
with experimentally determined values. When materials exhibiting elastic
hysteresis are in sliding contact then a similar expression must appear in the
friction term. If the materials are well lubricated then, as in rolling contact,
elastic hysteresis is the major cause of friction.
Consider for example the sliding of a rigid hemisphere on a flat sheet of
well-lubricated rubber. The deformation experienced by the rubber will
sensibly be the same as if a rigid sphere of the same radius were rolled over it
under the same load. Thus the friction force due to elastic hysteresis will be
the same in each case and the theory developed in chapter 8 can be used for
sliding contacts. It should be noted that this term is analogous to the plough-
ing term discussed earlier, but now the friction is due to elastic hysteresis
rather than the force required to deform the material plastically. References
are occasionally made to elastic hysteresis in metals. However, it is extremely
doubtful whether this has any effect on friction, and the energy losses referred
to are probably due to plastic deformation, analogous to the losses occurring
during a fatigue cycle.

4.9 DISCUSSION OF THE VARIOUS FRICTION THEORIES

Initially, if we limit the discussion to the friction of metals, we can ignore any
effects due to elastic hysteresis, which are negligibly small for metals. The
'ploughing' theory is nothing more than a particular case of the 'interaction
of asperities' theory in which some simplifying assumptions have been made.
Thus we need consider only the Bowden and Tabor 'adhesion' theory and
the 'interacting asperity theory'.

91
Let us now consider those features of the two theories that are similar. The
deformation in both cases is plastic, that is, the area of contact is defined by
the flow stress, and the major sources of frictional energy losses are due
to plastic deformation. Both theories include the effects due to a surface
film of shear strength equal to or lower than the bulk shear strength of the
softer of the two metals in contact. The only major difference between the
theories appears to be that the Bowden and Tabor theory purports to rely
solely on 'adhesion' as the surface interaction mechanism. However, as we
have seen, the theory also covers those situations where an oxide film covers
the metal surface in which case 'adhesion' is limited to those parts of the sur-
face where the oxide film is penetrated. Where the surface films are not pene-
trated then the relevant shear strength in equation 4.5 is the shear strength
of the film, but it is unlikely that the surface interaction is adhesive in these
cases. Thus it would appear that the Bowden and Tabor theory must im-
plicitly include the possibility of interlocking asperities. The interlocking
asperity theory does not exclude the possibility of adhesion, in fact this pos-
sibility is explicitly included and leads to the conclusion that the stresses on
individual asperities can be tensile over part of the contact, that is, that the
contribution of an individual asperity to the total normal load can be neg-
ative, again over part of the contact. Thus the two theories are not in conflict
and in confirmation of this conclusion we have the fact that the asperity
interaction theory can give exactly the same expression for the friction co-
efficient as the Bowden and Tabor theory. If the 'ploughing' term is included
in the Bowden and Tabor theory then good agreement between the two
theories is obtained for all junction angles.
The 'interlocking asperity' theory appears to be the more satisfactory of
the two and offers the possibility of further extension to include the rubbing of
surfaces with a realistic asperity height distribution resulting in both elastic
and plastic interactions, with work-hardening effects considered. However,
although somewhat less realistic, the Bowden and Tabor theory does give
friction coefficients of the correct magnitude over a wide range of conditions,
and has the advantage of being analytically much simpler than the asperity
interlocking theory.
When materials exhibiting elastic hysteresis losses are rubbed then the
only difference between this situation and the metallic case is in the form of
frictional energy losses. The theory presented in chapter 8 for rolling friction
can be readily adapted to the case of a given height distribution of hemi-
spherical asperities.

REFERENCES

I. G. Amontons. Hist. Acad. R. Soc., (1699), (Paris), 206.

2. F. P. Bowden and D. Tabor. The Friction and Lubrication of Solids. Pt II. Oxford
University Press, (1964).

92
3. S. Timoshenko. Strength of Materials. Pt II (3rd Edition). Van Nostrand-Reinhold,
New York, (1956).
4. S. Timoshenko. Strength of Materials. Pt I (3rd Edition). Van Nostrand-Reinhold,
New York, (1955).
5. D. H. Buckley. Friction, Wear and Lubrication in Vacuum. NASA, SP-277, (1971).
6. A. P. Green. Friction between unlubricated metals, a theoretical analysis of the
junction model. Proc. R. Soc., 228, A, (1955), 191.
7. C. M. Edwards and J. Halling. An analysis of the plastic interaction of surface
asperities and its relevance to the value of the coefficient of friction. J. Mech. Engng
Sci., 10, (1968), WI.
8. I. V. Kragelskii. Friction and Wear. Butterworths, London, (1965).

93
5
Wear
5.1 INTRODUCTION

Wear occurs as a natural consequence when two surfaces with a relative

motion interact with each other. Although our understanding of the various
mechanisms of wear is now improving, no reliable and simple quantitative
law comparable to that for friction has been evolved. This is not surprising
since the wear process involves many diverse phenomena, interacting in a
largely unpredictable manner. Furthermore, whereas the coefficients of
friction of most materials lie between 0.1 and 1.0, corresponding wear rates
can vary over many orders of magnitude.
Wear is often thought of as a wholly harmful phenomenon, but this is not
so. For instance, the wear experienced in 'running-in' is beneficial and many
forming operations, for example, machining and filing, rely on wear mechan-
isms. When wear is harmful then the rates of wear of sliding components
can be reduced, but not eliminated, by lubrication, by careful design, and by
material selection, and each of these aspects will be considered.

5.1.1 Definition

Wear is familiar to everyone and probably we all feel that we understand what
is meant by wear. However, the formulation of a precise and all embracing
definition of wear is difficult. A committee of the Institution of Mechanical
Engineers decided on the following definition: 'the progressive loss of sub-
stance from the surface of a body brought about by mechanical action'.

94
Kragelskii 1 defines wear as: 'the destruction of material produced as a result
of repeated disturbances of the frictional bonds'. Neither definition is perfect.
For instance, the first appears to eliminate spark erosion as a form of wear,
and the second perhaps places too much emphasis on fatigue effects in wear.

5.2 TYPES OF WEAR

In order to study and gain a better understanding of wear, it is essential to

recognise that several distinct and independent mechanisms are involved.
In his excellent 'Survey of Possible Wear Mechanisms', Burwell 2 lists four
mechanisms
(1) Adhesive wear
(2) Abrasive wear
(3) Corrosive wear
(4) Surface fatigue
He also includes a fifth classification under the heading 'Minor Types' of
wear, which covers erosion, cavitation and impact chipping. We will follow
the same classifications, but will also separately consider phenomena such as
fretting, which arise due to a combination of the above mechanisms.

5.2.1 Adhesive Wear

It has been shown earlier, (chapter 4), that macroscopically smooth surfaces
are rough on an atomic scale, and that when two such surfaces are brought
together contact is made at relatively few isolated asperities. As a normal
load is applied the local pressure at the asperities becomes extremely high.
The yield point stress is exceeded, and the asperities deform plastically,
until the real contact area has increased sufficiently to support the applied
load. In the absence of surface films the surfaces would adhere together, but
very small amounts of contaminant prevent adhesion under purely normal
loading. However, relative tangential motion at the interface acts to disperse
the contaminant films at the points of contact, and cold welding of the junc-
tions can take place. Continued sliding causes the junctions to be sheared and
new junctions to be formed. As we have seen in the previous chapter, this
naturally leads to a simple theory of friction with the coefficient of friction
equal to the shear strength divided by the yield pressure.
This model also leads naturally to the formulation of a mechanism of
wear, but not to a precise quantitative law of wear. The amount of wear de-
pends on the position at which the junction is sheared. If shear takes place at
the position of the interface then wear is zero. If shear takes place away from
the interface then metal is transferred from one surface to the other. With
further rubbing some of the transferred material is detached to form loose
wear particles by one of the mechanisms described later.

95
The transfer of material from one surface to another by adhesion has been
studied by several investigators. For example, Kerridge and Lancaster 3
used a radioactive pin of 60:40 brass, rubbing against a ring of tool steel.
The transfer of material was demonstrated by placing a photographic film
in contact with the tool steel after rubbing and obtaining an autoradiograph
of the transferred brass from the wear track.

Law of Adhesive Wear

Having described the mechanism of adhesive wear we can now develop a
simplified law of adhesive wear. Here we follow the approach of Archard 4 ,
although other workers have derived similar expressions.
Assume that the contact is made up of a number of similar asperities each
of radius a. The area of each contact is 1ta 2 and each contact supports a
load of p 0 1ta 2 , where Po is the yield pressure. The surfaces will pass completely
over each asperity in a sliding distance of 2a and we will assume that the
wear fragment produced at each asperity is hemispherical in shape and of
volume 2/37ta 3 .
Then the total wear volume Q, per unit distance of sliding is given by
2 3
Q = ~ 1"7ta
2a
2
1 2 1ta
= 1" ~ 1ta =- x n
3
where n is the total number of contacts. But each contact supports a load of
Po 1ta 2 , therefore

or
w
n1ta 2 = -
Po
therefore
w
Q=-
3po
This equation has been derived assuming that all asperity encounters produce
a wear particle. If only a fraction k of all encounters produce wear particles
then the equation becomes

Q=kw (5.1)
3po
where k is the probability of an asperity contact producing a wear particle.
All quantities in equation 5.1 are measurable except k, and it is this factor

96
that represents the uncertainty in the equation. k must be found for different
combinations of sliding materials and for different conditions of rubbing.
Relationships similar to equation 5.1 have also been developed by Holm 5
and by Burwell and Strang 6 • This equation leads to three 'laws' of wear.

(1) The volume of wear material is proportional to the distance of travel.

(2) The volume of wear material is proportional to the load.
(3) The volume of wear material is inversely proportional to the yield
stress, or the hardness, of the softer material. ·

The first 'law' is found to be true for a wide range of conditions. A number
of investigators have demonstrated the validity of the second law over a
limited range of load. For instance, figure 5.1, taken from the results of Burwell

0 0 079 em Oia cylinder

0 120° Cone

Figure 5.1 Wear rate as a function of load

and Strang 6 , shows the rate of wear plotted against load for steel rubbing
against steel. However, if the loading is increased further, plots of the form
shown in figure 5.2 are obtained. These show k/H, the adhesive wear co-
efficient plotted against the average pressure (that is, load over apparent
area of contact) for steels of different hardnesses. It can be seen that k remains
constant up to a pressure of about H/3 where His the hardness of the steel,

97
24 24

20 20

f 16 H/3
f'
16
Q 0 H/3
"0' I 30'
I
I
1:t 12
I
"'6 12
:t
'
\c I '
\c
I 8

I
4

(a) (b)

Figure 5.2 The variation of wear coefficient with apparent

pressure for steel: (a) Brine /I hardness 223; (b) Brine /I hardness
430

and above this pressure k, and hence the wear rate, increase rapidly. It is
found that at these higher loadings, large scale welding and seizure occur.
Similar results have been obtained for other metals, although the average
pressure at which k begins to increase is often lower than H/3. Under normal
loading H/3 is the pressure at which the plastic zones under individual
asperities begin to interact, and an increase in pressure above H /3 causes the
whole surface to become plastic so that the real area of contact is no longer
proportional to load. This condition is attained at a lower normal pressure
than H/3 when tangential (frictional) forces are present.
The third 'law' is supported by much experimental work, notably that of
Kruschov 7 , but as will be shown later, material properties other than hard-
ness, may also be important in determining the wear rate.

Rowe's Modified Adhesion Theory

Rowe 8 has modified the simple adhesion wear theory presented above to
include the effect of surface films. Rowe starts with the Archard equation for
wear

or
Q = k'A

98
He then points out that in lubricated systems, including systems lubricated
by surface films, k' must be related to both properties of the lubricant and
those of the sliding metals. The volume of adhesive wear must be related to
the metal-metal contact area Am, that is
Q = kmAm
where km is a constant for the sliding metals and independent of lubricant
properties or of surface films. A parameter {3 is now introduced such that

{3 =Am
A
{3 is called the fractional surface film defect and is characteristic of the lubri-
cant. For instance a poor lubricant would allow more metal-metal contact
than a good lubricant, and therefore a poor lubricant would have a higher {3
than a good lubricant. Substituting for Am we have

but

A=-
w
Po
Therefore

Q=kf3w (5.2)
m Po

Now this equation contains km, characteristic of the sliding metals, and {3,
characteristic of the lubricant (or contaminating film). Rowe introduced one
further refinement into this theory, similar to that introduced by Bowden and
Tabor into the adhesive friction theory. He argued that the appropriate value
for the yield pressure is that value obtained under combined shear and normal
stresses rather than that under a static normal load. The relationship between
these quantities is

where p = normal pressure

Po = flow pressure under static load
s= shear stress
ex = a constant.
From friction theory (see chapter 4)
s = fJ.P

Therefore

99
and therefore the wear equation becomes

(5.3)

5.2.2 Abrasive Wear

The term 'abrasive wear' covers two types of situation. In both cases wear is
occasioned by the ploughing-out of softer material by a harder surface. In
the first instance a rough hard surface slides against a softer surface. In the
second case abrasion is caused by loose hard particles sliding between
rubbing surfaces. The action of a file or emery paper against a softer metal is
an example of the first type of abrasion. There are two requirements here-
that one surface must be harder than the other, and that the hard surface
must be rough. For instance, the wear rate of carbon against tungsten
carbide is very low if the surface of the tungsten carbide has been smoothed
sufficiently (C. L.A. 40 micron). This type of wear has been largely eliminated
from modern machinery, due to a greater awareness of the importance of
surface finish and the availability of instruments for the routine measurement
of surface roughness.
However, the second mechanism is still of great importance. Not only do
most mechanisms work in an environment containing much airborne dust
and dirt, but also the products of corrosive wear are more often than not
abrasive in character. Particles of hard metal produced by the adhesive
wear mechanism can also cause abrasion. However, airborne dust and grit is
the largest source of abrasive particles and the practical solution here is to
exclude the dust by adequate sealing and filtration.
The fact that dust from the atmosphere is a major cause of abrasive
wear probably explains the large variability in wear measurements. Thus,
where attempts are being made to measure wear due to causes other than
abrasion, Cattaneo and Starkman 9 have suggested that the lowest wear
rates obtained are more appropriate than average values. They suggest that
increases from the minimum values are due to insufficient precautions
against ingress of abrasive particles. Statistical analysis appears to support
their ideas.

Abrasive Wear-Quantitative Relationship

In order to obtain a quantitative expression for abrasive wear we shall assume
a simplified model, in which one surface consists of an array of hard conical
asperities all with the same semi-angle 0. The second surface is softer and
flat.
Consider a single asperity creating a track through the softer surface as
shown in figure 5.3. In traversing unit distance, the asperity displaces a

100
Figure 5.3 Abrasive wear by a conical
indenter

volume of material rd. But d = r cot fJ, therefore the volume displaced by one
asperity in unit distance = r2 cot fJ.
We assume that the material has yielded under the normal load and there-
fore the asperity supports a load of nr 2 p0 /2 where Po is the yield pressure of
the softer material.
If there are n asperity contacts

total normal load W = T

nnr p
2

and the total volume displaced in unit distance is Q where Q = nr 2 cot fJ.
Eliminating n

Q = 2W cot fJ (5.4)
1tPo
It is interesting to note that this equation is of exactly the same form as the
adhesive wear equation (equation 5.1) and therefore the 'laws' of wear listed
under 'Law of Adhesive Wear' which were formulated from equation 5.1
apply equally well to abrasive wear.
The derivation of equation 5.4 has been based on an extremely simple
model. No account has been taken of the distribution of asperity heights and
shapes. Material will tend to build up in front of the asperities, altering the
conditions as it does so. As discussed in the next section parameters such as
Young's modulus will be important in certain situations, but are not con-
sidered in the simple theory. However, an equation of the form
Q = kaW (5.5)
H

where H is the hardness of the softer material and ka the abrasive wear con-
stant, is found to cover a wide range of abrasive situations.
The above derivation is appropriate to two-body abrasive wear. In the case
of loose abrasive particles causing wear on rubbed surfaces, that is, three-
body abrasion, the same form of equation will hold, but ka will be lower,
since in this case many of the particles will tend to roll rather than slide.
As wear due to abrasion proceeds, some blunting of the hard asperities or
particles will occur, thus reducing the wear rate. However, an abrasive grit,

101
which is brittle, can fracture causing a resharpening of the edges of the particle
and an increase in wear rate.
Many investigators have confirmed that hardness is the most important
parameter in abrasive wear. For example, Kruschov 7 plotted resistance to
wear against hardness for a range of annealed pure metals, obtaining a linear
relationship. He also found that prior work-hardening of the pure metals
had no effect on the wear rate. These and other experiments have led to the
conclusion that during abrasion a metal surface work-hardens to a maximum
value, and it is this value of hardness which is appropriate when considering
abrasion resistance.
Richardson 10 has shown that if wear resistance is studied as a function of
the ratio ofthe hardness H mfH a, where H m is the hardness ofthe metal surface
and Ha the hardness of the abrasive, then the wear resistance increases rapidly
as HmfHa becomes greater than 0.8. The region of abrasive wear where
HmfHa < 0.8 is known as 'hard abrasive wear', and where HmfHa > 0.8 as
'soft abrasive wear'. In the 'soft' region the abrasive wear does not cease until
HmfHa is much greater than unity. It has been suggested that the criterion for
abrasive wear to be negligible is equality of the yield stresses of the metal and
abrasive. Thus for contact between metals, abrasive wear should be negligibly
low when the hardnesses are equal. However, if an abrasive non-metal is
in rubbing contact with a metal, abrasive wear is appreciable until HmfHa is
considerably higher than unity. This becomes apparent if we consider the
relationship between hardness and yield stress, as measured by a diamond
indenter
H = CPo
where His hardness, Po is yield stress and c is a constant. For a metal, c is
nearly 3 (see chapter 3). However, for a non-metal, with a much lower Young's
modulus than a metal, elastic yielding around the indenter reduces the con-
straint around the plastic zone of the hardness indenter. Then c is less than 3,
and can be as low as 1.2 to 1.3. Thus, equivalence of yield stress between
metal and non-metallic abrasive, and hence negligible abrasive wear, requires
H mf H a ratios of up to 2.5.
In the above discussion, based on hardness as the parameter of prime im-
portance in abrasive wear, the value of Young's modulus has been introduced
in connection with the relationship between hardness and yield stress.
Young's modulus may have a more direct bearing on abrasive resistance.
Oberle 11 has pointed out that if, in the presence of abrasive particles, a
surface can elastically deform sufficiently to allow the particle to pass, then
permanent damage to the surface will be avoided. As an example of this,
water-lubricated rubber bearings for propeller shafts in ships operating in
sandy waters have proved more abrasion resistant than materials with a
higher value of Young's modulus, such as bronze. This indicates that the
important parameter is the limit of elastic strain, that is, the yield stress
divided by Young's modulus. Oberle has listed (table 5.1) various materials

102
TABLE 5.1 RATIO OF HARDNESS TO MODULUS OF ELASTICITY
FOR VARIOUS MATERIALS 11

Modulus of Brinell Hardness/elastic

elasticity, hardness modulus, H/E
Material Condition E, lbf/in 2 number, H (a)

Alundum (Al 20 3 ) Bonded 14 X 106 2000 143 x 10- 6

Chromium plate Bright 12 1000 83
Grey iron Hard 15 500 33
Tungsten carbide 9% Cobalt 81 1800 22
Steel Hard 29 600 21
Titanium Hard 17.5 300 17
Aluminium alloy Hard 10.5 120 11
Grey iron As cast 15 150 10
Structural steel Soft 30 150 5
Malleable iron Soft 25 125 5
Wrought iron Soft 29 100 3.5
Chromium metal As cast 36 125 3.5
Copper Soft 16 40 2.5
Silver Pure 11 25 2.3
Aluminium Pure lO 20 2.0
Lead Pure 2 4 2.0
Tin Pure 6 4 0.7

in descending values of H/E. No controlled tests have yet been made to

examine this relationship, but the trends indicated in the table appear to
agree with general experience of wear resistance. It should be pointed out,
however, that Spurr and Newcombe 12 interpreted their results to indicate
that wear resistance increases with increasing elastic modulus, in direct
disagreement with the work of Oberle.

5.2.3 Fatigue Wear

Rolling Contact
Adhesive and abrasive wear mechanisms depend on direct contact between
solids and they produce a wear pattern that is progressive from the start of
rubbing. If the surfaces can be separated by a lubricating film (and abrasive
particles excluded), then these wear mechanisms cannot operate. This is the
situation in well-designed rolling element bearings, where it is found that a
fatigue mechanism of failure takes place. For this case, although direct con-
tact does not occur, the opposing surfaces experience large stresses, transmit-
ted through the lubricating film during the rolling motion. The nature and
magnitude of the stresses can be found using the Hertz 13 equations. These
show that the maximum compressive stres!;es occur at the surface, but the
maximum shear stresses occur some distance below the surface as illustrated
in figure 5.4. As rolling proceeds, the directions of the shear stresses for any
103
Drstonce below
surface

Figure 5.4 The variation of shear stress with distance

below the surface

element change sign. Fatigue failure is dependent on the amplitude of the

reversed shear stresses and, if in rolling contact these are above the endurance
limit, failure will eventually occur.
The position of failure in a perfect material subjected to rolling contact will
be defined by the position of maximum reversed shear stress obtainable
using the Hertz equations. If, superimposed on the rolling contact, there is
also some sliding contact, then the position of failure moves nearer to the
surface. However, materials are rarely perfect and the exact position of ulti-
mate failure will be influenced by inclusions, porosity, microcracks and
other factors.
Rolling contact fatigue wear is characterised by the formation of large
wear fragments after a critical number of revolutions. Prior to this critical
point, negligible wear takes place and a rolling contact bearing will operate
normally until the wear particles are detached, when the useful life of the
bearing is terminated. This is in marked contrast to the wear experienced in
sliding bearings, where due to adhesion and/or abrasion, wear causes a
gradual deterioration from the start of running. From this we can see that the
amount of material removed by fatigue wear in rolling contact is not a very
useful parameter. Much more relevant is the useful life in terms of number of
revolutions or time at a given speed. The definition of life of a bearing used
by manufacturers is the number of revolutions which will be reached or
exceeded by 90 per cent of similar bearings. Testing of large numbers of

104
rolling bearings has shown that the life N, as defined above is inversely
proportional to the cube of the applied load W. That is
W3 X N =Constant (5.6)
The position (measured from the surface) of the maximum shear stress in
pure rolling is proportional to (WR) 113 for a ball and to (WR) 112 for a
cylinder 13 , and the following figures are presented as a typical example. For
a ball 1.0 em in diameter, subjected to a load of 1000 MN/m 2 the maximum
shear stresses occur at a depth of 0.12 mm from the surface of the ball.
The above discussion has been related to rolling elements with little surface
traction. There are many mechanisms where rolling and sliding motions are
combined, such as in hypoid gear teeth. Such mechanisms also fail by fatigue,
but the increased traction causes the maximum shear stresses to be nearer the
surface.
In practical applications the stresses transmitted by the oil films can be
extremely high and EP lubricants are often necessary. In this case severe wear
by adhesion is prevented at the expense of a low rate of wear due to corrosion,
as shown later. This wear rate can be sufficiently low for eventual failure to be
caused by fatigue. However, if some corrosive element such as water contami-
nates the lubricant, accelerated failure due to the combination of high stresses
and corrosion, stress corrosion, can result.

Sliding Contact
When sliding surfaces make contact via asperities we have seen that wear by
adhesion and abrasion can take place. However, it is conceivable that asperi-
ties can make contact without adhering or abrading and can pass each other,
leaving one or both asperities plastically deformed. After a critical number of
such contacts an asperity would fail due to fatigue, producing a wear frag-
ment. It is difficult to prove directly that fatigue is a major cause of wear in any
given set of conditions. Kragelskii attaches much importance to this mode of
wear, as is implied by his definition of wear, given earlier in this chapter. He
interprets many of his results and those of other investigators in terms of
fatigue. Lancaster 14 has suggested that the low wear experienced when
rubbing electrographite against metals may take place by fatigue, and
Archard and Hirst 15 have suggested that the metal transferred by adhesion is
finally detached by a fatigue process. A model of a low wear regime presented
by workers at IBM 16 and, rather unfortunately, described as 'Zero Wear',
invokes fatigue and corrosion as the only mechanisms responsible for wear in
this regime.
The factor k in equation 5.1 has been interpreted as the probability of an
asperity contact producing a wear particle, without any physical explanation
as to the method of their production. Also, although the adhesive-wear
theory can explain transferred wear particles, it does not explain how
loose wear particles are formed. In particular, the occurrence of wear of the

105
harder of the two rubbing surfaces is difficult to understand in terms of the
adhesion theory.
All these points can be explained on the assumption that wear is a fatigue
process. The k factor is now interpreted by assuming that a wear particle is
produced when an asperity has experienced a sufficient number of contacts
and deformations to produce a fatigue fracture. When this occurs, a loose
wear particle is produced and of course, this mechanism explains the pro-
duction of wear particles from both the harder and the softer of the two rub-
bing surfaces. The fatigue mechanism does not exclude the possibility of
transfer by an adhesive mechanism and therefore it appears that most of the
wear phenomena can be explained at least qualitatively in terms of fatigue.
Kragelskii 1 has attempted to produce a quantitative fatigue theory of
wear. He uses the results ofTavernelli and Coffin 17 who have shown that for a
wide range of materials the plastic strain produced on each fatigue cycle is
related toN, the number of cycles to failure, by the equation

[2;;iiJ = N

where eP is the plastic strain amplitude per cycle and erail is the plastic strain
at failure in a tensile test. However, Kragelskii appears to have used the total
strain rather than the plastic strain in his derivation, and in addition he
makes a rather obscure estimate of the strain amplitude in terms of asperity
geometry. Thus no satisfactory wear theory in terms of fatigue has yet been
developed although the mechanism appears to be more realistic than that of
the adhesion theory. In the opinion of the authors, it will be necessary to
develop such a theory in order to advance our understanding of wear pheno-
mena.

5.2.4 Corrosive Wear

When rubbing takes place in a corrosive environment, either gaseous or

liquid, then surface reactions take place and reaction products are found on
one or both surfaces. These reaction products are commonly poorly adherent
to the surfaces, and further rubbing causes their removal. This process is
then repeated. Thus corrosive wear requires both corrosion and rubbing.
The rate of growth of, say, an oxide film on a steel will decrease exponentially
with time, and therefore unless the oxide film is removed by rubbing, the
metal-to-oxide reaction will rapidly become negligibly small. The detailed
operation of corrosive wear is extremely complex. The reaction products
depend on the exact composition of the environment. For instance, small
quantities of water vapour in air cause the reaction product to be the hydr-
oxide rather than the oxide. Mechanisms operating in an industrial environ-
ment or near the coast will generally corrode more rapidly than those
operating in 'clean' air.

106
The most common liquid environments are aqueous and here small
amounts of dissolved gases, commonly oxygen or carbon dioxide, influence
corrosion. Corrosion is also influenced by the relative electropotential of
rubbing metals.
The presence of a lubricant usually protects the surfaces from the corrosive
environment. However, it is not uncommon for corrosive elements to be
dissolved in lubricants, for example, water in oil, and also lubricants can be
degraded in time and become progressively more corrosive.
Some mechanisms, notably hypoid gears, are required to work under
conditions where severe adhesion is possible, that is, at a pressure greater
than H/3. To prevent catastrophic metallic welding in these situations
lubricants with mildly corrosive additives are used, that is, EP additives.
Where corrosion is a major cause of wear, there is usually a complex inter-
action between various mechanisms. The original wear of surface films can
be due to either adhesion or abrasion. Since many commonly occurring films,
notably iron oxides, are abrasive, the corrosion and abrasion will combine.
High contact stresses can cause enhanced corrosion locally, leading to
pitting. It is well known that internal stresses in metals, caused during form-
ing operations, cause stress-corrosion cracking when in a corrosive atmos-
phere. This, combined with surface rubbing, can result in catastrophic
wear, as can the corrosion of a single-phase in a two-phase alloy bearing.
Corrosion is not always a deleterious phenomenon. Oxide films and other
corrosion products prevent adhesion of metal asperities and metallic wear in
vacuum, where oxide films cannot form, is generally very high.

5.2.5 Fretting

The wear phenomenon known as fretting is not a distinct mechanism, but it

is convenient to treat it separately.
Fretting occurs where low amplitude vibratory motion takes place between
two metal surfaces loaded together. This is a common occurrence, since
most machinery is subjected to vibration, both in transit and in operation.
Examples of vulnerable components are shrink fits, bolted parts, and splines.
Basically fretting is a form of adhesive wear, the normal load causing
adhesion between asperities and the vibrations causing rupture as described
earlier. Most commonly, fretting is combined with corrosion, in which case
the wear is known as fretting corrosion. In air the corrosion product is oxide,
and the characteristic fine reddish-brown powder produced from steels is
known as 'cocoa'. Here the initial wear debris is oxidised. The oxide particles
are abrasive and because of the close fit of the surface, cannot readily escape.
Further oscillatory motion causes abrasive wear and oxidisation, and so on.
Fretting wear can cause the formation of surface stress raisers and, if the
vibratory stresses are sufficiently high, fatigue cracks are propagated leading
to complete failure.

107
5.2.6 Erosion
The term erosion is usually taken to cover that form of damage experienced
by a solid body, when a fluid, which may contain solid particles, impinges on
to the surface of the body. Until fairly recently erosion has been considered a
minor source of wear, and many reviews of wear refer to it only briefly, if at all.
However, with the continuing development of mechanisms operating at high
speed, together with the need for materials with high strength-to-density
ratios, erosion has gained more prominence. This is exemplified by carbon-
fibre reinforced plastic turbine blades, whose leading edges require special
coatings to increase erosion resistance.

Erosion by Solid Particles

When a stream of solid particles is directed at a surface it is found that the
wear rate is dependent on the angle of incidence of the particles, and that the
wear rates for ductile and brittle solids follow different curves 18 as shown in
figure 5.5. This suggests two different mechanisms of erosion, dependent on
whether the eroded solid is ductile or brittle.
In the case of a ductile solid it is thought that the mechanism near to the
peak is similar to that of abrasion. At angles close to 90° for a ductile material

Angle of ottock 0

Figure 5.5 Dependence of rate of erosion on angle of

attack of impinging particles

108
a fatigue mechanism is probably predominant. For brittle materials surface
cracks are formed which link to give the wear particles. In both forms of wear,
ductile and brittle, the wear rate is proportional to the kinetic energy of the
impinging particles, that is, to the square of the velocity of the particles.

Fluid Erosion
When small drops of liquid are incident on the surface of a solid at high
speeds (1000 m/s)· very high pressures are experienced, exceeding the yield
strength of most materials. Thus plastic deformation or fracture can result
from single impact, and repeated impact leads to pittipg and erosive wear.

Cavitation Erosion
Cavitation erosion arises when a solid and fluid are in relative motion, and
bubbles formed in the fluid become unstable and implode against the surface
of the solid. Damage by this process is found in such components as ships'
propellers and centrifugal pumps.
The stability of a bubble is dependent on the difference in pressure between
the inside and outside of the bubble, and the surface energy of the bubble.
This last factor is a measure of the energy released by collapse of the bubble,
and of the potential damage at collapse. Thus a reduction of surface tension
of the liquid reduces damage, as does an increase in vapour pressure.

Spark Erosion
When an electric spark occurs between two surfaces permanent damage in
the form of metal removal and deposition results. This is a well-known
problem in the field of electrical contacts. When the contacts are rubbing,
as in the case of a commutator, the sparking damage on the copper commuta-
tor can then cause excessive wear of the brush by abrasion. Damage can be
cumulative since, if a particular commutator bar is recessed slightly and
sparking takes place, then the brush-bar contact will deteriorate causing more
severe sparking and even melting of the copper.
Spark erosion is used as a method of metal removal or cutting where
normal machining is difficult, for example, for very hard metals, and also
in cases where it is important to restrict the amount of subsurface damage,
for example strain-free machining of metal single crystals.

5.2.7 Size of Wear Particles

We have now reviewed the various wear mechanisms but have not con-
sidered the factors influencing the size of the wear particles produced.
Rabinowicz 19 has suggested that, if a wear particle is to be formed, the
elastic energy stored prior to detachment must be greater than the energy

109
for the new surface created. This approach leads to a criterion for whether or
not an asperity contact produces a wear particle, and thus to a value of k.
However, as is shown below, the theory is open to criticism.
Rabinowicz considers the energy stored in a hemispherical particle ad-
hering to one surface, after yielding has taken place. The stored energy per
unit volume of such a particle is p;/2E, where Po is the yield pressure, E is
Young's modulus, and the volume of the particle is ~1tr 3 . The surface energy
of newly created surfaces, if the hemisphere is detached along its diametral
plane, is 21tr 2 y, where y is the surface energy per unit area. Then according
to Rabinowicz

or
6Ey
r>-
p;
If Po is replaced by H/3 for a metal, then
54Ey
r>--
H2
The exact value of the constant has been arrived at by assuming a hemi-
spherical shape for the particle. For particles of other shapes this equation
can be replaced by
KEy
r>-- (5.7)
H2
where K is ~54 but depends on the particle shape.
Rabinowicz then goes on to say that H/E is roughly constant, and this
leads to
K'y
r>- (5.8)
H
Rabinowicz interprets this to mean that only wear particles above a certain
critical size, dependent on yjH, can become detached. However, examination
of table 5.1 indicates that HIE can vary over a wide range and that equation
5. 7 is more realistic than equation 5.8. Furthermore, Levy, Lindford and
Mitchell 20 argue that, in a rubbing situation, the value of r given in equation
5.7 should be regarded not as the minimum, but as the maximum value
obtainable. It is certainly difficult to see how r could ever exceed the right-
hand side of equation 5.7, and it is highly probable when other forms of
energy, such as the kinetic energy of the rubbing surfaces, are available that
wear particles will be detached before r has reached the critical size. Thus, the
relevant relationship becomes
KEy
r<-- H2 (5.9)

110
This suggestion is supported by the experimental work of Aghan and
Samuels 21 , who show that under abrasive polishing conditions, the wear
debris consists of particles several orders of magnitude smaller than the size
predicted by Rabinowicz.

5.2.8 The IBM Engineering Model for Wear

The expressions presented in the previous sections have attempted analyti-
cally to relate wear to conditions of sliding, such as load and speed, and
material properties, such as hardness and Young's modulus. None of these
expressions has gained universal acceptance, and because of this, workers
at the IBM Endicott Laboratory 16 have attempted to obtain empirically,
expressions for wear which could be directly applied to the design of rubbing
components. Such expressions have been obtained and appear to have wide
applicability. The work at IBM can be divided into two parts. In the first,
expressions are obtained which define the amount of rubbing which can be
accommodated before the depth of wear is greater than the initial surface
finish. In the second, expressions are presented which can be used to predict
wear greater than the depth of the original surface finish. The two parts are
entitled Zero Wear and Measurable Wear.

Zero Wear
First, a unit of sliding called a 'pass' is defined. This is the unit of sliding in
which sliding has been experienced over a distance equal to the width of the
apparent contact in the direction of sliding. For example, in the pin-on-
cylinder wear apparatus shown in figure 5.6, a single pass would be of length

Figure 5.6 The concept

of a pass in the IBM
model of wear. In one
complete revolution the
cylinder experiences one
pass and the pin experi-
ences 2rtr/l passes

111
I and for a complete revolution of the cylinder 2rtr/l passes would be ex-
perienced.
It has been found that the condition for zero wear (wear Jess than the
surface finish) in a single pass is

rmax < 0.54 r, (5.10)

where rmax is the maximum shear stress experienced and r, is the shear
stress necessary to cause yielding.
For zero wear in 2000 passes the condition is

rmax<y,r, (5.11)

where y, is a constant which is experimentally determined and can have

values of either 0.54 or 0.20.
The condition for zero wear for M passes is related to the condition for
zero wear in 2000 passes by the expression

(5.12)

The IBM workers have found that these expressions are valid for a large
range of materials, both metals and plastics, for various geometries, for
various lubricants and for different surface finishes.
The simplicity of the expressions is somewhat surprising. They appear to
be independent of surface finish, but it should be remembered that zero
wear is already defined in terms of the original surface finish. The fact that
y, takes only two values 0.54 and 0.20 has no theoretical explanation as yet,
but the choice of y, seems to be related to the tendency for material transfer,
0.54 being appropriate with a lower tendency for transfer and 0.20 with a
higher tendency for transfer. Palmgren's equation 5.6 is similar to equation
5.12 relating the load Pat which a roller bearing is operated toN, the number
of cycles to failure. That is

If the loads are converted to stress using the Hertz equation, then a ninth-
power relationship results.
All the necessary material constants appear to be contained in rmax•
which is dependent on the friction force, that is, on the hardness (or Young's
modulus) and on the shear strength at the junctions. Parameters such as
speed and temperature do not appear to be taken into account in this model.
However, these affect the material constants implicit in rmax. For instance
it was found that with certain plastics at very high rubbing speeds, the results
did not agree with predictions. However, it was found also that the high
speeds had caused melting of the plastics, thus modifying the material
properties. Using the corrected values for r, and rmax agreement between
predicted and experimental results was obtained.

112
Measurable Wear
Here the amount of wear Q is related to the number of passes M, and also
to the amount of energy E, dissipated in the form of wear during each pass.
The appropriate relationship relating these quantities was found to be the
differential relationship

dQ = (~)M dE+(:~ )E dM (5.13)

Two types of wear are distinguished, that involving severe transfer and that
involving only a moderate degree of transfer. The second type is more
common in practical systems and will be treated here.
After some simplifying assumptions, the following equation is obtained

d[ Q/) 912 ] = C dM (5.14)

(rmax

where Cis some constant for the system and I is the magnitude of sliding in a
single pass that is, the length of the apparent contact area in the direction of
sliding. Q is obtained by substituting the expression for rmax and I into the
equation and integrating. The constant C must in general be obtained
experimentally by determining the amount of wear for a particular number of
passes. However, Bayer 22 has indicated that the zero wear and measurable
wear models can be combined to give an analytical expression for C. In this
case the value of}', (either 0.54 or 0.20) must be found experimentally, as must
the condition of transfer, high or low.
The IBM workers have presented, in a series of papers, methods of applying
both the zero wear and measurable wear models to practical design problems.

5.3 VARlO US FACTORS AFFECTING WEAR

In the following sections we will examine a number of factors which can cause
considerable variation in the wear rates of rubbing surfaces. Although it is
convenient to consider these factors under different headings, we will see
that they interact and it is difficult to separate them one from another. For
instance, high temperatures at the surface will be generated by high loads and
speeds. The temperature influences surface film formation and can cause
changes in the surface structure and hardness. A number of the factors
considered below are treated in more detail in chapter 6.

5.3.1 Effect of Surface Films

Since both friction and wear are essentially surface phenomena the presence
and properties of films covering rubbing surfaces have a critical effect.
We will first consider wear under conditions where no surface films are
present, that is, in high vacuum.

113
Wear Under Vacuum
A comprehensive study of friction and adhesion under a vacuum of
10- 8 mmHg was carried out by Bowden and Rowe 23 . They found that under
the action of normal and tangential forces, junction growth occurred,
unimpeded by the presence of surface films. Thus welds were made over
comparatively large areas of contact resulting in high friction coefficients.
In order for tangential, that is, rubbing, motion to proceed these welded
bonds must be broken. As discussed earlier, the wear taking place depends
on the exact position of the fracture relative to the junction, but obviously,
the larger the area of the junctions, the larger the wear rate. In fact most
metals wear catastrophically under these conditions. These experiments of
Bowden and Rowe were carried out largely in order to investigate funda-
mental properties. However, the high wear experienced under vacuo is now
a problem of practical significance in high-flying aircraft, missiles and space-
craft. A great deal of work has now been accomplished aimed at under-
standing friction and wear under high vacuum, and also at providing prac-
tical bearings under these conditions. Much of this work is summarised by
Bisson 24 • Conventional lubrication using oils and greases is of course
impossible over long periods. Possible solutions to the problem are the use
of low shear strength metal films, plastics, or solid lubricants. A fairly
recent development has been the use of hexagonal close-packed metal as a
bearing material in space environments, but this will be discussed more fully
later.

Oxide Films

Most metals are covered by an oxide film and even after cleaning by machin-
ing or grinding acquire a film of oxide of between 5 and 50 molecular layers
in five minutes or less. When an oxide film covering an asperity is removed
by rubbing in a normal atmosphere the clean metallic surface will be covered
by a monomolecular layer almost instantaneously. This must be taken into
account in all considerations of metals rubbing under atmospheric con-
ditions.
Unless the loads are very small, the oxide film does not prevent inter-
metallic contact, as can be seen from the results of Wilson 25 who measured
the electrical contact resistance between various rubbing metal pairs,
metallic contact being indicated by a sharp drop in the contact resistance.
Where asperities make metallic contact, localised welding can and does take
place, but as has been shown earlier (chapter 4), the oxide film prevents the
junction growth which occurs in vacuo, and thus reduces both the friction
and the wear rate.
Many investigators have found that the wear of metals can be divided into
two regimes, mild wear and severe wear. In the mild wear region, occurring

114
at lower loads, the contact resistance is high, the wear debris is fine and
consists mainly of metal oxide, and the rubbed surfaces become polished.
In the severe wear region, at higher loads, the contact resistance is low, and
wear debris includes coarse metallic particles and the rubbed surfaces are
rough. This type of result is represented by figure 5.7, taken from Archard and
Hirst 15 .

10

-;;;
E
.t:.
2
. u
c:
~ ;;;
~
u
10- 1 l2
c:
0
u

Contact resostance

10- 11 " - - - - - - - - - ' - - - - - - - - - - L . . - - - - - - ' 1 0 - 3

0·1 10 100
Load (kg)

Figure 5.7 The variation of wear rate and contact resis-

tance for 60/40 brass pin

The properties of the oxide film are also important. For instance, the hard
brittle oxide formed on aluminium provides poor protection against heavy
wear. Barwell 26 presents an unexpected result as follows: steel was rubbed on
steel in a vacuum of 10- 4 mmHg; the wear was lower than when air was
admitted to the apparatus. The interpretation of these results is that a thin,
tough and tenacious oxide film was formed on the steel at 10- 4 mmHg
whereas when air was admitted the oxide film formed was thicker but less
protective.
115
In studies of the wear of copper slip-rings it has been found that under the
negative brush, the rate of copper oxidation was enhanced but the resultant
film was poorly adherent, giving higher wear rates. Under the positive brush,
the rate of copper oxidation was reduced, the film was then tough and ad-
herent giving low wear rates. At higher currents, high wear rates were re-
corded under the positive brush, apparently as a result of complete cathodic
reduction of the oxide film.
In this section we have considered the effect of oxide films on adhesive
wear, but of course wear due to a combination of chemical attack and rubbing
is defined as corrosive wear. Oxide films will also have a large influence on
abrasive wear since many metal oxides are hard and when present in the
form of wear debris act as abrasive particles.

Boundary Lubrication
The term 'boundary lubrication' refers to the situation where an oil film is
present between two rubbing surfaces but its thickness is insufficient to
prevent asperity contact through the film.
Many mechanisms operate totally under conditions of boundary lubrica-
tion. Others are designed to operate under full hydrodynamic lubrication, but
as the oil film thickness is a function of speed, (see chapter 10) during starting
up and running down the film thickness will be insufficient to provide
complete separation of the surfaces, and the conditions of boundary lubrica-
tion will be obtained. We may initially consider the action of a boundary
lubricant in much the same way as that of an oxide film, that is, it limits, but
does not prevent, metallic contact at asperities, and inhibits junction growth.
Further, a lubricant can reduce wear by limiting the access of a corrosive
liquid or gas when a bearing is operating in a hostile environment.
The effectiveness of a boundary lubricant is less dependent on the rheo-
logical properties of the lubricant than it is on its chemical properties.
It is found that liquid fatty acids are more effective than alcohols with a
comparable chain length. It is generally accepted that this is due to the more
reactive fatty acids being chemically absorbed on the metal surface, and it has
been found that one or two molecular layers are effective in reducing wear
by a factor as great as 1000. However, metals that do not react with the fatty
acid are lubricated equally well by an alcohol of similar chain length. Thus
we see that the effectiveness of boundary lubrication is increased in those
cases where a solid metallic soap film is formed over the metal surface.
Under the very high pressures developed in hypoid gears, organic boundary
lubricants are ineffective since they break down due to the high local tempera-
tures at the contacts. To counteract this extreme pressure, additives, for
example, organic chlorine or sulphur compounds, are added to the lubricant.
These are stable at normal temperatures but react with the metal at the hot
spots, forming metal chloride or sulphide films which inhibit welding of the
asperities and reduce adhesive wear to acceptable levels.

116
Solid Lubricants
Solid lubricants can be applied to bearing surfaces by means of an adhesive
such as a resin, or they may be introduced in powder form between two
rubbing metals at low loads, in which case they adhere to the metal surfaces,
eventually forming a fairly coherent covering. Also a solid lubricant in a
fine powder form may be added to a liquid lubricant where again the rubbing
surfaces eventually acquire a coating of the solid lubricant. The action of the
solid lubricant is added to the oil in a motor car engine, 'running-in' will
films, that is, it acts to reduce the number of metallic contacts and inhibits
junction growth. Furthermore, it provides a low shear strength interface.
The effectiveness of reducing wear is demonstrated by the fact that if a
solid lubricant is added to the oil in a motor car engine, 'running-in' will
take almost twice as long as under conventional lubrication. 'Running-in'
consists of removal by wear of the larger asperities produced by machining
operations. Solid lubricants are used to reduce wear in those situations
where a conventional lubricant (oil or grease) cannot be used-in high
vacuum (space craft) or at high temperatures. They are also useful in prevent-
ing wear for short periods in emergencies, such as the complete loss of oil
from a car engine.

Other Surface Layers

The wear between metals may be reduced by coating the surfaces with a thin
layer of metal which has higher wear resistance than the substrate metal.
This is obviously sound economic practice since many of the metals found
to be wear resistant are expensive, but the cost of an effective film thickness
can be very low.
Rhodium and chromium are both hard, and electroplated coatings of
these metals have been used successfully in the protection of cylinder liners,
crankshafts and similar applications. The wear resistant properties of these
films have been demonstrated by Moore and Tabor 27 . Very successful hard
facing and wear resistant coatings are deposited using spraying and fusing
techniques. These coatings, based largely on cobalt, chromium and iron have
good hot-hardness properties and are also highly corrosion resistant.
Films of soft metals such as indium and lead have been used on harder
metal substrates. An example of their effective use has been in lubricating
the dies used in deep drawing operations.
Besides these methods involving the deposition of a surface layer on to a
substrate, there are a number of techniques where the surface of the substrate
is subjected to chemical action to increase the wear resistance of the surface.
Examples are the phosphating and sulfinuzing processes for steels. Ferrous
surfaces can also be hardened by nitriding and carburising. Also hard surface
layers can be produced by heat treatment. For instance, large cast-iron rolls
for rolling mills are quenched so that the higher rate of cooling at the surface

117
produces a hard 'white iron' surface whereas the centre is composed of Jess
brittle 'grey iron'.
Finally, mention should be made of the fact that the rubbing action itself
produces changes in the surface layers. It has been found that prior work-
hardening has no effect on abrasive wear resistance as the rubbing action
work-hardens the surface to a maximum value. References to modified
surfaces in rubbed ferrous materials are common, although the exact form
of the modified material has rarely been analysed. A material referred to as a
hard 'white etching' layer is found on rubbed cast-irons and steels. The forma-
tion of these layers is probably associated with hot spot temperatures being
sufficiently high to take the carbides into solution, followed by rapid cooling.

5.3.2 Effect of Temperature

The temperature of rubbing surfaces influences the wear in three major ways.
It can alter
(a) the properties of the rubbing materials
(b) the form of the surface contaminating film
(c) the lubricant properties.
We will consider these in turn.
In general the hardness of metals is temperature dependent, the higher the
temperature, the lower the hardness. Thus as the tendency for asperities to
adhere and the wear rate increase with decreasing hardness, in the absence
of other effects, they also increase with increasing temperature. This effect
has been demonstrated by M. J. Hordon 28 • In order to counteract this
effect it is necessary to use metals with high hot-hardness for bearing materi-
als operating at high temperatures. Metals commonly used at high tempera-
tures include tool steels, and alloys with base composition of cobalt, chrom-
ium and molybdenum. However, at temperatures above about 850 °C, it is
necessary to use cermets or ceramics.
A quite different effect of temperature is a temperature-induced phase
change causing the properties of a bearing material t'l? alter radically. An
example of this is described in detail in chapter 6.
In normal atmospheres, most metals are covered by an oxide film. The
exact form and thickness of film is dependent on the temperature of forma-
tion, and again this is discussed in chapter 6.
Another effect of frictional heating, when rubbing ferrous materials in
normal atmospheres, is reported by Welsh 29 • He found that at low loads wear
was high but at higher loads the wear rate dropped to a very low value. A
metallographic study indicated that this fall in the wear rate was due to the
formation of a hard surface layer, which Welsh attributed to interaction with
the atmospheric nitrogen at the temperatures generated under the higher
loads.

118
If the temperature of operation of a bearing lubricated by an oil is raised,
deterioration is caused, first by oxidation of the oil and then by thermal
degradation. This places a limit beyond which organic fluids are ineffective
in reducing wear, and at high temperatures alternative lubricants must be
employed, for example, solid lubricants, such as graphite and molybdenum
disulphide.
Oxidation and thermal degradation cause irreversible changes in the lubri-
cating properties of an oil. In general, under conditions of boundary lubri-
cation, greatest protection is provided when the lubricant is solid. If the
temperature of such a lubricant is raised beyond the melting point, wear rate
increases but the change is reversible, since the lubricant again provides
protection when the temperature is lowered. A fatty acid lubricating reactive
metal surfaces will provide good wear resistance beyond its melting point
due to the formation of the metallic soap.

5.3.3 Effect of Load

An increase in load causes an increase in the frictional force, and hence a

temperature rise which produces the effects discussed earlier. Also, an in-
crease of load can cause a transition from mild wear to severe wear. This
occurs at about the point when

where W = load
Aa = apparent area of contact
H =hardness
and this has been demonstrated by Burwell and Strang 6 (see figure 5.2).
The transition from mild to severe wear is attributed to the interaction of
the plastic zones beneath the contacting asperities 30 . During mild wear, the
situation is as shown in figure 5.8, where there is no interaction between the
plastic zones. As the load is increased, the plastic zones interact as in figure
5.9, and the subsurface regions become entirely plastic. Beyond this point
Amonton's law is no longer obeyed, and severe wear occurs.

Figure 5.8 The independent plastic zones under the

contacting asperities of a plane slider under relatively
low loads

119
Figure 5.9 The interation of the plastic zones of a
plane slider under high loads

5.3.4 Effect of Compatibility

Rabinowicz 31 has suggested that metal pairs with low metallurgical compati-
bility will exhibit low friction and low wear. He defines 'metallurgically
compatible' metals as those which show a high degree of mutual solubility.
Thus compatibility ratings can be obtained from binary phase diagrams.
Rabinowicz has attempted to correlate adhesion, friction and wear results
from twelve different published studies, with compatibility of the metal pairs
used. He found zero correlation of compatibility with adhesion, a positive
correlation with friction and a greater correlation with wear.
This suggests that metal pairs for boundary lubricated or unlubricated
sliding should be chosen to have low mutual solubilities, in order for the wear
rate to be low. Certainly it has been generally accepted that in selecting metals
for bearings, similar metals, that is, metals with 100 per cent compatibility,
should be avoided.

5.3.5 Crystal Structure

Hexagonal metals give much lower friction coefficients when rubbing against
themselves than do face centred cubic and body centred cubic metals. This
is discussed in some detail in chapter 6 and will not be considered further here.

5.4 EXPERIMENTAL ASPECTS

Laboratory wear tests can be listed conveniently under two broad headings.
First, there are the tests in which specimens with convenient geometries are
run against each other; parameters which influence wear are varied one
at a time, and attempts are then made to correlate wear rate with the various
parameters. Tests under this heading range from simple comparative wear
studies to closely controlled experiments aimed at a better understanding of
fundamental wear mechanisms. Such simplified tests often produce results
which have little direct relevance to the complex wear situations met in
practice. Therefore in order to study the more practical problems of wear it is

120
more usual to perform a second type of test, which simulates the practical
situation. Often in this type of test the conditions simulated are the worst
wear conditions that are likely to be met in practice, and in some tests the
conditions are greatly accentuated in order to reduce testing time. Results
from tests of this type have limited applicability, and are unlikely to provide
any fundamental information.

5.4.1 Wear Apparatus

There are a number of specimen geometries in common use for simple wear
tests, and these are listed below.
The four-ball arrangement is one in which a single ball is rotated under
load against a cluster of three balls. This is mainly used as a comparative
test in assessing the performance of lubricants.
In the pin-on-disc type of machines, a stationary cylindrical pin, either
hemispherically, conically or flat ended is loaded against a rotating flat
disc or annulus. Alternatively two annuli, one stationary and one rotating,
may be used.
The pin-on-cylinder geometry again has a stationary pin but here its end
bears on the surface of a rotating cylinder. In this arrangement the cylinder
may also have a linear movement along its axis so that the pin is always
rubbing on virgin cylinder surface and the wear track is a spiral. The crossed-
cylinder apparatus has a stationary cylinder loaded with its cylindrical
surface against a rotating cylinder, with the cylinder axes at right angles.
Again, the rotating cylinder may have a superimposed linear motion.
In order to simulate combinations of sliding and rolling two driven discs
are loaded against each other. By varying the relative speeds of the discs
pure rolling or any combination of sliding and rolling can be studied.
Each of the above arrangements has its own characteristic advantages and
disadvantages. For example, the flat-ended pin-on-disc and the annulus-on-
annulus retain their macroscopic geometry during a run, but the pin-on-disc
is assymetric in that the disc material is open to the atmosphere over the
major part of the run. In the annulus-on-annulus both surfaces are bearing
continuously and exposure to the atmosphere is limited. Also, in this arrange-
ment it is more difficult for wear debris to escape. The crossed-cylinder
apparatus is probably the simplest apparatus to build. The macroscopic
geometry changes as wear proceeds, the size of the wear scar being represent-
ative of the total wear. An example of a crossed-cylinder apparatus is shown
in figure 4.2.

5.4.2 Precautions Necessary in Wear Measurement

Wear measurements are notoriously variable. Experiments on the same
machine and in the same laboratory can give widely differing results at
different times. Such variations are almost always due to lack of adequate

121
control of the conditions under which the surfaces are rubbing. For example,
the ingress of extraneous grit may cause the wear rate to increase, small
amounts of water vapour may alter the oxide film growth and handling of
specimens can deposit contaminating films of grease. Thus surfaces must be
prepared and cleaned in a reproducible way. Atmospheric conditions must
be constant, and ingress of abrasive material must be prevented. An extreme
example of these controlled conditions is in measuring wear under space
environments. Here the specimens are prepared mechanically, electro-
polished, and cleaned by electron bombardment and run under pressures of
10- 12 mmHg. An example of such a rig is shown in figure 4.3b.

5.5 WEAR PREVENTION

When two surfaces in contact slide relative to each other, wear will inevitably
take place. The only way to prevent wear is to separate the surfaces. This
implies continuous running under hydrodynamic conditions. When a bear-
ing is run down or started then some asperities will make contact. Under
these conditions wear can be kept low by the use of adequate boundary
lubricants or solid lubricants and by material selection. Hard materials wear
less rapidly than the soft ones and the rubbing of similar materials should be
avoided. Surfaces should be smooth, extraneous particles must be excluded
from bearings by filtration of the air and lubricants, and corrosive atmos-
pheres must be excluded where possible.
In practical situations wear is inevitable and in many cases bearings are so
designed that one member has a very low wear rate whereas the mating
member is considered replaceable, and has a higher wear rate. For example,
the crank shaft in an internal combustion engine is costly to replace, and it is
therefore made from a hard steel and is supported in relatively cheap bearing
shells, of much softer metals (lead-tin, copper-lead, aluminium-tin alloys).
Use of a softer metal bearing has additional advantages. It can deform easily
to redistribute local high loads which might be caused by shaft distortion
or misalignment, and extraneous abrasive particles can be absorbed. Even
under extreme conditions, such as total loss of lubricant, because of the low
melting point of many bearing materials damage to the shaft can be avoided
over a short period of time.

5.6 APPLICATION OF WEAR RELATIONSHIPS TO DESIGN

One ultimate objective of wear studies is to be able to design components

(in this context design includes material selection), that will run under a given
set of conditions with acceptable known values of wear rate. None of the
theories presented earlier enables bearing life to be predicted from a know-
ledge of load, speed, material properties and bearing geometry, but some
ways of using present knowledge to give design criteria are described below.

122
5.6.1 Application of Archard's Equation

If we wished to apply Archard's 4 wear equation 5.1 to a bearing operating

under conditions of boundary lubrication, then that specific bearing type
would have to be run and the wear rate measured in order to find k for those
particular conditions. If subsequently the conditions were changed, say by an
increase of load, and k was found to be the same as before, then Archard's
equation could be applied at all intermediate loads. In other words, having
found k experimentally over a range of conditions, we can use Archard's
equation to interpolate but not to extrapolate. Changes of geometry, speed
or load could each cause variations ink. However, we can use this equation in
a qualitative way, for example, an increase of load will generally cause an
increase in wear rate, an increase in material hardness will generally reduce
the wear rate.

5.6.2 Applications of the IBM Model

The IBM models for wear, described earlier, are an attempt to improve this
situation. In a number of publications the IBM team have described how the
wear models can be applied to practical situations. The following example is
taken from one of these papers 32 , and applies the 'zero wear' model to the
design of the steel-on-steel cam and follower shown in figure 5.1 0. The prob-
lem in this case is to ensure a satisfactory wear performance, that is, depth of
wear is less than the original surface finish, for 106 revolutions of the cam
under a load of 1 oz. Using the Hertz theory the maximum shear stress is

Figure 5.10 Cam and

follower to be designed
for 'zero' wear

123
found to be 7.95 x 10 3 lbf/in 2 . The value of/, the dimension of the contact
area in the direction of sliding is found, again from Hertz theory, to be 0.002 in.
Now because the follower is in constant contact, as opposed to the cam
which is in contact only over 2rtr/l of a revolution

M cam = 106 passes

Mrollower = 6.28 X 10 9 passes

The value of y, must be determined experimentally, but once determined

this value is appropriate for this combination of materials and lubricant
for all loads. We now assume that}', has been determined and that in this case
it had the value 0.54.
The values of yield points in shear are

!y(cam) = 4.0 X 104 lbfjin 2

ry(fol) = 1.50 X 10 5 lbf/in 2

By substituting the above values in equation 5.12 the endurance limits can
be calculated as

2 X 103)1/9
( !;,1- }'!y(fol) = 1.552 X 104 lbfjin 2
(fol)

for the follower, and

2 X 103)1/9
( M }'!y(cam) = 1.09 X 104 lbfjin 2
(cam)

for the cam. It can be seen that both these endurance limits are greater than
rmax, and therefore the design as specified is adequate.
If the endurance limits had been calculated as lower than rmax then it
would be necessary either to redesign the bearing and recalculate the endur-
ance limits or, as an alternative approach, to accept a different criterion for
acceptable wear and use the IBM Measurable Wear Model in order to
calculate the wear rate for the system.

5.6.3 The P-V Factor

An approach that has been widely used for bearings operating under bound-
ary lubrication conditions, is to specify the P- V factor for a particular
bearing. This factor is the product of load per unit area and speed, and is
found empirically for each type of bearing. The use of the P- V factor is
considered more fully in chapter 6.

124
5.7 AN EXAMPLE OF WEAR IN PRACTICE-WEAR OF AN
I.C. ENGINE

Prior to running, the various pairs of bearing surfaces in a new engine are
not 'mated together'. There may be slight initial misalignments and there will
certainly be 'high spots' on all surfaces. Bearing clearances will initially be
small and therefore the cooling flow of oil is low and this, together with the
initially higher friction, leads to higher than normal bearing temperatures.
'Running-in' is the process of removing high spots and mating the bearing
surfaces; also the surfaces work-harden and become more wear resistant.
During this period wear is higher than in normal running as the 'high spots'
cause more asperity contacts, the misalignments reduce apparent contact
areas and increase bearing pressures, and the higher temperatures usually
cause higher wear rates.
If 'running-in' proceeds satisfactorily the 'high spots' are removed by
adhesive wear and by plastic deformation. The wear gradually increases the
real area of contact until it equals the design contact area, and the con-
tact pressure drops. Frictional losses decrease during this period and bearing
clearances increase, thus reducing the surface temperatures. The wear rate
decreases due to all the above causes until it reaches the normal steady-state
wear rate for the engine. However, if the initial misalignments are sufficiently
high the apparent bearing surface area might be so low as to cause the pressure
to be higher than one-third of the hardness. In this case, the wear rate is
excessively high as shown by figure 5.2 and the real contact area may not
grow sufficiently to reduce the pressure to less than H/3, that is, to reduce the
wear rate to normal levels, before large-scale damage or even seizure has
occurred.
The wear rate during 'running-in', even when misalignments are minimal,
is higher than during normal running. Thus a greater quantity of wear
debris is produced and this, together with small amounts of casting sand and
machining swarf which are often present in new engines, is filtered out of the
lubricating oil to avoid excessive abrasion.
After the 'running-in' period, the two major sources of wear are abrasion
and corrosion. Abrasive wear is kept to a minimum by the exclusion of foreign
particles, and by efficient oil filtration. Corrosion is thought to be a major
mechanism of wear of cylinders and piston rings in car engines although
modern oil formulations have reduced corrosion considerably. The presence
of sulphur and additives in the petrol may be one cause of corrosion, but the
major cause is carbonic acid formed from the combustion products, C0 2
and water. Burwell 2 describes the situation as follows:
A car is started from cold in the morning and driven to work, where it is
parked and cools down, possibly to sub-zero temperatures. In the evening the
process is repeated. Thus two cycles from cold to hot to cold are repeated
daily even though the total distance driven might be small. Each time the

125
engine is cooled the carbonic acid condenses on the cylinder walls, producing
abrasive corrosion products (e.g., Fe 2 0 3 ). The next start-up wears away these
corrosion products producing additionally some abrasion. This type of
mechanism has been confirmed by comparative wear tests in which two
engines were run for equal times under similar conditions, other than that
one was run continuously, and the other was run intermittently and allowed
to cool after each run. The cylinder wear of the intermittently run engine was
many times higher than that of the engine run continuously. Analysis of the
oil from the engines also confirmed the importance of the corrosive mechan-
ism during intermittent running.
During continuous running it has been found that hydrodynamic con-
ditions exist between the piston and cylinder in all regions except those close
to extreme piston positions, i.e., BDC and TDC where the rubbing speed is
zero and lubrication is in the boundary region. This is confirmed by wear
measurements on the cylinder wall from which it is found that wear is greater
at the extreme ends of the stroke, and that wear at TDC is greater than that
at BDC. This is due to the lower temperature and better availability of oil at
the bottom of the stroke.

5.8 CONCLUSIONS

The above example of the piston/piston-ring interface in an internal-

combustion engine illustrates the way in which the wear of a very common-
place mechanism can be due to the combination, in varying degrees, of many
of the diffurent types of wear described in this chapter. Material at the interface
is worn away by adhesion, abrasion, corrosion and fatigue. In addition, the
relative importance of these wear mechanisms varies for different positions on
the cylinder due to differences in environment, speed, load, temperature and
efficiency of lubrication. This single example serves to show that the process
of wear is indeed a very complex one, and perhaps explains why attempts to
predict wear rates in practical situations have, so far, met with only limited
success.

REFERENCES
I. I. V. Kragelskii. Friction and Wear. Butterworths, London, (1965).
2. J. T. Burwell. Survey of possible wear mechanisms. Wear, 1, (1957), 119-141.
3. M. Kerridge and J. K. Lancaster. The stages in a process of severe metallic wear.
Proc. R. Soc., 236A, (1956), 250-264.
4. J. F. Archard. Contact and rubbing of flat surfaces. J. appl. Phys., 24, (1953),
981-988.
5. R. Holm. Electric Contacts. Almgvist & Wicksells, Uppsala, (1946).
6. J. T. Burwell and C. D. Strang. Metallic wear. Proc. R. Soc., 212A, (1952), 470-77.
7. M. M. Kruschov. Resistance of metals to wear by abrasion; related to hardness.
Instn mech. Engrs Conf. Lubrication and Wear, London, (1957), 655-59.

126
8. C. N. Rowe. Some aspects of the heat of absorption in the function of a boundary
lubricant. Trans. Am. Soc. Lubric. Engrs, 9, (1966), 100-111.
9. A. G. Cattaneo and E. S. Starkman. A.S.M. Symposium on Mechanical Wear,
Cleveland, Chapter IV, 1950.
10. R. C. D. Richardson. The wear of metals by relatively soft abrasives. Wear, 11,
(1968), 245-75.
II. T. L. Oberle. Properties influencing wear of metals. J. Metals, 3, (1951 ), 438-9.
12. R. T. Spurr and T. P. Newcombe. The friction and wear of various materials sliding
against unlubricated surfaces of different types and degrees of roughness. Instn.
mech. Engrs. Conf. Lubrication and Wear, London, (1957), 269.
13. H. Hertz. Z. reine angew. Math. 92, (1881), 155.
or see
S. P. Timoshenko and J. N. Goodier. Theory of Elasticity. McGraw-Hill, New
York, (1970), p. 409.
14. J. K. Lancaster. The relationship between the wear of carbon brush materials and
their elastic moduli. Br. J. appl. Phys. 14, (1963), 497-505.
15. J. F. Archard and W. Hirst. The wear of metals under unlubricated conditions.
Proc. R. Soc., 236A, (1956), 397-410.
16. R. G. Bayer, W. C. Clinton, C. W. Nelson and R. A. Schumacher. Engineering
model for wear. Wear, 5, (1962), 378-91.
17. J. F. Tavernelli and L. F. Coffin. A compilation and interpretation of cyclic-strain
fatigue tests. Trans. Am. Soc. Metals, 51, (1959), 438-53.
18. J. G. A. Bitter. A study of erosion phenomena. Wear, 6, (1963), 51~9.
19. E. Rabinowicz. Friction and Wear of Materials. Wiley, New York, (1965).
20. G. Levy, R. G. Linford and L. A. Mitchell. Wear behaviour and mechanical
properties; the similarity of seemingly unrelated approaches. C.E.G.B. Report No.
RD/B/N2250, (1972).
21. R. L. Aghan and L. E. Samuels. Mechanisms of abrasive polishing. Wear, 16,
(1970), 293-301.
22. R. G. Bayer. Prediction of wear in a sliding system. Wear, 11, (1968), 319-31.
23. F. P. Bowden and G. W. Rowe. The adhesion of clean metals. Proc. R. Soc.,
233A, (1956), 429-42.
24. E. E. Bisson, Friction and bearing problems in the vacuum and radiation environ-
ments of space. Advanced Bearing Technology, NASA, SP-38, (1964), 259-287.
25. R. W. Wilson. Influence of oxide films on metallic friction. Proc. R. Soc., 212A,
(1952), 450-52.
26. F. T. Barwell. Wear of metals. Wear, 1, (1957-8), 317-32.
27. A. J. W. Moore and D. Tabor. C.S.I.R. (Australia), Tribophysics Report A46,
(1942).
28. M. J. Hordon. Adhesion of metals in high vacuum. Adhesion or Cold Welding of
Materials in Space Environment. A.S.T.M., STP 431, (1967), 109-127.
29. N. C. Welsh. Frictional heating and its influence of wear of steel. J. appl. Phys.,
28, (1957), 960~8.
30. M. C. Shaw. Lubrication, friction and wear under gross plastic deformation.
Friction and Wear Interdisciplinary Workshop, NASA Lewis Research Center,
Cleveland, Ohio, TM-X-52748, (Nov. 1968).
31. E. Rabinowicz. Compatibility criteria for sliding metals. Friction and Lubrication
in Deformation Processing. Am. Soc. mech. Engrs, (1966), N.Y., 90-102.
32. R. G. Bayer, W. C. Clinton, T. C. Ku, C. W. Nelson, R. A. Schumacher, J. L. Sirico
and A. R. Wayson. Applying the wear model to design problems. A.S.M.E.
Design Eng. Conf., Paper No. 66 MD-12, (May, 1966).

127
6
Tribological Properties
of Solid Materials
6.1 INTRODUCTION

In previous chapters on friction and wear theories we have described the

basic interactions between moving surfaces, and the effects of these inter-
actions on friction coefficients and wear rates. In the present chapter the aim
is to explain the effects of material properties and environment on the
observed behaviour and, in particular, to describe the behaviour of materials
which have good tribological properties.
Undoubtedly the most effective way to reduce friction and wear is to
completely separate the solid surfaces by means of a film of a viscous fluid, as
described in the chapters on hydrodynamic and hydrostatic lubrication. As
long as such a fluid film can be maintained the coefficient of friction will be
very low, of the order of 0.003 or less, and wear will be entirely eliminated.
Under these circumstances the properties of the solid surfaces are only of
secondary importance. However the properties of the surfaces do become
very important under the two sets of circumstances described below.

( 1) If a film of fluid lubricant cannot effectively separate the solid surfaces,

as during the stopping and starting of hydrodynamically lubricated
equipment, then conditions of boundary lubrication obtain and, as this
implies, metal-to-metal contact can occur at contacting asperities. We
therefore need to use materials which will give acceptable tribological

128
behaviour during these periods of metal-to-metal contact. In the first
part of this chapter we therefore describe the tribological properties of
clean metal surfaces, and of surfaces covered by naturally occurring
films such as oxides, and discuss the criteria which give such surfaces
acceptable tribological properties.
(2) If for some reason-which may be technical or economic-fluid
lubrication is not considered to be suitable for a particular application,
we must use solid surface materials which have inherently good tribo-
logical properties. In the second part of this chapter we shall discuss the
properties and uses of such materials.
Throughout the chapter we shall consider the effects of environmental
variables, such as temperature and atmosphere. We shall indicate the limits
of usefulness of the various materials in terms of these variables, and the
operational parameters such as load and speed.

6.2 TRIBOLOGICAL PROPERTIES OF METALS

6.2.1 Clean Metal Surfaces

We have already seen, in chapters 4 and 5, that the friction coefficient and
wear rate on rubbing together clean metal surfaces can become very high, and
in many cases gross seizure of the opposing surfaces can occur. This has been
shown by the modified Bowden and Tabor theory of friction to be the result
of the plastic deformation of the contacting asperities under the combined
normal and tangential stresses.
Under most normal conditions steps are taken to minimise the amount of
metal-to-metal contact but, as we have seen, this can rarely be eliminated and
care must therefore be taken in material selection to reduce the ensuing
friction and wear. Some of the material properties which must be considered
are described below.

Hardness

Many metals show a transition from mild to severe wear when the nominal
contact pressure, that is, the load divided by the apparent area of contact,
becomes greater than some fraction of the hardness. For many metals this
fraction has a value of about one-third. This transition is generally attributed
to the interaction of the plastic zones beneath contacting asperities, as shown
in figure 6.1, so that gross plastic deformation can take place. To avoid this
effect it is clearly desirable to choose materials which have a hardness several
times greater than the apparent contact pressure.

129
(a) Lighf loads

(b) Heavy loads

Figure 6.1 Plastic zones be-

low contacting asperities

Mutual Solubility
There is some evidence to suggest that metals exhibiting a high degree of
mutual solubility will have poor tribological properties. One analysis by
Rabinowicz 1 of published friction, adhesion and wear results for different
metal couples has shown zero correlation of solubility with adhesion, a
positive correlation with friction, and a greater correlation with wear.
This suggests that metal pairs for boundary lubricated or unlubricated
sliding should be chosen to have low mutual solubilities. Certainly it is
generally accepted that similar metals, that is, metals with 100 per cent
mutual solubility, should be avoided.

Crystal Structure
The Bowden and Tabor theory indicates that perfectly clean surfaces which
are able to deform plastically will exhibit seizure before gross sliding can
occur. There is however one important class of metals which does not behave
in this way and which will give reasonable friction coefficients, (JL ~ 0.2-0.4)
and low wear rates even when absolutely clean. These are the metals which
have hexagonal close packed crystal structures and which plastically deform
by slip on a single slip plane, the basal plane. It is generally thought that the

130
good tribological properties of these materials are explained by this limited
slip behaviour, as the plastic deformation of the type invoked in the Bowden
and Tabor theory requires slip on several different slip planes. Without this
it is not possible to get the continuous junction growth which is the basis of the
theory.
Although this type of behaviour was of no more than theoretical interest
a few years ago, recent developments have created a demand for strong
materials which can be relied upon to run together under completely clean
conditions. Materials of this type have recently been used as hinges on the
doors of spacecraft, and they are also likely to find increasing use in inacces-
sible areas such as the interior of nuclear reactors.
Before leaving the subject of clean metals it is again worth mentioning the
use of soft metal films on hard substrates which has already been described
in chapter 5. These films are also finding increasing use in vacuum chambers
and space environments. An interesting metal in this context is lead, which as
a film on a hard substrate, is an excellent lubricant under vacuum conditions,
although poor in normal atmospheres. This difference is thought to be due to
the continuous conversion of lead to lead oxide under atmospheric con-
ditions.

6.2.2 Surfaces in Normal Atmospheres

Under normal atmospheric conditions most metals are covered by oxide

films and clean metal surfaces will acquire such a film 5 to 50 molecules thick
within a few seconds. Thus if the oxide is removed from an asperity by
rubbing it will almost certainly have been reformed before the asperity makes
another contact with the opposing surface.
The presence of an oxide film can be very beneficial, but unless the loads
are small the oxide cannot completely prevent intermetallic contact. As a
result of this, it has been found that the wear behaviour of many metals under
normal atmospheric conditions, can be divided into two regimes. In the mild
wear regime, which occurs at lower loads, the electrical contact resistance is
high, the wear debris is fine and consists mainly of oxide, and the metal
surfaces have a burnished appearance. In the severe wear regime, which
occurs at higher loads, the contact resistance is low, the wear debris includes
coarse metallic particles, and the surfaces are rough. The transition between
these two regimes is clearly at the load at which large scale penetration of the
oxide films, and consequently metallic adhesion, take place.
The properties of the particular oxide films are important. For example
the oxide formed on aluminium is hard and brittle and offers little protection
whereas, under suitable conditions, the films formed on steels are tough and
tenacious and offer excellent protection. This latter fact has been utilised in
nuclear reactors, where plain stainless steel bearings are used at elevated
temperatures under which conditions they form suitable oxide films.

131
6.2.3 Effects of Temperature
It is clear that the temperature of the rubbing surfaces can have a major
influence on all the properties described above. We can consider them in
turn.

Hardness
The hardness of metals generally decreases with increasing temperature and it
can be seen that, in the absence of other effects, wear rates will increase with
increasing temperature. To counteract this effect it is necessary to use materi-
als of high hot-hardness, such as tool steel and alloys based on cobalt,
chromium, and molybdenum, for high temperature bearings.

Mutual Solubility
The mutual solubility of metal pairs is a function of temperature, and although
this effect has not been investigated, it is therefore quite possible that increases
in solubility with increasing temperature could adversely effect tribological
properties.

Crystal Structure
Temperature-induced phase changes can have a profound effect on tribo-
logical properties. An excellent example of this is cobalt which at 417 "C
changes from its low temperature hexagonal close packed structure to a
face centred cubic structure. Wear experiments by Buckley 2 on cobalt, the
results of which are illustrated in figure 6.2, show that the wear rate at 350 oc
is one hundred times greater than that at 280 'C. This is undoubtedly due
to the combined ambient and frictional heating transforming the interface
material into the face centred cubic regime. This enables junction growth to
take place, due to the increased number of operative slip systems, as men-
tioned in the foregoing section headed 'crystal structure'.
Changes of temperature have a marked effect both on rates of oxidation
and on the types of oxide formed. This can clearly have a marked influence on
tribological properties. This has been strikingly demonstrated by Kragelskii 3
by rubbing pure Armco iron against itself at different speeds (and hence
different interface temperatures). At low speeds the wear rate was high, with
evidence of much adhesion, but at higher speeds the wear rate fell by almost
three orders of magnitude and the surfaces became smooth and polished.
The temperature of the interface at the transition speed was calculated to
be about 1000 'C. To prove that this was a temperature rather than a speed
effect the experiment was repeated at very low speed while high current
pulses were passed through the contacts to raise them to the same tempera-
ture. The wear rate was again very low and the surfaces were polished.

132
Wear rote
35xi0- 7 mm 3/m

04

0 100 200 300 400

Ambient temperature oc

Figure 6.2 Friction coefficient and wear rate for

cobalt on cobalt in vacuum at various temperatures-
pressure 10- 7 N/m 2 , sliding speed2 m/s, load9·81 N

6.3 SELF-LUBRICATING MATERIALS

In this section we shall describe materials which have intrinsically good tribo-
logical properties, but before describing specific examples it is worth consider-
ing why we use self-lubricating materials.

6.3.1 Advantages of Self-lubricating Materials

(a) Solid lubricants operate over a greater temperature range than fluids.
Oils become thin and decompose or oxidise at high temperatures; at low
temperatures they become very viscous, and eventually may solidify.
(b) Solid lubricants give better surface separation than liquid boundary
lubricants at high loads and low speeds.
(c) Many solid lubricants are much more chemically stable than liquids
and they can be used in environments such as strong acids, solvents,
and liquid gases.
(d) Solid lubricants are usually very clean and can be used in environ-
ments where cleanliness is essential, such as in food-processing equip-
ment.
(e) Solids can often be used to provide permanent lubrication for parts of
equipment which are inaccessible after assembly.
(I) By using solid lubricants, designs can be simplified by eliminating
complicated passageways and oil-circulating equipment.
(g) Solid lubricants are quite stable in very radioactive environments,
whereas oils and greases are degraded.

133
(h) Solid lubricants may be much more convenient than oils or greases.
An outstanding example of this is the use of self-lubricating bushes on
modern cars, with the consequent reduction in the time spent on periodic
maintenance.

6.4 TYPES OF SOLID LUBRICANT

Besides the hexagonal close packed metals and the soft metal films which we
have already described, the available solid lubricants can be divided into three
groups, lamellar solids, other inorganic solids, and plastics. We shall describe
each of these three groups in turn.

6.4.1 Lamellar Solids

A lamellar solid is one in which the atoms are bonded together in parallel
and comparatively widely spaced sheets. The two best known and most
widely used examples are graphite and molybdenum disulphide, which have
the crystal structures shown in figures 6.3 and 6.4. Under many circumstances
these are both excellent lubricants, as are other lamellar solids such as tung-
sten disulphide, cadmium chloride and cadmium iodide. However not all
lamellar solids act as lubricants, and there is as yet no theory which will tell
us whether or not a particular lamellar solid will be suitable.
The lamellar solids which can lubricate effectively do have certain charac-
teristics in common. The most important characteristic seems to be the
ability to form a strongly adherent transferred film on to the surface being

Figure 6.3 The crystal structure of

graphite

134
e Molybdenum
0 Sulphur

Figure 6.4 The crys-

tal structure of moly-
bdenum disulphide

lubricated. After the initial running in period, during which this film IS
formed, the interface therefore consists of lubricant on lubricant.
A second characteristic is that the material at both surfaces develops a
preferred orientation which is shown schematically in figure 6.5. This
orientation reduces the mechanical interaction between the surfaces, as can

Figure 6.5 Schematic illustration of the

surface orientation developed during the
rubbing of lamellar solid lubricants

135
be shown by reversing the direction of motion when the friction coefficient
increases significantly.
Although a great amount of experimental work has been performed on
graphite and molybdenum disulphide there is as yet no universally accepted
explanation of their behaviour. In the following paragraphs we shall briefly
describe some of the suggested explanations.

Graphite
The earliest suggested explanation for the properties of graphite was that due
to Bragg4 , who suggested that the shear strength parallel to the atomic
sheets was very low. This would allow the sheets of atoms to slide over each
other rather like a pack of playing cards, and hence enable the graphite to
act as a boundary lubricant. This explanation was accepted until the Second
World War, when it was found that graphite brushes in electrical generators
of high-flying aircraft wore out very rapidly. A systematic investigation of this
effect by Savage 5 showed that graphite had very poor tribological properties
in the absence of condensable vapours. This effect is very marked and the
amount of vapour needed for effective lubrication can be very low: for exam-
ple exposing dry graphite to a pressure of 400 N/m 2 of water vapour de-
creases the wear rate by three orders of magnitude, and the friction co-
efficient by a factor of five. The effectiveness of different vapours varies
enormously, and while nitrogen has no lubricating effect at atmospheric
pressure many organic vapours are effective at very low pressures, as shown
in figure 6.6.

008 J

E I
'E
E
.,
\
\
0
c
.,c "0 0
c
.,c
& 0 g 0
.c &
.,0 0 c., :cu Q; 0

:;:: 004 -
0.
'§- ·---
Cl.
c
~
!!
c
-:;.
\
~ ~

.8
~
0
u

0
0133
\
I 33 13 3
~
133 1330
Pressure Ntm2

Figure 6.6 Effects of condensable vapours on the wear of

graphite 9

136
Alternative explanations of this effect have been advanced. The first
explanation suggests that the vapours penetrate between the layers and
reduce the shear strength to a very low level. The second, and more widely
supported, suggestion is that the vapours saturate the surface forces, and
particularly the edge forces, on individual crystallites. The adhesion between
neighbouring crystallites is therefore very low so that they can easily slide
over each other. The surface layer of graphite, rather than the individual
crystals, can thus be sheared by low forces.
This second explanation has been supported by the work of Midgely and
Teer 6 and the work of Arnell, Midgely and Teer 7 , who have shown that
even nearly amorphous mechanical carbons and brittle pyrolytic carbons
exhibit the same low friction behaviour as graphite. Such behaviour is not
compatible with a low shear strength theory.

Molybdenum Disulphide
Unlike graphite, molybdenum disulphide, MoS 2 , appears to be an intrin-
sically good lubricant and in fact its friction coefficient is lower under high
vacuum conditions than in the presence of water vapour. Again there are
alternative explanations of this inherent low friction. The first explanation
suggests that molybdenum disulphide is a true low shear strength solid,
with shear taking place between the adjacent layers of the sulphur atoms
shown in figure 6.4. The second explanation, which is similar to that advanced
for graphite, is that the edge forces on the crystals are saturated by oxygen
atoms. Unlike the vapours on graphite, however, the oxygen is not volatile,
but persists up to the decomposition temperature of the solid.
Lamellar solids can be used in several different forms, the most common
ones being described below:
( 1) As dry powders or dispersions in fluids. The oldest method of using
these lubricants, and one which is still used extensively with MoS 2 is
to rub the surfaces to be lubricated with dry powders. Burnished films
of MoS 2 tend to adhere better than graphite, and the former is therefore
often applied as a dispersion in a solvent which subsequently evaporates
to leave a film of dry powder. Dispersions and dry powders are used
extensively to facilitate the assembly of close fitting parts, to lubricate
metal working components such as wire drawing dies, and as parting
agents for screw threads.
(2) Solid Blocks. Graphite and graphite carbons are often bonded to-
gether into solid blocks which can, for example, be made into solid
thrust bearings. Such carbon bearings were used for many years as the
clutch release bearings of cars. Solid carbon bearings are generally
made from a mixture of finely divided coke and a carbon binder such as
pitch, which is then fired to a very high temperature. This heat treatment
graphitises the entire mixture, and by varying the heat-treatment tem-
perature and time, a large range of materials can be made. These range

137
from highly crystalline electrographite, which is generally used for low
load applications such as electrical generator brushes, to almost
amorphous mechanical carbons, which have high strength and are
commonly used for thrust bearings.
(3) Bonded Films. Solid lubricants are frequently bonded to metal surfaces
by using organic resin binders, and such films have wear lives two orders
of magnitude higher than those of dry powder films. Because of their
excellent tribological properties bonded MoS 2 films are used extensively
throughout industry. The main drawback of these films, compared with
a dry powder film, is that the decomposition of the resin imposes a limit
to the maximum running temperature, and therefore to the load and
speed.
(4) Metal Composites. The thermal limit mentioned in the previous
paragraph can be overcome by incorporating the lubricant in a metal
matrix. The relative proportions of lubricant and metal in these films
are quite critical as can be seen from figure 6.7. Too much lubricant
weakens the composite, whereas too little provides an insufficient supply
of lubricant.
These composites provide an excellent illustration of the fundamental
principle which lies behind the design of bearings, that is, to provide
high load-carrying capacity with a low shear strength. If the compo-
sites contain insufficient lubricant they have excellent load-carrying
capacity, but the shear strength is close to that of the pure metal. Such

20

"'E
E 15

"~
E
0
>
h
~ 10
0
~

05

10 15
We1ghf percent tv'()S 2

Figure 6.7 Wear of powder metallurgy specimen

containing molybdenum disulphide 10

138
composites therefore have friction coefficients close to those of the pure
metals. Furthermore the lubricant cannot provide adequate surface
coverage and the wear rate is also high. On the other hand, if we have
excess lubricant, the surface shear strength is very low, but the compo-
site structure is weakened, and the load-carrying capacity is also low.
(5) Grease Additions. The major use of MoS 2 is as an additive to greases
and oils. Such filled greases are very effective during running-in opera-
tions and increase the effectiveness of lubrication under heavy loads.
They are also used to prevent seizure of threaded components.

6.4.2 Inorgan!c Solids for High Temperature Lubrication

Apart from the lamellar solids already described many inorganic solids have
been tried as solid lubricants with varying degrees of success. It is not the
purpose of this chapter to present an exhaustive list of the successful lubri-
cants, but two of the inorganic solids which are very promising as high tem-
perature lubricants are described below.
(a) Lead monoxide. This is a poor lubricant at temperatures below ap-
proximately 250 °C, but above this temperature it is a better lubricant
than MoS 2 and it retains excellent properties at temperatures up to
approximately 650 oc. Lead monoxide can be used in ceramic bonded
films which have excellent wear properties, and such a film at 650 oc
has better properties than a resin bonded MoS 2 film at room tem-
perature. This material has an upper temperature limit of approxi-
mately 700 oc, as the coating softens at higher temperatures.
(b) Calcium Fluoride. For lubrication at temperatures above 700 oc
one of the most effective lubricants so far discovered is a ceramic
bonded film of calcium fluoride. This retains excellent properties at
temperatures above 1000 oc, and it has better wear properties at high
temperatures than either molybdenum disulphide or lead oxide at their
optimum temperatures.

6.4.3 Plastics

Plastics are used successfully in many tribological applications, the most

widely used being nylon, fluorocarbons such as polytetrafluoroethylene
(PTFE), and phenolic laminates. In addition to the advantages already
cited for self lubricating materials, plastic bearings have the following ad-
vantages
(1) They absorb vibrations well and are quiet in operation.
(2) They readily deform to conform to mating parts: machining tolerances
and accuracy of alignment are therefore less critical than for metal
parts.

139
(3) They are easily formed into complicated shapes, either by machining or
moulding.
(4) They are very cheap.
Plastics are used in many forms, such as solid plastic, resin bonded films,
and as composites impregnated with other substances to give improved
tribological, physical and mechanical properties. There is an enormous
range of materials available and the aim here is simply to describe the pro-
perties of the base materials and to indicate the relative advantages of the
various forms.

6.5 TRIBOLOGICAL PROPERTIES OF PLASTICS

6.5.1 Friction of Plastics

It has been shown by King and Tabor 8 that, for most plastics, the friction
force between a plastic surface and a steel slider is as predicted by the simple
Bowden and Tabor theory of friction, that is, the friction force is approxi-
mately equal to the bulk shear stress of the plastic multiplied by the effective
area of contact. It is not, however, correct to state specific values of friction
coefficients for plastics. Such materials are usually viscoelastic and their
deformation is therefore dependent on strain rate. As this implies, the friction
coefficients of plastics can vary very significantly with such parameters as
sliding speed and surface roughness. However, the coefficients of friction of
most plastics, on metals and on themselves, are generally in the range
0.2-0.4. PTFE is an outstanding exception to this generalisation in that the
coefficient of friction for PTFE sliding on itself can be as low as 0.05 which is
the lowest known value for any solid. There is as yet no firm explanation of
this low friction coefficient, although it is generally ascribed to the intrinsic-
ally low adhesion between PTFE molecules. PTFE has long rigid molecules in
which carbon atoms are effectively shielded by the surrounding fluorine
atoms. The actual adhesion forces between molecules are therefore thought
to be very low, so that the surface molecules can roll or slide over each other.
On the other hand the mechanical interlocking of the molecules in the
bulk material gives it a relatively higher hardness and bulk shear strength.

6.5.2 Wear of Plastics

With the exception ofPTFE, the friction coefficients of plastics are not particu-
larly low, but the main advantage of plastics is that they wear at low and
reasonably predictable rates. This enables a designer to select, with some
accuracy, a bearing material which will give the desired lifetime under speci-
fied conditions of load and speed. The rate of wear of a plastic bearing is
obviously a function of load and speed, and it is found that we can define a
design criterion, the P- V factor, for plastic bearings as described below.

140
6.5.3 The P- V Factor
The derivation of the P- V factor is based on the reasonable assumption that
the rate of wear will be proportional to the rate of energy dissipation at the
sliding interface. Starting from this assumption we shall derive the relation-
ship between wear rate and the P- V factor for the two basic bearing con-
figurations.
(i) Flat Bearings. If we have a flat bearing surface of area A, as shown in
figure 6.8, subjected to a normal load W, then the amount of energy

Lood W

-
Shdmg veloctry v

,
Beom>g ,,
area A ,'
,,
,,

Figure 6.8 Schematic diagram offiat bearing

for derivation of P-V factor

dissipated on sliding a distance dx is given by JL W dx, where JL is the

coefficient of sliding friction. The rate of energy dissipation is therefore
given by
dx
J1 x W x dt = JLWV

where V is the sliding speed at the interface. If we then assume that the
volume wear rate Qis proportional to the rate of energy dissipation we
have, for constant JL
Q oc wv
The factors we are normally interested in are the depth of wear, or the
rate of linear wear normal to the sliding surface. We can see that the
depth of wear at any time is simply equal to the volume of wear divided
by the total contact area. Therefore

rate of linear wear = ~

141
and
rate of linear wear oc : V

or
rate of linear wear oc PV
(ii) Sleeve Bearings. Figure 6.9a shows a circular bearing subjected to a
normal load W, which is supported over the shaded bottom half of the
bearing. We assume that we have a normal reaction R per unit area at
this interface, and to find the relationship between W and R we consider
the equilibrium of vertical forces. We see on figure 6.9b that the reactive

Shaff

Load supported
an shaded half
(a) at beanng

(b)

Figure 6.9 Schematic diagram of solid lubricated

journal bearing and forces acting

force over an element of surface of length ds will be given by R dsl,

where I is the axial length of the bearing, and the vertical component of
this force will be
R dsl cos fJ
but
D
ds = 2 dfJ
142
Therefore the total vertical reaction is given by

lRID f
+n/2

cos() d() =RID

-n/2

This is equal to the applied vertical load W so that

w
R=-
Dl
But the rate of energy dissipation per unit area at the interface = JlR V.
Therefore the total rate of energy dissipation at the interface
rtD
= JlRV X 2/
W rtD
= Jl Dl v2 1
1t
=JlWV-
2
Therefore
volume wear rate oc JlWVrt

We are normally interested in the radial wear rate and we can see from
figure 6.10b that this is given by
volume wear rate
rtDl

Therefore
. JlWVrt
radtal wear rate oc 1t(i[

or
radial wear rate oc JlP V
where P is the load per unit projected area.
We can see from the above derivations that the same P-V factor applies
to each type of bearing, provided that P is taken as the load per unit pro-
jected area.
To a good approximation it is possible to specify a certain P-V factor
which must not be exceeded for a certain life expectancy. The use of a single
factor is open to some criticism since it assumes the same sensitivity of wear
rate to changes of P and V. This is frequently not the case and a more exact
practice is to give plots of acceptable wear rates on pressure velocity diagrams

143
"'E
'z

10- 1
velocity m/sec

Figure 6.10 Typical limiting P-V curve of PTFE

based material for wear rates of 2511 m in 100 hours

as shown in figure 6.10. It can be seen from these diagrams that the P- V
factor for a particular wear rate is reasonably constant except at the extremes
of load and speed.
The P- V limit of a bearing is that P- V factor at which the bearing would
fail rapidly due to melting or thermal decomposition.
It is also found that the steady state wear rate, Q, for a plastic bearing is
proportional to the P-V factor over much of the usable P-V range. We can
therefore define a wear coefficient K by the equation
Q=KPV
It can be seen from this equation that, if the wear rate is known for one
P- V value, K can be calculated and the wear rates for other P- V factors
determined.
It should be pointed out that the use of the P- V factor is not confined to
plastic bearings; similar factors can be given for other types of bearing and,
in particular, for resin bonded solid lubricaht films.

144
6.5.4 Factors Influencing the Wear of Plastic Bearings

(1) Lubrication. The performance of plastic bearings can be improved by

lubrication. Periodic lubrication can raise the P- V limit by several
hundred per cent, and under continuous lubrication conditions the
operation is limited only by the mechanical strength of the plastic.
(2) Temperature and Heat Dissipation. It has already been stated that the
P- V limit of a bearing is reached when the interface starts to melt or
thermally decompose. The P- V limit is therefore clearly affected by the
ambient temperature. It is also affected by the temperature rise at the
sliding interface and hence by the rate of heat dissipation from the inter-
face. In general the thermal conductivities of plastics are low and some
technique has to be adopted to assist the dissipation of heat. Such
tecl)niques include the addition of thermally conducting fillers, such as
metal powders, and the use of the plastic as thin liners inside metal
sleeves. In addition it is possible to use bearings intermittently at
increased P-V levels, allowing the bearing to cool between operating
periods.

6.5.5 Filled and Reinforced Plastics

All types of plastic bearings are used either by themselves or with a large
variety of fillers or reinforcements. The additives are used for several dif-
ferent purposes, as listed below.
(1) Improvement of Mechanical Properties. One disadvantage of plastic
bearings, in comparison with metals, is their lack of rigidity and strength.
These properties can be appreciably improved by various additions;
by far the most common being chopped glass-fibre. The addition of
glass fibre increases tensile strength, rigidity, and creep resistance, and
hence enables reinforced bearings to operate to higher P- V values.
In the case of phenolic laminate bearings the properties of the bearing
are of course affected by those of the laminating material. Typical
laminating materials are paper, linen, canvas, and woven glass-fibre.
In general the coarser materials such as canvas give the highest strength
and toughness, whereas the finer materials have better machining
properties and can be used for precision components.
(2) Improvement of Thermal Properties. We have already discussed the
use of metal fillers to improve thermal conductivity and heat dissipa-
tion. An additional disadvantage of plastic materials is that they
generally have high thermal expansion coefficients compared with
metals. To allow for this discrepancy, plastic bearings have often to be
made to much looser fits than other bearings. This discrepancy is again
reduced by filling or reinforcing the plastic, thus enabling the bearings
to be made to finer tolerances.

145
(3) Improved Friction and Wear. The most common additives to improve
the friction and wear properties of plastic bearings are graphite and
molybdenum disulphide. These can improve P- V values by an order of
magnitude, and wear rates by two or three orders of magnitude. They
also have the additional advantage of improving, to some extent, the
thermal and mechanical properties of the materials.

REFERENCES

I. E. Rabinowicz. Proc. Am. Soc. mech. Engrs/Am. Soc. Lubric. Engrs Lubrication
Conference,(October 1970).
2. D. H. Buckley. Cobalt, 38, (1968), 20-28.
3. I. V. Kragelskii. Friction and Wear. Butterworths, London, (1965), p. 91-2.
4. W. L. Bragg. Introduction to Crystal Analysis. Bell, London, ( 1924).
5. R. H. Savage. J. appl. Phys. 19, (1948), I.
6. H. W. Midgely and D. G. Teer. Nature, 189, (1961), 735.
7. R. D. Arnell, J. W. Midgely and D. G. Teer. Proc. lnstn. mech. Engrs, 179, 3j,
(1966), 115.
8. R. F. King and D. Tabor. Proc. Phys. Soc., 65B, (1953), 728.
9. R. H. Savage and D. L. Schaefer. J. appl. Phys., 27, (1956), 136.
10. R. L. Johnson, M.A. Swikert and E. E. Bisson. NACA TN.2027, (1950).

FURTHER READING

E. R. Braithwaite. Solid Lubricants and Surfaces. Pergamon Press, Oxford, (1964).

E. E. Bisson and W. J. Anderson. Advanced Bearing Technology. NASA SP-38, (1964).
F. J. Clauss. Solid Lubricants and Self Lubricating Solids. Academic Press, London,
(1972).

146
7
Friction Instability
7.1 INTRODUCTION
Unwanted vibrations which may arise during the operation of machines
are costly in terms of reduction of performance and service life, sometimes
endangering equipment and personnel. This chapter is concerned with those
vibrations occurring through the agency of friction forces at the sliding
parts. It introduces methods of study, examines the characteristics, and
considers the prevention or alleviation of the vibrations.
The reader will have encountered such vibrations manifested, for instance,
by the squealing of brakes, the creaking of hinges and the ringing of wine
glasses. He may also have witnessed the jerkiness which may occur when a
large mass is driven at slow speeds along slideways. All these involve friction
in a dominant role. They are most prevalent at low sliding speeds and often
they produce not only noise, but also severe wear and dimensional in-
accuracy. The indirect effects may be serious, as in textile machinery where
oscillations can lead to objectionable patterning in the cloth 1 .
The engineer concerned generally with vibrations requires a comprehen-
sive knowledge of theory; for our purpose, the study of frictional vibrations
at an elementary level, a short review of the basic ideas is a useful preliminary.
A system is said to be in a state of vibration when one or more of the
variables describing its state changes in a periodic manner. The most simp1e
example is the linear oscillator, figure 7.1, which may be represented, when
free from external influences, by an ordinary differential equation of the form
d2 x
dt2 + w;x = 0 (7.1)

147
X

Figure 7.1 Linear oscillator

where xis a system variable and w. is a parameter of the system. The solution
of equation 7.1, representing the unforced motion, is
(7.2)
where A and B are constants determined by the initial state of the system.
Equation 7.2 shows that the motion is sinusoidal, of amplitude A and
frequency w. radjs.
In those situations where account has to be taken of energy-dissipative
forces, a more realistic linear model is of the type
d 2x dx 2
dt 2 + 2(w. dt + w.x = 0 (7.3)

where ( is a constant parameter of the system. The solution of equation 7.3

appears in three forms. For I(I < 1
(7.4)
for I'I > 1

and for I'I = 1

(7.6)
where A, A 1 , and A 2 are constants determined by the initial state of the
system.
The parameters w. and ( are often called the undamped natural frequency
and damping ratio respectively. If the damping ratio is negative then the
system is unstable, displaying divergent motion whenever disturbed.
More complex linear systems, such as those having several degrees of
freedom, are represented by differential equations of higher order whose
solutions are combinations of the expressions 7.4, 7.5 and 7.6.
In the case of forced motion, the equation 7.3 is replaced by an equation
of the type

(7.7)

148
where f{t) represents an external disturbance independent of the motion.
The solution of equation 7.7 is the sum of the complementary function and
the particular integral. The complementary function is the solution of the
homogeneous equation 7.3, takes the form 7.4, 7.5 or 7.6 and, for ( > 0,
represents a decaying, or transient motion. The particular integral is a
solution of equation 7. 7 which does not depend on the initial conditions and
represents the steady-state motion. When f(t) is a periodic function the
particular integral will be periodic also. Often f(t) is sinusoidal and applying
standard mathematical procedures the particular integral is found to be
sinusoidal also, of the same frequency but differing in amplitude and phase
from the forcing function. The basic features of the transient and forced
oscillations are shown in figures 7.2a and b.
The mathematical models that have been discussed are representative of
many types of linear systems unJer varying conditions and it is worthwhile
reviewing them now from a more physical viewpoint. We shall consider a

o<{<l {;;J:l
Marginally stable, Stable. underdamped. Stable, overdamped.
constant amplitude dec0y1ng OSCIIIOtian non-oscillatory
oscillation
X

?,;. -1
Unstable, osc1llatory Unstable, non- OSCillatory
(a)

Tronstent penod Steady state

(b)

Figure 7.2 System responses: (a) transient motions,

(b) forced oscillations

149
specific example, the mass-spring-dashpot system subjected to an external
force F(t) and a displacement Y(t) as indicated in figure 7.3.
If F(t) and Y(t) are identically zero, then the element is a passive one with
no energy source to sustain oscillations. If F(t) or Y(t) are made to vary in a
prescribed way by an external source, then the response to these disturbances
consists of transient and forced motions. The former is of the type (7.4), (7.5)
or (7.6), and the latter is represented by the particular integral solution of an
equation similar to equation 7.7. If F(t) or Y(t) is periodic, then the system
will settle into a state of forced oscillation, energy being drawn from the
source from which originates F(t) or Y(t). A common example is foundation
vibration caused by unbalance of an engine.

Figure 7.3 Damped linear system subjected to

disturbances

If Y(t) again varies with time and F(t) is related to the motion of the mass,
then it is possible that oscillations occur even though the displacement Y(t)
is non-periodic. The system is said then to be in a state of self-excited oscilla-
tion. The energy necessary to sustain the oscillations is drawn from the
source providing the displacement Y(t), and energy is dissipated by the
dash pot and possibly, as in the case of friction, by the force F(t).
In both forced and self-excited oscillations, an energy source is available;
in the latter case the frequency is greatly influenced by the parameters of
the system itself. One example of self-excited vibrations is aircraft wing
flutter where the aerodynamic forces are affected by wing deflections.
Another, which we shall consider in detail later, involves the stick-slip of a
mass on slideways, the drive motor supplying the energy and slideway
friction providing forces which depend on the motion of the mass.
To conclude this review of vibration theory it is pertinent to emphasise
that many systems are nonlinear and are represented by non-linear equations.
Because the principle of superposition used in linear theory no longer holds,
analytical treatment is difficult. Methods of investigation appropriate to the
non-linearity resulting from the frictional forces will be considered in
section 7.3. A non-linear oscillation may be so jerky that the variation of
velocity, say, with time, is more like a saw-tooth than a sinusoidal wave,
and then it may be called a relaxation oscillation.

150
7.2 CHARACTERISTICS OF FRICTIONAL VIBRATIONS
In this study, we shall concentrate attention on systems having a single degree
of freedom and exclude those having several degrees of freedom where
friction may act as a coupling agent.
To gain a basic understanding we shall consider the system represented
by the model shown in figure 7.4. Suppose that the support A is driven at
constant speed. Then if the bearing resistance F is zero the mass M will
attain eventually the same speed as A although there may be a period of
transient oscillation influenced by the dashpot. If F is not zero and varies
with position or velocity, for instance, then under some conditions oscilla-
tions persist. We shall make the simplifying assumption that the friction
varies with speed either as shown in figure 7.5a or as shown in figure 7.5b
and these will be referred to as the stiction and negative gradient cases
respectively.

A M

-v
k
,,,,,,,,,,,, - F

Figure 7.4 Linear drive system

Fnct1on force Fnction force

Fe t-----
0 Slip velocoty 0 Slip velocoty

(a) (b)

Figure 7.5 Friction-velocity characteristics: (a) stiction

case; (b) negative gradient case

7.2.1 Stiction Case

Let the initial state of the system be such that the spring is uncompressed
and M is stationary. When A moves there will be no movement of M until
the drive force, that is the spring force plus the dashpot force, reaches a
value equal to the stiction force F,. As soon as M starts to move, the friction
151
falls rapidly to the value Fe so that there is an unbalanced force F 5 - Fe
causing a sudden acceleration. The velocity of M increases until the drive
force has fallen to the value Fe. If the mass velocity is then greater than that
of the support A, the spring force continues to fall and the mass decelerates.
If the damping is low, then eventually M comes to rest and the cycle of events
is repeated. Acceleration, velocity and displacement curves for the mass
are shown in figure 7.6, and it is clear that the motion is of the stick-slip type.

Acceleration

veloCity

Displacement

Figure 7.6 Stick-slip oscillation

7.2.2 Negative Gradient Case

Suppose that the mass is moving steadily and that it then encounters a small
disturbance causing a sudden reduction in speed. This causes the drive force
to grow and when it exceeds the friction force the mass will accelerate,
possibly to a speed above the drive speed. This phase is followed by a decelera-
tion phase, which in turn is followed by acceleration and so on. If the drive
damping coefficient is low compared with the friction-velocity gradient,
then the amplitude of the velocity swings increases at least until either
reversal occurs or velocity ranges having positive friction gradients are
entered.
The analysis for the stiction case will be covered in a later section. The
analysis for the negative gradient case is straightforward providing that we
consider velocity variations small enough to allow linearisation of the
friction-velocity curve.

152
The equation of motion for the mass M is

Md X
2
+ f(dX- v) + k(X- vt) = -F (7.8)
dt 2 dt
where X is the displacement of the mass M
v is the drive velocity
f is the dashpot coefficient
k is the spring constant
F is the frictional resistance.
Providing that v is constant, equation 7.8 may be written in terms of the
spring extension, x, as
(7.9)

For small variations about the velocity v, the frictional force may be expressed
as
F = F v - A.(dX
dt
- v) (7.10)

From equations 7.9 and 7.10 it follows that

d -x 2
-dt + kx = - ( Fv
+ f dx dx)
- A.-
M dt
dt 2
that is
d 2x dx
M dt2 + (f - A.) dt + kx = - F, (7.11)

The complementary function solution of equation 7.11 is of the form of

equation 7.4, 7.5 or 7.6 with the damping ratio given by
f-A.
(7 .12)
( = 2(Mk) 112 •

If the drive damping is low and the friction-velocity gradient is high then
f- A. is negative, the damping ratio is negative and the motion following
a disturbance is divergent, being oscillatory with increasing amplitude if
-1<(<0

7.2.3 Oscillation Characteristics

We shall anticipate some of the results of later analyses to give a qualitative
appreciation of the problem.

Effect of Drive Speed

Many experimenters have observed that oscillations are usually worst at
low speed. If the drive velocity is high then the oscillation decays. If the
153
velocity is low, the oscillations become more severe and saw-toothed. The
lowest drive speed for which the oscillation decays is often called the critical
speed.

Oscillation Amplitude
In the sinusoidal oscillations the velocity-time curve is symmetrical about
the mean velocity. This is not true when sticking occurs, as in case 7.2.1,
and at speeds well below the critical speed, peak velocities many times the
average velocity are reached during the slip period. This influences the choice
of friction model parameters for use in analysis.

Oscillation Frequency
The oscillation frequency cannot be equated to the natural frequency of the
drive system itself. However, if the oscillations are not violent the drive
frequency is a useful guide, and may be used when selecting the friction model.

7.3 REVIEW OF ANALYTICAL METHODS

Linear systems are those which may be represented by linear differential

equations. A consequence of linearity is the superposition property, namely
that linear combinations of the homogeneous equation of the form of
equation 7.3, are also solutions. Also the responses r 1(t), r 2 (t), of the system to
disturbances / 1 (t), / 2 (t), may be combined in the form a 1 r 1 (t) + a2 r2 (t)
to generate the response to a disturbance aif1(t) + a~ 2 (t). Thus the form of
the response is unaffected by the magnitude of the disturbance. In particular
a linear system when disturbed from a steady operating state will display
either a transient response always or a divergent response always, irrespective
of its initial state and the magnitude of the disturbance. Moreover, in the
case of an unstable oscillatory linear system, the vibration amplitude increases
indefinitely with time. Non-linear systems are often difficult to analyse and
understand because of the absence of the superposition property and the
fact that the whole character of the response may be dependent upon the size
of the disturbance and the initial operating state. Also, an unstable non-linear
system may oscillate with a constant amplitude. Methods of analysis have
been developed which can be applied to any linear system, but it is necessary
to select the method of attack appropriate to a given non-linear problem.
We shall now discuss briefly the methods which are in common use.

7.3.1 Perturbation Methods

Useful information can often be obtained by studying the behaviour of the
system in the neighbourhood of the operating state. Providing that the
non-linearities are continuous, tangent approximations may be used to

154
yield linear equations. Such an analysis may be sufficient to confirm that a
system is stable or unstable. In the latter instance, the analysis may suggest
design modifications for stabilising the system. This approach is feasible
when studying the effect of negative gradients of the friction-velocity curve
as we did in section 7.2.2. Where the character of the non-linearity is given
in graphical form, a tangent may be drawn at the operating point to give a
straight-line approximation; where it is expressed analytically, a Taylor
series for the non-linear function, with only the linear terms retained, is
used to represent the non-linearity. This method is not suitable for discontin-
uous non-linearities such as the stiction characteristics of the example of
section 7.2.1. Instead the method outlined next may be tried.

7.3.2 Piecewise Linear Approximation

We can often represent the non-linear characteristic with sufficient accuracy
by a series of linear expressions, each one valid within a finite region of the
input variable. The idea of this approximation is illustrated graphically
in figure 7.7. The calculation of the transient response to a disturbance
proceeds in a step-by-step manner, each step being of finite size. The final
state of one step is used as the initial state for the solution of the linear
equations of the next step interval. Sometimes it may be possible to draw
conclusions about not only the transient behaviour but also the steady state.
This approach is used in section 7.5.

y•l (x)
Non -lrnear characterrst1c

Approxrmot1on

0 X

Figure 7.7 Piecewise linear approximation

7 .3.3 Phase-Plane Analysis

The phase-plane method is applicable to the study of the transient response
of unforced systems described by ordinary second-order differential equations
with constant coefficients. So far we have illustrated the system response by
plotting displacement versus time. For a positioning system governed by a
second-order differential equation, the velocity and displacement together
specify the state and we can show how the state varies by plotting curves,

155
known as trajectories, in the velocity-displacement plane. In terms of the
velocity and displacement variables, the second-order equation is trans-
formed to a first-order equation which may then be solved either directly
or graphically to yield the trajectories. The method is particularly useful
when the non-linearity is a function of either velocity or displacement. Here
it will be demonstrated by a simple linear example.
Suppose that the drive system is equipped with low friction slideways
so that equation 7.9 is replaced by
d2 x dx
M dt 2 + f dt + kx = 0 (7.13)

Writing y = dx/dt equation 7.13 becomes

dy
My dx + fy + kx = 0 (7.14)

Thus at any point (x, y) of the phase plane, the slope, dy/dx, of the trajectory
may be calculated and the trajectory constructed graphically. The equation
for the line joining points of equal trajectory slope s, called an isocline, is
given by
(Ms + f)y + kx = 0
which is a straight line passing through the origin. Once the isoclines have
been drawn the approximate trajectory, starting from a given initial state,
can be rapidly constructed. Figure 7.8a shows some isoclines and a typical
trajectory for the case where the drive system is underdamped. The trajectory
spirals into the origin. In the case f = 0, the response of the system is contin-
uous oscillation and the trajectory is closed as shown in figure 7.8b. For
comparison, the corresponding displacement-time curves are shown.
If we need to calculate the time to travel from state A to state B, we can use
the relation
B

t = Jtdx
A
(7.15)

7.3.4 Describing Function Methods

The stability of linear systems may be studied using frequency response

methods 2 , developed by Nyquist and others. The methods are extremely
powerful and hinge on the fact that sinusoidal inputs, such as would occur
during steady oscillation, produce sinusoidal responses when applied to
linear elements. Non-linear effects, however, cause distortion such as when
a saturating amplifier clips the peaks of a sinusoidal input voltage. Con-
sequently the Nyquist method cannot be applied directly to non-linear

156
X

Phose- plane Dtsplocement • ttme

trajectory curve

(a)

01splocement- time
y X curve
Phose ·plane
traJectory

(b)

Figure 7.8 System responses: (a) under-damped case,

(b) zero-damping case

systems such as those involving friction. If appreciable inertia is present,

the higher frequency harmonics representing the distortion may be attenuated
leaving an almost sinusoidal response, and the frequency response method
can be modified, as in the describing function approach 3 , to yield useful
information about the approximate amplitude and frequency of the oscilla-
tion. The reader who wishes to consider the use of this method should
consult the references quoted 2 •3 .

7.3.5 Simulation

Analogue computers have been used in the study of systems involving

friction 4 •5 . The non-linearities present can be represented with reasonable
accuracy by standard computer elements. Precise data regarding the friction
characteristics, both static and dynamic, are unlikely to be available to the
system designer, but simulation permits rapid examination of the perform-
ance for a range of likely frictional conditions.

157
7.4 FRICTIONAL FORCE MODELS

The analysis of frictional oscillations requires models for the system elements
and the frictional forces. In electrical, mechanical or fluid mechanisms,
substantial progress has been made in meeting this need. Frictional forces
on the other hand, depend on many factors in ways which are not yet
quantified and which, in many cases, are subject to large variations in the
industrial environment. Consequently the analyst must select a model which
he believes is representative and of sufficient detail to meet the object of his
study, whether it be the elimination of troublesome oscillations or the
prediction of critical velocities and frequency and amplitude characteristics.
Since frictional force varies considerably with normal pressure between
the mating surfaces, geometry changes in some mechanisms may produce
significant changes in frictional force, and successful analysis may depend
more on taking into account the normal pressure variations than including
changes in the coefficient of friction. Here we confine our attention to those
systems where the normal force is constant, and are concerned with the
stability of steady sliding motion.
It is evident that the coefficient of friction will vary from one point of a
slideway to another. However, it is unlikely that such changes will be other
than small and irregular. Consequently, although they may cause a succession
of small disturbances to the motion, they are unlikely to be a major cause of
prolonged oscillation. Of course, the disturbances in velocity which they
initiate may be followed by oscillatory motion if the system is unstable.
Changes in frictional forces which are related to velocity or acceleration
changes are important. It is worthwhile emphasising that frictional vibration
usually occurs at low sliding speeds although less severe oscillations have
been observed at high speeds 6 . Where hydrodynamic lubrication exists,
frictional forces often increase the damping. So we concern ourselves with
models valid in the neighbourhood of zero velocity.
We have already met the basic models in section 7.2, figures 7.5a and b.
In the first model, the friction force is assumed to drop instantaneously
as slip starts and then remain constant. In the second, the friction decrease
is gradual. In both cases, the friction during stick is supposed to match
instantaneously the drive force tending to cause motion until this force
increases to a value just exceeding F., at which instant slip commences.
Also the frictional force during slip depends on the instantaneous value of
velocity.
A simple extension is to combine these models in various ways as shown
in figures 7.9a and b. At greater speeds the force resisting motion is likely
to be greater so the model may be modified to that shown in figure 7.9c. Of
course the discontinuities present in these models will not occur in practice,
but their piecewise linear nature eases the analysis. Alternatively, we select
an f-v curve with a continuous gradient during slip and represented by an

158
F F
F, F,

Sl1p veloc1ty

-F, -F,
(a) (b)

F F
F, F,

0 Shp veloCity

~ -F,

(c) (d)

Figure 7.9 Simplifiedmodelsfor the friction-velocity

characteristic

expression such as
(7.16)
Such a model is sketched in figure 7.9d. It need not be put into analytical
form if the phase-plane method is used.
A prolonged stick phase normally results in a greater value of the friction,
F., at the start of slip than is the case when the stick time is short 7 . The effect
varies considerably with the slide conditions, particularly with the lubricant.
A typical variation is shown in figure 7.10, and some researchers have
represented the effect by an expression such as
(7.17)
At this stage in our discussion, the position looks favourable since there is an
amount of data describing the variation of frictional force with velocity.
Unfortunately most of this data has been obtained from steady state measure-
ments, under conditions of constant velocity for each friction force recorded.
During an oscillation, however, the velocity is not constant and it would
be preferable to use data obtained from experiments made under conditions
of velocity variations and oscillation frequencies similar to those occurring

159
F

0 t,

Figure 7.10 Effect of stick time

on static friction

in the vibration study. Studies by Bell and Burdekin 8 , have shown that for
typical machine-tool slideway conditions the slope of the f -v curve for
dynamic conditions, that is during oscillation, differs from that recorded
under steady-state conditions. However the concept of a model such as
that in figure 7.5b may still be of use in the analysis of frictional oscillations
providing that the slope is measured under the appropriate dynamic condi-
tions; this approach has been used with some success by Bell and Burdekin 9 •
Another dynamic effect is hysteresis, for in some conditions it has been
observed that the f -v curve is not single-valued t. The slope during accelera-
tion may differ from that during deceleration, as is indicated by the simplified
model shown in figure 7.11. Another method which might account for
dynamic effects is to regard the friction as a function of both velocity v and
acceleration a in the simplified form
F = F. + At v + A2 a (7.18)
where At and A2 are constants.

Figure 7.11 Simplified hystere-

sis effect in the friction-velocity
model

160
7.5 ANALYSIS OF STICK-SLIP OSCILLATIONS

We have seen in section 7.2 that a friction-velocity characteristic having a

negative gradient will give rise to oscillations if the drive system damping is
insufficient, and the analysis was straightforward. Here we wish to study the
behaviour of the same drive systems, figure 7.4, when the friction-velocity
relationship displays a stiction characteristic, besides the negative gradient,
as shown in figure 7.9a. Derjaguin et a/ 10 have presented an extensive treat-
ment of this situation, including the effect on the oscillations of the dependence
of friction on the time of stationary contact.

7.5.1 Description of the Motion

Before the analysis, we will discuss the main features of the motion, starting
from a condition where the mass is at rest, the drive spring is uncompressed
and the drive end commences to move with constant velocity v. The friction
force will be assumed independent of the time of stationary contact.
The spring is compressed until its compressive force plus the dashpot
force reaches a value equal to the stiction force F •. The mass starts to slip
and it is assumed that instantaneously the friction force drops to the value Fe.
Let us measure the time, t, from that instant. Then the unbalanced force
acting on the mass at t = 0 is F. - Fe and hence the acceleration at t = 0 is
(F. - Fe)! M. The slip velocity increases until eventually the spring force
drops sufficiently for deceleration to commence. If the deceleration phase
ends without the velocity falling to zero, then sticking does not reoccur,
and subsequently the mass velocity tends to the drive velocity, v, in the manner
shown in figure 7.12c.
If the velocity falls to zero during the deceleration phase, then there are
two possibilities for the next phase, either reversal or sticking. In this treat-
ment we shall exclude the former, since it does not occur 10 if F. < 3Fe
which is normally the case. If sticking occurs then again the drive spring is
compressed until the mass slips, with the consequent drop in friction, and the
system is then in the same state as at the beginning of the first slip phase.
Thus the cycle is repeated and so on.
Summing up, the criterion for this theory is that stick-slip occurs if the
velocity during slip falls to zero following an initial sudden acceleration
from rest caused by the drop in friction F. - Fe.

7.5.2 Analysis

Let x denote the spring extension. Then since the drive velocity v is constant
dx dX
-=--V (7.19)
dt dt

161
and
d 2x d2 X
(7.20)
dt 2 = dt 2

The equation of motion for the mass is

d2 X [ dXJ
M dt2 = - f dx
dt- kx- Fe- .Adt (7.21)

dx
dt

Col
dx
df

(b)

dx
dt

(c)

Figure 7.12 Velocity vari-

ations (velocity= v + (dx/dt)):
(a) velocity during slip falls
below zero; (b) stick-slip oscil-
lation ensues; (c) increased
damping, velocity remains posi-
tive and oscillation decays

162
that is
2
Md -x
dt
2
+ f dx
-dt + kx = - [F
c
- A. (dx
-
dt
+ v)] (7.22)

that is
d 2x dx
M dt 2 +(f-A.) dt + kx = -[Fe- A.v] (7.23)

In terms of the damping ratio, (, and the undamped natural frequency, w",
given respectively by
f-A.
(7.24a)
( = 2(Mk) 112

and

(7.25b)

the equation of motion becomes

1 d 2x 2( dx
--+ --+x=-S (7.26)
w; dt 2 wn dt
where
F - A.v
s= ~
wn
2 (7.27)

The solution of equation 7.26 is found by adding the particular integral

solution, x = - S to the complementary function solution which is of the
form of equation 7.4. It may be written as

x = -S + e-<"'"'{A cos[wn(l - ( 2) 112 t] + B sin[wn(l - ( 2) 112 t]} (7.28)

from which we find

dx
dt = e-<"'"'{[ -(w"A + wn(l - (2)1i2B]cos[wn(l - (2)1i2t]

- [wn(l - ( 2) 112 A + (wnBJsin[wn(l - ( 2) 112 t]} (7.29)

At t = 0 the following conditions apply

dx
kx + f-
dt
= -F s (7.30)
and
dx
-+ v = 0 (7.31)
dt

163
From equations 7.28, 7.29, 7.30, 7.31, we obtain

x = -S + e-~w"'[(2v 1 - b)cos(wnyt) + ;. (2( 2 v1 - (b- V1 )sin(wnyt)J

(7.32)

and

dx
-
dt
= w . [
e-'w"t
n n Y
.
-v 1 cos(w yt)- (v 1 - 6 sm(w
n
yt) J (7.33)

where b =(F.- Fc)/k, y =(I - () 112 and v 1 = vjwn. The acceleration

during slip is

(7.34)

A plot of dxjdt versus t, from equation 7.33, takes the form shown in figure
7.12a. If it intersects the line (dxjdt) = - v then sticking occurs and the actual
motion is stick-slip as shown in figure 7.12b, where PQ represents the stick
phase. If there is no intersection, the motion is as shown in figure 7.12c.
Examining the curves in figure 7.12 we see that dxjdt overshoots zero,
which would be its steady value in the absence of stick -slip, and then under-
shoots. It is the severity of this undershoot which is important. The under-
shoot is increased if the initial acceleration is increased.
For a given sliding velocity, v, the critical acceleration, and hence the
critical value of the difference F, - Fe, is determined by the condition
that the curve of undershoot touches the line (dxjdt) = - v. In other words,
(d 2 xjdt 2 ) is zero when (dxjdt) = - v. From equation 7.34, (d 2 xjdt 2 ) is zero
when

(7.35)

Note that the lowest value of t satisfying equation 7.35 occurs at the first
overshoot, whereas we require the second lowest value, say T1 •
A second relation between T 1 and c5jv 1 is obtained from equation 7.33 and
the condition (dxjdt) = -r. We shall denote the important parameter c5jv 1
by P. In terms of the system parameters

p = _F::.. _s_-_F_:_c (7.36)

Mvwn

These relations may be solved numerically, for a given value of (, to yield

the critical value of P. The report by Derjaguin et af1° presents a curve
showing the relation between ( and the critical value of P, and also examines
the duration of the slip and stick stages.

164
7.5.3 Phase-Plane Analysis

The phase-plane is useful for a graphical study and shows clearly the
importance of the relative values of the initial acceleration at the start of
slip and the sliding velocity. By differenriating equation 7.26 with respect to t
we find

I d 2 (dx) 2( d (dx) (dx) (7.37)

w; dt 2 dt + wn dt dt + dt = O

In terms of non-dimensional phase-plane coordinates

1 dx
u = --
v dt
1 d 2x
W=--
VWn dt 2

equation 7.37 becomes

dw
w du + 2(w + u = 0 (7.38)

A trajectory satisfying equation 7.38 and the critical condition of touching

the line AA, where u = -1, is sketched in figure 7.13a. It is drawn for a value
of ( which is positive and less than unity.
More generally the initial state is represented by a point on the line AA.
If it lies below P, at Q say, then for the same value of(, the trajectory will not
touch AA and stick -slip is absent. If the initial state is above P, say at R,
then the trajectory intersects AA and stick-slip is present. Hence, for a given
value of(, it is the value of the parameter w1 = 0 which determines whether or
not stick -slip occurs. In terms of the system parameters
Fs- Fe
wr=o =
Mvwn
=P
Now consider the motion following a transient disturbance from a state
of steady sliding. If the disturbance causes the point (u, w), representing the
state, to pass into the cross-hatched region, then a period of sticking follows
and stick -slip will ensue if the steady sliding speed is less than the critical
speed.
An alternative use of the phase-plane method is to plot trajectories in
the (x, dx/dt) plane as in figure 7.13b.
Note that if ( > I the trajectory is as shown in figure 7.13c and stick-slip
does not occur.

165
A

IJ

A
Col

dx
df

dx
df

Slip

X
0 0 X

-v -v
SliCk lniliOI period
of SliCking
(b) (c)

Figure 7.13 Phase-plane trajectories: (a) critical velocity

condition (u, w) plane; (b) stick-slip trajectory (x, dxjdt)
plane; (c) over-damped case

7.5.4 Geometry Effects

The analysis given above is concerned with situations where the friction
force varies with speed. In other situations, friction forces may vary also with
geometrical changes. These will not be discussed in this book except to say
that they feature prominently in investigations of brake squeal and associated
topics' t.tz.

7.5.5 Zero Damping Case

To conclude this chapter with an example, we use the theory presented
earlier to study the motion of the drive system when both the driving damping
and the friction force variation with velocity are negligible with the con-
sequence that the damping ratio ( is zero.

166
Equation 7.26 becomes

(7.39)

At t = 0, the conditions are

kx =-F. (7.40)
and
dx
-+
dt
v=O (7.41)

The solution of equation 7.39, satisfying conditions 7.40 and 7.41 is

x =- ~c -0 cos(ront)- v 1 sin(ront) (7.42)

which leads to

(7.43)

and

(7.44)

The maximum slip velocity is reached at a time -r given by

0
tan( ron -r) = - -
v,
that is
tan( ron -r) = - P (7.45)
From equation 7.45 we obtain

sin(wn -r) = (1 +~ 2 ) 12
and
1
cos(wn -r) = - (1 + p2)lf2
Substituting these expression in equation 7.43 gives the maximum slip
velocity v.m as

v.m = v + (ddx)
t max (7.46)
= v[ 1 + (1 + p2)ti2J

167
The spring deflection, x, at the instant of maximum slip velocity is - FJk.
From equation 7.43 the duration of slip r 51 is obtained as
(7.47)
and equation 7.42 establishes that the spring deflection at the end of slip is
(F. - 2F0 )/k. If F. - 2Fc > 0 then there is a tendency for velocity reversal to
occur. If we assume that this would result in an immediate reversal of the
friction force to at least a magnitude of F c then velocity reversal will not
actually start unless F.> 3Fc. We shall ignore such an improbable case.
Hence at the end of slip, sticking occurs, and during this stage
dx
- =-v
dt
Consequently if the duration of sticking is r 1 , then the spring deflection at the
end of sticking will be given by

(7.48)

Friction Independent of Stick Time

Assuming that the friction force can rise instantaneously from Fe to F. then
the duration of the stick stage satisfies

that is

(7.49)

The total period is

The principal features of the oscillation are shown in figure 7.14a.

Friction Dependent on Stick Time

The above analysis requires modification if the friction force at the end of
the stick period grows with the duration of sticking, t, according to some
relation
F = F(t) (7.50)
If the duration r 1 of the stick period is low then
F(r 1 ) < Fs

168
X

0~-

(a)

dx
dl

0 X

No sttcktng
1occurs dunng
steady state
OSCIIIOIIOn

(b) (c I

Figure 7.14 Oscillations of drh·e system, zero damping

case: (a) stick-slip oscillations, Fs independent of stick
time; (b) stick-slip oscillations, F, dependent on stick time,
stick phase tends to finite size; (c) raised velocity, Fs de-
pendent on stick time, stick phase decreases and is absent
after several cycles

169
and it follows that the acceleration at the start of the second period of slip
is less than at the first, and the oscillation is of smaller amplitude. The stick
period duration following the first slip, r 1 , satisfies

F( r 1 ) = - k [ F.-k 2Fc - vr 1 J (7.51)

In general, the duration r" of the stick period following the nth period of
slip satisfies

F(rn) = -{F(rn1~ 2Fc _ vrn] (7.52)

The oscillation converges into a steady oscillation if

F(rn)- F(rn_ 1)--+ 0 as n -+CO
Hence from equation 7.52
F(r 00 ) = -F(r 00 ) + 2Fc + kvroo
that is
(7.53)
If F(t) is known, equation 7.53 may be used to establish a critical velocity,
above which ! 00 is zero signifying that the stick-slip character of the oscilla-
tion disappears after sufficient cycles of alternate slipping and sticking.
The mass velocity then varies sinusoidally between limits (0, 2v). The trajec-
tories for velocities below and above the critical velocity are shown in figures
7.14b and c.

7.6 FURTHER ANALYSIS

The previous section has presented methods of analysis which are applicable
to drive systems described by linear second-order differential equations with
frictional characteristics as in figure 7.9a. We are often concerned with systems
of higher order or more complex friction models. Hydraulic drives, for
example, because of compressibility effects, are often described by third or
higher order differential equations. In such cases we are entering regions of
difficulty beyond the scope of this chapter but some suggestions may be
helpful. We shall consider first a third-order system.
In section 7.5, we used for the stiction case the criterion that stick-slip
persists if sticking reoccurs following slip from an initial sticking state.
Now for a third-order system the state is dependent on three variables, not
the mass velocity and spring deflection only as was the case for the second-
order drive system. Consequently the state at the end of the stick period
following the first step is not necessarily identical with the initial stick state.

170
In such a case, one possible procedure is to approximate the system by a
second-order equation for a preliminary attack, establishing the approximate
critical velocity before refining the value using analogue or digital computers.
If the problem is to study stability in a small neighbourhood of a steady
sliding velocity, and the friction-velocity curve is linear within this region,
then the Routh-Hurwitz stability criterion may be applied. The governing
equation should be arranged in the form
d 3z d 2z dz
ao dt3 + al dt2 + a2 dt + a3 z = A (7.54)

where z is the dependent variable, A is constant, and a0 , a 1, a2 and a 3 are

coefficients which depend on the drive parameters and the local gradient of
the f -v curve. Then by the Routh- Hurwitz criterion the system will be
stable providing a0 , a 1, a2 and a 3 are all positive and satisfy the inequality
(7.55)
If the criterion is satisfied, then a small disturbance to the steady sliding
motion will not result in prolonged oscillation. For a statement of the
Routh-Hurwitz criterion applicable to higher order systems the reader
should consult, for instance, reference 2.
In situations where it is desirable to take into account the varying gradient
of the f -v curve, then the phase-plane method may be used 13 , provided
that the drive system is described by a second-order differential equation
similar to equation 7.7. The combination of friction non-linearity and higher-
order drive system equations usually requires the application of simulation
methods, describing function methods or advanced theory of non-linear
vibrations.

7.7 ELIMINATION OF STICK-SLIP

In this section guide lines are suggested for the redesign of the system to
free it from oscillation troubles. First, note that the trouble stems from the
interaction of the drive dynamics with the friction characteristics of the
bearing surfaces. Consequently we may consider altering one or both of the
drive and bearing systems.
One approach is to make the frictional forces so small that their variation
with velocity is of no importance. This may be achieved by using hydrostatic
bearings, ones specially coated with low-friction material, or those using
rolling elements. Such bearings are described elsewhere in this book. Although
there are applications where the designer is already considering their use
for other reasons, in others he may wish to avoid them on grounds of expense
and complication. Also, low-friction bearings do not provide the damping
that plain bearings may furnish outside the troublesome velocity region.
A second approach notes that for the situations we have considered the

171
undesirable feature of the f-v curve is a negative gradient or stiction effect.
One way which has been used successfully with machine-tool drives 14 , is to
select the slideway lubricant from among those mineral oils with polar
additives so that the f-v curve gradient is approximately zero or even
slightly positive for low velocities.
Looking now at the effect of the drive design parameters it is seen
immediately that one way of ensuring stability, or reducing oscillation
amplitude, is to increase the drive damping. This may be achieved by a
direct method such as adding a dashpot device, or indirectly as in servo-
mechanisms practice where velocity feedback improves stability 15 . Added
damping may make the system sluggish in velocity regions where the friction
gradient is positive. If the designer is prepared to add complexity he may
introduce non-linear damping so that the damping coefficient is high in the
dangerous speed regions and low elsewhere.
Normally, a very stiff transmission system is chosen so that excessive
deflections under load are avoided. Decreasing the stiffness usually causes
the stick -slip amplitude to increase but it should be remembered that the
drive damping ratio, for a fixed amount of damping is reduced when the
stiffness is increased. P.robably the major effect of increasing transmission
stiffness results from the consequent increase in natural frequency. When the
speed fluctuates slowly the .f -v curve, at low speeds, may display a strong
negative gradient characteristic whereas at high frequencies of speed fluctua-
tion the gradient may be effectively zero 9 . Also, in situations where the growth
of friction forces during the stick stage of an oscillation is important, the
increased stiffness is beneficial, since it decreases the stick time.

REFERENCES

I. H. Catling. Stick-slip friction as a cause of torsional vibration in textile drafting

rollers. Proc. Instn mech. Engrs, 174, No. 17, (1960), 575-86.
2. B. Porter. Stability Criteria for Linear Dynamical Systems. Oliver & Boyd,
Edinburgh, ( 1967).
3. H. E. Merritt. Hydraulic Control Systems. Wiley, New York, (1967).
4. A. J. Healey and J.D. Stringer. Dynamic characteristics of an oil hydraulic constant
speed drive. Proc. lnstn mech. Engrs, 183, Pt I, No. 34, (1968-9), 683-93.
5. J. Parnaby. Electronic analogue computer study of non-linear effects in a class of
hydraulic servomechanisms. J. mech. Engng Sci., 10, No.4, (1968), 346-59.
6. J. B. Hunt, I. Tor be and G. C. Spencer. The phase-plane analysis of sliding motion.
Wear, 8, (1965), 455-65.
7. S. Kato, N. Sato and T. Matsubayashi. Some considerations on characteristics of
static friction of machine tool slideway. Trans. Am. Soc. mec-h. Engrs, Series F.,
J. Lubric. Tech., 94, No.3, (July, 1973), 234-47.
8. R. Bell and M. Burdekin. Dynamic behaviour of plain sideways, Proc. Instn
mech. Engrs, 181, Pt I, No.8, (1966-7), 169-84.
9. R. Bell and M. Burdekin. A study of the stick-slip motion of machine tool feed
drives. Proc. lnstn mec·h. Engrs, 184, Pt I, (1969-70).

172
10. B. V. Derjaguin, V. E. Push and D. M. Tolstoi. A theory of stick-slip sliding of
solids. Proc. Conf Lubric. and Wear., lnstn mech. Engrs, (1957), 257--{)8.
II. R. T. Spurr. A theory of brake squeal. Proc. lnstn mech. Engrs, No. I, (1961--{)2),
33-52.
12. R. P. Jarvis and B. Mills. Vibrations induced by dry friction. Proc. lnstn mech.
Engrs, 178, Pt I, No. 32, (1963--M), 847-66.
13. B. R. Dudley and H. W. Swift. Friction relaxation oscillations. Proc. R. Soc.,
(1949), 849-61.
14. M. E. Merchant. Characteristics of typical polar and non-polar lubricant additives
under stick-slip conditions. J. Lubric. Engrs, 2, (June 1946), 56--{)1.
15. P. L. Taylor. Servomechanisms. Longman, Harlow, (1969).

173
8
The Mechanics of
Rolling Motion

8.1 INTRODUCTION

It is not without significance to society that the principle of rolling as a

method of translation dates from antiquity. None the less, this method of
translation may be properly claimed to be the product of man's ingenuity,
and indeed it is known that such an advanced society as the Incas did not
discover this principle.
In view of the very long history of usage of this method of translation, the
lack of useful experimental data concerning it is somewhat surprising. There
is undoubtedly a widespread awareness among engineers of the convenience
and efficiency exhibited by rolling systems. Thus we find a whole catalogue
of engineering situations to which rolling motion has been applied, namely,
wheels of all types, ball- and roller-bearings of many designs, variable speed
friction drives, belt drives (which may be considered as the rolling of a
pulley along a continuous belt), knife edges which constitute a small radius
rolling contact, gear teeth which function by a process of combined rolling
and sliding, and others.
These developments suggest that any rolling situation should be analysed
from three standpoints where any combination of the following effects may
be contributory.

174
8.1.1 Pure Rolling or Free Rolling
This represents the most elemental form of rolling motion and is probably
most nearly approached in the case of a cylinder or ball rolling without
constraint in a straight line along a plane.

8.1.2 Rolling with Combined Applied Surface Tractions

These conditions obtain in the case of a driven or braked wheel, where the
applied forces induce normal and tangential tractions in the contact zone
between the wheel and its track.

8.1.3 Rolling with Regions of Slip in the Contact Zone Necessitated by the
Geometry of the System
A ball rolling around the inner race of a ball-bearing may, to a first approxi-
mation, be considered to rotate about some instantaneous axis of rotation
necessitated by the geometric conformity between the ball and its track,
figure 8.1. From elementary kinematic considerations it then follows that
surface particles in the contact zone remote from this axis of rotation must

instoa~eu
axis

Figure 8.1

necessarily slip. None the less, it should be recognised that such slip may in
part be suppressed because of the ability of the materials to sustain elastic
deformation.
In a similar way, a ball rolling around a curved track is subjected to a
twisting moment which induces a degree of slip between the ball and its
track.

8.2 FREE ROLLING

Since all bodies are made from deformable materials some deformation must
ensue when they come into contact under load. For bodies having a geometry
defined by circular arcs, such contact occurs over a relatively restricted area,
but it is imperative that we appreciate that 'line contact'-the contact between
two cylinders or a cylinder and a plane-and 'point contact' -the contact

175
between two spheres or a sphere and a plane-still necessitate some finite
area of contact. The shape and size of the area of contact will depend upon
such factors as the individual geometry, the load, and the deformation
characteristics of the materials. Thus a cylinder rolling along a plane results
in a rectangular area of contact of width 2a, figure 8.2a, whereas a sphere
rolling along a plane results in a circular contact area, figure 8.2b. Both these
solutions are special cases of the well known hertzian analysis from which we
might postulate the general case shown in figure 8.2c. For this case the
contact would be elliptical in shape defined by the two semi-axes a and b,
(see chapter 3).
As body I in figure 8.2b rolls along body 2, it will be recognised that

(a)

®
'
' .!~_ ' I
2o
'1'7\•
\V2_
Contact-,- (b)
area I

' '
®-l2b b_
A> 'd7'i. 2o
~ ~
Conre~
area , I
(c I
( R 12 - radous of body I on plane 2 etc I

PLANE 2 PLANE I

Figure 8.2 Types of contact areas

176
material ahead of the line of centre in both bodies is being compressed,
while material to the rear of this line is having this stress state relieved.
For these conditions hertzian analysis applied to the elastic deformation of
such bodies results in the following observations. (See chapter 3 for an
elementary treatment of hertzian contact).

8.2.1 Cylinder on a Plane

The local pressure p in the contact area due to an applied load P is

p = 2P
rr.al
(I - xz)l/2
a2
that is, a semicircular pressure distribution with a maximum pressure at the
centre of the contact zone of value 2P jrr.al (figure 8.3). The value of the contact
semi-width a is given by

a = [4PR (·~ - vf + I - v~)]' 12

rr./ £, Ez
where vis Poisson's ratio and E is Young's modulus, suffixes 1 and 2 relate to
the bodies shown in figure 8.3.

~ 2a --j

Figure 8.3

Consider an elemental strip t5x at a distance x from the centre of the contact
zone. During the forward compression of the material in bodies 1 and 2, this
pressure gives rise to a resisting moment M about the centre of the contact
zone where

f
a
2Pa
M = plxdx =-
3rr.
0

177
Under ideal conditions the cylinders would be subjected to an equal and
opposite moment arising from the pressure distribution in the rear part of
the contact region. For the present, however, we will neglect this effect so
that if the cylinder rolls a distance x we note that the elastic work done due
to the forward compression alone would be

¢ = Mx = 2Pax
R 3nR
Let us now postulate that the rearward recovery does not replace the
whole of this elastic work of forward compression. This might be explained
by an energy dissipation due to the elastic hysteresis loss occurring in the
complex straining of the material which must occur during the rolling
process. If this loss is defined by a coefficient t: we can see that the work
dissipation during rolling a distance x will be t:</J and, if this loss represents the
rolling resistance, it follows that
2Pax
Fx = t:</J = t : - -
3nR
where F is the force required to overcome this resistance. The ratio F / P may
be specified as the coefficient of rolling resistance A., having an analogous
meaning to the coefficient of sliding friction Jl, hence

A. = !::. = ~ !!!__
P 3nR

= ~
3
x ___!:__
nR
x [4PR (1 - vi + l - v~)] 12
(8.1)
nl E1 E2

It will be noted that A. depends on the geometry of the roller, the applied load
and the elastic constants of the two bodies in contact. When one body is
very rigid, as for a rubber cylinder rolling along a steel plane, £ 2 has a very
large value with respect toE 1 and consequently the term 1 - v~/E 2 may be
neglected in any computation.
Although the foregoing provides a quantitative statement about the
resistance to free rolling of a cylinder, several qualifying observations must
be made.
(a) The hysteresis loss factor t: used in the expression for A., is not the same
as that which would be obtained from a simple tensile test. This arises
because in the rolling process, the elastic deformation of any particular
element of material passing through the contact region is a very complex
process in relation to the relatively simple uniform deformation pattern
in simple tension. Consideration of these two processes of deformation
suggests that the hysteresis loss factor for a rolling cylinder would be
about three times the loss factor obtained in a simple tension test.
(b) The hysteresis loss factor has here been assumed constant, whereas

178
some evidence exists to suggest that at larger strains it may vary with
the degree of straining. Furthermore, little is known about the effect
of strain rate on hysteresis, so that the foregoing analysis would suggest
that the free rolling resistance is independent of velocity.
(c) For a cylinder rolling along a plane it is well established that the
maximum stress condition exists below the surface of contact, (chapter 3).
Thus at high loads it is probable that this high stress region could
reach the condition for plastic flow giving the contact as shown in
figure 8.4. As rolling proceeds, any element of material just below the

Elastic

Elast1c :Pi~;Sfc' Elastic

Po1nts of ' , .. • • _.I
mox1mum stress Elosflc

Figure 8.4

surface passes through an elastic-plastic-elastic condition. This

process will lead to an appreciable plastic energy dissipation, even
though the actual contact surfaces are still at an elastic stress level. It
has been shown that, under certain loading conditions, such a situation
induces residual stresses so that after several cycles of rolling a condition
of shake-down to a purely elastic state may be achieved. Thus, in situa-
tions where any plastic deformation occurs, the resistance to rolling will
be increased by the plastic work dissipation, in addition to any elastic
hysteresis effects.
(d) In this analysis it has been assumed that the materials in contact are
perfectly elastic and provide a finite, time independent, hysteresis loss.
With many materials this is not the case, since there exists a relaxation
time and recovery to the rear of the contact zone may be incomplete.
Furthermore, such effects are known to be both time and temperature
dependent, but the results of analyses based on such visco-elastic
behaviour cannot be treated in this volume.

8.2.2 Sphere on a Plane

For this case the pressure distribution in the contact region is given by

p = 3N
21ta 2
(1 - .C.)I/2
a2

179
and the radius of the contact circle by

a=
3 (I---vf+I-- -v~)]'i
[4-NR 3
£ E 1 2

To derive the resisting moment arising from the forward compression of

material, we again consider an elemental strip <5x, figure 8.5. The total vertical
force on such a strip due to the pressure distribution is then given by

3N
-
~a
-2 dx
+ (al _ xl)l/2

J [I - (x 2 +2
a
i)]' 12
dy =3N3 (a 2 -
4-a
x 2 ) dx
_ (al _ xl)l/l

Thus the resultant resisting moment M is

a

M =
3N
4a 3
J(a 2 2
- x )x dx = ~
3Na
0

Again the work done in forward compression in rolling a distance x is

¢ = Mx = 3Nax
R 16R
and introducing a hysteresis loss factor for this condition 1:: one obtains
3Nax
Fx = ~:¢ = £--
16R
where
. F 3 w
).=-=----- (8.2)
N 16 R

Roll1ng

~ I
0

Contact
zone

Fi,qure 8.5

180
In this equation the contact size a is defined by

a= -NR
4
[3 (1--E-vf+. 1--E -v~)]
1 2
1 '3
(8.3)

The same general observations which were made for the cylinder rolling
along a plane still apply, except that the hysteresis loss factor for this con-
dition is only about twice that which will be obtained for simple tension.

8.2.3 General Elliptic Contact

For these conditions the pressure distribution over the ellipse of contact
is given by

p = 3N
2rcab
[1 - xz - i]l/2
a2 b2
For the special case of a = b this reduces to the circular contact condition
of the preceding case. The size of the contact ellipse is defined by the semi-
axes a and b given by

a- - k[ 3N
-
a :EP
(1-El--vf+1--Ez-d)] 1 3
'

b_- k[3N
b-
:Ep
(1---vf+ -Ez-v~)]
El
1- 1'3

where
1 1 1 1
:E =-+-+-+-
p Rll R12 R21 R22
In these formulae ka and kb are functions of the four principle radii of curva-
ture of the contacting bodies and are discussed in chapter 3.
In this case the total vertical force on an elemental strip t:5x, figure 8.6, is
given by
3N b(l- xf2ta2)1t2 ( xz i ) l t 2
f = - dx 1- - - - dy
2rcab a2 b2
- b(l- x2fa2)1f2

Thus
a
3Na
M = f fxdx=-
16
0
and

181
Figure 8.6

Hence
, F 3 ca
1'.=-=-- (8.4)
N 16 R 12
where a is given by

a
= [3N (~
ka :E E +
~)]1!3
E (8.5)
p I 2

In this case 1: would be in the range two to three times the value which will be
obtained from a simple tension test. When a is approximate ly equal to b the
value will be near to 2 whereas when b is very much greater than a the con-
ditions will approach those of the rolling cylinders and a value nearer to 3
will be more appropriate.

8.2.4 Heathcote Slip

In the foregoing arguments the contact conditions have been such as to allow
the assumption that the contact zone lies substantially within a single plane.
In many engineering situations, such as ball-bearing s of all types, this assump-
tion is no longer tenable. Consider a ball rolling along a grooved track as
shown in figure 8.7. The contact area will be an ellipse and provided the
subtended angle of the contact at the centre of the rolling element {J is less
than 30' it is found that the dimensions a and b are still predicted with
reasonable accuracy by hertzian formulae. For this situation, as rolling
proceeds, we must consider the rotation to occur about an instantaneou s
axis of rotation such as AA. As far as the contact ellipse is concerned we then

182
A A

I I
I I I

8L
I I 1 I

~2b
Figure 8.7 Heathcote-type slip

see that three regimes of relative slip must occur; the magnitude of slip
being defined by the product of the angular velocity and the distance of the
particular contact point in the xz plane from the axis AA.
The force F producing motion is now seen to consist of two components.
One component accommodates the hysteresis loss, using the arguments
discussed above, and the other component exists to accommodate the
sliding friction losses. The latter term may be considered as a resisting mo-
ment about the axis AA, arising from the summation of these frictional
effects. The method of carrying out such a summation is discussed in a later
section, since this effect is complicated by one further factor.
In most engineering applications dealing with steel bodies the size of the
contact ellipse and the height of the axis AA above 0 will be very small,
that is, several orders of magnitude less than the radius of the ball. Thus the
actual sliding velocities occurring in the contact zone during rolling will be
very small and furthermore, because of the rolling motion, the sliding
material rapidly passes through the contact region. This means that the
magnitude of relative displacement between particles in the ball and the
track is also very small (typically in the region of 10- 2 mmjs for steels).
Such small relative displacements can, to some degree, be accommodated by
a tangential surface straining of the ball and track materials without the
requirement for any slip. Thus one finds that in problems of this type some
of the sliding indicated in the simplified picture of figure 8.7 does not occur,
and that within the contact zone there are definite areas within which no
slip occurs. These areas are naturally most prone to occur in the regions where
the slip in figure 8.7 would tend to be very small, that is, adjacent to the
instantaneous axis of rotation AA. This somewhat complicated situation,
giving rise to areas of sticking and areas of slipping, is covered by the subject
of microslip and will now be dealt with in greater detail.

183
8.3 MICROSLIP IN ROLLING

8.3.1 Rolling Cylinders

Consider two rolling cylinders subject to a normal load and in which a

tangential traction is transmitted across the contact zone, figure 8.8a. This
could apply to a cylinder subjected to a driving torque, as in a railway
driving wheel, or subjected to a braking torque, in which case the tangential
traction would be in the opposite sense. It will be tacitly assumed that the
width of the contact zone and the form of the pressure distribution are those
predicted by hertzian analysis and will be unaffected by the presence of

Cylinder CD
subjectea to
driving torQue M
M=T,R

(o)

PluM length Pressure

distribution p
D
P=/p.dx
·o
-a I 0 I Jt

~
~Shear stress
distribution
1 1
I I r1t,-pp

I 0 I Jt tl

II Tlu:t length I T " /rJt, dx

1 -a
I

eJt 1 " linear surface

strain in body(D
eJt.z= tinear surface
strain in body ®
Compressoon e..,.- -tlz2·ux

(b)

Figure8.8

184
tangential tractions. If full sliding occurs over the contact region, the total
surface traction T will be 11P and at any point in the contact region the
tangential surface shear stress r xz will be given by J1P where J1 is the appro-
priate coefficient of sliding friction, figure 8.8b.
In the following solution the cylinders will be considered as having unit
axial length, so that the tractions T and P are in the form of load per unit
length. From contact theory it can readily be shown that within the contact
region such a situation produces a linear strain ex given by

J1 x
e =--
X 2R
(8.6)

It will be noted that within the contact region ex is a linear function of x

with a slope of J1/2R. Outside the contact region a more complicated distribu-
tion occurs as shown in figure 8.8b, but its exact form need not concern us
in the following solutions. This represents the solution for a rolling cylinder
where T = J1P.
We now consider the case when T < J1P and to do so we must consider the
<;:ffect of the resultant surface strains ex on the probability of relative motion
between surface particles on the two cylinders. To do this we must obtain an
expression for the velocity of discrete particles incorporating the displace-
ments due to surface straining.
Consider a particle distance X from the origin of coordinates in the un-
strained state and which, due to straining, is displaced a further distance u
in the x direction. The coordinate position of the particle will then be defined
by x, that is
X= X+ u
whence its velocity dx/dt is
dx dX du dx
-=-+······X-
dt dt dx dt
if we assume that u is time invariant, that is, that steady state conditions
prevail.
· Since strains are small, xis approximately equal to X, so that (dx/dt) ~
(dX /dt). But dX fdt is the velocity of the particle if straining is neglected,
that is, the rigid body velocity of the particle U x· The strain in the x direction
e; is given by dufdx so that we may write the resultant velocity of the particle,
say Vx as
(8.7)
If we now consider coincident particles in the surfaces of the two cylinders
we recognise from Newton's third law that the traction on the two bodies
must be in opposition but equal in magnitude so that at any point

185
The strain pattern in figure 8.8 is seen to satisfy this criterion. Thus the
velocities of the two coincident particles on the surfaces of the bodies in
contact may be written
vxl = uxl(l +ex!) (8.8)
Vxz = Uxz(l + exz) = Uxz(l- ex!)
Self-evidently these two velocities can only be identical in areas where
ex 1 and hence ex 2 are constant, since by definition U xl and U xz are constant
rigid body approach velocities. This therefore defines the basic condition
for areas in which the tendency to slip may be accommodated by surface
straining of the bodies. When Vx 1 = Vx 2 equations 8.8 result in the relation-
ship

or
~u
-=-2e
U xl (8.9)

where ~U is the difference between the two rigid body circumferential linear
speeds and U is the mean circumferential linear speed.

u = uxl + ux2 (8.10)

2
To summarise we note that when Tis less than JlP there may exist regions
of contact where no relative slip occurs, provided that within these regions the
surface strain due to the tangential tractions are of a constant value, and that
the circumferential velocities of the two cylinders will be different at points
remote from the contact zone, that is, where surface strains are not present.
An acceptable solution to this problem may now be provided by the argu-
ment shown diagramatically in figure 8.9. Figure 8.9a shows the normal
pressure distribution according to hertzian theory, while figure 8.9b shows
the tangential traction distribution when T' = JlP and r xz = JlP throughout
the contact region. For this condition the strain e~ is given by equation 8.6
throughout the contact region and is shown in figure 8.9c. The strains outside
the contact region are also shown but are not essential to the argument. Now
suppose a traction T" is applied over a contact zone of width 2a, the dis-
tribution and resultant strains are obviously as shown in figures 8.9d and
8.9e, when T" is in the opposite sense to T' for each body. If we now add the
traction in figures 8.9b and 8.9d we get a resultant traction T = T' + T"
for each body, and this is distributed as shown in figure 8.9f. Adding there-
sultant strains gives the result shown in figure 8.9g; the constant strain arising
because the slope of the strain distribution is the same in both figures 8.9c
and 8.9e. This must be so, since equation 8.6 shows that the slope of the strain
distribution J1/2R is independent of the value of either the applied traction or

186
(h)

(i)
"'".?
I
I

(j)

Figure 8.9

the size of the contact area. Examining figures 8.9f and 8.9g shows that we
now have a possible solution to our problem. The resultant value of T is
obviously less than J1P. In the region where the strain is constant, that is, the
region in which no slip occurs according to equation 8.9, we see that rxz is
less than J1P, whereas outside this region 'xz = J1P and in these areas slip must
occur.
This solution therefore appears to satisfy the essential features of the pro-
blem. However, if we examine the outcome in greater detail we find that in the
forward region of slip the inconsistency arises that the slip is in the same
direction as the traction J1p, thereby contravening the basic laws of friction.

187
It is therefore necessary to eliminate this possibility by the simple device of
eliminating the forward areas of slip by making c = a - ex, so that our final
solution is as shown in figures 8.9h to 8.9k. Here a tangential traction T < J.I.P
gives rise to an area of no slip (sticking) of size 2cx, and a region of slip at the
rear of the contact zone. Furthermore, the constant strain in the region of
sticking is given by Figure 8.9k

but since x" = x - c for the system of coordinates used and c denotes the
location of the centre of the stick area

J.l.
=--c
2R
In this region Vx 1 = Vx 2 so that equation 8.9 immediately yields the relation
between the circumferential velocities of the two bodies
AU J.1.
-=-2e
U xl
=+-c
- R
(8.11)

The ± sign in this equation indicates the direction of the applied traction.
The positive sign will apply when the traction is in the direction of rolling as
for body 1 in figure 8.1 Oa, and the negative sign will apply when the traction is
in the opposite direction to the direction of rolling, figure 8.10b. Combining
the values of the tangential traction

0 0
~- ~
~
® I I
~ ® I I
I I I I
t
Tens1on ''
'\ .
\

tr-----(
I •

I Slip SliCk ·I 1- Slip ·I· SliCk I

(a) Body (j)dnven (b) Body (j) braked

Figure 8.10

188
but T'1 = JJ.P = area of figure 8.9b. Since T" is the same form as JlP but over a
contact width 2cx rather than 2a, it follows that T" = JJ.P(cx/a) 2 = area of
figure 8.9h. Hence

(8.12)

that is

(8.13)

since ex = a - c.
Substituting c from equation 8.11 in equation 8.13 then produces
the relationship between the tangential traction and the so called creep
velocity !1.U

(8.14)

or in non-dimensional form

!1.U R = 1- (1 - I:__)t/2 (8.15)

UJla JlP
We therefore see that by considering the linear strain pattern which arises
from the applied tangential traction, we may identify areas of sticking and
areas of slipping in the contact zone between two rolling cylinders. We note
that due to the requirements of the elementary laws of friction this sticking
zone must always lie at the front part of the contact zone. This latter require-
ment is only true for the contact of materials having similar elastic properties 2 .
For materials having dissimilar elastic properties a solution of rather greater
complexity has been proposed which reduces to the foregoing solution when
the elastic constants are the same 3 .
The validity of equation (8.15) has been proved by experiments in which the
creep velocity !1.U has been measured for a range of values ofT 1 , P, U and R 4 .
Typical results are shown in figure 8.11 where the non-dimensional groups
!1.U R/U Jla and T !JlP are seen to agree with equation 8.15 when a steel cylinder
rolls along a steel track.

8.3.2 Rolling Spheres

Consider two spheres rolling together while subject to a normal load and a
transmitted tangential surface traction. The solution for the rolling cylinder
may be applied approximately to this case by considering the sphere to be
represented by a row of elemental cylinders of axial length by embracing the

189
~Il
<J:§- 05

025

02 04 06 0·8 1·0
T
pf>

Figure 8.11 An experimental and theoretical comparison

of the microslip between rolling cylinders

contact circle of the sphere, figure 8.12. The pressure distribution for a
sphere is (chapter 3)
p = JN [ 1 _ x2 _ y2Jtt2
(8.16)
2na 2 a2 a2
At any point y the contact width will extend from - (a 2 - y 2 ) 112 to
+ (a 2 - y 2 ) 112 whence the normal load on the cylinder element 15y is then

that is
3N 2
f.= 4a3 X dy

The normal load per unit length of each of the elemental cylinders is then
given by
p = fz = 3N X2
Dy 4a 3
which is distributed over a contact width of 2X where
x2 = a2- y2

190
Row of roll1ng
elemental
cylinder
of length 8y

Contact area

Figure 8.12 Method of utilising the 'strip theory of

microslip'

The total transmitted tangential traction F may obviously be considered as

being the summation of the individual tangential tractions per unit length T,
which are applied to each of the cylindrical elements by. Except for the case
when F = pN, and thus T = pP for each element by, we have as yet no
method for defining the value of T acting on any given element by for the
conditions when F is less than pN. We may resolve this difficulty by consider-
ing the velocity pattern in such a situation. Provided that the contact zone is
small with respect to the ball size, it is seen that the rigid body approach
velocity for every one of the elemental slips by will be the same, that is,

for body 1

for body 2

Substituting these values in equation 8.11 which relates to the velocity dif-
ference and the constant strain in the stick area in a rolling cylinder, defines
the location of the centre of the stick area as follows

191
Therefore

(8.17)

where cp = wdw 2 and K = J1/R using

1 1 1
- = - + -
R R1 R2
that is, the equivalent radius of the system (chapter 3).
It will be noted that this result suggests that c is the same for all the ele-
mental cylinders by although its magnitude still requires a value for¢, as yet
undefined. We may however obtain a further set of relationships by con-
sidering the equilibrium of the total system having a contact area as shown
in figure 8.13. For strips by at the extremity of the contact circle it is seen that

-F
r

STICk
areas

Figure 8.13

no-stick areas exist so that in these regions the tractions per unit length on
the elemental slip by, TsL• is given by
3N 2
TsL = JlP = J1 4a3 X

3N 2 2
TsL = J1 4a 3 (a - y ) (8.18)

From this equation it is seen that the value of TsL is a function of y having a
zero value at y = a. For other strips when c has a real physical value, the

192
stick area is at the leading edge of the contact zone and the traction, TsT• over
such an elemental strip dY is, using equation 8.13

1ST = JLP[ 2 ; - ;
2
2 J
substituting for P and X then gives
3jLNC
1ST = 4a3 [2(a2 - y2)t;2 - c]
(8.19)
Hence in the elemental strips having stick areas defined by c the tangential
traction per unit length is again a function of the y coordinate of the parti-
cular strip.
Referring to figure 8.13 we can easily identify they values where equations
for TsL and TsT should apply. Self-evidently, the line x = c will intersect the
circular contact boundaries at values Y = ±(a 2 - c 2 ) 112 so that for y values
outside these limits the equation for TsL applies while within these limits TsT
must be used.
Returning to our original problem of two rolling spheres subjected to a
normal load N and a tangential traction F, we see that equilibrium requires
that
a (a2-c2)1/2

F = 2 f
(a2 -c2)1/2
JSL dy +2 f0
TsT dy

f
(a2-c2)1fl
3Nc
(a 2 - y 2 ) dy + 2 x - 3 [2(a2 - y2)t/2 - c]dy
4a
0

These integrations are straightforward and may be used to replace c in

equation 8.17 thus producing a relationship between the total transmitted
traction F and the ensuing angular velocity ratio 4J. The pattern of stick areas
is of course that already shown in figure 8.13.
Experimental results for steel balls rolling on steel tracks show good
agreement with the predictions of this theory 5 · 6 . Further evidence for the
validity of these arguments has been provided by photoelastic studies of
rolling spheres which demonstrate the form of contact stress distribution
shown in figure 8.9f,

8.3.3 Sphere Rolling in a Grooved Track

Heathcote Slip
This topic has been mentioned previously (section 8.2.4) and it will be recalled
that it arises due to the geometric conformity between the ball and its track.
We may, however, now consider the implications of the surface straining in
mitigating such slip requirement by again utilising the simple solution fot
rolling cylinders.

193
Consider the elliptic contact area to be represented by a row of cylindrical
elements of axial length by and let the angular velocities of the ball and track
be w 1 and w 2 respectively. Because of the curvature of the contact zone we
now see that the rigid-body approach velocities will be given by
U 1 = w 1(R 1 - z)
U2 = w 2(R 2 + z)
where z is the vertical coordinate position of the particular element by at a
distance y from the origin of coordinates 0, (figure 8.14). If R, is the actual
N

0 UXz
wz

Figure 8.14 The geometry and kinematics of conforming

contacts

conforming radius of the ball and its track after deformation (see Chapter 3),
we may write
2 = R, _ (R; _ yz)112

and as y and z are small with respect to R, in most practical cases this may
be reduced since
y2)1/2
-
z=R c -R c( 1 -R,

y2 y2
=R -R
c c
+--···=-
2R, 2R,
For any given elemental cylinder by we may then substitute in equation 8.11
and thus obtain

194
Dividing this equation by w 2 R 1 and putting w 1 /w 2 = ¢gives

(¢ - RR) - ~
1
-y2- (l
2R,R 2
+ ¢)
(8.20)
±Kc = ( R ) 2
+~
+ _Y_(1- ¢)
¢
R1 2R,R 2
It will be apparent that¢ will have a value near to R 2 /R 1 so that the two
terms in the numerator of equation 8.20 are comparable, whereas in most
practical situations y is very much less than
2 R,, so that the second term in
the denominator may be neglected with respect to the first term. Equation
8.20 then becomes

(8.21)

This equation shows that for any elemental cylinder by at a coordinate

value y the instantaneous value of c is a quadratic function of y since ¢ is
independent of y. Furthermore it is seen that when c = 0, that is, sticking
occurs throughout the contact zone at that value of y, the values of y must
be given by
¢ _ R2 )
2 ( Rt
y = 2R,R 1 1+¢
Let these particular values of y be identified by y.
The patterns of sticking and slipping in the contact zone must then follow
the form shown in figure 8.15, which shows some typical patterns for high and
low geometric conformity between the ball and its track. It will be noted
that the degree of slipping is very much greater when the conformity is high.
Since equation 8.21 contains two unknowns, namely c and¢, the complete

Stick

St~k:
area

ocoo ~

(a) Contact pattern (b) Contact pattern

for low conformtfy for high con for mit y

Figure 8.15 Stick-slip patterns in conforming contacts

195
solution of this problem requires that the value of the tractions for each
particular elemental cylinder must be inserted and the equilibrium of the
system must then be considered. The value of the traction for any cylindrical
element is of course given by equations of the type 8.18 and 8.19 and the
equilibrium conditions (figure 8.14) are
resolving horizontally

F = f T SI. +f T ST
taking moments about 0
FR 1 = f TsL z +f TsT z
where the integrals are taken over the limits of the appropriate sticking and
slipping regions, taking due account of their directions.
It is not possible to obtain simple analytical solutions to this problem, but it
is worth noting that at very high conformity the stick areas become very
small and slip must occur throughout most of the contact zone. For the limit-
ing conditions where slip occurs everywhere in the contact zone, it can easily
be shown that
F 0.08J1b 2
N Ri
This method of solution may also be adopted to deal with a ball rolling around
a conforming circular track such as occurs in ball thrust bearings, although
such solutions necessitate the use of a digital computer. The resultant stick-
slip patterns of the type shown in figure 8.16 occur.

Slip
areas SliCk area

Corfacf zone

Figure 8.16 Stick-slip patterns in ball thrust bearings

196
We are now in a position to make a qualitative analysis of the stick-slip
patterns which could occur due to an element rolling along 1:t conforming
track and also subjected to either a driving or braking traction, (figure 8.17).
From figure 8.17 we may surmise that if a body is rolling along a conform-
ing track while subjected to a braking traction, the two microslip effects,
(figures 8.17a and 8.17c), will tend to increase the slip at the extremities of the
contact ellipse and reduce the slip near to the centre of the contact, that is,
slip is reduced in the region of high local contact pressure. This has the effect
of reducing wear due to the microslip, and has an optimum condition when
the slip in the high pressure region is eliminated. A detailed presentation of
these arguments may be found in references 8 and 9.

(a)

St1ck/Siip pattern due tractions

St1ck /Slip pollern due conformity

Decrease af
in high
slip~
pressure reg10ns
-
(o)·dcl
® (b)+(c)
Increase of slip
in h.igh pressure
reg10ns

Figure 8.17 Combinations of slip patterns due to

different effects

197
The method of solution proposed in the foregoing is approximate in that
it neglects any interaction between adjacent elemental cylinders. It does,
however, lead to reasonable predictions of the microslip effects for a wide
range of rolling contact situations, and should therefore enable designers to
ascertain the probable effects which would occur in such cases.

8.4 TYRE-ROAD CONTACTS

A very common example of rolling wheels subjected to tangential tractions

where Tis generally less than f.J.N, occurs in the wheels of cars and locomo-
tives. Whether such wheels are driving or being braked, the control of the
vehicle usually demands that macroscopic or limiting slip does not occur.
All such wheels should therefore clearly demonstrate the microslip effects
which have previously been discussed. With steel locomotive-wheels the
preceding arguments will provide reliable predictions of the microslip
behaviour. With pneumatic-tyred wheels the problem is more complex
and cannot be treated in detail in this text. None the less, even with such
wheels the general form of the microslip and traction relationship is similar
to that previously described for solid bodies in rolling contact.
Consider the contact patch between a tyre and the road. As shown in
figure 8.18 we may represent the total frictional tractions by an equivalent
force F ~ f.l.N acting through some point such as X. This system may be

Mr I
u~-+F.
X
0 X -

Figure 8.18 Resolution of the frictional

effects in the tyre-road contact

198
resolved into the three components acting at some origin of coordinates 0
where
(a) Fx is the driving or braking traction which will produce the type of
microslip ef:fects already described.
(b) F, is the transverse tangential traction which will also produce trans-
verse microslip. This farce is the 'cornering force', which as the name
implies is the force present when cornering and producing the change
of direction of the vehicle.
(c) M. is a torque called the 'self-aligning torque' which is endeavouring
to straighten the course of the vehicle. It is self-evident to the driver
through the feel of the steering wheel, and will only exist when the
vehicle is cornering. Clearly when the vehicle travels along a straight
course both F, and M. are zero.
Since Fx, F, and M. all arise from the frictional tractions the resultant
microslip velocities V x, V, and w. must be in the opposite sense to the
frictional tractions and clearly for a straight course V, = w. = 0. The form of
the relationship is similar to that discussed in earlier sections of this chapter.
Figure 8.19 shows the relationship between V x and F x and the similarity to
figure 8.11 is apparent. As Fx --... JJ.N macroslip, that is, skidding, will ensue.

Figure 8.19

Returning to the more general case, consider a car being driven around a
left-hand turn. The force system and the slip velocity components are as
shown in figure 8.20 and the path of motion lags behind the centre-line of the
tyre by the slip angle 0. This slip angle may be interpreted as the angular
microslip arising due to the self-aligning torque M •. The relationships
between F 1 , M. and (}are then most vividly seen in the form shown in figure
8.21 10 • Here we see that in the region ABC, as we get increased cornering
force F,, the self-aligning torque increases and the slip angle increases-our
common driving experience. In the region CD there is very little change in the
199
y

"~ f;
J'\
,.V._x_ __

Veh1cle
path

Figure 8.20

- ---- .... .. .
~ Position depends on pN
Cornering
force

'\ Effect of
~ increos1ng N
I
I
I
I
c ,,
.. ,
Self ohgn1ng torque Mr

Figure 8.21 The relation between the

major variables during cornering-
Gough Plot

self-aligning torque as the cornering force and slip angle increase, that is, we
have lost the 'feel' at the steering wheel. In the region DE, which is close to the
point where full skidding occurs, there is a reversal in the feel at the wheel
as the car drifts around the corner. It is also worth noting that the effect ofthe
friction coefficient p. and the normal load N are fairly modest in the region
ABC but become critical as the pointE is approached.
The nature of the road surface and the tread pattern on the tyre become
increasingly important in wet conditions. In such cases it is possible for
continuous films of fluid to form between the tyre and the road and in the
limit leads to loss of control due to 'aquaplaning'. This phenomenon is then
more a problem in lubrication than dry friction, and is briefly mentioned in
chapter 11. It is in effect an example of elastohydrodynamic lubrication
within the regime where film thickness is defined by equation 11.5.

200
REFERENCES

1. F. P. Bowden and D. Tabor. The Friction and Lubrication of Solids Part II. Oxford
University Press, ( 1964), 288.
2. J. Poritsky. J. app/. Mech., Trans. Am. Soc. mech. Engrs, 72, (1950), 191.
3. J. A. Jefferis and K. L. Johnson. Proc. lnstn mech. Engrs, 182, (1968), 281.
4. J. Halling and S. K. Sen Gupta. Wear, 24, (1973), 127.
5. K. L. Johnson. J. app/. Mech., Trans. Am. Soc. mech. Engrs, 80, (1958), 339.
6. J. Halling. Wear, 1, (1964), 516.
7. D. J. Haines and E. Ollerton. Proc. lnstn. mech. Engrs, 177, (1963), 95.
8. J. Halling. J. mech. Engng Sci., 6, (1964), 64.
9. K. L. Johnson (Ed. J. B. Bidwell). Rolling Contact Phenomena, Elsevier, New York,
(1962), 6.
10. V. E. Gough. Auto. Engr, (April, 1954).

201
9
Lubricant Properties
and Testing
9.1 INTRODUCTION
The term lubricant generally suggests oil or grease simply because they are
the most common lubricants in use; but they are not exclusive and, in fact,
any fluid can be used as a lubricant in the right circumstances. In modern
applications a very wide range of fluids are used as lubricants. Air or gas
bearings are now quite common, there are numerous examples of the use
of water as a lubricant and there is an increasing use of process fluids, one
rather exotic example being the use of liquid sodium as a lubricant in nuclear
reactors. In some situations solid lubricants are used but the main emphasis
in this chapter is on fluid lubricants.
The most important single property of a lubricant is its viscosity, and a
substantial portion of this chapter is devoted to viscosity measurement and
viscous behaviour in general. In the limited space available it is impossible
to consider all fluids in detail and, because of their importance, the remainder
of the chapter concentrates on oils and greases for which there are standard
tests laid down for the evaluation of many properties, including viscosity.

9.2 VISCOSITY
The viscosity of a fluid may be defined qualitatively as its resistance to flow,
and liquids are often described as 'thick' or 'thin' which is really an arbitrary,
visual assessment of their viscosity.

202
The foundations of modern viscous theory were laid down in the seven-
teenth century by Newton who proposed a method of quantifying the
viscosity of a fluid. Newton based his postulations on a model of fluid flow.
which considered the flow to be equivalent to a large number of thin layers of
fluid sliding over each other rather like a pack of playing cards, as shown
in figure 9.1.
The internal friction or viscosity of the fluid gives rise to shear stresses,
r between the relative sliding layers which act in such a way that they tend
to retard the faster moving layer and accelerate the slower layer. Newton

r y
r

Figure 9.1 Newtonian model offluid flow

postulated that the viscous shear stresses were directly proportional to the
shear strain rate. It can be shown that the rate of shear is equal to the velocity
gradient, dujdy and hence
du
roc-
dy
or
du
r = 11- (9.1)
dy

The constant of proportionality, 1J, is sometimes referred to as the viscosity,

coefficient of viscosity or absolute viscosity but the title dynamic viscosity
has been adopted in modern practice and it is recommended that this name
be used to avoid confusion.
Fluids whose viscous behaviour is described by equation 9.1 are referred
to as newtonian fluids, and most common fluids, particularly pure fluids
and fluids which have a simple molecular structure, fall into this category.
Fluids which do not behave in this manner are referred to as non-newtonian
fluids. Examples of non-newtonian fluids are greases, multiphase fluids
and simple fluids under extreme conditions, and their behaviour is discussed
in more detail later in the chapter. In the SI system the unit of dynamic

203
viscosity is 1 Ns/m 2 but in practice the commonly used unit is the poise (P)
which is a CGS unit.

1 poise = 1 dyne sjcm 2 = 1 gjcm s

and
1 poise = 0.1 Nsjm 2

The poise is quite a large unit and it is even more common to use its
submultiple, the centipoise (cP) where

1 Centipoise = 1 X 10- 2 poise

The dynamic viscosity of different fluids covers a wide range of values as
may be seen from the following examples. The dynamic viscosity of air is
of the order of 0.02 cP while water has a viscosity of the order of 1 cP.
Lubricating oils have viscosities in the range 2-400 cP and molten bitumen
up to 700 cP.
Instruments which are used to measure the viscosity of a fluid are known
as viscometers and many of these do not give a direct measure of dynamic
viscosity but determine the value rtf p where p is the density of the fluid.
The same ratio occurs frequently in fluid flow problems and is called the
kinematic viscosity of the fluid for which the symbol v is conventionally used
that is
v = rt/P (9.2)
The SI unit of kinematic viscosity is 1 m 2 /s but in practice the commonly
used unit is the stoke (S), which is a CGS unit.
1 Stoke= 1 cm 2 js = 1 X 10- 4 m 2 js
The stoke being a large unit it is more common to use its submultiple,
the centistoke (cS).
The viscosity of a fluid is not a constant, it is a function of temperature
and pressure and viscous properties vary from one fluid to another. The
newtonian model of fluid flow is adequate for the definition of dynamic
viscosity but it does not explain the factors which give rise to the viscous
effects. To get a better understanding of this we must consider the fluid in
greater detail, in fact, we must look at a fluid on a molecular scale.
All fluids consist of a large number of molecules which behave in a very
complex manner, but there are two main features of their behaviour which
contribute towards the viscous effects
(a) There is an attractive force between all molecules of the fluid. The
force of attraction between any two molecules depends upon the
distance separating them and decreases very rapidly as the separation
increases.

204
(b) The molecules of a fluid are in continuous random motion regardless
of any bulk movement of the fluid, and the mean velocity ofthe molecules
increases with temperature.
If we look again at the newtonian model of fluid flow (figure 9.1) we must
accept that a fluid does not actually separate into a number of layers since
there is a continuous movement of molecules throughout the fluid. However,
the model is still a reasonable approximation for steady flows, since the
quantity of fluid and the conditions in each hypothetical layer will not vary
with time. If we consider two adjacent layers of fluid it is clear that the inter-
molecular attraction forces between the molecules in the two layers will
give rise to an overall force which will oppose any relative sliding motion
of the two layers. This effect is equivalent to a shear stress acting at the inter-
face.
If two adjacent layers of fluid are moving at different velocities the momen-
tum of the fluid in each layer is not the same, and consequently the mass
transfer between layers is accompanied by momentum transfer. When
molecules from a slow moving layer pass into an adjacent layer moving at a
higher velocity they tend to reduce the momentum of that layer, which is
equivalent to applying a shear force opposing the motion. Similarly, mol-
ecules from the fast layer tend to increase the momentum of the slow layer,
which is equivalent to applying a shear force in the direction of motion.
The viscous properties of a fluid are the result of the combined effects of
the intermolecular forces and the momentum transfer. The relative contribu-
tion from each of these two sources depends upon the type of fluid concerned.
In liquids the molecules are close together and quite slow moving, with the
result that the intermolecular forces dominate their viscous behaviour.
On the other hand gas molecules are widely spaced and have higher velocities
than those in liquids and it is the momentum transfer which is predominant.
Viscosity data are available from many sources but a useful collection of
data are presented in the Engineering Sciences Data Sheets 1- 3 covering
common liquids, gases and petroleum products. These also include extensive
references to other sources of information.

9.2.1 Effect of Temperature of Viscosity

The viscosity of a liquid is due almost entirely to intermolecular forces.
As the temperature is increased the liquid expands, the molecules move
farther apart and the intermolecular forces decrease with the result that the
viscosity decreases. Examples of dynamic viscosity-temperature curves
for common liquids are shown in figure 9.2, and for typical lubricating oils
in figure 9.3. Since the variation in density of a liquid with temperature is
small the curves of figures 9.2 and 9.3 also represent the variation of kine-
matic viscosity.
It is convenient to have a mathematical equation relating viscosity and

205
Carbon lerrochloride

-20 0 20 40 60 80 100 120 140 160

Temp ("C)

Figure 9.2 Examples of viscosity-tempe rature re-

lationships for liquids at atmospheric pressure

1000

a:v 100

""' 10
5 SAE 50
SAE 20
SAE 5W
20 50 100 150
Temp ("C)

Figure 9.3 Typical viscosity-tem-

perature relationships for mineral
oils at atmospheric pressure

temperature, and several empirical formulae have been proposed for this
purpose. None of these empirical equations are universally applicable but
one such equation which is simple and gives reasonable accuracy for liquid
lubricants is
log '7 = A + B/T (9.3)
where A and Bare constants and T is the absolute temperature.
With gases, momentum transfer is the dominant contribution to their
viscosity. As the temperature of a gas is raised, the velocity of the molecules
is increased. This gives an increase in momentum transfer and consequently
the dynamic viscosity. Therefore, the effect of temperature on the viscosity
of a gas is opposite to that for a liquid. The dynamic viscosity of air as a

206
function of temperature is shown in figure 9.4. At constant pressure the
density of a gas is inversely proportional to the absolute temperature and
therefore the kinematic viscosity of a gas is more sensitive to temperature
than the dynamic viscosity. Useful information and further references on
viscosity-temperature characteristics are to be found in the Engineering
Sciences Data Sheetsl. 2 •3 •

1xlo-~760:92
Temp ("CI

Figure 9.4 Viscosity-temperature

curve for air

The variation of viscosity with temperature is a very important feature of a

lubricant. For lubricating oils the viscosity index (V.I.) has come into common
use and gives an approximate indication of the effect of temperature on
kinematic viscosity. The viscosity index classification was devised by Dean
and Davies 4 in 1929 and was based on two standard oil series. The standard
oils chosen were Pennsylvanian oils and Gulf Coast oils. The Gulf Coast
oils were thought to have the greatest variation of viscosity with temperature
and were assigned a viscosity index of 0, while the Pennsylvanian oils were
thought to have the least variation and were assigned a viscosity index of 100.
Following this, any other oil should have a V.I. between 0 and 100.
The method of determining the V.I. of an oil sample is shown dia-
gramatically in figure 9.5. The kinematic viscosity of the oil sample is

~
"',:. u---
8 I
!!! H - - - -
> I
E I VI=IOO
E I
~ I
,;;: I
I
100 210
Temp ("F)

Figure 9.5 Method of viscosity

index classification

207
determined at 100 oF and 210 °F. Two reference oils are chosen, one from
each standard series such that they both have the same kinematic viscosity
as the sample at 210 °F. The viscosity index of the sample is then determined
from the formula
L-U
V.I. = L _ H x 100 (9.4)

where L, U and Hare as shown in figure 9.5.

The viscosity index is based on an arbitrary scale and many anomalies
can arise in its use. It has the advantage that it is simple, but because of its
deficiencies it can only be regarded as an indication of the rate of change of
viscosity with temperature. With improved refining, the use of additives
and synthetic lubricants, V.I. values well outside the range 0-100 are now
encountered. Most commericallubricating oils have V.I. in the region of 100,
while automobile engine oils have V.I. of the order of 160 because of the
wide temperature range over which they must operate.
A great deal of work has been devoted to the viscosity-temperature
characteristics of petroleum products and it is now possible to determine with
reasonable accuracy the kinematic viscosity of a product at any temperature
knowing only its viscosity at two different temperatures. Two graphical
methods have been devised for this purpose known as the ASTM (American
Society for Testing and Materials) 5 method and the 'Refutas' method 6 .
Both these methods are based on the Walther equation
log 10 llog 10 (v +a) I = b + c log 10 T (9.5)
where a, b and c are constants which are dependent on the constitution
of the oil.

9.2.2 Effect of Pressure on Viscosity

When the pressure of a liquid or gas is increased the molecules are forced
closer together. This increases the intermolecular forces and the viscosity
increases. Typical viscosity-pressure characteristics are shown in figure 9.6.
With mineral oils the changes of viscosity with pressure in general only
become significant when the pressure exceeds approximately 2 x 107 Njm 2 •
When the pressure rises to 3.5 x 10 7 Njm 2 the viscosity of a typical mineral
oil is approximately twice that at atmospheric pressure. As the pressure
is further increased, the rate of change of viscosity increases, until at very
high pressure mineral oils cease to behave as liquids and become more like
a waxy solid. The variation of viscosity with pressure is very important in
situations which involve elastohydrodynamic lubrication, that is, the
lubrication of heavily loaded contacts such as gears and rolling bearings.
It is convenient to have a mathematical relationship between viscosity
and pressure and many empirical equations have been suggested for this
208
100,000

10.000

1~8
Vi
~
100
" 50 14 Oxl07 N/m 2
10 5xl07 N/m 2
7 Oxl0 7 N/m 2
10 3 5xl0 7 N/m 2
Armospheric
5
-30 0 50 100
Temp ("C)

Figure 9.6 Effect of pressure on viscosity of

SAE 40 oil

purpose. None of these are universally applicable but one such equation
which gives reasonable accuracy for liquids and is convenient for mathe-
matical manipulation is
(9.6)
where 11 is the dynamic viscosity at pressure p, '1o is the dynamic viscosity
at atmospheric pressure and oc is the pressure exponent of viscosity.
Useful data and further references on viscosity-pressure characteristics
are given in Engineering Sciences Data Sheets.

9.2.3 SAE Classification of Lubricating Oils

In the automobile industry lubricating oils are commonly classified by
their SAE (Society of Automotive Engineers) number. In general, the higher
the SAE number, the greater is the viscosity of the oil, but the number does
not specify the actual viscosity, it simply denotes that the viscosity-tempera-
ture curve lies within a band which is specified at temperature of 0 °F or
210 °F. Multigrade oils are given a double number. For example, an oil
designated SAE 5w/30 means that at 0 °F the oil falls in the SAE 5 grade and
at 210 °F it falls in the SAE 30 grade. In the classification lubricants are
divided into two types, crankcase oils and transmission oils. Typical viscosity-
temperature curves for a range ofSAE oils are shown in figure 9.7 and further
details of the method of classification may be obtained from the Society of
Automotive Engineers Handbook.

9.2.4 Non-newtonian Behaviour

For a newtonian fluid a graph of shear stress against rate of shear is a straight
line passing through the origin as shown in figure 9.8. For such a fluid the
viscosity is given by the slope of the line and therefore its viscosity can be
determined by a single measurement to obtain one point on the graph. Any

209
10,000
1000
500
'C 100
50
" 20
10
5

Temp (•C)

Figure 9. 7 Typical viscosity-tem-

perature graphs for mineral oils at
atmospheric pressure

Shear rote

Figure 9.8 Shear stress-shear rate

relationship for a newtonian fluid

fluid which does not behave in the manner described above is classified
as a non-newtonian fluid and it is clear that such fluids do not have a single
value of viscosity. In connection with non-newtonian fluids it is common
to use the term apparent viscosity, which is simply the ratio of shear stress
to shear rate at a specified shear rate. The departure from newtonian be-
haviour can take many forms. It may simply be a departure from the linear
shear stress-shear rate curve or the stress may vary with the length of time
that shearing has taken place or the fluid may behave in a pseudo-solid
manner. Before discussing these different forms of behaviour it is useful to
consider the factors which give rise to these non-newtonian effects.
Non-newtonian behaviour is in general a function of the structural
complexity of a fluid. Liquids such as water, dilute solutions, benzene and
light oils are newtonian. Such liquids are considered to have a loose molecular
structure which is not affected by shearing action. Highly dispersed suspen-
sions of solid or liquid particles may also behave as newtonian fluids if the

210
structure is loose, allowing the shear to take place in the fluid separating the
particles.
The most common non-newtonian fluids are suspensions in which the
suspended particles form a structure which interferes with the shearing of the
suspension medium in a manner dependent upon the shear rate. Examples of
non-newtonian fluids are water-oil emulsions and polymer-thickened oils,
and an extreme example is greases which are a concentrated suspension of
particles in an oil base. Non-newtonian behaviour can be broadly classified
under the following headings.

Plastic or Bingham Fluids

These are fluids which require a significant shear stress before flow or
deformation commences, but once this shear stress has been exceeded the
shear stress is proportional to the shear rate as shown in figure 9.9.
After being deformed a Bingham plastic retains its shape. This may or
may not be a desirable property. In the case of modelling clay such a property
is essential if the model is not to collapse, but in the case of, for example,
paint, it would be very undesirable since it would retain all the brush marks.

Figure 9.9 Shear stress-shear rate

relationship for a plastic or Bingham
fluid

Thixotropy
Thixotropy is a time-dependent loss of consistency when the fluid is sheared.
Ideally, the process is reversible in the sense that the viscosity returns to its
original value when the shear is removed and sufficient recovery time is
allowed, as shown in figure 9.10. This is referred to as a temporary viscosity
loss but in some instances the viscosity never returns to its original value and

211
?'
~
>
c
l"
8.. Low shear rate
!'t Med1um shear rate
H1gh shear rate

Time

Figure 9.10 Thixotropic behaviour

this is a permanent viscosity loss. It is thought that thixotropic fluids have a

structure which is being continuously broken down under the shearing action
and which, at the same time, rebuilds itself. The breakdown of the structure
progresses with time, giving a reduction in apparent viscosity until a stage
is reached where the structure is being rebuilt at the same rate and the
apparent viscosity attains a steady value. Ideally the structure returns to its
original form when the shear is removed, but a permanent loss of viscosity
is experienced when this does not occur.

Pseudo-Plastic Flow
This is a 'thinning' of the fluid as the shear rate is increased as shown in
figure 9.11. Pseudo-plastic fluids are usually composed of long molecules
which are randomly orientated with no connected structure. Application of a
shear stress tends to align the molecules giving a reduction in the apparent
viscosity. In general the process is reversible.

Shear rate

Figure 9.11 Pseudo-plastic behaviour

212
Dilitancy
This is a 'thickening' of the fluid when the shear rate is increased, as shown in
figure 9.12. Fluids which exhibit dilitancy are usually suspensions having
a high solid content and their behaviour can be related to the arrangement
of the particles. In the unstressed state the particles adopt a close-packed
formation which gives a minimum volume of voids, but when the stress is
applied the particles move to an open-packed formation dilating the voids.
This leads to a situation where there is insufficient fluid to fill them and gives
an increased resistance to flow. The same phenomenon can be observed when
walking on wet sand where the pressure of one's foot produces a dry patch .

..

Figure 9.12 Non-newtonian be-

haviour referred to as di/itancy

Elasticity
Elasticity is where the application of a constant stress simply produces a
constant deformation without initiating flow. When the stress is removed
the fluid recovers completely and returns to its undeformed shape. There is
a limiting value of shear stress which can be applied, that is, an elastic limit;
once this is exceeded flow takes place and the elastic effects may be swamped
by the viscous effects. Fluids which exhibit significant elastic behaviour are
referred to as viscoelastic fluids.

9.3 MEASUREMENT OF VISCOSITY

The importance of viscosity is reflected in the number of instruments which

have been devised to measure this property, and a thorough description of
all the different instruments would require a book in itself. However, the
instruments can be conveniently classified into three main groups, according
to the physical measurements employed

213
(i) Instruments in which the viscosity is determined from the flow of
the fluid.
(ii) Instruments in which the viscosity is determined from the motion
of a solid object through the fluid.
(iii) Rotational viscometers, in which the viscosity is determined by
shearing the fluid between two relatively rotating surfaces.
The instruments vary widely in their complexity and applicability. In
general, the simple instruments are only suitable for measuring the viscosity
of fluids in which the non-newtonian effects are small or negligible. When
non-newtonian effects are significant, more complex instruments must be
used and rotational viscometers are best suited for this purpose. The main
types of viscometer and their uses are described briefly in the following
sections. Some of these instruments have facilities built-in for controlling
and measuring the temperature of the test fluid. Where these facilities are
not provided it is usual to locate the viscometer in a temperature-controlled
liquid bath. The majority of viscometers require visual monitoring and the
bath liquid must be transparent. The choice of bath liquid depends upon the
range of temperature over which measurements are to be made. Water is
suitable for the range 0 oc to 90 ac. For higher temperatures clear mineral
oils are usually used and ethyl alcohol or acetone allow measurements to be
made down to -54 oc.

9.3.1 Capillary and Efflux Viscometers

Instruments in which the viscosity is determined from the flow of the fluid
can be subdivided into two types, capillary viscometers and efflux visco-
meters.

Capillary Viscometers
In capillary viscometers the viscosity is determined from the time taken for a
given volume of fluid to flow through a capillary tube. For satisfactory opera-
tion the flow in the capillary must be laminar and the measurements are
based on Poiseuille's Law for steady viscous flow in a pipe which takes the
form
TC d 4
flp t
(9.7)
1'/ = 128/q

where d and I are the diameter and length of the capillary respectively, !lp is
the pressure difference across the capillary and q is the volume of fluid
flowing through the capillary in time t.
Equation 9.7 is valid only for fully developed laminar flow and in some
cases it is necessary to make corrections to allow for 'end' effects. There are
two main corrections which may be made. First there is the kinetic energy

214
correction, which makes allowance for the fact that some of the available
pressure difference is taken up in accelerating the fluid at entry to the capillary,
thus reducing the pressure difference available to overcome the viscous
stresses. Second, the rapid changes in velocity at inlet and exit also influence
the flow and act in such a way that they effectively increase the length of the
pipe. The magnitude of this effect is a function of capillary diameter. When
the two corrections are incorporated equation 9.7 takes the modified form

rrd4 ~p t mpq
(9.8)
1'/ = 128(1 + nd)q - 128rr(l + nd)t

where m and n are empirical constants. m normally lies in the range 1 to 1.5
and the British Standard Specification 188 7 recommends a value of n = 3.
To minimise end effects the capillary tube should have a length-diameter
ratio of 300 or greater and the Reynolds number (Re) for the flow should not
exceed 1000 where

pdv
Re=- (9.9)
1'/

v being the mean velocity of flow, given by

(9.10)

Capillary viscometers may be subdivided into two types, absolute visco-

meters and relative viscometers. With absolute viscometers the viscosity is
calculated from the fundamental equation 9. 7. These viscometers are designed
to minimise end effects and their accuracy is mainly dependent on the
accuracy with which the capillary dimensions can be measured. The best
known absolute viscometer is the Bingham viscometer 8 . The majority of
capillary viscometers in common use are relative viscometers. These are
calibrated with liquids of known viscosity in order to obtain the viscometer
constant and, in general, give greater accuracy than absolute viscometers
since dimensional errors are eliminated. Relative viscometers are usually
gravity operated, in that the pressure difference is provided by the hydro-
static head of the test fluid which is dependent on the density of the fluid.
If ~his the hydrostatic head, ~P = pg ~hand equation 9.7 becomes

or
1t d4 g ~h t
v = = Bt (9 .11)
128/q

215
where B is a constant for a particular viscometer. For low viscosity fluids
where the flow time is short it may be necessary to correct for end effects
and then
v = Bt + Cjt (9.12)
where C is of a similar form to the coefficient of 1/t in equation 9.8 and its
numerical value is normally supplied with the instrument.
The Ostwald U-type viscometer and its modified forms are by far the most
commonly used relative viscometers. The design shown in figure 9.13 is
that adopted by the British Standards Institution and is known as the BS/U.

Capillary tube

Figure 9.13 Ostwald U-tube vis-

cometer

In the right-hand limb the reservoir is situated above a vertical capillary

tube, which is connected to the receiving bulb at a lower level in the left-hand
limb by aU-tube. In use the fluid level is first adjusted so that the meniscus
is at mark A. The fluid is then drawn into the right-hand limb until the
meniscus is about 0.5 em above Band then released. The time taken for the
meniscus to fall from B to C is noted and the viscosity calculated from equa-
tions 9.11 or 9.12.
There are eight recommended sizes of BS/U viscometers which cover a
range from 1-1500 centistokes. Detailed instructions and information are
given in BS 188 7 which also lays down the method of calibration using
freshly distilled water.

216
EJflux Viscometers

With effiux viscometers the viscosity is determined from the time taken for a
given volume of liquid to discharge under gravity through a short tube orifice
in the base of the instrument. There are three types of effiux viscometer in
common use, the Redwood, Saybolt and Engler viscometers. The three
instruments all have the form shown in figure 9.14, and differ only in the
quantity of fluid discharged and the dimensions, particularly the dimensions
of the orifice. The Redwood viscometer 9 is commonly used in the United

A- Contomer for test llqutd

8- Plug
C - Test llqutd level setttng
0-Dtschorge orifice
E- Water Jacket
F- Water agitator
G- Heater

Figure 9.14 Typical arrangement of an ejjlux

viscometer

Kingdom while the Saybolt viscometer is used mainly in Europe and the
ussR. The three types of effiux viscometer are all empirical in that they use
arbitrary scales, the viscosities being quoted in terms of the effiux time, for
example, Redwood seconds or Saybolt seconds. In the case of the Engler
instrument the viscosity is quoted in Engler degrees, this being the time of
effiux divided by that for water at the same temperature. The viscosity
measurements from these instruments can be converted to kinematic viscosity
units at the same temperature and a conversion table is given in table 9.1.
217
TABLE 9.1 VISCOSITY CONVERSION TABLE

Kinematic Saybolt
Viscosity Engler Redwood No. 1 Universal
(centistokes) (degrees) (seconds) (seconds)

2 1.14 31 32
4 1.31 36 39
6 1.48 41 46
8 1.66 46 52
10 1.84 52 59
12 2.02 58 66
14 2.22 65 74
16 2.44 71 81
18 2.65 78 90
20 2.88 86 98
25 3.46 105 120
30 4.08 124 142
35 4.71 144 164
40 5.35 164 187
45 5.99 184 209
50 6.65 204 232
60 7.92 245 279
70 9.24 285 325
80 10.60 326 371
90 11.9 368 419
100 13.2 406 463
150 19.9 620 700
200 26.8 820 940
250 33.0 1010 ll60
300 40.0 1230 1410
400 53.0 1640 1870
500 66.0 2040 2320
600 79.0 2430 2800
700 93.0 2820 3250
800 105 3250 3700
900 118 3650 4200
1000 133 4100 4750
1500 199 6100 7000
2000 260 8100 9200
2500 325 10100 11600
3000 400 12300 14000
4000 530 16100 18500
5000 660 20000 23000

This Table may be used for the approximate conversion of viscosity

units at the same temperature.

218
9.3.2 Falling-Body Viscometers

The most common type of falling-body viscometer is the falling-sphere

viscometer shown in figure 9.15. In this instrument the viscosity is determined
from the terminal velocity of a steel sphere falling through the test fluid.
When the sphere is falling at its terminal velocity v through the fluid it is in
equilibrium under three forces, its weight, the buoyancy force and the drag
force, that is
weight = buoyancy + drag
II Gu•de tube

Gloss tube

Sphere
T1mmg marks

Test ilqu•d

Figure 9.15 Falling-sphere vis-

cometer

The weight and buoyancy forces are readily determined from the size of
sphere and the densities of the sphere and the liquid. For a sphere of radius r
moving at velocity v through an infinite fluid, Stokes derived the expression
drag = 6rtrJrv
Therefore, for equilibrium
1nr 3 p.g = 1nr 3 pg + 6nrJrv
or
2 r 2 (p. - p)g
rJ =----- (9.13)
9 v
where Ps and p are the densities of the sphere and the liquid respectively.
Stokes formula only applies to a sphere moving through an infinite fluid
and it is sometimes necessary to apply a correction factor, F, to equation 9.13
to allow for the finite tube diameter, such that

219
Numerous correction factors have been proposed but a useful one is the
Faxen factor given by

F = 1 - 2.104(d/D) + 2.09(d/D) 3 - 0.9(d/D) 5

where d is the diameter of the sphere and D is the internal diameter of the
tube.
It is normal practice to use a steel sphere and the terminal velocity is
determined by timing the descent over a known distance. It is important
that the ball is released in the centre of the tube and that it does not have any
attached air bubbles. BS 188 7 recommends a maximum ratio of ball-tube
diameter of0.1 and that the ball velocity should not exceed 1 cm/s. Working
within the recommendations, viscosities in the range 10 to 2000 poise can
be measured. For visual timing the test liquid must be clear enough for the
sphere to be visible. If this is not the case the timing may be carried out
electrically, and then of course it is no longer necessary to use a glass tube.
Using a metal tube and a remote method of releasing the ball, the falling-
sphere viscometer can be very easily adapted to measure viscosity at pressures
greater than ambient.
There are a number of other types of falling-body viscometers. The rolling-
sphere viscometer consists of a ball rolling down an inclined tube filled with
test fluid. In this type, the ball diameter is only slightly less than the tube
diameter and the viscosity is a function of the terminal velocity. With very
small clearances the rolling-sphere viscometer has been used to measure
the viscosity of gases and can also be used for high-pressure measurements.
For reliable results these instruments must be calibrated.
The falling coaxial-cylinder viscometer consists of two coaxial cylinders
with their axes vertical. The outer cylinder is clamped and the clearance
space between the two cylinders is filled with the test fluid. In operation the
inner cylinder is released and falls under the action of gravity. The viscosity
is calculated from the expression

(9.14)

where W is the weight of the inner cylinder plus any additional weights, I is
the length of the outer cylinder, Do and D; are the diameters of the outer and
inner cylinders respectively and v is the speed of descent. This type of visco-
meter is particularly useful for measuring high viscosities and has the
facility for varying the shear rate which is useful for non-newtonian fluids.
The band viscometer is similar in operation to the falling-cylinder visco-
meter. A thin band is located in a narrow gap between two parallel shear
blocks as shown in figure 9.16. The gap is filled with the test fluid and when
the band is released it falls under the action of gravity, its motion being oppos-
ed by the shear stresses induced in the test fluid.

220
Bond

~Clearnc
~ filled
wlfh test llqu1d

Figure 9.16 Band viscometer

The viscosity is determined from the speed of descent using the following
equation
Wv
(9.15)
1J = 2bhd
where W is the applied load, vis the velocity of descent, b is the width of the
shear blocks, d is the length of the shear blocks in the direction of sliding and
h is the thickness of one of the shear films. This type of instrument is useful
for non-newtonian measurements and has been used for a wide range of
viscosities, but low viscosity fluids are difficult to retain in the shear gap.

9.3.3 Rotational Viscometers

Rotational viscometers consist essentially of two elements, one of which is

stationary while the other is rotated by an external drive, the space between
the two members being filled with test fluid. The viscosity measurements
are made either by applying a fixed torque and measuring the speed of
rotation, or by driving the rotating element at a constant speed and measuring
the torque required or the reaction torque on the stationary member.
There are two main types of rotational viscometer commonly known as
the rotating-cylinder viscometer and the cone and plate viscometer.
The rotating-cylinder viscometer consists of two concentric cylinders with
the annular clearance filled with test fluid. One of the cylinders, usually the
inner one, is driven while the other is stationary as shown in figure 9.17.
There are a number of commercial rotating-cylinder viscometers available
and these are usually supplied with cylinders of different diameters giving
different radial clearances. These can be driven at varying speeds allowing
a range of viscosities and shear rates to be accommodated. The main shearing
takes place in the annular clearance, but in some cases the shear forces on
the ends of the rotating cylinder may be significant and they must then be
allowed for or taken into account by calibration with a known liquid. The
cone and plate viscometer is shown in figure 9.18 and consists of a plane
surface and a conical surface, either of which may be rotated. The clearance

221
Figure 9.17 Basic arrange-
ment of rotating-cylinder vis-
cometer

Figure 9.18 Basic arrangement of cone and

plate viscometer
space is filled with the test fluid and the included angle of the cone is generally
large to ensure a substantially constant shear rate in the clearance spact-.
The cone and plate arrangement has been employed to study properties
other than viscosity. For example, the Weissenberg rheogonio•neter 10
which is basically a cone and plate viscometer can be used to measure elastic
effects in viscoelastic fluids. When a viscoelastic fluid is sheared in the instru-
ment, normal stresses are generated and the resulting forces tend to separate
the two surfaces. These forces are a measure of the elastic properties of the
fluid.
When using rotational viscometers problems can arise due to shear
heating. The energy dissipated in shearing goes to heating the test fluid
and if the temperature rise is significant, ambiguous results may be obtained.

9.3.4 High-Pressure Viscometers

The general description of viscometers presented in the preceding sections
covers only the types in common use and is by no means a comprehensive
treatment of the subject. Numerous other instruments have also been devised,

222
mainly for investigations which are outside the scope of those described.
Many of these have been devised to measure extremely high or low vis-
cosities which fall outside the range of normal instruments, others have been
specially designed to investigate non-newtonian phenomena and some
instruments have been developed to operate with very small quantities of
test fluid for situations where supplies are limited.
The choice of instrument to be used for measuring the viscosity of a
particular fluid is to some extent based on experience, but for newtonian
fluids the simpler types of instrument such as the capillary, efflux or falling-
sphere viscometers are usually adequate. For non-newtonian fluids the most
common instruments used are the rotational viscometers.
In heavily loaded situations such as occur in gear teeth and rolling bearings
the pressures in the loaded region can be extremely high and a thorough
knowledge of the high-pressure properties of a lubricant are essential.
Viscosity measurement at high-pressure presents many problems and is
usually regarded as a specialist field. The vast majority of viscometers have
been designed for operation at ambient pressure and cannot be conveniently
adapted for use at high pressures. Of the common instruments the falling-
sphere viscometer can be modified for this purpose as described in section
9.3.2. For high-pressure use, very careful design is required to ensure safety
and this usually results in poor accessibility. The necessity for an external
pressurisation system can also make such an instrument very expensive.
Another type of instrument which offers many advantages for high-
pressure viscometry is the disc machine shown diagramatically in figure 9.19.
This type of machine does not require an external pressurisation system
but instead the high pressures are generated internally in an elastohydro-
dynamic film developed between two relatively moving surfaces. In its

.Load

Q
0 coo .. ,o.. """"'
,to heav:ly loaded
contact regH>n

+ "'
t
Figure 9.19 Basic ar-
rangement of a disc
machine

223
simplest form the disc machine consists essentially of two discs with in-
dependent drives. In operation the discs are set in motion, the test fluid is
introduced between the discs in the region of minimum clearance and the
discs are then loaded together and separated only by a thin elastohydro-
dynamic film. By driving the discs at different speeds and measuring the
film thickness and the viscous torque on the discs, the viscosity can be
determined. Sophisticated measuring techniques are required for the disc
machine and its use as a viscometer is still being perfected, but it is a very
versatile instrument which can be used for non-newtonian measurements.

9.4 LUBRICATING OILS

Originally, lubricating oils were derived solely from animal and vegetable
sources, but this changed with the discovery of petroleum oils and now the
great majority of lubricating oils come from petroleum, that is, mineral or
hydrocarbon oils. One of the reasons for this change is that animal and
vegetable oils are more chemically reactive than mineral oils and deteriorate
mor~ rapidly in use. However, animal and vegetable oils are very 'oily' and
have good emulsifying properties and are still used in certain applications
where these properties are desirable, for example textile oils, compressor
oils and metal-cutting oils.
Mineral oils are complex substances containing a wide range of mineral
hydrocarbons and it is impossible to assess their performance simply from
chemical tests. Consequently, oils must be submitted to many tests to measure
their physical characteristics. For comparison purposes the tests must
produce reproducible results and therefore it is essential that they are carried
out under specified conditions. In general, the test methods applied are those
laid down either by the Institute of Petroleum 9 in the United Kingdom or by
the American Society for Testing and Materials in the USA 11 . These are
usually referred to as the IP or ASTM methods. The tests laid down cover
a wide range of lubricating oil properties including such features as rate of
deterioration, inflammability, compatibility and suitability for use under
extreme conditions and some of the main features are described below.

9.4.1 Properties of Mineral Oils

Relative Density
The relative density of a substance is the density of that substance divided
by the density of water at the same temperature and pressure. In the petroleum
industry the relative density of an oil is determined at 60 oF or may be
corrected to that temperature using standard tables or formulae; most
mineral oils have relative densities in the range 0.85 to 0.95. The density of a
224
fluid is required for flow rate calculations and for the conversion of kinematic
viscosity to dynamic viscosity and the density of an oil in g/cm 3 is very
nearly numerically equal to the relative density.

Specific Heat and TheriiUll Conductivity

The specific heat and thermal conductivity of an oil are important parameters
when the oil acts as a coolant or heat transfer medium. Most mineral oils have
specific heats in the range 0.44 to 0.48 and thermal conductivity of approx-
imately 3 x 10- 4 calfcm soc. Neither of these quantities is very sensitive to
temperature changes.

Acidity and Alkalinity

Even well-refined oils contain traces of weak organic acids, but in general
these are harmless. More important are the acids which are formed in use by
combustion or oxidation and acids introduced by contamination. The
acidity of mineral oils is expressed as a neutralisation number, this being the
weight of potassium hydroxide in milligrams required to neutralise one gram
of oil.
In some lubricants alkalinity is introduced to give them special properties.
Most modern engine oils are alkaline to allow them to neutralise fuel
combustion products. In some soluble cutting oils alkalinity is introduced to
protect ferrous metals from corrosion and to stabilise the emulsion. Alkalinity
is expressed as a base number which is the number of milligrams of potassium
hydroxide which are equivalent to the alkali present in one gram of oil.
(Tests IP 1 and IP 139)

Oxidation Stability
Mineral oils are not very reactive in the chemical sense, but they can oxidise
when exposed to oxygen or air at elevated temperatures and this has a
strong influence on the life of an oil. The rate of oxidation is a function of
temperature and the operating conditions. The products of oxidation vary
but in general they consist of acidic compounds, sludge and lacquers, which
may cause the oil to become corrosive, to increase in viscosity and to deposit
insoluble products on working surfaces. (Tests IP 48, 56, 114, 148, 157)

Flashpoint
Flashpoint may be defined as the lowest temperature at which the vapour
given off by an oil, when heated in a standard piece of apparatus, ignites
momentarily on the application of a flame. The flash point is a measure of the
fire hazard and is also useful in determining whether an oil has been contam-
inated. (Tests IP 34, 35 and BS 2839)

225
Foaming
Excessive foaming of a lubricant due to churning can lead to inadequate
lubrication and other problems. This is particularly true for crank-case oils
and hydraulic oils since the pump cannot deal efficiently with the frothy
mixture. (Test IP 146)

Pour Point

The pour point is the temperature at which oil ceases to flow freely under
specified conditions and this imposes a lower limit on the oil working
temperature. (Test IP 15)

Demulsibility

Mixtures of pure mineral oil and pure water separate out quite quickly but
if the mixture is contaminated separation takes much longer and a stable
emulsion may be formed. In many industrial applications oils come in
contact with steam and water and it is essential that the oil should separate
rapidly, allowing it to be recirculated. The most common test is to pass
saturated steam into a measured volume of oil under standard conditions
and to observe the time taken for the resulting emulsion to separate into
oil and water layers.
Some industrial oils are deliberately manufactured so that they will form
stable emulsion. The so-called soluble oils used in metal cutting are an
example. (Test IP 19)

Extreme-Pressure Properties
A number of machines are used for evaluating and comparing lubricants
under extreme-pressure conditions. All these machines involve heavily
loaded regions where sliding or combined sliding and rolling take place
and the parameters measured include friction, wear and seizure load. Due to
the differences in the various machines, correlation of results is poor and
lubricants can really only be compared using the same machine. A commonly
used machine is the four-ball tester shown in figure 9.20.

9.4.2 Additives and Solid Lubricants

Most modern lubricating oils have chemical compounds added to them to

improve the characteristics of the straight mineral oil. Additives are included
for many purposes and some of the main ones are described below.

226
'Load

Retoimng cup

Figure 9.20 Basic arrangement of four-ball

testing machine

Viscosity-index Impro11ers
These are added to an oil to reduce the rate of change of viscosity with
temperature and are usually high molecular weight polymers. There is a
tendency for the polymers to be broken down when subjected to high rates
of shear which will give a general reduction in both viscosity index and
viscosity. Some viscosity-index improvers also act as pour-point depressants.

Pour-point Depressants
When mineral oils are cooled, waxy crystals are precipitated, eventually
combining to form a rigid structure and this process determines the pour
point. Pour-point depressants delay the formation of a rigid structure by
forming a coating on the waxy crystals. Complex polymers and other types
of products are used as pour-point depressants.

Oxidation Inhibitors
Oxidation inhibitors prevent or reduce the rate of formation of oxidation
products such as acidic products and insoluble compounds. By implication,
oxidation inhibitors also act as corrosion inhibitors. Different types of
oxidation inhibitors are used, depending on whether they are required for
high or low temperature applications.

Detergents or Dispersants
These are used mainly in engine oils to keep the engine clean by holding
insoluble material in suspension and preventing the formation of sludge or
deposits. For high temperature conditions organo-metallic salts are the
usual type of additive, while polymer compounds are mostly used for low
temperature conditions.
227
Extreme Pressure Additives
Under conditions of extremely high pressures and temperatures straight
mineral oils are not capable of maintaining a lubricant film between surfaces.
These conditions are frequently met with in gears and extreme pressure addi-
tives allow a mineral oil to operate satisfactorily. The most common extreme
pressure additives are compounds of sulphur, chlorine and phosphorus
and these react with metal surfaces to form coatings which exhibit plastic
flow and have low shear resistance.

Cor.~in Inhibitors
These are alcohols, esters, amines, organic acids or soaps which are adsorbed
on, or react with, metal surfaces to form protective films.

Anti-foam Additives
These are usually silicone based compounds which minimise foaming in
situations where this might present problems.

Emulsifiers
These are used to stabilise oil-water emulsions and are particularly useful for
metal-cutting oils, fire resistant hydraulic fluids and similar fluids. Typical
emulsifiers are petroleum sulphonates and metallic soaps.

Solid Lubricants
These materials are lubricants in their own right and are used as dispersions
in mineral oils for special applications. At high temperatures they are advan-
tageous in that they are non-volatile and can provide a lubricant film when
the base oil has disappeared. Graphite, molybdenum disulphide and PTFE
are examples of solid lubricants (see chapter 6).

9.5 GREASES

A grease is a stabilised mixture of a liquid lubricant and a thickening agent

and may include additives to improve or impart particular properties. In the
majority of greases the liquid lubricant is a mineral oil but synthetic lubri-
cants such as silicone are sometimes used, particularly in greases which have
to operate over a wide temperature range such as in aircraft applications. An
indication of the typical working temperature range for different liquid bases
is given in table 9.2, but the range may be extended by the use of additives.

228
TABLE 9.2 USEFUL TEMPERATURE
RANGE OF DIFFERENT GREASES

Useful
temperature
Liquid base range

Mineral oils -30 OC-100 oc

Silicone -40 °C-150 oc
Diester and polyester -60 "C- 90 "C

The thickening agents include metallic soaps, clays and silica. The main
metallic soap thickeners are calcium, lithium and sodium soaps in the form of
fibres. In general there is no lower temperature limit for the soaps, but at
higher temperatures structural changes may occur, and since this temperature
is usually below the upper limit for the liquid component, it imposes an
upper temperature limit on the grease. With clay and silica thickeners there
appears to be no practical lower or upper limit to their use and they can
certainly be used up to several hundred degrees celcius.
The structure of greases is such that they are normally self-supporting in
the static state and this property allows greases to act both as a lubricant and
a seal against contamination. When sheared, greases behave as a viscous
fluid but are very non-newtonian and their properties at any one time are
strongly dependent on their previous usage. Many standard tests for greases
are laid down by both the Institute of Petroleum and the American Society
for Testing and Materials and some of the main ones are described below.

Consistency

In this test a standard cone is allowed to penetrate a sample of grease main-

tained at 25 oc. The consistency is measured in terms of the depth of pene-
tration (in tenths of a millimetre) after a period of five seconds. Usually, two
tests are made. The first test is made on unworked grease, that is, grease
straight from a container, but these results tend to be unreliable, since some
working occurs when the sample is removed from the container. A more
reliable second test is made on the grease after it has been subjected to a
period of working under standard conditions. (Tests IP 50, 167, ASTM.
0.217)
The National Lubricating Grease Institute (NLGI) of America has class-
ified greases in grades depending on their consistency and these gradings,
shown in table 9.3, are widely used. Consistency gives only a broad indication
of performance. Clearly a soft grease, say grade l, will set up a lower resis-
tance-to-motion in a bearing than a hard grade 5 grease, but two greases of
the same grade may give completely different bearing performance.

229
TABLE 9.3 NLGI CLASSIFICATION OF
GREASES

Worked penetration range.

(Tenths of a millimetre
Grade number at 25 oq

()()() 445-475
{)() 400-430
0 355-385
I 310-340
2 265-295
3 220-250
4 175-205
5 130-160
6 85-115

Oxidation Stability
Resistance to oxidation is assessed by oxidising thin layers of grease in a
bomb filled with oxygen under pressure and heated. The degree of oxidation
is determined from the fall in pressure over a given time. The test has some
correlation with the storage life of greases but is not reliable for predicting
service life. (Tests IP 142, ASTM. 0.942)

Drop Point
The drop point is the temperature at which a drop falls from the orifice of a
standard cup and this gives an indication of the transition from solid to
fluid state. In practice, drop points are usually considerably higher than
normal maximum service temperatures. (Tests IP 31, 132, ASTM. 0.2265,
BS 894)

Oil Separation

There is a natural tendency for oil to separate out from greases and this is
often referred to as 'bleeding'. When grease is stored in a container it is quite
common to see some free oil on the surface or at the bottom, but the basic
structure of the grease is not changed if the amount of free oil is small. Oil
separation tests are carried out by measuring the loss of oil from a sample of
grease, supported on a gauze or filter paper, when acted on by gravitational
or centrifugal forces. Such tests give an indication of the stability of a grease
but seem to have little relevance to the service of grease in bearings. (Tests
IP 121, ASTM. 0.1742)

230
Mechanical Stability
When greases are subjected to severe mechanical working, such as in rolling
bearings, their consistency may change considerably. To evaluate this char-
acteristic prolonged worked penetration tests may be carried out.

Extreme-Pressure Properties
For heavy duty service a grease must have good EP properties to protect
against frictional damage to the bearing surfaces. The EP properties of greases
are normally evaluated using the Timken Wear and Lubricants Tester or a
four-ball machine (see figure 9.20). (Tests ASTM. 0.2266)

Bearing Performance Tests

The foregoing simple laboratory tests do not provide reliable information on
the performance of a grease in service, aqd for this purpose it is necessary to
build test rigs in which test bearings are run under suitable conditions.
Leading rolling-bearing manufacturers have developed their own perform-
ance tests for evaluating greases and these involve a wide variety of test
bearings and conditions. A standard form of bearing rig test which has more
general application is also used. (Test IP. 168)

Grease Additives
Additives are often included in greases to impart special qualities. Examples
are corrosion inhibitors, oxidation inhibitors and EP additives. Solid lub-
ricants such as graphite or molybdenum disulphide are also used to reduce
wear and prevent seizure.

REFERENCES

I. Introductory memorandum on the viscosity of liquids. Engineering Sciences Data

Item No. 65001, Instn mech. Engrs.
2. Approximate data on the viscosity of some common liquids. Engineering Sciences
Data Item No. 66024, Instn mech. Engrs.
3. A guide to the viscosity of liquid petroleum products. Engineering Sciences Data
Item No. 67015, Instn mech. Engrs.
4. E. W. Dean and G. H. B. Davis. Viscosity variations of oils with temperature.
Chern. metal/. Engng, 36, (1929), 618-19.
5. Charts Published by American Society for Testing and Materials.
6. Refutas Chart available from Baird and Tatlock, London.
7. Determination of the viscosity of liquids in CGS units. BS 188, (1957), British
Standards Institution.

231
8. E. C. Bingham. Fluidity and Plasticity. McGraw-Hill, New York, (1922).
9. J.P. Standards for Petroleum and its Products. Institute of Petroleum, London.
(Annually).
10. K. Weissenberg. Principles of Rheological Measurement. Nelson, London, (1949).
II. ASTM. Standards on Petroleum Products and Lubricants. American Society for
Testing and Materials. Philadelphia. (Annually).

232
10
Hydrodynamic
Lubrication
10.1 INTRODUCTION
The introduction of a film of fluid between components with relative motion
forms the solution of a vast number of tribo10gical problems in engineering.
In chapter 9 we have seen how lubricants may be supplied to the contact at
high pressure. However, in many cases the viscosity of the fluid and the geo-
metry and relative motion of the surfaces, may be used to generate sufficient
pressure to prevent solid contact without any external pumping agency. If
the bearing is of a convergent shape in the direction of motion, the fluid
adhering to the moving surface will be dragged into the narrowing clearance
space, thus building-up a pressure sufficient to carry the load. This is the
principle of hydrodynamic lubrication, a mechanism which is essential to the
efficient functioning of the whole of modern industry. Motor vehicles, loco-
motives, machine tools, engines of all types, domestic appliances, aircraft,
surface and underwater vessels, gearboxes, pumps and spacecraft are only a
small part of an almost endless list of equipment and machines which rely
heavily on hydrodynamic films for their operation. Although usually so
beneficial, hydrodynamic films sometimes occur in situations where they are
undesirable or even dangerous. For example, care has to be taken to prevent
the formation of such a film of water between the pantograph of an electric
locomotive and the conductor in wet weather, and the tread pattern of a
motor tyre is an attempt to prevent 'aquaplaning', which is the build-up of a
hydrodynamic film between the tyre and the road, resulting in a loss of grip.

233
10.1.1 Notation
B Width of bearing
C Radial clearance of journal bearing
e Eccentricity
F Friction force per unit width
f Coefficient of friction
H1 hjh0
h Film thickness
hi Inlet film thickness
hm Reference film thickness ( = h0 for inclinded slider, = hi - h0 for
Rayleigh step, = C for journal bearing)
h0 Outlet film thickness for sliders
Minimum film thickness for rollers
Ji h/hm
h* Film thickness at point of maximum pressure
K Load capacity factor for inclined slider
log.(l + n) 2
n2 n(2 + n)
L Length of bearing
L1 L 2 Length of inlet and outlet sections of Rayleigh step bearing
n ~- 1
ho
N Rotational speed
p Pressure (gauge for liquid bearings; absolute for gas bearings)
Pa Atmospheric pressure
Ps Pressure at step of Rayleigh bearing
p PIP.
ph2
p'
617UL
q Reduced pressure ( = ~ (1 - e- ~P)
R Radius
s Sommerfeld number ( = ~ LD(~Y)
Inlet temperature
Outlet temperature
Surface velocity in tangential direction
u Fluid velocity in tangential direction
v Surface velocity in normal direction
w Load
X Coordinate along bearing
.X x/L
y Coordinate in axial direction
z Coordinate across film

234
Viscosity-pressure coefficient
Viscosity-temperature coefficient
Eccentricity ratio ( = ~)
Yf Absolute viscosity
(J Angle
A
.
Beanng number
(= l2Yf
(u I + u 2) ~ L )
2
mPa
p Density
Pi Density at inlet
Po Density at outlet
p'
Po
Pi
Shear stress
Attitude angle

10.2 THEORY
The variation of lubricant pressure in the bearing is described by the
Reynolds equation, which is derived and fully presented in the appendix. The
solution of the full equation involves great difficulties, but we will use a much
simplified version in the first instance, by making the following assumptions:
(1) The fluid is incompressible.
(2) The fluid is newtonian (the shear stress is directly proportional to the
shear strain rate).
(3) The fluid properties remain constant; effects due to variation in temp-
erature and pressure being neglected.
(4) Inertia and turbulence effects are negligible.
(5) The solid bodies remain rigid.
(6) The film is of sufficiently small thickness that the fluid pressure can be
considered constant through the thickness of the film, (but not of course
along its length).
(7) The bearing is infinitely wide.
It will be shown later how some of these assumptions may be relaxed to
adequately describe certain situations, although there are many cases where
the errors induced by these assumptions are negligible.

10.2.1 Longitudinal Motion (Sliding)

In the situation where one surface slides over another (in the x direction)
with no normal motion and a fluid between them, if the above assumptions
are made the Reynolds equation reduces to
dp
dx
=
12"
(u, + u2)(~
2 h3
(10.1)

235
where h* is the film thickness at the position of maximum pressure where
dp/dx = 0. U 1 and U 2 are the velocities in the x direction of the lower and
upper surfaces, respectively, and must be arranged such that the film thickness
h is invariant with time.
Consider two flat plates having velocities Ua and Ub as shown in figure
lO.la. If the system is resolved into its component parts a and b, we can see that
a is simply a rigid body motion of both surfaces and the fluid at a uniform

Col~

<<'
""' -<-\ U a
<' --u.
\\\\\\\\\\

A
\\\\\\\\\\
- - U0 -U0

(b)

0
0
(c)

l V2

v//////////1

~X

Figure 10.1

velocity U b, while B represents a system where h is invariant with time, to

which equation 10.1 may be applied putting U 1 = Ua - Ub and U 2 = 0.
Thus the Reynolds equation becomes for two sliding surfaces
dp
(10.2)
dx

236
Any situation with longitudinal velocities may be dealt with in the same way.
For example, two discs rolling on fixed axes with circumferential velocities
U a and U b, as in figure 10.1 b, produce a Reynolds equation

(10.3)

since in this case h is clearly invariant with time and equation 10.1 applies.

10.2.2 Normal Motion (Squeeze)

In this case there •is no relative sliding of the surfaces ( U 1 = U 2 = 0), but
there is movement normal to the surfaces. Consider the two parallel fiat
plates in figure 10.1 c with respective normal velocities V 1 and V 2 . Common-
sense tells us that a pressure will be developed in the fluid if V 1 - V 2 is
positive, and that the fluid will flow outwards from the point of maximum
pressure. The Reynolds equation confirms this, since making the same as-
sumptions as before, it becomes

(10.4)

where x* is the coordinate of the position of maximum pressure (dpjdx = 0).

This situation is often called 'squeeze film lubrication'.

10.2.3 Combined Longitudinal and Normal Motion

The build-up of pressure in a bearing where both types of relative motion are
present can be found by a simple summation of the two effects, thus

dp
dx
= 1211 (U 1 +
2
U 2 )(h- h*) 12 (V _ V)(x- x*)
hJ + 11 2 1 hJ (10.5)

To find the actual pressure distribution it is necessary to integrate the

equation. Two unknown quantities will then be present, the integration
constant and the value of x*. These are determined by the incorporation of
two relevant boundary conditions. The above equation may be applied to any
pair of surfaces, such as those shown in figure 10.2, provided that the appro-
priate velocity components are resolved to obtain the appropriate values of
U 1 , U 2 , V 1 and V 2 •

237
+

Figure 10.2

10.3 APPLICATION OF REYNOLDS EQUATION TO SLIDING

BEARINGS

We have discussed hydrodynamic action on the basis of a convergent

clearance space through the length of the bearing, figure 10.3 illustrating some
of the shapes which satisfy this condition for successful operation. All of these
forms and many others occur in practice, either because they are manufac-
tured in that way, or due to subsequent wear or deformation. However,
fortunately for the bearing designer, in his analysis of many different slider
profiles published in 1918, Lord Rayleigh concluded that there is little to

\
~ \'~ \ \'~
~ I '.,\ \\
~
\ \1..:, I'.\\"
(a) (b) (c)

~ ~ ~
\\ :1' ' .:.. \\\ \ \1~ \~
(d) (e) (f)

Figure 10.3 Slider bearing shapes

238
choose between the various forms of bearing, given the same inlet and outlet
film thicknesses. The slider of figure 10.3c, often called the Rayleigh step,
actually produces the highest peak pressure 1•

10.3.1 Plane-Inclined Slider

By far the most common form of lubricated slider bearing is the plane-
inclined pad illustrated in figure 10.3a. As an example of the application of
the Reynolds equation to slider bearings, we will determine the pressure
distribution and load capacity for such a configuration. The bearing and
notation are illustrated in figure 10.4.

Figure 10.4 Slider bearing notation

We can use equation 10.1 since there is no normal motion present, putting
V 1 = - u, V 2 = o.

This equation must be integrated with respect to x to yield the pressure

distribution

p = -617U J(h ~ 3 h*) dx

The film thickness can be expressed at any point as

h = ho (1 + nx)
L

239
where

where A is a constant of integration. It will be noted that h* is the value of h

where dpjdx = 0, that is, where the pressure has a maximum value. We have
two unknowns h* and A, which must be found by the introduction of two
boundary conditions: p = 0 at x = 0, and x = L. Note that pressures are
expressed as gauge pressures, that is, p = 0 represents ambient pressure.
Substitution of these two conditions gives

0 = - 6f/ U [ - -L
n
+ -h* -L + A
2h 0 n
J
and

0
= -
6
flU 1
[-L/n h*
+ n + 2h 0 (1
L/n
+ n) 2 + A
J
The solution of these two simultaneous equations yields

h* = 2h (1 + n)
o 2 +n
and
L
A=---
n(2 + n)
These can now be substituted in the pressure distribution equation to give

6f/ u L [ n I( I) ]1-
( 10.6)
p=~ (2+n{1+n:y
A further integration of the pressure gives the normal load capacity of the
bearing per unit width
x=L

W = J pdx ( 10.7)
x=o

240
For the inclined slider the load capacity is given by

W = 6'7UL2 [log.(l + n) _
h; n2
2
n(2 + n)
J (10.8)

or

where
K = log.(l + n) _ 2
n2 n(2 + n)
The maximum load capacity depends on the value of K, which in turn
depends on the inlet-outlet film thickness ratio. By putting dKjdn = 0
we can find the value of n for optimum load capacity. This occurs when n is
approximately 1.2, or the inlet film thickness is 2.2 times that at the outlet. In
this condition K = 0.0267. The load capacity is, however, very insensitive
to n-if n is reduced to 0.6 the value of K only falls to 0.0235, while if n is
increased to 2.0, K is only reduced to 0.0246.
Similar values of this optimum ratio apply to the other slider shapes.
When we remember that in the great majority of engineering applications, the
minimum film thickness will be in the order of 0.02 mm to 1 mm, it will be
appreciated that the angles, curvatures, steps, etc., which transform a rigid
surface into an efficient bearing are extremely small and often present manu-
facturing problems. This accounts for the popularity of the pivoted pad
bearing, which is free to take up its own angle to the other surface. This
bearing type and some of its derivatives are illustrated in figure 10.5.
If the pressure distribution for a plane-inclined slider is plotted, a curve

---
\\\\\\\\\\\\\\ \\\\\\\\\
\ \ \ \ \,...._._

\\\\\\\\\\\\ \\\\\\\\\\\\\\\
=-- =-----
Figure 10.5 Se!Fstabilising sliders

241
such as that of figure 10.6 is obtained. Clearly the centre of pressure lies to the
rear of the centre of the pad. Indeed, we know that

h* = 2h 0 (l + n)
2+n
If we substitute h* = h0 [l + (nx* /L)] we obtain an expression for the posi-
tion of maximum pressure
x*
L 2+n
x* I L varies from !- when n = 0 to 0 as n -+ oo. In other words, x* always
lies in the rear half of the pad. For equilibrium of the pivoted pad, the centre
of pressure must coincide with the pivot position. It is possible, for a given
set of conditions, to calculate exactly the correct position of the pivot, but
this is usually a futile exercise, since we do not have complete control over the

Figure 10.6

conditions in which the bearing operates. Moreover, a pad pivoted centrally

to allow motion in either direction functions quite satisfactorily, although
theory indicates that it should never generate a film. The explanation for this
is that the increases in pressure and temperature in the film distort the pad
to a convex curved shape, which is then self-stabilising about the centre
pivot.

10.3.2 The Rayleigh Step Bearing

It can be shown, using the calculus of variations, that it is the step bearing that
has the greatest load capacity of all the slider shapes. The bearing and its
notation are shown in figure 10.7. Equation 10.1 can be applied to each
section of the bearing in turn, U 1 = - U, U 2 = 0. After integration the
equation becomes

P= -6~u(h h3 h*)x +A

since (h - h*)jh3 is constant for each section.

We shall see when the pressure profile is produced for the step bearing in
figure 10.7 that, since the curve is discontinuous at the location of the maxi-

242
L,

-. u

Figure 10.7 Step bearing notation

mum pressure, h* must be simply regarded as a constant of integration,

having no physical significance, that is, h* =F h;, h* =F h0 •
The boundary conditions in this case are p = 0 at x = 0 and p = P. at
x = L 2 for the region where h = ho; p = P. at x = L 2 and p = 0 at x = L
for the region where h = h;. The resulting pressure distribution is shown in
figure 10. 7, the maximum pressure occurring at the step.
The load capacity is given by
L
W =p s X-
2
which results in the following expression when the value of p., obtained by
substituting the boundary conditions into the expression for p, is inserted
31JU LL 2 (L - L 2 )(a - l)
w = 3 2 (10.9)
(L 2 a + L- L 2 )ho
where a = h;/h 0 • For optimum performance a = 1.866 and L 1/L 2 = 2.549.
With these optimum parameters
617UU
w = 1;2 = 0.0342
0

which is an improvement on the load capacity of the inclined slider (0.0342

as compared with 0.0267).

10.4 CONTACTS IN THE FORM OF NON-CONFORMING

DISCS

Many engineering situations may be represented with some accuracy by two

cylinders or a cylinder and a plane. We immediately think of rolling element
bearings, the deformed pivoted pad and many others, including gear teeth
contacts which operate by a combined rolling and sliding motion.

243
The Reynolds equation simplified by the assumptions of section 10.2 may
be used to predict the film pressures when one or both the elements have the
form of non-conforming cylinders rolling and/or sliding over each other, as
illustrated in figure 10.8.

(b)

Figure 10.8 Equivalent cylinder

10.4.1 The Equivalent Cylinder

The cylinder and plane shown in figure 10.8b are more easily analysed than
the two-cylinder situation of figure 10.8a. However, the two cylinders of
radius R 1 and R 2 may be represented by an equivalent cylinder of radius R
and a plane, where
R 1R 2
R = ---=----=---
Rl + R2
and the surface velocities of the two components and the film thickness
variation in case a are preserved in case b. The two situations may be con-
sidered identical for the purposes of calcuation, except for the calculation of
horizontal components of pressure, as we shall see later in this chapter.

10.4.2 Cavitation

The Reynolds equation 10.1 may be used to predict the performance of the
bearing by the insertion of values for U 1 and U 2 and the film thickness in
terms of the distance through the bearing
h = h0 + R( 1 - COS (}) (10.10)
(See figure 10.9 for illustration and notation.)
It will be immediately obvious from both the figure and equation 10.10
that in this case, unlike the sliders previously dealt with, the convergent
portion of the bearing is followed by an equally long divergent portion
before ambient pressure is once aga.in achieved. If we apply our boundary
conditions as for the sliders, namely p = 0 at (} = - n/2 and + n/2, then a
pressure distribution results which is antisymmetrical about the horizontal
axis, indicating zero load capacity, (curve a of figure 10.9). This implies that

244
u,

Figure 10.9

there exist in the liquid, negative pressures of the same magnitude as the
positive pressures. In the vast majority of cases the liquid is unable to sustain
these negative pressures and hence, in this region, the film will rupture and
the air, which is almost inevitably present dissolved in the oil, will appear as
bubbles. This phenomenon, known as cavitation, and well-known to anyone
with experience of bearing operation, results in the majority of the clearance
space being composed of alternate fingers of liquid and air across the width
of the bearing. This phenomenon has received considerable attention,
especially by Floberg 2 and Dowson\ and for the student of hydrodynamic
design has two consequences. If the pressure falls to the vapour pressure of
the lubricant and low pressure air bubbles are allowed to form due to boiling
of the liquid, their subsequent collapse against solid surfaces can create such
enormous local pressures that damage in the form of pitting (cavitation eros-
ion) can result. This form of caviton~ although all too familiar to the hydrau-
lics engineer, does not often occur in bearing practice. The other consequence
is that it is no longer adequate to use the simple boundary condition that
p = 0 when (} = n/2 at the outlet from the film. Experimental measurement
indicates the pressure in the fluid has the form illustrated in figure 10.9
curve b, falling to the saturation pressure of the liquid at some point A in
the divergent zone. As the liquid is usually drawn from an ambient pressure
source, the saturation pressure may be regarded as ambient (p = 0).
The representation in analytical terms of what happens at this boundary
245
has been the subject of some discussion and controversy. However, there is
now almost universal acceptance of one form of this boundary condition.
To the left of A (figure 10.9) the volume rate of flow of fluid has two com-
ponents: [( U 1 + U 2 )/2]h due to the motion of the surfaces and an additional
flow due to the pressure gradient dpfdx. To the right of A the flow is simply
[( U 1 + U 2 )/2]h. Therefore for continuity at A dpfdx = 0 at A. This leads
us to the Reynolds cavitation boundary condition, which agrees well with
experience and states that the pressure curve terminates with zero gradient
at some unknown position in the divergent part of the film.
dp
p = - = 0 at () = () 2 (10.11)
dx

It has been shown by Dowson 3 that this condition can in some cases be
complicated by the existence of small subambient negative pressures for a
small distance before the pressure becomes ambient. The effect of such areas
is, however, very small and may be neglected.

10.4.3 Load Capacity

Due mainly to the more complex boundary condition at the outlet when the
liquid cavitates, the solution of even our simplified Reynolds equation to
give the force components on the cylinders is no longer as simple as in the
previous analyses. Martin 4 and Purday 5 performed the analysis for a cylinder
of parabolic shape, having a film thickness given by
x2
h = ho + 2R
The film shape for such a cylinder differs significantly from the circular shape
only in regions of large film thickness which contribute little to the perform-
ance. Floberg in 1961 presented a solution for a circular cylinder, showing
that the normal load capacity is given by
R
w= cx:(U 1 + u 2)'1 h (10.12)
0

The difference between Floberg's result and that for a parabolic cylinder with
the same ho and R is negligible. The value of ex: depends upon the ratio Rjh 0
as shown in table 10.1. Also shown in the table are values of(J 2 , the position
of the cavitation boundary. The values of ex: given are from calculations of the
normal force on the plane surface, the value of ex: for the disc being different
only at very low values of Rfho anc~ then by only a small amount. From the
table it can be seen that, for most practical cases, ex: may be given the approxi-
mate value of2.45. We can also see that as Rfho increases, the cavitation boun-
dary approaches the position of minimum film thickness.

246
TABLE 10.1

Rjh 0 C( oo2
10 1.399 11.31
102 2.215 3.82
103 2.411 1.22
104 2.442 0.385
105 2.447 0.122
106 2.451 0.039

10.5 THE JOURNAL BEARING

We come now to the most familiar and widely used form of hydrodynamic
bearing-the journal bearing. A rotating shaft is supported in a bush which
completely or partially surrounds it with a small clearance as in figure
10.1 Oa. If a load is applied to the journal it will be displaced from the centre,
thus forming, as it rotates, a convergent clearance space which is conducive
to the building-up of a lubricating film to support the load.
As the pressurised film is created, the journal moves round the bearing in
the same sense as the rotation until it reaches an equilibrium position as
shown in figure l0.10b. Note that the line of centres of the journal and the
bush does not coincide with the line of action of the load, but is displaced
by an angle qy. The distance between the centres, the eccentricity e, divided
by the radial clearance of the bearing C, is called the eccentricity ratio c.
Obviously c = 0 represents concentricity, and c = 1 represents contact
between the two surfaces. If the bearing is unwrapped, the form of the
clearance can be clearly seen, as in figure 10.1 Oc.
The film thickness at any point may be written as
h = C( 1 + c cos 0) (10.13)
providing CjR ~ 1. This is almost invariably the case, typical values of C/R
ranging from 0.0004 to 0.004. The film thickness may be substituted into our
Reynolds equation 10.1, together with U 1 = U, U 2 = 0 and x = RO
dp
dO =
617UR[
----cz (1
1 h*
+ c cos 0) 2 + C(1 + c cos 0) 3
J (10.14)

The solution of this equation was first achieved by Sommerfeld 6 by means

of the Sommerfeld transformation
1 - c2
1 + c cos 0 = 1
- n cosy
but the resulting equation for p does, of course, contain two unknown
constants, which require evaluation by the substitution of two boundary
conditions.
247
(a)

(b)

(c)

Figure 10.10 (a) Journal bearing,

(b) journal bearing notation, (c)
development of journal bearing
clearance

10.5.1 Boundary Conditions

Unlike the slider bearings we considered previously, where we usually know

that the pressures at the inlet and outlet of the film are ambient, the journal
bearing, being a cyclic device, has a continuous film of fluid round its circum-
ference. Of course in the bearing of finite length, the lubricant can flow out

248
and in from the bearing sides, but for the infinitely long journal bearing pres-
ently being considered this is precluded. Therefore the condition which must
be satisfied is that
Pe = Pe+z•
This is not much help for analysis of the bearing and we need to know the
pressure at some point. In practice this datum pressure is often the pressure
of the lubricant supply where it is introduced into the clearance of the bearing.
This supply may or may not be close to ambient pressure and, although
always located in the lower pressure area of the bearing, it need not coincide
with f)= 0.
Sommerfeld assumed a pressure Po at f) = 0 and f) = 27t. This produces
the so-called 'full Sommerfeld solution', with a pressure distribution given
by equation 10.15 and illustrated in figure 10.11.
611URe sin 0(2 + e cos 0)
(10.15)
P - Po = C 2 (2+ e2)(1 + e cos 0) 2
As with the discs of section 10.4.2, because there is a divergent clearance
present, impossibly large negative pressures can be predicted by these
boundary conditions. In a few cases it is possible that these negative pressures
are eliminated by a sufficiently high p0 , but in general the predicted distri-
bution has the form shown in figure I 0.11, with the negative shaded portion.

Figure 10.11

Unlike the case of the discs, the inclusion of the negative pressure, even with
Po = 0, does not imply zero load capacity, since the positive and negative
pressure areas occur in opposite halves of the bearing, and therefore both tend
to move the journal in the same direction. Since the negative pressures
would assist the bearing to support the load, the full Sommerfeld solution
will overestimate the load capacity.
We have seen in section 10.4.2 that it is closer to practical experience to
249
dismiss these large negative pressures, and to use the Reynolds cavitation
boundary condition that
dp
p = dO = 0 at 0 = 02 1t < e < 21t
The inclusion of this condition increases the difficulty in solving the Reynolds
equation 10.14, but the pressure distribution that is finally achieved is
shown in figure 10.11.

10.5.2 Partial Journal Bearings

Since almost half the clearance of the full journal bearing is occupied by low
pressure or cavitated fluid, there are many cases where, provided that the
load is in an approximately constant direction, this part of the bearing may be
dispensed with altogether. Indeed, this is often a distinct advantage because,
although the low pressure region contributes little to the load capacity, it
does add to the viscous drag, as we shall see later. The result is the partial
journal bearing, as illustrated in figure 10.12.
w
w

(b)

(a)

Figure 10.12 Partial journal bearings

The partial journal bearing has the same form of film thickness as the full
journal bearing, and may be analysed using the Reynolds equation 10.14.
The boundary conditions, however, are different for partial bearings.
The cyclic form is no longer present and usually the inlet boundary condition
is known to be that the pressure is ambient at the beginning of the bearing,
(p = 0 at 0 = a). At the outlet it will either be sufficient to put the pressure
equal to ambient at the end of the arc (0 = a + p), as in figure 10.12a, or the
Reynolds boundary condition will have to be adopted, as in figure 10.12b.
Which condition is applicable will depend on the condition of operation and
the length of the divergent clearance space.

250
10.5.3 Load Capacity of Journal Bearings

As the load on the journal is increased from zero, the centre of the journal
will move into a more and more eccentric position, with decreasing minimum
film thickness. If the Reynolds equation is applied using the full Sommerfeld
solution, assuming no cavitation, theory indicates that the journal moves at
right angles to the load line; as shown in figure 10.13. Experience, however,
tells us that this is not a very realistic solution. In practice the locus of the
journal centre is very different from the full Sommerfeld prediction, and
the Reynolds boundary condition must be used adequately to describe what
happens.

Full
Sommerfeld

Exper1mentol
curve

Figure 10.13 Locus of shaft

centre within clearance circle
with increasing load

It is often useful to express the load capacity of a journal in terms of the

Sommerfeld number, S, a dimensionless parameter which is given by

S =wLD YJN (R)

C
2
(10.16)

and which depends upon the eccentricity of the bearing. Rearranging

equation 10.16

(10.17)

This gives the load per unit width of the journal in terms of the Sommerfeld
number and the properties of the fluid and the bearing. Values of the Sommer-
feld number for different eccentricity ratios, assuming the Reynolds boundary
condition and ambient pressure at 0 = 0, are presented by Pinkus and
Sternlicht 7 . These are shown in table 10.2.
It will be appreciated that, since the angle 4J between the line of centres and
load line varies with the operating conditions, even if the bearing is designed

251
TABLE 10.2

FjR
E s tP we
0 00 71
0.1 0.247 69
0.2 0.123 67 2.57
0.3 0.0823 64 1.90
0.4 0.0628 62 1.53
0.5 0.0483 58 1.32
0.6 0.0389 54 1.20
0.7 0.0297 49 1.10
0.8 0.0211 42 0.962
0.9 O.ot14 32 0.721
0.95 0.00605 23 0.568
1.0 0

so that the supply is located at 0 = 0 for one particular case, it will not be so
for other loads, speeds and viscosities. In practice we find that the perform-
ance is virtually unaffected by the position of the inlet within ±30° of 0 = 0.
Indeed, there is remarkably little difference between the performance of the
full journal bearing and the 180°, or even 150°, partial bearing. However,
if the inlet should move a considerable distance into the convergent zone,
the load capacity will be reduced.
From the table we can see that the Sommerfeld number falls, and hence
the load capacity rises rapidly as the eccentricity ratio approaches unity.
Indeed, the load capacity virtually doubles as e increases from 0.8 to 0.9.
However, in practice we cannot approach e = 1 too closely, since in this
region the films are becoming so small that surface roughness and other
irregularities of manufacture begin to have a significant effect on the per-
formance.

10.6 VARIABLE VISCOSITY-THE REDUCED PRESSURE

CONCEPT

At this stage we may usefully relax one of the assumptions stated in section
10.2. It is well known that the viscosity of most fluids is markedly dependent
on the pressure, and in many engineering situations extremely high pressures
can occur, for example, in the restricted contacts between gear teeth and
between rolling elements and their tracks. For such situations we must
include this change of viscosity in our theory. A widely used relationship
for the effect of pressure on viscosity, especially for lubricating oils, is
(10.18)

252
where 1'/o is some reference viscosity and IX is known as the pressure exponent
of viscosity. Remembering that in our Reynolds equation the pressure
gradient, and therefore the instantaneous pressur'!, is a function of viscosity
and that viscosity itself increases with pressure, we can see that this effect
rapidly escalates the pressure gradient, and therefore the pressure, to very
large values. Because the viscosity is now a function of pressure the integra-
tion of Reynolds equation as it stands is no longer a simple matter. We
therefore introduce a parameter q, known as the reduced pressure, which is
given by

(10.19)

It should be noted that q has a maximum value of 1/IX, and if we substitute

for p from this relationship into equation (10.5), we obtain

dqdx = 121'/o U +2 U(h~ - h*) + 121'fo(Vz -

1 2 (x - x*)
VI) -h-3- (10.20)

We can see that equation 10.20 is of exactly the same form as the original
Reynolds equation 10.5, except that p is now replaced by q. Since 1'/o is a
constant, equation 10.20 can be readily integrated. Thus, any pressure
distribution found for the constant viscosity case can be used for a fluid where
the viscosity is defined by equation 10.18, by the substitution

1
p = -loge(l - 1Xq) (10.21)
IX

As an example of this procedure, we may consider the case of two discs

rolling with some degree of sliding. For a constant viscosity we obtain the
pressure distribution shown in figure 10.14.
If the fluid has a viscosity which depends exponentially on pressure, the

\
~varible vtscos1ty

Constant V1scos1ty
(a=O)

Figure 10.14 Pressure distribution with

variable viscosity

253
solution is transformed using equation 10.21 to give the new pressure distri-
bution. It will be noted that in the latter case the pressure is tending to an
infinite value, although the load remains finite, but this does not, of course,
occur. What happens in practice is that the very large pressures produce
deformation of the bodies, which redistributes the pressure over a finite area.
This situation is dealt with in the next chapter under the title 'elastohydro-
dynamic lubrication.'

10.7 SHEAR STRESSES AND TRACTION IN HYDRODYNAMIC

FILMS

Although one of the functions of lubrication is the reduction of friction

forces, however thick the lubricant film may be there will still remain tan-
gential forces opposing the motion. Considerable reduction of the co-
efficient of friction can be effected by the provision of a fluid film, typically
from 1 to 0.001, but even such relatively small frictional forces will result in
the dissipation of energy and consequent loss of efficiency of the machine.
It is usually, therefore, the objective of the designer to reduce these forces
to as low a value as possible.
To evaluate the drag on the solid boundaries of a lubricated contact such
as that illustrated in figure 10.15, we use the definition of newtonian vis-
cosity
du
1: = '1- (10.22)
dz
where 1: is the shear stress at the surface and du/dz is the rate of shear. The
drag per unit width is then given by integrating the shear stress along the
length of the bearing at the top and bottom surfaces
L

F1 = J11(:~) z=0 dx
0
(10.23)

F2 = - J'1 (du)
L -
dz z=h dx
0

Consideration of the system will readily tell us the direction of the friction
forces. F 1 will act in the positive x direction to oppose the motion, while the
force on the top surface will be trying to drag it in the direction of the velocity,
the negative x direction.
For the typical velocity distribution shown in figure 10.15 clearly du/dz is
not necessarily the same for each boundary and, therefore, the frictional drag
on the two surfaces will be different. The value of dujdz will depend on the

254
Figure 10.15

physical properties of the fluid and the type, geometry and velocity of the
bearing surfaces. Some particular bearing types are analysed in the following
sections.

10.7.1 Friction in Sliding Bearings

The velocity distribution across the film in a bearing consists of two parts,
as shown in figure 10.15. Fluid theory tells us that there is a linear distribu-
tion due simply to the parallel velocity of the surfaces, and also a parabolic
distribution due to the pressure gradient. The velocity of the flow at a trans-
verse coordinate z is

u =- u( 1- ~) - ;'7 (h - z) :~
To get the velocity gradient we differentiate with respect to z, remembering
that we have assumed that the pressure does not vary across the thickness
of the film
du U 1 dp
- =- - - (h - 2z)-
dz h 2'7 dx
On the lower surface z = 0
u h dp
---- (10.24)
h 2'7 dx
and on the upper surface z = h

( du) U h dp (10.25)
dz z = h = h + 2'7 dx
These velocity gradients are evaluated by substitution of the film shape and
the pressure distribution. For example, the inclined slider shown in figure
10.16 has a film thickness

h = ho (1 + nx) L

255
~u

Figure 10.16

So that for the lower surface, substituting equation 10.24 in equation 10.23,
we get an expression for the friction force per unit width of the bearing
L

F = f ('1hU - ~ dp) dx
1 2 dx
0

17 u ~ log.(1 + nx)]L L
= [ n L - ! fh dp dx (10.26)
ho 2 dx

l
0
0

17U L 1 1 dh
-log (1 + n) - - [ (hp)'- Jp-
=
nh 0 e 2 ° 0L dx dx
Note that since p = 0 at x = 0 and x = L
(hp)~ = 0
For small inclinations the term p dhjdx is the component of the normal
pressure on the inclined surface resolved parallel to the other surface. If
the total resultant force on the inclined surface in this direction is denoted by
Px then
Px nUL
F1 = 2 + nho log.(1 + n) (10.27)

Similarly
Px 17UL
F2 = 2- nho log.(l + n)
The value of Px can be found by the triangle of forces P, Wand Px. Since IX
is a very small angle P = W and IX = Pxi W = nh 0 / L. Therefore, the surface

256
drags per unit width, since we have already evaluated W (equation 10.8)
are given by
F1 = YfUL [4loge(l + n) __6_]
ho n 2+n
(10.28)
F 2 = _ YfUL [2loge(1 + n) __6_]
h0 n 2+n
Similar expressions can be derived for other forms of sliding bearings.
Notice that the drag on the moving lower surface is greater than that on the
upper surface. Care must be taken, therefore, when experimental measure-
ment of friction are made, that the value obtained is the one relevant to any
subsequent operations.

10.7.2 Friction in Rolling/Sliding Discs

For rolling and sliding discs, such as those represented in figute 10.17 by an
equivalent cylinder and a plane, the same equation (10.26) for the drag on the
plane will apply, with slight modification of the coordinates

F1 = fn/ (

e,
2
-
h dp
-
2R dO
- - (U1
Yf
h
-
)
U 2 ) R dO (10.29)

where 0 1 is the inlet and rr./2 is the outlet. Note that for(} > (} 2 (the cavitation
boundary)
dp
p = 0 and - = 0
d(}

(10.30)

F,
<::: ' <:::: ""' <:: \ <:( \ \ \ ' \ ,,.., \ \ ____..X
E3 •
u,
Figure 10.17

257
The first term in this equation is zero, since the pressure is zero at the two
boundary positions. The second term represents the resultant pressure
force on the cylinder in the x direction, P". The last term requires the sub-
stitution of the film shape and subsequent integration. If this is undertaken
we find that

(10.31)

where A is a function of R/ho

Once again the forces on the two individual elements are different, except
in the case of rolling without slip (U 1 = U 2 ), when the rolling friction is
given by P"/2 on each component. Purday's analysis 5 gives a value for A of
2.84, using () 1 = -rr/2 as the inlet boundary, but this neglects the viscous
drag from fluid in the cavitated region. If the maximum possible contribu-
tion from this region is included, then A has a value of 3.48 for cases where
R/ho is large.
The value of P" is given by

P" = 4.5tJ(U 1 + (R)1/2

U 2 ) ho (10.32)

for the usual large values of R/ho. For the plane of course P" = 0. If the real
situation consists of two discs, clearly there will be a horizontal force on each
member. If the two discs are of radius Ra and Rb respectively and R is the
equivalent cylinder radius, then it can be shown that

10.7.3 Friction in Journal Bearings

The friction torques acting in the journal bearing may be evaluated in the
same way as for the other bearing forms.

258
where the subscripts j and b denote the journal and bush respectively. If the
full Sommerfeld condition (see section 10.5.1), including negative pressures,
is used then these equations, like those for load capacity dealt with earlier,
are relatively simple to solve, giving the following expressions for the surface
friction torques per unit axial length
4,.VrrR 2 (1 + 2e 2 )
FiR = - C (2 + ez)(l _ ez)t;z
4'1UrtR 2 (1 - e2 ) 112 (10.33)
FbR = C 2 + ez

A solution using the more relevant Reynolds boundary condition (see

section 10.5.1) is more difficult to achieve. Values of the friction on the journal
using this boundary condition are presented by Pinkus and Sternlicht 7 in the
form
e sin cjJ 2rr 2 S
- -2- - (1 - ez)t/2
(10.34)

where S is the Sommerfeld number

S = 11N
W LD (R)C 2

as in equation 10.16. Values of FiR/WC taken from this reference are

found in table 10.2. The value of this parameter falls with increasing eccen-
tricity ratio and hence, since R/C is a constant. F/W, which may be regarded
as a form of coefficient of friction, also decreases with increasing eccentricity.
Figure 10.18a illustrates the forces on the bearing components from the
lubricant film, while figure 10.18b shows these forces reversed, that is, the
force on the film. If we consider the equilibrium of the oil film by considering
the forces acting on it, the moments due to the friction are FiR and FbR in
the directions shown, assuming R = Ri ~. ~ There is an additional

w w

(b)

Figure 10.18 (a) Friction forces on bearing com-

ponents, (b) forces on lubricant film

259
couple due to the normal loads Won the journal and bush being displaced
from each other by a perpendicular distance e sin ¢. For equilibrium
(10.35)
This gives an equation for Fh which, when written with that for Fi, yields a
pair of expressions having a pattern which by now is becoming familiar
(compare with equations 10.27 and 10.31).
WC~: sin¢ 2rc 2 SWC
FiR = - 2 - (1 _ ~:2)1;
( 10.36)
WC~: sin¢ 2rc 2 SWC
FbR = - 2 + (1 _ ~:2)1;
Note that the friction on the journal, which is of course the value used
for calculation of power loss, is always greater than the friction on the bush,
except for the concentric case. We must remember this if friction is being
measured experimentally in the usual way, namely by the reaction torque
on the stationary bush.

10.8 FINITE LENGTH BEARINGS

In the analysis so far we have considered the bearing to be infinitely wide and
the forces calculated have been for unit width of such a bearing. In practice
the lubricant will flow across the width of the bearing and out of the ends,
thus modifying such parameters as the pressure distribution, load capacity
and friction forces. This does not mean that the earlier results have no value,
but that care must be taken to apply them only in situations where the side
flow does not make them too inaccurate. Also the performance of a finite
bearing is often expressed in terms of the infinite bearing and side-leakage
factors.Thus, for example, the load capacity
Wrinite = (side-leakage factor) x Jt;nfinite

The analysis developed so far will be referred to as the one-dimensional

case, since the Reynolds equation contained only one derivative dp/dx,
precluding flow and pressure gradients in the y direction. The new form of
the Reynolds equation is ·

!_(h3 ap) + !___ (h3 ap)

ax ax ay ay
= 121'J(u I +2 u2) ah
ax (10.37)

The integration of this equation is a formidable task. Much effort has been
expended on the production of analytical and analogue solutions, although
in many cases the recent increase in availability of high-speed digital com-
puters has made numerical methods more attractive. Fortunately we need
not concern ourselves with the mechanics of solution, but will investigate the
significance of the results for particular bearing forms.

260
10.8.1 The Finite Slider Bearing

If we consider the slider bearing illustrated in figure 10.19a, we can see that
the moving surface will drag the lubricant into the bearing, but that unlike
the infinitely wide case some of the lubricant will flow out of the bearing
sides, as shown in figures I 0.19b and c, and therefore the pressure generated
in the narrowing clearance will be reduced. Obviously, the magnitude of
this effect will depend upon the ratio of the width of the bearing to its length,
the narrower the bearing the more its pressure will be reduced and vice versa.

(a)

(b) (c)

Figure 10.19 Side leakage in slider bearing

Fortunately, the conclusion drawn by Reynolds that the actual shape of

the slider surface is relatively unimportant is still valid, and so if we investi-
gate the case of the plane inclined slider, this will give a good indication of
what happens for other shapes.
The notation and the bearing are shown in Figure 10.19a. As before

h h{I + ~)
=

and the Reynolds equation is

!__ (h 3 op) + _£_ (h 3 op) = 6Yfu dx

dp (10.38)
ax ax oy oy
261
The pressure distribution resulting from the solution of equation 10.38 is a
function of both x and y and will have the form illustrated in figure 10.20a,
as compared with 10.20b, which is the one-dimensional case. Along the
edges of the bearing, y = 0 and y = B, the pressure is of course ambient.
The load capacity of the bearing is naturally reduced by side leakage. The
extent to which this is so can be seen from figure 10.2la, which is based on

(a) (b)

Figure 10.20 Pressure distribution in finite and in-

finite slider bearing

results derived numerically by Jakobsson and Floberg 8 , Hays 9 and Muskat

et aP 0 . For a pad of conventional dimensions with B ~ L, the load capacity
is only about a half that of the infinite slider, and until we make the width
about five times the length or more, the one-dimensional solution is plainly
inadequate.
When we consider the friction forces we must remember that, since the
load capacity is reduced, the surfaces must come closer together to carry
the same load and therefore the friction coefficient is likely to rise. This is
indeed the case, as the curves of figure 10.21 b show.

10.8.2 Finite Rolling and Sliding Disc~

The inclusion of finite width will affect the performance of discs in the same
manner as that for the slider. There is an additional complication in that the
cavitation boundary will no longer fall at a constant coordinate () 2 , but will
be curved. This can be seen in figure l0.22a, which represents a typical
pressure distribution between a finite cylinder and a plane as calculated by
Dowson and Whomes 11 . The cavitation boundary occurs at a decreasing
value of() as we depart from the centre line, until at the edge it coincides with
the position of minimum film thickness(() = 0).
Once again the load capacity will reduce as the bearing becomes narrower.
However, because the pressure is very slow to build-up through the bearing,
and the effective load carrying region is a narrow strip near the position of
minimum film thickness as shown in figure 10.22b, the load capacity of the
discs is relatively unaffected by side leakage for all but very narrow discs and
very low values of R/h 0 • In fact, it is as though the performance of the bearing
were defined by the dimensions of the load carrying region. Side leakage

262
300

200) ·-

100) \ . --- - --
\
70) r-...
"-...
50) \ 1'-- 8/L

~
025
...JI-c:o
.._
8/L [\..
~ 30) \ ...._
~
N
0\ ~
V.l
~ 20) \ --·- f-·
04

\
05

10)
""- ....
-
07
10
2

r~ l.--:t:d
4

--~
00

I~ I I I -I J:: 5 -~
............
.......... ~

0 2 3 4 5 6 0 2 3 4 5 6
n n

(a) (b)

Figure 10.21 (a) Load capacity of finite slider bearing 7 , (b) friction in finite slider bearing
Cen!re -line
__c:t
beonng

Col

carrymg reg1on
(b)

Figure 10.22 Pressure distributions between finite cylinder and

plane

factors for load capacity as calculated by Dawson and Whomes are shown
in figure 10.23a. For the most common practical cases where R/ho > 104 ,
the load capacity suffers a reduction of 10 per cent or more only for discs
with length-diameter ratios less than 0.2.
The coefficient of friction behaves in a more complicated way, since it
depends upon the relative amounts of rolling and sliding. A typical set of
curves to illustrate how the coefficient of friction is affected by the geometry
is shown in figure 10.23b for R/ho = 10 3 . We can see that while rolling
friction falls slightly as the discs become narrower, for large values of slip
the friction coefficient rises as the width of the disc falls.

10.8.3 The Finite Journal Bearing

It is implicit in the infinite journal bearing that the system is effectively a
closed one, the same lubricant being used continuously. In the finite journal
bearing this is of course not so, since the lubricant escapes from the ends and
must be made up by a supply. In some cases the bearing is completely
immersed so that fluid will flow in from the edges in the low pressure zone.
More commonly, however, a discrete supply is incorporated at some position
in the low pressure half. The fluid may enter the bearing in a variety of
ways-a single hole, multiple holes, a slot-each ofwhich will give a slightly
different performance.

264
LIR

(a)

.....
c
Q
v
.E
0
c.,
¥
8u

Sl1p 1%)

(b)

Figure 10.23 Load capacity and friction for a

finite cylinder and plam,l 1

Because the fluid is permitted to flow sideways, we have the familiar

reduction in load capacity compared with that of the infinite journal bearing.
Of course, the greater the length-diameter ratio, the greater the load capacity.
Also as the eccentricity becomes larger, the load capacity approaches that
of the infinite bearing. Typical curves based on data from Fuller 12 are shown
in figure 10.24a. Because of difficulties in manufacture and alignment,
journal bearings seldom have a length-diameter ratio of more than one.
From the curves we can see that for practical bearings a considerable
reduction in load capacity will be experienced, except when the eccentricity
ratio is very large. Because ofthe reduced load capacity and hence the smaller
film thickness, the coefficient of friction is much higher for the finite bearings,
especially at low eccentricity ratios. The variation of coefficient of friction
with length-diameter ratio and eccentricity ratio is shown in figure 10.24b.

265
LIO
(a)

1000..--------.-------.--

....
~ 10~-4

O·IL__ _ _ _ _ _O_L-5_ _ _ _ _ _ _1._,·0-

co
LID
(b)

Figure 10.24 (a) Load capacity of finite journal

bearing 12 , (b) friction in finite journal bearing

266
The Short Bearing Theory
We have seen that the Reynolds equation 10.39 for the finite journal is
difficult to solve analytically

_a_ (h 3 op) !_ (h 3 op) = 611 u ah (10.39)

R 2 oe ae + ay ay R ae
In an attempt to circumvent these difficulties, Ocvirk and Dubois 13
considered a bearing where pressure gradients around the circumference
were very small compared with those along the length. This has the effect of
reducing equation 10.39 to

(10.40)

Just as we initially considered an infinitely long bearing, this represents an

infinitely short bearing. The results are only applicable to situations where
the pressure gradients in the direction of motion are not large. For instance
the short bearing theory is most unsuitable to describe the performance
of discs where very high pressure gradients are experienced, and is not
recommended for the majority of slider bearings. For short journal bearings
running at low eccentricities, however, very useful results can be obtained
very simply.
Integrating equation 10.40 twice with respect to y gives

The boundary conditions are that p = 0 at y = ± L/2 where y is measured

from the centre line of the bearing. This yields

Of course this equation gives a circumferential pressure distribution at any

value of y which is proportional to dhjd() and dhjd(;l is negative for half of the
circumference, as shown in figure 10.25. For practical purposes we only

Figure 10.25

267
consider positive pressures by simply deleting the contribution from the
range () = n to 2n. The load capacity then becomes
_rJUO s 2 2 21;2
W- 4e2 (1 _ s 2 ) 2 [n (l - s ) + 16s ] (10.41)

and the friction coefficient

2n 2 CS
f = R(l _ s2)1;2 (10.42)

where Sis the Sommerfeld number as before. To get some appreciation of the
range of applicability of the short bearing theory, reference should be made to
figure 10.26. We can see that for bearings of the more common proportions,
L/D ~ 1, the short bearing theory gives good results for moderate eccen-
tricity ratios (s ~ 0.6).

- Numencal methods
---- Short beanng theory

0 2
LID

Figure 10.26 Application of short bearing theory

10.9 THERMAL EFFECTS

Throughout our discussion of hydrodynamic lubrication there has been one

principle underlying all our arguments: namely, that pressure is generated
by the lubricant being drawn into a narrowing clearance space. It is at first
somewhat surprising, therefore, to find that two parallel flat surfaces with
relative longitudinal motion, such as those shown in figure 10.27, do generate

v///// // /J
Inlet
\ \ b~'u \\\ \~ 1 0utle

Figure 10.27 Parallel surface bearing

268
a pressure within the film, in spite of the absence of any convergence. This
means that such a device, which is called a parallel surface bearing, will carry
a load. Why should this be so? Clearly if we look at our usual Reynolds
equation for such a slider

dp = 6 u(h - h*)
dx 'I 3 h
since h has the same value everywhere

dp =0
dx
resulting in no pressure build-up. The discrepancy between theory and prac-
tice comes from two of our initial assumptions which are no longer applicable:
that the fluid is incompressible and that the fluid properties remain constant.
As the lubricant is drawn into the bearing by the moving surface, it will be
subjected to shear due to the velocity gradient across the film. The energy
involved in this shearing will appear as heat, raising the temperature more
and more as it passes through the gap. This has two effects. The first is to
reduce the viscosity of the fluid, a typical viscosity-temperature curve
having been shown in chapter 9. The second is to try to increase the volume
of the fluid. It is this latter effect which generates the pressure. The fluid is
unable to expand, due to the restricted clearance and therefore the pressure
rises. This effect is sometimes called the thermal wedge.
In order to analyse this system we must return to the Reynolds equation in
a fuller form than we have used so far. Neglecting side flow

j_
dx
(ph12'1 dp)
dx
= i_ (pUh)
dx
3

2
(10.43)

Integrating this with respect to x gives

ph 3 dp pUh
-~=+A
12,., dx 2
or
dp 6,.,u 12A,.,
-=~+
dx h1 ph 3
and integrating again, remembering that h is a constant

6U
p=hl f '7dx+~
12A f 'I
pdx+B (10.44)

where A and B are constants.

In order to solve this equation it is necessary to know the viscosity and
density at any point along the bearing, which in turn requires a knowledge
of the temperature variation through the contact. To assess this with com-
plete accuracy is obviously a very difficult task, but a good approximation

269
is to assume that the energy dissipation is constant along the bearing and
that all this energy goes to raising the temperature of the lubricant, giving a
linear temperature distribution.
The density-temperature relationship for most fluids may be written
P = Pi + cxr(t - ti) + /3 (t 1 - ti + ·· ·
where Pi is the inlet density. In fact the third and subsequent terms account
for a very small percentage of the density change and so
p = Pi + tx (t 1 - tJ
Since the relationship between density and temperature is now linear and the
temperature variation along the bearing is also linear, the density variation
can be expressed as
X
P = Pi + L (Po - p;)

where Po is the density at the outlet.

We find that the contribution of viscosity variation is much smaller than
that of density change. This is partly due to the fact that the decrease in
viscosity with a temperature rise is to some extent offset by the increase in
viscosity with pressure rise. Therefore in our approximate analysis we shall
consider the viscosity to be constant.
If p' = pjpi equation 10.44 becomes

The boundary conditions are that p = 0 at inlet and outlet and so

(10.45)

where

p, = 1 + -CXr (10 - (.)

Pi I

We will choose some typical values so that we can appreciate the magnitude
of this effect
cx 1 for mineral oils :::,: -0.00065/"C
Pi for mineral oils :::,: 0.9 gjcm 3

270
The value of p' will depend upon the temperature difference between the
inlet and the outlet as shown in figure 10.28a. Since many lubricants boil
at about 100 oc, the maximum temperature difference which can occur is
about the same value. If we use t 0 - ti = 100, the pressure distribution has
the form of figure 10.28b, where p' is a dimensionless pressure. The maximum
value of p' is about O.ot 1. For a plane-inclined slider p;,ax ~ 0.042.

XIL
(a) (b)

Figure 10.28 Thermal effects in parallel surface bearing

Since the shapes of the pressure distributions in the two cases are very
similar, the relative load capacities can be assessed from the values of
maximum pressure. We can see therefore that the parallel surface bearing has
a load capacity approximately 1/3.5 that of the corresponding inclined
slider. Of course it is very seldom that the temperature rise through a bearing
is 100 oc. For most practical bearings it is more of the order of2-20 oc with a
correspondingly lower load capacity. Therefore for bearings where the power
dissipated in viscous friction is only sufficient to raise the temperature of the
lubricant by a few degrees, we are justified in neglecting thermal effects.

10.10 GAS-LUBRICATED BEARINGS

Throughout this chapter on hydrodynamic lubrication we have used the

words 'lubricant' and 'fluid' without indicating their nature, and although we
probably think automatically of water, oil, grease, and similar substances,
there is no reason why the fluid should not be a gas, such as air. Our physical
feel for the viscosity of an oil between our fingers, but not for the same effect
with air, is due to our fingers lack of sensitivity at viscosities below about
5 cP, rather than to the absence of this property.
As long ago as 1854 Hirn 14 proposed air as having great advantages as a
lubricant for bearings, although it was not until 1897 that Kingsbury 15
demonstrated the viability of such a proposal. Today gas bearings are very
much a part of our industrial life, doing useful service in machine-tool
spindles, inspection and measuring instruments, turbomachinery, gyro-
scopes, dental drills and many other more specialised applications.
The hydrodynamic gas bearing operates on exactly the same principle
271
as the liquid counterpart, namely the drawing of the fluid into a convergent
clearance with the consequent rise in pressure. However, if we look at any of
the equations we have derived for pressure distribution in hydrodynamic
bearings, we find that, at least in the first approximation, the pressure is
proportional to the lubricant viscosity. Since the viscosity of a gas is much
lower than that of a liquid, the pressures and hence the load capacities of gas
bearings are much less than with a liquid. The frictional drag is reduced in
roughly the same proportion, so that values of friction force in gas bearings
are very low. The coefficient of friction, however, is comparable with, and
indeed often slightly greater than with liquid lubrication.
To the purist 'hydrodynamic' and 'gas' are a contradiction. For this reason
gas bearings which generate their own pressure should be called 'aero-
dynamic' or 'self-acting'. We will consider the advantages and disadvantages
of self-acting gas bearings.
Advantages
(1) Very low friction.
(2) If air is used the lubricant is readily available and free.
(3) The lubricant will not contaminate the bearing surfaces.
(4) Unlike a liquid which boils with quite modest temperature rises and
freezes solid at low temperatures, a gas can be used at both very high and
very low temperatures.
Disadvantages
(1) Low load capacity. Pressures in self-acting gas bearing are typically
10 5 Njm 2 as compared with 10 7 N/m 2 in a liquid bearing.
(2) Susceptibility to instability. All fluid bearings exhibit a tendency to
instability in the form of self-indiced vibrations under certain conditions,
but gas bearings are rather more prone to this. There are, however,
design methods to overcome this difficulty. Hydrodynamic instability
is dealt with in more detail in section 10.11.
(3) Surface finish and machining accuracy of the bearing components
must be rather better than for liquid bearings. This is because lubricant
films and clearances are typically, but not always, thinner than those
experienced in liquid lubricated bearings.
In addition to the foregoing specific advantages and disadvantages we
must bear in mind other differences which are introduced by the change of
lubricant from liquid to gas.
Since the energy dissipated by the friction forces will be very low, there will
be little temperature rise through the bearing. We can therefore in most cases
consider the process to be· isothermal without much loss in accuracy. The
effects of compressibility will become important if the pressures generated are
at all high.
There is no cavitation in gas bearings. All films are continuous even though
the pressure falls below ambient.

272
10.10.1 Gas Bearing Theory

We employ a form of the Reynolds equation to describe the situation

~(h 3 op) + i_(h 3 op)= 12 (V1 + V2)~( h) (10.46)

ox Pox oy P oy Yf 2 ox P

Notice that we have to retain the density p as a variable. The relationship

between pressure and density will be in general polytropic, namely

!!_ = constant (10.47)

pn

In our consideration of gas bearings we must use

p = absolute pressure
since clearly if we regarded p = 0 as ambient pressure, equation 10.47 would
imply zero density in this condition, which is not true. Substituting for p
from 10.47 into equation 10.46 and cancelling the constant gives

If we consider the case of the infinitely wide bearing, implying no variation

in the y direction, and we assume that isothermal conditions prevail that is,
n = 1, then equation 10.48 reduces to

~ ( h3 dp) = 12 (U 1 + U z) ~ ( h)
dx p dx Yf 2 dx p

which on integration becomes

or
dp = 12 (U 1 + V 2 )[ph + CJ
dx '7 2 ph 3

If as usual we define the coordinate of the position where the pressure is a

maximum (dpjdx = 0) by x*, then

dp = 1217 (U 1 + V 2 ) [ph -(ph)*]

dx 2 ph 3
and for isothermal conditions when pjp =constant, this may be written
dp = 12 (U 1 + V 2 ) [h- (p*/p)h*] (10.49)
dx '7 2 h3

273
This equation is identical to equation 10.3 for incompressible fluids, except
for the inclusion of the density ratio. We can conclude therefore that for
bearings where the pressures generated are not high and there is little
consequent compression of the gas (p ~ p*), the bearing will behave exactly
as the liquid bearing, the pressure and load capacity being in the ratio of the
viscosities, that is
Wgas '1 gas 1
~iqud = '1Iiquid ~ 1()(){)
Therefore for gas bearings which do not produce large pressures, we may use
the solution developed for incompressible lubrication.
It is usual in gas bearing analysis to consider the equation in terms of
dimensionless variables, putting
- p X
p=- x=-L
Pa
where Pa is ambient pressure and hm is some reference film thickness. This
transforms equation 10.49 into
dp = 12 (U 1 + U 2 ) -~ [11 - (p*/p)/1*] (10.50)
dx 11 2 h;, Pa /1 3
The term
2 (U I + Vz) L
1 '1 2 ~ mPa
is often referred to as the bearing number or compressibility number and is
denoted by A. This is a very significant parameter to which we will refer
repeatedly in our consideration of the various gas bearing forms which
follows.

10.10.2 Gas-Lubricated Slider Bearings

Inclined Slider
If we consider the plane-inclined slider illustrated in figure 10.29a, we may
solve equation 10.50 by integrating and inserting the boundary condition
that p = 1 (p = p3 ) at inlet and outlet. The pressure profile produced is
shown in figure 10.29b. We can see that the maximum pressure for a gas
bearing occurs at a smaller film thickness than for the incompressible lubri-
cant, so that the centre of pressure is always further towards the trailing
edge of the bearing.
If we investigate the effect of bearing number on the pressure profile, we
get a family of curves such as those derived by Gross 16 and presented in

274
(a)

(b)

Figure 10.29

figure 10.30a. Notice that for A-+ 0 the pressure distribution approximates
=
to that for an incompressible fluid. At A oo the peak pressure occurs at
the end of the bearing and· in this case
Pmax h;
Pa ho
The variation of load capacity of the bearing is shown in figure 10.30b.
For low bearing numbers we can see that the maximum load capacity occurs
when hJho is between 2 and 3, but for higher bearing numbers the optimum
ratio increases.
Algebraic expressions can be produced for the load capacity of an inclined
slider for the extreme cases A -+ 0 and A -+ oo, but no general formula is
available to cover the whole range of bearing numbers. The magnitude of the
load capacity may be seen from figure 10.30b, or by reference to table 10.3.
The friction force on each component is given by
2F hm =log. li; + 3(/i;- 1) ~
'IBL(Ul + U2) li;- 1- A LBpa
where the addition sign is to be used for the faster moving surface.

Step Slider Bearing

We have seen that for incompressible fluids the step bearing is theoretically
the most efficient geometrical shape for a slider. It is reasonable therefore to
investigate the performance of such a bearing when lubricated with a gas.
The analytical procedure is exactly the same as that for the liquid lubricated
bearing, namely the application of Reynolds equation to each section of the

275
(a)

(b)

Figure 10.30 (a) Pressure distribution in a slider lubricated with a

compressible fluid 16 ; (b) load capacity of gas lubricated slider 16

TABLE 10.3 LOAD CAPACITY OF PLANE-INCLINED GAS

LUBRICATED SLIDER (neglecting side leakage)
(After Gross 16 )

~ 0.5
1.5

0.01091
2.0

0.01323
3.0

0.01232
4.0

0.01034
6.0

0.007241
1.0 0.020172 0.02640 0.02063 0.02063 0.01448
5.0 0.0957 0.1234 0.1201 0.1023 0.07243
10.0 0.1486 0.2124 0.2252 0.1980 0.1435
50.0 0.1942 0.3367 0.5618 0.6424 0.5943

276
bearing, having a pressure Pstep at the change of clearance. The mass flow
in each section may then be equated to give a value for Pstep· The resulting
pressure distribution has typically the form of the curves in figure 10.31.
We can see that for low bearing numbers the profile is almost triangular,
approaching the incompressible case (see figure 10.7). As A increases the
pressure in the inlet region is slow to build up, compared with the triangular
profile while that in the outlet is even slower to decay. The net effect is to

A~o
-------.-------------
A ~20

pt 0
t : 'lcL L L L L L L L j
\ J,~ \ \ \\ ( ( (: \ \ \ \ \ \ \~ \
Figure 10.31 Gas lubricated step bearing 16

increase the area under the curve and therefore more load is supported. Thus
for the same step pressure the gas film will usually carry more load than the
incompressible film. The load capacity of the bearing can be calculated and
typical results are illustrated in figure 10.32a.
The optimum values of step height and ratio L 1 /L 2 can be deduced from
the curves presented in figure 10.32b.

10.10.3 Gas-Lubricated Journal Bearings

During the past decade the gas journal bearing has become an indispensable
part of our industrial life. The low friction energy dissipation, together with
its other advantages, has led to its incorporation in many different forms in
widely differing applications.

277
3.------------,-
(a)
i!
L2-
-I

20 25

I 0.-------r-- ------,----- ,------,

(b)

0 5 10 15 20

Figure 10.32 Gas lubricated step bearing 16

The gas journal bearing operates in the same way as its liquid counterpart
except for the important difference that there is no cavitation zone, the film
being continuous. A typical pressure distribution is shown in figure 10.33
(cf. figure 10.11) for a 360° bearing. The subambient portion of the pressure
curve will of course add to the load capacity since it applies a suction to the
top of the shaft which opposes the applied load.
As with all self-acting gas bearings the load capacity of the bearing is well
below that of the comparable liquid bearing, but in the case of the full gas
journal bearing, the theoretical load capacity is seldom the practical limit
of operation. This is more often dictated by the onset of instability, the most
prevalent form of which is known as 'half-speed whirl' and which will be
discussed in more detail in section 10.11.

278
w

Figure 10.33 Pressure distri-

bution in gas lubricatedjournal
bearing

An investigation into the load capacity of the gas journal bearing will
reveal that once again we have the problem of a Reynolds equation to which
there is no direct analytical solution. Many methods have been adopted in an
attempt to solve this problem, analytical and semi-analytical solutions
being produced for very high and very low bearing numbers, but for the
general case numerical methods are essential. The most significant contribu-
tion in this field is the work of Raimondi 1 7 who used relaxation methods to
produce set~ of load-bearing number and attitude angle-bearing number
curves for various length-diameter ratios. Typical of these are figure 10.34a
and 10.34b, which show load and attitude angle, the angle between the load
load and the line joining the shaft and bush centres, for a full journal bearing
having a length-diameter ratio of unity.

Partial Journal Bearings

In liquid lubrication we have seen that the use of a partial journal bearing
can reduce the frictional torque on the shaft without much affecting the
load capacity. In gas lubrication, however, the load capacity is reduced
significantly if the low pressure section of the circumference is removed
because the suction of the subambient pressure does contribute to the load
capacity. The friction will also be reduced. The most common reason for the
use of the partial gas journal bearing is the suppression of half-speed whirl.
This will be discussed in section 10.11.

10.10.4 Other Gas Bearing Types

We have dealt with the gas bearing slider and journal bearing, but many other
types of gas bearing are frequently used in industry. A few of the more im-
portant ones are described below.
279
lOr-.,~

e08

(a)

(b)

Figure 10.34 Gas lubricated journal bearing: (a) load

capacity; (b) attitude angle

The Tilting-pad Journal Bearing

The tilting-pad journal bearing is illustrated in figure 10.35a. The pads, of
which there may be any number, are mounted on pivots so that the pads may
adopt their own angle to the shaft surface. Although it is obviously more
complicated to manufacture and assemble, the load capacity is somewhat
reduced and design data is not so readily available, it does have some

280
Prvot
Pod

(a) {b)

Cc I

(e)

Figure 10.35 (a) Tilting-pad journal bearing; (b) pivoted

sector bearing; (c) spiral groove thrust bearings; (d) spiral
groove journal; (e) foil bearing

distinct advantages over the full 360° journal bearing. It is almost entirely
free from half-speed whirl and therefore may be used at high speeds, where this
instability precludes the use of full journal bearings. If, as is usual, the pads
are permitted to pivot axially as well as circumferentially, they are capable
of absorbing shaft misalignment. Any foreign matter or debris may escape
from the bearing clearance by way of the gaps between the pads.
This type of bearing has been applied with great success to rotating turbo-
machinery.
A derivative of the tilting-pad bearing is the 'window pad' bearing or
'nutcracker' bearing, illustrated in figure 10.35b, which combines the load
capacity of the full journal bearing with the whirl resistance introduced by
the tilting pad.

281
Spiral Groove Bearings
This form of bearing, originally developed by Whipple 18 • 19 is used as both a
thrust and journal bearing. Figure 10.35c illustrates its use as a thrust plate.
In the surface of the plate is cut a series of spiral grooves to a depth of about
0.01-0.05 mm. When the plate or its mating surface is rotated in the correct
direction, the gas is dragged along the grooves into the bearing until it meets
the end of the groove. Because its exit is restricted a pressure is generated to
support the load.
The journal bearing counterpart, sometimes called the herringbone
grooved bearing, is illustrated in figure 10.35d. In this case gas is pumped by
the grooves from the ends of the bearing into the centre.

Foil Bearings
In this form of bearing a flexible band is wrapped partially round a shaft as in
figure 10.35e. If the band is stationary and the shaft rotating, we have a type
of journal bearing, whereas in some cases the shaft may be stationary and
the band moving as happens, for example, in magnetic tape machines to guide
the tape over the recording head. In either case we have a converging clear-
ance space which is sufficient to generate a gas film between the two com-
ponents.

10.11 HYDRODYNAMIC INSTABILITY

In the preceding work we have considered the bearing to be running steadily

with no variation of position, film thickness, angle, and other factors, with
time. In practice this is not always the case, because hydrodynamic bearings
are subject to instability.
In general we can differentiate between two forms of instability. The first
type, which is often called synchronous whirl when found in journal bearings,
is caused by a periodic disturbance outside the hearing such that the bearing
system is excited into resonance. Since the lubricant film can be considered
an elastic system, even though the stiffness is not constant, it will be capable
of such a resonance. For example, if the speed of a rotating shaft in a journal
bearing is progressively increased, at some stage an out-of-balance or similar
periodic force will produce resonance of the bearing. On further increasing
the speed we may pass through the resonant frequency. Both thrust and
journal bearings are subject to instability, but we will restrict our discussion
to journal bearings, since it is here that the problem is most frequent and
troublesome.
The other form of instability is induced in the lubricant film itself and is
called 'half-speed whirl'. The reason for this is that the displacement of the
shaft centre under a load is not in the direction of the load (see figure 10.1 0).
If, in figure 10.36, the shaft centre is displaced from 0 to P, the restoring

282
w

Figure 10.36

force due to the lubricant film pressure does not act along OP, but has a
component W sin 4J which will cause the shaft to move in a circumferential
direction and to whirl; that is, at the same time as the shaft rotates about its
own centre P, the shaft centre tends to rotate about the bearing centre 0.
If the whirl takes place at half the rotational speed of the shaft, this will
coincide with the mean rotational speed of the lubricant, since the circum-
ferential velocity of the lubricant varies from zero at the bush to the shaft
speed at the shaft. Because the film shape and fluid then have no relative
circumferential motion, no hydrodynamic film is produced and the lubrica-
tion fails, often with disastrous consequences. This phenomenon is en-
countered in all full journal bearings, but is especially prevalent in gas
bearings, because of their low damping capabilities, because the attitude
angle 4J is in general higher than for the comparable liquid bearing and
because of their frequent use in high speed applications.

10.11.1 The Prediction and Suppression of Instability

Synchronous Whirl
In order to investigate the synchronous instability of any rotorjbearing
system, it is necessary to evaluate the characteristics of that system: shaft
inertia and flexibility, stiffness and damping characteristics of the bearing
films, etc.A model for such a system is shown in figure 10.37a. The analysis
presents problems since the bearing properties themselves change with the
rotational speed and whirling speed.
If the rotating part is not perfectly balanced, the locus of the shaft centre
will be a circle or some other closed orbit about the steady equilibrium po-
sition. The size of the orbit will depend on the bearing stiffness relative to the
rotor stiffness in the way indicated in figure l 0.37b. If the whirl orbit remains
the same in successive rotations, the whirl is stable. This will be the case until
a 'critical' speed of the system is reached. At this point the rotational fre-
quency of the rotor unbalance and the natural frequency of the system will
be in resonance and the whirl orbit will increase. If the rotational speed is
283
(a)

Very flexible Moderately flexible

beanngs beanngs

1st mode ~ =----=-= B ~ -- . . _y____ _

2ndmode ~-
:!~ ~-= "'-~ ~ - -
.

3rdmode
~-:
0 , ~ 'v
(b)

Figure 10.37 (b) (After MTI/ Rensselaer Polytechnic Institute Gas Bearing
Design Manual)

kept at approximately this value, the magnitude of the whirl will increase
until either it reaches a stable position with a large amplitude limited by the
damping capacity, or the orbit increases until failure takes place. However, if
these critical speeds are passed through rapidly, the whirl orbit does not have
time to grow sufficiently for failure to occur. A typical case is shown in
figure I 0.38.
It is usual procedure therefore to design the bearing system so that the
critical speeds do not coincide with the most commonly used running speeds.
This may be done either by increasing the bearing stiffness so that the critical
speeds are very high, or reducing the stiffness so that the critical speeds are
quickly passed through and normal operation takes place where the attenua-
tion is large. The stiffness may be increased by reducing the bearing clearance,
but this often imposes too stringent manufacturing and assembly conditions.
Another means of suppressing or allowing for whirl is the introduction
of extra damping int0 the system. This is often done by flexibly mounting

284
Low accelerat1on

.,I
ut
::>

!i
&I

F1rst cnr1cal Second en r ICC I Des treble

speed speed operat1ng
speed

Figure 10.38 ( Ajier MT/j Rensselaer Polytechnic Institute

Gas Bearing Des~qn Manual)

the bearing housings in rubber '0' rings or metal diaphragms. This intro-
duces another element into the analysis, which must be allowed for as
illustrated in Figure 10.39.
It is clear from the above considerations that rotating parts to be sup-
ported in hydrodynamic bearings should be balanced as carefully as possible
in an attempt to eliminate or at least minimise synchronous instability.

Fi,qure 10.39

Half-Speed Whirl
This form of instability is induced in the hydrodynamic film itself and is
independent of rotor balancing. It is always present in hydrodynamic
bearings, but its amplitude only becomes large enough to be apparent at a
certain threshold rotational frequency. The system will not pass through this
danger region with a further increase in rotational speed; on the contrary
this will only accelerate failure.
The value of the threshold speed is dealt with by Marsh 20 . Hydrostatic

285
bearings are also subject to this form of instability, if the rotational speed is
sufficiently high for the hydrodynamic effects to become important. The
rotational speed will once again be in the region of 1.5 to 2 times the lowest
critical whirl speed and the actual whirl speed will be half the rotational
speed for a liquid bearing and in the range 0.25 to 0.45 the rotational speed for
a gas bearing. In this case it is termed fractional-frequency whirl.
The most important ways of preventing or minimising half-speed whirl are
by the introduction of damping and by interfering with the circumferential
symmetry of the bearing.
If extra damping is incorporated into the system the energy generated by
the whirling can be partially destroyed by the damping medium. This is often
done by flexibly mounting the bearings as described earlier.
Many devices are employed to break the symmetry of the full journal
bearing, although there is often the price to pay of reduced load capacity,
especially with gas bearings. Some of these methods are illustrated in
figure 10.40. The methods incorporated range from the simple axial groove,
figure 10.40a, cut in the low pressure half of the bearing to the tilting-pad
arrangement of figure 10.40c. The 'lemon' bearing has deliberate ovality

(O)

AXIal groove

O~
b)

Port1a1
beonng

(c l Tilling pod
beonng

(dl Lemon
beonng

Figure 10.40 Instability-inhibiting devices

286
of the bore. Bearings can have two, three, four, or any number of lobes and
indeed often accidental out-of-roundness is sufficient to delay the onset
of whirl. While the partial bearing of figure I0.40b, and tilting-pad bearing of
figure 10.40c are practically proof against half-speed whirl, the others will
raise the threshold speed, hopefully beyond the operating range.
The more heavily loaded the bearing, the higher the threshold speed, since
the higher the eccentricity the more closely the attitude angle approaches
zero (see figure 10.13). The problem is therefore acute in vertically mounted
spindles, where often there is negligible loading.

REFERENCES

I. Lord Rayleigh. Notes on the theory of lubrication. Phil. Mag., 35, No. 205, (Jan.
1918), 1-12.
2. L. Floberg. Lubrication of two cylindrical surfaces considering cavitation. Trans.
Chalmers University of Technology, No. 234, Institute of Machine Elements,
(1961 ).
3. D. Dowson. Investigation of cavitation in lubricating films supporting small loads.
Proc. Conf. on Lubrication and Wear. Paper 49, lnst. meek Engrs, (1957).
4. H. M. Martin. Lubrication of gear teeth. Engineering, 102, (1916), 199.
5. H. F. P. Purday. Streamline Flmr. Constable, London, (1949).
6. A. Sommerfeld. Zur hydrodynamischen theorie der schmiermittelreibung. · Z.
angew. Math. Phys., 50, (1904), 97-155.
7. 0. Pinkus and B. Stern Iicht. Theory of Hydrodynamic Lubrication. McGraw-Hill,
New York, (1961).
8. B. Jakobsson and L. Floberg. The rectangular plane pad bearing. Trans. Chalmers
University of Technology, No. 203, Institute of Machine Elements, ( 1958).
9. D. I. Hays. Plane sliders of finite width. Trans. Am. Soc. Lubric. Engrs, I, No. 2,
(1958).
10. M. Muskat, F. Morgan, and M. W. Meres. The Lubrication of plane sliders.
J. appl. Phys., II, (March, 1940).
II. D. Dowson and T. L. Whomes. Side leakage factors for a rigid cylinder lubricated
by an isoviscous fluid. Proc. lnstn me(k Engrs, 181, 30, (1967), 165-70.
12. D. D. Fuller. Theory and Practice ofLubricationfor Engineers. Wiley, New York,
(1966).
13. G. B. Dubois and F. W. Ocvirk. Analytical derivation and experimental evaluation
of short-bearing approximation for full journal bearings. NACA Rep. 1157, ( 1953).
14. G. Hirn. 'Surles principaux phenomenes qui pn!sentent les frottements mediats'.
Bull. Soc. ind. Mulhouse, 26, (1854), 188-277.
15. A. Kingsbury. Experiments with an air-lubricated journal. J. Am. Soc. nav. Engrs,
9, (1897), 267-92.
16. W. A. Gross. Gas Film Lubrication. Wiley, New York, (1962).
17. A. A. Raimondi. A numerical solution for the gas-lubricated journal bearing of
finite length. Trans. Am. Soc. Lubric. Engrs, 4, (1961), 131-55.
18. R. T. P. Whipple. Theory of the spiral grooved thrust bearing with liquid or gas
lubricant. Atomic Energy Research Est., Harwell, Berks., T/R 622, (1951).
19. R. T. P. Whipple. Herringbone pattern thrust bearing. Atomic Energy Re~arch
Est., Harwell, Berks., T/M 29, (1951).
20. H. Marsh. The stability of aerodynamic gas bearings. Ministry of Aviation Report,
(1964).

287
11
Elastohydrodynamic
Lubrication
11.1 HIGHLY LOADED CONTACTS

In the previous chapter we discussed the formation of a hydrodynamic film

of lubricant to support a normal load without examining the effects of the
size of this load or, more usefully, the value of the load per unit area. We now
look more closely at 'highly loaded' contacts, where loads act over relatively
small contact areas. Such contacts are to be found in the so-called 'line
contacts' of gear teeth and roller bearings and the 'point contact' of ball-
bearings. As the contact areas in the latter cases are typically only about one-
thousandth of those occurring in such situations as journal bearings, the
mean pressures will be about one thousand time greater. We may appreciate_
that such high pressures will affect the behaviour so that the hydrodynamic
solutions which were used to study journal and pad bearings will have to be
modified. Indeed we shall find that these high pressures can lead both to
changes in the viscosity of the lubricant and elastic deformation of the
bodies in contact, with consequent changes in the geometry of the bodies
bounding the lubricant film.
If, for example, we apply the hydrodynamic equations of the earlier
chapter to the contact between a pair of gear teeth with
w= 10 5 Njm '1 = 3.6 X 10- 3 Nsjm 2
U1 + U2 = 10 mjs R = 0.05 m

288
the predicted film thickness according to equation 10.12 is 0.044 J.lm. Since
it is virtually impossible to produce gear teeth smooth or straight to this order
of accuracy, this thickness of film will not prevent metallic contact of the
surface asperities thus producing disastrous wear of the contacting surfaces.
In practice the system operates quite satisfactorily with an adequate lubri-
cant film, which leads us to the conclusion that our simple hydrodynamic
theory is no longer applicable, calling for a re-examination of the assumptions
made in section 10.2.
The work of Bell 1 indicates that any non-newtonian behaviour of the
lubricant will have a detrimental effect on the film thickness so that, while in
some cases the lubricant will certainly cease to be newtonian, this can in no
way explain the increase in film thickness. Clearly side leakage will also act
to reduce the film thickness.
While we have seen in the last chapter that temperature rise through the
bearing (the thermal wedge) does introduce additional load capacity, we
must also remember that this effect is very small and cannot possibly account
for the discrepancy.

11.1.1 Notation
b Half the hertzian contact width
E Young's modulus
. 1 1 - vi 1 - v~
E' G1ven byE' = - E - + - £ -
' 2
Rolling friction per unit width
Sliding friction per unit width
Film thickness
Minimum film thickness in Martin analysis
h/hm
Thermal conductivity
Half length of the cylinder
Pressure

Po Maximum hertzian pressure = JWE'

rtR
q Reduced pressure
0.2W312
Maximum pressure in Martin analysis= (1JU)' 12 R
Pressure at which the viscosity is increased by a factor e (2. 718) = 1/a.
Radius
. U 1 + U2
Rollmg speed = 2
Surface velocities
Load per unit width
Pressure viscosity exponent

289
y Temperature viscosity exponent
17 Absolute viscosity
v Poisson's ratio
</J Ellipticality factor

11.1.2 Variable Viscosity

When we introduce the change of viscosity with pressure, we find a significant

increase in film thickness. The effect of variable viscosity has been examined in
section 10.6 for a lubricant with an exponential viscosity-pressure character-
istic, using the reduced pressure concept.
1
p = - log.(1 - cx.q) (10.21 repeated)
(X

Figure 10.14 shows that the pressure in the variable viscosity case rises more
quickly and reaches a higher value than for the isoviscous fluid and hence the
film thickness will be increased. Of course when q becomes equal to 1/cx. the
predicted pressure is infinite, but if we limit the solution to situations which
avoid this, we can assess the effect of variable viscosity. Blok 2 performed this
analysis using an exponential pressure-viscosity relationship and predicted
a film thickness two-and-a-halftimes that predicted by the isoviscous Martin
solution. Other people, using different relationships, have produced similar
values.
Although this effect is beneficial, it is insufficient to explain the discrepancy
between theory and practice. There must be yet another effect contributing
to the generation of larger film thicknesses than those predicted by Martin.

11.1.3 Elastic Deformation

Because the fluid pressure must be very large where high specific loads are
encountered, it is no longer reasonable to assume that the bearing surfaces
will remain undeformed. If we consider the case of gear teeth or rolling
element bearings, we may represent the contact by a plane and equivalent
cylinder as in section 10.4.1. Under these conditions the cylinder will flatten
locally at the contact so that it has the form illustrated in figure 11.1. The

...- , /HerfZIOn
1' \............- pressure
I \
I I

~·'
/? !/)~
Figure 11.1

290
parallel section will have the effect of increasing the load carrying area and
thus increasing the film thickness.
If we use hertzian theory to examine the deformation in the case of dry
contact, we find that the distribution of pressure normal to the surface is a
half-ellipse, having a maximum value (WEj1tR) 112 , the length of the flattened
portion being 8R(W/21t) 1' 2.
It is this resulting elastic deformation which allows films of practical size
in highly loaded contacts. This condition is called elastohydrodynamic
lubrication.

11.2 ELASTOHYDRODYNAMIC THEORY

To produce results of use to the designers of highly loaded bearings, we are

faced with the simultaneous solution of the Reynolds equation, the elastic
deformation equation and the equation relating viscosity and pressure. If
isothermal conditions no longer prevail, we must also incorporate the energy
equation for the film and the conduction equation for heat passing to and
from the solids. The cavitation boundary condition must be employed (see
section 10.4.2).
With the resulting complexity, it is hardly surprising that it was not until
the introduction of the high-speed digital computer that a full solution
became a practical proposition. However, with some inspired assumptions,
a very good approximate solution was obtained by Grubin 3 as early as 1949.
During the next decade solutions of limited range were produced by Petru-
sevich4 and Weber and Saalfeld 5 , but it was not until 1959 (when Dowson
and Higginson 6 obtained a solution of the inverse hydrodynamic problem
involving the solution of the Reynolds equation to give the geometry to
produce a specified pressure distribution), that the elastohydrodynamic
situation really yielded to analysis. Since that time many solutions of con-
stantly increasing range and accuracy have been achieved, notably by
Dowson and Higginson 7 •8 , Archard, Gair and Hirst 9 , and Herrebrugh 10 .
Inclusion of temperature variation increases the computation necessary,
but solutions have been successfully assomplished by Sternlicht, Lewis and
Flynn 11 , Cheng and Sternlicht 12 , and Dowson and Whitaker 13 . A solution
which allows for inlet heating due to shearing of the lubricant has been
produced by Greenwood and Kauzlarich 14.

11.2.1 Results

Film Shape and Pressure Distribution

Results from some of the foregoing analyses permit us to plot the shape of
the distorted cylinder and the pressure distribution for a given set of con-
ditions. Typically these will have the form illustrated in figure 11.2.
291
I
Pressure , - - , P
" \ Jl
I ,II
I ·.1
I ~
I I
I I,
r:
-
I ----
/i
.
Figure 11.2
If we consider first the film shape, we see the convergent inlet zone, followed
by a virtually parallel section. At outlet there is a constriction which can
amount to a reduction of up to 25 per cent of the parallel film thickness.
The pressure distribution is also illustrated. The curve shown represents
a low speed situation, where the pressure is very close to the hertzian 'dry'
distribution. There is a slight build-up before the parallel zone and towards
the outlet we see the pressure spike and also the cavitation boundary. The
spike can be shown theoretically to exist, although because it is extremely
narrow, experimental verification of its existence is very difficult. Its presence
is important because of the possibility of high subsurface stresses. Of course,
the area under the pressure curve must be the same as that under the hertzian
semi-ellipse.
If the speed is increased, the pressure distribution will depart more and
more from the hertzian as shown in figure 11.3.

"'.~;reasing
\IOCIIy

Figure 11.3 Effect ofvelocity on pressure

distribution 8

292
Film Thickness
The most important criterion for determining the success of the lubrication
of a contact is the size of the minimum film thickness. Therefore elastohydro-
dynamic analyses have usually been aimed at producing a formula which
gives the minimum film thickness in terms of the operating conditions. The
results are inevitably produced in numerical form, but an empirical fit can be
obtained to the results of Dowson and Higginson by a power law of the form
(11.1)
where
1 1- vi 1 - v~
E'=-E-+-E-
t 2

There are two rather surprising aspects of this equation: the film thickness is
not very sensitive to changes in load or elasticity. This can be explained if we
consider that the build-up of pressure occurs in the convergent inlet zone.
Increasing the load or decreasing the elastic modulus leads to greater
deformation, a longer parallel zone and hence a larger load carrying area with
approximately the same film thickness.
If we substitute the numerical v~ lues of our gear example into the equation
we now find that the predicted minimum film thickness is 0.5 Jlm, which is a
comfortably large practical film, bearing in mind the magnitude of surface
roughness and manufacturing inaccuracies.

11.2.2 Dimensionless Groups

It is often convenient to express the film thickness equation in terms of dimen-
sionless groups, but there are so many ways of doing this, each with its own
protagonists, that seldom will the reader find two publications which use the
same form. Only three such groups are necessary to describe any situation,
although Dowson and Whitaker in their original work use four. We will use a
system which is based on the work ofGreenwood 15 and Johnson 16 , since it
lends itself to an appreciation of the physical significance of each group.
We may express the relationship between the parameters in terms of a power
law using three dimensionless groups thus
Wh =A[(1JU)ti2R]a[(1JVRE')ti2Jb (11.2)
17U R tXW 312 W
where the number A and the indices a and bare constant for any particular
area of operation. This formulation may be modified, for purposes which will
become apparent, to
h/4.91]UR = B[~/0.2W3iJa(E')tb
W IX (1JU)ti2R 1tR (1JU)ti2R (11.3)

where the numerical constant has been adjusted appropriately.

293
If we examine the terms, we can readily identify four quantities
(1) The minimum film thickness for an isoviscous lubricant and rigid
solids. (The Martin solution, section 10.4.3.)
rJVR
hm=4.9w

(2) The maximum lubricant pressure in the Martin analysis

0.2W 312
qm = (rJU)II2R

(3) The pressure at which the viscosity is increased by a factor 2. 718 (e)
1
qa =-a

(4) The maximum pressure in dry hertzian contact

Po
= (WE')t/2
rtR

Equation 11.3 now may be written

e.h.l. film thickness
rigid isoviscous film thickness

= B x (pr~sue to in~ras.e vi~costy by e times)a

maximum ng1d Isoviscous pressure

(
maximum hertzian pressure )b
x maximum rigid isoviscous pressure
or in symbols
}!_ = B(qa)a(Po)b (11.4)
hm qm qm
The pressure ratio qJqm may be taken to indicate the effect of variable
viscosity on the film thickness, while the pressure ratio pJqm may be identified
as the effect of elasticity.
Unfortunately no universal values may be ascribed to B, a and b to cover
all operating conditions. We will therefore identify four distinct regions of
operation for which the appropriate values must be used.

/soviscous Lubricant, Rigid Solids

This is of course the Martin solution
a=b=O B=1
giving h/hm = 1.

294
Elastic Deformation much more Significant than the Pressure Viscosity
Effect
A solution for this case due to Herrebrugh 10 gives
a=O b = -0.8 B = 1.1
Thus the film thickness equation becomes
h
-=1.1-
(qm)O·B ( 11.5)
hm Po

Pressure Viscosity Effect much more Significant than Elastic Deformation

In this case, which is contrary to the previous one, it is the elasticity term
which disappears, giving values, according to Blok 17 , of
a= -f b=O B= 0.99
Therefore the equation which must be used is
!!_ = 0.99(qm)2/3 (11.6)
hm qa.

Both Elastic Deformation and Pressure Viscosity Effects are Important

This is the region for which Dowson and Higginson's results as stated in
equation 11.1 are appropriate. Rearranging equation 11.1 gives
a = -0.54 b = -0.06 B = 1.35
The film thickness equation therefore becomes
h _
- - 1.35-
(qm)0·54(qm)0·06
- ( 11. 7)
hm qa. Po
Clearly it is vital to use the formula appropriate to the operating conditions,
by identifying the particular regime in which they lie. This is best achieved
by means of a 'Johnsonchart' 16 , such as figure 11.4, in which lines of constant
film thickness are plotted on a graph having axes qmfqa. and qm!Po· Results
in each of the four regions described above have been used, together with
intelligent interpolation, to produce a map of elastohydrodynamic situations.
Approximate boundaries between the regions are indicated, We see that the
lines are horizontal in some places, indicating a dependence on qmfq" alone,
in other places are vertical, indicating that the film thickness depends only
on qmfPo• and between these two regions the full elastohydrodynamic solu-
tion pertains; both qmfqa. and qmfPo affecting the film thickness.
Therefore before attempting to analyse an elastohydrodynamic problem
it is essential to compute qmfqa. and qm!Po to determine in which regime the
system lies. It is then possible to identify which of the above formulae will
lead to the most satisfactory solution.
295
IO-,.~;_=¥<"'+k1
h=IOO I I I
/-I
Gr~b1n I ,I / "'\

----- -~ '7:/7:, /)
l
-.r~s,o
100"""=-+--------+-_-=..:_-=.
ii=20 I
J-1I I

~
Blok
___ l;,IO ? /_i· /
I
<)
1

I
/ ' "\

\ I
-
- /
h=5.L
'\I ', ~
1/
IOj=.:t-~"'vl,+
1\ I ~

ii=2 ! \ / /.'
Weber and Saalfel\ /
1
1~+-r
ii= I \ II

Marlin
;-+-----i
011----+------+-17<---t----+- +--+--+--
L~
/ /
ii·l /

I
01~-· 0·1 10 100 1000
Numen col soluflons ~,/Po
Extrapolated and
opprox1mote sotut1ons

Figure 11.4 Regimes of e/astohydrodynamic lubrication 16

11.3 COMPARISON OF THEORY AND EXPERIMENT

As so much attention has been given to the calculation of the film thickness
in elastohydrodynamic lubrication, it is not surprising that many efforts
have been made to confirm the theory by direct measurement of the film.
This has usually been done, not in the actual gear or bearing contact, but in a
disc machine with its simpler geometry and steadier running conditions.
The disc machine consists of usually two, but occasionally four, circular
discs which are loaded together, at least one being driven (see figure 11.5).
Even in this simple configuration the determination of film thickness is no
simple matter, since it is typically of the order of 1 Jlm or less. Many methods
have been employed, but the most successful have been the measurement of
the electrical capacitance of the film and the X-ray transmission technique.

296
w
w

(a) (b)

Figure 11.5 (a) 2-disc machine; (b) 4-disc machine

Capacitance measurement, which first yielded results in the work of

LewickP 8 , was refined by Crook 19 , incorporating a more accurate shape of
the deformed cylinders, and was used with great precision by Dyson, Naylor
and Wilson 20 in 1965.
The most significant results using X-ray transmission are those by Sibley
and Orcu(t 21 in 1961.
The general agreement between the theoretical and experimental results is
very satisfactory. Figure 11.6 shows the results by Crook and Dyson et a/
for the same oil showing excellent agreement with the theoretical predictions
of Dowson and Higginson.
We would expect that the lower limit of film thickness would be given by
li = 1.0, the Martin solution, but values well below this can be seen. This can
perhaps be explained by the inadequacy of the inlet boundary condition,
which assumes that p = 0 at a distance R before the centreline in the Martin
theory (see reference 11 of chapter 10).
The general picture can be seen more clearly on the 'Johnson' chart of
figure 11.7, in which certain values of film thickness are selected and plotted
with coordinates qmfqa and qmfPo· The order of the results is correct, although
the vertical scatter of the points appears to indicate that h is a function of
qmfPo rather than either qmfqa or qmfqa and qmfPo· Indeed Greenwood 22
points out that a good approximation is given by

!!_ = 2.81 (qm)0·8

hm Po
which, apart from the different numerical constant, agrees well with the
prediction of Herrebrugh (equation 11.5). Certainly if we plot the film thick-
ness against qmfPo the depedence is no less strong than against qmfqa. How-

297
I.e:; 9-("t
A ~0 .v.
0 + a Crook
0 101-
~
E
/ + Dyson Naylor

/
+
~ • and w11son

0 ••
a• "'
0 0
+
+ oowson and
H1gg1nson
~0 00
• 0
0 0

0 1 ' - - - - _ j _ ' - - - - - - ___ ______L _____ ·- -~'

0 2 10 10 2
Theoretical h

Figure 11.6 Comparison of experimental values of film

thickness with theoretical prediction 22

ever, this can be explained if we write out the three film thickness equations
11.5, 11.6 and 11.7 in terms of the practical quantities

h = 2.32(1JU)0 -6 R 0 · 6 w-o·zE'- 0 . 4 1X 0 (11.5 rearranged)

h = 1.66(1JU) 2 i 3 R 1i 3 W 0 £' 0 1X 213 (11.6 rearranged)

h = 2.65('JU) 0 "7 R 0 "43 W_ 0 _13 E' 0 . 03 1X 0 . 54 (11.7 rearranged)

The most appreciable differences between these equations are the indices
of E and IX. These are just the properties which most experimental workers,
using conventional steel and oil, have not varied and it is therefore hardly
surprising that over these ranges all equations give similar results.
The shape of the film was investigated by Crook 19 using capacitance
methods on a four-disc machine. The central disc was made of glass on which
was evaporated a thin chromium electrode. As the electrode passed through
the contact the potential was measured and a trace such as that in figure 11.8
produced. The essential features as predicted by theory are clearly shown.

298
IcY -

ii•l0·2
...•...
·~
10 - h•4 .. 08c--, r-- 1\
1>•2·04 •
.......

\
a,
•c
•
o.
0
•

-~ 01 I
··---
10
Q../Po

;; 1·02 2·04 4 08 10· 2

Dyson et ol
Crook
•
o
•c

Figure 11.7 Comparison of experimental and

theoretical values offilm thickness on the Johnson
chart 22

The pressure distribution is one of the most difficult measurements to

make. Kannel 23 deposited a thin strip of manganin on the surface of the
disc and, as this passed through the load carrying region, it produced the
records shown in figure 11.9. The local peak in the pressure curve just before
the rapid fall at outlet is taken as evidence of the existence of the pressure
spike. Even in this experiment the width of the pressure transducer was
several times the theoetical width of the spike.

Figure 11.8 Cathode-ray oscillo-

scope trace 19

299
Pressure

11111!!1 c=::l
Directton of rolling

Figure 11.9

11.4 TRACTION
In elastohydrodynamic lubrication, as in all lubrication mechanisms,
surface tractions will be present. If the two surfaces are moving with the same
speed (pure rolling in the case of cylinders), energy will be required to com-
press the lubricant as it enters the contact, the only possible source of this
energy being the motion of the surfaces. This will therefore produce a
retarding force on each of the surfaces, which we shall call the rolling
friction, FR.
If the surfaces are moving with different speeds then naturally the slower
surface will attempt to retard the faster, while the faster surface will exert an
equal force on the slower attempting to accelerate it. This force we shall call
the sliding friction, F5 • Hence the retarding force on each surface may be
written
for the slower surface F R - F s
for the faster surface FR + F5

11.4.1 Calculation of Frictional Forces

As in all lubricated contacts the tractions are produced by the velocity
gradients at the surfaces on which they act. Typical velocity profiles are
shown in figures 11.1 Oa and b.
In pure rolling the velocity gradient in the parallel region, while having a
non-zero value, is too small to be visible in the figure and is negligible for all
practical purposes, indicating that the major contribution to the rolling
friction comes from the convergent inlet region.
The frictional forces on the surfaces are

F1 = f (~ ::)y=o dx

F=2 - J(~ du)

dy y=h
dx

300
(a)

(b)

W1th shd1ng

Figure 11.10 Velocity distribution in

E.H.L. films: (a) pure rolling; (b) with
sliding

where the integrals are taken between limits which embrace the region of
significant friction. For isothermal conditions these equations become

The first term in each can be identified as the rolling friction since it remains
when u2 = u,.
The second term in each equation is the sliding friction. Since the value of
this integral is inversely proportional to h, the contribution of the entry zone
will be small compared to that of the parallel region and there will be little
loss in accuracy if the integral is rewritten

FS_(U2-U,)fd
- I] X
hp
integrated over the parallel region where hP is the parallel film thickness.
This may further be simplified

Fs (U 2 -U 1 )_
= I]
hp
301
where ij is an effective mean viscosity. An effective viscosity is necessarily
introduced, since the actual viscosity at any position will vary with the
pressure and the temperature.
A typical graph of friction-sliding speed for the faster surface is shown in
figure 11.11. Figure 11.11a showing constant rolling speed with different
lo~ds and 11.11 b showing constant load with different rolling speeds. There
are several points which we should note from the graphs.
( 1) At constant rolling speed F R is virtually unaffected by load, but rises
with increasing rolling speed. This confirms the correlation between
FR and hp since hp depends on W0 " 13 and U0 "43 •
(2) For low sliding speeds F 8 is proportional to sliding speed, indicating
that ij is unaffected by the small amounts of sliding.
(3) At constant rolling speed Fs increases with load. This is to a small
extent due to the reduction of hP with load, but to a much larger extent
to the increase of ij with pressure.
(4) At constant load F5 decreases with increasing rolling speed. This is
because of the increase in hP with U.
(5) As sliding is increased the traction rises to a maximum and with further
increase will fall drastically. This is accounted for largely by the decrease
in r; due to the temperature rise within the film resulting from the viscous
shearing of the lubricant. It might be expected that the decrease in
viscosity would promote a reduction in film thickness leading to a rise
in Fs to offset the above effect, but it must be remembered that the film
thickness is controlled by the inlet viscosity rather than the actual
viscosity within the contact.

Effective Viscosity

The heat generated in the contact is proportional to the product of the shear
stress and the rate of strain. Of the two ways in which it is carried away-
conduction to the surfaces and convection in the moving lubricant -Crook
has shown that, at least for metallic components, the conduction effect is the
predominant one 24 .
Dowson 2 5 develops an expression for r; assuming a hertzian pressure distri-
bution, conduction as the only means of heat transfer, and considering the
parallel region only

where y is the temperature coefficient of viscosity, K is the thermal con-

ductivity, a is the pressure coefficient of viscosity and 1J s is the viscosity of
the lubricant at the temperature of the surfaces at ambient pressure.

302
5

u1 ;Uz =1000 rev/m1n

E
2
.,:J 3
e- 45kN/m
{3.

21 kN/m

9 6 kN/m
2 8kN/m

05 10
Slip 2 !U,-u2 1
U1+Uz
(a)

Rollong speed
(RS)" U 1+Uz
2

0 02 04 06 08 10
Slip 2 (U,-Uz)
u,..-u2
(b)

Figure 11.11 Effect of load and speed on the traction

in rolling contact 32

303
This approximation is valid for values of
(U2-U1)2 1
l'fxY 8K ~

where l'fx is the viscosity at the temperature of the surfaces and at a pressure
appropriate to the value of x.

Non-newtonian Effects
Since it has been observed that thermal effects do not account entirely for
the reduction in traction at high rates of slip, additional rheological factors
must be sought. Indeed most lubricants are non-newtonian in that they pos-
sess shear elasticity in addition to viscosity. When the time taken to traverse
the contact is very small, the elastic deformation becomes significant com-
pared with the viscous deformation. In this case the apparent viscosity will
be larger than the viscosity as calculated earlier. This effect is important when
the transit time approaches the 'relaxation time' of the lubricant. The 'relax-
ation time' is defined by 1]/G, where G is the elastic shear modulus, and is the
time taken for the stress produced by a suddenly applied shear strain to fall
to 1/e of its initial value. For oils this is in the order of 10- 4 seconds 24 .

11.5 THREE-DIMENSIONAL SOLUTIONS

11.5.1 Side Leakage

We have so far confined our examination of the elastohydrodynamic problem
to two dimensions. The rollers of our analysis have been operating as
sections of infinitely long cylinders, permitting no flow or variation of
pressure in the axial direction. In practice we find that this does not result
in a great loss of accuracy for the cylinder and plane situation. The effective
load carrying region has approximately the width of the hertzian ellipse in
the direction of motion and therefore the ratio of the axial length-effective
width is very large, even though the nominal length-diameter ratio is small.
The way in which this affects the side leakage has been explained in chapter
10 (see figures 10.19a and 10.22b ).

11.5.2 Point Contact

In nominal point contact, such as that between a sphere and a plane, not
only is there variation of film thickness in a direction parallel to the rolling
axis, but also the effective load carrying region will be circular. In this case
therefore a new solution which permits flow and variation of the parameters
parallel to the rolling axis is necessary.

304
No complete solution of this case is available but Archard and Cowking 26
present an approximate solution of the form

~ = t. 4 (~Rr · 74(~r ·7 4(£:2r·07

where W is the total load on the sphere.
Experimental verification of this formula is not altogether conclusive.
27
Archard and Kirk's experiments suggest that

~ oc (cx~)o·s7uOR-03B

while Gohar 28 , using optical interference to measure the film thickness,

presents the empirical formula
h_
(~U
--1 .28---
R)o· 7(aw)o·49(ER2)o·I
R W R2 W
Interference methods offer an interesting picture of the film shape. Figure
11 .12 is taken from the work ofGohar and Cameron and shows the familiar 29
parallel region, which is now approximately circular, and the restriction at
exit which extends in a semicircle round the outlet side of the contact.
We have so far considered a sphere and a plane. For other point contact
situations, such as a ball rolling in a groove of different radius, the above
equations do not pertain. The new model will be in the form of a flat plane
and a body with different radii of curvature in the two principal directions,
Rx and RY, the area of contact being elliptical.

I Qron<Je
Purple

I Green
Yellow

/ Gr~:
Yellow

Figure 11.12 Interference pattern in E.H.L.

point contact 29

305
Archard and Cowking define an ellipticality parameter

<P= ( 1+-~ 2 R )- 1
3 RY
where Rx is the radius of curvature in the direction of motion and RY is that
measured parallel to the rolling axis. This permits the pressure in the point
contact to be related to that in the line contact by
q(point) = cjJq(line)
where the effective radius in the line contact is equal to Rx and the same load
per unit length is used .. Theory suggests that h oc ¢ 213 while experimental
results on a crossed cylinders machine indicate that h oc ¢ 0 ' 93 . For the case
of the sphere and plane Rx = RY and cjJ = 0.6.

11.6 FATIGUE FAILURE

For many years it has been observed that in the rolling lubricated contact of
metal surfaces under high load, failure takes place by a characteristic fatigue
process known as pitting. A crack starts at or near the surface and propagates
inwards at an acute angle to the direction of motion, resulting in the eventual
removal of a particle of the metal, leaving a characteristic pit on the surface.
Although the initiation of the cracks remains something of a mystery, it is
generally accepted that they propagate by means of a mechanism proposed
by Way in 1935 30 . A crack, once formed, is filled with lubricant. As the crack
passes through the contact zone, the very high pressures are communicated to
the liquid in the crack, producing stresses which propagate the crack. The
most important work in this field has been carried out by Dawson 31 . He
found that an increase in film thickness increases fatigue life and that for the
same film thickness the life is increased by a decrease in the surface roughness.
This work indicates the importance of good surface finish in suppressing
fatigue and emphasises the need to maintain continuous unbroken lubricated
films.

REFERENCES
l. J. C. Bell. Lubrication of rolling surfaces by a Ree-Eyring fluid. Trans. Am. Soc.
mech. Engrs, 5, 160.
2. H. Blok. Discussion. Gear Lubrication Symposium. Part I, The Lubrication of
Gears. J. Inst. Petrol., 38, (1952), 673.
3. A. N. Grubin and I. E. Vinogradova. Central Scientific Research Institute for
Technology and Mechanical Engineering, Book No. 30, Moscow. (D.S.l.R.
Translation No. 337), (1949).
4. A. I. Petrusevich. Fundamental conclusions from the contact-hydrodynamic theory
of lubrication. Izo. Akad. Nunk SSSR (OTN) 2, ( 1951 ), 209.
5. C. Weber and K. Saalfeld. Schmierfilm bei walzen mit verformung. Z. angew. Math.
Mech., 34, Nos. (l-2). (1954), 54.
6. D. Dowson and G. R. Higginson. A numerical solution to the elastohydrodynamic
problem. J. mech. Engng Sci., 1, No. I, (1959), 6.

306
7. D. Dowson and G. R. Higginson. The effect of material properties on the lubrica-
tion of elastic rollers. J. mech. Engng Sci., 2, No. 3, (1960), 188.
8. D. Dowson and G. R. Higginson. New roller bearing lubrication formula. Engineer-
ing, 192, ( 1961 ), 158.
9. G. D. Archard, F. C. Gair and W. Hirst. The elastohydrodynamic lubrication of
rollers. Proc. R. Soc., 262A, (1961), 51.
10. K. Herrebrugh. Solving the incompressible and isothermal problem in elasto-
hydrodynamic lubrication through an integral equation. Trans. Am. Soc. mech.
Engrs, 90, Series F, (1968), 262.
II. B. Sternlicht, P. Lewis and P. Flynn. Theory of lubrication and failure of rolling
contact. Trans. Am. Soc. mech. Engrs, 83, Series D, No.2, (1961), 213.
12. H. S. Cheng and B. Sternlicht. A numerical solution for the pressure, temperature
and film thickness between two infinitely long, lubricated rolling and sliding
cylinders, under heavy loads. Trans. Am. Soc. mech. Engrs, 87, Series D, (1965), 695.
13. D. Dowson and A. V. Whitaker, A numerical procedure for the solution of the
elastohydrodynamic problem of rolling and sliding contacts lubricated by a
newtonian fluid. Proc. lnstn mech. Engrs, 180, Pt. 3B, (1965-66), 57.
14. J. Greenwood and J. J. Kauzlarich. Inlet shear heating in elastohydrodynamic
lubrication. Trans. Am. Soc. mech. Engrs, Paper 72-Lub-21, (1972).
15. J. Greenwood. Presentation of elastohydrodynamic film thickness results. J. mech.
Engng Sci., 11, No.2, (1969), 128.
16. K. L. Johnson. Regimes of elastohydrodynamic lubrication. J. mech. Engng Sci.,
12, No. I, (1970), 9.
17. H. Blok. Discussion of paper by E. McEwen. J. lnst. Petrol., 38, (1952), 673.
18. W. Lewicki. Some physical aspects of lubrication in rolling bearings and gears.
Engineer, 200 (176), (1955), 212.
19. A. W. Crook. Elastohydrodynamic lubrication of rollers. Nature, 190, (1961) 1182.
20. A Dyson, H. Naylor and A. R. Wilson. The measurement of oil-film thickness in
elastohydrodynamic contacts. Proc. lnstn mech. Engrs, 180, Pt. 3B, (1965-66), 119.
21. L. B. Sibley and F. K. Orcutt, Elastohydrodynamic lubrication of rolling contact
surfaces. Trans. Am. Soc. mech. Engrs, A, No.2, (1961), 234.
22. J. A. Greenwood, A re-examination of elastohydrodynamic film thickness measure-
ments. Wear, 15, (1970), 281.
23. J. W. Kannel. Measurement of pressure in rolling contact. Proc. lnstn mech.
Engrs, 180, Pt. 3B (1965-66), 135.
24. A. W. Crook. The lubrication of rollers. IV. Measurements of friction and effective
viscosity. Phil. Trans. R. Soc., 255A, (1963), 281.
25. D. Dowson and G. R. Higginson. Elastohydrodynamic Lubrication. Pergamon,
London, (1966).
26. J. F. Archard and E. W. Cowking. Elastohydrodynamic lubrication at point
contacts. Proc. lnstn mech. Engrs, 180, Pt. 3B, (1965-66), 47.
27. J. F. Archard and M. T. Kirk. Lubrication at point contacts. Proc. R. Soc., 261A,
(1961 ), 532.
28. R. Gohar. Oil film thickness and rolling friction in elastohydrodynamic point
contact. Trans. Am. Soc. mech. Engrs, J. Lub. Tech., 371, (July, 1971).
29. R. Gohar and A. Cameron. The mapping of elastohydrodynamic contacts. ASME/
ASLE Lubrication Conf., Minneapolis, Preprint No. 66, LC21, (1966).
30. S. Way. Pitting due to rolling contact. J. appl. Mech., 2, (1935), 49.
31. P. Dawson. The pitting of lubricated gear teeth and rollers. Power Transmission,
30, No. 351, 208.
32. D. Dowson and T. L. Whomes. Effect of surface quality upon the traction character-
istics of lubricated cylindrical surfaces. Proc. lnstn mech. Engrs, 182, Pt. I, No.
14, (1967-68).

307
12
Hydrostatic Lubrication
12.1 EXTERNALLY PRESSURISED BEARINGS

We have seen in chapter 10 that hydrodynamic or self-acting bearings only

operate effectively if the two following features exist in their construction:
(a) a film thickness which varies in a preferred manner;
(b) relative tangential movement between the members of the bearing.
Externally pressurised, or hydrostatic, bearings originally attracted attention
because of their capability of working ·in the complete absence of these pre-
requisites, that is, they can operate with a uniform film thickness and no
relative tangential motion. Thus, for hydrostatic bearings both ohjox,
ohjoz and U and V may be put equal to zero in the Reynolds equation of
appendix A l which then reduces to the following form
azp azp
azx + oz2 = 0 (12.1)

This is Laplace's equation in two dimensions which must be solved for the
particular geometry of the bearing under consideration. This is clearly a
much simpler proposition than the original complete form of Reynolds
equation. Closed form solutions are easily obtained for a number of simple
shapes, such as long rectangular, circular, conical and spherical bearings 1 ,
since these may be described by one coordinate. More complicated shapes
may be tackled with series solutions or complex variable methods, but the
current widespread availability of digital computers enables numerical
solutions for practically any shape of bearing to be obtained in a fairly routine

308
manner. It is also of importance that great confidence can be placed in the
solutions obtained, since the boundary conditions in hydrostatic bearings
are usually well prescribed; for example problems of cavitation rarely occur.
It must be noted at the outset, however, that a practically useful solution of
Laplace's equation is only obtained if the pressure on at least one boundary
of a bearing is maintained higher than the pressure at the other boundaries.
This immediately implies that the lubricating fluid must be supplied to a
hydrostatic bearing at a comparatively high pressure from an external
source, whereas hydrodynamic bearings require only that fluid be supplied
at a pressure sufficient merely to ensure its presence in the bearing.
Although the matter will be pursued in more detail in the sections which
follow, it is obvious that the general level of pressure in an externally pressur-
ised bearing should increase as the load on the bearing increases. This in-
crease in pressure is, in the vast majority of cases, accompanied by a reduction
in the film thickness in the bearing so that a finite stiffness is exhibited. It
will be shown that it is usually necessary to place a control or compensating
device between the bearing and the source of high pressure lubricant, to
ensure that a satisfactory performace results. Clearly this means that the
behaviour of a given shape of hydrostatic bearing is determined as much by
the characteristics of the control elements which are used, as by the details of
its internal construction.

12.2 GENERAL DESCRIPTION OF HYDROSTATIC BEARINGS

A section through a typical hydrostatic thrust bearing is shown in figure 12.1.

The lower member ofthe bearing consists of two distinct regions, a number of
lands or sills which are separated from the upper member of the bearing by a
comparatively thin film of lubricating fluid, and a recess or pocket which has
a depth considerably greater than the film thickness over the lands. As
fluid is supplied to the recess at a high pressure it is clear that a particular
pressure profile exists over the whole area of the bearing, and that the

Load

<< << « < << l(

< < ( < < ( ( ((

Pump

(a) Constant flow system (b) Constant pressure system

Figure 12.1 Typical arrangements of hydrostatic bearing systems

309
integral of this pressure distribution must equal the load which is applied
to the bearing.
It is obvious that this pressure distribution can only be maintained if fluid
is supplied to the recess at a rate equal to the rate at which it escapes over
the lands of the bearing. The simplest way of supplying fluid to the recess of a
hydrostatic bearing is by means of a 'constant flow' system of the type shown
in figure 12.1a. In this arrangement a high pressure pump is assumed to
deliver fluid from a reservoir at a constant rate of flow regardless of the pres-
sure which exists in the recess. The general characteristics of this type of
system are discussed in some detail in section 12.4, but it is more usual in
fact for hydrostatic bearings to be operated in a 'constant pressure' system as
indicated in figure 12.1 b.
Two additional items are necessary for the successful application of this
system, a pressure control valve, and a compensating element. The purpose
and characteristics of the various types of compensating elements will not
be described until sections 12.5 and 12.6 respectively; it is sufficient to say
that they merely provide a resistance to flow. In this system the pressure con-
trol valve ensures that fluid is delivered to the compensating element at a
constant pressure regardless of the rate at which fluid flows through it.

12.3 VISCOUS FLOW THROUGH RECTANGULAR GAPS AND

CIRCULAR CAPILLARY TUBES

It was shown in section 12.1 that the shape of the pressure distribution in a
hydrostatic bearing operating with a uniform film thickness in the absence of
relative motion may be found by solving Laplace's equation 12.1. However, in
order to complete the analysis of any particular bearing it is necessary to
relate the pressure distribution in the fluid within the actual bearing itself to
either
(a) the flow rate of fluid delivered by the pump in the case of a constant flow
system, or
(b) the characteristics of the compensating element in the case of a con-
stant pressure system.
This may be achieved by deducing the rate of flow of fluid through the bearing
from its dimensions and the pressure distribution within it by means of
elementary viscous flow theory.
Referring to figure 12.2, consider the fluid film of uniform thickness hand
constant dynamic viscosity '1 which extends over an area specified by the x
and z coordinates shown. Let us assume that the pressure distribution in this
area has been determined so that the pressure profile in the direction of x
at a section specified by Z 0 is as shown, and consider the situation at the
point X 0 , Z0 • Here, the gradient of the pressure profile in the x direction

310
Penphery of
fluid film
Pressure prof1le
m plane of z 0

Figure 12.2 Viscous flow in a thin fluid film

due to pressure gradients

is (opjox)x 0 ,z 0 • The volumetric flow rate of fluid, (qx)x 0 ,z 0 , through the cross-
hatched area of height h and width c5z in the direction of x at the position
X 0 , Z 0 is given by viscous flow theory 1 as

(12.2)

The negative sign arises because the flow of fluid is always in the direction
in which the pressure decreases, at least for a stationary bearing. Similar
reasoning reveals that the volumetric flow rate of fluid (q.)x z in the direction
0
of z at the point X 0 , Z 0 is given by o.

Although the pressure gradients in the x and z directions are related to

each other at every point in the film by Laplace's equation, it is useful to note
that the rate of flow of fluid under viscous conditions in, say, the x direction is
dependent on the pressure gradient in the x direction but not on the pressure
gradient in the z direction.
It is convenient to introduce the properties of one of the most usual types
of compensating element in this section, since viscous flow theory 1 also
yields a simple expression for the volumetric flow rate of fluid through a
straight length of pipe or capillary of circular cross-section. It may be intui-
tively concluded that the rate of decrease of pressure along the length of the
pipe is uniform. The volumetric flow rate qc of a fluid of dynamic viscosity Yf
through a pipe of diameter d and length 11 when a pressure difference p2 - p 1
is maintained over the length of the pipe is given by

qc =
rtd 4 (P 2 - P1 )
U81J -,-, ---- (12.3)

311
The negative sign of equation 12.2 has been dispensed with here since the
direction of flow along the pipe is obvious.
These expressions will be used in the sections which follow, to pursue the
analysis and discussion of the characteristics of hydrostatic bearings.

12.4 LONG RECTANGULAR THRUST BEARINGS IN A

CONSTANT FLOW SYSTEM

In order to derive the important characteristics and appreciate the salient

features of a constant flow hydrostatic bearing system the very simple arrange-
ment shown in figure 12.3a will be considered in this section; the plane
upper member of this bearing has, of course, been omitted in this schematic
representation. There are two features which make this system particularly
easy to analyse, the absence of a compensating element and the fact that the
shape of the bearing leads to an extremely straightforward solution of
Laplace's equation, provided that the film thickness remains uniform.

/1/l//l//////

~F1Im thickness. II

Recess pressure. p,

(a) (b)

Figure 12.3 Long rectangular thrust bearing

It is first assumed that the length of the bearing is much greater than the
width of the lands in the direction of the x axis, that is, L ~ I. This leads to
the simplification that most of the fluid which is supplied to the bearing by
the pump leaves by flowing from the recess over the lands in the direction
of the x axis, and that comparatively little fluid flows out of the ends of the
recess in the z direction. We therefore ignore those parts of the bearing which
extend beyond the length of the recess. This is equivalent to assuming that
there is infinite resistance to flow in the direction of the z axis at the ends of
the recess.
It is also generally postulated in the study of hydrostatic bearings that the
pressure of the fluid is uniform over the whole area of the recess. This is
justified by the fact that as the depth of the recess in a hydrostatic bearing is
about one hundred times greater than the mean film thickness of fluid over

312
its lands, the resistance to the flow of fluid within the recess is much less than
that over the lands.
We now assume that a pressure Pr above ambient exists in the recess and
set up the x-z axes as shown in order to apply Laplace's equation to this
geometry. Following the above introduction it is evident that there can be no
change of pressure over the lands in the direction of the z axis, so that
equation 12.1 reduces to

which on integration becomes

p =Ax+ B
where A and B are constants. The boundary conditions for this expression
are that p = Pr at x = 0 and p = 0 at x = I; substitution of these in the
above gives

A = - !!.I_ and B = Pr
I
so that the pressure distribution over both lands is given as

This is immediately recognised as a linear or uniform pressure gradient, so

that
0P Pr
ox I
and we are now able to use equation 12.2 to determine the rate of flow of fluid
from the bearing. Thus we put

(:~)X 0 .Z 0 -1
=

since this does not vary with x, and determine the volumetric flow rate
through an elemental channel of width (Jz as

q
X
= {-h3 (-!!.I_)}
12'7 I
(Jz

Since the expression in parenthesis is not a function of z we may integrate the

flow rate qx between z = 0 and L to obtain the total flow rate of fluid out of
one side of the recess as
Now we make the important statement of continuity that twice this quantity
must equal the total rate of flow of fluid into the bearing from the pump, thus

Q = 2 h3Lpr
12'11
or

(12.4)

This is an important expression since it relates the recess pressure to the

other parameters of the bearing. Since the pressure distribution in the bearing
is now known to be of the form shown in figure 12.3b, it may be easily inte-
grated to yield the total load, W, supported by the bearing. The load sup-
ported by the incremental length of land bx is simply pL bx, so that the load
supported by the whole bearing is
I

P.bL + 2 fP.(I- y)L dx

0

The first term in the above expression represents the load supported by the
uniform pressure in the recess, while the second represents the load carried
by the lands. Completion of the integration yields the load capacity of the
bearing as
W= P.L(b +I)
and eliminating Pr from this expression with the aid of equation 12.4 gives

W = ~ (t)[L(b + /)] (12.5)

The physical interpretation of these relationships is that because the bear-

ing is supplied at a constant rate, the recess pressure must rise to overcome the
increased resistance to flow which is brought about by a decrease in film
thickness.
The rate of change of load capacity with film thickness is known as the
stiffness of the bearing and is an important parameter which may be evalu-
ated by differentiating equation 12.5 with respect to h.

aa: = - ~:t (t)[L(b + l)J (12.6)

The negative sign indicates that the load capacity increases as the film
thickness decreases, and that the stiffness increases very rapidly with reduced
film thickness because of the presence of the fourth power of the film thickness
in the denominator.

314
The expressions obtained here are very simple but in practice greater
complication arises because of the inability to operate more than one bearing
from a constant flow system and from the need to use bearings of more com-
plicated shape. The first of these complications will be discussed in the
following two sections, while the second will be considered in section 12.8.

12.5 THE NEED FOR COMPENSATION IN MULTIBEARING

ARRANGEMENTS

It is clear, from studying the bearing geometry and pressure distribution

shown in figure 12.3, that a uniform film thickness can only be maintained if
the line of action of the applied load is perfectly coincident with the centreline
of the bearing. This limitation does not greatly affect the hydrostatic theory
which has been presented in section 12.4, but it is obviously of great practical
importance, since it implies that a simple hydrostatic thrust bearing of the
type so far discussed provides little, if any, resistance to tilting. In figure
12.4a the upper member of the bearing is shown in an inclined attitude
following the displacement of the line of action of the load to the right of the
bearing's centreline. The film thickness over the left-hand half of the bearing
therefore diverges so that the resistance to flow decreases progressively
across the width of the land. Without resorting to the analysis involved in the
solution of this problem, it may be deduced that the pressure gradient de-
creases progressively over the left-hand land. By similar reasoning the pres-
sure gradient over the right-hand land must therefore increase with the result

r
that the pressure profile shown in figure 12.4b is produced in the bearing.
Centre line

,.~
-4:>.»>>>>.>,
j :
on~• m •ooo

I )), \~
Reduced res.stoncP ___- ---
to flow ~ .,~'71 ------Increased res1stonce
to flow

It
'
Constant
flow

(a) M1sa11gned beanng

Decras•ng~u
gradient Increasing pressure
gradient

(b) Resultant pressure profile

Figure 12.4 Effect of non-uniform film thickness on

pressure distribution in bearing

315
The modified pressure distribution shown in figure 12.4b provides a
moment about the central longitudinal axis of the bearing tending to oppose
the moment due to the offset load. Unfortunately the tilting capacity which
is developed by this means is normally very small in bearings of practical
proportions, and it is therefore usual to employ bearings in pairs to support
offset loads, as shown in figure 12.5.
For this arrangement to provide a significantly large resistance to tilting
it is clearly necessary for the pressures in the recesses of the two bearings to be
different, and in the case shown p2 must be greater than p 1 • This condition

Recess pressure. Pz

Pump

(a) 1nd1v1dual pumps. Pz>Pi

~, > ,,,,,, ~
1 Load

Pump

(b) $1ngle pump. Pz =p,

Figure 12.5 Multibearing systems

may be easily realised by using two separate constant flow systems of the
type previously dealt with, as shown in figure 12.5a. Assuming that the actual
variation of the film thickness in each bearing will now be small because of the
enhanced tilting resistance of the arrangement, ~t follows from equation 12.5
that p 2 will in fact be greater than p 1 , as the average film thickness in the right-
hand bearing will be less than that in the left-hand one. Thus both the load
applied to the bearings and the moment produced by the displacement of its
line of action from the centreline of the bearings, will be supported by this
arrangement.

316
It must be noted that merely connecting a single constant flow pump to the
two bearings, as shown in figure 12.5b, does not produce an effective arrange-
ment because the pressures p 1 and p 2 would always be equal to each other
regardless of the variation of film thickness in the bearings, provided of
course that the resistance to flow in the conduits which supply the recesses is
small.
Although the principle of using a separate constant flow pump to supply
each bearing in a multibearing system is technically sound, it suffers from
the obvious disadvantage that a high-pressure pump must be provided for
each bearing.* This solution to the problem of carrying offset loads may
be justifiable in installations which include bearings of large dimensions,
but in machinery of more modest size it is both practically and economically
out of the question for there to be as many pumps as bearings. A method has
been devised which makes it possible to operate a number of bearings from a
single high-pressure pump by introducing compensating elements into a
constant pressure, rather than a constant flow, system. This important
modification is discussed at length in the next section.

12.6 CHARACTERISTICS OF COMPENSATED BEARINGS

The fundamental features of a constant pressure system used in conjunction

with a single hydrostatic bearing have already been illustrated in figure
12.1 b. Following the discussion in the previous section it will be appreciated
that a more typical arrangement would involve a number of bearings as
shown, for example, in figure 12.6a. In this figure three separate thrust
bearings, which need not necessarily be of the same size as shown, are supplied
with fluid from a single high-pressure pump through individual compensating
elements. It is important to note that the pressures in the three recesses may
assume different values if the upper member of the bearing becomes inclined;
this contrasts with the case shown in figure 12.5b and will be explained in due
course.
In this section the principle of operation and an elementary analysis of a
single compensated bearing of the long rectangular shape treated in section
12.4 will be considered as part of a constant pressure system. First, it is clear
from the discussion in section 12.4 that the pressure in the recess of the
bearing shown in figure 12.1 b must increase as the film thickness decreases,
even though fluid is delivered to the compensating element at a constant
supply pressure. Now, the flow resistance provided by the compensating
element must be overcome by the pressure difference which is established
across it. From considerations of continuity the rate of flow of fluid through

* It is in fact possible to construct control valves and mechanical flow dividers which ensure
that fluid is supplied at a rate which is more or less constant. Any number of these devices may
therefore be connected between an equal number of bearings and a single high pressure pump.
These then act effectively as constant flow sources.

317
o Load

<,(~Jl
Recess

Supply pressure. p,

(a) A mult1beonng, compensated. constant -pressure system

1n wh1ch P3>P2 >P1

<<<<t<<<<t<<< _l
Th
Recess pressure Pr

Supply pressure, p, Compensarong

elemenr

(b) Schemaric diagram of a compensared bearong

Figure 12.6 Compensated bearings

the compensating element must equal the rate at which fluid flows out of the
recess. It will be shown below that the way in which the recess pressure, film
thickness and characteristics of the compensating element are related
accounts for the satisfactory operation of a constant pressure hydrostatic
bearing system.
Compensating elements, which are also known as control elements or
restrictors, are usually, but not exclusively, of two types
(a) capillary tubes
(b) orifices.
The characteristics of the former element have already been described in
section 12.3 and therefore in the first instance we will consider the analysis
of the long rectangular bearing of figure 12.3 in a capillary-compensated
constant pressure system.

12.6.1 Capillary Compensation

This system is shown schematically in figure 12.6b where the pressure drop
across the capillary tube is clearly Ps - p,, and it therefore follows from

318
equation 12.3 that the rate of flow of fluid through it is

nd 4 (~)
1281] 1,
Similarly regardless of the way in which fluid is supplied to the recess, it
follows from section 12.4 that fluid flows out of it at a rate
h3Lpr
61]1

By continuity these two rates of flow must be the same and equal to the rate of
flow which is necessary to operate the bearing, hence
nd 4 h3 L
1281]/, (ps - Pr) = 61]/ Pr

or

( 12.7)

where
K = 3nd 4 /
64/,L
Having thus determined the recess pressure in terms of the bearing's
dimensions, its load capacity may now be found from the expression develop-
ed in section 12.4, that is
W = PrL(b +I)

substituting for Pr from equation 12.7 gives

W = _p._L_(b--=------+_1)_ (12.8)
(1 + h 3 /K)

This expression indicates that the load capacity of the bearing does in
fact increase as the film thickness decreases. The physical explanation of this
behaviour is that the pressure drop across the capillary decreases to maintain
continuity of flow as the recess pressure increases in response to a reduction
in the bearing's film thickness. This same principle is used very effectively
in fluid gauging equipment and in the construction of pneumatically-
operated process controllers.
Following the procedure of section 12.4 we differentiate expression 12.8
with respect to h to evaluate the stiffness of the bearing

aw = - K3
ah (
1 + h3 /K
h )2
p.L(b + l) (12.9)

319
Now, both of the expressions for Wand oW/oh assume a more compli-
cated form than the corresponding ones which were derived for a constant
flow system in section 12.4, but certain common features may be discerned,
for example, the presence of the factor L(b + l). This observation will be
pursued in section 12.8.
Before considering the characteristics of compensated bearings further, it is
necessary to introduce the method of dimensionless presentation, which is
conveniently and conventionally used in the discussion of hydrostatic bear-
ings. Although this refinement could have been avoided in view of the sim-
plicity of the bearings under discussion, it is nevertheless introduced here
since a knowledge of it is desirable when more complicated systems are
considered.
We may define the ratio of the recess pressure, p., to the supply pressure,
p., as a dimensionless recess pressure, p, and the ratio of the film thickness,
h, to some datum or design thickness, h0 , as 1i so that expression 12.7 be-
comes, in dimensionless form
- 1
(12.10)
p = 1+ (h~/K)li 3

If the bearing normally operates at the design film thickness h0 then the
value of the dimensionless recess pressure p is an important parameter
which is known as the pressure ratio r of the bearing, that is, the ratio of the
recess pressure to the supply pressure at the design film thickness h0 is the
bearing's pressure ratio r. Thus putting p = r and 1i = 1 for these conditions
in expression 12.10 gives
I
r = ----=--
1 + (h~/K)
from which
1- r
(h 3 /K) = - -
o r
( 12.11)

Substituting for (h~/ K) from equation 12.11 in equation 12.10 gives

(12.12)

This is an important general expression which relates the dimensionless

recess pressure of any capillary compensated bearing, not just one of the
shape shown in figure 12.3, to its dimensionless film thickness for given values
of pressure ratio.
It may be seen from figure 12.3b that the load capacity of the bearing will
attain its greatest value when the pressure in its recess equals the supply
pressure, consequently it is convenient to define a dimensionless load, W,

320
as the ratio of the load actually carried by the bearing under a given set of
conditions to the load it would carry if the recess pressure rose to the supply
pressure, that is, the maximum load. We know that the load capacity, W,
is given by
W = p,L(b +I)
and remembering that the rec.ess pressure may be derived from expression
12.12 it follows that

(12.13)

It is of interest that the dimensionless recess pressure given by expression

12.12 is identical with this value of dimensionless load.
Similarly, it is possible to express the stiffness of the bearing in dimension-
less form by dividing both sides of expression 12.9 by the factor

which has the dimensions of stiffness; the dimensionless stiffness aw;ali

then follows as
aw
aw oh
ar (p.L~o+ /))

~ _ Jh;ho X (, + (~)n' r
putting h2 = h;li 2 and substituting for h~/K from equation 12.11 yields the
dimensionless stiffness as

(12.14a)

Since it is understood that the load capacity of a bearing increases with

decreasing film thickness the negative sign in the above expression is often
discarded by defining the absolute value of the dimensionless stiffness,
S, asS= -oWjoli. Similarly the dimensional form follows from equation
12.9 asS = -oWjoh.
It is of interest to use expression 12.12 to eliminate the pressure ratio of

321
the bearing r from expression 12.14a to yield the following alternative expres-
sion for the stiffness of the bearing
aw 3
----ar = - Fz p(l - p) (12.14b)

It must be noted that in this expression the dimensionless pressure p

is the pressure which exists in the recess at the particular value of dimension-
n
less film thickness for which the stiffness is being evaluated. This dimension-
less pressure is of course found from expression 12.12 in each case.
Finally, the total flow rate necessary to operate the bearing at any value of
film thickness may be expressed non-dimensionally as a proportion of the
rate of flow which would occur if the recess pressure became equal to the
supply pressure at the design film thickness. As the flow rate Q in the general
case is given by

and Q has been defined as

Q /(hi~f·)
substitution of the above value of Q gives Q = pn 3 but p is itself a function
n
of as given by equation 12.12, thus

(12.15)

It is interesting to note that this expression reveals the dimensionless flow

rate at the design film thickness to be equal to the pressure ratio of the bearing,
that is, Q = r when h = h0 • This agrees of course with the definition of Q.
It has been shown that the important parameters of a capillary compen-
sated bearing of simple shape may be expressed in dimensionless form. It is
worth recalling at this stage that the dimensional values of these parameters
are easily determined in a given case as follows

recess pressure Pr = (p.)P

load capacity W = (p. L(b + l)) W
. (p.L(b + l)\c
stiffness s=-
aw
oh = ho r
flow rate
3
Q = ( h Lp 6 )-
111 • Q (12.16)

322
It is clear from the foregoing that the compensating element of a constant
pressure system simply provides a resistance to flow between the pressure
control valve and the bearing's recess. Because the flow through a capillary
compensator is proportional to the pressure drop across it (expression 12.3),
it is recognised as a linear device. Although the rate of flow through an
orifice is proportional to the square root of the pressure difference across it,
hence its classification as a non-linear element, it too finds application as a
compensating element. The characteristics of an orifice-compensated bearing
are considered in the next section.

12.6.2 Orifice Compensation

It is shown in textbooks on fluid mechanics 2 that the rate of flow of an in-

compressible fluid through a short, sharp-edged orifice of diameter d and
discharge coefficient cd is
'!!_ C d2(2(p2 - p,))''2
4 d p

where p is the density of the fluid and (p 2 - p 1 ) the relevant pressure dif-
ference. If the capillary compensator which has been considered as the
compensating element in the bearing of figure 12.6b is now replaced by an
orifice then the condition of continuity of flow becomes

'!!_ C d2(2(p. - P.))''2

4 d p

The investigation ofthis bearing may be facilitated by analysing it from the

outset in terms of dimensionless quantities, thus in the nomenclature of the
previous section we put Pr = PPs and h = 1ih0 in the above expression, which
then reduces to
(12.17)
where, in this case
K = 91t2 (Cd~2'1l)
2pps h0 L
The dimensionless recess pressure is obtained by evaluating the positive
root of the equation for p which is explicit in (equation 12.17)

This expression is comparable to equation 12.10 and may be further general-

ised by using the fact that p must be equal to the pressure ratio of the bearing

323
r when the dimensionless film thickness li is unity. Substituting these con-
ditions in equation 12.17 gives

K=--
'2 (12.18)
1- r
and the expression for the dimensionless pressure ratio becomes

p= 2 ~6 c~ ,)[(1 + 4C ~ ')n Y' 1]

6
2
_
(12.19)

It is again useful to note that this expression is true for all orifice controlled
bearings and not merely for the kind shown in figure 12.3b.
Consideration of the manner in which the dimensionless stiffness of a
capillary-compensated bearing, expression 12.14a, was derived reveals that
o Wjoli = opjoTi and it follows that the dimensionless stiffness of an orifice-
compensated bearing is obtained by differentiating equation 12.19 with
respect to 1i

(12.20a)
This is a much more complicated expression than the equivalent one for a
capillary-compensated bearing given in expression 12.l4a but it too may be
conveniently simplified by first removing the expressions under the square
root signs by using equation 12.19. After some further simplification the
above expression reduces to

aw __ ~
ali -
-(1 - P)
1i P 2 - P (12.20b)

As in expression 12.14b the pressure p in this expression must be deter-

mined from expression 12.19 for the particular film thickness Fr under con-
sideration.
The dimensionless flow variable Q for an orifice-compensated bearing is
defined in exactly the same way as for a capillary-restricted bearing. From
equation 12.19

Q= 2~3 C~ ,)[(1 + 4C ~ ')n6Y'2- 1] (12.21)

Since the dimensionless parameters p, o Wjoli and Qhave been determined,

the dimensional values of Pro W, oW joh and Q for an orifice-controlled
bearing may be evaluated immediately from expression 12.16. This is true
even though the expressions of 12.16 were derived initially for a capillary-
controlled bearing.

324
12.6.3 Constant Flow Compensation

Although it may seem that a constant flow system of the type discussed in
section 12.4 is not describable in terms of a supply pressure, it is usual for the
delivery pressure of the pump which operates it to be limited to some value,
say p,. Thus, if the recess pressure in the bearing's recess is Pr at some oper-
ating film thickness h then, as before, a dimensionless recess pressure may be
defined as PriPs and a dimensionless film thickness as hjh 0 , where h0 is
the film thickness under design conditions. Now from expression 12.4 we
may write
- ( Q1Jl ) 1
p = 6 Lp,h~ Ji3
The pressure ratio of the bearing may again be defined as the ratio r of
the recess pressure to the supply pressure at the design film thickness, that is,
when li = 1. Thus from the above

r = ( 6Q1Jl) (12.22)
Lp,h~
and
r
p = Ji3 (12.23)

The dimensionless stiffness of the bearing then follows immediately

from equation 12.23 as
aw 3r
(12.24a)
ali fi4
or in terms of the recess pressure
aw 3p
ar --,; (12.24b)

It is, of course, a trivial matter to show that the dimensionless flow rate Q
is a constant and is given by
(12.25)
The dimensional values of this bearing's parameters may be found by
use of expressions 12.16.

12.7 COMPARISON OF CHARACTERISTICS

It is significant, and of course convenient, that the expressions for the
recess pressure, load capacity, stiffness and flow rate which were derived in
dimensionless form in the previous section for capillary, orifice and constant
flow compensated bearings are functions of only two variables, the dimen-
sionless film thickness li and the pressure ratio r. In this section these dimen-
sionless parameters will be discussed at some length to ensure that a clear

325
appreciation of the characteristics of the various types of compensation is
gained. Because the dimensional values of the variables of interest may be
obtained for any particular bearing of the long rectangular shape which has
so far been considered, by use of expressions 12.16, this discussion will be
presented entirely in terms of the dimensionless variables and, to avoid
repetition, the adjective dimensionless will be dispensed with in this section.

12.7.1 Choice of Pressure Ratio

Because the stiffness of a bearing determines the deflection which occurs in a

machine following the imposition of a load, it is recognised as one of the most
important parameters. It is therefore reasonable and convenient to assume
that it is desirable to maximise the stiffness of a bearing at some particular
combination of operating conditions. Now as the stiffness of a capillary
compensated bearing is given by equation 12.14b it follows that for a given
film thickness the stiffness is a function of only the recess pressure. By dif-
ferentiating this expression with respect to p and equating to zero it is
revealed that the stiffness will be a maximum when p = 0.500. Substitution
of this value of p in expression 12.14b gives the maximum stiffness of a
capillary compensated bearing at the film thickness fi as

s max
= 0.75
Ji (12.26)

Similarly, differentiating expression 12.20b with respect top shows that the
stiffness of an orifice-compensated bearing is a maximum when
p2 - 4p +2= 0
The positive root of this expression is 2 - (2) 112 = 0.586 and substitution
of this value of pin expression 12.20b gives the following expression for the
maximum stiffness of an orifice compensated bearing at the film thickness fi
- 1.0296
smax= -Ji- (12.27)

It is clear from expression 12.24b that the stiffness of a bearing with con-
stant flow compensation does not exhibit a maximum, but increases mono-
tonically with recess pressure.
Without sacrificing too much generality, we may now fix the film thickness
at unity so that the preferred values of recess pressure derived above then
become the pressure ratios r of the bearing. By evaluating the stiffness of a
capillary, orifice and constant flow controlled bearing for various values of
pressure ratio from expressions 12.14b, 12.20b and 12.24b respectively, the
curves of figure 12.7 may be drawn. The maximum values of the stiffness of
capillary and orifice compensated bearings are shown at the preferred
values of pressure ratio and it is noticeable that the stiffness of both types of

326
14 Constant flow Maximum value is
S•3r I 0296 at r• 0 586
12

10 Onf1ce
Ill)
,; 08 S•6r(~-:_')
~
~ 06

04
Capillary
02 S•3r(l·r)

0
0 02 04 0·6 08 10
Pressure ratio, r

Figure 12.7 Effect of pressure ratio on the stiffness of a

simple hydrostatic thrust bearing with constant flow,
capillary and orifice compensation

bearing is relatively insensitive to variations in pressure ratio of, say, ±0.05

about these values. The stiffness of a bearing with constant flow compensa-
tion changes rapidly with pressure ratio as mentioned earlier.
An important practical observation is that under the most favourable
conditions the stiffness of an orifice controlled bearing is about 33 per cent
greater than that of the best equivalent capillary-controlled bearing, but of
course the higher recess pressure necessary increases the flow rate of fluid
required to operate the former bearing.

12.7.2 Effects of Changes in Film Thickness

In the last section the performance of the various types of bearing was con-
sidered and the effects of varying the pressure ratio while the film thickness
was maintained constant at the design value of unity resulted in the curves of
figure 12.7 being drawn and discussed. Assuming that the preferred values of
pressure ratio of 0.5 and 0.586, for capillary and orifice-controlled bearings
respectively, would be used at the design film thickness of unity, it is of interest
to study the behaviour of the bearings when the film thickness departs from
the design value in response to changes in loading.
First considering a capillary-controlled bearing which operates at a
pressure of 0.5 under design conditions, the variation of recess pressure and
hence load capacity may be obtained by putting r = 0.5 in expressions 12.12
or 12.13. The curve for p and W drawn in figure 12.8a then results, and it is
clear that the recess pressure rapidly approaches the supply pressure once
the film thickness decreases below, say, 0.6. Alternatively this observation
may be interpreted as indicating that a high proportion of the maximum
load capacity of the bearing is developed once the film thickness becomes less
than about 0.6.

327
14

12

10
~
u 08 5. ~ 0 75/ii
c:
0
IV) ~ ........ / mo•
IQ" 06
...... _ _ Swhenr~05atfiI·O

04
--
0·2

0~-r,
0 0 2 04 0 6 0 8 I0 I 2 1-4 I6 I8 20
1i
(a) Capillary campens!i.r~0·5 a! 1?~·0

1-4

1·2 s~ I 0296 a! i~l 0

10
I~
u 08
c:
0
IV)
IQ" 06

04

02

o+-~.,r
0 0 2 04 0 6 0 8 I0 I 2 1-4 I 6 18 20

(b) Orrfice compenso!ron, r-0 586 of i~ 1·0

30
- - - r ~0·25 a! 1~ I 0
2 5 --- r~o 333 at i~1 0

1~ 20
u
c:
0 15
kl)
rQ"
10

05

o~-r
0 02 04 06 08 1·0 12 1·4 16 1·8 20
j,
(c) Constant flow compensation, r~O 25 and 0·333 a! i~1 0

Figure 12.8 Characteristics of compensated bearings

328
The way in which the stiffness changes with film thickness may be studied
by putting r = 0.5 in expression 12.14a which then becomes

By substituting various values of 1i in this expression the curve of S drawn

in figure 12.8a is obtained. Naturally, as given by equation 12.26, the stiffness
has a value of0.75 at 1i = 1, but it is of importance that as the film thickness
decreases from this value the stiffness reaches a maximum before decreasing
rapidly as the film thickness approaches zero. By differentiating the above
expression with respect to 1i and equating to zero, the value of 1i at which this
maximum occurs is found to be 0.7937 (actually (!-) 113 ). Substitution of this
value of 1i into the above expression gives the value of the maximum stiffness
as 0.837, but this behaviour will only arise in a bearing which has a pressure
ratio of0.5 at a film thickness of unity.
It is of interest to note that the value of the recess pressure under these
conditions as found from equation 12.12 is 0.666 (actually 2/3). This figure is
often incorrectly quoted in the literature as the value of pressure ratio which
yields the maximum value of stiffness for a capillary-compensated bearing;
this anomaly is explained below.
Consideration of expression 12.26 shows that a hyperbolic relationship
exists between the maximum value of stiffness ( -oWfoh)max• and the film
thickness li. This relationship is also plotted in figure 12.8a, but it must be
realised that a bearing only exhibits this characteristic if its compensating
element is continually altered to ensure that the recess pressure assumes a
value of0.5 at all values of.film thickness. Thus the true maximum value of the
stiffness attainable at the film thickness of 0.7937, or for that matter at any
other film thickness, is given by expression 12.26. On substituting 1i = 0.7937
in this expression the maximum stiffness at this film thickness is shown to be
0.944. This is of course greater than the value of 0.837 which would be ob-
tained at this particular value of film thickness when the bearing's compensat-
ing element is chosen merely to give a recess pressure of0.5 at a film thickness
of unity only.
Since there is considerable confusion in the published treatments and
explanations of the characteristics which have been plotted in figure 12.8a,
it is worth summarising the above discussion as follows: if the recess pressure
of a capillary-compensated bearing is adjusted to be 0.5 at all values of film
thickness, then its stiffness is given by the hyperbolic relationship of ex-
pression 12.26, that is, the stiffness increases rapidly as the film thickness is
reduced. However, if the recess pressure is chosen to be 0.5 at a film thickness
of unity, say, the stiffness of the bearing is described by expression 12.14a and
a reduction in film thickness will at first cause an increase in stiffness until a
maximum is reached at a film thickness of 0.7937, but further decreases in
film thickness result in a rapidly decreasing stiffness. In all cases the greatest

329
possible stiffness which can be developed by a capillary-compensated bearing
is that given by the hyperbolic relationship of expression 12.26.
In the case of an orifice-compensated bearing a similar approach may be
made and the variation of recess pressure and load capacity with film thick-
ness may be investigated, for example, by putting r = 0.586 in expression
12.19. The curve for p and W drawn in figure 12.8b then results, and it is
noticeable that as the film thickness decreases a large proportion of the
maximum load capacity of the bearing is developed even more rapidly than
in the case of a capillary-controlled bearing.
The variation of stiffness with film thickness may be determined by putting
r = 0.586 in expression 12.20a, or by using equation 12.20b in conjunction
with equation 12.19, and the curve of S drawn in figure 12.8b results. Again a
maximum in this curve is reached before the stiffness decreases rapidly with
film thickness for values of li less than, say, about 0.8. Differentiation of
expression 12.20a with respect to Ti, with r put equal to 0.586, yields an in-
tractable expression for the value of Ti at which the maximum of the stiffness
curve occurs. However, a study of figure 12.8b reveals that this occurs at a
film thickness of approximately 0.9 and that the maximum in stiffness is
about 1.09.
The recess pressure under these conditions may be found from expression
12.19 as 0.691, a figure which similarly appears incorrectly in the literature as
the value of pressure ratio which produces maximum stiffness in an orifice-
compensated bearing.
Consideration of the hyperbolic relationship of expression 12.27 shows that
the maximum stiffness smax attainable at this film thickness is slightly higher
at 1.148.
In parallel with the discussion of the behaviour of a capillary compen-
sated bearing it is now clear that the maximum stiffness smax of an orifice
compensated bearing will only be attained if its recess pressure is con-
tinually adjusted, as the film thickness varies, to the value of 0.586. If the
compensating element is merely chosen to yield a pressure ratio of 0.586
at a film thickness of unity, then the stiffness will increase to the maximum
of 1.090 as the film thickness decreases to about 0.900, but if the recess pres-
sure were then adjusted to 0.586 at this film thickness, the stiffness would
increase to the value of 1.148 given by the hyperbolic relationship of ex-
pression 12.27.
Fallowing the considerations of section 12. 7.1 it is clear that the choice
of pressure ratio to be used in the discussion of the characteristics of a
bearing with constant flow compensation is not as obvious as it was for the
other two types of bearing which have been dealt with. For purposes of
comparison, however, it follows from expression 12.24b that at a film thick-
ness of unity the stiffness of a bearing with constant flow compensation may
be made equal to that of a capillary controlled bearing, that is 0.75, by taking
a pressure ratio of 0.25. The variation of recess pressure and load capacity
then follows from expression 12.23 and is plotted in figure 12.8c along with

330
the corresponding stiffness for the same pressure ratio as given by expression
12.24b.It must be noted that these curves are terminated at the film thickness
where the recess pressure reaches a value of unity.
Similarly it is also useful to consider the behaviour of the bearing at a
pressure ratio of 0.333 which gives approximately the same stiffness as an
orifice compensated bearing, about 1.0, at a film thickness of unity. The
variation of recess pressure and stiffness with film thickness at this pressure
ratio is also shown in figure 12.8c.
Although these choices of pressure ratio may seem a little arbitrary, they
yield characteristics which may be compared with those of the other two
types of bearing; this is done in the next section.

12.7.3 General Comparison of Types of Compensation

In comparing the three types of bearing, by studying their behaviour over a
range of film thickness, some values must clearly be assigned to their re-
spective pressure ratios. Following the developments of the previous section
the pressure ratio of capillary- and orifice-controlled bearings have been
taken as 0.500 and 0.586 respectively, because these values give the corres-
ponding values of Smax at a film thickness of unity. The behaviour of a constant
flow compensated bearing will be studied at the two pressure ratios of
0.250 and 0.333 which, as has been shown in the previous section, produce
values of stiffness equal to the maximum obtainable stiffness of capillary and
orifice-controlled bearings respectively at a film thickness of unity.
Since the curves which describe the behaviour of the three types of bearing
under these conditions have already, with the exception of the relationships
for flow rate, been presented they will now be drawn together and discussed
as a whole.
In figure 12.9a the variation of recess pressure and load capacity is shown
over a range of film thickness for the four cases mentioned above. As has
been noticed previously an orifice-controlled bearing develops a larger
proportion of its maximum load capacity than a capillary-controlled bearing
at all values of film thickness, but otherwise there is little difference between
these two curves. However, the load capacity of the bearings which operate
with constant flow control at pressure ratios of 0.250 and 0.333, increase
much more rapidly with decreasing film thickness than either of the bearings
which work in a constant pressure system, but since the recess pressure
should not by definition exceed unity, the decrease in film thickness is severely
limited. Thus if the film thickness of a bearing with a pressure ratio of 0.25
decreases below about 0.63 then the operating pressure of the pump which
supplies the fluid would exceed the practical limit imposed on it.
This consideration must be applied to the other characteristics of interest
and the variation of flow rate with film thickness shown in figure 12.9b for
bearings with constant flow control are similarly terminated when the film
thickness reaches the value at which the recess pressure attains a value of one.

331
Constant flow r=0·333 ot h= 1·0
Constant flow r =0·250 ot h= 1·0
08 Orifice r=0586ot 'h=I·O
Capillary r=0·500 ot }i=I·O
I~ 06
g
0
IQ. 0-4

02

ot-~ 0 0·2 04 0 6 0·8 1·0 1·2 1-4 1·6 1·8 2·0

fi
(o) Load capac1ty

10

08

Capillary, r=0·500 at h= 1·0

06 Orifice, r =0· 586 at fi = 1·0
Constant flow r=0·333 at h= 1·0
04 Constant flow, r=0·250 at h•I·O
~-'
02

0+-~r.,
0 0·2 0·4 06 0·8 1·0 1·2 14 1·6 1·8 2·0
fi
(b) Flow rate

14

1·2 Constant flow r=0·333 at ii = 1·0

Constant flow r=0·250 at h=I·O
1·0 r=0586 at ii=l 0
r =0·500 at }i =1·0
08
Ill)

06

0-4

02

0~-.r,
0 02 04 06 0·8 10 1·2 14 16 18 20
ii
(c) Stiffness

Figure 12.9 Comparison of bearing characteristics

Above these film thicknesses the flow rates are, of course, constant as would
be expected with this form of control. It is also noticeable that there is
comparatively little difference between the flow rate of fluid necessary to
operate capillary- and orifice-controlled bearings, but it is again of practical
importance that as the film thickness decreases, more fluid is returned

332
directly from the constant pressure control valve, figure 12.1 b, to the reser-
voir of the system.
Several important features are apparent in figure 12.9c, although greater
stiffness is provided by orifice rather than capillary compensation at film
thicknesses of about unity, there is a more rapid decrease with decreasing
film thickness in the former case. It may therefore be useful in a practical
case to utilise the more uniform stiffness of a capillary controlled bearing.
This possibility must be balanced against the previously observed con-
sideration that a bearing with orifice control exhibits a higher load capacity
than a capillary-compensated bearing over a wide range of film thickness.
There are two striking features revealed by the curves for the bearings with
constant flow control drawn in figure 12.9c. It is first clear that the stiffness
becomes progressively larger at an increasing rate as the film thickness
decreases, and second that the magnitude of their stiffness is considerably
greater than that of either an orifice- or a capillary-controlled bearing at
film thicknesses of just less than unity. It must be recalled that since the recess
pressure cannot exceed unity, the range of operation of a bearing with
constant flow control is limited and therefore the very rapid increase in
stiffness shown in this figure must terminate at film thicknesses of about 0.63
and 0. 70 for pressure ratios of 0.25 and 0.333 respectively as indicated in
figure 12.9b. Nevertheless, it may be useful in some applications to make use
of the very high stiffness which may be obtained with bearings of this type,
provided that the often severe limitation on the variation of film thickness
is tolerable.
It should be remembered that the preceding discussion of the various
types of externally pressurised bearings has been based on particular values
of pressure ratios which, although chosen for a sound reason, clearly deter-
mine the characteristics of the bearings. Since changes in the pressure ratios
affect the characteristics of the bearings, the observations which have been
made in this section could perhaps be questioned by considering the be-
haviour of bearings at other values of pressure ratio. However, the general
features and important tendencies of the characteristics of externally pres-
surised bearings, at values of pressure ratio which would be likely to be used
in practice, have been introduced and discussed.

12.8 FLOW, LOAD AND POWER FACTORS FOR OTHER

SHAPES OF BEARING

The foregoing discussions have been based on the treatment of the long
rectangular thrust pad of figure 12.3, and it is clear that bearings of other
shapes may be similarly analysed by solving Laplace's equation 12.1 for the
appropriate plan shape and boundary conditions. However, it will be shown
in this section that considerable rationalisation of the presentation of such

333
solutions may be made by the introduction of load, flow and power factors
which depend only on the shape of a bearing and not its actual physical
dimensions.

12.8.1 Flow, Load and Power Factors for a Long Rectangular Pad

Regardless of the means of compensation it has been shown that the load
capacity of a long rectangular pad of the type drawn in figure 12.3 is given by
W= L(l + b)pr
Noting that the overall area of the pad A is L(b + 2/) we may rewrite the
above as
W = (Apr)a (12.28)
where a= (I + b)/(21 + b), the load factor for a long rectangular pad, or
- 1 + (b/1)
(12.29)
a = 2 + (bjl)

Clearly (b/1) may vary from nearly zero for wide lands, to large values for
narrow lands and the variation of a over a useful range of the ratio (bj[) is
shown in figure 12.1 0.
Similarly the rate of flow of fluid necessary to operate the pad is given by
expression 12.4 as

Q= (h;r)~
Again, the factor (L/61) may be recognised as the dimensionless coefficient
which introduces the shape of the pad into the expression for flow rate and
thus we may write

( 12.30)
where

(12.31)

ij is known as the flow factor for a long rectangular thrust bearing and its
variation for a particular range of the ratio (L//) is also shown in figure 12.10.
It is interesting to note that the load capacity of any bearing of this form
is given by expression 12.28 and that this involves only the overall area of the
bearing A, the recess pressure p., and the particular value of load factor a
corresponding to the shape of the bearing as defined by the ratio (b/1).
Similarly the flow rate through the bearing is given by expression 12.30.
Thus the behaviour of a large range of bearings of different sizes and shapes
may be deduced from these expressions in conjunctil n with the curves of
figure 12.10.

334
l b l
//;;>?
~ Li!>(b+2ll

15

,.._
10 10 0
a v
09 .E
'"' ~
~ 08
f for Lll •20 5 &
.E
"0 07 •I
8 •I
-'
06 •5
0
2 4 6 8 10
bll

'"'
J2
€ 2
~
0
0:

Figure 12.10 Load, flow and power factors for a long

rectangular thrust bearing

It is often necessary to consider if a thrust bearing has the most suitable

proportions or shape for the particular application in mind. What is con-
sidered to be suitable varies of course from application to application, but
it is often appropriate to consider how the pumping power necessary to
operate a bearing for a given load per unit overall area, varies with changes
in its shape. Now the required power P is simply the product of the flow
rate and supply pressure, hence from expression 12.30

p = c:;)pq
and noting from expression 12.28 that

--
P,-
1
A xpii
- (w)
we may write

P = h3
IJ p
~ (w)2(!)
A a2
335
Remembering that the dimensionless recess pressure p would be chosen
independently on the considerations of section 12.7.1, the power required
to operate the bearing for a given load per unit area WI A and fixed values
of film thickness h and fluid viscosity '1 is, as far as the geometric shape of the
bearing alone is concerned, dependent only on the factor ii/a 2 which is
termed the power factor J thus

P__ h'1 ~p (W)

3
A
! 2
(12.32)
where
J = -q2 (12.33)
a
Combining expressions 12.29 and 12.31 with 12.33 shows that the power
factor for a long rectangular pad is given by

J= (61L)(2 ++ b/1b/1)
I
2
(12.34)

The variation of this parameter for a few values of the ratio L/1 is shown as a
function of b/1 in figure 12.10. The curves plotted in figure 12.10 enable the
designer of a long rectangular pad immediately to deduce the effects of
changes in shape on the behaviour of the bearing.

12.8.2 Flow, Load and Power Factors for a Circular Pad

A very useful form of hydrostatic thrust bearing in the form of an annulus is
shown in figure 12.11 a. Fluid is supplied to the central recess of radius R 1 ,
and it flows radially outward to the periphery of the bearing at radius R 2 •

r;l
' R2
Radtus rafto R• R 11R2

•
7
,,
6
g
.E
10 5 ~
3:
0
08 4 a.
I<:J "'c
3 ,.,. "
~ 06
.E €
"'"0_J
04 2 .E
3:
0
02 G:
Q
0+--.---.--,--.--.---.--.--+
01 02 03 04 05 06 07 08 09
Rod1us rotto R

Figure 12.11 Load, flow and power factors

for a circular thrust bearing

336
If the uniform clearance between the pad and its plane mating member is h,
the fluid viscosity Yf and the recess pressure p, it may be shown 3 that the closed
form solution of Laplace's equation in the case of this particular shape
yields the following expressions for load capacity W and flow rate Q

where R = RtfR 2 , the radius ratio of the bearing which is sufficient to des-
cribe its shape completely.
Recognising 1tR~ as the overall area of the bearing and putting

(12.35a)

and
7t 1
q= - X -,------- (12.35b)
6 loge(l/R)

the above expressions for Wand Q may be written as

w= (Ap,)a
and

These latter two expressions are identical to expressions 12.28 and 12.30
respectively, which were derived for a long rectangular pad but of course the
values of a and q must be correctly chosen. It follows that the power con-
sumption of a circular thrust bearing is given by expression 12.32, provided
of course that the power factor J is given the value

(12.35c)

The variation of the factors a, q and J for a circular pad are given over a
wide range of radius ratio R in figure 12.11 b, and it is interesting to observe
that a minimum exists in the power factor at a radius ratio of about 0.54.
This type of deduction is often useful during the design stage when the pro-
portions of a bearing are being chosen.

337
12.8.3 Load, Flow and Power Factors for a Typical Bearing of more
Complicated Shape

It is a relatively simple matter to obtain closed form expressions for the load,
flow and power factors for the long rectangular and circular types of thrust
bearing which have been dealt with so far. However, there are many practi-
cally useful plan forms which do not yield to simple analysis and the three
factors must then be evaluated by more advanced methods, for example,
with the aid of a conducting sheet analogue or by digital computation.
An example of this class of bearing is the square thrust pad of side 4/ shown
in figure 12.12a. This pad contains four equi-spaced recesses of side 2s and
its overall area A is 16f2. Assuming that the pressures in the four recesses all

Overall are~ 151 2

(a) Plan of beanng

14

12

,..._
10 10 .:
g
!!
:;;
08 8 3:
0
a.
u
c
0
It> ,.;
06 6
v?5
0
v
!! !!
u
0 3:
0 .9
..J 04 4
lL

02 2

5/(

(b) Factors of beanng

Figure 12.12 Load, flow and power factors

for a square thrust bearing with four equispaced
square recesses

338
attain the same value p., it follows that its load capacity will be given by
expression 12.28, its flow rate by expression 12.30 and its power consumption
by expression 12.32, provided that the appropriate values of ii, q and f
respectively are used in these expressions. Now it is clear that the flow
patterns and pressure distribution in this bearing will be of a complicated
form and no accurate closed form expressions for these three factors may be
derived. Nevertheless since these parameters have been evaluated by an
advanced method 4 they may be presented in graphical form as functions of
the ratio (s/1) which alone is capable of defining the shape of the bearing.
These curves are drawn in figure 12.12b and the performance of any size of
bearing of this shape may be deduced immediately by entering the appro-
priate values of ii, q and f from these curves into expressions 12.28, 12.30 and
12.32 respectively. It is again of practical importance that the power con-
sumption, for a given load per unit overall area, will be a minimum when
the ratio (s/1) has a value of approximately 0.4.
The load, flow and power factors for many other useful shapes of thrust
bearing are available in the literature but some care is necessary in using them
since certain constants are not always included in the definitions of these
parameters, for example the constant rt/6 of expression 12.35b is sometimes
included in expression 12.30 rather than in q. The use of load, flow and
power factor offers a considerable amount of simplification and, more
importantly, generalisation in even these complicated cases; an example of
their application to the case of a single tilted thrust bearing is presented in
the following section.
Finally since the general expressions for the load capacity, stiffness and
flow rate given in expression 12.16 have been used in the discussion of
compensation in section 12.7 it is of interest to relate these parameters to the
load and flow factors which have been introduced here.
It will be recalled from expressions 12.16 that the load capacity of a long
rectangular bearing is given by

W = p.L(b + I)W
Introducing the load factor ii and overall area A

W = p.AiiW

or

C~)=aW
The parameter (Wfp.A) is an alternative form of dimensionless load and
is often used in work on hydrostatic bearings. It may be interpreted as the
effectiveness of a bearing because it is equal to the ratio of the load actually
carried by a bearing to the load which would be carried if its overall area A

339
were subjected to the supply pressure Ps· Thus putting W' = (Wfp.A) the
relationship between the two dimensionless load parameters is found to be

W'=aW (12.36a)

and remembering from section 12.6.2 that W = fi

W' = afi (12.36b)

Similar reasoning shows that the dimensionless stiffness defined in ex-

pression 12.16 is related to W' by

(12.37a)

or

(12.37b)

The dimensionless flow rate is related to the flow factor by

Q' = (h~ oPs )= ijQ ( 12.38)

'1

12.9 SLIDING EFFECTS IN THRUST BEARINGS

So far our attention has been limited to the operation of hydrostatic thrust
bearings under static conditions, that is, where there is no relative motion of
the two bearing surfaces. However, in practice, bearings are required to
operate under dynamic conditions and we must now consider the effects
introduced by relative motion of the bearing surfaces.
Any general relative motion of the bearing surfaces can be divided into
two components, a normal velocity component and a sliding component and
the effects of these may be considered separately. Normal velocities occur
when a bearing is subjected to a time-dependent load and have the effect of
modifying the pressure distribution in the bearing-this is termed squeeze
action. The resulting changes in pressure give rise to a force which is pro-
portional to the squeeze velocity and opposite in direction to the motion of
the bearing surfaces. Hydrostatic bearings therefore possess inherent viscous
damping. A thorough treatment of this feature of hydrostatic bearings is
beyond the scope of this book, but in the majority of applications dynamic
loads are of secondary importance and are often ignored. However, sliding
effects cannot be so readily disposed of.

340
12.9.1 Friction in Hydrostatic Thrust Bearings

The simple circular thrust pad shown in figure 12.13a is very suitable for
supporting a rotating member such as a shaft. When the shaft is stationary
the lubricant flows radially outwards from the recess but rotation introduces
a circumferential velocity. The circumferential velocity varies linearly
across the lubricant film, as shown in figure l2.13a and this shearing action

--t--
w

~k (a)

Comb1ned velOCity d1stnbut10n

(b)

Figure 12.13 Effects ofsliding velocity in thrust bearings

gives rise to shear stresses which oppose the rotation. For a newtonian
lubricant the shear stress r is given by 2

where rJ is the dynamic viscosity of the lubricant and ouj oy is the velocity
gradient. In this case, at any general radius r the. velocity gradient does not

341
vary in the circumferential direction and is equal to wrjh; the shear stress at
radius r is given by
wr
T = '1-
h

This shear stress acts on the surface of the rotating member and produces
a torque which opposes the motion, that is, a friction torque. Normally the
recess depth is much bigger than the bearing clearance and therefore the
shear stresses in the recess area are usually negligible compared with those
over the land area. To evaluate the friction torque we consider a circumfer-
ential element, radius r, width Jr. The shear torque produced by this element
is (r x 2nr Jr x r) and integrating this over the land area gives the total
friction torque T

( 12.39)

For thrust bearings used in slideway applications the motion is linear

rather than rotary and this introduces some additional effects. In the rotary
example the motion of the surfaces was in a direction at right angles to
the direction of flow, but with linear sliding the motion is in the same direc-
tion as the flow. Let us consider the example of a one-dimensional thrust
pad in which the plane mating surface is sliding in a direction normal to
the axis of the recess. In the static case with no sliding the velocity distri-
bution across the lubricant film is parabolic as shown in figure 12.13b.
Clearly, the velocity gradient on the surface of each land is not zero, and
therefore shear stresses act on these surfaces and produce forces tending
to drive them in the direction of flow. However, because of symmetry the
forces on the two lands are equal and opposite and the net shear force on the
bearing is zero. When sliding takes place a linear velocity distribution is
superimposed on the parabolic distribution so that the resulting velocity
distribution is as shown in figure 12.13b. It is important to note that the
velocity distributions for the two lands are now different and a net friction
force is now generated which opposes the motion of the sliding surface.
Since the system is linear we can claculate the friction force by considering
only that factor which brings it about, namely the linear velocity distribution
resulting from the sliding action. The shear stress due to the sliding action is
given by

342
and since this is constant over the land area, the total friction force F is
given by

(12.40)

where A1 is the total land area. The above equation again assumes that the
recess depth is large compared with the film thickness so that shear forces over
the recess area can be ignored.
Although the arguments have been developed for a one-dimensional
bearing, equation 12.40 can be applied to bearings of any shape.

12.9.2 Effect of Sliding on Flow Rate

In addition to the friction effects, sliding also modifies the flow in the bearing
clearance. Under stationary conditions the flows from the two sides of the
one-dimensional bearing (figure 12.13b) are equal. When sliding takes place
a shear flow is superimposed on the static flow. If the total static flow rate into
the bearing per unit length is Q then the static flow rate over each land is Q/2.
The shear flow rate is equal to the mean shear velocity multiplied by the flow
area, and for the linear velocity distribution the shear flow per unit length of
bearing is Uh/2. With sliding the shear flow on the upstream land of the
bearing reduces the static flow so that the net outflow over this land is
Q/2 - Uh/2. On the downstream side the shear flow increases the existing
flow giving Q/2 + Uh/2. While this situation exists, the total flow into the
bearing is

(i + ~h)+ (i- ~h)= Q

that is, the inflow is the same as it was under static conditions. Since the shear
flow is dependent on the sliding velocity it is possible to have the situation
where Uh/2 = Q/2 and the flow from the upstream side of the bearing is then
zero. If the sliding velocity is increased beyond this point, the outflow from
the upstream side becomes negative that is, the flow is inwards towards the
recess. This can only happen if there is sufficient lubricant present on the
sliding surface approaching the bearing to provide the necessary inflow.
If the approaching surface is dry or carries insufficient lubricant the above
arguments no longer apply, so that the conditions on the upstream land are
modified with the result that the lubricant film no longer extends to the
edge of the bearing. This phenomenon is known as film recession, and the
extent of the recession increases with the sliding velocity. In the presence
of film recession the load capacity of the bearing is reduced and, more
importantly, there is the danger of air being entrained into the recess with
the risk that the bearing could collapse. This must of course be prevented
and a simple method is to introduce a low pressure source of lubricant at

343
the upstream edge of the bearing to supply the required inflow and thereby
prevent air entrainment.
Film recession is not peculiar to one-dimensional bearings, it can occur
with any shape of bearing, but the criterion for the onset of film recession in
other cases is more complicated than it is with the one-dimensional bearing.

12.9.3 Non-parallel Operation of Thrust Bearings

It is convenient and common practice to design hydrostatic thrust bearings

to operate with a uniform film thickness, that is, with the two bearing
surfaces parallel, since this simplifies both the manufacture and analysis of
performance predictions. However, it is highly unlikely that parallel opera-
tion is ever achieved in practice. It is reasonably straightforward to produce
thrust bearing surfaces which are flat, but with manufacturing tolerances and
assembly errors it is very difficult to ensure that the bearing surfaces are
parallel when assembled, and even if they are, structural deflections due to
loading and thermal effects will occur in operation. Deflections of the bearing
surfaces can lead to complex variations in bearing clearance and these in
turn can produce significant changes in the bearing properties under both
static and sliding conditions.
A simple but common example of non-uniform clearance is that which
occurs due to tilt, that is, where both surfaces are plane but inclined relative
to each other as shown in figure 12.14. In order to assess the possible effects
of tilt on the performance of a bearing we shall consider a simple circular
thrust pad and, initially, limit our attention to the static operation. When
the bearing surfaces are parallel the pressure distribution is symmetrical
about the bearing's centre but this is modified in the presence of tilt. Over the
recess area, tilt has negligible effect and the pressure remains essentially

Pressure d1sfnbut1on

Figure 12.14 £fleet of tilt on pressure

distribution

344
uniform. Over the land where the flow takes place through a converging
clearance the pressures are increased and where the clearance is diverging
the pressures are decreased. The unsymmetrical pressure distribution
produces a moment which tends to force the bearing into a parallel state,
but the magnitude of this moment is so small that its effect can usually be
neglected, however, the accompanying changes in load capacity and flow
rate may be very significant.
It has been shown that the load capacity and flow rate of a parallel thrust
bearing can be expressed in the form 5
w= (p,A)a

Q= (\h3)ij
and these equations can be extended to tilted operation. For parallel con-
ditions aand q are functions of the bearing geometry only, which is described
entirely by the radius ratio, in this case, expression 12.35
(1 - Rz)

ao = 2 loge(*)

1t

where the subscript o has been introduced to refer to parallel conditions.

ij
When tilt occurs a0 and 0 are modified and the effect can be described in
terms of a tilt parameter

where rx is the angle of tilt (in radius) and h is the effective clearance at the
centre of the bearing as shown in figure 12.14. In the presence of tilt, a and
ijare both greater than the corresponding values for parallel surfaces, as
shown in figures 12.15a and b, the changes being dependent on the radius
ratio. The increase in ll is greatest for small radius ratios and can be up to
ij
15 per cent while changes in can be as large as 100 per cent and increase with
radius ratio.
The data of figures 12.15a and b allow a bearing to be designed specifically
for tilted operation, but this is not normal practice at present and it is more
useful to consider how the operation of a bearing designed for parallel
conditions may be modified by the inevitable occurrence of tilt: these effects
are demonstrated in figures 12.15c- f. It should be noted that we are now
considering the effects of tilt on the bearing system as a whole, and these are
dependent on the pressure ratio whereas the load and flow coefficients for

345
116

(a)

I 12

,.,•
108
'"'

104

100
0
a
22

(b)

20

18

'~
,.,.
' 16

14

Figure 12.15 Characteristics of tilted pads

the pad are not. The curves given in figures l2.l5c-f are for a capillary-
compensated bearing supplied from a constant pressure source, but similar
curves would be obtained with orifice compensation. Except for small
radius ratios, the central clearance is always less than the parallel design
clearance (figure l2.l5c). These changes in clearance are generally small, but
when they are combined with the effects of tilt they can produce dangerously
small minimum clearances even for modest values of the tilt parameter
as shown in figure l2.l5d. Changes in the design flow rate are generally

346
r•O·~

100
?'? }R·025
~
(c)

095

<:.
~ 090

085

080

0 02 04 06 08 10

10

(d)
08

06

OA

02

0 02 04 06

Figure 12.15

smaller than might be expected from the variation in flow parameter but the
increase in flow can still be as much as 20-30 per cent (figure 12.15e). The
static stiffness of a tilted bearing depends upon the combination of radius
ratio, pressure ratio and tilt parameter and may be greater or smaller than
the design stiffness.
The data of figure 12.15 are derived for a tilted circular thrust pad but they
are indicative of the effects which tilt will have on the static behaviour of
other shapes of bearing.

347
1·3

(e)

12

1·1
033
033
~-05

04 - 06 08 10
a
R=O 25
R=O 50
R=075

13

(f)

12

<R
\;) II

09

0 02 0 4 - 06 08 10
a
R=O 25
R=O 50
R•O 75

Figure 12.15

348
When sliding is introduced, the effects of tilt are more pronounced. The
non-uniform clearance leads to the generation of hydrodynamic pressures
which are superimposed on the hydrostatic pressures 6 and these can make a
significant difference to the bearing. When the tilt and sliding produce a
converging lubricant film, positive hydrodynamic pressures are generated
which enhance the load carrying capacity. Therefore the same load can be
supported with a smaller recess pressure, but this in turn leads to a greater
pressure drop across the compensating restrictor with a consequent increase
in flow, which is accompanied by an increased film thickness. When the tilt
and sliding lead to a diverging lubricant film, the hydrodynamic action
reduces the pressure over the lands and decreases the load capacity. In order
to support the same load the recess pressure must be increased, the flow rate
consequently decreases as does the film thickness.
It is possible for the pressure on the lands to fall locally to the vapour
pressure of the lubricant (approximately absolute zero for oil) as a result of the
hydrodynamic action, and this produces cavitation which further modifies
the bearing behaviour. The characteristics of a tilted sliding bearing may
also be further modified by boundary recession unless precautions are taken
as already mentioned. A more detailed treatment of the behaviour of tilted
sliding bearings is given in references 6 and 7.

12.10 HYDROSTATIC JOURNAL BEARINGS

An obvious and practically useful development of the types of hydrostatic

thrust pads which have been described so far is their application to the
support of rotating circular shafts. The principles of operation of hydro-
static journal bearings are the same as those of thrust bearings but their
behaviour, analysis and design are complicated by two main factors

(i) Except for the unloaded, concentric position the fluid film thickness in
the bearing is not uniform.
(ii) Shaft rotation results in comparatively high relative velocities being
established between the shaft and bearing. This considerably alters the
pressure distribution within the bearing.

Hydrostatic journal bearings offer two advantages over all other types of
fluid film journal bearings: their load capacity and radial stiffness are high
even at zero speed, and very small frictional torque exists during running.
The fact that the frictional torque is zero at zero shaft speed is unique, and
hydrostatic journal bearings are often chosen for particular applications by
virtue of this fact alone.
The evolution of journal bearing configurations and the behaviour of some
important types will be described in this section.

349
12.10.1 Evolution of Journal Bearings

It has been recognised that two main classes of hydrostatic journal bearings
exist
(i) multipad 8 bearings which consist of one or more thrust pads with curved
surfaces:
(ii) multirecess 9 bearings in which individual pads are almost indistinguish-
able from each other.
A single-pad journal bearing is shown in figure 12.16a. This consists of a
simple thrust bearing shaped to accommodate the curved surface of the shaft,
but obviously severe limitations exist on the direction in which radial loads
may be applied. A typical multipad bearing is shown in figure 12.16b and,
although a useful load capacity is provided for all radial directions, two

(a) (b)

(c) (d)

Figure 12.16 Evolution of hydrostatic journal bearings

undesirable features of this design are apparent. First, the drainage grooves
between adjacent pads are regions of low pressure and therefore contribute
nothing to the bearing's load capacity. Second, proper analysis, and there-
fore good design, is difficult since the flow patterns and pressure distribution
in the lands of the bearing are difficult to predict even when the shaft is
stationary and concentric.
An improvement may be made in the first of these features by developing
the configuration shown in figure 12.16c. The absence of drainage grooves
in this bearing results in a continuous land being formed between the
recesses, so that a larger effective area of high pressure is formed and the flow

350
rate required to operate the bearing is reduced. When this type of bearing
is taken to an extreme the shape of figure 12.16d is obtained with the following
accrued advantages
(i) The relatively large pockets form regions of high pressure which
provide good load capacity and stiffness.
(ii) Because the lands are narrow compared with their length simple one
dimensional viscous flow theory, section 12.4, may be used to describe
the flow of fluid within the bearing.
(iii) Following the assumption of (ii) above and the reasoning of section
12.9.2, it becomes possible to deal with the effects of shaft rotation on
the flow of fluid within the bearing in a simple manner.
In the following sections a brief treatment of the types of journal bearings
introduced above will be given.

12.10.2 Journal Bearings with Isolated Pads-Multipad Bearings

The method of analysing bearings of this general type, figure 12.16b, is

introduced in this section by considering the bearing of figure 12.17 which
consists of N equi-spaced, isolated pads of identical shape and overall area
A. Assume that a load W is applied to the non-rotating shaft at some angle
<P to the vertical line which has been drawn through the centreline of pad I,
and that the pads are numbered consecutively in the clockwise direction,
the angular pitch of the pads thus being 2n/ N. It may further be assumed that
the centre line of the shaft will move to some position within the bearing

Figure 12.17 Hydrostatic journal

bearing with N isolated pads

351
specified by the eccentricity ratio e and the attitude angle rx. If the radial
clearance between the shaft and bearing in the unloaded concentric position
is c then the eccentricity ratio e in the loaded condition may be defined as
the ratio of the distance between the centrelines of the shaft and journal to
the clearance c.
Study of the geometry of the bearing shows that the decrease in the original
film thickness c along the centreline of the ith pad is given by

Assuming for the moment that the stiffness of each pad is given by expres-
sion 12.37a as

oW=
ah
(psA)(oW')
c afi

where a is the load factor for the individual pads and opjofi depends on the
type of compensation adopted. It will be recalled that opjofi may be obtained
in the general case from expressions 12.14b, 12.20b and 12.24b for capillary,
orifice and constant flow compensation respectively.
Thus the additional radial load W; produced by the ith pad because of the
movement of the shaft is simply given by

w= I
(aw)
oh i
x bh.
I

or

W; = Ps
( -c-A)_(op)
a ofi X ce cos [2rc
N (1. - 1) - rx J (12.41)

Referring to figure 12.17 it is seen that the components of this force in the
direction of and at right angles to the line of centres are W; cos [(2rc/ N)(i - 1)
- rx] and W; sin [(2rc/N)(i - 1) - rx] respectively. The total loads produced
by all N pads in these two directions, Wx and ~ respectively, follow im-
mediately as

wX 2: ~
N
i=I
cos [2rc
- (i - 1) -
N
!X J
and

~ = 2:N ~sin
i=l
[2rc
- (i - 1) - rx
N
J
352
Substituting for »-;from equation 12.41 yields

Wx = (-p·A)_
ace ~ (ap) cos [21t
L... . l)- aJ
- (1- - 2
(12.42)
c ali i;, N
and

~ = (PscA)aceJ (~)co{
1
(i _ 1) - a }i{~ J
(i - 1) - a

It will be appreciated that as the value of(ap;ali) will vary from pad to pad,
the quantities within the summation signs of expression 12.42 must be
evaluated at length in the general case. However, it may be shown that if
the eccentricity ratio e is less than about 0.2 or 0.3, then there is comparatively
little variation in (ap;ali) which may therefore be given the value it assumes
at zero eccentricity. Noting that

LN cos 2 {21t (i -
- 1) - a } = -N
;;, N 2
and

J, cos{~ (i- l)- a}sin{~ (i- l)- a}= 0

for all values of N and a, expressions 12.42 reduce to

W----ee
x c
_(psA) aN (ap)
2 ali
and wy = 0.
Thus the shaft deflects along the direction of the applied load and it follows
that the load angle 4> must equal the attitude angle a. It follows also that the
zero eccentricity stiffness of the bearing, So, may be deduced from the above
expressiOn as

s wx
ce (p·A)
0 c aN2 (ap)
=
ali
=

and a nondimensional zero eccentricity stiffness, S0 , may be defined as

(12.43)

The flow rate of fluid Q, necessary to operate a multi pad bearing is simply
found by applying expression 12.38 to each of the N pads and forming a
sum such as

353
where Qi must be evaluated for each pad from expressions 12.15, 12.21 and
12.25 for capillary, orifice and constant flow compensation respectively.
The benefit of the generalised description of hydrostatic bearings which has
been developed in this chapter is now apparent since expression 12.43
enables the zero eccentricity stiffness of a hydrostatic journal bearing with
any number of individual pads of any shape to be determined for any type
of compensation. It is useful to note from expression 12.43 that the zero
eccentricity stiffness of a journal bearing which consists of N equi-spaced
and identical pads is simply N /2 times the radial stiffness of one of the pads.
The above treatment of multi pad journal bearings is necessarily brief and
many important features have been discounted. For example it is more
difficult to evaluate the load factor a for a pad with curved surfaces than it is
for a plane pad of the same projected shape, because the pressure in the fluid
film acts radially and not perpendicularly to the plane shape of the pad. At
high values of eccentricity ratio e the non-uniformity of the film thickness
cannot be ignored and it is recognised that the load factor of each pad will
be dependent on the eccentricity ratio and the attitude angle of the shaft.
Finally, the effects of shaft rotation can also be important, so that in principle
a solution of the Reynolds equation, rather than Laplace's equation, must
be obtained for each pad before the behaviour of the complete journal
bearing is determined.
In spite of these and other complications the simple analysis which has been
presented in this section is adequate for many purposes.

12.10.3 Journal Bearings with Interacting Pads-Multirecess Bearings

The analysis and design of multirecess hydrostatic journal bearings, in which
no drainage grooves exist between adjacent recesses, have been considered
by many workers 10 . Although this classification includes bearings of the
type shown in figure 12.16c only the more highly developed and practically
useful shape with narrow lands shown in figure 12.16d will be considered
here. A rigorous analysis involves the simultaneous determination, for every
given shaft position, of as many recess pressures as there are recesses, and
consequently the results of this form of inverse analysis are generally pre-
sented graphically. It is not possible to deal with this approach here, and
therefore the results of a more restricted analysis of a bearing with four
recesses will be presented and discussed 11 .
In figure 12.18 the general form of the type of bearing under discussion
is shown along with the dimensions required to specify its geometry. The
directions of shaft rotation and applied load are indicated, and of importance
are the definitions of the load angle ¢ and attitude angle ex implied in this
diagram. The position of the shaft is thus given by the eccentricity ratio e
and the attitude angle ex. An important geometric variable for this type of
bearing is the aspect ratio m which is defined as m = lrl!Dlc. This parameter
may be taken as a measure of the ratio of the circumferential flow rate be-

354
...L./ I. l, .I
Figure 12.18 Hydrostatic journal bearing with four
recesses
tween adjacent recesses to the axial flow rate from each recess, and clearly de-
termines the degree of interaction between the recesses. Two other parameters
are necessary to describe the bearing, that is, the pressure ratio rand the speed
variable w.
The pressure ratio r is defined as the ratio of the pressures in the recesses of
the bearing to the supply pressure at zero eccentricity and is related to the
bearing's dimensions and fluid properties through the following expressions
r 2A.
For orifice compensation 112
(1 - r) 7t

.
For capt.11ary compensatton r 2A.
-- = -
1- r 1t
where
A. - 3(21t)t'zCddz'fl
- (pp.)lf2c3D
for orifices and
31td4 l
A.= 32l,c4 D
for capillaries. Here '1 and pare the dynamic viscosity and density respectively
of the working fluid, c the radial clearance, d the diameter of the orifice or
capillary, Cd the discharge coefficient of the orifice, I, the length of each
capillary restrictor, and p. the supply pressure.
Similarly the speed variable w is given in terms of the above dimensions,
and parameters, for all types of compensation as
61t1fN. II,
w = --'=--
czp.
where N. is the speed of shaft rotation in revolutions per unit time.

355
By writing down a continuity of flow equation for each recess of the bearing
shown in figure 12.18 it is possible, with some assumptions, to derive closed
form expressions for the changes in recess pressures which occur when the
shaft assumes a small eccentricity in response to an applied load. Integra-
tion of these changes in pressure results in the following expression for the
radial stiffness of the bearing about the concentric position.

or, in dimensionless form

(12.44)

where
11(-r-) +-+2m
a=- 11
4 1- r 2
for orifice compensation, or

11(r)
a=- 11
- - +-+2m
2 1- r 2
for capillary compensation, and
11
a= l +2m

for constant flow compensation.

It will be seen that the dimensionless stiffness So is a function of aspect
ratio m and pressure ratio r, and for a non-rotating shaft its variation takes
the form shown in figure 12.19 for the common cases of orifice and capillary
compensation.
It is clear from these curves and expression 12.44 that the stiffness of a
hydrostatic journal bearing with four recesses increases with shaft speed,
decreases with increases in aspect ratio, is a function of pressure ratio and is
in general greater for orifice compensation than capillary compensation. It
may also be shown that in the presence of shaft rotation the attitude angle
r:x is not equal to the load angle cjJ but is related to it as follows

r:x =tan
_ 1 (3rtan cjJ +w) ( 12.45)
3r - w tan cjJ

It is also clear from figure 12.19 that the nearer the aspect ratio approaches
the practically unobtainable value of zero the greater will be the stiffness of
the bearing. For each aspect ratio a pressure ratio may be chosen which
yields the greatest value of stiffness and the locus of this optimum value of

356
14 Locus of maximum
stiffness

Locus of maxtmum
st1ffness 10
10

08 08
IV)

06 06
IV)

04 04

02 02

02 04 06 0 8 10 0 2 04 0 6 08 I0
Pressure rotto r Pressure rotto r

(a) Capillary compensation (b) Onf1ce campensat1on

Figure 12.19 Variation of stiffness of a four-recess journal hearing at zero

speed

stiffness and pressure ratio is shown for both cases of compensation in figure
12.19.
The flow rate necessary to operate a four-recess bearing, Q, depends only
on the flow of fluid in the axial direction and for small values of eccentricity
this may be expressed as

Q, = (n~3 P.}
The results which have been discussed above were obtained on the assump-
tion that the eccentricity ratio is small, and it is clearly necessary to know
the limitations of this restriction 12 . In figure 12.20 the behaviour of a bearing

01
-
--- Inverse solutton
Smoll eccentnc1ty solutton

..
Q
02 Lines of constant dimenstonless
load at 1ncrements of 0 I
03
~
?:' 04
g
cQ) 05 Locus of shaft
centre at w • "'b • I A 22
u
u
w 06

Figure 12.20 Behaviour of a hydrostatic

journal bearing at various values of shaft speed

357
with an aspect ratio of 1.345 is shown for a load angle of ¢ = 45o and a
pressure ratio of r = 0.5. Two sets of curves are drawn in this diagram which
show the displacement of the shaft at several speeds of shaft rotation for
different values of a dimensionless load W, defined here as
-w = w
-=--cc:-----::--
D(l. + l)p,
as given first by the small displacement analysis described above and second
by the more rigorous inverse analysis which is in principle applicable at high
values of eccentricity ratio. The shaft speeds considered are described in
terms of fractions of an optimum speed variable, W 0 , which is related to the
rate at which energy is dissipated in the bearing, but details of its significance
are not of importance in this limited discussion. For convenience the at-
titude angle is measured from the line of loading and not from the centre-
line of a circumferential land as before. The full lines represent the inverse
solution and the broken lines the small eccentricity solution.
Up to an eccentricity ratio of about 0.3 very little discrepancy exists
between the two solutions in either the prediction of eccentricity ratio or
attitude angle. However, the small eccentricity solution predicts a constant
stiffness and attitude angle for each shaft speed while gradual variations are
revealed by the inverse solution. It is significant that the small eccentricity
solution gives either an overestimate or an underestimate of the load capacity,
depending on the shaft speed. Thus, neither the increasing stiffness at zero
speed or the decreasing stiffness at the optimum speed variable are predicted.
It is emphasised that the manner in which the load capacity changes with
eccentricity depends on the pressure ratio, speed variable and load angle.
However, the comparison between the two methods of solution presented
here for one set of circumstances is generally true.
Many other features of the detailed operation of this type of bearing have
been studied in the extensive literature on this topic which may be consulted
if required.

12.11 OTHER TYPES OF HYDROSTATIC BEARINGS

Although several types and shapes of hydrostatic bearing have been described,
analysed and discussed in this chapter, it is obvious that these represent only
a small selection of the many useful configurations which have been developed
to meet the requirements of various particular applications. It has been
shown, however, that the behaviour of bearings of practically any shape
and complication may be conveniently expressed in the form of dimensionless
factors, which may then be used to derive the dimensional characteristics of a
wide range of actual bearings. The determination of the behaviour of a
hydrostatic bearing system must of course involve the characteristics of the
control or compensating devices used, but again a formal presentation of the

358
cardinal features of three important methods of compensation have been
presented in dimensionless form.
It must be remembered, however, that many features which are sometimes
of importance in hydrostatic bearings have not been considered at all;
these include dynamic response, cavitation, energy dissipation, active
compensation, inherent compensation, and the effects of non-uniform film
thickness in bearings of complicated shape.

REFERENCES

I. H. C. Rippel. Design of hydrostatic bearings- Pt I. Machine Design, (Aug. 1963),

108.
2. R. H. F. Pao. Fluid Mechanics. Wiley, New York, (1965).
3. D. D. Fuller. Theory and Practice of Lubrication for Engineers. (Ch. 3.) Wiley, New
York, (1956).
4. A.M. Loeb and H. C. Rippel. Determination of optimum proportions for hydro-
static bearings. Trans. Am. Soc. Lubric. Engrs, 1 (2), (1958), 241.
5. R. B. Howarth. Effects of tilt on the performance of hydrostatic thrust pads. Proc.
Jnstn mech. Engrs, 185, (1970-71), 51/71.
6. R. B. Howarth and M. J. Newton. Investigation of the effects of tilt and sliding on
the performance of hydrostatic thrust bearings. Instn mech. Engrsflnstn Prod.
Engrs Conf. Externally Pressurized Bearings, London, (1971).
7. M. J. Newton and R. B. Howarth. Film recession in hydrostatic thrust bearings.
Proc. Jnstn mech. Engrs, 187, (1973) 57/73.
8. H. C. Rippel. Design of hydrostatic bearings-Pt 9. Machine Design, (Nov. 1963),
199.
9. H. C. Rippel. Design of hydrostatic bearings-Pt 10. Machine Design, (Dec. 1963),
!58.
10. P. B. Davies. A general analysis of multirecess hydrostatic journal bearings. Proc.
Jnstn mech. En_qrs, 184, Pt. I, ( 1970).
II. P. B. Davies and R. Leonard. The dynamic behaviour of multirecess hydrostatic
journal bearings. Proc. Jnstn mech. Engrs, 181, Pt. 3L, (1969-70), 139.
12. P. B. Davies. Modes of failure in multirecess hydrostatic journal bearings. Proc.
lOth Int. M.T.D.R. Conf. Pergamon Press, (1969), 425.

359
13
Selection of
Tribological Solutions
13.1 INTRODUCTION

Before considering tribological solutions a designer must first consider the

possibility of eliminating the tribological difficulty by using. an alternative
design. Designs with fewer moving parts not only eliminate some of the
tribological headaches, but invariably provide a cheaper and more elegant
solution to the overall problem. In chapter I this principle was illustrated by
a rather graphic example, but many other less glamorous examples can be
quoted. The use of elastometers and flexible linkages to provide a range of
oscillatory motions without rubbing interfaces is an established practice,
while the recently developed Wankel engine clearly illustrates the same basic
principle.
Despite the designer's best endeavours, however, there are still many
residual tribological contacts and this chapter indicates some of the guide
lines in choosing the most appropriate tribological solution to any particular
problem. In a sense this chapter demonstrates the relevance of much in the
foregoing chapters, so that the reader may obtain a perspective view of the
authors' intentions.
From the designer's standpoint he will be interested in achieving a tribo-
logical solution to the particular problem which satisfies the basic operational
parameters of load capacity and speed within a specified environment. It is
therefore necessary to consider the effects of load and speed within any

360
specified environment and to identify those physical conditions which place
limits on the performance.
In practice the majority of tribological problems in industry will arise
within an atmospheric environment. None the less special problems are
created in systems operating in high vacuums, nuclear reactors and some
chemical plants, because of the absence of self-generating oxide films. The
tribology in these so called 'hostile environments' presents special problems
which are not dealt with in this text, although the earlier chapters will have
indicated why they arise and, more usefully, how they might be solved. In
what follows we shall consider the effects of load and speed in defining the
limits of performance of the various tribological solutions assuming a normal
atmospheric environment.

13.2 LOAD AND SPEED

The full ranges of load and speed are too complex to be considered in detail
since they have both magnitude and direction, each of which may be constant
or vary in some prescribed pattern. There are four broad classes of loading,
while the velocity has eight combinations due to the additional effect of
combined rolling and sliding as occurs between mating gear teeth. The full
pattern of load and speed options may be classified as shown in figure 13.1,
which also shows some typical examples from the thirty-two possible
combinations of load and speed patterns. When it is recalled that no attempt

Four types of load E1ght types of speed

Journal beonng Gear teeth Con rod bearing Com Cor wheel

Load and speed Load. speed Load vanes Load and Load. speed and
constant 1n and roll/shde 1n mogn1tude speed vary roll/slide ore
mogn1tude and all vary and dtrection m magmtude constant 1n
dtrectton steady store

Figure 13.1 Types of load and speed variation

361
has been made to define the wide range of patterns of load and speed varia-
tions, we can see how these two variables alone can lead to an enormous
range of tribological situations.
In the following it will generally be assumed that the load and speed are
constant in every sense, that is, we shall concern ourselves with steady-state
arguments. None the less the effects of load and speed variations can be
inferred by extrapolation of the ideas presented, particularly when one
recalls the importance of the contact history of discrete particles. Consider
the journal bearing shown in figure 13.2 where the load is constant in magni-
tude but oscillates between the limits ±rx and produces the pressure distribu-
tion shown. The stress history of particle 1 on the shaft and particle 2 on the
housing will be as shown, and will clearly depend on the relative value of rx.
w

Pressure
d1strtbutton

pP
p

Shoff

o;
>
.S!
V>
V>
~
CD
a
(j,

®
a-t-a·
HOUSing

Figure 13.2 Oscillating constant norma/load; case

where oscillations are at same speed as shaft

Since the total effect will be the sum of all such contact histories of the
particles of the shaft and the housing, diagrams such as this give a real feel
for the effects of varying either the load or the speed. Indeed in the case
shown it is seen that for particles on the housing, the variation in the direction
of the load reduces the contact stress and therefore the frictional and thermal

362
effects which would occur, as compared with the effects when the load
direction does not vary. It is from this type of argument that we can see that
the service conditions for a connecting rod bearing are somewhat less severe
than they would be for a bearing in which the same load was applied in a
constant direction.
Using our accumulated knowledge we shall now consider the types of
limits on load and speed which are likely to arise in tribological contacts.
In the following the emphasis is on the shapes of the load-speed limitations;
the actual position of each boundary would be defined by such factors as the
maximum allowable stress, temperature and wear rate. In all these diagrams
the load and speed are plotted logarithmically to indicate the shapes of the
actual limits and also to cover several orders of magnitude of the variables.

Strength Considerations (figure 13.3a)

Considering the geometry of the contact and the materials to be used we can
define the maximum load which may be safely applied to the contact. Thus
for a cylindrical or spherical contact the results from chapter 3 define our
safe working load, while with other geometries even simple calculations will
yield the required answer.

Inertia Characteristics (figure 13.3h)

Since the components of any tribological contact are in motion they are
subjected to stress fields arising from their inertia. This defines a maximum
speed above which failure could ensue. A simple example of this limitation
arises with rotating shafts which can be made to burst due to the induced
centrifugal stress at high rotational speeds. With more complex rolling
bearings the effects of centrifugal and gyroscopic forces can also impose
maximum speeds above which operation becomes unsatisfactory and
failures may occur due to bursting of the cage and similar phenomena.

Frictional Instabilities (figure 13.3c)

At very low operating speeds some tribological devices will suffer from the
type of instability discussed in chapter 7. For operation with such systems the
designer must either provide a satisfactory low speed combination or
restrict the system from such a low speed operation.

Heat Sources (figure 13.3d)

The heat release due to friction may be distributed between both the moving
and stationary contacting bodies. This has been discussed in chapter 3 so
that the thermal limit, as identified by a maximum allowable temperature for

363
llnW

ln.v In V
Stregth limit lnert1o force l1m1t lnstob1l1ty 11m11
(a) (b) (c)

~'
~·
ln.W

~
lnV lnV lnV
Thermal hmi r Wear hm1t Fot1gue l1m1t
(d) (el (f)

lnW lnW ln.W

HydrostatiC, .,.
------<
, , ...,
,, 'a
~'HydronamiC
,,
In V ln.V ln. V
Hydrodynamic f1lm E.HL film Hydrostatic + Hydrodynamic
limit limit lim1t
(g) (h) (I)

Figure 13.3 Load and speed limits on performance

the two components, would be as shown in figure 13.3d, where the load-
speed relationships for the two cases are based on equations 3.46 and 3.47
in chapter 3.

Wear Limit (figure 13.3e)

Many wear mechanisms such as adhesive or abrasive wear have been seen in
chapter 5 to be reasonably represented by the product of load and velocity.
The limit to define a specified wear value is therefore as shown in figure 13.3e.
This limit i~ often defined by the product of pressure and velocity, the P- V
limit, since the load per unit area, that is, pressure, is a more useful design
parameter than the load, which is only meaningful when associated with the
area over which it operates.

364
Fatigue Limit (figure 13.3/)
Fatigue limits are particularly relevant where such mechanisms lead to
failure, as in rolling contact bearings. In such bearings the fatigue will depend
on both the stress level and the frequency of loading. Practical tests show that
the relationship between load W and life L in revolutions is g\ven by W 3 L
for ball bearings and W4 L for roller bearings. Since L is clearly related to
speed V one can define the fatigue limit by W 3 V = Constant for ball-
bearings as shown in figure 13.3f.

Hydrodynamic Films (figure 13.3g)

When load is carried by virtue of the pressure generated by hydrodynamic
action, it has been seen in chapter 10 that the film thickness is some function
,v
of jW. In such cases the operational limit is clearly defined by the need for a
continuous film to cai:ry the applied load at the given speed using the appro-
priate lubricant. This immediately defines the limit of operation of such
systems by the line V /W is a constant. At the higher sliding velocities the
heat released due to viscous shearing will result in a reduction in the vis-
cosity of the fluid. This will result in a departure from linearity of the bounding
limit as shown in figure 13.3g.

Elastohydrodynamic Films (figure 13.3h)

From the results of chapter II it will be noted that this type of hydrodynamic
film is less dependent on the load than the pure hydrodynamic film and this
leads to a limit for film formation of the type shown in figure 13.3h.

Hydrostatic Films (figure 13.3t)

In chapter 12 it was shown that by using an external pumping system a load
carrying fluid film could be created, the maximum load carrying capacity
being entirely dependent on the external pressure available. This leads to a
limiting characteristic of the type shown in figure 13.3i by the line a, the slight
reduction at high speeds being again due to reductions in viscosity due to
frictional heating. In such devices it is also clear that at higher speeds some
additional hydrodynamic load carrying capacity may be generated so that
the total load capacity will be the sum of both the hydrostatic and hydro-
dynamic effects as shown in figure 13.3i.
The foregoing has not attempted to provide numerical values for the
various limits, since these clearly depend on such factors as the geometry,
the materials and the properties of the lubricant. None the less, for any
particular system the information in the preceding chapters of this book
would allow such calculations to be carried out. One interesting application
of these ideas is their use in determining the characteristics of various types
of journal bearing.

365
Stot1c strength limit

Max
speed
lim1t
Max
speed
11mit

ln.V ln. V In V
(rev/m1nl (rev/min} (rev/min)

Dry rubb1ng bg Rolling con toe t bg Hydrodynamic bg

Figure 13.4 Limits on three types ofjournal bearing

Hydrodynom IC bearing
Rolling contact beonng
Dry rubb1ng beonng

. -. ' . .
. -. .,
-~
20in dia
---------

10
rev/mm

Figure 13.5 Comparison of three bearing types

366
13.3 THE SELECTION OF JOURNAL BEARINGS
We shall consider three widely used types of journal bearings, the hydro-
dynamic bearing, the ball bearing, and the dry rubbing bearing based on
PTFE compounds. Using the limits described in figure 13.3 it is readily seen
that the basic characteristics for each of these bearing types will have the
form shown in figure 13.4. If we now calculate each of these limits for a given
diameter of shaft we may plot the ensuing characteristic on a load-speed
graph as shown in figure 13.5, the load and speed ranges being chosen to
cover almost the complete range of engineering applications. From such a
plot we can clearly see the ad vantages in load carrying capacity of ball bear-
ings over other types at speed of 1-2000 rev/min, which explains why they
are so often employed in this type of situation, for example, small electric
motors and the like. With large diameter shafts at the same order of speed,
the enhanced load capacity of hydrodynamic bearings is clearly demon-
strated. This explains why such bearings are so often used in large steam
turbines, etc. In such cases it is also common practice to use a form of auxiliary
hydrostatic jacking to preclude metallic contact at the lower speeds during
start up and stopping of such equipment. Bearing selection charts of this
kind have been constructed for a wide range of shaft sizes for both journal
and thrust bearings and are published by the Institution of Mechanical
Engineers. (Design Data, Item 65007, 66023).

13.4 MATCHING OF TRIBOLOGICAL SOLUTIONS

Finally one should always try to match the characteristics of the chosen
tribological solution to the particular problem in hand. Thus with large
gearbox bearings the load-speed characteristics might be of the form shown
in figure 13.6, and it is clear that the characteristics of the hydrodynamic

ln.W

,' Gearbox
.'./bearing
,;I' choroctenstic

7\T-;;e
, beonng
, ,' characteristic
, ,
, ,,

, ... ··'
ln. V
(rev/m~n)

Ffqure 13.6
367
bearing are a very good match for this requirement. With steam turbines the
load is substantially constant at all speeds, so that it would appear that a
form of hydrostatic bearing might be the most appropriate choice in such
situations.

13.5 CONCLUSIONS

The foregoing might be considered as the author's attempt to justify the

content of the preceding chapters, and to show the essential unity of the
subject of tribology. The most important result from the introduction of this
word is that we are made to concentrate on the essential problem of carrying
load, and to do this we should select the most appropriate tribological
solution. These solutions arise from the studies of engineering and science
in many disciplines as has been seen from the contents of this book, but this
should not detract from the essential unity of the subject which is concerned
with the function to be performed.

REFERENCES

I. M. J. Neale. Tribo/ogy Handbook. Butterworth, London, (1973).

2. M. J. Neale and A. B. Crease. Rubbing bearings for aircraft, a survey of applications,
materials and needs. M.O.D. Report P.E.S. and T.M., July, 1972.
3. Engineering Science Data. Items 68018, 66023, 65007, 67033, Instn mech. Engrs.

368
Appendix A
Reynolds Equation
The equations of motion for a viscous fluid are basic to the study of lubricant
films. These are normally referred to as the Navier-Stokes equations'
which, for a newtonian fluid, take the form

pDu =
Dt
X _ op + j_ [ ( 2 ou _ ~
ox ox 11 ox 3
L\)]
+ :y [1'/G; + ;:) J+ :z ['1( ~ + :)~ J
(A.l)

+-a ['1 (au

OX
- aw)J +-aya ['1(av-az+aw)J
az +ax- ay-
where
o a a a a
-=-+u-+v-+w-
Dt at ax ay az
L\ = au + av + OW
ax ay az
X, Y, Z = components of body forces in coordinate directions
u, v, w = components of velocity in coordinate directions

369
There is no general solution to the Navier-Stokes equations but in most
lubricated situations the conditions are such that the equations can be greatly
simplified. The usual situation involves two surfaces moving relative to each
other and separated by a film of lubricant as shown in figures A.l and A.2.

<< <<<<< "'"" ""'""' <<<<<<<

t -
Va
Uo

Figure A.J

Figure A.2
- u.

In order to simplify the Navier-Stokes equations a number of justifiable

assumptions must be made as listed below
(a) The flow is laminar; no vortices and no turbulence anywhere in the
lubricant film.
(b) The body forces are zero or at least negligible compared with the
viscous forces, that is, X = Y = Z = 0.
(c) Inertia forces are negligible compared with the viscous forces, that is,
Du Dv Ow
-=-=-=0
Dt Dt Dt
(d) The clearance is small compared with the dimensions of the lubricant
film in the x and z directions. This allows the curvature of the fluid film
to be ignored and also rotational velocities may be replaced by trans-
lational velocities.
(e) At any location (x, z) the pressure, density and viscosity are constant
across the lubricant film, that is
op = op = OIJ = 0
oy oy oy
370
(f) No slip at the bearing surfaces, that is, at the bearing surfaces the
velocity of the lubricant is identical with the surface velocity.
(g) Compared with the two velocity gradients ou/oy and owjoy all other
velocity gradients are negligible. This is a justifiable assumption since
u and w are usually much greater than v, and y is a much smaller dimen-
sion than x and z.
With these assumptions the Navier-Stokes equations reduce to

o2u 1 op
(A.2a)
oy 2 =~ox

02W 1 op
(A.2b)
oy2 '1 oz

Integrating equation A.2a with respect to y and using the boundary con-
ditions u = ua when y = 0 and u = ub when y = h gives

1 op (h - y) y
u = - - [y(y-
211 OX
h)J + -h- u a +-h ub (A.3a)

Integrating equation A.2b with respect to y and using the boundary con-
ditions w = 0 when y = 0 and w = 0 when y = h gives

1 op
w = - - [y(y - h)] (A.3b)
2'1 oz

Equation A.3b is the velocity distribution for a fluid flowing under laminar
conditions through a small clearance h as a result of the pressure gradient
op/oz. In fluid mechanics this type of pressure induced flow is referred to as
Poiseuille flow. In equation A.3a the first term on the right-hand side rep-
resents the Poiseuille flow while the two remaining terms describe the
velocity distribution which results from the motion of the bearing surfaces
alone, that is with constant pressure. In fluid mechanics this type of flow is
described as Couette flow.
The flow of lubricant must satisfy the continuity requirements expressed
by the continuity equation which, for steady flow, takes the form

Integrating the continuity equation across the lubricant film

f
h

-
ox
- d y+
o(pu) f h

- d y+
o(pv)
-
oy
f h

- - d y-
o(pw)
-
oz
- 0 (A.4)
0 0 0

371
Substituting u and w from equation A.3 into equation A.4 and completing
the integration yields

!___
OX
(ph3'1 oxop) + oz~ (ph3'1 ozop) = 6(Ua- Ub) o(ph)
ox
o(Ua + Ub)
+ 6ph ox + l2p(V,- V,)

If the bearing surfaces are inelastic in the direction of sliding Ua and Ub will
be independent of x, that is

and the above equation reduces to

!___
OX
(ph3'1 OXop) + ~OZ (ph3IJ OZop) = 6(U a
- u) o(ph)
b OX
+ 12p(V, - V,) (A.S)

Equation A.5 is the general form of Reynolds equation which is fundamental

to analysis of all lubrication problems.
In some situations it is possible to simplify Reynolds equation. If for
example it is reasonable to assume that the lubricant i,s incompressible,
that is, the density is constant, then equation A.5 reduces to

-OX(h
a -- 3

'1 OX
a --
ap) +- (h 3
op) = 6(U
OZ IJ OZ a
- U ) oh
b OX
- + 12(V.b - V)
a
(A.6)

If, in addition, it may be assumed that the viscosity ofthe lubricant does not
change then equation A.6 can be further simplified to

(A.7)

Some complex situations exist where it is difficult to ascribe the correct

values to the components of velocity. The difficulty arises because there may
exist velocity components in the y direction even though there is no normal
motion of the bearing elements. (Consider, for example, the case of two
rolling cylinders where, at all positions except the point of closest approach,
there is a component of the surface velocity in they direction.) We overcome
this problem by rewriting the Reynolds equation so that these components
are automatically included. If we consider a case such as that shown in
figure A.3, we see that there is no normal motion of the two solids and the
components of velocity are

in the x direction U a, U b cos(arc tan ::) ~ Ub

372
YL )( <<<<<<<<<<<<<<<

Figure A.3

m . 0,
. t he y d"1rect10n . (arctan dx
Ub Sin
dh) ~ ub dx
dh

The right-hand side of equation A.7 becomes

We may therefore rewrite the Reynolds equation for the case where there is
no normal motion of the bodies and the film thickness is invariant with time
(see chapter 10) as

a (h 3 op) a (h 3 op)
- - - + - - - =6(V +V)- oh
ox 11 OX 11oz oz I 2 OX

where U 1 and U 2 are the actual surface velocities, not the components.
If the two solids are moving in the y direction, the effect of this can be
added into the equation, thus

(A.8)

where V2 and V1 are the respective velocities in the y direction of the two
solids.

Sliding Motion
If we neglect pressure variation in the z direction, assume constant viscosity
and have pure sliding (VI = v2 = 0), the Reynolds equation becomes

d ( 3 dp)
dx h dx = 617(V, + V2)
oh
ox
373
Integrating with respect to x gives

h3 :~ = 617( U 1 + U 2 )h + constant

If dpjdx = 0 at some film thickness h*

0 = 617( U 1 + U 2 )h* + constant
or
constant = - 617( U 1 + U 2 )h*
Therefore

(A.9)

Normal Motion

and integrating with respect to x

dp
h3 dx = l2J7(V2 - V1 )x + constant
If x = x* where the pressure is a maximum that is, where dp/dx = 0, then
constant = -l2J7(V2 - V1 )x*
and

dp = 12 (V - V) (x - x*) (A.lO)
dx 11 2 1 h3

REFERENCES

I. 0. Pinkus and B. Sternlicht. Theory of Hydrodynamic Lubrication, McGraw-Hill,

New York, (1961).

374
Appendix B
Some Typical Problems
Example 1
A sample of a surface profile consists of 100 triangular asperities having a
constant flank angle IX and maximum peak valley heights of 1, 2, 3, ... , etc.
microns. Assuming that the valleys have a point distribution, that is, all
valleys lie at the same level, find the position of the centre-line and calculate
the C.L.A. value of the profile. Plot the all-ordinate distribution of this profile
and hence determine the R.M.S. and C.L.A. values. Does the order in which
the peaks occur affect your answer?
Solution
Let the centre-line be at distance h above the level of the valleys. By definition
of the centre-line h will be given by
h = total area enclosed by profile
profile length
or
100
L z 2 tan IX
h = 0~
2 L z tan IX
I

I 00.
[ z 2 tan IX dz 100
~ ----,-,10""0,_--- ~ - ~-tm
3
2 J z tan
0
IX dz

375
100
2 I (z - h) 2 tan rx
C.L.A. = ___ h_loo___ ----
2 L z tan rx
I
100

2 I (z- h)2 dz 16(100)3

~ ------;-;IO"'O_ _ _ = 81(100)2 = 19.75 J.lm
2 J z dz
0

For this surface the all-ordinate distribution curve is basically a right-angled

triangle. The centre-line of the distribution is at height 100/3 J.lm. This is
converted to a probability curve by making the area unity, that is, by using a
scaling factor 1/100. The first moment m 1 of one-half of the probability distri-
bution curve about the centreline is
2/3
2 J ztj.t(z) dz
0
that is,
C.L.A. = ~?
100 = 19.75 J.lm X

The second moment m2 about the centre-line is

2/3
m2 = J[z t/t(z)dz]
2 1' 2

- 1;3

Therefore
R.M.S. = 23.6 J.lm

Example 2
Consider the profiles of surfaces with (i) rectangular, and (ii) triangular height
distributions. Derive the standardised distribution functions for both cases,
and show that C.L.A. = (yi3/2) R.M.S. for the rectangular distribution and
C.L.A. = (J6/3) R.M.S. for the triangular distribution. Also show that for a
gaussian distribution C.L.A. = (2/rr) 112 R.M.S.
Solution
The distribution functions arc given by
. I
(I) t/t(z) = 2(3)' 2(J

.. ,J, I
(u) 'l'(z) = -- . - - ·· ·-
z
(6)"2(J 6(J2

I z
=(6)1!2(J
- - +6(J2
- - (6) 112 (J < z < 0

376
For the rectangular distribution
(3)'12u
1 (3)1/2
C.L.A. =2 f z 2(3)1i2u = -2- u
0

For the triangular distribution

C.L.A. ~ 2 Tz("
0
6 )~"u - 6;,) dz ~ !} u
J'
(6)1!2u 0

R.M.S. = J [z
0
2
( 6):' 2 u - 6; 2 )dz + J
-(6)1/2u
z 2 ((6):' 2 u + 6; 2 dz)
2
= u

For the gaussian distribution

1/t(z) = 1 e-z2/2u2
u(27t)1/2

C.L.A. - 2
-
f~ z x u( 21t) 112 e
1 -z2f2u2
dz -
- (2)1/2
~ u
0

Therefore
2)1/2
C.L.A. = (~ R.M.S.

Example 3
Discuss briefly the advantages and disadvantages of profilometry. The bear-
ing area of a surface represents the ratio of the real area of contact to the
nominal area; show that for surfaces with randomly distributed asperities
this is given by the line density 1/ L, where L is the total length of the surface
profile and I is that part of L which lies in regions of real contact.

Solution
Consider the surface to consist of closely placed parallel profiles dividing
the whole area into n differential strips in both the x and y directions.

377
If the line density of the ith strip in the X direction is denoted by Xi, then
for infinitesimally narrow strips of areas Ai the mean line density of the whole
surface in the x direction will be given by
n
IXjAj
X = =_:o:.__
::._i
n

This is clearly equal to the ratio of the real contact area a to the nominal
area A, that is, X= ajA. Similarly, the mean line density in they direction
will be given by
n
L l';Aj a
f i=O__
= ::.____::"
A

For surfaces with randomly distributed asperities, the line density of contact
of any profile lj L will be equivalent to the mean line density of contact, hence
- _ I a
X=Y=-=-
L A

Example 4
The contact of a rough surface and a smooth plane produces circular contact
spots of radius a. If a particular asperity carries a load P uniformly distributed
over the contact spot, find the vertical displacement at any point inside the
contact area and show that the maximum displacement is given by
2(1 - v2 )P
W=
naE

Solution
Referring to equation 3.13, the displacement of a point D due to the elemental
load acting on area dA is given by (figure B.l)

The length of the chord AB is 2a cos() and</> varies from 0 to n/2 so that

378
Figure B.l

since a sin () = r sin ¢1 we have

w=
4(1 - v2)P
2 E
n a
I(
n/2

1-
r2 .
2
a
Sill ¢1
)1/2
d¢1
0

The maximum displacement occurs at the centre of the contact area, that is
at r = 0, thus
2(1 - v2 )P
wmax = naE

ExampleS
A surface contact model consists of a smooth plane in contact with a rough
plane covered with spherical asperities of the same radius {3 and having a
uniform distribution of peak heights. If the asperity area density is '1 derive
expressions for the real area of contact A and the total load Pas functions of
the normal approach () for cases of both plastic and elastic modes of asperity
deformation. Show that A oc P for the plastic case and A oc P 415 for the
elastic case.

Solution
The uniform distribution of asperity peaks is given by
1
1/!(z) = 2(3)1i2a

For a plastic deformation mode

(3)1/la

A = 2nf31Ja I
d
(z - d) 2(3)1112 a dz

379
where a is the nominal area.
(3)1/2"

P = 2rtf31JaH f (z - d) 2(3)1112 0' dz

d

where H is the flow pressure (constant). Hence

1
A =- P or A oc. P
H

For an elastic deformation mode

(3)112"

A = 1t1Jf3a f (z - d) 2(3)\' 2 a dz
d

_ 1t1Jf3a 112 _ 2 _ 1t1Jf3a J2

- 4(3)1/20' ((3) 0' d) - 4(3)1/20'

and

f
(3)1/2"

p = ~ 3 .,,.,
,n112aE'
(z - d)
3/2 1 adz
2(3)1/2
d

4 11112 E' 4 a112 E'

= '11-' a (( 3)112a _ d)stz = '11-' a J512
15(3) a112 15(3) 112 a

or
A = constant x P415
A oc. p4!5

Example6

Two contacting discs of the same radius R roll under a normal load with
angular velocities w 1 and w 2 where w 2 > w 1 • One of the discs is assumed to
have a smooth surface while the surface of the other disc consists of spherical
asperities having equal heights and the same radius p and assumed to
deform plastically under a constant flow pressure H. If J1 is the coefficient of
friction and J is the mechanical equivalent of heat, find the surface tempera-
tures produced at a single contact spot when the load produces a normal
approach J between the two discs.

380
Solution
The contact spot may be considered as a stationary heat source with respect
to disc 1 and a moving heat source, having a sliding velocity (w 2 - w 1 )R,
with respect to disc 2.
The total rate of heat Q generated at the contact spot is given by
Q = JlPg(w 2 - w 1 )R
J
Now, if;, is that part of Q supplied to disc 1, then the heat supplied to disc 2
will be (1 - ;,)Q.
Using equations 3.40 and 3.42 we have

and
0 - 0.318(1 - A.)Q
z - a312(ClpcV)'t2

For a plastically deforming asperity the area of the contact spot A is given by
A = 2n{Ji5 = rra 2
thus, the radius of the contact spot is
a= (2/J(i)'l2
and the load supported by the contact spot is
P = 2rr{Ji5H
The surface temperatures can now be obtained as
2;,1mfibH g(w 2 - w 1 )R rri.JlH g(2{Jb) 112
e, = 4(2{Ji5)'12JCt,. - = - 4ClJ (w2- w,)R

and

02
= 0.318(1 - ;,)J1(2n{Ji5H)g(w 2 - w 1)R
(2{Jb) 314 J[Ct 2 p 2 c 2(w 2 - w 1)R] 112

_ 0.318rr(1 - i.)(2[Jc5)' 14 J1H.!f_ [ _ ] 112

- (Clzp2c2)l12j (wz w,)R

Example 7
A solid brass cylinder of 10 mm diameter with a conical end supports a
load of 1 kg and rubs against a smooth steel disc of 100 mm diameter,
rotating in air at 1 revolution per second. Describe qualitatively the expected
variation of friction and wear with time of rubbing.

381
Solution Hint
Initially the pressure over the contact is greater than H /3. See sections 4.1.3,
4.4.1, and 5.2.la.

ExampleS
Discuss the relative merits of the Edwards and Halling, and the Bowden and
Tabor friction theories.
Use the Bowden and Tabor theory to plot the friction coefficient 11 against
c (where c is the ratio of the critical shear stresses for surface film and sub-
strate) for different junction angles. Compare the results with those plotted
in figure 4.12.

Solution Hint
The Bowden and Tabor theory states that the friction force is the sum of an
adhesion term and a ploughing term. The adhesion term is given in section
4.5.1 and the ploughing term in section 4.7.

Example 9
Show from first principles that the adhesive wear rate is proportional to
load and inversely proportional to the hardness of the softer of the two rub-
bing materials.
Discuss the effect of surface films on the wear rate of metals.

Solution Hint
See section 5.2.la and 5.2.1b.

Example 10
Explain why 'wear rate' is not a meaningful parameter in describing the
useful life of rolling element bearings.

Solution Hint
See section 5.2.3a.

Example 11
Derive a relationship between the coefficient of friction and the wear rate for
two body abrasion.

Solution Hint
See sections 4. 7 and 5.2.2a.

382
Example 12

In a given rubbing situation it is found that the maximum allowable shear

stress for zero wear in 2000 passes is r. Using the IBM model for zero wear
obtain a graphical relationship between the number of passes and the maxi-
mum allowable shear stress.

Solution Hint
See section 5.2.8a.

Example 13

A plastic journal bearing 20 mm in diameter and 100 mm long carries a

radial load of 150 N. The wear rate of the bearing reaches the maximum
permissible level at a rotational speed of 500 rev/min.

(a) What is the maximum radial load which can be carried at 750 rev /min.?
(b) How should the dimensions of the bearing be changed to enable it to
carry the original load at 800 rev/min.?

Solution Hint
See section 6.5.3.
(a) 100 N.
(b) The length should be increased to 160 mm.

Example 14

A machine drive system is represented by figure 7.4, with values M = 300 kg

and k = 15 x 104 Nm- 1 . The slideways coefficient of friction J1. is measured
at various constant speeds, and the results show that, approximately, J1.
varies linearly from 0.13 at 0.002 ms- 1 to 0.12 at 0.004 ms- 1 . Dynamic
data for the friction variation is not available and the design is to be based on
the constant speed data.
Derive the equation of motion for the driven mass, at a drive speed
v = 0.003 ms- 1 , and calculate the minimum value of the dash pot coefficient
in Nm- 1 s if the overall damping ratio for the system, relevant to small tran-
sient oscillations, is to be at least 0.2.

Solution
The gradient of friction-velocity curve = - 14 700 Nm- 1s. The equation of
motion, measuring X from the initial position of the mass with the spring

383
undeflected, is

2
dt- 0.003) - 15
300 ddtX2 = - f (dX X dX)
104 (X- 0.0031)- ( 412- 14 700df

f- 14 700
( = 2(300 X 15 X 104 ) 112
f = 17 400 Nm- 1 s

Example 15

The system of example 14 is redesigned and the effective stiffness is increased

to a value k = 12 x 10 6 Nm- 1 . It is decided to obtain more realistic data
for the friction resistance offered by the slideways and dynamic tests are
carried out at various speeds and frequencies of oscillation. During oscillation
it is observed that, approximately, the friction resistance varies linearly with
speed and a sample of the results for the negative gradient, A, of the friction
velocity curve is as follows

(i) v = 0.002 ms- 1 , 20Hz, A= 3000 Nm- 1 s

(ii) v = 0.002 ms- 1 , 40Hz, A= 2000 Nm- 1 s
(iii) v = 0.004 ms- 1 , 20Hz, A = 2000 Nm- 1 s
(iv) v = 0.004 ms- 1, 40Hz, A = 1000 Nm -Is

Using linear interpolation between the values of the sample, calculate the
approximate value of the dashpot coefficient f to give the overall damping
ratio 0.2, in regard to small amplitude transient oscillations occurring at a
speed of 0.003 ms- 1 .

Solution
The undamped natural frequency for the redesigned system IS 32 Hz.
Assuming that oscillations occur at this frequency

(i) at v = 0.002 ms- 1 A= 2400 Nm -Is

(ii) at v = 0.004 ms- 1 A= 1400 Nm- 1 s

Interpolating for

v = 0.003 ms- 1
f- 1900
( = 2(300 X 12 X 106 ) 112
f = 25 900 Nm- 1 s

384
Example 16

A drive system similar to that shown in figure 7.4 with friction characteristics
as in figure 7.5a has the following parameter values

F.= 50 N
Fe= 30 N
M = 20 kg
k = 5 x 104 Nm- 1
A.=O
f = 400 Nm- 1s
The static friction, F., is assumed to be independent of the time of sticking.
(a) Determine whether or not stick-slip occurs at a speed v = 4 x 10- 3
ms- 1•
(b) Estimate the minimum speed above which stick-slip does not occur.

Solution
(a) For the parameter values and at a speed of 4 x w- 3 ms- 1, the values
of(, P, y (see section 7.5) are
f-A. /(
( = 2(Mk)l/2 = 0.2 w
n
= -M = 50 rad s- 1

y=~ 2 =0.98

From equation 7.35, w" yT1 = 3rr/2. Inserting these values in equation
7.33 shows that the slip velocity is negative at time T~> proving that
sticking will occur, and hence stick -slip oscillations will ensue.
(b) Trial values for v, using the method of part a show that the approxi-
mate critical velocity is 7.3 x 10- 3 ms- 1 .

Example 17

Two cylinders of equal weight Wand axial length I are placed at the top of a
plane inclined to the horizontal at an angle~- It may be assumed that the
plane is rigid and that the two cylinders are defined by diameters D 1 and
D2 , elastic properties E 1 v1 and E 2 v2 and hysteresis loss coefficients t: 1 and
t: 2 • Determine the ratio of the distances rolled by the two cylinders after any
time interval if both start from rest.
If the two cylinders in the above case have solid rubber tyres, one with
t: = 0.10, the other with t: = 0.05, but E and v are the same for both rubbers,
what would be the ratio of their radii if they travelled the same distance in
the same time?

385
Solution
Accelerating torques on cylinders are
( W sin fJ - F 1)R 1 = I 1 ¢1
where I 1 = (3/2)WRi
(W sin fJ- F 2 )R 2 = I 2 cij2

where I 2 = (3/2)WR~ also

4WR 1 (1 - vi}l 12
a1 =-- x ----
nl £1
4WR 2 (1 - v~) 12
a 2 = - - x ---=--
nl £1
From equation 8.1
2 Wa e
F - - -1-1
1 - 3 nR
I

2 Wa e
F - - -2-2
2 - 3 nR 2

Distance rolled from rest is c/>R = ¢Rt after t sees. Thus ratio of distance
rolled by two cylinders is

. 0--2 ( -e-
sm 1 )
a
3 nR 1 1
. 0 --
sm 2 (-
£2a)
3 nR 2 2

where a 1 and a2 are defined above.

If S 1 = S 2 and e 1 adR 1 = e2 a2 /R 2 , substituting gives

~=4
R2

Example 18
A sphere of diameter D and density p is made from a material of modulus E,
Poisson's ratio v and hysteresis loss coefficient e. The sphere rolls along a
horizontal plane of the same material with a linear velocity v. If it is assumed

386
that the heat released due to rolling friction is equally distributed between
the sphere and the plane what would be the temperature rise of the plane?
Assume that the thermal properties of the material are conductivity IX,
mechanical heat equivalent J, and specific heat c. What can you say about
the underlying assumption above as time elapses?

Solution
In equation 8.2

F=2_Wea
16 R
where

3
a=-WR - -
(1 - v2)1/3
2 E
Rate of working is Fv, so for the plane
1 Fv
Q=--
2 J
From equation 3.42 for moving heat source

(} = 0.318Q
a(a!Xpcv) 112
Since

these equations yield

_ 0.125R 2
f)----
e(pv)
---
112

] QIXC

Substituting for a gives

e ( p2v3RsE )1;16
(} = 0.092 X J X IX3C3(1 - V2)

After one revolution the temperature of the sphere is higher than the plane so
that the assumption is less valid as further distance is travelled.

Example 19
A 10 em diameter steel cylinder of axial length 1 em is subjected to a normal
load of 200 N and a resisting torque of 0.5 Nm. It rolls along a steel plane
at a uniform velocity of0.1 mjs. Calculate the discrepancy in distance travelled
in 100 seconds over what would have been expected from purely geometric

387
considerations. State the percentage of the contact area subjected to micro-
slip under these conditions. The coefficient of friction is 0.15, E = 200 x 10 9
Nm- 2 and v = 0.3 for steel.

Solution
Normal load P = 200 N
Tangential traction T = 10 N
T 10
J.l.P 30
From equation 8.12

(;y = 1 ;

; = GY 12
= 0.815

that is, 81.5 per cent of contact width is stick area so that 18.5 per cent of con-
tact width is slip area. Now
a = (8PR(l - v2))t;2
rt/E

= (8 X 200 X 0.05 X 0.91) 112 = l.08 X 10 _ 4 m

1t X 0.01 X 200 X 10 9
In equation 8.14
!1U
u
0.15 X 1.08
-------=---=-=---- [ 1 -
0.05
X 10- 4
(1 - w12]

= 6.16 X 10- 3
Due microslip
LlU = 6.16 X 10- 3 X 0.1
therefore
!is= !1Ut = 0.616 >< 10- 3 x 100 = 0.0616 m
Cylinder travels 0.0616 m less than 10m in 100 sec.

Example 20
A 1 em diameter steel ball rolls around the inner race of a ball-bearing having
a groove radius of 0.6 em and an overall diameter to the bottom of the
groove of 6 em. The normal load on the ball is 100 N and the torque applied
to the ball gives rise to an angular velocity ratio between the ball and the

388
race of value ¢. Determine the size of the contact zone and the value of ¢
when c = 0, that is, no slip occurs along the lines where y is one-quarter the
value of the major diameter of the contact ellipse. For this condition sketch
the shape of the stick and slip zones in the contact ellipse. Use E = 200
x 10 9 Nm- 2 , v = 0.3 and 11 = 0.2.

Solution
Rll = R,2 = 0.5 X w- 2 m
R2! = 3.0 X w- 2 m
R22 = -0.6 X w- 2 m
From equation 3.23
B - A = 1.0 X I0 2
B + A = 1.333 x 10 2
Hence
1.0
cosy = 1.33 = 0.75

y = 41°24'
From figure 3.15
ka = 2.08
Using equation 3.22 gives
a = 7.73 X l0- 4 m b = 2.24 x 10- 4 m
From equation 3.24
2 X 0.5 X 0.6 X l0- 4
R =
c (0.5 + 0.6) lO 2
= 0.546 X w- 2 m
From equation 8.21
y2
J1 (l/J- 6 )- 2RcR (I + l/J)
± Kc = +- c = ---...,....,----,.,----
- R 11 (¢ + 6)
Putting c = 0 and y = b/2 and Rc = 0.546 x 10- 2 gives a value of¢ as
6.00275/0.99725. Inserting this value of¢, 11 = 0.2 and R 11 = 0.5 x 10- 2 m
shows variation of c withy so that the stick-slip pattern may then be con-
structed.

Example 21
A circular hydrostatic thrust pad, 15 em diameter has a central circular recess
7.5 em diameter. The pad is designed to support a mass of 2500 kg with a
uniform clearance of0.05 mm. The lubricant is incompressible with dynamic
viscosity 0.1 poise and is supplied to the pad from a constant pressure source

389
at 5 x 10 6 Njm 2 via an external capillary restrictor. Calculate the volu-
metric flow rate of lubricant and the static stiffness of the pad under these
conditions. If, in operation, the bearing surfaces are not parallel, but inclined
at an angle of 5 x 10- 4 radians, calculate the new flow rate and the minimum
clearance between the bearing surfaces.

Solution
For Parallel Operation First, calculate a and q from equations 12.35 and
then p, from equation 12.28. The flow rate and stiffness can then be calculated
from equations 12.30 and 12.37b.
The relationship between flow rate and pressure drop in a capillary takes
the form Q = K 11P where K is a constant which can be calculated from the
parallel conditions.

For Tilted Operation From equation 12.35

w (a)
P, = Aa

Also
p h3
Q = K(ps - pr ) = _I _
I]
q

hence
P,
Pr = - - - - - - - , . - (b)
1 + (aRY !!_
K 11 a3
After substituting the known values equations a and b give two expressions
for p, as functions of a (q and a are functions of a). Using the data of figures
12.15a and 12.15b the two graphs of p, against iX can be plotted and the inter-
section defines the values of p, and a under tilted condition. The new flow
rate may now be calculated from the pressure drop across the restrictor. The
solution for a gives the central clearance and allows the minimum clearance
to be calculated.

Example 22

An orifice-controlled hydrostatic journal bearing of the type shown in

figure 12.18 supports a shaft of 8 em diameter, has an overall length of 6 em,
a recess length of 4 em and the width of the lands which separate the recesses
is 2 em. Fluid is supplied to the bearing at a pressure of 10 Njmm 2 and in the
unloaded condition the recess pressures are 6 Njmm 2 • If the radial clearance

390
between the shaft and bearing is 0.04 mm find the zero eccentricity stiffness
of the bearing. Estimate the rate of flow required to operate the bearing if the
viscosity of the fluid is 300 cP.

Solution
From section 12.10.3, the aspect ratio of bearing m is 0.25 and pressure
ratio r is 0.6.
From expression 12.44, or directly from figure 12.19, the dimensionless
stiffness of the bearing S0 , is 1.10 and the dimensional stiffness follows as
1.1 x 10 6 Njmm.
The flow rate required is 5.4 cm 3 js.

Example 23

A step bearing 20 em long, 50 em wide, moving at 50 cm/s is lubricated with

an oil of viscosity 10 x 10- 3 N sjm 2 • Design the bearing such that it has
optimum proportions to carry a load of 200 N. For this condition calculate
the coefficient of friction and the power consumption. If the bearing as
designed above were lubricated with a gas of viscosity 10- 5 N sjm 2 , calculate
the bearing number A.

Solution
From section 10.3.2 we have the ratio L 1 /L 2 • The smaller film thickness may
be calculated from
617UL
w =~0.342
0

and hence the step height from the optimum value of a

L 1 = 14.35 em L 2 = 5.63 em
step height = 0.28 mm

Surface tractions may be calculated from equations 10.24 and 10.25 since
we know dpjdx for each section. The traction on the lower surface is equal
to the traction on the upper plus the pressure at the step multiplied by the
step area. Why?

Coefficient of friction = 0.067

Power consumption = 0.67 W

For gas bearing A ~ 0.0018 assuming an atmospheric pressure of 1 bar.

This is so low that we could use incompressible theory to describe the action
of this bearing.

391
Example 24
A rotor weighing 2000 N, rotating at 3000 rev/min is supported symmetrically
by two hydrodynamic journal bearings. The journal shaft is 6 em diameter
and each bearing bush is 6 em long. Oil is supplied having a viscosity of
2 x 10- 3 N sec/m 2 . Can the bearing be designed to run such that the mini-
mum film thickness does not fall below 10 J.lm? What if the minimum film
thickness is permitted to fall to 7 J.lm? If either of these conditions is possible,
select a suitable clearance.

Solution
The only parameter we can change is the radial clearance C. The Sommerfeld
number S = (ryN /W)LD(R/C) 2 , the maximum permissible eccentricity s
depends on the clearance, but s also depends on the Sommerfeld number.
Since this precludes an analytical solution, we must guess a value of clearance
and see how the calculated eccentricity compares with the permissible value.
We may write S = constant/WC 2 and with our guessed value of C we
calculate the maximum permissible eccentricity. The value W in the expres-
sion is the load on a length L of an infinitely long bearing, while our bearing
has L/ D = I. We must therefore use chart 10.24a to ascertain the side leakage
factor. To achieve a load/bearing of 100 N we will have to have conditions
which would permit an infinitely long bearing to carry a larger load. There-
fore, W > 1000 N. CalculateS and hence, from table 10.2, find t:. This will
probably not be the same as the initially assumed value. What can be con-
cluded from comparison of the values? Select another clearance and repeat.
You will find that it is not possible to achieve the first condition, but that
the second may be obtained with a maximum radial clearance of about
26 J.lm.

Example 25
A steel strip 60 em wide is being transported on rollers 10 em diameter
rotating at 600 rev/min. At any given instant the strip is supported by 10
rollers which may be assumed to carry 100 N each. If each contact is fully
lubricated by oil of viscosity 5 x 10- 3 N s/m 2 and the strip moves at 50 cm/s,
calculate the minimum thickness of the oil film at each roller, the total side-
ways force on the strip and the total rate of energy input to the rollers.
Would an increase in viscosity decrease the energy consumption?

Solution
Find U 1 and U 2 and substitute in equation 10.12 to find h0 • Note that we
have to assume a value of ho to determine r:x from table 10.1. Do not forget
that the width is 60 em.
ho = 3.8 J.lm
392
Find Px from equation 10.32 and the frictional forces from 10.31. Once again
do not forget the width.
Px=3.17N
I'J(U 1 - U 2 )A = 0.12 N
F 1 = -3.18 N
F 2 = -3.16 N

Total force on steel = 31.6 N

Energy = Number of rollers x torque x angular velocity ~ 100 J/s.

Example 26
Two rollers each of length 50 em, diameter 40 em and rotational speed
30/n rev/min are loaded together. Two lubricants are available
lubricant A '1 = 10 X 10- 3 Ns/m 2 (X= 0.2 X 10- 7 m 2/N
B '1 = 12 X 10- 3 Ns/m 2 (X= 8 X 10- 7 m 2 /N
If the two rollers are made from steel, calculate the elastohydrodynamic
film thickness for the following combinations of load and lubricant
(1) 2.5 X 10 5 N, A
(2) 2.5 X 10 3 N, A
(3) 104 N, B,
(4) when one of the rollers is replaced by a rubber roller of the same di-
mensions, the load is 104 N and the lubricant is A. For steel E = 208
x 10 9 N/m 2 ,,u=0.3;forrubberE= 10 x 10 7 N/m 2 ;J..1=0.5.

Solution
Before calculating the film thickness, it is necessary to decide in which regime
the situation lies. For each case calculate qm, qa and Po (section 11.22).
Evaluate qm/qa and qm/Po· It will be found that each case lies in a different
regime. The dimensionless film thickness may then either be calculated by
means of the appropriate equation from 11.4-11.7 or more simply may be
read off the chart (figure 11.4). Film thicknesses are
(I) 0.072 Jlm
(2) 0.20 Jlm
(3) 0.059 Jlm
(4) 0.049 Jlm

393
Author Index

Abbott, E. J. 24, 25, 39 Clauss, F. J. 146

Aghan, R. L. Ill, 127 Clinton, W. C. 127
Amonton, G. 6, 72,92 Coffin, L. F. 106, 127
Anderson, W. J. 146 Coulomb, C. A. 6
Archard, G. D. 291,307 Cowking, E. W. 306, 307
Archard, J. F. 39, 71, 96, 105, 115, 123, Crease, A. B. 368
126, 127, 305, 306, 307 Crook, A. W. 297, 298, 307
Arnell, R. D. 137, 146
da Vinci, Leonardo 4
Harwell, F. T. 115, 127
Davies, G. H. B. 207, 231
Bayer, R. G. 113, 127
Davies, P. B. 359
Bell,J.C. 289,306
Davison, C. St C. 14, 15
Bell, R. 160, 172
Dawson, P. 307
Bingham, E. C. 232
Dean, E. W. 207,231
Bisson, E. E. 114, 127, 146
Derjaguin, B. V. 161, 173
Bitter, J. G. A. 127
Dowson, D. 14, 245, 262, 287, 291, 293,
Blok, H. 290, 295, 306
295,297,302,306
Bowden, F. P. 79, 82, 85, 86, 87, 89, 91,
Dubois, G. B. 267,287
92, 99, 114, 127, 129, 140, 201
Bragg, W. L. 136, 146 Dudley, B. R. 173
Braithwaite, E. R. 146 Dyson, A. 297, 307
Buckley, D. H. 93, 132, 146
Burdekin, M. 160, 172 Edwards, C. M. 87,93
Burwell, J. T. 95, 97, 119, 125, 126 EI-Refaie, M. 39

Cameron, A. 307 Firestone, F. A. 24, 25, 39

Catling, H. 172 Floberg, L. 245, 246, 262, 287
Cattaneo, A. G. 100, 127 Flynn, P. 291, 307
Cheng, H. S. 291, 307 Fuller, D. D. 265, 287, 359

394
Gair, F. C. 291, 307 Lewis, P. 291
Gohar, R. 307 Lindford, R. G. 110, 127
Goodier, J. N. 71, 127 Loeb, A. M. 359
Gough, V. E. 201
Green, A. P. 87, 93 Marsh, H. 285, 287
Greenwood, J. A. 71, 291, 293,297,307 Martin, H. M. 246, 287, 290, 297
Gross, W. A. 274,287 Matsubayshi, T. 172
Grubin, A. N. 291, 306 Merchant, M. E. 173
Meres, M. W. 287
Haines, D. J. 201 Merritt, H. E. 172
Halling, J. 14, 39, 71, 87, 93,201 Meyers, N. 0. 24, 39
Harris, J. H. 15 Midgeley, J. W. 137, 146
Hays, D. I. 262, 287 Mills, B. 173
Healey, A. J. 172 Mitchell, L.A. 110, 127
Heathcote, H. L. 182, 193 Moore,A.J.W. 117,127
Herrebrugh, K. 291,295,297, 307 Morgan, F. 287
Hertz, H. 48, 103, 123, 127 Muskat, M. 262,287
Higginson, G. R. 291,293, 295,297,
306, 307 Naylor, H. 297, 307
Hirn, G. 271, 287 Neale, M. J. 368
Hirst, W. 105, 115, 127, 291, 307 Nelson, C. W. 127
H.M.S.O. 15 Newcomb, T. P. 103, 127
Holm, R. 97, 126 Newton, I. 7
Hooke, R. 7 Newton, M. J. 359
Hordon, M. J. 118, 127 Nuri, K. A. 71
Howarth, R. B. 359 Nyquist, H. 156
Hunt,J.B. 172
Oberle, T. L. 102, 127
IBM 105, Ill Ocvirk, F. W. 267, 287
Ollerton, E. 201
Jakobsen, B. 262, 287 Orcutt, F. K. 297, 307
Jarvis, R. P. 173
Jefferis, J. A. 20 I Palmgren, A. 112
Johnson, K. L. 201, 293, 307 Pao, R. H. F. 359
Johnson, R. L. 146 Parnaby, J. 172
Jost, H. P. 10, II Peklenik, J. 39
Petroff, N. P. 7, 15
Kannel, J. W. 299, 307 Petrusevich, A. I. 291, 306
Kato, S. 172 Pinkus, 0. 251,259, 287, 374
Kauzlarich, J. J. 291, 307 Poritsky, J. 201
Kerridge, M. 96, 126 Porter, B. 172
King, R. F. 140, 146 Purday, H. F. P. 246, 258, 287
Kingsbury, A. 271, 287 Push, V. E. 173
Kirk, M. T. 305, 307
Kragelskii, I. V. 91, 93, 95, 105, 106, Rabinowicz, E. 109, 120, 127, 130, 146
126, 132, 146 Raimondi, A. A. 279, 287
Kruschov, M. M. 98, 102, 126 Rayleigh, Lord 238, 287
Ku, T. C. 127 Reason, R. H. 39
Reynolds, 0. 0. 7, 15, 261
Lancaster, J. K. 96, 105, 126, 127 Richardson, R. C. D. 102, 127
Leonard, R. 359 Rippel, H. C. 359
Levy, G. 110, 127 Rowe, C. N. 98, 127
Lewicki, W. 297, 307 Rowe, G. W. 114, 127

395
Saalfeld, K. 291, 306 Tavernelli, J. F. 106, 127
Samuels, L. E. Ill, 127 Taylor, P. L. 173
Sato, N. 172 Teer, D. G. 137, 146
Savage, R. H. 136, 146 Timoshenko, S. 71, 93, 127
Schaeffer, D. L. 146 Tolstoi, D. M. 173
Schumacher, R. A. 127 Torbe, I. 172
Sen Gupta, S. K. 201 Tower, B. 7, 15
Shaw, M. C. 127
Sibley, L. B. 297, 307
Sirico, J. L. 127 Vinogradova, I. E. 306
Sommerfeld, A. 247, 249, 287
Spencer,G.C. 172
Spurr, R. T. 103, 127, 173 Way, S. 306, 307
Starkman, E. S. I 00, 127 Wayson, A. R. 127
Sternlicht, B. 251,259, 287,291, 307, Weber, C. 291, 306
374 Weissenburg, K. 232
Stokes, G. 7 Welsh, N.C. 118, 127
Strang, C. D. 97, 119, 126 Whipple, R. T. P. 282, 287
Stringer,J.D. 172 Whitaker, A. V. 291,293, 307
Swift, H. W. 173 Whitehouse, D. J. 39
Swikert, M.A. 146 Whomes, T. L. 262, 287, 307
Williamson, J. B. P. 39,71
Tabor, D. 71, 79, 82, 85, 86, 87, 89, 91, Wilson, A. R. 297, 307
92, 117, 127, 129, 140, 146, 201 Wilson, R. W. 114, 127

396
Subject Index

abrasive wear I 00, 125 centre-line average roughness (C.L.A.)

adhesion 77 definition 22
adhesion theory of friction 79, 81 centre of pressure, plane inclined slider
Amonton's laws 72 242
aquaplaning 200 coefficient of friction 185
Archard's law of wear 96, 123 compatability, metallurgical 120
area of contact, real 73 complementary function 149
asperities, interaction of 87 compressibility number 274
asperity interlocking 77 conformity 175, 193
autocorrelation functions of surfaces connecting-rod bearing 363
components of 35 contact, ellipse of 181
definition of 32 of curved rough surfaces 67
power spectrum 35 of surfaces in sliding 41, 68
contact region 185
bearing area curve 24 contact zone 175; see also area of
bearing load capacity; see under load contact
capacity contaminant films 83
bearing number, gas lubrication 274 control 9
Bingham fluid behaviour 211 cornering force 199
boundary lubrication 116 Couette flow 371
braking traction 199 creep velocity 189
critical speed of shafts 283
crystal structure 120, 130
calcium fluoride as a high temperature
lubricant 139 damping ratio 148
carburising 117 deformation of surfaces 40
cavitation 244 criterion for 65
cavitation boundary condition 246 elastic 61
centre-line, definition of 22 plastic 62

397
density-temperature relationship for films, soft 117
fluids 270 finite length bearings 260
describing function 156 finite journal bearing 264
design, application of wear finite rolling and sliding discs 262
relationships to 122 finite slider bearings 261
design data 367 foil bearing 282
dilitancy 213 forced oscillations 149
dimensionless groups, four-ball tester 226
elastohydrodynamic 293 free rolling 175
disc machine 223, 296 fretting; see under wear
discs 243 friction 72
displacements of surfaces in contact, adhesion theory of 77
horizontal 47, 48 kinetic 72
vertical 47, 48 laws of 6, 72
distribution, gaussian 26 measurement of 74
rectangular (uniform) 28 sliding 183, 30 I
triangular (linear) 29 static 72
distribution function 26, 29, 30 variation of, with speed, for
kurtosis 31 negative gradient case 152
median of 30 for stiction case 151
moments of 27 friction coefficient 185
skewness of 30 friction force 72
standard deviation of, definition of 26 friction force models 158
variance 26 dynamic effects in 160
driving traction 199 hysteresis effect in 160
friction instability 363
eccentricity ratio 247 frictional oscillations, amplitude of 154
economics I 0 effect of drive speed on 153
effective viscosity 302 frequency of 154
elastic deformation of lubricated
surfaces 290 gas lubrication 271
elastic hysteresis 78, 178 bearing number 274
elastic work 178 graphite 136
elasticity of lubricants 213 greases 228
elastohydrodynamic limit 365 additives to 231
elastohydrodynamic lubrication 200, bearing performance tests on 231
288 et seq. consistency of 229
electrical contact resistance 114 drop point of 230
elemental cylinder 189 extreme pressure properties of 231
elliptic contact 181 mechanical stability of 231
end effects in finite length bearings 260 oil separation of 230
energy losses 9 oxidation stability of 230
environment 10
equivalent cylinders 244 half-speed whirl 282, 285
erosive wear; see under wear hard coatings 117
externally pressurised bearings; see hardness 97, 100
hydrostatic lubrication effect of, on friction and wear 129
Heathcote slip 182, 193
fatigue in elastohydrodynamic Hertz equations 103
lubrication 306 hertzian contact 177, 291
fatigue limit 365 pressure distribution of 49, 56
film shape 292, 298 stress distribution of 50
film thickness, elastohydrodynamir 293 three-dimensional 56
measurement of 296 two-dimensional 48

398
high temperature solid lubricant 139 lubrication; see self-lubricating materials
historical survey 4 lubrication, boundary 116
hydrodynamic limit 365 high temperature 139
hydrodynamic lubrication 233 et seq.
hydrostatic limit 365 macroroughness 18
hydrostatic lubrication 308 et seq. maximum load 363
hydrostatic journal bearings, effect of metallic surfaces 16
rotation on 356 layers in 16
evolution of 350 maps of 18, 19
multipad 351 roughness of 17
radial stiffness of 353 waviness in 18
multirecess 354 microroughness 18
radial stiffness of 356 microslip 184
hydrostatic thrust bearings, capillary mineral oils 224
compensation of 318 acidity of 225
characteristics of 325 additives to 226
compensation in, need for 315 alkalinity of 225
constant flow compensation of demulsibility of 226
312,325 extreme pressure properties of 226
control of; see compensation in- flashpoint of 225
effects on, of changes in film foaming of 226
thickness 327 oxidation stability of 225
effects of sliding in 340 pour point of 226
flow, load and power factor of 334 relative density of 224
non-parallel operation of 344 SAE classification of 209
orifice compensation of 323 specific heat of 225
pressure ratio of 320, 326 thermal conductivity of 225
restrictors in; see compensation in- molybdenum disulphide 137
stiffness of 314
Navier-Stokes equation 369
inertia effects 363 newtonian behaviour 203
inlet position, journal bearing 252 nitriding 117
instability, hydrodynamic 282 non-newtonian behaviour 209
interference methods for film thickness in fluids 304
measurement 305 normal approach, between a sphere and
inverse hydrodynamic solution 291 a plane 60
isocline 156 between rough surfaces 61
definition of 59
'Johnson' chart 295 normal motion 237
journal bearing 247 optimum film thickness ratio of plane
gas lubricated 277 inclined slider 241
junction growth 82 oxide films 114
lamellar solids 134 parabolic cylinder 246
lead monoxide as a high temperature partial journal bearing 250
lubricant 139 gas lubricated 279
load, effect on wear of 119 particular integral 149
load capacity, discs 246 perturbation methods 154
journal bearing 251 phase-plane analysis 155
plane inclined slider 241 phosphating 117
Rayleigh step 243 piecewise-linear approximation 155
lobed bearing 287 planned obsolescence 8
locomotive wheels 198 plastic deformation 78
lubricants, elasticity of 213 plastic fluid behaviour 211

399
plasticity index 66 self-lubricating materials 133
plastic work 179 advantages of 133
plastics 139 types of 134
effects of lubrication on 145 shear stress, critical 80
effects of temperature on 145 in fluids 254
filled and reinforced 145 short bearing theory 267
friction of 140 side leakage, elastohydrodynamic 304
P-V factor of 141 side leakage factor 260
for flat bearing 141 skidding 200
for sleeve bearing 142 sliding friction 183, 301
wear coefficients of 144 elastohydrodynamic 300
wear of 140 slider bearing, gas lubricated 274
ploughing 79, 89 plane inclined 239
pneumatic tyres 198 Rayleigh step 242
point contact, elastohydrodynamic 304 slip angle 199
Poiseuille flow 371 slip areas 189
pressure distribution, soft films 117
elastohydrodynamic 292 solid lubricants 117
experimental measurement of 299 solubility 130
of plane inclined slider 240 Sommerfeld number 251
of Rayleigh step bearing 243 Sommerfeld solution for journal
pressure spike in elastohydrodynamic bearing 249
lubrication 292 Sommerfeld transformation 247
profilometry 19, 20 spiral groove bearing 282
pseudo-plastic behaviour 212 squeeze motion 237
P-Vfactor 124, 141 statistical properties of surfaces 25
all-ordinate distribution of 25
Rayleigh step bearing 242 height distribution of 25
real area of contact 61 step bearing, gas lubricated 275
of rough surfaces 62 stick areas 189
reduced pressure 253 stick-slip 150, 151
regimes of elastohydrodynamic elimination of 171
lubrication 295 stick-slip oscillation, analysis of 161
relaxation time of fluids 304 zero damping case for 166
Reynolds equation, for combined sliding effect of stick time on 168
and squeeze motion 237 stick-slip pattern 195
for gas lubrication 273 stress distribution of surfaces in
for normal motion 374 contact 41
for sliding motion 235, 373 due to a single normal load 42
for squeeze motion 237 due to a single tangential load 43
road surface 200 due to a single uniformly distributed
rolling contact, wear of 103 load 44,45
rolling friction 300 isochromatics of 46, 52, 53
rolling resistance 178 sulfinuzing 117
root-mean square value (R.M.S.) of surface, distribution of slope on 37
roughness, definition of 22 metallic; see under metallic surface
rough surfaces, real area of contact surface asperities 17
of 62 area distribution of 24
Routh-Hurwitz criterion 171 distribution of curvatures of 36
Rowe's wear theory 98 distribution of peaks of 35
running-in 125 distribution of slope of 37
surface contact, effect of tangential
self-aligning torque 199 tractions on 54
self-excited oscillation !50 stick-slip phenomena in 55

400
surface films 98, 113, 131 capillary 214
surface parameters 22 cone and plate 221
measurement of 38 efflux 217
surface roughness 73 Engler 217
surface strains 185 falling cylinder 220
surface texture 18 falling sphere 219
methods of assessment of 19 Ostwald 216
surface valleys 22, 24 Redwood 217
distribution of 35 relative 215
surface velocities 236 rolling sphere 220
surfaces; see asperities; autocorrelation rotating cylinder 221
functions of surfaces; centre Saybolt 217
line, definition of; defor- viscosity 202
mation of surfaces; displace- dynamic 203
ments of surfaces in contact; index 207
statistical properties of kinematic 204
surfaces; stress distribution measurement of 213
of surfaces in contact units of 204
synchronous whirl 282, 283
wear 9, 94 et seq.
tangential traction 184 abrasive 100, 125
Talysurf 20 law of 100
temperature effects 132 adhesive 95
induced phase change 118 law of 96
on hardness 118 cavitation erosion 109
on stability of oil 119 corrosive 106, 125
thermal effects, due to moving heat definition of 94
source 69 effect of temperature on 118
due to sliding surface contact 68 erosive 108
due to stationary heat source 68 by fluids 109
in hydrodynamic bearings 268 by solid particles 108
thermal limits 363 fatigue 103
thermal wedge 269 sliding contact I 05
thixotropy 211 fretting I 07
tilting pad bearing, gas lubricated 280 IBM Engineering model of Ill,
traction, elastohydrodynamic 300 123
fluid 254 IBM Engineering model of zero
of journal bearings 258 Ill
of rolling and sliding discs 257 IBM Engineering model of
of slider bearings 256 measurable 113
tractions 175 measurement of 120
trajectory !56 mild 114
transient oscillations 149 prevention of 122
tread pattern 200 rolling contact 103
tribology I Rowe's theory of 98
tyre-road contacts 198 severe 115
spark erosion I 09
undamped natural frequency 148 under vacuum 114
wear limit 364
vacuum, wear under 114
wear particles, size of I 09
variable viscosity 252, 290
whirling of shafts 283
visco-elastic fluids 213
damping of 284
viscometer, absolute 215
band 220 yield pressure 79, 96
Bingham 215 Young's modulus 102

401