William 

Murphy · Jonathan Black
Garth Hastings Editors

Handbook of
Biomaterial
Properties
Second Edition
Handbook of Biomaterial Properties
William Murphy • Jonathan Black
Garth Hastings
Editors

Handbook of Biomaterial
Properties
Second Edition
Editors
William Murphy Jonathan Black
Department of Biomedical Engineering Principal: IMN Biomaterials
University of Wisconsin King of Prussia, PA, USA
Madison, WI, USA

Department of Orthopedics and

Rehabilitation
University of Wisconsin
Madison, WI, USA

Garth Hastings
Staffordshire University (Emeritus)
Lyme, Staffordshire, UK

ISBN 978-1-4939-3303-7 ISBN 978-1-4939-3305-1 (eBook)

DOI 10.1007/978-1-4939-3305-1

Library of Congress Control Number: 2016933771

© Springer Science+Business Media New York 2016

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the
editors give a warranty, express or implied, with respect to the material contained herein or for any errors
or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer Science+Business Media LLC New York
Foreword

Progress in the development of surgical implant materials has been hindered by the
lack of basic information on the nature of the tissues, organs and systems being
repaired or replaced. Materials’ properties of living systems, whose study has been
conducted largely under the rubric of tissue mechanics, has tended to be more
descriptive than quantitative. In the early days of the modern surgical implant era,
this deficiency was not critical. However, as implants continue to improve and both
longer service life and higher reliability are sought, the inability to predict the
behavior of implanted manufactured materials has revealed the relative lack of
knowledge of the materials properties of the supporting or host system, either in
health or disease. Such a situation is unacceptable in more conventional engineering
practice: the success of new designs for aeronautical and marine applications
depends exquisitely upon a detailed, disciplined and quantitative knowledge of ser-
vice environments, including the properties of materials which will be encountered
and interacted with. Thus the knowledge of the myriad physical properties of ocean
ice makes possible the design and development of icebreakers without the need for
trial and error. In contrast, the development period for a new surgical implant, incor-
porating new materials, may well exceed a decade and even then only short term
performance predictions can be made.
Is it possible to construct an adequate data base of materials properties of both
manufactured materials and biological tissues and fluids such that in vitro simula-
tions can be used to validate future implant designs before in vivo service? While
there are no apparent intellectual barriers to attaining such a goal, it clearly lies in
the distant future, given the complexity of possible interactions between manufac-
tured materials and living systems.
However, a great body of data has accumulated concerning the materials aspects
of both implantable materials and natural tissues and fluids. Unfortunately, these
data are broadly distributed in many forms of publication and have been gained
from experimental observations of varying degrees of accuracy and precision. This
is a situation very similar to that in general engineering in the early phases of the
Industrial Revolution. The response then was the publication of engineering hand-
books, drawing together, first in general publication and later in specialty versions,

v
vi Foreword

the known and accepted data of the time. In this spirit, we offer this 2nd Edition of
the Handbook of Biomaterial Properties
Biomaterials, as manufactured for use in implants, do not exist usefully out of
context with their applications. Thus, a material satisfactory in one application can
be wholly unsuccessful in another. In this spirit, the Editors have given direction to
the experts responsible for each part of this Handbook to consider not merely the
intrinsic and interactive properties of biomaterials but also their appropriate (and in
some cases inappropriate) applications as well as their historical context. The
experts have in some cases added significant content specific to each class of mate-
rial. For example, content is included on not just bulk properties but also surface
modifications of titanium, which has become quite important in orthopedic implant
design. It is hoped that the results will prove valuable, although in different ways, to
the student, the researcher, the engineer and the practicing physician who uses
implants.
A handbook like this necessarily becomes incomplete immediately upon publi-
cation, since it will be seen to contain errors of both omission and commission.
Such has been the case with previous engineering handbooks: the problem can only
be dealt with by providing new, revised editions. This 2nd Edition provides updated
insights, data, citations, and topic areas not found in the 1st Edition. The Editors
would appreciate any contributions and/or criticisms which the users of this hand-
book may make and promise to take account of them in future revisions.
Introduction

It is a feature of any developing science and its accompanying technology that infor-
mation relating to different aspects is scattered throughout the relevant, and sometimes
not so relevant literature. As the subject becomes more mature, a body of information
can be categorized and brought together for the use of practitioners. In providing this
Handbook of Biomaterial Properties the Editors believe that the latter stage has been
reached in several parts of the vast field of biomaterials science and engineering. This
2nd Edition of the Handbook provides an updated body of information for a subset of
chapters, as well as new chapters that represent biomaterials fields that have matured
in the intervening time between the 1st and 2nd Editions.
Many of the properties of the synthetic materials have been available for some
time, for example those of the various metallic alloys used in clinical practice have
been specified in various International, European and National Standards and can be
found by searching. In the case of polymeric materials, while the information is in
commercial product literature and various proprietary handbooks, it is diverse by
the nature of the wide range of materials commercially available and the search for
it can be time consuming. The situation is much the same for ceramic and composite
materials: there the challenge is finding the appropriate properties for the specific
compositions and grades in use as biomaterials.
However, when information is sought for on materials properties of human tis-
sues, the problem is more acute as such data are even more scattered and the meth-
ods for determination are not always stated or clearly defined. For the established
worker this presents a major task. For the new researcher it may make establishing
a project area a needlessly time consuming activity. The biomaterials bulletin boards
(on the Internet) frequently display requests for help in finding characterization
methods and/or reliable properties of natural materials, and sometimes the informa-
tion is actually not available. Even when it is available, the original source of it is
not always generally known.
In approaching their task, the Editors have tried to bring together into one source
book the information that is available. To do this they have asked for the help of
many colleagues worldwide to be contributors to the Handbook. It has not been pos-
sible to cover all the areas the Editors had hoped. Some topics could not be covered,

vii
viii Introduction

or the information was judged to be too fragmentary or unreliable to make it worth

including. This is inevitably the sort of project that will continue to be incomplete;
however, new information will be provided as more experiments are done and as
methods for measurement and analysis improve. The aim has been to make this
Handbook a ready reference which will be consulted regularly by every technician,
engineering and research worker in the fields of biomaterials and medical devices.
We have tried, not always successfully, to keep the textual content to a minimum,
and emphasize tabular presentation of data. However, in some cases it has been
decided to include more text in order to establish the background of materials prop-
erties and use, and to point to critical features in processing or production which
would guide the worker looking for new applications or new materials. For exam-
ple, in polymer processing, the need to dry materials thoroughly before fabrication
may not be understood by those less well versed in production techniques. In
another example, the relationship of synthetic biomaterials to vocal fold or corneal
tissue repair may not be obvious by those outside of this specialized area, so text is
included to clarify the connection of these natural biomaterials to synthetic bioma-
terials used as devices.
It is hoped that the Handbook will be used and useful, not perfect but a valuable
contribution to a field that we believe has matured sufficiently to merit such a pub-
lication. The fields embodied in the 1st Edition of The Handbook have now further
matured to justify this updated 2nd Edition - inevitable technological improvement
in materials design and characterization have led to updated data, and tissue engi-
neering applications have created a need for reference data in new tissue areas (e.g.
ligament, cornea, vocal fold). The Handbook is divided into synthetic and natural
materials and the treatment is different in each part. More background was felt to be
needed for many of the synthetic materials, since processing and structural varia-
tions have a profound effect on properties and performance. Biological performance
of these materials depends on a range of chemical, physical and engineering proper-
ties and the physical form can also influence in vivo behavior. We have not attempted
to deal with issues of biological performance, or biocompatibility, but have dealt
with those other features of the materials which were felt to be relevant to them as
potential biomaterials. Only materials having apparent clinical applications have
been included.
The biological materials have more dynamic properties since, in vivo, they
respond to physiological stimuli and may develop modified properties accordingly.
The treatment of their properties has been limited to those determined by well char-
acterized methods for human tissues, with a few exceptions where data and other
species are deemed to be applicable and reliable. Those properties determined
almost totally in vitro may not be directly predictive of the performance of the living
materials in vivo, but are a guide to the medical device designer who wishes to
determine a device design specification. Such a designer often finds it hard to realize
the complexity of the task of dealing with a non-engineering system. What really
are the parameters needed in order to design an effectively functioning joint endo-
prosthesis or a heart valve? Do tissue properties measure post explantation assist? Is
individual patient lifestyle an important factor? There is immediately a degree of
Introduction ix

uncertainty in such design processes, and total reliability in performance cannot be

given a prospective guarantee. However, the more we learn about the materials and
systems of the human body and their interaction with synthetic biomaterials, the
closer we may perhaps become to the ideal ‘menotic’ or forgotten implant which
remains in ‘menosis’—close and settled union from the Greek μϵνω—with the
tissues in which it has been placed.
Two final comments:
First, although the Editors and contributors frequently refer to synthetic and, in
some cases, processed natural materials as ‘biomaterial,’ nothing herein should be
taken as either an implied or explicit warrantee of the usefulness, safety or efficacy
of any material or any grade or variation of any material in any medical device or
surgical implant. Such determinations are an intrinsic part of the design, develop-
ment, manufacture and clinical evaluation process for any device. Rather, the mate-
rials listed here should be considered, on the basis of their intrinsic properties and,
in many cases, prior use, to be candidates to search as biomaterials: possibly to
become parts of successful devices to evaluate, direct, supplement, or replace the
functions of living tissues.
Second, the Editors earlier refer to absences of topics and of data for particular
synthetic or natural materials. While this may be viewed, perhaps by reviewers and
users alike, as a shortcoming of the Handbook, we view it as a virtue for two
reasons:
• Where reliable data are available but were overlooked in this edition, we hope
that potential contributors will come forward to volunteer their help for hoped for
subsequent editions.
• Where reliable data are not available, we hope that their absence will prove both
a guide and a stimulus for future investigators in biomaterials science and
engineering.
Indeed, this 2nd edition of the Handbook provides new data and citations in more
than 10 chapters, new contributors, and also includes new chapters in areas that have
more recently matured (e.g. the vocal fold). We are hopeful that this new Edition
adds value to the 1st Edition, and that future Editions will address inevitable “short-
comings” in this Edition
The Editors, of course, welcome any comments and constructive criticism.

Madison, WI, USA William Murphy

King of Prussia, PA, USA Jonathan Black
Lyme, Staffordshire, UK Garth Hastings
Contents

Foreword ........................................................................................................... v
Introduction ....................................................................................................... vii
Contributors ...................................................................................................... xv

PART I
A1 Cortical Bone ........................................................................................... 3
J. Currey
A2 Cancellous Bone ...................................................................................... 15
Christopher J. Hernandez
A3 Dentin and Enamel ................................................................................. 23
K.E. Healy
B1 Cartilage................................................................................................... 37
J.R. Parsons
B2 Fibrocartilage .......................................................................................... 45
V.M. Gharpuray
B3 Ligament and Tendon ............................................................................. 55
Connie S. Chamberlain and Ray Vanderby
B4 Skin and Muscle ...................................................................................... 63
A.F.T. Mak and M. Zhang
B5 Brain Tissues ........................................................................................... 67
S.S. Margulies and D. F. Meaney
B6 Arteries, Veins and Lymphatic Vessels .................................................. 77
X. Deng and R. Guidoin

xi
xii Contents

B7 The Intraocular Lens .............................................................................. 103

Traian V. Chirila and Shuko Suzuki
C1 Blood and Related Fluids ....................................................................... 115
V. Turitto and S.M. Slack
C2 The Vitreous Humor ............................................................................... 125
Traian V. Chirila and Ye Hong
C3 The Cornea .............................................................................................. 135
Traian V. Chirila and Shuko Suzuki

PART II

1a Metallic Biomaterials: Introduction ...................................................... 151

H. Breme, V. Biehl, Nina Reger, and Ellen Gawalt
1b Metallic Biomaterials: Cobalt-Chromium Alloys ................................ 159
Gopinath Mani
1c Metallic Biomaterials: Titanium and Titanium Alloys ....................... 167
H. Breme, V. Biehl, Nina Reger, and Ellen Gawalt
1d Dental Restoration Materials ................................................................. 191
Jonathan Black and Garth Hastings
2 Composite Materials .............................................................................. 205
L. Ambrosio, G. Carotenuto, and L. Nicolais
3 Thermoplastic Polymers In Biomedical Applications: Structures,
Properties and Processing ...................................................................... 261
S.H. Teoh, Z.G. Tang, and Garth W. Hastings
4 Biomedical elastomers ............................................................................ 291
J.W. Boretos and S.J. Boretos
5 Oxide Bioceramics: Inert Ceramic Materials
in Medicine and Dentistry ...................................................................... 339
J. Li and G.W. Hastings
6 Ceramic Materials Testing and Fracture Mechanics .......................... 353
D. Daily
7 Properties of Bioactive Glasses and Glass-ceramics ............................ 447
L.L. Hench and T. Kokubo
8 Wear ......................................................................................................... 455
M. LaBerge and J.D. Desjardins
9 Degradation/resorption in Bioactive Ceramics in Orthopaedics ........ 495
H. Oonishi and H. Oomamiuda
Contents xiii

10 Corrosion of Metallic Implants.............................................................. 509

M.A. Barbosa
11 Carbons .................................................................................................... 549
A.D. Haubold, R.B. More, and J.C. Bokros

PART III
1 General Concepts of Biocompatibility .................................................. 563
D.F. Williams
2 Soft Tissue Response ............................................................................... 571
J.M. Anderson
3 Hard Tissue Response............................................................................. 581
T.O. Albrektsson
4 Immune Response ................................................................................... 593
K. Merritt
5 Cancer ...................................................................................................... 607
M. Rock
6 Blood–material Interactions .................................................................. 621
S.R. Hanson
7 Soft Tissue Response to Silicones........................................................... 631
S.E. Gabriel
8 Vocal Folds ............................................................................................... 645
Joel Gaston and Susan L. Thibeault

Index ................................................................................................................. 659

Contributors

T.O. Albrektsson Handicap Research, Institute for Surgical Sciences, University

of Gothenburg, Gothenburg, Sweden
L. Ambrosio Department of Materials and Production Engineering, University of
Naples Federico II Institute of Composite Materials Technology CNR, Piazzale
Technio, Naples, Italy
J.M. Anderson Department of Pathology Case Western Reserve University,
University Hospitals of Cleveland, Cleveland, OH, USA
M.A. Barbosa INEB-Ma do Campo Alegre, Porto, Portugal
V. Biehl Lehrstuhl für Metallische Werkstoffe, Universität des Saarlandes,
Saarbrücken, Germany
Jonathan Black Principal: IMN Biomaterials, King of Prussia, PA, USA
J.C. Bokros Medical Carbon Research Institute, Austin, TX, USA
J.W. Boretos Consultants for Biomaterials, Rockville, Maryland, USA
S.J. Boretos Consultants for Biomaterials, Rockville, Maryland, USA
J. Breme Lehrstuhl für Metallische Werkstoffe, Universität des Saarlandes,
Saarbrücken, Germany
G. Carotenuto Department of Materials and Production Engineering, University
of Naples Federico II Institute of Composite Materials Technology CNR, Piazzale
Technio, Naples, Italy
Connie S. Chamberlain Department of Orthopedics and Rehabilitation, Wisconsin
Institute for Medical Research, University of Wisconsin-Madison, Madison,
WI, USA

xv
xvi Contributors

Traian V. Chirila Queensland Eye Institute, South Brisbane, QLD, Australia

J. Currey Department of Biology, York University, York, YOl 5DD, UK
X. Deng Laboratorie de Chirugie Exp Agriculture Services, Universite Laval,
Quebéc, G1K 7P4, Canada
J.D. DesJardins Department of Bioengineering, Rhodes Clemson University,
Clemson, SC, USA
S.E. Gabriel Division of Rheumatology, Mayo Clinic, Rochester, MN, USA
Joel Gaston Department of Biomedical Engineering, University of Wisconsin—
Madison, Madison, WI, USA
Ellen Gawalt Department of Chemistry and Biochemistry, Bayer School of
Natural and Environmental Sciences, Mellon, Pittsburgh, PA, USA
V.M. Gharpuray Department of Bioengineering, Clemson University, Clemson,
SC, USA
R. Guidoin Laboratorie de Chirugie Exp Agriculture Services, Université Laval,
Quebéc, Canada
S.R. Hanson Division of Hematology/Oncology, Emory University, AJ, Atlanta,
GA, USA
Garth Hastings The Biomaterials Programme Institute of Materials Research and
Engineering, National University of Singapore, Singapore, Singapore
A.D. Haubold Medical Carbon Research Institute, Austin, TX, USA
K.E. Healy Department of Biological Materials, Northwestern University,
Chicago, IL, USA
L.L. Hench Division of Material Chemistry Faculty of Engineering, Imperial
College Department of Materials, Kyoto University, Sakyo-ku Kyoto, Japan,
London, UK
Christopher J. Hernandez Sibley School of Mechanical and Aerospace
Engineering, Cornell University, Ithaca, NY, USA
Meinig School of Biomedical Engineering, Cornell University, Ithaca, NY, USA
Ye Hong Cooper Vision, Pleasanton, CA, USA
T.M. Keaveny Department of Mechanical Engineering Etcheverry Hall, University
of California at Berkeley, Berkeley, CA94720, USA
T. Kokubo Division of Material Chemistry Faculty of Engineering, Imperial
College Department of Materials, Kyoto University, Sakyo-ku Kyoto, Japan,
London, UK
Contributors xvii

M. LaBerge Department of Bioengineering, Rhodes Clemson University,

Clemson, SC, USA
R.E. Levine Musculoskeletal Research Centre, University of Pittsburgh, Pittsburgh,
PA, USA
J. Li Centre for Oral Biology, Karolinska Institute, Huddinge, Sweden
A.F.T. Mak Rehabilitation Engineering Centre Hong Kong Polytechnic, Hunghom,
Kowloon, Hong Kong
Gopinath Mani Department of Biomedical Engineering, University of South
Dakota, Sioux Falls, SD, USA
S.S. Margulies Department of Bioengineering, University of Pennsylvaniaa,
Philadelphia, PA, USA
D.F. Meaney Department of Bioengineering, University of Pennsylvania,
Philadelphia, PA, USA
K. Merritt Stoneridge Dr., Gaithersburg, MD, USA
R.B. More Medical Carbon Research Institute, Austin, TX, USA
L. Nicolais Department of Materials and Production Engineering, University of
Naples Federico II Institute of Composite Materials Technology CNR, Piazzale
Technio, Naples, Italy
H. Oomamiuda Department of Orthapaedic Surgery Artificial Joint Section and
Biomaterial Research Laboratory, Osaka-Minami National Hospital, Osaja, Japan
H. Oonishi Department of Orthapaedic Surgery, Artificial Joint Section and
Biomaterial Research Laboratory, Osaka-Minami National Hospital, Osaka, Japan
J.R. Parsons Orthopaedics-UMDNJ, 185 South Orange A venue University
Heights, Newark, NJ, USA
Nina Reger Department of Chemistry and Biochemistry, Bayer School of Natural
and Environmental Sciences, Mellon, Pittsburgh, PA, USA
M. Rock Department of Orthopaedics, Mayo Clinic, Rochester, MN, USA
S.M. Slack Department of Biomedical Engineering, University of Memphis,
Memphis, TN, USA
Shuko Suzuki Queensland Eye Institute, South Brisbane, Queensland, Australia
Z.G. Tang BIOMAT Centre, National University of Singapore, Singapore,
Singapore
S.H. Teoh Institute of Materials Research & Engineering – IMRE, National
University of Singapore, Singapore, Singapore
xviii Contributors

Susan L. Thibeault Department of Surgery, University of Wisconsin–Madison,

Madison, WI, USA
Department of Biomedical Engineering, University of Wisconsin–Madison,
Madison, WI, USA
Department of Communication Sciences and Disorders, University of Wisconsin–
Madison, Madison, WI, USA
V. Turitto Department of Biomedical Engineering, University of Memphis,
Memphis, TN, USA
Ray Vanderby Jr. Department of Orthopedics and Rehabilitation, Wisconsin
Institute for Medical Research, University of Wisconsin-Madison, Madison, WI,
USA
Department of Biomedical Engineering, Wisconsin Institute for Medical Research,
University of Wisconsin-Madison, Madison, WI, USA
D.F. Williams Department of Clinical Engineering, Royal Liverpool University
Hospital, UK
S.L. -Y. Woo Musculoskeletal Research Center, University of Pittsburgh,
Pittsburgh, PA, USA
M. Zhang Rehabilitation Engineering Centre, Hong Kong, Polytechnic, Hungham,
Kowloon, Hong Kong
Part I
Chapter A1
Cortical Bone

J. Currey

A1.1  Composition

A1.1.1  Overall

The main constituents are the mineral hydroxyapatite, the fibrous protein collagen,
and water. There is some non-collagenous organic material.
Highly mineralized bone (petrosal bones of some non-human mammals) has
little organic material (8% in the horse petrosal to 3% in the tympanic bulla) [3].
(Almost certainly human ear bones will be somewhere near or in this region, though
they seem not to have been studied.)

A1.1.2  Organic

The main organic component is collagen. Most is Type I, but there are small amounts
of Type III and Type VI, found in restricted locations [4]. Slowly heated collagen
shrinks at a particular temperature, giving an indication of the stability of the mol-
ecules. Bone collagen in men has a shrinkage temperature of about 61.5°–63.5°C up
to the age of about 60, but about 60°C over that age. Bone from women showed
much greater variability [5]. About 10% of the bone organic material is non-­
collagenous, mainly non-collagenous protein, NCP. The main ones are listed below.
They have supposed functions that change rapidly.
• Osteocalcin (OC), or bone Gla protein (BGP)
• Osteonectin (ON), or SPARC
• Osteopontin (OPN) or secreted phosphoprotein I (SPPI)

J. Currey (*)
Department of Biology, York University, York YOl 5DD, UK

© Springer Science+Business Media New York 2016 3

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_1
4 J. Currey

Table A1.1  Composition of Cortical Bone

Water Organic Ash Source
Mass % 12.0 28.1 59.9 [1]
Volume % 23.9 38.4 37.7 [1]
Volume % 15.5 41.8 39.9 [2]

Table A1.2  Density of Cortical Bone

Wet bone 1990 kg m-3 [1]

• Bone sialoprotein (BSP)

The relative amounts of these proteins can vary greatly. Ninomiya et al. [6]
report far more osteocalcin (31 times) in cortical bone than in trabecular bone, and
far more osteonectin (29 times) in trabecular bone than in cortical bone.

A1.1.3  Mineral

The mineral has a plate-like habit, the crystals being extremely small, about 4 nm by
50 nm by 50 nm. The mineral is a variant of hydroxyapatite, itself a variant of cal-
cium phosphate: Ca10(PO4)6(OH)2 [7]. The crystals are impure. In particular there is
about 4–6% of carbonate replacing the phosphate groups, making the mineral tech-
nically a carbonate apatite, dahllite, and various other substitutions take place [8].

A1.1.4  Cement line

The cement line round Haversian systems (secondary osteons) contains less calcium
and phosphorus, and more sulphur than nearby parts of bone. This may indicate the
presence of more sulphated mucosubstances, making the cement line viscous [9].

A1.2  Physical Properties

A1.2.1  Density

A1.2.2  Electromechanical behavior

Strained bone develops electrical potential differences. These used to be attributed

to piezoelectric effects. However, the size of the piezoelectric effects is small com-
pared with those produced by streaming potentials [10]. Furthermore, there were
A1  Cortical Bone 5

various anomalies with the potentials generated, which did not always accord with
theory. The consensus now is that ‘SGPs’ (stress-generated potentials) are over-
whelmingly caused by streaming potentials [10, 11]. Scott and Korostoff [12] deter-
mined, amongst other things, the relaxation time constants of the stress generated
potentials, which varied greatly as a function of the conductivity and viscosity of the
permeating fluid. As an example of their findings: a step-imposed loading moment
which produced a peak strain of 4 × 10-4 induced an SGP of 1.8 mV, yielding a value
of the SGP/strain ratio of 4500 mV. The SGP decayed rapidly at constant strain,
reaching zero within about one second. For more detail, the complex original paper
must be consulted.

A1.2.3  Other Physical Properties

Behari [10] gives a useful general review of many ‘solid state’ properties of bone,
both human and non-human, many of which are not dealt with here. These proper-
ties include the Hall effect, photo-electric effects, electron paramagnetic resonance
effects and so on.

A1.3  Mechanical Properties

A1.3.1  General

There is a great range for values in the literature for many reasons. Amongst
these are:
(a) Different treatment of specimens  Drying bone and then re-wetting it pro-
duces some small differences [13], as does formalin fixation [14]. Testing bone dry
produces results quite different from those in wet bone; dry bone is stiffer, stronger,
and considerably more brittle. Very small samples produce values for stiffness and
strength less than those from larger samples [15, 16]. High strain rates generally
produce a higher modulus of elasticity, a higher strength [17], and a greater strain to
failure than specimens tested at low strain rate.
(b) Different age and health of donors  Age may affect intrinsic properties.
Osteoporotic bone may differ from ‘normal’ bone in ways other than the fact that it
is more porous; there is evidence that the collagen is different from that in similar-­
aged non-osteoporotic subjects [18]. Bone from osteogenesis imperfecta patients
has a higher proportion of Type III and Type V collagen compared with Type I col-
lagen, than bone from normal subjects [19]. Bone collagen from osteopetrotic sub-
jects is in general older than that from normal subjects, and has correspondingly
different properties [5].
6 J. Currey

(c) Differences between bones, and sites in the bones  The ear bones (ossicles)
and portions of the temporal bones (petrosals) are highly mineralized, and will
undoubtedly be stiffer and more brittle than others (though they seem not to have
been investigated in humans). Long bones differ along their length and around their
circumference. The distal femur is less highly mineralized and weaker in tensile and
compressive static loading, and at any level the posterior part is similarly less min-
eralized and weaker [20].
The values reported below should be considered paradigmatic, that is, to be valid
for a well-performed test on bone obtained from a middle aged person with no dis-
ease. Other values are reported in such a way as to make it clear how some property
is a function of other features of the specimen.

A1.3.2  Stiffness

(a) General  There are two ways of testing bone: mechanically by relating stresses
to strains; ultrasonically, by subjecting the bone to ultrasound and measuring the
velocity of the sound. From a knowledge of the density one can then obtain a stiff-
ness matrix. If this is inverted it becomes a compliance matrix, the reciprocal of the
individual terms of which are equivalent to the so-called technical moduli derived
by mechanical testing [21]. Reilly and Burstein [22] give mechanical values, and
Ashman et al. [23] give ultrasonic measurements. Reilly and Burstein [22] assumed
transverse isotropy (that is, symmetry around the longitudinal axis of the bone),
while Ashman et al. [23] assumed orthotropy (that is, that the values for stiffness
could be different in the longitudinal, radial and tangential directions).
Reilly and Burstein [22] give values for Young’s modulus at a number of inter-
mediate angular orientations, but they do not form a very uniform set.
(b) Tensile modulus versus compressive modulus  Reilly et al. [24] tested femo-
ral specimens specifically to determine whether the value for Young’s modulus was
different in tension and compression. A paired Student’s ‘t’ test showed no signifi-
cant difference between the compressive and tensile moduli at the 95% confidence
level. Calculations on their data show the the 95% confidence interval ranged from
compression modulus 1.72 GPa higher to tension modulus 0.27 GPa higher. The
load-deformation traces showed no change of slope going from compression into
tension and vice versa.
(c) Very small specimens  The bending modulus of very small specimens was 6.62
GPa [5].
(d) Locational variations: Metaphysis versus diaphysis  Young’s modulus has
been determined in three-point bending for extremely small plates (7 mm by 5 mm
by (about) 0.3 mm) from the femoral metaphyseal shell and from the diaphysis of
the same bones [16].
A1  Cortical Bone 7

Table A1.3  Mechanical Properties

Femur Tension [23] Femur Tension [22] Femur Compression [22]
Elastic moduli (GPa):
E1 12.0 12.8 11.7
E2 13.4 12.8 11.7
E3 20.0 17.7 18.2
Shear moduli* (GPa):
G12 4.5 – –
G13 5.6 3.3 –
G23 6.2 3.3 –
Poisson’s ratios:
ν12 0.38 0.53 0.63
ν13 0.22 – –
ν23 0.24 – –
ν21 0.42 0.53 0.63
ν31 0.37 0.41 0.38
ν32 0.35 0.41 0.38
Subscript 1: radial direction relative to the long axis of the bone, 2: tangential direction, 3: longi-
tudinal direction.
* Shear values are included under tension for convenience.

Table A1.4  Locational Variations in Modulus

Location Longitudinal (GPa) Transverse (GPa) Source
Metaphysis 9.6 5.5 [16]
Diaphysis 12.5 6.0 [16]

The differences between these values and those reported by Reilly and Burstein
[22] are probably attributable not to the difference in testing mode, since bending
and tension tests from the same bone generally give similar values for Young’s
modulus, but to the very small size of the specimen, and to the rather low density of
the specimens.
(e) Compression; effect of mineral  The compressive behavior of cubes, relating
the properties to the density of the specimens gives, using ρa (fat-free mass divided
by anatomical volume, g cm-3) as the explanatory variable:
Young’s modulus (GPa) = 3.3ρa2.4 for compact bone [25].
The higher values of ρa were of the order of 1.8 g cm-3(=1800 kg m-3); this equa-
tion [25] predicts a value of 13.5 GPa for such a specimen. Multiple regression
analysis showed that the dependence of Young’s modulus on density was caused by
8 J. Currey

Table A1.5  Moduli of Osteons

Modulus (GPa) Longitudinal Osteons ‘Alternate’ Osteons Source
Tension 11.7 5.5 [26]
Compression 6.3 7.4 [27]
Bending 2.3 2.6 [28]
Torsional* 22.7 16.8 [29]
* Values for an 80-year-old man excluded.

the effect of porosity on density, and that, in these specimens, the effect of mineral
content was insignificant.

(f) Single secondary osteons  Ascenzi and co-workers [26–29] distinguish two
types of secondary osteon: ‘longitudinal’ osteons, whose collagen fibres have a basi-
cally longitudinal orientation, and ‘alternate’ osteons, whose fibres have markedly
different courses in neighboring lamellae. (This difference is a contentious issue.)
N.B.: These studies of Ascenzi and co-workers [26–29] are widely quoted, so
beware of some apparent anomalies (apart from changes in nomenclature between
papers). The bending modulus is remarkably low compared with the tension and
compression moduli. The torsional (shear) modulus is remarkably high, compared
both with the shear modulus values obtained by others (above), and with the tension
and compression values. Torsional moduli are expected, on theoretical grounds, to
be less than the tension and compression moduli. Furthermore, the large differences
between the tension and compression moduli have not been reported elsewhere.]

(g) Strain rate effects  Calculations [30], incorporating data from non-human as
well as human material, predict that Young’s modulus is very modestly dependent
upon strain rate:

E = 21402(strain rate (s -1 ))0.050 MPa

[N.B. statements about strain rate effects in bone are suspect unless it is clear that
the workers have taken machine compliance into account!]

(h) Viscoelastic-damage properties  Viscoelastic time constant (the value τ (s) in

the equation):

Î (t ) = b1 exp[to - t  t ] + b 2

where the betas are parameters, t is time (s), to is time at which the specimen is held
at a constant stress below the creep threshold: 6.1 s [31]. For reference, its value in
bovine bone: 3.6 s.
A1  Cortical Bone 9

Table A1.6  Strength of Cortical Bone [22]

Mode Orientation Breaking Strength (MPa) Yield Stress (MPa) Ultimate Strain
Tension Longitudinal 133 114 0.031
Tangential 52 – 0.007
Compression Longitudinal 205 – –
Tangential 130 – –
Shear 67 – –

Table A1.7  Locational Variations in Strength

Location Longitudinal (MPa) Transverse (MPa) Source
Metaphysis 101 50 [16]
Diaphysis 129 47 [16]

Table A1.8  Strength of Osteons

Strength (MPa) Longitudinal Osteons ‘Alternate’ Osteons Source
Tension 120 102 [26]
Compression 110 134 [27]
Bending 390 348 [28]
Torsional* 202 167 [29]
* Values for an 80 year old man excluded.

A1.3.3  Strength

(a) Overall
(b) Combined loading  Cezayirlioglu et al. [32] tested human bone under com-
bined axial and torsional loading. The results are too complex to tabulate, but should
be consulted by readers interested in complex loading phenomena.

(c) Metaphysis versus diaphysis  Same specimens as reported for modulus above
(Table A1.4) [16]. ‘Tensile’ strength calculated from the bending moment, using a
‘rupture factor’ to take account of the non-uniform distribution of strain in the
specimen.
(d) Effect of mineral  Keller [25], using the same specimens as above, provides the
following relationship:

Strength = 43.9 r a2.0 ( MPa )

10 J. Currey

[N.B.: The effect of mineralization, as opposed to density, is possibly of impor-

tance here; the original paper must be consulted.]

(e) Single secondary osteons  The same nomenclature applies as for moduli of
osteons (Table A1.5).
[N.B. The bending strengths and torsional strengths seem very high, even bearing
in mind that no allowance has been made in bending for non-elastic effects.]

(f) Strain rate effects  Bone will bear a higher stress if it is loaded at a higher strain
(or stress) rate. Carter and Caler [17] found an empirical relationship that failure
.
stress (σf (MPa)) was a function of either stress rate((s ) ) or strain rate (Î) :

s f = 87(s )0.053

.
s f = 87 (Î) 0.055

N.B. These relationships imply an increase of 44% in the failure stress if the
stress rate is increased one thousandfold. This relationship has been found to be
roughly the same in other, non-human, mammals.

(g) Creep  Creep threshold (the stress below which no creep occurs): 73 MPa [31].
The equivalent value for bovine bone is 117 MPa [31]. Specimens in tension or
compression were held at particular stresses [33]. The time (seconds) to failure is
given as a function of normalized stress (stress/Young’s modulus (MPa/MPa)):

Tension : Time to failure = 1.45 ´ 10 -36 (normalized stress)-15.8

Compression : Time to failure = 4.07 ´ 10 -37 (normalized stress)-17.8

(h) Fatigue  Some workers report the log of the number of cycles as a function of
the applied stress levels, some report the log cycle number as a function of log stress
levels, and some report log stress levels as a function of log cycle number. [The last
seems wrong, since the applied stress can hardly be a function of the number of
cycles the specimen is going to bear, but it is frequently used in fatigue studies. It is
not possible simply to reverse the dependent and independent axes because the
equations are derived from regressions with associated uncertainty.] The variation
between the results for different testing modes is considerable.
Carter et al. [34] report on the effect of Young’s modulus of elasticity and poros-
ity in their specimens. They find that Young’s modulus is positively associated with
fatigue life, and porosity is negatively associated:
A1  Cortical Bone 11

Table A1.9  Effect of Remodeling [35]

Property Primary Osteons Haversian Osteons
Tensile Strength (MPa) 162 133
Ultimate Strain 0.026 0.022
Young’s modulus (GPa) 19.7 18.0

log Nf = -2.05 log Dso (S.E. 0.599)

log Nf = -4.82 log Dso + 0186
. E (S.E. 0.387)
log Nf = -2.63 log Dso - 0.061 P (S.E. 0.513)
log Nf = -4.73 log Dso + 0160
. E - 0.029 P (S.E. 0.363)
where Nf: number of cycles to failure; Δσ0: initial stress range (these experiments
were carried out under strain control, so stress range decreased as damage spread
and the specimens became more compliant); E: Young’s modulus (GPa); P: porosity (%).
Incorporating Young’s modulus into the equation has a marked effect in reducing
the standard error; porosity has a much less strong effect.
[N.B. Many workers normalize their data in an effort to reduce the effect that
variations in Young’s modulus have in increasing the scatter of the results.]
Choi and Goldstein [15] provide alternate, somewhat higher values.

(i) Effect of remodeling  Vincentelli and Grigorov [35] examined the effect of
Haversian remodelling on the tibia. The specimens they reported were almost
entirely primary or Haversian, with few specimens having a scattering of secondary
osteons. [Unfortunately they probably (it is not clear) allowed their specimens to
dry out, so it is not sure that bone in vivo would show the same behavior. However,
their results are similar to those found in nonhuman specimens.]

Additional Reading
Cowin, S.C. (ed.)(1989) Bone Mechanics Boca Raton: CRC Press.
A more rigorous, less chatty and less biologically, oriented approach than the
following books by Currey and by Martin and Burr. The chapters on mechanics (2,
6 and 7), written by Cowin himself, are particularly authoritative.
Currey, J.D. (1984) The Mechanical Adaptation of Bones Princeton: University
Press.
Out of print, new edition in preparation. Tries to deal with all aspects of mechani-
cal properties of bone as a material and of whole bones. Not overly technical.
Written from a general biological perspective, thus, does not concentrate on human
material.
12 J. Currey

Martin, R.B. and Burr, D.B. (1989) Structure, Function and Adaptation of
Compact Bone New York: Raven Press.
There are not many values of mechanical properties here, but the treatment of the
biology of bone, and of fatigue of bone tissue, is excellent and the discussion of
remodeling, although now somewhat out of date, is a very good introduction to this
intellectually taxing topic.
Nigg, B.M. and Herzog, W. (eds)(1994) Biomechanics of the Musculoskeletal
System John Wiley: Chichester.
Deals with many aspects of biomechanics, including locomotion, with an empha-
sis on human material. There is a full treatment of the measurement of many biome-
chanical properties.

References

1. Gong, J.K., Arnold, J.S. and Cohn, S.H. (1964) Composition of trabecular and cortical bone.
Anat. Rec., 149, 325–331.
2. Biltz, R.M. and Pellegrino, E.D. (1969) The chemical anatomy of bone I. A comparative study
of bone composition in sixteen vertebrates. J. Bone Joint Surg., 51A, 456–466.
3. Lees, S. and Escoubes, M. (1987) Vapor pressure isotherms, composition and density of
­hyperdense bones of horse, whale and porpoise. Con. Tiss. Res., 16, 305–322.
4. Keene, D.R., Sakai, L.Y. and Burgeson, R.E. (1991) Human bone contains type III collagen,
type VI collagen, and fibrillin: Type III collagen is present on specific fibers that may mediate
attachment of tendons, ligaments, and periosteum to calcified bone cortex. J. Histochem.
Cytochem., 39, 59–69.
5. Danielsen, C.C., Mosekilde, Li., Bollerslev, J. et al. (1994) Thermal stability of cortical bone
collagen in relation to age in normal individuals and in individuals with osteopetrosis. Bone,
15, 91–96.
6. Ninomiya, J.T., Tracy, R.P., Calore, J.D., et al. (1990) Heterogeneity of human bone. J. Bone
Min. Res., 5, 933–938.
7. Lowenstam, H.A. and Weiner, S. (1989) On Biomineralization, Oxford University Press,
New York.
8. McConnell, D. (1962) The crystal structure of bone. Clin. Orthop. Rel. Res., 23, 253–68.
9. Burr, D.B., Schaffler, M.B. and Frederickson, R.G. (1988) Composition of the cement line and
its possible mechanical role as a local interface in human compact bone. J. Biomech., 21,
939–945.
10. Behari, J. (1991) Solid state bone behavior. Prog. Biophys. Mol. Biol., 56, 1–41.
11. Martin, R.B. and Burr, D.B. (1989) Structure, Function, and Adaptation of Compact Bone,
Raven Press, New York.
12. Scott, G.C. and Korostoff, E. (1990) Oscillatory and step response electromechanical phenom-
ena in human and bovine bone. J. Biomech., 23, 127–143.
13. Currey, J.D. (1988) The effects of drying and re-wetting on some mechanical properties of
cortical bone. J. Biomech., 21, 439–441.
14. Sedlin, E.D. (1967) A rheological model for cortical bone. Acta Orthop. Scand., (Suppl. 83),
1–78.
15. Choi, K. and Goldstein, S.A. (1992) A comparison of the fatigue behavior of human trabecular
and cortical bone tissue. J. Biomech., 25, 1371–1381.
16. Lotz, J.C., Gerhart, T.N. and Hayes, W.C. (1991) Mechanical properties of metaphyseal bone
in the proximal femur. J. Biomech., 24, 317–329.
A1  Cortical Bone 13

17. Carter, D.R. and Caler, W.E. (1985) A cumulative damage model for bone fracture. J. Orthop.
Res., 3, 84–90.
18. Bailey, A.J., Wotton, S.F., Sims, T.J. et al. (1993) Biochemical changes in the collagen of
human osteoporotic bone matrix. Con. Tiss. Res., 29, 119–132.
19. Bateman, J.F., Chan, D., Mascara, T. et al. (1986) Collagen defects in lethal perinatal osteo-
genesis imperfecta. Biochem. J., 240, 699–708.
20. Saito, S. (1983) (Distribution of the X-ray density, compressive and tensile breaking strength
in the human femoral shaft) Die Verteilung von Dichte, Druck und Festigkeit im menslichen
Femurschaft. Anat. Anzeiger Jena, 154, 365–376.
21. Cowin, S.C. (1989) Bone Mechanics, CRC Press, Boca Raton.
22. Reilly, D.T. and Burstein, A.H. (1975) The elastic and ultimate properties of compact bone
tissue. J. Biomech., 8, 393–405.
23. Ashman, R.B., Cowin, S.C., Van Buskirk, W.C. (1984) A continuous wave technique for the
measurement of the elastic properties of cortical bone. J. Biomech., 17, 349–361.
24. Reilly, D.T. and Burstein, A.H. (1974) The mechanical properties of cortical bone. J. Bone
Joint Surg., 56A, 1001-1022
25. Keller, T.S. (1994) Predicting the compressive mechanical behavior of bone. J. Biomech., 27,
1159–1168.
26. Ascenzi, A. and Bonucci, E. (1967) The tensile properties of single osteons. Anat. Rec., 158,
375–386.
27. Ascenzi, A. and Bonucci, E. (1968) The compressive properties of single osteons. Anat. Rec.,
161, 377–391.
28. Ascenzi, A., Baschieri, P., and Benvenuti, A. (1990) The bending properties of single osteons.
J. Biomech., 23, 763–771.
29. Ascenzi, A., Baschieri, P., and Benvenuti, A. (1994) The torsional properties of single selected
osteons. J. Biomech., 27, 875–884.
30. Carter, D.R. and Caler, W.E. (1983) Cycle-dependent and time-dependent bone fracture with
repeated loading. J. Biomech. Eng., 105, 166–170.
31. Fondrk, M., Bahniuk, E., Davy, D.T. et al. (1988) Some viscoplastic characteristics of bovine
and human cortical bone. J. Biomech., 21, 623–630.
32. Cezayirlioglu, H., Bahniuk, E., Davy, D.T. et al. (1985) Anisotropic yield behavior of bone
under combined axial force and torque. J. Biomech., 18, 61–69.
33. Caler, W.E. and Carter, D.R. (1989) Bone creep-fatigue damage accumulation J. Biomech., 22,
625–635.
34. Carter, D.R., Caler, W.E., Spengler, D.M. et al. (1981) Uniaxial fatigue of human bone. The
influence of tissue physical characteristics. J. Biomech., 14, 461–70.
35. Vincentelli, R., and Grigorov, M. (1985) The effect of haversian remodeling on the tensile
properties of human cortical bone. J. Biomech., 18, 201–207.
Chapter A2
Cancellous Bone

Christopher J. Hernandez

A2.1  Microstructure

Cancellous bone (also referred to as trabecular bone or spongy bone) is a porous

cellular solid consisting of platelike and rodlike struts called trabeculae. The size
and arrangement of trabeculae vary among species and within regions of the skel-
eton and change with age. Average trabecular thickness can be as great as 300 μm
but, in elderly human tissue, ranges from 100 to 200 μm [1]. The orientation of
trabeculae within cancellous bone varies, resulting in considerable specimen-
to-­specimen heterogeneity. At the continuum level (specimens 3–5 mm in smallest
dimension) the density of cancellous bone is measured as the mass of the specimen
(wet after removing the marrow) divided by specimen volume and is referred to as
the “apparent density.” The apparent density of human cancellous bone typically
ranges from 0.05 to 1.1 g/cm3. The apparent density of cancellous bone is not to be
confused with the “tissue density” which expresses the density of individual tra-
beculae. The volume fraction of human cancellous bone (expressed in the bone
literature as BV/TV) ranges from 5 % to 60 %. The surface-to-volume ratio of
human cancellous bone (BS/TV) is related to bone volume fraction in the follow-
ing manner [2]:

0.70
BS æ BV ö
= 8.84 ç ÷
TV è TV ø

C.J. Hernandez (*)

Sibley School of Mechanical and Aerospace Engineering, Cornell University,
219 Upson Hall, Ithaca, NY 14853, USA
Meinig School of Biomedical Engineering, Cornell University, 219 Upson Hall,
Ithaca, NY 14853, USA
e-mail: [email protected]

© Springer Science+Business Media New York 2016 15

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_2
16 C.J. Hernandez

A2.2  Tissue Composition and Ultrastructure

At the tissue level cancellous bone is a ceramic polymer composite with composi-
tion similar to that of cortical bone. By mass, bone tissue is 65 % mineral (primarily
an impure hydroxyapatite), 25 % organic (primarily type I collagen but also includ-
ing non-collagenous proteins), and 10 % water. By volume bone tissue is 36 %
mineral, 55 % organic, and 9 % water [3]. These proportions vary with age of the
individual as well as length of time the tissue has been present in the body (the “tis-
sue age”) and differ among species. In general tissue density is highly conserved
and shows a small range within a species (1.6–2.0 g/cm3 in humans) [3, 4]. Hetero­
geneity within bone tissue is present, however, and is caused primarily by bone
remodeling. Bone remodeling is a process in which bone cells remove and replace
bone tissue at discrete locations on bone surfaces. Structures known as cement lines
mark the boundaries of previous remodeling activity within trabeculae and divide
more recently remodeled (younger) tissue from non-remodeled (older) regions of
the tissue that are typically stiffer and more mineralized. The large surface-­
to-volume ratio of cancellous bone (as compared to cortical bone) enables greater
turnover of bone tissue during remodeling, reducing the proportion of older tissue,
possibly explaining why cancellous bone has been reported to have 10–14 % lower
tissue density than cortical bone [5].

A2.3  Mechanical Properties

The mechanical properties of cancellous bone at the continuum scale (specimens

3–5 mm in smallest dimension) are reported here. Mechanical properties of cancel-
lous bone can vary among species and among anatomical sites within individuals.
Mechanical properties can also be altered with aging and in the presence of disease
states.

A2.3.1  Elastic Modulus and Strength

The stiffness and strength of cancellous bone are determined primarily by apparent
density [6, 7]. When loaded in uniaxial compression or tension cancellous bone
displays a slightly nonlinear stress-strain curve [8] and yield is commonly deter-
mined using the 0.2 % offset criteria. Cancellous bone is anisotropic due to both
microstructure and tissue anisotropy.
The Young’s modulus and shear modulus of cancellous bone are determined
primarily by tissue apparent density (Table A2.1) and can vary over tenfold among
regions of the skeleton within the same individual. Hence, there is no one value for
an elastic modulus of cancellous bone and cancellous bone mechanical properties
must therefore be estimated from apparent density. Empirically derived power law
Table A2.1  Mechanical properties of cancellous bone are strongly influenced by density. Regression models relating cancellous bone mechanical properties
to apparent density (wet) are provided. Mechanical properties are reported on axis (in the direction of primary trabecular orientation). Results are provided as
mean ± SD and (95 % CI) when available. Regression equations were achieved using linear regression on log-transformed data
Y = A × XB
3
Mechanical property n Age Apparent density (g/cm ) A B r2
A2  Cancellous Bone

Young’s modulus (MPa)a 142 67 ± 15 0.27 8920 1.83 0.88

(61–68) (0.09–0.75) (7540–10,550) (1.72–1.94)
Shear modulus (MPa)b 42 Bovine NRb 2.09 1.90 0.84
σy (MPa, tension)a 52 65 ± 14 0.29 31.35 1.67 0.86
(61–68) (0.25–0.34) (23.81–41.29) (1.58–1.77)
σy (MPa, compression)a 68 65 ± 14 0.27 ± 0.17 51.15 1.85 0.91
(61–68) (0.24–0.32) (41.53–62.99) (1.72–1.99)
σult (MPa tension)c 22 54 ± 11 0.19 ± 0.04 13.3 1.07 0.47
(27–82) (0.09–0.29)
σult (MPa compression)c 22 54 ± 11 0.17 ± 0.04 33.2 1.53 0.68
(27–82) (0.07–0.27)
τult (MPa)b 15 Bovine NRb 72.74 1.86 0.64
KQ (MPa/m1/2, crack oriented 166 82 ± 6.8 Range: 0.25–1.10 g/cm3 0.67 1.58 0.62
⊥ to trabeculae)d (65–99)
KQ (MPa/m1/2, crack oriented 169 82 ± 6.8 Range: 0.25–1.10 g/cm3 0.54 1.60 0.56
|| to trabecular orientation)d (65–99)
NR not reported
a
Regression models include cancellous bone from vertebrae, proximal tibia, and proximal femur. More predictive regression models for each anatomical region
are available [9, 10]
b
Determined in the bovine proximal tibia through torsion testing [11]
c
Vertebral trabecular bone [12]
d
Specimens from human femoral head and equine vertebrae were pooled. Measures of KIC were not achieved due to elastic-plastic behavior associated with
trabecular bending/torsion [13]
17
18 C.J. Hernandez

models are useful for general prediction of specimen Young’s modulus and shear
modulus, although linear regressions are appropriate when the range in apparent
density is small (for example, if only one region of the skeleton is considered). The
Poisson’s ratio of cancellous bone is difficult to measure and poorly understood and
is typically estimated (common estimates range from 0.1 to 0.3). Cancellous bone
strength is strongly correlated with Young’s modulus. The ultimate strength of can-
cellous bone is strongly correlated with yield strength (in vertebral cancellous bone
ultimate strength is 20 % greater than yield strength [14]). Yield strain and ultimate
strain are not correlated with apparent density. Yield strain of cancellous bone within
a region of the skeleton shows little interindividual variability, but differences among
skeletal sites have been noted. Compressive yield strains of human cancellous bone
have been reported to range from 0.70 to 0.85 (across different regions of the skele-
ton). Yield strains in tension are always lower than those in compression (reported to
range from 0.60 to 0.70). Ultimate strain is more variable for unknown reasons.
Trabecular alignment and microstructure have been shown to influence mechani-
cal properties of cancellous bone. Trabecular alignment is the primary cause of
cancellous bone anisotropy. When loading is applied at 90° to the primary trabecu-
lar orientation, cancellous bone Young’s modulus is 40–60 % smaller and the ulti-
mate strength is 30–45 % smaller [15]. Interestingly yield strain in the transverse
directions is not different from that on axis. The effects of trabecular alignment on
Young’s modulus can be described using a fabric tensor and a quadratic Tsai-Wu
criteria has been used successfully to describe a multiaxial failure envelope for can-
cellous bone [16]. More complicated multiaxial failure criteria have been used to
address some shortcomings of the Tsai-Wu criteria [17, 18].
Tissue material properties also influence cancellous bone Young’s modulus and
strength. Compressive Young’s modulus and strength of bone (cancellous and corti-
cal bone pooled) is related to bone volume fraction (BV/TV) and the tissue degree
of mineralization (α, inorganic mass/bone mass) in the following manner [19]:
2.58
æ BV ö
E ( GPa ) = 84 ç ÷ a 2.74
è TV ø
1.92
æ BV ö
s ult ( MPa ) = 794 ç ÷ a 2.79 .
è TV ø
The regression exponents applied to tissue degree of mineralization are greater than
those on bone volume fraction, demonstrating that bone strength is more sensitive to
variation in tissue degree of mineralization than to variation in bone volume fraction.

A2.3.2  Viscoelastic and Fatigue Properties

Cancellous bone displays viscoelastic properties. The strain rate has only a small
effect on the Young’s modulus and strength of cancellous bone under uniaxial load-
ing; Young’s modulus and strength are related to strain rate to the power of 0.06 [6].
Hydraulic stiffening of the marrow does not influence Young’s modulus and strength
A2  Cancellous Bone 19

until strain rates exceed 1 s−1 [6]. Creep deformation in cancellous bone follows the
pattern of a rapid primary phase, slow secondary phase, and rapid tertiary phase
[20]. The creep rate of bovine cancellous bone is related to normalized stress (σ/E0,
where E0 is the initial Young’s modulus) as follows [20]:
17.65
de c æs ö
= 2.21 ´ 1033 ç ÷ .
dt è E0 ø

Human vertebral cancellous bone submitted to low-magnitude, compressive creep
loading (σ/E0 = 1500 με or less) has been shown to have nonlinear viscoelastic prop-
erties such that residual strains persist up to ten times longer than the period of
constant loading [21].
Under fatigue loading cancellous bone displays an S-N curve such that the num-
ber of cycles to failure (Nf) is related to normalized stress (σ/E0) as follows [22]:

N f = 4.57 ´ 10 -18 ( s / E0 )
-8.54
.

A2.3.3  Fracture Toughness

Relatively little is known regarding resistance to crack growth in cancellous bone.

Resistance to crack growth in cancellous bone specimens has been assessed with
linear elastic fracture mechanics approaches (Table A2.1), but care must be taken in
interpreting the quantitative values because large deformations in trabeculae at the
crack tip prevent assessment of KIC (see [13] for a discussion of the utility of linear
elastic fracture mechanics in continuum specimens of cancellous bone).

A2.3.4  Post-Yield and Damage Behavior

Tissue damage in cancellous bone impairs mechanical performance during subse-

quent loading. Reductions in Young’s modulus have been observed in specimens of
cancellous bone submitted to as little as 0.4 % apparent strain (well below yield
strain) [23]. Loading in compression causes reductions in Young’s modulus [24–26]
and strength [24, 25] that are related to the maximum applied strain experienced by
the specimen. Tissue damage in the form of microscopic and sub-microscopic
cracks is also related to the maximum applied strain and subsequent reductions in
Young’s modulus and strength. Relatively small amounts of microscopic tissue
damage (less than that characterized as “naturally occurring”) have been associated
with 50–60 % reductions in strength and the use of over 90 % of cancellous bone
fatigue life [24, 26]. Hence, microscopic tissue damage accumulated in cancellous
bone in vivo may contribute to bone failure. Microscopic tissue damage is generated
more rapidly following changes in loading mode (between compression and shear
for example) [27]. Despite the presence of microscopic damage, residual strains
20 C.J. Hernandez

following loading are typically small; upon removal of load cancellous bone recov-
ers as much as 70–94 % of the applied strain [24, 25, 28].

A2.4  Tissue-Level Mechanical Properties

The mechanical properties of cancellous bone tissue (the constituents of individual

trabeculae) remain poorly understood. Mechanical testing of individual trabeculae
is challenging and requires many assumptions that may limit the accuracy of the
results. The Young’s modulus of individual trabeculae typically averages 3–6 GPa
[29]. In contrast, the Young’s modulus of human cancellous bone tissue assessed
through nanoindentation is, on average, 10–18 GPa [30–32], and is similar to the
range to tissue-level Young’s moduli estimated using finite element models of can-
cellous bone microstructure [33]. Relatively little is known regarding tissue-level
anisotropy, tissue-level viscoelasticity, fatigue, and fracture toughness.

Additional Reading

Cowin S. Bone Mechanics Handbook. In. 2 ed. Boca Raton: CRC Press; 2001.
This book provides a complete review of bone mechanical properties and inter-
actions with bone cell biology.
Currey JD. Bones: Structure and Mechanics. Princeton, NJ, USA: Princeton
University Press; 2002.
This is a review that includes thorough discussion of non-human bone mechani-
cal properties and function.

References

1. Liu XS, Sajda P, Saha PK, Wehrli FW, Bevill G, Keaveny TM, Guo XE (2008) Complete volu-
metric decomposition of individual trabecular plates and rods and its morphological correla-
tions with anisotropic elastic moduli in human trabecular bone. J Bone Miner Res 23:223–235
2. Fyhrie DP, Fazzalari NL, Goulet R, Goldstein SA (1993) Direct calculation of the surface-to-­
volume ratio for human cancellous bone. J Biomech 26:955–967
3. Robinson RA (1975) Physicochemical structure of bone. Clin Orthop 53:263–315
4. Galante J, Rostoker W, Ray RD (1970) Physical properties of trabecular bone. Calcif Tissue
Res 5:236–246
5. Gong JK, Arnold JS, Cohn SH (1964) Composition of trabecular and cortical bone. Anat Rec
149:325–332
6. Carter DR, Hayes WC (1977) The compressive behavior of bone as a two-phase porous struc-
ture. J Bone Joint Surg 59-A:954–962
7. Keaveny TM, Morgan EF, Niebur GL, Yeh OC (2001) Biomechanics of trabecular bone. Annu
Rev Biomed Eng 3:307–333
8. Morgan EF, Yeh OC, Chang WC, Keaveny TM (2001) Nonlinear behavior of trabecular bone
at small strains. J Biomech Eng 123:1–9
A2  Cancellous Bone 21

9. Morgan EF, Bayraktar HH, Keaveny TM (2003) Trabecular bone modulus-density relation-
ships depend on anatomic site. J Biomech 36:897–904
10. Morgan EF, Keaveny TM (2001) Dependence of yield strain of human trabecular bone on
anatomic site. J Biomech 34:569–577
11. Garrison JG, Gargac JA, Niebur GL (2011) Shear strength and toughness of trabecular bone
are more sensitive to density than damage. J Biomech 44:2747–2754
12. Kopperdahl DL, Keaveny TM (1998) Yield strain behavior of trabecular bone. J Biomech
31:601–608
13. Cook RB, Zioupos P (2009) The fracture toughness of cancellous bone. J Biomech

42:2054–2060
14. Crawford RP, Cann CE, Keaveny TM (2003) Finite element models predict in vitro vertebral
body compressive strength better than quantitative computed tomography. Bone 33:744–750
15. Goulet RW, Goldstein SA, Ciarelli MJ, Kuhn JL, Brown MB, Feldkamp LA (1994) The rela-
tionship between the structural and orthogonal compressive properties of trabecular bone.
J Biomech 27:375–389
16. Cowin SC (1985) The relationship between the elasticity tensor and the fabric tensor. Mech
Mater 4:137–147
17. Bayraktar HH, Gupta A, Kwon RY, Papadopoulos P, Keaveny TM (2004) The modified super-­
ellipsoid yield criterion for human trabecular bone. J Biomech Eng 126:677–684
18. Wolfram U, Gross T, Pahr DH, Schwiedrzik J, Wilke HJ, Zysset PK (2012) Fabric-based
Tsai-Wu yield criteria for vertebral trabecular bone in stress and strain space. J Mech Behav
Biomed Mater 15:218–228
19. Hernandez CJ, Beaupre GS, Keller TS, Carter DR (2001) The influence of bone volume frac-
tion and ash fraction on bone strength and modulus. Bone 29:74–78
20. Bowman SM, Keaveny TM, Gibson LJ, Hayes WC, McMahon TA (1994) Compressive creep
behavior of bovine trabecular bone. J Biomech 27:301–310
21. Yamamoto E, Paul Crawford R, Chan DD, Keaveny TM (2006) Development of residual
strains in human vertebral trabecular bone after prolonged static and cyclic loading at low load
levels. J Biomech 39:1812–1818
22. Haddock SM, Yeh OC, Mummaneni PV, Rosenberg WS, Keaveny TM (2004) Similarity in the
fatigue behavior of trabecular bone across site and species. J Biomech 37:181–187
23. Morgan EF, Yeh OC, Keaveny TM (2005) Damage in trabecular bone at small strains. Eur
J Morphol 42:13–21
24. Hernandez CJ, Lambers FM, Widjaja J, Chapa C, Rimnac CM (2014) Quantitative relation-
ships between microdamage and cancellous bone strength and stiffness. Bone 66:205–213
25. Keaveny TM, Wachtel EF, Kopperdahl DL (1999) Mechanical behavior of human trabecular
bone after overloading. J Orthop Res 17:346–353
26. Lambers FM, Bouman AR, Rimnac CM, Hernandez CJ (2013) Microdamage caused by
fatigue loading in human cancellous bone: relationship to reductions in bone biomechanical
performance. PLoS One 8, e83662
27. Wang X, Niebur GL (2006) Microdamage propagation in trabecular bone due to changes in
loading mode. J Biomech 39:781–790
28. Fyhrie DP, Schaffler MB (1994) Failure mechanisms in human vertebral cancellous bone.
Bone 15:105–109
29. Guo XE (2001) Mechanical properties of cortical bone and cancellous tissue. In: Cowin SC
(ed) Bone Mechanics Handbook. CRC Press, Boca Raton, pp 10.11–10.23
30. Kim G, Cole JH, Boskey AL, Baker SP, van der Meulen MC (2014) Reduced tissue-level stiff-
ness and mineralization in osteoporotic cancellous bone. Calcif Tissue Int 95:125–131
31. Norman J, Shapter JG, Short K, Smith LJ, Fazzalari NL (2008) Micromechanical properties of
human trabecular bone: a hierarchical investigation using nanoindentation. J Biomed Mater
Res A 87:196–202
32. Rho JY, Roy ME, Tsui TY, Pharr GM (1999) Elastic properties of microstructural components
of human bone tissue as measured by nanoindentation. J Biomed Mater Res 45:48–54
33. Bayraktar HH, Morgan EF, Niebur GL, Morris GE, Wong EK, Keaveny TM (2004) Comparison
of the elastic and yield properties of human femoral trabecular and cortical bone tissue.
J Biomech 37:27–35
Chapter A3
Dentin and Enamel

K.E. Healy

A3.1 Introduction

A3.1.1 Structure of human dentition:

The permanent adult human dentition normally consists of 32 teeth, of which 16 are
located in the mandible and 16 in the maxilla. There are 4 incisors, 2 canines, 4
premolars and 6 molars for the upper and lower dentition. The incisors are used for
cutting food, the canines for tearing, the premolars for grasping, and the molars for
grinding (i.e., masticating). There is a generic heterogeneous structure for these
teeth, where enamel forms an exterior layer over the underlying dentin. From the
cervix to the apex of the root, the exterior of the dentin is covered by cementum to
which the periodontal ligament attaches the tooth to alveolar bone. Dental enamel is
dense, highly mineralized, hard, and brittle. It contains prism-like structures that
span from the enamel surface to the junction of enamel and dentin, the dentino-
enamel junction (DEJ). The prisms are comprised of hydroxyapatite crystallites and
contain very little organic matrix. These properties make dental enamel an excellent
material for cutting and masticating food (i.e., processes that involve friction and
wear). In contrast, dentin is not as hard as enamel, but it is tougher. Dentin is a
heterogeneous material and can be thought of as a composite structure containing
four major components: dentin matrix; dentinal tubules; mineral (i.e., carbonate
containing hydroxyapatite); and, dentinal fluid. The dentinal tubules (~45 000 per
mm2) are formed during development of the dentin matrix and are distributed

K.E. Healy (*)

Department of Biological Materials, Northwestern University,
311 E. Chicago Ave., Chicago, IL 60611-3008, USA

© Springer Science+Business Media New York 2016 23

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_3
24 K.E. Healy

throughout the dentin matrix in a somewhat uniform manner. The dentin matrix
mineralizes in an anisotropic fashion, where a highly mineralized tissue, peritubular
dentin, surrounds the dentinal tubules. The mineralized tissue between the dentinal
tubules and peritubular dentin is referred to as intertubular dentin. Histological
examination has revealed that intertubular dentin is less mineralized than peritubular
dentin. Furthermore, the matrix and mineral content of root dentin is different from
coronal dentin. A good review of the structure of teeth can be found in Waters [1].

A3.2 Composition

Table A3.1 Basic Constituents of Human Dentin and Enamel*

Enamel Dentin
Weight % Volume % Weight % Volume %
Mineral (density, 3000 kg m-3) 96 90 70 50
Organic (density, 1400 kg m-3) 1 2 20 30
Water (density, 1000 kg m-3) 3 8 10 20
* Adapted from [1–3].

Table A3.2 Major Elemental Composition of Surface and Bulk Dental Enamel
Enamel Mean wt% (range or Dentin Mean wt% (range or Source,
standard deviation, ±) standard deviation, ±) Comments
Ca 37.4 ± 1.0 – [4]#
37.1 ± 0.2 (26.7–47.9) 26.9 ± 0.2 (21.8–31.3) [5]#
36.3 ± 0.1 (27.7–42.0) 27.6 ± 0.1 (24.7–31.5) [5]‡
P 17.8 ± 0.2 13.5 ± 0.1 [5]●, age >25yrs
17.68 ± 0.2 [6]¤
Na 0.72 ± 0.008 (0.42–1.03) 0.72 ± 0.008 (0.26–0.87) [5]#
0.72 ± 0.008 (0.49–0.88) 0.64 ± 0.001 (0.55–0.75) [5]‡
Cl 0.28 ± 0.01 0.05 ± 0.004 [5]#, age >25yrs
0.32 ± 0.01 0.072 ± 0.022 [7]#
K 0.026 ± 0.001 0.02 ± 0.001 [5]‡, age <25yrs
Mg 0.39 ± 0.02 (0.13–0.77) 0.74 ± 0.02 (0.25–0.94) [5]#
0.39 ± 0.004 (0.24–0.48) 0.76 ± 0.004 (0.58–0.89) [5]‡
CO3 3.2 (2.4–4.2) 4.6 (4–5) [2,3]†
# Neutron activated gamma-ray spectrometric analysis.
‡ Atomic absorption spectrophotometry.
● Colorimetic assay.
† Average compiled from the literature.
* Neutron activated gamma-ray spectrometric analysis (Na, Cl, Al, Mn, Ca, and P), atomic
absorption spectrophotometry (K, Mg, Zn, Cu, and Fe), or a fluoride-specific electrode (F).
¤ Atomic absorption spectrophotometry (Ca), and colorimetric method (P).
A3 Dentin and Enamel 25

Table A3.3 Trace Elemental Composition of Surface and Bulk Dental Enamel
Surface Enamel Whole Enamel
Mean (range) Median Median Source,
At.# μg/g μg/g Mean (range) μg/g μg/g comments
S 16 281 (530–130) 270 [8]†,[9]‡
F 9 752 (1948–25) 666 293 (730–95) 200 [8]†,[9]‡
123.8 ± 7.9 [10]◊
Zn 30 893 (5400–61) 576 199 (400–91) 190 [8]†,[9]‡
276 ± 106 [4]#
263.42 ± 14.8 [10]◊
Mg 12 745 (3600–115) 576 1,670 (3,000–470) 1,550 [8]†,[9]‡
AI 13 343 (2304–16) 202 12.5 (70–1.5) 5.6 [8]†,[9]‡
Sr 38 204 (7632–9) 36 81 (280–26) 56 [8]†,[9]‡
93.5 ± 21.9 [4]#
111.19 ± 9.86 [10]◊
Fe 26 138 (1404–18) 68 4.4 (21–0.8) 2.6 [8]†,[9]‡
2.77 [7]*
Si 14 70 (504–1.3) 40 [8]†,[9]‡
Mn 25 59 (468–2.6) 33 0.28 (0.64–0.08) 0.26 [8]†,[9]‡
0.54 ± 0.08 [4]#
0.59 ± 0.04 [10]◊
Ag 47 32 (396–0.2) 2 0.35 (1.3–0.03) 0.16 [8]†,[9]‡
0.56 ± 0.29 [10]◊
Pb 82 24 (79–1.2) 18 3.6 (6.5–1.3) 3.6 [8]†,[9]‡
Ni 28 23 (270–0.4) 9 [8]†,[9]‡
Ba 56 22 (432–0.8) 7 4.2 (13–0.8) 3.4 [8]†,[9]‡
Se 34 18 (72–2.9) 16 0.27 (0.5–0.12) 0.22 [8]†,[9]‡
Li 3 14 (58–0.3) 10 1.13 (3.4–0.23) 0.93 [8]†,[9]‡
Sb 51 8 (90–0) 3 0.13 (0.34–0.02) 0.11 [8]†,[9]‡
Ga 31 6 (32–0) 5 [8]†,[9]‡
Sn 50 9.3 (72–0.9) 5.8 0.21 (0.92–0.03) 0.14 [8]†,[9]‡
Ge 32 7.6 (39.6–0.5) 4.0 [8]†,[9]‡
B 5 5.3 (13.0–0.8) 3.6 5.0 (39–0.5) 2.4 [8]†,[9]‡
Cu 29 4.20 (81–0.1) 0.45 [8]†,[9]‡
0.26 ± 0.11 [4]#
1.38 [7]*
Br 35 3.1 (14.0–0.4) 4.1 1.12 (2.6–0.32) 0.93 [8]†,[9]‡
4.6 ± 1.1 [4]#
Cd 48 2.7 (7.6–0.6) 1.8 0.51 (2.4–0.03) 0.22 [8]†,[9]‡
Y 39 1.8 (9.3–0) 0.9 0.007 (0.17–<0.01) <0.01 [8]†,[9]‡
Ti 22 1.6 (24.5–0.1) 0.6 0.19 (4.4–<0.1) <0.1 [8]†,[9]‡
V 23 1.4 (14.4–0.1) 0.5 0.017 (0.03–0.01) 0.02 [8]†,[9]‡
La 57 1.4 (7.2–0) 0.8 [8]†,[9]‡
Be 4 1.3 (6.1–0) 1.2 [8]†,[9]‡
Cr 24 1.1 (4.7–0.2) 0.7 3.2 (18–0.1) 1.5 [8]†,[9]‡
26 K.E. Healy

Table A3.3 (continued)

Surface Enamel Whole Enamel
Mean (range) Median Median Source,
At.# μg/g μg/g Mean (range) μg/g μg/g comments
1.02 ± 0.51 [10]◊
Rb 37 0.6 (4.0–0.1) 0.4 0.39 (0.87– 0.17) 0.32 [8]†,[9]‡
Zr 40 0.6 (1.9–0) 0.3 0.1 (0.57– <0.02) O.Q7 [8]†,[9]‡
Ce 58 0.6 (6.1–0) 0 0.07 (1.9– 0.02) 0.07 [8]†,[9]‡
W 74 0.24 ± 0.12 [8]†,[9]‡
Co 27 0.2 (2.7–0) 0.1 [8]†,[9]‡
0.13 ± 0.13 [10]◊
Pr 59 0.2 (4.7–0) 0 0.027 (0.07– <0.01) 0.03 [8]†,[9]‡
Cs 55 0.1 (1.9–0) 0 0.04 (0.1– <0.02) 0.04 [8]†,[9]‡
Mo 42 0.1 (0.5–0.04) 0.04 7.2 (39– 0.7) 6.3 [8]†,[9]‡
I 53 0.05 (4.7–0) 0.05 0.036 (0.07– 0.01) 0.03 [8]†,[9]‡
Bi 83 0.001 (0.04–0) 0 0.006 (0.07– <0.02) 0.02 [8]†,[9]‡
Nd 60 0.045 (0.09– <0.02) 0.05 [8]†,[9]‡
Nb 41 0.28 (0.76– <0.1) 0.24 [8]†,[9]‡
Au 79 0.02 ± 0.01 [4]#
‡ Whole enamel from premolars of young patients (age<20 yrs), determined by spark source mass
spectroscopy.
† Surface enamel (depth of analysis 42 ± 8.5 μm) from premolars of young patients (age<20yrs),
determined by spark source mass spectroscopy.
# Bulk enamel from premolars of 14–16 yrs male and female patients, selected population of
Stockholm Sweden, determined by neutron activated gamma-ray spectrometric analysis. Standard
deviation, ±.
◊ Neutron activated gamma-ray spectrometric analysis.
* Neutron activated gamma-ray spectrometric analysis (Na, Cl, AI, Mn, Ca, and P), atomic absorp-
tion spectrophotometry (K, Mg, Zn, Cu, and Fe), or a fluoride-specific electrode (F).

Table A3.4 Significant Differences in Trace Element Composition of Whole Human Enamel for
High and Low Caries Populationst†
High Caries Low Caries
At.# (Mean ± SE), μg/g (Mean ± SE), μg/g Source
F 9 82.1 ± 7.99 125.7 ± 11.23 [11]
Sr 38 104.1 ± 9.14 184.0 ± 14.68 [11]
Mn 25 1.57 ± 0.24 0.87 ± 0.15 [12]
Zr 40 0.27 ± 0.1 0.16 ± 0.09 [11]
Cu 26 0.71 ± 0.2 0.17 ± 0.04 [12]
† Determined by spark source mass spectroscopy
A3 Dentin and Enamel 27

Table A3.5 Ca/P Molar Ratio of Human Enamel and Dentin

Enamel Ca/P molar ratio Dentin Ca/P molar ratio Source, comments
1.58 [4]#
1.61 1.54 [5]†, ●, age >25 yrs
1.58 1.58 [5]*, ●, age >25 yrs
1.65 13]**
1.64 [6]¤
1.61 [14]**
# Neutron activated gamma-ray spectrometric analysis.
† Ca determined by neutron activated gamma-ray spectrometric analysis
* Ca determined by atomic absorption spectrophotometry
● P determined by colorimetic assay.
** Determined by energy dispersive X-ray analysis.
¤ Determined by atomic absorption spectrophotometry (Ca), and by the colorimetric method (P).

Table A3.6 Crystallite Size and Lattice Parameters of the Apatite in Human Enamel and Dentin*
a-axis (nm) c-axis (nm) Width (nm) Thickness (nm) Source, Comments
Enamel
0.9445 0.6885 [2]#
0.9440 0.6872 [15]†
0.9441 0.6880 68.4 ± 3.4 26.3 ± 2.2 [6]†, ● ±S.D.
0.9446 0.6886 [16]†
68.3 ± 13.4 26.3 ± 2.19 [17]‡,● ± S.E.
Dentin
0.9434 ± 0.6868 ± [18]‡
0.0007 0.0009 29.6 ± 3.7 3.2 ± 0.5 [19]●, intertubular
dentin
36.55 ± 1.45 10.33 ± 7.91 [20]●, mixed carious
and sound dentin
* Asymmetric hexagonal crystal with the thickness of the crystal less than the width.
† X-ray diffraction method of determination.
● High resolution transmission electron microscopy.
# Data from [2], average compiled from the literature.
28 K.E. Healy

Table A3.7 Elastic Moduli and Viscoelastic Properties of Human Dentin and Enamel
Incisors Canine Pre-molars Molars Source, Comments
E: Dentin
11.0 (5.8) [21]t,†,‡
13 (4) 14 (6) 14 (0.7) 12 (2) [22]Crown, c,†
9.7 (2) 12 (3) 9.0 (2) 7.6 (3) [22]Root, c,†
10.16 [23]b,||
10.87 [23] b,dehyd., ||
9.49 [23] b, re-hyd, ||
E: Enamel
84.3 (8.9) [24]Cusp, c, ||
77.9 (4.8) [24]Side, c, ||
48 (6) 46 (5) [22]Cusp, c,‡
33 (2) 32 (4) [22]Axial (side), c, ^
9.7 (3) [22]Axial (side), c, ||
12 (3) [22]Occlusal, c, ||
E,(∞): Dentin
12 [25]c, constant strain,
hydrated,^,‡
H1(t): Dentin
0.38 [25] c, constant
(0.136) strain, hydrated,^,‡
E: modulus of elasticity (GPa); Er (∞): relaxed modulus (GPa); H1(t): distribution of relaxation
times (GPa); c: compression; t: tension; b: three-point bending.
|| Applied load approximately parallel to either the long axis of the enamel rods or dentinal tubules.
^ Applied load approximately perpendicular to either the long axis of the enamel rods or dentinal
tubules.
† Applied load with respect to either the long axis of the enamel rods or dentinal tubules was variable.
‡ Type of tooth unknown or various teeth used for measurement; data are tabulated under molar.
Note: standard deviations are given in parentheses.

Table A3.8 Mechanical Properties of Human Enamel

Incisors Canine Pre-molars Molars Source, comments
Stress at 353 (83) [24]Cusp, c, ||
Proportional 336 (61) [24]Axial(side), c, ||
Limit (MPa) 194 (19) 224 (26) [22]Cusp, c,†
183 (12) 186.2 (17) [22]Axial (side), c, ^
70.3 (22) [22]Axial (side), c, ||
98.6 (26) [22]Occlusal, c, ||
91.0 (10) [22]Incisal edge, c,†
Tensile Strength (MPa) 10 (2.6) [26]†
Compressive 384 (92) [24]Cusp, c, ||
Strength (MPa) 372 (56) [24]Axial (side), c, ||
288 (48) 261 (41) [22]Cusp, c,†
253(35) 239 (30) [22]Axial (side), c, ^
(continued)
A3 Dentin and Enamel 29

Table A3.8 (continued)

Incisors Canine Pre-molars Molars Source, comments
94.5 (32) [22]Axial (side), c, ||
127 (30) [22]Occlusal, c, ||
220 (13) [22]Incisal edge, c,†
c: compression; hyd: hydrated; dehyd: dehydrated; re-hyd: re-hydrated.
|| Fracture or applied load approximately parallel to the long axis of the enamel rods.
^ Fracture or applied load approximately perpendicular to the long axis of the enamel rods.
† Applied load with respect to either the long axis of the enamel rods or dentinal tubules was variable.
‡ Type of tooth unknown or various teeth used for measurement; data are tabulated under molar.
Note: standard deviations are given in parentheses.

Table A3.9 Mechanical Properties of Human Dentin

Incisors Canine Pre-molars Molars Comments
Stress at 167 (20.0) [24]c
Proportional 124 (26) 140 (15) 146 (17) 148 (21) [22]c
Limit (MPa) 86 (24) 112 (34) 110 (38) 108 (39) [22]c
110.5 (22.6) [23]b, hyd., ||
167.3 (37.5) [23]b, dehyd, ||
103.1 (16.8) [23]b, re-hyd, ||
158 (32) [17]
154 (23) [17]
Tensile 52 (10) [26]hyd, †,‡
Strength 37.3 (13.6) [23]hyd, ||
(MPa) 34.5 (11.1) [23]dehyd, ||
37.3 (9.0) [23]re-hyd, ||
39.3 (7.4) [21]hyd,†, ‡
Compressive 297 (24.8) [24]Crown
Strength 232 (21) 276 (72) 248 (10) 305 (59) [22] Crown
(MPa) 233 (66) 217 (26) 231 (38) 250 (60) [22]Root
295 (21) [23]Crown,‡
251 (30) [23]Crown, ‡
Shear Strength 134 (4.5) [27]Oil, Cervical
(MPa) root, ^, ‡
Flexural 165.6 (36.1) [23]hyd, ||
Strength 167.3 (37.5) [23]dehyd, ||
(MPa) 162.5 (25.4) [23]re-hyd, ||
hyd: hydrated; dehyd: dehydrated; re-hyd: re-hydrated
|| Applied load approximately parallel to the long axis of the dentinal tubules
* Applied load approximately perpendicular to the long axis of the dentinal tubules;
‡ Type of tooth unknown or various teeth used for measurement; data are tabulated under molar;
† Applied load with respect to either the long axis of the dentinal tubules was variable.
‡ 95% confidence intervals.
Note: standard deviations are given in parentheses.
30 K.E. Healy

Table A3.10 Toughness, Fracture Toughness, and Work of Fracture of Human Dentin and Enamel
Incisors Canine Pre-molars Molars Source, comments
Fracture
Toughness,
Kc (MNm-3/2)
[28]*
Enamel 0.97(0.09) 1.00(0.23) Maxillary,
cervical, †
1.27(0.09) 0.7(0.08) Mandibular,
cervical, †
Toughness (MJm-3)
Dentin 62.7 (6.2)† [27] Root, shear,
oil storage, ^, ‡
2.4 (1.1) [17]Tension,
crown, hydr., ||
Work of Fracture
(102 Jm-2) Dentine 2.7 (1.6) [29],^
5.5 (1.7) [29] ||
Enamel 1.9(0.56) [29], ^
0.13(.065) [29] ||
|| Applied load approximately parallel to either the long axis of the enamel rods or dentinal
tubules.
^ Applied load approximately perpendicular to either the long axis of the enamel rods or dentinal
tubules.
† Applied load with respect to either the long axis of the enamel rods or dentinal tubules was
variable.
‡ Type of tooth unknown or various teeth used for measurement; data are tabulated under
molar.
* Microindentation method used. Load was 500 g with a Vickers’ indenter.
Note: standard deviations are given in parentheses.
A3 Dentin and Enamel 31

Table A3.11 Hardness of Fracture of Human Dentin and Enamel (see notes for units)
Incisor Pre-molar Molar Source, comments
Enamel 365 (35) [30] >90% incisors,®, †
393 (50) [30]‡, molars and premolars,®, †
385 (5.8) [31]⨂ † ‡
367 (17) [32] ||,⨂, incisors, premolars
327 (34) [32]^,⨂, incisors, premolars
Dentin
25–81.7 [33]∆, ||, [34]a
97.8 [33]a, calculated for zero tubule density
44.5–80.9 [14]◊, ||, [34]a
100 [14]a, calculated for zero tubule density
75 (0.8) [31]⨂,†,‡
a
Inverse correlation between hardness and dentinal tubule density.
|| Applied load approximately parallel to either the long axis of the enamel rods or dentinal tubules.
^ Applied load approximately perpendicular to either the long axis of the enamel rods or dentinal
tubules.
† Applied load with respect to either the long axis of the enamel rods or dentinal tubules was variable.
‡ Type of tooth unknown or various teeth used for measurement; data are tabulated under molar.
* Microindentation method used. Load was 500 g with a Vickers’ indenter.
® Knoop hardness test using 500 g load.
⨂ Knoop microhardness test using 50 g load.
∆ Knoop microhardness test using 100 g load.
◊ Micromdentation method used. Load was 50 g with a Vickers’ indenter.

Table A3.12 Permeabilitya of Human Dentin

Periphery(μl cm-2 min-1) Center (μl cm-2 min-1) Source, comments
36.4 (13.1)‡ 14.3 (7.0)† [33], unerupted third molars,
a
Fluid filtration rate.
‡ Sound human dentin, average of 4 samples, 4 readings per sample.
† Sound human dentin, average of 4 samples, 1 reading per sample.
32 K.E. Healy

Table A3.13 Wetability of Human Enamel

Contact Angle, θ(deg)
Surface Tension, Ground Source
Liquid γLV (dynes/em) In situ enamel enamel Comments
Polar
Water 72.4 [35] 25.4 [36]†
72.8 36 [37]
72.6 40.0 (0.1) [38]* n=330
Glycerol 63.7 [35] 44.7 [36]†
63.4 55 [37]
63.4 45.6(0.2) [38]* n=50
Formamide 58.5 [35] 28.0 [35]†
58.2 24 [36]
58.2 37.6 (0.1) [37]* n=50
Thiodiglycol 53.5 [34] 30.8 [36]†
54.0 43 [37]
54.0 27.6 (0.2) [38]* n=60
Non-polar
Methylene iodide 51.7 [35] 48.6 [36]†
50.8 50 [37]
50.8 38.1 (0.1) [38] n=50
S-Tetrabromoethane 49.8 [35] 38.3 [36]†
47.5 40 [38]
1- Bromonaphthalene 44.6 34 [38]*, n=50
44.6 16.1 (0.1)
o-Dibromobenzene 42.0 22 [37]
Propylene carbonate 41.8 [35] 31.8 [36]†
1-Methyl-naphthalene 38.7 20 [36]
Dicyclohexyl 32.7 [35] 12.2-spread [37]
33.0 7 [36]†
n-Hexadecane 27.6 [35] spreading [37]
27.7 spreading [36]†
* Plane ground enamel surfaces, measurements from 46 erupted and unerupted teeth, mixed
location (molars, premolars, incisors). Parentheses: standard error
† in situ contact angle measurements on human enamel, average of mean values for 4 teeth
(maxillary or mandibular incisors).
A3 Dentin and Enamel 33

Table A3.14 Wetability of Human Dentin [38]

Surface Tension, γLV Ground Dentin Contact
Liquid (dynes/em) Angle, θ (deg) Comments
Polar
Water 72.6 45.3 (0.2) *, n=100
Glycerol 63.4 44.6 (0.1) *, n=50
Formamide 58.2 37.6 (0.2) *, n=50
Thiodiglycol 54.0 33.6 (0.3) *, n=50
Non-polar
Methylene iodide 50.8 36.7 (0.3) *, n=50
1-bromo-naphthalene 49.8 16.8 (0.2) *, n=50
* Plane ground dentin surfaces, measurements from 46 erupted and unerupted teeth, mixed loca-
tion (molars, premolars, incisors). Parentheses: standard error.

Table A3.15 Critical Surface Tensions (γc) of Human Enamel and Dentin
Critical Surface Tension, γc
(dynes cm-1) Source, Comments
Enamel 46.1 (40.0 – 55.6)a [38]*, calculated from polar and
Ground surface non-polar liquids
In situ enamel, γcd 45.3 ± 70.2b [39]∆, calculated from polar liquids,
In situ enamel, γcd 32.9 ± 4.7 [38]∆, calculated from non-polar liquids
In situ enamel, γcd 32 [37]†, calculated from non-polar liquids
Dentin 45.1 (40.7 – 51.1)a [38]*, calculated from polar and
non-polar liquids
a
Range of values from different test liquids.
b
Standard deviation.
* Plane ground dentin surfaces, measurements from 46 erupted and unerupted teeth, mixed loca-
tion (molars, premolars, incisors). Parentheses: standard error.
∆ In situ measurements from 76 test subjects: 29 female and 47 male. Measurements made on
teeth with intact pellicle (i.e., biofilm). γcp only calculated from glycerol and thiodiglycol.
† Average of 4 teeth from 2 subjects. γcd calculated from non-polar liquids.

A3.3 Final Comments

The quality of data presented can be inferred from the standard deviations or stan-
dard error associated with the mean values. In some cases the error can be attributed
to either small sample populations or specimen preparation. Where possible, either
the number of specimens used or the number of replications of a measurement was
reported. The reader should use this information as a guideline of the quality of
data. When data are reported for small sample populations, then these data were
usually the only source for a given physical property. In review of the literature,
specimen preparation appears to have had the most influence on the precision and
accuracy of data. Sample collection and storage conditions (e.g., dehydration, cross-
linking agents, exogenous contamination) need to be taken into consideration when
34 K.E. Healy

utilizing the information tabulated. Additional sources of error are dependent on the
analytical technique or test method used to make the measurement. It is more diffi-
cult to discern the influence of the instrumentation on the reliability of the
measurements. However, confidence of the accuracy was judged based on the use
of adequate control samples with known physical properties (e.g., correction of
mechanical data). In light of these comments, data in the literature were deemed
most accurate and appropriate for this handbook when the following conditions
were met: the sample population was large; non-destructive specimen preparation
and storage conditions were used; and, multiple replications of measurements on a
single sample were performed.
There are significant omissions in the data available in the literature. Most nota-
ble, is the lack of quantitative analysis of the organic phase of dentin and enamel,
and determination of the viscoelastic properties of dentin. The lack of data is attrib-
uted to the technical difficulty required to make such measurements and the hetero-
geneous nature of the dentin, which imparts large variations in these data depending
on anatomical location. Other significance absences are the lack of electrical and
thermal properties. Finally, vacancies in the tables provided demonstrate omissions
in available data.

Additional Reading

Carter, J.M., Sorensen, S.E., Johnson, R.R., Teitelbaum, R.L. and Levine, M.S.
(1983) Punch Shear Testing of Extracted Vital and Endodontically Treated Teeth.
J. Biomechanics 16(10), 841–848.
Utilized a miniature punch shear apparatus to determine shear strength and
toughness perpendicular to the direction of dentinal tubules. Dentin harvested from
the cemento-enamel junction to one-third the distance to the root apex. Strengths:
novel measurements, precise measurements, defined specimen location, defined ori-
entation of testing. Limitations: tooth type not defined for ‘constrained’ tests, teeth
stored in mineral oil prior to testing.
Driessens, F.C.M., and Verbeeck, R.M.H. (1990a) The Mineral in Tooth Enamel
and Dental Carries. In Biominerals, F.C.M and Verbeeck, R.M.H. (eds), CRC Press,
Boca Raton, Florida, pp. 105–161.
Driessens, F.C.M., and Verbeeck, R.M.H. (1990b) Dentin, Its Mineral and
Caries, In Biominerals, F.C.M and Verbeeck, R.M.H. (eds), CRC Press, Boca Raton,
Florida, pp. 163–178.
An authoritative text on biominerals with an excellent review of the properties of
enamel and dentin. An excellent supplement to this handbook.
Glantz, P-O. (1969) On Wetability and Adhesiveness. Odontologisk Revy, 20
supp. 17, 1–132.
Comprehensive assessment of the wetability of human enamel and dentin.
Strengths include using multiple probe liquids on numerous teeth.
A3 Dentin and Enamel 35

Korostoff, E., Pollack, S.R., and Duncanson, M.G. (1975) Viscoelastic Properties
of Human Dentin. J. Biomedical Materials Res., 9, 661–674.
Measured some viscoelastic properties of human radicular dentin under constant
strain. Linear viscoelastic theory applied. Strengths: unique examination of visco-
elastic properties, defined orientation of dentinal tubules, storage conditions and
testing environment well controlled. Limitations: large scatter in H1(t), mixed data
for different teeth.
Marshall, G.W. (1993) Dentin: Microstructure and Characterization. Quintessence
International, 24(9), 606–616.
A Review of the microstructure and characterization of dentin.
Waters, N.E. (1980) Some Mechanical and Physical Properties of Teeth.
Symposia of the Society for Experimental Biology, 34, 99–135.
Concise review of mechanical and physical properties of teeth. Good paper for
anatomy of enamel and dentin.

References

1. Waters, N.E. (1980) Some mechanical and physical properties of teeth. Symp. Soc. Exp. Biol.,
34, 99–135
2. Driessens, F.C.M. and Verbeeck R.M.H. (1990) The mineral in tooth enamel and dental caries.
In: Biominerals, F.C.M. and Verbeeck, R.M.H. (eds), CRC Press, Boca Raton, Florida,
pp. 105–161
3. Driessens, F.C.M. and Verbeeck, R.M.H. (1990) Dentin, its mineral and caries, In: Biominerals,
Driessens, F.C.M. and Verbeeck, R.M.H. (eds), CRC Press, Boca Raton, Florida, pp 163–178
4. Söremark, R. and Samsahl, K. (1961) Gamma-ray spectrometric analysis of elements in nor-
mal human enamel. Arch. Oral. Bio., Special Suppl., 6, 275–283.
5. Derise, N.L., Ritchey, S.J. and Furr, A.K. (1974) Mineral composition of normal human
enamel and dentin and the relation of composition to dental caries: I Macrominerals and com-
parison of methods of analyses. J. Dental Res., 53(4),847–852
6. LeGeros, R.Z., Silverstone, L.M., Daculsi, G. et al. (1983) In vitro caries-like lesion formation
in F-containing tooth enamel. J. Dental Res., 62(2), 138–144
7. Lakomaa, E-L. and Rytömaa, I. (1977) Mineral composition of enamel and dentin of primary
and permanent teeth in Finland. Scand. 1. Dent. Res., 85, 89–95.
8. Cutress, T.W. (1979) A preliminary study of the microelement composition of the outer layer
of dental enamel. Caries Res., 13, 73–79.
9. Losee, F.L., Cutress, T.W. and Brown, R (1974) Natural elements of the periodic table in
human dental enamel. Caries Res., 8, 123–134.
10. Retief, D.H., Cleaton-Jones, P.E., Turkstra, J. et al. (1971) The quantitative analysis of sixteen
elements in normal human enamel and dentine by neutron activation analysis and high-
resolution gamma-spectrometry. Arch. Oral Bio., 16, 1257–1267.
11. Curzon, M.E.J. and Losee, F.L. (1977) Dental caries and trace element composition of whole
human enamel: Eastern United States. J. Amer. Dental Assoc., 94, 1146–1150.
12. Curzon, M.E.J. and Losee, F.L. (1978) Dental caries and trace element composition of whole
human enamel: Western United States. J. Amer. Dental Assoc., 96, 819–822.
13. Kodaka, T., Debari, K., Yamada, M. et al. (1992) Correlation between microhardness and
mineral content in sound human enamel. Caries Res., 26, 139–141.
14. Panighi, M. and G’Sell, C. (1992) Influence of calcium concentration on the dentin wetability
of an adhesive. J. Biomed. Mater. Res., 26, 1081–1089.
36 K.E. Healy

15. Holcomb, D.W. and Young, R.A. (1980) Thermal decomposition of human tooth enamel.
Calcif Tiss. Intern., 31, 189–201
16. Sakae, T. (1988) X-Ray diffraction and thermal studies of crystals from the outer and inner
layers of human dental enamel. Archs. Oral Bio., 33(10), 707–713.
17. Huang, T.-J.G., Schilder, H. and Nathanson, D. (1992) Effects of moisture content and end-
odontic treatment on some mechanical properties of human dentin. J. Endodontics, 18(5),
209–215
18. Kerebel, B, Daculsi, G. and Kerebel, L.M. (1979) Ultrastructure studies of enamel crystallites.
J. Dental Res., 58(B), 844–851.
19. Jervøe, P. and Madsen, H.E.L. (1974) Calcium phosphates with apatite structure. I. Precipitation
at different temperatures. Acta Chem. Scand., A28, 477–481.
20. Daculsi, G., Kerebel, B. and Verbaere, A (1978). (Méthode de mesure descristaux d’apatite de
la dentine humanie en microscopie électronique en transmission de Haute Résolution)(Fr.)
(Method of measurement of apatite crystals in human dentin by high resolution transmission
electron microscopy), Comptes Rendu Acad. Sci. Paris, Sér. D., 286, 1439.
21. Voegel, J.C. and Frank, R.M. (1977) Ultrastructural study of apatite crystal dissolution in
human dentine and bone. Jour. Bioi. Buccale, 5, 181–194.
22. Lehman, M.L. (1963) Tensile strength of human dentin. J. Dent. Res., 46(1), 197–201
23. Stanford, J.W., Weigel, K.V., Paffenbarger, G.C. et al. (1960) Compressive properties of hard
tooth tissues and some restorative materials. J. American Dental Assoc., 60, 746–756.
24. Jameson, M.W., Hood, J.A.A. and Tidmarsh, B.G. (1993) The effects of dehydration and rehy-
dration on some mechanical properties of human dentine. J. Biomech., 26(9), 1055–1065.
25. Craig, R.G., Peyton, F.A. and Johnson, D.W. (1961) Compressive properties of enamel, dental
cements, and gold. J. Dent. Res., 40(5), 936–945.
26. Korostoff, E., Pollack, S.R and Duncanson, M.G. (1975) Viscoelastic properties of human
dentin. J. Biomed. Mater. Res., 9, 661–674.
27. Bowen, R.L. and Rodriguez, M.S. (1962) Tensile strength and modulus of elasticity of Tooth
Structure and Several Restorative Materials. J. American Dental Assoc., 64, 378–387.
28. Carter, J.M., Sorensen, S.E., Johnson, R.R., et al. (1983) Punch shear testing of extracted vital
and endodontically treated teeth. J. Biomech., 16(10), 841–848.
29. Hassan, R, Caputo, A.A. and Bunshah, R.F. (1981) Fracture toughness of human enamel.
J. Dent. Res., 60(4), 820–827.
30. Rasmussen, S.T., Patchin, R.E., Scott, D.B. et al. (1976) Fracture properties of human enamel
and dentin. J. Dent. Res., 55(1), 154–164.
31. Caldwell, R.C., Muntz, M.L., Gilmore, R.W. et al. (1957) Microhardness studies of intact
surface enamel. J. Dent. Res., 36(5), 732–738.
32. Remizov, S.M., Prujansky, L.Y. and Matveevsky, R.M. (1991) Wear resistance and microhard-
ness of human teeth. Proc. Inst. Mech. Eng., Part H: J. Eng. in Med., 205(3), 201–202.
33. Davidson, C.L., Hoekstra, I.S. and Arends, J. (1974) Microhardness of sound, decalcified and
etched tooth enamel related to the calcium content. Caries Res., 8, 135–144.
34. Pashley, D.H., Andringa, H.J., Derkson, G.D. et al. (1987) Regional variability in the perme-
ability of human dentin. Arch. Oral Biol., 32(7), 519–523.
35. Pashley, D.H., Okabe, A. and Parham, P. (1985) The Relationship between dentin microhard-
ness and tubule density. Endod. Dent. Traumatol., 1, 176–179.
36. Baier, R.E. and Zisman, W.A. (1975) Wetting properties of collagen and gelatin surfaces, in
‘Applied Chemistry at Protein Interfaces’, vol. 145, Advances in Chemistry series (ed.
R.F. Gould), American Chemical Society, Washington DC, pp. 155–174.
37. Jendresen, M.D., Baier, R.E. and Glantz, P-O. (1984) Contact angles in a biological setting:
Measurements in the human oral cavity. J. Coli. Interface Sci., 100(1), 233–238.
38. Baier, R.E. (1973) Occurrence, nature, and extent of cohesive and adhesive forces in dental
integuments. in: Surface Chemistry and Dental Integument’s. Lasslo, A. and Quintana, R.P.
(eds), Thomas, Springfield, IL pp. 337–391.
39. Glantz, P-O. (1969) On wetability and adhesiveness. Odontologisk Revy, 20 supp. 17, 1–132.
40. Jendresen, M.D. and Glantz, P-O. (1980) Clinical adhesiveness of the tooth surface. Acta
Odontol. Scand., 38, 379–383.
Chapter B1
Cartilage

J.R. Parsons

B1.1 Introduction

B1.1.1 Articular cartilage

Articular or hyaline cartilage forms the bearing surfaces of the movable joints of the
body. Hyaline cartilage also exists in tissues of the larynx, tracheal tube rings, rib
and costral cartilage, nasal septum and in the growth plates of long bones. As a
bearing surface, this tough, resilient tissue displays exceptional mechanical and
tribologic properties due exclusively to the unique interaction of the constituents of
the tissue extracellular matrix. Usually, the phenotypic cells (chondrocytes) of car-
tilage make up less than 10% of the total volume of the tissue and have not been
considered to contribute to the mechanical properties of the tissue. The extracellular
matrix consists of a tight collagen fiber network which contains and constrains a
highly hydrophilic gel of aggregated proteoglycan macromolecules. Collagen
accounts for approximately 50% of the dry weight of the tissue, the remainder being
proteoglycans and cellular material. In the fully hydrated state, water contributes
60% to 80% of the wet weight of the tissue. Mechanically, intact normal articular
cartilage behaves as a linear viscoelastic solid. This behavior is the result of viscous
drag of fluid through the tissue in concert with the intrinsic properties of the extra-
cellular matrix. Further, fluid exudation across the cartilage surface in response to
physiologic loading is thought to play a significant role in the lubrication of joints.
The importance of articular cartilage as a bearing surface has led to extensive
mechanical and tribologic studies of this tissue.

J.R. Parsons (*)

Orthopaedics-UMDNJ 185 South Orange A venue
University Heights, Newark, NJ 07103–2714, USA

© Springer Science+Business Media New York 2016 37

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_4
38 J.R. Parsons

B1.1.2 Fibrocartilage

Fibrocartilage contains a higher dry weight percentage of collagen and less proteo-
glycan than does articular cartilage. Consequently, in the hydrated state, fibrocarti-
lage contains less water. Fibrocartilage is generally considered to be tougher and
somewhat less resilient than articular cartilage. In humans, fibrocartilage is found in
the meniscus of the knee joint, the annulus fibrosus of the intervertebral disc and in
the temperomandibular joint. The mechanical behavior of fibrocartilage has not
been studied to the extent of articular cartilage; however, it is not as satisfactory a
load bearing material.

B1.1.3 Elastic cartilage

Elastic cartilage is more elastic and resilient in nature than is either hyaline or fibro-
cartilage. This is the result of lower collagen content coupled with the presence of
elastin fibers. Elastic cartilage is found in the human epiglottis, external ear struc-
tures and eustachian tube. As elastic cartilage is not a major structural component
of the musculoskeletal system, the mechanical properties of this tissue have been
largely ignored.

B1.2 Composition
Table B1.1 Cartilage Composition
Water (wt%) Organic (wt%) Source
Articular Cartilage 60–80 20–40 [1]
Type II Collagen — 15–20
Other collagen — <2
Proteoglycan — 10
Fibrocartilage 74 26 [2]
Type I Collagen — 20
Other collagen — <1
Proteoglycan <1

B1.3 Mechanical Properties of Articular Cartilage

B1.3.1 Compression (Table B1.2)

Compressive cartilage properties have often been examined using creep indenta-
tion, confined compression or free/unconfined compression methods. Indentation
techniques permit in situ testing without the necessity of special specimen prepara-
tion as with tensile testing. However, the extraction of intrinsic mechanical
B1 Cartilage 39

Table B1.2 Compressive Properties of Articular Cartilage

Location Value (MPa) Source
Un relaxed (initial):
Creep modulus (indentation):
Femoral head 1.9–14.4 [7]
Young’s modulus (indentation):
Patella 2.25 [8]
Young’s Modulus (confined compression):
Tibial plateau 5.1–7.9* [9]
11.6* [6]
Bulk Modulus (confined compression):
Tibial plateau 31–56 [9]
25* [6]
Young’s Modulus (unconfined compression):
Various 8.4–15.3 [7]
Relaxed (equilibrium):
Aggregate modulus (Indentation):
Patella 0.3–0.6 [4]
0.3–1.5 [10]
Femoral condyle 0.4–1.0 [10]
0.5–0.7 [11]
Carpo-metacarpal joint 0.4–0.8 [12]
Young’s modulus (confined compression):
Patella 0.7 [8]
Tibial plateau 0.7* [6]
Bulk modulus (confined compression):
Tibial plateau 9.1* [6]
* Calculated from other measurements.

parameters from creep indentation data is analytically complex [3, 4]. Confined
compression or unconfined compression tests require preparation of cylindrical
cored specimens of tissue and underlying bone. With unconfined compression, the
free draining tissue edges and low aspect ratio, layered nature of the test specimen
may introduce error. Compression of a laterally confined specimen by a porous
plunger produces uniaxial deformation and fluid flow. Confined compression creep
data has been analyzed to yield an aggregate equilibrium compressive modulus and
permeability coefficient [5] and uniaxial creep compliance [6].

B1.3.2 Tensile (Table B1.3)

Tensile properties for human articular have been determined by cutting standard
tensile specimens from the cartilage surface and performing constant strain rate,
creep or stress relaxation tensile tests. Test results are strongly influenced by
40 J.R. Parsons

Table B1.3 Tensile Properties of Articular Cartilage

Location Value (MPa)* Source
Relaxed (equilibrium):
Young’s modulus: [13]
Femoral condyle,
Surface zone 10.5
Subsurface zone 5.5
Middle zone 3.7
Un relaxed (100%/min):
Young’s modulus: [7]
Femoral condyle,
Surface zone 200–400
Middle zone 40–175
Strength: [7]
Femoral condyle,
Surface zone 20–35
Middle zone 11–25
* All measurements parallel to collagen direction.

Table B1.4 Shear Properties of Articular Cartilage

Location Value (MPa) Source
Relaxed (equilibrium):
Shear modulus:
Patella,
Middle zone 0.25 [14]
Tibial plateau 2.6 [6]
Unrelaxed (initial):
Shear modulus:
Tibial plateau 4.1 [6]
5.1–7.9 [ 9]

collagen volume fraction and orientation and are largely insensitive to proteoglycan
content [7] Collagen volume fraction and orientation is highest in the cartilage sur-
face layer. Collagen content and orientation diminishes in subsequent lower layers.

B1.3.3 Shear(Table B1.4)

Shear properties for articular cartilage have been determined through torsional
creep, stress relaxation and torsional dynamic tests of excised cartilage disks. Creep
and dynamic shearing of rectangular cartilage specimens between plates has been
conducted on animal tissue (usually bovine) but not human tissue. When torsional
shear strains remain small, the observed shear properties are flow independent. That
is, under small strain conditions, fluid flow is negligible and viscoelastic behavior
can be attributed strictly to the collagen/proteoglycan extracellular matrix.
B1 Cartilage 41

B1.3.4 Poisson’s ratio

Poisson’s ratio has been calculated directly from tensile tests (v = 0.37–0.50) [10]
and indirectly from torsional shear and confined compression creep data (v = 0.37–
0.47) [6, 9]. More recently, the relationship between Poisson’s ratio, n, aggregate
modulus, Ha, and permeability, k, have been established for cartilage indentation
testing based on biphasic (fluid and porous solid) constitutive theory [15]. Using a
complex numerical solution and curve fitting scheme, Poisson’s ratio can be
extracted from indentation data, resulting in values of v = 0.00–0.30 [11, 12, 16].
However, care must be exercised in interpreting such indirect measures of Poisson’s
ratio as unexpected results can arise; e.g. v = 0.0.

B1.3.5 Permeability

The porous solid matrix of articular cartilage permits the movement of interstitial
water in response to a pressure gradient. Flow of water through and across the tissue
is largely responsible for the viscoelastic character of cartilage. Flow is related to
tissue permeability through the hydraulic permeability coefficient, k, as defined by
Darcy’s law. The permeability coefficient has been measured in flow chambers
where a known pressure gradient produces flow across a cartilage layer of known
thickness and area (k = 4.0 – 17.0 x 10-16m4/Ns)[7]. However, such experiments have
demonstrated significant decreases in permeability coefficient with increasing pres-
sure gradient, increasing compressive tissue strain and with increasing proteoglycan
content. Evoking biphasic constitutive theory with numerical solutions and/or curve
fitting routines permits an indirect determination of the permeability coefficient
from confined compression creep data and creep indentation data (k = 5.2 – 21.7 ×
10-16m4/Ns)[11, 12].

B1.3.6 Articular cartilage tribologic properties

Healthy articular cartilage has remarkable tribologic properties. Under high load
conditions the tissue displays extremely low frictional coefficients and virtually
undetectable wear. The dynamic coefficient of friction, μd, has been measured in
whole joints using Stanton pendulum or other pendulum techniques where the joint
forms the pendulum pivot (μd = 0.015–0.04)[17–19]. The coefficients of friction of
cartilage plugs bearing on other materials has been determined for human and ani-
mal tissue but these sorts of experiments have little relevance for actual in situ car-
tilage behavior and are not reported here.
The lubrication of articular cartilage remains a subject of continuing debate and
no one lubrication mechanism can be clearly identified. Both fluid film and bound-
ary lubrication are thought to play primary roles in joint lubrication and the domi-
nance of one or the other probably depends on loading and velocity conditions.
Further as cartilage is a relatively soft viscoelastic material, elastohydrodynamics
42 J.R. Parsons

may discourage fluid film breakdown and thus promote hydrodynamic lubrication.
Exudation of fluid across the cartilage surface in response to an advancing load has
also been suggested to aid lubrication.
No reliable wear tests have been performed on human articular cartilage bearing
surfaces under physiologic conditions.

B1.4 Fibrocartilage Mechanical Properties

Human fibrocartilage tensile mechanical properties have been determined by cutting

standard tensile specimens either from the knee meniscus or from single or multiple
lamella from the annulus fibrosus and performing constant strain rate tensile tests.
Test results are strongly influenced by collagen volume fraction and orientation and
are largely insensitive to proteoglycan content. Annulus fibrosus has also been tested
in confined compression, permitting derivation of an aggregate compressive modulus
and permeability. Data are reported in Section B2.

B1.5 Elastic Cartilage Mechanical Properties

No reliable data are available for human tissue.

Additional Reading

Freeman, MAR (ed.) (1979) Adult Articular Cartilage, 2nd ed., Pitman Medical
Publishing Co, Kent, UK.
Although now somewhat out of date, this classic text forms the basis for current
thinking on cartilage biochemistry, physiochemistry, biomechanics and tribology.
The volume of original data, found nowhere else, is truly impressive.
Mow V.C., Holmes, M.H. and Lai, W.M. (1984): Fluid transport and mechanical
properties of articular cartilage. A review. J. Biomech., 17:377–394.
This survey article provides an historical perspective of cartilage mechanics
research and leads the reader through the modern biphasic theory of cartilage
mechanics at the material level. References provided are particularly useful in
developing a bibliography of the important classic studies in this field.
Mow, V.C. and Ratcliffe, A (eds)(1993): Structure and Function of Articular
Cartilage, CRC Press, Boca Raton.
This up-to-date monograph is perhaps the best current work on the subject.
Details from many of the references in this section (below) can be found in the sec-
tion on cartilage biomechanics.
B1 Cartilage 43

References

1. Maroudas, A. (1979) Physiochemical properties of articular cartilage. in Adult Articular

Cartilage, 2nd ed., M.A.R. Freeman (ed.), Pitman Medical Publishing Co, Kent, UK,
pp. 215–290.
2. Fithian, D.C., Kelly, M.A. and Mow, V.C. (1990) Material properties and structure-function
relationships in the menisci. Clin. Orthop. Rel. Res., 252, 19–31.
3. Mak, A., Lai, W.M. and Mow, V.C. (1987) Biphasic indentation of articular cartilage: Part I,
Theoretical analysis. J. Biomech., 20, 703–714.
4. Mow, V., Gibbs, M.C. and Lai, W.M., et al. (1989) Biphasic indentation of articular cartilage:
Part II, A numerical algorithm and experimental study. J. Biomech., 22, 853–861.
5. Mow, V., Kuei, S.C. and Lai, W.M. (1980) Biphasic creep and stress relaxation of articular
cartilage in compression: Theory and experiments. J. Biomech, Eng., 102, 73–84.
6. Hayes, W. and Mockros, L.F. (1971) Viscoelastic properties of human articular cartilage.
J. Appl. Physiol., 31, 562–568.
7. Kempson, G. (1979) Mechanical properties of articular cartilage. In Adult Articular Cartilage,
2nd ed., Freeman M.A.R., Editor, Pitman Medical Publishing Co. Ltd., Kent, England,
pp. 333–414.
8. Sokoloff, L. (1966) Elasticity of aging cartilage. Fed. Proc., 25, 1089–1095.
9. Hori, R. and Mockros, L.F. (1976) Indentation tests of human articular cartilage. J. Biomech.,
9, 259–268.
10. Armstrong, C. and Mow, V.C. (1982) Variations in the intrinsic mechanical properties of
human articular cartilage with age, degeneration and water content. J. Bone Joint Surg., 64A,
88–94.
11. Athanasiou, K., Rosenwasser, M.P., and Buckwalter, J.A., et al. (1991) Interspecies compari-
son of in situ intrinsic mechanical properties of distal femoral cartilage. J. Orthop. Res., 9,
330–340.
12. Ateshian, G., Gardner, J.R., Saed-Nejad, F. et al. (1993) Material properties and biochemical
composition of thumb carpometacarpal joint cartilage. Trans. Orthop. Res. Soc., 18, 323.
13. Akizuki, S., Mow, V.C. and Muller, F., et al. (1986) Tensile properties of human knee joint
cartilage: Part I, Influence of ionic concentrations, weight bearing and fibrillation on the tensile
modulus. J. Orthop. Res., 4, 379–392.
14. Zhu, W., Lai, W.M. and Mow, V.C. (1986) Intrinsic quasilinear viscoelastic behavior of the
extracellular matrix of cartilage. Trans. Orthop. Res. Soc., 11, 407.
15. Mak, A. (1986) The apparent viscoelastic behavior of articular cartilage- The contributions
from the intrinsic matrix viscoelasticity and interstitial fluid flows. J. Biomech. Eng., 108,
123–130.
16. Akizuki, S., Mow, V.C., Lai, W.M., et al. (1986) Topographical variation of the biphasic inden-
tation properties of human tibial plateau cartilage. Trans. Orthop. Res. Soc., 11, 406.
17. Charnley, J. (1960) The lubrication of animal joints in relation to surgical reconstructions by
arthroplasty. Ann. Rheum. Dis., 19, 10–19.
18. Little, J., Freeman, M.A.R., and Swanson, S.V. (1969) Experience on friction in the human
joint. In Lubrication and Wear in Joints, Wright V., Editor, Sector Publishing, London, UK.,
pp. 110–114.
19. Unsworth, A., Dawson, D. and Wright, V. (1975) Some new evidence on human joint lubrica-
tion. Ann. Rheum. Dis., 34, 277–285.
Chapter B2
Fibrocartilage

V.M. Gharpuray

B2.1  Introduction

The human menisci and intervertebral discs perform several important mechanical
functions in the human body. The ability to perform these functions and conse-
quently their intrinsic biomechanical properties are dependent on the interaction of
the constituents of these structures. Both the menisci and intervertebral discs have a
fibrocartilaginous structure that consists of two distinct phases: a fluid phase con-
sisting of mainly water and dissolved electrolytes, and a solid phase composed of
highly oriented collagen fibers, cells, proteoglycans and other proteins. As with all
other biological materials, both menisci and discs exhibit non-linear viscoelastic
and anisotropic properties. The non-linear stiffness or elasticity of the structure is
imparted by the collagen fibers and to a lesser extent by osmotic pressures within
the tissue which are generated by the degree of hydration [1, 2]. The viscoelastic or
energy dissipation properties are a result of fluid flow within and through the struc-
tures and also of molecular relaxation effects from the motion of long chains of
collagen and proteoglycans [3]. Anisotropy is a consequence of the orientation and
concentration of collagen fibers within the proteoglycan gel.

B2.2  Structure and Composition

A normal adult human knee contains two menisci – the lateral and the medial,
whose average lengths are 38 and 45 mm, and average volumes are 2.9 and 3.45 cm3
respectively [4]. At the femoral articulating surface of each meniscus, for a depth of

V.M. Gharpuray (*)

Department of Bioengineering 401 Rhodes Eng. Res. Ctr., Clemson University,
Clemson, SC 29634-0905, USA

© Springer Science+Business Media New York 2016 45

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_5
46 V.M. Gharpuray

approximately 100 µm, fine collagen fibrils (mainly Type II) are randomly oriented
to form a woven mesh [5, 6]. Beneath this surface layer, larger rope-like bundles
(approximately 100 µm in diameter) of Type I collagen are arranged predominantly
in the circumferential direction. In the posterior half of the medial meniscus how-
ever, the fibers are not as highly oriented [6, 7]. A few radial fibers may also be seen
interspersed within the circumferential fibers [6–8]. At the tibial surface is another
articulating layer, in which the principal orientation of the fibers is radial.
Collagen and other proteins make up the organic content of the meniscus, while
dissolved electrolytes make up the inorganic content (which is negligibly small).
The collagen is primarily Type I (98%) with small amounts of Type II, III and
Type V [9], and comprises up to 25% of the wet weight and 90% of the dry weight
in human material (Table B2.1). 10% of the dry weight and up to 2% of the wet
weight are due to non-collagenous proteins, which consist predominantly of proteo-
glycans, and smaller amounts of structural glycoproteins, cell membrane bound
receptors and intercalated membrane glycoproteins. These proportions do not
appear to vary with location in the menisci [10].
There are 23, approximately cylindrical, intervertebral discs in the human spine
that account for 20–30% of its overall length [13]. The cross sectional shape varies
with level: roughly elliptical, rounded triangular and kidney shaped in the cervical,
thoracic and lumbar regions respectively [14]. The cross sectional shape is some-
times quantified by a shape index (S) defined by S = 4πA/C2 where A= cross sec-
tional area, and C =circumference of the cross section (Table B2.2). Most discs are
wedge-shaped in sagittal section with the anterior height greater that the posterior
height. The cross sectional area increases with level such that lumbar discs are
larger than cervical discs.
The intervertebral disc contains two distinct regions, the nucleus pulposus and
the annulus fibrosus. The nucleus occupies about 50% of the volume of the disc, and
contains mostly water and small amounts of randomly oriented collagen fibers, cells
and non-collagenous proteins [19]. The annulus is a tough ring-like structure that
surrounds the nucleus. Highly oriented collagen fibers (primarily Types I and II)
form a laminate structure in the annulus, with approximately 20–25 laminae ori-
ented alternately at approximately +(60–70)° and –(60–70)° to the spinal axis [14,
20, 21]). Type IX collagen is believed to cross link the Types I and II fibers, and
provide some resistance to circumferential tears. Collagen content and fiber orienta-
tion is highest in the outermost layers, and both decrease as the nucleus is approached
from the periphery of the disc (Table B2.3) The water content in the disc varies with
position in the disc: it is highest in the nucleus and lowest in the outermost layers of
the annulus [22, 23]. It has also been shown that the water content varies with cir-
cumferential position [4], and is higher in the posterior of the disc. The nucleus

Table B2.1 Composition Wet weight Dry weight

of the menisci
Water Collagen NCP* Collagen NCP
60–70%a 15–25%a 1–2%a 70–90%a,b 8–20%b,c‡
* NCP = non-collagenous proteins; ‡ 21.9% in neonates
decreasing to 8.1 % between 30–70 years. a: [8];b: [11];c: [12].
B2 Fibrocartilage 47

Table B2.2  Shape and size of the adult intervertebral disc

Shape Posterior/ Cross Sectional Height
Disc Level Index (Sa) Anterior Heighta Area (mm2) (mm)
L5-S1 0.885 0.35
L4-L5 0.897 0.51 1714b 11–12b,c
L3-L4 0.866 0.55 1662b 10.4b
L2-L3 0.866 0.61 1859b 9.75b
Ll-L2 0.825 0.68 1640b 8.83b
T12-Ll 0.844 0.75
Tl1-T12 0.856 0.80
Tl0-T11 0.885 1.11
T9-Tl0 0.879 0.74
T8-T9 0.919 0.88 4–6c
T7-T8 0.878 0.81
T6-T7 0.898 0.84
T5-T6 0.935 1.07
T4-T5 0.868 0.97
T3-T4 0.836 0.72
T2-T3 0.870 0.74
Tl-T2 0.815 0.76
C7-Tl 0.785 0.62 1292* 6.00d
C6-C7 0.708 0.82 1152* 5.67d
C5-C6 0.828 0.44  949* 5.50d
C4-C5 0.825 0.47  892* 5.25d
C3-C4 0.870 0.50  827* 5.00d
C2-C3 0.893 0.56  732* 4.75d
* Computed from data in (18). a: (13);b: (15,16);c: (17);d: (18).

Table B2.3  Composition of the intervertebral disc in young adults

Tissue Water Content % Collagen Content %
Nucleus Pulposus 85–95a 2–5b
Annulus Fibrosus 85 (innermost layer) 5 (innermost layer)
65 (outermost layer) 21 (outermost layer)
a: [23]; b: [24].

loses water and becomes more fibrous and desiccated with age, causing the bound-
ary between the annulus and the nucleus to become less clear [19, 23].

B2.3  Hydraulic Permeability and Drag Coefficients

Experimental data suggest that water is capable of flowing through both meniscal
and discal tissues, and is dependent on a material property of the tissue called
hydraulic permeability [25], which may be modeled by Darcy’s Law as:
48 V.M. Gharpuray

AD P
Q=k (B2.1)
h

where k = hydraulic permeability coefficient of the tissue; Q = volume rate of

fluid flow; A = area across which fluid flow occurs; h = thickness of the tissue; and
∆P = pressure gradient across the thickness h that causes fluid flow.
The diffusive drag coefficient K is related to the permeability coefficient by

(f f ) 2
K= (B2.2)
k

where Φf is the porosity of the tissue and is defined as the ratio of interstitial fluid
volume to total tissue volume.
For human meniscal (annulus) tissue, k, the permeability is 2.5 × 10-16 m4/Ns
[22], about one third of that reported for bovine tissue: 8.1 ×10-16 m4/Ns [26]. There
is no significant variation in the permeability coefficient with location of the speci-
men. The porosity (Φf) of both tissues is approximately 0.75, and therefore the drag
coefficient, K, is very high and ranges from 1014 to 1015 Ns/m4.

B2.4  Elastic Properties

Under quasi-static loading, or in conditions under which ‘short-term’ loading

responses are expected to occur [27], both meniscal and discal tissues may be mod-
eled as linear elastic and orthotropic. Under a constant load rate, the non-linear
behavior may be described by an exponential stress-strain relationship given by

s = A[e Be - 1] (B2.3)

where A and B are constants for the given material. The constant B is proportional
to the tangent modulus (i.e., dσ/dϵ), and sometimes a third constant C is defined as
C = A*B, and is the tangent modulus as σ → 0 [8].
The macroscopic tensile strength of the entire meniscus was studied by Mathur
et al. [4] by gripping the horns of the meniscus, and stretching it to failure. The
results suggested that the medial meniscus was significantly weaker than the lateral
meniscus (ultimate loads of 247 N and 329 N respectively), and that the mode of
failure was not by transverse cracking, but predominantly by oblique (medial) or
spiral (lateral) tearing.
Strength and modulus of the meniscus vary with different locations and with dif-
ferent orientations of the specimen due to structural and compositional changes
(Table B2.4). For loading parallel to the fibers, it appears that the meniscus may be
stronger in the anterior location, and that the lateral meniscus may be stronger than
the medial meniscus. This may be explained in part by the fact that the fiber orienta-
tion is more random in the posterior part of the medial meniscus.
B2 Fibrocartilage 49

Table B2.4  Tensile strength Location

of meniscal tissue (MPa)a,*
Meniscus Orientation† Anterior Central Posterior
Lateral Parallel 10.37 6.31 6.87
Perpendicular  0.80 0.88 0.54
Medial Parallel — 3.36 5.86
Perpendicular — 0.85 1.23
a: Averaged from data in [6].
* Tissues were fixed in formalin before testing.
† Either parallel or perpendicular to the circumferential direction.

Table B2.5  Non-linear parameters and tensile modulus of menisci

Tensile
modulusa
Meniscus Location A B C (MPa)
Medial Anterior 1.6 28.4 42.4 159.6
Central 0.9 27.3 23.7  93.2
Posterior 1.4 20.1 25.2 110.2
Lateral Anterior 1.4 28.8 30.2 159.1
Central 2.1 31.9 55.7 228.8
Posterior 3.2 27.5 67.5 294.1
a: Slope of the stress-strain curve in the linear portion after the toe region.

A similar trend is seen in the tensile modulus of meniscal specimens oriented

parallel to the circumferential direction (Table B.2.5), and if a power law is used as
the constitutive equation (equation B2.3), the coefficients A and C show an identical
pattern [7, 10, 28].
The properties of intervertebral discs are more complex than those of the menisci,
since properties vary with disc level, and discs must withstand loads and moments
in three orthogonal directions. (Table B2.6).
As with the meniscus, strength and modulus of discal tissue vary with location
and orientation of the specimen (Table B2.7). Lin et al. [15, 16] have however
shown that elastic moduli of annular specimens are independent of disc level.

B2.5  Viscoelastic Behavior

Finally, the rate dependent properties are usually modeled by a three-parameter

solid which consists of a spring (m2) and a dashpot (h) in parallel connected to
another spring (m1) in series. Viscoelastic properties may also be expressed in terms
of the dynamic modulus G*. A sinusoidal displacement of the form u = uo eiωt is
applied to the specimen (this is usually a torsional strain), and the resulting force
response F = Fo eiωt+δ is measured. Here ϕ = circular frequency, i = √(–1) and δ is
50 V.M. Gharpuray

Table B2.6  Mechanical properties of the intervertebral disc

Stiffness (N/mm or Nm/rad)
Loading mode Level Strength (N) Initiala Averageb Finale
Compression L5-S1 5574d 1448d 3511d
L4-L5 5128d 306d, 413e 2405d, 721c
L3-L4 5351d 1352d 2756d
L2-L3 4905d 439d, 461e 3160d, 997c
Ll-L2
Flexion Lumbar 46f 8451f
Extension Lumbar 74f
Lateral bending Lumbar 64f 704f
Axial torsion Lumbar 157f 604f
Compression Lumbar 800g
g
Posterior shear Lumbar 102 148g
Anterior shear Lumbar 91g 123g
Lateral shear Lumbar 113g 169g
a: Slope of the toe region of the load displacement curve.
b: Average slope of the load displacement curve.
c: Slope of the load displacement curve excluding the toe region.
d: [29]; e: [30]; f: [31]; g: [32].

Table B2.7  Mechanical properties of the annulus fibrosus

Layer Location
Property Specimen orientation Average Inner Middle Outer
Tensile Modulus Horizontal 3.54c
Parallel to fibers 3.41 410a
Perpendicular to lamellae 0.16b
Ultimate stress Horizontal
Parallel to fibers 110a
Perpendicular to lamellae 0.187b
a: MPa, Specimen 2×2.5 mm cross section, 6.5 mm length [33].
b: MPa, Specimen 7.5×2.5 mm cross section, 4.5 mm length [33].
c: N/mm, Specimen 2×1.5 mm cross section, 15–25 mm length [34].

the phase angle shift between the applied displacement and the measured force. The
dynamic modulus is than obtained as

F F0 i d
G* = = e = G¢ + iG¢¢ (B2.4)
u u0
where G′ and G″ are the loss and storage moduli respectively. In some cases, it may
be more convenient to express viscoelastic properties in terms of the magnitude of
the dynamic shear modulus and the phase angle shift as
B2 Fibrocartilage 51

Table B2.8 Viscoelastic Specimen orientation |G*| (MPa) δ (degrees)

properties of meniscal tissue
Circumferential 36.8 16.7
Axial 29.8 19.4
Radial 21.4 20.8

Table B2.9 Viscoelastic m1 (MPa) m2 (MPa) h (GPas)

properties of the intervertebral
10–13 13–40 65–280
disc* [36, 37]
* Ranges.

¢¢ G¢¢
| G | * = (G¢ 2 + G 2 ); d = tan -1 ( ) (B2.5)
G¢

The anisotropic viscoelastic properties in shear of the meniscus have been deter-
mined by subjecting discs of meniscal tissue to sinusoidal torsional loading ­[35] (Table
B2.8). The specimens were cut in the three directions of orthotropicsymmetry, i.e.
circumferential, axial and radial. A definite correlation is seen with the orientation of
the fibers and both the magnitude of the dynamic modulus |G*| and the phase angle δ.
The viscoelastic properties of the human intervertebral disc have been modeled [36,
37] using the three-parameter solid. The parameters were obtained by fitting experimen-
tally obtained creep curves to analytical equations using linear regression (Table B2.9).

B2.6  Discussion

Since it is nearly impossible to carry out meaningful experiments in vivo on the human
disc or meniscus, the properties reported above have been obtained from cadaveric tis-
sue. Test specimens were obtained from autopsy material (10–48 hours after death), and
were either tested immediately or stored frozen for varying periods of time before test-
ing. Statistical analyses of these data show high standard deviations and errors may be a
consequence of aging and degeneration, diurnal changes or surgical interventions) cause
a subsequent change in mechanical properties. Further, it has been well documented that
the mechanical properties of collagenous tissues change with storage medium, storage
temperature, time after death and ‘preconditioning’ state [38, 39]. Both the disc and the
meniscus contain highly oriented collagen fibers, and location and orientation of the test
specimen can cause significant changes in the test results. Additional factors such as sex,
diet, and level of activity also play a relatively minor role in this variation.

Additional Reading

Ghosh, P. (ed.) (1988) The Biology of the Intervertebral Disc, Vol I and II, CRC
Press, Boca Raton.
52 V.M. Gharpuray

One of the most comprehensive texts available about the intervertebral disc. It is
written from the biological perspective, and contains exhaustive information about
each component of the disc. Volume I includes chapters on disc structure and
development, vasculature, innervation, collagen and non-collagenous proteins.
Volume II contains information on nutrition and metabolism, mechanics, pathology
and disease states.
Mow, V.C., Arnoczky, S.P. and Jackson, D.W. (1992) Knee Meniscus: Basic and
Clinical Foundations. Raven Press, New York.
This monograph is designed to serve as a comprehensive reference for clinicians
and researchers interested in the meniscus. It includes chapters on gross anatomy,
structure and function of the menisci and their mechanical behavior, pathological
disorders, clinical and surgical methods of treatment and meniscal disorders.
Mow, V.C. and Hayes, W.C. (1991) Basic Orthopaedic Biomechanics, Raven
Press, New York.
This book is aimed at teaching senior engineering students or orthopaedic resi-
dents the fundamental principles of biomechanics of the musculoskeletal system.
The book contains several chapters on the mechanics of joints, and the properties
and functions of joint tissues. The chapter devoted to articular cartilage and the
meniscus includes a review of collagen-proteoglycan interactions, and how these
directly affect the mechanical behavior of the tissue. The biphasic and the triphasic
theories for the viscoelastic properties are also discussed.
White, A.A. and Panjabi, M.M., (1990) Clinical Biomechanics of the Spine,
J.B. Lippincott Company, Philadelphia.
An excellent reference book for an engineer or a physician interested in the spine.
Each topic is written from the viewpoint of a biomechanician and the topics covered
include kinetics and kinematics of vertebral joints, pathological disorders of the
spine and their surgical management. Chapter 1 contains an introductory section on
the intervertebral disc that describes its structure, function and biomechanics.

References

1. Armstrong, C.G. and Mow, V.C. (1982) Variations in the intrinsic mechanical properties of
human articular cartilage with age, degeneration and water content. J. Bone Joint Surg., 64A,
88–94.
2. Mow, V.C., Holmes, M.H. and Lai, W.M. (1984) Fluid transport and mechanical properties of
articular cartilage: A review. J. Biomech., 102, 73–84.
3. Hayes, W.C. and Bodine, A.J. (1978) Flow independent viscoelastic properties of articular
cartilage matrix. J. Biomech., 11, 407–419.
4. Mathur, P.D., McDonald, J.R. and Ghormley, R.K. (1949) A study of the tensile strength of the
menisci of the knee. J. Bone Joint Surg., 32A, 650–654.
5. Aspden, R.M., Yarker, Y.E. and Hukins, D.W.L. (1985) Collagen Orientations in the meniscus
of the knee joint. J. Anat., 140, 371–380.
6 Bullough, P.G., Munuera, L., Murphy, J. and Weinstein, A.M. (1970) The strength of the
menisci of the knee as it relates to their fine structure. J. Bone Joint Surg., 52B, 564–570.
B2 Fibrocartilage 53

7 Fithian, D.C., Kelly, M.A. and Mow, V.C. (1990) Material properties and structure-function
relationships in the menisci. Clin. Orthop. Rei. Res., 252, 19–31.
8 Mow, V.C., Zhu, W. and Ratcliffe, A. (1991) Structure and function of articular cartilage and
meniscus, in Basic Orthopaedic Biomechanics, (eds V.C. Mow and W.C. Hayes), Raven Press,
New York, pp. 143–198.
9 Eyre, D.R. and Wu, J.J. (1983) Collagen of fibrocartilage: A distinctive molecular phenotype
in bovine meniscus. F. E. B. S. Letters, 158, 265–270.
10 Fithian, D.C., Zhu, W.B., Ratcliffe, A., Kelly, M.A. and Mow, V.C. (1989b) Exponential law
representation of tensile properties of human meniscus. Proceedings of the Institute of
Mechanical Engineers. The Changing Role of Orthopaedics, Mechanical Engineering
Publications Limited, London. pp. 85–90
11 Ghosh, P. and Taylor, T.K.F. (1987) The knee joint meniscus: A fibrocartilage of some distinc-
tion. Clin. Orthop. Rei. Res., 224, 52–63.
12 Ingman, A.M., Ghosh, P. and Taylor, T.K.F. (1974) Variation of collagenous and non-­
collagenous proteins of human knee joint menisci with age and degeneration. Gerontology, 20,
212–223.
13 Pooni, J.S., Hukins, D.W.L., Harris, P.F., Hilton, R.C. and Davies, K.E. (1986) Comparison of
the structure of human intervertebral discs in the cervical, thoracic and lumbar regions of the
spine. Surg. Radiol. Anat., 8, 175–182.
14 Hirsch, C., Inglemark, B-H. and Miller, M. (1963) The anatomical basis of back pain. Acta
Orthop. Scand., 33, 2–17.
15 Lin, H.S., Liu, Y.K. and Adams, K.H. (1978) Mechanical response of the lumbar intervertebral
joint under physiological (complex) loading. J. Bone Joint Surg., 60A, 41–55.
16 Lin, H.S., Liu, Y.K., Ray, G. and Nikravesh, P. (1978) System identification for material prop-
erties of the intervertebral joint. J. Biomech., 11, 1–14.
17 Taylor, J.R. (1975) Growth of human intervertebral discs and vertebral bodies. J. Anat., 120,
49–68.
18 Panjabi, M.M., Summers, D.J., Pelker, R.R., Videman, T., Friedlander, G.E. and Southwick,
W.O. (1986) Three-dimensional load-displacement curves due to forces on the cervical spine.
J. Ortho. Res., 4, 152–161.
19 Inoue, H. (1981) Three-dimensional architecture of lumbar intervertebral discs. Spine, 6,
139–146.
20 Panagiotacopulos, N.D., Knauss, W.G. and Bloch, R. (1979) On the mechanical properties of
human intervertebral disc materials. Biorheology, 16, 317–330.
21 Marchand, F. and Ahmed, A.M. (1988) Investigation of the laminate structure of lumbar disc
annulus fibrosus. Trans. Orthop. Res. Soc., 13, 271.
22 Best B.A., Guilak, F., Setton, L.A., Zhu, W., Saed-Nejad, F., Ratcliffe, A., Weidenbaum, M.
and Mow, V.C. (1994) Compressive mechanical properties of the human annulus fibrosus and
their relationship to biochemical composition. Spine, 19, 212–221.
23 Gower, W.E. and Pedrini, V. (1969) Age-related variations in protein-polysaccharide from
human nucleus pulposus, annulus fibrosus and costal cartilage. J. Bone Joint Surg., 51A,
1154–1162.
24 Lyons, G., Eisenstein, S.M. and Sweet, M.B.E. (1981) Biochemical changes in intervertebral
disc degeneration, Biophysics Acta, 673, 443.
25 Lai, W.M. and Mow, V.C. (1980) Drag-induced compression of articular cartilage during a
permeation experiment. Biorheology, 17, 111–123.
26 Proctor, C.S., Schmidt, M.B., Whipple, R.R., Kelly, M.A. and Mow, V.C. (1989) Material
Properties of the normal medial bovine meniscus. J. Ortho. Res., 7, 771–782.
27 Eberhardt, A.W., Keer, L.M., Lewis, J.L. and Vithoontien, V. (1990) An analytical model of
joint contact. J. Biomech. Eng., 112, 407–413.
28 Fithian D.C., Schmidt, M.B., Ratcliffe, A. and Mow, V.C. (1989a) Human meniscus tensile
properties: Regional variation and biochemical correlation. Trans. Orthop. Res. Soc., 14, 205.
54 V.M. Gharpuray

29 Brown, T., Hansen, R.J. and Torra, A.J. (1957) Some mechanical tests on the lumbosacral spine
with particular reference to the intervertebral discs. J. Bone Joint Surg., 39A, 1135–1164.
30 Hirsch, C. and Nachemson, A. (1954) New observations on the mechanical behavior of the
lumbar discs. Acta Orthop. Scand., 23, 254–283.
31 Schultz, A.B., Warwick, D.N., Berkson, M.H. and Nachemson, A.L. (1979) Mechanical prop-
erties of human lumbar spine motion segments Part I. J. Biomech. Eng., 101, 46–52.
32 Berkson, M.H., Nachemson, A. and Schultz, A.B. (1979) Mechanical properties of human
lumbar spine motion segments Part I. J. Biomech. Eng., 101,53–57.
33 Marchand, F. and Ahmed, A.M. (1989) Mechanical properties and failure mechanisms of the
lumbar disc annulus. Trans. Orthop. Res. Soc., 14, 355.
34 Galante, J.O. (1967) Tensile properties of the human lumbar annulus fibrosus. Acta Orthop.
Scand., (Suppl. 100), 1–91.
35 Chern, K.Y., Zhu, W.B. and Mow, V.C. (1989) Anisotropic viscoelastic shear properties of
meniscus. Adv. Bioeng., BED-15, 105–106.
36 Burns, M.L. et al. (1984) Analysis of compressive creep behavior of the vertebral unit sub-
jected to uniform axial loading using exact parametric solution equations of Kelvin solid
models Part I. J. Biomech., 17, 113–130.
37 Kazarian, L.E. and Kaleps, I. (1979) Mechanical and physical properties of the human inter-
vertebral joint. Technical Report AMRL-TR-79-3, Aerospace Medical Research Laboratory,
Wright Patterson Air Force Base, OH
38 Black, J. (1976) Dead or alive: The problem of in vitro tissue mechanics. J. Biomed. Mats.
Res., 10, 377–389.
39 Black, J. (1984) Tissue properties: Relation of in vitro studies to in vivo behavior, in Natural
and Living Biomaterials, Ed. G.W. Hastings and P. Ducheyne, CRC Press, Boca Raton,
pp. 5–26.
Chapter B3
Ligament and Tendon

Connie S. Chamberlain and Ray Vanderby Jr.

B3.1 Structure

Ligaments and tendons (T/L) are a hierarchical structure of dense, parallel

connective tissue bands that are hypovascular and hypocellular and span a joint con-
necting bone to bone and bone to muscle, respectively. T/Ls are initially assembled
from cross-linked tropocollagen molecules aggregated progressively into microfi-
brils, subfibrils, fibrils, and fibers (Fig. B3.1). Collagen fibers, which display a wav-
iness or crimp pattern in an unloaded state, are contained in fiber bundles, called
fascicles. The entire T/L is then organized into a collection of fascicles and sur-
rounded by a more vascular connective tissue called the epiligament/epitenon. The
paratenon surrounds the epiligament/epitenon and facilitates gliding along contigu-
ous structures. Altogether, this hierarchical structure provides the T/L with high
tensile force and resilience while preventing damage and separation of the fibers
under mechanical stress.

C.S. Chamberlain (*)

Department of Orthopedics and Rehabilitation, Wisconsin Institute for Medical Research,
University of Wisconsin-Madison, 1111 Highland Avenue, Madison, WI 53705, USA
e-mail: [email protected]
R. Vanderby Jr.
Department of Orthopedics and Rehabilitation, Wisconsin Institute for Medical Research,
University of Wisconsin-Madison, 1111 Highland Avenue, Madison, WI 53705, USA
Department of Biomedical Engineering, Wisconsin Institute for Medical Research, University
of Wisconsin-Madison, 1111 Highland Avenue, Madison, WI 53705, USA

© Springer Science+Business Media New York 2016 55

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_6
56 C.S. Chamberlain and R. Vanderby Jr.

Fig. B3.1 Hierarchical structure of the tendon/ligament. The tropocollagen cross-linked mole-
cules are arranged in microfibrils, subfibrils, fibrils, fibers, and fascicles. The fascicles are sur-
rounded by the epitenon/epiligament and paratenon

Fig. B3.2 Composition of

the T/L

B3.2 Composition

Ligaments and tendons consist of few cells and abundant extracellular matrix. The
cellular component, primarily consisting of fibroblasts/tenocytes, comprises 20 %
of the total tissue volume whereas the ECM accounts for the remaining 80 %
(Fig. B3.2) [1, 2]. Approximately 70 % of the ECM consists of water, which con-
tributes to modulating cellular function and viscoelastic (i.e., time dependent)
behavior. The remaining 30 % of the ECM is comprised of solids, such as collagen
and ground substance. The ground substance (non-fibrous component of the matrix)
primarily comprises hyaluronan, glycoproteins, and proteoglycans and modulates
tissue metabolism, provides shock absorption, decreases internal friction, and binds
water. The fibrous component, comprising 75–80 % of T/L dry weight, provides
B3 Ligament and Tendon 57

strength and a support framework and consists primarily of type I collagen. Lesser
amounts of types III, VI, V, XI, and XIV collagens are also present.

B3.3 Normal T/L Function

The ligament primarily serves a mechanical function to passively stabilize and

guide joints through a normal range of motion. It offers virtually no resistance to
normal joint motion but stiffens to become a “check rein” against non-physiologic
joint motions that could damage articular structures. Tendons with muscles move or
stabilize articular joints. T/Ls must therefore exhibit nonlinear mechanical behavior
that is relatively compliant under low loads and quickly stiffens under higher loads.
This passive elastic behavior is accompanied by a time-dependent (or viscous)
behavior (e.g., if we stretch and “warm up” before exercise our joints move with
less resistance).
The mechanical behavior of a T/L provides a clear understanding of its normal
function as well as injury mechanisms. The strength of a T/L under a uniformly dis-
tributed load is largely determined by the cross-sectional size of the tissue and the
rate of loading. For instance, T/Ls exhibit increased strength and stiffness with
increased rate of loading. Structural properties of T/Ls are typically gathered via
tensile tests at a constant rate of deformation and displayed on a force-elongation
curve (Fig. B3.3a). From the force-elongation data, a nominal stress-strain relation-
ship is determined to provide mean mechanical properties of the tissue (i.e., normal-
ized structural properties; Fig. B3.3b). Here nominal stress (σ) is the force (F) divided
by the original cross-sectional area of the ligament (A0) and nominal strain (ε) is the

Fig. B3.3 Example of force-elongation (a) and stress-strain (b) curves of the ligament tested to
failure. (b) Stress-strain curve of a ligament contains the three primary regions (in red), including
the toe, linear, and failure regions. The toe region indicates a nonlinear increase in load as the tissue
elongates. The linear region indicates the area of greatest stiffness whereas the failure region indi-
cates that the collagen fibers have disrupted and failed. F = force; A0 = ligament cross-sectional
area; Li = instantaneous length after loading; L0 = original length
58 C.S. Chamberlain and R. Vanderby Jr.

stretched ligament length (Li) minus its original length (Lo) and divided by its original
length (Lo). Often this result is multiplied by 100 to be expressed as a percent. The
calculated stress-strain curve further identifies the commonly reported elastic modu-
lus (slope of the linear region), ultimate tensile strength (σu), and strain (εu).
The resulting stress-strain or force-elongation curves separate the tissue behavior
into three regions, called the toe, linear, and failure regions (Fig. B3.3b). Within the
toe region, collagen fibers of unstressed T/L are composed in a sinusoidal (“crimp”)
pattern. The mechanism causing collagen fiber crimp is not well understood but the
crimped fibers provide minimal resistance to movement over an initial range of
strains. The crimp region is very compliant. As the T/L becomes loaded, the fibers
straighten and the crimp pattern is lost [3, 4]. As loading continues, T/L stiffness
increases and greater force is required to produce an equivalent increment of strain.
Once the strain increases to 1.5–4 % [5], a nearly linear stiffness is observed (i.e.,
the linear region). As strain is further increased beyond the linear range (8–10 %
strain), fiber disruption and ultimately complete and irreversible failure occur [4, 6].
Irreversible damage begins approximately half way through the linear region, after
which the original length (Lo) can no longer be recovered with unloading, and fibril-
lar damage is observed with electron microscopy [7]. Once entering the linear
region, ultimate failure occurs after a predictable increment of strain (~9.3 %) [8].
Viscoelastic behavior of T/Ls aids and protects joint function. Viscoelasticity
adds time dependence to the above elastic phenomena. It manifests itself by allow-
ing easier joint motion after stretching, by increasing the T/L strength to resist
impact forces, and by absorbing some of the energy in dynamic loadings [5]. These
behaviors can be quantified via creep and relaxation testing. If a T/L is held at a
constant deformation (or strain), it exhibits relaxation which means that the load (or
stress) decreases over time (Fig. B3.4a). T/L also exhibits creep where deformation

Fig. B3.4 Viscoelastic properties of ligaments are demonstrated by load relaxation (a) and creep
(b). Load relaxation is demonstrated by loading the ligament (below the linear region) and main-
taining constant strain over an extended period. Load decreases rapidly the first 6–8 h, and then
continues at a low rate (a). Creep takes place when the ligament is loaded (below the linear region)
at a constant level over time. As a result, deformation increases quickly at first and then continues
at a low rate
B3 Ligament and Tendon 59

(or strain) increases over time under constant tensile force (Fig. B3.4b). As a T/L is
subjected to increased rate of loading, the linear portion of the stress-strain curve
becomes steeper, indicating greater stiffness and a greater tissue load before reach-
ing the failure strain.
In vitro ligament and tendon testing is often conducted as a composite of bone-
ligament-bone and muscle-tendon-bone, respectively, because bone blocks are easy
to grip and this minimizes gripping artifacts in the tissue. However, the maximum
load, stiffness, and elongation values can then represent the average properties of a
composite and may vary depending on the material behavior of the bone and enthe-
sis, as well as the size, shape, and orientation of the sample [9]. Mechanical proper-
ties of the human ligaments and tendons from various joints are included in
Table B3.1.

Table B3.1 Mechanical properties of the lower and upper limb ligaments and tendons
Tissue Modulus (MPa) UTS (MPa) Strain at UTS (%) Source
Knee ligament
Anterior cruciate 65–541 13–46 9–44 [10, 11]
Posterior cruciate 109-248 24–36 10–29 [12, 13]
Knee tendon
Patellar 143–660 24–69 14–27 [10, 14–17]
Ankle ligament
Lateral collateral 216–512 24–46 13–17 [18]
Medial collateral 54–321 16–34 10–33 [18, 19]
Ankle tendon
Achilles 65 24–61 24–59 [20]
Palmaris longus 2310 ± 620 91 ± 15 NA [19]
Other
Semintendinous 362 ± 22 89 ± 5 52 ± 3 [10]
Gracilis 613 ± 41 112 ± 4 34 ± 2 [10]
Shoulder ligament
Inferior glenohumoral 30–42 5–6 8–15 [21, 22]
Capsule 32–67 8–21 NA [22]
Spine ligament
Posterior longitudinal NA 21–28 11–44 [23]
Ligamentum flavum NA 1–15 21–102 [23]
Anterior longitudinal 286–724 8–37 10–57 [23, 24]
Supraspinal NA 9–16 39–115 [23]
Interspinal NA 2–9 39–120 [23]
Forearm ligament
Carpal joint 23–119 NA NA [25]
Palmar radioulnar 39 ± 18 5.7 ± 1.7 51 ± 24 [26]
Dorsal radioulnar 52 ± 33 8±5 61 ± 29 [26]
Tissues separated into bundles or tested in multiple studies are listed as a range of values. Data
determined by one bundle or one study are expressed at mean ± S.E.M.
60 C.S. Chamberlain and R. Vanderby Jr.

B3.4 Injured T/L Function

T/L injury results from sudden, prolonged, or excessive forces that exceed the
elastic or recoverable limits of the tissue. When subfailure damage occurs, the tis-
sue does not recover its original length (Lo) after removal of the deforming load.
Ligaments incapable of returning to their original length after elongation are con-
sidered lax and no longer provide the same passive restraints to joint motion and
the same proprioception. If tendon length changes, optimal muscle strength occurs
at different joint positions. Once the ultimate strain is exceeded, the collagen fibers
reach a point of failure and all will rupture. The functional recovery of T/Ls after
injury is a slow and incomplete process. After 2 years of healing, MCLs only
reached 80 % of their control values for strength [8]. During the remodeling phase
of healing, viscoelastic properties are significantly compromised, and scars tend to
stress-relax and maintain load less efficiently than normal ligament [27–29].
Remodeling ligament scars also creep twice the amount as normal ligaments dur-
ing cyclic and static loads [30]. This would increase joint laxity and decrease pro-
prioception after healing. Long-term biomechanical recovery of ligament is
dependent on a number of factors, including the size of the initial gap, proximity
of the torn ligament ends, and joint movement during healing. Tendon’s functional
recovery is also dependent on length, but the bigger concern is scarring down dur-
ing healing to restrict joint motion. Tendon movement is necessary early to prevent
adhesions but load must be very low because of mechanical compromise. Many
strategies have attempted to recapitulate original properties after healing, including
surgical repair, physical therapy protocols, grafting, gene therapy, and tissue engi-
neering, but thus far, healing does not regenerate the native tissue. This issue con-
tinues as a focus for future investigations.

B3.5 Conclusions

Ligament and tendon injuries are a frequent occurrence with significant morbidity.
Appropriate clinical management requires an understanding of damaged T/L
mechanical function and its capacity for self-repair. The composition and hierar-
chical organization of these tissues facilitate its important function. One-
dimensional, elastic properties such as the ultimate tensile strength, modulus, and
strain are commonly gathered and pertinent for comparing normal and compro-
mised tissue and evaluating the efficacy of different healing treatments. In addition,
viscoelastic properties provide metrics to further characterize tissue behavior [31].
These are typically one-dimensional, mean behaviors of normal or healing tissues.
Defects, scaring, and tissue flaws are really three-dimensional, localized phenom-
ena. Alternative methods are needed that locally characterize these regions of com-
promise both in vitro and in vivo. Recently, ultrasound-based methods (example
[32]) have been explored as a possible approach to measure local tendon stress and
B3 Ligament and Tendon 61

strain in a dynamic, time-dependent manner. Such data would further enhance our
knowledge of T/L behavior and provide a tool to evaluate new treatment
strategies.

References

1. Frank CB (2004) Ligament structure, physiology and function. J Musculoskelet Neuronal

Interact 4(2):199–201
2. Nordin M, Lorenz T, Campello M (2001) Biomechanics of tendons and ligaments. In: Nordin
M, Frankel VH (eds) Basic biomechanics of the musculoskeletal system. Lippincott Williams
& Wilkins, Philadelphia, PA, pp 102–125
3. Hirsch G (1974) Tensile properties during tendon healing. A comparative study of intact and
sutured rabbit peroneus brevis tendons. Acta Orthop Scand Suppl 153:1–145
4. Woo SLY, An KN, Arnoczky DVM, Fithian D, Myers B (1994) Anatomy, biology, and biome-
chanics, of the tendon, ligament, and meniscus. In: Simon SR (ed) Orthopaedic basic science.
AAOS, Rosemont, IL, p 52
5. Viidik A (1973) Functional properties of collagenous tissues. Int Rev Connect Tissue Res
6:127–215
6. Sharma P, Maffulli N (2006) Biology of tendon injury: healing, modeling and remodeling.
J Musculoskelet Neuronal Interact 6(2):181–190
7. Provenzano PP, Alejandro-Osorio AL, Valhmu WB et al (2005) Intrinsic fibroblast-mediated
remodeling of damaged collagenous matrices in vivo. Matrix Biol 23(8):543–555
8. LaCroix AS, Duenwald-Kuehl SE, Lakes RS et al (2013) Relationship between tendon stiff-
ness and failure: a metaanalysis. J Appl Physiol 115(1):43–51
9. Woo SLY, Hollis JM, Adams DJ et al (1991) Tensile properties of the human femur-anterior
cruciate ligament-tibia complex—the effects of specimen age and orientation. Am J Sports
Med 19(3):217–225
10. Butler DL, Grood ES, Noyes FR et al (1984) Effects of structure and strain-measurement
technique on the material properties of young human tendons and fascia. J Biomech
17(8):579–596
11. Noyes FR, Grood ES (1976) The strength of the anterior cruciate ligament in humans and
Rhesus monkeys. J Bone Joint Surg Am 58(8):1074–1082
12. Race A, Amis AA (1994) The mechanical properties of the two bundles of the human posterior
cruciate ligament. J Biomech 27(1):13–24
13. Prietto MP, Bain JR, Stonebrook SN et al (1988) Tensile strength of the human posterior cruci-
ate ligament (PCL). Transactions of the Orthopaedic Research Society 13:195
14. Bechtold JE, Eastlund DT, Butts MK et al (1994) The effects of freeze-drying and ethylene
oxide sterilization on the mechanical properties of human patellar tendon. Am J Sports Med
22(4):562–566
15. Blevins FT, Hecker AT, Bigler GT et al (1994) The effects of donor age and strain rate on the
biomechanical properties of bone-patellar tendon-bone allografts. Am J Sports Med
22(3):328–333
16. Butler DL, Kay MD, Stouffer DC (1986) Comparison of material properties in fascicle-bone
units from human patellar tendon and knee ligaments. J Biomech 19(6):425–432
17. Johnson GA, Tramaglini DM, Levine RE et al (1994) Tensile and viscoelastic properties of
human patellar tendon. J Orthop Res 12(6):796–803
18. Siegler S, Block J, Schneck CD (1988) The mechanical characteristics of the collateral liga-
ments of the human ankle joint. Foot Ankle 8(5):234–242
19. Regan WD, Korinek SL, Morrey BF et al (1991) Biomechanical study of ligaments around the
elbow joint. Clin Orthop Relat Res 271:170–179
62 C.S. Chamberlain and R. Vanderby Jr.

20. Paulos LE, Cawley PW, France EP (1991) Impact biomechanics of lateral knee bracing. The
anterior cruciate ligament. Am J Sports Med 19(4):337–342
21. Bigliani LU, Pollock RG, Soslowsky LJ et al (1992) Tensile properties of the inferior
glenohumeral ligament. J Orthop Res 10(2):187–197
22. Itoi E, Grabowski JJ, Morrey BF et al (1993) Capsular properties of the shoulder. Tohoku
J Exp Med 171(3):203–210
23. Pintar FA, Yoganandan N, Myers T et al (1992) Biomechanical properties of human lumbar
spine ligaments. J Biomech 25(11):1351–1356
24. Neumann P, Keller TS, Ekstrom L et al (1992) Mechanical properties of the human lumbar
anterior longitudinal ligament. J Biomech 25(10):1185–1194
25. Savelberg HH, Kooloos JG, Huiskes R et al (1992) Stiffness of the ligaments of the human
wrist joint. J Biomech 25(4):369–376
26. Schuind F, An KN, Berglund L et al (1991) The distal radioulnar ligaments: a biomechanical
study. J Hand Surg Am 16(6):1106–1114
27. Woo SLY, Inoue M, Mcgurkburleson E et al (1987) Treatment of the medial collateral ligament
injury. 2. Structure and function of canine knees in response to differing treatment regimens.
Am J Sports Med 15(1):22–29
28. Clayton ML, Miles JS, Abdulla M (1968) Experimental investigations of ligamentous healing.
Clin Orthop Relat Res 61:146–153
29. Chimich D, Frank C, Shrive N et al (1991) The effects of initial end contact on medial collat-
eral ligament healing—a morphological and biomechanical study in a rabbit model. J Orthop
Res 9(1):37–47
30. Thornton GM, Leask GP, Shrive NG et al (2000) Early medial collateral ligament scars have
inferior creep behaviour. J Orthop Res 18(2):238–246
31. Duenwald SE, Vanderby R Jr, Lakes RS (2009) Viscoelastic relaxation and recovery of tendon.
Ann Biomed Eng 37(6):1131–1140
32. Duenwald S, Kobayashi H, Frisch K et al (2011) Ultrasound echo is related to stress and strain
in tendon. J Biomech 44(3):424–429
Chapter B4
Skin and Muscle

A.F.T. Mak and M. Zhang

B4.1 Introduction

Early studies [11] of the material properties of human skin and muscle are largely
suspect due to problems of inappropriate tissue handling, preservation and speci-
men preparation. Recent efforts have focused on methods which can determine
properties in situ in living individuals or on very freshly excised tissues. Among the
in vivo testing methodologies, indentation has proven to be the most popular,
although it sums up the contributions of various tissue layers [1, 3, 4, 6, 7, 9]. The
load-displacement curve obtained during indentation depends in decreasing degree
upon each of the tissues beneath the indentor. The derived properties, in addition,
can be expected to vary with anatomical site, subject age and external environmen-
tal conditions (temperature, relative humidity, etc.). Additional results have been
obtained in vivo through the use of Doppler ultrasound techniques [2, 5].

B4.2 In-Vivo Mechanical Properties

B4.2.1 Doppler Results

Krouskop et al. [5, 8] applied Doppler ultrasound techniques to measure the point-
to-point biomechanical property of the human skin and subcutaneous musculatures.
Tests of the forearms and legs suggested that the elastic moduli are strongly
dependent on the contraction status of the muscles. A 16-fold increase (from 6.2 kPa

Data are provided from indentation and ultrasound measurement techniques only.
A.F.T. Mak (*) • M. Zhang
Rehabilitation Engineering Centre, Hong Kong Polytechnic, Hunghom, Kowloon, Hong Kong

© Springer Science+Business Media New York 2016 63

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_7
64 A.F.T. Mak and M. Zhang

Table B4.1 Apparent Elastic Anatomical Location Elastic Modulus (kPa)(std dev.)
Moduli of Relaxed Above
Anterior 57.9 (31.1)
Knee Tissues [5]*
Lateral 53.2 (30.5)
Posterior 141.4 (79.1)
Medial 72.3 (45.5)
Superficial 117.6 (63.0)
Underlying 59.1 (74.0)
* Determined by Doppler ultrasonic techniques.

to 109 kPa) in the modulus was reported at a 10% strain as muscle contraction
changed from minimal to maximum. Malinauskas et al. [5] used the same technique
to examine the stump tissues of above-knee amputees and found that the average
modulus was significantly higher in posterior tissues than in other locations (Table
B4.1). They found, additionally, that superficial tissues were stiffer than deeper
structures. Note that the ages and sex of the subjects of these two studies [2, 5] were
not reported.

B4.2.2 Indentation results

Ziegert and Lewis [9] measured the in vivo indentation properties of the soft tissues
covering the anterior-medial tibiae. A preload of 22.4 N was used with indentors of
6 to 25 mm in diameter. The observed load displacement relationship were essen-
tially linearly elastic. The structural stiffness was noted to vary by up to 70%
between sites in one individual and up to 300% between individuals. Unfortunately,
the thicknesses of overlying tissues were not determined at the different sites for the
individuals studied.
Lanir et al. [3] measured the in vivo indentation behavior of human forehead skin
with pressures up to 5 kPa. The observed behavior was linearly elastic and calcu-
lated stiffnesses were 4 to 12 kPa.
Bader and Bowker [1] studied the in vivo indentation properties of soft tissues on
the anterior aspects of human forearms and thighs by applying constant pressures of
11.7 and 7 kPa respectively. Tissue thickness was measured by using a skinfold cali-
per and Poisson’s rato was assumed to be 0.3. With these data, the stiffness of fore-
arm and thigh tissue were calculated to be, respectively, 1.99 and 1.51 kPa.
Vannah and Chlidress [7] applied similar techniques to measure the human calf,
but confined the limb within a shell. They noted that stress relaxation occurred
within one second of load application and no preconditioning effect was noted.
Torres-Morenos et al. [6] performed a similar study, working through ports in quad-
rilateral sockets of three above-knee amputees. However, they found the mechanical
properties of the tissues to be significantly non-linear, with site and rate dependen-
cies, as well as being strongly influenced by muscular activity.
Mak et al. [4] (Table B4.2) measured the in-vivo indentation properties of
the below knee tissues of young adults (N=6) between the ages of 25 and 35.
B4 Skin and Muscle 65

Table B4.2 Initial and Relaxed Elastic Moduli of Tissues Around Proximal Human Tibiae [4]*
Initial Elastic Modulus Relaxed Elastic Modulus
Anatomical Location (Ein) (kPa)(std dev.) (Eeq) (kPa)(std dev.)
Medial
Relaxed 102.6 (8.6) 99.8 (9.2)
Contracted 147.3** (15.8) 142.9** (16.7)
Con./Relax. 1.44 1.43
Lateral†
Relaxed 132.9 (7.2) 130.1 (7.9)
Contracted 194.3** (24,7) 188.4** (23.0)
Con./Relax. 1.46 1.45
* By indentation; ** Different from relaxed (p <0.001); † Between tibia and fibula.

A 4 mm diameter indentor was used, with a final indentation of about 5 mm.

The fixed indentation was then maintained for 2–3 seconds to observe the difference
between initial and relaxed (equilibrium) properties. The tests were done with the
knee in 20° of flexion and were repeated with and without muscular contraction.
The Poisson’s ratio was assumed to be 0.5 for initial measurements and 0.45 for
relaxed (equilibrium) measurements.

Additional Reading

Bader, D.L. and Bowker, P. (1983) Mechanical characteristics of skin and underly-
ing tissues in vivo. Biomaterials, 4, 305–8
This paper describes an indentation experiment to investigate in vivo the bulk
mechanical properties of the composite of skin and underlying tissues on the ante-
rior aspects of human forearms and thighs by applying constant pressures. Significant
variations in tissue stiffness with sex, age and body site were also demonstrated.
Malinauskas, M., Krouskop, T.A. and Barry, P.A. (1989) Noninvasive measure-
ment of the stiffness of tissue in the above-knee amputation limb. J. Rehab. Res.
Dev., 26(3), 45–52
The paper reports a noninvasive technique to measure the mechanical properties
of the bulk soft tissues by a pulsed ultrasonic Doppler system. An ultrasonic trans-
ducer was used to measure internal displacement resulting from external acoustical
perturbations. Measurements were made at four sites of 8 aboveknee residual limbs.
The Young’s moduli were found in a range of 53–141 kPa. Superficial tissue had a
significantly higher modulus than the tissue beneath.
The repeatability test indicated an acceptable repeatibility. An improved device
can possibly be a useful tool in prosthetic fitting and CAD socket design.
Rab, G.T. (1994) Muscle, in Human Walking (2nd ed.) (eds J. Rose, J.C. Gamble),
Williams & Wilkins, Baltimore, pp. 101–122.
66 A.F.T. Mak and M. Zhang

A concise description of the active properties of muscle tissue, with direct application
to the development of forces within the human gait cycle.
Reynolds, D. and Lord, L. (1992) Interface load for computer-aided design of
below-knee prosthetic sockets. Med. & Biol. Eng. & Comput., 30, 419–426.
The authors investigated the bulk tissue behaviour of the below-knee amputee’s
residual limb. An assessment of Young’s modulus was made by matching the inden-
tation experimental curves with the curves produced by the finite element modelling
of the indentation into a layer of tissue with idealized mechanical properties. In vivo
tests, conducted at four sites of a below-knee amputee’s limb (patella tendon, pop-
liteal, and anterolateral regions) found the local moduli to be 145, 50, 50 and 120
kPa respectively. The effect of muscle tension on the measured indentation response
was also investigated. The results showed that the stiffness increased with muscle
contraction.

References

1. Bader D.L. Bowker, P. (1983) Mechanical characteristics of skin and underlying tissues in vivo.
Biomaterials 4:305–8
2. Krouskop, T.A, Dougherty. D.R and Vonson, F.S. (1987) A pulsed Doppler ultrasonic system
for making noninvasive measurements of the mechanical properties of soft tissue. I. Rehab.
Res. Dev., 24(1), 1–8.
3. Lanir, Y., Dikstein, S., Hartzshtark, A., et al. (1990) In-vivo indentation of human skin. Trans.
ASME (J. Biomech. Eng.), 112, 63–69.
4. Mak, A.F.T., Liu, G.H.W. and Lee, S.Y. (1994) Biomechanical assessment of below-knee
residual limb tissue, J. Rehab. Res. Dev., 31, 188–198.
5. Malinauskas, M., Krouskop, T.A and Barry, P.A (4) (1989) Noninvasive measurement of the
stiffness of tissue in the above-knee amputation limb. I.Rehab. Res. Dev., 26(3), 45–52.
6. Torres-Morenos, R., Solomonidis, S.E. and Jones, D. (1992) Geometric and mechanical charac-
teristics of the above-knee residual limb. Proc. 7th. World Cong. Int. Soc. Prosthetics Orthotics,
pp. 149.
7. Vannah, W.M. and Chlidress, D.S. (1988) An investigation of the three dimensional mechani-
cal response of bulk muscular tissue: Experimental methods and results, in Computational
Methods in Bioengineering (eds R.L. Spilker and B.R Simon), Amer. Soc. Mech. Eng.,
New York, pp. 493–503
8. Yamada, H. (1970) (ed. F.G. Evans) Strength of Biological Tissues, Williams & Wilkins,
Baltimore.
9. Ziegert, J.C. and Lewis, J.L. (1978) In-vivo mechanical properties of soft tissue covering bony
prominence, Trans ASME (J. Biomech. Eng.), 100, 194–201.
Chapter B5
Brain Tissues

S.S. Margulies and D.F. Meaney

B5.1  Introduction

The brain is organized into the cerebrum, brain stem, and cerebellum. The cerebrum
consists of two cerebral hemispheres, basal ganglia, and the diencephalon. The
hemispheres contain the cerebral cortex and underlying white matter, and are asso-
ciated with higher order functioning, including memory, cognition, and fine motor
control. The basal ganglia, contained within the hemispheres, controls gross motor
function. The diencephalon is much smaller than the cerebrum, contains the thala-
mus and hypothalamus, and is associated with relaying sensory information and
controlling the autonomic nervous system. The brainstem contains the mesence­
phalon, pons and the medulla oblangata. The smallest segment of the brain, the
mesencephalon, is located below the diencephalon and is thought to play a role in
consciousness. Muscle activation, tone and equilibrium is controlled in the pons and
cerebellum located below the mesencephalon, and respiratory and cardiac processes
are governed by the medulla oblongata, located directly beneath the pons.
The brain contains grey matter and white matter substances that are easily distin-
guished upon gross examination. Grey matter contains a densely packed network of
neural cell bodies and associated glial cells, whereas white matter contains myelin-
ated axonal tracts, relatively few neuronal cell bodies, and a supporting environment
of glial cells. The entire brain is surrounded by cerebrospinal fluid contained within
an extensive ventricular system that occupies approximately one-tenth of the total
brain volume. The ventricular system supports the brain as well as the spinal cord, and
provides nutrients to and removes waste products from the central nervous system.

S.S. Margulies (*)

Department of Bioengineering, University of Pennsylvania,
105D Hayden Hall, Philadelphia, PA 19104-6392, USA
D.F. Meaney
Department of Bioengineering, 105E Hayden Hall, University of Pennsylvania,
Philadelphia, PA 19104-6392, USA

© Springer Science+Business Media New York 2016 67

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_8
68 S.S. Margulies and D.F. Meaney

Understanding the response of the complex brain structure to thermal, electrical,

or mechanical stimuli necessitates a thorough investigation of the physical proper-
ties of each brain component. This task is still in its infancy, and therefore the mate-
rial cited in this chapter is accompanied by several caveats. First, most properties
cited are for whole brain. However, where available, properties of the white matter,
gray matter, cerebrum, cerebellum, and brainstem are noted. Second, extensive
studies have been conducted on brain tissue from a broad range of species. To pro-
vide the reader with information most similar to human tissue, only primate data
(nonhuman and human) have been included. Exceptions occur only when there is no
primate data available, and are noted. Finally, because the brain is a highly perfused
organ and its properties may differ between in vivo and in vitro conditions, and
with the post mortem time period before testing. To facilitate comparison between
tests, information regarding in vivo/in vitro conditions, post mortem time, and test
procedure is included with the data.

B5.2  Composition
Table B5.1  Brain Tissue Composition
Water (wt%) Ash (wt%) Lipid (wt%) Protein (wt%)
Whole brain 76.3–78.5 (77.4) 1.4–2 (1.5) 9–17  8–12
Grey matter 83–86 1.5 5.3  8–12
White matter 68–77 1.4 18 11–12
Approximate overall ranges given. Values in parentheses indicate averages.
Source: [1–3].

B5.2.1  Mass [2]

Adult male whole brain (20–30 yr): 1400 g

(cerebrum 1200 g, cerebellum, 150 g)
Adult female whole brain (20–30 yr): 1200 g
Adult male whole brain (90 yr): 1161 g

B5.2.2  Dimensions and shape [2]

The adult brain with the brain stem is approximately half an ellipsoid.

Diameter: Vertical Transverse Anteroposterior

Male: 13 cm 14 cm 16.5 cm
Female: 12.5 cm 13 cm 15.5 cm
B5  Brain Tissues 69

B5.2.3  Density (adult) in kg/m3

Brain 1030–1041 [2]

Grey matter 1039 [1, 3]
White matter 1043 [1, 3]

B5.3  Mechanical Properties

B5.3.1  Bulk modulus

Excised brain samples: 2.1 × 106 kPa (independent of frequency) [4].

B5.3.2  Poisson’s ratio

Using the relationship v = (3K – E)/6K

where K is the bulk modulus and E is the elastic modulus, and considering that
K is 4–5 orders of magnitude larger than E, brain is approximated as an incompress-
ible material (v = 0.5) [5].

B5.3.3  Elastic and shear moduli

See Table B5.2.

B5.3.4  Creep modulus

See Table B5.3.

B5.4  Electrical Properties

(no primate data available)

B5.4.1  Electrical conductivity temperature coefficient [14]

Cow and pig whole brain ( ∆σ/σ)/∆T = 3.2 °C-1.

 70

Table B5.2  Elastic Modulus and Shear Modulus of Normal Brain at 37°C (typical values, not averages)
Species
Testing technique (post mortuum time) Frequency (Hz) Analysis format Results (kPa) G1 (kPa) G2 (kPa) Source
Free vibration Rhesus monkey 31 E* = E1 + i E2 E1 = 91.2 30.3 18.0 [6]
(15 min) E2 + 53.9
Free vibration Human (6–12 hr) 34 E* = E1 + i E2 E1 = 66.8 22.3 8.7 [6]
E2 = 26.3
Harmonic shear Human white matter 9–10 G* = G1+ i G2 Min: G* = 0.75 + i 0.3 0.75–1.41 0.3–0.6 [7]
(10–62 hr) Max G* = 1.41 + i 0.6
Driving point Rhesus (in vivo) 80 Theoretical G1 = 19.6 19.6 11.2 [7, 8]
impedance approx of data G2 = 11.2
G* = G1 + i G2
Quasi-static Rhesus monkey live 0 Static elastic E1 = 10–60 live and dead 10–60 [9]
expansion (in vivo) dead Modulus, E1 E1 = 40–120 fixed
of balloon (5–45 min) fixed
within tissue (formaldehyde)
Sudden Human (acceleration compare tissue Shear modulus G = l.l7–2.19 1.7 [10]
acceleration duration ≈ displacement kPa (average = 1.7) Kinematic (Tissue
of a cylinder 17.5 ms.) with that of a viscosity ν = 14–124  cm2/s temp
filled with tissue Voigt solid (average = 89) unknown)
cylinder
S.S. Margulies and D.F. Meaney
Harmonic shear Human (<3 hr) 5–350 G* = G1+ i G2 G1 G2 7.6–33.9 2.76–81.4 [11]
5  7.6  2.8
15  8.4  3.5
35 11.7  5.2
85 19.3 13.4
105 21.4 18.0
B5  Brain Tissues

225 29.0 45.9

350 33.9 81.4
Harmonic shear Human (<3 hr) G* = G1+ i G2 G1 G2 [11]
Grey matter 2–10
Axis 1 10.6 1.5
Axis 2  6.3 1.5
Axis 3  4.1 1.4
average  7.0 1.5 7.0 1.5
white matter 2–10
Axis 1  7.7 2.6
Axis 2  7.0 3.2
Axis 3  7.3 3.5
average  7.3 3.1 7.3 3.1
* Assumed tissue is incompressible (v = O.5), therefore G* = E*/3.
71
 72

Table B5.3  Creep Modulus of Normal Brain at 37°C (typical values, not averages)
Testing
technique* Species Model Results Source
Compression Rhesus monkey J(t) = C1 + C2 In(t) C1 = 2.97 kPa-1 [6]
C -1
(15 min) 2 = 0.18 kPa
t >0.1s
Compression Human (6–12 hr) J(t) = C1 – C2 In(t) C1 = 2.45 kPa-1 [6]
C2 = 0.18 kPa-1
t >0.1s
Compression Human (6–12 hr) Nonlinear solid K1 = 25.77 kPa [12]
1/ 2 K2 = 20.46 kPa,
æ k22 + 4s 0 k3 ö
s e- m t ç s 0 ( ) k K3 = 104.04 kPa
ϵ( t ) = 0 + + - 2 ÷ . 1 - e- m t
( )
k2 çç k1 2 k3 2 k3 ÷÷ C1 = 651.8 kPa s
è ø σo = 3.44, 4.82, 6.89 kPa
K2
Where m =
C1
Compression Human (6–12 hr) Nonlinear fluid Kl = 74.41 kPa [12]
1/ 2
s 0 C1C2 C1 ( C22 + 4s 0C3 ) 1 é 2 1/ 2
K2 = 20.67 kPa
Î (t ) = + - + (C2 + 4s 0C3 ) - C2 ùúû C1 = 1266.38 kPa
K 2 2 K1C3 2 K1C3 2C3 êë C2 = 36599.7 kPa s
æ C1 - mt ö s 0 é C1 ù - mt
C3 = 1.38 kPa S2
ç t + e ÷ + . ê1 + ú (1 - e ) σo = 3.44, 4.82, 6.89 kPa
è K1 ø K1 ë C2 û
S.S. Margulies and D.F. Meaney
Compression Human (6–12 hr) Hyperelastic with material dissipation C1 = 6.89 kPa [12]
. C2 = 17.23 kPa
s 11 é 2 1 ù é C2 ù B1 l
= l - ú ê1 + ú+ . B1 = 0.55 kPa S2
2C1 êë l û ë C1l û C1 l 3 λ = stretch ratio
[l 2 - 6l 3 + 1] κ = stretch rate
B5  Brain Tissues

Fluid infusion Cat Poroelastic Grey matter [13]

¶e G = 2 kPa
= k ( 2G + l ) Ñ 2 e λ = 90 kPa
¶t
G = shear modulus κ = 7.5 × 10-9 kPa-m2/s
λ = Lame constant White matter
K = permeability G = 0.9 kPa
λ = 40 kPa
κ = 5 × 10-9 kPa-m2/s
* All tests consisted of a load applied rapidly and then held constant.
73
74 S.S. Margulies and D.F. Meaney

Table B5.4  Electrical conductivity of brain tissues

Temp (°C) Conductivity (s) (mS/cm) Species Source
Whole brain 37 1.7 cow, pig [14]
Cerebrum 39 1.38–1.92 rabbit [15]
Cerebellum 39 1.17–1.64 rabbit [15]
Cortex body 3.1 rabbit [16]
Cortex 37 4.5 cat [17]
White matter body 1.0 rabbit [18]
White matter 37 2.9 cat [17]

B5.5  Thermal Properites

Table B5.5  Thermal properties of normal, unperfused human brain tissue

Conductivity, Diffusivity, Specific heat,
Temp (°C) (W/m°K) (cm2/s × 103) (J/g °K) Source
Whole brain 5–20 0.528 1.38 ± 0.11 – [19]
Whole brain 37 0.503–0.576 – – [20]
Cortex 37 0.515 1.47 – [21]
5–20 0.565 1.43 ± 0.09 3.68 [19]
White matter 5–20 0.565 1.34 ± 0.10 3.60 [19]
Values A–B indicate approximate range, values A ± B indicate mean ± standard deviation.

B5.6  Diffusion Properties

5.6.1  Sucrose (feline brain) [22]

D ~ 3.2–3.8 × 10-6 cm2/sec matter

~1.3–1.9 × 10-6 cm2/sec white matter

B5.6.2  Small ions [23]

D ~ 8.6 × 10-6 cm2/sec

5.6.3  Large molecules (>150 angstrom): [23]

Hindered diffusion occurs

B5  Brain Tissues 75

B5.7  Comments

The list of primate brain tissue properties presented in this chapter highlights the
numerous areas where data are either unavailable or largely incomplete. With a
renewal of interest in modeling the normal and pathological response of the brain to
various stimuli, it is possible that some of the missing tissue property data will be
generated in the near future. Investigators should be cautioned, however, that
because brain tissue properties vary with environmental factors, measurements
made under unphysiologic conditions may differ from those of living tissue. As an
example, the brain is highly vascularized, and the role of blood flow, volume, and
pressure on tissue behavior remains to be determined.
The caution exercised by the experimentalist in generating new data should be
matched by sound skepticism on the part of the investigators who are developing ana-
lytical or computational models of brain functional and structural response. Although
the ability to calculate detailed responses has improved greatly in the past decade, these
sophisticated models are limited by the available experimental data used to develop
and validate the models. To create a realistic representation of normal or pathological
response of the brain, it is essential that the model parameters be based on measured
tissue properties and that any conclusions drawn from the models be validated with
measured response data.Therefore, it is clear that future experimental studies are
needed to determine the properties and response of living primate brain tissue.

Additional Reading

Cooney, D.O. (1976) Biomedical Engineering Principles: An introduction to fluid,

heat and mass transfer processes, Marcel Dekker, New York.
Provides more detailed examples in bioheat transfer and pharmacokinetics which
may be useful in modeling heat and mass transfer in the brain parenchyma.
Fung, Y.C. (1993) Biomechanics: Mechanical properties of living tissues,
2nd ed., Springer-Verlag, New York.
Fung, Y.C. (1990) Biomechanics: Motion, flow, stress and growth, Springer-­
Verlag, New York.
Fung, Y.C. (1965) Foundations of Solid Mechanics, Prentice Hall, Englewood Cliffs.
These works describe both basic principles of mechanics and their specific
­applications in biomechanics. A review of the constitutive property relationships for
biological tissues included throughout these texts may be particularly helpful
for applying the material-property information listed previously.
Lih, M.L. (1975) Transport Phenomena in Medicine and Biology, John Wiley &
Sons, New York.
A concise review of the principles used in modeling the transport phenomena in
several biological systems including examples of heat and mass transfer.
Nolte, J. (1988) The Human Brain, 2nd. ed., C.V. Mosby, St Louis
Provides a more detailed review of the structure and function of the different
brain regions.
76 S.S. Margulies and D.F. Meaney

References

1. Duck, F.A. (1990) Physical Properties of Tissue, Academic Press, New York.

2. ICRP (1975) Report of the Task Group on Reference Man, ICRP Publication 23, International
Commission on Radiological Protection, Pergamon Press, Oxford, pp. 212–215; 280–281.
3. Woodard, H.Q. and White, D.R. (1986) The composition of body tissues. Brit. J. Radiol., 59,
1209–1219.
4. McElhaney, J.H., Roberts, V.L. and Hilyard, J.F. (1976) Handbook of Human Tolerance,
Japanese Automobile Research Institute, Tokyo, pp. 151.
5. Fung, Y.C. (1993) Biomechanics: Mechanical properties of living tissues, 2nd ed., Springer-­
Verlag, New York.
6. Galford, J.E. and McElhaney, J.H. (1970) A viscoelastic study of scalp, brain, and dura.
J. Biomech., 3, 211–221.
7. Fallenstein, G.T., Hulce, V.D. and Melvin, J.W. (1969) Dynamic mechanical properties of
human brain tissue. J. Biomech., 2, 217–226.
8. Wang, H.C. and Wineman, A.S. (1972) A mathematical model for the determination of visco-
elastic behavior of brain in vivo - I: Oscillatory response. J. Biomech., 5, 431–446.
9. Metz, H., McElhaney, J. and Ommaya, A.K. (1970) A comparison of the elasticity of live,
dead, and fixed brain tissue. J. Biomech., 3, 453–458.
10. Ljung, C. (1975) A model for brain deformation due to rotation of the skull. J. Biomech., 8,
263–274.
11. Shuck, L.Z. and Advani, S.H. (1972) Rheological response of human brain tissue in shear.
J. Basic Eng., Trans ASM E, 94, 905–911.
12. Pamidi, M. and Advani, S. (1978) Nonlinear constitutive relations for human brain tissue.
J. Biomech. Eng., 100, 44–48.
13. Basser, P. (1992) Interstitial pressure, volume and flow during infusion into brain tissue.
Microvascular Res., 44, 143–165.
14. Osswald, K. (1937), (Measurement of the conductivity and dielectric constants of biological
tissues and liquids by microwave) (Ger.) Messung der Leitfahigkeit und Dielektrizitatkonstante
biologischer gewebe und Flussigkeiten bei kurzen Wellen. Hochfrequentz Tech. Elektroakustik,
49, 40–49.
15. Crile, G.W., Hosmer, H.R. and Rowland, A.F. (1922) The electrical conductivity of animal
tissues under normal and pathological conditions. Am. J. Physiol., 60, 59–106.
16. Ranck, J.B. and Be Merit, S.L. (1963) Specific impedance of rabbit cerebral cortex. Exp.
Neurol., 7 144–152.
17. Freygang, W.H. and Landaw, W.M. (1955) Some relations between resistivity and electrical
activity in the cerebral cortex of the cat. J. Cell. Comp. Physiol., 45, 377–392.
18. van Harreveld, A., Murphy, T. and Nobel, K.W. (1963) Specific impedance of rabbit’s cortical
tissue. Am. J. Physiol., 205, 203–207.
19. Cooper, T.E. and Trezek, G.J. (1972) A probe technique for determining thermal conductivity
of tissue. J. Heat Transfer, Trans. ASME, 94, 133–140.
20. Bowman, H.F. (1981) Heat transfer and thermal dosimetry. J. Microwave Power, 16,

121–133.
21. Valvano, J.W., Cochran, J.R. and Diller, K.R. (1985) Thermal conductivity and diffusivity of
biomaterial measured with self-heating thermistors. Int. J. Thermophys., 6, 301–311.
22. Rosenberg, G.A., Kyner, W.T. and Estrada, E. (1980) Bulk flow of brain interstitial fluid under
normal and hypermolar conditions. Am. J. Physiol., 238, f42–f49.
23. Nicholson, C. (1985) Diffusion from an injected volume of substance in brain tissue with
arbitrary volume fraction and tortuosity. Brain Research, 33 325–329.
Chapter B6
Arteries, Veins and Lymphatic Vessels

X. Deng and R. Guidoin

B6.1  Introduction

Blood and lymphatic vessels are soft tissues with densities which exhibit nonlinear
stress-strain relationships [1]. The walls of blood and lymphatic vessels show not
only elastic [2, 3] or pseudoelastic [4] behavior, but also possess distinctive inelas-
tic character [5, 6] as well, including viscosity, creep, stress relaxation and pres-
sure diameter hysteresis. The mechanical properties of these vessels depend largely
on the constituents of their walls, especially the collagen, elastin, and vascular
smooth muscle content. In general, the walls of blood and lymphatic vessels are
anisotropic. Moreover, their properties are affected by age and disease state. This
section presents the data concerning the characteristic dimensions of arterial tree
and venous system; the constituents and mechanical properties of the vessel walls.
Water permeability or hydraulic conductivity of blood vessel walls have been also
included, because this transport property of blood vessel wall is believed to be
important both in nourishing the vessel walls and in affecting development of ath-
erosclerosis [7–9].
The data are collected primarily from human tissue but animal results are also
included in places for completeness. Among the three kinds of vessels, the arte-

X. Deng (*) • R. Guidoin

Laboratorie de Chirugie Exp Agriculture Services,
Room 1701 Services Building Université Laval, Quebéc G1K 7P4, Canada

© Springer Science+Business Media New York 2016 77

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_9
78 X. Deng and R. Guidoin

rial wall has been extensively investigated while studies of lymphatic vessels are
very rare.

B6.2  Morphometry of the Arterial Tree and Venous System

Detailed measurements of the number and size of blood vessels in the living
body are very difficult to perform, so reliable information is scarce. Moreover,
data collected on vessels in one tissue or organ are not applicable to another.
Thus one should be cautious in using morphometric data; only the data for large
vessels are reliable.
The aorta is tapered, but most other arteries can be considered to have a constant
diameter between branches. The rate of taper varies from individual to individual,
presumably, between species. However, in the dog, the change of aortic cross
sectional area can be described by the exponential equation:

A = Ao e( - b x  Ro ) (B6.1)

Table B6.1  Morphometric and Related Properties of the Human Systemic Circulation*
Diameter Wall thickness Length Blood velocity Reynolds
Vessel (mm) (mm) (cm) (cm/sec) number
Ascending aorta 32 1.6 5–5.5 63 3600–5800
Arch of aorta 25–30 4–5
Thoracic aorta 20 1.2 16 27 1200–1500
Abdominal 17–20 0.9 15
aorta
Femoral artery 8 0.5 32
Carotid artery 9 0.75 18
Radial artery 4 0.35 23
Large artery 2–6 20–50 110–850
Capillaries 0.005–0.01 0.05–0.1 0.0007–0.003
Large veins 5–10 15–20 210–570
Vena cava 20 11–16 630–900
* Source [11–17]. Note: The Reynolds numbers were calculated assuming a value for the viscosity
of the blood of 0.035 poise.
B6  Arteries, Veins and Lymphatic Vessels 79

Table B6.2  Morphometries of the Pulmonary Arterial System*

Zone Number of branches Diameter (mm) Length (mm)
Proximal 1.000 30.000 90.50
3.000 14.830 32.00
8.000 8.060 10.90
2.000 x 10 5.820 20.70
6.600 x 10 3.650 17.90
2.030 x 102 2.090 10.50
6.750 x 102 1.330 6.60
2.290 x 103 0.850 4.69
Intermediate 5.861 x 103 0.525 3.16
1.756 x 104 0.351 2.10
5.255 x 104 0.224 1.38
1.574 x 105 0.138 0.91
Distal 4.713 x 105 0.086 0.65
1.411 x 106 0.054 0.44
4.226 x 106 0.034 0.29
1.266 x 107 0.021 0.20
3.000 x 108 0.013 0.13
* Source: Adapted from [18].
Note: The data were obtained from a 32-year-old woman who had been free of respiratory disease
and died of uremia. For the purpose of description, the pulmonary arterial tree was divided into
three zones.

where A is the cross sectional area of the aorta, Ao and Ro are the respective cross
sectional area and radius at the upstream site, x is the distance from the upstream
site, and β a taper factor, which varies between 0.02 and 0.05 [10]. In man, the taper
is found not to be as smooth as implied by the above equation; thus values of β are
unavailable.
Morphometric and related data are given in Tables B6.1, B6.2 and B6.3.

B6.3  Constituents of the Arterial Wall

B6.3.1  Normal arterial wall

The main constituents of normal human arterial tissues from young adult subjects
(20–39 years) are listed in Table B6.4 [20]. The major part of the dry matter in the
arterial wall consists of proteins such as elastin and collagen. Because the impor-
tance of elastin and collagen in the mechanical properties of arterial wall, the com-
position of media and adventitial layers in terms of collagen, elastin, smooth muscle
and ground substance is listed in Table B6.5 for three different arterial tissues [1].
Table B6.3  Morphometric and Related Properties of the Canine Systemic Circulation*†
Ascending Descending Abdominal Femoral Vena cava, Pulmonary
Site aorta aorta aorta artery Carotid artery Arteriole Capillary Venule inferior artery, main
Internal 1.5 1.3 0.9 0.4 0.5 5 x 10-3 6 x 10-4 4 x 10-3 1.0 1.7
diameter, (1.0–2.4) (0.8–1.8) (0.5–1.2) (0.2–0.8) (0.2–0.8) (1–8 x 10-3) (4–8 x 10-4) (1–7.5 x 10-3) (0.6–1.5) (1.0–2.0)
di (cm)
Wall thickness, 0.065 – 0.05 0.04 0.03 2 x 10-3 1 x 10-4 2 X 10-4 0.015 0.02
h (cm) (0.05–0.08) (0.04–0.06) (0.02–0.06) (0.02–0.04) (0.01–0.02) (0.01–0.03)
h/di 0.07 – 0.06 0.07 0.08 0.4 0.17 0.05 0.015 0.01
(0.055–0.084) (0.04–0.09) (0.055–0.11) (0.055–0.095)
In vivo length 5 20 15 10 15 0.15 0.06 0.15 30 3.5
(cm) (10–20) (0.1–0.2) (0.02–0.1) (0.1–0.2) (20–40) (3–4)
Cross-section 2 1.3 0.6 0.2 0.2 2x10-5 3x10-7 2x10-5 0.8 2.3
area (cm2)
Total vascular 2 2 2 3 3 125 600 570 3.0 2.3
cross-section at
each level (cm2)
Blood velocity 0.2 1.05 0.55 1.0 – 0.75 0.07 0.35 0.25 0.7
(peak) (ms-1) (0.4–2.9) (0.25–2.5) (0.5–0.6) (1.0–1.2) (0.15–0.4)
Blood velocity 0.2 0.2 0.15 0.1 – (5–10 x 10-3 ) 2–17 x10-4 2–5 x10-3 – 0.15
(mean) (ms-1) (0.1–0.4) (0.1–0.4) (0.08–0.2) (0.1–0.15) (0.06–0.28)
Peak Reynolds 4500 3400 1250 1000 – 0.09 0.001 0.035 700 3000
number, R̂ e
* Source: Adapted from [19].
† Normal values for canine cardiovascular parameters. An approximate average value, and then the range, is given whe re possible. All values are for the dog except those for
arteriole, capillary, and venule, which have only been measured in smaller mammals.
B6  Arteries, Veins and Lymphatic Vessels 81

Table B6.4  Composition of Normal Arterial Tissue*

Aorta Femoral artery Brachial artery
Components (human) (human) (human)
Organic Dry matter 28.0† 25.3 26.3
Nitrogen 4.1 3.5 –
Total lipids 1.680 – –
Cholesterol 0.290 0.135 0.185
Inorganic Total ash 0.730 0.675 0.670
Calcium 0.070 0.147 0.144
Total PO4 0.375 – –
* Source: Adapted from [20].
† Values are expressed in percentage of wet tissue weight.

Table B6.5  Composition of Human Arteries at In Vivo blood Pressure

Thoracic aorta Plantar artery Pulmonary artery
Media
Smooth muscle 33.5 ± l0.4† 60.5 ± 6.5 46.4 ± 7.7
Ground substance 5.6 ± 6.7 26.4 ± 6.4 17.2 ± 8.6
Elastin 24.3 ± 7.7 1.3 ± 1.1 9.0 ± 3.2
Collagen 36.8 ± 10.2 11.9 ± 8.4 27.4 ± 13.2
Adventitia
Collagen 77.7 ± 14.1 63.9 ± 9.7 63.0 ± 8.5
Ground substance 10.6 ± 10.4 24.7 ± 2.6 25.1 ± 8.3
Fibroblasts 9.4 ± 11.0 11.4 ± 2.6 10.4 ± 6.1
Elastin 2.4 ± 3.2 0 1.5 ± 1.5
* Source: Adapted from [1].
† (Mean ± S.D.)

The collagen in adventitia and media is mostly Type III, some Type I, and a trace of
Type V while the collagen of the basal lamina is Type IV [1].
Table B6.6 lists the constituents of additional arteries (canine), and the the ratio
of collagen to elastin [21].
Composition changes of arterial tissues with age
The composition of normal human arterial tissues is altered with age in many
aspects. Table B6.7 lists the observed changes in human aorta, pulmonary and fem-
oral arteries [20]. There is a tendency that both the dry matter and nitrogen content
of arterial tissues decreases with age. However, the relative quantity of collagen [22]
and elastin [23, 24] in the arterial wall remains almost unchanged with age. Below
the age of 39, the wall of human thoracic aorta has 32.1 ± 5.5% elastin, between the
age 40–69, the wall contains 34.4 ± 9.3%, and from 70–89, the elastin content is
36.5 ± 10.1 [24].
 82

Table B6.6  Arterial Wall Constituents, and Ratio of Collagen to Elastin*

Percentage of wet tissue Percentage of dry defatted tissue C
Extracted Collagen+
Artery n† H 2O fat+H2O Collagen Elastin elastin E
Coronary 9 63.2 ± 1.0‡ 71.5 ± 1.4 47.9 ± 2.6 15.6 ± 0.7 63.5 ± 2.7 3.12 ± 0.21
Aorta, ascending 9 73.8 ± 0.6 74.0 ± 0.5 19.6 ± 1.2 41.1 ± 2.1 60.7 ± 2.2 0.49 ± 0.04
Carotid 6 71.1 ± 0.1 71.2 ± 0.1 50.7 ± 2.1 20.1 ± 1.0 70.8 ± 2.5 2.55 ± 0.13
Aorta, abdominal 10 70.4 ± 0.4 70.8 ± 0.3 45.5 ± 1.7 30.1 ± 1.7 75.6 ± 1.8 1.58 ± 0.15
Cranial mesenteric, 10 70.8 ± 0.5 71.6 ± 0.4 38.1 ± 1.7 26.5 ± 1.7 64.6 ± 1.8 1.51 ± 0.15
proximal
Cranial mesenteric, 9 71.4 ± 0.4 72.0 ± 0.4 37.4 ± 1.4 22.4 ± 1.5 59.8 ± 1.6 1.72 ± 0.11
distal
Cranial mesenteric, 10 69.5 ± 0.6 73.1 ± 0.7 36.1 ± 1.5 21.8 ± 0.9 57.9 ± 1.7 1.69 ± 0.10
branches
Renal 9 70.4 ± 0.7 70.8 ± 0.7 42.6 ± 1.6 18.7 ± 1.8 61.3 ± 2.1 2.46 ± 0.27
Femoral 10 68.0 ± 0.3 68.1 ± 0.3 44.5 ± 1.4 24.5 ± 1.6 69.0 ± 2.1 1.89 ± 0.14
* Source: [21] by permission.
† Number of specimens.
‡ Specimens slightly dehydrated owing to unavoidably long dissection. Mean ± standard deviation. All percentage values refer to w/w.
X. Deng and R. Guidoin
B6  Arteries, Veins and Lymphatic Vessels 83

Table B6.7  Variation of Normal Human Arterial Tissue Composition with Age*
Aorta
Age Acid
group Dry Nitro- Total Choles- Total Cal- Total soluble Potas
(years) matter gen lipids terol ash cium PO4 PO4 sium
10–19 29.5† 4.38 1.23 0.15 – 0.03 0.25 0.16 0.055
30–39 28.5 4.03 1.75 0.28 0.81 0.09 0.40 0.29 0.040
50–59 28.0 3.67 1.90 0.48 1.55 0.21 0.54 0.41 0.039
70–79 28.0 3.38 – 0.71 2.80 0.39 – – 0.033
Pulmonary artery Femoral artery
Age
group Dry Nitro- Choles- Potas Dry Nitro Choles- Total Cal-
(years) matter gen terol Calcium ium Matter gen terol ash cium
10–19 26.5 3.91 0.12 0.025 0.033 26.9 3.84 0.11 0.59 0.14
30–39 25.7 3.71 0.17 0.028 0.031 24.4 3.30 0.15 0.71 0.18
50–59 24.9 3.45 0.22 0.027 0.026 22.8 2.83 0.23 1.15 0.40
70–79 23.0 3.25 – 0.060 0.025 25.3 2.90 0.55 3.17 1.07
* Source: Adapted from [20]; see for additional age group values.
† Values expressed in percentage of wet tissue weight.

Table B6.8  Effect of Hypertension on Composition of Arteries*

dry wt Elastin (E) Collagen (C) E and C
Group N‡ (mg) Wt(mg) (% total wt) WT(mg) (% total wt) (% total wt)
III C† 2 5.25 1.76 33.5 0.69 13.1 46.6
5.58 2.20 39.0 0.83 14.3 53.3
III H† 2 7.00 2.86 41.6 1.44 20.6 62.2
6.50 2.78 42.8 1.06 16.3 59.1
IIC 2 4.96 2.23 44.9 0.72 14.5 59.4
5.26 2.14 40.6 0.68 12.8 53.4
IIH 2 8.12 2.94 36.2 1.68 20.7 56.9
7.50 2.70 36.0 1.52 20.3 56.3
IC 2 4.36 1.91 43.8 1.12 25.7 69.5
5.56 2.09 37.6 1.24 22.3 59.9
IH 2 8.88 3.08 34.7 2.08 23.4 58.1
10.89 3.50 32.1 2.33 21.4 53.5
P<0.0l# P<0.00l P>0.2 P<0.01 P=0.2 P>0.8
* Source[25] by permission: †: N = number of animals (rats); C = normotensive; H - hypertensive.
#: Normotensive vs. hypertensive
84 X. Deng and R. Guidoin

Table B6.9  Effect of Atherosclerosis on Comparison of Arteries*

Collegen +
H2O (g/kg Collagen Elastin Elastin Collegen
Site wet wt) (% dry wt) (% dry wt) (% dry wt) Elastin
Carotid 740±3† 50.6±0.8 28.0±0.3 78.6±0.8 1.81±0.04
Control (n = 11)
Diet (n = 7) 764±7## 49.8±1.6 33.1±1.3## 84.2±1.9## 1.61±0.07#
Iliac 742±3 46.2±0.6 33.0±0.7 79.2±1.1 1.41±0.03
Control (n = 16)
Diet (n = 5) 724±6## 37.8±2.0 25. l± 1.2## 66.0±2.3## 1.49±0.02#
* Source: (26) by permission; canine subjects.
† means ± SE.
## Significantly different, P < 0.01; # Significantly different, P < 0.05.

Changes in elastin and collagen content due to hypertension

Experimental observation by Wolinsky [25] showed that the absolute amounts of
both elastin and collagen contents increased in hypertensive rats; however, the per-
centage of these elements remained essentially constant (Table B6.8)

Table B6.10  Composition of Normal Human Venous Tissue*

Femoral vein Vena cava, inferior
Dry matter 28.0† 26.1
Nitrogen 4.08 –
Cholesterol 0.076 0.083
Total ash 0.590 –
Calcium 0.058 0.012
Potassium – 0.065
* Source: adapted from [27].
† Percentage of wet tissue weight.

Changes in elastin and collagen content due to atherosclerosis

Table B6.9 lists the changes in elastin and collagen contents of canine carotid and
iliac arteries due to dietary atherosclerosis [26]. In the iliac site the ratio of collagen
to elastin was increased, while the ratio in the carotid site was decreased.

B6.4  Constituents of the venous wall

B6.4.1  Normal venous wall

The main constituents of normal human venous tissue are listed in Table B6.10.
B6  Arteries, Veins and Lymphatic Vessels 85

Table B6.11  Composition of Normal Human Venous Tissue*

Femoral vein
Age group Dry matter Nitrogen Cholesterol Total ash Calcium
0–9 31.1 † 4.63 0.058 0.525 0.051
20–29 27.4 3.88 0.071 0.586 0.053
40–49 24.7 3.34 0.064 0.542 0.065
60–69 23.2 3.06 0.087 0.555 0.075
70–79 21.8 2.92 0.087 0.600 0.083
Vena cava, inferior
Age group Dry matter Nitrogen Cholesterol Calcium Potassium
0–9 30.9 – – – –
20–29 26.8 – 0.082 0.011 0.072
40–49 24.5 – 0.091 0.011 0.053
60–69 21.7 – 0.097 0.010 0.048
70–79 22.7 – 0.097 0.011 0.048
* Source: [27] by permission.; see source for additional age groups.
† Percentage of wet tissue weight.

Table B6.12  Mechanical Properties of Layers of the Vascular Wall*

Young’s Relative
modulus extensibilitya Maximum Tensile strength
Material (dynes/cm2) (cm) extension (%) (dynes/cm2)
Elastin 3 to 6 × 106 10 100–220 3.6 × 106 to
4.4 × 107
Collagen 1 × 109 to 0.03 5.50 5 × 107 to
2.9 × 1010 5 × 109
Smooth muscle
Relaxed 6 × 104 – 300 –
Contracted 1 × 105 to 2.3 300 –
12.7 × 106
* Source: Adapted from [28]; Fiber 10 em in length, 1 mm2 cross section, sustaining a load
of 30 g.

B6.4.2  C
 hanges with age in composition of normal venous
tissues

The changes with age in the composition of normal venous vessels are listed in
Table B6.11.
86 X. Deng and R. Guidoin

B6.5  Mechanical Properties of Arteries

The blood vessel wall consists of three layers: the intima, media, and adventitia. The
intima contains mainly the endothelial cells that contribute little to the strength of
the blood vessels. The media and the adventitia contain smooth muscle cells, elastin
and collagen. Elastin is the most ‘linearly’ elastic biosolid material known. Unlike
elastin, collagen does not obey Hooke’s Law. However, collagen is the main load
carrying element of blood vessels. Table B6.12 lists the mechanical properties of
tissues composing the blood vessel wall.

B6.5.1  Static mechanical properties of arteries

Studies of arterial wall mechanics have clearly established the anisotropic nature of
arteries [1, 29, 30]. In vivo pressurized arteries are deformed simultaneously in all
directions. But, experimental studies [31] have demonstrated that arteries deform
orthotropically. Therefore, arterial deformations may be examined in three orthogo-
nal directions, namely, the longitudinal, circumferential, and radial directions.
There are nine elastic parameters: the three elastic moduli, Eθ, Ez and Er; and six
Poisson’s ratios, σrθ,σθr,σzθ,σθz,σrz andσzr. As far as hemodynamics is concerned, how-
ever, among the three elastic moduli the circumferential one is most important.
The cicumferential elastic modulus is termed the incremental modulus and deter-
mined by the following equation (B6.2):

Dp i 2 (1- s 2 ) R 0 R i 2
Eq = (B6.2)
DR 0 ( R 0 2 - R 2i )

Table B6.13  Elastic Modulus of Human Arteries*

Arterial segment Eθ (×106 dynes/cm2)
Thoracic aorta 4.0–10
Abdominal aorta 4.0–15
Iliac artery 8.0–40
Femoral artery 12–40
Carotid artery 3.0–8.0
* Source: Adapted from [33,34]; young subjects (≤ 30 years) at a
transmural pressure of 100 mmHg.
Note: The circumferential elastic moduli were calculated assuming
that the arterial walls have a Poisson’s ratio (the ratio of transverse to
longitudinal strain) of 0.5.
B6  Arteries, Veins and Lymphatic Vessels 87

Table B6.14  Hydrodynamic Properties of the Aorta*

Thoracic aorta
Age (yrs) Zo (mmHgS cm3) Vp (m/s) E0† (106 N/m2)
30–39 0.13 5.8 0.56
40–49 0.17 7.4   0.8
50–59 0.19 9.0 1.13
60–69 0.19 10.0 1.25
70–79 0.22 12.4 1.87
80–89 0.22 13.6  2.2
Abdominal aorta
Age (yrs) Zo (mmHgS cm3) Vp (m/s) E0† (106 N/m2)
30–39 0.31 7.9 0.8
40–49 0.53 10.5 1.22
50–59 0.53 11.4 1.3
60–69 0.51 12.0 1.5
70–79 0.68 14.5 1.75
* Adapted from [35] by permission;
† Poisson’s ratio of arterial wall assumed to be 0.5.

where Δpi is the transmural pressure increment, Ro and Ri are the respective external
and internal diameter of the vessel, ΔRo is the change in the external diameter due
to Δpi, and σ the Poisson’s ratio (the ratio of transverse strain to longitudinal strain).
The detailed technique for measuring the circumferential incremental elastic modu-
lus of the arteries was described by Bergel [32].
Table B6.13 presents the circumferential incremental elastic modulus of human
arterial walls from young adults (≤ 35 years) at a transmural pressure of 100 mmHg.
Experimental data by Bader [34] demonstrated that the circumferential elastic
modulus of human thoracic aorta increased almost linearly with age. Table B6.14
gives the variation with age in the elastic moduli for human thoracic and abdominal
aorta at a pressure of 100 mmHg.
It should be mentioned that all the circumferential elastic moduli given in Table
B6.14 are based on the assumption of a Poisson’s ratio of 0.5. This is not strictly
true when large strains are considered [36]. Patel et al. [37] measured Poisson’s
ratios for the aorta in living dogs at a transmural pressure of about 110 mmHg, as
well as the circumferential, longitudinal and radial incremental elastic moduli and
determined that individual values vary between 0.29 and 0.71.
88 X. Deng and R. Guidoin

B6.5.2  C
 ompliance, pressure cross-sectional area relationship,
and retraction

By measuring the static elastic properties of human thoracic and abdominal aortas
in vitro, Langewouters et al. [35] proposed the following empirical relationship
between the cross-sectional area of the lumen (A) and the pressure in the vessel (p):

ïì 1 1 æ p - p0 ö üï
A( p) = Am í + tan -1 ç ÷ý (B6.3)
îï 2 p è p1 ø þï

in which Am, Po and p1 are three independent parameters that are defined in Equation
(B6.4) below.
Another important mechanical property of blood vessels in the compliance.
Langewouters et al. [35] defined the ‘static’ compliance as the derivative of equa-
tion (B6.3) with respect to pressure

Cm Am
C ( p) = ; Cm = (B6.4)
æ p - p0 ö
2
p p1
1+ ç ÷
è p1 ø
where Cm and Am are the maximum compliance and the maximum crosssectional
area of the vessel, respectively; Po is the pressure at which aortic compliance reaches
its maximum; and p1 is the half-width pressure, i.e. at p0 ± p1 , aortic compliance is
equal to Cm/2. According to Langewouters et al. [35], the ‘static’ compliance values
of human thoracic aorta at 100 mmHg range from 1.9 to 17 x 10·3 cm2/mmHg; and
0.6 to 4.4 x 10-3 cm2/mmHg for abdominal aorta.
Table B6.15 lists the relative wall thickness and the retraction on excision for a
variety of blood vessels. The retraction of a vessel is the amount by which a segment
of vessel shortens on removal from the body, expressed as a percentage of the length
of the segment in situ. The relative wall thickness is the ratio of wall thickness to
mean diameter of the vessel.

B6.5.3  Tensile properties of human arteries

Table B6.16 presents typical data for the tensile properties of arterial tissues from
Yamada [43]. The test specimens of tissues were strips each with a reduced middle
region 10 mm length, 2–3 mm in width, and a length to width ratio of 3:1. In the
tables:
1. Tensile breaking load per unit width (g/mm) = ultimate tensile strength
(g/mm2) x thickness (mm)
tensile breaking load ( g )
2. Ultimate tensile strength ( g mm 2 ) =
cross - section area of the test section
B6  Arteries, Veins and Lymphatic Vessels 89

Table B6.15  Relative Wall Thickness and Retraction on Excision of Various blood Vessels*
Vessel Species γx% Retraction % Source
Thoracic aorta Dog 10.5 32 [2]
Abdominal aorta Dog 10.5 34 [2]
Femoral artery Dog 11.5 42 [2]
Carotid artery Dog 13.2 35 [2]
Iliac artery Dog – 40 [2]
Carotid artery Dog 13.6 – [38]
Carotid artery Cat 14.5 – [38]
Carotid artery Rabbit 11.2 – [38]
Carotid sinus Dog 20.0 – [38]
Carotid sinus Cat 16.0 – [38]
Carotid sinus Rabbit 12.0 – [38]
Thoracic aorta Dog 14.0 – [39]
Abdominal aorta Dog 12.0 – [39]
Femoral artery Dog 13.0 – [39]
Pulmonary artery Man 2.0 – [40]
Abdominal vena cava Dog 2.3 30 [41]
Thoracic aorta Man 6–9 25–15 [33]†
Abdominal aorta Man 8–13 30–17 [33]†
Femoral artery Man 12–19 40–25 [33]†
Carotid artery Man 2–15 25–18 [33]†
Source: Adapted from [42].
† Measurement from [33] of young (<35) and old (>35) subjects respectively.

Table B6.16  Tensile Properties of Human Coronary Arterial Tissue* (Longitudinal Direction)
Tensile Breaking Ultimate Tensile Ultimate
Age (yrs) Load/Unit width (g/mm) Strength (g/mm2) Elongation (%)
10–19 85±3.1 140±3.0 99±2.4
20–39 82±1.8 114±9.3 78±1.6
40–59 82±1.8 104±4.7 68±3.5
60–79 79±2.9 104±4.7 4.5±3.8
Adult 81 107 64
(average)
* Adapted from [43].

3. Ultimate percentage elongation ( % ) = breaking elongation ( mm )
´100
original length of the specimen ( mm )

Table B6.16 lists the tensile data for human coronary arterial tissue in the longi-
tudinal direction.
90 X. Deng and R. Guidoin

Other arteries have similar properties [43]. Yamada [43] provides the tensile
properties of animal tissues in various tables.

B6.5.4  Dynamic mechanical properties of arteries

The most direct way to study arterial viscoelasticity is to determine the response of
the test tissue to oscillatory stresses. If the arterial wall is conceived to be repre-
sented by a simple Kelvin-Voigt model consisting of a spring and a dashpot in paral-
lel, the dynamic elastic component and the viscous component of a vessel can be
expressed as

ED = E cos f (B6.5)
hw = E sin f

where E is the complex dynamic elastic modulus that is identical to the incremental
elastic modulus under static stresses; ϕ is the phase lag of the strain behind the
stress (in the case of circumferential direction, it is the phase lag of diameter behind
the pressure); ED is the dynamic elastic modulus of the vessel; and ηω is the viscous
retarding modulus (η is the viscous constant and w the angular frequency). For mea-
suring the circumferential dynamic mechanical properties, the test vessel is usually
subjected to an oscillatory pressure. The pressure oscillations are in a sinusoidal
form. In circumferential direction, E can be calculated from equation (B6.2) with
the recorded diameter of the blood vessel and the oscillatory pressure [44].
Table B6.17 lists the dynamic mechanical properties of different arteries at a
frequency of 2.0 Hz at a mean pressure of 100 mmHg. In this table, EP is the circum-
ferential pressure-strain modulus defined as

Table B6.17  Circumferential Dynamic Mechanical Properties of Different Human Arteries*

Vessel Ep (dynes-cm-2 x 10-6) Φ(radians) Vp (m/s) Source
Ascending aorta 0.8 – 6.0 [29]
Carotid 6.2 – 17 [29]
Carotid 0.4 – 4.6+ [45]
Thoracic aorta 0.6–1.0 0.12 6–9+ [33]++
Abdominal aorta 0.7–1.5 0.1 6–8+ [33]++
Femoral 2.5–7.0 0.15 13–18 [33]++
Carotid 2.5–3.0 0.1 – [33]++
Pulmonary artery 0.1–0.16 – 2.3–2.9 [29]
* Adapted from [42].
Elastic modulus and pulse-wave velocity values collected from the literature. Those values for the
PWV marked thus: + were measured, the others have been calculated from dynamic elasticity
measurements; ++ measurements for young (<35) and old (>35) subjects respectively. Φ = phase
difference between wall stress and strain at 2 Hz. Ep is defined in eq (B6.6). The mean pressure was
100 mmHg.
B6  Arteries, Veins and Lymphatic Vessels 91

E p = DPi Ro  DRo (B6.6)

in which ΔPi is the pressure increment, Ro is the external diameter of the vessel and
ΔRo the external diameter change. Ep is essentially a reciprocal of the compliance of
an artery and differs from elastic modulus in that it defines the stiffness of the total
artery. Nevertheless, it is an indication of ED.
Another important parameter listed in Table B6.17 is the pulse-wave velocity Vp
(PWV) that can be calculated from the Moens-Korteweg equation:

V p2 = Eh 2 Rp (B6.7)

where E = elastic modulus of the wall, h = wall thickness, R = mean radius of the
vessel, ρ = density of blood.
There is general understanding that the dynamic elastic modulus is not strongly
frequency dependent above 2–4 Hz and that it increases from the static value at
quite low frequencies. Bergel [42] provides additional values for canine vessels.

B6.5.5  Creep and stress relaxation

When a subject is suddenly strained and then the strain is maintained constant after-
ward, the corresponding stresses induced in the subject decrease with time, this
phenomenon is called stress relaxation. If the subject is suddenly stressed and then
the stress is maintained afterwards, the subject continues to deform, this phenome-
non is called creep. Creep and stress relaxation are another two important phenom-
ena in the arterial viscoelasiticity. Langewouters et al. [46] studied the creep
responses of human thoracic and abdominal aortic segments. The pressure in the
segments was changed in steps of 20 mmHg between 20 and 180 mmHg. Aortic

Table B6.18  Creep Constants of Human Thoracic and Abdominal Aortas*

Parameter Thoracic aorta (n = 35) Abdominal aorta (n = 16)
Range Mean S.D. Range Mean S.D.
α1 0.05–0.13 0.076 0.017 0.05–0.12 0.078 0.017
α2 0.03–0.15 0.102 0.028 0.07–0.15 0.101 0.025
t1 (S) 0.31–1.43 0.73 0.29 0.33–0.81 0.61 0.12
t2 (S) 5.9–23.5 14.0 4.1 4.6–17.7 12.1 3.4

( 10 -3 cm 2
1.5–12.3 5.1 2.5 0.5–3.3 1.5 0.78
C
mmHg )
Age 30–78 63 14 30–78 58 15
* Adapted from [46] by permission.
α1, α2, creep fractions; t1, t2, time constants; C, compliance; S.D., standard deviation.
92 X. Deng and R. Guidoin

creep curves at each pressure level were described individually by a constant plus
biexponential creep model (C-model, [47])

é æ -t ö æ -t öù
Aic = DA ê1 - a1 exp ç i ÷ - a 2 exp ç i ÷ú (B6.8)

êë è t1 ø è t2 ø úû

where A = aortic cross-sectional area; Aci = sample value of aortic creep response at
time ti; δA = change in aortic area upon a 20 mmHg pressure step; t = time; i =
sample number; a1,a2 = creep fraction; τ1,τ2 = time constant. Table B6.18 lists the
creep fractions and time constants for all aortic segments [46].
Stress relaxation relations for human arteries are not available; however, Tanaka
and Fung [48] studied the stress relaxation spectrum of the canine aorta. They
expressed the stress history with respect to a step change in strain in the form:

T (t ,1) = G(t ).T (e) (1), G(0) = 1 (B6.9)

where G (t) is the normalized relaxation function of time; T(e) (l) is a function of
strain l, called elastic response. This is the tensile stress instantaneously generated
in the aortic tissue when a step elongation, l, is imposed on the specimen.
If the relaxation function is written as:

1é
1 + ò s (t ) e - t t dt ù
µ
G (t ) = (B6.10)
A ê
ë 0 ûú

in which

A = é1 + ò s (t ) dt ù
µ

ëê 0 ûú

is a normalized factor, then the spectrum S (t) is expected to be a continuous func-

tion of relaxation time t. A special form of S (t) is proposed

S (t ) = c  t for t1 £ t £ t2
(B6.11)
= 0 for t1 < t1 , t > t2

Values for segments of the canine aorta are given in [48].
B6  Arteries, Veins and Lymphatic Vessels 93

Table B6.19  Comparison of elastic Moduli between Venous and arterial Segments*
Pressure Extension ratio Incremental venous elastic Carotid artery Incremental
(cm H2O) modulus modulus
λθ λz Eθ (dynes/cm2 Ez (dynes/cm2 Eθ (dynes/cm2 Ez (dynes/cm2
x 106) x 106) x 106) x 106)
Canine Jugular Vein
10 1.457 1.481 15 ± 3* 1.2 ± 0.18† 7.62 5.16
25 1.463 1.530 47 ± 6† 4.4 ± 0.35† 8.39 7.15
50 1.472 1.597 88 ± 7† 11.8 ± 2.1 9.51 10.69
75 1.478 1.646 98 ± 7† 46 ± 13* 10.37 13.97
100 1.482 1.675 117 ± l0† 67 ± 25 10.92 16.24
125 1.484 1.686 134 ± 24† 89 ± 32 11.16 17.17
150 1.484 1.686 171 ± 9† 113 ± 13† 11.16 17.17
Human Saphenous Vein
10 1.357 1.169 0.27 ± 0.12† 1.61 ± 0.32* 5.30 0.017
25 1.417 1.206 0.65 ± 0.13† 2.03 ± 0.39* 5.82 0.328
50 1.500 1.266 1.89 ± 0.41† 2.75 ± 0.78 6.00 0.735
75 1.561 1.325 9.85 ± 1.6 3.18 ± 0.76 9.66 1.80
100 1.602 1.381 15.0 ± 2.6 3.56 ± 0.58 12.77 3.15
125 1.621 1.430 20.4 ± 1.6* 3.98 ± 0.96 14.79 4.59
150 1.621 1.470 25.1 ± 7.5 4.75 ± 1.2 15.51 5.93
* Adapted from [50] by permission.
* p < 0.05 for the comparison between the venous and carotid moduli.
† P < 0.01 for the comparison between the venous and carotid moduli.

B6.6  Mechanical Properties of veins

B6.6.1  Static mechanical properties of veins

The structure of the venous walls is basically similar to that of the arterial walls. The
main difference is that they contain less muscle and elastic tissue than the arterial
walls, which raises the static elastic modulus two to fourfold [49]. Because the
venous walls are much thinner than the arterial wall, they are easily collapsible
when they are subject to external compressions.
Table B6.19 lists the static incremental elastic moduli of the canine jugular vein
and human saphenous vein. For the purpose of comparison, the static increment
elastic modulus of the canine carotid artery segments are also presented in the table.
This comparison is of interest because in some arterial reconstructive surgeries, a
vein is used as a substitute for an artery.
Sobin [51] obtained data on the mechanical properties of human vena cava from
autopsy material. The data may be expressed by the following equation:
94 X. Deng and R. Guidoin

Table B6.20  Tensile Properties of Human Venous Tissue*

Age group
Vein Direction 20–39 yr 40–50 yr 60–69 yr adult
average
Tensile Breaking Load per Unit Width (g/mm)
Inferior vena cava L 102.0 87.0 68.0 89.0
T 245.0 224.0 224.0 232.0
Femoral L 159.0 149.0 149.0 153.0
T 211.0 217.0 224.0 216.0
Popliteal L 116.0 116.0 116.0 116.0
T 180.0 197.0 158.0 182.0
Ultimate Tensile Strength (kg/mm2)
Inferior vena cava L 0.15 0.11 0.08 0.12
T 0.36 0.28 0.27 0.31
Femoral L 0.24 0.21 0.20 0.22
T 0.32 0.31 0.29 0.31
Popliteal L 0.20 0.17 0.15 0.18
T 0.31 0.29 0.20 0.27
Ultimate Percentage Elongation
Inferior vena cava L 98.0 77.0 70.0 84.0
T 58.0 47.0 44.0 51.0
Femoral L 97.0 72.0 72.0 82.0
T 79.0 67.0 56.0 70.0
Popliteal L 112.0 97.0 81.0 100.0
T 77.0 77.0 77.0 77.0
* Adapted from [43].L = longitudinal; T = transverse.

T = a ECl[ a( E 2 - E *2 )]
E = (l 2 - 1)  2 (B6.12)
T * = a CE * l *

where λ. is the ratio of the changed length of the specimen divided by the reference
length of the specimen, C and a are material constants, and E* is the strain that cor-
responds to a selected value of stress S*. The product of the constant αC is similar
to the elastic modulus, provided that the modulus is defined as the ratio, S*/E*,
where S* = T*/λ*. Fung [1] provides typical values of these constants obtained
experimentally; however, no universal constants have been discovered.

B6.6.2  Tensile properties of veins

The tensile properties of human venous tissues are presented in Table B6.20. For the
testing method and definitions of the terms in the table, please refer to Section B6.5
on Mechanical properties of arteries.
B6  Arteries, Veins and Lymphatic Vessels 95

Table B6.21  Water Permeability of Iliac Vessels*

Vessel Pressure (mmHg) Filtration rate
(cm/sec x 106)
Iliac artery 0–20 0
25–80 2.58
100 4.08
200 6.42
Iliac vein 0–20 6.94
21–40 9.72
41–60 11.67
61–80 12.22
81–100 12.78
* Adapted from [56].

B6.7  Mechanical Characteristics of Lymphatic Vessels

The problem concerning the ontogenesis of the lymphatic vessels is still not com-
pletely solved. However, most of the evidence indicates that the large lymphatic
vessels are derived from the veins [52]. Therefore, lymphatics can be considered
modified veins. According to the reports by Ohhashi et al. [53, 54], the circumfer-
ential elastic modulus of the bovine mesenteric lymphatics ranged from 4.2 x 104 to
2.7 x 105 dynes/cm2 at a pressure range from 0 to about 20 mmHg, and the elastic
modulus of canine thoracic duct is about 2.0 x 105 dynes/cm2. These values are less
than those of veins obtained by Bergel [55]. Therefore, the lymphatics are more
distensible than the veins.

B6.8  Transport Properties of blood Vessels

Under transmural pressures, fluid or plasma will flow across the walls of blood ves-
sels. On one hand, convective fluid motion through the blood vessel wall plays a
very important role in nourishing the vessel walls, on the other hand, it is involved
in atherogenesis by promoting the transport of macromolecules such as lipoproteins
into the arterial wall [55–58], possibly through leaky endothelial cell junctions in
regions of high endothelial cell turnover [59]. Table B6.21 lists the filtration proper-
ties of excised, presumably normal, human iliac blood vessels.
96 X. Deng and R. Guidoin

B6.9 Effect of Age, Hypertension and Atherosclerosis

on blood Vessels

B6.9.1  Age
Two well known changes accompany aging of the cardiovascular system are dila-
tion of thoracic aorta [60] and increased thickness of arterial wall [61]. Arterial
walls become less distensible with aging [34, 62, 63]. Both dry matter and nitrogen
content of artery tissue show a tendency to decrease with age in large and medium
sized arteries [20]. But the relative quantity of collagen [22] and elastin [23, 24] in
the arterial wall remains essentially unchanged. With aging the arterial wall becomes
progressively stiffer. Bader [34] found that the circumferential elastic modulus of
human thoracic aorta increased linearly with age. At 100 mmHg, the static circum-
ferential modulus of ‘young’ (<35 years) human thoracic aorta averaged 7.5 x 106
dynes/cm2, and for the ‘old’ (>35 years) the average was 16.6 x 106 dynes/cm2 [33].
Young peripheral arteries tend to have a greater viscosity (ƞω) than the older ones
[33]. Despite the overall increase in stiffness, the arterial wall itself is considerably
weaker than the younger/older one [43].

B6.9.2  Hypertension

Several studies have shown that the water content of human, rat and dog arteries is
increased in hypertension, and this increased water content may be associated with
an increased wall thickness [64, 65]. Due to the limitations in studying samples
from human subjects, animal models (mainly rats) have been employed. Mallov
[66] found that the aorta from hypertensive rats had more smooth muscle than nor-
mal aorta. Greenwald and Berry [67] reported increased elastin and decreased col-
lagen content in the aorta from spontaneous hypertensive rats when compared with
the normal aorta. Wolinsky [25] observed an increase in the absolute amounts of
both medial elastin and collagen contents in hypertensive rats. However, the relative
percentage of these elements remained essentially constant. Experimental studies
[67–69] showed an increase in vessel stiffness with the development of hyperten-
sion. This increase in vessel stiffness results in a smaller vessel diameter for a given
distending pressure, i.e. a decrease in the distensibility [70].

B6.9.3  Atherosclerosis

It is generally accepted that substantial changes in the arterial wall occur with ath-
erosclerosis in man. In human atherosclerotic arteries, it appears that there may be
an absolute increase in collagen and a decrease in muscle fibers when compared
B6  Arteries, Veins and Lymphatic Vessels 97

with normal arteries [28]. In canine iliac artery the ratio of collagen to elastin was
found to be increased, while the ratio in carotid was decreased [26]. The elastic
moduli of the diseased aorta and common iliac arteries are several times higher than
those reported [33] for normal arteries. The most popular model used to study the
effect of atherosclerosis on arterial wall properties is the rabbit subject to a high
cholesterol diet. Cox and Detweiler [26] have shown that in the iliac arteries from
high cholesterol fed greyhounds, collagen and elastin contents are decreased, but
the ratio of collagen to elastin is increased. In carotid arteries from the treated ani-
mals, the elastin content is increased and the collagen to elastin ratio is decreased.
Their results also show that the elastic modulus of the iliacs from the cholesterol fed
animals is higher than that of the normal iliacs while the treated carotids are
unchanged. By using a rabbit model, Pynadath and Mukherjee [71] found that cho-
lesterol feeding had no effect on the longitudinal dynamic elastic modulus of the
aorta, but the circumferential one was affected significantly. After six weeks of
feeding, the circumferential dynamic elastic modulus increased from the normal
value of 2.7 x 106 dynes/cm2 to 4.0 x 10 6 dynes/cm2. This increase showed a remark-
able correlation with the cholesterol content in the aortas. Although animal models
shed some light on the effects of cholesterol feeding, since atherosclerotic changes
in the arterial wall are so closely related to aging, it is difficult to separate the effect
of atherosclerosis from those of aging. The effects of atherosclerosis on the mechan-
ical properties of the arterial wall remain unclear. Confusion with the effects of
aging and other forms of arteriosclerosis such as medial calcification make interpre-
tation of the results difficult.

B6.10  Final Comments

Mechanical properties of the arteries from human and various animals have been
extensively studied. However, literature on lymphatic vessels is very scarce. The
data on the circumferential elastic modulus of the lymphatic vessels obtained by
Ohhashi et al. [53, 54] seem to be too low considering that the lymphatics are origi-
nating from the veins.
The overall viscoelastic properties of a large blood vessel such as the aorta are
known to be nonlinear [2, 44]and anisotropic [37]. But due to the fact that the blood
vessel wall is incompressible [72] and deforms orthotropically [31], the mechanical
properties of blood vessels can be described mainly by six coefficients: an elastic
and a viscous moduli in the longitudinal, circumferential and radial directions.
Among them, only the moduli in the circumferential and longitudinal directions
have been studied widely. Much fewer data in the radial direction can be found in
literature. In calculation of the circumferential elastic moduli, it was usually
assumed that the Poisson’s ratio was 0.5 that is not strictly true when large strains
are considered [36]. In fact, the measured data [37, 73] show that the Poisson’s ratio
is about 0.3, not 0.5 as would be predicted for an isotropic material. But the Poisson’s
98 X. Deng and R. Guidoin

ratio was almost constant with respect to circumferential strain and pressure in both
relaxed and constricted canine carotid arteries [73].
To our best knowledge, water diffusion properties of blood vessels have been
studied extensively, but their electrical and thermal properties are still unknown.

Acknowledgement  This work was supported by the Medical Research Council of Canada (Grant
MT-7879). The assistance of Y. Marois, M. King and Y. Douville in preparation of this material is
gratefully acknowledged.

Additional Reading

Canfield, T.R. and Dorbin, P.B. (1987) Static properties of blood vessels, in
Handbook of Bioengineering (eds R.  Skalak and S.  Chien), McGraw-Hill Book
Company, New York, pp. 16.1–16.28.
The authors discuss the mechanical behavior of arteries and the mathematical
method required for quantification of such data. The discussion is entirely con-
cerned with the elastic or pseudoelastic behavior of blood vessels. It should be
emphasized that the arterial wall also exhibits inelastic properties, such as viscosity,
creep, stress & relaxation and pressure-diameter hysteresis. Very few data on
mechanical properties of blood vessels are presented.
Dorbin, P.B. (1983) Vascular mechanics, in Handbook of Physiology, Vol. 3 The
Cardiovascular System, (eds J.T. Shepherd and F. Abboud), Amer. Physiol. Soc.,
Washington, DC, Section 2, pp. 65–102.
This chapter reviews the essential concepts of vascular mechanics and its meth-
ods of quantification. Some of the important controversies are discussed, and further
research areas are pointed out. Detailed information is provided on the structural
and mechanical changes of arteries with age. The effect of vascular disorders such
as arterosclerosis and hypertension on the mechanical behavior of blood vessels is
discussed as well. Extensive addition literature sources are provided.
Schneck, D.J.(1995) An outline of cardiovascular structure and function, in The
Biomedical Engineering Handbook,(ed. J.D. Bronzino), CRC Press, Boca Raton,
pp. 3–14.
The cardiovascular system is described as a highway network, which includes a
pumping station (the heart), a working fluid (blood), a complex branching configu-
ration of distributing and collecting pipes and channels (the blood vessels), and a
sophisticated means for both intrinsic (inherent) and extrinsic (antonomic and endo-
crine) control. Data on both the arterial and venous systems are tabulated. However,
no detailed sources are provided for the data listed. This is a very suitable reference
for biomedical engineers.
Hargen, A.R. and Villavicenco, J.L. (1995) Mechanics of tissue/lymphatic sup-
port, in The Biomedical Engineering Handbook, (ed J.D.  Bronzino), CRC Press,
Boca Raton, pp. 493–504.
B6  Arteries, Veins and Lymphatic Vessels 99

From an engineering point of view, the authors discuss the lymphatic system as
a drainage system for fluids and waste products from tissues. Basic concepts of
lymphatic transport along with clinical disorders are discused, although briefly.
Extensive additional sources are cited.

References

1. Fung, Y.C. (1993) Biomechanics, Mechanical Properties of Living Tissues, 2nd edn, Springer-
Verlag, New York, pp. 321–391.
2. Bergel, D.H. (1961) The static elastic properties of the arterial wall. J. Physiol., 156,
445–457.
3. Patel, D.J. and Vaishnav, R.N. (1972) The rheology of large blood vessels, Cardiovascular
Fluid Dynamics, D.H. Bergel (ed.), Academic Press, New York, Vol. 2, pp. 1–64.
4. Fung, Y.C., Fronek, K. and Patitucci, P. (1979) Pseudoelasticity of arteries and the choice of its
mathematical expression. Am. J. Physiol., 237, H620–H631.
5. Remington, J.W. (1955) Hysteresis loop behavior of the aorta and other extensible tissues. Am.
J. Physiol., 180, 83–95.
6. Alexander, R.S. (1971) Contribution of plastoelasticity to the tone of the cat portal vein. Circ.
Res., 28, 461–469.
7. Anitschkov, N. (1933) Experimental arteriosclerosis in animals, in Arteriosclerosis,
E.V. Cowdry, Editor, Macmillan, New York, pp. 298–299.
8. Wilens, S.L. and McCluskey, R.T. (1954) The permeability of excised arteries and other tis-
sues to serum lipid. Circ. Res., 2, 175–182.
9. Baldwin, A.L., Wilson, L.M. and Simon, B.R. (1992) Effect of pressure on aortic hydraulic
conductance. Arteria. Thromb., 12, 163–171.
10. Fung, Y.C. (1984) Biodynamics, Circulation, Springer-Verlag, New York, p. 77.
11. Brecher, G.A. (1956) Venous Return, Grune & Stratton, New York, pp. 30–67.
12. Helps, E.P.W. and McDonald, D.A. (1954) Observation on laminar flow in veins. J. Physiol.,
124, 631–639.
13. Spencer, M.P. and Denison, A.B. (1963) Pulsatile blood flow in the vascular system, in
Handbook of Physiology, Section 2, Circulation, W.F.  Hamilton and P.  Dow, Editors, Am.
Physiol. Soc., Washington, pp. 839–864.
14. Fry, D.L. and Greefield, J.C. Jr. (1964) The mathematical approach to hemodynamics, with
particular reference to Womersley’s theory, in Pulsatile blood Flow, Attinger, E.O. (ed),
McGraw-Hill, New York, pp. 85–99.
15. Maggio, E. (1965) Microhemocirculation, Observable Variables and Their Biological Control,
C.C. Thomas, Springfield.
16. Whitmore, R.L. (1968) Rheology of the Circulation, Pergamon Press, Oxford, pp. 99–108.
17. Morse, D.E. (1979) The normal aorta: embryology, anatomy, and histology of the aorta, in The
Aorta, J. Lindsay, Jr. and J.W. Hurst, Editors, Grune & Stratton, New York, pp. 15–37.
18. Singhal, S., Henderson, R., Horsfield, K. et al. (1973) Morphometry of human pulmonary arte-
rial tree. Circ. Res., 33, 190–197.
19. Fung, Y.C. (1984) Biodynamics, Circulation, Springer-Verlag, New York, pp. 77–165.
20. Kirk, J.E. (1962) Arterial and arteriolar system, biochemistry, in Blood Vessels and Lymphatics,
D.l. Abramson, Editor, Academic Press Inc., London, pp. 82–95.
21. Fischer, G.M. and Llaurado, J.G. (1966) Collagen and elastin content in canine arteries
selected from functionally different vascular beds. Circ. Res., 19, 394–399.
22. Kanabrocki, E.L., Fels, I.G. and Kaplan, E. (1960) Calcium, cholesterol and collagen levels in
human aorta. J. Gerontal., 15, 383–387.
23. Lansing, A., Alex, M. and Rosenthal, T.B. (1950) Calcium and elastin in human arteriosclero-
sis. J. Gerontal., 5, 112–119.
100 X. Deng and R. Guidoin

24. Kraemer, D.M. and Miller, H. (1953) Elastin content of the abdominal fraction of human aor-
tae. Arch. Pathol., 55, 70–72.
25. Wolinsky, H. (1970) Response of the rat aortic media to hypertension. Morphological and
chemical studies. Circ. Res., 26, 507–523.
26. Cox, R.H. and Detweiler, D.K. (1979) Arterial wall properties and dietary atherosclerosis in
the racing greyhound. Am. J. Physiol., 5, H790–H797.
27. Kirk, J.E. (1962) Venous system, biochemistry, in Blood Vessels and Lymphatics,
D.I. Abramson, Editor, Academic Press Inc., London, pp. 211–213.
28. Strandness, D.E. and Sumner, D.S. (eds) (1975) Hemodynamics for Surgeons, Grune &
Stratton, New York, New York, pp. 161–205.
29. Patel, D.J., Greenfield, J.C. and Fry, D.L. (1964) In vivo pressure-length-radius relationships
of certain vessels in man and dogs, in Pulsatile blood Flow, E.O. Attinger, Editor, McGraw-
Hill, New York, pp. 293–302.
30. Cox, R.H. (1975) Anisotropic properties of the canine carotid artery in vitro. J. Biomech., 8,
293–300.
31. Patel, D.J. and Fry, D.L. (1969) The elastic symmetry of arterial segments in dogs. Circ. Res.,
24, 1–8.
32. Bergel, D.H. (1958) A photo-electric method for the determination of the elasto-viscous
behavior of the arterial wall. J. Physiol., 141, 22–23.
33. Learoyd, B.M. and Taylor, M.G. (1966) Alteration with age in the viscoelastic properties of
human arterial walls. Circ. Res., 18, 278–292.
34. Bader, H. (1967) Dependence of wall stress in the human thoracic aorta on age and pressure.
Circ. Res., 20, 354–361.
35. Langewouters, G.J., Wesseling, K.H. and Goedhard, W.J.A. (1984) The static elastic proper-
ties of 45 human thoracic and 20 abdominal aortas in vitro and the parameters of a new model.
J. Biomech., 17, 425–435.
36. Bergel, D.H. (1960) The Visco-elastic Properties of the Arterial Wall. PhD thesis, University
of London.
37. Patel, D.J., Janicki, J.S. and Carew, T.E. (1969) Static anisotropic elastic properties of the aorta
in living dogs. Circ. Res., 25, 765–779.
38. Rees, P.M. and Jepson, P. (1970) Measurement of arterial geometry and wall composition in
the carotid sinus baroreceptor area. Circ. Res., 26, 461–467.
39. Gow, B.S. and Taylor, M.G. (1968) Measurement of viscoelastic properties of arteries in the
living dog. Circ. Res., 23, 111–122.
40. Reid, L. (1968) Structural and functional reappraisal of the pulmonary artery system, in
Scientific Basis of Medicine, Annual Reviews, Vol. 8, pp. 289–307.
41. Yates, W.G. (1969) Experimental studies of the variations in the mechanical properties of the
canine abdominal vena cava, in SUDAAR Report 393, Stanford University, California.
42. Bergel, D.H. (1972) The properties of blood vessels, in Biomechanics, Its Functions and
Objectives, Y.C. Fung, Editor, Prentice-Hall, Englewood Cliffs, N.J., p. 110 and p. 131.
43. Yamada, H. (1970) Strength of Biological Materials, F.G. Evans, Editor, Williams and Wilkins,
Baltimore, pp. 106–277.
44. Bergel, D.H. (1961) The dynamic elastic properties of arterial wall. J. Physiol., 156,

458–469.
45. Arndt, J.O., Klauske, J. and Mersch, F. (1968) The diameter of the intact carotid artery in man
and its change with pulse pressure. Pfülger’s Arch,, 30, 230–240.
46. Langewouters, G.J., Wesseling, K.H. and Goedhard, W.J.A. (1985) The pressure dependent
dynamic elasticity of 35 thoracic and 16 abdominal human aortas in vitro described by a five
component model. J. Biomech., 18, 613–620.
47. Langewouters, G.J. (1982) Visco-Elasticity of the Human Aorta in vitro in Relation to Pressure
and Age. PhD thesis, Free University, Amsterdam, The Netherland.
48. Tanaka, T.T. and Fung, Y.C. (1974) Elastic and inelastic properties of the canine aorta and their
variation along the aortic trees. J. Biomech., 7, 357–370.
B6  Arteries, Veins and Lymphatic Vessels 101

49. Attinger, E.O. (1967) Modelling of pressure-flow relations in arteries and veins (Abstract).
Biorheology, 4, 84.
50. Wesly, R.L.R., Vaishnav, R.N., Fuchs, J.C.A. et al. (1975) Static linear and nonlinear elastic
properties of normal and arterialized venous tissue in dog and man. Circ. Res., 37, 509–520.
51. Sobin, P. (1977) Mechanical Properties of Human Veins. M.S. thesis, University of California,
San Diego, California.
52. Susznyák, I., Földi, M. and Szabó, G. (1967) Lymphatics and Lymph Circulation: Physiology
and Pathology, Pergamon Press, London, pp. 33–50.
53. Ohhashi, T., Azuma, T. and Sakaguchi, M. (1980) Active and passive mechanical characteris-
tics of bovine mesenteric lymphatics. Am. J. Physiol., 239, H88–H95.
54. Ohhashi, T. (1987) Comparison of viscoelastic properties of walls and functional characteris-
tics of valves in lymphatic and venous vessels. Lymphatic, 20, 219–223.
55. Bergel, D.H. (1964) Arterial viscoelasticity, in Pulsatile blood Flow, E.O. Attinger, Editor,
McGraw-Hill, New York, pp. 275–292.
56. Wilens, S.L. and McCluskey, R.T. (1952) The comparative filtration filtration properties of
excised arteries and veins. Am. J. Med. Sci., 224, 540–547.
57. Tedgui, A. and Lever, M.J. (1985) The interaction of convection and diffusion in the transport
of 131I-albumin within the media of rabbit thoracic aorta. Circ. Res., 57, 856–863.
58. Tedgui, A. and Lever, M.J. (1987) Effect of pressure and intimal damage on 131I-albumin and
14
C-sucrose spaces in aorta. Am. J. Physiol., 253, H1530–H1539.
59. Weinbaum, S., Tzeghai, G., Ganatos, P. et al. (1985) Effect of cell turnover and leaky junctions
on arterial macromolecular transport. Am. J. Physiol., 248, H945–H960.
60. Bazett, H.C., Cotton, F.S., Laplace, L.B. et al. (1935) The calculation of cardiac output and
effective peripheral resistance from blood pressure measurements with an appendix on the size
of the aorta in man. Am. J. Physiol., 113, 312–334.
61. Anderson, J.R. (1980), Muir’s Textbook of Pathology, Arnold, London, pp. 360–395.
62. Nakashima, T. and Tanikawa, J. (1971) A study of human aortic distensibility with relation to
atherosclerosis and aging. Angiology, 22, 477–490.
63. Mozersky, D.J., Summer, D.S., Hokanson, D.E. et al. (1972) Transcutaneous measurement of
the elastic properties of the human femoral artery. Circulation, 46, 948–955.
64. Tobian, L. (1960) Interrelationship of electrolytes, juxtaglomerular cells and hypertension.
Physiol. Rev., 40, 280–312.
65. Peterson, L.H. (1963) Systems behavior, feed-back loops and high blood pressure research.
Circ. Res., 12, 585–596.
66. Mallov, S. (1959) Comparative reactivities of aortic strips for hypertensive and normotensive
rats to epinephrine and levarterenol. Circ. Res., 7, 196–201
67. Greenwald, S.E. and Berry, C.L. (1978) Static mechanical properties and chemical composi-
tion of the aorta of spontaneously hypertensive rats, a comparison with the effects of induced
hypertension. Cardiovas. Res., 12, 364–372.
68. Feigl, E.O., Peterson, L.H. and Jones, A.W. (1963} Mechanical and chemical properties of
arteries in experimental hypertension. J. Clin. Invest., 42, 1640–1647.
69. Bandick, N. and Sparks, H. (1970) Viscoelastic properties of the aortas of hypertensive rats.
Proc. Soc. Exp. Bioi. Med., 134, 56–60.
70. Greene, M.A., Friedlander, R., Boltax, A.J. et al. (1966) Distensibility of arteries in human
hypertension. Proc. Roy. Soc. Exp. Biol., 121, 580–585.
71. Pynadath, T.I. and Mukherjee, D.P. (1977) Dynamic mechanical properties of atherosclerotic
aorta: A correlation between the cholesterol ester content and the viscoelastic properties of
atherosclerotic aorta. Atherosclerosis, 26, 311–318.
72. Carew, T.E.R., Vaishnav, R.N. and Patel, D.J. (1968) Compressibility of the arterial wall. Circ.
Res., 23, 61–68.
73. Dobrin, P.B. and Doyle, J.M. (1970) Vascular smooth muscle and the anisotropy of dog carotid
artery. Circ. Res., 27, 105–119.
Chapter B7
The Intraocular Lens

Traian V. Chirila and Shuko Suzuki

B7.1 Introduction

Although the existence of the intraocular crystalline lens in the eye was recognized
by the scholars of the Hellenistic period (about 2000 years ago), the actual role of
the lens in vision was properly understood much later. Truly scientific approaches
to the lens measurements and properties began to be applied only in the nineteenth
century [1]. For instance, a simple property—the weight of the human intraocular
lens—was first reported in 1883 [2].
The intraocular crystalline lens is positioned between the aqueous humor and
vitreous body of the eye. It is usually referred to as “the lens,” which is rather con-
venient considering that the term “intraocular lens” (with the acronym IOL) is now-
adays used exclusively to designate an artificial device implanted after surgical
removal of a cataractous (opaque) natural lens.
The lens refracts the light which enters the eye through the pupil and focuses it
on the retina. Its main functions are the following: (1) provides refractive power to
the optical system of the eye; (2) provides the accommodation necessary for normal
vision; (3) maintains its own transparency; and (4) absorbs UV radiation and blue
light, both deleterious to the subsequent ocular segments. These functions are all
important, but the lens’ contribution to the process of accommodation is crucial for
normal vision. Ocular accommodation is the ability of the eye to adjust the focal
length from far to near through changes in the shape of the lens. The consequence
of losing this ability with age is known as presbyopia, and is one of the major causes
of the need for visual correction in the middle-aged humans. Although the existing
theories attempting to explain presbyopia are still debated [3, 4], many investigators
believe that this condition is related to changes within the lens, especially its age-

T.V. Chirila (*) • S. Suzuki

Queensland Eye Institute, 140 Melbourne Street, South Brisbane, QLD 4101, Australia
e-mail: [email protected]

© Springer Science+Business Media New York 2016 103

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_10
104 T.V. Chirila and S. Suzuki

induced stiffening. As a result, the stiffness of the human lens has been extensively
investigated, and serves as an important parameter in the modeling of the
accommodation and presbyopia. The topic has been critically appraised and supple-
mented with improved experimental techniques in a recent study [5].
The lens is a biconvex body similar to a flattened globe. For descriptive pur-
poses, it has two poles (anterior and posterior), an equator, and therefore two diam-
eters (polar diameter, or thickness of the lens, and equatorial diameter).
The lens is composed of epithelial cells which become anuclear and elongated as
they are displaced further toward the center. Because of the enormous length finally
attained by these cells, they are referred to as lens fibers. The lens is surrounded by
a transparent acellular capsule of variable thickness. A proper epithelium underlies
the capsule along the anterior side and equator, but not under the posterior capsule.
The superficial layers of cells and fibers constitute the lens cortex, while the lens
nucleus is situated in the center. The fibers are continuously formed throughout life
and the new fibers cover the old ones which are displaced toward the nucleus.

B7.2 Compendium of Physicochemical Characteristics

of the Human Lens

Currently, there are significantly more compositional and physical property data
available on animal lenses than on the human lens. The spread of the data measured
for the human lens, compiled here from various sources, results from large varia-
tions in methodologies from one research group to another, and in the individual
characteristics of the tissue from one human donor to another. Even the determina-
tion of a straightforward property like the water content has led to variable results
(Tables B7.1 and B7.2), likely due both to nonuniform distribution of the water in
the lens and to the variability of methods employed for measurements. While data
on the inorganic content of the human lens (Table B7.3) are generally in agreement,
there was a larger variation in reporting the organic content. This is presumably due
to the greater sensitivity of the organic metabolism to age and disease. The most
reliable results are included in Table B7.4.
Tables B7.5, B7.6, B7.7, and B7.8 provide some key dimensional properties of
the lens, while Tables B7.9 and B7.10 are focused on perhaps its most important
feature, the optical properties.
Mechanical characteristics of the lens required sophisticated procedures for their
measurement and the data are likely difficult to reproduce. The currently available
mechanical properties of the human lens are summarized in Tables B7.11, B7.12,
B7.13, B7.14, B7.15, B7.16, B7.17, and B7.18. Although the lens was traditionally
regarded as an elastic material, most investigators agree now that it is in fact a vis-
coelastic solid. The first rheological measurements on human lenses were reported
B7 The Intraocular Lens 105

Table B7.1 Water content of the normal human lens

Method Value (%) Source
n.a. 65a [6]
Vacuum dehydration 68.6 ± 4.3b; 63.4 ± 2.9c [7]
Microsectioning 52.5–66.2c,d; 72.5–90d,e [8]
Raman microspectroscopy 69 ± 4b; 65 ± 4c [9]
Freeze drying 68c; 80e; 75f [10]
Raman microspectroscopy 58.0 ± 4.7 (<70 year)c [11]
63.0 ± 2.8 (>70 year)c
85.3 ± 9.4e
80.9 ± 8.3f
± Standard deviation
a
Whole lens
b
Cortex
c
Nucleus
d
Age 62–68 years
e
Outermost anterior cortex
f
Outermost posterior cortex

Table B7.2 Water content of cataractous human lenses

Method Value (%) Source
Drying 67.6a; 75.4b [12]
Freeze-drying 79c,d; 83.5c,e; 78d,f; 87.5e,f; 68d,e,g [11]
Drying 63.8 (<60 year); 67.7 (>80 year) [13]
a
No sclerosis, average age 64.7 years
b
Advanced sclerosis, average age 70.8 years
c
Outermost anterior cortex
d
Primary nuclear cataract
e
Subcapsular cataract
f
Outermost posterior cortex
g
Nucleus

by Itoi et al. [30], who found an apparent elastic modulus of 10–100 kPa and a
loss tangent (phase shift) of 0.3–0.4. More recent studies using rheometry [31]
confirmed the increase of stiffness with age and its variation with the distance from
the center of the lens. We have to mention, however, that the magnitude of the
values reported for the viscoelastic parameters of the human lens varies from one
laboratory to another. Table B7.19 comprises the most recent results determined by
rheometry.
106 T.V. Chirila and S. Suzuki

Table B7.3 Inorganic ion content of the normal adult human lens
Ion Representative value Source
Sodium 91 mg/100 g wet wt [6]
Potassium 170 mg/100 g wet wt [6]
Calcium 1.4 mg/100 g wet wt [6]
Magnesium 0.29 mg/100 g wet wt [6]
6.2 μg/g dry wt [14]
Zinc 21 μg/g dry wt [6]
25 μg/g dry wt [14]
Copper <1 μg/g dry wt [6]
0.6 μg/g dry wt [14]
Manganese 0.2 μg/g dry wt [14]
Iron 0.4 μg/g dry wt [14]
Rubidium 6.8 μ/g dry wt [14]
Chloride 35.3 mg/100 g wet wt [6]
Phosphate 25 mg/100 g wet wt [15]
Sulfate 24 mg/100 g wet wt [15]
pH 7.3–7.7 [16]

Table B7.4 Organic content of the human lensa

Component Representative value Source
Proteins 30 % of lens (young); 35 % of [17]
lens (old)
Ascorbic acid 30 [14]
Glutathioneb 170 (normal lens) [18]
52 (cataractous lens) [18]
46–150 [14]
200–450 [15]
Taurine 10 [14]
6.7 [15]
Alanine 11.9 [14]
Glycine 5.9 [14]
Glutamic acid 50 [14]
Serine 5.9 [14]
Urea 28.2 [14]
Inositol 462 [15]
Cholesterol 1.4 mg/lens [14]
Phospholipids: cortex 600–725c [14]
Phospholipids: nucleus 450–650c [14]
a
Expressed as mg per 100 g wet weight of lens, unless otherwise specified
b
There is a large variation in the reported data on glutathione content
c
Variation with age
B7 The Intraocular Lens 107

Table B7.5 Dimensional variation with age of the human lens [14]
Dimension Value (mm)
Polar diameter (lens thickness) 3.5–5
Equatorial diameter 6.5–9
Anterior radius of curvature 8–14
Posterior radius of curvature 4.5–7.5

Table B7.6 Dimensional variation with age of the equatorial diameter

and thickness of the human lens [19]
Age (years) Equatorial diameter (mm) Thickness (mm)
22 8.90 3.77
31 9.22 3.60
41 9.16 3.62
50 9.37 3.78
62 8.99 4.22
83 8.77 5.31

Table B7.7 Thickness of a 35-year-old human lens capsule [14]

Location Value (μm)
Anterior pole 14
Anterior, maximum 21
Equator 17
Posterior pole 4
Posterior, maximum 23

Table B7.8 Weight, volume, and density of the human lens in adult life [20]
Age interval (years) Weight, mean (mg) Volume, mean (mm3) Density (g/cm3)a
20–30 172.0 162.9 1.034
30–40 190.3 177.3 1.048
40–50 202.4 188.1 1.061
50–60 222.3 205.4 1.072
60–70 230.1 213.0 1.082
70–80 237.1 218.3 1.091
a
Calculated at the beginning of a decade

Table B7.9 Refractive index of the human lens

nucleus as a function of age (adapted from [21])
Age (years) Refractive index
14 1.4325
82 1.4175
108 T.V. Chirila and S. Suzuki

Table B7.10 Transmissivity of the human lens as a function of age [22]

Transmission of radiation (%)
Age (years)
Wavelength (nm) 25 54 82
350 1.2 1.2 1.2
400 4.8 4.8 4.8
450 38.0 30.6 21.7
500 70.0 42.7 30.0
700 75.0 51.7 37.0

Table B7.11 Variation with age of tensile modulus (Young’s

modulus of elasticity) of the decapsulated human lens [23]a
Modulus (kPa)
Age (years) Polar Equatorial
At birth 0.85 0.75
20 1.0 0.75
40 1.5 1.1
63 3.0 3.0
a
Determined by the spinning method

Table B7.12 Variation with age of the shear modulus of the human lens [4]
Shear modulus, G (Pa)a
Age interval (years) Cortex Nucleus
<30b 98.3 ± 64.5 39.0 ± 13.8
>60c 2040 ± 710 17,400 ± 4900
± Standard deviation
a
Determined from penetration measurements
b
Six samples
c
Twelve samples

Table B7.13 Hardness of cataractous human lenses [24]a

Age interval
(years) Mean force (N) Number of lenses
<60 0.80 13
61–70 0.87 20
71–80 1.12 31
>80 1.38 27
a
Measured by the force necessary to cut the lens in a guillotine
B7 The Intraocular Lens 109

Table B7.14 Mechanical properties of the human anterior lens capsule [25]a
Age (years) Property Value
20 Tensile modulus (MPa) 5.6
50 4.0
80 1.5
<20 Ultimate tensile stress (MPa) 2.3
>70 0.7
–b Elongation (%) 29
–b Poisson’s ratio 0.47 ± 0.5
± Standard deviation
a
Determined from the volume–pressure relationship upon distension with iso-
tonic saline
b
Found to be independent of age

Table B7.15 Variation with age of the mechanical properties of the human anterior lens capsule [26]
Ultimate strain Ultimate stress Ultimate elastic
Age (MPa) (MPa) modulus (MPa)
7 months 108 17.5 44.8
98 years 40 1.5 4.4

Table B7.16 Variation with age of the mechanical properties of the human posterior lens capsule [27]
Age Ultimate strain (MPa) Ultimate stress (MPa) Ultimate elastic modulus (MPa)
1 101 16.9 55.7
94 34 1.1 5.4

Table B7.17 Force of contraction for maximum

accommodation of the human lens [28]a
Age (years) Force (mN)
25 1.2
35 10.9
45 12.8
55 11.4
a
Determined from the stress-dioptric power relationship
110 T.V. Chirila and S. Suzuki

Table B7.18 Spring constants of human lens and zonules [29]a

Spring constant at 10 % Age interval (years)
elongation (mN) 2–39 40–70
Lens, polar 13.5 ± 0.80 25.4 ± 0.24
Lens, equatorial 12.3 ± 0.65 36.5 ± 0.23
Zonulesb 0.38 ± 0.32 0.65 ± 0.85
± Standard deviation
a
Spring constant is defined here as S′ = F/(∆1/10), where F is total force and ∆l/10 is elongation
b
Determined on specimens zonule-lens-zonule, after the excision of the ciliary muscles

Table B7.19 Viscoelastic properties of the human lens [32]a

Lens Storage modulus, Loss modulus, Dynamic viscosity, Loss tangent,
region G′ (Pa) G″ (Pa) η′ (Pa s) δ (°)
Nucleusb 11.00 ± 4.67 24.91 ± 10.98 0.33 66.17
Cortexc 12.40 ± 6.85 25.98 ± 12.62 0.35 64.49
± Standard deviation
a
Measured for postmortem lenses in a controlled strain rheometer at 75 Hz
b
Average results from 26 samples
c
Average results from eight samples

Table B7.20 Electrical properties of the isolated human lens

Property Value Source
Potential difference (mV)a 7 [33]
Short-circuit current density (μA/cm2)b 5 [33]
Resistance (kΩ cm2)c 1.5 [33]
Relative permittivity 39.1 ± 0.12 (nucleus)d,e [34]
38.1 ± 0.13 (nucleus)e,f [34]
55.6 ± 0.17 (cortex)d,e [34]
52.1 ± 0.16 (cortex)e,f [34]
Conductivity (S/m) 0.31 ± 0.005 (nucleus)d,e [34]
0.33 ± 0.005 (nucleus)e,f [34]
0.73 ± 0.07 (cortex)d,e [34]
0.72 ± 0.005 (cortex)e,f [34]
a
Anterior side positive
b
Reflects sodium transport from the posterior to the anterior lens side and is expressed as a current
density
c
Calculated as the ratio between potential difference and current density
d
Measured at 20 °C
e
Range given for 95 % confidence interval
f
Measured at 37 °C
B7 The Intraocular Lens 111

Large variations in the electrical properties of animal lenses have been reported,
but it seems that the only measurements performed on human lenses are those
shown in Table B7.20.
Additional Reading
Bellows, J.G. (ed.) (1975) Cataract and Abnormalities of the Lens, Grune &
Stratton, New York.
A valuable collection of 42 contributions on the lens, its pathology, and surgery,
written by known experts such as Barraquer, Bellows, Choyce, Girard, Hockwin,
Kaufman, Rosen, and Yanoff. The first five introductory chapters present historical
aspects, development, and characterization of the lens. However, most of the book
is dedicated to cataract and its treatment.
Spector, A. (1982) Aging of the lens and cataract formation, in Aging and Human
Visual Function (eds R. Sekuler, D. Kline and K. Dismukes), Alan R. Liss, Inc.,
New York, pp. 27–43.
A brief but comprehensive account of the changes which take place in the com-
position and metabolism of the lens during aging and cataractogenesis.
Duncan, G. and Jacob, T.J.C. (1984) The lens as a physicochemical system, in
The Eye, vol. lb, 3rd edn (ed H. Davson), Academic Press, Orlando, FL,
pp. 159–206.
This text develops some topics usually neglected in other books, including the
structural order in the lens, optical properties of the lens, role of lens membranes,
and electrolyte transport and distribution in the lens.
Cotlier, E. (1987) The lens, in Adler’s Physiology of the Eye, 8th edn (eds
R.A. Moses and W.M. Hart), C.V. Mosby Co., St. Louis, pp. 268–290.
A systematic presentation of the anatomy, biochemistry, and physiology of the
lens.
Moses, R.A. (1987) Accommodation, in Adler’s Physiology of the Eye, 8th edn,
(eds R.A. Moses and W.M. Hart), C.V. Mosby Co., St. Louis, pp. 291–310.
A thorough exposition of all aspects of the mechanism of accommodation and
the role of the lens in vision. A text, by now classic, on a topic much more complex
than it appears.
Jones, W.L. (1991) Traumatic injury to the lens. Optom. Clin., 1, 125–42.
This review article analyzes the effects of concussive trauma to the eye, empha-
sizing the types of injuries to the lens. The mechanical response of the anterior and
posterior segments of the eye to external forces is also described.
Zampighi, G.A. (2006) The lens, in The Biology of the Eye, (ed J. Fischbarg),
Elsevier, Amsterdam, pp. 149–79.
A concise but informative chapter on relevant aspects of the human lens, includ-
ing its cellular architecture and the mechanisms of the molecular processes and
fluxes.
Beebe, D.C. (2011) The lens, in Adler’s Physiology of the Eye, 11th edn (eds
L.A. Levin et al.), Elsevier, Edinburgh, pp. 131–63.
Written by an expert, and beautifully illustrated, this text covers the current
knowledge on the human lens and is supported by almost 500 references.
Recommended sources for ophthalmic terminology:
112 T.V. Chirila and S. Suzuki

Cassin, B., Solomon, S.A.B. and Rubin, M.L. (1990) Dictionary of Eye
Terminology, 2nd edn, Triad Publishing Co., Gainesville, FL, 286 pp.
Stein, H.A., Slatt, B.J. and Stein, R.M. (1992) Ophthalmic Terminology. Speller
and Vocabulary Builder, 3rd edn, Mosby-Year Book Inc., St. Louis, MO, pp. 3–33,
243–257.
Myles, W.M. (1993) Ophthalmic etymology. Surv. Ophthalmol., 37, 306–9.

References

1. Grom E (1975) History of the crystalline lens. In: Bellows JG (ed) Cataract and abnormalities
of the lens. Grune & Stratton, New York, pp 1–28
2. Smith P (1883) Diseases of crystalline lens and capsule. 1. On the growth of the crystalline
lens. Trans Ophthalmol Soc UK 3:79–99
3. Burd HJ, Judge SJ, Flavell MJ (1999) Mechanics of accommodation of the human eye. Vision
Res 39:1591–1595
4. Heys KR, Cram SL, Truscott JW (2004) Massive increase in stiffness of the human lens
nucleus with age: the basis for presbyopia? Mol Vis 10:956–963
5. Wilde GS (2011) Measurement of human lens stiffness for modelling presbyopic treatments.
Ph.D. Thesis, University of Oxford, 219 pp. http://www.eng.ox.ac.uk/civil/publications/the-
ses/wilde.pdf. Accessed 12 Feb 2014
6. Kuck JFR (1970) Chemical constituents of the lens. In: Graymore CN (ed) Biochemistry of the
eye. Academic, London, pp 183–260
7. Fisher RF, Pettet BE (1973) Presbyopia and the water content of the human crystalline lens.
J Physiol 234:443–447
8. Bours J, Fodisch HJ, Hockwin O (1987) Age-related changes in water and crystalline content
of the fetal and adult human lens, demonstrated by a microsectioning technique. Ophthalmic
Res 19:235–239
9. Huizinga A, Bot ACC, de Mul FFM, Vrensen GFJM, Greve J (1989) Local variation in abso-
lute water content of human and rabbit eye lenses measured by Raman microspectroscopy. Exp
Eye Res 48:487–496
10. Deussen A, Pau H (1989) Regional water content of clear and cataractous human lenses.
Ophthalmic Res 21:374–380
11. Siebinga I, Vrensen GFJM, de Mul FFM, Greve J (1991) Age-related changes in local water
and protein content of human eye lenses measured by Raman microspectroscopy. Exp Eye Res
53:233–239
12. Salit PW (1943) Mineral constituents of sclerosed human lenses. Arch Ophthalmol
30:255–258
13. Tabandeh H, Thompson GM, Heyworth P, Dorey S, Woods AJ, Lynch D (1994) Water content,
lens hardness and cataract appearance. Eye 8:125–129
14. Harding JJ, Crabbe MJC (1984) The lens: development, proteins, metabolism and cataract. In:
Davson H (ed) The eye, vol lb, 3rd edn. Academic, Orlando, FL, pp 207–492
15. Paterson CA (1985) Crystalline lens. In: Duane TD, Jaeger EA (eds) Biomedical foundations
of ophthalmology, vol 2, 2nd edn. Harper & Row, Philadelphia, Chapter 10
16. Kuck JFR (1970) Metabolism of the lens. In: Graymore CN (ed) Biochemistry of the eye.
Academic, London, pp 261–318
17. Davson H (1990) Physiology of the eye, 5th edn. Macmillan, London, Chapter 4
18. Dische Z, Zil H (1951) Studies on the oxidation of cysteine to cystine in lens proteins during
cataract formation. Am J Ophthalmol 34:104–113
B7 The Intraocular Lens 113

19. Strenk SA, Semmlow JL, Strenk LM, Munoz P, Gronlund-Jacob J, DeMarco JK (1999) Age-
related changes in human ciliary muscle and lens: a magnetic resonance imaging study. Invest
Ophthalmol Vis Sci 40:1162–1169
20. Scammon RE, Hesdorfer MB (1937) Growth in mass and volume of the human lens in postna-
tal life. Arch Ophthalmol 17:104–112
21. Moffat BA, Atchison DA, Pope JM (2002) Explanation of the lens paradox. Optom Vis Sci
79:148–150
22. Lerman S (1987) Chemical and physical properties of the normal and ageing lens: spectro-
scopic (UV, fluorescence, phosphorescence, and NMR) analyses. Am J Optom Physiol Optics
64:11–22
23. Fisher RF (1971) The elastic constants of the human lens. J Physiol 212:147–180
24. Heyworth P, Thompson GM, Tabandeh H, McGuigan S (1993) The relationship between clini-
cal classification of cataract and lens hardness. Eye 7:726–730
25. Fisher RF (1969) Elastic constants of the human lens capsule. J Physiol 201:1–19
26. Krag S, Olsen T, Andreassen TT (1997) Biomechanical characteristics of the human anterior
lens capsule in relation to age. Invest Ophthalmol Vis Sci 38:357–363
27. Krag S, Andreassen TT (2003) Mechanical properties of the human posterior lens capsule.
Invest Ophthalmol Vis Sci 44:691–696
28. Fisher RF (1977) The force of contraction of the human ciliary muscle during accommodation.
J Physiol 270:51–74
29. van Alphen GWHM, Graebel WP (1991) Elasticity of tissues involved in accommodation.
Vision Res 31:1417–1438
30. Itoi M, Ito N, Kaneko H (1965) Visco-elastic properties of the lens. Exp Eye Res 4:168–173
31. Weeber HA, Eckert G, Pechhold W, van der Heijde RGL (2007) Stiffness gradient in the crys-
talline lens. Graefe’s Arch Clin Exp Ophthalmol 245:1357–1366
32. Schachar RA, Chan RW, Fu M (2011) Viscoelastic properties of fresh human lenses under 40
years of age: implications for the aetiology of presbyopia. Br J Ophthalmol 95:1010–1013
33. Platsch KD, Wiederholt M (1981) Effect of ion substitution and ouabain on short circuit cur-
rent in the isolated human and rabbit lens. Exp Eye Res 32:615–625
34. Gabriel C (1996) Compilation of the dielectric properties of body tissue at RF and microwave
frequencies. Final Technical Report AL/OE-TR-1996-0004, Prepared for Brooks Air Force
Base, Texas, 16 pp. www.dtic.mil/cgi-bin/GetTRDoc?AD = ADA303903. Accessed 27 Feb
2014
Chapter C1
Blood and Related Fluids

V. Turitto and S.M. Slack

C1.1—Introduction

This section provides data for several human biological fluids including blood,
plasma or serum, cerebrospinal (CS) fluid, lymph, synovial fluid, and tear fluid. The
material presented here was gleaned from a variety of sources, with emphasis placed
on the most recently published work, and includes physicochemical properties
(Table C1.1), cellular compositions (Table C1.2), concentrations of inorganics
(Table C1.3), organics (Table C1.4), and major proteins (Table C1.5). In addition,
various properties of the major proteins are presented in Table C1.7, while Tables
C1.8 and C1.9 contain information regarding the components of the coagulation
and complement cascades, respectively. Because of the variability in values for
many properties of biological fluids, in many cases a normal singular range of such
values is listed. In all cases, the data are those compiled for normal human adults
and, where possible, differences with respect to gender are included. It must be
stressed that fluid properties can readily change as a result of disease, aging, or drug
ingestion.
The following equations can be used to estimate blood volume (BV, mL), eryth-
rocyte volume (EV, mL), and plasma volume (PV, mL) from the known body mass
(b, kg) with a coefficient of variation of approximately 10%:

V. Turitto (*)
Department of Biomedical Engineering, University of Memphis, Memphis, TN 38152, USA
S.M. Slack
Department of Biomedical Engineering, University of Memphis,
Campus Box 526582, Memphis, TN 38152-6502, USA

© Springer Science+Business Media New York 2016 115

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_11
116 V. Turitto and S.M. Slack

Table C1.1 Physiochemical properties [1–3]

Property Whole blood Plasma (serum)
Dielectric constant 8.0–8.5 –
Freezing point Depression (°C) 0.557–0.577 0.512–0.568
Osmolality (mosmlkg) – 276–295
pH 7.38–7.42 7.39–7.45
Refractive index – 1.3485–1.3513
Relative viscosity 2.18–3.59 1.18–1.59
Specific gravity 1.052–1.061 1.022–1.026
Specific conductivity (S/cm) – 0.0117–0.0123
Specific heat (cal/g/°C) 0.87 0.94
Surface tension (dyne/cm) 55.5–61.2 56.2

Property Synovial fluid CS fluid Tear fluid

Dielectric constant – – –
Freezing Point Depression (°C) – 0.540–0.603 0.572–0.642
Osmolality (mosmlkg) 292–300 290–324 309–347
pH 7.29–7.45 7.35–7.70 7.3–7.7
Refractive index – 1.3349–1.3351 1.3361–1.3379
Relative viscosity > 300 1.020–1.027 1.26–1.32
Specific gravity 1.008–1.015 1.0032–1.0048 1.004–1.005
Specific conductivity (S/cm) 0.0119 – –
Specific heat (cal/g/°C) – – –
Surface tension (dyne/cm) – 60.0–63.0 –
The refractive index, specific gravity, and surface tension were measured at 20°C, the specific
conductivity at 25°C, and the relative viscosity at 37°C. The specific gravity is that relative to
water. The viscosity of serum is slightly less than that of plasma due to the absence of fibrinogen.
Blood viscosity depends strongly on shear rate and hematocrit and the value given in Table C1.1 is
that at high shear rates(>200s-1) and normal hematocrits (40–45%). Blood is a non-Newtonian fluid
and exhibits increased viscosity with decreasing shear rate. Correlations relating blood viscosity to
hematocrit, shear rate, and protein content have been described in the literature [4, 5]. The reader
is referred to several excellent publications for further details regarding factors affecting blood
viscosity [4, 6–11].

These equations, relating BY, PY, and EY to body weight, are taken from
Lentner [12].

Males (M) BY = 41.0 x b + 1530 Females (F) BY = 47.16 x b + 864

PY = 19.6 x b + 1050 PY = 28.89 x b + 455
EY = 21.4 x b + 490 EY = 18.26 x b + 409

Additional correlations relating these volumes to body weight and surface area
are available from the same source.
C1 Blood and Related Fluids 117

Table C1.2 Cellular composition of biological fluids [12]

A. Whole blood
Whole blood:
Cell type Cells/μL Cell size (μm) Half-life in circulation
Erythrocytes 4.6–6.2 x 106 (M) 7–8 25 ± 2 days
4.2–5.2 x 106 (F)
Leukocytes
Neutrophils 3000–6500 10–15 6–8 hours
Eosinophils 50–250 10–15 8–12 hours
Basophils 15–50 10–15 ?
Monocytes 300–500 12–20 1–3 days
Lymphocytes 1000–3000 7–8 Variable
Platelets 1.5–3.5 x 105 2–4 3.2–5.2 days
Reticulocytes 2.3–9.3 x 1()4 7–10 –
Synovial fluid:
Cell type Celis/μL
Leukocytes 4–5
Monocytes 35–40
Lymphocytes 15–16
Synovial cells 2–3
The variability in the half-life of circulating lymphocytes is a result of the many subsets of this cell
type, e.g., B-cells, helper and suppressor T-cells, etc. Cerebrospinal fluid also contains ~ 1–5 cells/
μL, primarily lymphocytes.

Table C1.3 Inorganic content of various fluids [1]

Compound Whole blood Plasma (serum) Synovial fluid
Bicarbonate 19.1–22.7 25–30 –
Bromide 0.033–0.074 0.043–0.093 –
Calcium 2.42 2.12–2.72 1.2–2.4
Chloride 77–86 100–108 87–138
Copper (μM) 11.3–19.5 13–22 –
Fluoride (μM) 5.3–23.7 – –
Iodine (μM) 0.2–1.34 0.30–0.47 –
Iron 7.5–10.0 0.01–0.027 –
Magnesium 1.48–1.85 0.7–0.86 –
Phosphorous (total) 10.1–14.3 2.87–4.81 –
Potassium 40–60 3.5–4.7 3.5–4.5
Sodium 79–91 134–143 133–139
Zinc 0.076–0.196 0.011–0.023 –
(continued)
118 V. Turitto and S.M. Slack

Table C1.3 (continued)

Compound Cerebrospinal fluid Tear fluid Lymph
Bicarbonate 18.6–25.0 20–40 –
Bromide 0.018–0.048 – –
Calcium 1.02–1.34 0.35–0.77 1.7–2.8
Chloride 119–131 110–135 87–103
Copper (μM) 0.13–0.37 – –
Fluoride (μM) 55 – –
Iodine (μM) – – –
Iron 0.0003–0.0015 – –
Magnesium 0.55–1.23 – –
Phosphorous (total) 0.442–0.694 – 2.0–3.6
Potassium 2.62–3.30 6.6–25.8 3.9–5.6
Sodium 137–153 126–166 118–132
Zinc – – –
Concentrations are in mM, unless otherwise specified.

Table C1.4 Organic content of various fluids [13–15]

Species Whole blood Plasma (serum) CS fluid
Amino acids (mg/L) 48–74 20–51 10–15
Ammonia (mg/L) 0.26–0.69 0.22–0.47 0.14–0.26
Bilirubin (mg/L) 2–14 2–8 <0.1
Cholesterol 1.15–2.25 1.7–2.1 –
Creatine (mg/L) 3–5 1.3–7.7 4.6—19
Creatinine (mg/L) 10–20 5.6–10.5 6.5–10.5
Fat, neutral 0.85–2.35 0.25–2.6 trace
Fatty acids 2.5–3.9 3.5–4.0 trace
Glucose 630–870 650–966 430–640
Hyaluronic acid – – –
Lipids, total 4.45–6.1 2.85–6.75 0.01–0.02
Total nitrogen 30–41 12–14.3 0.16–0.22
Nonprotein nitrogen 0.26–0.50 0.14–0.32 0.11–0.20
Phospholipid 2.25–2.85 2.0–2.5 0.002–0.01
Urea 0.166–0.39 0.18–0.43 0.14–0.36
Uric acid (mg/L) 6—50 30.5–70.7 1.1–6.3
Water 830–865 930–955 980–990
(continued)
C1 Blood and Related Fluids 119

Table C1.4 (continued)

Species Synovial fluid Tear fluid Lymph
Amino acids (mg/L) – – –
Ammonia (mg/L) – 50 –
Bilirubin (mg/L) – – 8
Cholesterol – – 0.34–1.06
Creatine (mg/L) – – –
Creatinine (mg/L) – – 8–89
Fat, neutral – – –
Fatty acids – – –
Glucose – 0.025 1.36–1.40
Hyaluronic acid 3.32 – –
Lipids, total – – –
Total nitrogen 0.084–4.0 1.58 –
Nonprotein nitrogen 0.22–0.43 – 0.13–1.39
Phospholipid – – –
Urea 0.15 0.33–1.4 –
Uric acid (mg/L) 39 – 17–108
Water 960–988 982 810–860
Concentrations are in mg/mL, unless otherwise specified.

Table C1.5 Major protein content of various fluids [12, 14]

Protein Plasma (serum) CS fluid1 Synovial fluid
Albumin 37.6–54.9 155±39 6–10
α1-Acid glycoprotein 0.48–1.26 1.85±0.74 –
(orosomucoid)
α1-Antitrypsin 0.98–2.45 7.0±3.0 0.78±0.017
β2-Microglobulin 0.58–2.24 0.1–1.9 –
Haptoglobin 2.24±1.5 0.1
Type 1.1 1.45±0.34
Type 2.1 2.06±0.67
Type 2.2 1.74±0.70
Ceruloplasmin 0.09–0.51 0.88±0.21 0.043±0.016
Transferrin 1.52–3.36 8.42±3.5 –
C1 Inhibitor 0.15–0.35 – –
α2-Macroglobulin 1.45–4.43 4.64±1.84 0.31–0.21
IgA 0.7–3.12 2.26±0.95 0.62–1.15
IgG 6.4–13.5 13.9±6.6 1.47–4.62
IgM 0.56–3.52 – 0.09–0.22
Fibrinogen 2–4 0.65 –
Lysozyme – – –
Fibronectin 0.09–0.25 – –
Hemopexin 0.53–1.21 – –
(continued)
120 V. Turitto and S.M. Slack

Table C1.5 (continued)

Protein Synovial fluid Tear fluid
Albumin 3.94 15–26.7
α1-Acid glycoprotein – –
(orosomucoid)
α1-Antitrypsin 0.015 –
β2-Microglobulin – –
Haptoglobin – –
Ceruloplasmin 0.04 –
Transferrin – –
C1 Inhibitor – –
α2-Macroglobulin – –
IgA 0.04–0.80 –
IgG 0.04–0.62 7.8
IgM trace –
Fibrinogen – –
Lysozyme 1–2.8 –
Fibronectin – –
Hemopexin – –
1
Protein concentrations in CS fluid are given in mg/L.
All others have units of mg/mL.

Table C1.6 Fluid volumes [16]

Fluid Volume Male (mL) Volume Female (mL)
Whole blood 4490 3600
Erythrocytes 2030 1470
Plasma 2460 2130
Cerebrospinal fluid 100–160 100–160
Tear fluid 4–13 4–13
C1 Blood and Related Fluids 121

Table C1.7 Properties of major plasma proteins [16, 17]

Plasma
concentration Molecular
Protein (mg/mL) weight (Da) p1 S1 D2
Prealbumin 0.12–0.39 54 980 4.7 4.2 –
Albumin 38–52 66 500 4.9 4.6 6.1
α1 - Acid Glycoprotein 0.5–1.5 44000 2.7 3.1 5.3
(Orosomucoid)
α1 - Antitrypsin 2.0–4.0 54 000 4.0 3.5 5.2
α2 - Macroglobulin 1.5–4.5 725 000 5.4 19.6 2.4
α2 - Haptoglobin
Type 1.1 1.0–2.2 100 000 4.1 4.4 4.7
Type 2.1 1.6–3.0 200 000 4.1 4.3– –
6.5
Type 2.2 1.2–2.6 400 000 – 7.5 –
α2 - Ceruloplasmin 0.15-.60 160 000 4.4 7.08 3.76
Transferrin 2.0–3.2 76 500 5.9 5.5 5.0
Hemopexin 0.56–0.89 57 000 5.8 4.8 –
Lipoproteins 5.5–6 140 000–20 000 – – 5.4
000
IgA (Monomer) 1.4–4.2 162 000 – 7 3.4
IgG 6–17 150 000 6.3– 6.5– 4.0
7.3 7.0
IgM 0.5–1.9 950 000 – 18–20 2.6
C1q 0.05–0.1 459 000 – 11.1 –
C3 1.5–1.7 185 000 6.1– 9.5 4.5
6.8
C4 0.3–0.6 200 000 – 10.0 –
Fibrinogen 2.0–4.0 340 000 5.5 7.6 1.97
Plasma concentration Half-life
Protein (mg/mL) E2803 V204 CH2O5 (days)
Prealbumin 0.12–0.39 14.1 0.74 – 1.9
Albumin 38–52 5.8 0.733 0 17–23
α1 - Acid glycoprotein 0.5–1.5 8.9 0.675 41.4 5.2
(orosomucoid)
α1 - Antitrypsin 2.0–4.0 5.3 0.646 12.2 3.9
α2 - Macroglobulin 1.5–4.5 8.1 0.735 8.4 7.8
α2 - Haptoglobin
Type 1.1 1.0–2.2 12.0 0.766 19.3 2–4
Type 2.1 1.6–3.0 12.2 – –
Type 2.2 1.2–2.6 – – –
α2 - Ceruloplasmin 0.15-.60 14.9 0.713 8 4.3
Transferrin 2.0–3.2 11.2 0.758 5.9 7–10
Hemopexin 0.56–0.89 19.7 0.702 23.0 9.5
(continued)
122 V. Turitto and S.M. Slack

Table C1.7 (continued)

Plasma concentration Half-life
Protein (mg/mL) E2803 V204 CH2O5 (days)
Lipoproteins 5.5–6 – – – –
IgA (Monomer) 1.4–4.2 13.4 0.725 7.5 5–6.5
IgG 6–17 13.8 0.739 2.9 20–21
IgM 0.5–1.9 13.3 0.724 12 5.1
C1q 0.05–0.1 6.82 – 8 –
C3 1.5–1.7 – 0.736 – –
C4 0.3–0.6 – – – –
Fibrinogen 2.0–4.0 13.6 0.723 2.5 3.1–3.4
1
Sedimentation constant in water at 20°C, expressed in Svedberg units.
2
Diffusion coefficient in water at 20°C, expressed in 10-7 cm2/s.
3
Extinction coefficient for light of wavelength 280 nm traveling 1 em through a 10 mg/ml protein
solution.
4
Partial specific volume of the protein at 20°C, expressed as ml g-l
5
Carbohydrate content of the protein, expressed as the percentage by mass.

Table C1.8 Proteins involved in blood coagulation [19]

Plasma concentration Relative molecular Biological
Protein (μg/mL) weight, Mr (Da) half-life t1/2 (hr)
Fibrinogen 2000–4000 340 000 72–120
Prothrombin 70–140 71 600 48–72
Factor III (tissue factor) – 45 000 –
Factor V 4–14 330 000 12–15
Factor VII trace 50 2–5
Factor VIII ~0.2 330 000 8–12
Factor IX ~5.0 330 000 24
Factor X ~12 58 800 24–40
Factor XI 2.0–7.0 160 000 48–84
Factor XII 15–47 80 000 50–60
Factor XIII ~10 320 000 216–240
Protein C ~4.0 62 000 10
Protein S ~22 77 000 –
Protein Z ~2.9 62 000 60
Prekallikrein 35–50 85 000 –
High molecular 70–90 120 000 –
weight kininogen
α1 - Protease inhibitor 2500 55 000 –
Antithrombin III 230 ± 23 58 000 67
C1 Blood and Related Fluids 123

Table C1.9 Proteins in the compliment system

Protein Serum concentration Relative molecular weight, Sedimentation constant
(mg/L) Mr (Da) S2oW(10-13S)
C1q 70 ± 14 459 000 11.1
C1r 39 ± 2 83 000 7.5
C1s 36 ± 3 83 000 4.5
C2 27 ± 5.6 108 000 4.5
C3 1612 ± 244 185 000 9.5
C4 498 ± 151 200 000 10.0
C5 153 ± 29 185 000 8.7
C6 50.9 ± 8 128 000 5.5
C7 4–60 121 000 6.0
C8 43.2 ± 6.5 151 000 8.0
C9 57.5 ± 12.7 71 000 4.5
Factor B 275 ±55 92 000 5–6
Factor D trace 24 000 3.0
Properdin 28.4 ± 5 220 000 5.4
C1 inhibitor 158 ± 14 100 000 –
Factor H 525 ±58 150 000 6.0
Factor I 38.6 ± 5.5 88 000 5.5

Additional Reading

Ditmer, D.S. (ed.) (1961) Blood and Other Body Fluids, Federation of American
Societies for Experimental Biology, Washington, D.C.
This text provides a thorough compilation of the physical properties and compo-
sition of numerous biological fluids. Unlike the Geigy Scientific Tables, this book
also reports data for many non-human species. However, citations and some mea-
surement techniques are somewhat outdated.
Kjeldsberg C.R. and Knight J.A. (eds) (1993) Body Fluids: Laboratory
Examination of Amniotic, Cerebrospinal, Seminal, Serous & Synovial Fluids, 3rd
ed., American Society of Clinical Pathologists, Chicago.
An excellent source of information, especially for a clinician or medical tech-
nologist. Includes numerous color photographs of fluids and cells. Discusses abnor-
mal amounts or types of specific proteins and cells in fluids as potentially diagnostic
of disease states.
Lentner, C. (ed.) (1984) Geigy Scientific Tables, Ciba-Geigy, Basle.
This is the most comprehensive source of information available on properties
and composition of body fluids. Volumes 1 and 3 provide extensive data, generally
in tabular form, on fluid content (as well as measurement technique), related to
gender, age and disease state.
124 V. Turitto and S.M. Slack

References

1. Ditmer, D.S. (ed.) (1961) Blood and Other Body Fluids, Federation of American Societies for
Experimental Biology, Washington, D.C.
2. Fullard, R.J. (1988) Current Eye Research, 7, 163–179.
3. Chmiel, H. and Walitza, E. (1980) On the Rheology of Blood and Synovial Fluids, Research
Studies Press, New York.
4. Barbanel, J.C., Lowe, G.D.O. and Forbes, C.D. (1984), The viscosity of blood. in Mathematics
in Medicine and Biomechanics, G.F. Roach (ed.), Shiva Publications, Nantwich, p. 19.
5. Begg, T.B. and Hearns, J.B. (1966) Components in blood viscosity: The relative contribution
of hematocrit, plasma fibrinogen and other proteins. Clinical Science, 31, 87–93.
6. Whitmore, R.L. (1968) Rheology of The Circulation, Pergamon Press, New York.
7. Merrill, E.W. (1969) Rheology of blood. Physiology Reviews, 49, 863–867.
8. Harkness, J. (1971) The viscosity of human blood plasma: Its measurement in health and dis-
ease. Biorheology, 8, 171–193.
9. Lowe, G.D.O., Barbanel, J.C. and Forbes, C.D. (eds) (1981) Clinical Aspects of Blood Viscosity
and Cell Deformability, Springer-Verlag, New York.
10. Lowe, G.D.O and Barbanel, J.C. (1988) Plasma and blood viscosity, in Clinical Blood
Rheology, G.D.O. Lowe (ed.), CRC Press, Boca Raton, pp. 11–44.
11. Schmidt-Schonbein, H. (1988) Fluid dynamics and hemorheology, in Clinical Blood Rheology,
G.D.O. Lowe (ed.), CRC Press, Boca Raton, pp. 129–220.
12. Lentner, C. (ed.) (1984) Geigy Scientific Tables, Ciba-Geigy, Basle.
13. Bicks, R.L. (1993) Hematology: Clinical and Laboratory Practice, Mosby, St Louis.
14. Sokoloff , L. (ed.) (1978) The Joints and Synovial Fluid, Academic Press, New York.
15. Hermens, W.T., Willems, G.M. and Visser, M.P. (1982) Quantification of Circulating Proteins:
Theory and Applications Based on Analysis of Plasma Protein Levels, Martinus Nijhoff, The
Hague.
16. Colman, R.W., Hirsh, J., Marder, V.J., et al. (eds) (1993) Hemostasis and Thrombosis,
Lippincott, Philadelphia.
17. Schultze, H.E. and Heremans, J.F. (1966) Nature and Metabolism of Extracellular Proteins,
Elsevier, Amsterdam.
18. Bing, D.H. (ed.) (1978) The Chemistry and Physiology of the Human Plasma Proteins,
Pergamon Press, Boston.
19. Stamatoyannopoulos, G., Nienhuis, A.W., Majerus, P.W., et al. (eds) (1994) The Molecular
Basis of Blood Diseases, W.B. Saunders, Philadelphia.
Chapter C2
The Vitreous Humor

Traian V. Chirila and Ye Hong

C2.1 Introduction

The vitreous humor, also termed vitreous body, vitreus, or vitreous, is a clear and
transparent mass (gel or liquid or a mixture of both) that fills the posterior cavity of
the eye in vertebrates, between the lens and the retina.
The mammalian vitreous humor can be defined as an avascular, virtually acellu-
lar, highly hydrated gel located in the vitreous cavity of the eye and consisting of a
dilute network of heterotypic collagen fibrils surrounded by a mixture of glycosami-
noglycans where hyaluronan is the predominant component. In humans, the vitre-
ous humor is perceived as a gel-like body that can provide an adequate support for
the retina, allows the diffusion of metabolic solutes, and allows the light to reach the
retina. Historically, there are two differing concepts on the nature of vitreous humor.
A significant amount of evidence supports the view that the vitreous humor is basi-
cally an extracellular matrix. Another model has been developed in which the vitre-
ous body is considered as a specialized, but simple, connective tissue. The two
concepts are not yet reconciled; therefore the structure and role of the vitreous
humor are usually regarded from both points of view. It is generally accepted that
this gel-like material possesses a unique macromolecular organization in the form
of a double-network system consisting of a scaffold of randomly spaced rodlike
collagen fibrils surrounded by and entangled with a network of very large coiled-up
macromolecules of hyaluronic acid (hyaluronan). The double-network model
explains satisfactorily most of the properties of the vitreous body, as well as its
remarkable mechanical stability, although it probably overestimates the importance
of hyaluronan. The natural vitreous gel displays true viscoelastic properties which

T.V. Chirila (*)

Queensland Eye Institute, 140 Melbourne Street, South Brisbane, QLD 4101, Australia
e-mail: [email protected]
Y. Hong
Cooper Vision, Inc., Pleasanton, CA 94588, USA

© Springer Science+Business Media New York 2016 125

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_12
126 T.V. Chirila and Y. Hong

enable it to resist sudden compression shocks, offering much the best protection for
the retina against contusion trauma. It is generally believed that the hyaluronan
network imparts the latter feature, while the collagen network is responsible for the
plasticity and tensile strength of the vitreous humor. The above considerations illus-
trate the understanding of the supramolecular organization of the vitreous humor as
presented it in the first edition of this book [1]. Since that time, there was not much
progress in this understanding [2]. However, our appreciation of the physiological
role and functions of the vitreous humor in the eye has improved significantly [3, 4].
Further research revealed its central role in major diseases of the eye such as dia-
betic retinopathy, retinal vein occlusion, age-related macular degeneration, nuclear
sclerotic cataract, and primary open-angle glaucoma [5].
Biochemically, the vitreous humor consists of collagen types I, V, IX, and XI,
non-collagenous proteins (opticin, fibrillin-1, VIT1, fibronectin, proteoglycans, and
others), and glycosaminoglycans [3, 6]. There has been considerable controversy
regarding the existence of direct interactions between hyaluronan and the collagen
fibrils, and the role of the former in maintaining the long-range spacing in the col-
lagen network. There is no clear evidence for such interactions, but some experi-
ments have suggested [7] that about 6 % of the total hyaluronan may be involved in
maintaining the vitreous gel in a distended state. However, it is rather believed [3]
that, due to indirect interfibrillar interactions, the collagen network has sufficient
mechanical resilience to support itself in solution. It was suggested that the vitreous
humor plays only a minor role in the regulation of intraocular pressure [8].
The role of the vitreous humor and its physiological functions in the human eye
can be summarized as such: (a) contributes to the growth of the eye and to maintain-
ing its volume, elasticity, and resilience, and provides protection during mechanical
trauma; (b) contributes to the total transparency of the ocular pathways; (c) serves
as a support for the intraocular lens and contributes to the accommodation; (d) acts
as a repository and transport conduit for the substances involved in the metabolism
of the surrounding ocular tissues; (e) acts as a barrier to biomolecules, biomacro-
molecules, and cells, and as an inhibitor of inflammation and neovascularization; (f)
prevents the onset of certain types of cataract by protecting the intraocular lens
against oxygen-induced damage; and (g) plays a significant role in the etiology of a
number of major ocular diseases.

C2.1.1 Compendium of Physicochemical Characteristics

of the Human Vitreous Humor

As is the case with many other structural elements of the eye, there are presently
much more data on animal vitreous humor than on the human counterpart. The
numerical data tabulated here cover almost everything reported so far on the analy-
sis of the human vitreous. However, in many cases it is not possible to select a most
reliable single value; dependable values representing a range are thus provided in
several of the following tables, illustrating the large variability in the experimental
methodology from one laboratory to another, in collecting and processing the
C2 The Vitreous Humor 127

samples, and in the individual, inherent characteristics of samples as such. Also,

considering that the vitreous samples processed and subjected to analysis were all
postmortem, differences are expected in some characteristics as compared to the
antemortem vitreous humor.
Being a very loose tissue, albeit well structured in a highly specialized way, the
vitreous body becomes a homogeneous fluid during processing for measurements;
therefore, the resulting data for its properties illustrate rather the behavior of a
fluid consisting mainly of water and containing minute amounts of inorganic and
organic components. The dimensional characteristics, bulk chemical composi-
tion, and optical properties are seemingly not affected by the homogenization
process. However, some investigators considered the separate existence of gel and
liquid fractions in the vitreous.
Over the past decades, the vitreous body was perceived as a typical viscoelastic
material. Accounts of the earlier rheological studies of the mammalian vitreous
humor have been summarized elsewhere [32, 33]. Since the first edition of this
book, a number of rheological studies [34–39] have been reported on the animal
vitreous, but almost none regarding the human vitreous. We can say that its charac-
terization by rheometry (Table C2.9) is still in its infancy, perhaps due to the large
sample variability. Indeed, in ten postmortem eyes, between the ages of 21 and 65
years, a range of 2–38 Pa was measured experimentally for the storage modulus G′
[41], with no clear dependence upon age and with sample-to-sample variation
observed even in two humors harvested from the same donor.

Table C2.1 Physical properties of the human vitreous humor

Property Value Source
Volume 3.9 mL [9]
Weight 3.9 g [9]
Water content 99.7 % [10]
99 % [11]
pH 7.5 [12]
7.4–7.52 [13]
7–7.3 [14]
Osmolality 288–323 mOsm/kg [15]
Osmotic pressure
(freezing-point depression) −0.554 to −0.518 °C [12]
Density 1.0053–1.0089 g/cm3 [16]
Intrinsic viscosity 3–5 × 103 cm3/g [17]
Dynamic viscosity 1.6 cP [18]
Refractive index 1.3345 [19]
1.3345–1.337 [9]
128 T.V. Chirila and Y. Hong

Table C2.2 Inorganic ion content of the human vitreous humor

Ion Representative value Source
Sodium 2.714–3.542 mg/cm3 [23]
135–151 mEq/L [24]
3.15 g/kg water [11]
2.603–5.805 mg/cm3 [14]
Potassium 130–470 μg/cm3 [23]
0.15 g/kg H2O [11]
4.2–7.2 mEq/L [24]
308–788 μg/cm3 [14]
Calcium 56–106 μg/cm3 [23]
6.0–8.0 mg/100 mL [24]
14–76 μg/cm3 [14]
Phosphate 0.1–3.3 mEq/dm3 [23]
Chloride 3.155–5.140 mg/cm3 [23]
108–132 mEq/L [24]
4 g/kg water [11]
3.477–7.621 mg/cm3 [14]
Bicarbonate 1.2–3.0 g/kg water [11]

Table C2.3 Organic content of the human vitreous humor

(low-molecular-weight components)
Component Representative value Source
Lipids 2 μg/mL [25]
Glucose 17–105 mg/dL water [23]
37–180 mg/100 mL [24]
30–70 mg/dL water [11]
Lactic acid 70 mg/dL water [11]
Urea 24–172 mg/dL water [23]
Creatinine 0.3–3.0 mg/dL water [23]
Citrate 1.9 mg/dL water [11]
Pyruvic acid 7.3 mg/dL water [11]
Ascorbic acid 36 mg/100 g [26]
C2 The Vitreous Humor 129

Table C2.4 Organic content of the human vitreous humor (high-molecular-weight components)
Component Representative value (μg/cm3) Source
Proteinsa 280–1360 [25]
450–1100 [27]
Hyaluronan 100–400 [27]
42–399 [14]
65–210 [28]
Versicanb 60 [29]
Collagen 40–120 [27]
30–532 [14]
Albumin 293 ± 18 [30]
Immunoglobulin (IgG) 33.5 ± 3 [30]
α1-Antitrypsin 141 ± 2.9 [30]
α1-Acid glycoprotein 4 ± 0.7 [30]
(±) represents standard deviation
a
Total protein content
b
A high-molecular-weight proteoglycan based on chondroitin sulfate

Table C2.5 Variation with age of total protein

content in the liquid fraction of human vitreous [27]
Age range (years) Protein (mg/cm3)
10–50 0.4–0.6
50–80 0.7–0.9
>80 0.9–1.0

Table C2.6 Axial length of the human vitreous body

during maturation [20]a
Age (years) and gender Axial length (mm)
<13, male 10.48
<13, female 10.22
>13, male 16.09
>13, female 15.59
a
The axial growth of the vitreous body is essentially
completed by the age of 13 years
130 T.V. Chirila and Y. Hong

Table C2.7 Gel and liquid volume of the human vitreous as a function of age
(adapted from [21])a
Age (years) Gel volume (cm3) Liquid volume (cm3)
Birth 1.6 0
5 3.3 0
10 3.5 0.7
20 3.9 0.9
30 3.9 0.9
40 3.9 0.9
50 3.5 1.3
60 3.2 1.6
70 2.8 2.0
80 2.5 2.3
90 2.2 2.6
a
The liquid vitreous appears first in childhood and by the seventh decade it
occupies half of the vitreous [21, 22]

Table C2.8 Transmission of radiation through the human

vitreous humor (adapted from [31])
Wavelength (nm) Transmittance (total, %)
300 0
325 76
350 82
400 90
500 97
600 98
700 98

Table C2.9 Rheological characteristics of the human vitreous humor [40]

Anterior Region in the vitreous
Parameter Central Posterior
Residual viscosity 1.4 2.2 4.9
ηm (Pa s)
Internal viscosity 0.3 0.35 0.5
ηk (Pa s)
Relaxation time 0.38 0.30 1.61
τm (s)
Retardation time 0.27 0.41 0.46
τk (s)
Elastic compliance,
instantaneous, Jm (m−2 N−1) 0.1 0.3 0.3
Elastic modulus,
internal, Gk (Pa) 2.5 1.3 1.2
C2 The Vitreous Humor 131

Additional Reading

Balazs, E.A. (1968) The molecular biology of the vitreous, in New and Controversial
Aspects of Retinal Detachment (ed A. McPherson), Harper & Row, New York,
pp. 3–15.
This is a landmark paper on the nature of the vitreous body, describing the
“mechanochemical” (or “double-network”) model. This model explains satisfacto-
rily the correlations between some properties of the vitreous (composition, rheol-
ogy, volume, cell population, transparency) and the physicochemical principles
governing its stability (frictional interaction, expansion/contraction, the excluded-
volume concept, and the molecular-sieve effect).
Berman, E.R. and Voaden, M. (1970) The vitreous body, in Biochemistry of the
Eye (ed C.N. Graymore), Academic Press, London, pp. 373–471.
A comprehensive summary of knowledge at that time on animal and human vit-
reous humor, including development, chemical composition, metabolism, and aging
effects.
Shields, J. A. (1976) Pathology of the vitreous, in Current Concepts of the
Vitreous Including Vitrectomy (ed K.A. Gitter), C.V. Mosby Co., St. Louis,
pp. 14–42.
This book chapter presents competently the pathologic vitreous, including devel-
opmental abnormalities, inflammation, hemorrhage, effects of trauma, systemic dis-
eases, and degenerative processes.
Gloor, B.P. (1987) The vitreous, in Adler’s Physiology of the Eye, 8th edn (eds
R.A. Moses and W.M. Hart), C.V. Mosby Co., St. Louis, pp. 246–267.
A concise description of all aspects of the vitreous humor, including properties,
development, anatomy, structure, biochemistry, metabolism, and pathology.
Sebag, J. (1989) The Vitreous. Structure, Function, and Pathology, Springer-
Verlag, New York.
This is probably only the second single-authored book in this century to be dedi-
cated entirely to the topic of vitreous humor, written by the leading scholar in the
field. It is a well-structured and updated compendium. The first half of the book is
dedicated to structure, properties, and physiology of the vitreous. Pathology of the
vitreous is analyzed in the other half from a biological angle. Although a clinician,
the author manages to avoid typical clinical descriptions and to provide a text which
integrates the basic scientific knowledge for both clinicians and scientists.
Williams, G.A. and Blumenkranz, M.S. (1992), Vitreous humor, in Duane’s
Foundations of Clinical Ophthalmology, vol. 2 (eds W. Tasman and E.A. Jaeger),
J.B. Lippincott Co., Philadelphia, chapter 11.
This chapter (27 pages) presents the modern concepts in the pathophysiologic
mechanisms of vitreous diseases, and in the clinical conditions involving the vitre-
ous (detachment, macular holes and membranes, diabetes, proliferative vitreoreti-
nopathy, hyalosis, amyloidosis). Aspects such as separation of the vitreous from the
retina and traction of the vitreous by hypocellular gel contraction are explained
according to the most recent findings.
132 T.V. Chirila and Y. Hong

Lund-Andersen, H., Sebag, J., Sander, B. and la Cour, M. (2006) The vitreous, in
The Biology of the Eye (ed J. Fischbarg), Elsevier, Amsterdam, pp. 181–94.
A concise but very informative text covering all aspects of the vitreous humor.
Lund-Andersen, H. and Sander, B. (2011) The vitreous, in Adler’s Physiology of
the Eye, 11th edn (eds L.A. Levin et al.), Elsevier, Edinburgh, pp. 164–81.
The most recent chapter on the vitreous humor to appear in the traditional text-
book, covering the modern views on this part of the eye, and emphasizing biophysi-
cal and physiological aspects.
Sebag, J. and Green, W.R. (2013) Vitreous and vitreoretinal surface, in Retina,
5th edn (ed S.J. Ryan), vol. 1, part 2, Elsevier, Amsterdam, pp. 482–516.
A concise, update, and beautifully written account of the current knowledge of
the vitreous humor, providing a balanced presentation of the structure, anatomy,
physiology, and pathology of the vitreous and of its interface with the retina. This
chapter in the classical monumental treatise displays also remarkable graphics.
Recommended sources for ophthalmic terminology:
Cassin, B., Solomon, S.A.B. and Rubin, M.L. (1990) Dictionary of Eye
Terminology, 2nd edn, Triad Publishing Co., Gainesville, FL, 286 pp.
Stein, H.A., Slatt, B.J. and Stein, R.M. (1992) Ophthalmic Terminology. Speller
and Vocabulary Builder, 3rd edn, Mosby-Year Book Inc., St. Louis, MO, pp. 3–33,
183–198, 275–278.
Myles, W.M. (1993) Ophthalmic etymology. Surv. Ophthalmol., 37, 306–9.

References

1. Chirila TV, Hong Y (1998) The vitreous humor. In: Black J, Hastings G (eds) Handbook of
biomaterial properties. Chapman & Hall, London, pp 125–131
2. Sebag J (2009) Vitreous: the resplendent enigma. Br J Ophthalmol 93:989–991
3. Bishop PN (2000) Structural macromolecules and supramolecular organization of the vitreous
gel. Prog Retin Eye Res 19:232–244
4. Kleinberg TT, Tzekov RT, Stein L, Ravi N, Kaushal S (2011) Vitreous substitutes: a compre-
hensive review. Surv Ophthalmol 56:300–323
5. Holekamp NM (2010) The vitreous gel: more than meets the eye. Am J Ophthalmol
149:32–36
6. Angi M, Kalirai H, Coupland SE, Damato BE, Semeraro F, Romano MR (2012) Proteomic
analyses of the vitreous humour. Mediat Inflamm 2012; Article ID 148039 (7 pages).
doi:10.1155/2012/148039
7. Bishop PN, McLeod D, Reardon A (1999) Effects of hyaluronan lyase, hyaluronidase, and
chondroitin ABC lyase on mammalian vitreous gel. Invest Ophthalmol Vis Sci
40:2173–2178
8. Collins R, van der Werff TJ (1980) Mathematical models of the dynamics of the human eye.
Lect Notes Biomath 34:5–6
9. Redslob E (1932) Le corps vitré. Masson & Cie, Paris, pp 299–305
10. Duke-Elder WS (1929) The physico-chemical properties of the vitreous body. J Physiol
68:155–165
11. Nordmann J (1968) Chimie. In: Brini A, Bronner A, Gerhard JP et al (eds) Biologie et chirurgie
du corps vitré. Masson & Cie, Paris, pp 95–167
C2 The Vitreous Humor 133

12. Mörner CT (1894) Untersuchung der Proteïnsubstanzen in den licht-brechenden Medien des
Auges. Z Physiol Chem 18:233–256
13. Gala A (1925) Observations on the hydrogen ion concentration in the vitreous body of the eye
with reference to glaucoma. Br J Ophthalmol 9:516–519
14. Lee B (1994) Comparative rheological studies of the vitreous body of the eye. Ph.D. Thesis,
University of Pennsylvania, 1992. U.M.I./Bell & Howell Co., Ann Arbor, MI, pp. 102,
138–152
15. Sturner WQ, Dowdey ABC, Putnam RS, Dempsey JL (1972) Osmolality and other chemical
determinations in postmortem human vitreous humor. J Forensic Sci 17:387–393
16. Visser-Heerema J (1936) Über das spezifische Gewicht der bei der Operation von
Netzhautablösungen gewonnenen Flüssigkeit. Arch Augenheilkd 109:543–561
17. Berman ER, Michaelson IC (1964) The chemical composition of the human vitreous body as
related to age and myopia. Exp Eye Res 3:9–15
18. Shafer DM (1965) Intraocular injections as adjuncts to other retinal detachment procedures.
In: Schepens CL, Regan CDJ (eds) Controversial aspects of the management of retinal detach-
ment. Little, Brown & Co., Boston, pp 186–204
19. Guggenheim I, Franceschetti A (1928) Refraktometrische Untersuchungen des Glaskörpers
von Kaninchen und Mensch (unter physiologischen und pathologischen Bedingungen). Arch
Augenheilkd 98:448–482
20. Larsen JS (1971) The sagittal growth of the eye. III. Ultrasonic measurement of the posterior
segment (axial length of the vitreous) from birth to puberty. Acta Ophthalmol 49:441–453
21. Balazs EA (1992) Functional anatomy of the vitreus. In: Tasman W, Jaeger EA (eds) Duane’s
foundations of clinical ophthalmology, vol 1. J.B. Lippincott Co., Philadelphia, Chapter 17
22. Balazs EA, Denlinger JL (1982) Aging changes in the vitreus. In: Sekuler R, Kline D,
Dismukes K (eds) Aging and human visual function. Alan R. Liss, Inc., New York, pp 45–57
23. Naumann HN (1959) Postmortem chemistry of the vitreous body in man. Arch Ophthalmol
62:356–363
24. Coe JI (1969) Postmortem chemistries of human vitreous humor. Am J Clin Pathol
51:741–750
25. Swann DA (1980) Chemistry and biology of the vitreous body. Int Rev Exp Pathol 22:1–64
26. Süllmann H (1951) Chemie des Auges. Tabul Biol 22:1–119
27. Balazs EA, Denlinger JL (1984) The vitreus. In: Davson H (ed) The eye, vol la, 3rd edn.
Academic, Orlando, FL, pp 533–589
28. Grimshaw J, Kane A, Trocha-Grimshaw J, Douglas A, Chakravarthy U, Archer D (1994)
Quantitative analysis of hyaluronan in vitreous humor using capillary electrophoresis.
Electrophoresis 15:936–940
29. Theocharis DA, Skandalis SS, Noulas AV, Papageorgakopoulou N, Theocharis AD, Karamanos
NK (2008) Hyaluronan and chondroitin sulfate proteoglycans in the supramolecular organiza-
tion of the mammalian vitreous body. Connect Tissue Res 49:124–128
30. Clausen R, Weller M, Wiedemann P et al (1991) An immunochemical quantitative analysis of
the protein pattern in physiologic and pathologic vitreous. Graefe’s Arch Clin Exp Ophthalmol
229:186–190
31. Boettner EA, Wolter JR (1962) Transmission of the ocular media. Invest Ophthalmol
1:776–783
32. Chirila TV, Hong Y, Dalton PD, Constable IJ, Refojo MF (1998) The use of hydrophilic poly-
mers as artificial vitreous. Prog Polym Sci 23:475–508
33. Swindle-Reilly KE, Ravi N (2010) Designing hydrogels as vitreous substitutes in ophthalmic
surgery. In: Chirila T (ed) Biomaterials and regenerative medicine in ophthalmology.
Woodhead Publishing Ltd/CRC, Oxford/Boca Raton, pp 339–373
34. Nickerson CS, Karageozian HL, Park J, Kornfield JA (2005) Internal tension: a novel hypoth-
esis concerning the mechanical properties of the vitreous humor. Macromol Symp
227:183–189
134 T.V. Chirila and Y. Hong

35. Nickerson CS, Park J, Kornfield JA, Karageozian H (2008) Rheological properties of the vitre-
ous and the role of hyaluronic acid. J Biomech 41:1840–1846
36. Zimberlin JA, McManus JJ, Crosby AJ (2010) Cavitation rheology of the vitreous: mechanical
properties of biological tissue. Soft Matter 6:3632–3635
37. Sharif-Kashani P, Nishida K, Kavehpour HP, Schwartz SD, Hubschman JP (2013) Effect of cut
rates on fluidic behavior of chopped vitreous. Retina 33:166–169
38. Abdelkawi SA, Abdel-Salam AM, Ghoniem DF, Ghaly SK (2013) Vitreous humor rheology
after Nd:YAG laser photo disruption. Cell Biochem Biophys 68:267–274
39. Watts F, Tan LE, Wilson CG, Girkin JM, Tassieri M, Wright AJ (2014) Investigating the micro-
rheology of the vitreous humor using an optically trapped local probe. J Opt 16; Article ID
#015301 (7 pp)
40. Lee B, Litt M, Buchsbaum G (1992) Rheology of the vitreous body. Part I: Viscoelasticity of
human vitreous. Biorheology 29:521–533
41. Nickerson CS (2006) Engineering the mechanical properties of ocular tissues. Ph.D. Thesis,
California Institute of Technology, 2005, pp. 57–59. http://resolver.caltech.edu./
CaltechETD:etd-03172005-145045. Accessed 11 Dec 2013
Chapter C3
The Cornea

Traian V. Chirila and Shuko Suzuki

C3.1 Introduction

It would be perhaps unreasonable to rank various elements or segments of the

human eye according to a criterion based on their level of importance. The eye
works as an eminently complex system where each part plays a definite role in the
process of maintaining normal vision. Yet, a tendency toward emphasizing the func-
tional prominence of the cornea can be noticed in the ophthalmic literature. A pos-
sible reason may be the fact that the cornea is the first layer of the eye, in direct
contact with the external environment, which results in functional tasks different
from most of the other ocular elements. Indeed, the cornea acts as the essential bar-
rier that protects the eye against physical and chemical injuries. Other functions of
the cornea include the following: (a) it assures almost 100 % transmission of visible
light and about 70 % of the total dioptric power of the human eye; (b) it protects the
posterior ocular elements against damage from UV radiation by absorbing the rays
between the wavelengths 200 and 300 nm; (c) it withstands the intraocular pressure
(IOP) exerted from within the eye; (d) it functions as an impervious barrier against
microbes and other pathogens. The intricate architecture of the cornea, comprising
five discrete layers (epithelium, Bowman’s layer, stroma, Descemet’s membrane,
and endothelium, as counted from outside toward inside), all within 500–600 μm of
tissue, may be both a consequence and an indication of the diversity of the func-
tional requirements the cornea must meet. As it is now in the vertebrate eye—a
transparent, clear, smooth, elastic, tough, and relatively thick multilayered tissue—
the cornea appears as an ideal result of evolution and the epitome of an expedient
structure-to-function relationship.

T.V. Chirila (*) • S. Suzuki

Queensland Eye Institute, 140 Melbourne Street, South Brisbane, QLD 4101, Australia

© Springer Science+Business Media New York 2016 135

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_13
136 T.V. Chirila and S. Suzuki

The corneal epithelium (about 10 % of the total corneal thickness) consists in 5

to 7 layers of specialized cells, which are interconnected. About 90 % of the corneal
thickness is constituted by the stroma (substantia propria). This layer consists of
water (about 78 % by weight [A6]), collagen fibrils, proteoglycans as ground sub-
stance, and a few specialized cells (keratocytes). The collagen fibrils are organized
in lamellae in a highly specific fashion, imparting considerable mechanical strength
to the stroma. The stroma manifests permanently a tendency to swell in water, due
to the hydrophilicity of proteoglycan components. A cornea with a stroma that
swells beyond the normal value becomes opaque and, if untreated, leads to reduced
vision or blindness. The stroma rests upon Descemet’s membrane, a strong and
highly elastic membrane. The endothelium (or the posterior epithelium) is a single
layer of hexagonal cells. The endothelium performs the crucial task of maintaining
the normal hydration of the stroma, through its “leaky barrier-and-pump” function,
with a complicated and not fully understood mechanism. This process puts extraor-
dinary metabolic demands upon the endothelial cells.
The cornea is able to maintain indefinitely its integrity and transparency. Upon
pathogenic attack or due to external insults, its defense capability may be overpow-
ered and the ensuing disruption of any of the processes governing the cornea’s per-
formance will lead to reduced vision. The cornea can be afflicted by a large variety
of disorders, of which many are able to cause loss of vision despite treatment [A5,
1–4]. It can be affected by pathologic disorders (inflammations, dystrophies, and
degenerations), as well as by traumatic disorders (caused by mechanical, thermal, or
chemical injuries). Modern ophthalmology provides a large range of treatment pro-
cedures, but in case of opaque or distorted corneas, only the transplantation of donor
corneal grafts (known as penetrating keratoplasty) can restore vision. In fact, the
first successful full-thickness allograft was performed in the cornea [5], and today
the penetrating keratoplasty is still one of the most successful procedures for organ
transplantation. However, this procedure has a poor outcome in cases of chemical
burns, vascularized traumatic injuries, and a number of pathologic conditions. The
alternative for such cases is either prosthokeratoplasty, a procedure where the dam-
aged cornea is replaced with an artificial cornea, known also as a keratoprosthesis,
or transplantation of tissue-engineered constructs [6–8].

C3.2 Compendium of Physicochemical Characteristics

of the Cornea

It is rather surprising that, in spite of the prominence given to its study, the human
cornea has been considerably less investigated than the animal corneas in some
respects, perhaps less than any ocular element, which is rather peculiar considering
that excised tissue rims from postmortem corneas are routinely discarded in the eye
banks. Experimental data concerning certain properties of the human cornea, such
as its composition or electrical properties, are actually available to a little extent,
C3 The Cornea 137

and the investigators must rely upon information abundantly available for the cor-
nea of a variety of other vertebrate species. Table C3.1 shows information available
on the composition of the human cornea, as summarized by Ehlers and Hjortdal
[A6] (likely as rounded-off values). The fact that our literature search did not result
in more data confirms the above statement.
Another peculiarity of the study of the human cornea is the great attention and
amount of research dedicated to its dimensional characteristics (diameter, thickness,
geometry). However, this aspect has a justification. For instance, the corneal diam-
eter is an important tool in identifying the congenital and developmental anomalies
of the cornea, and for the design and fitting of contact lenses, intraocular lenses, and
capsular tension rings. The early literature on the measurement of the corneal diam-
eter has been reviewed by Martin and Holden [9]; more recent results, with an
emphasis on the corneas of non-Caucasian subjects, have been summarized by
Mashige [10]. As such, the concept of “corneal diameter” is poorly defined and still
debated, and consequently the reported estimations differ to some extent between
investigators. Table C3.2 presents a selection of data for the corneal diameter.
Similarly, the measurement of the thickness of the human cornea triggered an
inordinate amount of investigation. Since the corneal thickness is very sensitive to

Table C3.1 Composition of the human corneal stroma [A6]

Component % (wt)
Water 78
Collagen 15
Other proteins 5
Keratan sulfates 0.7
Chondroitin sulfate 0.3
Salts 1

Table C3.2 Diameter of the corneaa

Value (mm) Remarks Source
12.89 ± 0.60b,c HCDd [9]
11.64 ± 0.49b,c HVIDe [9]
11.71 ± 0.42b,f Horizontal WTWg diameter [11]
11.46 ± 0.48b,h Horizontal WTW diameter [12]
10.63 ± 0.63b,h Vertical WTW diameter [12]
11.05 ± 0.52b,h Average (horizontal, vertical) [12]
a
early literature reviewed in [9] shows a variation for horizontal diameters between 10.25 and
13.5 mm; the recent literature reviewed in [10], with emphasis on non-Caucasian corneas, shows
a variation between 10 and 14.6 mm for the HVID and between 10 and 11 mm for the VVID;
b
± standard deviation; c measured in 50 persons aged between 20 and 29 years, using catoptric
imaging and closed-circuit television; d horizontal corneal diameter; e HVID: horizontal visible iris
diameter; VVID: vertical visible iris diameter; f measured in 370 right eyes and 373 left eyes (all
Caucasians), using the Orbscan II topography system; g “white-to-white” diameter; h measured (in
triplicate) in 78 postmortem eyes, using a Castroviejo caliper
138 T.V. Chirila and S. Suzuki

an abnormal hydration of the stroma, and also related to the intraocular pressure in
the eye, it became a diagnostic tool for certain pathological conditions, such as
corneal degenerations. Exact measurement of the corneal thickness is also of great
significance in the assessment of patients for corneal refractive surgery. The mea-
surement of this quantity is associated with some difficulties; for instance, the thick-
ness of postmortem corneas is much higher than in living persons due to uncontrolled
hydration. Also, the thickness is not uniform throughout the cornea: it is the thinnest
in the central region and becomes thicker toward the periphery. How important is
this dimension is illustrated in the amount of research that has been reviewed in no
less than four major papers [10, 13–15]. Table C3.3 contains representative values
for the central corneal thickness (CCT).
An essential function of the cornea is to allow the formation of an image on the
retina, and one of the prerequisites for this task is the existence of a curvature of the
corneal surface(s) that can refract light sufficiently to focus on the retina. Corneal cur-

Table C3.3 Central corneal thickness in human subjects

Method Value (μm)a Source
Optical techniqueb 565 ± 2.3c [16]
559 ± 4.5d [16]
571 ± 7.1e [16]
Optical techniqueb 507 ± 4.2f [17]
Haag-Streit pachometer 520 ± 18 (right eye)g [18]
524 ± 20 (left eye)g [18]
534 (at wakening)h [19]
507 (at bedtime)h [19]
Specular techniqueb 515 ± 33i [20]
Ultrasound techniques 514.6 ± 38.4 (right eye)j [21]
516.2 ± 37.8 (left eye)j [21]
542 ± 33k [22]
545-556l [23]
Optical low-coherence reflectometry 519.6 ± 1.2 (day 1)m [24]
(OLCR) 519.9 ± 0.9 (day 2)m [24]
523.8 ± 0.6 (day 3)m [24]
Orbscan pachometer 596 ± 40k [22]
(scanning slit principle)
Specular microscopy 566-578l [23]
528.3 ± 5.5n [25]
Ultrasound biomicroscopy 550-560l [23]
a
± standard error; b custom-made device; c measured in 224 corneas (from 125 persons) aged
between <15 and >74 years; d measured in 55 corneas aged below 25 years; e measured in 28 cor-
neas aged above 65 years; f measured in 44 persons aged between 18 and >35 years; g measured in
150 eyes (from 113 persons) aged between <15 and >74 years; h measured in 9 persons daily for 1
month; i measured in 115 persons aged between 10 and 90 years; j measured in 10 persons aged
between 23 and 44 years; k measured in 20 persons, mean age 33.3 ± 9 years; l measured in 31 per-
sons (62 eyes), with the range given for 95 % confidence interval; m measured in one person for 3
consecutive days, 20 times each day; n measured in 40 persons (215 measurements)
C3 The Cornea 139

vature, expressed commonly as the radius of curvature, has been of interest to vision
scientists for centuries. A system of methodology and instrumentation, covered by the
term “keratometry,” has been developed to measure the curvature of the central cornea.
The refractive power of the cornea can be calculated from its radius of curvature, thick-
ness, and refractive index. Representative values of the corneal radius of curvature and
other derived parameters for the corneal optic system are summarized in Table C3.4.
Another parameter of significance for the light refraction through the cornea,
intrinsic to the tissue as such, is the refractive index. Its measurement has a recorded
history dating back to the nineteenth century, even before the introduction of refrac-
tometers. A refractive index of 1.376, which is commonly used in applications, has
originated in the value of 1.3763 measured in human cornea by Matthiesen in 1891.
However, as seen in Table C3.5, it has been seldom that refractive indices in the
vicinity of 1.376 were reported. There is also substantial variability in the measured
values, likely due to variation in tissues, subjects, and procedures. A similar level of

Table C3.4 Geometrical and optical properties of the human corneaa

Property Value Source
Radius of curvature (mm)b 7.88 ± 0.045c,d [26]e
7.796 (right eye) [26]f
7.746 (left eye) [26]f
7.7 (anterior surface)g [27]
6.8 (posterior surface)g [27]
7.249 to 8.993 (horizontal meridian)h [28]
6.637 to 8.001 (vertical meridian)h [28]
7.74 ± 0.26i,j [29]
7.73 ± 0.29i,k [29]
7.66 ± 0.26i,l [29]
7.55 ± 0.28i,m [29]
5.62 to 7.22 (posterior, vertical meridian)n [30]
7.84 (anterior, averaged)o [31]
6.42 (posterior, averaged)o [31]
8.03 ± 0.29c,p (males) [32]
7.91 ± 0.28c,p (females) [32]
Refractive power (D) 43.05r [27]
First focal length (mm) –23.227r [27]
Second focal length (mm) 31.031r [27]
a
for a review of the literature between 1975 and 2001 see [10], where the extreme values were reported
in the ranges of 7.10 to 8.75 mm (in the horizontal meridian) and 7.06 to 8.66 mm (in the vertical merid-
ian); b for a conversion table radius of curvature to refractive (dioptric) power see Contact Lens Spectrum
(2012) at http://www.clspectrum.com/printarticle.aspx?articleID=107267 (accessed 03/05/2014);
c
± standard error; d measured in 12 eyes in living subjects; e [26]; f [26]; g measured in 2 persons; h mea-
sured in 26 persons, aged 19 to 36 years; i ± standard deviation; j measured in 27 emmetropic persons; k
measured in 28 lowly myopic persons; l measured in 32 moderately myopic persons; m measured in 18
highly myopic persons; n measured in 120 children, aged 6 to 17 years; o measured in 25 persons, aged
60 to 80 years; p measured in 249 persons (498 eyes), aged between 18 and 29 years, all of Negroid race
(African); r calculated from experimental values for radius of curvature and thickness, and assuming a
refractive index of 1.376
140 T.V. Chirila and S. Suzuki

Table C3.5 The refractive index of the human corneal tissue

Value Remarks Source
1.3569 Whole cornea [33]a
1.377 Whole cornea, age 55 years [27]b
1.3721 Whole cornea, age 2 days [27]b
1.3763 Whole cornea [27, 33]c
1.3751 Whole cornea, at 656.3 nm (red light) [34]d
1.3818 Whole cornea, at 486.1 nm (blue light) [34]d
1.401 ± 0.005e Corneal epitheliumf [35]
1.380 ± 0.005e Anterior corneal stromag [35]
1.373 ± 0.001e Posterior corneal stromag [35]
1.411 ± 0.001 Stromal collagenh [36]
1.365 ± 0.003 Extrafibrillar ground substanceh [36]
1.372g, i Stroma (from one enucleated eye) [37]
1.404g, j Stroma (from one enucleated eye) [37]
1.378h, i Stroma [37]
1.399h, j Stroma [37]
1.389 ± 0.004e, h, k At 1270 nm in 3 postmortem tissue specimens [38]
1.3970 ± 0.001e, l Central epithelium [39]
1.3946 ± 0.001e, l Nasal peripheral epithelium [39]
1.3940 ± 0.001e, l Temporal peripheral epithelium [39]
1.369 ± 0.008e, m Corneal stroma [40]
1.373 ± 0.006e, n Corneal stroma [41]
a
[33]; b [27]; c [27, 33]; d [34]; e ± standard deviation; f measured in 10 eyes, aged between 19 and
27 years; g measured in ex vivo tissue; h measured in postmortem eye(s); i measured immediately
after removing the epithelium; j measured 35 min after removing the epithelium; k group refractive
index (ng) resulting from a low-coherence source; l measured in 10 persons, aged between 22 and
30 years; m measured during keratomileusis surgery (mechanical microkeratome) in 36 patients,
aged between 18 and 56 years; n measured during keratomileusis surgery (femtosecond laser) in
115 patients, aged between 18 and 74 years

data spread can be noticed with respect to the measurement of radiation transmis-
sion through the cornea (Table C3.6).
Although the electrical properties of the ocular elements have been extensively
investigated in the rabbit eye [47], there is little amount of such data on the human
eye. Table C3.7 contains the available information regarding the human cornea.
Like any tissue in our body, the cornea undergoes changes with the passage of
time. However, this tissue appears to be remarkably stable and to withstand elevated
physical stress for long intervals, in most of the cases for our natural life duration.
This dimensional stability, which is essential for normal vision, is a result of the
inherent mechanical properties of the corneal tissues, and it is believed that the latter
are ultimately determined by the properties and integrity of the stroma and the under-
lying Descemet’s membrane. The biomechanics of the cornea is a rather well-
established field. However, there are large discrepancies between the results of the
mechanical characteristics measured for the human cornea. For instance, the varia-
tion in the reported experimental values for the elastic modulus covers three orders
C3 The Cornea 141

Table C3.6 Transmission of ultraviolet and visible radiation through the human cornea
Transmittance (%)
Wavelength (nm) Source: [42]a,b [43]b,c [44]b,d [45]b,e [46]f
270 0 0 − <0.01 −
290 0 0 − 0.1 −
300 0 ~7 − ~2 −
310 58 ~8 − ~8 27
325 63 33 − ~10 −
350 72 50 − ~12 −
400 81 58 − ~15 74
500 90 64 89.2 − −
600 92 71 93.7 − −
700 94 78 95.4 − −
a
measured in 9 enucleated eyes, aged between 4 weeks and 75 years, on isolated cornea; b extrap-
olated from the transmission spectra; c measured on a 24-year-old cornea; d measured in 8 postmor-
tem eyes, aged between 22 and 45 years, on the whole eyes; e measured in 20 corneal rims discarded
form the eye bank; f measured in the central region of 4 postmortem corneas (aged 35 to 67 years)
after removal of the epithelia; - not measured

Table C3.7 Electric properties of the human corneal tissue

Property Value Source
Relative permittivity 62.6 ± 0.13a, b [48]
53.0 ± 0.22a, c [48, 49]
Conductivity (S m-1) 1.00 ± 0.005a, b [48]
1.05 ± 0.01a, c [48, 49]
0.4c [50]
0.26 ± 0.03b [51]
Transendothelial potential 200-500d, e [52-54]
difference (μV)
Endothelial electrical resistance (Ω 12.4 ± 0.9d, f [52]
cm2)
10.6d [54]
10.26d [54]
Transendothelial short circuit 33.0 ± 3.5d, f [52]
current density (μA cm-2)
a
range given for 95 % confidence interval; b measured at 20 °C; c measured at 37 °C; d postmortem
corneas; e posterior side is negative; f ± standard error

of magnitude! Diverse explanations have been advanced in order to explain

such variation, including the following: (a) differences in methodology between
laboratories; (b) nonuniform quality of the postmortem corneal specimens due to
uncontrolled hydration; (c) nonlinearity of the stress-strain relationship in the corneal
tissue; (d) occasionally, the measurements (e.g., elastic modulus) can be dramatically
affected by the contribution of the adjacent sclera, a strong and highly elastic tissue,
if the in vitro test specimens were not adequately isolated or when the measurements
Table C3.8 The mechanical properties of the human corneal tissuea
142

Method Remarks Young’s modulus (MPa) Ultimate strength (MPa) Elongation (%) Source
Uniaxial tensile test Cornea strips 57 ± 4.1b, c 12.7 ± 0.6b, c 35 ± 3b, c [55]
Central cornea strips 0.34d [56]
Central cornea strips 4.1e [56]
Central cornea strip with 0.282 ± 0.159f [57]
Bowman’s layer
Central cornea strip without 0.245 ± 0.088f [57]
Bowman’s layer
Peripheral strips 3.81 ± 0.4g [58]
h
Central strips 1.3 [59]
Inflation Anterior hemisphere of globe 0.37i [60]
Whole globe 1.11j [61]
Descemet’s membrane 5k [62]
Whole globe 2.87 ± 0.38b, l, m [63]
Whole globe 8.55 ± 0.83b, l, n [63]
Whole globe 19.5 ± 0.98b, l, o [63]
Excised cornea 0.79 ± 0.22p [64]
Descemet’s membrane 11.8 ± 1b, q 1.72 ± 0.19b, q 31.2 ± 1.4b, q [65]
Cornea with scleral ring 0.624r, s [66]
Cornea with scleral ring 0.777r, t [66]
Cornea with scleral ring 0.961r, u [66]
Compression Whole globe 0.036v [67]
Ultrasonic Whole globe in saline 5.3 ± 1.1w [68]
Whole globe in 15 % 20 ± 1.0x [68]
dextran solution
T.V. Chirila and S. Suzuki
 C3

Rheometry Central cornea strips (horizontal 3y [69]

cut)
Central cornea strips (Vertical 1z [69]
cut)
Central cornea strips (Diagonal 0.3aa [69]
The Cornea

cut)
Optical methods Living subjects 0.0245 ± 0.0057bb [70]
Living subjects 9.03 ± 0.42cc [71]
Living subjects 2.48 ± 091dd [72]
AFM indentation Anterior basement membrane 0.0075 ± 0.0042ee [73]
Descemet’s membrane 0.050 ± 0.0178ff [73]
Bowman’s layer 0.1098 ± 0.0132gg [74]
Anterior stroma 0.0331 ± 0.0061hh [74]
Anterior stroma 1.14-2.63ii [75]
Anterior stroma 0.25 ± 0.21jj [76]
Posterior stroma 0.1 ± 0.06kk [76]
a
values are mean, or mean ± standard deviation, except in some cases as further specified; b mean ± standard error; c measured in 11 corneas, aged 14 to 76 years;
d
intraocular pressure (IOP) = 30 mm Hg; e IOP = 10 mm Hg; f measured in 5 pairs of eyes, aged 59 to 82 years, IOP = 38 mm Hg; g measured in 10 corneal rings
left from keratoplasty; h measured in 5 enucleated eyes, IOP = 38 mm Hg; i at a strain of ~ 3.3 × 10-2 cm/cm for enucleated corneas; j measured in 5 pairs of eyes,
aged 39 to 67 years, IOP = 19 mm Hg; k aged 45 to 90 years, IOP = 11 mm Hg; l measured in 8 eyes, aged 64 to 88 years; m IOP = 2–10 mm Hg; n IOP = 10–25 mm Hg;
o
IOP = 25–100 mm Hg; p measured in 12 corneas, aged 20 to 69 years, IOP = 15 mm Hg; q measured in 7 Descemet’s membranes, aged 81.3 ± 12.3 years;
r
pressure rate = 37.5 mm Hg/min, IOP = 30 mm Hg; s measured in 4 corneas, aged 50 to 64 years; t measured in 6 corneas, aged 65 to 79 years; u measured in 13
corneas, aged 80 to 95 years; v measured in 2 enucleated eyes, IOP = 35 or 40 mm Hg; w measured in 6 eyes, aged 75 to 82 years; x measured in 6 eyes, aged 78 to
84 years; y measured in 4 corneas; z measured in 3 corneas; aa measured in 7 corneas; bb measured in 28 persons in both eyes, aged 16 to 66 years using photokera-
tometry and topographic pachometry, IOP = 10 mm Hg; cc measured in 29 persons, aged 17 to 66 years, using slit-lamp, photokeratoscopy, topographic pachometry,
and Goldmann applanation tonometry, IOP = 30 mm Hg; dd measured in 5 persons, aged 27 to 80 years, using Reichert ocular response analyzer (ORA); ee measured
in 6 corneas, aged 58 to 67 years; ff measured in 5 corneas, aged 53 to 68 years; gg measured in 6 corneas, aged 19 to 73 years; hh measured in 5 corneas, aged 19 to
73 years; ii measured in 4 corneas; jj measured in 6 pairs of corneas, aged 39 to 89 years; kk measured in 6 pairs of corneas, aged 56 to 88 years
143
144 T.V. Chirila and S. Suzuki

are carried out on living subjects. Table C3.8 presents a compilation, hopefully
exhaustive, of the mechanical property measurements of the human cornea.

Additional Reading

[A1] Maurice, D.M. (1962) The cornea and sclera, in The Eye. Vol. 1: Vegetative
Physiology and Biochemistry (ed H. Davson), Academic Press, New York,
pp. 289-368.
This was the defining text to many generations of corneal researchers, written by
one of the greatest scholars on the topic.
[A2] King, Jr., J.H. and McTigue, J.W. (eds) (1965) The Cornea World Congress,
Butterworths, Washington, 754 pp.
A book containing the proceedings of the First World Congress on the Cornea,
held in Washington, D.C. in 1964, attended by the most illustrious corneal surgeons
and investigators of the period, such as Joachim Barraquer, José Barraquer, Bietti,
Cardona, Castroviejo, DeVoe, Dohlman, Franceschetti, Maumenee, Refojo, Rycroft,
and many others. The book covers the entire knowledge on the cornea up to that
time. It is still a valuable reference book for all cornea scholars.
[A3] Sherrard, E.S. (1977) The cornea: structure and transparency, in Scientific
Foundations of Ophthalmology (eds E.S. Perkins and D.W. Hill), Heinemann
Medical Books, London, pp. 29-35.
A concise yet very informative text, explaining in accessible terms the structure
and function of the cornea.
[A4] Fehér, J. (1996) Pathophysiology of the Eye, Vol. 3: Transparency and
Refraction of the Cornea, Akadémiai Kiadó, Budapest, 230 pp.
This book presents in organized fashion many aspects of the cornea including its
development, anatomy, structure, physiology, and new approaches to certain inter-
relations between the parts of the cornea.
[A5] Foster, C.S, Azar, D.T. and Dohlman, C.H. (eds) (2005) Smolin and Thoft’s
The Cornea: Scientific Foundations and Clinical Practice, 4th edn, Lippincott
Williams & Wilkins, Philadelphia, 1323 pp. + 189 plates.
This monumental treatise does not need any recommendation, as its importance in
the study of the cornea stood the test of time for generations of corneal investigators.
[A6] Ehlers, N. and Hjortdal, J. (2006) The cornea: epithelium and stroma, in
The Biology of the Eye (ed J. Fischbarg), Elsevier, Amsterdam, pp. 83–111.
A well-balanced description of essential aspects of the corneal properties and
function. Highly recommended as an introductory meaningful text for corneal
investigators.
[A7] Fischbarg, J. (2006) The corneal endothelium, in The Biology of the Eye
(ed J. Fischbarg), Elsevier, Amsterdam, pp. 113–25.
A significant contribution to the understanding of the corneal endothelium and
the mechanism of its pump-barrier function, written by the proponent of the latest
theory on the transendothelial fluid transport.
C3 The Cornea 145

[A8] Dawson, D.G., Ubels, J.L. and Edelhauser, H.F. (2011) Cornea and sclera,
in Adler’s Physiology of the Eye, 11th edn (eds L.A. Levin et al.), Elsevier,
Edinburgh, pp. 71–130.
The most recent chapter on the cornea in a traditional textbook, beautifully writ-
ten and illustrated, and covering the latest views on the structure and function of this
part of the eye.
[A9] Krachmer, J.H., Mannis, M.J. and Holland, E.J. (eds) (2011) Cornea, 3rd
edn, vol. I & vol. II, Mosby Elsevier, St. Louis, MO, 1913 pp.
A comprehensive, state-of-the art coverage of the topic throughout 171 chapters
organized in two volumes.

Recommended sources for ophthalmic terminology:

Cassin, B., Solomon, S.A.B. and Rubin, M.L. (1990) Dictionary of Eye Terminology, 2nd edn,
Triad Publishing Co., Gainesville, FL, 286 pp.
Stein, H.A., Slatt, B.J. and Stein, R.M. (1992) Ophthalmic Terminology. Speller and Vocabulary
Builder, 3rd edn, Mosby-Year Book Inc., St. Louis, MO, pp. 3–33, 135-55, 261-5.
Myles, W.M. (1993) Ophthalmic etymology. Surv. Ophthalmol., 37, 306–9.

References

1. Grayson, M. (1979) Diseases of the Cornea, C.V. Mosby Co., St. Louis.
2. Donaldson, D.D. (1980) Atlas of External Diseases of the Eye. Vol. III: Cornea and Sclera, 2nd
edn, C.V. Mosby Co., St. Louis.
3. Coster, D.J. (2002) Cornea, Fundamentals of Clinical Ophthalmology Series, BMJ Books,
London.
4. Holland, E.J. and Mannis, M.J. (2002) Ocular Surface Diseases. Medical and Surgical
Management, Springer-Verlag, New York.
5. Zirm, E. (1906) Eine erfolgreiche totale Keratoplastik. Graefes Arch. Ophthalmol., 64, 580-93.
6. Chirila, T.V., Hicks, C.R., Dalton, P.D., Vijayasekaran, S., Lou, X., Hong, Y., Clayton, A.B.,
Ziegelaar, B.W., Fitton, J.H., Platten, S., Crawford, G.J. and Constable, I.J. (1998) Artificial
cornea. Prog. Polym. Sci., 23, 447-73.
7. Evans, M.D.M. and Sweeney, D.F. (2010) Synthetic corneal implants, in Biomaterials and
Regenerative Medicine in Ophthalmology (ed T. Chirila), Woodhead Publishing Ltd, Oxford &
CRC, Boca Raton, pp. 65-133.
8. Proulx, S., Guillemette, M., Carrier, P., Auger, F.A., Germain, L., Giasson, C.J., Gaudreault,
M. and Guérin, S.L. (2010) Tissue engineering of human cornea, in Biomaterials and
Regenerative Medicine in Ophthalmology (ed T. Chirila), Woodhead Publishing Ltd, Oxford &
CRC, Boca Raton, pp. 150-192.
9. Martin, D.K. and Holden, B.A. (1982) A new method for measuring the diameter of the in vivo
human cornea. Am. J. Optom. Physiol. Opt., 59, 436-41.
10. Mashige, K.P. (2013) A review of corneal diameter, curvature and thickness values and influ-
encing factors. S. Afr. Optom., 72, 185-94.
11. Rüfer, F., Schröder, A. and Erb, C. (2005) White-to-white corneal diameter. Normal values in
healthy humans obtained with the Orbscan II topography system. Cornea, 24, 259-61.
12. Khng, C. and Osher, R.H. (2008) Evaluation of the relationship between corneal diameter and
lens diameter. J. Cataract Refract. Surg., 34, 475-9.
146 T.V. Chirila and S. Suzuki

13. Mishima, S. (1968) Corneal thickness. Surv. Ophthalmol., 13, 57-96.

14. Doughty, M.J. and Zaman, M.L. (2000) Human corneal thickness and its impact on intraocular
pressure measures: a review and meta-analysis approach. Surv. Ophthalmol., 44, 367-408.
15. Ehlers, N. and Hjortdal, J. (2004) Corneal thickness: measurement and implications. Exp. Eye
Res., 78, 543-8.
16. von Bahr, G. (1948) Measurements of the thickness of the cornea. Acta Ophthalmol., 26,
247-65.
17. Maurice, D.M. and Giardini, A.A. (1951) A simple optical apparatus for measuring the corneal
thickness, and the average thickness of the human cornea. Br. J. Ophthalmol., 35, 169-77.
18. Hansen, F.K. (1971) A clinical study of the normal human central corneal thickness. Acta
Ophthalmol., 49, 82-9.
19. Manchester, P.T. (1970) Hydration of the cornea. Trans. Am. Ophthalmol. Soc., 68, 425-61.
20. Olsen, T. and Ehlers, N. (1984) The thickness of the human cornea as determined by a specular
method. Acta Ophthalmol., 62, 859-71.
21. Reinstein, D.Z., Silverman, R.H., Rondeau, M.J. and Coleman, D.J. (1994) Epithelial and
corneal thickness measurements by high-frequency ultrasound digital signal processing.
Ophthalmology, 101, 140-6.
22. Marsich, M.M. and Bullimore, M.A. (2000) The repeatability of corneal thickness measures.
Cornea, 19, 792-5.
23. Tam, E.S. and Rootman, D.S. (2003) Comparison of central corneal thickness measurements
by specular microscopy, ultrasound pachymetry, and ultrasound biomicroscopy. J. Cataract
Refract. Surg., 29, 1179-84.
24. Böhnke, M., Masters, B.R., Wälti, R., Ballif, J.J., Chavanne, P., Gianotti, R. and Salathé, R.P.
(1999) Precision and reproducibility of measurements of human corneal thickness with rapid
optical low-coherence reflectometry (OLCR). J. Biomed. Opt., 4, 152-6.
25. Bovelle, R., Kaufman, S.C., Thompson, H.W. and Hamano, H. (1999) Corneal thickness mea-
surements with the Topcon SP-2000P specular microscope and an ultrasound pachymeter.
Arch. Ophthalmol., 117, 868-70.
26. von Helmholtz, H. (1924) Helmholtz’s Treatise on Physiological Optics (ed J.P.C. Southall),
vol. 1 (trans. from the 3rd German edition), Optical Society of America, Rochester, NY,
pp. 9-10.
27. Gullstrand, A. (1924) Procedure of the rays in the eye – Laws of first order, in Helmholtz’s
Treatise on Physiological Optics (ed J.P.C. Southall), vol. 1 (trans. from the 3rd German edi-
tion), Optical Society of America, Rochester, NY, pp. 301-58.
28. Chapman, H.C. and Brubaker, A.P. (1893) The radius of curvatures of the cornea. Proc. Acad.
Nat. Sci. Philadelphia, 45, 349-60.
29. Carney, L.G., Mainstone, J.C. and Henderson, B.A. (1997) Corneal topography and myopia.
A cross-sectional study. Invest. Ophthalmol. Vis. Sci., 38, 311-20.
30. Garner, L.F., Owens, H., Yap, M.K.H., Frith, M.J. and Kinnear, R.F. (1997) Radius of curva-
ture of the posterior surface of the cornea. Optom. Vis. Sci., 74, 496-8.
31. Vojniković, B., Gabrić, N., Dekaris, I. and Jurić, B. (2013) Curvature analysis of the corneal
front and back surface. Coll. Antropol., 37 (Suppl. 1), 93-6.
32. Esenwah, E.C., Azuamah, Y.C. and Ekwe, C. (2013) An investigation into the radius of curva-
ture of the cornea of young adults in relation to the conformity of an 8.6 mm base curve contact
lens. Int. J. Health Sci. Res., 3, 1-5.
33. Patel, S. (1987) Refractive index of the mammalian cornea and its influence during pachome-
try. Ophthal. Physiol. Opt., 7, 503-6.
34. Millodot, M. and Newton, I.A. (1976) A possible change of refractive index with age and its
relevance to chromatic aberration. Graefe’s Arch. Clin. Exp. Ophthalmol., 201, 159-67.
35. Patel, S., Marshall, J. and Fitzke III, F.W. (1995) Refractive index of the human corneal epi-
thelium and stroma. J. Refr. Surg., 11, 100-5.
36. Meek, K.M., Leonard, D.W., Connon, C.J., Dennis, S. and Khan, S. (2003) Transparency,
swelling and scarring in the corneal stroma. Eye, 17, 927-36.
C3 The Cornea 147

37. Patel, S., Alió, J.L. and Pérez-Santonja, J. (2004) Refractive index change in bovine and
human corneal stroma before and after LASIK: a study of untreated and re-treated corneas
implicating stromal hydration. Invest. Ophthalmol. Vis. Sci., 45, 3523-30.
38. Lin, R.C., Shure, M.A., Rollins, A.M., Izatt, J.A. and Huang, D. (2004) Group index of the
human cornea at 1.3-μm wavelength obtained in vitro by optical coherence domain refractom-
etry. Opt. Lett., 29, 83-5.
39. Vasudevan, B., Simpson, T.L. and Sivak, J.G. (2008) Regional variation in the refractive-index
of the bovine and human cornea. Optom. Vis. Sci., 85, 977-81.
40. Patel, S., Alió, J.L., Amparo, F. and Rodriguez-Prats, J.L. (2011) The influence of age on the
refractive index of the human corneal stroma resected using a mechanical microkeratome.
Cornea, 30, 1353-7.
41. Amparo, F., Patel, S., Alió, J.L., Rodriguez-Prats, J.L. and Moreno, L.J. (2012) Relationship
between patient age and refractive index of the corneal stroma during refractive surgery
assisted by femtosecond laser flap creation. Cornea, 31, 751-5.
42. Boettner, E.A. and Wolter, J.R. (1962) Transmission of the ocular media. Invest. Ophthalmol.,
1, 776-83.
43. Lerman, S. (1984) Biophysical aspects of corneal and lenticular transparency. Curr. Eye Res.,
3, 3-14.
44. Beems, E.M. and van Best, J.A. (1990) Light transmission of the cornea in whole human eyes.
Exp. Eye Res., 50, 393-5.
45. Kolozsvári, L., Nógrádi, A., Hopp, B. and Bor, Z. (2002) UV absorbance of the human cornea
in the 240- to 400-nm range. Invest. Ophthalmol. Vis. Sci., 43, 2165-8.
46. Doutch, J.J., Quantock, A.J., Joyce, N.C. and Meek, K.M. (2012) Unltraviolet light transmis-
sion through the human corneal stroma is reduced in the periphery. Biophys. J., 102, 1258-64.
47. Gabriel, C., Sheppard, R.J. and Grant, E.H. (1983) Dielectric properties of ocular tissue at
37 °C. Phys. Med. Biol., 28, 43-9.
48. Gabriel, C. (1996) Compilation of the dielectric properties of body tissues at RF and micro-
wave frequencies, Final Technical Report AL/OE-TR-1996-0004, Prepared for Brooks
Air Force Base, Texas, www.dtic.mil/cgi-bin/GetTRDoc?AD = ADA303903 (accessed
02/27/2014), 16 pp.
49. Gabriel, S., Lau, R.W. and Gabriel, C. (1996) The dielectric properties of biological tissues:
III. Parametric models for the dielectric spectrum of tissues. Phys. Med. Biol., 41, 2271-93.
50. Gabriel, C. and Gabriel, S. (1996) Compilation of the dielectric properties of body tissues at
RF and microwave frequencies, Final Report AL/OE-TR-1996-0037, Prepared for AFOSR/NL
Bolling AFB DC 20332-0001, http://niremf.ifac.cnr.it/docs/DIELECTRIC/Report.html
(accessed 03/03/2014), 10 pp.
51. Oksala, A. and Lehtinen, A. (1959) Comparative studies on the electrical conductivity of aque-
ous humour, vitreous body, cornea and sclera. Acta Ophthalmol., 37, 388-94.
52. Wigham, C. and Hodson, S. (1981) Bicarbonate and the trans-endothelial short circuit current
of human cornea. Curr. Eye Res., 1, 285-90.
53. Hodson, S., Wigham, C., Williams, L., Mayes, K.R. and Graham, M.V. (1981) Observations
on the human cornea in vitro. Exp. Eye Res., 32, 353-60.
54. Hodson, S. and Wigham, C. (1983) The permeability of rabbit and human corneal endothe-
lium. J. Physiol., 342, 409-19.
55. Andreassen, T.T., Simonsen, A.H. and Oxlund, H. (1980) Biomechanical properties of kerato-
conus and normal corneas. Exp. Eye Res., 31, 435-41.
56. Hoeltzel, D.A., Altman, P., Buzard, K. and Choe, K. (1992) Strip extensiometry for compari-
son of the mechanical response of bovine, rabbit, and human corneas. J. Biomech. Eng., 114,
202-15.
57. Seiler, T., Matallana, M., Sendler, S. and Bende, T. (1992) Does Bowman’s layer determine the
biomechanical properties of the cornea? Refr. Corneal Surg., 8, 139-42.
58. Zeng, Y., Yang, J., Huang, K., Lee, Z. and Lee, X. (2001) A comparison of biomechanical
properties between human and porcine cornea. J. Biomech., 34, 533-7.
148 T.V. Chirila and S. Suzuki

59. Wollensak, G., Spoerl, E. and Seiler, T. (2003) Stress-strain measurements of human and por-
cine corneas after riboflavin–ultraviolet-A-induced cross-linking. J. Cataract Refract. Surg.,
29, 1780-5.
60. Woo, S.L.-Y., Kobayashi, A.S., Schlegel, W.A. and Lawrence, C. (1972) Nonlinear material
properties of intact cornea and sclera. Exp. Eye Res., 14, 29-39.
61. Kobayashi, A.S., Staberg, L.G. and Schlegel, W.A. (1973) Viscoelastic properties of human
cornea. Exp. Mech., 13, 497-503.
62. Jue, B. and Maurice, D.M. (1986) The mechanical properties of the rabbit and human cornea.
J. Biomech., 19, 847-53.
63. Hjordtal, J.Ø. (1996) Regional elastic performance of the human cornea. J. Biomech., 29,
931-42.
64. Bryant, M.R. and McDonnell, P.J. (1996) Constitutive laws for biomechanical modeling of
refractive surgery. J. Biomech. Eng., 118, 473-81.
65. Danielsen, C.C. (2004) Tensile mechanical and creep properties of Descemet’s membrane and
lens capsule. Exp. Eye Res., 79, 343-50.
66. Elsheikh, A., Wang, D., Brown, M., Rama, P., Campanelli, M. and Pye, D. (2007) Assessment
of corneal biomechanical properties and their variation with age. Curr. Eye Res., 32, 11-9.
67. Schwartz, N.J., Mackay, R.S. and Sackman, J.L. (1966) A theoretical and experimental study
of the mechanical behavior of the cornea with application to the measurement of intraocular
pressure. Bull. Math. Biophys., 28, 585-643.
68. Wang, H., Prendiville, P.L., McDonnell, P.J. and Chang, W.V. (1996) An ultrasonic technique
for the measurement of the elastic moduli of human cornea. J. Biomech., 29, 1633-6.
69. Jayasuriya, A.C., Ghosh, S., Scheinbeim, J.I., Lubkin, V., Bennett, G. and Kramer, P. (2003).
A study of piezoelectric and mechanical anisotropies of the human cornea. Biosens.
Bioelectron., 18, 381-7.
70. Sjøntoft, E. and Edmund, C. (1987) In vivo determination of Young’s modulus for the human
cornea. Bull. Math. Biol., 49, 217-32.
71. Edmund, C. (1988) Corneal elasticity and ocular rigidity in normal and keratoconic eyes. Acta
Ophthalmol., 66, 134-40.
72. Glass, D.H. (2008) Characterization of the biomechanical properties of the in vivo human
cornea, Ph.D. Thesis, Ohio State University, https://etd.ohiolink.edu/ap/10?0::NO:10:P10_
ACCESSION_NUM:osu1211920104 (accessed 02/27/2014), p. 60.
73. Last, J.A., Liliensiek, S.J., Nealey, P.F. and Murphy, C.J. (2009) Determining the mechanical
properties of human corneal basement membranes with Atomic Force Microscopy. J. Struct.
Biol., 167, 19-24.
74. Last, J.A., Thomasy, S.M., Croasdale, C.R., Russell, P. and Murphy, C.J. (2012) Compliance
profile of the human cornea as measured by atomic force microscopy. Micron, 43, 1293-8.
75. Lombardo, M., Lombardo, G., Carbone, G., De Santo, M.P., Barberi, R. and Serrao, S. (2012)
Biomechanics of the anterior human corneal tissue investigated with atomic force microscopy.
Invest. Ophthalmol. Vis. Sci., 53, 1050-7.
76. Dias, J., Diakonis, V.F., Kankariya, V.P., Yoo, S.H. and Zierbath, N.M. (2013) Anterior and
posterior corneal stroma elasticity after corneal collagen crosslinking treatment. Exp. Eye Res.,
116, 58-62.
Part II
Chapter 1a
Metallic Biomaterials: Introduction

H. Breme, V. Biehl, Nina Reger, and Ellen Gawalt

1a.1 Introduction

Metallic biomaterials are one of the most commonly used biomaterial groups along
with ceramics, synthetic polymers, and naturally derived products. The utility of
these metallic materials is based largely on the formation of a thin but protective
oxide layer. The oxide layer forms upon exposure to oxygen and re-forms within
milliseconds after damage [1]. This layer reduces corrosion in vivo, one of the
major requirements of a robust biomaterial. The other requirements include biotol-
erability, bioadhesion, biofunctionality (bioactivity), and processability. In practice
each metal or alloy, i.e., titanium alloys, stainless steel, and Co-Cr alloys, has their
own advantages for different applications based on mechanical, chemical, and bio-
functional properties. Generally, metallic biomaterials are used for structural appli-
cations such as implants, pins, and bone scaffolding due to their excellent mechanical
properties such as Young’s modulus, tensile strength, ductility, fatigue, and wear
resistance. However, they can be used for unloaded, purely functional devices such
as cages for pumps, valves, and heart pacemakers. The first generation of metallic
biomaterials was designed for minimal toxicity. The second generation has been
designed for functionality at both at the mechanical and molecular level to enhance
integration of the material into the biological environment and increase longevity of
the implant. The third generation has focused not only on functionality but also on
regeneration of the surrounding tissue in conjunction with the bioactive material [2].

H. Breme • V. Biehl
Lehrstuhl für Metallische Werkstoffe, Universität des Saarlandes, Saarbrücken, Germany
N. Reger • E. Gawalt (*)
Department of Chemistry and Biochemistry, Bayer School of Natural and Environmental
Sciences, Mellon 337, 600 Forbes Avenue, Pittsburgh, PA 15282, USA
e-mail: [email protected]

© Springer Science+Business Media New York 2016 151

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_14
152 H. Breme et al.

In the following chapters the metallic biomaterials are characterized in terms of

their composition, physical and mechanical properties, and their corrosion and
biological behavior.

1a.2 Alloys and Their Applications

Most metallic biomaterials in use today are alloys with Ti, Fe, Co, Cr, Ni, V, Al, and
Ta as common components. Each alloy has different mechanical properties detailed
herein making them preferred materials for different applications. Stainless steel
316 L (SS316L), Co-Cr, Ti6Al4V, and Nitinol are among the most common [3].
SS316L is a strong material that is used for small fixation devices such as pins,
screws, and plates. It was originally used for prosthetics such as hip and knee
replacements but the market size for SS316L prosthetics has been consistently
shrinking due to the prevalent use of lighter alloys. The material is composed of
66 % Fe, 19 % Cr, 9 % Ni, 3 % Mo, and 2 % Mn with Si and C. The oxide surface
of SS316L is predominately composed of Fe2O3 and Cr2O3 [4]. This surface is sus-
ceptible to pitting and corrosion, which is the largest cause of failure [5]. Subsequent
to the corrosion, Ni ions may be released into the system. If the Ni ions collect in
organs, toxicity becomes an issue [6]. Therefore, alloys where Ni is replaced with
nitrogen are being introduced into the arena [7].
Cobalt-chromium (Co-Cr) alloyed with W, Mo, or other metals are used for both
weight-bearing and non-weight-bearing applications. These lightweight materials
are common in hip and knee prosthetics. Wear and fatigue are an issue with Co-Cr
alloys. Therefore, metal-on-metal applications are limited and Co-Cr is used in
metal-polymer applications or portions of an implant. This class of alloys is also
common in arterial stent applications [8]. As a result of the wear issues, Cr ion tox-
icity questions have led to a reduction in usage.
Titanium and its alloys are the most common metallic biomaterials. Ti6Al4V is
used in prosthetic devices including knees, hips, and vertebrae. The material is light
with a favorable Young’s modulus [9]. The oxide surface consists mainly of TiO2
and Al2O3 in approximately the same percentage as the bulk. TiO2 is an extremely
stable and inert oxide under physiological and electrochemical conditions coupled
with a high coefficient of friction making the material an ideal prosthetic and for use
in metal-on-metal applications. Even though there are no detectable amounts of V
present on the surface, concerns with V toxicity have been addressed and Ta or Nb
has been used as replacements. Nitinol is a shape memory alloy that is used in stent
and orthodontic applications. The surface of this 50–50 alloy is generally consid-
ered to be TiO2. In these applications, the inert oxide surface is important because
integration into the surrounding tissue in its applications is not critical and in some
cases undesirable. Due to these benefits, Ti alloys dominate the metallic biomateri-
als market [3].
1a Metallic Biomaterials: Introduction 153

1a.3 Mechanical Properties

The largest application for metallic biomaterials is in orthopedics. In these

applications, stress shielding, tensile strength, elongation at fracture, fracture tough-
ness, and fatigue strength are all important factors. For example, stress shielding is
an issue because the Young’s modulus of bone is 30 GPa while the modulus of the
alloys is much higher, i.e., 100–110 GPa for Ti6Al4V. This mismatch can lead to
fractures, failure, and eventual replacement of the prosthetic. Fatigue strength is
extremely important in weight-bearing applications wherein repeated applied loads
can cause structural damage. Another key property is fracture toughness. It is a
measure of a material to resist fracture once a crack has formed and is therefore an
essential measure in both weight- and non-weight-bearing applications. Specific
strength data for each metal can be found in their respective chapters.

1a.4 Corrosion and Inertness

The primary corrosion products of metallic implants are mainly responsible for the
biotolerability of the implanted alloy and their sustained interactions. The outer
oxide layer of the metals and alloys provides protection from corrosion [5, 6].
However, the oxide layer does associate with hydroxyl groups and/or bound water
that can facilitate corrosion either through electrochemical reactions or through
reactions with biological fluids (in vivo or simulated). This can result in pitting or
crevice corrosion and changes in surface composition, which can ultimately lead to
metal ion release. This type of corrosion is particularly important in biological sys-
tems where the ions can be transported to various organs and due to an enrichment
of the metal, toxicity may ensue.
The primary corrosion products of the most abundant elements in metallic implant
materials vary due to their thermodynamic stability. While the oxides or hydroxides
of Al, Cr, Nb, Ta, Ti, and V are stable due to a more negative heat of formation than
that of water, the oxides and hydroxides of Co and Ni are unstable because of a less
negative heat of formation than that of water [10]. The interaction between the oxide
or hydroxide and body fluid is increased if the heat of formation for the oxide or
hydroxide is increased. Therefore, the thermodynamically stable corrosion products
have a low solution product and a low solubility in the body fluid. This is directly
demonstrated by the pKa values (negative logarithm) of the solution product of the
primary corrosion products [11]. While Ti-, Ta-, Nb-, and Cr-oxides have pKa values
of >14, i.e., hydrolysis cannot play a role, Co-, Fe-, and Ni-oxides possess negative
pKa values which cause considerable solubility. In spite of a high negative heat of
formation for Fe2O3 and Fe- and Cr-hydroxide negative pKa values and a high solubil-
ity are reported [11]. A remarkable solubility of Cr in serum was observed [12], while
titanium is practically insoluble due to the formation of the thermodynamically stable
oxide TiO2. It is this stability that favors Ti as a biomaterial over other materials.
154 H. Breme et al.

Further, the oxide layers of Ti, Ta, and Nb are semi- or nonconductive oxides.
These oxides are able to prevent to a great extent an exchange of electrons and
therefore a flow of ions through the tissue [13] due to their isolating effect. This
isolating effect may be demonstrated by the dielectric constants of the different
metal oxides. While TiO2 (rutile), Fe2O3, and Nb2O5 have constants even higher than
that of water, A12O3, Cr2O3, and Ta2Os have a lower isolating effect and a higher
conductivity. For Ni- and V-oxides, dielectric constants are not available because of
their high conductivity [14]. As a result of these electrochemical properties, it is
clear why Ti alloys are much less susceptible to pitting and crevice corrosion than
SS316L.

1a.5 Biofunctionality

The new generation of metallic biomaterials must be biofunctional not just biotoler-
able. A biofunctional material will integrate with the existing surrounding tissue or
induce tissue growth de novo. This begins with cell adhesion, spreading, prolifera-
tion, and eventually tissue growth. However, the surface oxide layers are inert
towards the surrounding tissue, protein- and environment-reducing integration, and
potential functionality [15]. Therefore, the chemical bonding of a metallic implant
with the tissue, which is observed between bioactive ceramics like hydroxyapatite
and bone, seems to be limited and improbable, and the adhesion strength between
the bone and the metal will have a primarily mechanical character. While metallic
biomaterials, in general, do not induce de novo tissue formation they support adhe-
sion and spreading of many cell types including fibroblast, endothelial, and osteo-
blast cells. Cell adhesion and growth have been studied in many systems with
varying surface properties on the molecular, nano, micro, or macroscale.
Osseointegration of orthopedic alloys has been studied under many conditions,
surface pretreatments, and modifications. For example, the slow formation of cal-
cium phosphate mineral, the mineral component of bone, occurs on TiO2, Nitinol,
and SS316L both in vivo and in Hanks’ buffered saline solution without surface
modification [1]. In another study, the bone ingrowth behavior of miniplates of cp-Ti
and SS316L was investigated by the implantation of these plates on the legs of
Hanford minipigs. The miniplates were fixed to the legs of the pigs by screws. After
an 8-week exposure time, in the animals where Ti plates had been used, a new forma-
tion of bone could be observed in close contact to the surface of the plates. In contrast
to this result, when SS316L was used, there was less new bone formation and unde-
sired granulated tissue was found between the metallic surface and surrounding
bone. This granulated tissue at the interface bone/implant has the disadvantage that
it is not supplied with blood and unable to transfer or sustain forces, so that an even-
tual loosening of the implant will take place [16]. Successful growth of bone in close
contact with metal and metal alloys has also been reported in similar investigations
in which the contact area tissue/implant has been studied in detail [3, 5–7, 17–23].
1a Metallic Biomaterials: Introduction 155

Surface modifications are used to improve the functionality and integration of

metallic materials. Many reported modifications include bioactive molecules,
organic thin films, or metallic thin films. These modifications can improve soft tis-
sue growth and osseointegration [1–3, 5, 7–9, 15, 17, 24–28]. Molecular and chemi-
cal coatings are increasingly important but outside the scope of this handbook.
Physical modification of the oxide surface at the nano- and microscale increases cell
adhesion. A Ti alloy femoral implant with nanomodifications, i.e., nanotrabecular
and nanotuft-like structures, to the surface was implanted into a rat. The removed
implant had a greater Ca phosphate content and a higher bone-to-implant contact
percentage after 4 weeks compared to the rough surface control [29–31].
Additionally, TiO2 nanotubes grown on the surface of Ti dental and tibial implants
showed an increase in osseointegration compared to traditional smooth and unmod-
ified Ti surfaces [29, 30, 32]. Therefore, adding nanostructures to microsmooth and
microrough bone implants shows an improvement in early healing and long-term
osseointegration [1, 29].
Two clinically relevant surface modifications, which improve biofunctionality,
include surface roughness and surface porosity [28, 29, 33]. Surface roughness can
affect cell adhesion and subsequent integration. Surface roughness is clinically used
to improve cell adhesion and subsequent soft tissue integration and osseointegration
in prosthetics. This influence was investigated using cylinders of Ti and Ti alloys
which were implanted on the legs of rabbits. A measurable adhesion of the Ti alloys
could be observed only on implants with a surface roughness of >22 μm [33]. With
increasing roughness the adhesion strength is improved. Meanwhile, smooth, highly
polished surfaces reduce soft tissue attachment. This is important in reducing scar
tissue formation on some portions of implants or on parts of implants where sliding
is necessary [28, 34, 35].
Porosity of the implant improves cell adhesion and eventual bone ingrowth into
an implant. In addition to the improved fixation, the porous implant has two other
advantages: its Young’s modulus is decreased, which provides better transmission
of the functional load and stimulation of new bone formation, and the damping
capacity of the implant is increased. The shear stress generated by the functional
loading is decreased because, similar to the thread of a screw, the load at this inter-
face causes a normal stress perpendicular to the inclined area and a lower shear
stress which is effective in the inclined area. An important example of this surface
modification is the porous head of a metallic hip implant. In this application the
decrease in load decreases potential fracturing along with the improved bone fixa-
tion [33].
In contrast to the Ti alloys, the adhesion of bioglass is not dependent on the sur-
face roughness [33]. These results show that the growth of the bone and the tissue
in close contact to Ti and its alloys with formation of a strong bond must have a
more biomechanical than chemical-bioactive character. Consequently, it was shown
that by an increase of the surface area, e.g., by drilling holes in the contact area of
the implanted cylinders to the bone, the tear-off force necessary was increased.
However, if the supplementary surface was taken into consideration, the adhesion
156 H. Breme et al.

strength was not increased. A Ti5A12.5Fe implant coated with hydroxyapatite had
a maximum adhesion strength of 1.97 N/mm2 [33].
The shape of non-weight-bearing biomaterials is also important. For example,
those consisting of nanostructures of Al and Ti have shown to enhance cell adhe-
sion and proliferation. The metals were formed into nanofibers and nanospheres.
Both nanoshapes provided a platform for Ca deposition. This step is critical to
long-term osseointegration of the implant material into the body. However, the
nanofibers provided a superior platform for bone formation when compared to the
nanospheres [9, 36, 37].
Traditionally, dental implants have outpaced the orthopedic implants in terms of
longevity, biointegration, and functionality lasting up to 25 years. These materials
include metals and ceramics. For example, implants made from different materials
inserted into the jaws of dogs varied in their behavior. With Ti implants the bone
grew in close contact to the metal surface. In contrast, when SS316L was used as the
implant material, a fibrous encapsulation, which separated the implant from the sur-
rounding tissue, was formed. In Ti implants instead of this fibrous encapsulation, an
intercellular substance appeared. A similar unfavorable behavior was found in the
case of dental implants of a Co-Cr alloy in dogs [38]. As in orthopedic research,
recent dental studies have emphasized the surface properties of the materials. Due
to the oxide nature of the implant surfaces, there are hydroxyl and μ-oxo groups on
the surface. The presence of these hydrophilic groups increases mesenchymal stem
cell adhesion and differentiation leading to enhanced integration. Combining micro-
roughness with these hydrophilic groups provided the best environment for osseo-
integration in vivo [25].

1a.6 Future Directions

Metallic biomaterials are used in orthopedic, craniofacial, dental, and stabilizing

implants each year affecting millions of patients. This multibillion dollar industry
will continue to increase and diversify as the population ages and obesity-related
health issues proliferate causing the need for more implants and innovation.
Improving the lifetime of the biomaterial in the next generation is of paramount
importance. In order to do this, the material’s toxicity must be reduced. Toxicity
may be changed by varying the alloy content, i.e., exchanging nitrogen for nickel in
SS316L. Alternately, the surface corrosion could be reduced through surface pas-
sivation or modification such as plasma nitriding or enhancing the oxide layer.
Continuing focused efforts on integration with the surrounding tissue and bone is
important. For example, porous implants have been successful at increasing bone
fixation while chemical modification of the surface will continue to be an important
avenue for improving healing by delivering drugs, peptides, proteins, and growth
factors to the wound site. This may lead to attraction and proliferation of cells and
eventual growth of bone and tissue at the site or in some cases attraction of Ca and
phosphate and the growth of hydroxyapatite. Biodegradable metals represent a
1a Metallic Biomaterials: Introduction 157

recent category of interest in biomaterials. Their use in applications from bone

scaffolds to arterial stents is being evaluated. Issues with stability and degradations
rates can hinder their performance and must be addressed in the future. Overall,
metallic biomaterials are and will continue to be in high demand due to their strength
parameters, which make them ideal for structural biomaterials as prosthetics, scaf-
folds, pins, screws, and plates.

References

1. Minagar S, Wang J, Berndt CC, Ivanova EP, Wen C (2013) Cell response of anodized nano-
tubes on titanium and titanium alloys. J Biomed Mater Res A 101A(9):2726–2739
2. Mantripragada VP, Lecka-Czernik B, Ebraheim NA, Jayasuriya AC (2013) An overview of
recent advances in designing orthopedic and craniofacial implants. J Biomed Mater Res A
101A(11):3349–3364
3. Niinomi M, Nakai M, Hieda J (2012) Development of new metallic alloys for biomedical
applications. Acta Biomater 8:3888–3903
4. Raman A, Dubey M, Gouzman I, Gawalt ES (2006) Formation of Self-assembled monolayers
of alkylphosphonic acid on the native oxide surface of SS316L. Langmuir 22:6469–6472
5. Antunes RA, Lopes de Oliveria MC (2012) Corrosion fatigue of biomedical metallic alloys:
mechanisms and mitigation. Acta Biomater 8:937–962
6. Singh R, Dahotre NB (2007) Corrosion degradation and prevention by surface modification of
biometallic materials. J Mater Sci Mater Med 18:725–751
7. Talha M, Behera CK, Sinha OP (2013) A review on nickel-free nitrogen containing austenitic
stainless steels for biomedical applications. Mater Sci Eng C 33:3563–3575
8. Zhang K, Liu T, Li J, Chen J, Wang J, Huang N (2014) Surface modification of implanted
cardiovascular metal stents: from antithrombosis and antirestenosis to endothelialization.
J Biomed Mater Res A 102A(2):588–609
9. Guduru D, Niepel M, Vogel J, Groth T (2011) Nanostructured material surfaces—preparation,
effect on cellular behavior, and potential biomedical applications: a review. J Artif Organs
34(10):963–985
10. Kubashewski O, Evans EC, Alcock CB (1967) Metallurgical thermochemistry. Pergamon
Press, London
11. Steinemann SG, Perren SM (1984) Titanium alloys as metallic biomaterials. In: Proc. of the
5th world conf. on titanium, vol 2, pp. 1327–1334
12. Zitter H (1976) Schädigung des Gewebes durch metallische Implantate. Unfallheilkunde
79:91
13. Zitter K, Plenk H, Strassl H (1980) Tissue and cell reactions in vivo and in vitro to different
metals for dental implant. In: Heimke G (ed) Dental implants. C. Hanser, Miinchen, p 15
14. Ferguson AB, Akahashi Y, Laing PG, Hodge ES (1962) Characteristics of trace ions released
from embedded metal implants in the rabbit. J Bone Joint Surg 44:323–336
15. Pound BG (2014) Passive films on metallic biomaterials under stimulated physiological condi-
tions. J Biomed Mater Res A 102A:1595–1604
16. Breme J, Steinhäuser E, Paulus G (1988) Commercially pure titanium Steinhäuser plate–screw
system for maxillo facial surgery. Biomaterials 9:310–313
17. Lyndon JA, Boyd BJ, Birbilis N (2014) Metallic implant drug/device combinations for con-
trolled drug release in orthopaedic applications. J Control Release 179:63–75
18. Krekeler G, Schilli W (1984) Das ITI-Implantat Typ H: Technische Entwicklung,
Tierexperiment und klinische Erfahrung. Chirurgische Zahnheilkunde 12:2253–2263
19. Kirsch A (1980) Titan-spritzbeschichtetes Zahnwurzel-implantat unter physiologischer
Belastung beim Menschen. Dt ZahnärztlZ 35:112–114
158 H. Breme et al.

20. Schröder A, van der Zypen E, Sutter F (1981) The reaction of bone, connective tissue and
epithelium to endosteal implants with titanium- sprayed surface. J Maxillofac Surg 1981(9):15
21. Brånemark PI, Adell R, Albrektsson T, Lekholm U, Lundkvist S, Rockier B (1983)
Osseointegrated titanium fixtures in the treatment of edentulousness. Biomaterials 4:25
22. Schröder A, Stich H, Strautmann F, Sutter F (1978) Über die Anlagerung von Osteozement an
einem belasteten Implantatkörper. Schweiz Monatsschr Zahnheilk 4:1051–1058
23. Kydd WL, Daly CH (1976) Bone–titanium implant response to mechanical stress. J Prosthet
Dent 35:567–571
24. Golish SR, Anderson PA (2012) Bearing surfaces for total disc arthroplasty: metal-on-metal
versus metal-on-polyethylene and other biomaterials. Spine J 12:693–701
25. Bose S, Tarafder S (2012) Calcium phosphate ceramic systems in growth factor and drug
delivery for bone tissues engineering: a review. Acta Biomater 8:1401–1421
26. Gittens RA, Scheideler L, Rupp F, Hyzy SL, Geis-Gerstorfer J, Schwartz Z, Boyan BD (2014)
A review on the wettability of dental implant surfaces: II. Biological and clinical aspects. Acta
Biomater 10:2907–2918
27. Purnama A, Hermawan H, Couet J, Mantovani D (2010) Assessing the biocompatibility of
degradable metallic materials: state-of-the-art and focus on the potential of genetic regulation.
Acta Biomater 6:1800–1807
28. Richards RG, Moriarty TF, Miclau T, McClellan RT, Grainger DW (2012) Advances in bioma-
terials and surface technologies. J Orthop Trauma 26(12):703–707
29. Gittens RA, Olivares-Navarrete R, Schwartz Z, Boyan BD (2014) Implant osseointegration
and the role of microroughness and nanostructures: lessons for spine implants. Acta Biomater
10:3363–3371
30. Bjursten LM, Rasmusson L, Oh S, Smith GC, Brammer KS, Jin S (2010) Titanium dioxide
nanotubes enhance bone bonding in vivo. J Biomed Mater Res A 92:1218–1224
31. Tsukimura N, Ueno T, Iwasa F, Minamikawa H, Sugita Y, Ishizaki K et al (2011) Bone integra-
tion capability of alkali- and heat-treated nanobimorphic Ti-15Mo-5Zr-3Al. Acta Biomater
7:4267–4277
32. Thakral GK, Thakral R, Sharma N, Seth J, Vashisht P (2014) Nanosurface—the future of
implants. J Clin Diagn Res 8(5):7–10
33. Schmitz HJ, Gross V, Kinne R, Fuhrmann G, Strunz V (1988) Der Einfluβ unterschiedlicher
Oberflächenstrukturierung plastischer Implantate auf das histologische Zugfestigkeitsverhalten
des Interface. DVM-Vortragsreihe Implantate 7
34. Hayes JS, Richards RG (2010) The use of titanium and stainless steel in fracture fixation.
Expert Rev Med Dev 7:843–853
35. Hayes JS, Richards RG (2010) Surfaces to control tissue adhesion for osteosynthesis with
metal implants: in vitro and in vivo studies to bring solutions to the patient. Expert Rev Med
Dev 7:131–142
36. Webster TJ, Ergun C, Doremus RH, Siegel RW, Bizios R (2000) Enhanced functions of osteo-
blasts on nanophase ceramics. Biomaterials 21(17):1803–1810
37. Webster TJ, Gutwein LG, Tepper F (2008) Increased osteoblast function on nanofibered alu-
mina. In: 26th annual conference on composites, advanced ceramics, materials, and structures
B: ceramic engineering and science proceedings. John Wiley & Sons, Inc., pp. 817–824
38. Strassl H (1978) Experimentelle Studie über das Verhalten von titanbeschichteten Werkstoffen
hinsichtlich der Gewebekompatibilität im Vergleich zu anderen Metallimplanteten. Teil 1.
Osterr Z Stomatol 75(4):134–146
Chapter 1b
Metallic Biomaterials: Cobalt-Chromium
Alloys

Gopinath Mani

1b.1 Introduction

Cobalt-chromium (CoCr) alloys have been extensively used for making various
orthopedic, dental, and cardiovascular implants and devices. These alloys possess
superior mechanical properties with high resistance to corrosion, wear, and fatigue.
The biocompatibility and blood compatibility of these alloys have also been well
demonstrated. In this chapter, the four types of CoCr alloys (ASTM F75 alloy,
ASTM F799 alloy, ASTM F90 alloy, and ASTM F562 alloy) that are currently used
for biomedical implants and devices are discussed. Also, the other types of CoCr
alloy (ASTM F563, ASTM F1058, and Havar) that have been considered and inves-
tigated for biomedical implants and devices are discussed in this chapter. The chap-
ter is divided into subsections based on these individual alloys.

1b.2 ASTM F75

ASTM F75 is a cast cobalt-chromium-molybdenum (Co-28Cr-6Mo) alloy. The uni-

fied numbering system (UNS) number for this alloy is R30075. The chemical com-
position of F75 alloy is provided in Table 1b.1. The different commercial names of
F75 alloy are (a) Vitallium (Howmedica, Inc.); (b) Haynes Stellite 21 (Cabot Corp.);
(c) Protasul-2 (Sulzer AG); and (d) Zimaloy (Zimmer). This alloy possesses excel-
lent corrosion resistance even in highly corrosive chloride environments. This is
mainly due to its bulk composition (Table 1b.1) and the surface chromium oxide
(Cr2O3) layer.

G. Mani (*)
Department of Biomedical Engineering, University of South Dakota,
4800 North Career Avenue, Sioux Falls, SD 57107, USA
e-mail: [email protected]

© Springer Science+Business Media New York 2016 159

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_15
160 G. Mani

Table 1b.1 Chemical composition (wt.%) of F75 alloya

Co Cr Mo W Ni Mn Si Fe C N P Al S
58.9– 27.0– 5.0– 0.2 2.5 1.0 1.0 0.75 0.35 0.25 0.02 0.30 0.01
69.5 30.0 7.0
a
Values were taken from Refs. [1–6]

This alloy is made into the required shape of the final device by a process known
as investment casting [1–4]. During this process, initially, a wax mold is made rep-
licating the shape of the final device. Then, the mold is coated with a ceramic fol-
lowed by melting the wax to leave a ceramic mold. After that, the alloy is melted at
a temperature range of 1350–1450 °C, and poured into the ceramic mold. Once the
metal is solidified, the ceramic mold is broken away, and the metal is further pro-
cessed to obtain the final device.
The way how the casting conditions are carried out can produce different micro-
structural features and can adversely affect the mechanical properties of the alloy
[1–4]. Firstly, as-cast F75 alloy typically consists of a Co-rich matrix (alpha-phase)
with carbides (Co23C6, Cr23C6, and Mo23C6) in interdendritic spaces and grain
boundaries. The relative amounts of alpha and carbide phases in the alloy are
expected to be 85 % and 15 %, respectively. However, if there are any non-
equilibrium conditions happened during cooling, the interdendritic regions become
rich in solutes such as Cr, Mo, and C, along with carbides, and the dendrites will be
rich in Co with lack of Cr. This situation is considered to be unfavorable since the
regions with lack of Cr can act as an anode with respective to the rest of the alloy
microstructure. However, this situation can be improved by solution annealing at
1225 °C for 1 h following casting. Secondly, the solidification step in the casting
process not only creates dendrites but also large-size grains. This is not favorable
since large-size grains can significantly decrease the yield strength of the alloy.
Thirdly, defects may occur in the alloy during casting. For instance, if a ceramic
mold piece is broken and entrapped in the alloy during its solidification, it could
lead to stress concentrations and initiate cracks at the site where the ceramic piece
is present. Such defects can cause fatigue fracture of the device under in vivo condi-
tions. For the similar reasons, it is vital to avoid any porosity (macro- or micro-
pores) in the alloy during casting conditions.
To circumvent the above-mentioned limitations and to improve the microstruc-
tural features and mechanical properties of the alloy, techniques such as hot isotac-
tic pressing (HIP) have been developed [1–4]. In HIP, a fine powder of F75 alloy is
compacted and sintered together under a pressure of 100 MPa at 1100 °C for 1 h,
and then forged to obtain the final shape. The alloy prepared using HIP typically
showed a much smaller grain size (approximately 8 μm), which not only provided
higher yield strength but also improved fatigue properties when compared to those
of as-cast alloy. The mechanical properties of as-cast and HIP F75 alloys are pro-
vided in Table 1b.2. F75 alloy is commonly used to make total hip arthroplasty
components (femoral stems, modular head, and acetabular cup liner) and total knee
arthroplasty components (metal femoral and tibial components). Also, this alloy is
commonly used to make dental implants such as partial dentures and bridgeworks.
1b Metallic Biomaterials: Cobalt-Chromium Alloys 161

Table 1b.2 Mechanical properties of F75 alloya

Young’s Yield Tensile Fatigue
modulus strength strength strength
F75 alloy types (GPa) (MPa) (MPa) (MPa)
As-cast/annealed 210 448–517 655–899 207–310
Powder metallurgy/hot 253 841 1277 725–950
isotactic pressing
a
Values were taken from Refs. [1–5]

Table 1b.3 Chemical composition (wt.%) of F799 alloya

Co Cr Mo Ni Mn Si Fe C N
58–59 26.0–30.0 5.0–7.0 1.0 1.0 1.0 1.5 0.35 0.25
a
Values were taken from Refs. [1–6]

1b.3 ASTM F799

ASTM F799 is a thermomechanical cobalt-chromium-molybdenum (Co-28Cr-

6Mo) alloy. The UNS number for this alloy is R31537. The chemical composition
of F799 alloy is provided in Table 1b.3. The alloy is mechanically processed by hot
forging at 800 °C after casting to obtain the final shape of the device [1–4]. The
thermomechanical processing is the main difference between this alloy and F75
alloy. The processing is carried out in several steps to obtain the final desired shape,
with different temperatures used at each step. Although a higher temperature can
produce greater deformation, it can also lead to loss of strengthening. Hence, only
during the initial stages of processing a higher temperature is applied to allow easier
deformation followed by processing at lower temperature to induce cold working.
By doing this, the final product of the alloy is obtained with very high strengths.
The microstructure of F799 alloy shows a more worked grain structure when
compared to that of as-cast F75 alloy [1–4]. Also, a hexagonal close-packed (HCP)
phase is formed through a shear-induced transformation of face-centered cubic
(FCC) matrix to HCP platelets. The mechanical properties of F799 alloy are pro-
vided in Table 1b.4. As can be seen from this table, the yield and tensile strengths of
this alloy are significantly greater than that of as-cast and HIP F75 alloys. Similar to
F75 alloy, F799 alloy is also used for the different components of orthopedic and
dental implants and devices.

1b.4 ASTM F90

ASTM F90 is a wrought cobalt-chromium-tungsten-nickel (Co-20Cr-15W-10Ni)

alloy. The UNS number for this alloy is R30605. The chemical composition of
F90 alloy is provided in Table 1b.5. The commercial name of F90 alloy is
162 G. Mani

Table 1b.4 Mechanical properties of F799 alloya

F799 alloy Young’s modulus Yield strength Tensile strength Fatigue strength
type (GPa) (MPa) (MPa) (MPa)
Hot forged 210 896–1200 1399–1586 600–896
a
Values were taken from Refs. [1–5]

Table 1b.5 Chemical composition (wt.%) of F90 alloya

Co Cr W Ni Mn Si Fe C P S
45.5–56.2 19.0–21.0 14.0–16.0 9.0–11.0 2.0 0.4 3.0 0.15 0.04 0.03
a
Values were taken from Refs. [1–6]

Table 1b.6 Mechanical properties of F90 alloya

Young’s modulus Yield strength Tensile strength Fatigue strength
F90 alloy types (GPa) (MPa) (MPa) (MPa)
Annealed 210 448–648 951–1220 Not available
Cold worked to 210 1606 1896 586
44 %
a
Values were taken from Refs. [1–5]

Haynes 25 (L605). In this alloy, tungsten (W) and nickel (Ni) are added to improve
the machinability and fabrication properties. The yield and tensile strengths of F90
alloy in annealed state are equivalent to those of F75 alloys. However, when the F90
alloy is cold worked to 44 %, the yield and tensile strengths significantly increased
when compared to those of F75 alloys. The mechanical properties of F90 alloys in
annealed and cold worked states are provided in Table 1b.6. F90 alloy is commonly
used to make cardiovascular implants such as stents and mechanical heart valves.
Also, this alloy is used in orthopedic applications such as cerclage cables and
guide rods.

1b.5 ASTM F562

ASTM F562 is a wrought cobalt-nickel-chromium-molybdenum (Co-35Ni-20Cr-

10Mo) alloy. The UNS number for this alloy is R30035. The chemical composition
of F562 alloy is provided in Table 1b.7. This alloy is also known as MP35N. The
“MP” in the name refers to multiple phases in its microstructure. This alloy can be
heat treated and cold worked to produce a high-strength alloy with a controlled
microstructure [1–4].
Under equilibrium conditions, the pure solid cobalt has a face-centered cubic
(FCC) crystal structure above 419 °C and a hexagonal close-packed (HCP) crystal
structure below 419 °C. When cobalt is alloyed with other elements for making the
1b Metallic Biomaterials: Cobalt-Chromium Alloys 163

Table 1b.7 Chemical composition (wt.%) of F562 alloya

Co Cr Mo Ni Mn Si Fe C P Ti S
29.0–38.8 19.0–21.0 9.0–10.5 33.0– 0.15 0.15 1.0 0.025 0.015 1.0 0.010
37.0
a
Values were taken from Refs. [1–6]

Table 1b.8 Mechanical properties of F562 alloya

Young’s modulus Yield strength Tensile strength Fatigue strength
F562 alloy types (GPa) (MPa) (MPa) (MPa)
Hot forged 232 965–1000 1206 500
Cold worked and 232 1500 1795 689–793
aged
a
Values were taken from Refs. [1–5]

F562 alloy, the processing involves cold working to 50 %, which greatly supports
the transformation of FCC to HCC crystal structure. The HCP phase exists as plate-
lets within the FCC grains to significantly strengthen the alloy. The alloy can also
be further strengthened by a heat treatment of 430–650 °C. This heat treatment
produces cobalt molybdenum (Co3Mo) precipitates on the HCP platelets. Hence,
this alloy is truly a multiphase material. The higher strength of the alloy is mainly
due to the combination of a variety of processing methods such as cold working,
solid solution strengthening, and precipitation hardening. The mechanical proper-
ties of F562 alloy in hot-forged, and cold-worked and aged forms are provided in
Table 1b.8. As can be seen from this table, the tensile strength of cold-worked and
aged F562 alloy is greater than that of all the other types (F75, F799, and F90) of
CoCr alloy that are used for biomedical applications. F562 alloy is used for cardio-
vascular stents, lead wires for pacemakers, and heart valve components.

1b.6 Other Types of CoCr Alloy Investigated

for Biomedical Applications

Although the above-discussed four types (F75, F799, F90, and F562) of CoCr alloy
are currently used for making various biomedical implants and devices, the few
other types of CoCr alloy such as ASTM F563, ASTM F1058, and UNS R30004
have also been investigated for biomedical applications. Each of these alloys (F563,
F1058, and UNS R30004) has been shown to produce no harmful effects such as
cytotoxicity, systemic toxicity, hemolysis, and pyrogenicity in intracutaneous irrita-
tion, intramuscular implantation, and skin sensitization tests [1].
164 G. Mani

1b.7 ASTM F563

ASTM F563 is a wrought cobalt-nickel-chromium-molybdenum-tungsten-iron

(Co-20Ni-20Cr-3.5Mo-3.5W-5Fe) alloy. The UNS number for this alloy is R30563.
The chemical composition of F563 alloy is provided in Table 1b.9.
F563 alloy can be significantly strengthened by cold working, and cold working
and aging. The mechanical properties of the alloy are provided in Table 1b.10. F563
alloy was initially made in the form of bars, wires, and forgings for implant manu-
facture. However, this specification was withdrawn in 2005 since it was determined
that this alloy is no longer being used for implants.

1b.8 ASTM F1058

ASTM F1058 is a wrought cobalt-chromium-iron-nickel-molybdenum (40Co-20Cr-

16Fe-15Ni-7Mo) alloy. This alloy is available in two grades, grade 1 (UNS No.
R30003) and grade 2 (UNS No. R30008). The chemical composition of grades 1
and 2 of F1058 alloy is provided in Tables 1b.11a and 1b.11b, respectively.
The grade 1 of F1058 alloy is commonly known as Elgiloy. This alloy has been
used for making certain components (springs) of artificial heart. The grade 2 of
F1058 alloy has been used for making carotid artery clamps for use in neurosurgery
and vascular surgery. The tensile and yield strengths of cold-worked and aged
F1058 wire for the different diameters of the wire are provided in Table 1b.12.

1b.9 UNS R30005

UNS R30005 is commonly known as Havar. The chemical composition of Havar is

provided in Table 1b.13. The mechanical properties of Havar are provided in
Table 1b.14. As can be seen from the table, the yield and tensile strengths of cold-
rolled or cold-rolled and aged Havar are greater than those of most of the other
CoCr alloy types that are used for biomedical applications.

Table 1b.9 Chemical composition (wt.%) of F563 alloya

Co Cr Mo W Ni Mn Si Fe C Ti S
Balance 18.0– 3.0–4.0 3.0–4.0 15.0– 1.0 0.50 4.0–6.0 0.05 0.5– 0.010
22.0 25.0 3.50
a
Values were taken from Refs. [1, 2, 5]
1b Metallic Biomaterials: Cobalt-Chromium Alloys 165

Table 1b.10 Mechanical properties of F563 alloya

Young’s modulus Yield strength Tensile strength
F563 alloy types (GPa) (MPa) (MPa)
Fully annealed 230 276 600
Cold worked or cold worked and 230 827–1172 1000–1310
aged (medium hard–hard)
Cold worked and aged (extra hard) 230 1310 1586
a
Values were taken from Refs. [1, 5, 7]

Table 1b.11a Chemical composition (wt.%) of F1058 alloy (grade 1)a

Co Cr Mo Ni Mn Si Fe C P Be S
39.0– 19.0– 6.0– 14.0– 1.5– 1.20 Balance 0.15 0.015 0.10 0.015
41.0 21.0 8.0 16.0 2.5
a
Values were taken from Refs. [1, 5, 8]

Table 1b.11b Chemical composition (wt.%) of F1058 alloy (grade 2)a

Co Cr Mo Ni Mn Si Fe C P Be S
39.0– 18.5– 6.5– 15.0– 1.0– 1.20 Balance 0.15 0.015 0.001 0.015
42.0 21.5 7.5 18.0 2.0
a
Values were taken from Refs. [1, 5, 8]

Table 1b.12 Mechanical properties of F1058 alloya

F1058 alloy wire diameters (mm) Yield strength (MPa) Tensile strength (MPa)
0.02–0.12 – 2275
0.12–1.00 1450 2000
1.00–1.50 1380 1965
1.50–2.00 1380 1895
2.00–2.50 1345 1895
2.50–3.00 1275 1860
3.00–3.50 1240 1860
a
Values were taken from Refs. [1, 5, 7]

Table 1b.13 Chemical composition (wt.%) of Havara

Co Cr Mo W Ni Mn Fe C
42.5 19.5 2.4 2.7 12.7 1.6 Balance 0.2
a
Values were taken from Refs. [1, 9, 10]
166 G. Mani

Table 1b.14 Mechanical properties of Havara

Havar types Yield strength (MPa) Tensile strength (MPa) Hardness
Annealed 483 965 25
Cold rolled 1724 1862 50
Cold rolled and aged 2069 2275 60
a
Values were taken from Refs. [1, 9, 10]

References

1. Davis JR (2003) Handbook of materials for medical devices. ASM International, Materials
Park
2. Park JB, Kim YK (2003) Metallic biomaterials. In: Park JB, Bronzino JD (eds) Biomaterials
principles and applications. CRC Press, Boca Raton, pp 1–20
3. Brunski JB (2004) Metals. In: Ratner BD, Hoffman AS, Schoen FJ, Lemons JE (eds)
Biomaterials science: an introduction to materials in medicine. Elsevier Academic Press,
London, pp 137–153
4. Agrawal CM, Ong JL, Appleford MR, Mani G (2014) Introduction to biomaterials basic the-
ory with engineering applications. Cambridge University Press, New York
5. Breme J, Biehl V (1998) Metallic biomaterials. In: Black J, Hastings G (eds) Handbook of
biomaterial properties. Springer, New York, pp 135–213
6. Klarstrom D, Crook P (2001) Cobalt alloys: alloy designation system. In: Buschow KHJ, Cahn
R, Flemings M, Ilschner B, Kramer E, Mahajan S, Veyssiere P (eds) Encyclopedia of materi-
als: science and technology. Pergamon, Headington Hill Hall, pp 1279–1280
7. Chen Q, Thou G (2015) Biomaterials: a basic introduction. CRC Press, Boca Raton
8. Narushima T, Ueda K, Alfirano (2015) Co-Cr alloys as effective metallic biomaterials. In:
Niinomi M, Narushima T, Nakai M (eds) Advances in metallic biomaterials tissues, materials
and biological reactions. Springer, New York, pp 157–178
9. Robinson M (2005) Havar®, a Co-Cr biocompatible alloy for medical implants, encyclopedia
of materials: science and technology. Elsevier, pp 1–6
10. Robinson M (2004) Havar®, a new, old Co-Cr biocompatible alloy for implants. In: Shrivastava
S (ed) Medical device materials proceedings of the materials & processes for medical devices
conferences. ASM International, Materials Park, pp 324–328
Chapter 1c
Metallic Biomaterials: Titanium
and Titanium Alloys

H. Breme, V. Biehl, Nina Reger, and Ellen Gawalt

1c.1 Composition
Table 1c.1 Comparison of international standards for titanium and titanium alloys (Refs. 1, 2a)
International
United Organization for United
Germany Kingdom Standardization States Japan
Alloy DIN BS France AIR ISO ASTM JIS
Ti-1 17850 2TA1 9182 T-35 5832/II F67/Grade 1 H4600
Ti-2 17850 2TA2 9182 T-35 F67/Grade 2
Ti-3 17850 2TA6 9182 T-50 F67/Grade 3
Ti-4 17850 2TA6, 2AT7 9182 T-60 F67/Grade 4
2AT8, 2AT9
Ti6Al4V 17851 2TA10, 9183T-A6V 5832/III F136-13 H4607
2TA11
2TA13,
2TA28
TA56, TA59
Ti5Al2.5Fe 17865 BS7252–10 – 5832/X – –
Ti6Al7Nb 17851 S94–081 5832/XI Fl295 –
a
Ti-l–Ti-4 = commercially pure unalloyed titanium; others: direct chemical composition

H. Breme • V. Biehl
Lehrstuhl für Metallische Werkstoffe, Universität des Saarlandes, Saarbrücken, Germany
N. Reger • E. Gawalt (*)
Department of Chemistry and Biochemistry, Bayer School of Natural and Environmental
Sciences, Mellon 337, 600 Forbes Ave., Pittsburgh, PA 15282, USA
e-mail: [email protected]

© Springer Science+Business Media New York 2016 167

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_16
168 H. Breme et al.

Table 1c.2 Chemical composition of commercially pure (cp)-titanium (wt%) (Refs. 1–3)
cp-Ti Fe max. O approx. N max. C max. H max. Ti
Grade 1 0.2 0.18 0.03 0.1 0.015 Balance
Grade 2 0.3 0.25 0.03 0.08 0.015 Balance
Grade 3 0.3 0.35 0.05 0.1 0.015 Balance
Grade 4 0.5 0.40 0.05 0.1 0.015 Balance

Table 1c.3 Chemical composition of (α + β)-titanium alloys (wt%) (Ref. 1)

Others
Alloy Al V Fe Nb Ta O N C H Single Sum Ti
Ti6Al4V 5.5–6.5 3.5–4.5 0.25 – – 0.13 0.05 0.08 0.012 0.1 0.4 Balance
Ti5Al2.5Fe 4.5–5.5 – 2.0–3.0 – – 0.2 0.05 0.08 0.015 0.1 0.4 Balance
Ti6Al7Nb 5.5–6.5 – 0.25 6.5–7.5 0.5 0.2 0.05 0.08 0.009 – – Balance

Table 1c.4 Chemical composition (wt%) of β and near-β titanium alloys (Refs. 4–7)
Alloy Al Mo Zr Ta Nb Sn Ti
Til5Mo5Zr3Al 3.8 15 5 77
Til2Mo5Zr5Sn 12 5 4.5 Balance
Ti30Nb 30 70
Ti30Ta 29.6 Balance

1c.2 Physical Properties

Table 1c.5 Physical Young’s modulus 102.7 GPa

properties of cp-Ti grade 1
Density 4.51 g/cm3
(Refs. 8–11)
Melting point 1670 °C
Boiling point 3260 °C
Transformation temperature α–β 890 °C
Crystal structure >882 °C β bcc
<882 °C α cph
Magnetic properties Paramagnetic
Heat of transformation 67 kJ/kg
Thermal neutron-capture cross 5.8 × 10–22 cm2
section
Specific heat at 15 °C 0.52 kJ/kg K
Heat of fusion 419 kJ/kg
Thermal conductivity at room 17 W/mK
temperature
Thermal expansion coefficient 9.6 × 10–6 °C−1
between 0 and 315 °C
Specific heat resistivity at 20 °C 0.5 μΩm
1c Metallic Biomaterials: Titanium and Titanium Alloys 169

Table 1c.6 Physical properties of (α + β)-titanium alloys (Refs. 4, 8–9, 12–14)

Property Ti6Al4V Ti5Al2.5Fe Ti6Al7Nb
Young’s modulus (GPa) 100–110 110 105
Density (g/cm3) 4.43 4.45 4.52
Microstructure A1–A9 A1–A9 A1–A9
Transformation temperature (°C) 990 ± 15 950 ± 15 1010 ± 15
Thermal conductivity at room temperature (W/mK) 6.6 – 7
Coefficient of thermal expansion 9.5 9.3 –
between 30 and 200 °C (×10–6 K−1)
Specific heat at 20 °C (kJ/kg K) 0.56 – 0.54
Specific electrical resistivity at 20 °C (52 mm2/m) 1.66 – –

Table 1c.7 Influence of alloying elements and heat treatment on Young’s modulus (Refs. 5,
15–18)
Alloy Heat treatment Young’s modulus (GPa)
Ti30Ta As rolled 69
1 h/1000 °C/H2O 63
Ti30Nb As rolled 80
Til5Mo5Zr3Al ELIa Solution heat treated at 840 °C 75
Solution heat treated at 740 °C 88
Solution heat treated at 740 °C 113
+ aged at 600 °C
a
ELI = extra-low interstitial

1c.3 Processing of cp-Ti and Ti Alloys

Provided that the following characteristics of titanium are taken into consideration,
almost all processing procedures are possible:
1. High affinity to oxygen, nitrogen, and hydrogen gases
2. High reactivity to all metals to produce intermetallic compounds
3. Relatively low Young’s modulus and therefore backspringing
4. Relatively low thermal conductivity
5. Tendency to stick to tools

1c.3.1 Hot Working and Heat Treatment

Titanium and titanium alloys are fabricated into semifinished products by conven-
tional methods such as forging, rolling, pressing, and drawing. When Ti materials
are heated, care must be taken to avoid an excessive adsorption of oxygen, nitrogen,
and hydrogen. Therefore, heating and annealing should take place in a neutral or
170 H. Breme et al.

slightly oxidizing atmosphere. During heating in a gas-fired furnace, direct contact

with the flame must be avoided because of the risk of hydrogen pickup and local
overheating. In a short heating period, oxygen pickup is restricted to the surface
area. This surface zone must be removed by chemical or mechanical methods.
Hydrogen is able to penetrate the matrix rapidly; therefore, a reducing atmosphere
must be avoided. The hot working temperature depends on the alloy composition
and should be selected to obtain the best mechanical properties and grain structure
(A1–A9 according to ETTC2, Ref. 17). Table 1c.8 summarizes the deformation
temperatures for the various Ti materials. Table 1c.9 lists the temperature ranges
and recommended annealing times for stress relieving, soft annealing, and solution
treating with age hardening. When the cross section is very small, annealing is
favorably carried out in a high-vacuum furnace. Prior to this annealing treatment,
the oxide film must be removed from the surface to avoid diffusion of oxygen into
the material.

1c.3.2 Working of Sheet

At room temperature cp-Ti grades 1 and 2 can be worked very well, and grades 3
and 4 only moderately well. Titanium alloys, because of their high yield/tensile
strength ratio, can be worked only under certain conditions.

Table 1c.8 Deformation Alloy Deformation temperature (°C)

temperatures for various
cp-Ti Grade 1 650–870
titanium materials
(Refs. 19, 20) cp-Ti Grade 2 650–870
cp-Ti Grade 3 700–900
cp-Ti Grade 4 700–930
Ti6Al4V 700–1100

Table 1c.9 Recommendation for the heat treatment of cp-Ti and Ti alloys (Refs. 2, 8)
Solution annealing + age
Alloy Stress relief Soft annealing hardening
Ti grade 1 15 min–4 h at 480–595 °C/air 6 min–2 h at −
Ti grade 2 650–750 °C/air
Ti grade 3
Ti grade 4
Ti6A14V 3 min/mm Min. 1 h 15 min–1 h at
1–4 h at 480–650 °C/air max. 4 h at 955–970 °C/H2O
700–850 °C +2 h at 480–595 °C/H2O
air or furnace cool
Ti5A12.5Fe 3 min/mm Min. 10 min 15–1 h at 800–920 °C/H2O
15 min–4 h at 540–650 °C/air max. 4 h at +2–4 h at 480–600 °C/air
720–845 °C
air or furnace cool
1c Metallic Biomaterials: Titanium and Titanium Alloys 171

For deep drawing, special coatings in the form of polymer foils have proved to
be effective. At high temperatures colloidal graphite and common hot press greases
with graphite or molybdenum disulfide additives have been successful. The Fe-,
Ni-, and Cr-contents should be limited to 0.05, 0.1, and 0.33 wt%, respectively, to
allow during a short annealing treatment (1–5 min at 750–850 °C) a grain growth
producing an average grain size of 50–70 μm. Due to this grain growth a deforma-
tion by twinning, with a resulting increased deep drawability, occurs (Ref. 19).
Superplastic forming is a material-saving and cost-reducing process for manufac-
turing parts of a complicated shape because it can be carried out together with diffu-
sion welding in a single operation. Fine-grained alloys like Ti6Al4V and Ti5Al2.5Fe
can be used for superplastic deformation. Other special deformation processes such
as stretch forming, spinning, or explosion forming are also possible.

1c.3.3 Descaling

The average thickness of the oxide layer on the surface of cp-Ti as a function of
temperature can be found in Table 1c.10 and the composition of the oxide layer on
cp-Ti and Ti alloys can be found in Table 1c.11. The oxide layer on the surface of
thick-walled pieces generated during deformation and/or heat treatment is removed
by sandblasting and/or pickling. The workpiece is treated in an aqueous solution of
20 wt% HNO3 and 2 wt% HF.
Thin-walled pieces are merely pickled in an electrolytic solution or salt bath. It
is important that not only the surface layer of oxide but also the underlying oxygen
enriched diffusion zone is removed. Otherwise, the machinability and the service
life of turning and milling tools would be negatively affected.

Table 1c.10 Thickness of Temperature

oxide formed on the surface (°C) Measurable thickness (mm)
of cp-Ti as a function of
315 None
temperature (Ref. 2)
425 None
540 None
650 <0.005
705 0.005
760 0.008
815 <0.025
870 <0.025
925 <0.05
980 0.05
1040 0.10
1095 0.36
172 H. Breme et al.

Table 1c.11 Nature of oxides formed on cp-Ti and titanium alloys (Ref. 22)
Oxide
Alloy TiO2 Al2O3 Nb2O5 MoO3/MoO2 ZrO2
cp-Ti ×
Ti6Al4V × ×
Ti5Al2.5Fe × ×
Ti6Al7Nb × × ×
Ti15Mo5Zr3Al × × × ×

1c.3.4 Machining

The machining of titanium materials presents no difficulties provided that the

following properties are taken into account:
1. Relatively low thermal conductivity which may cause high thermal stresses at
the cutting edge of the tool
2. Low Young’s modulus which applies pressure to the tool
3. Tendency to stick to the tool
Titanium materials must be machined at low cutting speeds, at a relatively high
feed rate, and with an ample supply of coolant (sulfur-containing oil; mixture of
tetrachloride carbon, molybdenum sulfide, and graphite; 5 % aqueous solution of
sodium nitrite, 5–10 % aqueous solution of water-soluble oil or sulfurized chlori-
nated oil). The cutting tools should be sharp and mounted as rigidly as possible.
Recommended parameters for turning and milling are given in Table 1c.12. Since
titanium dust and chips can easily catch fire, safety precautions must be taken.
Threads should be cut on a lathe, as thread-cutting discs are subject to seizure.
Sawing causes no difficulties if a high blade pressure is used and the coolant supply
is sufficient. Coarse-toothed blades (4 teeth per inch) are recommended.
For grinding, aluminum oxide (5–10 m/s) and silicon carbide (20–30 m/s) can be
used.

1c.3.5 Soldering and Brazing

Immediately before soldering and brazing, the oxide layer present on the surface of
titanium material must be removed. For direct applications using a torch, alumi-
num–zinc and tin–zinc solders are suitable. The higher temperatures required for
brazing present the difficulty of avoiding the formation of intermetallic phases. As
with almost all metals, titanium forms brittle intermetallic phases in the fusion zone.
The only exception is silver, so that this metal forms one of the main constituents of
brazers. The sources of heat used for brazing and soldering are the acetylene torch,
high-frequency induction coils, infrared radiation, an inert-gas-shielded arc with
graphite or tungsten electrodes, furnaces with an argon atmosphere (min. 99.99 %
1c Metallic Biomaterials: Titanium and Titanium Alloys 173

Table 1c.12 Recommendations for the cutting and milling of cp-Ti and Ti alloys (Ref. 8)
Cutting speed Feed (mm/ Cutting Relief
Alloy (m/min) rev) Depth of cut (mm) angle angle
Rough cutting
cp-Ti TC 30–75 0.2–0.4 0–6° 6–8°
HSS 7.5–4.0 0.1–1.25 6–15° 5–7°
>2.5
Ti alloys TC 15–25 0.2–0.4 0–6° 6–8°
HSS 3–15 0.1–0.4 6–15° 5–7°
Forecutting
cp-Ti TC 60–100 0.1–0.4 6–6° 6–8°
HSS 18–50 0.075–0.2 6–15° 5–7°
0.75–2.5
Ti alloys TC 20–50 0.1–0.4 6–6° 6–8°
HSS 5–15 0.075–0.2 6–15° 5–7°
Finish cutting
cp-Ti TC 60–100 0.075–0.3 0–15° 6–8°
HSS 20–50 0.05–0.1 5–6° 5–7°
0.1–0.75
Ti alloys TC 20–70 0.075–0.3 0–15° 6–8°
HSS 9–15 0.05–0.1 5–6° 5–7°
Milling
cp-Ti TC 25–30 0.07–0.15
HSS 50–60 0.1–0.2 >1.25 face cutter – –
Ti alloys TC 7.5–20 0.07–0.2 >2.5 gear cutter
HSS 15–30 0.1–0.2
TC hard metal (tungsten carbide)
HSS high-speed steel

and/or a dew point below −50 °C), and high-vacuum furnaces. If brazing is not
performed under vacuum or in a controlled atmosphere, fluxes are necessary to
dissolve the oxide layers and prevent a pickup of gases.

1c.3.6 Welding

The inert-gas-shielded arc processes (TIG and MIG) are mainly used for fusion weld-
ing. In special cases resistance, ultrasonic, electron beam, diffusion, and laser welding
are applied. With the cp-Ti grades the weld attains mechanical properties approximat-
ing those of the base metal. A slight decrease in ductility may occur with high tensile
grades. Under passivating conditions, titanium welds have the same corrosion resis-
tance as the base metal. On the contrary, in reducing media the weld may be subjected
to a more severe corrosive attack than the base metal. During the welding operation,
the weld, the heat-affected zone, and the underside of the weld are shielded from the
atmosphere. The filler rod used is an uncoated wire of the same grade or of a grade
174 H. Breme et al.

with a lower hardness than the base metal. Careful preparation of the joint is neces-
sary; that is, surface impurities must be removed by grinding or pickling in order to
avoid porosity. Even fingermarks can produce a hardening of the weld. A single layer
can weld sheets up to 2.5 mm thick. In order to avoid local oxygen concentrations
oxidation products, such as those found at the tip of the electrode, must be cut off. The
effectiveness of the inert gas is responsible for the welding rate. The optimum argon
flow rate has proved to be about 6–8 l/mm. After welding, the appearance of a dark
blue or gray oxide layer indicates an insufficient inert gas shielding and an embrittle-
ment of the weld due to oxygen and/or nitrogen pickup. The hardness of a good weld
may exceed that of the fully recrystallized base metal by a maximum of 50 VHN. If,
after a slight grinding of the surface, a hardness test should give a higher value, the
weld must be completely removed because of embrittlement.
Electron beam welding is particularly suitable for titanium materials. It offers
many advantages such as very narrow seams and small heat-affected zones, weldabil-
ity of thick diameters, high welding speed, and reproducibility of even complex welds.
Titanium materials can be spot welded without any particular preparation under
similar conditions to those of stainless steel. Using flat-tipped electrodes, spot weld-
ing can be performed without inert gas. A hardening of the zone by up to 50 VHN
compared with the base metal is regarded as normal and does not diminish the
strength of the joint. Seam and flashbutt welding are also possible if an argon atmo-
sphere is used.
Diffusion welding is of particular importance for titanium materials because these
materials are more amenable to a homogeneous band in the solid state than other
metals. After welding, the joint zone shows a higher temperature under high vacuum
or, in an inert atmosphere, a microstructure very similar to that of the base metal.

1c.4 Mechanical Properties

Table 1c.13 Mechanical properties of commercially pure titanium (Ref. 3, 21)

Tensile
yield Ultimate Ratio
strength tensile yield/ Elongation Reduction Bend radius (105°)
(0.2 %) strength tensile at fractureb of areab Brinell for sheet thickness
Titaniuma (MPa) (MPa) strength (%) (%) hardness (T)
>1.8–
<1.8 mm 4.75 mm
Grade 1 170 240 0.71 24 30 110–170 3T 4T
Grade 2 275 345 0.80 20 30 140–200 4T 5T
Grade 3 380 450 0.84 18 30 140–200 4T 5T
Grade 4 483 550 0.88 15 25 200–275 5T 6T
a
Condition: Sheet, as rolled
b
Minimum values
1c Metallic Biomaterials: Titanium and Titanium Alloys 175

Table 1c.14 Influence of a cold deformation on the mechanical properties of commercially pure
titanium (Ref. 23)
Tensile yield Elongation
Condition strength Ultimate tensile Ratio yield/ at fracture
cp-Titanium (%) (0.2 %) (MPa) strength (MPa) tensile strength (%)
Ti grade 1 30 555 635 0.87 18
40 560 645 0.87 16
55 605 710 0.85 15
60 620 725 0.86 14
65 640 730 0.88 14.5
Ti grade 2 30 605 680 0.89 18
40 645 740 0.87 17
50 680 780 0.87 16
60 685 795 0.86 16
65 692 810 0.85 16.5

Table 1c.15 Mechanical properties of β and near-β-titanium alloys (experimental alloys) (Refs. 6,
15, 24–26)
Tensile yield Ratio yield/ Elongation
strength Ultimate tensile tensile at fracture Reduction
Alloy (0.2 %) (MPa) strength (MPa) strength (%) of area (%)
Til5Mo5Zr3Al 838 852 0.98 25 48
Ti30Nb 500 700 0.71 20 60
Ti30Ta 590 740 0.80 28 58

Table 1c.16 Influence of heat treatment on the mechanical properties of β- and near-β-titanium
alloys (Refs. 6, 15)
Tensile
yield Ultimate Ratio
strength tensile yield/ Elongation
(0.2 %) strength tensile at fracture Reduction
Alloy Condition (MPa) (MPa) strength (%) of area (%)
Til5Mo5Zr3Al SHT at 840 °C 870 882 0.99 20 83.2
SHT at 740 °C 968 975 0.99 16.9 64.5
SHT at 740 °C 1087 1099 0.99 15.3 57.5
+ aged
at 600 °C
45 % cold worked 1284 1312 0.98 11.3 43.8
+ aged
at 600 °C
Ti30Ta Annealed at 1100 °C 650 800 0.81 8 42
Annealed at 1200 °C 660 800 0.83 8 38
Annealed at 1300 °C 665 800 0.83 8 30
Annealed at 1400 °C 680 800 0.83 6 18
SHT solution heat treatment
 176

Table 1c.17 Mechanical properties of (α + β)-titanium alloys (Refs. 2, 27–29)

Bend radius
Condition sheet Tensile yield (105°) for sheet
as rolled strength Ultimate tensile Ratio yield/ Elongation at Reduction thickness (T)
Alloy thickness (mm) (0.2 %) (MPa) strength (MPa) tensile strength fracture (%) of area (%) Hardness <2 mm <2–5 mm
Ti6A14V −6 834 937 0.92 19 46 310 9T 10T
6–100 951 1117 0.90 17 60 310
Ti5A12.5Fe −6 780 860 0.91 8 25 310 9T 10T
Ti6A17Nb Extruded 811– 869– 0.93– 7–13 24–44
hot 952 1008 0.94
rolled 943– 1016– 0.93 11–16 40–55
+ hot 1008 1086
forged
H. Breme et al.
1c Metallic Biomaterials: Titanium and Titanium Alloys 177

Table 1c.18 Influence of a solution treating and ageing on the mechanical properties of Ti6Al4V
Tensile
yield Ultimate
strength tensile Ratio yield/ Elongation
(0.2 %) strength tensile at fracture Reduction
Condition (MPa) (MPa) strength (%) of area (%)
Sheet ≤ 12.5 mm 15–60 min 1070 1140 0.94 8 20
at 800–920 °C/H2O + 2–4 h
480–600 °C/air

Table 1c.19 Influence of a plasma nitriding (PVD) on the mechanical properties of Ti6Al4V (Ref. 30)
Tensile yield Elongation
strength (0.2 %) Ultimate tensile Ratio yield/ at fracture
Treatment (MPa) strength (MPa) tensile strength (%)
Untreated 809 894 0.90 20
Vacuum annealed 815 924 0.88 21
20 h/850 °C
Plasma nitrideda 805 914 0.88 20
20 h/850 °C/N2
Plasma nitrideda 880 984 0.89 20
20 h/850 °C/NH3
a
Plasma nitriding at 20–40 kW

Table 1c.20 Fracture toughness of Ti-alloys (Refs. 6, 31)

Alloy Condition Fracture toughness KIC (N/mm3/2)
Ti6Al4V Annealed 1740
Solution treated + annealed 2020
Ti5Al2.5Fe Annealed (2 h/700 °C/air) 1225
Solution treated + annealed 1785
(1 h/900 °C/H2O/2 h/700 °C/air)
Til5Mo5Zr3Al Solution treated at 740 °C 4580
Solution treated at 740 °C 2430
+ annealed at 600 °C
40 % cold worked 980
+ annealed at 600 °C
178 H. Breme et al.

1c.5 Fatigue

Table 1c.21 High cycle fatigue strength σB and rotating

fatigue strength σR of pure titanium and titanium alloys
(Wöhler curves) (Refs. 15, 32–37)
Alloy R σB (MPa) R σR (MPa)
cp-Ti −1 230–280 −1 200
Ti6Al4V −1 610–625 −1 500–660
Ti5Al2.5Fe −1 580 −1 450–550
Ti6Al7Nb −1 500–600 −1 450–600
cp-Nb −1 270 −1 150
cp-Ta −1 410 −1 200
Ti30Ta −1 – −1 400

Table 1c.22 Rotating bending fatigue tests of unnotched and notched titanium alloys (Ref. 31)
Stress Fatigue strength for
concentration alternating tensile stresses
Alloy R Condition factor (>107 cycles) (MPa)
Ti6Al4V −1 1.0 725
Ti5Al12.5Fe −1 Wrought + annealed 1.0 725
−1 Wrought + solution 3.6 300
−1 Treated + annealed
−1 Cast + HIP 3.6 300
−1 Cast + HIP 1.0 450

Table 1c.23 High cycle fatigue strength of hip endoprostheses of titanium alloys, measured in
0.9 % NaCl solution at 37 °C. Testing conditions similar to DIN 58840 (simulated loosened shaft,
50 mm) (Refs. 34, 36, 38)
Maximum load in 2 × 107 cycles (kN)
Alloy 0.9 % NaCl (f = 2 Hz)
Hot wrought Ti6Al4V 6.5–8.0
Wrought Ti5Al2.5Fe 8
Ti6Al7Nb 3.5–6.0

Table 1c.24 Influence of the mean stress Sm on the fatigue strength of Ti6Al4V (Ref. 39)
Fatigue strength (MPa)
Sm (MPa) R Notch factor Kf at 107 cycles
0 −1 1 400
2.82 250
250 −0.1 1 300
0.33 2.82 125
500 0.33 1 250
0.66 2.82 100
750 0.7 1 125
0.81 2.82 80
1c Metallic Biomaterials: Titanium and Titanium Alloys 179

Table 1c.25 Influence of the Notch

notch factor on the fatigue factor Kf Stress ratio R Fatigue strength (MPa)
strength of Ti6Al4V (Refs.
1 −1 400
40–43)
1.7 −1 300
3.7 −1 150
6.0 −1 100

Table 1c.26 Influence of interstitial elements on the rotating bending strength of Ti6Al4V (Ref. 44)
Chemical composition of the base alloy (wt%)
Al V Fe C N O H Ti
6.03 3.96 0.10 0.02 0.009 0.1 0.005 balance
Notch factor Fatigue strength
Composition Heat treatment Kf R (MPa)
Base alloy 1 h 815 °C/furnace 1 –1 610
600 °C air 3 –1 210
Base alloy 1 h 855 °C/furnace 1 –1 510
+ 0.07 % N 600 °C air 3 –1 180
Base alloy 1 h 870 °C/furnace 1 –1 550
+ 0.2 % O 600 °C air 3 –1 210
Base alloy 1 h 840 °C/furnace 1 –1 560
+ 9.2 % C 600 °C air 3 –1 230
Base alloy 1 h 930 °C/furnace 1 –1 580
+ 0.07 % N 600 °C air 3 –1 240
+ 0.2 % O
+ 0.2 % C

Table 1c.27 Influence of texture and test directions on the rotating bending fatigue strength of
Ti6Al4V (fine equiaxed microstructure in rolled plates) (Ref. 45)
Fatigue strength
(MPa) test direction
Type of texture and method of production R Text direction at 107 cycles
Basal (0002 tilted out of the rolling plane –1 Rolling direction 625
by about 30°) cross rolling in lower
(α + β)-field
Transverse (0002 aligned parallel to the –1 Rolling direction 630
rolling direction) unidirectional in the Transverse 590
higher (α + β)-field, 950 °C direction
Basal/transverse (both are present) –1 Rolling direction 720
Unidirectional roll at about 930 °C Transverse 690
direction
180 H. Breme et al.

Table 1c.28 Influence of heat treatment (annealing and precipitation hardening, respectively) on
the fatigue strength of Ti6Al4V (Ref. 43, 46)
Condition Yield strength (MPa) Tensile strength (MPa)
As annealed 900 955
As hardened 1100 1195
Fatigue strength (MPa)
Condition sm Kf R at 107 cycles
As annealed 0 1 −1 510
0 3.3 −1 300
As hardened 0 1 −1 600
0 3.3 −1 280
As annealed 200 1 −0.3 425
200 3.3 −0.01 200
As hardened 200 1 −0.5 600
200 3.3 0 200
As annealed 300 1 0.14 400
300 3.3 0.23 165
As hardened 300 1 −0.22 550
300 3.3 −0.2 190

Table 1c.29 Influence of the beta field heat treatment on the fatigue strength of Ti6Al4V (Ref. 47)
Fatigue strength (MPa)
Heat treatment R at 107 cycles
0.5 h 1010 °C/AC + 2 h 700 °C/AC 0 525
5 h 1000 °C/AC + 2 h 700 °C/AC 0 620
0.5 h 1010 °C/H2O + 2 h 700 °C/AC 0 750
5 h 1010 °C/H2O + 2 h 700 °C/AC2 0 650

Table 1c.30 Influence of the surface treatment on the rotating bending fatigue (fine lamellar
microstructure, produced by annealing in 15 min/1050 °C/H2O + 1 h/800 °C/H2O) (Ref. 48)
Fatigue strength (MPa)
Surface treatment R at 107 cycles
Electrically polished −1 680
Mechanically polished (7 μm) −1 750
Mechanically polished (80 μm) −1 605
Mechanically polished (80 μm) + 1 h 500 °C −1 550
Mechanically polished (180 μm) −1 600
Mechanically polished (80 μm) + 1 h 800 °C −1 450
1c Metallic Biomaterials: Titanium and Titanium Alloys 181

Table 1c.31 Influence of the surface treatment on the rotating bending fatigue of Ti6Al4V (fine
equiaxed microstructure produced by rolling at 800 °C/H2O + 1 h/800 °C (HP)) (Ref. 48)
Surface treatment R Fatigue strength (MPa)
Electrically polished −1 610
Shot peened −1 710
Shot peened + 1 h 500 °C −1 390
Shot peened + 1 h 500 °C −1 800
20 μm surface removed
Shot peened + 20 μm surface removed −1 820

Table 1c.32 Influence of surface working on the rotating bending of Ti6Al4V (Ref. 40, 49)
Surface working Notch factor (Kf) R Fatigue strength (MPa)
Mechanically polished 1 −1 620
Mechanically polished 1 −1 660
+ cold roll bent
Ground 1 −1 540
Mechanically polished 2.02 −1 330

Table 1c.33 Influence of plasma nitriding (PVD) on the rotating bending fatigue of Ti6Al4V
(Ref. 30)
Maximum bending stress
Treatment R at 107 cycles (MPa)
Untreated −1 600
Vacuum annealed 20 h/850 °C −1 370
Plasma nitrideda 20 h/850 °C/N2 –1 470
Plasma nitrideda 20 h/700 °C/NH3 –1 550
Solution treated 1 h/940 °C/vac. –1 530
+ Ar cooled
Solution treated 1 h/940 °C/vac. –1 500
+ Ar cooled + plasma nitrideda
at 20 h/770 °C/N2
a
Plasma nitriding at 20–40 kW
182 H. Breme et al.

1c.6 Corrosion and Wear

Table 1c.34 Electrochemical data for titanium and titanium alloys in 0.1 M NaCl under different
conditions (Refs. 24, 50–54)
Corrosion potential Passive current density breakdown potential
Alloy Ecorr (mV) Ip (μA/cm2) Eb (mV)
cp-Ti
pH 7 −628 2.5
pH 2 −459 5–10 >1500
Ti5Al2.5Fe
pH 7 −529 0.68
pH 2 −567 0.71 >1500
Ti6Al4V
pH 7 −510 0.92
pH 2 −699 0.69 >1500
Ti6Al7Nb
pH 7 −368 0.53 >1000
Ti30Ta
pH 7 −419 0.3 >1500

Table 1c.35 Polarization current (i) and polarization resistance (Rc) of titanium and titanium alloys
in pure saline at 37 °C (Ref. 50) and in 0.9 % NaCl with a stable redox system [Fe(CN)64VFe(CN)63]
(Ref. 55)
Rc (kΩ cm2)
2
Material i (μA/cm ) Pure saline Saline + redox
cp-Ti 0.010 1000 714
Ti6Al4V 0.008 1250 455

Table 1c.36 Repassivation time in 0.9 % NaCl and breakdown potential in Hanks’ solution of
cp-Ti and Ti alloys (Refs. 24, 32, 56, 57)
Breakdown potential (mV) Repassivation Time (ms)
Alloy vs. calomel electrode −500 mV +500 mV
cp-Ti 2400 43 44.4
cp-Ta 2250 – –
Ti30Ta >1500 41.7 47.5
Ti40Nb >1500 44.6 43.4
Ti6Al4V >2000 37 41
Ti5Al2.5Fe 110–130 120–160
1c Metallic Biomaterials: Titanium and Titanium Alloys 183

Table 1c.37 Electrochemical data for anodic titanium and Ti6Al4V at 37 °C in different solutions
(de-aired) versus standard calomel electrode (SCE) (Ref. 58)
Corrosion potential Passive current density Breakdown potential
Alloy vs. SCE Ep (mV) Ib (μA/c−2) E (mV) Solutiona
cp-Ti −440 to 490 1.0–3.0 1300 1
−94 to 140 0–1.0 1750 2
−94 5.0–9.0 1950 3
Ti6Al4V −200 to 250 0.9–1.0 1155–1240 1
−240 to 300 0.8–1.5 1900 2
−180 to 250 0.9–2.0 1550 3
a
1 = Ringer’s solution; 2 = Hanks’ solution; 3 = 0.17 M NaCl solution

Table 1c.38 Electrochemical data for cp-Ti and Ti alloys after 7 days in artificial saliva (Ref. 59)
Alloy Corrosion potential (mV) Current densities
Icorr (nA/cm2) Ipass (μA/cm2)
cp-Ti −260 21 9.5
Ti6Al4V −230 31 10
Ti6Al7Nb −220 32 9
Ti5Al2.5Fe −180 30 6

Table 1c.39 Repassivation NaCl pH = 7.4 HCl pH = 3 Repassivation time tc

time of titanium and titanium
(shortcut alloy) (activated alloy) (ms)
alloys in contact with
different metallic materials cp-Ti cp-Ti –
(Ref. 60) Ti30Ta 37.7
Ti6Al4V 41.8
cp-Nb 480.0
Co30Cr6.5Mo 38.7
Co28Cr5Mo 38.4
X3CrNiMol812 1000.0
Ti30Ta cp-Ti 43.0
Ti30Ta 48.6
Ti6Al4V 39.0
cp-Nb 1080.0
Co30Cr6.5Mo 44.1
Co28Cr5Mo (4200)
X3CrNiMol812 1000.0
Ti6Al4V cp-Ti 44.0
Ti30Ta 34.2
184 H. Breme et al.

Table 1c.40 Influence of the surface treatment on the fretting corrosion behavior of Ti6Al4V (Ref. 61)
Material combination Total weight loss (μg) Ti (μg) V (μg)
Untreated–untreated 2423 3925 78.5
Untreated–nitrogen ion implanted 1295 1260 31.2
Untreated–PVD coated with TiN 1002 902 15.0
Untreated–plasma ion nitrided 807 716 6.4
PVD-PVD 713 470 8.5
Plasma ion nitrided–plasma ion nitrided 273 87 0.5
Testing conditions: plate screw system (micromotion = 100), 14 days in calf serum solution (1 Hz
for 1,200,000 cycles)

Table 1c.41 Influence of the surface treatment on the wear behavior of Ti6Al7Nb as a result of a
pin-on-disk test (Ref. 62)
PVD coated with Oxygen diffusion hardened (ODH)
Property 3 μm TiN layer (30 μm hardened surface)
Wear factor (10−7 mm3/Nm) 2.111 1.353
Coefficient of friction 0.078 0.051
Surface roughness Rz (μm) 0.159 0.330
Wetting angle (°) 47 49

Table 1c.42 Volumetric wear rate of Ti6Al4V and Ti6Al7Nb

under different sliding speeds and normal load (Ref. 63 )
Sliding speed Load (N) Volumetric wear (mm3 N/mm)
Ti6Al4V Ti6Al7Nb
1 mm/s 3 4.45 × 10−3 5.48 × 10−3
−3
6 4.24 × 10 9.64 × 10−3
−3
10 11.08 × 10 13.12 × 10−3
−3
15 mm/s 3 20.35 × 10 22.06 × 10−3
−3
6 37.98 × 10 38.10 × 10−3
−3
10 51.55 × 10 57.35 × 10−3
−3
25 mm/s 3 27.32 × 10 31.38 × 10−3
−3
6 42.15 × 10 45.28 × 10−3
10 54.21 × 10−3 57.62 × 10−3
1c Metallic Biomaterials: Titanium and Titanium Alloys 185

Table 1c.43 Influence of ion implantation of nitrogen on the wear properties of commercial cp-Ti
and Ti6A14V (Ref. 64)
Material Friction couple Total wear (mg) Friction coefficient
cp-Tia Untreated–untreated 632.3 0.48
Nitrided–nitrided 54.3 0.10
4 h/940 °C/N2:H2 = 2:1
Ti6A14Vb Untreated–untreated 600.0 0.46
Nitrided–nitrided 40.1 0.10
4 h/940 ° C/N2: H2 = 2:1
Nitrided–nitrided 92.3 0.12
6 h/800 °C/N2:H2 = 1:1
a
Friction distance = 1257 m
b
Friction distance = 1885 m

Table 1c.44 Rate of formation of corrosion products for

cp-Ti in Hanks’ solution during current-time-tests (Ref. 65)
Metal converted into
Polishing method compound (ng/cm2 h)
Mechanically polished 4.1
Chemically polished 3.5

1c.7 Biological Properties

Table 1c.45 Biocompatibility of cp-Ti and Ti alloys, survival rate of L132 cells
incubated with powders (Ref. 66)
Alloy Powder concentration (mg/L) Survival rate of cells
cp-Ti >400
Ti6Al4V >400
Ti5Al2.5Fe >400 >80 %
Ti30Ta >400
Ti30Nb >400

Table 1c.46 Influence of the implantation time (in vivo) on the surface roughness and peak-to-valley
(P–V) height of Ti6A14V femoral heads (Ref. 67)
Implantation time
Before
Position implantation 85 months 110 months 124 months
Ra (nm) P–V Ra (nm) P–V Ra (nm) P–V Ra (nm) P–V
Anterior 43 ± 10 370 ± 72 58 ± 50 746 ± 509 250 ± 147 2044 ± 1178 86 ± 81 812 ± 763
Posterior 41 ± 6 591 ± 333 150 ± 125 2281 ± 1842 114 ± 96 1175 ± 778 142 ± 131 1045 ± 890
Medial 51 ± 14 411 ± 159 44 ± 29 649 ± 259 118 ± 69 1224 ± 731 412 ± 11 401 ± 125
Lateral 52 ± 9 364 ± 68 71 + 55 722 ± 474 117 ± 106 1195 ± 1009 40 ± 16 527 ± 156
186 H. Breme et al.

1c.8 Nitinol: Shape Memory

Table 1c.47 Properties of Nitinol (shape memory, Ni45Ti) alloy (Refs. 50, 68–71)
Below
Above (=austenitic) (= martensitic)
Transition temperature (−200 to 110 °C)
Density (g/cm3) 6.45
Melting point (°C) 1310
Young’s modulus (GPa) 83 28–41
Tensile yield strength (0.2 %) (MPa) 195–690 70–140
Ultimate tensile strength (MPa) 640–1380 103–862
Ratio yield/tensile strength 0.75–0.68 0.33–0.16
Elongation at fracture (%) 1–15 up to 60
Thermal expansion coefficient (°C−1) 1.1 × 10−7 6.6 × 10−6
Specific heat (Cal/g °C) 0.20
Electrical conductivity (S/m) 1–1.5
Corrosion potential (mV)
in 0.1 M NaCl pH 7 −431
pH 2 −518
Passive current density (A/cm2)
in 0.1 M NaCl pH 7 0.44
pH 2 0.61
Breakdown potential (mV)
in 0.1 M NaCl pH 7 890
pH 2 960

Table 1c.48 Resulting oxide layer thickness on Nitinol stents

using different preparation techniques (Ref. 72)
Surface treatment Oxide layer thickness (Å)
Electropolished 34 ± 14
Passivated in 10 % HNO3 30 ± 1
Air aged at 450 °C 240 ± 70
Heat treated in a NO2−/NO3− solution 911 ± 270
500 °C

Table 1c.49 Ion release from Nitinol incubated with L132 cell culture (Ref. 73)
Ni (ppm) Ti (ppm)
3 days 6 days 3 days 6 days
hp-Ni 6.599 ± 0.037 11.364 ± 0.034 n.m. n.m.
cp-Ti n.m. n.m. 0.001 0.002
NiTi 0.081 ± 0.006 0.176 ± 0.008 0.004 0.006
hp-Ni high-purity nickel, n.m. not measured
1c Metallic Biomaterials: Titanium and Titanium Alloys 187

Table 1c.50 Survival rate of L132 cells incubated with

Nitinol powder (Ref. 74)
Metal Powder concentration (μg/mL) Survival rate of cells
Nitinol 400 71.2 %

References

1. Cverna F, Horesh J, Whittle S, Yuko I (2001) Worldwide guide to equivalent nonferrous metals
and alloys, 4th edn. ASM International, Materials Park, OH
2. Donachie MJ (2000) Titanium: a technical guide, 2nd edn. ASM International, Materials Park, OH
3. Standard specification for unalloyed titanium, for surgical implant applications (UNS R50250,
UNS R50400, UNS R50550, UNS R50700) (2006) ASTM International, West Conshohocken, PA
4. Brunette DM, Tengvall P, Textor M, Thomsen P (2001) Titanium in medicine: material sci-
ence, surface science, engineering, biological responses and medical applications, 1st edn.
Springer, New York
5. Zhou YL, Niinomi M, Akahori T (2004) Effects of Ta content on Young’s modulus and tensile
properties of binary Ti-Ta alloys for biomedical applications. Mater Sci Eng A 371:283–290
6. Semlitsch M, Staub F, Weber H (1985) Development of a vital, high-strength titanium
aluminium-niobium alloy for surgical implants. In: Proc. fifth European CONF. ON
BIOMATERIALS, Paris, 4–6 Sept 1985
7. Mausli P-A, Steinemann SG, Simpson JP (1984) Properties of surface oxides on titanium and
some titanium alloys. In: Proc. of the sixth world conference on titanium, vol 3, pp. 1759–1764
8. Material properties of titanium and titanium alloys (DIN 17869) (1990) Beuth Veriag, Germany
9. Microstructure standard for titanium alloy bars, Technical Committee of European Titanium
Producers, Publication ETTC2 (1979)
10. Supra Alloys, Inc. (2013) Aerospace materials specifications for titanium and titanium alloys.
http://www.supraalloys.com/specs.php
11. MatWeb Material Property Data. Titanium, Ti. http://www.matweb.com/search/DataSheet.asp
x?MatGUID=66a15d609a3f4c829cb6ad08f0dafc01
12. Eliaz N (2012) Degradation of implants, 2012th edn. Springer, New York
13. Titanium alloys—Ti6Al7Nb properties and applications. AZO materials. http://www.azom.
com/properties.aspx?ArticleID=2064
14. MatWeb Material Property Data. Titanium IMI 367 (Ti-6Al-7Nb). http://www.matweb.com/
search/datasheet.aspx?matguid=71fff43e6722453c8c9783d017d66977&ckck=1
15. Breme J, Wadewitz V, Burger K (1988) Verbund Titanlegierungen/Al2O3-Keramik fur dentale
Implantate—Entwicklung geeigneter Legierungen. In: Ondracek G (ed) Verbundwerkstoffe—
Stoffverbunde. DGM, pp. 123–130
16. Breme J, Biehl V, Schulte W, d'Hoedt B, Donath K (1993) Development and functionality of
isoelastic dental implants of titanium alloys. Biomaterials 14(12):887–892
17. Akahori T, Niinomi M, Fukui H, Ogawa M, Toda H (2005) Improvement in fatigue character-
istics of newly developed beta type titanium alloy for biomedical applications by thermo-
mechanical treatments. Mater Sci Eng C 25:248–254
18. Kim SE, Jeong HW, Hyun YT, Lee YT, Jung CH, Kim SK, Song JS, Lee JH (2007) Elastic modu-
lus and in vitro biocompatibility of Ti-xNb and Ti-xTa alloys. Met Mater Int 13(2):145–149
19. Zwicker U (1974) Titan und Titanlegierungen. Springer, New York
20. Lee W, Lin C (1998) High-temperature deformation behaviour of Ti6Al4V alloy evaluated at
high strain-rate compression tests. J Mater Process Technol 75:127–136
21. Titanium and titanium wrought alloys forgings (hammer and drop forgings)—technical speci-
fication (DIN 17864) (1990) Beuth Veriag, Germany
22. Long M, Rack HJ (2008) Titanium alloys in total joint replacement—a materials science
perspective. Biomaterials 19:1621–1639
188 H. Breme et al.

23. Ramsdell JD, Hull ED (1960) Characteristics of cold-rolled and annealed Ti. Bureau of Mines
Rep. of Inv. 5656
24. Breme J, Wadewitz V (1989) Comparison of Ti-Ta, Ti-Nb alloys. J Oral Max Implants 4(2):
113–118
25. Niinomi M (1998) Mechanical properties of biomedical titanium alloys. Mater Sci Eng A243:
231–236
26. Shuh A, Bigoney J, Hönle W, Zeiler G, Holzwarth U, Forst R (2007) Second generation (low
modulus) titanium alloys in total hip arthroplasty. Materwiss Werkstofftech 38(12):1003–1007
27. Stemlitsch M, Staub F, Weber H (1985) Development of a vital, high-strength titanium
aluminum-niobium alloy for surgical implants. In: Proc. fifth European conf. on biomaterials,
Paris, 4–6 Sept 1985
28. Breme J, Zhou Y, Groh L (1995) Development of a titanium alloy suitable for an optimized
coating with hydroxyapatite. Biomaterials 16:239–244
29. Titanium and titanium alloy strip, sheet and plate—technical conditions of delivery (DIN
17860) (1990) Beuth Veriag, Germany
30. Lanagan J (1988) Properties of plasma nitrided titanium alloys. In: Proc. of the sixth world
conf. on titanium, vol 4, pp. 1957–1962
31. Borowy KH, Kramer KH (1984) On the properties of a new titanium alloy (TiA15Fe2.5) as
implant material. In: Proc. of the fifth world conf. on Ti, vol 2, pp. 1381–1386
32. Thull R (1979) Eigenschaften von Metallen, fur orthopadische Implantate und deren Priifung.
Orthopädie 7:29
33. Breme J, Schmidt H-J (1990) Criteria for the bioinertness of metals for osseo-integrated
implants. In: ed. Heimke G (ed) Osseo-integrated implants, CRC press, vol 1, pp. 31–80
34. Semlitsch M, Weber H (1992) Titanlegierungen fur zementlose Huftprothesen. In: Hipp E,
Gradinger R, Asherl R (eds) Die zementlose Hiiftprothese. Demeter Verlag, Grafelfingen, pp 18–26
35. Semlitsch M, Panic B (1994) 15 years of experience with test criteria for fracture-proof
anchorage stems of artificial hip joints. In: Buchhorn GH, Willert H-G (eds) Technical princi-
ples, design and safety of joint implants. Hogrefe & Huber Publishers, Seattle, pp 23–36
36. Papakyriacou M, Mayer H, Pypen C, Plenk H Jr, Stanzl-Tschegg S (2000) Effects of surface treat-
ments on high cycle corrosion fatigue of metallic implant materials. Int J Fatigue 22:873–886
37. Narayan R (2009) Biomedical materials, 2009th edn. Springer, New York
38. Kunze E (1988) Vergleichende Untersuchungen zum Langzeit-Ermii-dungsverhalten von
Hiiftgelenkprothesen an Luft und in NaCl-Losung. Metall 2:140–145
39. RMI-Titanium (1969) Reactive Metals Inc., Niles, OH
40. Weinberg IG, Hanna IE (1957) An evaluation of the fatigue properties of Ti and Ti-alloys.
TML Rep. 77
41. Mote WM, Frost RB (1958) The engineering properties of commercial Ti-alloys. TML Rep. 92
42. Weltzin RB, Koves G (1968) Impact fatigue testing of Ti alloys. J Mater 3:469–480
43. Properties and selection of metals. In: ASM metals handbook, vol 1. ASM International (1961)
44. Dittmar CB, Bauer GW, Evers D (1957) The effect of microstructural variables and interstitial
elements on the fatigue behavior of Ti and commercial Ti alloys, Mallory, Sharen Titanium
Corp. AD-110726WADC-TR 56–304, 95
45. Liitjering G, Gysler A (1984) Fatigue. In: Proc. of the fifth conf. on Ti, pp. 2065–2083
46. Hempel M, Hillnhagen E (1966) Dauerschwingfestigkeit der technischen Titanlegierungen
TiA15Sn2.5 und TiA16V4. Arch Eisenhiittenwesen 37:263–277
47. Riidinger K, Fischer D (1984) Relationship between primary alpha content, tensile properties
and high cycle fatigue behaviour of Ti6Al4V. In: Proc. of the fifth conf. on Ti, pp. 2123–2130
48. Wagner L, Gerdes C, Liitjering G (1984) Influence of surface treatment on fatigue strength of
Ti6Al4V. In: Proc. of the fifth conf. on Ti, vol 4, pp. 2147–2154
49. Liu HW (1956) Effect of surface finish on the fatigue strength of TiA16V4- alloy. Dept. of
Theoretical and Applied Mechanics, University of Illinois, T.U.A.M. Rep. 533
50. Geis-Gerstorfer J, Weber H, Breme J (1988) Electrochemische Untersuchung auf die
Korrosionsbestandigkeit von Implantatlegierungen. Z Zahnarztl Implantol 4(1):31–36
1c Metallic Biomaterials: Titanium and Titanium Alloys 189

51. Titanium alloys; chemical composition (DIN 17851) (1990) Beuth Veriag, Germany
52. Luiz de Assis S, Wolynec S, Costa I (2006) Corrosion characterization of titanium alloys by
electrochemical techniques. Electrochim Acta 51:1815–1819
53. Kesteven J, Kannan MB, Walter R, Khakbaz H, Choe H (2015) Low elastic modulus Ti–Ta
alloys for load-bearing permanent implants: enhancing the biodegradation resistance by elec-
trochemical surface engineering. Mater Sci Eng C 46:226–231
54. Cai Z, Nakajima H, Woldu M, Berglud A, Bergman M, Okabe T (1999) In vitro corrosion
resistance of titanium made using different fabrication methods. Biomaterials 20:183–190
55. Breme J (1988) Titanium and titanium alloys, biomaterials of preference. In: Proc. of the sixth
world conf. on Ti, vol 1, pp. 57–68
56. Fraker AC, Ruff AW et al (1983) Surface preparation and corrosion. Behaviour of titanium
alloys for surgical implants. In: Luckey HA, Kubli F (eds) Titanium alloys in surgical implants.
ASTM STP796, pp. 206–219
57. Breme J, Wadewitz V, Ecker D (1986) Untersuchungen zum Einfluf* des Geftigezustandes auf das
Korrosionverhalten der Implantatlegierung TiA15Fe2.5. Z./Zahnarztl. Implantologie, Bd. II 32–37
58. Gagg C (1988) Corrosion characteristics of Til5V3Cr3Sn3Al (Til5-3) alloy in a physiological
environment
59. Mareci D, Ungureanu G, Aelenei DM, Mirza Rosca JC (2007) Electrochemical characteristics
of titanium based biomaterials in artificial saliva. Mater Corros 58(11):848–856
60. Wadewitz V, Breme J (1989) Titan-Legierungen fur dentale Implantate. Z Zahnarztl Implantol
V:116–120
61. Maurer A, Brown SA, Merrit K (1992) Effects of surface treatments on fretting corrosion of
Ti6A14V. In: Proc. of the fourth world biomaterials congress, p. 200
62. Streicher RM, Weber H, Schon R, Semlitsch M (1991) Wet behaviour of different ceramic
surfaces in comparison to TiN and OHD treated Ti6A14V alloy paired with polyethylene. In:
Vincenzini P (ed) Ceramics in substitutive and reconstructive surgery. Elsevier, Amsterdam,
pp 197–205
63. Fellah M, Assala O, Labaïz M, Dekhil L, Iost A (2013) Friction and wear behavior of Ti-6Al-4V
and Ti-6Al-7Nb biomaterial alloy. J Biomater Nanobiotechnol 4:374–384
64. Hu Y-S, Dai D-H, Dong Y-L (1988) The ion nitriding of titanium materials and its applications.
In: Proc. of the sixth world confer, on titanium, vol 4, pp. 1801–1804
65. Mears DC (1977) Metals in medicine and surgery. Int Met Rev, Review 218, 119–155
66. Hildebrand HF (1993) Biologische Aspekte beim Einsatz von Implantatwerkstoffen. DGM
Hochschulseminar, Saarbriicken, Germany
67. Easter TL, Graham RM, Jacobs JJ, Black J, LaBerge M (1994) Clinical performance of
Ti6A14V femoral components: wear mechanisms vs. surface profile. In: Proc. of the 20th
annual meeting of the society for biomaterials, p. 185
68. Ebel B (1990) Zur Biokompatibilitat von Implantatwerkstoffen. KfK 4476
69. Tautzenberger P, Stockel D (1987) Vergleich der Eigenschaften von Thermobimetallen und
Memory-Elementen, 1, 26–32
70. Nitinol Technical Properties. Johnson Matthey Medical Components (2015) http://jmmedical.
com/resources/221/Nitinol-Technical-Properties.html
71. Thompson SA (2000) An overview of nickel-titanium alloys used in dentistry. Int Endod
J 33:297–310
72. Trépanier C, Tabrizian M, Yahia L, Bilodeau L, Piron DL (1997) Effect of modification of oxide
layer on NiTi stent corrosion resistance; Nitinol: Freemont, CA. http://www.nitinol.com/media/
reference-library/033.pdf
73. Rocher P, El Medawar L, Hornez JC, Traisnel M, Breme J, Hildebrand HF (2004) Biocorrosion
and cytocompatibility assessment of NiTi shape memory alloys. Scr Mater 50:255–260
74. El Medawar L, Rocher P, Hornez JC, Traisnel M, Breme J, Hildenbrand HF (2002)
Electrochemical and cytocompatibility assessment of NiTiNOL memory shape alloy for orth-
odontic use. Biomol Eng 19(2–6):153–160
Chapter 1d
Dental Restoration Materials

Jonathan Black and Garth Hastings

1D.1  Amalgams

1d.1.1  Composition of alloys

Table 1d.1  Chemical composition of dental amalgams (wt%) (Ref. 1, 2)
Alloy powder* Ag Sn Cu Zn Hg
LCS 66–73 25–29 <6 <2 0–3
HCB 69 17 13 1 –-
HCSS, HCSL 40–60 25–30 15–30 –- –-
The mercury concentration after amalgamization is <50 %.
* LCS = low-copper spherical.
HCB = high-copper blended.
HCSS = high-copper single-composition spherical.
HCSL = high-copper single-composition lathe-cut.

1d.1.2  Physical properties

Table 1d.2  Physical and Thermal expansion coefficient 25 × 10-6/K

mechanical properties of
Thermal conductivity 23 W/mK
amalgams (Ref. 2)
Ultimate tensile strength ≥60 MPa
Ultimate compression strength ≥300 MPa
Fracture toughness KIC <1 MPa m1/2

J. Black (*)
Principal: IMN Biomaterials, King of Prussia, PA, USA
G. Hastings
The Biomaterials Programme Institute of Materials Research and Engineering,
National University of Singapore, Singapore, Singapore

© Springer Science+Business Media New York 2016 191

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_17
192 J. Black and G. Hastings

1d.1.3  General properties and processing

General properties of dental amalgams (Ref. 1, 2)

By definition amalgam is an alloy of mercury with one or more other metals.
Dental amalgams are produced by mixing an alloy powder with mercury. Usually
the alloy powder is called ‘amalgam’ and the composition of amalgam denotes the
composition of the powder. The two basic physical shapes of the alloy powder are:
1. irregular shaped particles (length 20–120 μm, width 10–70 μm, thickness 10–35
μm) produced by lathe cutting.
2. spherical particles (diameter < 30 μm) produced by atomization.
Amalgams for dental restorations are classified by their copper content in two
basic types:
1 . low-copper type, < 6 wt% Cu, used since the late 19th century.
2. high-copper type (non-gamma-2 amalgam), >6 wt% Cu, used since 1960.
The high-copper type is available in two alloy groups, classified by the mixture
of different powder shapes:
2.1 mixed alloys consisting of
2.1.1 2/3 conventional lathe cutting + 1/3 spherical powder of Ag–Cu eutecticum
(total composition see Table 1d.1 composition of eutecticum: Ag30Cu)
2.1.2 1/3 conventional lathe cutting powder + 2/3 spherical powder (Ag25Sn15Cu)
2.2 one-component amalgams consisting of
2.2.1 ternary alloys, spherical shape
2.2.2 ternary alloys, irregular shape
2.2.3 quaternary alloys, non-spherical shape
During amalgamization the powder reacts with the mercury, producing different
phases, depending on the amalgam powder:
Ag3 Sn + Hg → Ag2 Hg3 + Sn7 −8 Hg + Ag3 Sn
1. low copper type:
γ phase γ 1 phase γ 2 phase γ 3 phase
2. high copper type, mixed alloy (a(i)), reacts in two steps:

step one : Ag3 Sn + Hg → Ag2 Hg3 + Sn7 −8 Hg + Ag3 Sn

γ phase γ 1 phase γ 2 phase unreacted

step two : Sn 7 −8 Hg − Ag − Cu → Cu6 Sn5 + Ag2 Hg3
γ 2 phase eutetic beads

The result of the amalgamization is a microstructure consisting of unreacted Ag3Sn
and Ag–Cu surrounded by a layer of Cu6Sn5 and the γ1 matrix. The microstructure
of the one-component non-gamma-2 amalgams is similar to that of the mixed alloy
except that the Cu6Sn5 particles are decomposed in the γ1phase and form no layer.
1d  Dental Restoration Materials 193

Processing of dental amalgams

For dental restoration (fillings) the mercury is mixed with the (amalgam) alloy
immediately before application. The alloy is delivered in powder or pellet form. The
amount of mercury after amalgamization must be below 50 wt%. Trituration is
accomplished manually in a mortar with a pestle or more often automatically using
a capsule of mercury with a given weight and an alloy pellet. The amalgam is placed
in the cavity of the tooth in small portions which are pressed with a force of
40–50 N. On the surface of the filling a mercury-rich phase should appear, allowing
a good bonding to the following portion. When the cavity is completely filled, the
mercury-rich phase must be removed, and the filling can be modelled to the desired
shape. 24 hours after the amalgamization the filling must be polished in order to
achieve a smooth surface, resulting in a low corrosion rate.

1d.1.4  Mechanical properties

Table 1d.3  Mechanical properties of dental amalgams (Ref. 2)
Ultimate compression Creep (%) (7) days after
strength (MPa) amalgamization, pressure 38 MPa)
After
Amalgam* After 1 h setting
LCL 120–170 380–450 2.5–3.5
LCS 140–180 380–450 0.3–1.5
HCB 120–330 410–460 0.2–1.7
HCSS, HCSL ternary alloy 230–320 460–540 0.002–0.3
HCSS, HCSL quaternary 210–410 430–480 0.06–0.1
alloy with indium
* LCL = low-copper lathe-cut. LCS = low-copper spherical. HCB = high-copper blended. HCSS
= high-copper single-composition spherical. HCSL = high-copper single-composition lathe-cut.

1d.1.5 Corrosion and wear

Table 1d.4  Average passive current density, range of passivity, corrosion and breakdown potential
of a non-gamma-2 amalgam in artificial saliva (Ref. 3)
Passive Current Corrosion Breakdown
Density (μA Range of Passivity Potential Potential
Alloy cm-2) (mV vs. SCE) (mV vs. SCE) (mV vs. SCE)
Non-gamma-2 1.5 -300 – +300 -400 300
amalgam

Table 1d.5 Repassivation rates in artificial saliva versus standard calomel electrode (Ref. 3)
Material Corrosion Potential Potential (mV) after
(mV) 1h 2h 36 h
Non-gamma-2 amalgam -400 -408 -382 -290
194 J. Black and G. Hastings

Corrosion of amalgams

1. Corrosion of conventional amalgams

The heterogeneous structure of conventional amalgams is responsible for high cor-
rosion. The γ2 phase is the most active phase in electrochemical corrosion because
it is less noble than γ and γ1 Therefore an anodic dissolution occurs. The corrosion
products of the γ2 phase are
1.1 Sn2+ ions, in the presence of saliva SnO2 and Sn(OH)6CI
2.2 the released Hg reacts with particles of the γ phase.

2. Corrosion of the high copper type amalgams

Because of the absence of the γ2 phase, the Cu6Sn5 phase is the electrochemically
based phase. However, the total corrosion current in the non-gamma-2 amalgams is
much lower than in the conventional types. Therefore, the amount of corrosion
products is much lower than in gamma-2 alloys. An additional advantage of these
amalgams is the lack of mercury during corrosion.
Generally, to achieve a low corrosion in amalgams certain requirements must be
fulfilled:
• a polished surface
• no contact with gold (this would lead to a high corrosion of the amalgams, setting
mercury free and resulting in a reaction between gold and mercury).

Table 1d.6  Rate of release Rate of Hg Release

of mercury vapour per unit Alloy* (ng/min mm2)
area of different types of
LCL New 0.011
almagam dipped in an
isotonic NaCl solution (pH = Old 0.011
6) (Ref. 4) HCB New 0.133
Old 0.017
HCSS, New 0.022
HCSL Old 0.022
In-containing New 0.222
one component Old 0.350
LCL = low-copper lathe-cut. HCB = high-copper blended.
HCSS = high-copper single-composition spherical. HCSL =
high-copper single-composition lathe-cut. New = tested
within 2 months after amalgamization. Old = tester after 1.5
years after amalgamization.
Table 1d.7  Chemical composition of high gold-containing dental cast alloys (wt%) (Ref. 5, 6, 8)
Au + Pt
Type* Metals Au Pt Pd Ir Rh Ag Cu Zn Ta In Re Fe Sn
1d  Dental Restoration Materials

HGC-1 88.6 87.5 1.0 0.1 11.5

HGC-2 80.5–81.2 75.7–79.3 0.3–1.4 1.6–3.3 12.3–15.0 4.1–5.5 0.4–1.0 0–0.1
HGC-3 78.0–78.5 74.0–74.4 0–2.4 2.0–3.5 0.1 9.6–13.5 7.0–11.5 0.9–1.0
HGC-4 75.5–80.0 65.5–71.0 4.4–12.9 0.0–2.0 0.1 0–1.1 10.0–14.0 8.2–10.0 0.5–4.0
HGC-l-C 95.0–97.0 80.0–85.0 5.0–11.0 3.3–4.4 0.2 1.6 3.0–5.0
HGC-2-C 95.0 70.0 7.5 15.0 0.5 2.0 5.0
HGC-3-C 98.0–99.0 82.6–86.0 9.7–10.4 0–2.2 0.1–0.3 0–1.6 1.0–2.0
HGC-4-C 82.9–97.4 73.8–84.4 8.0–9.0 5.0–8.9 0.1 1.2–9.2 0.3–4.4 0–2.0 0–0.2 1.5–2.5 0–0.2 0–0.2 0.5 0.8
HGC = high gold-containing (Au + Pt metals > 75 wt%, Pt metals = Pt, Pd, Ir, Rh, Re, Os).
1 = soft, 2 = medium hard, 3 = hard, 4 = extra hard.
c = ceramic alloy (bonding with ceramic is possible).
195
Table 1d.8  Chemical composition of low gold-containing dental cast alloys (wt%) (Type 4: extra hard) (Ref. 5, 7, 8)
Au + Pt
Type* metals Au Pt Pd Ir Ag Cu Sn Zn In Ga Re Ta Fe Co
LGC-4 48.0–66.7 40.0–62.2 0–4.4 09.9 0.1 23.3–35.0 7.0–12.0 0–1.5 0.4–3.5 0–5.0

LGC-4-C 74.8–89.8 43.0–55.1 29.0–38.5 0.1–0.2 0–19.5 0–0.3 0–0.5 0–9.0 0–1.5 0–.2 0–0.1 0–0.2 0–2.8
* LGC-4 = low gold-containing, extra hard (60 wt% ≤ Au + Pt metals ≤ 75 wt%).
LGC-4-C = low gold-containing, extra hard, fusible (C = ceramic alloy (a bonding with ceramic is possible).
Table 1d.9  Chemical composition of AgPd and Pd-alloys (wt%) (Ref. 5, 8)
1d  Dental Restoration Materials

Au + Pt
Type metals Au Pt Pd Ir Ag Cu Sn Zn In Ru Ga Ge
AgPd-1 29.5 2.0 27.5 70.0 0.2 0.3
AgPd-4 29.5–40.0 ≤2.0 27.4–39.9 0.1 52.0–58.5 0–10.5 ≤2.0 1.5–4.0 ≤2.0
Pd-4-C 52–88 0–17 ≤1.0 25–70 7.2–38.0 0–11.6 1.9–7.5 ≤2.0 0–4.0 ≤0.8 0–7.2 ≤0.5
Key: see Table 1d.7
197
198 J. Black and G. Hastings

1D.2  Noble Metals

Table 1d.10  Physical Properties of precious dental alloys (Ref. 5, 8)

Melting Mean Coefficient Young’s
Density Temperature of Expansion Modulus
Alloy (g/cm3) (interval) (°C) (106/K) 25–600° C (GPa)
HGC-1 17.2 1030–1080
HGC-2 16.1–16.4 900–1040 92–95
HGC-3 15.6–15.8 900–975
HGC-4 15.6–16.8 900–1000 98–109
HGC-l-C 18.3–18.6 1090–1370 14.1–14.8
HGC-2-C 17.3 1285–1370 13.6
HGC-3-C 18.4–19.5 1045–1220 14.2–14.7 100–105
HGC-4-C 16.7–18.1 900–1260 14.0–16.8 102–113
LGC-4 12.1–14.1 770–1065 94–106
LGC-4-C 14.0–14.8 1150–1315 13.8–14.8
AgPd-1 11.1
AgPd-4 10.6–11.1 950–1150
Pd-4-C 11.2–12.2 1100–1290 14.0–15.4 122–126
Key: see Table 1d.7.

1d.2.1  Composition of alloys

1d.2.2  Physical properties

1d.2.3  Processing of Precious Metal Alloys (Ref. 2, 5, 8)

Casting

Precious metal alloys are normally cast by means of the lost wax process. The well-­
known method of wax modelling is applied. For the commonly used casting proce-
dure, centrifugal and vacuum pressure casting, the alloys can be heated by the
following methods:
• resistence
• propane/oxygen torch (reducing flame zone)
• HF induction
• electrical arc
The alloys are melted in graphite or ceramic crucibles. After removal of the cru-
cible the alloys can if required be hardened. After casting or brazing the alloys are
descaled. Mechanical cleaning can be carried out with rotating tools, ceramic grind-
ing wheels or rubber polishers.
1d  Dental Restoration Materials 199

Heat Treatment

Depending on the type of alloy and its application, the dental alloys are heat treated.
After the casting the alloys are quenched. A homogenization at 700°C followed by
rapid cooling should be carried out in order to decompose grain segregations, espe-
cially in alloys containing platinum. After a cold deformation the alloys should be
stress-relieved. Precipitation hardening is performed by:
• slow cooling from 700°C to room temperature
• cooling from 450 to 250°C in 30 minutes, followed by quenching
• heating between 350 and 450°C for 15 minutes, followed by quenching
Table 1d.12 gives the recommended heat treatments for various noble metal
alloys used in dental restoration.

Brazing

Brazing can be carried out with a torch or in a furnace. For larger surfaces furnace
brazing is preferable. The best strength properties are obtained with a solder gap of
0.05 to 0.2 mm between the surfaces. Table 1d.11 shows the chemical composition
and the brazing temperatures of various filler metals.

Bonding with Ceramics

Precious metal alloys are cast by the lost wax process. After removal of the crucible with
carbide tools the castings are sand-blasted with alumina (100–150 μm, pressure 2 bar) to
roughen the surfaces and to provide by increasing the surface an improvement in the adhe-
sion strength. After cleaning with water and hot steam an additional cleaning of the alloys
is performed by annealing for 10 minutes at 980°C. The bonding procedure takes place in
the temperature range of 800–900°C. After bonding the surface must be carefully cleaned
in order to provide good corrosion resistance. The final step of the process is a polishing
operation with rotating cotton or wool buffers and a small amount of polishing paste.

Table 1d.11  Chemical composition (wt%) and brazing temperatures of various filler metals (Ref.
5, 8)
Filler Brazing
Metals for Temp.
Brazing (°C) Au Pt Pd Ir Ag Cu Zn In Re
HGC, 700–840 50–73 ≤19 ≤1.0 ≤0.1 8.0–28 0–9.0 6.0–14 ≤2.0 ≤0.1
LGC
AgPd 760–820 73 0.9 1.0 0.1 13.0 12.0
HGC-C, 700–1120 50–73 ≤1.9 ≤1.0 ≤0.1 10.0 0–5.0 12.0– ≤2.0 ≤0.1
LGC-C –28.0 14.0
Pd-4-C 1030 50–73 ≤1.9 ≤1.0 10.0 3.0– 12.0– ≤2.0
–1120 –28.0 5.0 14.0
Key: see Table 1d.7.
200 J. Black and G. Hastings

Table 1d.12  Recommended heat treatments for various noble metal alloys (Ref. 5, 8)
Oxidizing without
Precipitation Hardening Soft Annealing Vacuum
Time Temperature Time Temperature Time Temperature
Alloy (min) (°C) (min) (°C) (min) (°C)
HGC 15 400 15 700–800
HGC-C 15 500–600 15 950 10 960–980
LGC 15 400–500 15 700–800
LGC-C 15 600 15 950 10 980
AgPd 15 550
Pd-4-C 15 600 15 950 10 980
Key: see Table 1d.7.

1d.2.4  Mechanical properties

Table 1d.13  Mechanical properties of high gold-containing dental cast alloys (Ref. 5, 8)
Ultimate
Tensile Yield Tensile Ratio Yield/ Elongation
Strength (0.2 Strength Tensile at Fracture Hardness
Type %) (MPa) (MPa) Strength (%) HVHN
HGC-­1 80 170 0.47 45 55
HGC-2 180–240 370–390 0.49–0.61 35–45 95–110
HGC-3 s 330–350 460 0.72–0.76 35–40 145
h 350–390 550–590 0.64–0.66 20–23 170–190
HGC-4 s 300–420 500–580 0.60–0.72 15–37 155–195
h 540–780 710–870 0.76–0.90 5–18 225–295
HGC-l-C s 90–130 220–280 0.41–0.46 29–38 60–75
h 105–140 230–300 0.46–0.47 27–38 70–90
HGC-2-C s 230 400 0.58 20 105
h 240 410 0.59 18 125
HGC-3-C s 370–420 460–515 0.80–0.82 8–15 150–160
h 470–490 530–590 0.83–0.89 6–9 185–200
HGC-4-C s 380–480 530–580 0.72–0.83 7–14 150–200
h 470–600 550–650 0.85–0.92 3–6 220–230
Key: see Table 1d.7.
s = soft, h = hardened.
1d  Dental Restoration Materials 201

Table 1d.14  Mechanical properties of low gold-containing dental cast alloys (Ref.: 5, 8)
Ultimate
Tensile Yield Tensile Ratio Yield/
Strength Strength Tensile Elongation at Hardness
Type (0.2%) (MPa) (MPa) Strength Fracture (%) VHN5
LGC-4 s 310–400 480–510 0.65–0.78 18–43 155–170
h 555–830 640–890 0.87–0.93 3–13 220–275
LGC-4C s 310–590 570–790 0.54–0.75 11–26 180–250
h 550–700 710–900 0.77–0.78 6–18 235–285
Key: see Table 1d.7.
s = soft, h = hardened

Table 1d.15  Mechanical properties of AgPd and Pd-alloys (Ref.: 5, 8)

Tensile Yield Ultimate
Strength (0.2 Tensile Ratio Yield/ Elongation at Hardness
Type %) (MPa) Strength (MPa) Tensile Strength Fracture (%) VHN5
AgPd-1 80 230 0.35 33 55
AgPd-4 285–595 510–950 0.56–0.63 3–31 140–310
Pd-4-C 340–630 630–900 0.54–0.70 8–30 180–285
Key: see Table 1d.7.

1d.2.5. Corrosion and wear

Table 1d.16  Polarization current (i) and polarization resistance (Rc) of gold
in 0.9% NaCl and in 0.9% NaCl with a stable redox system [Fe(CN)64- /
Fe(CN63-] at 37 °C corresponding to the potential of the body fluid of 400
mV (Ref. 7, 12)
Material i (μA/cm2 Rc (kΩcm2)
0.9% NaCl saline + redox
Au 0.009 1100 0.28

Table 1d.17  Average passive current density, range of passivity, corrosion and breakdown
potential of a high gold containing alloy (HGC) in artifîcial saliva (Ref. 3)
Range of
Passive Current Passivity Corrosion Potential Breakdown Potential
Alloy Density (μA cm-2) (mV vs. SCE) (mVvs. SCE) (mVvs. SCE)
HGC 1.5 -100 – +400 -137 400

Table 1d.18  Repassivation rates in artifical saliva versus standard calomel electrode (Ref. 3)
Material Corrosion Potential Potential (mV) after
(mV) 1h 2h 36h
HGC -137 -48 -26 -26
202 J. Black and G. Hastings

1D.3  CoCr-Alloys

See Chapter 1b, CoCr-alloys.

1D.4  NiCr-Alloys (Ref. 9, 10, 11)

NiCr-alloys are difficult to classify because of the wide range of the chemical com-
position, as shown in Table 1d.19. The NiCr-alloys in dentistry are generally used
for porcelain veneered and unveneered crowns, fixed and removable partial dentures
and bridgework. As the processing is similar to that of the CoCr-alloys, it is not
described here (see Chapter 1b.3). The requirements for these specific applications
determine the chemical composition. The corrosion resistance of the NiCr-alloys is
provided by the chromium content which produces a passive oxide layer on the
surface. Beryllium is added as a solid solution strengthener and supports the self-­
fluxing at the porcelain veneering temperature. It is also responsible for the good
chemical bonding to the porcelain. As it lowers the melting temperature, beryllium
also improves the stability.
Aluminium also produces a passive oxide layer, aids in the bonding to the porce-
lain and strengthens the alloy due to the precipitation of AlNi3. Silicon lowers the
melting temperature and, like manganese, acts as a deoxidizer. Molybdenum and
niobium are added to improve corrosion resistance and, like iron, are used to adapt
the thermal expansion coefficient to the coefficient of the porcelain.
The wide composition range results in an equally wide range of physical (Table
1d.20) and mechanical properties (Table 1d.21). The high rigidity and strength of
these alloys as compared to that of the precious metal alloys make them suitable for
the production of small prosthetic devices.

Table 1d.19  Chemical composition (wt%) of the NiCr-alloys used in dentistry Ref. 11)
Ni Co Fe Cr Mo Nb Ti W Be Ga Si C Others
58 0 0 12 0.5 0 0 0 0 0 0 ≤0.5 Al, Ce, B,
–- –- –- –- –- –- –- –- –- –- –- Mn, Sn, Y, V,
82 2 9 26 16 7 3 4 1.5 7.5 3 Ta, La, Cu

Table 1d.20  Physical and mechanical properties of NiCr-alloys

used in dentistry (Ref. 11)
Melting interval 940–1430 °C
Young’s modulus 170–220 GPa
Density 7.8–8.6 g/cm3
Mean coefficient of linear thermal 13.9–15.5 10-6/K
expansion between 25–600 °C
1d  Dental Restoration Materials 203

Table 1d.21  Mechanical properties of NiCr-alloys used in dentistry (Ref. 12)

Tensile Yield Strength (0.2%) (MPa) Elongation at Fracture (%) Hardness (VHN5)
255–800 3–25 160–395

References

1. Cahn, R.W., Haasen, P. and Kramer, E.J. (eds) (1992) Materials Science and Technology, Vol.
14, VCH.
2. Combe, E. (1984) Zahnärztliche Werkstoffe, Hanser.
3. Elagli, K., Traisnel, M. and Hildebrand, H.F. (1993) Electrochemical Behaviour of Titanium
and Dental Alloys in Artificial Saliva. Electrochimica Acta, 38(14), 1769–1774. 881.
4. Berglund, A. (1993) An in vitro and in vivo study of the release of mercury vapor for different
types of amalgam alloys. J. Dent. Res. 72 (5), 939–945.
5. Product information, Degussa, Germany.
6. DIN 13906–1 (=EN 21562, ISO 1562) (1990). Beuth.
7. DIN 13906–2 (invalid, replaced by DIN 28891) (1990). Beuth.
8. Product information, Wieland Edelmetalle, Germany.
9. Fraker, A.C., Corrosion of Metallic Implants and Prosthetic Devices, in Metals Handbook, 9th
Ed., Vol. 13: Corrosion, p. 1351.
10. Encyclopedia of Materials Science and Engineering, Vol. 2, (1986) Pergamon Press,
pp. 1057–1058.
11. NiCo-alloy producing industries.
12. Zitter, H. and Plenk, H. (1987) The Electrochemical Behaviour of Metallic Implant Materials
as an Indicator of their Biocompatibility. J. of Biomedical Materials Research, 21, 881.
Chapter 2
Composite Materials

L. Ambrosio, G. Carotenuto, and L. Nicolais

2.1  Types of Composites and Component Materials

Composites are combined materials created by the synthetic assembly of two or

more components – a selected reinforcing agent and a compatible matrix binder – in
order to obtain specifie and advanced characteristics and properties. The compo-
nents of composite do not dissolve or otherwise merge completely into each other,
but nevertheless do act in concert. The components as well as the interface between
them can usually be physically identified, and it is the behavior and properties of the
interface that generally control the properties of the composite. The properties of a
composite cannot be achieved by any of the components acting alone.
The composites can be classified on the basis of the form of their structural com-
ponents: fibrous (composed of fibers in a matrix), laminar (composed of layers of
materials), and particulate (composed of particles in a matrix). The particulate class
can be further subdivided into flake (flat flakes in a matrix) or skeletal (composed of
a continuous skeletal matrix filled by a second material). In general, the reinforcing
agent can be either fibrous, powdered, spherical, crystalline, or whiskered and either
an organic, inorganic, metallic, or ceramic material.

2.2  Fibre Types and Properties

A summary of the most important reinforcing filaments and their properties is

­presented in Tables 2.1–2.7.

L. Ambrosio (*) • G. Carotenuto • L. Nicolais

Department of Materials and Production Engineering, University of Naples Federico II
Institute of Composite Materials Technology CNR, Piazzale Technio, 80, 80125 Naples, Italy

© Springer Science+Business Media New York 2016 205

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_18
206 L. Ambrosio et al.

Table 2.1  Compositions and properties of various glasses (4)

Compound A C E R S
SiO2 72.0 64.6 52.4 60.0 64.4
Al2O3, Fe2O3 1.5 4.1 14.4 25.0 25.0
CaO 10.0 13.4 17.2 9.0 –
MgO 2.5 3.3 4.6 6.0 10.3
Na2O, K2O 14.2 9.6 0.8 – 0.3
B2O3 – 4.7 10.6 – –
BaO – 0.9 – – –

Table 2.2  Properties of fiberglass (2)

Grade of class
Property A B E S
Physical properties
Specific gravity 2.50 2.49 2.54 2.48
Mohs gravity – 6.5 6.5 6.5
Mechanical properties
Tensile strength, psi×106 (MPa)
At 72°F (22 °C) 440 440 500 665
(3033) (3033) (3448) (4585)
At 700°F (371 °C) – – 380 545
(2620) (3758)
At 1000°F (538 °C) – – 250 350
(1724) (2413)
Tensile modulus elasticity at 72°F (22 °C), – 10.0 10.5 12.4
psi×106 (GPa)
(69.0) (72.4) (85.5)
Yield elongation, % – 4.8 4.8 5.7
Elastic recovery, % – 100 100 100
Thermal properties
Coefficient of thermal linear expansion, 4.8 4.0 2.8 3.1
in./in./°Fx10-6 (m/m/°C)
(8.6) (7.2) (5.0) (5.6)
Coefficient of thermal conductivity, – – 72 –
Btu- in./hr/sq ft/°F (watt/motor K)
Specific heat at 72°F (22 °C) – 0.212 0.197 0.176
Softening point, °F(°C) 1340 1380 1545
Electrical properties
Dielectric strength, V/mil – – 498 –
Dielectric constant at 72 °F (22 °C)
At 60 Hz – – 5.9–6.4 5.0–6.4
At 106 Hz 6.9 7.0 6.3 5.1
Dissipation (power) factor at 72 °F (22 °C)
At 60 Hz – – 0.005 0.003
(continued)
2  Composite Materials 207

Table 2.2 (continued)
Grade of class
Property A B E S
At 106 Hz – – 0.002 0.003
Volume resistivity at 72 °F (22 °C) and 500 V – – 1015 1016
DC, ½-cm
Optical properties
Index of refraction – – 1013 104
Acoustic Properties
Velocity of sound, ft/sec (m/sec) – – 17,500 19,200
(5330) (5850)

Table 2.3  Thermal Properties of Kevlar 49 (8)

Property Value
Long-term use at elevated temperature in air, °C (°F) 160(320)
Decomposition temperature, °C (°F) 500 (932)
Tensile strength, MPa (Ksi)
At room temperature for 16 months No strength loss
At 50 °C (122°F) in air for 2 months No strength loss
At 100 °C (212°F) in air 3170
At 200 °C (392°F) in air 2720
Tensile modulus, GPa (103 Ksi)
At room temperature for 16 months No modulus loss
At 50 °C (122°F) in air for 2 months No modulus loss
At 100 °C (212°F) in air 113.8 (16.5)
At 200 °C (3920 F) in air 110.3 (16.0)
Shrinkage, %/°C (%°F) 4 x 10-4 (22 x 10-4)
Thermal coefficient of expansion, 10-6 cm/cm °C
Longitudinal, 0–100 °C (32–212°F) -2
Radial, 0–100 °C (32–212°F) +59
Specific heat at room temperature, J/g°C (Btu/lb°F) 1.42 (0.34)
Thermal conductivity at room temperature, J cm/sec m2°C (Btuin/hr ft2
°F)
Heat flow perpendicular to fibers 4.110 (0.285)
Heat flow paraliel to fibers 4.816 (0.334)
Heat of combustion, kJ/g (Btu/lb) 34.8 (15,000)

Table 2.4  Properties of Specific gravity 1.44

Kevlar 29 (2)
Tensile strength (GPa) 2.76
Tensile modulus (GPa) 58
Elongation (%) 4.0
Filament diameter (μm) 12.1
208 L. Ambrosio et al.

Table 2.5  Typical Properties of Boron fibers and of other Commercially Available Reinforcement
Filaments (1)
Diameter Manufacturing Average strength Density Modulus
Filament material (μm) technique N/m2 (Ksi) (g/cm3) (GN/m2)
Boron 100–150 CVD 34 (500) 2.6 400
SiC-coated boron 100–150 CVD 31 (450) 2.7 400
SiC 100 CVD 27 (400) 3.5 400
B4C 70–100 CVD 24 (350) 2.7 400
Boron on carbon 100 CVD 24 (350) 2.2 —
Al2O3 250 Melt withdrawal 24 (350) 4.0 250
Beryllium 100–250 Wire drawing 13 (200) 1.8 250
Tungsten 150–250 Wire drawing 27 (400) 19.2 400
‘Rocket Wire’ 50–100 Wire drawing 41(600) 7.9 180
AFC-77

Table 2.6  Carbon fibre precursors (2)

Precursor material Carbon content (%wt) Carbonization yield (%wt)
Cellulose fibre 45 10–15
Polyacrylonitrile 68 40
Lignin 70 50
Hydrocarbon pitch 95 85–90

Table 2.7  Properties of some carbon fibers (2)

Tensile Tensile Elonga- Carbon
Fibre Fibre Trade Diameter Specific strength modulus tion content
precursor type name (μm) gravity (GPa) (GPa) (%) (%wt)
Hydrocaron Carbon Kureha 10.5 1.6 1.03 – 2.5 99.5
pitch
Lignin Carbon 10–15 1.5 0.6 – 1.5 90
Cellulose Graphite Thornel 6.6 1.67 2.0 390 0.6 99.9
Polyacrylonit Carbon Graphil 8 1.76 3.2 230 – –
rile HT
Polyacrylonit Graphite Graphil 8 1.87 2.4 330 — —
rile HM

2.2.1  Glass Fibers

Glass fibers are the most common of all reinforcing fibers for polymeric matrix
composites. Their main advantages are low cost, high tensile strength, high chemi-
cal resistance and good insulating properties. On the other hand they display a low
tensile modulus, a relatively high density comparated to the other fibers, a high
2  Composite Materials 209

sensitivity to wearing and a low fatigue resistance. Depending on the chemical

­composition of the glass they are commercially available in different grade: A, C, E,
R, S. At one time ‘A’ or alkali glass was quite common as the basic material for
glass fibre production. Today this has been virtually completely superseded by ‘E’
or electrical grade glass. E-glass is a very low alkali content borosilicate glass which
provides good electrical and mechanical properties, coupled with good chemical
resistance. Another glass produced in commercial quantities for fibers production is
the C-glass, a special chemical resistant glass. This is used in the manufacture of
surfacing tissues to provide additional chemical resistance over E-glass. For specifie
application ‘R’ and ‘S’ glasses are available as fibers. These are high strength
glasses used mostly for aerospace applications.

2.2.2  Aramid Fibers

The most common aramid fiber available is the Kevlar 49. These fibers are com-
posed of a highly oriented crystalline polymer and present the highest tensile
strength/weight ratio. On the other hand the disadvantages that they present are
the low compressive strength, difficulty of manufacturing and a sensitivity to
ultraviolet light and water. However, Kevlar fibers find applications in sporting
goods.

2.2.3  Boron Fibers

These fibers are characterized by an extremely high tensile modulus coupled with a
large diameter, offering an excellent resistance to buckling that contributes to a high
compressive strength of the composites. Their high cost is due to the processing
operations. For this reason boron fibers find application only in aerospace and mili-
tary contexts.

2.2.4  Graphite Fibers

Carbon fibers owe their success in high performance composites to their extremely
high tensile modulus/weight and tensile strength/weight ratios, high fatigue strength
and low coefficient of thermal expansion, coupled with a low ratio of cost to perfor-
mance. Carbon fibers are commercially available with a variety of moduli ranging
from 270 GPa to 600 GPa. They are produced following two different processes,
depending on the type of precursors.
210 L. Ambrosio et al.

2.3  Matrix Materials

Polymer matrix resins bind the reinforcing fibers and fabrics together in composite
structures. Resins also act as sizing, load distributors, and vibration dampeners in
the composite structure. A wide variety of thermoset and thermoplastic resins are
used in polymer composites. A summary of the most important materials used as
composite matrix and their properties is presented.

2.4  Thermoplastic Matrix

Although many commercial applications for filled thermoplastics exist, use of ther-
moplastics in advanced composites is still in the developmental stage. In fact, the
industry is still split on the use of thermoplastics for advanced composites. According
to some, it is difficult to improve the rigidity, stability and thermal and chemical
resistance of the thermoset resins currently available. Others believe that the major-
ity of the thermoplastic resins available today not only possess the high-­service
temperature characteristics required, but also hold the potential of quicker and more
economical processing once the existing problems are resolved. The various ther-
moplastics and their characteristics are described in the following paragraphs:
Polyamides  commonly referred to as nylons, are produced by condensation
between diamines and diacids or by self-condensation of an amino acid. Like many
other resins, the exact chemistry of polyamides can vary and, therefore, so can the
final properties. However, all polyamides have low-service temperatures and low
melting points.
Polyamideimides  are marketed by Amoco Chemicals under the trade name Torlon. A
polyamideimide is also a good candidate for the production of thermoplastic prepregs.
Polycarbonates  are noted for an extremely high-impact strength in an u­ nreinforced
state. However, compared to such other crystalline polymers as PEEK, a polycar-
bonate does not bond well to reinforcing fibers. It also has poor chemical resistance,
which can be greatly improved by alloying it with thermoplastic polyesters.
Although only a small quantity of polycarbonate with carbon fiber reinforcement is
used today, primarily for interior aircraft structures, polycarbonate is potentially one
of the more promising matrix resins for advanced composites, especially in alloyed
versions such as General Electric’s Xenoy and Bayer’s Makroblend.
Polyetheretherketones  (PEEK) have excellent properties for use in advanced poly-
mer composites, including low flammability, low smoke and toxic gas emission, and
broad chemical and solvent resistance. PEEK possesses a continuous service tempera-
ture of 200 °C to 240 °C (392 °F to 464 °F) and has a very high melting point of 334 °C
(633 °F). PEEK also provides excellent abrasion resistance at its service temperature,
radiation resistance, excellent fatigue and wear resistance and a relatively low specific
gravity of 1.32. It is possible to process PEEK on conventional extrusion and molding
equipment, and its highly crystalline nature responds well to fiber reinforcement.
2  Composite Materials 211

Polyetherimides  (PEI) have been commercially introduced by General Electric as

ULTEM, and the company states that it could be a possible matrix for carbon fiber.
Polyether sulfones  (PES) are also amorphous in structure. Although PES has poor
resistance to solvents, it possesses several valuable properties. The resin is capable
of providing thousands of hours of service at temperatures up to 180 °C (356 °F)
and has very good load-bearing properties. It is dimensionally stable up to 200 °C
(392 °F) and, like PEEK, possesses excellent flame resistance and favorable pro-
cessing characteristics. In addition to thermoset polyimides, several thermoplastic
polyimides are offered for high-temperature applications. Although difficult to pro-
cess, these resins maintain favorable performance characteristics up to a higher tem-
perature, 371 °C (700 °F), than their thermoset counterparts.
Polyphenylene sulfides  (PPS), partially crystalline polymers, are produced by the
reaction of p-dichlorobenzene and sodium sulfide. This polymer has metallic-like
properties and responds well to reinforcement. PPS possesses good creep and good
moisture resistance and a low coefficient of thermal expansion.

2.5  Thermosets Matrix

The basic difference between thermoset and thermoplastic resins is the reaction
heat. A thermoset resin is cured by the application of heat and often by the addition
chemicals called curing agents. Once cured, the material is infusible, unsoluble and
can softened or reworked with the addition of heat.
A thermoplastic, on the other hand, is capable of being repeatedly softened by
addition of heat and hardened by decreasing temperature. The change occurring
thermoplastic resin with the addition of heat is primarily physical, not chemical.
This different provides one major advantage for thermoplastics: any scrap from
fabrication can be reused.
Thermoset resins can vary greatly with respect to service temperature, solvent
resist and other important characteristics. A description follows of the various ther-
moset resin their basic characteristics.

2.6  Vinyl Ester Resins

Vinyl ester resins are thermosetting resins that consist of a polymer backbone with
acrylate or methacrylate termination. The backbone component of vinyl ester resins
can derived from an epoxide resins, polyester resins, urethane resin, and so on, but
those base epoxide resins are of particular commercial significance.
Vinyl ester resins are produced by the addition of ethylenically unsaturated
carbo acids (methacrylic or acrylic acid) to an epoxide resin (usually of the bisphe-
nol epichlorohydrin type). The reaction of acid addition to the epoxide ring (esteri-
fication exothermic and produces a hydroxyl group without the formation of
212 L. Ambrosio et al.

by-­products. Appropriate diluents and polymerization inhibitors are added during

or after esterification. Epoxide resins that have been used to produce vinyl ester
resins include:
• bisphenol A types (general-purpose and heat-resistant vinyl esters)
• phenolic-novolac types (heat-resistant vinyl esters)
• tetrabromo bisphenol A types (fire-retardant vinyl esters)
Vinyl ester resins contain double bonds that react and crosslink in the presence
of free radicals produced by chemical, thermal or radiation sources.

2.7  Epoxide Resins

Epoxide are materials which contain two or more glycidyl groups per molecule. The
uncured resins range from free flowing liquids to high melting solids, which can be
cross-linked by reaction with an appropriate curing agent.
Typical curing agents include primary and secondary amines, polyamides and
organic anhydrides. Other curing agents used are the catalytic curing agents, such
as the boron trifluoride complexes. No by-products are evolved during cure. The
resultant cured resins are generally hard thermoset materials with excellent mechan-
ical, chemical and electrical properties.
These materials can be conveniently divided into six classes of resins:
• bisphenol A based (e.g. diglycidyl ether of bisphenol-A (DGEBA))
• glycidyl esters
• glycidyl amines (e.g. tetraglycidyl amine of 4,4-diamino-diphenyl-methane)
• novolacs brominated resins (e.g. diglycidyl ether of tetrabromo-bisphenol A)
• cycloaliphatic resins (e.g. Tetraglycidyl ether of tetraphenylene ethane)

2.8  Diluents

Diluents are added to epoxide resins primarily to lower viscosity and thus to improve
handling characteristics. They also modify the cured properties of the resin. Diluents
can be divided into two classes: (a) the reactive diluents and (b) the non-reactive
diluents.

2.8.1  Reactive Diluents

• Butane-l, 4-diol diglycidyl ether

• n-Butyl glycidyl ether (nBGE)
• Glycidyl methacrylate
2  Composite Materials 213

• Phenyl glycidyl ether (PGE)

• 2-Ethylhexyl glycidyl ether (2EHGE)
• Iso-octyl glycidyl ether (IOGE)
• Diethylene glycol monobutyl glycidyl ether
• Cresyl glycidyl ether (CGE)
• p-t-Butylphenil glicidyl ether
• Epoxide 7 (C8–C10 glycidyl ether)
• Epoxide 8 (C12–C14 glycidyl ether)
• Dibromocresyl glycidyl ether (BROC)
• Dibromophenyl glycidyl ether (DER 599)

2.8.2  Non Reactive Diluents

• Benzyl alcohol
• Furfuryl alcohol
• Dibutyl phthalate (DBP)

2.9  Curing Agents for Epoxide Resins

These are known variously as curing agents, hardeners, activators or catalysts. They
are required to convert liquid and solid epoxide resins into tough infusible thermo-
set polymers. The curing agents promote this curing reaction by opening the epox-
ide ring and become chemically bound into the resin in the process. Each of the
curing agents for epoxide resins will now be discussed in turn.

2.9.1  Amine Curing Agents

Amine curing agents may be primary or secondary amines, aliphatic, alicyclic or

aromatic. The reaction with an epoxide resin is an addition reaction where the amine
links directly with an epoxide group to form a combined polymer, with hydroxyl
groups formed during the reaction. The amines commonly used are the following
• Ethylenediamine (EDA)
• Trimethylhexamethylenediamine (TMD)
• Diethylenetriamine (DTA)
• 2-Hydroxyethyldiethylenetriamine (T)
• Dipropylenetriamine (DPTA)
• Triethylenetetramine (TETA)
• Tetraethylenepentamine (TEPA)
• Diethylaminopropylamine (DEAPA)
214 L. Ambrosio et al.

• Dimethylaminpropylamine (DMAPA)
• m-Xylylenediamine (mXDA)
• N-Aminoethylpiperazine (AEP)

2.9.2  Anhydride Curing Agent

These consist of organic anhydrides and are used in roughly stoichiometric propor-
tions with epoxide resins. The anhydride commonly used are the following:
• Phthalic anhydride (EPA)
• Tetrahydrophthalic anhydride (THPA)
• Methyltetrahydrophthalic anhydride (MTHPA)
• Endomethylenetetrahydrophthalic anhydride (NA)
• Hexahydrophthalic anhydride (HHPA)
• Methylhexahydrophthalic anhydride (MHPA)
• Trimellitic anhydride (TMA)
• Pyromellitic dianhydride (PDMA)
• Dodecenylsuccinic anhydride (DDSA)

2.9.3  Accelerators for Anhydride Cured Systems

Various accelerators can be used with epoxy/anhydride systems to promote cure.

Some accelerators in use are:
• Benzyldimethylamine (BDMA)
• Tris(dimethylaminomethyl)phenol
• 1-Methylimidazole (DY 070 from Ciba-Geigy)
• N-Butylimidazole
• 2-Ethyl-4-mathylimidazole
• Triamylammonium phenate (DY 063 from Ciba-Geigy)

2.9.4  Polyamide Curing Agents

The polyamides used to cure epoxide resins are all reactive compounds with free
amine groups. They may be amidopolyamines, aminopolyamides or imidazolines.
They are mostly used in coating systems and in adhesive formulations. Some sup-
pliers of polyamide curing agents are listed below:
• Ancamide (Anchor Chemical Co, Ltd)
• Araldite (Ciba-Geigy Plastics & Additives Co.)
2  Composite Materials 215

• Versamid, Genamid, Synolide (Cray Valley Products Ltd)

• Plastamid (Croda Resin Ltd)
• Grilonit (Grilon (UK) Ltd)
• Beckopox (Hoechst AG(Reichhold Albert Chemie AG))
• Euredur (Schering Chemicals Ltd)
• Epikure (Shell Chemicals UK Ltd)
• Casamid (Thomas Swan & Co. Ltd)
• Uracure (Synthetic Resins Ltd)
• Thiokol (Thiokol Chemicals Ltd)
• Wolfamid (Victor Wolf Ltd)

2.9.5  Other Curing Agents

Several other types of curing agent are used with epoxide resins for laminating
applications or in moulding compounds. Examples of these curing agents are:
• Dicyandiamide (Dicy)
• Boron trifluoride complexes
• Boron trifluoride monoethylamine
• 2-Ethyl-4-methylimidazole
• N-n-Butylimidazole

2.10  Polyester Resins

The basic materials used to make a polyester resin are a dibasic organic acid or
anhydride and a dihydric alcohol.

2.10.1  Catalysts or Initiators

Catalyst or initiators for unsaturated polyester resin system consist of organic per-
oxides. Some commercially available catalysts are:
• Diacyl peroxides (benzoyl peroxide, 2,4-dichlorobenzoyl peroxide, dilauroyl
peroxide, diacetyl peroxide).
• Ketone peroxides (methyl ethyl ketone peroxide, cyclohexanone peroxide, acet-
ylacetone peroxide, methyl isobutyl ketone peroxide).
• Hydroperoxides (t-butyl hydroperoxide, cumene hydroperoxide).
• Dialkyl and diaralkyl peroxides (dicumyl peroxide, di-t-butyl peroxide, t-butyl
cumyl peroxide, 2,5-dimethyl-2,5-bis(t-butylperoxy)hexane).
216 L. Ambrosio et al.

• Peroxyesters (t-butyl peroxybenzoate, t-butyl peroxydiethylacetate, t-butyl

peroxyester, t-butyl peroxyisononanoate, t-butyl peroctoate, dit-butyl
­
­diperoxyphthalate, t-buthyl peroxypivalate).
• Perketals (2,2-bis(t-butylperoxy)butane, l,l-bis(t-butylperoxy)cyclohexane,
l,l-bis(t-butylperoxy)-3,3,5-trimethylcyclohexane).

2.10.2  Accelerators or Promotors

These are materials which when used in conjunction with an organic peroxide cata-
lyst increase the rate at which that peroxide breaks down into free radicals. Some of
the commercially available accelerators are:
• Cobalt accelerators: cobalt siccatolate, naphthenate or octoate
• Manganese accelerators: manganese salts
• Vanadium accelerators
• Tertiary amine accelerators: dimethylaniline (used to accelerate diacyl peroxide
catalysed system); diethylaniline (used to accelerate benzoyl peroxide catalysed
system); dimethyl-p-toluidine (used to accelerate benzoyl peroxide catalysed
system).

2.11  Laminate Properties

Current attitudes regarding composite materials emphasize the relationship of struc-

tural performance to the properties of a ply. A ‘ply’ is a thin sheet of material con-
sisting of an oriented array of fibers, embedded in a continuous matrix material.
These plies are stacked one upon other, in a definite sequence and orientation, and
bonded together yielding a laminate with tailored properties. The properties of the
laminate are related to the properties of the ply by the specification of the ply thick-
ness, stacking sequence, and the orientation of each ply. The properties of the ply
are, in turn, specified by the properties of the fibers and the matrix, their volumetric
concentration, and geometrie packing in the ply. Generally, the material is pre-
formed and can be purchased in a continuous compilant tape or sheet form which is
in a chemically semicured condition. Fabrication of structural items involves using
this ‘prepreg’ material, either winding it on to a mandrel or cutting and stacking it
on to a mold, after which heat and pressure or tension is applied to complete the
chemical hardening process.
The basis for engineering design of a such material is then the properties of a
cured ply or lamina as it exists in a laminate. This ply is treated as a thin two-­
dimensional item and is mechanically characterized by its stress–strain response to:
(i) loading in the direction of the filaments, which exhibits a nearly linear response
up to a large fracture stress; (ii) loading in the direction transverse to the filament
2  Composite Materials 217

orientation, which exhibits a significantly decreased moduli and strength, and (iii)
the response of the material to an in-plane shear load.
By the contrast with isotropic metallic materials, an oriented ply, in the form of
a thin sheet, is anisotropic and requires four elastic (plane stress) constants to spec-
ify its stiffness properties in its natural orientation

σ 1 = Q11ε1 + Q12ε1
σ 2 = Q12ε1 + Q22ε2 (2.1)
σ 6 = Q66ε6
where σ6=τ12 and є6=τ12 or in matrix form

σ 1 Q11 Q11 0 ε1
σ 2 = Q11 Q11 0 . ε2 (2.2)
σ 3 0 0 Q11 ε3

where the plane stress stiffness moduli are

Q11 = E11 / (1 −ν 12 ν 21 )
Q22 = E22 / (1 −ν 12 ν 21 )
(2.3)
Q12 = ν 21 E11 / (1 −ν 12 ν 21 )
Q6 = G12

where vij is the Poisson ratio, defined as -є/єj.
If, however, the ply is rotated with respect to the applied stress or strain direction
additional moduli appear, which results in the direction-indicated shear coupling
rotation simple extension

σ 1 Q11* Q12* Q16* ε1

σ 2 = Q12* Q22* Q26* . ε2 (2.4)
σ 3 Q16* Q26* Q66* ε3

where

Q11* = U11 + U 2 cos ( 2θ ) + U3 cos ( 4θ )

Q22* = U1 + U 2 cos ( 2θ ) + U3 cos ( 4θ )
Q12* = U 4 + U3 cos ( 4θ )
(2.5)
Q66* = U 5 + U3 cos ( 4θ )
Q16* = −1 / 2U 2 sin ( 2θ ) − U3 sin ( 4θ )
Q26* = −1 / 2U 2 sin ( 2θ ) − U3 sin ( 4θ )

218 L. Ambrosio et al.

The invariants Ui to the rotation are

8 ( 11
U1 = 1 3Q + 3Q22 + 2Q12 + 4Q66 )

U2 = 1 (Q − Q )
2 11 22

U3 = 1 ( Q + Q − 2Q − 4Q ) (2.6)
8 11 22 12 66

U4 = 1 ( Q + Q + 6Q − 4Q )
8 11 22 12 66

U5 = 1 ( Q + Q − 2Q − 4Q )
8 11 22 12 66

In addition, lamination can result in up to 18 elastic coefficients and increased
deformational complexities, but the additional coefficients can all be derived from
the four primary coefficients using the concept of rotation and ply-stacking
sequence. These complications are the result of geometric variables. If the laminate
is properly constructed, the in-plane stretching or stiffness properties can still be
specified by four elastic coefficients. We shall consider laminates of this nature.
Note that both short and continuous fibers are handled in the same manner. These
calculations, while tedious, are analytically simple. The ‘plane stress’, the Qij terms,
are employed because lamination neglects the mechanical properties through the
ply thickness. These stiffnesses are sometimes regrouped into new constants called
‘invariants’, the Ui terms, for analytical simplicity. To compute the properties of the
laminate one then sums the ply (hk) properties through the thickness of the laminate,
weighted by the thickness (hk) of each oriented ply
N
Aij = ∑(Qij )k hk
k=1
For a balanced (same number of ±θ) and symmetrical system (+θ or − θ at same
distance above and below the midplane) the laminate solution is

A11 = U1 + U 2 cos(2θ ) + U3 cos(4θ )

A22 = U1 + U 2 cos(2θ ) + U3 cos(4θ )
(2.7)
A12 = U 4 − U3 cos( 4θ )
A66 = U 5 − U3 cos( 4θ )

Note the inverted terms Ajj yield the required elastic properties of the laminate in
terms of the individual ply properties E11, E12 and G12.

(
E11 = A11 A22 − A212 / A22 )
E22 = (A11 A22 − A 2
12 )/ A
11 (2.8)
ν 12 / E11 = A12 / ( A11 A22 − A212 ) G12 = A66

2  Composite Materials 219

These calculations have been thoroughly tested and agree closely with
e­ xperiment. The circles and squares are the experimental points and the lines are the
theoretical predictions for a nylon fiber reinforced rubber. The angle ply laminate is
predicted from the ply properties. The ply properties are in turn correlated with the
transformation equations and the micromechanics. The micromechanics employed
in this demonstration are based upon the ‘self-consistent method’ developed by Hill
(8). Hill rigorously modeled the composite as a single fiber, encased in a cylinder of
matrix, with both embedded in an unbounded homogeneous medium which is mac-
roscopically indistinguishable from the composite. Hermann (9) employed this
model to obtain solution in terms of Hill’s ‘reduced moduli’. Halpin and Tsai (10)
reduced Hermann’s solution to simpler analytical form and extended its use for a
variety of filament geometries

E11 = E f V f + EmVm
ν 12 = ν f V f + ν mVm (2.9)
p / pm = (1 + ηξ V f ) / (1 − ηV f )

where

η = ( t f / t m − 1) / ( p f / pm + ζ )
ξ = ( e / d ) ; ζ G12 = 1ζ G 23 = 1 / ( 3 − 4ν m ) (2.10)
p = E22 , G12 , G23 ; p f = E f , G f ; pm = Em , Gm

These equations are suitable for single calculation and were employed p­ reviously
for the single ply and angle ply properties. The short fiber composite properties are
also given by the Halpin-Tsai equations where the moduli in the fiber orientation
direction is a sensitive function of aspect ratio (1/d) at small aspect ratios and has
the same properties of a continuous fiber composite at large but finite aspect ratios.
If the ply is used in the construction of a balanced and symmetrical 0/90 laminate
and is mechanically tested, bilinear stress/strain curve is obtained, and the stiffness
is the sum, through the thickness of the plane stress stiffness of each layer. As the
laminate is deformed each ply possesses the same in-plane strain, and when the
strain on the 90 layers in the laminate prevents the 90° layer from carrying its share
of the load, Qij(90°) = 0. This load is transferred to the unbroken layers, the 0° layers
for our illustration, and results in a loss of laminate stiffness or modulus. Continual
loading will ultimately produce a catastrophic failure of the laminate when the
strain capability of the unbroken, 0°, layers is exceeded. For a 0/90 construction,
employing glass/epoxy material, the ratio of the ultimate failure stress to the crazing
stress is 6.1. Note a change in stiffness as the 90° and then the 45° layers fail, and
the correspondence of the theoretical ultimate strength of 356 MPa with the experi-
mental results of 346 MPa. While the strain for transverse ply failure is constant
from laminate to laminate, the stress required to craze the system as well as cause
final failure is a function of laminate geometry, because the construction of the
220 L. Ambrosio et al.

laminate specifies the stiffness properties (crazing stress = stiffness x allowable

transverse ply strain). It must be noticed that the area under the stress/strain curve is
proportional to the impact energy. Therefore, lamination permits the engineer to
tailor a fixed prepreg system to meet the conflicting stress/strain demands at differ-
ent points in a structure. A further point, the crazing stress threshold is generally at
or below the creep fracture or fatigue limit for all classes of composites (for glass/
epoxy the fatigue limit lies between 0.25 and 0.30 of static ultimate strength). Boron
and graphite are fatigue insensitive filaments, thus no fatigue damage is realized
below first ply failure.
Thus, the material properties of a laminate are specified in terms of the ply engi-
neering moduli, E11, E22, ν12 and G12; the engineering strains to failure, ϵ1, ϵ2 and ϵ6;
and the thermal expansion coefficients, ϵ1 and ϵ2.

2.12  Composite Fabrication

Various fabrications used in the reinforced plastics industry are discussed below:

2.12.1  Hand Lay-Up and Spray-Up Procedures

In one of the simplest and labor intensive procedures, pigmented, catalysed resin is
applied to the surface of the mold. This gel coat in room temperature lay-up tech-
niques will end up on the surface of the finished composite (FRP). Catalysed resin-­
impregnated mat is then applied over the gel coat and this and subsequent layers are
brushed or rolled to assure good contact between layers and to remove any entrapped
air. This procedure is continued until the desired thickness of the composite is attained.
The assembled composite may be cured at room temperature or at elevated tem-
peratures for faster cycles. This procedure, which was originally called contact mold-
ing, may be upgraded by the application of a vacuum or pressure bag placed over a
Cellophane film on the final layer to reduce void formation in the composite. The
laminate may also be built up by a spray-up process in which a mixture of chopped
glass strands and catalysed resin is sprayed on the gel coat instead of resin-­
impregnated mat. In any case, the inner surface will be less smooth than the first layer
formed by the gel coat. Tanks, boats and pipe may be fabricated by this technique.

2.12.2  Centrifugal Casting

Fiber-reinforced plastic pipe (FRP) can be produced by rotating a mixture of

chopped strand and catalysed resin inside a hollow mandrel. Because of differences
in specific gravity, there is a tendency for these composites to be less homogeneous
than those produced by other techniques.
2  Composite Materials 221

In addition to being available as continuous filaments and staple fibers in mats,

fiberglass textiles are also available as biaxial, triaxial, knitted and three dimen-
sional braided patterns. Many different resin matrices are in use but the emphasis in
this chapter will be on unsaturated polyester and epoxy resins. While the strength
and stiffness are controlled primarily by the reinforcements, the resinous matrix
contributes to thermal conductivity and flexibility. The ultimate properties of these
composites are based on a harmonious contribution of both the continuous and dis-
continuous phases.

2.12.3  Matched Die Molding

Matched die molding of a premix of chopped glass, roving, and catalyzed resin is
used for relatively large scale production of reinforced articles. Uncured dough-like
compositions are called bulk molding compounds (BMC). Uncured resin-­
impregnated sheets are called sheet molding compound (SMC). These compounds
are supplemented by thick molding compounds (TMC) and XMC.  TMC is pro-
duced continuously on a machine that resembles a 2 roll mill. XMC, in which the
continuous impregnated fiber are arranged in an X-pattern, is produced on a fila-
ment winding machine.
Autoclave Molding  Autoclave molding, the process of curing thermoset resins at
elevated temperature and pressure in an inert environment, has an important role in
the fabrication of continuous fiber reinforced thermoplastics. While most compa-
nies view thermoplastics as an alternative to traditional autoclave long-cycle pro-
cessing, they have come to accept the following reasons for the autoclave processing
of thermoplastic matrices:
• Availability
• High temperature and pressure capability
• Reduced tooling requirements
• Uniform pressure distribution.
The dominant reason for the autoclaves’ role in thermoplastics production is its
availability. Aerospace first- and second-tier contractors, who conducted much of
the developmental work in parts fabrication, all have autoclaves on hand.
While it may seem defeatist to use the autoclave for thermoplastic resins which
undergo no chemical reaction and lend themselves to rapid fabrication, autoclave
use offers other advantages.
In the industry, many fabricators own autoclaves capable of processing high tem-
perature thermosets (e.g. polyimides) at operating temperatures up to 800 °F and
pressures of 150 to 200 psi; which can also accommodate high temperature thermo-
plastics such as PEEK, which is normally processed in the 700 to 750 °F range. One
disadvantage, particularly for production-sized autoclaves, is the inability to finely
control cool-down rate – a critical process step for the semicrystalline thermoplas-
tics. This situation can be improved by implementing integrally cooled tooling.
222 L. Ambrosio et al.

Where high consolidation pressures are required, the autoclave can reduce tool-
ing costs by eliminating the need for matched metal tooling, when compared to a
molding process (e.g. compression molding or thermoforming). Less expensive
tooling aids can be used for consolidation via the autoclave’s pressurized environ-
ment. The autoclave also provides uniform pressure over the part’s area and elimi-
nates pressure distribution concerns associated with a matched tool in a press.
Thermoforming  Thermoforming offers vast potential for high volume thermoplas-
tic composite parts fabrication. There are many thermoforming variations; but by
basic definition, thermoforming is the heating of a reinforced thermoplastic matrix
sheet or kit above the softening temperature, followed by forcing the material
against a contour by mechanical (e.g. matched tooling, plug) or pneumatic (e.g. dif-
ferential air or hydraulic) means. The material is then held and cooled sufficiently
for shape retention, and removed from the mold. Thermoforming implies only those
processes applicable to thermoplastic resins, and is often used in the same context
as the term ‘compression molding’ which also applies to thermosets. In this section,
the process is defined as the preheating of the lay-up or reconsolidated sheet, fol-
lowed forming via a matched mold.
Phillips Petroleum has applied this technology to fabric reinforced Ryton (PPS)
materials. A conveyorized infrared oven is used to rapidly preheat the lay-up to 600
°F in two to three minutes. The charge is quickly transferred to a preheated mold in
a fast closing press for part forming. Total cycle times of one to three minutes are
feasible with this automated approach.
High production rates can be achieved using thermoforming technology.
However, it is difficult to form high quality continuous fiber reinforced thermoplas-
tic parts with demanding geometries, due to the restricted movement of the fiber.
Du Pont’s long discontinuous fiber sheet products with PEEK and J-2 resins
provide easier fabrication of complex shapes. This product form is particularly
attractive to helicopter manufacturers for the press forming of highly contoured
secondary structure parts.
Transfer molding  Transfer molding is used for the manufacture of small compo-
nents and is particularly useful with multi-cavity tools and where small inserts are
to be moulded in. Materials used are polyester and epoxide dough moulding com-
pounds, although a new liquid resin injection technique is reported.
Heated steel molds, preferably hard chrome plated, are used, which may be of
multicavity design. Tooling costs are higher than for compression moulding since
appropriate gates and runners must be included in the mould.
A pre-weighed quantity of DMC is placed in a heated transfer pot by hand. A
punch or ram compresses the material and causes it to flow into the heated tool cav-
ity where it cures. The tool is mounted between the platens of a press.
Factors to be considered with transfer molding are transfer and tool clamping
pressures and transfer time. To reduce transfer time and increase overall efficiency
the molding compound may be pre-heated in an oven or high frequency pre-heater
such as a micro-wave oven.
2  Composite Materials 223

Mold temperatures range from 155 to 170 °C both for polyester and epoxide
resin compounds, with molding pressures ranging from 5 to 100 MPa depending on
the type of compound to be processed, mold design and temperature. Cure time in
the mold (excluding pre-heat time) is usually of the order of 10–30 s per millimetre
of wall thickness for both types of compound.
Injection molding  Injection molding, a technique used extensively for the process-
ing of thermoplastic materials, has also been developed to process thermosetting
resin systems. Due to high mould costs it is generally only suitable for the large
scale production of small-to-medium sized components. Materials processed in this
way are polyester and epoxy DMC and also phenolics, ureas, melamines and diallyl
phthalate moulding compounds. These latter materials are generally more difficult
to process than either polyester or epoxy DMC.
Thermoset moldings produced by injection moulding are used widely in the
electrical and automotive fields, thus large production runs are common.
Injection molding has advantages over both compression and transfer molding in
that the process is more automated and far higher production rates can be achieved.
Although mould costs are higher than for compression molding, overall finished
component costs are generally lower. With small weight components, scrap from
runners can be high compared with compression molding but for large mouldings
this becomes relatively insignifiant. Injection molding is also better for thick parts
since, with the pre-heating of the DMC before injection into the mold, shorter mold-
ing cycles are possible.
While injection molding machines designed specifically for processing thermo-
set materials are available, a number of manufacturers offer replacement screw and
barrel assemblies and stuffer hoppers to convert conventional thermoplastic injec-
tion moulding machines to process thermosets.
Molding compound is transferred in the cold state by pressure from the material
hopper into the main injection chamber. Here it can be preheated before injection
into the heated mould tool. Injection, through a special nozzle, can be either by ram
or screw pressure. If screw feed is used, the screw must be of the type designed to
process thermosets as opposed to thermoplastics.
Early machines were designed with vertical clamping pressure on the mold but
today horizontal machines are mostly used. Since thermosetting materials are liq-
uid until gelation occurs, clamping pressure has to be maintained on the mould
until the resin has cured. Unless this is done, excessive flash will form. Heated
matched metal molds are used, which may be of multi-cavity design. These molds
must be designed for use with thermosetting resins, taking into account the fact
that thermoset moldings are harder, more rigid and less easily deformed than
thermoplastics.
A typical temperature sequence for injection molding DMC would be: feed hop-
per and feed zone – ambient temperature; metering section 50–60 °C; nozzle 80–90 °C;
mould temperature 135–185 °C for polyester DMC or 160–22 °C for epoxy DMC;
injection pressure 80–160 MPa. Cure time is generally of the order of 10–20 s
per millimetre of wall thickness. Very little finishing of moldings is necessary.
224 L. Ambrosio et al.

Where fully automatic molding machines are used, hydraulic ejection with
p­ erhaps a ‘joggle’ facility is necessary, since thermosets have a tendency to stick in
the mold.

2.12.4  Filament winding

Filament winding is a technique used for the manufacture of pipes, tubes, cylinders
and spheres and is frequently used for the construction of large tanks and pipework
for the chemical industry. By suitable design, filament wound structures can be
fabricated to withstand very high pressures in service. In general, products fabri-
cated by filament winding have the highest strength to weight ratios and can have
glass contents of up to 80 % by weight.
The process is suitable for use both with polyester and epoxide resin systems and
a variety of fibres including glass, carbon, aramid and metals, providing that these
materials are available in continuous filament lengths. Glass fibre is by far the most
common reinforcement used and will be used as the example in the description of
the process.
Filament winding is basically a simple process, although numerous modifica-
tions have been developed to improve product quality. Moldings can be produced
by either a wet lay-up process or from prepreg.
In recent years filament winding has been extended to the continuous production
of pipe using a continuous steel band mandrel. In this way continuous lengths of
pipe can be produced, with diameters ranging from about 0.3 to 3.5 m.

2.12.5  Wet lay-up

With the wet lay-up process glass rovings are drawn through a resin bath to impreg-
nate them with resin. The impregnated rovings are then wound under tension round
a rotating mandrel. Generally the feed head supplying the rovings to the mandrel
traverses backwards and forwards along the mandrel.
The mandrel, which may be segmented for large diameter pipes, is generally
wrapped with a release film, such as Cellophane, prior to wrap- ping with glass and
resin. The mandrel may incorporate some means of heating the resin system, such
as embedded electric heaters, or provision for steam heating. Alternatively, the fully
wrapped mandrel and laminate may be transferred to a curing oven to effect cure.
In order to provide a resin-rich, corrosion resistant inner lining to the pipe, the
mandrel may be wrapped with a surfacing tissue followed by one or two layers of
chopped strand mat or woven tape prior to filament winding. This first layer is usu-
ally allowed to cure partially before winding commences to prevent the resin from
being squeezed out into the main laminate.
2  Composite Materials 225

The winding angle used during construction of pipes or tanks depends on the
strength/performance requirements and may vary from longitudinal through helical
to circumferential. Often a combination of different winding patterns is used to give
optimum performance. Accurate fibre alignment is possible.
For pipe construction, steel mandrels are generally used. However, where cylin-
ders or spheres are to be made, an alternative material has to be used so that it can
be removed once the resin system has cured. In these cases the mandrel can be made
from wax, a low melting metal alloy, or an inert plaster held together with a water
soluble binder. Clearly, in these cases the mandrel can only be used once. Material
choice for the mandrel will depend on the cure cycle needed for the resin system.
In addition to winding with continuous filament rovings, machines have been
developed which permit winding with tapes or slit-width chopped strand mat and
woven rovings. These reinforcements may be used alone or combined with continu-
ous filament rovings. Thus considerable design flexibility exists for the production
of large simple shapes.
Improved chemical resistance can be achieved by the use of a thermoplastic or
synthetic rubber liner. If the liner is sufficiently rigid it can be supported on a light
frame and used as the mandrel, if not then it can be wrapped round the mandrel first.
A grade of polypropylene is available which has a woven glass cloth partially rolled
into one side to improve adhesion of the resin system (Celmar).
Dunlop has recently developed a new process for pipe production to produce
pipes in the range 200–2000 mm diameter. Essentially the process is similar to con-
ventional filament winding. A mandrel, suitably coated with release agent, is
wrapped with an epoxide resin impregnated glass tape over this is wound a 150 mm
high-strength steel strip angled to give 50 % overlap. Epoxide resin system is
applied to the steel strip to ensure that each layer is fully encapsulated. From three
to 13 layers of steel may be applied to satisfy different pressure ratings. The pipe is
finished by wrapping with further resin impregnated glass tape and the resin system
cured. Pipe produced in this way has excellent corrosion resistance coupled with a
high strength/weight ratio. It is said to provide up to 50 % weight saving over con-
ventional steel pipe.
Various processes are available for the ‘on-site’ construction of large filament
wound storage tanks. By manufacturing these on site, transport problems are over-
come and integral structures can be produced. With the various processes either
horizontal or vertical mandrels are first constructed from preformed GRP sheets.
These are then wrapped with resin impregnated glass rovings.
Filament wound vessels can be produced from prepreg tapes and rovings. This
technique is often used with carbon fibre to reduce fibre damage during the winding
operation and to permit the use of resin systems which cannot be handled by wet
lay-up techniques. Here, it is essential to use a heated mandrel to melt the resin and
hence displace air and consolidate and cure the laminate. Resin content of the lami-
nate can be controlled more accurately with prepreg since the prepreg can be made
with exactly the right resin content. The use of prepregs also makes for cleaner
operation.
226 L. Ambrosio et al.

Filament winding has been used to provide a protective laminate on the outside
of steel pressure pipes where external corrosion can take place. An example of this
use is in the protection of the splash zone of steel riser pipes used on sea based oil
and gas production platforms. Here, care has to be taken in the design of such a
composite structure since the coefficient of expansion of the filament wound glass
wrap can be lower than that of the steel core. If such a composite structure is pro-
duced using a heat cured resin system (say 120 °C cure) and then subjected to sub-
zero temperatures in use, the steel pipe can shrink away from the laminate and
permit entry of water by capillary action. Thus the object of wrapping the pipe to
prevent corrosion can be defeated, since corrosion can then still take place under the
laminate. once the bond has broken it can never be remade.

2.12.6  Centrifugal Moulding

Centrifugal moulding or casting is a method used for making cylindrical objects

with uniform wall thickness. It is mainly used for the production of large diameter
pipes, up to 5 m in diameter, from either polyester or vinyl ester resin systems,
although epoxide resin systems may also be used. Pipes produced in this way are
void-free and smooth on both the inner and outer surfaces. Threaded sections can be
molded into the external wall if required. More recently, a system has been devel-
oped for producing tapered and curved poles by centrifugal casting.
Molds need to be bored and polished to a mirror finish and of sufficient strength
to withstand, without distortion, the high G-force exerted during spinning. A steam
jacket or other means of heating may be built into the mould to cure the resin sys-
tem, or alternatively hot air may be blown through the mold.
Release agents used include silicones, bake-on PTFE types or PVA, although
silicones are generally preferred. If the exterior of the finished pipe is to be painted
then a silicone release agent should not be used. Choice of resin system must depend
on application and a heat cured resin may be used, particularly where chemical
resistance is required. In addition, the process lends itself to the use of several dif-
ferent resin systems incorporated in the one pipe, such as an abrasion resistant outer
skin, a general purpose centre and a chemically resistant inner skin. The resin sys-
tem used should have some degree of flexibility to give good impact resistance to
the pipe, coupled with good chemical resistance. To achieve both of these require-
ments in the one resin system may require a compromise in properties.
Various methods of fabrication can be used. With large diameter moulds, glass
and resin can be applied by the hand lay-up technique using a slow gel resin system
so that the whole mold can be coated and the mould spun before gelation of the
resin takes place. Here, it is necessary to tailor the reinforcement to shape to avoid
overlaps.
Alternatively, the reinforcement may be wrapped round a mandrel, inserted into
the mould and then unwound onto the mould surface. The mandrel is removed before
resin is injected into the mold. With this technique a faster gelling resin system may
be used. Woven fabrics and chopped strand mat are suitable reinforcements.
2  Composite Materials 227

A third method, capable of being fully automated, is generally preferred. Here,

resin and glass are applied to the mold surface utilizing a travelling feeder arm fitted
with a chopper and spray gun, which passes slowly backwards and forwards through
the mold while the mold is rotated. The laminate can be built up in layers of
0.5–1 mm thickness per pass once the reinforcement and resin system have been
placed in the mold it is rotated at up to 2500 revolutions per minute (rpm) depending
on mold diameter, the larger the mould diameter the slower the speed. For example,
with a 2 m diameter mould a rotation speed of about 180 rpm is used. At this speed
the mould surface is rotating at about 68 km/h (42.4 mph).
After the main resin system has gelled, a chemically resistant topcoat may be
applied while the mold is still rotating. Mold rotation is continued until all the resin
has cured.
In all of the above techniques, a relatively flexible glass reinforced resin pipe is
produced, with properties similar to some of those made by filament winding. To
produce stiffer pipes a modification to the spray technique has been developed and
is in use commercially. Stiffer pipes, particularly in the larger diameter range, offer
considerable advantages in handling and installation and maintain their shape dur-
ing installation.
With this modified process, which is fully automatic, the mold is coated with
release agent and rotated at a suitable speed. The feeder arm is designed in such a
way that it can deliver programmed amounts of resin, chopped glass and a filler
such as sand, to the mold surface as it moves in and out of the mold. In this way a
layer of abrasion resistant sand filled resin system can be applied to the mold sur-
face over this is applied a layer of glass reinforced resin. By suitable design of the
chopper unit, fibre orientation can be controlled. Next a layer of sand filled resin is
applied followed by a further layer of glass reinforced resin. Depending on the type
and size of pipe to be produced, several more layers of GRP alternated with filled
resin may be added. Finally, the inner surface is coated with a suitable chemically
resistant layer of resin, which may be lightly filled or reinforced, to give a smooth
corrosion resistant lining.
The equipment used can be programmed to feed materials to the mould to build
up the pipe wall thickness at a rate of between 0.5 and 1 mm per pass and to com-
pact the materials by centrifugal force throughout the whole production cycle once
the final layer of resin has been applied, hot air is passed through the mold to assist
curing of the resin system. The mould is cooled to ambient temperature to assist
removal of the finished pipe, which is pushed out using a hydraulically operated
ram. Using this technique, rigid hard wearing pipes can be produced, tailored to
meet end user requirements. This is a simplified version of the pipes produced com-
mercially, to illustrate pipe construction. In practice several more layers of glass
and sand filled system may be incorporated in the pipe wall. A range of bends and
joints is also available to meet most needs.
In the Usui process for producing tapered pipes, the glass fibre reinforcement is
wound round a tapered mandrel to make a preform. This is inserted into the mould
in the centrifugal machine and the mandrel removed. Resin is poured in and the
mold tilted to a pre-determined angle and then rotated until the resin has cured.
228 L. Ambrosio et al.

To produce a curved tapered pipe, a flexible mold is used. This is bent to the
required shape once the preform has been placed in position. It is also claimed that
base plates can be simultaneously molded on by this method. These base plates are
first preformed and then inserted into the mold where they become firmly bonded into
the pole. Typical applications include poles for street lighting, flag poles and aerials.

2.12.7  Continuous Sheet Manufacture

For the purposes of this book, only those processes for which polyester resins and,
to a limited extent, epoxide resins are used will be described. It should be noted,
however, that the decorative laminates used in building and transport applications
are in the main manufactured from melamine faced phenolic resin/paper laminates.
Several patented processes exist for continuous sheet production, all of which are
similar in broad principle.
Resin and glass reinforcement are sandwiched between two sheets of release
film, such as Melinex, Mylar or Cellophane and passed through rollers to consoli-
date the laminate before curing in an oven. Resin is applied to the release film either
by spray or trickle process, care being taken to ensure that application is uniform.
Glass reinforcement is laid in the resin and a second layer of release film applied.
This sandwich is passed through a series of rollers to expel all air bubbles and con-
solidate the laminate to the correct thickness. During the next stage the laminate
sandwich is either passed directly through an oven to produce flat sheet or through
rollers or dies and then an oven to produce corrugated sheet. Once cured the sheet
is trimmed to the required width and cut into suitable lengths. Depending on the
process, corrugations may run longitudinally or transversely. Production speeds of
up to 12 m/min are possible.
To produce clear sheeting the refractive index of the resin system must match
that of the glass reinforcement. For this reason special resins have been developed
which match the refractive index of E-glass. For translucent sheeting A-glass may
be used but, due to its low refractive index, it is unsuitable for use in transparent
sheeting. In any case today A-glass is rarely found.
Generally the glass reinforcement used consists of chopped strand mat with a
soluble polyester powder binder or chopped rovings deposited directly into the resin
film. However, for certain applications woven fabrics may be used. With the latter
and to a much lesser extent with chopped strand mat, the glass cloth may be drawn
through a resin bath and excess resin removed between doctor blades or rollers,
before placing between two layers of release film and processing as before.
With high quality sheeting a surfacing tissue may be used to ensure a resin-rich
finish. Such sheeting must be installed with the resin-rich surface exposed to the
weather.
Resin systems are generally specifically formulated for each machine, since gel
time and viscosity must suit the particular operating conditions of the machine. Resin
systems used include those suitable for producing clear fire retardant sheeting.
2  Composite Materials 229

By the correct choice of resin system, sheeting can be manufactured which will
not yellow to any extent after exposure to tropical weather conditions for several
years. However, to ensure that this is the case the resin system must be chosen with
care and must be fully cured. Also the release film must be removed before installa-
tion and the laminate should contain not less than 75 % by weight of resin. In other
cases the resin content of the laminate may fall between 65 and 75 % by weight.

2.12.8  Pultrusion

Pultrusion is a technique used for producing continuous fibre reinforced sections in

which the orientation of the fibres is kept constant during cure. The process is suit-
able for use with both polyester and epoxide resin systems, reinforced with glass,
carbon or synthetic fibres. An infinite number of profiles can be produced using
appropriate dies and includes rods, tubes and flat and angled sections. All profiles
have high strength and stiffness in the lengthwise direction, with fibre content gen-
erally around 60–65 % by volume.
The reinforcements used consist of continuous fibres such as glass rovings or
continuous carbon fibre tows, woven rovings or chopped strand mat or a combination
of the two, depending on the strength and rigidity required in the molded profile.
Two processes are available which use liquid resin systems. In the first the rein-
forcement is drawn through an impregnating bath containing catalysed resin. For
this process, a resin system with a long pot-life at room temperature is necessary.
The reinforcement is then drawn through a heated die which removes excess resin,
determines the cross-sectional shape and cures the resin system.
In the second process the reinforcement, accurately positioned and under ten-
sion, is drawn through a heated metal die where impregnation of the fibres and cure
of the resin system takes place. Here, by the use of appropriate resin injection equip-
ment, a short pot-life system can be used. Typical resin injection pressures are
between 0.1 and 0.5 MPa. To speed up cure, the reinforcement may be pre-heated
to about 100°C before passing through the die. Production rates of I m/min can be
achieved. By careful design of the pulling mechanism, consistent profiles can be
produced with no bending or twisting of the fibres. With some resin systems a tun-
nel oven may be required after the die to give a suitable post cure.
Apart from the wet processes it is also possible to make pultruded sections from
prepregs. The forming procedure is the same as that used with wet resin systems.
The prepreg is drawn through a heated die which melts the resin, compresses the
prepreg into the required shape and cures the resin. This is a somewhat cleaner pro-
cess than that using a resin bath.
It is reported that sandwich panels are being produced in the USA by pultrusion.
In this process a plywood core is completely encased in a 3 mm thick glass polyester
skin, resin penetrating the plywood during production to give increased bond
strength and moisture resistance.
230 L. Ambrosio et al.

2.13  Mechanical Properties

A summary of the mechanical properties of the most important matrix materials and
their composites with different reinforcing fibers are presented in Tables 2.8–2.39.

Table 2.8  Properties of a Typical Filled and Unfilled Polypropylene (1)

PP 40% PP 40% PP 40% PP 30%
Properties PP Talc CaCO3 Glass Graphite
Tg°C 170 168 168
Heat deflection temperature, 55 100 80 160 120
1.82 MPa °C
Maximum resistance to 100 120 110 135 125
continuous heat °C
Coefficient of linear expansion 9 6 4 3 3
cm/cmxl0-5 °C
Tensile strength, MPa 35 32 26 82 47
% Elongation 150 5 15 2 0.5
Flexural strength MPa 48 60 45 100 62
Compressive strength, MPa 45 52 35 64 55
Notched Izod impact J/m 42 27 42 90 56
Hardness Rockwell R90 R100 R88 R105 R100
Specific gravity 0.90 1.25 1.23 1.22 1.04

Table 2.9  Properties of a Typical PEEK Resin (1)

PEEK 30% PEEK 30%
Properties PEEK Glass Graphite
Tg°C 334 334 334
Heat deflection temperature, 1.82 MPa °C 165 282 282
Maximum resistance to continuous heat °C 150 270 270
Coefficient of linear expansion cm/cmxl0-5 °C 5.5 2.1 1.5
Tensile strength, MPa 100 162 173
% Elongation 40 2 2
Flexural strength MPa 110 255 313
Notched Izod impact J/m 150 110 70
Hardness Rockwell R123 R123 R123
Specific gravity 1.32 1.44 1.32
2  Composite Materials 231

Table 2.10  Properties of a Typical Polyetherimide Resins (1)

PEI 10% PEI 20% PEI 30% PEI 30%
Properties PEI Glass Glass Glass Graphite
Tg°C 216 216 216 216 216
Heat deflection temperature, 195 200 205 210 210
1.82 MPa °C
Maximum resistance to 165 170 175 180 180
continuous heat °C
Coefficient of linear expansion 5 4 3 2 2
cm/cm×10−5 °C
Tensile strength MPa 105 114 138 169 216
% Elongation 7 5 4 3 2
Flexural strength MPa 144 193 205 225 283
Compressive strength MPa 140 155 162 175 220
Notched Izod impact J/m 55 60 85 110 75
Hardness Rochvell M110 M116 M120 M125 M127
Specific gravity 1.3 1.35 1.45 1.5 1.4

Table 2.11  Properties of Typical Polycarbonate Sheets (1)

PC 10% PC 30% PC 40% PC 40% Polyester
Properties PC Glass Glass Glass Graphite Carbonate
T °C g 150 150 150 150 150 160
Heat deflection 139 142 144 144 146 150
temperature,1.82 MPa °C
Maximum resistance to 125 130 130 130 130 135
continuous heat °C
Coefficient of linear 7 2 1 1 1 8
expansion cm/cm×10-5 °C
Tensile strength MPa 65 65 135 165 165 73
% Elongation 110 6 3 3 2 90
Flexural strength MPa 93 105 155 93 240 240
Notched Izod impact J/m 130 110 90 100 90 300
Hardness Rockwell M70 M75 M92 R118 R119 M85
Specific gravity 1.2 1.28 1.4 1.35 1.35 1.2
232 L. Ambrosio et al.

Table 2.12  Properties of Polyphenylene Sulfide Resins (1)

PPS 40%
Properties PPS PPS 40% Glass Graphite
Tg °C 290 290 280
Heat deflection temperature, 1.82 MPa °C 133 260 260
Maximum resistance to continuous heat °C 120 240 240
Coefficient of linear expansion cm/cm×10-5 °C 5 2 1
Tensile strength MPa 65 135 160
% Elongation 2 2 1.5
Flexural strength MPa 95 185 210
Compressive strength MPa 95 160 180
Notched Izod impact J/m 25 80 55
Hardness Rockwell R123 R123 R123
Specific gravity 1.3 1.65 1.45

Table 2.13  Properties of a Heat deflection temperature, 1.82 MPa °C 174

Typical Polyarylate (1)
Maximum resistance to continuous heat °C 150
Coefficient of linear expansion cm/cm×10-5 °C 6.5
Tensile strength MPa 68
% Elongation 50
Flexural strength MPa 74
Compressive strength MPa 93
Notched Izod impact J/m 210
Hardness Rockwell R125
Specific gravity 1.2

Table 2.14  Properties of Typical Polyamide-Imide Plastics (1)

PAI 30 %
Properties PA Glass PAI 30 % Graphite
Tg °C 275 275 275
Heat deflection temperature, 1.82 MPa °C 275 275 275
Maximum resistance to continuous heat °C 260 260 260
Coefficient of linear expansion cm/cm×10-5 °C 3.6 1.8 2.0
Tensile strength MPa 150 195 205
% Elongation 13 6 6
Flexural strength MPa 200 315 315
Compressive strength MPa 258 300 300
Notched Izod impact J/m 135 105 45
Hardness Rockwell E78 E94 E94
Specific gravity 1.39 1.57 1.41
Table 2.15  Properties of Typical Polysulfones (1)
Modified
2  Composite Materials

Polysulfone Polysulfone Polysulfone Polyether SIIlfone PES 20 PES 20 % Modified Polysulfone

Properties (Udel) Gloss 30 % Graphite 30 % (Victrex) % Glass Graphite Polysulfone 30 % Glass
Heat deflection 190 198 190 200 210 210 150 150
temperature, 1.82 MPa
°C
Maximum resistance to 170 175 175 185 200 200 150 150
continuous heat °C
Coefficient of linear 6 2.5 2.5 5.5 2 1 4 5
expansion cm/cmx10-5
°C
Tensile strength MPa 70 100 100 138 127 190 43 115
% Elongation 5 115 115 85 2 115 50 2
Flexural strength MPa 106 200 215 120 175 250 85 175
Compressive strength 176 95 175 95 150 150 125 150
MPa
Notched Izod impact J/m 64 58 64 110 75 75 150 75
Hardness Rockwell M69 M95 M80 M88 M98 R123 R117 M80
Specific gravity 1.25 1.5 1.36 1.4 1.5 1.5 1.35 1.5
233
Table 2.16  Properties of typical Nylons (1)
Nylon-6 Nylon-6 Nylon 66 Nylon 66 Nylon 66 Nylon 66
30 % 30 % 30 % 30 % 40 % 50 %
Properties Nylon 6 glass graphite Nylon 66 glass graphite clay mica
Tg °C 226 215 215 265 265 265 265 215
Heat 78 210 215 75 250 260 190 230
deflection
temperature,
1.82 MPa °C
Maximum 65 190 205 100 225 240 150 170
resistance to
continuous
heat °C
Coefficient of 8 4 5 8 2 2 3 3
linear
expansion
cm/cmx 10-5
°C
Tensile 62 138 205 82 180 227 75 90
strength MPa
% Elongation 30 5 3 60 4 3 9 9
Flexural 96 150 135 103 180 170 160 150
strength MPa
Compressive 55 130 155 55 110 88 50 85
strength MPa
Notched Izod R119 M85 M80 M85 M85 R120 M80 M80
impact J/m
Hardness 1.13 1.38 1.28 1.14 1.37 1.35 1.4 1.4
Rockwell
Specific 96 275 315 103 275 330 205 400
gravity

Table 2.17  Properties of typical Nylons (1)

Nylon Nylon Nylon Nylon 6–12 Nylon Nylon
Properties 69 6–10 6–12 35 % Glass 11 12 Aramid
Tg °C 205 220 210 210 192 177 275
Heat deflection 55 60 69 216 150 146 260
temperature, 1.82 MPa °C
Maximum resistance to 60 70 75 200 140 135 150
continuous heat °C
Coefficient of linear 8 8 8 6 10 8 3
expansion cm/cmx10-5 °C
Tensile strength MPa 58 60 50 145 55 55 120
% Elongation 80 125 200 4 200 225 5
Flexural strength MPa 40 40 44 80 40 42 172
Compressive strength MPa 100 90 90 150 80 80 207
Notched Izod impact J/m 60 60 60 96 96 110 75
Hardness Rockwell R111 R105 M78 M93 R108 R105 E90
Specific gravity 1.09 1.08 1.08 1.35 1.04 1.01 1.2
Table 2.18  Typical properties of Discontinuous Graphite-Fiber-Reinforced Thermoplastic
Composites (8)
Nylon Polysulfone Polyester Polyphenilene ETFE 30
Property 66 30 % 30 % 30% sulfide 30%C %C
Specific gravity 1.28 1.32 1.47 1.45 1.73
Water absorption (24 hr), % 0.5 0.15 0.04 0.004 0.015
Equilibrium, % 2.4 0.38 0.23 0.1 0.24
Mold shrinkage, % 1.5–2.5 2–3 1–2 1 1.5–2.5
Tensile strength, ksi (MPa) 35 19 20 27 15
(241) (131) (138) (186) (103)
Tensile elongation, % 3–4 2–3 2–3 2–3 2–3
Flexural strength, ksi (MPa) 51 25.5 29 34 20
(351) (176) (200) (234) (138)
Flexural modulus of 2.9 2.05 2.0 2.45 1.65
elasticity, Msi (GPa) (20) (14.1) (13.8) (16.9) (11.4)
Shear strength ft-Ib/in (J/m) 13 7 – – 7
(89.6) (48.1) (48.2)
Izod impact strength ft-Ib/in 1.5 1.1 1.2 1.1 4–5
(J/m) (80.1) (58.7) (64.1) (58.7) 213–267)
Thermal deflection 495 365 430 500 465
temperature at 264 psi (1.82 (257) (185) (221) (260) (241)
MPa), °F
Coefficient of linear thermal 1.05 0.7 0.5 0.6 0.8
expansion, in/in., °Fx1O-5 (1.89) (1.26) (0.9) (1.08) (1.44)
(m/m/°CxlO · 5)
Thermal conductivity, 7.0 5.5 6.5 5.2 5.6
BTU-in/hr ft2 °F (W/m°C) (12.1) (9.5) (11.2) (9.0) (9.7)
Surface resistivity, Ω/m2 3–5 1–3 2–4 1–3 3–5

Table 2.19  Typical properties of Discontinuous Graphite-Fiber-Reinforced Thermoplastic

Composites (8)
Polypropilene Polycarbonate VF2-TFE
Property 30% C 30% C 20% C
Specific gravity 1.06 1.36 1.77
Water absorption (24 hr), % — — 0.03
Equilibrium, % — — —
Mold shrinkage, % 5 2 2.5–3.5
Tensile strength, ksi (MPa) 5.4 (37.2) 11.5 (79.2) 12.3 (84.7)
Tensile elongation, % 2.2 2.1 3–4
Flexural strength, ksi (MPa) 6.7 (46.2) 17.1 (118) 17.2 (119)
Flexural modulus of elasticity, Msi 0.60 (4.1) 1.08 (7.44) 1.10 (7.58)
(GPa)
Shear strength ft-lb/in (J/m) — — 7.5 (51.7)
Izod impact strength ft-lb/in (J/m) 0.7 (37.4) 3.0 (160.2) 2.6 (138.8)
Thermal deflection temperature at 264 295 (146) 293 (145) 248 (120)
psi 1.82 MPa), °F
Coefficient of linear thermal 3.0 (5.4) 1.6 (2.8) 1.6 (2.8)
expansion, in./in.,
°F×10-5(m/m/°C×10-5
Thermal conductivity, BTU-in/hr ft2 3.0 (5.7) — —
°F (W/m°C)
Surface resistivity, Ω/m2 3–5 — —
236 L. Ambrosio et al.

Table 2.20  Properties of a Typical Styrene Polymer (1)

Properties PS PS 30% Glass
Heat deflection temperature, 1.82 MPa °C 90 105
Maximum resistance to continuous heat °C 75 95
Coefficient of linear expansion cm/cm×10-5 °C 7.5 4.0
Tensile strength MPa 41 82
% Elongation 1.5 1.0
Flexural strength MPa 83 117
Compressive strength MPa 90 103
Notched Izod impact J/m 21 20
Hardness Rockwell M65 M70
Specific gravity 1.04 1.2

Table 2.21  Properties of Typical Polyester Resins (1)

PET PET PET PBT PBT PBT
30% 45% 30% 30% 45% 30%
Properties PET Glass Glass Glass PBT Glass Glass Glass
Tg°C 255 255 255 245 245 245 245
Heat deflection 220 225 225 85 210 210 215
temperature,
1.82 MPa °C
Maximum resistance 200 210 210 80 200 200 205
to continuous heat
°C
Coefficient of linear 6.5 2 3 6 2.5 2 3
expansion cm/
cm×10-5 °C
Tensile strength MPa 58 165 175 50 220 90 155
% Elongation 100 3 115 100 3 3 2
Flexural strength 110 175 260 100 175 140 215
MPa
Compressive 90 155 105 98 145 105 100
strength MPa
Notched Izod impact 35 75 75 40 50 50 70
J/m
Hardness Rockwell M97 R118 R125 M72 M90 M80 R120
Specific gravity 1.35 1.6 1.4 1.34 1.5 1.7 1.41
2  Composite Materials 237

Table 2.22  Properties of Typical Polyimides (1)

Properties PI PI 40 % graphite
Tg °C 330 365
Heat deflection temperature, 1.82 MPa °C 315 360
Maximum resistance to continuous heat °C 290 810
Coefficient of linear expansion cm/cm×10-5 °C 5 4
Tensile strength MPa 96 44
% Elongation 9 3
Flexural strength MPa 165 145
Compressive strength MPa 240 125
Notched Izod impact J/m 83 38
Hardness Rockwell E70 E27
Specific gravity 1.4 1.65

Table 2.23  Properties of DAP (1)

Properties DAP Fiber glass filled Mineral filled
Heat deflection temperature, 1.82 MPa °C 155 225 200
Maximum resistance to continuous heat °C 100 210 200
Coefficient of linear expansion cm/cm×10-5 °C — 3 3.5
Tensile strength MPa 27.6 50 45
% Elongation 4.6 4 4
Flexural strength MPa 62.0 90 65
Compressive strength MPa 150 205 170
Notched Izod impact J/m 14 50 16
Hardness Rockwell E115 E84 E61
Specific gravity 11.3 1.80 1.75

Table 2.24  Properties Of Typical Amino Plastics (1)

Properties Cellulose-filled UF Fberglass-filled MF
Heat deflection temperature, 1.82 MPa °C 175 200
Maximum resistance to continuous heat °C 100 175
Coefficient of linear expansion cm/cm×10-5 °C 4.0 1.6
Tensile strength MPa 65 50
% Elongation 0.7 0.7
Flexural strength MPa 90 130
Compressive strength MPa 275 310
Notched Izod impact J/m 16 100
Hardness Rockwell M120 M120
Specific gravity 1.50 1.9
238 L. Ambrosio et al.

Table 2.25  Properties of fiberglass Composites with Different Thermosets (1)

Diallyl
Properties phthalate Epoxy Phenolic Polyester Polyamide
Heat deflection temperature 225 150 200 200 350
1.8 MPa °C
Maximum continuous use 210 140 175 160 310
temperature °C
Coefficient of linear expansion 3 2 2 2.5 1.3
cm/cmx10-5 °C
Tensile strength, MPa 50 83 41 70 44
% Elongation 4 4 1.5 1 1
Flexural strength, MPa 90 103 172 172 145
Compressive strength, MPa 205 100 200 200 300
Notched Izod impact J/m 50 25 175 200 300
Hardness Rockwell E84 M105 M110 M50 M118
Specific gravity 1.8 1.9 1.5 2 1.6

Table 2.26  Properties of the Most Common Resin for High Performance Composites
Coefficient
Tensile Flexural Max service of thermal Water
strength modulus Density temperature expansion absorption
Materials (MPa) (MPa) (gcm-3) (°C) (10-50 °C-1) (24 h%)
Epoxy 35–85 15–35 1.38 25–85 8–12 0.1
Polyimide 120 35 1.46 380 9 0.3
PEEK 92 40 1.30 140 6–9 0.1
Polyamide/imide 95 50 1.38 200 6.3 0.3
Polyether/imide 105 35 1.27 200 5.6 0.25
Polyphenylene/ 70 40 1.32 75 9.9 0.2
sulfide
Phenolics 50–55 10–24 1.30 50–175 4.5–11 0.1–0.2

Table 2.27  Properties of Polyester Composites Reinforced by Continuous and Chopped Fiberglass
Tensile Flexural Transverse Transverse
Continuous Chopped Strength Strength Tensile Strength Flexural Strength
Filament (%) Glass (%) MPa MPa MPa MPa
75 0 690 1200 24 35
65 10 660 1135 27 90
45 20 570 980 60 155
25 20 500 810 95 200
15 60 410 680 125 260
Table 2.28  Typical properties of cured polyester resins
2  Composite Materials

Cast resin properties Laminate properties

Flexural Tensile Tensile
strength strength Modulus Flexural Tensile Tensile
Resin (MPa) (GPa) (GPa) % Elongation HDT(˚C) % glass Strength(MPa) Strength(MPa) modulus(GPa)
Orthophthalic 100 65–75 3.2 2.0–4.0 55–110 30 150 90 7
Isophthalic 140 70–75 3.5 3.5 75–130 30 230 120 8
Neo-pentylglycol 130 70 3.4 2.4 110 30 170 90 7
Isophthalicl/ 130 60 3.4 2.5 90–115 30 160 90 7
neopentylglycol
HET acid 80 40–50 3.2 1.3–4.0 55–80 30 150 85 7
Isophthalic/HET 85 55 3.2 2.9 70 30 150 90 7
acid
Bisphenol A 130 60–75 3.2 2.5–4.0 120–136 30 170 90 7
Chlorinated 110 50–60 3.4 1.2–4.8 55–80 30 140 90 7
paraffin
Isophthalic/ 90 60 2.0 4.8 50 30 140 100 7
chlorinated paraffin
239
240 L. Ambrosio et al.

Table 2.29  Characteristics of Epoxy resins (8)

Resin Description Characteristics and end uses
Epocryl® A neat dimethacrylate ester of a Nominal 1 000 000 cp (1 kPa sec)
Resin 12 low-molecular weight viscosity designed for molding,
bisphenol A epoxy resin adhesives, and electrical prepreg
Epocryl® A neat diacrylate ester of a Base resin for fonnulation of
Resin 370 low-molecular-weight bisphenol UV-cure inks and coating
A expoxy resin
CoRezyn VE-8100 A dimethacrylate ester of an Nominal 100 cP (5 dPa sec)
Derakane® 411-C-50 intermediate-molecular- weight viscosity designed for chemical-
Epocryl® bisphenol A epoxy resin resistant FRP applications: hand
Resin 321 containing 50 wt% styrene lay-up and filament winding
Corrolite 31–345 A dimethacrylate ester of a Nominal 500 cP (5 dPa sec)
CoRezyn VE-8300 high-molecular-weight bisphenol viscosity designed for Chemical-
Derakane® 411–45 A epoxy resin containing 45 wt% resistant FRP applications: hand
Epocryl® Resin styrene lay-up and filament winding
Derakane® 470–36 A methacrylate ester of a Nominal 200 cP (2 dPa sec)
phenolicnovolac epoxy resin viscosity designed for solvent
containing 36 wt% styrene resistance and high-service-
temperature FRP applications
Epocryl® Resin A methacrylate ester of a Nominal 500 cP (5 dPa sec)
low-molecular-weight bisphenol viscosity designed for solvent
A epoxy resin containing 40 wt resistance and high-service-
% styrene temperature FRP applications.
Exhibits high tensile strengths and
elongation with high modulus.
Designed for filament winding and
hand lay-up applications
Derakane® 51OA40 A dimethacrylate ester of a Nominal 350 cP (3.5 dPa sec)
brominated bisphenol A epoxy viscosity designed to impart fire
resin containing 40 wt% styrene retardancy for chemical-resistant
FRP applications

Table 2.30  Shear Properties of Composites of Kevlar 49 Fiber in Epoxy Resins (8)
Shear Strain Secant Shear
Cure Cycle Shear Failure at Failure Modulus at 0.5%
Epoxy System hours/°C Stress MPa Stress % Shear Strain MPa
(weight ratio (hours/°F) (CV) (CV) (CV)
XD 7818/ERL 2.5/80 + 2/160 21.4 (2.6) 1.35 (2.2) 1884 (3.9)
4206/Tonox 60–40 (2.5/176 + 2/320)
DER 332/ Jeffamine 24/60 29.4 (2.0) 173 (2.3) 1923 (4.7)
T-403 (100/39) (24/140)
ERL 2256/Tonox 16/50 + 2/120 23.0 (8.6) 1.49 (2.2) 1775 (0.9)
60–40 (100/25/28.3) (16/122 + 2/248)
Epon 826/RD2/ 3/60 + 2/120 23.4 (6.3) 1.91 (6.5) 1520 (3.9)
Tonox 60–40 (3/140 + 2/248)
(100/25/28.3)
XB 2793/Tonox 2/90 + 2/160 21.9 (0.3) 1.69 (2.9) 1600
60–40 (100/25.6) (2/194 + 2.320)
(continued)
Table 2.30 (continued)
Shear Strain Secant Shear
Cure Cycle Shear Failure at Failure Modulus at 0.5%
Epoxy System hours/°C Stress MPa Stress % Shear Strain MPa
(weight ratio (hours/°F) (CV) (CV) (CV)
XD 7818/XD 5/80 + 3/120 39.7 (0.9) 2.43 (2.5) 1852 (1.7)
7575.02/XD (5/176 + 3/248)
7114/Tonox 60/DAP
(50/50/45/14.1/14.1)
XD 7818/XD 5/60 + 3/120 31.9 (3.4) 1.91 (4.5) 1850 (1.5)
7114/Tonox LC
(100/45/50.3)

Table 2.31  Properties of a Heat deflection temperature at 1.8 MPa (°C) 96

typical PMMA Sheet* (1)
Maximum resistance to continuous heat (°C) 90
Coefficient of linear expansion cm/cm/°Cx10-5 7.6
Tensile strength (MPa) 72
Percent elongation 5
Flexural strength (MPa) 110
Compressive strength (MPa) 124
Notched Izod impact(J/m) 74
Hardness Rockwell M93
Specific gravity 1.19
*Polymethyl methocrylate (PMMA) is largely used for bio-
medical applications, optical fibers and cultured marble.

Table 2.32  Properties of Typical Polyftuorocarbons (1)

Properties PFTE PCTFE PVDF PVF
Heat deflection temperature 1.8 MPa (°C) 100 100 80 90
Maximum resistance to continuous heat (°C) 250 200 150 125
Coefficient of linear expansion cm/cm °C x 10-5 10 14 8.5 10
Tensile strength, MPa 24 34 55 65
Flexural strength MPa 50 60 75 90
Notched Izod impact J/m 160 100 150 100
% Elongation 200 100 200 200
Hardness, Rockwell D52 R80 R110 R83
Specific gravity 2.16 2.1 1.76 1.4

Table 2.33  Properties of typical Filled PTFE (1)

Unfilled 15% 25% 15% 60%
Properties PTFE Glass Glass Graphite Bronze
Thermal conductivity mW/MK 0.244 0.37 0.45 0.45 0.46
Tensile strength MPa 28 25 17.5 21 14
% Elongation 350 300 250 250 150
Notched Izod impact J/m 152 146 119 100 75
Coefficient of friction, 3.4 MPa load 0.08 0.13 0.13 0.10 0.10
Wear factor 1/pPa 5013 280 26 102 12
Shore durometer hardness 51D 54D 57D 61D 70D
Specific gravity 2.18 2.21 2.24 2.16 3.74
242 L. Ambrosio et al.

Table 2.34  Properties of Glass transition temperature (°C) -25

Typical LDPE Plastics (1)
Coefficient of linear expansion cm/cm°Cxl0-5 15
Tensile strength MPa 20
%Elongation 350
Shore hardness 47D
Specific gravity 0.925

Table 2.35  Properties of a typical Filled and Unfilled HDPE (1)

30% Glass-Filled
Properties HDPE HDPE
Melting point (°C) 130 140
Heat deflection temp. at 1.82 MPa (°C) 40 120
Maximum resistance to continuous heat (°C) 40 110
Coefficient of linear expansion cm/cm °Cxl0-5 10 5
Tensile strength MPa 27 62
% Elongation 100 1.5
Flexural strength MPa — 76
Compressive strength MPa 21 43
Notched Izod impact J/m 133 64
Hardness Rockwell D40 R75
Specific gravity 0.95 1.3

Table 2.36  Thermal and Property Value

Electrical Properties of
Thermal conductivity (46 volume % fiber)
Kevlar 49 Fabric/Epoxy
Composites (8) Across fabric layers, W/m °K 0.22
Parallel to warp, W/m °K 0.91
Thermal coefficient of expansion (20–100°C) 0
Dielectric constant (58 volume % fiber)
Perpendicular at 9.3x109 Hz (room temperature) 3.3
Parallel at 9.3x109 Hz (room temperature) 3.7
Perpendicular (48 volume % fiber) at 106 Hz 4.1
Dielectric strength (48 volume % fiber), V/mm 24.4
(V/mil)
Volume resistivity (48 volume % fiber), 1 2 -cm 5x1015

Surface resistivity (48 volume % fiber), 1 2 cm 5x1015

Arc resistance (48 volume % fiber), seconds 125

Table 2.37  Property of Unidirectional Thornel 300-Kevlar 49/Epoxy Hybrid Composites (8)
Tension Compression Flexure
Ratio Stress at Stress at Stress at Stress at Short- Short-
Thornel Ultimate Ultimate 0.02% 0.02% Ultimate Ultimate 0.02% 0.02% Ultimate Ultimate beam- beam
to Specific Modulus Modulus stress stress offset offset stress stress offset offset stress stress shear shear
Kevlar gravity GPa (Msi) MPa (ksi) (ksi) (ksi) MPa (ksi) MPa (ksi) sMPa (ksi) stress stress
100/0 1.60 145 (21.1) 1565 (227) 678 (98.4) 1007 (146) 1605 (223) 1606 (233) 91 (13.2)*
75/25 1.56 120 (17.4) 1281 (186) 469 (68.8) 938 (136) 1248 (181) 1358 (197) 76 (11.0)
50/50 1.51 108 (15.7) 1213 (176) 413 (59.9) 688 (99.8) 827 (120) 1103 (160) 56 (8.1)
0/100 1.35 77 (11.2) 1262 (183) 182 (26.4) 286 (41.5) 339 (49.2) 634 (91.9) 49 (7.1)
244 L. Ambrosio et al.

Table 2.38  Properties of Epoxy Resin Composites with Different Reinforcing Fibers
Aramid
Properties E-Glass S-Glass Kevlar 49 Graphit Boron
Thermal conductivity W/mK 0.9 1.1 0.9 5 1
Linear expansion cm/cm°C×10-5 1.2 1.1 1 2 1
Tensile strength MPa 450 700 800 700 1600
Elastic modulus MPa 24 000 30 000 33 000 60 000 207 000

Fracture toughness MPa m 1 2 22 25 34 18 35

Specific gravity 2.1 2.0 1.4 1.6 2.1

Table 2.39  Properties of Fiberglass-reinforced Polyester Composites with Different Fabrication

Techniqes (1)
Filament
Preform Cold-­ Spray-­ wound
Properties SMC BMC mat molding press up (epoxy) Pultruded
Glass content % 22 25 30 25 40 55 60
Heat deflection 225 225 205 190 190 190 175
temp.
Maximum 180 175 185 180 185 150 200
resistance to
continuous heat °C
Coefficient of linear 1.0 1.0 1.4 1.4 1.6 4 5
expansion °Cxl0-5
Tensile strength 90 48 110 110 95 120 80
KPa
% Elongation 1 0.5 1.5 1.5 1.0 2.0 2.0
Flexural strength 165 100 220 190 150 1250 1000
KPa
Compressive 80 30 150 125 135 400 340
strength MPa
Notched Izod 640 240 800 560 425 2660 2750
impact J/m
Hardness Rockwell H75 H95 H70 H70 H70 M110 H96
Specific gravity 1.9 1.9 1.9 1.6 1.5 1.9 1.8

2.14  Antioxidants and Effect of Environmental Exposure

Many of the polymeric matrices will require some type of antioxidants to improve
aging properties. The most common primary antioxidants are hindered phenols
such as butylated hydroxytoluene (BHT). Typical low toxic antioxidants are
reported in Tables 2.40–2.42.
The toxicity of commonly used polymer stabilizers and additives are classified in
the following table.
2  Composite Materials 245

Table 2.40 Typical Primary antioxidants Butylated hydroxy toluene (BHT)

antioxidants of low
(usually hindered
toxicity (3)
phenols)
Thioester antioxidants Dilaryl thiodipropionate (DLTDP)
(usually derivates of Distearyl thiodipropionate (DSTDP)
thiodipropionic acid)
Phosphite Distearyl pentaerythritol diphosphite
antioxidants
(usually derivates of Tris (nonylphenyl) phosphite
aromatic
phosphites)

Table 2.41  Toxicity of commonly used polymer stabilizers and additives (3)
Stabilizer Toxicity
Fatty acid derivates of Low toxicity. Used in non-toxic medical
calcium, zinc applications
and magnesium
Barium fatty acid Moderately toxic
compounds
Lead & cadmium Highly toxic. Not recommended in the US for use in medical
derivates applications (cadmium pigments considered of low toxicity in
England)
Amines (antioxidants) Generally toxic, with aromatic amines showing carcinogenic
tendencies. Newer types less toxic
Butylated hydroxy toluene Considered non-toxic as also used in foods. Recently investigated
(BHT) and found non-carcinogenic
(antioxidant)
Octyl tin compounds Only class of tin compound classified as of low toxicity and used
in medical applications e.g. di-(n-octyl)tin maleate polymer

Table 2.42  Toxicity data on some common plasticizers used in plastic manufacture (3)
Plasticizer Toxicity
Adipates To date animal experiments indicate possible carcinogenicity
Glycolates Generally of low toxicity levels. However studies underway as
commercial form are phthalyl derivates
Phosphates Generally cause irritation to skin and mucous membranes
Phthalates Although commercially used in medical devices, environmental
effects and toxicological properties continually under investigation
Epoxidized soya bean oil Chelating type of plasticizer with low toxicity

2.15  The Radiation Stability of Commercial Materials

The radiation resistance of common polymeric materials used as matrix for com-
posite are shown in Tables 2.43–2.52. Generally, polystyrene and urethane rubber
have the most resistance.
246 L. Ambrosio et al.

Table 2.43  The radiation resistance of common materials used as matrix of polymeric composites (3)
Material Stability effect
ABS Stable for single dose of 2.5 Mrad
Polyamides Suitable for single doses of 2.5 Mrad level
Polyethylene Stable under ordinary conditions at 2.5 Mrad
Polypropylene Embrittles – newer variations more resistant
Poly(vinyl chloride) Withstands single dose radiation cycle – but discolors – some HCl
liberated
Polystyrene Most radiation – stable of common polymers
Poly(tetrafluoroethylene) Poor resistance to radiation – copolymers less affected
Polysulfone Stable under ordinary conditions at 2.5 Mrad
Polyacetals Embrittles – discolors
Polyurethane Stable under ordinary conditions at 2.5 Mrad
Polymethylmethacrylate Embrittles – discolors
Rubber natural Stable under ordinary conditions when properly compounded
Rubber butyl Poor stability at low radiation levels
Rubber silicones Stable under ordinary conditions at 2.5 Mrad
Urethanes Excellent radiation resistance

Table 2.44  The radiation resistance of common materials used as matris of polymeric composites (3)
Material Thermoplastics Stability
Acrylonitrile/Butadiene/Styrene(ABS) Good
Cellulosics esters more stable Fair Undergoes chain scission,
Fluoinated ethylene propylene (FEP) Fair Copolymers more resistant than
Homopolymer
Polyacetal Poor Embrittles
Polyamides aromatic Excellent
Polyamides aliphatic Fair Hardens as levels increased
Polycarbonates Good Yellow – mechanical
properties unchanged
Polyesters (aromatic) Good
Polyethylene Good lowers melt flow
Polymethylmethacrylate Poor Degrades-turns brown
Polyphenylene sulfide Good
Polyproplyene Fair Improved stability if properly
stabilized
Polysulfone Excellent Yellow natural color
Polystyrene Excellent
Polytetrafluoroethylene Poor Acid evolved
Polyvinylchloride homopolymer Good If properly stabilized
Polyvinylchloride Copolymer Fair HCl evolved – turns brown
Styrene/ Acrylonitrile (SAN) Good More resistant than ABS
2  Composite Materials 247

Table 2.45  The radiation resistance of common materials used as matrix of polymeric composites (3)
Material Thermosetting resin Stability
Epoxies Excellent Very stable with the use of aromatic curing
agen
Phenol or urea formaldehyde Good
Polyimides Excellent
Polyesters Good
Polyurethanes Excellent

Table 2.46  The radiation Material elastomers Stability

resistance of common
Polyisobutylene (butyl) Poor
materials used as matrix of
polymeric composites (3) Natural Good
Urethanes Excellent
Nitrile Good
Polyacrylic Poor
Styrene-butadiene Good
EPDM Good
Silicones Good

Table 2.47  The radiation Textiles Stability

resistance of polymers (3)
Polyesters Excellent
Cellulosics Poor
Nylon Fair
248 L. Ambrosio et al.

Table 2.48  Synergism between a UV absorber and a thermal antioxidant (6)

Time to embrittlement of
low-density polyethylene (in
Additive (0.4 pph) UV stabilizer Antioxidant hours)
None None 400
Octylphenyl salicylate None 1600
None Tri (nonylphenyl) 1800
phosphite
Octylphenyl salicylate Tri (nonylphenyl) 7000
phosphite
2-Hydroxy-4-n- None 2000
octoxybenzophenone
None Tri (nonylphenyl) 1000
phosphite
2-Hydroxy-4-n- Tri(nonylphenyl) 8500
octoxybenzophenone phosphite

Table 2.49 Stability Hours to react with

of hydrocarbon polymers Material 10 cc of O2 at 140°C
with bound phenolic
Low-density polyethylene:
antioxidants (6)
Uninhibited 3
Reacted with diazooxide 14
High-density Polyethylene:
Commercially stabilized 175
Reacted with diazooxide 411
Polypropylene
Uninhibited <1
Reacted with diazooxide 31

Table 2.50  Additives incorporated into natural rubber and as bound antioxidants (6)
Hours to absorb 1 % by wt. of O2
Additive Before extraction After extraction
N,N’-Diethyl-p-nitrosoaniline 39 30
p-Nitrosodiphenylaniline 60 53
p-Nitrosophenol 31 30
2,6-Diter-butyl-p-cresol 47 4
Table 2.51  Summary of effects of moisture and ambient aging on epoxy composites (7)
Flexural strength, MN/m2(ksi)
2  Composite Materials

Temperature 24-h H2O 6-week 20-week 52-week

Orient K(°F) Control boil Retention % humidity Retention ambient Retention % ambient Retention %
B/5505 boron/epoxy
[0] 297 2070 1950 94 2120 103 2180 105 2280 110
(75) (300) (283) (308) (316) (330)
[0] 450 1730 393 23 683 39 1030 59 910 53
(350) (251) (57) (99) (149) (132)
[0±45] 450 862 545 63 641 74 882 102 807 94
(350) (125) (79) (93) (128) (117)
A-S/3501 graphite/epoxy
[0] 297 1680 1680 100 1680 100 1850 110 1620 96
(75) (244) (244) (244) (268) (235)
[0] 450 1300 434 34 386 30 703 54 593 46
(350) (188) (244) (56) (102) (86)
[0±45] 450 676 365 54 276 41 545 81 386 57
(350) (98) (53) (40) (79) (56)
249
250 L. Ambrosio et al.

Table 2.52  Summary of thermal aging of epoxy and polyimide system (7)
Aging Test
Material temperature temperature Aging Retention of
system Orientation K(°F) K(°F) time, h tensile strength, %
B/E [O] 394 (250) 450 (350) 1000 99
[CP] 94
[O] 450 (350) 450 (350) 1000 100
[CP] 100
G/E [O] 394 (250) 450 (350) 1000 94
[CP] 100
[0] 450 (350) 450 (350) 1000 100
[CP] 100
G/PI [O] 505 (450) 505 (450) 1000 98
[CP] 92
[O] 561 (550) 561 (550) 1000 87
[CP] 100

2.16  Polymers Aging

2.17  Composite Materials in Medicine

In recent years, carbon fiber has been recognized as a material with many exciting
applications in medicine. Several commercial products utilize carbon fiber as a rein-
forcing material which serves to enhance the mechanical properties of the poly-
meric resin systems in which it is included. The attractive feature of carbon
reinforced polymer for this application is that the orientation and fiber content can
be varied in the implant to provide the mechanical property orientation necessary
for good function. The carbon fiber can be distributed in matrix material to provide
strength in only those locations and directions where it is needed. The implant must
be designed in a way that fatigue failure does not occur and the matrix material is
not attacked by the physiological environment. The matrix materials used are listed
in Table 2.53.
These polymers were combined with unsized carbon fiber, into ±15° laminated,
test specimens approximately 2.5 cm×7.5 cm×0.3 cm. Testing was performed in
three point bending giving the results in Table 2.55.
Because a composite hip will be subjected to the physiological environment in
use, an accelerated test was performed to evaluate changes in properties. In this test,
the samples were immersed in 0.9 % saline solution maintained at 90°C for one
week; the results are shown in Table 2.56.
Blood compatible materials are essential to circulatory support devices.
Numerous materials have been considered for use in prosthetic devices.
2  Composite Materials 251

Table 2.53  Polymeric materials used as matrix for carbon fibers composites (3)
Polymer Polymer type Commercial name and manufacturer
Polysulfone Thermoplastic UDEL MG-11, Union Carbide, Dallas, TX
Poly-methyl methacrylate Thermoplastic PMMA I.V. 0.4, Rohm & Haas, Philadelphia,
PA
Epoxy (low viscosity) Thermoset Stycast 1267
Epoxy (high viscosity) Thermoset C-8W795 & H.R. 795, Hysol Corp.,
Los Angeles, CA

Table 2.54  Mechanical properties of carbon fibers used in carbon prosthesis

Fibre- Density Diameter Tensile Elastic Strain to α(/10-6
type (g/cm3) (μm ) strength (MPa) modulus (GPa) failure (%) K)
T300 1.75 7 3430 230 1.5 -1.5
HM 35 1.79 6.7 2350 358 0.6 -0.5

Table 2.55 Typical Polymer Ultimate strength (MPa) Modulus (GPa)

Mechanical Properties
PMMA 772 55
of Polymer-Carbon
Composites (3) Polysulfone 938 76
Epoxy Stycast 535 30
Epoxy Hysol 207 24

Table 2.56  Accelerated Test Data (3)

Strength (MPa) Modulus (GPa)
Matrix Before After Before After
Polysulfone 807 723 77 67
PMMA 687 594 76 73
Epoxy (Stycast) 535 323 30 21

2.17.1  Carbons In Heart Valve Prostheses

Carbons are widely used in prosthetic heart valves, as a result of their favorable
mechanical and biological properties. Pyrolytic carbons, deposited in a fluidized
bed, have high strength, and high fatigue and wear resistance. Compatibility with
blood and soft tissue is good.
252 L. Ambrosio et al.

Table 2.57  Representative Mechanical Properties of LTI, Glassy, and VaporDeposited Carbons (3)
Glassy Vapor deposited LTI carbon with
Property carbon carbon LTI carbon silicon (5–12%)
Density, g/cm3 1.5 1.9 1.9 2.1
Crystallite size, Å 30 10 35 35
Flexural strength 1000 psi 20 80 70 85
Young’s modulus, 106 psi 4.0 2.5 3.0 4.0
Strain to fracture, % 1.0 5.0 2.0 2.0
Fatigue limit/fracture strength 1.0 1.0 1.0 1.0
Strain energy to fracture, 1 12 7 9
100 psi

Table 2.58  Mechanical Properties of Suture Materials (3)

Suture Yield Breaking Yield Breaking Modulus of Specific work
stress stress strain strain (%) elasticity of rupture
(GPD) (GPD) (%) (N/Tex)×10-2
Dexon 0.80 6.30 1.9 22.6 55 6.63
Vicryl 0.97 6.55 1.8 18.4 67.5 5.46
Mersilene 1.20 4.20 2.7 8 53 1.32
Silk 1.33 3.43 1.9 11.5 79.0 2.36
Nurolon 0.34 3.80 1.6 18.2 21.0 2.80
Ethilon 0.41 6.25 2.2 33 20.0 8.96
Prolene 0.52 5.14 1.2 42 58.5 14.69

2.17.2  Wound Closure Biomaterials

Virtually every operation requires the use of materials to close the wound for subse-
quent successful healing. The material must retain adequate strength during the
critical period of healing; it should also induce minimal tissue reaction that might
interfere with the healing process. The complexity involved in wound healing calls
for different types of wound closure materials.

2.18 Metal Matrix Composites

The metal matrix composites can be described as materials whose microstructure

comprises a continuous metallic phase into which a second phase (ceramic materi-
als) has been artificially introduced during processing, as reinforcement.
2  Composite Materials 253

2.18.1 Matrix Materials

The most common matrices are the low-density metals, such as aluminum and alu-
minum alloys, and magnesium and its alloys. Some work has been carried out on
lead alloys, mainly for bearing applications, and there is interest in the reinforce-
ment, for example, of titanium-, nickel- and iron- base alloys for higher-temperature
performance. However, the problems encountered in achieving the thermodynamic
stability of fibers in intimate contact with metals become more severe as the poten-
tial service temperature is raised, and the bulk of development work at present rests
with the light alloys.

2.18.2 Reinforcements

The principal reinforcements for metal matrices include continuous fibers of car-
bon, boron, aluminum oxide, silica, aluminosilicate compositions and silicon car-
bide. Some ceramic fibers are also available in short staple form, and whiskers of
carbon, silicon carbide and silicon nitride can be obtained commercially in limited
quantities. There is also interest in the use of refractory particles to modify alloy
properties such as wear and abrasion resistance. In this case, particle sizes and vol-
ume fractions are greater than those developed metallurgically in conventional
alloys, and incorporation of the particles into the metal is achieved mechanically
rather than by precipitation as a consequence of heat treatment. Most metal-matrix
composites consist of a dispersed reinforcing phase of fibers, whiskers or particles,
with each reinforcing element ideally separated from the next by a region of metal.
A summary of properties of the most important metal matrix composites is pre-
sented in Tables 2.59–2.69.

Table 2.59  Summary of Mechanical Properties of A13 Aluminum-28 v/o

Thornel- 50 Composite (5)
Property Value
Ultimate tensile strength 730 MN/m2 (106 000 psi) at 20 °C
660 MN/m2 (95 000 psi) at 500 °C
Rule-of-mixtures strength 700 MN/m2 (101 000 psi) at 20 °C
550 MN/m2 (80 000 psi) at 500 °C
Transverse tensile strength ~83 MN/m2 (~12000 psi)
Tensile elastic modulus, E 145 GN/m2 (21.0x106 psi)
Shear modulus, G (calculated) 55 GN/m2 (7.9x106 psi)
Density 2.4 g/cm3 (0.0805 Ib/in.3)
Strength-to-density ratio 2.4x106 cm (1.25x106 in.)
Modulus-to-density ratio 620x106 cm (248xl06 in.)
Poisson’s ratio, m (calculated) 0.306
254 L. Ambrosio et al.

Table 2.60  Summary of Transverse Tensile Strengths of Various Aluminum- Graphite Composite
Systems (5)
Composite Average
Number
Fiber Matrix (MN/m2) (psi) High (psi) Low (psi) of tests
Thornel-50 Al-12Si 26 3777 6500 433 9
Courtaulds 220 Al 42 6117 8690 3760 20
Courtaulds 356 Al 70 10,008 14,600 5500 26
Courtaulds HM Al -10Mg 29.5 4280 4500 3600 5
Whittaker-Morgan 356 Al 50 7300 11,300 4100 5
Whittaker-Morgan 7075 Al 21 3040 5100 400 5

Table 2.61  Uniaxial Tensile Data for Aluminum-Silicon Alloy-Thornel-50 Composite Thermally
Cycled 20 Times from -193 to +500°C (5)
Sample number Ultimate tensile strength (psi) Rule-of-mixture strength (%)
C7 103 000 103
C8 100 000 99
C9 100 000 00
C10 99 000 99
Average 101 000 100

Table 2.62 Transverse Transverse strength

Tensile Strengths for 356
Sample number (MN/m2) (psi)
Aluminum-Courtaulds HM
Graphite Composite (5) 808A 91 13 100
808A 88 12 700
808A 74 10 700
808A 68 9900
808B 79 11 500
837A 67 9700
837A 67.5 9800
837A 76 11 100
837B 67 9700
837B 72 10 400

Table 2.63  Corrosion Behavior of Aluminum–Graphite Composite for 1000hr (5)

356 aluminum 356 aluminum–25 v/o Thornel-50
Environment (23°C) (50°C ) (23°C ) (50°C )
Distilled water Nil Nil 1.2 1.2
3.5% NaCl solution 1.1 4.9 4.7 9.8
2  Composite Materials 255

Table 2.64  Tensile Properties of Al2O3-Whisker-Nickel Composites at 25 and 1000 °C(5)

Type of Test Whisker volume Composite Strength-to-density
composite temperature (°C) fraction (v/o) strength (MN/m2 ratio (106 cm)
Continuous 25 22 1230 1.63
25 51 1050 1.68
25 39 1350 2.0
1000 16 282 0.114
1000 21 495 0.665
1000 21 495 0.67
1000 29 759 1.08
Discontinuous 25 28 621 0.845
25 19 1180 1.52
25 11 938 1.14
1000 17 451 0.542
1000 28 106 0.144
1000 10 269 0.33
1000 20 618 0.80

Table 2.65  Off-Axis Tensile Properties of Ti-6Al-4V-28 v/o SiC (5)

Average Strength (ksi)
Filament orientation Ultimate tensile Proportional Elastic modulus Poisson ‘s
(degrees) strength limit (106 psi) ratio
0 142 117 36 0.275
15 135 117 35 0.277
30 113 104 32 0.346
45 107 75 31 0.346
90 95 53 28 0.250

Table 2.66  Properties of Ti-6Al-4V-50 v/o Borsic Composites (5)

Elastic Elastic
Tensile Failure modulus modulus Coefficient
Temperature Orientation strength strain (106 psi) (106 psi) of expansion
(°F) (degrees) (ksi) (min./in.) Tensile Flexure (10-6/°F)
70 0 140 3340 41.5 34.4 2.50
70 15 100 3220 36.8 33.3 —
70 45 66 4220 31.2 31.8 —
70 90 42 3130 29.8 31.2 3.17
500 0 119 — — 33.2 2.80
700 0 107 — — 32.4 —
850 0 109 — — 31.5 3.17
850 15 86 — — 29.9 —
850 45 53 — — 27.6 —
850 90 35 — — 24.4 3.64
256 L. Ambrosio et al.

Table 2.67  Summary of Mechanical Properties of Magnesium-Graphite Composites (5)

Strength/ Modulus/
Strength Strength E (GN/ Density density density
Composite (psi) (MN/m2) E (10-6 psi) m2) (gm/cm2) (10-6cm) (10-6 cm)
Mg-42v/o 65 000 450 26.6 184 1.77 2.5 1000
Thornel-75
Mg-ZK60A 50 000 345 6.5 45 1.80 1.9 250

Table 2.68  Room-Temperature Properties of Lead-Graphite Composites (5)

Modulus
Strength of elasticity Density Strength/density Modulus/density
Composite lb/in.2 10-6lb/in.2 Ib/in. 3 10-6 in. 10-6in.
Pure lead 2 000 2.0 0.41 0.005 4.9
Lead-base bearing 10 500 4.2 0.35 0.03 12.0
(75Pb-15Sb-l0Sn)
Lead-graphite 41 104 000 29.0 0.270 0.385 107.0
vol% Thornel-75
Fibers
Lead-graphite 35 72 000 17.4 0.28 0.26 62.3
vol% Courtaulds
HM

Table 2.69  Summary of Mechanical Properties of Zinc and Zinc-Graphite Composite (5)
Modulus of
Strength elasticity Density Strength density Modulus/density
System (psi) (10-6 psi) lb/in.3 10-6in. (10-6 in.)
Z-35 v/o 110 900 16.9 0.191 0.58 88.5
Thornel/75
Alloy AG40A 41 000 10.0 0.240 0.17 41.7

2.19 Ceramic Matrix Composites

Composite structures in ceramics have been developed for two major reasons. First,
they provide a means to enhance dramatically the performance of the so-called
functional ceramics; these are systems where electrical, dielectrical, piezoelectric or
sensitizing properties are greatly amplified by appropriate composite design.
Secondly, they are used to avoid or diminish the brittle behaviour of structural
ceramic systems.
A summary of properties of the most important ceramic matrix composites is
presented in Tables 2.70–2.74.
2  Composite Materials 257

Table 2.70  Glass and Glass ceramics suitable as matrices (4)

Maximum use
Matrix type Major constituent Minor constituent temperature
Glass
7740 Boro-silicate B2O3, Si02 Na2O,Al2O3 6000
1723 A1203, MgO, CaO, Si02 B2O3, Bao 7000C
Alumino-silicate
7930 High silica SiO2 B2O3 11500C
Glass ceramics
LAS I Li20, A1203, MgO, Si02 ZnO, ZrO, BaO 1000°C
LAS II Li20, Al2O3, MgO, SiO2,Nb2O5 ZnO, Zr02, BaO 11000C
LAS III Li2O, A1203, MgO, Si02,Nb2O3 ZrO2 12000C
MAS MgO, A1203, Si02 Ba0 12000C
BMAS BaO, MgO, Al2O3, Si02 – 12500C

Table 2.71  Properties of Property Si3N4 C/Si3N4

silicon nitride and carbon/
Bulk density (g/cm3) 3.44 2.7
silicon nitride (4)
Fibre content (vol.%) – 30
Bending strength (MPa) 473±30 454±42
Young’s modulus (GPa) 247±16 188±18
Fracture toughness (MPa m1/2) 3.7±0.7 15.6±1.2
Work of fracture (J/m2) 19.3± 4770±770

Table 2.72  Room temperature strengths of RBSN* and SiC/RBSN (4)

Axial strength (MPa)
Test 0 % Fiber 23 % Fiber 40 % Fiber
Four point bend 107 ± 26 539 ± 48 616 ± 36
Three point bend — 717 ± 80 958 ± 45
Tensile — 352 ± 73 536 ± 20
* RBSN - Reaction bonded silicon nitride.
258 L. Ambrosio et al.

Table 2.73  Properties of brittle fibre/SiC matrix composites (4)

Comp.density Fracture toughness 4 Point bending
Reinforcement Matrix Vol % % Th. (MPa m1/2) strength, MPa
SiC Fibers SiC 45 70–77 – 213–230
39.5 68–75 224–410
SiC Cloth SiC 41.6 75–90 – 419–437
37.9 73–89 187–217
SiC chopped SiC 25.4 51–81 – 90–177
fibers 21.9 71–77 – 50–94
SiC cloth plain SiC 35.5 62–83 1.8–3.6 72–107
weave satin 46.3 65–85 – 71–196
weave 50.2 68–84 – 44–97
SiC chopped SiC 16.8 69–82 – 61–106
fibers 24.3 68–76 – 74–98
25.4 51–81 – 90–177
SiC cloth SiC 41–45 64–90 – 107–476
SiC fibers SiC 39–57 68–77 – 38–410
SiC fibers SiC – >90 >25 320
C fibers SiC – >90 >25 530

Table 2.74  Room temperature of some unreinforced ceramics and ceramic matrix composite (4)
Material Flexural strength MPa Fracture toughness MPa m1/2
Ai2O3 550 4–5.0
SiC whiskers/Al2O3 800 8.7
SiC 500 4.0
SiC fibers/SiC 750 25.0
ZrO2 200 5.0
SiC/ZrO2 450 22.0
Borosilicate glass  60 0.6
SiC fibers/borosilicate glass 830 18.9
Glass ceramic 200 2.0
SiC fibers/glass ceramics 830 17.0
Reaction bonded Si3N4 260 2–3.0
SiC whiskers/reaction bonded 900 20
Si3N4
Hot pressed Si3N4 470 3.7–4.5
SiC whiskers/hot pressed Si3N4 800 56

References

1. Seymour, R.B. (1990) Polymer Composites, VSP, Utrecht, The Netherlands.

2. Weatherhead, R.G. (1980) FRP Technology, Applied Science Publishers, London.
3. Biocompatible Polymers, Metals, and Composites, edited by M.  Szycher, Technomic
Publishing Co. Inc.
2  Composite Materials 259

4. Islam, M.U., Wallance, W. and Kandeil, A.Y. (1985) Artificial Composites for High
Temperature Applications, Noyes Data Corporation, Park Ridge, New Jersey, USA.
5. Kreider, Kenneth G. (1974) Metallic Matrix Composites, Vol. 4, Academic Press, New York
and London.
6. Hawkins, W.L. (1984) Polymer Degradation and Stabilization, Springer-Verlag.
7. Environmental Effects on Advanced Composite Materials, symposium presented at the
seventy-­eighth Annual Meeting ASTM, Philadelphia, 1975.
8. Hill, R. (1964) J. Mech. Phys. Solids, 12, 199.
9. Hermann, J.J. (1970) Proc. Konigl. Nederl. Akad. Weteschappen Amsterdam, B70, 1.
10. Halpin, J.C. and Pagano, N.J. (1964) J. Compos. Mater., 3, 720.
Chapter 3
Thermoplastic Polymers In Biomedical
Applications: Structures, Properties
and Processing

S.H. Teoh, Z.G. Tang, and Garth W. Hastings

3.1 Introduction

In general thermoplastic polymers are made up of long linear chain molecules

which exhibit large scale chain mobility and deformation under shear forces above
their softening temperature. This change is reversible. Above this temperature the
thermal motions of the chain segments are sufficient to overcome inter- and intra-
molecular forces. At room temperature the material is a viscoelastic solid. Their
behaviour is dependent on chain morphology, structure, crystallinity and the types
of additives added (often to aid processing). The materials can easily be processed
into different type of products and are considered to be the most important class of
plastic materials commercially available. The proeessability of this class of plastics
is a key characteristic for developing biomedical applications.
Nine potential biomedical applications areas have been identified (Jones and
Denning, 1988):
1. Membranes in extracorporeal applications such as oxygenators;
2. Bioactive membranes e.g., controlled release delivery systems and artificial
cells;
3. Disposable equipment e.g., blood bags and disposable syringes;
4. Sutures and adhesives including biodegradable and non-biodegradable materials;
5. Cardiovascular devices such as vascular grafts;
6. Reconstructive and orthopaedic implants;
7. Ophthalmic devices such as corneas and contact lens;

S.H. Teoh (*) • G.W. Hastings

Institute of Materials Research & Engineering – IMRE Block S7, Level 3, Room 17B,
National University of Singapore, Singapore, 10 Kent Ridge Crescent, 119260, Singapore
Z.G. Tang
BIOMAT Centre, National University of Singapore, Singapore 119260, Singapore

© Springer Science+Business Media New York 2016 261

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_19
262 S.H. Teoh et al.

8. Dental restorative materials including dentures;

9. Degradable plastic commodity products.
This section focuses on 12 thermoplastic polymers which have found wide appli-
cation in the above. Each part deals with one polymer or one group of polymers of
structural similarity. The content includes the chemical structure, structure-property
relationships, tables of physical, mechanical and thermal properties, and processing
conditions of each candidate thermoplastic. Some properties can be predicted from
the structural characteristics of the polymers.
In general, for a given polymer, higher molecular weight tends to improve
mechanical properties, but the increase in the resistance to flow in the fluid state
makes processing more difficult and costly. A wide range of molecular weight is
generally more appropriate for processing of each polymer. The effect of branching
in otherwise linear molecules is significant. Short but numerous branches irregu-
larly spaced may reduce considerably the ability of portions of linear chains to form
crystal-like domains, and the corresponding polymer will display a lower stiffness,
a good example being the highly branched low-density polyethylene of lower degree
of crystallinity than the less branched high density product. Tacticity of polymer
molecules greatly affects crystallinity and stiffness. For example, commercial poly-
propylene is usually about 90~95% isotactic, and is stiff, highly crystalline and with
a high melting point, whereas the atactic polypropylene is an amorphous somewhat
rubbery material of little value. Within the range of commercial polymers, the
greater the amount of isotactic material the greater the crystallinity and hence
the greater the softening point, stiffness, tensile strength, modulus and hardness.
The inter-molecular and intra-molecular forces also influence the properties of
polymers. The hydrogen bonds or van der Waal’s and other dispersion forces
between adjacent molecules produce a large increase in melting temperature. A high
energy barrier to molecular rotations hinders the ability of molecules to take up the
required conformations to form crystals. Polymethylmethacrylate (PMMA), and
polycarbonate exist in an amorphous state and are completely transparent. The crys-
tallinity is controlled by both structural factors and processing conditions. From the
processing standpoint, the higher the crystallinity, the bigger the shrinkage observed
after product processing. Thermoplastic polymers exist in semi-crystalline and
amorphous states. The ratio of these two states affects material properties strongly
and can be characterized using X-ray analysis and by observing the thermal behav-
iour of the polymers. Amorphous thermoplastics are normally transparent and do
not have a fixed melting temperature like that of the semi-crystalline thermoplas-
tics. They are also less resistant to solvent attack. Semi-crystalline thermoplastics,
because of the presence of crystallites, are more fatigue and wear resistant. A typi-
cal case is polyacetal which has more than 20 years of in vivo experience as an
occluder in the Björk–Shiley tilting disc mechanical heart valves.
A main requirement for a polymeric candidate is its biocompatibility with bio-
logical tissues and fluids. Biocompatibility will depend on the polymer intrinsic
chemical nature and the additives present. It is a complex issue not dealt with here.
It is not always possible to distinguish the medical-grade polymers from the
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 263

conventional polymers. They may come from a batch intended for general purposes,
but are selected on the basis of clean condition or trace element analysis or
mechanical properties. Subsequent processing requires clean room conditions and
care to avoid any contamination. There is still some inherent uncertainty about con-
stituents unless there has been complete disclosure and/or only a ‘pure’ polymer is
used. With new developments in polymeric biomaterials, the situation should
improve.
It is hoped that the following sections will be of value to researchers in science
and engineering and to clinical practitioners who are engaged in the development
and material selection of new thermoplastic polymers for biomedical applications.

3.2 Polyethylene

Commercially, polyethylene is produced from ethylene in various densities (from

linear low to ultra high). There are four quite distinct processes to the preparation
of commercial polymers of ethylene: (a) high pressure processes, (b) Ziegler pro-
cesses, (c) the Phillips process, (d) the Standard Oil (Indiana) process. High pres-
sure polymers (British patent 471590, 1930) are of the lower density range for
polyethylene (0.915–0.94 g/cm3) and usually also of the lower range of molecular
weights. Until the mid- 1950s, all commercial polyethylenes were produced by
high pressure processes. These materials were branched materials and of moderate
number average molecular weight, generally less than 50 000. Ziegler polymers
(Ziegler, 1955) are intermediate in density (ca. 0.945 g/cm3) between the high pres-
sure polyethylenes and those produced by the Phillips and Standard Oil processes.
Phillips polymers have the highest density of any commercial polythylenes (ca.
0.96 g/cm3) Ziegler and Phillips processes produce polymers at lower temperature
and pressures with a modified structure giving a higher density, greater hardness
higher softening points. The Standard Oil process gives a density of about 0.96 g/
cm3) similar to the Phillips materials. Processes, such as a gas phase process devel-
oped by Union Carbide for making linear low density polyethylene (LLDPE), were
aimed to produce polyethylenes with short chain branch but no long chain branch.
High pressure polymers have more branching and even with side chains as long as
the main chain. In contrast, the high density polyethylene (HDPE) produced by the
Ziegler or Phillips methods has only 3 to 5 branches per 1000 C-atoms and the
linear low density PE has very few branches. The weak point in the chain which is
sensitive to degradative environmental effects is located at the branching site. The
amount of branches in polyethylene also influences the crystallinity of polyethyl-
ene. A higher degree of crystallinity and associated denser packing leads to higher
density and larger crystals (LDPE< 1 μm; LLDPE, 2–4 μm; HDPE, 2–8 μm). The
crystallinity is increased with slower cooling rate. Only the HDPE and the ultra
high molecular weight polyethylene (UHMWPE) find extensive medical
applications.
264 S.H. Teoh et al.

Chemically, polyethylene is inert and there are no effective solvents at room

temperature. However, polyethylene is subject to oxidation and halogenation.
Chemicals like nitric acid produce oxidative deterioration and affect mechanical
properties of polyethylene. The environmental oxidation of polyethylene happens at
high temperature, under ultra-violet light and/or high energy irradiation, e.g.,
gamma irradiation. Polyethylene should be kept from contact with halogenating
agents and environments. The lower molecular weight polyethylene may be dis-
solved at high temperature and swollen by chemicals such as benzene and xylene.
The resistance to environmental stress cracking (ESC) increases with molecular
weight, (copolymers being more resistant than homopolymers).

3.3 Polypropylene

Polypropylene is an addition polymer of propylene. The chemical structure of poly-

propylene is often described as the repeating unit of 2-methyl ethylene. During
polymerization, the CH3 groups characteristic of this olefin can be incorporated
spatially into the macromolecule in different ways. The resulting products have dif-
ferent properties and are classified as a. isotactic polypropylene, where the CH3
groups are on the same side of the main chain; b. syndiotactic polypropylene, where
the CH3 groups are symmetrically arranged on the two sides of the main chain; c.
atactic polypropylene, where the CH3 groups are randomly distributed in the spatial
relationship to the main chain.
The atactic polymer is an amorphous somewhat rubbery material of little value,
whereas the isotactic polymer stiff, highly crystalline and with high melting point.
Commercial polymers are usually about 90–95% isotactic. Within the range of
commercial polymers, the greater the amount of isotactic material, the greater the
crystallinity, and hence the greater the softening point, stiffness, tensile strength,
modulus, and hardness. The properties of the polymer will depend on the size and
type of crystal structure formed in its construction.
Molecular weights of polypropylenes are in the range Mn = 38 000–60 000 and Mw
= 220 000–700 000 with the values of Mw/Mn = 5.6–11.9, higher than those encoun-
tered normally in polyethylene. The high molecular weight polymer from propylene
was introduced in 1954 by G. Natta using a modified Ziegler process, and commer-
cialized in 1957 by Montecatini under a trade name Moplen. This was followed in
1983 by the Spheripol process; in 1988, the Valtec process; and in 1990, Himont
process. The greatest influence of molecular weight and molecular weight distribu-
tion is on the rheological properties. Rheological investigations show that polypro-
pylene deviates more strongly from Newtonian behaviour than does polyethylene.
The effect of shear rate on the apparent melt viscosity is greater for polypropylene
Although polypropylene and polyethylene are similar structurally,. polypropyl-
ene has a lower density around 0.90 g/cm3 and a higher Tg and Tm. The higher melt-
ing point of polypropylene gives the option for autoclave sterilization. The chemical
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 265

resistance of polypropylene is similar to high density polyethylene, but it is more

susceptible to oxidation, chemical degradation and crosslinking (irradiation, violet
light and other physical means) than polyethylene. Polypropylene is better in creep
resistance and in resisting environmental stress cracking than polyethylene.

3.4 Polyurethane

Polyurethanes are block copolymers containing blocks of low molecular weight

polyethers or polyesters linked together by a urethane group. The variety of linkages
in polymers results from the further reaction of urethane groups with isocyanates
and of isocyanates with amines, water, or carboxylic acids.
Attention in this section will be focussed on the thermoplastic polyurethane
elastomers. These polymers are based on three monomers: (1) an isocyanate
source, (2) a macroglycol or carbonate, and (3) a chain extender, or curing agent.
The isocyanates can be either aromatic or aliphatic. Although the aliphatic based
polyurethanes are more expensive, and inferior in physical properties they do not
show the embrittlement, weakening, and progressive darkening of the aromatic
equivalents.
The final physical and biological properties of the polyurethanes depend prin-
cipally on the type of macroglycol used in the synthesis. The polyether-based
polyurethanes are less sensitive to hydrolysis, and are thus more stable in vivo. The
polycaprolactone-based polyurethanes, due to their quick crystallization, can be
used as solvent-activated, pressure-sensitive adhesives. For medical applications,
the polyether-based polyurethanes, particularly those based on polytetramethylene
ether glycol (PTMEG) have been used. Among chain extenders, there are two
choices: either difunctional or multifunctional monomers. For the production of
linear elastomers, only difunctional chain extenders are used, of these, diols and
diamines are by far the most important. The chain extenders for the thermoplastic
polyurethanes must be linear diols, among which, 1,4-butane diol has been chosen
for medical applications. This chain extender produces thermoplastic polyure-
thanes with high physical properties, excellent processing conditions and clear
polymers.
Polyurethane elastomers are a mixture of crystalline (hard segment) and amor-
phous domains(soft segment), and the hard segments are considered to result from
contributions of the diisocyanate and chain extender components. They significantly
affect mechanical properties, particularly modulus, hardness and tear strength. Soft
segments therefore affect the tensile strength and elongation at yield and break.
Polyurethanes are sensitive to strong acids, strong alkalis, aromatics, alcohols,
hot water, hot moist air and saturated steam. The hydrolytic stability of polyure-
thanes in applications must be considered carefully. However, polyurethanes are
resistant to weak acids, weak alkalis, ozone, oxygen, mineral grease, oils and petro-
leum. There are doubts for the oxidation stability of polytetramethylene ether glycol
266 S.H. Teoh et al.

based polyurethanes. Polycarbonate urethane is a promising substitute with good

oxidation stability.
The thermoplastic polyurethanes are characterized by the following properties:
a. high elongation at break and high flexibility (also at low temperature), b. low
permanent deformation on static and dynamic loading, c. favourable friction and
abrasion performance, d. high damping power, e. high resistance to high energy and
UV radiation, and f. plasticizer free,

3.5 Polytetrafluoroethylene

Polytetrafluoroethylene, PTFE, is the polymerization product of tetrafluoroethylene

discovered in 1938 by R.J. Plunkett of Du Pont. The polymer is linear and free from
any significant amount of branching. The highly compact structure leads to a mol-
ecule of great stiffness and results in a high crystalline melting point and thermal
stability of the polymer.
The weight average molecular weights of commercial PTFE are in the range 400
000 to 9 000 000. The degree of crystallinity of the polymer reaches 94%. The prop-
erties of PTFE moldings are considerably influeneed by the processing conditions
and polymer grades. After processing, cooling conditions determine the crystallin-
ity of the molding. Slow cooling leads to higher crystallinity which affects the phys-
ical properties as well as mechanical and thermal properties.
PTFE is a tough, flexible material of moderate tensile strength with excellent
resistance to heat, chemicals and to the passage of an electric current. The poly-
mer is not wetted by water and absorption is not detectable. The permeability to
gases is low, the water vapour transmission rate being approximately half that of
low density polyethylene and polyethylene terephthalate. It has the lowest coef-
ficient of friction of all solids and the dynamic and static coefficients of friction
are equal, i.e. stick-slip does not occur. Abrasion resistance is low. The thermal
stability of PTFE is excellent up to 300 °C but it is degraded by high energy
radiation.
Phase transition behaviour precludes the use of the conventional molding
methods, and PTFE can be processed by employing a process similar to that of
metallurgical sintering. In 1963, Shinsaburo Oslinge of Sumitomo Industries in
Japan discovered a process for expanding PTFE during extrusion. The e-PTFE has
been considered for fabrication of vascular grafts.
Apart from its good slip and wear characteristics the advantages of PTFE
are: a. almost universal chemical resistance, b. insolubility in all known sol-
vents below 300 °C, c. high thermal stability, d. continuous service temperature
range -270 to 260 °C, e. low adhesion, f. low coefficient of friction, g. outstand-
ing electrical and dielectric properties, h. resistant to stress cracking and weath-
ering, but limited use in structural components because of the low modulus of
elasticity.
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 267

3.6 Polyvinylchloride

Commercial PVC polymers are largely amorphous, slightly branched molecules

with the monomer residues arranged in a head-to-tail sequence. The molecular
weights for most commercial polymers are in the range of Mw = 100 000–200 000,
Mn = 45 000–64 000 although values may be as low as 40 000 and as high as 480
000 for the weight average molecular weight. The ratio of Mn/Mw is usually about 2
for the commercial material although it may increase with the higher molecular
weight grades.
The polarity and strong inter-chain attraction gives a higher hardness and
stiffness than polyethylene. Thus PVC has a higher dielectric constant and power
factor than polyethylene, although at temperatures below the glass transition tem-
perature the power factor is still comparatively low (0.01–0.05 at 60 Hz) because of
the immobility of the dipole. PVC is mainly used in a plasticized form. There are
many materials that are suitable plasticisers for PVC. They have similar solubility
parameters to PVC, i.e., about 19.4 MPa½ and are also weak proton acceptors and
may be incorporated by mixing at elevated temperatures to give mixtures stable at
room temperature.
The release of low molecular weight plasticizer has resulted in polymeric plasti-
cisers being developed, but esters are still widely used and are effective in plasticisa-
tion. (Black, 1992; Brydson, 1982, and Park and Lakes, 1992).
Characteristic properties are:

3.6.1 Unplasticized PVC

a. high mechanical strength, rigidity and hardness, b. low impact strength in unmod-
ified form, c. translucent to transparent (depending on method of manufacture), d.
good electrical properties in the low voltage and low frequency range, and e. high
chemical resistance.

3.6.2 Plasticized PVC

a. flexibility adjustable over a wide range, b. depending on type of plasticiser, tough-

ness very temperature dependent, c. translucent to transparent, d. good electrical
properties in the low voltage and low frequency range, e. chemical resistance is
dependent on the formulation and very dependent on temperature, and f. the poly-
mers contain less than 1 ppm vinyl chloride monomer.
268 S.H. Teoh et al.

3.7 Polyamides

Chemically, the polyamides may be divided into two types: a. those based on
diamines and dibasic acids, and b. those based on amino acids or lactams (Chapman
and Chruma, 1985). Commercial use of nylons is dominated by two products, one
from each type, nylon 66 and nylon 6 from Є-caprolactam.
Aliphatic polyamide is linear and easy to crystallize but crystallinity varies
widely with conditions. Crystalline content may be 50–60% by slow cooling and
10% by fast cooling. High interchain attraction in the crystalline zones and flexibil-
ity in the amorphous zones leads to polymers which are tough above their apparent
glass transition temperatures (Brydson, 1982).
Polyamides have excellent fibre-forming capability due to interchain H-bonding
and a high degree of crystallinity which increases strength in the fibre direction
(Park and Lakes, 1992). Polyamides are hygroscopic and lose strength in vivo. The
amorphous region of polyamide chains is sensitive to the attack of water. The greater
the degree of crystallinity, the less the water absorption and hence the less the effect
of humidity on the properties of the polymer. The reversible absorption of water is
associated with a change in volume and thus of dimensions.
The mechanical properties of moulded polyamide materials depend on molecu-
lar weight, crystallinity and moisture content. In the dry, freshly molded state, all
polyamide grades are hard and brittle. When conditioned they are tough and wear
resistant. High melting points result in good mechanical properties up to tempera-
tures in the region of 120–150 °C.
They are only soluble in a few solvents (formic acid, glacial acetic acid, phenols
and cresols), of similar high solubility parameter. Nylons are of exceptionally good
resistance to hydrocarbons. Esters, alkyl halides, and glycols have little effect on
them. Alcohols can swell the polymers and sometime dissolve some copolymers.
Mineral acids attack the nylons but the rate of attack depends on the type of nylon
and the nature and concentration of the acid. Nitric acid is generally active at all
concentrations. The nylons have very good resistance to alkalis at room temperature.
Resistance to all chemicals is more limited at elevated temperature (Brydson, 1989).
Generally, polyamides are characterized by: a. high strength, stiffness and
hardness, b. high heat distortion temperature, c. high wear resistance, good slip and
dry-running properties, d. good damping capacity, e. good resistance to solvents,
fuels and lubricants, f. non-toxicity, g. good processability, h. aliphatic polyamides
are partially crystalline and thus opaque, and i. moisture content impairs mechanical
properties and affects dimensions of moldings.

3.8 Polyacrylates

Polyacrylates are based on acrylic acid, methacrylic acid, and their esters.
Among them, polymethylmethacrylate (PMMA) and polyhydroxy ethylmethacry-
late (PHEMA) have found wide applications as biomedical materials. The clinical
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 269

history of polyacrylates began when it was unexpectedly discovered that the frag-
ments of PMMA plastic aircraft canopies stayed in the body of the wounded without
any adverse chronic reactions (Jones and Denning, 1988; Park and Lakes, 1992).
In normal conditions, PMMA is a hard transparent material. Molecular weight is
the main property determinant. High molecular weight PMMA can be manufactured
by free radical polymerization (bulk, emulsion, and suspension polymerisation).
Bulk polymerization is used for cast semi-finish products (sheet, profiles and even
tubes), and the cast polymer is distinguished by superior mechanical properties and
high surface finish (Brydson, 1982 and Domininghaus, 1993). Cast material has a
number average molecular weight of about 106 whilst the Tg is about 106°C. The
extensive molecular entanglement prevents melting below its decomposition tem-
perature (approx. 170°C).
An amorphous polymer, PMMA has a solubility parameter of about 18.8 MPa½
and is soluble in a number of solvents with similar solubility parameters. Solvents
include ethyl acetate (δ: 18.6 MPa½), ethylene dichloride (δ: 20.0 MPa½), trichloro-
ethylene (δ: 19 MPa½), chloroform (δ: 19 MPa½), and toluene (δ: 20 MPa½). The
polymer is attacked by mineral acids but is resistant to alkalis, water and most aque-
ous inorganic salt solutions. A number of organic materials although not solvents
may cause crazing and cracking (e.g. aliphatic alcohols).
The characteristic properties of PMMA are, a. high hardness, stiffness and
strength, b. homopolymers are brittle, copolymers are tough, c. scratch-resistant, high
gloss surface capable of being polished, d. water-white transparency, copolymers
exhibit inherent yellowish color, e. high heat distortion temperature, f. good electrical
and dielectric properties, g. resistant to weak acids and alkaline solution as well as to
non-polar solvents, grease, oils and water, h. susceptible to stress cracking, i. flam-
mable, j. good processability and machinability, k. rather low resistance to creep at
temperature only slightly above room temperature, and l. high melt viscosity due to
the high chain stiffness caused by restricted rotations about the C-bonds in the back-
bone chains

3.9 Polyacetal

Polyacetal can be divided into two basic types, acetal homoploymer and acetal
copolymer. Both homopolymer and copolymer are available in a range of molecular
weights (Mn = 20 000–100 000). The homopolymer is a polymer of formaldehyde
with a molecular structure of repeated oxymethylene units (Staudinger, 1932).
Large-scale production of polyformaldehyde, i.e. polyacetal, commenced in 1958 in
the USA (US Patent 2 768 994, 1956) (British patent 770 717, 1957). Delrin (1959)
was the first trade mark for this polymer by Du Pont Company. The copolymers
were introduced by the Celanese Corporation of America, and the first commercial
product named Celcon (1960). One of the major advantages of copolymerization is
to stabilize polyacetal because the homopolymer tends to depolymerize and elimi-
nate formaldehyde. The most important stabilization method is structural modifica-
tion of the polymer by, for example, copolymerization with cyclic ether.
270 S.H. Teoh et al.

As can been seen polyacetal has a very simple structure of a polyether. Unlike
polyethylene, polyacetal has no branching, and its molecules can pack more closely
together than those of polyethylene. The resultant polymer is thus harder and has a
higher melting point than polyethylene (175°C for homopolymer), exhibiting a high
crystallinity (77–85%).
No effective solvents have been found for temperatures below 70°C. Swelling
occurs with solvents of similar solubility parameter (δ: 22.4 MPa½). However, poly-
acetal should be kept away from strong acids, strong alkalis, and oxidizing agents.
Water can not degrade it but may swell it or permeate through it and affect the
dimensions of its products. Prolonged exposure to ultra-violet light will induce sur-
face chalking and reduce the molecular weight of the polymers. Polyacetals, both
homopolymer and copolymer are also radiation sensitive. The radiation damage
threshold is estimated at 0.5 Mrad with 25% damage at 1.1 Mrad (Szycher, 1991).
Generally, the polyacetals have the following characteristics, a. high tensile
strength, shear strength, stiffness, and toughness, b. predictable stress/ strain rela-
tionships, c. predictable dimensional behavior, d. chemical and corrosion resistance,
e. abrasion resistance, f. light weight and good appearance, g. acceptability for food
contact application (most grades), h. ease of processing, and i. competitive costs.
The enormous commercial success of the polyacetals is owed to their very high resis-
tance to creep and fatigue. The acetal resins show superior creep resistance to the nylons.

3.10 Polycarbonate

Polycarbonate, PC, is a linear thermoplastic based on the bis-phenol A dihydroxy

compound. In 1898 Einhorn reacted dihydroxybenzenes with phosgene in solution
in pyridine (Brydson, 1982), and production began in both Germany and USA in
1958. General purpose polycarbonate is a linear polyester of carbonic acid in which
dihydric phenols are linked through carbonate groups, while standard grades are
made from bisphenol A and phosgene (Carhart, 1985).
The rigid molecular back bone of bis-phenol A PC leads to high melting tem-
perature (Tm = 225–250°C) and high glass transition temperature (Tg = 145°C). The
polymer does not show any crystallinity. After annealing polymer between 80 and
130°C there is a small increase in density and hence there must be a decrease in free
volume, and a large drop in impact strength; impact strength may be reduced by
annealing crystallization and aging (Brydson, 1982).
The limited crystallinity contributes to the toughness of PC. Highly crystalline sam-
ples prepared by heating for prolonged periods above their Tg or by precipitation from
solution are quite brittle. Although of good impact strength and creep resistance tensile
strain of 0.75% or more produces cracking or crazing. The refractive index of PC lies
in the range of 1.56 to 1.65 (higher than PMMA and silicone rubber) and the transpar-
ency of 85 to 90% is reached in the region of visible light (Domininghaus, 1993).
The chemical resistance of PC is poor and hydrolysis of aliphatic PC s is more
prominent than that of bis-phenol A PC s. There is resistance to dilute (25%) min-
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 271

eral acids and dilute alkaline solutions other than caustic soda and caustic potash.
Where the resin comes into contact with organophilic hydrolysing agents such as
ammonia and the amines, the benzene rings give little protection and reaction is
quite rapid. The absence of both secondary and tertiary C-H bonds leads to a high
measure of oxidative stability. Oxidation takes place only when thin films are heated
in air to temperatures above 300°C.
Typical properties include: a. low density, b. high strength, stiffness, hardness
and toughness over the range from -150 to + 135°C unreinforced and from -150 to
+145°C when reinforced, c. crystal clear transparency, high refractive index, high
surface gloss, d. can be colored in all important shades, transparent, translucent or
opaque with great depth of color, e. good electrical insulation properties which are
not impaired by moisture, f. high resistance to weathering for wall thicknesses
greater than 0.75 mm, g. high resistance to high energy radiation, and h. self-
extinguishing after removal of the ignition source.
The main disadvantages are: a. processing requires care, b. limited chemical
resistance, c. notch sensitivity and susceptibility to stress cracking.

3.11 Polyethylene Terephthalate

Polyethylene terephthalate, PET, is a thermoplastic polyester made by condensation

reaction of ethylene glycol with either terephthalic acid or dimethyl terephthalate
(Margolis, 1985). By the end of the 1920s J.R. Whinfield and J.T. Dickson discov-
ered PET (BP 578079). It was first commercialized by Du Pont in 1930 (Brydson,
1982) as Dacron®, followed by ICI with Terylene® Films and blow-molded articles
have become very important commercially.
The average molecular weights are distributed from 15 000 to 20 000. The physi-
cal properties of PET are largely determined by the degree of crystallinity, which
varies between 30 and 40% depending on the processing conditions. The rate of
crystallization of PET is considerably less than that of polyacetal (POM) and
HDPE. The growth rate of the spherulites is only 10 μm/mm for PET compared with
400 μm/mm for POM and 5000 (μm/mm for HDPE (Domininghaus, 1993). To
achieve better crystallinity, the mould temperature should be equivalent to that for
maximum growth. This point is about 175°C, higher than for POM and HDPE. Rapid-
crystallization agents, nucleating agents, reduce the process cycle time and permit
lower mould temperatures below 100°C, leading to very fine spherulites and hard
stiff mouldings. Extrusion and rapid quenching below the temperature at which
most crystallization occurs (between 120 and 200°C), produces amorphous materi-
als and this may be followed with uniaxial orientation for fibres and biaxial orienta-
tion for films. The orientation is carried out at 100–120°C, the glass transition
temperature, Tg being 86°C.
The permeability of water vapor through PET is higher than that of polyolefins
but lower than that of polycarbonate, polyamide, and polyacetal. Antioxidants are
necessary to prevent to the oxidation of polyether segments in thermoplastic
272 S.H. Teoh et al.

polyester elastomer. Chemical resistance of PET is generally good to acids, alkalis,

and organic solvents.
Typical properties for partially crystalline PET include, a. high strength and stiff-
ness, b. favorable creep characteristics in comparison with POM, c. hard surface
capable of being polished, d. high dimensional stability, e. good electrical, mediocre
dielectric properties, and f. high chemical resistance except to strong acid and alka-
line solution.

3.12 Polyetheretherketone

Polyetheretherketone, PEEK, is a polymer combining stiff conjugated aromatic

groups and flexible ether segments. It was first prepared in the laboratory in 1977
and then marketed in 1978 by ICI, under the trade name of Victrex (Brydson, 1982).
The distribution of aromatic rings and polar flexible groups in the chain affects
the glass transition temperature, such that PEEK has a Tg around 145°C and Tm ca.
335°C, PEK (polyether ketone) Tg ca. 165°C and Tm ca. 342°C. Normally the chain
stiffness and bulkiness of aromatic rings make it difficult for these polymers to
crystallize and although they invariably remain mainly amorphous (Mascia, 1989),
PEEK is a partially crystalline thermoplastic. The maximum crystallinity of 48% is
achieved from the melt at 256°C and by subsequent conditioning of moldings at
185°C (Domininghaus, 1993). PEEK polymers are capable of melt processing
(Brydson, 1982). Other specific features are excellent gamma radiation resistance,
and good resistance to environmental stress cracking. PEEK shows excellent chemi-
cal resistance and can be used in aggressive environments.
Generally, the main characteristics of this material include, a. high tensile and
flexural strength, b. high impact strength, c. high fatigue limit, d. high heat distor-
tion temperature (315°C for 30 glass reinforced), e. good electrical properties over
a wide range of temperature, f. favourable slip and wear properties, g. high chemical
resistance, h. high resistance to hydrolysis, j. high resistance to radiation, k. low
flammability, very low gas and smoke emission, and 1. easy processing, no thermal
after-treatment of injection moldings.

3.13 Polysulfone

Polysulfone is a polymer which has properties matching those of light metals (Park
and Lakes, 1992). The first commercial polysulfone was introduced in 1965 by
Union Carbide as Bakelite Polysulfone, now called Udel®. In 1967 3M offered
Astrel 360 referred to as a polyarylsulfone. In 1972 ICI introduced a polyethersul-
fone Victrex®. A high toughness polysulfone was released in the late 1970s by
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 273

Union Carbide. Although the commercial polymers are linear and most have regular
structures they are all principally amorphous. The backbone aromatic structure
leads to high values of the glass transition temperature between 190 and 230°C. The
Union Carbide materials have a secondary transition at -100°C and the ICI polymer
at -70°C. Typical Mn values are ca. 23 000. Commercial materials are described
variously as polysulfones (Udel), polyarylsulfones (Astrel), polyether sulfones or
polyarylethersulfones (Victrex) (Brydson, 1982).
The polymer is manufactured from bisphenol A and 4, 4-dichlorosulphonyl
sulfone by multi-step condensation. The most distinctive feature of the backbone
chain of those polymers is the diphenylene sulfone group. The sulphur atom in
each group is in its highest state of oxidation and tends to draw electrons from
the adjacent benzene rings, hence resisting any tendency to lose electrons to an
oxidizing agent. Polysulfones thus show outstanding oxidation resistance. The
aromatic nature of the diphenylene sulphone can absorb considerable energy
applied as heat or radiation and so resists thermal degradation. The diphenylene
sulfone group thus confers on the entire polymer molecule the inherent charac-
teristics of thermal stability, oxidation resistance, and rigidity at elevated
temperatures.
The potential for energy dissipation confers good impact strength and ductility
down to -100°C with high elongation to break and tensile strength. Under most
conditions, impact properties rival those of bisphenol A polycarbonate. Unlike
polycarbonate, however, polysulfone can exhibit excellent resistance to hydrolysis
or reduction of molecular weight even at elevated temperatures. Tests on the hydro-
lysis stability of polysulfones have been carried out up to 10 000 hours without
observed loss of molecular weight.
The polymers are stable in aqueous inorganic acids, alkalis, salt solutions,
aliphatic hydrocarbons, and paraffin oils, are transparent, capable of steam steril-
ization, and free from taste and smell. They should not come in contact with
ketones, aromatic solvents, chlorinated hydrocarbons, and polar organic solvents.
They may show stress crazing on exposure to steam or water. A polyethersulfone,
however, exhibited no crazing even after 300 hours and retained 90% of initial
tensile impact strength. For a thermoplastic material, creep is low at moderate
temperatures but is significant at temperatures approaching the glass transition.
However, the wear properties of this material are not as good as PE and POM
(Teoh et al., 1994).
Generally polysulfone has the following characteristic properties, a. high strength,
stiffness and hardness between -100 and +150°C short-term to 180°C, b. high ther-
mal stability and heat distortion temperature, c. crystal clear (slightly yellowish)
transparency, d. high processing temperature, e. high melt viscosity, f. high chemical
resistance, g. susceptibility to stress cracking with certain solvents, h. high resistance
to β-, γ-, X- and IR-radiation, i. high transmittance for microwaves, and j. high flame
resistance and low smoke development.
Table 3.1 Chemical structures of thermoplastic polymers in biomedical applications
274
S.H. Teoh et al.

(continued)
3 Thermoplastic Polymers In Biomedical Applications: Structures, Properties… 275

Table 3.1 (continued)

 276

Table 3.1 (continued)

S.H. Teoh et al.
 3

Table 3.2 Properties of thermoplastic polymers in biomedical applications

a. Physical properties
Physical properties Unit PE PP PU PTFE PVC PA
Density g/cm3 0.954–0.965 0.90–0.915 1.02–1.28 2.10–2.20 1.16–1.70 1.02–1.15
Water absorption % 0.001–0.02 0.01–0.035 0.1–0.9 0.01–0.05 0.04–0.75 0.25–3.5
Solubility parameter MPa½ 16.4–16.6 16.3 16.4–19.5 12.6 19.4–21.5 23.02
Refractive index, nD20 1.52–1.54 1.47–1.51 1.5–1.65 1.35–1.38 1.52–1.57 1.52–1.57

Physical properties Unit PMMA POM PC PET PEEK PS

Density g/cm3 1.12–1.2 1.40–1.42 1.2–1.26 1.31–1.38 1.29–1.49 1.13–1.60
Water absorption % 0.1–0.4 0.2–0.4 0.15–0.7 0.06–0.3 0.15–0.51 0.14–0.43
Solubility parameter MPa½ 18.58 22.4 19.4–19.8 21.54 20.2 20.26–22.47
Refractive index, nD20 1.49–1.51 1.48 1.56–1.60 1.51 1.56–1.67

b. Mechanical properties
Mechanical property Unit PE PP PU PTFE PVC PA
Bulk modulus GPa 0.8–2.2 1.6–2.5 1.5–2 1–2 3–4 2.4–3.3
Tensile strength MPa 30–40 21–40 28–40 15–40 10–75 44–90
Elongation at break % 130–500 100–300 600–720 250–550 10–400 40–250
Young’s modulus GPa 0.45–1.3 1–1.6 0.0018–0.009 0.3–0.7 1.0–3.8 1.4–2.8
Elastic limit MPa 20–30 20–33 28–40 15–30 23–52 40–58
Endurance limit MPa 13–19.6 11–18.2 21–30 9–18 13.8–31.2 22–31.9
Fracture toughness MPa m½ 2.2–4 1.7–2.1 0.1–0.4 2.5–3 1–4 1.8–2.6
Hardness MPa 60–90 60–100 50–120 27–90 70–155 100–160
Thermoplastic Polymers In Biomedical Applications: Structures, Properties…

Compressive strength MPa 30–40 30–45 33–50 30–60 32–80 60–100

Poisson’s ratio 0.4–0.42 0.4–0.45 0.47–0.49 0.44–0.47 0.37–0.43 0.38–0.42
Shear modulus GPa 0.18–0.46 0.4–0.6 0.0008"4–0.003 0.11–0.24 0.7–1.1 0.52–0.9
(continued)
277
Table 3.2 (continued)
278

b. Mechanical properties
Mechanical property Unit PMMA POM PC PET PEEK PS
Bulk modulus GPa 3–4.8 4–5.6 2.8–4.6 3–4.9 4–4.5 3.8–4.6
Tensile strength MPa 38–80 70–75 56–75 42–80 70–208 50–100
Elongation at break % 2.5–6 15–80 80–130 50–300 1.3–50 25–80
Young’s modulus CPa 1.8–3.3 2.55–3.5 2–2.9 2.2–3.5 3.6–13 2.4–2.9
Elastic limit MPa 35–70 65–72 53–75 50–72 12–60 58–70
Endurance limit MPa 19.3–38.5 28–42 29.2–41.3 30–43.2 33–36 34.8–42
Fracture toughness MPa m½ 0.8–1.3 1–1.5 2.5–3.2 1.2–2 2.3–2.5 1.3–2
Hardness MPa 100–220 110–220 110–180 97–210 100–120 180–240
Compressive strength MPa 45–107 70–80 100–120 65–90 80–120 72–100
Poisson’s ratio 0.4–0.43 0.38–0.43 0.39–0.44 0.38–0.43 0.38–0.43 0.38–0.42
Shear modulus GPa 0.6–1.2 0.79–1 0.95–1.05 0.83–1.1 1.2–1.4 0.8–1

c. Thermal properties
Thermal property Unit PE PP PU PTFE PVC PA
Service temperature in °C 90–130 140 80–130 300 55–100 130–200
air without mechanical
loading (short-term)
Service temperature in air °C 70–100 100 60–80 250 50–85 70–120
without mechanical
loading (long-term)
Minimum service °c −63 to −53 −123 to −23 −123 to −23 −263 to −253 −43 to −28 −60 to −50
temperature
Melting(Tm)/decomposing °c 125–135 160–180 180–250* 322–327 150* 220–267
(Td*) ranges
Glass transition °c −113 to −103 −30 to −3 −73 to −23 20 to 22 −23 to 90 20 to 92
S.H. Teoh et al.

temperature T
 3

c. Thermal properties
Thermal property Unit PE PP PU PTFE PVC PA
Softening temperature °C 40–50 70–100 100 40–110 80–200
Specific heat J/g.K 1.95–2.20 1.70–2.35 0.4–1.76 1.00–1.01 0.85–1.80 1.26–1.8
Thermal expansion 106/K 100–200 80–200 150–210 100–150 60–210 80–150
Thermal conductivity W/m K 0.42–0.52 0.12–0.24 0.29–1.80 0.19–0.25 0.13–0.26 0.23–0.29

Thermal property Unit PMMA POM PC PET PEEK PS

Service temperature in °C 76–108 110–140 160 180–200 300 160–260
air without mechanical
loading (short-term)
Service temperature in °C 65–98 90–110 135 100 250 150–200
air without mechanical
loading (long-term)
Minimum service °c −123 to −73 −123 to −73 −133 to −123 −133 to −38 −123 to −103 −123 to −73
temperature
Melting(Tm)/decomposing °c ~170* 164–175 225–250 245–255 335 >500*
(Td*) ranges
Glass transition °c 106–115 -13–75 145 67–127 144 167–230
temperature Tg
Softening temperature °c 70–115 110–163 138–148 70–185 140–315 150–216
Specific heat J/gK 1.28–1.5 1.40–1.46 1.17–130 1.05–1.60 1.5–1.6 1.1–1.30
Thermal expansion 106/K 62–105 90–125 40–75 50–120 15–47 53–58
Thermoplastic Polymers In Biomedical Applications: Structures, Properties…

Thermal conductivity W/mK 0.10–0.19 0.22–1.1 0.14–0.22 0.15–0.34 0.25–0.92 0.13–0.28

279
Table 3.3 Processing conditions for thermoplastc polymers in biomedical applications
a. Physical properties
Thermoplastics Process Special process Pre-treatment Remarks
1. Polyethylene Injection moulding Film processing No pre-drying PE-UHMW in solid or porous form has been used in biomedical
Extrusion for PE-LD treatment except study and application. Its most outstanding properties are wear or
Blow film extrusion Rotational hygroscopic abrasion resistance, excellent impact resistance, and fatigue
Flat film extrusion moulding for additives are resistance.
Blow moulding PE-powders added.
Thermoforming Block, sheet, Stabilizers and
Compression tube, profile, and antioxidants are
moulding film processings needed for specific
for PE-HD and processing.
PE-HD-UHMW
Powder sintering
technology
for PE-UHMW
2. Polypropylene Injection moulding Extrusion: blown No pre-drying Polypropylene has exceptionally high flex life, excellent
Extrusion film, flat film, treatment except environmental stress cracking resistance, excellent wear
Blow moulding sheet, tube, hygroscopic resistance, higher temperature resistance (withstanding steam
Compression monofilaments additives’ are sterilization), and low cost. Fiber applications such as suture,
moulding Injection: added. sewing ring, braided ligament, skin and abdominal patches.
Thermoforming Long-lasting Stabilizers and Promising applications in thin-wall packaging competing with
integral hinges antioxidants are polystyrene. Polypropylene has lower specific heat and the flow
Biaxially oriented needed for specific properties are more sensitive to temperature and shear rate. The
packaging film processing. mold shrinkage is lower than polyethylene but higher than with
Tapes polystyrene. It has higher melt strength is important for extrusion
blow molding of hollow objects. Lower molecular weight grades
are suited for extrusion of monofilaments and injection molding
of thin-walled articles. Cold forming may be done at room
temperature (rolling), and forging, pressure forming, rolling and
stamping at temperatures below the crystallite melting region
(150 to 160°C). Film processing especially in oriented form
competing with polyethylene.
3. Polyurethane Injection moulding Sheet extrusion Polyurethane, Characteristics high flexibility and high impact resistance, and
Extrusion Shape extrusion especially excellent biocompatibility. Film forms of polyurethane have been
Cast or blow film aliphatic type is used in fabrication of vascular graft and patches, heart valve
extrusion hygroscopic; the leaflets, blood pumps, diaphragms for implantable artificial heart,
Fiber processing pellets must be and carriers for drug delivery. Elastomeric fibers (Spandex) made
dried before from polyurethane copolymer have been used in surgical hoses.
extrusion. Unfavorable processing conditions will induce residual stresses in
the products which impair the biostability of polyurethane-based
prostheses. To avoid residual stresses in polyurethane tubes, an
upper limit of drawn down ratio of 1.5:1.0 are recommended for
the appropriate stretching during extrusion. If water bath for tube
processing is too cold, residual stresses are also induced. A
recommended temperature for the water bath is between
21–27°C.
4. Polytetrafluorethylene Sintering High temperature PTFE is a Exceptional chemical resistance, temperature resistance, and
Ram extrusion sintering process hydrophobic radiation and weathering resistance.
Paste extrusion for parts, sheets, polymer and Outstanding electrical properties as insulator or dielectric
Coating followed plates pre-drying is material.
by sintering Ram extrusion not necessary. Low adhesion and low coefficient of friction.
Impregnation for rods, tubes, Exceptional flame resistance.
profiles, wire Expanded PTFE (Gortex) has been used in fabrication of blood
coatings, and vessel prostheses.
fibers
Insulating films,
crucibles
Expanded tubular
form
(continued)
Table 3.3 (continued)
5. Polyvinylchloride Extrusion Blown film Proper stored Plasticized PVC favours calendering, while unplasticized PVC
Calendering Flat film polyvinylchlorde prefers extrusion.
Injection molding Sheets can be used Injection moulding is difficult for both plasticized and
Extrusion blow Tubes and without pre-drying. unplasticized PVCs except careful control of processing
molding profiles Premixing of conditions or special design of machine.
Stretch blow Cable sheathing ingredients will Characteristic flame resistance.
molding Bristles and be considered for Decomposition happens at high temperature. Overheat in
Compression mono filaments plasticized PVc. processing should be avoided.
molding Tubular, sheet, plate, and film forms of PVC have been used in
Sintering medical devices such as blood bags and catheters. Implants of
PVC are not encouraged.
6. Polyamides Injection molding Injection: Polyamides, Excellent friction properties and good wear and abrasion
Extrusion thin-wall articles, especially resistance.
Extrusion blow engineering aliphatic types Excellent hydrocarbon resistance.
molding components are hygroscopic. Films are used for packaging. Fiber form is employed as suture
Extrusion: Pre-drying is materials.
bristles, needed before
packaging, tapes, processing, and
fiber, wire, film, also precaution
sheet, tubes, will be considered
profiles, during and after
sheathing process.
Polyacrylates can Excellent transparency, good scratch resistance.
easily pick up Good processability and machinability.
moisture from Monomer and polymer powder casting to produce bone cement.
environment. Hydrogel has been used to fabricate contact lens.
Pre-drying is Monomer-polymer doughs is used for processing dentures.
necessary.
7. Polyacrylates Injection molding Primary forms: Polyacrylates can Excellent transparency, good scratch resistance. Good
Extrusion sheet, rod, and easily pick up processability and machinability. Monomer and polymer powder
Compression tube Casting moisture from casting to produce bone cement. Hydrogel has been used to
molding from monomer environment. fabricate contact lens. Monomer-polymer doughs is used for
Thermoforming for optical Pre-drying is processing dentures.
properties necessary.
Extrusion from
thermoplastic
resins to produce
sheet. Injection
moulding for
small complex
parts
8. Polyacetal Injection molding Injection Polyacetal is less Oustanding creep and fatigue resistance.
Extrusion moulding hygroscopic than Good toughness and impact resistance.
Blow molding Extrusion nylon. Excellent strength for engineering application.
Compression However, acetal Processing temperature must be carefully controlled.
molding polymer must be Fiber and film forms of polyacetal are not available.
stored in dry place. Polyacetal, Delrin, has been used in disc of mechanical heart
valves.
9. Polycarbonate Injection molding Injection and Polycarbonate can Applications have been found in consideration of toughness,
Extrusion extrusion pick up enough rigidity, transparency, self-extinguishing characteristics, good
Blow moldings Films: extruded moisture to electrical insulation and heat resistance.
Thermoforming and solvent cast, deteriorate quality Polycarbonate has been used in the manufacture of contact lenses.
Hot bending uniaxially of products.
oriented Pre-drying is
amorphous and necessary.
partially,
crystalline
Tube, rod, profile,
sheet: extrusion
(continued)
Table 3.3 (continued)
10. Polyethyleneterephthalate Injection moulding Fiber process Polyethyleneter- Characteristic crystallization.
Extrusion Uniaxially ephthalate is Both amorphous and crystallized products can be made through
Blow molding oriented tapes hygroscopic. control of crystallization.
Films, packaging Pre-drying is The benefits from PET products are their stiffness, warp
film, sheet, necessary. resistance, and dimension stability.
articles Fiber form of PET has been used to fabricate blood vessels and
Biaxially oriented by-pass prostheses.
film and sheet Suture made from PET.
Dacron® sewing ring and medical fabrics.
11. Polyetheretherketone Injection molding Injection molding PEEK is Characteristic high strength at high temperature.
Compression for articles hydrophobic Excellent resistance to hydrolysis and environmental stress
molding Extrusion: polymer but cracking.
Extrusion filmIcast and pre-drying is Carbon fiber/PEEK composite investigated in bone fracture
Composite oriented necessary for fixation.
monofilament quality control.
wire covering
Composite:
carbon fiber/
PEEK composite
12. Polysulphone Injection molding Injection molding Polysulphone is Good rigidity, creep resistance, and toughness.
Extrusion for articles hygroscopic and Hydrolysis resistance and can undergo repeated steam
Blow molding Extrusion: film predrying is sterilization.
Thermoforming and sheet which required to avoid Transparent products can be made and used in medical field.
can be vapor formation Hollow fiber and membrane devices have been used in
thermoformed during processing. hemodialysis.
Carbon fiber/polysulphone composite has been used for bone
fracture fixation.
 3

Table 3.4 Trade names of thermoplastic polymers in biomedical applications

1. Polyethylene (PE)
PE-LD and PE-HD
Eltex(Solvay, BE) Fertene Maplen (Hirnrnont, IT) Ladene (SABlC, Saudi Arabia)
Finathene (Fina, BE) Rurniten (Rurnianca, IT) Escorene (Exxon Chern, US)
Alathon, Sclair (Du Pont Canada Inc., CA) NeoZex (Mitsui, JP) Fortiflex (Soltex Polymer, US)
Boalen (Petrirnex, CS) Nipolon (Toyo Soda, JP) Norchern (USI, US)
Hostalen (Hoechst, DE) Novatec (Mitsubishi, JP) Paxon (Allied Signal Corp., US)
Lupolen (BASF, DE) Hi Zex (Mitsol, JP) Microthene, Petrothene (US I Chern., US)
Vestolen (Huls, DE) Mirason (Mitsui Polychern., JP) HiFax (Hirnont, US)
Ertileno (Union ERT, ES) Sholex (Showa Denko, JP) Marlex (Phillips Petrol., US)
Natene (Rhone-Poulenc, FR) Surnikathene (Surnitorno Chern., JP) Super Dylan (Arco Chern., US)
Lacqtene (Atochern, FR) Suntec (Asahi Chern., JP) Tenite (Easrrnan Chern., US)
Carlona (Shell, GB) Staflene (Nisseki Plastic Chern., JP) Hipten (Hernijska Ind., YU)
Rigidex, Novex (BP Chemicals, GB) Yukalon (Mitsubishi Petroleum, JP) Okiten (INAOKI, YU)
Tipolen (Tiszai Vegyi Kombinat, HU) Starnylan (DSM, NL)
Eraclene (EniChern Base, IT) Ropol (Chern. Kornb. Borcesti; RO)
PE-HD-UHMW
Hostalen GUR (Hoechst AG, DE) Lupolen UHM (BASF AG, DE)
PE-LLD
Eltex (Solvay, BE) Rurniten (Rurnianca, IT) Escorene (Exxon Chern., US)
Alathon, Sclair (Du pont Canada, CA) Mirason (Mitsui Polychern., JP) Marlex TR 130 (Phillips Petroleum, US)
Novapol LL (Novacor Chern., CA) Novatex (Mitsubishi Chern., JP) Microthene, Petrothene (USI Chern., US)
Thermoplastic Polymers In Biomedical Applications: Structures, Properties…

Lupolen (BASF, DE) Surnikathene (Surnitorno Chern., JP) Norchern (USllExxon Chern., US)
Lotrex (Orkern Norsolor SA, FR) Ultzex (Mitsui Petrochemical, JP) Rexene (EI Paso Chern., US)
News (Neste Oy Chern., FI) Starnylex (DSM, NL) Tenite (Eastman Chern., US)
Visqueen (lCI, GB) Ladene (SABlC, Saudi Arabia)
285

Eraclear (EniChern., IT) Dowlex (Dow Chemical, US)

(continued)
Table 3.4 (continued)
286

2. Polypropylene (PP)
Asota, Ecofelt (Chemic Linz, AT) Eltex (Solvay, FR) Starnylan (DSM, NL)
Marlex (Phillips Petroleum Co., BE) Carlona (Shell, GB) Bicor (Mobil Chern., US)
Istrono (Chern., Werke J. Dirnitrow, CS) Propathene (ICI, GB) Extrel, Twistlock, Vistalon(Exxon
Chern, US)
Tatren (Petrirnex, CS) Biofol (Chern. Kombinat. Tisza, HU) Fortilene (Soltex Polymer Corp. US)
Hostalen PP (Hoechst, DE) Afax, Moplen, Valtec (Hirnont, IT) Liteplate S (Hercules, US)
Platilon (Deutsche A TOchern., DE) Bifax (Showa denko. JP) Profax (Hirnont, US)
Trovidur (HulsTroisdorf, DE) Eperon (Kanegafuchi Chem., JP) Rexene (EI Paso Chem., US)
Ultralen (Lonza Werke, DE) Noblen (Mitsubishi Petrochemical, JP) Tenite (Eastman Chemical, US)
Vestolen P (Huls, DE) Novatec (Mitsubishi Chem., JP)
Apryl, Lacqtene P (Atochem., FR) PolyPro, Sunlet (Mitsui Petrochem., JP)

3. Polyurethane (PU)
General
Ucefix, Ucellex (UCB, BE) Europolymers (Avalon Chemical Co., GB) Urafil (Akzo-Wilson-Fiberfil, US)
Fabeltan (Tubize Plastics, BE) Jectothane (Dunlop Holdings, GB) Hi-Tuff (J.P. Stevens & Co., US)
Caprolan, Elastolen, Elastolan (Elastogran Pemullex (Pemu Chemolimpex, HU) Esteloc, Estane, Roylar
Polyurethane, DE) (B.F. Goodrich Chemical Co., US)
Desmopan (Bayer, DE) Uthane (Urethanes India, IN) Irogran, Plastothane (Morton
Thiokol, US)
Lurollex (Lehmann u. Voss Co., DE) Pelprene (Toyobo Co., Resins Div., JP) Proplastic (Pro Lam, US)
Oldopren (Busing u. Fasch & Co., DE) Durane (Swanson, US) Q-Thane (Quinn & Co., US)
Durelast (B & T polymers, GB)
Medical special
Angiollex (Abiomed Danvers, MA) Mitrathane (PolyMedica Burlington, MA, US) Corplex (Corvita Miami, FL, US)
Biomer (Ethicon Somerville, NJ, US) Pellethane (Dow Chemical La Porte, TX, US) Unithane 80F (NICPBP Beijing,
S.H. Teoh et al.

China)
 3

3. Polyurethane (PU)
Cardiothane (Kontron Everett, MA, US) Surethane (Cardiac Control Palm Coast, FL, US) Corethane (Corvita Miami, FL, US)
Chronollex (PolyMedica Burlington, MA, US) Tecollex (Thermedics Inc Woburn, MA, US) PU 10 (Univ. NSW, Australia)
Hemothane (Sams, Div 3M Ann Arbor, MI, US) Toyobo TM5 (Toyobo Co. Osaka, Japan)
4. Polytetralluoroethylene (PTFE)
Hostallon TF (Hoechst, DE) Neollon (Daikin Ind., JP) RT duroid (Rogers Corp., US)
Pamllon (Norton Pampus, DE) Polyllon (Daikin Kogyo Co., JP) Rulon (Dixon Ind. Corp., US)
Forallon (A TOCHEM, FR) Ferrotron, F1uorosint (Polymer Corp., US) Salox (Allegheny Plastics, US)
Sorellon (Pechiney U.K., FR) F1uoroloy (F1urocatbon, US) Tellon (Du Pont de Numours, US)
Gallon (Plastic Omnium, FR) FluoroMet (Raymark, US) Turcite (W.S. Shamban & Co., US)
Fluon (ICI, GB) Goretex (W.L. Gore Assoc., US) Tygallor (American Cyanamid
Aerospace, US)
Algollon (Montedison, IT) Halon (Ansimont, US) Xylon (Whitford Corp., US)

5. Polyvinylchloride (PVC)
Benvic (Solvay, BE) Corvic, Vynide, Welvic (ICI, GB) Rosevil (Chern. Kombinat Borzesti, RO)
Plastilit, Selchim, Solvic (Solvay & Cie., BE) Ongrovil (Barsodi Veggi Komb., HU) Ensolite (Uniroyal Chern., US)
Vipopham (Lonza, CH) Ravinil, Sicron, Vipla, Viplast (EniChem, IT) Ethyl (Ethyl Corp., Polymer Div., US)
Astralon, Trocal, Trosiplast, Trovidur, Vestolit Vixir (S.I.R. (Montedison), IT) Fiberloc, Geon (B.F. Goodrich, US)
(Huls-Troisdorf, DE)
Decelith (VEB Ellenburger Chemiewerk, DE) Hishiplate (Mitsubishi Plastics Ind., JP) Pliovic, Vinacel, Vycell (Goodyear, US)
Vinnol (Wacker Chemie, DE) Kureha, Viclon (Kureha Chern. Ind., JP) Vygen (General Tire & Rubber Co., US)
Genopak, Genotherm, Hostalit (Hoechst, DE) Vinychlon (Mitsui Toatsu Chern., JP) Vynaloy (B.F. Goodrich Chern. Co., US)
Thermoplastic Polymers In Biomedical Applications: Structures, Properties…

Armodur (Rhone-Poulenc, FR) Vinylfoil (Mitsubishi Gas Ind., JP) Hipnil (Hemijska Industria, YU)
Bipeau, Orgavyl, Polycal (ATOCHEM., FR) Varian (DSM, NL) Jugotherm, Juvinil (Jugovinil, YU)
Ekavyl (PCUK, FR) Norvinyl, Pevikon (Norsk Hydro, NO) Zadrovil (Polikem, YU)
Carina, Duraftex (Shell Intern. Chern Co., GB) Oltivil (Chern. Komb. Pitesti, RO)
287

(continued)
Table 3.4 (continued)
288

6. Polyamide (PA)
PA 6
Fabenyl (Tubize Plastics, Maranyl (ICI, GB) Akulon (Engineering Plastics of AKZO
Plastics, NL)
Grilon (Ems Chemie, CH) Latamid (L.A.T.I., IT) Capron (Allied Signal Engn. Plastics, US)
Durethan B (Bayer, DE) Nivionplast (EniChem, IT) Plaskon (Plaskon Moldings Div., US)
Ultramid B (BASF, DE) Renyl (Snia Technopolimeri, IT) Zytel (Du Pont de Nemours, US)
Orgamide (ATOCHEM, FR) Amilan, Amilon (Toray Ind., JP)
Technyl C (Rhone-Poulenc Specialites Chimiques, FR) Torayca (Toray Ind., JP)
PA 66
Durethan A (Bayer, DE) Leona (Asahi Chemical Ind., JP) Minion (Du Pont de Nemours, US)
Technyl A (Rhone-Poulenc Specialites Chimiques, FR) Torayca (Toray Ind., JP) Zytel (Du Pont de Nemours, US)
Maranyl A (ICI, GB) Akulon (Engineering Plastics of
AKZO Plastics, NL)

7. Polyacrylates
Umaplex (Synthesia, CS) Asterite (ICI, GB) Sumipex (Sumitomo Chern. Co., JP)
Acrifix (Rohm, DE) Diakon (ICI, GB) Casoglas (Casolith, NL)
Lucryl (BASF, DE) Perspex (ICI, GB) Acrylite (Cy/Ro Industries, US)
Degaplast (Degussa, DE) Unilok (British Vita Co., GB) Lucite (Du Pont de Nemours, US)
Deglas, Dewglas (Degussa, DE) Vetredil (Vetril, IT) Corian (Du Pont de Nemours, US)
Dewoglas (Degussa, DE) Vedril (Montedison, IT) Crofon (Du Pont de Nemours, US)
Paraglas (Degussa, DE) Acrypanel (Mitsubishi Rayon Co., JP) Electroglas (Glasftex Corp., US)
Plexidur, Plexiglas (Rohm, DE) Delmer, Depet (Asahi Chem. Ind., JP) Exolite (Cyro Industries, US)
Resarit (Resart, DE) Eska (Mitsubishi Rayon, JP) Gardlite (Southern Plastics Co., US)
Vestiform (Huls, DE) Palapet (Kyowa Gas Chem., JP) Oroglas (Rohm & Haas Co., US)
Altuglas (Altulor, Orekem, FR) Shinkolite (Mitsubishi Rayon Co.,JP) Swedcast (Swedlow, US)
S.H. Teoh et al.
 3

Table 3.4 (continued)

8. Polyacetal (POM)
Homopolymers
Delrin (Du Pont, US) Tenal (Asahi Chemical Ind., JP)
Copolymers
CeIcon, Hostaform (Hoechst, DE) Duracon, Alkon, and Kematal (Daicel Polyplastic Co., JP) Ultraform (BASF, DE)

9. Polycarbonate (PC)
SparJux (Solvay & Cie., BE) Sinvet (EniChem, IT) Lexan (General Electric Plastics, US)
Durolon (Policarbonateos do Brazil, BR) Novarex (Mitsubishi Chem. Ind., JP) MerIon (Mobay Chemical Corp .. US)
Makrolon (Bayer, DE) Panlite (Teijin Chemicals, JP) Polycarbafil (Akzo-WilsonFiberfil, US)
Orgalan (A TOCHEM, FR) Xantar (DSM, NL) Polygard (Poly tech, US)
Star-C (Ferro Eurostar, FR) Calibre (Dow Chemical Corp., US) Stat-Kon (LNP Corp., US)
Royalite (British Vita Co., GB) Ekonol (Carborundum Co., US)

10. Polyethyleneterephthalate
(PET)
Crastin (Ciba Geigy, CH) Melinar, Melinite (lCI, GB) Pelion (Mobay Chem. Corp., US)
Grilpet (Ems Chemie, CH) Amite (Akzo Engng. Plastics, NL) Petra (Allied Signal, US)
Impet (Hoechst, DE) Etar (Eastman Chem. Intern., US) Ropet (Rohm & Hass Co., US)
Ultradur (BASF, DE) Mindel (Amoco Performance Products, US) Rynite (Du Pont de Nemours, US)

11. Polyetheretherketone
(PEEK)
Thermoplastic Polymers In Biomedical Applications: Structures, Properties…

Victrex PEEK (ICI, UK)

12. Polysulphone
U. Polysulphone Stabar (ICI, UK) Udel (Amoco Performance Products, US)
289

Sumilik FST (Sumitorno,

Chern. Co., JP)
290 S.H. Teoh et al.

References

Black, J. (ed.) (1992), in Biological Performance of Materials: Fundamentals of Biocompatibility,

Second edition, Marcel Dekker. Inc, New York.
Brydson, J.A (ed.), Plastics Materials, Butterworths Scientific, fourth edition, 1982; fifth edition,
1989.
Carhart, RO. (1985) Polycarbonate, in Engineering Thermoplastics, Properties and Applications,
Margolis, J.M. (ed.), Chapter 3, pp. 29–82..
Chapman, RD. and Chroma, J.L. (1985) Nylon plastics, in Engineering Thermoplastics, Properties
and Applications, Margolis J.M. (ed.), pp. 83–122.
Charrier, J.M. (ed.) (1990) Polymeric Materials and Processing: Plastics, Elastomers and
Composites, Hanser.
Domininghaus, H. (ed.) (1993) Plastics for Engineers, Materials Properties, Applications,
Car-Hanser Verlag.
Goodman, S.B. and Fomasier, V.L. (1992) Clinical and experimental studies in the biology
of aseptic loosing of joint arthroplasties and the role of polymer particles, in Particulate
Debris from Medical Implants: Mechanisms of Formation and Biological Consequences,
ASTM ATP 1144, K.R St John, (ed), American Society for Testing and Materials, Philadelphia,
pp. 27–37.
Harper, C.A (eds) (1992) Handbook of Plastics, Elastomers, and Composites, McGraw-Hill.
How, T.V. (1992) Mechanical properties of arteries and arterial grafts, in Cardiovascular
Biomaterials, Hastings, G.W. (ed.), Springer-Verlag, London, pp.1–35.
Jones, AJ. and Denning, N.T. (1988) in Polymeric Biomaterials: Bio- and Ecocompatible Polymers,
A Perspective for Australia, Department of Industry, Technology and Commerce.
Lilley, P.A, Blunn, G.W., and Walker, P.S. (1993) Wear performance of PEEK as a potential pros-
thetic knee joint material, in 7th International Conference on Polymers in Medicine and
Surgery, 1–3 September 1993, Leeuwenhorst Congress Center, Noordwijkerhout, The
Netherlands, pp. 320–326.
Margolis, J.M. (1985) Engineering Thermoplastics: Properties and Applications, Dekker,
New York.
Mascia, L. (1989), in Thermoplastics: Materials Engineering, Second Edition, Elsevier Applied
Science, London and New York.
McMillin, C.R (1994) Elastomers for biomedical applications, Rubber Chem. and Tech. 67,
417–446.
Park, J.B. and Lakes, RS. (1992) Biomaterials, an Introduction, Second Edition, Plenum Press,
New York and London.
Rubin, 1.1. (ed.) (1990), Handbook of Plastic Materials and Technology, John Wiley & Son.
Staudinger, H. (1932) Die Hochmolekularen Organischer Verbindungen, Julius Springer.
Stokes, K., McVenes, R., and Anderson, J.M. (1995) Polyurethane elastomer biostability,
J. Biomaterials Applications, 9, 321–355.
Szycher, M. (1991) in Blood Compatible Materials and Devices: Perspectives Towards the 21st
Century, Sharma, C.P. and Szycher, M. (eds), Technomic Publishing CO., Inc., Lancaster,
Basel, pp. 33–85.
Teoh, S.H., Lim, S.c., Yoon, E.T., and Goh, K.S. (1994a) A new method for in vitro wear assess-
ment of materials used in mechanical heart valves, in Biomaterials Mechanical Properties,
ASTM STP 1173, H.E. Kambic and AT. Yokobori, Jr. (eds), American Society for Testing and
Materials, Philadelphia, pp.43–52.
Teoh, S.H., Martin, R.L., Lim, S.c., et al. (1990) Delrin as an occIuder material, ASAIO
Transactions, 36, M417–421.
Watson, M., Cebon, D., Ashby, M., Charlton, C., and Chong, W.T. (eds) (1994) Cambridge
Materials Selector V2.02, National University of Singapore, Granta Design Ltd.
Wenz, L.M., Merritt, K., Brown, S.A, and Moet, A (1990) In-vitro biocompatibility of poly-
etheretherketone and polysulphone composites, J. Biomed. Maters Res., 24, 207–215.
Ziegler, K.E. (1955) Angew. Chem. 67(426),541.
Chapter 4
Biomedical elastomers

J.W. Boretos and S.J. Boretos

4.1 Introduction

Elastomers are described as materials that possess pronounced elasticity and

rebound. They can be tough, relatively impermeable to air and water and exhibit
resistance to cutting, tearing and abrasion. Often they are modified by compounding
to increase their hardness and strength. Or, conversely, they can be soft, compliant
and absorbent to water if the need exists. In some instances their properties can
closely simulate that of the tissues which they must contact. As biomedical materi-
als they may have originated from commercial formulations or been custom
designed from basic chemistry. Those that have been judged as biocompatible have
made significant contributions towards the development of successful medical
devices. Literally, every basic elastomer has been evaluated at some time for its pos-
sible suitability in contact with the body. This would include such materials as natu-
ral rubber, styrene rubber, polybutyl rubber, silicone rubber, acrylate rubber,
Hypalon®, polyurethanes, fluorinated hydrocarbon rubbers, polyvinyl chloride,
thermoplastic vulcanizates and others. Of these, only special medical grade formu-
lations of silicone, polyurethane, polyvinyl chloride and thermoplastic elastomer
have continued to be commercially successful.
There are important differences between materials and differences among simi-
lar materials within a given generic class. These differences are due to the chemical
composition of the polymer, the molecular configuration of the polymer and the
presence of functional groups. For instance, polyurethanes of a polyester base were
initially tried and found unstable for implantation whereas polyether based polyure-
thanes were decidedly more stable. Elastomers with aromatic structures behave dif-
ferently than the polymer having aliphatic structure. Not every material is suitable

J.W. Boretos (*) • S.J. Boretos

Consultants for Biomaterials, 6 Drake Court, Rockville, Maryland 20853, USA

© Springer Science+Business Media New York 2016 291

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_20
292 J.W. Boretos and S.J. Boretos

for every application. Some have been found to perform successfully under static
conditions but fail or perform undesirably under dynamic situations. Often, the
design of a device and the demands upon it will determine if the elastomer chosen
is the proper selection. Therefore the material and its use are inseparable. They must
be evaluated together. Merely passing an array of physical and biological tests do
not confer success. Biocompatibility is an essential element of medical grade elas-
tomers. A set of compatibility tests determine the general physiological acceptabil-
ity of an elastomer. These consist of passing USP Class VI tests. Additional testing
may be required depending upon the device, its area of application and the time it is
in contact with tissues. A Master File is often registered by the manufacturer of the
basic elastomer with the FDA to attest to its properties, composition and response
to biological testing. Demands on medical device manufacturers have never been
more stringent. Regulatory pressures, more indepth testing, the threat of litigation
plus the constraints of health care cost containment are affecting all aspects of the
design and development process and the availability of some biomedical elasto-
mers. A variety of elastomeric materials are available to meet the design challenges
presented by medical devices. However, there is still a need for even better
materials.
The elastomers that are listed here should be considered in light of their suitabil-
ity for a specific application. The properties tables should serve as a guide to design
options for those in the early stages of the development process. Keep in mind that
these properties listed in the tables and the compatibility standings are only indica-
tive of the performance characteristics that an elastomer may exhibit.

4.2 Types of Elastomer

Biomedical elastomers can be classified as to whether they are thermoplastic or

thermosetting in nature. Thermoplastic biomedical elastomers are gaining in com-
mercial importance and in some cases replacing traditionally used vulcanized ver-
sions. Thermosetting elastomers are irreversibly crosslinked and have had the
longest history of medical use. Both groups will be described citing representative
medical elastomers that are either commercially available or that may replace elas-
tomers that have been recently withdrawn from the market.

4.2.1 Thermoplastic elastomers

Thermoplastic elastomers (TPEs) are a special class of materials that process simi-
larly to other thermoplastic polymers, yet possess many of the desirable properties
of thermoset elastomers. Some TPEs are elastomeric alloys consisting of cross-
linked particles of rubber surrounded by a thermoplastic matrix. Others consist of
block copolymers and are typified by polyurethanes and styrene polymers.
4 Biomedical elastomers 293

Depending upon which thermoplastic elastomer is chosen, physical properties

can vary over a wide range. They can be either hard, or soft, flexible or stiff, elastic
or rigid. For the most part, they are smooth to the touch, yet will form tight seals to
surfaces they contact. They can be processed using conventional techniques and
equipment and in automated modes. Medical applications consist of such examples
as pacemaker lead wire coatings, artificial hearts, and catheters. A wide variety of
sundry uses have contributed to patient care and consists of bulbs and bladders,
serum caps and tubes, cushions, diaphragms, electrical connectors, flexible medical
wire coatings, gaskets, needle shields, pharmaceutical closures, seals, stoppers, tub-
ing, and valves. Most of the TPEs can be sterilized using gas, steam and radiation
with very little change in their molecular structures or properties (Table 4.13).

Thermoplastic vulcanizates
Thermoplastic vulcanizates are a separate class of thermoplastic elastomers (TPEs)
with Santoprene® as the representative biomedical elastomer.

Santoprene®
This thermoplastic vulcanizate is an olefin based elastomer; an elastomeric alloy. It
is totally synthetic and does not contain any natural rubber thereby avoiding many
of the allergic reaction problems associated with natural rubber latex. It exhibits
outstanding flex-fatigue resistance, low temperature flexibility (-40 °C) and resis-
tance to tearing and abrasion. Its resistance to plastic deformation under tensile and
compression stress is another of its features. Santoprene® is reported to be superior
to natural rubber in some situations and replaces silicone elastomers in others. It has
found use in peristaltic pump tubing, syringe plungers, seals, and caps, tracheal and
enteral tubing, vial closures and pump seals, disposable anesthetic hoses, intrave-
nous delivery systems, and other hospital devices. Santoprene® has met USP Class
IV requirements for in vivo biological reactivity and conforms to the Tripartite
Biocompatibility Guidance standards. However, the manufacturer does not recom-
mend Santoprene® for use as part of human implants. The material may be injection
molded, extruded, blow molded and thermoformed. For details on physical proper-
ties, processing and biocompatibility see Tables 4.1, 4.2, 4.13 and 4.14.

Copolyester ether elastomer

Ecdel™
This copolyester ether TPE is essentially polycyclohexanedimethyl-
cyclohexanedicarboxylate (PCCE). It is reported to possess the chemical resistance,
toughness and inertness yet exhibits elastic flexibility over a broad temperature
range. Ecdel™ is an unusual elastomer since it has a crystalline structure. Quenching
during molding can reduce its crystallinity and impart increased clarity. The material
is being used for uniquely designed intravenous bags with built-in bottle necks and
fasteners. The material can be injection or blow molded and extruded into film or
sheet; but only Ecdel™ 9967 may be processed into tubing. This TPE is also manu-
factured under the name CR3 by Abbott Labs (Tables 4.2, 4.12, 4.13, and 4.14).
294 J.W. Boretos and S.J. Boretos

Polyurethane-based elastomers
Polyurethanes are another class of TPEs. They are a large family of chemical com-
pounds that can consist of ether-based, ester-based, polycarbonate-based or poly-
propylene-based varieties. A number of copolymers are also included;. polyurethanes
are combinations of macroglycols and diisocyanates that have been polymerized
into tough and elastic materials. TPE polyurethanes have been used for peristaltic
pump tubing, parenteral solution tubing and catheters. The tables list the majority of
those that are commercially available. Among others are those either of limited sup-
ply, available for proprietary use only or that have been successful, but recently
discontinued such as:
• Hemothane Sarns Div. of 3M. Restricted to proprietary use.
• Biomer Ethicon, Inc. No longer available through this source.
• Surethane Cardiac Control Systems, Inc. Redissolved Lycra® thread. Some for-
mulations may have a few percent PDMS blended with it. Limited availability.
• Pellethane™ 2360 Dow Chemical, USA. This material is no longer available for
medical implant use (see also Pellethane™).
• Angioflex ABIOMED, Danvers, Mass. Restricted to proprietary use.
• Cardiothane Kontrol, Inc. A silicone-urethane interpenetrating polymer network.
Limited availability.
Internationally, polyurethanes for medical use have been developed by Italy,
China and Japan.

Biospan®
This TPE is a segmented polyurethane and is reported to be not significantly differ-
ent from biomer in chemistry and molecular weight. It is a polytetra-methyleneoxide-
based aromatic polyurethane urea with mixed aliphatic and cycloaliphatic diamine
chain extenders. A copolymer of diisopropylamino-ethyl methacrylate and decyl
methacrylate are added as a stabilizer. The material is supplied as 25% solids in
dimethyl acetamide solvent (Tables 4.3, 4.12, 4.13, and 4.14).

Biospan®-S
This is a silicone modified analog of Biospan® with a different stabilizer. It pos-
sesses a silicone-rich surface to enhance thromboresistance while maintaining the
bulk properties of Biospan® (Tables 4.3, 4.12, 4.13, and 4.14).

Biospan®-D
This is another version of Biospan® with surface modification by an oligomeric
hydrocarbon covalently bonded to the base polymer during synthesis. The additive
has a pronounced effect on the polymer surface chemistry but little effect on the
bulk properties of the base polymer according to the manufacturer (Tables 4.3, 4.12,
4.13, and 4.14).
 4 Biomedical elastomers

Table 4.1 Typical Properties of Thermoplastic Vu1canizates

Property
Durometer Tensile Modulus
Specific hardness strength, Elongation, ASTM Tear strength Compression set,
gravity shore psi percent D-412 pli,die C percent
Product and ASTM ASTM ASTM ASTM ASTM ASTM
Manufacturer Product no. D-792 D-2240 D-412 D-412 psi % D-624 D-395
Santoprene® 281– 45 0.97 45A 435 300 175 100 80 11
Rubber, 281– 55 0.97 55A 640 330 250 100 108 23
281– 64 0.97 64A 1,030 400 340 100 140 23
Advanced 281– 73 0.98 73A 1230 460 520 100 159 26
Elastomer 281– 87 0.96 87A 2,300 520 1,010 100 278 29
Systems 283– 40 0.95 40 D 2,750 560 1,250 100 369 32
295
 296

Table 4.2 Typical Properties of Copolyester Elastomers, PCCE

Property
Durometer Tensile Modulus Tear
Specific hardness strength, Elongation, ASTM strength Compression set,
gravity shore psi percent D-412 pli,die B percent
Product ASTM ASTM ASTM ASTM ASTM ASTM
and Manufacturer Product no. D-792 D-2240 D-412 D-412 psi % D-624 D-395
Ecdel™ 9965 -
Elastomer, 9966 1.13 95 A 3,500 380 16,000 100 100 40
Eastman 9967 135
Chemical Co.
J.W. Boretos and S.J. Boretos
Table 4.3 Typical Properties of Polyurethane-based Elastomers
Property
Tensile Modulus
Specific Durometer strength, Elongation, ASTM Tear strength Compression set,
gravity hardness shore psi percent D-412 pli,die C percent
Product and ASTM ASTM ASTM ASTM ASTM ASTM
Manufacturer Product no. D-792 D-2240 D-412 D-412 psi % D-624 D-395
Biospan Biospan® 75A 6000 850 575 100
4 Biomedical elastomers

segmented Biospan® D 70 D 6000 1000 550 100

polyurethane, Biospan® S 70 D 5500 1050 450 100
The Polymer
Technology
Group, Inc.
Hydrothane™, Dry 93 A 7800 580
Poly Medica Very dry 95 A 5800 475
Biomaterials, Inc. Wet 85 A 5600 500
MF-5000 1.15 50 A 3000 500 300 100
Medicaflex™, MF-5001 1.15 55 A 3000 300 100
MF-5040 1.15 60 A 5000 700 300 100
Advanced MF-5041 300 100
Resin MF-5056 1.15 65 A 5000 750 500 100
Technology MF-5057 550 100
MF-5062 1.14 60 A 5000 800 500 100
2363-55D 1.15 55 D 6900 390 2500 100 650 25
Pellethane™ 2363-55DE 1.15 53 D 6500 450 2300 100 600 30
2363 series, 2363-65D 1.17 62 D 6460 2900 100 1100 30
2363-75D 1.21 76 D 5810 380 1470
2363-80A 1.13 81 A 5200 550 880 100 470 25
297

(continued)
Table 4.3 (continued)
298

Property
Tensile Modulus
Specific Durometer strength, Elongation, ASTM Tear strength Compression set,
gravity hardness shore psi percent D-412 pli,die C percent
Dow Chemical 2363-80AE 1.12 85 A 4200 650 890 100 420 30
Co. 2363-80A 1.30 81 A 6860 670 970 100
R0120
2363-90A 1.14 90 A 5850 500 1700 100 570 25
2363-90AE 1.14 90 A 6000 550 1475 100 540
PolyBlend™ PB1000-650 65 D to 75 D6500 350 5300 100 - -
1000 and PB1100-55 1.02 55 A 2150 800 135 100 140 55-66
PolyBlend™ PB1100-60 1.02 60 A 2400 210 100 150 50-60
1100,
Poly Medica PB1100-75 1.02 75 A 3250 575 420 100 240 45-50
Biomaterials, Inc. PB1100-80 1.02 80 A 4600 590 555 100 330 25-30
EG60D 1.09 51 D 7829 363 2000 100
Tecoflex®, EG60D-B20 1.32 55 D 7484 370
EG65D 1.10 60 D 8074 335 2500 100
EG65D-B20 - 63 D 6986 321
Thermedics, EG68D 1.10 8686 332
Inc. EG72D 1.11 67 D 7739 307 3400 100
EG80A 1.04 72 A 5640 709 400 100
Thermedics, EG80A-B20 1.24 73 A 5571 715
Inc. EG85A 1.05 77 A 6935 565 700 100
EG85A-B20 1.25 83 A 5282 632
EG85A-B40 1.51 84 A 5093 559
EG93A 1.08 87 A 7127 423 1100 100
J.W. Boretos and S.J. Boretos
 EG100A 1.09 94 A 8282 370 1800 100
EG100A-B20 1.29 93 A 7104 369
EG100A-B40 1.54 96 A 5607 360
1055D 1.16 54 D 9600 350 2500 100
Tecothane® 1065D 1.18 64 D 10 000 300 3200 100
1074A 1.10 75 A 6000 530 530 100
1075D 1.19 75 D 8300 240 3600 100
4 Biomedical elastomers

Thermedics, 1085A 1.12 85 A 7000 450 800 100

Inc. 1095A 1.15 94 A 9400 400 1600 100
2055D 1.36 55 D 9000 360 2700 100
2065D 1.38 67 D 8500 300 3100 100
2074A 1.30 77 A 5500 580 510 100
2075D 1.40 77 D 7600 230 3000 100
2085A 1.32 87 A 6600 550 800 100
2095A 1.35 97 A 8200 450 1600 100
5187 1.20 87 A 6000 500 750 100 500 12
Texin™ 5265 1.17 65 D 6000 460 3300 100 1200 20
5286 1.12 86 A 6000 550 700 100 500 16
Miles, Inc. 5370 1.21 70 D 6000 180 4500 100 900
DP7-3002 88 A 2208 579 815 100 399 —
DP7-3003 — 50 D 3714 458 1049 100 564
DP7-3004 55 D 4783 392 1766 100 819
299
Table 4.4 Typical Properties of Polycarbonate-based Polyurethane
300

Property
Tensile Modulus
Specific Durometer strength, Elongation, ASTM Tear strength pli,
gravity hardness shore psi percent D-412 die C
Product and ASTM ASTM ASTM ASTM ASTM
Manufacturer Product no. D-792 D-2240 D-412 D-412 psi % D-624
PC-3555D 1.15 60 D 7000 350 1500 100
Carbothane™ PC-3555D-B20 1.36 57 D 8300 380 1600 100
PC-3572D 1.15 71 D 8500 300 4100 100
Thermedics, PC-3572D-B20 1.35 71 D 8400 310 3400 100
Inc. PC-3575A 1.15 73 A 4400 500 380 100
PC-3575A-B20 73 A 3500 600 410 100
PC-3585A 1.15 84 A 6500 390 640 100
PC-3585A-B40 1.68 89 A 3800 521 700 100
PC-3595A 1.15 95 A 6500 520 900 100
PC-3595A-B20 1.36 96 A 8300 390 1100 100
Chronoflex™ Chronoflex™ 70 A 7500 500 700 100
AR, Poly AR
Medica
Bio-materials, Inc.
Corethane® TPE 55D 1.211 55 D 7000- 365- 1850- 100
8500 440 2200
and TPE 75D 1.216 75 D 7000- 255- 5300- 100
9100 320 5700
TPE 80A 1.179 80 A 400- 770-
490 1250
Corhesive™, Corhesive™ 1.179 80 A 6500- 400- 770- 100
Corvita Corp. (cured) 7500 900 1250
J.W. Boretos and S.J. Boretos

Texin™ 5370, 5370 1.21 70 D 6000 180 4500 100 900

Miles, Inc.
4 Biomedical elastomers 301

Table 4.5 Typical Properties of Polypropylene-based Elastomers

Property
Durometer Tensile Tear
Specific hardness strength, Elongation, strength Compression
gravity shore psi percent pli,die B set, percent
Product
and Product ASTM ASTM ASTM ASTM ASTM
Manufacturer no. D-792 D-2240 D-412 D-412 D-624 ASTMD-395
Sarlink® 3260 0.95 60 A 870 619 183 42
medical grade
DSM
Thermoplastic
Elastomers, Inc.

Hydrothane™
Hydrothane™ is a TPE hydrogel belonging to the polyurethane family of polymers.
Hydrothane™ is an aliphatic material with water absorption capabilities ranging
from 5 to 25% by weight while still maintaining high tensile strength and elonga-
tion. Because of its water absorption capacity, Hydrothane™ is reported to be bac-
teria-resistant and lubricious. The polymer can be processed by conventional
extrusion and injection molding techniques. It can also be dissolved in dimethyl
acetamide solvent to produce a 25% solids solution suitable for dip-coating and
other solution processing techniques (Tables 4.3, 4.12, and 4.13).

Medicaflex™
The Lambda series of Medicaftex is a polyurethane-based TPE polymer that exhib-
its low modulus characteristics with high tear strength and abrasion resistance.
Those listed in the tables have passed USP Class VI compatibility tests and have
been used as replacements in some natural rubber latex and silicone rubber applica-
tions. The polymer has been applied to uses such as catheters, tubing and films
where softness, low durometer hardness, low modulus or high elongation are needed
(Tables 4.3, 4.12, and 4.13).

Pellethane™ polyurethane elastomers

The 2363 series Pellethane™ TPE elastomers have a wide range of durometer hard-
ness and are noted for their high tensile and tear strength and abrasion resistance.
Chemically they are classed as polytetramethylene glycol ether polyurethanes. The
ether series is the most widely used for medical applications although polyester ver-
sions of Pellethane™ are useful for some applications. None of these polymers have
the disadvantage of containing plasticizers which can migrate out of the polymer
over time resulting in reduction in physical properties. Medical tubing made from
Pellethane™ polymer is widely used. These TPEs are unaffected by ethylene oxide
gas, gamma radiation and electron beam sterilization procedures. Pellethane™ can
be processed by injection molding and extrusion. For details on physical properties,
processing and biocompatibility (Tables 4.3, 4.12, and 4.13).
302 J.W. Boretos and S.J. Boretos

Table 4.6 Typical Properties of Plasticized Polyvinyl Chloride

Property
Tensile Modulus
Specific Durometer strength, Elongation, ASTM
gravity shore psi percent D-412
Product and ASTM ASTM ASTM ASTM psi %
Manufacturer Product no. D-792 D-2240 D-412 D-412
3511TX-02 1.12 35 A 1110 525 235 100
Elastichem™ 4011TX-02 1.16 40 A 1300 500 266 100
PVC, 5011TX-02 1.16 50 A 1650 465 426 100
Colorite 5511TX-02 1.18 55 A 1790 465 455 100
Plastics Co. 6011TX-02 1.18 60 A 1936 465 488 100
7011TX-02 1.21 70 A 2667 400 952 100
7511TX-02 1.22 75 A 3000 360 1400 100
8011TX-02 1.23 80 A 3646 330 2025 100
0-1234 1.21 58 A 1400 400 600 100
Ellay™ PVC, 0-1290 1.26 83 A 2750 275 1500 100
0-1541 1.23 81 A 2400 300 1400 100
0-1554 1.21 70 A 2000 400 950 100
Ellay, Inc. 0-2112 1.24 82 A 2650 320 1200 100
0-2129 1.24 83 A 2670 310 1500 100
0-2202 1.54 75 A 2360 270 1190 100
0-2609 1.20 68 A 1950 410 800 100
0-2610 1.24 83 A 2460 295 1450 100
0-2623 1.24 82 A 2550 325 1350 100
0-2631 1.19 65 A 1800 390 650 100
0-3110 1.22 74 A 2100 355 1000 100
0-3115R 1.20 68 A 1900 400 800 100
0-3119 1.22 75 A 2150 350 1100 100
0-3138R 1.22 75 A 2200 350 1075 100
0-3140R 1.25 87 A 2850 330 1600 100
0-3147 1.28 95 A 3100 250 2350 100
0-3149R 1.23 78 A 2400 340 1150 100
0-3154 1.19 65 A 1750 410 725 100
0-3155R 1.20 68 A 1850 390 780 100
0-3166R 1.25 85 A 2700 320 1650 100
0-3195 1.27 90 A 2950 280 2210 100
0-3200 1.18 60 A 1600 450 525 100
Ellay™ PVC, 0-3201 1.21 70 A 2000 340 800 100
0-3224R 1.21 77 A 2300 345 1100 100
0-3231R 1.26 88 A 3000 280 1800 100
Ellay, Inc. 0-4106R 1.25 85 A 2650 300 1600 100
0-4108 1.25 85 A 2750 300 1600 100
0-4109R 1.25 85 A 2800 310 1700 100
0-4113 1.31 100 A 3960 184 3200 100
(continued)
4 Biomedical elastomers 303

Table 4.6 (continued)

Property
Tensile Modulus
Specific Durometer strength, Elongation, ASTM
gravity shore psi percent D-412
Product and ASTM ASTM ASTM ASTM psi %
Manufacturer Product no. D-792 D-2240 D-412 D-412
0-4114 1.20 67 A 1900 400 780 100
0-4115 1.26 87 A 2800 295 1650 100
0-4116R 1.27 90 A 2950 265 2100 100
Ellay, Inc. 0-4120 1.21 68 A 2180 400 830 100
0-4121 1.23 81 A 2550 325 1400 100
0-4122 1.33 110 A 4500 135 4180 100
0-4124R 1.28 95 A 3050 250 2200 100
0-4125 1.24 80 A 3150 355 1310 100
0-4129 1.18 63 A 1670 430 620 100
0-4132 1.21 70 A 2000 395 900 100
0-4135 1.23 80 A 2550 320 1260 100
0-4140 1.23 80 A 2500 330 1250 100
0-4150 1.26 88 A 2900 290 2200 100
0-5210C 1.26 82 A 2300 225 1250 100
BB-69 1.23 78 A 2200 340 1150 100
EH-222C 1.21 70 A 2050 365 1100 100
ES-2967ZPH 1.22 75 A 2300 360 1200 100
Geon® PVC, 121AR 1.4 2800 380
213 1.4 2205 379 1010 100
B. F. Goodrich
Co. 250x100 1700- 430-460 400-500 100
1850
Multichem™ 6014 1.15 60 A 1640 540 400 100
PVC, 7014 1.19 70 A 2040 600 625 100
Colorite 8014 1.22 80 A 2100 500 1000 100
Plastics Co. 8514 1.24 85 A 2250 530 880 100
3300-45 NT 1.13 45 A 1100 480 325 100
Teknor™ 3300-50 NT 1.14 50 A 1220 460 370 100
PVC, 3300-55 NT 1.16 55 A 1500 520 100
3300-60 NT 1.17 60 A 1550 450 560 100
Teknor Apex 3300-68 NT 1.18 68 A 1850 430 690 100
Co. 3300-75 NT 1.20 75 A 2150 420 900 100
3300-80 NT 1.21 80 A 2400 1,320 100
3300-85 NT 1.23 85 A 2800 380 1,560 100
3300-90 NT 1.25 90 A 3100 340 2,100 100
3310-50 NT 1.35 50 A 1000 430 330 100
3310-55 NT 1.35 55 A 1100 410 400 100
3310-60 NT 1.35 60 A 1300 400 480 100
(continued)
304 J.W. Boretos and S.J. Boretos

Table 4.6 (continued)

Property
Tensile Modulus
Specific Durometer strength, Elongation, ASTM
gravity shore psi percent D-412
Product and ASTM ASTM ASTM ASTM psi %
Manufacturer Product no. D-792 D-2240 D-412 D-412
3310-65 NT 1.35 65 A 1500 390 590 100
3310-70 NT 1.35 70 A 1770 380 700 100
3310-75 NT 1.35 75 A 1900 370 800 100
3310-80 NT 1.35 80 A 2200 360 1,050 100
3310-85 NT 1.35 85 A 2500 340 1,470 100
3310-90 NT 1.35 90 A 2900 330 1,900 100
Teknor Apex 90A471R-60NT 1.16 60 A 1500 450
Co. 90A471R-65NT 1.17 65 A 1750 440
90A471R-70NT 1.18 70 A 1900 430
90A471R-75NT 1.20 75 A 2150 420
90A471R-80NT 1.23 80 A 2690 380
90A471R-85NT 1.23 85 A 2800
90A471R-90NT 1.27 90 A 3350 360

Notice Regarding Long-Term Medical Implant Applications

The Dow Chemical Company does not recommend Pellethane™ elastomers for long-term
medical implant applications in humans (more than 30 days). Nor do they recommend the
use of Pellethane™ elastomers for cardiac prosthetic devices regardless of the time period
that the device will be wholly or partially implanted in the body. Such applications include,
but are not limited to, pacemaker leads and devices, cardiac prosthetic devices such as
artificial hearts, heart valves, intra-aortic balloon and control systems, and ventricular
bypass assist devices. The company does not recommend any non-medical resin (or film)
product for use in any human implant applications.

PolyBlend™ polyurethane
This TPE has been described as an aromatic elastoplastic polyurethane alloy. It pos-
sesses a low coefficient of friction, low extractables, and dimensional stability.
Hardness ranges from 65 to 75 Shore D. The material is classified for short-term (29
days or less) implantation. Clear and radiopaque formulations are available. Tubing
should be annealed at 80°C for four hours to reduce crystallinity (Tables 4.3, 4.4,
4.12, and 4.14).

Tecoflex® polyurethane
Tecoflex is an aliphatic polyether-based polyurethane that is available in clear and
radiopaque grades. They are reaction products of methylene bis (cyclohexyl) diiso-
cyanate (HMDI), poly (tetramethylene ether glycol) (PTMEG), and 1,4 butane diol
chain extender. The manufacturer claims that the aliphatic composition of Tecoflex®
4 Biomedical elastomers 305

eliminates the danger of forming methylene dianiline (MDA) which is potentially

carcinogenic. MDA can be generated from aromatic polyurethanes if they are
improperly processed or overheated. Tecoflex has been reported to crack under
stress when implanted, long-term, in animals. An advantage of Tecoflex is that it
softens considerably within minutes of insertion in the body. This feature can offer
patient comfort for short-term applications such as catheters and enteral tubes; it is
also reported to reduce the risk of vascular trauma (Tables 4.3, 4.12, and 4.13).

Tecothane®
Tecothane® is an aromatic polyether-based TPE polyurethane polymer. It has pro-
cessibility and biocompatibility characteristics similar to Tecoflex® except that it is
an aromatic rather than an aliphatic polyurethane. Tecothane® is synthesized from
methylene diisocyanate (MDI), polytetramethylene ether glycol and 1,4 butanediol
chain extender. By varying the ratios of the reactants, polymers have been prepared
ranging from soft elastomers to rigid plastics. The manufacturer of Tecoflex® and
Tecothane® point out that there is not much difference between medical-grade, ali-
phatic and aromatic polyether-based polyurethanes with regard to chemical,
mechanical and biological properties. However, they caution that with improper pro-
cessing of Tecothane® (e.g., high moisture content or steam sterilization) it is pos-
sible to form measurable amounts of methylene dianiline (MDA), a listed carcinogen.
The use of ethylene oxide or gamma radiation are suitable sterilizing agents that do
not affect the chemical or physical properties (Tables 4.3, 4.12, and 4.13).

Texin™
There are four basic polymer formulations of Texin polyurethane TPE that may be
suitable for medical applications. They range in hardness and flexural modulus.
Texin elastomers are produced by the reaction of diisocyanate with a high molecular
weight polyester or polyether polymer and a low molecular weight diol. The poly-
ethers (products 5286 and 5265) offer greater hydrolytic stability and stress crack
resistance. The polyesterbased polyurethane (product 5187) and the polyester poly-
urethane/ polycarbonate blend (product 5370) possess high impact strength and
high stiffness along with useful low-temperature properties. Texin is not recom-
mended for implants of greater than 30 days duration. Texin should not be sterilized
by autoclave or use of boiling water. Other advantages offered by Texin TPUs are
that plasticizers are not necessary to achieve flexibility, the amount of extractables
are low, and they possess high tensile strength, high tear strength, and high abrasion
resistance. Texin polyurethanes are hydroscopic and will absorb ambient moisture.
They can be processed by extrusion and injection molding if thoroughly dried
beforehand. As with all chemical systems, the proper use and handling of these
materials can not be over-emphasized (Tables 4.3, 4.12, and 4.13).
Texin™ 5370 is a blend of polyester-based polyurethane and polycarbonate. It
offers high impact strength and high stiffness. Steam sterilization or boiling should
be avoided (Tables 4.3, 4.12, and 4.13).
Table 4.7 Typical Properties of Styrene-based Thermoplastic Elastomers
306

Property
Durometer Tensile Modulus
Specific hardness strength, Elongation ASTM Tear strength Compression set,
gravity shore psi percent D-412 pli,die B percent
Product and ASTM ASTM ASTM ASTM ASTM ASTM
Manufacturer Product no. D-792 D-2240 D-412 D-412 psi % D-624 D-395
R70-001 0.90 50 A 1200 900 150 100 16
C-Flex®, R70-003 0.90 70 A 1280 760 340 100 25
R70-005 0.90 30 A 1400 950 100 100 11
R70-026 0.90 90 A 1830 650 1,010 100
Consolidated R70-028 0.90 35 A 990 800 120 100 13
Polymer R70-046 0.90 34 A 1320 940 110 100 135 12
Technologies, Inc. R70-050 0.90 48 A 1250 880 170 100 100 18
R70-051 0.90 74 A 1140 680 370 100 150 28
R70-058 0.94 70 A 2080 660 300 100 120 55
R70-057 0.92 40 A 1220 890 100 100 90 33
R70-068 0.93 50 A 1630 850 140 100 110 38
R70-072 0.90 60 A 1270 780 240 100 20
R70-081 0.90 45 A 1440 920 120 100 17
R70-082 0.90 61 A 1270 860 230 100 130 19
R70-085 0.90 50 A 1380 750 200 100 17
R70-089 0.90 45 A 1640 700
R70-091 0.90 50 A 1280 780 130 100
R70-116 0.90 30 A 1105 810 100 100 84 24
R70-190 0.90 5 A 270 1010 20 100
R70-214 0.90 18 A 450 780
D-2103 0.94 70 A 4300 880 400 300 205
J.W. Boretos and S.J. Boretos
Kraton®, D-2104 0.93 27 A 1700 1350 200 300 180
D-2109 0.94 44 A 950 800 300 300 160
Shell G-2701 0.90 67 A 1600 800 480 300 260
Chemical G-2703 0.90 63 A 1200 670 470 300 230
Co. G-2705 0.90 55 A 850 700 400 300 140 38
G-2706 0.90 28 A 850 950 130 300 140
G-2712 0.88 42 A 840 820 250 300 140
4 Biomedical elastomers
307
308 J.W. Boretos and S.J. Boretos

Polycarbonate-based polyurethanes
Carbothane™
This medical grade TPE polyurethane is the reaction product of an aliphatic diiso-
cyanate, a polycarbonate-based macrodiol, and a chain terminating low molecular
weight diol (Tables 4.4, 4.12, and 4.13).

ChronoFlex™ AR.
Available as a dimethyl acetamide solution, this segmented, aromatic, polycarbonate-
based TPE polyurethane was designed to mimic Ethicon Corporation’s Biomer. The
polymer is made from the addition of diphenylmethane 4,4’-diisocyanate to a poly-
carbonate diol followed by addition of a mixture of chain extenders and a molecular
weight regulator. The polymer is believed to be resistant to environmental stress
cracking such as that experienced by other polyurethanes coated onto pacemaker
leads (Tables 4.4, 4.12, and 4.13).

Coremer™
Specifically designed as an 80 Shore A durometer TPE, this is a diamine chain
extended version of Corethane®. Coremer™ solution cast films have a low initial
modulus and high flex fatigue life. Information as to long-term biostability is not
available at this time (Tables 4.4 and 4.13).

Corethane®
A polycarbonate TPE polyurethane that claims biostability is achieved through its
replacement of virtually all ether or ester linkages with carbonate groups. The soft
segment is composed of a polycarbonate diol formed by the condensation reaction
of 1,6-hexanediol with ethylene carbonate. The polycarbonate diol is converted to a
high molecular weight polyurethane by the reaction with 1,4-methylene bisphenyl
diisocyanate (MDI) and 1,4-butanediol. It is reported to be resistant to environmen-
tal stress cracking as experienced with insulation on pacemaker lead wires. The
polymer can be extruded, injection molded or compression-molded, and can be
bonded with conventional urethane adhesives and solvents (Tables 4.4, 4.12, 4.13,
and 4.14).

Corhesive™
Corhesive™ is a solvent-free, two-component reaction adhesive system for use
with polyurethanes, plasma treated silicones and certain metals (Tables 4.4, 4.12,
4.13, and 4.14).

Polypropylene-based elastomers Sarlink®

This is a polypropylene-based TPE that has been used as a replacement for medical
stoppers previously made from butyl rubber. Sarlink® has the characteristics typical
of rubber vulcanizates such as elasticity, flexibility, high coefficients of friction and
softness. Sarlink® combines gas impermeability without concern for contamination
of biological medium. Applications for medical grade Sarlink® are inserts on syringe
plungers, reusable injection caps, vacuum assisted blood sampling tubes, plus
4 Biomedical elastomers 309

flexible grade tubing. The number of stoppers produced from Sarlink annually num-
ber in the billions. The material can be injection molded, blow molded, extruded,
calendered, and thermoformed on standard processing equipment. It can be thermal
bonded or adhesive bonded (Tables 4.5, 4.12, and 4.13).

Polyvinyl chloride elastomers

Polyvinyl chloride polymer is polymerized from vinyl chloride monomers. It is a
hard material which can be made soft and flexible through the addition of a plasti-
cizer or a copolymer. As such, it resembles an elastomer and can be included with
other TPEs. Also optionally added to PVC are fillers, stabilizers, antioxidants and
others. A typical PVC plasticizer for medical products is di(2-ethylhexyl) phthalate
(also known as dioctyl phthalate, DOP). Some producers of PVC also offer non-
phthalate formulations. PVC has been used extensively for blood bags, blood tub-
ing, endotracheal tubes, catheters and fittings, urology tubes, intravenous tubing,
respiratory devices and dialysis sets. Leaching of the plasticizer can offer difficulties
if the application is not short-term. Medical grade PVC is available from
B.F. Goodrich under the name Geon® RX, Elastichem™ PVC, Ellay™ PVC,
Multichem™ PVC, Teknor™ PVC, AlphaGary and others. PVC polymers have also
been incorporated as additives to polyurethane to alter the properties of the latter.

Elastichem™ PVC.
This polyvinyl chloride compound family is highly elastomeric and exhibits a dry
non-tacky surface even at hardnesses as low as 40 Shore A durometer. Their rubber-
like resilience, high elongation and low permanent set and fatigue resistance offer
advantages over conventional formulations (Tables 4.6, 4.12, and 4.13).

Ellay™ PVC.
Compounds from Ellay Corp. are available with Shore hardness ranges from 55 A
to 100 A. The polymers have been applied to medication delivery systems, blood
collection, processing and storage, gastro-urological devices and collection sys-
tems. Product numbers ending in ‘R’ are special radiation resistant grades (Tables
4.6, 4.12 and 4.13).

Geon® PVC.
Geon® PVC is associated with vinyl examination gloves. For this use, Geon® recom-
mends a combination of Geon® 121 AR and 213. For a more ‘latex type’ feeling,
Goodtouch 250x100 is recommended. Typical film samples have passed patch insult
tests when worn against the skin for extended periods (Tables 4.6, 4.12 and 4.13).

Multichem™ PVC
This line of PVC polymers consist of alloys of PVC in combination with other poly-
mers. They display notable dynamic properties and resistance to migration and
extraction. These non-toxic PVC compounds (includes Multichem™ and
Elastichem™) have over 25 years of experience in the medical field (Tables 4.6,
4.12 and 4.13).
310 J.W. Boretos and S.J. Boretos

Teknor™ Apex PVC

This extrudable PVC has found use as tubing for blood transport and delivery sys-
tems, dialysis and enteral feeding systems, oxygen delivery systems, catheters, and
drainage systems. Product numbers containing an R are special radiation resistant
grades (Tables 4.6, 4.12 and 4.13).

Styrene-based elastomers
C-Flex®TPE.
C-Flex® thermoplastic elastomers are based on styrene/ethylene-butylene/styrene
block copolymers. C-Flex® polymers designated as ‘medical grade’ are clear and
can be processed using conventional extrusion and injection molding equipment.
They have been tested using Good Laboratory Practices and have successfully
passed USP Class VI, biocompatibility tests. Translucent versions have high
rebound values at ultimate elongation. Medical tubing, ureteral stents, blood pumps,
feeding tubes and nephrostomy catheters are successful uses of this material (Tables
4.7, 4.12 and 4.13).

Kraton®
Kraton® elastomer consists of block segments of styrene and rubber monomers and
are available as Kraton® D and G series. The D series is based on unsaturated mid-
block styrene-butadiene-styrene copolymers whereas the G series is based on sty-
rene-ethylene/butylene-styrene copolymers with a stable saturated midblock. Listed
among the attributes of both series are such features as low extractables, dimen-
sional stability, vapor and gas transmission properties, ease of sterilization, softness
and clarity. They exhibit elastomeric flexibility coupled with thermoplastic proces-
sibility (Tables 4.7, 4.12, 4.13).

4.2.2 crosslinked elastomers

Natural rubber
Natural rubber (cis-polyisoprene) is strong and one of the most flexible of the elas-
tomers. The material has been used for surgeon's gloves, catheters, urinary drains
and vial stoppers. However, because it has the potential to cause allergic reactions
thought to be due to the elution of entrapped natural protein, this elastomer is being
used less now than in the past. Safer substitutes are being selected.

Silicone elastomers
Silicone elastomers have a long history of use in the medical field. They have been
applied to cannulas, catheters, drainage tubes, balloon catheters, finger and toe
joints, pacemaker lead wire insulation, components of artificial heart valves, breast
implants, intraocular lenses, contraceptive devices, burn dressings and a variety of
4 Biomedical elastomers 311

associated medical devices. A silicone reference material has been made available
by the National Institutes of Health to equate the blood compatibility of different
surfaces for vascular applications. This material is available as a silica-free sheet.
Contact the Artificial Heart Program, NHBLI, NIH, Bethesda, Md. for further
information.
The silicone elastomers most commonly used for medical applications are the
high consistency (HC) and liquid injection molding (LIM) types. The former is
most often peroxide cured and the latter platinum cured although there are varia-
tions. Both materials are similar in properties. LIM offers greater advantages to the
medical device molder and is gaining in popularity. This form of silicone may
become the molder’s material of choice within the next few years.

High consistency (HC) silicone elastomer

High consistency silicone elastomer consists of methyl and vinyl substituted sili-
cones with aromatic and fluorinated functional groups in some formulations. For
the most part, they are peroxide crosslinked. Items are usually compression or trans-
fer molded (Tables 4.8).

Liquid injection molding (LIM) silicone elastomer

Liquid injection molding (LIM) with liquid silicone rubber (LSR) is fast becoming
the technique of choice for processing silicone elastomers. Modifications of con-
ventional injection molding equipment are required. For example, pumps to handle
two components being injected simultaneously are required. The heaters on the
injection barrel and nozzle are replaced by water cooled jackets. The mold is heated
in the range of 300 to 400°F. Because the (LSR) flows easily, injection pressures are
low (800 to 3000 psi). Elastomeric items cure rapidly in the mold (e.g., a 7 gram
part will crosslink in about 15 seconds at 350 °F). Many formulations rely on plati-
num as a crosslinker. Perhaps in the future, the majority of silicone rubber molded
parts will be made in this fashion. Appropriate equipment is commercially
available.
Tables 4.8, 4.9, 4.10 and 4.11 list the silicones made by Applied Silicone Corp.,
Dow Corning Corp., and NuSil Technologies. Table 4.12 lists their biocompatibility
status and Table 4.13 recommended sterilization methods. Dow Corning no longer
offers the following materials for general sale:
• Silastic MDX 4–4515
• Silastic MDX 4–4515
• Silastic Q7–2245
• Dow Corning Q7–2213
Further, they have discontinued the sale of all implant grade materials.

Other silicones
Silicones and polyurethanes have been used to produce denture liner materials and
maxillofacial prostheses. Most of these materials are silicone based, e.g., Flexibase,
312 J.W. Boretos and S.J. Boretos

Table 4.8 Typical Properties of High Consistency (HC) Silicone Elastomers

Property
Tear
Specific Durometer Tensile Elongation, strength
Product gravity hardness, strength, percent pli, die B
and ASTM shore ASTM psi ASTM ASTM ASTM
Manufacturer Product no. D-792 D-2240 D-412 D-412 D-624
40039 1.12 35 A 1600 1200 200
Applied 40040 1.15 50 A 1500 900 220
Silicone
Medical 40041 1.20 66 A 1200 900 260
Implant
Grade, 40042 1.20 78 A 1200 600 280
Applied 40043 1.12 23 A 1100 1500 160
Silicone
Corp. 40044 1.12 33 A 1600 1015 150
40045 1.15 51 A 1400 600 190
40046 1.20 66 A 1200 500 250
40063 1.20 70 A 1400 850 280
MED-2174 1.15 52 A 1200 715 200
NuSil MED-2245 1.13 41 A 1300 700 140
Silicone,
MED-4515 1.15 52 A 1350 450 90
NuSil MED 4516 1.21 72 A 1175 370 80
Technology
MED-4735 1.10 35 A 1310 1250 200
MED 4750 1.15 50 A 1350 810 230
MED 4755 1.14 57 A 1375 800 300
MED 4765 1.20 65 A 1100 900 240
MED-4770 1.17 70 A 1375 700 300
MDX4-4210 1.10 25 A 550 350 50
Silastic ® Q7-4535 1.10 33 A 1200 1015 160
Medical Q7-4550 1.14 51 A 1375 600 170
Materials,
Q7-4565 1.20 66 A 1000 550 210
Dow Corning Q7-4720 1.10 23 A 1200 1100 150
Corp. Q7-4735 1.10 35 A 1050 1200 200
Q7-4750 1.14 50 A 1300 900 230
Q7-4765 1.14 50 A 1300 900 230
Q7-4780 1.22 78 A 850 600 190
4 Biomedical elastomers 313

Table 4.9 Typical Properties of Liquid Injection Molding (LIM) Silicone Elastomers
Property
Tear
Specific Durometer Tensile Elongation, strength
Product gravity hardness, strength, percent pli, die B
and ASTM shore ASTM psi ASTM ASTM ASTM
Manufacturer Product no. D-792 D-2240 D-412 D-412 D-624
40023 1.11 10 A 500 750 80
Applied 40024 1.11 20 A 800 600 140
Silicone
Medical 40025 1.12 30 A 950 600 150
Implant
Grade, 40026 1.12 40 A 980 450 170
Applied 40027 1.13 50 A 1000 400 190
Silicone
Corp. 40028 1.13 60 A 1100 350 220
40029 1.10 30 A 900 300 80
40071 1.14 70 A 1200 350 220
40072 1.10 25 A 650 400 60
40082 1.10 40 A 900 250 110
NuSil MED-6210 1.04 50 A 1000 100 35
Silicone,
MED-6233 1.03 50 A 1200 300 75
NuSil MED-6382 1.13 45 A 400 200
Technology
MED-6820 1.05 40 A 750 125 25
Silastic® Q7-4840 1.12 40 A 950 425 150
Medical Q7-4850 1.14 50 A 1350 550 225
Materials,
Dow Corning Q7-6860 1.16 60 A 1300 450 250
Corp.

Molloplast-B, Prolastic, RS 330 T-RTV, Coe-Soft, Coe-Super Soft, Vertex Soft,

PERform Soft, and Petal Soft. Other custom made elastomers have been applied to
maxillofacial prostheses, e.g., Cosmesil, Silastic® 4-4210, Silastic® 4-4515, Silicone
A-102, Silicone A-2186, Silskin II, Isophorone polyurethane, and Epithane-3.
Denture liners with acrylic and silicone include Coe-Soft, Coe Super-Soft, Vertex
Soft, Molloplast-B and Flexibase.

Dispersions
Solvent solutions of polyurethane elastomers and silicone elastomers are given in
Table 4.10. These materials are helpful in casting thin films and odd or complex
shapes.
 314

Table 4.10 Typical Properties of Elastomeric Dispersions

Property
Specific Durometer Tensile strength, Elongation Modulus ASTM Tear strength
gravity hardness psi percent D-412 pli,die B
Product ASTM ASTM ASTM ASTM ASTM
and Manufacturer Product no. D-792 D-2240 D-412 D-412 psi % D-624
40000 1.10 35 A 1800 800 185
Applied Silicone 40001 1.18 32 A 1200 200
Medical Implant 40002 1.08 24 A 800 700 60
Grade, 40016 1.10 35 A 1800 800 185
Applied Silicone 40021 1.08 24 A 1000 100
Corp. 40032 1.19 40 A 500 120
MED-2213 1.13 Shore 00,82 1300 700 190 200 140
NuSil Silicone, MED-6400 1.08 32 A 1250 325 300 150
MED2-6400 1.08 32 A 1250 800
NuSil Technology MED-6600 1.10 20 A 1000 275 300 90
MED2-6600 1.10 25 A 1000 750 325 300
MED-6605 1.08 25 A 900 1000 75 175 100
MED3-6605 1.08 25 A 900 1000 100 200 125
MED-6607 1.10 40 A 900 650 - 130
MED-6640 1.12 30 A 1650 1100 150 100 280
MED2-6640 1.12 30 A 1750 100 100 275
MED2-6650 1.15 35 A 1100 750 200 300
Silastic ® Q7-2630 - Shore 00, 70 800 900 50 200 -
Medical Materials,
Dow Corning
Corp.
J.W. Boretos and S.J. Boretos
Product no. Form viscosity cp. Solvent System Cure System Chemical Type
Used
40000 35% solids, 2000 Xylene Platinum Methyl vinyl siloxane
1 part addition
40001 Phenyl vinyl
40002 32% solids, 500 Acetoxy siloxane Dimethyl siloxane
1 part
40016 27% solids, 2000 1,1,1 Platinum Methyl vinyl siloxane
4 Biomedical elastomers

1 part trichloroethane addition

40021 32% solids, 500 Xylene Acetoxy Dimethyl siloxane
1 part
40032 21 % solids, 800 1,1,1 Platinum Fluorovinyl methyl siloxane
1 part trichloroethane addition
MED-2213 15% solids, 7000 1,1,1 Dimethyl-methyl vinyl siloxane
1 part trichloroethane
MED-6400 35% solids, 600 Xylene Platinum
2 part, 1:1 addition
MED2-6400 25% solids, 2 part, 800 1,1,1 Vinyl methyl siloxane
1:1 trichloro-ethane
MED-6600 35% solids, 2 part 300 Xylene
MED2-6600 1600 1,1,1
trichloroethane
MED-6605 30% solids, 1 part 800 Xylene Acetoxy
MED3-6605 22% solids, 1 part 1250 1,1,1 Dimethyl siloxane
trichloroethane
MED-6607 33% solids, 1 part 5500 VM&P naphtha Oxime
MED-6640 25% solids, 2 part 7000 Xylene Methyl vinyl siloxane
MED2-6640 15% solids, 2 part 5000 1,1,1 Platinum
trichloroethane addition
MED2-6650 20% solids, 2 part 3000 Fluorovinyl methyl siloxane
315

Q7-263010% solids - Q7-2650 Acetoxy Dimethyl siloxane

316 J.W. Boretos and S.J. Boretos

Table 4.11 Typical Properties of Silicone Elastomeric Adhesives

Property
Adhesive
Durometer Tensile Tear strength
Specific hardness strength, Elongation, strength, (to silicone)
gravity Shore Psi Percent die B, pli Pli
Product
and Product ASTM ASTM ASTM ASTM ASTM
Manufacturer no. D-792 D-2240 D-412 D-412 D-624 pli
Applied 40064 18+
Silicone
Medical Medical 1.08 24 A 850 750 70 18
Implant Grade
Grade, RTV
Silicone
Adhesive
Applied Medical 950 770 18
Silicone Grade
Corp. High
Strength
RTV
Silicone
Adhesive
NuSil MED- 1.07 29 A 550 450
Silicone, 1137
NuSil
Technology
Silastic® Medical 1.06 29 A 450 400 30 20+
Medical Adhesive
Materials A
Dow 355 Medical Grade Pressure 1.40
Corning Sensitive
Corp.
Product no. Cure Conditions Comments
40064 Produces acetic acid. Cures @ Bonds silicones to each other and some
RT with atmospheric moisture, synthetics, metals.
20 to 60% RH.
Medical 24 hours @ 25°C, aged Bonding silicone to polyester, etc.
Grade RTV 24 hours @ RT. High strength bonds to
Silicone polyester, nylon, polyurethane
Adhesive and metals.
Medical
Grade High
Strength RTV
Silicone
Adhesive
MED-1137 Produces acetic acid. Bonding silicones to each other & some
Cure 3 days @ RT with synthetics/metals.
atmospheric moisture, 20 to When fully cured resembles some conventional
60% RH. silicone elastomers.
(continued)
4 Biomedical elastomers 317

Table 4.11 (continued)

Property
Adhesive
Durometer Tensile Tear strength
Specific hardness strength, Elongation, strength, (to silicone)
gravity Shore Psi Percent die B, pli Pli
Product
and Product ASTM ASTM ASTM ASTM ASTM
Manufacturer no. D-792 D-2240 D-412 D-412 D-624 pli
Medical Produces acetic acid, requires Bonding silicone rubber to itself. Useful for
Adhesive A 50% RH & 7 days to cure. cast films or parts from dispersions.
355 Medical Non-curing dispersion Adheres to skin for use with
Grade - becomes adhesive as solvent ileostomy and colostomy
Pressure evaporates. appliances.
Sensitive

4.3 Establishing Equivalence

Specific polymeric materials traditionally used for medical applications have been
recently withdrawn from the medical market. Silicone elastomers are among those
withdrawn. To maintain continued supply of vital implants, methods of determining
equivalence for withdrawn elastomers with new or existing ones has been adopted
by the FDA in the form of an FDA Guidance Document.

4.3.1 FDA Guidance document for substitution

of equivalent elastomers

The FDA will allow manufacturers to change sources of silicone elastomers (and
others) if they can show that the replacement material is ‘not substantially differ-
ent’ from materials described in existing approved applications. The device manu-
facturer is still required to certify that the processes of fabrication, cure and
sterilization it uses in the manufacture of its device are appropriate for the new
material and that the device will perform as intended. Premarket notification sub-
mission under section 510(k) of the Federal Food, Drug, and Cosmetic Act (21
USC 360(k) and 21 CFR 807.81(a)(3)(i), or a supplemental premarket approval
application under 21 USC 360(k) section 515 and 21 CFR 814.39 is necessary
when change could significantly affect the safety or effectiveness of the device.
These submissions are required to be submitted and approved before the device
may be marketed with the change.
There are a number of tests necessary for comparison of silicone elastomers as
indicated by ‘Guidance for Manufacturers of Silicone Devices Affected by
Withdrawal of Dow Corning Silastic® Materials’ (Federal Register, Vol. 58, No.
127, Tuesday, July 6, 1993/ Notices, 36207). They compare the physical, chemical
and biological properties of the bulk polymers as they are received from the supplier
and also compare the molded elastomer as it exists in the final medical device.
Table 4.12 Biocompatibility of Various Elastomers
318

Biocompatibility Status*
Product and Intracutaneous Systemic Skin
Classification Manufacturer Product no. Hemolysis Pyrogenicity Injection Injection Sensitization
Santoprene® 281-45 passed passed passed passed passed
Thermoplastic Rubber, 281-55 passed passed passed passed passed
elastomer 281-64 passed passed passed passed passed
Advanced 281-73 passed passed passed passed passed
Elastomer 281-87 passed passed passed passed passed
Systems 283-40 passed passed passed passed passed
PCCE Ecdel™ Elastomer, 9965 passed passed passed
copolyester Eastman Chemical 9966
Co.
elastomer Biospan® Segmented,
Polyurethane- Polyurethane, Biospan® passed passed passed passed passed
based The Polymer
elastomers Technology Group,
Inc.
Hydrothane™,
Poly Medica Hydrothane™
Biomaterials, Inc.
MP-5000
Medicaflex™, MF-5001 passed passed passed
MF-5040
Advanced Resin MF-5041
Technology MF-5056
MF-5057
J.W. Boretos and S.J. Boretos
 MF-5062
Pellethane™ Pellethane™
2363 series, 2363 series passed passed passed
Dow Chemical Co.
PolyBlend™ 1000 , PolyBlend™
and PolyBlend™ 1000 and
1100 PolyBlend™
4 Biomedical elastomers

Poly Medica 1100

Biomaterials, Inc.
Biocompatibility Status*
Intramuscular Tissue Cell
Product no. 10 days 30 days 90 days Culture Comments
281-45
281-55 passed passed passed Passed USP
Class VI
testing,
281-64 Tripartite
testing, mouse
embryo
281-73 toxicity testing
and Ames
281-87 Mutagenicity
testing.
283-40
9965 passed
9966
Biospan® passed
(continued)
319
Table 4.12 (continued)
320

Hydrothane™ See text for

status.
MF-5000 passed passed
MF-5001 passed passed Passed USP
Class VI
testing.
MF-5040 passed
MF-5041 passed passed
MF-5056 passed
MF-5057 passed
MF-5062
Peliethane™ Passed USP
Class VI
testing.
2363 series passed passed See text for
status.
PolyBlend™ See text for
status.
1000 and
PolyBlend™
Biocompatibility Status*
Classification Product and Product no. Hemolysis Intracutaneous Systemic Skin
Manufacturer Sensitization
Pyrogenicity Injection Injection Sensitization
Polyurethane based Tecoflex® EG60A passed passed passed passed
elastomers and EG80D
Tecothane® 1055D
1065D
J.W. Boretos and S.J. Boretos
 1074A passed passed passed
Thermedics, Inc. 1075D
1085A
1095A
Texin™,
Texin™ passed passed passed
Miles, Inc.
4 Biomedical elastomers

Polycarbonate based Carbothane TM, PC-3555D passed passed passed

polyurethanes Thermedics, Inc. PC-3572D
PC-3575A
PC-3585A
PC-3595A
ChronoFlex™ AR,
Poly Medica ChronoFlex™ passed passed passed passed
Biomaterials, Inc. AR
Coremer™, Coremer™
Corethane®, TPE 55D passed passed passed
and TPE 75D passed
TPE 80A passed passed passed passed
Corhesive™, Corhesive™, passed passed passed passed passed
Corvita Corp.
Polypropylene based Sarlink® medical Sarlink® medical
elastomers grade, DSM grade
Thermoplastic
Elastomers, Inc.
(continued)
321
Table 4.12 (continued)
322

Biocompatibility Status*
Intramuscular Tissue Cell
Product no. 10 days 30 days 90 days Culture Comments
EG60A
EG80D
1055D
1065D
1074A
1075D
1085A
1095A
Texin™ passed passed Passed USP
Class VI
testing.
See text for
status.
PC-3555D
PC-3572D
PC-3575A
PC-3585A
PC-3595A
ChronoFlex™ passed passed Passed USP
AR Class VI
testing.
Coremer™ See text for
status.
TPE 55D passed passed passed passed
TPE 75D passed
TPE 80A passed passed passed passed
J.W. Boretos and S.J. Boretos
Corhesive™ passed passed passed passed
Sarlink® See text for
medical grade status.
Biocompatibility Status*
Classification Product and Product no. Hemolysis Intracutaneous Systemic Skin
Manufacturer Pyrogenicity Injection Injection Sensitization
Polyvinyl chloride Elastichem™PVC, Elastichem™PVC,
elastomers Geon® PVC, B. F. PVC
4 Biomedical elastomers

Goodrich Co.
Plastics Co.
Ellay™ PVC Ellay™ PVC Geon® PVC passed passed passed
Ellay, Inc. Multichem™ PVC
Geon® PVC, B. F.
Goodrich Co.
Multichem™ PVC,
Colorite
Plastics Co.
3300-45 NT
Teknor™ PVC, 3300-50 NT
3300-55 NT
Teknor Apex Co. 3300-60 NT passed passed passed passed
3300-68 NT
3300-75 NT
3300-80 NT
Teknor™ PVC, 3300-85 NT passed passed passed passed
3300-90 NT
(continued)
323
Table 4.12 (continued)
324

Teknor Apex Co. 3310-50 NT

3310-55 NT
3310-60 NT
3310-65 NT
3310-70 NT
3310-75 NT
3310-80 NT
3310-85 NT
3310-90 NT
90A471R-60NT passed passed passed
90A471R-65NT passed passed passed
9OA471R-70NT passed passed passed
90A471R-75NT passed passed passed
90A471R-80NT passed passed passed
90A471R-85NT passed passed passed
90A471R-90NT passed passed passed
Biocompatibility Status*
Intramuscular Tissue Cell
Product no. 10 days 30 days 90 days Culture Comments
Elastichem™ See text for
status.
PVC
Ellay™ PVC passed passed Passed USP
Class VI
testing.
Geon® PVC See text for
status.
Multichem™ See text for
status.
J.W. Boretos and S.J. Boretos

PVC
3300-45 NT passed
3300-50 NT passed
3300-55 NT passed
3300-60 NT passed
3300-68 NT passed
3300-75 NT passed
3300-80 NT passed passed Passed USP
4 Biomedical elastomers

Class VI
testing.
3300-85 NT passed Passed USP
Class VI
testing.
3300-90 NT passed
3310-50 NT passed
3310-55 NT passed
3310-60 NT passed
3310-65 NT passed
3310-70 NT passed
3310-75 NT passed
3310-80 NT passed
3310-85 NT passed
3310-90 NT passed
90A471R-60NT passed Passed USP
Class VI
testing.
90A471R-65NT passed Passed USP
Class VI
testing.
(continued)
325
Table 4.12 (continued)
326

90A471R-70NT passed Passed USP

Class VI
testing.
90A471R-75NT passed Passed USP
Class VI
testing.
90A471R-80NT passed Passed USP
Class VI
testing.
90A471R-85NT passed Passed USP
Class VI
testing.
90A471R-90NT passed Passed USP
Class VI
testing.
Biocompatibility Status*
Classification Product and Product no. Hemolysis Intracutaneous Systemic Skin
Manufacturer Pyrogenicity Injection Injection Sensitization
R70-001 passed passed passed passed
Styrene-based elastomers C-Flex®, R70-003 passed passed passed passed
R70-005 passed passed passed passed
Consolidated R70-026 passed passed passed passed
Polymer
Technologies, Inc.
R70-028 passed passed passed passed
R70-046 passed passed passed passed
R70-050 passed passed passed passed
R70-051 passed passed passed passed
J.W. Boretos and S.J. Boretos

R70-058 passed passed passed passed

 R70-067 passed passed passed passed
R70-068 passed passed passed passed
R70-072 passed passed passed passed
R70-081 passed passed passed passed
R70-082 passed passed passed passed
R70-085 passed passed passed passed
R70-089 passed passed passed passed
4 Biomedical elastomers

R70-091 passed passed passed passed

R70-116 passed passed passed passed
R70-190 passed passed passed passed
R70-214 passed passed passed passed
D-2103 passed passed passed passed
Kraton®, D-2104 passed passed passed passed
D-2109 passed passed passed passed
Shell G-2701 passed passed passed passed
Chemical Co. G-2703 passed passed passed passed
G-2705 passed passed passed passed
G-2706 passed passed passed passed
G-2712 passed passed passed passed
Biocompatibility Status*
Intramuscular Tissue Cell
Product no. 10 days 30 days 90 days Culture Comments
R70-001 passed passed passed passed
R70-003 passed passed passed passed C-Flex testing
data is
available from
R70-005 manufacturer.
R70-026 passed passed passed passed
327

(continued)
Table 4.12 (continued)
328

R70-028 passed passed passed passed

R70-046 passed passed passed passed
R70-050 passed passed passed passed
R70-051 passed passed passed passed
R70-058 passed passed passed passed
R70-067 passed passed passed passed
R70-068 passed passed passed passed
R70-072 passed passed passed passed
R70-081 passed passed passed passed
R70-082 passed passed passed passed
R70-085 passed passed passed passed
R70-089 passed passed passed passed
R70-091 passed passed passed passed
R70-116 passed passed passed passed
R70-190 passed passed passed passed
R70-214 passed passed passed passed
D-2103 passed passed passed passed
D-2104 - passed - passed Passed USP
Class VI
testing.
D-2109 passed passed Passed USP
Class VI
testing.
G-2701 passed passed Passed USP
Class VI
testing.
J.W. Boretos and S.J. Boretos
G-2703 passed passed Passed USP
Class VI
testing.
G-2705 passed passed Passed USP
Class VI
testing.
G-2706 passed passed Passed USP
Class VI
4 Biomedical elastomers

testing.
G-2712 passed passed Passed USP
Class VI
testing.
Biocompatibility Status*
Classification Product and Product no. Hemolysis Intracutaneous Systemic Skin
Manufacturer Pyrogenicity Injection Injection Sensitization
Polydimethylsiloxane Applied Silicone Applied
Medical Implant Silicone
Grade, Applied Medical
Silicone Corp. Implant
Grade
NuSii Silicone, NuSil NuSii Silicone
Technology
MDX4-4210 passed passed passed passed
Silastic ® Q7-4535 passed passed passed passed
Medical Materials, Q7-4550 passed passed passed passed
Q7-4565 passed passed passed passed
Dow Corning Corp. Q7-4720 passed passed passed passed
Q7-4735 passed passed passed passed
Polydimethyl Silastic ® Q7-4750 passed passed passed passed
329

(continued)
Table 4.12 (continued)
330

siloxane Medical Materials, Q7-4765 passed passed passed passed

Q7-4780 passed passed passed passed
Dow Corning Corp. Q7-4840 passed passed passed passed
Q7-4850 passed passed passed passed
Q7-6860 passed passed passed passed passed
Medical
Adhesive A passed passed passed passed
355 Medical passed passed passed passed
Grade Pressure
Sensitive
Biocompatibility Status*
Intramuscular Tissue Cell
Product no. 10 days 30 days 90 days Culture Comments
Applied Applied
Silicone Silicone testing
data is
available from
manufacturer.
Medical
Implant
Grade
NuSii Silicone See text for
status.
MDX4-4210 passed passed passed See text for
status.
Q7-4535 passed passed See text for
status.
J.W. Boretos and S.J. Boretos
Q7-4550 passed passed See text for
status.
Q7-4565 passed passed See text for
status.
Q7-4720 passed See text for
status.
Q7-4735 passed passed See text for
status.
4 Biomedical elastomers

Q7-4750 passed See text for

status.
Q7-4765 passed See text for
status.
Q7-4780 passed See text for
status.
Q7-4840 passed See text for
status.
Q7-4850 passed passed See text for
status.
Q7-6860 passed See text for
status.
Medical See text for
Adhesive A status.
passed passed passed See text for
status.
355 Medical passed passed passed passed See text for
status.
Grade Pressure Sensitive
*Biocompatibility based on comparison with USP negative controls
Note: It is the user's responsibility to adequately test or determine that these materials are suitable or safe for any application.
331
332 J.W. Boretos and S.J. Boretos

Table 4.13 Sterilization Methods for Elastomers

Ethylene
Product Steam/autoclave Cobalt 60 oxide Cold solution
Biospan OK — OK —
Biospan D — — —
Biospan S — — — —
C-Flex R70-001 OK OK OK —
R70-003 OK OK OK —
R70-005 no OK OK —
R70-026 OK no OK —
R70-028 no OK OK —
R70-046 no OK OK —
R70-050 OK OK OK —
R70-051 OK OK OK —
R70-072 OK OK OK OK
R70-081 OK OK OK —
R70-082 OK OK OK —
R70-085 OK OK OK —
R70-089 NR OK OK —
R70-091 NR OK OK —
R70-116 no OK OK —
R70-190 no OK OK
R70-214 no OK OK —
Carbothane with caution — OK —
ChronoFlex — — — —
Coremer — OK OK OK
Corethane 80A — OK OK OK
55D — OK OK OK
75D — OK OK OK
Corhesive — — — —
Ecdel elastomers OK no OK —
Hydrothane — —
—
Kraton G-series — OK
D-series — — OK
Medicaflex no OK OK —
Natural rubber, gum OK OK OK
Natural rubber, latex with caution OK OK —
Pellethane no OK OK —
Poly blend — — „
Poly blend 1100 — — —
PVC Elastichem OK OK OK —
Ellay OK OK OK —
Geon — — —
Multi-Chem OK OK OK —
Teknor OK OK OK —
(continued)
4 Biomedical elastomers 333

Table 4.13 (continued)

Ethylene
Product Steam/autoclave Cobalt 60 oxide Cold solution
in general Flexible with caution OK —
PVC, OK
rigid PVC, no
Santoprene OK OK OK OK
Sarlink 3260 — —
Silicone High OK OK OK —
consistency
LIM OK OK OK —
Adhesives OK OK OK —
Dispersions OK OK OK —
Tecoflex with caution OK OK with caution
Tecothane — OK OK with caution
Texin no OK OK —
Caution: with some aromatic polyurethanes methylene dianiline (MDA) can be generated with
steam sterilization.

Table 4.14 Water Absorption of Various Elastomers

Water absorption,
percent (after 24
Classification Product and manufacturer Product no. hours) ASTM-D 570
Thermoplastic Santoprene® Rubber, 281-55 281-64 6.0
vulcanizate
Advanced Elastomer Systems 281-87 283-40 0.0
PCCE Ecdel™ Elastomer, 9965
copolyester 9966 0.4
elastomer Eastman Chemical Co. 9967
Corethane®, TPE 80A 1.2
Polycarbonate-based Corvita Corp. TPE 55D 0.9
TPE 75D 0.8
polyurethanes Corhesive™, 1.2
Corvita Corp.
Biospan® segmented Biospan® 1.5
Polyurethane-based polyurethane, The Polymer Biospan®D 1.3
Technology Biospan®S 1.5
elastomers Group, Inc.
PolyBlend™ 1000, PB1000-650 <
and
PolyBlend™ 1100 PB1100-55
PB1100-60 <0.4
Poly Medica PB1100-75
Biomaterials, Inc. PB1100-80
Silicone rubber 0.1-0.5
Silicone type A <0.2
adhesive
334 J.W. Boretos and S.J. Boretos

4.3.2 Equivalent silicone elastomers

Two manufacturers, NuSil Technology and Applied Silicone Corp., are providing
equivalent silicone materials for the Dow Corning products that have been with-
drawn. Tables 4.15 and 4.16 gives reported comparisons.

Table 4.15 Equivalent Silicone Elastomers for Existing Dow Corning Silicones
Dow Corning NuSilt‡ Silicone Applied ∆ Silicone Medical Grade Silicone
Silicone* Equivalent Equivalent Description
Medical MED-1137 40064 Medical RTV Adhesive, Acetoxy
Adhesive A System (see also Rehau, Table
4.16)
Q7-4535 MED-4535 40044 High Consistency, 35 Durometer,
Peroxide Cure
Q7-4550 MED-4550 40045 High Consistency, 50 Durometer,
Peroxide Cure
Q7-4565 MED-4565 40046 High Consistency, 65 Durometer,
Peroxide Cure
Q7-4720 MED-4720 40043 High Consistency, 20 Durometer,
Platinum Cure
Q7-4735 MED-4735 40039 High Consistency, 35 Durometer,
Platinum Cure
Q7-4750 MED-4750 40040 High Consistency, 50 Durometer,
Platinum Cure
Q7-4780 MED-4780 40042 High Consistency, 80 Durometer,
Platinum Cure
MDX4-4210 MED-42111 40072 Liquid Silicone, 25 Durometer,
Platinum Cure
40029 Liquid Silicone, 30 Durometer,
Platinum Cure
Q7-4840 MED-4840 40026 Liquid Silicone, 40 Durometer
Platinum Cure
Q7-4850 MED-4850 40027 Liquid Silicone, 50 Durometer,
Platinum Cure
Q7-4865 MED-4865 Liquid Silicone, 65 Durometer,
Platinum Cure
DC-360 MED-360 40047 Medical Grade Silicone Fluid,
1000 cps.
Specify 40073 Medical Grade Silicone Fluid,
350 cps.
viscosity 40074 Medical Grade Silicone Fluid,
20 cps.
* Dow Corning Corp., Midland, Ml. ‡ NuSil Silicone Technology, Carpinteria, CA
∆ Applied Silicone Corp., Ventura, CA Note: It is the user's responsibility to adequately test or
determine that these materials are suitable or safe for any application.
4 Biomedical elastomers 335

Table 4.16 Equivalent Silicone Elastomers for Withdrawn Dow Corning silicones
Dow Corning NuSilt‡ Silicone Applied ∆ Silicone Medical Grade Silicone
Silicone* Equivalent Equivalent Description
MDX4-4515 MED-4515 40045 50 Durometer, peroxide cure
MDX4-4516 MED-4516 40046 60 Durometer, peroxide cure
Q7-2245 MED-2245 40009 40 Durometer, platinum cure
Q7-2213 MED-2213 40076 Dispersion in 1, 1, 1
Rehau 1511¥ 40076 trichloroethane
Medical RTV adhesive,
acetoxy system
*Dow Corning Corp., Midland, Ml. ‡ NuSil Silicone Technology, Carpinteria, CA.
△ Applied Silicone Corp., Ventura, CA. ¥ Rehau AG and Co., Rehau, Germany.
Note: It is the user’s responsibility to adequately test or determine that these materials are suitable
or safe for any application.

4.4 Sterilization of Elastomers

4.4.1 Sterilization methods

Not all materials respond alike when subjected to various means of sterilization.
Some are heat sensitive, some will absorb sterilization fluids, some will be affected
by molecular changes when subjected to radiation sterilization and others will
absorb and hold irritating gases for extended periods of time. Table 4.13 gives ster-
ilization methods that have been judged most appropriate for each elastomer. The
consequences of using an inappropriate method can be loss in physical properties
and an adverse biological response.

4.5 Relevant ASTM Standards

Standard methods of testing elastomers used for medical applications are given by
specific ASTM test methods. Physical and biological tests are provided here to
serve as references for the data cited in the tables and listed in Table 4.17. They are
also designated in the FDA Guidance Document.

4.6 Biocompatibility

Table 4.12 on biocompatibility of various elastomers is intended to show the status

of in vitro and in vivo testing. The successful outcome of these tests can serve as
guides to potentially acceptable performance of an elastomeric product in a medical
device under development. However, the use of elastomeric products in medical
devices is the responsibility of the device manufacturer who must establish their
safety and efficacy with the FDA.
336 J.W. Boretos and S.J. Boretos

Table 4.17 Relevant ASTM Standards

D 395 Test Method for Rubber Property - Compression Set
D 412 Test Method for Vulcanized Rubber, Thermoplastic Rubbers and Thermoplastic
Elastomer - Tension
D 471 Test Method for Rubber Property - Effect of Liquids
D 570 Test Method for Water Absorption of Plastics
D 624 Test Method for Tear Strength of Conventional Vulcanized Rubber and
Thermoplastic Elastomer
D 638 Test method for Tensile Properties of Plastics
D 792 Test Method for Specific Gravity (Relative Density) and Density of Plastics by
Displacement
D 797 Test Methods for Rubber Property - Young's Modulus at Normal and Subnormal
Temperatures
D 1630 Test Method for Rubber Property - Abrasion Resistance (NBS Abrader)
D 1708 Test method for Tensile Properties of Plastics by Use of Microtensile Specimens
D 1790 Test method for Brittleness Temperature of Plastic Film by Impact
D 1938 Test method for Tear Propagation Resistance of Plastic Film and Thin Sheeting by
a Single-Tear Method
D 2240 Test Method for Rubber Property - Durometer Hardness
D 2702 Standard Practice for Rubber Chemicals - Determination of Infrared Absorption
Characteristics
D 3418 Test Method for Transition Temperatures of Polymers by Thermal Analysis
D 3593 Test Method for Molecular Weight Averages and Molecular Weight Distribution of
Certain Polymers by Liquid Size-Exclusion Chromatography (Gel Permeation
Chromatography, GPC) Using Universal Calibration
D 5023 Test Method for Measuring the Dynamic Mechanical Properties of Plastics Using
Three Point Bending
D 5026 Test Method for Measuring the Dynamic Mechanical Properties of Plastics in
Tension
E 355 Standard Practice for Gas Chromatography, Terms and Relationships
E 1356 Test Method for Glass Transition Temperatures by Differential Scanning
Calorimetry or Differential Thermal Analysis
F 604 Classification for Silicone Elastomers Used in Medical Applications
F 619 Standard Practice for Extraction of Medical Plastics
F 720 Standard Practice for Testing Guinea Pigs for Contact Allergens: Guinea Pig
Maximization Test
F 748 Standard Practice for Selecting Generic Biological Test Methods for Materials and
Devices
F 749 Standard Practice for Evaluating Material Extracts by Intracutaneous Injection in
the Rabbit
F 750 Standard Practice for Evaluating Material Extracts by Systemic Injection in the
Mouse
F 813 Standard Practice for Direct Contact Cell Culture Evaluation of Materials for
Medical Devices
F 895 Standard Practice for Agar Diffusion Cell Culture Screening for Cytotoxicity
F 981 Standard Practice for Assessment of Compatibility of Biomaterials (Non-porous)
for Surgical Implants with Respect to Effect of Materials in Muscle and Bone
4 Biomedical elastomers 337

4.7 Sources

• AlphaGaryAlphaGary, Leominster, MA
• Applied SiliconeApplied Silicone Corp., Ventura, CA
• Biospan®Polymer Technology Group, Inc., Emeryville, CA
• C-Flex®Consolidated Polymer Technologies, Inc., Largo, FL
• Carbothane™Thermedics, Inc., Woburn, MA
• ChronoFlex™PolyMedica Industries, Inc., Woburn, MA
• Coremer™Corvita Corp., Miami, FL
• Corethane®Corvita Corp., Miami, FL
• Corhesive™Corvita Corp., Miami, FL
• Ecdel™Eastman Chemical Co., Kingsport, TN
• Elastichem™Colorite Plastics Co., Ridgefield, NJ
• Ellay™Ellay, Inc., City of Commerce, CA
• Geon®B.F. Goodrich Co., Chemical Group, Cleveland, OH
• Hydrothane™PolyMedica Industries, Inc., Woburn, MA
• Kraton®Shell Chemical Co., Oak Brook, IL
• Medicaflex™Advanced Resin Technology, Manchester, NH
• Multichem™Colorite Plastics Co., Ridgefield, NJ
• Natural rubberExxon Chem. Co., Buffalo Grove, IL Goodyear Tire and Rubber
Co., Akron, OH
• NuSil SiliconeNuSil Technology, Carpinteria, CA
• Pellethane™Dow Chemical Co., Midland, MI
• PolyBlend™PolyMedica Industries, Inc., Woburn, MA
• Santoprene®Advanced Elastomer Systems, St Louis MO
• Sarlink®DSM Thermoplastic Elastomers, Inc., Leominster, MA
• SilasticDow Corning Corp., Midland, MI
• Tecoflex®Thermedics, Inc., Woburn, MA
• Tecothane®Thermedics, Inc., Woburn, MA
• Teknor™Teknor Apex Co., Pawtucket, RI
• Texin™Miles, Inc., Pittsburgh, PA
Chapter 5
Oxide Bioceramics: Inert Ceramic Materials
in Medicine and Dentistry

J. Li and G.W. Hastings

5.1 Introduction

Single oxide ceramics, e.g. aluminium oxide (A12O3, alumina) and zirconium
dioxide (ZrO2, zirconia), are bioceramics of an inert nature. An inert ceramic does
not form a bonding to bone similar to those bioceramics of bioactive nature. Alumina
bioceramics are in the pure aluminium oxide form, whereas zirconia bioceramics
are partially stabilized by additional oxides, e.g. yttrium oxide, calcium oxide or
magnesium oxide.
Oxide ceramics exhibit superior mechanical properties, corrosion and wear
resistance. Since the oxides are the highest oxidation state of the metal, they are
stable even in the most invasive industrial and biomedical environments. Alumina
and zirconia are utilized as load-bearing hard tissue replacements and fixation
implants in dentistry and surgery.

5.2 Short History

Although the use of alumina as implants can be traced back to the 1930s as described
by Hulbert et al. (1) (Table 5.1), the extensive use of alumina since the 1980s has
depended on new powder processing technology enabling grain size reduction of
the sintered ceramics from 10 micrometers down to 2 micrometers (Figure 5.1,
microstructure of alumina). This significantly improves the performance of the

J. Li (*)
Centre for Oral Biology, Karolinska Institute, Huddinge S 141–4, Sweden
G.W. Hastings
Institute of Materials Research and Engineering, Singapore, Singapore

© Springer Science+Business Media New York 2016 339

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_21
340 J. Li and G.W. Hastings

Table 5.1 Evaluation of oxide ceramic implants: alumina and zirconia

First suggestion of application of alumina ceramics
1932 in medicine Rock (2)
1963 First orthopaedic bone substitute application Smith (3)
1964 First dental implant of alumina Sandhaus (4)
1970 French hip prosthesis: Alumina ceramic ball and cup Boutin (5)
1974 German hip prosthesis Mittelmeier (6)
1977 28 mm alumina ball Shikita (7)
1981 Alumina total knee prosthesis Oonishi et al. (8)
1982 FDA approval for non-cemented alumina ceramic
cup and ball and CoCrMo-stem of Mittelmeier type
1986 First zirconia ball of 32 mm Lord et al. (9)
1993 First dental implant of zirconia Akagawa et al. (10)
1995 First zirconia dental post Meyenberg et al. (11)
1996 First zirconia inlay Johansson (12)

Figure 5.1 SEM micrograph of dense alumina, etched in boiling H3PO4 for 6 minutes to show the
microstructure.

alumina ceramic hip balls. Alumina and partially stabilized zirconia are currently in
extensive use as implants in consequence of their high strength, excellent corrosion
and wear resistance and stability, non-toxicity and biocompatibility in vivo. A summary
of alumina- and zirconia-based implants is presented in Table 5.2. The most estab-
lished example is in the total hip endoprosthesis with a combination of metallic
stem, ceramic ball and ultra high molecule weight polyethylene (UHMWPE)
5 Oxide Bioceramics: Inert Ceramic Materials in Medicine and Dentistry 341

Table 5.2 Biomedical applications of oxide ceramics

Materials Applications References
Alumina Hip ball & cup Clarke and Willmann (13)
Knee joint Oonishi et al. (8)
Bone screws
Dental implant Kawahara (14)
Dental crowns & brackets Sinha et al. (15)
Zirconia Hip ball Christel (16)
Dental implants Akagawa et al. (10)
Dental post, brackets and inlay (10, 11), Keith et al. (17)

acetabular cup. A ten year clinical success rate better than 90% is reported for the
cemented total hip endoprosthesis.
Dental implants of polycrystalline alumina were suggested by Sandhaus in
Germany (4). Type Tübingen was produced by Frialit in the 1970s. These devices
have not been generally accepted, due to the fracture failure of the implants, particu-
larly for those of polycrystalline type produced in the early 1970s. The single crystal
sapphire type, introduced in Japan by Kawahara in the 1970s (18) is, however, still
being used and a recent 10-year clinical follow-up report from Sweden showed a
92% success rate (19) for the single crystal dental implants.
Alumina and zirconia ceramics are also being used for alveolar ridge reconstruc-
tion (20), maxillofacial reconstruction, as ossicular bone substitutes (21), and in
ophthalmology (22), knee prosthesis (8), bone screws as well as other applications
as dental biomaterials, such as dental crown core, post, bracket and inlay (23, 24).

5.3 Material Properties and Processing

5.3.1 Materials properties

Although alumina is chemically more stable it is mechanically weaker than zirconia,

and the phase changes or transformation mechanisms in zirconia produce a unique
ceramic material having much higher strength and higher fracture toughness com-
pared with alumina and other ceramics. The excellent mechanical properties of
zirconia allow the design of hip balls of smaller diameter in order to reduce the
wear of the UHMWPE cup with expected increased long-term clinical performance
as a result.
The chemical stability of alumina is related to its phase stability, whereas the
phase changes of zirconia result in degradation in strength and wear resistance.
Release of substances from zirconia and alumina implants to the surrounding tissue
is very low and neither local nor systemic effects have been reported.
342 J. Li and G.W. Hastings

Aluminium oxide: alumina

Aluminium oxide is produced by heating its hydrates. At least seven forms of

alumina have been reported, but six of these forms have traditionally been desig-
nated ‘gamma alumina’. When heated above 1200°C, all other structures are irre-
versibly transformed to the hexagonal alpha-alumina, corundum, a close-packed
arrangement of oxygen ions. Thus alphaalumina is the only stable form above
1200°C and by far the most commonly used of structural ceramics. Alpha-alumina
is thermodynamically stable and is crystallographically identical with the single
crystal ruby and sapphire ceramics. Each aluminium ion is surrounded by six oxygen
ions, three of which form a regular triangle on one side, the other three form a simi-
lar triangle on the other side, with the two planes of the triangles being parallel and
the triangles being twisted 180° (25).

Physical and mechanical properties

Table 5.3 and 5.4. Resulting from a strong chemical bond between the Al and O
ions, as expected from the value of heat of formation (-400K cal/mol), Al2O3 has a
high melting point, the highest hardness among known oxides, and high mechanical
strength (26).

Table 5.3 Engineering Properties of Alumina and Zirconia (At 25 °C)*

Property Al2O3 ZrO2**
Physical
Crystallography Hexagonal Tetragonal***
a(Å) 4.76 3.64
c(Å) 13.0 5.27
Space group D6Ba P42/nmc
Melting point (°C) 2040 2680
Density (g/cm3) 3.98 6.08
Grain size (μm) 1–6 0.54
Hardness (GPa) 22 12.2
Modulus of elasticity, (GPa) 366 201
Poisson’s ratio 0.26 0.30
Thermal coefficient of 6.5 10.1
expansion 25–200 °C
Mechanical
Flexural strength (MPa) 551 1074
Compressive Strength (MPa) 3790 7500
Tensile strength (MPa) 310 420
Fracture toughness (MPa m1/2) 4.0 6–15
* Sources: refs 26, 44 and 45
** Zirconia presented is the yttria-partially stabilized material
*** Most of the medical-grade zirconia is partially stabilized tetragonal zirconia
5 Oxide Bioceramics: Inert Ceramic Materials in Medicine and Dentistry 343

Table 5.4 Properties of medical-grade ceramic materials according to the standards to the
standards and the manufacturer’s technical date – alumina and zirconia
Alumina
according to
ISO-6474 Zirconia
ASTM Frialit according to
F 603–83 bioceramic ISO/DIS Prozyr®
Property DIN 58 8353 alumina 13356 zirconia
Purity (%) >99.5 >99.5 >99.5* >95
Density (g/cm3) >3.9 >3.98 >6.0 6
Porosity (%) 0 ** 0 0
Grain size (μm) <4.5 >2.5 <0.6 <1
Microhardness (GPa) 23 23 — 13
Young’s modulus (GPa) 380 380 — 220
Flexural strength (MPa) >400 >450 >900 >920
Biaxial flexural strength (MPa) 250 — >550 —
Impact strength (cm MPa) >40 >40 124
Fracture toughness (MPa m1/2) 10
Wear resistance (mm3/h) 0.01 0.001 __
Corrosion resistance (mg/m2d) <0.1 <0.1 — —
* ZrO2+HfO2+Y2O3
** Not available.

Chemical properties

Alumina is chemically stable and corrosion resistant. It is insoluble in water and

very slightly soluble in strong acids and alkalies. Therefore, practically no release
of ions from alumina occurs at a physiological pH level, 7.4.

Wear resistance

Arising from the chemical stability and high surface finish and accurate dimensions,
there is a very low friction torque between the alumina femoral heads and the ace-
tabular cup, leading to a low wear rate. Combinations of ceramic head/UHMWPE
cup and ceramic head/ceramic cup were tested and compared to the metal head/
UHMWPE cup. The wear resistance of the ceramic head/UHMWPE cup combina-
tion over metal/UHMWPE has improved from 1.3 to 34 times in the laboratory and
from three to four times clinically (27, 28). No alumina wear particles from retrieved
ceramic/UHMWPE were found, whereas UHMWPE wear particles from microns
to millimetres in size were found in the retrieved surrounding tissues. However,
from the ceramic/ceramic combination, ceramic particles resembling ‘fine grains
and great fragments in the ranges from 0.5 to 10 micrometers diameter, with the
predominant size of about 1 micrometer’ were found in the surrounding tissue (29).
The advantage of ceramic/ceramic combination over ceramic/UHMWPE is, there-
fore, doubtful. For wear tests, we refer to ISO-6474 ASTM F-603.
344 J. Li and G.W. Hastings

Clinical performance

The fracture of ceramic balls in ceramic: UHMWPE combination has been virtually
zero. Fritsch and Gleitz (30) published a failure analysis on 4341 alumina ceramic
heads articulating with 2693 alumina ceramic and 1464 polymer sockets implanted
over 20 years (1974 to 1994), and concluded that the use of ball type neckless heads
brought the fracture rate close to zero. The success rate of 10 years foliow-up is
normally above 90% for the ‘elderly’ patient population. Stem and cup loosening
are the causes of failure, where the consistent wear debris from UHMWPE and bone
cement remain the problems.

Zirconium dioxide: zirconia

Zirconia ceramics are termed polymorphic because they undergo several transfor-
mations on cooling from a molten state to room temperature. It exhibits three well-
defined polymorphs, the monoclinic, tetragonal and cubic phases and a high pressure
orthorhombic form also exists. The monoclinic phase is stable up to about 1170°C
where it transforms to the tetragonal phase, stable up to 2370°C, while the cubic
phase exists up to the melting point 2680°C. A large volume change of 3 to 5%
occurs when zirconia is cooled down and transforms from the tetragonal to the
monoclinic phase.

Partially stabilized zirconia (PSZ) and tetragonal zirconia

polycrystals (TZP)

The volume change due to phase transformation is sufficient to exceed elastic and
fracture limits and causes cracking of the zirconia ceramics. Therefore, additives
such as calcia (CaO), magnesia (MgO) and/or yttria (Y2O3) must be mixed with
zirconia to stabilize the material in either the tetragonal or the cubic phase. PSZ is a
mixture of cubic and tetragonal and /or monoclinic phases, whereas TZP is 100%
tetragonal (phase diagram Figure 5.2). Both PSZ and TZP are suggested for medical
implant applications. Yttria-TZP ceramics have a strength and fracture toughness
approximately twice that of alumina ceramics used in the biomedical field. This
makes zirconia heads less sensitive to stress concentrations at the points of contact
with metal cones.

Physical and mechanical properties

Zirconia ceramics have a high density because of heavy zirconium ions, and a low
microhardness and elastic modulus, together with high strength and fracture tough-
ness compared to other ceramics including alumina. The superior mechanical strength
provides the possibilities for producing ceramic ball heads of size below 32 mm.
5 Oxide Bioceramics: Inert Ceramic Materials in Medicine and Dentistry 345

Figure 5.2(a) Part of the

equilibrium phase diagram
for the system ZrOz-
CaO. Css refers to the
cubic solid-solution phase,
Tss to the tetragonal
solid-solution phase, and
Mss to the monoclinic
solid-solution phase
(ref. 21).

Figure 5.2(b) Y2O3–ZrO2

phase diagram: the
addition of less than 5% of
Y2O3 to ZrO2 allows the
sintering of a fully
tetragonal material
(t=tetragonal phase;
m=monoclinic phase;
c=cubic phase) (ref. 16).
346 J. Li and G.W. Hastings

Fracture toughness mechanisms:

Garvie et al. were the first to realize the transformation toughening mechanism for
zirconia ceramics. Increase of both strength and fracture toughness can be obtained
by utilizing the tetragonal-monoclinic phase transformation of metastable tetrago-
nal grains induced by the presence of the stress field ahead of a crack (31). The
volume change and the shear strain developed in the martensitic reaction were rec-
ognized as opposing the opening of the crack and therefore acting to increase the
resistance to crack propagation.

Wear resistance and chemical stability:

The published results of in vitro wear tests demonstrated that zirconia has a supe-
rior wear resistance. Saikko (32) showed no wear of zirconia femoral heads on
his hip simulator wear test against 10.9 mm UHMWPE cup, and Praveen Kumar
et al. (33) demonstrated the high wear resistance of zirconia against UHMWPE
and the superiority of zirconia ceramics even over alumina ceramics in terms of
low wear and low friction. A significant reduction in the wear rate of zirconia
ball heads compared to the metal ball heads was reported on a pin-on-disc wear
test and on a hip simulator (34). However, there are two potential limitations for
the use of zirconia as bioceramics: degradation and radiation. It is known that the
phase transformation is accelerated in aqueous environment, but little is known
about how this phase transformation will occur in biological environment,
particularly under dynamic loadings. A warning against steam resterilization has
been issued in the UK. Radioactive U-235 impurity was detected in some ‘pure
zirconia’, both alpha- and gamma-irradiation were measured from zirconia fem-
oral balls. Although the radioactivity was low, more work is required to verify
this matter (13).

Clinical performance

The surface degradation of the zirconia balls due to the phase transformation under
loading seems to be a problem, although no significant change in mechanical
strength was reported in some long-term in vivo and in vitro studies (35, 36).
Seriously, catastrophic failure of modular zirconia ceramics femoral head compo-
nents after total hip arthroplasty was reported (37). Since zirconia femoral heads
have a short clinical history and few clinical results are available, more investigation
is required to eliminate the factors which impair the clinical stability of zirconia
ceramics under loading.
5 Oxide Bioceramics: Inert Ceramic Materials in Medicine and Dentistry 347

5.3.2 Materials processing

An advanced ceramic is processed in such a way that the structure of the materials
on different levels, including atomic, electronic, grain boundary, microstructural
and macrostructural, is under strict control. In the manufacturing processes, empha-
sis is placed on producing dense ceramics with a fine microstructure. However,
other factors such as chemical composition, the nature and distribution of the
impurities, crystal structure, grain size, and defects are also of importance to the
performance of the ceramic materials. Three basic processes are involved in the
production of fine ceramic components, namely: 1. powder technology, 2. densifi-
cation or sintering and 3. machining. Both alumina and zirconia hip balls are pro-
duced by compacting fined-grained powder (green bodies), and sintering at
1500–1700 °C and finally grinding or lapping to obtain a high surface finish and
sphericity (Ra<0.02 μm).

5.4 Biocompatibility of Oxide Bioceramics

No materials placed within a living tissue can be considered to be completely inert.

However, oxide bioceramics, by their very nature, do not suffer from corrosion or
degradation in biological environments, as metals or plastics do. Ceramics, having
molecular structures completely different from those of living tissues, are generally
stable inside the living body and provide a high degree of acceptance by the apposi-
tion to the surrounding tissue as shown by in vitro and in vivo studies Ichikawa et al.
observed no adverse soft tissue responses to zirconia and alumina implants after 12
months of implanation (38). Takamura et al. reported that alumina and zirconia did
not possess chronic toxicity to mice (39), whereas Steflik et al. found a biological
seal at the alumina dental implant and epithelium interface (40). However, oxide
bioceramics do not form a chemical bond to bone tissue and are therefore defined as
inert biomaterials. Oxide bioceramics are defined as inert biomaterials.
The ASTM standards (F 748/82, 763/82) and ISO standards No 10993 have set
the guidance for biological testing of biomaterials for orthopaedic application. The
materials should be tested in soft tissue as well as in hard tissue environments, for
both short-term and long-term experiments. A summary of recommended biologi-
cal testing is presented in Table 5.5. Both alumina and zirconia have shown non-
toxicity and good biocompatibility according to the tests. Testing results for zirconia
made by a French Company are shown in Table 5.6. Although some serious
problems occurred with zirconia balls, the basic biocompatibility of the zirconia
remains. Soft tissue and bone responses to zirconia and alumina were studied in our
lab: no adverse tissue reaction to these ceramics were found. The patterns of tissue-
materials interface after 1 month implantation in muscle and femur of rat are shown
in Figure 5.3.
348 J. Li and G.W. Hastings

Table 5.5 Guidance for Biologic Evaluation Tests of the Implant Device in Contact to
bone Tissue (According to ISO 10993–1:1992 (E))
Contact duration
A-limited B-prolonged C-permanent
Biological tests (>24 h) (<24 h to 30 days) (<30 days)
Cytotoxity x x x
Sensitisation x x x
Irritation/Intracutaneous x x x
Reactivity
Irritation/Intracutaneous x
Genotoxicity x x
Implantation x x
Chronic toxicity x
Carcinogenicity x
The related tests see ISO standards from No. 10993–1 to 10993–6

Table 5.6 Biological evaluations of zirconia ceramics (Prozyr®, Ceramiques Desmarquest,

France)
Biocompatibility Standard used Results
Short-term in vivo ASTM F 763/82 Very good
biocompatibility
In vitro biocompatibility ASTM F 748/82
Cell culture cytotoxicity PRS 90.702 Good cytocompatibility
Mutagenicity Ames test No mutagenic activity
Micronucleus test
Systemic injection acute ASTM F 750/82 According to standard
toxicity
Intracutaneous injection ASTM F 749/82 No irritation
ASTM F720/81
Sensitization Magnusson No sensitization

5.5 Applications

5.5.1 Orthopaedic applications

The dominating application of alumina and zirconia is as hip balls as well as cups
of total femoral prosthesis. The neckless hip balls are the most popular design. In
1981, Oonishi et al. (8) reported on the use of an alumina ceramic total knee pros-
thesis. High alumina ceramic middle ear implants (Frialit) are used clinically in
Europe since 1979 (21). An opthalmological implant device consisting of a combi-
nation of a single crystal alumina optional cylinder and a polycrystalline alumina
holding ring was introduced clinically in 1977 (22). Kawahara (12) has reported
extensively on single crystal alumina bone screws.
5 Oxide Bioceramics: Inert Ceramic Materials in Medicine and Dentistry 349

Figure 5.3(a) Optical micrograph of alumina and soft tissue interface.

Figure 5.3(b) Zirconia and bone interface 1 month after implantation. Arrows are pointing to the
interfaces.
350 J. Li and G.W. Hastings

Table 5.7 Ceramic manufacturers known for their bioceramic productions

Primary Secondary Trade Names
Manufacturer Country Materials Materials Al2O3 (ZrO2)
Astroment USA ZrO2 Al2O3
Ceraver France Al2O3 ZrO2
Cerasiv Germany Al2O3 ZrO2 Biolox
Desmarquest France ZrO2 (Proyzr)
HiTech France ZrO2 Al2O3
Kyocera Japan Al2O3 ZrO2 Bioceram
Metoxit Switzerland ZrO2
Morgan Matroc England ZrO2 Al2O3
NGK Japan ZrO2
Biocare Sweden Al2O3 Procera
Unitek USA Al2O3 Transcend 2000
Maillefer Switzerland ZrO2

5.5.2 Dental applications

Alumina and zirconia ceramics have been utilized for root analogue, endosteal
screws, blades and pin-type dental implants. The root and blade form dental implants
used during the 1970s tended to fracture after a few years in function (41, 42) (Brose
et al., 1987, Driskell, 1987). Although initial testing of these polycrystalline alu-
mina materials showed adequate mechanical strength, the long-term clinical results
demonstrated functional limitations related to material properties and implant
design. However, single crystalline alumina showed mechanical strength superior to
that of polycrystalline alumina. It allows a much higher load. One-stage dental
implants of single crystalline alumina are used clinically with a high success rate.
McKinnery (43) had also reported on single crystal alumina blade and screw dental
implants. Dental implants of zirconia have not been widely used clinically although
zirconia has a similar mechanical strength and a much higher fracture toughness in
addition to lower cost of production compared to single crystalline alumina. The
term dental implant is used only for materials in contact with bone and soft tissue
(14). Alumina and zirconia are also used in other dental applications, alumina
ceramic crowns, Procera® (23), zirconia dental post, (10) and recently a dental
inlay of zirconia was introduced (11). Orthodontic brackets made of oxide ceramics
were also produced, tested and used clinically. Unfortunately, tooth surface damage
was observed when the brackets were taken away (15). Modification of the debond-
ing technique is under developing.

5.6 Manufacturers and Their Implant Products

Clarke and Willmann (13) make a comprehensive summary about the bioceramic
manufacturers (Table 5.6). Some dental companies are included.
5 Oxide Bioceramics: Inert Ceramic Materials in Medicine and Dentistry 351

5.7 Problems and Future Prospects

Hip balls of polycrystalline alumina have a minimum size limitation to ca. 28 mm

due to strength limitations. A reduced ball size might have two positive effects on
the applications: reduced wear and better suitability (smaller) for Asian patients.
Although single crystalline alumina might overcome the strength limitation, the
cost of manufacturing is unreasonably high and in addition, some processing prob-
lems remain. Zirconia, on the other hand, has a high strength and high fracture
toughness, but it suffers from potential biodegradation. Therefore, the future
research and development will focus on the understanding of degradation mecha-
nisms of zirconia in the body and the improvement of stability of this material. Of
course, combinations, such as alumina/zirconia composite and even non-oxide
ceramic, such as nitrides and carbides, ought also to be investigated.

References

1. Hulbert, S.F., Bokros, J.C., Hench, L.L., Wilson, J. and Heimke, G. In Ceramics in Clinical
Applications, ed. by Vincenzini, P. Elsevier, Amsterdam, 1987, pp. 3–27.
2. Rock, M. German Patent 583 589, 1933.
3. Smith, L. Arch. Surg. 1963; 87: 653–661.
4. Sandhaus, S. British Patent 1083769, 1967.
5. Boutin, P. Presse Med. 1971; 79: 639.
6. Mittelmeier, H. Z. Orthop. Ihre Grenzgeb 1974; 112 : 27.
7. Shikita, T. Paper presented at the XIV World Congress of SICOT, Kyoto, Japan, October
15–20, 1978.
8. Oonishi, H., Okabe, N., Hamaguchi, T. and Nabeshima, T. Orthopaedic Ceramic Implants I,
1981; 11–18.
9. Lord, G. et al. Paper presented at the Harrington Arthritis Research Centre Symposium,
November 18–21, 1990.
10. Akagawa, Y. et al. J. Prosthet. Dent. 1993; 69: 599–604.
11. Meyenberg, K.H., Luthy, H. and Scharer, P. J. Esthet. Dent., 1995; 7(2): 73–80.
12. Johansson, B. Tandläkartidningen 1996; 14746–749.
13. Clarke, I.C and Willmann, G. In Bone implant Interface ed. H.V. Cameron,Hugh U., Mosby,
1994, pp. 222.
14. Kawahara, H. In Encyclopedic handbook of biomaterials and bioengineering ed. by Wise,
Donald L. et al. Marcel Dekker, Inc., New York, 1995, pp. 1469–1524.
15. Sinha, P.K., Rohrer, M.D., Nanda, R.S. and Brickman, C.D. American J. Orthodont &
Dentofacial Orthop, 1995; 108: 455–63.
16. Christel, P.S. In Condse Encyclopedia of Medical & Dental Materials, ed. by Williams, D.F.,
Pergamon Press, Oxford, 1990, pp. 375–379.
17. Keith, O., Kusy, R.P. and Whitley, J.Q. American J. Orthodont. & Dentofacial Orthop., 1994;
106(6): 605–614.
18. Kawahara, H. Orthopaedic Ceramic Implants 1,1981; 1–10.
19. Fartash, B. Single Crystal Sapphire Dental Implants: Experimental and Clinical Studies. PhD
thesis, Karolinska Institute, Sweden, 1996.
20. Hammer, N.B., Topazian, R.G., McKinney, P.V. and Hulbert, S.F. J. Dent. Res. 1973; 52:
356–361.
352 J. Li and G.W. Hastings

21. Jahnke, K. Biomaterials in Otology , Martinus Nijhoff, The Hague, 1984:205–209.

22. Polack, F.M. and Heimke, G. Ophthalmology, 1980; 87(7): 693–698.
23. Andersson, M. and Odén, A. Acta Odont. Scand., 1993; 51: 59–64.
24. Kittipibul, P. and Godfrey, K. American J. Orthodont & Dentofadal Orthopedics, 1995;
108(3): 308–315.
25. Heimke, G. In Metal and Ceramic Biomaterials Vol I Structure, ed. by Ducheyne, P. and
Hastings, G.W. CRC Press Inc., Boca Raton. Florida, 1984, pp. 41–42.
26. Miyayama. M. et al. In Ceramics and Glass, Engineered Materials Handbook Volume 4, ASM
International, 1991, pp. 748–757.
27. Griss, P. In Functional Behavior of Orthopaedic Biomaterials 11, Ch 2, 1984.
28. Jager M. and Plitz, W. Triobology of aluminium ceramics, Symposium of Biomaterials,
pp. 114–122, 1981.
29. Willert, H.G. et al. In Implant Retrieval: Material and Biological Analysis, ed. by Weinstein.
A., Gibbons, D., Brown, S. and Ruff, W. 1981.
30. Fritsch, E.W. and Gleitz, M. Clin. Orthop. & Related Res., 1996; 328: 129–136.
31. Garvie, R.C, Hannink, R.H. and Pascoe, R.T. Nature, 1975; 258: 703.
32. Saikka, V.O. Acta Orthop. Scand., 1995; 66(6): 501–506.
33. Kumar, P. et al. J. Biomed. Mater. Res., 1991; 25: 813–828.
34. Derbyshire, B. et af. Medical Eng. & Phys., 1994; 16(39): 229–36.
35. Shimizu, K. et al. J. Biomed. Mater. Res., 1993; 27: 729–734.
36. Cales, B., Stefani, Y. and Lilley, E. J. Biomed. Mater. Res., 1994; 28: 619–624.
37. Hummer, C.D., Rothman, R.H. and Hozack, W.J. J. Arthroplasty , 1995; 10(6): 848–850.
38. Ichikawa, Y. et al. J. Prosthet. Dent. 1992; 68: 322–6.
39. Takamura, K. et al. J. Biomed. Mater. Res., 1994; 28 : 583–589.
40. Steflik, D., McKinney Jr R.V. and Koth, D. In Bioceramics: Materials characteristics versus
in vivo behavior, ed. by Ducheyne, P. and Lemons, J.E. Ann. N. Y. Acad. Sci. 523, pp. 4–18.
41. Brose, M. et al. J. Dent. Res., 1987; 66: 113.
42. Driskel, T.D. J. Calif Dent. Assoc., 1987; 16–25.
43. Mckinney, R.V. and Koth, D.L. J. Prosthet. Dent., 1982: 47 : 69–84.
44. Park, J.B. and Lakes, R.S. Biomaterials: an Introduction, 2nd edition, Plenum Press, New York,
1992.
45. Bajpai, P.K. and Billotte, W.G. In The Biomedical Engineering Handbook, ed.by Bronzino,
J.D. CRC Press, 1995, pp. 552–580.
Chapter 6
Ceramic Materials Testing and Fracture
Mechanics

D. Daily

6.1  Introduction

6.1.1  Ductile and Brittle Behaviour

In general, metals deform by dislocation movement which is essentially a shearing

process as the planes of atoms slip over each other. The process is referred to as
“slip”. To be completely ductile there should be five systems in a crystal capable of
independent slip movement (this is known as the von Mises criterion). Most FCC
metals with a face-centred cubic lattice structure deform quite easily by slip, while
other systems less so. If the criterion cannot be met, the behaviour is that of semi-­
brittle or brittle solids.
Ductile behaviour is quite well known in inorganic solids. Single crystals of most
alkali metal halides can be deformed at room temperature, and it is also possible to
strain a pure single crystal of magnesium oxide by 10 %. To obtain NaCl as a
polycrystalline ceramic the powder is simply pressed to 100 % density at room tem-
perature in a die. This is impossible by sintering. Many ceramics become ductile at
high temperatures when slip planes which are inactive at room temperature become
activated eventually to meet the von Mises criterion. Ceramics also become deform-
able by grain boundary sliding at high temperature, and in creep both types of defor-
mation behaviour are possible. At room temperature both ceramics and glasses fall
into the brittle category and any significant deformation is accompanied by cracking.
This may be intergranular, i.e. within the grain boundaries in materials like alumina,
or within the grains as in zirconia.
It is very difficult to identify obvious properties which relate to strength of a
ceramic, but there are a number of factors to which one can point.

D. Daily
(deceased)

© Springer Science+Business Media New York 2016 353

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_22
354 D. Daily

6.1.2  Microstructure

6.1.2.1  Effect of Grain Size

-1
s g = s ¥ + s 1G 2

σg = strength of the ceramic, G = mean grain size. σ1 and σ∞ are material constants.
This is the Orowan–Petch relation first observed in steels in 1953 and holds for all
crystalline materials. It implies that the finer the grain size, the stronger the material,
although it fails for small grain sizes.
Carniglia [33] has identified a two-stage relationship in which the strength depen-
dence changed at a critical grain-size Go. Davidge and Evans enlarged on this [34]
and related the grain-size/strength dependence to the fracture mechanism. In the
larger grain size region, to which the above relations refer, fracture occurs as a con-
sequence of the extension of inherent flaws while for the lower grain sizes, plastic
flow is the dominant cause. The changeover in behaviour is a property of the material
considered, related to its purity, and environmental factors such as temperature. In
ceramics where there are large and small grains, an initial flaw in a large grain region
propagates until it reaches a finer region where the apparent fracture surface energy
is higher. Here it is held up until the stress becomes critical for the new region.
Producers of structural bio ceramic components devote much effort towards the
choice and quality of raw materials used in the manufacture of these devices. This is
of particular importance to load-bearing orthopaedic implant applications where
strength and fracture toughness properties are critical to the long-term performance
of the device. The need therefore is to produce high purity, dense ceramic with fine
grain crystalline microstructures, free from inclusions and/or impurities often intro-
duced through raw materials and production processes. The introduction inhomoge-
neity within the microstructure of the ceramic does have a detrimental effect on the
strength and fracture toughness properties of the device. Inclusion or defects as
small as ~10 μm in a sintered ceramic microstructure have contributed in vivo fail-
ure. It must also be mentioned that other factors such as device design (see Chap.
7—Ceramic hip joint endoprosthesis) and processing surface treatment, for example
grinding, honing and polishing, do influence the strength of the ceramic component
and ultimately the long-term success of a device.
The majority of ceramics currently used in the manufacture of orthopaedic
devices are high-purity zirconia toughened alumina (ZTA), alumina (>99.9 %), zir-
conia, and all contain grain size substantially sub-micrometre. In recent years vari-
ous ISO, BSI, ASTM and CEN Standards have been developed and published,
specifying the material performance requirements and testing methodologies for
implantable grade ceramics. Microstructural specifications including grain size are
stated in the ISO 6474: part 1 for alumina, ISO 6474: part 2 for ZTA and the ISO
13356 for zirconia. The ISO 6474 part 1 grain size specification for load bearing
orthopaedic implantable alumina is ≤2.5 μm although the grain size in commercial
grades has significant proportion of sub-micrometre microstructure. Zirconia (ISO
13356) specifies a grain size ≤0.4 μm; commercial grades are typically 0.3–0.4 μm.
6  Ceramic Materials Testing and Fracture Mechanics 355

ZTA forms a matrix of alumina and zirconia have two size specifications ≤1.5 μm
for the crystalline alumina phase and ≤0.6 μm for the zirconia, the typical grain size
0.5–0.8 μm for alumina phase and 0.1–0.2 μm for zirconia.

6.1.2.2  Strength of Individual Grains

Sapphire single crystals are stronger than polycrystalline alumina, so a crack would
be expected to follow the grain boundaries, so that the smaller the grains, the greater
the surface free energy of sapphire,—about 6–7 J/m, while for polycrystalline alu-
mina, this is about 30 J/m. It should therefore be harder to create new surfaces in
ceramic alumina than in sapphire. The only explanation for this is that in the
ceramic, the cracks are there already, as surface and volume flaws, while in sapphire
they need to be nucleated.
Pure cubic zirconia is a relatively weak material and normal sintering invariably
generates very large grains. One concludes that it is easier for a crack to go across a
grain rather than travel through the grain boundaries.

6.1.2.3  Multiphase and Multicomponent Structures

New considerations have to be introduced for these materials such as (1) the coher-
ency of the boundaries between different crystalline components; (2) the stress
imparted at the interfaces between components (and this can be severe enough to
crack the material) as a consequence of cooling components or phases of different
thermal expansion; (3) different elastic moduli of non-equilibrium of solutes within
phases once these can create internal stresses which oppose crack propagation.

6.1.3  Porosity

Porosity reduces the strength of ceramics. Approximately a 50 % decrease in

strength results from a 10 % increase porosity. Duckworth [35] proposed the follow-
ing relationship:

s = s 0 exp - bP
where σ = strength of the porous material, σ0 = strength of the fully dense material,
P = porosity, b is a constant.
Coble and Kingery measured the flexural strength of polycrystalline aluminas
for porosities between 5 % and 50 % porosity, and using the symbols above,
proposed the following empirical equation:
-s
0.6 P = exp
8000 (1 - p )

356 D. Daily

Knudsen [36] and Passmore et al. [37] derived formulae to describe the coincident
effects of grain size and porosity on alumina at both 25 and 1200 °C. These were of
the form:

s = k × exp ( -bP ) × G ( )-1+ cP

,

where c, k and b are constants, and G is the grain size. This relationship becomes
increasingly in error as the porosity approaches zero.
It is usually considered that the reduction of strength by porosity is caused by
stress concentrations around individual pores. In some instances pores can interact
with a moving crack front to produce a reduction of localised stress [37]. Other
porosity concepts include the assumption by Eudier [38] that the strength is lowered
by the simple reduction of load-bearing cross-sectional area, for which he proposed
the equation:
s = 1 - kP 2 / 3
s0

k is a constant approximately equal to 1.21.
Porosity will be reconsidered after the Griffith concept has been discussed.

6.1.4  Valency Type

In general covalent substances are stronger than ionic. Diamond, silicon carbide,
silicon nitride, cubic boron nitride (borazon), boron carbide are typical; they contain
very few flaws and usually have a rigid framework structure.
Alumina and silica are substances with both ionic and covalent character. At very
high stresses (e.g. at the tips of cracks) deformation by slip is possible causing a
dissipation strain energy.
The relationship between failure stress, surface energy and concentration of
flaws is a complicated one, but a number of approaches to this problem have been
made, giving rise to the science of fracture mechanics.

6.2  Strength Testing of Ceramics

6.2.1  Tensile Testing

In contrast with metals, stress does not cause deformation by slip in ceramics, and
eventual failure along the planes of maximum shear stress. Ceramics fail by the
propagation of cracks, and in particular when one of them reaches a critical length.
Ceramics also are 8–10 times stronger in compression than in tension, and so the
ceramic member is designed for use in compression, while is specified its strength
as a result of tensile test.
6  Ceramic Materials Testing and Fracture Mechanics 357

Metals are usually tested in tension on a tensile machine where a machined spec-
imen is trained and the strain is measured as the increase in separation of a pair of
marks divided by the original distance separating them (called the “gauge length”),
and the stress is determined from a load cell reading. This method of testing is not
convenient for ceramics for the following reasons.

6.2.2  Ceramic Machining

Machining ceramics to precise dimensions requires diamond grinding. This is a

costly operation, and a specially prepared piece may have to be made in a separate
operation. Consequently this may not be representative of the batch of material
about which information is required. The grinding operation itself can introduce
surface flaws which may impair the strength of the test piece.

6.2.3  Test Pieces

The test piece will probably be dumbbell shaped and carefully have avoided sharp
corners which would create a stress concentration. It would be supported by a ring,
and another ring would be used to supply the extension force.
The alignment of such a test piece would be critical, requiring a universal joint
between the extending crosshead of the machine and the test piece. Failure to do this
would cause a stress intensification where the ring makes contact.

6.2.4  Flexion and Flexure Testing

Direct tensile measurement is therefore not a practical proposition and testing is

performed by flexion of bars, rods, discs or other shapes, which easily allow repre-
sentative samples to be made. It will be shown that it is possible to normalise the
results of flexure tests to a tension equivalent. The theory behind flexure testing will
now be examined in some detail.

6.2.5  Beam Theory [12]

In the process of bending a material, cross section mm and pp (Fig 2.1) rotate with
respect to each other about axes perpendicular to the plane of bending, so that lon-
gitudinal fibres on the convex side are extended and longitudinal fibres on the con-
cave side are compressed.
358 D. Daily

Fig. 2.1  Beam Theory

r O

M m p
M
n n¢

s δ y s¢ x

m p
y

Fig. 2.2  Neutral Surface

n′
dA
y
n
x

There is also some intermediates surface which is neither extended nor com-
pressed, called the “neutral surface” represented by nn′. The radius of curvature r is
measured from the neutral surface nn′. Let ss′ be any surface at a distance y from the
neutral axis, then, from similar triangles,

nn¢ ss¢ nn¢ + d x d x

= = = (6.1)
r r + dy r + dy dy

dx dy
strain =Îx = =
nn¢ r
sx
From Hooke’s Law, =E (6.2)
Îx
dy
Therefore s x = E
r
This diagram represents the distribution of stresses. They increase numerically
with distance from nn′.
Let dA represent an elemental area of cross section at distance y from nn′

Ey
Force acting on dA = s x dA = dA (6.3)
r
6  Ceramic Materials Testing and Fracture Mechanics 359

Sum of all such forces over the cross section represent a couple resultant in the x
direction, therefore = ∅

Ey E

ò r
dA = òydA = f
r
(6.4)

Ey
Moment of force acting on dA, with respect to nn′ is dA × y .
r
Adding all the moments over the cross section, and equating to the moment of
the external forces M,

Ey 2 EI
dA = z = M (6.5)
r r

6.2.5.1  Rectangular Section Test Piece

1 M
This defines the relation = , where I = òy 2 dA (6.6)
r EI
Iz is the “moment of inertia of bending” (analogous to the expression used in describ-
ing a rotating body).

Fig. 2.3 Rectangular
Section Test Piece y

h/
2

dy
C z

h/
2

b z
360 D. Daily

For a rectangle,
h/2
bh 3
I z = 2 ò y 2 bdy = . (6.7)
12
0
For a circular section (diameter d)

p d4
Iz = . (6.8)
64

Eliminating r, we have,

My
sz =
Iz

For a rectangle,
h
2
bh 3
I z = 2 ò y 2 bdy = (6.9)
12
0
For a circular section (diameter d)

TTd 4
Iz = (6.10)
64

Eliminating r, we have,

My
sz =
Iz

Maximum tensile or compressive stress is in the outermost fibre. If nn′ is the middle
of the section, and the thickness = h
Mh
(s x )max = (6.11)
2I z

6.2.6  3 and 4 Point Bending

For a rectangular, four-point bending system,

P ( L - 2)
M=
2 2
6  Ceramic Materials Testing and Fracture Mechanics 361

Fig. 2.4  4 Point Bending p/

2
p/2

p/2 p/2
L
`

p ( L - l ) h 12
Stress in outermost fibre = × × × 3
2 2 2 bh
3 p(L - l)
=
2 bh 2
Frequently 1/4 point loading is used, i.e. where L = 4 a and  = 2a so that maximum
3Pa
stress = 2 . This is the failure stress (sometimes called modulus of rupture)
bh
3Pa ì 3 PL ü
sF - í= ý
bh 2 î 4 bh 2 þ

æ 16 Pa ö
ç For a cyliner s F =
è p d 3 ÷ø

For a three-point bend, let  = f

6 Pa 3PL
sF = = (6.12)
bh 2 2bh 2

æ 32 Pa ö
ç For a cylinder s F = (6.13)
è p d 3 ÷ø

6.2.6.1  Calculation of Deflections

6.2.6.2  Simple Cantilever

EI d2L
P (L - x) = M = = EI 2 *
R dx

Integrate,

dy P æ x2 ö
= ç Lx - ÷+ A
dx EI è 2 ø

362 D. Daily

L
O x

x,y
P
y

Fig. 2.5  Simple Cantilever

dy
Now A = Æ , because = Æ at the origin
dx
P ì Lx 2 x 3 ü
y= í - ý+ B
EI î 2 6þ

Now B = Æ because y = Æ when x = Æ at origin.
Depression at the free end is therefore

PL3
D max = (6.14)
3EI

For a curve y = f ( x ) at a specific point,

3
é1 + ( dy / dx )2 ù 2
1
R=ë 2 2
û =
d y / dx d y / dx 2
2

(In this instance)

6.2.6.3  Three-Point Bend

This is equivalent to two cantilevers symmetrically opposed. Point of application of

P equivalent to simple cantilever fixation point.
Let L be the length of the test specimen

( P / 2 )( L / 2 )
3
PL3
Dmax = = (6.15)
3EI 48 EI
6  Ceramic Materials Testing and Fracture Mechanics 363

L
2 2

p
p p
2 2

Fig. 2.6  3 Point Bending

6.2.6.4  Four-Point Bend

We require to determine:
(a) radius of curvature.
(b) maximum elevation of the centre of the beam above the support points.
(c) depression of ends below support points.
Let h be elevation above supports, then from the similar triangles OPC and CPQ
R - h ( / 2)
=
(  / 2 ) ( 2h )
Neglecting h cf. R,

2 2
= 2 Rh, h =
4 8R
The important aspect of the four-point bend test is that there is a circular arc with a
constant bending moment between the loading points (B and C in Fig. 7). Effectively
the specimen selects its own weakest point for failure within this region.

P ( L -  ) EI
Over BC, M = × =
2 2 R
2 EI
R=
( P / 2)( L - )
( P / 2 ) ( L -  ) 2
h=
16 EI
(6.16)
(  / 2 )
Let C represent the origin. Inclination to horizontal at c =
R
( P / 2)( L - ) 
= arcsin q q =
4 EI
Let right hand part of the beam contain the points (x, y).
Taking moments,

P ìï ( L -  ) üï d 2 y
M= í - x ý = 2 EI
2 îï 2 þï dx

364 D. Daily

h Q θ
p B C (x,y) p
2
l p l 2
2 2

p p
2 2

Fig. 2.7  4 Point Bend Test

dy ( P / 2 ) ìï ( L -  ) x - x üï
2

= í ý+ A
dx EI îï 2 2 þï

æ dy ö
At C, the origin, x = Æ, A = ç ÷ = q
è dx ø
dy ( P / 2 ) ïì ( L -  ) üï
= í( L -  ) x - x + × ý
2

dx 2 EI îï 2 þï

By integration,

( P / 2 ) ïì ( L -  ) x ( L -  ) × x ïü
y= í x- + ý+ B
2 EI ïî 2 3 2 ïþ

At C, x = f , y = f , therefore B = f . L -
To find the depression at the ends, let x = .
2
Depression of the ends below the origin,

( P / 2 ) ìï ( L -  ) ( L - ) üï
3 3

+ ( L - ) × ý
2
= í -
8 EI ï 2 6
î þï
For ¼ point, symmetrical loading, let L = 4 a , 1 = 2 a
6  Ceramic Materials Testing and Fracture Mechanics 365

( P / 2 ) æ 8a 3
8a 3 ö
= ç - + 8a 3 ÷
8 EI è 2 6 ø (6.17)
2 Pa
=
3EI
These deflection formulae may be used to determine the elastic modulus. At high
stresses the machine deflection becomes significant, and a blank run has to be per-
formed to determine this.

6.2.7  Test Configurations

6.2.7.1  Biaxial Flexure of Discs

RL is the loaded radius; RO is the outer radius; RS is the supported radius; t is the
plate thickness; P is the loading force; V is Poisson’s ratio.
Principal stresses in the plate are s r = s ¶

3P ìï RS (R32 - RL2 ) üï
sr = s¶ = í(1 + v )1n + (1 + v ) ý (6.18)
2p t 2 RL 2 R02
îï þï
These are effectively the failure stresses for the ceramic.
This test is especially valuable for ceramics in that:
(a) it tests simultaneously in two dimensions at right angles.
(b) there are no sharp edges in the stress field to initiate cracks.
(c) discs are easy to press, fire and polish.
(d) they take up little room in a furnace, and a batch of ten can expect to have iden-
tical firings.

Fig. 2.8  Biaxial Flexure

Disc Test

RL

Ro o

RS
366 D. Daily

Vertical guide
rods for ball

Ball bearing

Brazed-on Loading Ring

Disc under test

RL Hardened support ring

Rs
Rc

Fig. 2.9  Test configuration

Fig. 2.10  Forces Applied P

During test

Alternatively the hardened support ring may be replaced by three balls at the
apexes of an equilateral triangle. This overcomes the problem of flatness in discs.
RS becomes the radius of the circumcircle of the latter.
(Ref. Ph.D. Thesis, A.D. Sivill, University of Nottingham, 1974) [32].

6.2.7.2  Diametrical Compression Test [19, 29]

A disc specimen is compressed between two flat plates. Tensile splitting occurs
along the loaded diameter AB. It was developed originally in Brazil and Japan for
testing concrete about 1943. A biaxial stress distribution is set up in the disc, pro-
vided the thickness is small compared with the diameter.

2P
s max = (6.19)
DtTT
6  Ceramic Materials Testing and Fracture Mechanics 367

6.2.7.3  Brittle Ring Test [19, 31]

6W × ( OD + ID )
= . (6.20)
p × t ( OD - ID )
2

(Tensile stresses are indicated at W, X, Y, Z.)
This configuration is especially useful for high temperature testing in that it
avoids the use of testing jigs (at 1600–1700 °C it would be deforming as fast as the
test piece). The specimen undergoes fracture in four places, often showing separate
maxima on the testing machine. In some respects it resembles two three-point bend
tests superimposed above and below the horizontal diameter. The mathematical
derivation is tedious, but the test is capable of high reproducibility.

6.2.7.4  Theta Specimen [13, 14, 19]

This is also used for high temperature strength determinations. Photoelastic studies
indicate that central bar is in pure uniform tension. This pattern was selected out of
60 shapes, but the difficulty of precision shaping has restricted its application. This

Fig. 2.11  Brittle Ring Test W

I.D
w z
OD
y
t = thickness

Fig. 2.12  Force applied P

during Brittle Ring testing

D
368 D. Daily

should no longer be a problem exploiting the high reproducibility of the computer-­

machining technique of P. Barnes and D.F. Dailly, at the North Staffordshire
Polytechnic. The relation to evaluate the stress in the central bar is

s = kP / Dt (6.21)

σ = stress in bar, P = applied load, D = outer diameter, t = thickness: k is a con-

stant = 13.8 (Daniels) or 16.4 (Durelli).
K can be easily determined using a strain gauge and a knowledge of the elastic
modulus.

6.2.8  Significance of Test Results

To appreciate the problems associated with ceramic design it would be useful to

compare the elastic behaviour with those of metals. For a non-ferrous metal such as
copper (Fig. 14), the stress developed as a consequence of the strain can be repre-
sented in two stages. The initial region is Hookean, where the slope of the curve is
the elastic modulus, but beyond this it departs from linearity and the resultant stress
is proportionately less because of slip caused by dislocation movement on the
favourably oriented slip planes in the crystal lattice. In an FCC metal like copper,
with four slip planes per unit cell this process is easy. With hexagonal or body-­
centred cubic systems slip is less easy.
It is therefore usual to specify such metals in terms of the ultimate tensile strength
(UTS) and a 1 % or 0.1 % proof stress, the percentages referring to the amount of
permanent deformation resulting from the strain (Fig. 2.14).
For ferrous metals the situation is slightly different, in that a “yield point” is
displayed. Once again the dislocations are responsible in that a threshold stress has
to be supplied to overcome the stabilising effect of the interstitial carbon and nitro-
gen atoms which have anchored themselves in the strained region of the lattice
immediately adjacent to the dislocation core. We have therefore to specify the UTS
and the yield point (Fig. 2.15).
For a ceramic (Fig. 2.15) the stress–strain graph displays an entirely Hookean
behaviour, and this continues until the test-specimen breaks. However, the fracture
stress can vary over a wide range, and no ultimate tensile stress is able to be speci-
fied. For a design specification this introduces an entirely new concept. It must be
recognised that there is no realistic lower limit below which the test specimen can-
not fail, and the best that can be given is finite probability of failure under a given
tensile stress. It will further be shown that many factors are likely to influence the
value of the strength of a ceramic. These may emerge from the conditions under
which the test is performed, the size of test piece chosen, the smoothness of its sur-
face which is under tension and the particular configuration of the test. So that all
test results are comparable, there should be a normalisation of the results to unit
volume under pure tension.
6  Ceramic Materials Testing and Fracture Mechanics 369

Fig. 2.13 Stress/Strain
Graph of Non- Ferrous
Metal

Fig. 2.14  Ultimate Tensile

Strength of Mild Steel

6.2.9  Processing of Failure Data [7]

The following expression was given by Weibull for the probability of failure, and it
is especially applicable to brittle materials.
ìïæ -s - s ö m üï
Pf = 1 - exp íç u
÷ ý (6.22)
ïîè s 0 ø ïþ

370 D. Daily

Fig. 2.15 Stress/Strain
graph of Ceramic
(Hookean behaviour)
*
*
*
σ * CERAMIC

σu is a “threshold stress” below which the probability of failure is zero. The thresh-
old stress when compared to the mean stress is relatively small and typically set as
zero rather than given a finite number which may result in over estimation of the
probability of failure. It does have significance for glasses.
σ0 is a normalising factor.
σ is the stress in the material.
m is the “Weibull modulus” and it is a measure of the variability of σ. The higher
the value of m, the more consistent the material.
The function is valid for all stresses in the range s u £ s £ ¥ .
Failure probabilities are in the range 0 £ Pf < 1 .
For ceramic materials it is usual to find m in the range 5 < m < 20.
The curve can be fitted to the experimental points by the manipulation of the
values m, σu and σ0.
Stress/mean stress of batch.
(Ph.D. University of Nottingham 1973)
While this may be the best method of obtaining m, it is time-consuming and
laborious, but with the help of a few assumptions a very rapid procedure is possible,
particularly if a computer is employed.
Assume s u = 0. Let F be the probability of failure corresponding to an applied
stress σ.

ìï æ s ö m üï
Then F = 1 - exp í- ç ÷ ý
ïî è s 0 ø ïþ
m -- -- Takelogs
1 æs ö
Then = exp ç ÷
1- F è s0 ø
m -- -- Takelogs again
1 æs ö
ln =ç 0 ÷ ,
1- F è s ø

6  Ceramic Materials Testing and Fracture Mechanics 371

1.0
Firing batch C
Weibull modulus m = 20.6
Number of Specimens N = 49
0.9

0.8

0.7

0.6
Failure propability p

0.5

0.4

0.3

0.2

0.1

0.0
0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3
Stress/mean stress of batch

Fig. 2.16  Weibull distribution of the fracture data of a sintered batch of Reaction bonded silicon
nitride

1
= m ln s - m ln s 0 .
ln. ln. (6.23)
1- F
æ 1 ö
Weibull graph paper is ruled with ln.ln.ç ÷ as the ordinate units and one
è 1- F ø
­simply plots F against σ to obtain straight line graph of which m is the slope.
1 1
Note : when s = s 0 ln. ln. = 0 ln =1
1- F 1- F
372 D. Daily

exp -1 = (1 - F ) F = 1 - exp -1 = 0.632

This makes σ0 the value of σ when 63.2 % of failures have occurred.
If one fails to obtain a straight line, or there is a sharp gradient change, it is a sign
that more than one process is occurring during the failure, e.g. if there is an unex-
pected decrease in strength at the lower end of failure stresses it suggests a process-
ing fault, or a bimodal distribution (Fig. 2.22) suggests a wide disparity of particle
sizes (e.g. in abnormal grain-growth).

6.2.10  Ranking (or How We Arrive at the Value of F)

One must remember that the purpose of the Weibull statistics is to be able to take a
limited number of data from the breakage of test-specimens and obtain a reasonable
value for m. This will act as a control for our processing schedule. Twenty is a con-
venient number, but if the quantity of raw material is limited, even ten results can be
useful.
Consider a very large population of N data. If one value has to be chosen at ran-
dom, then each has a 1/N chance of being picked.
If the data is placed in ascending order, the chance of picking a value less than,
or equal to the ith number in the list is i/N.
Each value in the data list is therefore associated with a cumulative probability
of P = i/N.
Let the large circle contain N data (in a bag if the analogy is helpful). Let us
withdraw i′ data at random and arrange it in ascending order. (Think of i′ = 20 for a
convenient number.)
Assemble the data in order on the ladder diagram (Fig. 2.18) and represent it with
the full lines. Return the data to the bag, shake it and withdraw another 20 (i″).
Set these up in ascending order, and indicate them with the short dashed lines.
Return, shake and withdraw another 20 (i‴).

Fig. 2.17  “N” data ring

i′′
i′

i′′′′
i′′′
6  Ceramic Materials Testing and Fracture Mechanics 373

Fig. 2.18  Ladder Diagram

′′′
i20
i′ R20
′′′′
i20 ′′ 20
i20

i10′′′′
′
i10
′′′
i10 R10
i10

i9′′′ i9′
i9′′′′
i9′′ R9

i8′′′′ i8′′
i8′ R8
i8′′′
i7′′
i7′ R7
i7′′′′
i7′′′
i6′
i6′′′′ i′′′ R6
6
i6′′

i5′′′′ i5′′′
i5′ R5
i5′′

i4′′′ R4
i4′′′′ i4′′ i4′
i3′′
i3′′′′ i3′ R3
i3′′′
i2′′′′
i2′′′
i2′ R2
i2′′

i1′′′′ i1′′′
R1
i1′′ i1′

Set these up in ascending order, and indicate with the long dashed lines. Return,
shake and withdraw another 2(i″″) (dotted lines).
Repeat until i′ is obtained.
Each sequence of first values  i′, i″, i‴, .. .. .. .. i′
second values  i′, i″, i‴, .. .. .. .. i′
third values  i′, i″, i‴, .. .. .. .. i′
twentieth values  i′, i″, i‴, .. .. .. .. i′
is called a “rank”.
Each value is a binomial distribution of 20 values (or other number if chosen)
which will have a most probable value represented by the mean or median of the
distribution. There are formulae for calculating these or there are tabulated values
for convenient numbers. Procedure:
When an appropriate number of tests have been made,
1 . the results are arranged in sequence.
2. the table of “F” values for the number of tests is obtained (Table 6.1).
3. using Weibull paper, we plot F is ordinate and σ as abscissa the lowest value of
F is made to correspond with the lowest failure stress and so on.
4. using the least squares method, the best line is obtained through the points.
5. gradient of the line is m, the Weibull modulus.
374 D. Daily

Table 6.1 Median Rank Sample size

ranks (F50) [7]
5 10 20
1 12.95  6.70  3.41
2 31.38 16.23  8.25
3 50.00 25.86 13.15
4 68.62 35.51 18.06
5 87.06 45.17 22.97
6 54.83 27.88
7 64.49 32.80
8 74.14 37.71
9 83.77 42.63
10 93.30 47.54
11 52.46
12 57.37
13 62.29
14 67.21
15 72.12
16 77.03
17 81.95
18 86.85
19 91.75
20 96.59

By producing the line backwards it is possible to predict the failure probability at

any stress loading. This is probably the most satisfactory way to specify the strength
of a brittle material such as a ceramic.

6.2.11  Nature of the Test [7]

To compare three point bend specimens with those tested in pure tension, the fol-
lowing expression relates

1/ m
s B3 é v ù
= ê( m + 1) × T ú
2
(6.24)
sT ë vB 3 û

For a Weibull modulus of 10 and specimens of equal volume,

s B3
= 1.614 (6.25)
sT

6  Ceramic Materials Testing and Fracture Mechanics 375

0.999 DW05 m= 6.52

0.990
0.900

0.632

pf

0.100

0.010

0.001
1 10 100 1000
σf / MP a

Fig. 2.19  95 % alumina 1600 °C 120 min. 50 mm × 3 mm bars, ¼ point loading

0.999 WRING m= 13.34

0.990
0.900
0.632

pf
0.100

0.010

0.001
1 10 100 1000
σf / Kg

Fig. 2.20  Ground and refined alumina ring specimens

376 D. Daily

0.999 SRING m= 6.62

0.990
0.900

0.632

pf

0.100

0.010

0.001
1 10 100 1000
σf / Kg

Fig. 2.21  Grand alumina ring specimens

0.999 AR01 m= 2.43

0.990
0.900

0.632

pf
0.100

0.010

0.001
1 10 100 1000
σf / MPa

Fig. 2.22  AIS alumina, showing bimodal distribution

6  Ceramic Materials Testing and Fracture Mechanics 377

Comparison of four point with tension uses the following expression

1/ m
s B4 é V ù
= ê( m + 1) × T ú (6.26)
sT ë VB4 û

and for a Weibull modulus of 10 and equal volumes of test pieces

s B4
= 1.27 (6.27)
sT
Relating four point to three point under the same conditions gives

s B4
= 0.785 (6.28)
s B3

6.2.12  Volume of Test Piece [7]

Because brittle fracture depends on the largest crack in the stress field growing until
it reaches a critical length, the larger the volume of the test specimen the higher will
be the probability of a crack being present which is large enough to cause failure. It
is for this reason that one uses large test pieces, and chooses four-point loading
instead of three-point. The relationship is simple, i.e.

1
s v1 æ V2 ö m
ç ÷ (6.29)
s v2 è V1 ø

In a rearranged from, s V1 V1 = s V2 V2 . This expression will be used in the following
m m

chapter.

6.3  Strength of Solids [11]

6.3.1  Theoretical Strength of Crystals

The shear stress at which deformation should begin in an ideally perfect crystal was
calculated by Frenkel in 1926. Referring to Fig. 2.1 representing two adjacent
planes of atoms in a simple crystal structure, we see that if the upper plane were
displaced relative to the lower in the slip direction b, causing atoms A′, B′, C′, etc.
to lie over B, C, D, etc. respectively, slip would have taken place. The crystal would
remain perfect as the atoms had moved to equivalent lattice sites. The shear stress
378 D. Daily

necessary to produce this permanent plastic deformation is the critical shear stress
τc of the ideally perfect crystal.
To obtain its value we first consider the form of the stress displacement curve
(Fig. 2.2) measuring displacements indicated in Fig. 2.1.
Now τ = 0, when x = 0 and x = b;
The crystal is in equilibrium in both cases.
1
When x = b , the atoms in the upper layer can be seen to be in a position of
2
unstable equilibrium; the force in the slip direction is zero, for each atom is equally
attracted to the right and to the left.
1
Consequently τ = 0, also at x = b .
2
Further, we know from the elastic behaviour of crystals that the shear stress
increases linearly with strain within the elastic range. The slope of the displacement
curve at x = 0 and x = b must therefore comply with Hook’s law.

t =G (6.30)
g
where
Gx
g = x / a,t = (6.31)
a

Thus

dx
= G / a ( x  a)
dx
Now τ(x) must be a periodic function with wavelength b, and we therefore taken
the relation as typical,

æ 2p x ö
t = K sin ç ÷ (6.32)
è b ø
1
Other harmonic functions satisfying the boundary conditions at x = 0, b and b
2
could have been chosen, without affecting the result. As Eq. (6.32) must reduce to
Eq. (6.30) for small strains, one finds that

x ( 2p x )
t =G =K
a b

Re-presenting a very small angle, so that

Gb
K= (6.33)
2p a
6  Ceramic Materials Testing and Fracture Mechanics 379

The highest stress occurs when

x 1
2p = p
b 2
1
i.e. at x = × b , and is numerically equal to K. In practice a and b will not differ
4
from one another appreciably and one therefore obtains the result that the critical
shear stress of the ideally perfect crystal is

t c » G / 10 (6.34)
1
The maximum elastic shear strain occurs when x = b and is therefore equal to
4
1 b which, with a ≈ b, yields
,
4 a
g c » 25 % (6.35)

Perfect crystals necessary to check this theory can only be obtained in the form
of microscopically thin whiskers. They have, nevertheless, established the correct-
ness of the foregoing results. However, the critical shear stress of pure copper is
only about 20 kg/cm2, while G/10, Eq. (6.34) yields about 40,000 kg/cm2. The
strength of the ideal crystal is therefore about 2000 greater than the real one. In met-
als this discrepancy is accounted for by the presence of dislocations, while in non-
metals, and in particular in ceramic substances, the discrepancy is caused by
cracking, although both processes can be present to some degree in either class of
material.

6.3.2  Elastic Constants and Interatomic Forces [11]

Ae2 B
If we recollect the Born equation, which is u = + n and plot U, the potential
r r
energy against r, the interionic distance, the curve displays a minimum of energy,
resulting from a balance of attractive and repulsive forces. This minimum is at the
distance r0, which is the equilibrium interionic spacing in the lattice. (The first
term in the Born equation represents the Coulombic energy, and the latter repre-
sents the repulsion caused by quantum exchange energy as the electron shells of
adjacent ions approach.) Similar curves exist for covalent and van der Waals
bonds. The elastic constants can be estimated from the function U. A crude calcu-
lation illustrates the principle involved. The force between the atoms is du dr and
stress is

æ du ö 1
s » ç ÷× 2 (6.36)
è dr ø r0
380 D. Daily

dr = force
r0 = area of one atom plane
i.e. the force divided over the approximate area over which it operates. Now

dr
ds = cd Î= c (6.37)
r0

and thus

ds 1 æ ¶ 2 u ö
c = r0 ~ ç ÷ (6.38)
dr r0 è ¶r 2 ø

For a shear deformation, the Born repulsive term may be neglected, and thus for
ionic crystals u µ 1 r Substitution in Eq. (6.32) predicts a proportionality between
0
C44 and r0−4.
When we try to shear the lattice we have to overcome the peak of maximum
resistance. Over small displacements the stress-distance curve is approximately a
sine wave so we may write

P
s = s th sin ( r - r0 ) (6.39)
2a
where a is the atomic displacement corresponding to the theoretical strength, and
the material is assumed isotropic. ds
If E is the elastic modulus, then, E = , and
dÎ
ds ds r0s th P P
= r0 = cos ( r - r0 )
dÎ dr 2a 2a
For very small strains, r–r0 → 0, the cosine term → 1, so that

2 Ea
s th = (6.40)
P r0

To estimate the constant a, Orowan’s argument is that the work done in stressing
the material to σth, must be at least equal to the energy required to create two new
fracture faces. If γ0 is the surface energy per unit area,
r0 + 2a
2 Ea P
ò s dr = ò sin ( r - r0 ) dr = 2g 0 (6.41)
r0
P r0 2a

and this reduces to

4 Ea 2
= g0 (6.42)

P 2 r0
6  Ceramic Materials Testing and Fracture Mechanics 381

Eliminating a using Eq. (6.35) gives

1
æ Eg ö 2
s th = ç 0 ÷ (6.43)
è r0 ø

High surface energy and stiffness, together with a small lattice spacing, contrib-
ute to high strength, although a high fabricated density is also required. It has been
shown that α ≈ 0.14 r0 for ionic crystals. Note in the previous hypothetical approach
r
a = 0 substituting in Eq. (6.37), gives
4
Er0
g0 » (6.44)
125
-10
Substituting values for alumina where E = 4 ´ 10 N / m , r0 = 4 ´ 10 m gives
11 2

-2
g 0 » 1 × 2 J m (This is for a sapphire and not a polycrystal) and agrees with experi-
mental data.
In general, two lines of approach to the study of fracture have been made.
These are:
1. The Griffith’s approach, where the reduction of strain energy in the test sample
is equated to the increase of surface energy of the crack.
Formulae for fracture energy have been derived and practical evaluation methods
described.
2. The elastic continuum approach, where the intensified stresses at the edge of the
crack-tip are considered. From this the ‘Stress Intensity Factor’ can be defined,
the critical value of which is a materials property can be used in engineering
design. Equations from the two approaches can be related to provide a number
of useful testing methods to evaluate the fracture toughness of a material.

6.3.3  Griffith’s Theory [2]

6.3.3.1  Basic Derivation of Expression

Let a plate contain a thin elliptical crack of length 2c, represented by AOB in
Fig. 3.1, under a uniform stress σ. To find an expression for the relaxation of the
strain energy as the crack is formed, assume that the plate consists of a number of
independently operating linear elements. Behind the crack front there will be a
stress free zone, caused by the shrinkage of the elements. Let it be assumed that the
length of this zone, measured along each element from the line AOB representing
the centre of the crack is y0. This figure will obviously vary from zero at the crack
edges to y0 in the centre.
382 D. Daily

Fig. 3.1  Elliptical Crack

Diagram Y

σ
2

0
A B z

σ
2

Fig. 3.2  Stress Free Zone

AOB

yo

yo
(max)

uy

A 0 x B
C C

The original strain can now be calculated:

Strain imparted to each element is

my E
= (6.45)
y0 s /2

E
Therefore m y = y0
s /2

uy represent the y coordinates of the ellipse and y are lengths of the stress-free zone.
For the ellipse,

x2 my
2
b2 2
+
c2 b2
= 1, m 2
y =
c2
c - x2 ( ) (6.46)

2
æ E ö b
y02 × ç ÷ =
ès / 2 ø c
2 (
c2 - x 2 ) (6.47)

6  Ceramic Materials Testing and Fracture Mechanics 383

Consider an element at x = ∅

2
æ E ö
y02( max ) ç ÷ =b
2

ès / 2 ø
y02 c 2
y02( max )
(
= c2 - x 2 )

The locus of the lengths of y0 is a circle if

c2
=1
y02( max )

We may confirm this if we let x = c/2
y2
= 3 / 4 (Fig. 3.2)
y02( max )
y 3
=
y0( max ) 2

3
Since 0 x = c 2 , xy = , then 0 y = c . Which is a radius of the same circle.
2
Therefore y0 lies on a circle, and y0 = c - x
2 2
( )
In order to compute these relaxing stresses, we require to know the displacement
of the upper surface of the ellipse, given by the following expression:

æ E ö
my = (c 2
- x2 × ç)
çs ÷
÷ (6.48)
è 2 ø
Work done by the relaxation of the elements (με) is also the strain energy released,
and is given by,

UÎ = 1 sm y dx ( per element )
2
c
UÎ = 1 òsm y dx ( per quadrant )
2
0

c
UÎ = 2 òsm y dx ( for circle ) (6.49)
0
384 D. Daily

That is

s2 c
UÎ = ò (c )
- x 2 dx
2
(6.50)
E
0
c

ò (c )
- x 2 dx = 1 / 2 x (c )
- x 2 + 1 / 2c 2 is sin(x/a) + c
2 2

0
When x = ∅, expr. = ∅; when x = c, expr = Πc2/4.
Leaving the notion of independent elements and returning to the continuous
sheet, the elastic constraints must now be put in, so that the expression becomes
modified to

s2 c
UÎ =
E
(1 + r ) ( k + 1) ò (c 2
- x2 ) (6.51)
0
where k = 3–4r for plane strain,
(3 - r )
and k = for plane stress
(1 + r )
(r is Poisson’s ratio).
On integration, Eq. (6.51) gives

s 2 P c2
U= (1 + r ) ( k + 1) (6.52)
4E
This expression represents the total crack surface and it should be noted that the
expression is independent of the semi-minor axis of the ellipse b

6.3.4  Griffith’s Criterion [2]

According to the Griffith criterion, the surface energy, Us, needed to create new
surfaces, is supplied by the energy released, Uε, The total free energy, U, for the
cracked body is expressed as,

m = ms - mÎ + m (6.53)

where Ū is the component of the strain energy independent of the crack in the body
æ du ö s2
ç i.e. dc = Æ ÷ and is equal to 2E per unit volume for this system. In Griffith’s
è ø
terminology, U is the total potential energy for the system. In terms of the total
energy of the system

du
< 0 when the crack is unstable
dc

du
= 0 when the crack is in equilibrium
dc

6  Ceramic Materials Testing and Fracture Mechanics 385

du
> 0 when the crack isstable
dc
The surface energy per unit thickness, Us, equals 4cγ0. Substituting for Us and Uε
in Eq. (6.53) we have,

s 2 P c2
u = 4cg 0 - (1 + r ) ( k + 1) + m (6.54)
4E
du
For the equilibrium situation = Æ , and if the applied stress be equal to σf.
dc
du dus duÎ s 2P c
= - = 4g 0 - f (1 + r ) ( k + 1) = 0
dc dc dc 2E
Therefore

8g 0
sf = (6.55)
P c (1 + r ) ( k + 1)

For plane stress and plane strain the expression tales the form,

2 Eg 0
sf = Plane strain (6.56)

(1 + r ) P c
2

2 Eg 0
sf = plane stress (6.57)
Pc
The results of tests carried out by Griffith indicated that the theoretical predic-
tions were justified provided that there was little or no plastic flow in the region of
the crack tip. Because there is always some dissipation of energy due to plastic flow,
the Griffith expression tends to give a higher value than the true thermodynamic one.
The theory does bring out a very important point, namely that the strength of a
solid is dependent on the flaw size in a direction normal to the applied stress, and the
probability of failure will, in turn, depend on the probability of there being a flaw in
excess of the critical size.

6.3.5  Energy Release Rate

We define G as the energy release rate. This is either the strain energy release or the
potential energy release rate, and it is the energy released per unit extension of the
crack front per unit thickness of the body (rate in this instance relates to distance
rather than time).
386 D. Daily

In the opening mode (or mode 1, in Fig. 4.3), if the crack moves at both ends a
distance 2dc, relative to a crack size 2c,
From equation, at equilibrium,

du dus duÎ
= - = 0. (6.58)
dc dc dc
duÎ æ dus ö
G= = dus = 4g 0 dC (6.59)
dc çè dc ÷ø

For the case where the crack operates from one end only,

duÎ æ dus ö
G= = dus = 2ro dC (6.60)
dc çè dc ÷ø

where the crack size is critical (Fig. 3.2) c = c and G is equal to twice the thermody-
namic surface energy.
I.e. G = 4g 0 (for the double-ended crack) [1–5] and 2γ0 (for the single ended
crack).
This defines the strain-energy-release rate per unit width of crack front.
The conclusion of this analysis is that we can relate the failure stress σf with the
size of the crack c, provided we have some knowledge of γ0 the fracture energy, or
G, the strain energy release rate.
What follows is a discussion of the significance of G, followed by practical
­evaluation methods for γ0

reduction of stress increase in strain

after fracture after fracture
J L

th K
ow
Stress

gr
a ck
t cr
ou
th wt
h
wi
gro
with

o N M
Strain

Fig. 3.3  Strain Energy Release Rate

6  Ceramic Materials Testing and Fracture Mechanics 387

Fig. 3.4  Failure stress P

associated with Size of
crack
∇

δc 2c δc

6.3.6  Geometric Illustration of Compliance

An energy balance criterion has been developed by Davidge [3] using a simple geo-
metrical argument (Fig. 3.3). Suppose the stress/strain curve of the sheet containing
the crack is represented by OJ. Let the crack now grow in length from 2C to
2(C + dC), i.e. equal amounts at both ends. The sheet will then be less stiff. The
stress/strain curve is now represented by OKL. When deformation is by dead load-
ing (i.e. constant stress), the strain increases by JL When a rigid machine is used the
strain is constant and therefore the reduction of stress is given by JK. Let the frac-
ture energy per unit area be given by γo. As the crack grows, the energy required to
form new fracture faces is 4γodC.
The elastic strain energy is du.
The external work done on the sheet is dW.
The condition for fracture is therefore,

d ( w - u ) ³ 4g o dC (6.61)

We need to estimate W and U.
Note that the energy is proportional to the area under the graph, so that when
fracture occurs at constant strain, dW = ∅ and an amount of elastic strain energy,
proportional to OKJ is released (dU is negative).
When fracture occurs at constant stress, dW is proportional to JLMN and the
elastic strain energy increases proportional to OLM − OJN = OJL. Because JKL
becomes negligible in the limit, d(W − U) is again proportional to OJK.
388 D. Daily

Note that dW = 2dU; half the external work is absorbed as elastic energy, and half
is available to assist crack propagation.
For either type of loading, we can continue the discussion solely in terms of U.
The condition for fracture reduces to
du
³ 4g o (6.62)
dC

6.3.7  Analytical illustration of Compliance

Following Lawn and Wilshaw [1], another approach to the understanding of G is as

follows (Fig. 3.4):
A crack is made in sheet of brittle solid of unit thickness and one end is loaded
with a tensile force P. The test piece behaves according to Hooke’s law as follows:

d = Pl (6.63)

where δ is the elongation of the test piece and λ is the elastic compliance. The strain
energy is therefore,
d
1
uÎ = òPdd = Pd
2
o
(P is a function of δ and is area under curve)

1 2 1 d2
= P l= (6.64)
2 2 l

Fig. 3.5  Hookes’s Law

δ
6  Ceramic Materials Testing and Fracture Mechanics 389

Let us now consider two extreme loading configurations,

1. Constant force (i.e. dead-weight loading)
Let U be the strain energy in the test piece, and W be the work done by the load,
then, as the consequence of crack-growth,

d w » Pd (d ) = P 2dl (6.65)

1 2
du = P dl (6.66)
2
and the total change in mechanical energy is

-1 2
d ( -w + u ) » P dl (6.67)
2
2. Constant displacement

dW = O
1 (6.68)
d u = - P 2 dl
2
giving

1
d ( -w + u ) = - P 2dl (6.69)
2
It will be seen that the two results are identical, i.e. the mechanical energy
released during incremental crack extension is independent of the loading configu-
ration. We can therefore define a crack extension force, which is the strain energy
release rate with respect to the crack length in a test piece of unit thickness.
The strain energy release rate is a measure of available potential energy, and the
given the symbol G and is equal to

-d ( -w + u )
dC
and is equivalent to

æ ¶u ö
G = -ç ÷
è ¶c ød

Because W = ∅, for the constant strain situation (i.e. where the fixation is rigid).
390 D. Daily

6.3.7.1  Practical Methods [1]

From equations G =
1 2 dl
2
P
dc and 2 (
1 d2
l2 )dl , G can be deduced in two
dc
ways. At constant stress, all that is needed is the load, and a calibration for the com-
pliance. At constant strain, one needs to know the compliance, and then its rate of
change with the crack-length. Because of the negative sign the strain energy actu-
ally decreases under constant strain, but under constant stress it increases. For this
reason constant strain methods are preferable, because the crack configuration is
stable.
Davidge and Tappin [10] describe a practical method for compliance analysis
with a notched bar specimen. The “stiffness” k is determined from the load/deflec-
tion curve, and it is given by:

P = kd (6.70)

Stored energy at the instant of fracture is given by

PFd F
U=
2
æ ¶u ö
Now g c = - ç ÷ , fracture occurring at a fixed deflection, or
è ¶A ød
æ ¶u ö æ ¶K ö
gc = -ç ÷ ×ç ÷
è ¶k ød è ¶A ød

But,

æ ¶u ö
ç ¶k ÷ = d
2

è ø
and thus

æ ¶K ö
g c = -d 2 ç ÷/2 (6.71)
è ¶A ø

Experimentally, one has to measure the specimen stiffness k as a function of the

initial crack area A = 2bc. For each notch depth used, ¶K ¶A is obtained from the
slope of the curve at the appropriate value of A, as in Fig. 3.4. If we substitute these
values of ¶K ¶A in Eq. (6.70) with the experimental values of δF, a series of values
of γc is given for each notch depth.
The equivalence of the two equations is shown as follows:
Remembering that k(stiffness) =  1 (λ = compliance) and stating Hooke’s law
l
P = kδ (P is the applied force, δ is the deflection)
6  Ceramic Materials Testing and Fracture Mechanics 391

1 2 dl
G= P
2 dc
æ1ö
dç ÷
1 k 1 1 dk
= P2 è ø = - P2 2
2 dc 2 k dc
d 2 æ dk ö bd 2 æç ¶k ö÷
=- ç ÷ =
2 è dc ød 2 ç¶A ÷
è 2 ød
æ ö
d 2 ç ¶k ÷
=- ç ÷
2 ç A÷
ç¶ ÷
è 2 ød
(both sides of a single-ended crack)
Since
G = 2g o (at equilibrium)

G -d 2 æ ¶k ö
gc = = (6.72)
2 2 çè ¶A ÷ød

6.3.8  E
 xperimental Technique for Compliance:
Calculating of γ [10]

Five test bars of ∅.5 cm square section and 4 cm length are prepared as accurately
as possible (allowing for sintering shrinkage) with progressively increasing notch-­
depths, cut accurately with a diamond saw, so that the crack area increases conve-
niently in units of ∅. 1 cm of cross-sectional area. The bar is then subject to a
three-point bend test, and the stiffness is obtained from dividing the applied load by
the deflection obtained. It is advisable to find the average value of a number of tests,
particularly at low crack areas. This deflection may be measured directly by a trans-
ducer or in terms of movement of the chart, making corrections for the deformation
of the load cell of the testing-machine at comparable loads. Note that the first objec-
tive is not to extend the crack or break the specimen, but merely to obtain plots from
æ ¶k ö
which k and ç ÷ can be broken and δF obtained (at the point of fracture).
è ¶A ød
The tabulation is completed for each specimen, the stiffness against crack area
æ ¶k ö
curve drawn, and ç ÷ obtained for each specimen area, by measuring the gradi-
è ¶A ød
ents of the curve.
392 D. Daily

Fig. 3.6  Stiffness versus

Crack Area

δκ
δA δ

STIFFNESS κ
CRACK AREA A - 2bc
General form of κ versus A curve

Fig. 3.7  Stiffness Gradient 12

Curve
(PMMA)
10
(107 dyn/cm)

8
STIFFNESS κ

0
0 0.1 0.2 0.3 0.4 0.5
CRACK AREA A - 2bc (cm2)
κ versus A data for PMMA.

6.3.8.1  Compliance Results

From the knowledge of δF for each specimen, γc can be calculated [10].

6.3.9  Work of Fracture

This requires a crack to propagate in a specimen while its growth is under control.
The stronger the specimen the more difficult this becomes. It is essential to have:
(a) a very hard machine.
(b) the facility for very slow rates of strain.
6  Ceramic Materials Testing and Fracture Mechanics 393

Fig. 3.8  Alumina Ceramic 14

gradient Curve
(Al2O3)
12

10

STIFFNESS κ (109 dyn /cm)

8

0
0 0.1 0.2 0.3 0.4 0.5
CRACK AREA A - 2bc (cm2)

κ versus A data for alumina

Fig. 3.9  Glass Gradient

Curve (GLASS)

2.5
STIFFNESS κ (109 dyn/cm)

2.0

1.5

1.0

0.5

0
0 0.1 0.2 0.3 0.4 0.5

CRACK AREA A - 2bc (cm2)

κ versus A data for glass
394 D. Daily

Table 6.2  Experimental Tabulation Data for Material Testing

æ ¶k ö
-ç ÷
P δ k Ave k b c A = zbc è ¶A ø δF

Testing Speed Crosshead gears

Testing Speed Chart gears

Fig. 3.10 Specimen L
geometry. I = span;
b = breadth; d = crack depth
b

c
d T

Fig. 3.11 Load/deflection
curve. Pf = fracture load; PF
δ1 = fracture deflection; σF = 3 PF L
k = specimen stiffness; 2 bd
2

σf = fracture stress
γF U
=
2b (d-c)
LOAD P

σF
κ
DEFLECTION δ

Controlled cracking can only occur in brittle material when the notch depth is so
large that the test specimen is substantially weakened. One then arrives at a situation
where the residual strain energy in a specimen is less than that of the increase in
surface energy required when fracture occurs. Increasing the external strain then
goes directly into providing the energy for overcoming the surface and propagating
the crack.
Measure the area under the curve to obtain U Record P Calculate γF and δF
6  Ceramic Materials Testing and Fracture Mechanics 395

Fig. 3.12  Controlled load/ 12

deflection curve for
graphite
10 (GRAPHITE)
(c/d = 0.4)

LOAD (κ8)
6

0
1 2 3 4
−2
DEFLECTION (10 cm)

Fig. 3.13 Load/deflection 20
curves for
PMMA. Numbers on graph (PMMA)
refer to notch depth ratio
c/d. X indicates
catastrophic failure
15

0
LOAD P (κ8)

0.1
10 x

0.2
x

0.3
x
5
0.4
x
0.5
0.6
0.7
0.8
0.9
0
-1
10 cm
-1
DEFLECTION δ (10 cm)
396 D. Daily

Table 6.3  Experimental Tabulation for Fracture Mechanics

Work of fracture
L b d c Pδ σF u

Materials: Alumina test-bars L = 8  cm, b = 0.5  cm, d = 1.0  cm

These are cut with diamond saw to give c = 0.7, 0.8, 0.85, 0.9, 0.95 cm
Breaking is carried out at 0.005 cm/min. and the graph is automatically plotted by the recorder
Repeat with Perspex (i.e. polymethylmethacrylate)
L, b, and d similar, c = 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 cm

6.3.10  Significance of  γo, the Fracture Energy

Davidge [3] also draws attention to the fact that the γo in the Griffith equation is an
ideal value, requiring perfectly sharp cracks to propagate through a material and
creating planer fracture surfaces. Fracture in ceramics is not far from ideal, but there
are energy-dissipating processes in what is a relatively complex process. For exam-
ple fracture may be due to:
(a) a combination of transcrystalline and intercrystalline microstructure, manifest-
ing as steps in the fracture surface.
(b) Because of the high stresses near the crack-tip ( approaching…) there is a pos-
sibility that plastic flow may occur.
(c) Cracking branching in the microstructure causing subsidiary cracking.
All of these are energy-consuming processes which make the apparent fracture
energy γ1, somewhat higher than the thermodynamic surface energy γ0. The effective
surface energy γ1. is only a valid concept when plastic flow is localised near the crack
tip and would not apply to metals where plastic flow is the normal deformation mode.

6.3.11  Double Cantilever Bean [15]

Attach the test piece show above, to the Instron in the tensile mode, i.e. P (up) to the
load-cell and P (down) to the cross-head via a universal coupling. Follow the prog-
ress of the crack front (L) and measure 2d with a transducer system.

P up

y
2b

2δ
x w

L
Pdown

Fig. 3.14  Double Cantilever Beam

6  Ceramic Materials Testing and Fracture Mechanics 397

Bending moment is distributed along the beam, and is caused by the applied
force P, so that

M ( x ) = P ( L - x ) where 0 < x < L

Beam theory predicts that the strain energy U will be,

P 2 L2 æ wb3 ö
L
1
( )
2 EI ò0
U- M 2
x dx = ç where I = ÷ (6.73)
6E × I è 12 ø

(E is the elastic modulus and I the moment of inertia of bending)
Deflection of beam at point of application of force is obtained using Castigliano’s
theorem.

æ ¶U ö PL3
d =ç ÷ = (6.74)
è ¶P ø x = L 3EI

Strain energy is therefore

3EId 2
U= (6.75)
2 L3

By the conservation of energy

Pdd + g WdL + dU = 0 (6.76)

Workdone by external force Energy expended in new surface Change in stored energy

From above

dL dd
æ ¶U ö æ ¶U ö 9 EId dL 2
dU = ç ÷ +ç ÷ dd = - + Pdd
è ¶L ød è ¶d ø L 2 L4

9 EId 2 3Eb3d 2
Therefore g w = or g d = ( measuring d ) (6.77)
2 L3 8 L4

6 P 2 L2
( 3.24 ) and ( 3.27 ) give g P = ( measuringP ) (6.78)
Ew 2 b3
Both of these formulas can be used. A problem is to make the crack grow straight,
and to clearly see the tip of it. This will be a problem with opaque bodies, where one
cannot use total internal reflection to see the crack tip. The crack may be guided by
the use a rectangular grooved. Specimen
398 D. Daily

w2 w1
a
b

Fig. 3.15  Test Specimen

dS = g w1 dL
æ ¶U ö -9 EId
2
æ w2 b3 ö
ç ¶L ÷ = 4 ç where I =
12
and where a <<< b ÷ (6.79)
è ø 2L è ø
3Ew2 b d
3 2
6P L2 2
Thus g d = and g P =
8w1 L4 w2 w1b3 E

6.4  Stress Concentrators

6.4.1  Elliptical Cracks [1, 2]

An important historical precursor to the Griffith study was the street analysis of
Inglis (1913) of an elliptical hole in a uniformly stressed plate. This showed that the
local stresses about a sharp notch or corner could rise to several times that of the
applied stress. It thus became apparent that even submicroscopic flaws might be
potential sources of weakness in solids. In the limiting case a crack could be
regarded as an infinitesimally narrow ellipse.
Let us therefore examine the modifying effect of the hole on the distribution of
stresses in the solid plate. Let it be assumed that Hooke’s law applies everywhere in
the plate, and the dimensions b, c (Fig. 4.1) are small in comparison with those of
the plate. The problem then reduces to one of linear elasticity.
Beginning with the equation of an ellipse

x 2 y2
+ =1 (6.80)
c2 b2
One can show that at C the radius of curvature is:

b2
P= (6.81)
c
It is at the point C where the greatest concentration of stress occurs. The appar-
ently simple equation (6.81) is obtained as a result of a complex analysis [1–3, 5]

æ 2C ö é æ C ö 12 ù
s yy ( C ,q ) = s A ç 1 + = s ê
A 1+ ç ÷ ú. (6.82)
è b ÷ø êë è P ø úû

6  Ceramic Materials Testing and Fracture Mechanics 399

Fig. 4.1  Effect of Hole on σA

the Distribution of Stress
in a Solid

−2b O
C X

2c

σA

Fig. 4.2 Stress units of

Concentration 6
σA
5

4
Y
3

2 σ
yy
B
1
b σxx
o

O C X

It appears that stress concentration depends on the shape of the hole rather that
its size. The variation of local stress along OX is also of interest. Figure 4.2 illus-
trates the case where C = 3b. σyy drops from its maximum at C approach the value of
σA asymptotically for large values of x, while σxx rises to a sharp peak within a small
distance from the stress-free surface and subsequently drops to zero at high x. (It can
be appreciated that when C / b = 5, s yy = s max = 11 .)
400 D. Daily

For the case where b <<< C, this equation reduces to

1
s max 2C æC ö 2
= = 2ç ÷ . (6.83)
sA b èPø

(C/P) is “the Stress concentration factor”.
Rearranging this equation, and multiplying by the factor p , yields the follow-
ing equation.

1
s A p c = s max p P . (6.84)
2
in which we can call the L.H.S. K1 = s p C , and is known as the “stress intensity
factor”.
It is important to realise that K1 contains only the macroscopic quantities (the
external tensile stress) and the half crack-length, C, both of which are measurable.
It should be appreciated that the Eq. (6.95) implies a physical limit restricted by the
properties P and σmax without breaking at the tip of the crack, for P is given by the
microstructure of the material, and σmax cannot be larger than the internal molecular
strength, σM. There should exist therefore a specific material limit given by

1
K1C = s M p P (6.85)
2
the so-called “critical stress intensity factor”, for “fracture toughness”.
In Davidge’s treatment [3] the expression (6.85) is related to the Orowan value
for the theoretical strengths of solids, i.e. the latter value for σth is made equivalent
to the stress at the edge of the crack tip. For the failure this gives

E P
sf = × . (6.86)
4C ro

Now if the tip of the crack were infinitesimally small the stresses would be infi-
nitely large. It is reasonable to assign the value half an atomic spacing to P if the
crack passes between adjacent planes of atoms. Substitution in (6.38) gives a second
value for i.e.
1
æ Eg ö 2
sf = ç o ÷ (6.87)
è SC ø
The stress is somewhat lower than the Griffith stress at which therefore, fracture
should certainly occur and Orowan’s value for could only be accurate to within a
factor of two. Hence we conclude the Griffith criterion is adequate and sufficient.
A criticism of this solution is that the elastic stresses in the direction of the
applied force rapidly reach infinity as the crack becomes sharper, and so become
insensitive to the differences in the mode of cracking and the stress configurations.
6  Ceramic Materials Testing and Fracture Mechanics 401

6.4.2  K1 Values for Specific Crack Systems [2]

1. Internal crack in infinite body length of crack = 2a app. stress = σ.

K1 = s p a . (6.88)

2. Internal crack in a plate of finite width.

As above, only width = 2w.

é æ pa ö 2 ù
1

K1 = s ê2 w tan ç ÷ ú (6.89)
êë è 2W ø úû

Expanding the above expression we have,

é p 22 ù
K1 = s p a ê1 + a ú ± (6.90)
ë 24W û

as a/w → θ, becomes
3. Surface crack in a semi infinite body.

K1 = 1.12s p a (6.91)

Fig. 4.9a  Internal Crack in σ

plate with infinite body
length

2a

σ
402 D. Daily

Fig. 4.10a  Internal crack σ

in plate of finite width

2σ

2w

Fig. 4.12a  Surface Crack σ

4. Two symmetrical edge cracks in a plate of finite width.

1
é æpa ö æ p a öù 2
K1 = s ê2 w tan ç ÷ + q .2 w sin ç ÷ú . (6.92)
ë è 2w ø è w øû

5. Central penny-shaped crack in an infinite body.

K1 = 2s a / p (6.93)
6  Ceramic Materials Testing and Fracture Mechanics 403

Fig. 4.13a Two σ
Symmetrical edge cracks
in plate of finite width

2w

Fig. 4.14a  Central penny

shaped crack in an infinite
body

2a
A B

Fig. 4.14b  Plan view of

Penny Crack

A B
Plan

6. Surface crack in a three-point bend system.

Flexural strength in a system without a surface crack is given by

3 P1
a= . 2 (6.94)
2 bd

b = breadth, d = thickness, 1 = span.

404 D. Daily

Fig. 4.15a  Surface crack P

in 3pt bend System
d
a
P/ P/
2 2

When a surface crack is introduced, and 1 = 4d.

3P1 3 é ù
1 3 5
æaö 2 æaö 2 æaö 2
K1 = ê1.93 ç ÷ - 3.07 ç ÷ + 14.53 ç ÷ +ú (6.95)
2bd 2 ê èdø èdø èdø úû
ë
for ∅.2 < a/d < ∅.6.
In terms of the flexural strength, the expression (6.95) becomes

é æaö ù
K1s a ê1.93 - 3.07 ç ÷ + ú
ë èdø û

This application is continued in the testing methods latter on.

Function have been developed which satisfy the requirements of linear elasticity
theory for the stresses in the vicinity of a sharp slit. Their derivations are mathemati-
cally complex, and only the solutions are reported here. Reference can be made to
the following text-books and papers [1–5].

6.4.3  Crack Propagation [1, 2, 4, 5, 20]

A crack may propagate in three different modes as represented in Fig. 4.3. While the
opening mode is likely to predominate, the other modes do exist, and combinations
of mode 1 with others are possible. The other modes will only operate if the crack
or the test piece is not free to move into a favourable orientation for mode 1. The
modes are referred to as
1 . the opening mode
2. the shearing or sliding mode
3. the tearing mode
6  Ceramic Materials Testing and Fracture Mechanics 405

P
P P

P P

Fig. 4.3  Open Mode, shearing or Sliding Mode, Tearing Mode fracture of a Brittle Material

τxy σyy

y
σxx

τyx
r

Fig. 4.4  Stress Analysis at Crack Tip

6.4.4  K4 Application: Stress Analysis at Crack Tip

The stress distributions near to the end of a Griffith’s crack has been analysed as
follows:
406 D. Daily

6.4.4.1  Mode 1 Stresses

s xx ü =
ìcos q
ï 2 ( ) éêë1 - sin (q 2 ) sin (3q 2 )ùúû ( tensile )
ï K ïï
s yy ý =
ï 2p r
ícos q 2
ï
( ) éêë1 + sin (q 2 ) sin (3q 2 )ùúû ( tensile ) (6.96)
t xy þ =

ïsin q
ïî 2 ( ) cos (q 2 ) cos (3q 2 ) ( shear )

s zz = n (s xx + s yy ) ( P. strain ) ( tensile )
= q ( P. stress )
t zz = t yz = q ( shear )

6.4.4.2  Mode 1 Displacements

(1 +n ) éêë( 2 K - 1) cos (q 2 ) - cos ( 3q 2 )ùúû

K æ r ö 2
m=
2 E çè 2p ÷ø
1

(1 +n ) éëê - ( 2 K + 1) sin (q 2 ) + sin ( 3q 2 )ùûú

K æ r ö 2
v= (6.97)
2 E çè 2p ÷ø
W = ( vz / E ) (s xx + s yy ) ( Plane strain )
= q ( Plane strain )

These equations follow from those derived in Ref. [5] an example of which is

c
sx =s × cos (q / 2 ) (1 - sin (q / 2 ) sin ( 3q / 2 ) (6.98)
2r
σ represents the general stress in the material, and the trigonometrical terms are the
resolved components of it.

6.4.4.3  For Mode 2

Similar expressions also exist e.g.

s xx = 2 êë ( )( ( ) ( )
ì- sin q é 2 + cos q cos 3q ù
ï 2 2 úû
ïï
K
s yy = p
2p r
í
ï
2 ( ) ( ) ( )
sin q coss q cos 3q
2 2
t xy =
ïî 2 êë ( )( ( ) ( )
ï cos q é 1 - sin q sin 3q ù
2 2 úû
s zz = t xz = t yz = q for plane stress, (6.99)
= n (s xx + s yy ) and t xz = t yz = q for plane strain

6  Ceramic Materials Testing and Fracture Mechanics 407

6.4.4.4  For Mode 2 Displacements

(1 +n ) éëê( 2 K + 3) sin (q 2 ) + sin ( 3q 2 )ùûú

Kp æ r ö 2
m=
2 E çè 2p ÷ø
1
K æ r ö
( ) ( )
2
v= p ç - (1 + n ) éê( 2 K - 3 ) cos q + cos 3q ùú
2 E è 2p ÷ø ë 2 2 û
W = ( vz / E ) (s xx + s yy ) ( Plane stress ) = q ( Plane stress )

6.4.4.5  For Mode 3

t xz =
K III
2p r
( 2 ) , andt
- sin q yz =
K III
2p r
( 2)
cos q
(6.100)
s xx = s yy = s zz = t xy = q

The K values depend on the applied loading and crack geometry, and determine
the intensity of the local field.
It is also possible to express the above equations in polar coordinates.
The value of r has certain limitations, i.e. the equations are not valid if r → θ or
becomes large in relation to the length of the crack.
Where two crack modes operate simultaneously, the resolved components of
each in a specific direction are additive.

6.4.5  T
 he Application of K1 in terms of is a Materials
Property [2]

There is a direct relation between K1 and Eγ.

Because E and γ are materials properties then, by inference, K1 should also be.
The actual determination of has shown it to be structure sensitive and depends on
the energy losses incurred in its measurement. This problem will also be encoun-
tered with K1.
The equivalence of K and G parameters.
Extension of the crack involves:
1
(a) Linear opening in the y direction Strain energy = s yy × uy .
2
1
(b) Shear in x direction Strain energy = t xy × uz .
2
1
(c) Shear in z direction Strain energy = t zy × ux .
2
408 D. Daily

y
c δc

O
x

Fig. 4.5  The K1 a material property

Energy is additive, therefore total strain energy =

1
2
(s yy m y + t xy mz + t xy mz ) (6.101)

If crack extends from ∅ to δc, strain energy increase =
dc

(d U E ) = 2 òU E dx
o
dc
(6.102)
= ò (s yy m x + t zy m Lz ) dx
o
æ dm E ö dm E
In the limit ç ÷=
è d c ø dc
Let X be measured from the ∅ (closed position) where f < X < d c and let fail-
ure be in mode 1.
Work done by BOTH crack surfaces
dc
dU = 2 ò 1 (6.103)
2
o
Now σ is the stress normal to the crack surface before it opens or extends (i.e.
crack is in the closed position).
U is the normal displacement of the crack surfaces after the crack has opened.
dc
d UE 1
G= = ( in the limit ) òs udx. (6.104)
dc dc o

6  Ceramic Materials Testing and Fracture Mechanics 409

Derived from the K1 equation (see p. 49)

K1
s yy = cos(q / 2 ) × [1 + sin(q / 2 × sin ( 3q / 2 )] (6.105)
2p r

K1 r
m= (1 + u ) éë( 2k + 1) sin(q / 2 - sin ( 3q / 2 )ùû (6.106)
2E 2p
so that for σyy (σ directed towards the surface) q = f and r = x .
And for μ (μ displaced away from the surface) q = p and r = d c - x .
Substituting,

K1 K1 (d c - x )
s yy = m= (1 + v )( 2k + 2 ) (6.107)
2p x 2E 2
1
d Uc 1 K12
dc
(d c - x ) 2 dx
G= ( mode 1) = (1 + y )( k + 1) ò (6.108)
dc d c 2p E x
o
d c ×p
The value of the integral (Hahn, p. 149) is [5].
2
Giving

K12
G= (1 + Y )( k + 1) (6.109)
4E
At fracture, for plane strain and stress

K12c K2
G=
E E
( )
( plane stress ) G = 1c 1 - Y 2 ( plane stress ) (6.110)

By these simple expressions therefore, it is possible to relate fracture energy and
fracture toughness.

6.4.6  Crack Stability

In most sources of information, fracture mechanics tends to concern itself with

mode 1, and information on other cracking modes is not readily available. There are
two ways of regarding this problem:
(a) By subjecting a crack inclined at an angle to a uniform stress, and
(b) By subjecting a crack to a complex stress pattern.
A relatively clear analysis for (a) is given in Ref. [1], which is summarised below,
while the mathematically more complex (b) is given in Ref. [5].
It is equally possible to express the resolved stresses in terms.
Of KI, KII and KIII and polar coordinates, shown below (Fig. 4.6).
410 D. Daily

σ θθ

y τ rθ σ rr

r σ rr
σθθ

Fig. 4.6  KI, KII and KIII and Polar coordinates

For mode 1

K1
s qq = cos3 (q / 2 ) = K1 fqq (q )
2p r
K1
s rr = cos (q / 2 ) éë1 + sin 2 (q / 2 ) ûù = K1 frr (q ) (6.111)
2p r
K1
t rq = sin (q / 2 ) cos2 (q / 2 ) = K1 frq (q )
2p r
For mode 2

K II
s rr = sin (q / 2 ) éë1 - 3 sin 2 (q / 2 ) ùû = K II frr ( e )
2p r
K II
s qq = - 3 sinn (q / 2 ) cos2 (q / 2 ) = K II fqq ( e ) (6.112)
2p r
K II
t rc = cos (q / 2 ) [1 - 3 sin 2 (/ 2] = K II frq (q )
2p r
Fig. 4.7, is from Lawn and Wilshaw’s “Fracture of Brittle solids and is a graphi-
cal illustration of Figs. 4.32 and 4.33, with θ extended over the range −π to +π [1].
Fig. 4.8, from the same source, demonstrates the effect of adding these curves.
Assuming K I = K II , it is possible to see how an asymmetric function develops for
K when the f(θθ) are added.
An inclined crack in a loaded plate will be subject to two stresses and the stress
intensity factors are transformed accordingly as
Total normal component

K1¢ (q ) = K1 fqq1 + K II fqqII

Fig. 4.7  Fracture of Brittle 2
Solids: Mode 1 Curves
Mode I

f rr
1

fθθ

frθ

-1
-π 0 π
θ

Fig. 4.8  Mode 2 Curves 2

frr Mode II

1
frθ

-1
fθθ

-2
-π 0 π
θ

Fig. 4.8  Mode 1 and 2 2 G(0)/G(0)

4.0
Modes I + II

1 1.0
0.25

KII /KI = 0
0
–π 0 π
q
412 D. Daily

Fig. 4.9  Principle Stress

in Applied Field

I
S

II

SA

Total shear component

K II¢ (q ) = K1 fr1q + K II frIIq

(6.113)

¢
K III (q ) = f
Initially, when the new crack is still ≪ c, the original crack length, it will
move in a direction determined by the ratio K1/KII, but as it moves clear of the
crack it turns to move normal to the applied stress and is subsequently influenced
only by K1.
Deviant paths therefore are not permanent features but the system restores itself
and the path becomes orthogonal to the greatest principal stress in the applied
field.

6.5  Testing Methods for Determination of K1 [6]

6.5.1  Notched Bar (S.E.N.B.)

From the Griffith equation,

2 Eg K1
sf = = (6.114)
pc pc

where c = crack length

K1 = s f p c (6.115)

6  Ceramic Materials Testing and Fracture Mechanics 413

Fig. 5.1  Notched Bar P/2 P/2

Specimen
c
d
c

b
P

p can be incorporated into a factor Y along with geometric constants

K1c 3P1
sf = 1
= (6.116)
2bd 2
Y ×c 2

1
3P1Y c 2
K= where
2bd

Y = Ao + A1 ( c / d ) + A2 ( c / d ) + A3 ( c / d ) + A4 ( 4 / d )
2 3 4
(6.117)

The curves (Fig. 5.2) relate Y with c/d. It clearly shows that a c/d value between
0.1 and 0.2 is most favourable and contrasts with the work of fracture experiment.
Values of A coefficients

Loading A0 A1 A2 A3 A4
Type
1/w = 8 +1.96 −2.75 +13.66 −23.98 +25.22
1/w = 4 +1.93 −3.07 +14.53 −25.11 +25.80
Four point +1.99 −2.47 +12.97 −23.17 +24.80

This method of K1 determination depends on a knowledge of the crack depth,

which has been made by a saw-cut.
It presupposes that:
1. The rate of loading will be sufficiently fast, so that growth is not significant as
the machine is loaded.
2. The end of the saw-cut will generate a network of micro cracks, one of which
will grow and eventually fracture the specimen.
1
6 Mc 2
(Another way of looking at this formula is K = where M is the applied
bd 2
bending moment. This is useful for four-point bends)
The curves show the variation of Y with the ratio c/d.
414 D. Daily

Fig. 5.2  The Variation of 3.4

Y with the Ratio c/d
3.2

3.0

2.8

2.6
y
2.4

2.2

2.0

1.8 3 POINT, S/W-8

3 POINT, S/W-4
1.6
0 .1 .2 .3 .4 .5 .6
c/d

6.5.2  The Double Torsion Method [3, 16, 21, 24, 30]

6.5.2.1  Double Torsion Test Piece

Where a greater degree of control over the crack progress is required, then the dou-
ble torsion or double cantilever methods are preferred. They also lend themselves to
the study of stress corrosion effects.
If y is the displacement of the loading points, P is the applied load and a is the
crack length, from the compliance relationship,

y = P ( Ba + c ) (6.118)

Differentiating w.r.t. time,

dy dp PBda
= ( Ba + c ) + (6.119)
dt dt dt
da
Note that is the crack velocity, which at constant displacement
dt
dp
da (
Ba + c )
dy = f dt
is given by, = (6.120)
dt dt BP
The values for P and a are interrelated, when the specimen is subject to constant
displacement. When the crack has finished growing, we have Pf and af, hence
6  Ceramic Materials Testing and Fracture Mechanics 415

DOUBLE TORSION TEST-PIECE

Shape of crack profile

K1 = Pb′ 3 (1 + √−)
b P/2
P/2
bd3dn dn

P/2

d P/2

Direction of crack propagation

Diminished width to prevent crack running laterally

Fig. 5.3  Double Torsion Test Piece

P ( Ba + c ) = Pf ( Baf + c ) (6.121)

Substituting for (Ba + c) in eq. (6.120), we obtain,

da P dp
= 2f ( Baf + c ) (6.122)
dt p B dt

For a relatively large crack length, where c <<< B, the equation simplifies to

da p dp
= - 2f af (6.123)
dt p dt

The crack velocity can be found if the initial and final loads, the rate of load relax-
ation with time, and the final crack length can be measured. The specimen is loaded
to a suitable level (below K1) and for a fixed displacement, the load decay with time
is followed on the pen recorder coupled to the load-cell of the testing machine. The
final crack length is measured, probably with the aid of a dye penetrant in the crack.
Also we note that
1
é 3 (1 + u ) ù 2
K1 ( orK1c ) = P1 ( or P1c ) wm ê 3 ú (6.124)
ë wt t n û

1
é 3 (1 + u ) ù 2
K I = PWm ê 3 ú
ë wt t n û
416 D. Daily

Fig. 5.4  Crack Velocity

Test

W
P

tn
P/2 wm

Fig. 5.5  Crack Velocity

KIC
Curve

III
crack velocity da

KIt
dt

KId
LOG V

Vt
II

I
KO

VO

da
Figures 5.5 and 5.6 show the characteristic shapes of the curve when is plot-
dt
ted against K1. In the results for polycrystalline alumina, K1c is equal to 5.2 MN m−3/2.
The slope in region 1 is often referred to by the symbol “n” and called the “crack
susceptibility.” One can contrast the behaviour of alumina in moist air with that in
toluene (where air is excluded). In air the onset of stress corrosion is at a lower
stress level than in toluene and in vacuum it is depressed still further.
Region 2 is depressed in high corrosion situations [24] and the effect of Ca and
Mg ions in the alumina depresses the value of n. Increasing the purity of the alumina
increases n to the extent that measurement becomes very difficult because the initial
reading approaches K1c, although the early detection of the movement of the crack
by acoustic emission should simplify this measurement [25].
An important consideration in the double torsion technique is that the crack front
has a curved profile [16]. Allowance can be made for this in the calculation, but a
method which almost eliminates this to use very thin specimens.
6  Ceramic Materials Testing and Fracture Mechanics 417

Fig. 5.6  A Comparison of

the Crack Velocity of
Alumina Ceramics in Dry
and Humid Conditions

Alumina of various categories has been used for the manufacture of microcir-
cuits and these can be obtained in 50 mm squares which are very suitable for the
double torsion test. Some difficulty has been encountered in making an initial crack
in the edge to begin the process. The following technique has been successful. A
small cut (<1 mm) is made in the edge, and this is dipped symmetrically into a bath
of Silicone oil at 180–200 °C. When the edge with the crack has attained the tem-
perature, it is then plunged into cold water. With practice the thermal shock can be
used to make a short crack in the side of the specimen which can then be placed in
the equipment and tested.
Results of Double Torsion Experiments (Davidge [3])

6.5.3  Double Cantilever Beam

This technique was the original method for the investigation of fracture energy and
K1. The formula for K1 is

3.45 Pa é æ t öù
K1 = 3 ê1 + 0.7 ç a ÷ ú (6.125)
bt 2 ë è øû
418 D. Daily

The technique suffers in that it is dependent on a, the crack length. Glass and epoxy
resins etc. create no problems, because the crack can be seen due to total reflection
on the air surface within the crack. K1 for sapphire was also measured in this way.
But for opaque ceramics it is not very successful because the exact end of the crack
cannot be seen, unless the specimen is dismounted and a dye penetrant used to
reveal the end of the crack. This will alter the surface conditions and destroy the
value of the experiment.
Another difficulty is caused by the asymmetric loading of the specimen on a
conventional hard testing machine. The weight of the test piece will influence the
result. It can be used successfully in certain purpose built machines which put the
test piece in tension horizontally [15]. This technique also allows stress-corrosion
measurements to be made.
It has been said [28] that this technique measures the pure K1 fracture mode.
While this is probably true, the results do not differ significantly from those obtained
by the double torsion technique.
Because the DCB method depends on a knowledge of the crack length, an alter-
native method has been proposed which eliminates this, carefully removed and the
specimen dried before use. The moment arms were attached with high strength
epoxy resin which was cured at 70 °C. Small lugs were also incorporated in the
moment arms to meet with a transducer for measuring the separation of the crack
surfaces. Alumina proved to be a very difficult material to measure in this way.
When one attempted to work at constant strain, the crack did not extend even though
one approached within 90 % of K1c. Any further increase simply shattered the speci-
men which was clearly too stiff.

2t
a

Configuration for double cantilever beam

P

Fig. 5.7  Configuration for Double Cantilever Beam

6  Ceramic Materials Testing and Fracture Mechanics 419

Fig. 5.8  Crack Opening Measurement

Fig. 5.9 Specimen
Fracturing across Width
due to Surface
Imperfections

The crack velocity can be obtained from the measurement of the crack opening,
or from the depression of the loading points by a simple relation.
If the beam deflects through an angle θ, then tan θ = d/L = a/r = aM/EI. Hence
a = (EI/ML) ∙ d.
Like the DT, or DCB this method is easy to operate if there is significant stress
corrosion, but with increased purity of materials the stress intensity factor K1
approaches K1c as the regions I and II of the curve of Fig. 7.4 diminish and if one
420 D. Daily

attempts to allow the stress to relax from an initial constant strain it is more than
likely the specimen will fracture prematurely.
An alternative to this is to perform the experiment under constant stress. This can
be generated by means of a long lever arm on the end of which is a container of
zircon sand (chosen for its high density). This system, which admittedly appears
crude, is capable of very fine adjustment, and the stress actually on the moment
arms is read off from the testing machine chart as before. It proved quite possible to
control the growth of the crack by this method, but the values of K1 in those speci-
mens which fractured were twice as high as expected. This was because in the
constant loading method an equal quantity of energy is used to strain the specimen
as that used in extending the crack (p. 51). Before it is possible to accept these
results as reliable, a check for the influence of crack kinetic energy should be made
[1]. Attempts to use the constant moment method in this manner for pure alumina
gave results which were no better than the double torsion. A large number of plots
were made with K1 ranging from 2.2 to 2.5 MN m−3/2 over a velocity range from 10−8
to 10−5. The constant velocity region (II) was assumed to be about 10 m/s and this
had a particularly wide scatter of results. K1c was in the range 4.5–5 MN m–3/2.
A possible advantage of this configuration is that by adjustment of the position
of the test-specimen one can perform fracture experiments in the K3 mode.
Precautions have to be taken to prevent
(a) the specimen rotating back into the normal K1 mode, and
( b) the specimen fracturing across its width from a surface imperfection near the
point of attachment of the moment arms. Cladding with a thin piece of metal
attached with epoxy resin accomplishes this, and it does not significantly alter
the moment of inertia of bending of the specimen.

6.5.4  The “C” Specimen

An elegant method for the measurement of fracture toughness in the tensile mode
has been reported by Kapp et al. [26]. It is intended for use with hollow cylinder
geometries, and one can appreciate that it can be used in reverse, i.e. from a knowl-
edge of K it is capable of detecting critical sizes of flaw in a hollow cylinder geom-
etry. Much of the mathematical theory is available in Refs. [17, 18, 26], but the
paper of Kapp et al., presents the formula for K in a readily applicable form.

æ ö
ç P ÷é
K =ç 3 X / W + 1.9 + 1.1 a ù
1 ÷ë wû
ç BW ÷
è 2ø
é
( ) ( )
æ 1 - r1 ö ù F a
2
X ê1 + 0.25 1 - a ç
ë w è r2 ÷ø úû w (6.126)

( w ) = éêë( a w ) / (1 - a w ) ùé
( ) ( )
1 3
where F a
2 2
a a
2
a
3
ù
ú ê3.74 - 6.30 w + 6.32 w 2.45 w ú
ûë û
6  Ceramic Materials Testing and Fracture Mechanics 421

Limits:
For 0.45 ≤ a/w ≤ 0.55
For all r1/r2 values, Accuracy within ±1 %
For X/W either 0.5 or 0
For 0.2 ≤ a/w ≤ 1.0
For all r1/r2 values, Accuracy within ±1.5 %
For X/W either 0.5 or 0
For 0.2 ≤ a/w ≤ 1.0
For all r1/r2 values, Accuracy within ±3 %
For 0 ≤ X/W ≤ 1
An application of this method is found in Tan and Davidge’s paper [18]. This
paper also describes an alternative mode of gripping the specimen, and making it
more applicable for evaluating the quality of tube materials.

6.5.5  Short Rod, Short Bar or “Stub” Test Piece

The basis of this test is a short cylindrical or block specimen which has a notch
machined into it of such a shape that it leaves a “V” configuration behind in the
remaining ceramic. A basic requirement for this must be a very precise fabrication

Fig. 5.10 Short
Cylindrical or Block
Specimen

2θ

d
422 D. Daily

Fig. 5.11  The “C” shaped

Specimen indicating
geometric parameters

method for the initial cylinder or block, and a means of sawing in the notch to very
precise specifications after firing. Testing is carried out by opening up the notch and
this may be achieved in two ways.
1 . By pulling the specimen apart using a specially shaped tensile jig.
2. Inserting a thin inflatable bag into the notch and pressuring it until the test-piece
breaks.
The latter method is clearly quite attractive if expensive test machines are not
available.
The crack growth from the point of the V-shaped slot in the specimen is initially
stable, even for the most brittle materials. Thus a real crack is obtained during the
initial loading of the specimen and an ever increasing load is required for continued
crack growth until the crack reaches a critical length ac. Thereafter, the load to pro-
duce further crack growth decreases. ac depends upon the specimen geometry, but is
independent of the specimen material, provided only that it behaves in accordance
with linear elastic fracture mechanics. The equation for K1c is as follows:

K1c = AF Pc B (6.127)

Pc is the peak pressure on the notch face during the test.
B is the specimen diameter.
AF is a dimensionless constant dependent on the specimen geometry and loading
configuration.
6  Ceramic Materials Testing and Fracture Mechanics 423

For the rod specimens these are the ratios (related to B)

W / B = 1.5000 + 0.006
C / B = 0.03 + 0.002
a0 / B = 0.531 + 0.006
a p / B = 0.482 + 0.002

8 = 55.2 + 0.5°
For block specimens L/B = 0.870.
Figures 5.13 and 5.14 are the alternative configurations. Figure 5.15 is a section
through the notch.
For the ration given above, A = 7.51.
Typical dimensions are going to be determined by the width of the notch. For a
1 mm notch, one requires a 33.3 mm diameter.
Main reference is L.M. Barker in ASTM/STP 678 (p. 73 [6]). (The author dis-
cusses the use of a curved profile on the “V”. This might be achievable with com-
puter machining.)

6.5.6  Indentation Methods

One of the difficulties of obtaining fracture toughness data is that for the most part
specially prepared specimens with characteristic shapes have to be produced. It
would be very desirable to have a technique which would eliminate these difficulties
even at the expense of accuracy if it could be used to sort bodies into categories, so
that when a promising material was forthcoming, fewer refined test specimens
would have to be prepared. It is possible to do this with a hardness indenter which
is a standard piece of equipment in metallurgical and engineering laboratories.
Figure 5.17 represents the impression left by a Vicker’s diamond pyramid indenter
and showing cracks emanating from each corner.
Figure 5.16 is a longitudinal section through the diagonal of the pyramid base.
The advancing crack front is the semicircle of diameter 2D′ (surface extremities of
the crack) and of sub-surface depth D.
2a is the length of the diagonal indentation, and to be
Effective, D/a ≥ 2.
An expression K1c, given by Lawn and Swain [3] is

(1 - 2g )(aro )
1/ 2 1/ 2
æPö
K1c = çD÷ (6.128)
21/ 2 p 2 b è ø

where
P = indenter load.
D = crack depth.
424 D. Daily

Fig. 5.13  Short Bar

Specimen

Fig. 5.14  Short Bar

Specimen

Fig. 5.15  Plan view of

Short Bar specimen

γ = Poisson’s ratio.
ρo = mean contact pressure.
α = dimensionless constant determined by indenter geometry (for Vicker’s a = 2 / p ).
β = dimensionless constant determined by the deformation zone geometry (for
Vicker’s or Knoop indenter β ~ 2).
6  Ceramic Materials Testing and Fracture Mechanics 425

For a pointed indenter which leaves geometrically similar impressions in a

homogeneous specimens at all loads, the mean indentation pressure remains
invariant. And is equal to the hardness of the material. Equation (6.128) predicts
that a plot of P against D will be a Straight line for a given material. By measur-
ing the hardness (ρo) and the slope of P/D, the fracture toughness can be
calculated.
There is a second, simpler treatment, predicting a different relationship between
the parameters [27]

1 æ P ö
K1c = (6.129)

p 3 / 2 tany çè D1/ 2 ÷
ø

where ψ is the indenter cone half angle (Strictly it is only valid for conical indenters)
but it gives a good fit to crack growth obtained with a Vicker’s indenter. In this case
ψ is assumed to be 68°, the half angle between the faces. Analysis predicts that
3
P / D 2 will be a straight line and the hardness is not needed.
Other points about the methods:
1 . Crack length should be ten times the max-grain size.
2. Crack depth should be less than 1/3 specimen thickness.
3. For ceramics loading rate was 5.0 mm/min with as Vickers’s indenter.
4. Measurement of crack length was by travelling microscope with dye-penetrant
to render crack.
5. Data collected at four indenter loads with a min. Of three indents at each load.
For materials with high stress corrosion rates, the indenter may be covered with
paraffin oil otherwise tests are carried out in air. For transparent materials photogra-
phy is probably a better method to measure crack lengths.

Fig. 5.16  Illustration of P

parameters defining the
medium crack which forms
during indentation
2y

2a

2D'
426 D. Daily

Fig. 5.17 Indentation
cracks introduced by a
Vickers indenter. (a)
Schematic showing the
parameters used in the test.
(b) A typical crack pattern
obtained in Si3N4
(polarised light reflected
micrograph)

Fig. 5.18  Indentation cracks introduced by a Vickers indenter (a) schematic showing the param-
eters used in the test

6.6  Time Dependence of Strength [3]

6.6.1  Stress Corrosion

When a crack in a ceramic is exposed to the air, or other environment, the freshly
exposed, very reactive surfaces will undergo a chemical reaction with whatever is
nearest to it. If you consider that the bonds which hold the crystal together have
been disrupted, and there are the free unbounded electrons of the valency forces
available to react, then this effect is inevitable. If it were possible to compress a
6  Ceramic Materials Testing and Fracture Mechanics 427

crack before the atmosphere was able to attack it, the crack would heal up again, and
this can be demonstrated in glass. The net effect of the atmosphere is to stabilise the
crack in the open position. The most deadly corrosion agent is simply water vapour,
and the amount present in normal moist air is quite enough to cause damage.
Obviously more active agents, e.g. saline solution, present in animal tissues will be
even more dangerous. In effect the (OH)− groups from the water (or Cl− ions) satisfy
the released valency forces. The combined effect of stress in the material and the
activation energy of the chemical attack causes a crack to steadily grow until it
eventually reaches the critical Griffith flaw size, and if the stress is sustained, the
material will break.
Referring to the double torsion graph (Fig. 5.5), the linear region (I) is that for
sub-critical crack growth (or stress corrosion) and for a particular stress (on the
abscissa) the appropriate crack velocity can be read off on the ordinate. As the crack
grows the stress intensity factor increases, the velocity increases, and a catastrophic
situation ensues (i.e. when the stress intensity factor becomes critical).
Under constant stress σ time to failure is given by:
Cc
dc
t= ò
Ci
v
(6.130)

1
Ci is the initial crack size and Cc is the critical size, and K1c = g s Cc 2 .
1
Because generally K1 = g s C 2 , differentiation given

2 K1
dC = dK1 (6.131)
s 2g 2
Substituting for dc in Eq. (6.130).
K1 c
2 K3
t=
s g
2 2 ò
K1i
v
dK1 (6.132)

Referring to the double torsion graph (Fig. 5.5), we have,
Region I, log v = const.x log K1 i.e v = α1K1.
Region II, Constant velocity region v = α2.
Region III, A third term is possible, but is usually not considered, because crack is
reaching sonic velocities, and the specimen is in effect broken at the end of
region II.
Let K1* be the limiting value of K1 separating regions I and II

2 æ 1 K1 (1- n ) 1
K1 c
ö
then t( minimum ) = ç ò K1 dK1 + ò K dK ÷ (6.133)
s g
2 2 ç a1 k a2
1
÷
è 1 K1 ø
428 D. Daily

For a real situation, the limit of tolerable stress corrosion should be the end of
region I, so if we simply integrate this part of the equation we get,

2
t( min ) =
s 2g 2a1 ( n - 2 )
( K12i- n - K12*- n ) (6.134)

2-n 2-n
Furthermore if n is large (typically) 10, K 1i >> K 1* so that

2 K12i- n
t( min ) = (6.135)
s 2g 2a1 ( n - 2 )

1

Let a particular specimen have an initial crack length c1, then K1i = sg ci2 ; com-
bining with Eq. (6.135), (6.136) we see that tσn is constant (for this particular speci-
men). Thus the ration of the lifetimes tσi, tσj, at two subcritical stresses σi and σj, is
given by
n
æ si ö ts j
çç ÷÷ = (6.136)
ès j ø ts i

This relation allows the strength/probability/time diagram to be made, combining
the statistical and time-dependent properties of ceramics.

6.6.2  The SPT Diagram

The failure times under constant stress conditions are likely to be inconveniently
short compared with those measured a constant rate of strain. These times are
related by the expression

te = ( n - 1) ts (6.137)

(where n is the slope of the K1/v curve in region I).
Suppose a specimen fails under stress σεi, in testing time tεi then s e i = Ee i te i .
For another specimen failing under stress σεj, in time tεj, then s e j = Ee j te j .
for which

æ s ei ö e i te i
çç ÷÷ = (6.138)
è se j ø e j te j

Eliminating the time using Eqs. (6.137) and (6.138), we have,

n +1
æ s ei ö ei
çç ÷÷ = (6.139)
è se j ø ej

6  Ceramic Materials Testing and Fracture Mechanics 429

Fig. 6.1  A typical crack

Pattern obtained in Si3N4
(polarized light reflected
micrograph)

0.99

0.98

0.95

0.90
Survivel probability

0.80

0.50

0.10
0.05 1055 1045 1035 1025 105 15 0.15
0.01
0.001

150 200 250 300 350 400 450 500

Stress MNm-2

Fig. 6.2  S.P.T diagram for 95 % alumina

S.P.T. diagram for 95 % alumina [1, 3].

S.P.T. diagram for Co-WC alloy at different temperatures (Braiden et al. [23]).
This enables n to be estimated independently of the K1/v diagram.
430 D. Daily

6.6.2.1  Procedure

(1) About 100 test specimens (bars or discs) are prepared and the side which
becomes under tension is polished.
(2) Tests on five batches of 20 are conducted over a wide range of strain rates (each
batch having a constant rate). The Weibull modulus for each is obtained.
(3) A representative stress is chosen for each batch (usually the median) and n is
calculated between pairs of batches from the median stress and the straining
rate.
(4) Select a standard failure time (usually 1 s) and normalise the most convenient
Weibull spread to this time (using Eq. 6.137)?.
(5) On the basis of this, draw parallel Weibull slopes in decades of seconds until the
desired life of the component is reached. For convenience remember that
3–107 s is equivalent to 1 year.
For n = 10 each decade represents a reduction of strength of 21 %.

6.6.2.2  Application

This procedure is necessary when ceramic components are designed for use in cor-
rosive environments. Typical applications are hip-joints and grinding wheels. For
the hip prosthesis the corrosive environment is the body fluid which is approxi-
mately normal saline (e.g. tears). A projected life is 50 years, which makes a revi-
sion of the operation unlikely!, and the stress chosen as a basis for calculation is 10
times body weight (i.e. if one jumped down four stairs and landed on one leg.) The
calculation is pessimistic because this is an exceptional stress condition and not a
constantly applied one.
Apart from the knee these strictures do not apply to any other ceramic joint
replacement because they are not under stress.
In the case of grinding wheels, the normal moist air we live in is corrosive agent,
and the centrifugal force of its rotation creates the stress. Failure of the wheel cata-
strophic and fatal accident. Manufacturers of wheel should therefore specify the
maximum number of hours for safe usage.
One should be able to apply this reasoning to any kind of ceramic component
which will be under stress during its useful life. How do you regard ceramic engines
in cars, in lorries, or in aircraft!? Perhaps they may be regarded as disposable items
since they should become cheaper than conventional materials and thereby create
and environmental disaster with all the indestructible ceramic litter.

6.6.3  Proof Testing [3, 22]

The inherently brittle nature of ceramics makes their use in high stress situations
attended with some risk, and before use car be contemplated one has to make a very
careful assessment of this risk.
6  Ceramic Materials Testing and Fracture Mechanics 431

Typically the questions which might be asked are:

“Given the characteristics of its design, is it possible for a hit joint ball to last 40
years, subject to a maximum loading of 10× body weight (chosen to account for
impacts during stumbles or other unusual situations. Will it survive the corrosive
body environments)”.
“Will a ceramic–bladed jet–engine survive 1 flight, 5 flights, 20 flights etc. in ser-
vice across the Atlantic” What is quite certain is that only whole numbers are
acceptable, and that a fair number of flights are required before the engine can be
taken out of service to be operated economically.
“How many hours will a grinding wheel survive at its normal speed of rotation,
working in normal humid air before it must, for safety’s sake be taken out of
service”.
Putting these questions in another way would be to say “What is the chance of a
catastrophically large flaw being present in a highly stressed region of the ceramic”.
A rough method of improving their survival probability would be to simply place
all the specimens under a testing load and deliberately allow a few of them break to
remove those which had manufacturing defects. This is better that nothing, but it is
possible to be much more precise. Initially one must decide what is a realistic load
for the ceramic component. It is them overloaded for a finite length of time during
which the cracks will grow. Some, of course, will exceed the critical length and the
ceramic will fail, but it is possible to calculate a critical crack length under a proof-­
test load, at which the ceramic will be just short o the point of failure. This repre-
sents a “worst case” situation because most of the survivors will be far better that
this.
If we now go back to the realistic load, the critical cracks will now be longer than
they were before testing, and the service life will be the time required for the cracks
to grow by stress corrosion from the proof-test worst crack length to the working
load critical crack length. If it assumed that the specimens are being proof tested at
the top end the K1 region we can state that,

K1 p = Ys p cp < K1c (6.140)

In subsequent service

K1a = Ys a c p (6.141)

Thus

sa
K1a = K1c (6.142)
sp

2 K12c- n (s a / s p )
2-n

K ( min ) = (6.143)
s a2Ya21 ( n - 2 )

432 D. Daily

Fig. 6.3  S.P.T diagram for Co-WC alloy at different temperatures (Braiden et al. [23])

Choosing different ratios of σa/σp the stress, σa may be plotted against the time for
which it is applied (Eq. 6.143), on a log-log plot this must have a slope of −2. We
can superimpose on this diagram curves which relate the time in seconds to the
stress for a specific probability of failure. These can be obtained from the S.P.D.
diagram. It is possible to predict therefore, for a specific ratio of failures to survi-
vors, what the service time would be under a given stress, and also the proof testing
ratio σp/σa which would guarantee this performance.
Figure 6.3 shows that if we allow a one per thousand (10– 3) proof-testing failure
ratio, a service life of 5 ´ 10 7 s is guaranteed for a service stress of 30 MPa, if the
proof testing ratio had been 2.6/1.
(Check that if we allow one per hundred failures during proof-testing at a ratio of
3.6/1, the service life under the same stress is 1016 s. Since this is about 108 years
they would have to have been made before mankind appeared.)
For hip joint balls designed to operate with a conical fixation, σa is 9000 N
applied over the effective conical surface, and σp is 25,000 N. This should last about
50 years. Because there is some slight degree of failure when the load is removed,
the specimens are reloaded to the working stress level to ascertain they are still
sound.
6  Ceramic Materials Testing and Fracture Mechanics 433

6.7  Ceramic Hip Joints Endoprosthesis

The design of the ceramic head for the hip joint prosthesis requires the solution of a
number of problems, important among which are as follows:
1. Using a conical fixation to the metal stem creates a hoop stress which will be
severe if the cone is forced or impacted too violently into ball.
2. It is unlikely that the conical surfaces will be machined so precisely that stress
will not be concentrated at some point on the cone.1
3. A working stress limit on the ball is considered to be about ten times weight,
taking account of sudden impacts on one limb.
4. The effect of stress corrosion after 30–40 years may weaken the ceramic.

6.7.1  Finite Element Analysis (FEA)

The testing of every combination of a hip device design with multiple modular
heads and neck arrangements and sizes are not practical in terms of time or cost for
the pre-market approval process of a device. To overcome this issue often “the worst
case” prosthesis design/combination is identified and used as the bases for the
approval of the range of similar devices. The identification of the “worse case”
design may not always be obvious; however, it is generally accepted that femoral
heads with the largest off set of +12 mm [29] on the smallest stem in terms of diam-
eter produce the most extreme stress conditions in the neck of the femoral stem. In
particular with new device designs were there is no historical data to draw from then
Finite Element Analysis (FEA) can be a valuable tool to determine the worst case
design combination to test. When this chapter was written in the 1990s FEA was a
relatively recent computational technique; however, it is now widely used in medi-
cal devices to verify a design is “fit for purpose”. FEA is used to simulate the load-
ing conditions within a human body and therefore assisted in the development of
more realistic testing protocols.
In the original investigation, the aim was to answer questions raised (Sect. 6.7)
questions 1, 2 and 3. In the original computational process the optimum mesh of
152 elements was drawn and eight nodes per element were defined at which the
stresses were measured. In modern computational studies it is common to use 4000
[30] mesh units to represent the possible contact areas using a new range of analyti-
cal models; however, the original work describes the basic theory behind the analy-
sis therefore it was decided to retain much of the original texts.
Three load cases and three restraints were defined, as follows (Fig. 7.2).

 Accurately machined cone and socket sections from epoxy resin were analysed photoelastically.
1

However, careful the machining, a “high spot” which concentrated the stress was always revealed.
Much depends on the confidence one has in the machining operation.
434 D. Daily

6.7.1.1  Load Cases

Case (A) was chosen to represent the very small area of contact of the prosthesis
ball with a ceramic acetabular cup.
Case (B) was chosen to represent the wider distribution of the load when a less
rigid UHMWPE (ultra high molecular weight polyethylene) acetabular cup was
used.
Case (C) was chosen to represent the situation in the body, where the load was
not applied axially but spread over an arc between 20° and 55° to the cone axis. This
case is somewhat exaggerated and chosen to accommodate the computational sys-
tem. It serves to indicate that in a realistic situation where the load is applied at 25°
to the axis there will be more distortion than is represented by cases A and B.

6.7.1.2  Restraints

1. The load was transmitted to the shaft via the ball across the lower third of the
conical interface, i.e. at the end of the cone remote from the centre of the ball.
2. Load transmitted via the middle third, and
3. Load transmitted via the upper third, i.e. at the end of the cone nearest the
centre.

6.7.1.3  Results and Conclusions from FEA

The finite element system requires information about the physical properties of the
material e.g. modulus of elasticity, Poisson’s ratio etc. and the load applied, which
was 9000 N (i.e. about ten times body weight. The body in this instance is rather
large, but not untypical of the type of elderly patient needing a hip replacement) [8].
Because the ceramic heads were symmetrical, the problem was considerably
simplified, and the analysis was able to provide plane strain diagrams of the
following:
1. The highly magnified “displaced shape” of the ball under load (Figs. 7.3, 4, 5).
The diagrams were chosen to represent (“worst cases”). The diagrams appear to
display very large distortions, but the scale of these is very large, i.e. 1 cm on the
diagrams represents the order of micrometres.
2. Diagrams representing the magnitude and direction of the two principle stresses
at the nodal points in the xy plane.
3. Diagrams showing equal stress contour lines through the nodal points for the
maximum and minimum principle stresses and for the hoop stress which is in the
z direction.
6  Ceramic Materials Testing and Fracture Mechanics 435

A stress printout for all the nodes which surround each element was also pro-
vided, from which it was possible to derive the average stress at the centre of each
element.
The finite element analysis was able to point to several factors about the design
of the prosthesis shape, the manner of application of the load and to enable specifi-
cations for the alumina to be drawn.

6.7.1.4  Shape Factors

Examination of the diagrams for the principle stress (2) above and the equal stress
contour lines (3) show large tensile stresses at the base of the cone and at the cone
opening. This can be remedied by undercutting the base of the cone to turn it into a
dome, and by relieving the corner at the cone opening.

6.7.1.5  Application of Load

The displaced shape diagrams (1) and the stress contours (3) clearly show that the
application of the load over a larger area from the polyethylene cup produces a
higher concentration of large tensile stresses than is the case with the alumina ace-
tabular cup.

6.7.2  Weibull Equation

This involved the application of the four function Weibull equation [9, 32] as
follows:
The probability of failure of a test specimen can be represented by the standard
Weibull equation.

ìï é (s - s ) ù m üï
Pf = 1 - exp í- ê u ú ý (6.144)
ïî ë s 0 û ïþ

Another form of this equation is given by

ïì æ 1 ö æ s ö ïü
m m

Pf = 1 - exp í- ç ÷! ç ÷ ý (6.145)
ïî è m ø è s f ø ïþ
436 D. Daily

which employs the relation s f = s 0 (1 / m )! , where (1/m)! are the “gamma func-
tion” and is available in tables, and s f is the mean stress of the batch.
The failure stress of a specimen is also related to its volume, given by the expres-
sion (based on 6.144)

s f mV = s fv m v (6.146)

s fv is the tensile breaking stress of a test piece of reference volume v. The purpose
of the exercise is to evaluate this.
Combining the two we arrive at the four function Weibull equation

ì ö V m üï
m
ï æ1ö æ 1
Pf = 1 - exp í- ç ÷!m × ç ÷ × ×s ý (6.147)
ç ÷ v
ïî è m ø è s fv ø ïþ

The stress in each volume element dV is made up from the three principal
stresses so σmdV has to be applied three times over in each element, i.e.
( )
s 1m + s 2m + s 3m dV . Account has also to be taken of whether the stress is tensile or
compressive. Use is made of the Heaviside function H(σ). Its value is unity for posi-
tive tensile stresses, and −∝ for negative compressive stresses, ∝ is the modulus of
the ratio of the mean failure stresses per unit volume in uniaxial tension and com-
pression. In this exercise ∝ was conveniently given the value of −10 for the alumina
under compression.
The final expression for the failure probability of the specimen is given by

ì m ü éæ s öm æ s öm æ s ök ù
ï æ1ö æ 1 ö 1ï
Ptotal = 1 - exp í -ç ÷ !m ×ç
m
îï è ø è fr
s
÷ × ý ò êêçè H (s11 ) ÷ø + çè H (s22 ) ÷ø + çè H (s33 ) ÷ø úú dv ( Ref. 10 )
ø v ïþ r
ë û
(6.148)
where the “stress-volume integral” is taken over all the elements in the body.
From the finite element diagrams, the area of each element was calculated in
square millimetres by approximating each to a trapezoid. The volume of revolution
of each element about the centre line of the section was then calculated.
To construct Tables 6.4 and 6.5 the 152 elements were divided into eight separate
computer programmes to estimate the stress-volume-integral, the sum of which is
shown in the first column. For convenience of calculation the stress unit chosen was
10 MPa.
6  Ceramic Materials Testing and Fracture Mechanics 437

3.5

1014 P1 = 10-5 10-4 10-3 10-2 10-1

3.0
1012

1010 2.5
100 yr
Time in service s

108 10 yr
1yr
2.0
1min
106

1d
104
1h
R = 1.5
102
1min

1 2 3 4 5 6 8 10 20 30 40 50 60 80 100

Service stress MN m-2

Fig. 7.1  Finite element mesh—elements numbered (1–152)

Fig. 7.2  Loadcases (A, B and C) and restraints (1, 2 and 3)

438 D. Daily

Fig. 7.3  Displaced shapes of prosthesis head. Load case A. Constraints 1 and 3

Fig. 7.4  Displaced shape of prosthesis head. Load case B. Constraint 1

6  Ceramic Materials Testing and Fracture Mechanics 439

Fig. 7.4 (continued)

Fig. 7.5  Asymmetric load. Constraint 3

Table 6.4  Experimental Data for Notched Bar Test
440

Str/vol. Int Ref., vol. Weibull mod. Gamma for “m” value Failure prob. reqd.
Sv(-----)dv (v) (m) (1/m)! (Pr)
1.050E14
8.890E10 mm3 10.000 0.610 0.100
6.890E13 1000.000 ln(1 − Pr)
9.280E15 −0.105
1.030E12 −Gamma
1.140E10 −0.610
5.330E14
5.800E14
Total Total/Ref. V S/V Int × Gamma “A”/E−14 Req. str/10
1.057E16 1.057E13 −6.446E12 1.634 23.91 MPa
Str/vol. Int Ref. vol. Weibull mod. Gamma for “m” value Failure prob. reqd. (Pr)
Sv()dv (v) (m) (1/m)!
1.05E14
9.890E10 mm3 10.000 0.610 0.100
6.890E13 1000.000 In (1 − Pr)
9.280E15 −0.105
1.030E12 −Gamma
1.140E10 −0.610
5.330E14
5.800E14
Total Total/Ref. V S/V Int × −Gamma “A”/E−14 Req. str/10
1.057E16 1.057E13 −6.446E12 0.016 38.099 MP a
“A” = (S/V Int × Gamma)/log(1 − Pr)
With a Weibull modulus of 10 the required tensile strength of a 1 cc test specimen is 381 MPa for a 0.1 % failure probability
The required tensile strength is 239 MPa for a 10 % failure probability
D. Daily
Table 6.5  Survival Probability Vs. Stress
Str/vol. Int Ref., vol. Weibull mod. Gamma for “m” value
Sv(-----)dv (v) (m) (1/m)! Failure Prob. reqd. (P)
3.500E20
3.210E15 mm3 15.000 0.590 0.100
3.520E19 1000.000 ln(1 − Pr)
9.940E18 −0.105
1.460E17 −Gamma
1.140E10 −0.590
2.400E21
8.820E20
Total Total/Ref. V S/V Int × −Gamma “A”/E−14 Req. str/10
3.677E21 3.677E18 −2.170E18 4.856E−6 19.39 MPa
Str/vol. Int Ref. Vol. Weibull mod. Failure prob. reqd.
3.500E20
3.210E15 mm3 15.000 0.590 0.001
6  Ceramic Materials Testing and Fracture Mechanics

3.520E19 1000.000 In(1 − Pr)

9.940E18 −0.001
1.460E17 −Gamma
1.140E10 −0.590
2.400E21
8.820E20
Total Total/Ref. V S/V Int × −Gamma “A”/E−14 Req. str./10 26.449 MPa
1.057E16 3.677E18 −2.170E18 4.61E−8
“A” = (S/V. Int × –Gamma)/log(1 − Pr)
With a Weibull modulus of 15, the required tensile strength of a 1 cm test specimen is 264.5 MPa for a 0.1 % failure probability
The required tensile strength is 194 MPa for a 10 % failure probability
441
442 D. Daily

6.7.3  A Design solution to the Stem Fixation Problem

An obvious solution to the problems created by the conical fixation is to avoid this
if possible. A cylindrical fixation would have the following advantages:
Dangerous hoop stresses would not be created by impacting the ball in service or
during fixation.
It is easier and less expensive for the manufacturer to press solid spheres and
afterwards machine out the cylindrical fixation hole, rather than having to machine
out a precise conical hole or include it in the original pressing.
To obtain the degree of bonding required needs a very precise frictional fit
between the stem and the ball. This has been achieved by
(a) Covering the stem with a plastic sleeve and forcing it into the cavity.
(b) Diffusion bonding a biocompatible metallised layer of precise thickness into
the surface of the cylindrical cavity in the ceramic so that it becomes integral
with the ceramic. A titanium shaft is subsequently fitted into the cavity. This is
still under active development.

6.7.4  Summary and Conclusions

Testing methods for alumina have to be chosen so that

(a) a representative proportion of the material is actually tested
(b) they can all receive identical processing
(c) they can be easily polished
(d) they are as large as the facilities permit
The required parameters are the strength and Weibull modulus. These can be
used, along with the “stress-volume-integral” obtained from the results of finite ele-
ment analysis, to provide the factors in the four function Weibull equation. From
this the minimum strengths of the representative test pieces may be defined.
The clearly emerging results from this analysis are as follows:
(a) The required strength of the material is increased as the specified risk of failure
is reduced.
(b) The required strength is reduced if the Weibull modulus in increased.
A combination of high strength, high “n” value (making allowance for stress cor-
rosion projected over the length of time in service), and high Weibull modulus
should ensure a satisfactory materials performance.
6  Ceramic Materials Testing and Fracture Mechanics 443

Considering therefore the results in Tables 6.4 and 6.5, to ensure to safety of a

hip prosthesis, a value of 400 MPa for the flexural strength should be expected, hav-
ing made an appropriate deduction for stress corrosion, and expecting to attain a
Weibull modulus of at least 15 (which should be attainable after polishing). These
figures would demand a very fine grained alumina and is in line with current think-
ing. To convert the tensile strengths into experimental flexural strengths or modulae
of rupture requires them to be multiplied by factors given in (1.3) and (1.4) which
in turn depend on the Weibull modulus.
Comparing Fig. 7.3 with Figs. 7.4 and 7.5, it seem evident that load cases B and
C contribute to shear stress, which when resolved cause higher tensile stresses than
in load case A. This would indicate that the ceramic acetabular cup is safer to use
than the polyethylene.
It should be emphasised that this degree of rigour is required only for highly
stressed components. Many of the other applications of alumina (e.g. in the dental
field) do not require such high specifications.

6.7.5  Alternatives Approach to Design

With increased pressure from regulatory bodies demanding stringent testing

throughout the design, validation and manufacturing phases of health care product
development, alternative aids are emerging to assist in the pre-market product per-
formance evaluation studies. Digital Imaging Correlation [31] is one such aid using
digital cameras to quantitatively analysis the surface deformation of device under
load. This system is based on the ability of the high resolution cameras to track a
speckle pattern on the surface of an object as it deforms and provides a visual and
quantitative analysis of surface strains of materials and products undergoing stress
testing. Digital Imaging Correlation has been used in the analysis of prosthetic
stress shielding. Stress shielding occurs as the result of a discrepancy between the
elastic properties of the bone and the implant. The strain characteristics of differing
materials on the femur have been demonstrated in laboratories using strain gauges
and photoelastic techniques.
FEA can still present problems when modelling stress and strain complex geom-
etries, often associated with uncertainty surrounding boundary conditions and mate-
rial properties of the modelling. Digital Imaging Correlation can aid in the validation
of FEA theoretical computer models with a view to reducing the risks associated
with development of new increasing more complex device designs.
444 D. Daily

Fig. 7.6  Femoral head under axial loading, Image supplied by Lucideon

References

1. Lawn BR, Wilshaw TR (1975) Fracture of brittle solids. Cambridge University Press, London
2. Jayatilaka A de S (1979) Fracture of engineering brittle materials. Applied Science Publishers,
London
3. Davidge RW (1979) Mechanical behaviour of ceramics. Cambridge University Press,
Cambridge
4. Kerkhof F (1983) Handbook of ceramics. Schmidt, Freiburg im Breisgau
5. Hahn HG (1976) Bruchmechanik. Teubner, Stuttgart
6. Freiman SW (ed) (1979) Fracture mechanics applied to brittle materials. A.S.T.M., S.T.P. 678
7. Braiden P (1975) An introduction to Weibull statistics. A.E.R.E.R 7165
8. Maier HR, Staerk N, Krauth A (1983) In: Hastings GW, Williams DF (eds) Mechanical prop-
erties of biomaterials. Wiley, New York
9. Stanley P, Fessler H, Sivill AD (1973) An engineer’s approach to the prediction of failure prob-
ability of brittle components. Proc Br Ceram Soc 22:453–487
10. Davidge RW, Tappin G (1968) The effective surface energy of brittle materials. J Mater Sci
3:165–173
11. Feltham P (1966) Deformation and strength of materials. Butterworths, London
12. Timoshenko S (1976) Strength of materials, vol 1. Van Nostrand-Reinhold, New York, p 92
13. Durelli AJ, Morse S, Parks V (1962) The theta specimen for determining tensile strength of
brittle materials. Mater Res Stand 2(2):114–117
14. Daniel IM, Weil NA (1962) Effect of non-uniform stress fields. Studies of the brittle behaviour
of ceramic materials. Report AD 277659
15. Griffiths R, Holloway DG (1970) The fracture energy of some epoxy resin materials. J Mater
Sci 5:302–307
16. Evans AG (1972) A method for evaluating the time-dependent failure characteristics of brittle
materials and its application to polycrystalline alumina. J Mater Sci 7:1137–1146
6  Ceramic Materials Testing and Fracture Mechanics 445

17. Evans JRG, Stevens R (1984) The “C”-ring test for the strength of brittle materials. Trans Br
Ceram Soc 83:14–18
18. Tan SR, Davidge RW et al (1980) Strength-controlling flaws in beta-alumina. Trans Br Ceram
Soc 79:120–127
19. Mordfib L, Kerper MJ (1969) Strength testing of ceramics. Office of Aerospace Research,
Arlington, VA
20. Paris PC, Sih GC (1965) Stress analysis of cracks. ASTM Spec Tech Publ 381
21. Williams P, Evans AG (1973) A simple method for studying slow crack growth. J Test Eval
1(4):264–270
22. Evans AG, Wiederhorn SM (1973) Proof testing of ceramic materials—an analytical basis for
failure prediction. Int J Fract 10(3):379–373
23. Braiden PM, Davidge RW, Airey R (1977) Time dependent strength parameters for tungsten
carbides containing 6 or 16% cobalt at room and elevated temperatures. J Mech Phys Solids
25:257–268
24. Lach S, Dailly DF, Hastings GW (1982) Stress corrosion of debased alumina. Proc Br Ceram
Soc 31:191–200
25. Dalgleish BJ, Pratt PL, Rawlings RD, Fakhr A (1980) The fracture toughness testing of ceram-
ics and acoustic emission. Mater Sci Eng 45:9–20
26. Lawn BR, Swain MV (1975) Microfracture beneath point indentations in brittle solids. J Mater
Sci 10:113–122
27. Champonier RP (1979) A.S.T.M./S.T.P. No. 678, p 60
28. Shaw MC, Braiden PM, de Salvo GJ (1975) The disc test for brittle materials. J Eng Ind
97(1):77–87
29. Khan I, Naylor M, Gupta G (2013) Characterization of Orthopaedic Devices. In: Bandyopadhyay
A, Bose S (eds) Characterization of biomaterials. Elsevier, Amsterdam, pp 323–351
30. Cilingir A (2010) Finite element analysis of the contact mechanics of ceramic-on-ceramic hip
resurfacing prostheses. J Bionic Eng 7(3):244–253
31. Smith R (2015) The application of digital image correlation technology in healthcare. Lucideon
White Paper
32. Lach S (1981) Ph.D. Thesis, C.N.A.A.
33. Carniglia SC (1972) Working model for porosity effects on the uniaxial strength of ceramics.
J Am Ceram Soc 55(12):610–618
34. Davidge RW, Evans AG (1970) The strength of ceramics. Mater Sci Eng 6(5):281–298
35. Duckworth W (1953) Discussion of Ryshkewitch paper compression strength of porous sin-
tered alumina and zirconia. J Am Ceram Soc 36(2):68
36. Knudsen FP (1959) Dependence of mechanical strength of brittle polycrystalline specimens on
porosity and grain size. J Am Ceram Soc 42(8):376–387
37. Passmore EM, Spriggs RM, Vasilos T (1965) Strength-grain size-porosity relations in alu-
mina. J Am Ceram Soc 48(1):1–7
38. Eudier M (1962) The mechanical properties of sintered low-alloy steels. Powder Metall
5(9):278–290
Chapter 7
Properties of Bioactive Glasses
and Glass-ceramics

L.L. Hench and T. Kokubo

7.1 Definition of bioactivity:

A bioactive material is one that elicits a specific biological response at the interface
of the material which results in the formation of a bond between the tissues and the
material. A common characteristic of bioactive glasses, bioactive glass-ceramics,
and bioactive ceramics is that their surface develops a biologically active hydroxy
carbonate apatite (HCA) layer which bonds with collagen fibrils. The HCA phase
that forms on bioactive implants is equivalent chemically and structurally to the
mineral phase of bone. It is that equivalence which is responsible for interfacial
bonding1–3.

7.2 Bioactive Bonding

Bioactive materials develop an adherent interface with tissues that resist substantial
mechanical forces. In many cases the interfacial strength of adhesion is equivalent
to or greater than the cohesive strength of bone. The interfacial strength of a bioac-
tive implant bonded to bone is 15–40 times greater than the interfacial adherence of
non-bioactive materials (such as Al2O3) (Table 7.1 and Figure 7.1), tested in the
same animal model (rabbit tibia) (Figure 7.2)4.

L.L. Hench (*) • T. Kokubo

Division of Material Chemistry Faculty of Engineering, Imperial College Department
of Materials, Kyoto University, Sakyo-ku Kyoto, 606-01, Japan, Prince Consort Road,
London SW7 2BP, UK

© Springer Science+Business Media New York 2016 447

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_23
448 L.L. Hench and T. Kokubo

Table 7.1 Failure Loads of Bioceramics by Detaching (Pull-Off) Test

Failure Load (kg)
Materials 8 Weeks 24 Weeks Location of Fracture
Dense sintered alumina1,2 0.13 ± 0.02 Interface
Bioglass® 45S5-type glass1,2 2.75 ± 1.80 within material
Ceravital® KGS-type glass-ceramic3 3.52 ± 1.48 4.35 ± 1.45 within material
Cerabone® A-W glass-ceramic1,2 7.44 ± 1.91 8.19 ± 3.6 within bone
Dense sintered hydroxyapatite1,2 6.28 ± 1.58 7.77 ± 1.91 within material
Dense sintered β-3CaO • P2O54 7.58 ± 1.97 not specified
Natural polycrystalline calcite5 4.11 ± 0.98 within material
1.
T. Nakamura, T. Yamamuro, S. Higashi, T. Kokubo and S. Ito (1985) A New Glass- Ceramic
for Bone Replacement: Evaluation of its Bonding to Bone Tissue, J. Biomed. Maters Res. 19,
685–698.
2.
T. Nakamura, T. Yamamuro, S. Higashi, Y. Kakutani, T. Kitsugi, T. Kokubo and S. Ito, 1985, A
New Bioactive Glass-Ceramic for Artificial Bone, in Treatise on Biomedical Materials, 1,
T. Yamamuro, ed., Research Center for Medical Polymers and Biomaterials at Kyoto University,
Kyoto, Japan, pp. 109–17.
3.
S. Kotani, T. Yamamuro, T. Nakamura, T. Kisugi, Y. Fujmita, K. Kawanabe, T. Kokubo and
C. Ohtsuki (1990) The Bone-Bonding Behavior of Two Glass-Ceramics (KGS and A-W GC), in
Bioceramics, Vol. 2, G. Heimke, ed., German Ceramic Society, Cologne, pp. 105–112.
4.
S. Kotani, Y. Fujita, T. Kitsugi, T. Nakamura, T. Yamamuro, C. Ohtsuki and T. Kokubo (1991)
Bone Bonding Mechanisms of β-tricalcium Phosphate, J. Biomed. Maters. Res. 25, 1303–15.
5.
Y. Fujita, T. Yamamuro, T. Nakamura, S. Kotani, C. Ohtsuki and T. Kokubo (1991) The Bonding
Behavior of Calcite to Bone, J. Biomed. Maters. Res. 25, 1991–2003.
7.
T. Yamamuro (1993) A/W Glass-Ceramic: Clinical Applications, in Introduction to Bioceramics,
eds L.L. Hench and J. Wilson, World Scientific Publishing Co., London, 1993, pp. 89–104.

Fig. 7.1 Comparison of interfacial bond strengths of bioactive implants with non-bonding
implants (alumina) using ‘pull off’ detaching test [4, 5]
7 Properties of Bioactive Glasses and Glass-ceramics 449

Fig. 7.2 Schematic of ‘pull off’ detaching test for determining bone-implant bonding (based upon
T. Yamamuro, ref. 4)

7.3 Bioactive Compositions

Bioactive materials are composed of very specific compositional ranges of Na2O,

CaO, P2O5 and SiO2 due to the importance of these compounds in the in vivo forma-
tion of hydroxy carbonate apatite (HCA) bone mineral (Table 7.2) [1, 2]. All com-
positions either form a HCA layer on their surface or partially dissolve (resorb) as
HCA crystals are formed during the mineralization of osteoid. The rate of formation
of HCA and bone depends upon the composition of the material (Figure 7.3) with
bioactive glasses and glass-ceramics containing < 52% SiO2 being the most rapid.
The time difference in time dependence of interfacial bond strength (Figure 7.1) is
due to the different rates of growth of the interfacial HCA layer. The bioactivity
index of a materials (IB) is defined as:
IB = 100/t0.5bb
IB is obtained from Figure 7.3, and is proportional to the reciprocal of the time
required for one half (0.5) of the interface to be bonded to bone. IB values are shown
in Table 7.2 for the various bioactive implants.
450 L.L. Hench and T. Kokubo

Fig. 7.3 Time dependence of interfacial bone formation for various types of bioceramic implants

The compositional dependence of bonding of bone to various bioactive glasses

is illustrated in Figure 7.4 for the Na2O-CaO-SiO2 system, with a constant 6 weight
percent P2O5 and in Figure 7.5 for the CaO-P2O5-SiO2 system. Figure 7.4 also shows
iso IB values for the Na2O-CaO-P2O5-SiO2 system. When IB = 0 there is no interfacial
bond with bone; i.e., the material develops a non-adherent fibrous capsule and is
nearly bioinert.

7.4 Physical Properties

Table 7.2 summarizes the physical properties of the bioactive glasses, glassceram-
ics, and ceramics in clinical use, with references. The bioactive glasses are single
phase amorphous materials which have high IB values (rapidly form a bone bond)
but have low mechanical strength and toughness. These materials should be used in
particulate form (as powders), as coatings, or in low load bearing applications, as
listed in Table 7.3. Bioactive glass-ceramics are multi-phase materials with a fine,
homogeneous grain size and good mechanical strength and toughness5 and interme-
diate IB values. They can be used in moderate load bearing
1.
L.L. Hench and E.C. Ethridge, Biomaterials, An Interfacial Approach, p. 137, Academic Press,
New York, 1982.
2.
ö.H. Andersson, K.H. Karlsson, K. Kangasniemi, and A. Yli-Urpo, Models for Physical Properties
and Bioactivity of Phosphate Opal Glasses, Glastech. Ber., 61, 300–305 (1988).
Table 7.2 Composition and Mechanical Properties of Bioactive Ceramics Used Clinically
Sintered7,8,9
hydroxypatite Sintered 10,11
Bioglass®1 Glass-ceramic3 Glass-ceramic4 Glass-ceramic5 Glass-ceramic6 Ca10(PO4)6 β-3CaO⋅
Composition 45S5 S53P42 Ceravital® Cerabone® A-W Ilmaplant® Ll Bioverit® (OH)2 > 99.2% P2O5 > 99.7%
Na2O 24.5 wt% 22.6 wt% 5–10 wt% 0 wt% 4.6 wt% 3–8 wt%
K2 O 0 0.5–3.0 0 0.2 3–8 wt%
MgO 0 2.5–5.0 4.6 2.8 2–21
CaO 24.5 21.8 30–35 44.7 31.9 10–34
Al2O3 0 0 0 0 8–15
SiO2 45.0 53.9 40–50 34.0 44.3 19–54
P2O5 6.0 1.7 10–50 16.2 11.2 2–10
CaF2 0 0.5 5.0 F 3–23
B2O3 0
Phase Glass Glass Apatite Apatite Apatite Apatite Apatite Whitlockite
(Ca10(PO4)6 (O,F2)) (Ca10(PO4)6 (β-3CaO • P2O5)
(OH)2)
Glass β-Wollastonite β-Wollastonite Phlogopite
(CaO · Si02) ((Na1K)Mg3
(AlSiO10)F2)
Glass Glass Glass
Density (g/cm3) 2.6572 3.07 2.8 3.16 3.07
Hardness 458 ± 9.4 680 500 600
(Vickers) (HV)
Compressive strength (MPa) 500 1080 500 500–1000 460–687
Bending strength (MPa) 42(Tensile) 215 160 100–160 115–200 140–154
Young modulus (GPa) 35 100–150 118 70–88 80–110 33–90
Fracture toughness, KIC(MPa 2.0 2.5 0.5–1.0 1.0
m1/2)
Slow crack growth, n 33 12–27
Index of bioactivity lB12 12.5 3.8 5.6 7.5 (est) 3.1
452 L.L. Hench and T. Kokubo

Fig. 7.4 The compositional dependence of bone bonding to bioactive glasses (region A) contain-
ing 6 weight % P2O5. Soft tissue bonding occurs for compositions with IB values > 8 (see text).
Region B: non-bioactive compositions. Glasses in Region C are resorbable. (Based upon chapters
1 and 3 in ref. 1.)

Fig. 7.5 Compositional dependence of bioactivity for glasses in the CaO–P2O5–SiO2. (Based
upon T. Kokubo, ref. 5.)
7 Properties of Bioactive Glasses and Glass-ceramics 453

Table 7.3 Clinical Uses of Bioactive Glasses and Glass-Ceramics

Material Form Application Function
45S5 Bioglass® Bulk Endosseous Space filling and tissue bonding
alveolar ridge
maintenance
Bulk Middle ear Restore conductive hearing by
prostheses replacing part of ossicular chain
Powder Repair of Restore bone lost by periodontal
periodontal defects disease and prevent epithelial
down growth
Powder Fixation of revision Restore bone loss due to loosening
arthroplasty of hip prostheses
Cerabone® Bulk Vertebral Replace vertebrae removed in
(A/W glass-ceramic) prostheses tumor surgery
Iliac crest Replace bone removed for
prostheses autogenous graft
Coating Fixation of hip Provide bioactive bonding of
prostheses implant
S53P4 Bulk Orbital floor Repair damaged bone supporting
prostheses eye
Powder Cranial repair Repair bone lost due to trauma
3.
H. Bromer, K. Deutscher, B. Blenke, E. Pfeil and V. Strunz, Properties of the Bioactive Implant
Material Ceravital®, in Science of Ceramics, Vol. 9, 1977, pp. 219–223.
4.
T. Kokubo, Mechanical Properties of a New Type of Glass-Ceramic for Prosthetic Applications,
in Multiphase Biomedical Materials, T. Tsuruta and A. Nakajima, eds, VSP, Utrecht, Netherlands,
1989.
5.
G. Berger, F. Sauer, G. Steinborn, F.G. Wishsmann, V. Thieme, St Kohler and H. Dressel, Clinical
Application of Surface Reactive ApatitelWollastonite Containing Glass-Ceramics, in Proceedings
of XV International Congress on Glass, Vol. 3a, O.V. Mazurin, eds, Nauka, Leningrad, 1989,
pp. 120–126.
6.
W. Vogel and W. Holland, The Development of Bioglass® Ceramics for Medical Applications,
Angew Chern. Int. Ed. Engl. 26,527–544 (1987).
7.
M. Jarcho, C.H. Bolen, M.B. Thomas, J. Bobick, J.F. Kay and RH. Doremus, Hydroxyapatite
Synthesis and Characterization in Dense Polycrystalline Form, J. Mater. Sci. 11, 2027–2035
(1976).
8.
M. Akao, H. Aoki, and K. Kato, Mechanical Properties of Sintered Hydroxyapatite for Prosthetic
Applications, J. Mater. Sci. 16,809–812 (1981).
9.
G, Dewith, H.J.A. Van Dijk, N. Hattu and K. Prijs, Preparation, Microstructure and Mechanical
Properties of Dense Polyerystalline Hydroxyapatite,J. Mater. Sci. 16, 1592–1598 (1981).
10.
M. Jarcho, RL. Salsbury, M.B. Thomas and RH. Doremus, Synthesis and Fabrication of
β-tricalcium Phosphate (Whitlockite) Ceramics for Potential Prosthetic Applications, J. Mater. Sci.
14, 142–150 (1979).
11.
M. Akao, M. Aoki, K. Kato and A. Sato, Dense Polycrystalline β-tricalcium Phosphate for
Prosthetic Applications, J. Mater. Sci. 17, 343–346 (1932).
12.
L.L. Hench, Bioactive Ceramics, in Bioceramics: Materials Characteristics Versus In Vivo
Behavior, P. Ducheyne, J.E. Lemons, eds, Annuals of the NY Academy of Sciences, u523, 1988,
pp. 54–71
applications as bulk materials (Table 7.3). Polycrystalline bioactive ceramics, such
as synthetic hydroxyapatite (HA), have moderate strengths and relatively low IB
values and should be used as particulate or in nonload bearing applications. Compo-
454 L.L. Hench and T. Kokubo

Fig. 7.6 Thickness of interfacial bonding layers for various bioceramics

sitions with the highest IB values develop interfacial bonding layers (Figure 7.6)
composed of both hydrated silica gel layers and Ca, P-rich layers. Compositions
with low to moderate IB values form thinner bonding zones composed primarily of
Ca-P-rich compounds. Non-bonding implants have a non adherent fibrous tissue
layer at the implant interface.

References

1. Hench, L.L. and Wilson, J. (eds) (1993) Introduction to Bioceramics, World Scientific
Publishers, London and Singapore, pp. 1–24.
2. Gross, U., Kinne, R., Schmitz, H.J. and Strunz, V. (1988) The response of bone to surface
active glass/glass-ceramics. CRC Critical Reviews in Biocompatibility, 4, 2.
3. Yamamuro, T., Hench, L.L. and Wilson, J. (eds) (1990) Handbook of Bioactive Ceramics,
Vol 1: Bioactive Glasses and Glass-Ceramics, CRC Press, Boca Raton, FL.
4. Yamamuro, T. (1993) A/W glass-ceramic: clinical applications, in Introduction to Bioceramics
(eds L.L. Hench and J. Wilson), World Scientific Publishers, London and Singapore,
pp. 89–104.
5. Kokubo, T. (1993) A/W glass-ceramic: processing and properties, in Introduction to
Bioceramics (eds L.L. Hench and J. Wilson), World Scientific Publishers, London and
Singapore, pp. 75–88.
Chapter 8
Wear

M. LaBerge and J.D. Desjardins

8.1  Introduction

Biomaterials used in the fabrication of implants are subjected to wear. Wear of

­biomaterials and devices has been shown to be detrimental to their long term suc-
cess resulting in implant retrieval and revision. One of the most dramatic impacts
of the wear of biomaterials and its consequences is observed with artificial joints.
As stated by [1] wear has emerged as a central problem limiting the long-term lon-
gevity of total joint replacements. Ultra-high-molecular-weight polyethylene
(UHMWPE) wear debris has been shown by many authors to trigger an osteolytic
reaction which leads to implant loosening [2]. Wear is a process resulting in the
progressive loss of material involving many diverse mechanisms and phenomena
which are often unpredictable (Table 8.1). The wear process of materials is pre-
dominantly governed by their mechanical and/or chemical behavior. More often
than not, the wear processes listed in Table 8.1 do not act independently. However,
even though several wear mechanisms are involved, it is often the case that one
particular mechanism dominates (Dowson, 1981).
Unfortunately surface wear of an implant results from its use, and therefore, can-
not be avoided or eliminated. Because wear is a limiting factor in the successful
outcome and lifetime of an implant, it is of the utmost importance to characterize
the wear resistance of materials used in implant design, and the effect of the design
on wear. The volume of material removed from surfaces in specific tribosystems as
a result of wear processes has been described phenomenologically and estimated by
different models (Table 8.2). Several experimental wear studies have been con-
ducted to (1) predict the amount of material removed in specific conditions,(2)

M. LaBerge (*) J.D. Desjardins

Department of Bioengineering, 301 Rhodes Clemson University,
Clemson, SC 29634-0905, USA

© Springer Science+Business Media New York 2016 455

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_24
456 M. LaBerge and J.D. Desjardins

Table 8.1  Wear mechanisms

Wear mechanisms Definition
Adhesive wear Characterized by the transfer of material from one surface to another
surface during relative motion. This type of wear is a consequence of
adhesive forces acting at the junction of surface asperities. The transferred
fragments may either be permanently or only temporarily attached to the
other surface. Adhesive wear has been denoted as being the most
commonly detected mechanism of wear, unfortunately it is also the least
preventable (Dowson, 1981).
Abrasive wear Results from a hard asperity damaging or ploughing the surface of a softer
material. The presence of hard particles may be due to the original material
properties of one of the surfaces or loose debris particles which have
become entrapped between the two sliding surfaces and/or embedded into
one of the surfaces expediting abrasive wear. Generally, the resistance to
abrasion can be related to the hardness of the material, however, this
relationship is not directly proportional (Suh, 1986).
Delamination Involves material removal subsequently to plastic deformation, crack
wear nucleation, and propagation in the subsurface (Jahanmir and Suh, 1977;
Suh, 1986)
Fatigue wear Associated with cyclic stress variations and therefore, the lifetime of the
material is dependent on the number of cycles. Cyclic deformation of the
contacting surfaces leads to the initiation and propagation of microcracks
(Rowe, 1980). Subsurface crack initiation generally occurs in the region of
maximum shear stress which will depend upon the geometry of the materials
Fretting wear Generated by a relative oscillatory tangential movement of small
amplitudes (damage can be caused by movement with amplitudes as small
as 0.125 μm) which may occur between two surfaces in contact subjected
to vibration [7]
Corrosive wear Observed when the environment interacts chemically or electrochemically
with one or both of the surfaces. Therefore, the wear rate is dependent on
the environmental conditions affecting the chemical reactivity of the
surfaces. This type of wear mechanism is important for biomaterials since
they function in an extremely harsh environment, the human body [8].
Cavitation wear Occurs from the collapse of cavitation bubbles where the surrounding
liquid rushes to refill the void and collides with the material surface.
Craters form on the surface of a softer material while a harder surface will
experience cracks and spalls which come from subsurface damage.
Subsurface damage induced by cavitation is observed in the softer or
weaker phases of the material [9, 10]

compare the effect of different fabrication and sterilization processes on materials,

(3) produce wear debris to be used in biocompatibility studies, and (4) characterize
the behavior of a new material destined for biomedical applications. Overall wear
tests are primarily conducted to ascertain the basic mechanisms of wear for a par-
ticular combination of materials, or the more restrictive yet equally elusive determi-
nation of the rate of wear to facilitate the estimation of their 'clinical' life.
Experimental results are highly dependent on the geometry of the contact, the lubri-
cant, the tribological conditions including velocity and load, and material properties
[4]. Therefore, experimental protocols aimed at investigating the wear properties of
8 Wear 457

Table 8.2  Selected wear models

Wear process Model Specifications
Adhesive Archard equation: •  Use hardness as the only
weara [12] V F material property, even
Wad= = K N though K depends on
L H
various properties of both
materials
Wad = wear rate (worn volume per unit sliding •  Implies that wear rate is
distance) proportional to real contact
K = wear coefficient area in plastic contacts and
V = volume of wear may not be applicable for
L = sliding distance cases involving elastic
FN = normal load contacts
H = hardness of the softer material
[13] Hornbogen equation: •  B ased on a comparison of
Py E ′FN1.5 the strain occurring during
Wad = N 2 asperity interactions with
K IC 2 H 1.5
Wad = wear rate the critical strain at which
KIC = fracture toughness crack growth is initiated. If
Py = yield strength the applied strain is smaller
N = work hardening factor than the critical strain, the
E′ = equivalent elastic modulus wear rate is independent of
E′ = 1/[(1 − v12)/E1 + (1 − v22)/E2] toughness, and follows
Archard’s
Abrasive Rabinowicz model: •  Assumes that asperities of
wear [1] V KFN the harder surface are
Wab = = tan θ
L πH conical
Wab = wear rate (worn volume per unit sliding
distance)
K = wear coefficient
V = volume of wear
L = sliding distance
FN = normal load
H = hardness of the softer material
tan θ = weighted average of the tan θ values of
all the individual cones θ = average slope of
the asperities
[14] Zum Gahr model: •  C onsiders the processes of
fab cos ρ sin θ microcutting,
Wab = FN microploughing, and
K1K 2τ c cos (θ / 2 )  0.5 cos (θ − ρ )
  microcracking in the
abrasive wear of ductile
metals
Wab = wear rate •  This model includes the
fab = model factor (1 for microcutting) effects of work hardening
K1 = relaxation of normal and shear stress ductility homogeneity of
K2 = texture factor (1 for fcc metals) strain crystal anisotropy
tc = shear stress for dislocation movement
ρ = friction angle at abrasive-material
interface
(continued)
458 M. LaBerge and J.D. Desjardins

Table 8.2 (continued)
Wear process Model Specifications
Fatigue wear Halling model: •  Incorporates the concept of
[15] ηγ fatigue failure as well as
Wfa = K 2 FN
e1 H simple plastic deformation
Wfa = wear rate failure
η = line distribution of asperities
γ = constant defining particle size
e1 = strain to failure in one loading cycle
H = hardness of the softer material
K = wear coefficient
Corrosive Quinn model: •  E
 xplains wear in steel and
wear dA exp [ −Q / RcTc ] assumes that surface
[16, 17] Wcorr = c FN asperity layers formed
3e2 ρ 2 vH
tribochemically are
Wcorr = wear rate detached at a certain critical
ρ = density of material thickness
Ac = Arrhenius constant
Q = activation energy
Rc = gas constant
Tc = contact temperature
d = asperity contact diameter
v = sliding velocity
e = critical thickness of reaction layer
Ultrahigh-­ Wang model: •  F or ultrahigh-­molecular-­
molecular-­ d ( µ − µ0 )  1 1   sin 2α  weight polyethylene in
weight k = k′  −  × 1 −  lubricated multi-directional
polyethylene 2γ c X
 c X 0   2α  sliding
(UHMWPE) k′ = constant •  Assumes the occurrence of
wear [18] d = diameter of fibrils (mm) preferential molecular
μ = coefficient of friction alignment in the principal
μ0 = coefficient of friction for initiating direction of sliding and
surface failure rupture or splitting of the
Xc = cross-link density (mole/g) oriented molecules in the
X0 = critical cross-link density (mole/g) perpendicular direction
γc = C–C bond energy (Joule) associated with secondary
α = maximum cross-shear angle (radian) sliding
a
Note
• The wear factor (k) is a measure of the rate at which a given combination of materials wears in
the environment of the test. k is widely used for comparative purposes (mm3/Nm)
• According to Dowson [19], if the mean contact stresses are not too high the wear of polymer
against a hard surface (metal or ceramic) is obtained with fair accuracy with the relationship
V = kFL
• K is directly influenced by the roughness average (Ra) of the metallic counterface for the contact
UHMWPE–stainless steel in water under reciprocating pin-on-plate conditions given by the
relationship k = 4.0 × 10−5 Ra1.2 [20]
8 Wear 459

biomaterials should be designed to assess or predict their behavior in simulated

clinical conditions [21].

In Vitro Wear Testing

Although wear is a very complex process, apparatuses are available which allow for
the accumulation of data resulting in an estimate of the wear resistance of a combi-
nation of materials or a device. Preliminary material studies will commonly be per-
formed on laboratory wear benches while devices will be evaluated with simulators.
The classic arrangement of articulating surfaces in material wear testing involves an
upper bearing part that is attached to a pin, and a lower bearing material that is
attached to a specimen holding plate, or disk. The primary advantage of this con-
figuration is that any combination of candidate bearing materials can be incorpo-
rated into the testing setup. Typically, this involves either a spherical or a flat
material geometry on the pin, and a flat material geometry at the disk counterface.
The sacrificial material (which experiences the greatest amount of wear) can be
fabricated into either the pin or the disk; the decision regarding location of the sac-
rificial material depends on the particular type of analysis desired. Although it has
an important effect on the resulting wear, primarily because the material on the
surface of the pin is under constant load while the disk is loaded only periodically
in discrete regions, the importance of this decision is often overlooked. Historical
configurations of this pin-on-flat wear testing arrangement are most often referred
to as a “pin-on-disk” (POD) wear test. Many variations of this arrangement reported
in the literature include rotary pin-on-disk, reciprocating pin-on-disk, circularly
translating pin-on-disk, and multi-axis pin-on-disk configurations [22]. For clarity,
the translations and rotations of both the pin and disk have been described using
consistent terminology as follows: (1) translations are given with respect to a fixed
point of observation not on either the pin or the disk; circular translation does not
imply rotation about an internal axis; and (2) rotations are given with respect to the
central axis of either the pin or the disk being described, wherein the axis is normal
to the bearing surface.
One of the most basic wear testing configurations is the rotary pin-on-disk
method. In this method the pin is neither translated nor rotated, and the disk is
rotated without translation, producing a circular wear track on the flat specimen.
This configuration is often used to produce constant-velocity, unidirectional shear-
ing forces at the pin and disk surfaces. The pin surface is under continuous load and
unidirectional shear conditions, while only discrete locations of the flat specimen
are periodically loaded in unidirectional shear during each cycle. In this manner, the
pin specimen experiences the majority of the wear potential. Depending upon mate-
rial fabrication and geometric constraints, the candidate material under wear inves-
tigation can be fabricated into either pins or sheets; typically the softer material (the
polymer in a metal/polymer combination) is used as the upper bearing surface.
460 M. LaBerge and J.D. Desjardins

A similar type of wear testing configuration is the reciprocating pin-on-disk

method. In a reciprocating pin-on-disk wear test, the pin is linearly translated back
and forth along a single line trajectory without rotation, while the disk neither trans-
lates nor rotates [22]. This produces a linear wear track, with a stop/start condition
at either end of the wear track, and a variable (usually sinusoidal) velocity profile
along the trajectory. In this configuration, the pin surface is under continuous load
and reciprocating shear conditions, while the discrete locations of the flat specimen
of the wear track are periodically loaded in reciprocating shear. Similar to the rotary
pin-on-disk method, the pin specimen experiences the majority of the wear poten-
tial. Historically, this reciprocating pin-on-disk configuration has been extensively
used in all types of material-pair wear evaluations.
Many polymers have been shown to wear more severely under conditions that
produce multi-directional shear forces [23] than under unidirectional and recipro-
cating wear testing conditions. This has been theorized to be a result of a process
whereby the long molecular chains of the polymer reorient themselves along the
directions of the shear vector [24], and has been assessed using soft X-ray absorp-
tion spectroscopy to measure molecular orientation at UHMWPE surfaces [25].
Such findings have resulted in the introduction of a new class of pin-on-disk wear
testing systems that attempt to subject the material surfaces to multi-directional
shearing forces similar to those seen in total joint replacements [26–28]. The circu-
larly translating pin-on-disk (CTPOD) system was introduced as a method by which
the pin material could be subjected to continuous multi-directional shearing condi-
tions [29]. Multi-axis pin-on-disk wear testing incorporates one or two additional
degrees of motion into a standard reciprocating wear testing profile to produce spe-
cific cross-shear conditions [26, 28, 30]. The multi-axis wear testing devices utilize
a simplified rectangular path as a wear pattern [18, 26, 31–33], more representative
multi-directional shearing conditions [29] with shear vector orientations changing
gradually over a complete range of angles during constant sliding [32], or more
multi-axis cross-shear system incorporating the kinematics of bearing surfaces in
total joint replacements [22] Overall, the latter systems have provided valuable
insight into the implications of multi-directional motion on wear rates.
In general, most pin-on-disk wear testing configurations produce the highest
wear potential for the material that is placed at the pin. For this reason, the poly-
meric material is most often located there, with the metallic material acting as the
disk. Intrinsic to this configuration, however, is the fact that the polymer experi-
ences constant load across the entire contacting surface area of the pin, and wear is
measured by gravimetric analysis of material removal. This is contrary to the in vivo
condition of total knee joint bearing surfaces whereby the polymer is cyclically
loaded in discrete locations. The greater physiological relevance of the opposite
situation (metal pin on polymer flat) has been noted [34, 35] (Tables 8.3 and 8.4).
The standard procedure under the American Society for Testing and Materials
(ASTM F-732-00, 2011—Standard Test Method for Wear Testing of Polymeric
Materials Used in Total Joint Prostheses) describes the testing protocol for charac-
terizing the wear resistance of material combinations to be used in the design of
orthopaedic total joint replacements. Several investigators have used modified or
8 Wear 461

Table 8.3  Example of biomaterial combinations tribologically characterized with POD and
simulators
Material combinations Test apparatus
Stainless steel-UHMWPE Pin-on-disc; joint simulator [36–38]
Co-Cr-alloys-UHMWPE Pin-on-disc; joint simulator [38–43]
Titanium alloys-UHMWPE Pin-on-disc; joint simulator [40, 44, 45]
Alumina-UHMWPE Pin-on-disc; joint simulator [38, 42, 43, 46]
Zirconia-UHMWPE Pin-on-disc; joint simulator [37, 42, 47]
CoCrMo-CoCrMo Pin-on-disc; joint simulator [46, 48–52]
CoCrMo-Delrin Disc-on-flat [53]
PEEK-CoCrMo Pin-on-plate [54, 55]
CoCrMo-alumina Joint simulator [42]
Ti6A14V-alumina Joint simulator [42]
Alumina-alumina Wear and friction benches; joint simulator [46, 51,
56–59]
Elastomers-metal Reciprocating friction benches; joint simulator [22,
60–68]
Dental resins-enamel Pin-on-flat [69]
Modified UHMWPE-metal Pin-on-disc [70–72] Joint Simulator [73–75]

Table 8.4  In vitro measurement of wear machines

Wear apparatus type Comments
Pin-on-disc or pin-on-plate Useful in studying basic wear mechanisms. Steady and
well-controlled operating conditions. Fails to replicate the
reciprocating motion observed in joints
Reciprocating pin-on-disc Simulates the reciprocating motion observed in vivo and fatigue
(POD) loading (loading and unloading)
Multi-directional/multi-axis Incorporates one or two additional degrees of motion into a
pin-on-disc standard reciprocating wear testing profile to produce specific
cross-shear conditions
Joint simulators Improved understanding of the wear processes encountered in
(knee, hip, spine/disc prostheses by simulating mechanical conditions observed in
simulators) patients

adapted versions of this standard to assess the wear resistance of bearing surfaces
for orthopaedic applications [48, 60, 76–82]. Other ASTM methods and protocols
from the International Organization for Standardization (ISO) pertinent to the
­evaluation of wear performance of engineering materials and devices are listed in
Table 8.5.
Wear resistance is usually reported in terms of wear rate, either linear or
­volumetric, with different units such as volume lost per 106 cycles (MC), mass loss
per 106 cycles, or linear displacement per 106 cycles. A complete walking cycle is
represented by two steps. One cycle on a reciprocating pin-on-flat system is obtained
462 M. LaBerge and J.D. Desjardins

Table 8.5  In vitro friction and wear measurement standards

a
ASTM G40-13 Standard Terminology Relating to Wear and Erosion
ASTM Standard Test Method for Ranking Resistance of Materials to Sliding Wear
G77-05(2010) Using Block-on-Ring Wear Test
ASTM Standard Guide for Recommended Format of Wear Test Data Suitable for
G118-02(2015) Databases
ASTM G119-09 Standard Guide for Determining Synergism Between Wear and Corrosion
ASTM Standard Test Method for Linearly Reciprocating Ball-on-Flat Sliding
G133-05(2010) Wear
ASTM Standard Test Method for Ranking Resistance of Plastics to Sliding Wear
G176-03(2009) Using Block-on-Ring Wear Test—Cumulative Wear Method
ASTM G190-15 Standard Guide for Developing and Selecting Wear Tests
ASTM G206-11 Standard Guide for Measuring the Wear Volumes of Piston Ring Segments
Run against Flat Coupons in Reciprocating Wear Tests
ASTM Standard Test Method for Wear Testing of Polymeric Materials Used in
F732-00(2011) Total Joint Prostheses
ASTM Standard Guide for Gravimetric Wear Assessment of Prosthetic Hip
F1714-96(2013) Designs in Simulator Devices
ASTM Standard Practice for Fretting Corrosion Testing of Modular Implant
F1875-98(2014) Interfaces: Hip Femoral Head-Bore and Cone Taper Interface
ASTM Standard Practice for Characterization of Particles
F1877-05(2010)
ASTM Standard Practice for Gravimetric Measurement of Polymeric Components
F2025-06(2012) for Wear Assessment
ASTM Standard Test Method for Determining Femoral Head Penetration into
F2385-04(2010) Acetabular Components of Total Hip Replacement Using Clinical
Radiographs
ASTM F2423-11 Standard Guide for Functional, Kinematic, and Wear Assessment of Total
Disc Prostheses ASTM F2694-07(2013) Standard Practice for Functional
and Wear Evaluation of Motion-Preserving Lumbar Total Facet Prostheses
ASTM F2624-12 Standard Test Method for Static, Dynamic, and Wear Assessment of
Extra-Discal Single Level Spinal Constructs
ASTM Standard Practice for Functional and Wear Evaluation of Motion-­
F2694-07(2013) Preserving Lumbar Total Facet Prostheses
ASTM F2979-14 Standard Guide for Characterization of Wear from the Articulating
Surfaces in Retrieved Metal-on-Metal and other Hard-on-Hard Hip
Prostheses
ASTM Standard Guide for High Demand Hip Simulator Wear Testing of
F3047M-15 Hard-on-hard Articulations
ASTM D7596-14 Standard Test Method for Automatic Particle Counting and Particle Shape
Classification of Oils Using a Direct Imaging Integrated Tester
b
ISO Implants for surgery—Wear of total hip-joint prostheses—Part 1: Loading
14242-1:2014 and displacement parameters for wear-testing machines and corresponding
environmental conditions for test
ISO Implants for surgery—Wear of total hip-joint prostheses—Part 2: Methods
14242-2:2000 of measurement
(continued)
8 Wear 463

Table 8.5 (continued)
ISO Implants for surgery—Wear of total knee-joint prostheses—Part 1:
14243-1:2009 Loading and displacement parameters for wear-testing machines with load
control and corresponding environmental conditions for test
ISO Implants for surgery—Wear of total knee-joint prostheses—Part 2:
14243-2:2009 Methods of measurement
ISO Implants for surgery—Wear of total knee-joint prostheses—Part 3:
14243-3:2014 Loading and displacement parameters for wear-testing machines
with displacement control and corresponding environmental conditions
for test
ISO/TS Dental materials—Guidance on testing of wear—Part 2: Wear by two and/
14569-2:2001 or three body contact
ISO 16428:2005 Implants for surgery—Test solutions and environmental conditions for
static and dynamic corrosion tests on implantable materials and medical
devices
ISO 17853:2011 Wear of implant materials—Polymer and metal wear particles—Isolation
and characterization
ISO Implants for surgery—Wear of total intervertebral spinal disc prostheses—
18192-1:2011 Part 1: Loading and displacement parameters for wear testing and
corresponding environmental conditions for test
ISO Implants for surgery—Wear of total intervertebral spinal disc prostheses—
18192-2:2010 Part 2: Nucleus replacements
a
American Society for Testing and Materials, Philadelphia, PA (www.astm.org)
b
International organization for Standardization, Geneva, Switzerland (www.iso.org)

by two passes (return to starting point), while the other cycle on a rotating pin-on-­
disc system corresponds to one revolution. It is assumed that a normal individual
will make two million steps per year while an active subject may make more than
10 million steps [21] at a maximum frequency of 1 Hz. Investigators have also
reported wear rates as cubic millimeters (volume) per millimeter (sliding distance)
(mm3/mm). The volume is calculated by measuring the mass loss and using the
density of the polymer as a conversion factor. Tables 8.6 and 8.7 present a critical
selection of wear data available for biomaterial tribosystems useful to the orthopae-
dic design community. Both friction coefficient and wear rate are used as design
parameters. Unless independently monitored, friction coefficients are usually
acquired during wear tests. The static coefficient of friction is calculated using the
force required to initiate motion. The kinetic coefficient of friction may vary during
a test for a constant velocity and should be calculated from averaged force readings
during the duration of the test. The ASTM method G115–10 (2013) proposes a
guide for “Measuring and Reporting Friction Coefficients” which is designed to
assist investigators in the selection of an appropriate method for measuring the fric-
tional properties of materials.
Table 8.6  Friction coefficients of various implant materials from representative in vitro studies
464

Average (or range)

Material contact friction coefficient Testing apparatus Tribological conditions References
Stainless steel 0.07–0.13 Pin-on-disk •  Load = 3.45 M Pa [40]
(316LVM) + UHMWPE •  Velocity = 50 × 106 mm/year
•  Lub#: serum
•  Duration: 2 years test
•  28–32 °C
Stainless steel 0.078 Pin-on-disk •  Load = 3 MPa [37]
(316L) + UHMWPE •  Velocity = 60 mm/s
•  Lub: bovine serum, 40–50 ml
•  24–26 °C
Stainless steel (100CR6-­ 0.17 (a); 0.10 (b); 0.16 Ball-on-disk •  10 mm diameter ball [83]
German) + UHMWPE (c); 0.14 (d) vibrotribometer •  Oscillation = 10 Hz; 1.65 mm amplitude
(Chirulene-German) (Optimol •  Load = 50 and 300 N
SRV-German) •  Lub: (a) none; (b) human synovial fluid; (c)
yellow bone marrow; (d) red bone marrow
•  37 °C
Stainless steel + UHMWPE 0.03–0.09 Reciprocating •  Load = 445 N [38]
flat-on-flat •  Velocity = 100 cycles/min
•  Lub: bovine serum
•  Duration: 3.7 × 106 cycles
Stainless steel + UHMWPE 0.05 Dual hip simulator •  Velocity = 30 cycles/min [84]
(Charnley) •  Load = 250 kg
•  Duration: 1000 h
•  Lub: serum
•  Room temperature
Stainless steel (Ortron 0.034 Single-channel hip •  Range of motion = 30° [85]
90) + UHMWPE (ASTM F 648) joint simulator •  Lub: Deionized water at 37 °C
•  Load = 1–4 kN
•  Angular Velocity = 0.6–2.4 rad/s
M. LaBerge and J.D. Desjardins
 Average (or range)
Material contact friction coefficient Testing apparatus Tribological conditions References
Stainless steel 0.040 Single-channel hip •  Range of motion = 30° [85]
(316L) + UHMWPE joint simulator •  Lub: Deionized water at 37 °C
•  Load = 1–4 kN
8 Wear

•  Angular velocity = 0.6–2.4 rad/s
Stainless steel 0.03–0.09 Twelve-channel •  Conforming, flat-on-flat configuration [86]
(316L) + UHMWPE (ASTM F friction and wear •  Velocity = 100 cyc./min
648) machine FW-12 •  Pressure = 6.90 MPa
•  Duration: 3.7 × 106 cycles
•  Lub: bovine calf serum w/55 sodium azide
CoCrMo Dry 0.13 Lub: 0.21 Rolling-sliding •  Velocity = 25 m/min [46]
(Protasul-2) + UHMWPE apparatus •  Pressure = 30 N/cm2
•  Duration: 20 h
•  Lub: none or distilled water
•  Room temperature
Co-­Cr + UHMWPE 0.05–0.11 Pin-on-disk •  Load = 3.45 M Pa [40]
•  Velocity = 50 × 106 mm/year
•  Lub: serum
•  Duration: 2 years test
•  28–32 °C
Co–Cr + UHMWPE 00.07–0.25 Pin-on-flat •  Axial load = 223 N [87]
•  Duration 250,000 cycles
•  Lub: bovine serum
Co–Cr–Mo 0.08–0.15 Pin-on-flat •  Contact pressure = 4.8 MPa [42]
(Vitallium) + UHMWPE •  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  Lub: distilled, deionized H2O
•  37.1 °C
(continued)
465
Table 8.6 (continued)
466

Average (or range)

Material contact friction coefficient Testing apparatus Tribological conditions References
Co–Cr–Mo (ASTM 0.060–0.093 Reciprocating motion •  Maximum stress = 6 MPa [88]
F799) + UHMWPE (GUR 415) friction bench (line/ •  Lub: 86 % Glycerine
flat) •  Frequency = 1 Hz
•  Duration: 500,000 cycles
•  Sliding Dist. = 100 mm/cycle
Co–Cr–Mo (Muller) + UHMWPE 0.018–0.045 Durham hip function •  Range of motion = 20° [89]
simulator •  Lub: Carboxymethyl cellulose
•  Dynamic load = 2000 N
•  Frequency = 1 Hz
Co–Cr–Mo 0.057 Single-channel hip •  Range of motion = 30° [85]
(Vitallium) + UHMWPE (ASTM joint simulator •  Load = 1–4 kN
F 648) •  Angular velocity = 0.6–2.4 rad/s
•  Lub: deionized water at 37 °C
Co–Cr–Mo 0.038–0.063 Single-channel hip •  Range of motion = 30° [85]
(Zimalloy) + UHMWPE (ASTM joint simulator •  Load = 1.4 kN
F 648) •  Angular velocity = 0.6–2.4 rad/s
•  Lub: deionized water at 37 °C
Co–Cr–Mo (ASTM 0.044 Hip simulator machine •  Lub: synovial fluid [90]
F75) + UHMWPE •  Peak load = 150 kg
Co–Cr–Mo (ASTM F 0.052–0.070 Single-channel hip •  Range of motion = 30° [85]
799) + UHMWPE (ASTM F 648) joint simulator •  Load = 1.4 kN
•  Angular velocity = 0.6–2.4 rad/s
•  Lub: deionized water at 37 °C
CoCrMo + UHMWPE 0.06 Dual hip simulator •  Velocity = 30 cycles/min [84]
(Charnley-Muller) •  Load = 250 kg
•  Duration: 100 h
•  Lub: serum
•  Room temperature
M. LaBerge and J.D. Desjardins
 Average (or range)
Material contact friction coefficient Testing apparatus Tribological conditions References
Co–Cr–Mo + UHMWPE 0.05–01.1 Twelve-channel •  Conforming, flat-on-flat configuration [86]
friction and wear •  Velocity = 100 cyc./min
machine FW-12 •  Pressure = 6.90 MPa
8 Wear

•  Duration: 3.7 × 106 cycles
•  Lub: bovine calf serum w/55 sodium azide
Co–Cr–Mo + UHMWPE 0.06–0.07 Hip joint simulator •  Load 2.5 kN static load [91]
(a,c) •  Lub: (a) serum, (b) serum albumin, (c)
0.10–0.12 synovial fluid, (d) veronate buffer
(b,d)
Co–Cr–Mo + UHMWPE 0.04–0.06 Hip joint simulator •  Load 2.5 kN static load [92]
•   elocity: 30 cycles/min
V
•  Duration: 1.8 million cycles
•  Lub: serum
Co–Cr–Mo + UHMWPE 0.03–0.05 Pin-on-disk •  Load = 100 N [93]
•  Lub: Ringer’s solution
•  Velocity = 0.05 m/s
•  Duration: 48 h
Ti–6A1–4V + UHMWPE 0.04–0.26 Pin-on-flat •  Axial load = 223 N [87]
•  Duration: 250,000 cycles
•  Lub: bovine serum
Ti–6A1–4V + UHMWPE 0.05–0.121 Reciprocating flat •  Load = 445 N [38]
on-flat •  Velocity = 100 cycles/min
•  Lub: bovine serum
•  Duration: 4.1 × 106 cycles
Ti–6A1–4V (Ion 0.058 Single-channel hip •  Range of motion = 30° [85]
implanted) + UHMWPE (ASTM joint simulator •  Lub: deionized water at 37 °C
F 648) •  Load = 1–4 kN
•  Angular velocity = 0.6–2.4 rad/s
(continued)
467
Table 8.6 (continued)
468

Average (or range)

Material contact friction coefficient Testing apparatus Tribological conditions References
Ti–6A1–4VELI (ASTM F 0.123–0.133 Single-channel hip •  Range of motion = 30° [85]
136) + UHMWPE (ASTM F 648) joint simulator •  Lub: deionized water at 37 °C
•  Load = 1-4 kN
•  Angular velocity = 0.6–2.4 rad/s
Alumina + UHMWPE 0.06–0.10 Reciprocating •  Load = 223 N [38]
flat-on-flat •  Velocity = 60 cycles/min
•  Lub: bovine serum
Alumina + UHMWPE 0.056 Pin-on-disk •  Load = 3 MPa [37]
•  Velocity = 60 mm/s
•  Lub: bovine serum, 40–50 ml
•  24–26 °C
Alumina (Vitox) + UHMWPE 0.06–0.18 Pin-on-flat •  Contact pressure = 4.8 MPa [42]
•  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  Lub: distilled, deionized H2O
•  37.1 °C
Alumina + UHMWPE 0.06–0.25 Pin-on-flat •  Axial load = 223 N [87]
•  Duration: 250,000 cycles
•  Lub: bovine serum
Alumina + UHMWPE Dry: 0.16 Rolling-sliding •  Velocity = 25 m/min [46]
Lub: 0.05 apparatus •  Pressure = 30 N/cm2
•  Duration: 20 h
•  Lub: none or distilled water
•  Room temperature
Alumina (BIOLOX) + UHMWPE 0.022–0.062 Single-channel hip •  Range of motion = 30° [85]
(ASTM F 648) joint simulator •  Lub: deionized water at 37 °C
•  Load = 1–4 kN
•  Angular velocity = 0.6–2.4 rad/s
M. LaBerge and J.D. Desjardins
 Average (or range)
Material contact friction coefficient Testing apparatus Tribological conditions References
Alumina (ASTM F 0.050 Single-channel hip •  Range of motion = 30° [85]
603) + UHMWPE (ASTM F 648) joint simulator •  Lub: deionized water at 37 °C
•  Load = 1–4 kN
8 Wear

•  Angular velocity = 0.6–2.4 rad/s
Zirconia (Y–PSZ) + UHMWPE 0.049 Pin-on-disk •  Load = 3 MPa [37]
•  Velocity = 60 mm/s
•  Lub: bovine serum, 40–50 ml
•  24–26 °C
Zirconia (Zyranox) + UHMWPE 0.05–0.16 Pin-on-flat •  Contact pressure = 4.8 MPa [42]
•  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  Lub: distilled, deionized H2O
•  37.1 °C
Zirconia + UHMWPE (ASTM F 0.059 Single-channel hip •  Range of Motion = 30° [85]
648) joint simulator •  Lub: deionized water at 37 °C
•  Load = 1–4 kN
•  Angular velocity = 0.6–2.4 rad/s
Stainless steel (100CR6-­ 0.6 (a); 0.26 (b); 0.106 Ball-on-disk •  10 mm diameter ball [83]
German) + Stainless steel (c); 0.1 (d) vibrotribometer •  Oscillation = 10 Hz; 1.65 mm amplitude
(100CR6-German) (Optimol •  Load = 50 and 300 N
SRV-German) •  Lub: (a) none; (b) human synovial fluid; (c)
yellow bone marrow; (d) red bone marrow
•  37 °C
Co–Cr–Mo + Co–Cr–Mo 0.03–0.04 Pin-on-disk •  Load = 100 N [93]
•  Lub: Ringer’s solution
•  Velocity = 0.05 m/s
•  Duration: 48 h
(continued)
469
 470

Table 8.6 (continued)
Average (or range)
Material contact friction coefficient Testing apparatus Tribological conditions References
CoCrMo (Protasul-2) + CoCrMo Dry: 0.4 Rolling-sliding •  Velocity = 25 m/min [46]
(Protasul-2) Lub: 0.35 apparatus •  Pressure = 30 N/cm2
•  Duration: 20 h
•  Lub: none or distilled water
•  Room temperature
Co–Cr–Mo + Co–Cr–Mo 0.12–0.13 Hip joint simulator •  Load 2.5 kN static load [91]
(a,b,c) •  Lub: (a) serum, (b) serum albumin, (c)
0.22 (d) synovial fluid, (d) veronate buffer
Co–Cr–Mo + Co–Cr–Mo 0.13–0.14 Hip joint simulator •  Load 2.5 kN static load [92]
•  Velocity: 30 cycles/min
•  Duration: 1.8 million cycles
•  Lub: serum
Co–Cr–Mo (ASTM F75) 0.16 Hip simulator machine •  Lub: synovial fluid [90]
•  Peak load = 150 kg
CoCrMo + CoCrMo Serum and synovial fluid: Dual-hip simulator •  Velocity = 30 cycles/min [84]
(McKee-Farrar) 0.12 Saline: 0.22 •  Load = 250 kg
•  Duration: 1000 h
•  Lub: serum and synovial fluid or saline
•  Room temperature
M. LaBerge and J.D. Desjardins
 Average (or range)
Material contact friction coefficient Testing apparatus Tribological conditions References
CoCrMo + CoCrMo 0.13 Dual-hip simulator •  Velocity = 30 cycles/min [84]
(McKee-Farrar) •  Load = 250 kg
•  Duration: 1000 h
8 Wear

•  Lub: serum
•  Room temperature
Alumina + alumina 0.26–0.35 Pin-on-disk •  Load = 100 N [93]
•  Lub: Ringer’s solution
•  Velocity = 0.05 m/s
•  Duration: 48 h
Alumina + alumina Dry: 0.71 Lub: 0.09 Rolling-sliding •  Velocity = 25 m/min [46]
apparatus •  Pressure = 30 N/cm2
•  Duration: 20 h
•  Lub: none or distilled water
•  Room temperature
471
 472

Table 8.7  Wear rates of various material combination from in vitro studies
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Stainless steel 0.4 mm/year NA •  Pin-on-disk Smooth polymer, [40]
(316LVM) + UHMWPE •  Luba: bovine serum metal—scratches
•  Duration: 2 years test
•  Load = 3.45 MPa
•  Velocity = 50 × 106 mm/year
•  28–32 °C
Stainless steel 0.17–0.23 mm3/106 NA •  Reciprocating flat-on-flat Surface scratching [38]
(316L) + UHMWPE •  Load = 445 N
•  Velocity = 100 cycles/min
•  Lub: bovine serum
•  Duration = 3.7 × 106 cycles
Stainless steel (316L) + NA 27.7 × 10−7  mm3/Nm •  Pin-on-disk Original machine [37]
UHMWPE •  Lub: bovine serum, marks gone, new
40–50 ml wear marks
•  Load = 3 MPa
•  Velocity = 60 mm/s
•  24–26 °C
Stainless steel Machined UHMWPE: NA •  Annular disk on flat pin UHMWPE transfer [45]
(316L) + Machined 3.23  in./in. × 10−9 Molded •  Range of motion = 110° film
UHMWPE (HiFax 1900) UHMWPE: 1.70 in./ •  Sliding velocity = 43.3 in./
in. × 10−9 min
•  Stress = 500 psi
•  Lub: Ringer’s solution
Stainless steel + UHMWPE Max. depth of wear: NA •  Dual-hip simulator Evidence of brittle [84]
(Charnley) 0.15 mm •  Velocity = 30 cycles/min fracture
•  Load = 250 kg
•  Duration: 1000 g
•  Lub: serum
•  Room temperature
M. LaBerge and J.D. Desjardins
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Stainless steel (316L) UHMWPE: NA •  Twelve-channel friction and Quantification of [86]
0.20 mm3/106 cycles wear machine FW-12 wear separately from
UHMWPE 0.65 μm/year •  Conforming, flat-on-flat creep deformation.
8 Wear

configuration Adhesive/abrasive
•  Velocity = 100 cyc./min wear emphasizes over
•  Pressure = 6.90 MPa fatigue wear
•  Duration: 3.7 × 106 cycles
•  Lub: bovine calf serum
w/55 sodium azide
Stainless steel + UHMWPE UHMWPE: NA •  Ten station hip simulator Highly loaded region [94]
40 mg/106 cycles •  Range of motion = 46° of UHMWPE smooth
•  Lub: Bovine blood serum at and shiny, peeling,
37 ° C pitting
•  Load = Oscillating
0–2030 N
•  Frequency = 1 Hz
•  Duration: 1 × 106 cycles
Stainless steel + UHMWPE 1.62 × 10−7  mm3/Nm NA •  Ball-and-socket simulator Uniform superficial [44]
•  Ra < 0.016 scratches, occasional
•  Sterilized with deeper marks. Metal
2.5 Mrad g rad particles, acrylic
•  Lub: bovine serum with cement particles
sodium azide
•  Load = 2000 N
•  Speed = 100 mm/s
(continued)
473
 474

Table 8.7 (continued)
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Cast Co-Cr-Mo pins PE thickness change: NA •  Reciprocating pin-on-flat Abrasive wear of PE. [95]
(ASTM F-75) + UHMWPE 64 ± 13  μm •  Sterilized with 2.5 Mrad Transfer of PE on
(GUR 415 plate) •  Lub: deionized water Co–Cr pins.
•  36 MPa Line contact stress Oxidative wear of
•  Frequency = 2.1 Hz Co–Cr pins
•  Stroke length = 15 mm
•  Duration = 2 × 106 cycles
•  Final friction = 0.079 ± 0.001
Wrought Co-Cr pins PE thickness change: NA •  Reciprocating pin-on-flat Abrasive wear of [95]
(ASTM F-90) + UHMWPE 71 ± 25  μm •  Sterilized with 2.5 Mrad PE. Transfer of PE on
GUR 415 plate •  Lub: deionized water Co–Cr pins.
•  36 MPa Line contact stress Oxidative wear of
•  Frequency = 2.1 Hz Co–Cr pins
•  Stroke length = 15 mm
•  Duration = 2 × 106 cycles
•  Final friction = 0.101 ± 0.019
Co–Cr–Mo CoCrMo: 0.1 mg/20 h NA •  Rolling-sliding wear and NA [40]
(Protasul-2) + UHMWPE UHMWPE: 1 mg/20 h •  Friction apparatus
•  Velocity • 25 m/min
•  Pressure = 30 N/cm2
•  Duration: 20 h
•  Dry condition
•  Room temperature
Co-Cr-Mo (hot 0.5 mm/year NA •  Pin-on-disk Smooth polymer, [87]
pressed) + UHMWPE •  Lub: bovine serum metal—scratches
•  Duration: 2 years test
•  Load = 3.45 M Pa
•  Velocity = 50 × 106 mm/year
•  28–32 °C
M. LaBerge and J.D. Desjardins
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Co–Cr + UHMWPE 1.05 mg/106 cycles NA •  Pin-on-flat Transfer of PE to [42]
•  Lub: distilled, deionized Co-Cr Surface
water Adhesive wear
8 Wear

•  Contact pressure: 4.8 MPa

•  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  37.1 °C
Co-Cr-Mo Machined UHMWPE: NA •  Annular disk on flat pin UHMWPE transfer [45]
(passivated) + UHMWPE 3.23  in./in. × 10−9 Molded •  Range of motion = 110° film
(HiFax 1900) UHMWPE: 1.50 in./ •  Sliding velocity = 43.3 in./ Quantification of
in. × 10−9 min wear separately from
•  Stress = 500 psi creep deformation
•  Lub: Ringer’s solution
•  Twelve-channel friction and
wear machine FW-12
Co-Cr-Mo + UHMWPE UHMWPE: NA •  Conforming, flat-on-flat Adhesive/abrasive [86]
0.17 mm3/106 cycles configuration wear emphasizes over
0.55 μm/year •  Velocity = 100 cyc./min fatigue wear
•  Pressure = 6.90 MPa
•  Duration: 3.7 × 106 cycles
•  Lub: bovine calf serum
w/55 sodium azide
Co–Cr–Mo + UHMWPE UHMWPE: NA •  Ten-station hip simulator Highly loaded region [94]
68 mg/106 cycles •  Range of motion = 46° of UHMWPE smooth
•  Lub: bovine blood serum at and shiny
37 ° C Peeling
•  Load = Oscillating Pitting
0–2030 N
•  Frequency = 1 Hz
•  Duration: 1 × 106 cycles
(continued)
475
 476

Table 8.7 (continued)
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Co-Cr-Mo + UHMWPE Max. depth of wear: NA •  Dual hip simulator Evidence of brittle [84]
(Charnley–Muller) 0.08 mm •  Velocity = 30 cycles/min fracture
•  Load = 250 kg
•  Duration: 1000 h
•  Lub: serum
•  Room temperature
Co-Cr-Mo + UHMWPE 0.15  mm/1.8 × 106  cycles NA •  Hip joint simulator Creep, abrasion, [91]
(Charnley) •  2.5 kN static load adhesion. Max cup
•  Lub: bovine serum wear depth
•  Duration: 1.8 × 106 cycles
•  Velocity = 30 cycles/min
Co–Cr–Mo + UHMWPE 0.075  mm/1.8 × 106  cycles NA •  Hip joint simulator Creep, abrasion, [91]
(Charnley–Muller) •  2.5 kN static load adhesion. Max cup
•  Lub: bovine serum wear depth
•  Duration: 1.8 × 106 cycles
•  Velocity = 30 cycles/min
Co-Cr-Mo + UHMWPE 1.8 mg/105 cycles NA •  Knee joint simulator Creep and fatigue [41]
(Duo-Patella) •  700 lb peak load cracks evident
•  Velocity = 33 cycles/min
•  Duration: 105 cycles
•  Lub: double-spun bovine
serum
Co-Cr-Mo + UHMWPE 1.1 mg/105 cycles NA •  Knee joint simulator Creep and fatigue [41]
(Ewald) •  700 lb peak load cracks evident
•  Velocity = 33 cycles/min
•  Duration: 105 cycles
•  Lub: double-spun bovine
serum
M. LaBerge and J.D. Desjardins
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Co-Cr-Mo + UHMWPE 0.3 mg/105 NA •  Knee joint simulator Creep and fatigue [41]
(Spherocentric) •  700 lb peak load cracks evident
•  Velocity = 33 cycles/min
8 Wear

•  Duration: 105 cycles

•  Lub: double-spun bovine
serum
Co–Cr–Mo + UHMWPE 0.4 mg/105 cycles NA •  Knee joint simulator Creep and fatigue [41]
(Geomedic) •  700 lb peak load cracks evident
•  Velocity = 33 cycles/min
•  Duration: 105 cycles
•  Lub: double-spun bovine
serum
Co-Cr-Mo + UHMWPE 0.3 mg/105 cycles NA •  Knee joint simulator Creep and fatigue [41]
(Geometric) •  700 lb peak load cracks evident
•  Velocity = 33 cycles/min
•  Duration = 105 cycles
•  Lub: double-spun bovine
serum
Ti–6A1–4V + UHMWPE 0.47 mg/106 cycles NA •  Pin-on-flat Surface scratching [87]
•  Axial load = 223 N
•  Duration: 250,000 cycles
•  Lub: bovine serum
Ti–6A1–4V + UHMWPE 0.3 mm/year NA •  Pin-on-disk Abrasion by cement [40]
•  Lub: bovine serum particles
•  Duration: 2 years test
•  Load = 3.45 MPa
•  Velocity = 50 × 106 mm/year
•  28–32 °C
(continued)
477
 478

Table 8.7 (continued)
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Ti–6A1–4V PE thickness change: NA •  Reciprocating pin-on-flat Adhesive transfer of 95]
Pins + UHMWPE GUR 59 ± 12  μm •  Sterilized with 2.5 Mrad PE on cylinders.
415 plate •  Lub: deionized water Oxidative wear of
•  36 MPa Line contact stress Ti-6A1-4V
•  Frequency = 2.1 Hz Abrasive wear of PE
•  Stroke length = 15 mm
•  Duration = 2 × 106 cycles
•  Final friction = 0.112 ± 0.007
Ti-6A1-4V 0.9 mm/year NA •  Pin-on-disk No scratches present [40]
(nitrided) + UHMWPE •  Lub: bovine serum
•  Duration: 2 years test
•  Load = 3.45 MPa
•  Velocity = 50 × 106 mm/year
•  28–32 °C
Ti–6A1–4V Machined UHMWPE: NA •  Annular disk-on-flat pin UHMWPE transfer [45]
(passivated) + UHMWPE 0.84–2.07 in/in × 10−9 •  Range of motion = 110° film
(HiFax 1900) •  Sliding velocity = 43.3 in./
min
•  Stress = 500 psi
•  Lub: Ringer’s solution
Ti-6A1-4V Machined UHMWPE: NA •  Annular disk-on-flat pin UHMWPE transfer [45]
(nitrided) + UHMWPE 1.83 in/in × 10−9 •  Range of motion = 110° film
(HiFax 1900) •  Sliding velocity = 43.3 in./
min
•  Stress = 500 psi
•  Lub: Ringer’s solution
M. LaBerge and J.D. Desjardins
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Ti-6A1-4V (not Machined UHMWPE: NA •  Annular disk-on-flat pin UHMWPE transfer [45]
passivated) + UHMWPE 1.55 in/in × 10−9 •  Range of motion = 110° film
(HiFax 1900) •  Sliding velocity = 43.3 in./
8 Wear

min
•  Stress: 500 psi
•  Lub: Ringer’s solution
Ti-6A1-4V + UHMWPE 0.04–0.11 mm3/106 NA •  Reciprocating flat-on-flat Surface scratching [38]
•  Load = 445 N
•  Velocity = 100 cycles/min
•  Lub: bovine serum
•  Duration: 4.1 × 106 cycles
Ti-6Al-4V (implanted with NA 1.9 × 10−7  mm3/Nm •  Ball and socket simulator Uniform superficial [44]
nitrogen) + UHMWPE •  Ra < 0.016 scratches, occasional
•  Sterilized w/2.5 Mrad g rad deeper marks
•  Lub: bovine serum w/ Metal particles,
sodium azide acrylic cement
•  Load = 2000 N particles
•  Speed = 100 mm/s
Ti-6A1-4V NA 1.98 × 10−7  mm3/Nm •  Ball-and-socket simulator Uniform superficial [44]
(conventional) + UHMWPE •  Ra < 0.016 scratches, occasional
•  Sterilized w/2.5 Mrad g rad deeper marks
•  Lub: bovine serum w/ Metal particles,
sodium azide acrylic cement
•  Load = 2000 N particles
•  Speed = 100 mm/s
Alumina + UHMWPE NA 18.2 × 10−7  mm3/Nm •  Pin-on-disk Original machine [37]
•  Lub: bovine serum, marks worn,
40–50 ml smoother
•  Load = 3 M Pa
•  Velocity = 60 mm/s
•  24–26 °C
(continued)
479
 480

Table 8.7 (continued)
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Alumina + UHMWPE 0.26 mg/106 cycles NA •  Pin-on-flat NA [38]
•  Axial load = 223 N
•  Duration: 250,000 cycles
•  Lub: bovine serum
Alumina + UHMWPE 0.04 mg/106 cycles NA •  Pin-on-flat Minimal wear [42]
•  Lub: distilled, deionized
water
•  Contact pressure: 4.8 MPa
•  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  37.1 °C
Alumina + UHMWPE Alumina: 0.1 mg/20 h NA •  Rolling-sliding wear and NA [46]
UHMWPE: 0.1 mg/20 h friction apparatus
•  Velocity = 25 m/min
•  Pressure = 30 N/cm2
•  Duration: 20 h
•  Dry condition
•  Room temperature
Zirconia + UHMWPE 0.03 mg/106 cycles NA •  Pin on flat Minimal wear [42]
•  Lub: distilled, deionized
water
•  Contact pressure: 4.8 MPa
•  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  37.1 °C
M. LaBerge and J.D. Desjardins
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Solid yttria Zr02 PE thickness change: NA •  Reciprocating pin-on-flat Abrasive wear from [95]
pins + UHMWPE GUR 33 ± 13  μm •  Sterilized with 2.5 Mrad the surface roughness
415 plate •  Lub: deionized water characteristic and
8 Wear

•  36 MPa Line contact stress adhesive wear

•  Frequency = 2.1 Hz
•  Stroke length = 15 mm
•  Duration = 2 × 106 cycles
•  Final friction = 0.033 ± 0.005
ZrO2 surface on Zr-2.5Nb Pe thickness change: NA •  Reciprocating pin-on-flat Abrasive wear from [95]
pins + UHMWPE GUR 25 ± 20  μm •  Sterilized with 2.5 Mrad the surface roughness
415 plate •  Lub: deionized water characteristics and
•  36 MPa Line contact stress adhesive wear
•  Frequency = 2.1 Hz
•  Stroke length = 15 mm
•  Duration = 2 × 106 cycles
•  Final friction = 0.040 ± 0.008
Zirconia NA 10.7 × 10−7  mm3/Nm •  Pin-on-disk Original machine [37]
(Y-PSZ) + UHMWPE •  Lub: bovine serum, marks still visible
40–50 ml
•  Load = 3 MPa
•  Velocity = 60 mm/s
•  24–26 °C
Co–Cr–Mo + Co-Cr-Mo 2–10 × 10−9  mm/mm NA •  Disc-on-plate NA [51]
•  Lub: water
•  37 °C
Co-Cr-Mo Roller: 23 mg/20 h NA •  Rolling-sliding wear and NA [46]
(Protasul-2) + Co-Cr-Mo Slider: 23 mg/20 h friction apparatus
(Protasul-2) •  Velocity = 25 m/min
•  Pressure = 30 N/cm2
•  Duration: 20 h
•  Dry condition
•  Room temperature
481

(continued)
 482

Table 8.7 (continued)
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Co-Cr-Mo + Co-Cr-Mo Max. depth of wear: NA •  Dual-hip simulator Adhesive and [84]
(McKee-Farrar) 0.01 mm •  Velocity = 30 cycles/min abrasive types of
•  Load = 250 kg wear
•  Duration: 1000 h
•  Lub: serum
•  Room temperature
Co-Cr-Mo •  initial wear: 10–20 mm NA •  Stanmore Mk III hip Equal amount of wear [52]
(Protasul-2) + Co-Cr-Mo •  linear wear: simulator on both components
(Protasul-2) 2–4 mm/106 cycles •  37 mm diameter head Preferential wear
•  Frequency = 1/2 Hz direction, with
•  Load = 300–3500 N pronounced grooving
•  Duration: min. 2.5 × 106
movements/test
•  Lub: Ringer’s solution
w/30 % calf serum
Co–Cr-Mo + Co-Cr-Mo 0.013  mm/1.8 × 106  cycles NA •  Hip joint simulator Abrasion, scratching. [91]
(McKee-Farrar) •  2.5 kN static load Max cup wear depth
•  Lub: bovine serum
•  Duration: 1.8 × 106 cycles
•  Velocity = 30 cycles/min
Co-Cr-Mo (Protasul-­ •  I nitial wear: 10–20 mm NA •  Stanmore Mk III hip Equal amount of wear [56]
21WF) + Co-Cr-Mo •  Linear wear: simulator on both components
(Protasul-21WF) 2–4 mm/106 cycles •  28 and 32 mm heads Preferential wear
•  Frequency = 1/2 Hz direction, with
•  Load = 300–3500 N pronounced grooving
•  Duration: min. 2.5 × 106
movements/test
•  Lub: Ringer’s solution
w/30 % calf serum
M. LaBerge and J.D. Desjardins
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Alumina + alumina 1.2 × 10-7  mm/mm NA •  Disc-on-plate NA [46]
•  Lub: water
•  37 °C
8 Wear

Stainless steel (316 2.26 mm/year 3.4 × 10−5  mm3/Nm •  Pin-on-flat No creep. Particles [96]
S16) + PTFE •  Lub: bovine serum present
•  Ra = 0.0 mm
•  Load = 40 n/pin
•  Speed = 2 p rad/s
•  Sliding dist. = 0.24 m/s
Stainless steel + Delrin 500 31 mm/year NA •  Pin-on-flat NA [53]
•  Velocity = 100 mm/s
•  Load = 6.9 N/mm2
•  Lub: serum
•  Room temperature
•  4.8 years’ effective use
Stainless steel + Polyester 521 mm/year NA •  Pin-on-flat NA [53]
•  Velocity = 100 mm/s
•  Load = 6.9 N/mm2
•  Lub: serum
•  Room temperature
•  4.8 years’ effective use
Stainless steel + alumina (mg/cyc) SS: 176, 146, NA •  Five station hip simulator Corrosion of stainless [97]
(Protek) 212 (Helsinki) steel heads
A1203: 0.3, 2.1, 0.2 •  32 mm head
•  Frequency = 1.08 sH3
•  Load = 35 kN
•  Duration: 3 × 106 cycles
•  Lub: distilled, deionized
water
•  37 °C
(continued)
483
 484

Table 8.7 (continued)
Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Stainless steel + alumina (mg/cyc) SS: 50.9, 62.4, NA •  Five-station hip simulator Corrosion of stainless [97]
(Thackray) 46.0 (Helsinki) steel heads
A12O3: 0.8, 1.0, 0.7 •  22.2 mm head
•  Frequency = 1.08 sH3
•  Load = 35 kN
•  Duration: 3 × 106 cycles
•  Lub: distilled, deionized
water
•  37 °C
Co–Cr-Mo + alumina (mg/cyc) CoCrMo: 39.7, NA •  Five-station hip simulator NA [97]
(Link) 48.2. 94.0 (Helsinki)
A12O3: 1.5, 0.5, 0.0 •  32 mm head
•  Frequency = 1.08 sH3
•  Load = 35 kN
•  Duration: 3 × 106 cycles
•  Lub: distilled, deionized
water
•  37 °C
Co–Cr–Mo + alumina (mg/cyc) CoCrMo: 2.6, NA •  Five-station hip simulator NA [97]
(Howmedica) 4.7, 4.3 •  (Helsinski)
A12O3: 0.1, 0.0, 0.0 •  32 mm head
•  Frequency = 1.08 sH3
•  Load = 35 kN
•  Duration: 3 × 106 cycles
•  Lub: distilled, deionized
water
•  37 °C
M. LaBerge and J.D. Desjardins
 Material contact Wear rate Wear coefficient Tribological conditions Wear mechanisms References
Co–Cr-Mo + Delrin 550 37.5 mm/year NA •  Disk-on-flat NA [53]
•  Velocity = 106 mm/s
•  Load = 3.7 N/mm2
8 Wear

•  Lub: water
•  37 °C
Si3N4 + UHMWPE 0.27 mg/106 cycles NA •  Pin-on-flat NA [42]
•  Lub: distilled, deionized
water
•  Contact pressure: 4.8 MPa
•  Frequency = 1 Hz
•  Sliding dist. = 50 mm
•  37.1 °C
a
Lub lubricant
485
486 M. LaBerge and J.D. Desjardins

Baykal and collaborators have critically and systematically reviewed the litera-
ture of multi-axis pin-on-disc wear testing for the past decades with a focus on
UHMWPE tribological properties and relevant wear mechanisms. Data reviewed
have been extracted from 22 published studies and are presented as UHMWPE
average wear rate (mg/MC), average wear rate (mm3/MC), and average wear factor
(mm3/Nm) as well as contact area for bearing couples [77].

Linear Clinical Wear

Methods have been proposed by Griffith et al. [98] and Deavane et al. [99] to radio-
graphically measure linear wear in UHMWPE acetabular cups of total hip replace-
ments. The overall penetration of a femoral head (Tables 8.8 and 8.9) into a
UHMWPE acetabular cup is a consequence of both creep and wear, where creep is

Table 8.8  Clinical penetration into UHMWPE cups for metallic femoral heads
Reference UHMWPE penetration rate Clinical
[98] • 0.07 mm/year average 491 acetabular cups 8.3-year follow-up (range
(range 0.06–0.24 mm/year) 7–9 years)
[100] • 0.13 mm/year average (range Follow-up at least 9.5 years after implantation
0–0.39 mm/year) for 22 mm (227 of the 22 mm femoral head size; 98 of the
head size 28 mm size; and 60 of the 32 mm head size)
• 0.08 mm/year average (range
0–0.3 mm/year) for 28 mm
head size
• 0.10 mm/year average (range
0–0.32 mm/year) for 32 mm
head size
[101] • 0.21 mm/year (range 87 acetabular cups 9-year service life (range
0.005–0.6 mm/year) <1–17.5 years)
[102] • 0.022 mm/year (range Four explained acetabular cups with no
0.010–0.034 mm/year) apparent deterioration of the finish of the
femoral head 20 year mean service life (range
17–23 years)

Table 8.9  Clinical penetration into UHMWPE cups for alumina ceramic heads
Reference Femoral head Average penetration rate
[103] Alumina 0.098 mm/year (6-year follow-up; 28 mm cups)
[103] Alumina 0.072 mm/year
(g irradiated (6-year follow-up; 28 mm cups)
UHMWPE)
[104] Alumina 0.084 mm/year
[105] Alumina 0.080 mm/year (38 acetabular cups over a period of 6.7 years)
[106] Alumina 0.100 mm/year
8 Wear 487

predominantly observed for the first million loading cycles [19]. Atkinson et al.
[107] estimated that the total UHMWPE residual compression due to creep in total
hip replacements after x can be estimated by

Residual compression = 94 + 33 ( x − 1)µm

This relationship has served as a guide to the significance of UHMWPE creep
for total hip arthroplasty [19]. Another important issue in the tribological behavior
of polymers used as bearing surfaces is their viscoelastic-plastic character. For
example, due to this time-dependent stress–strain response, the measured hardness
varies continuously as a function of indentation time. Therefore, in certain condi-
tions the wear behavior of these materials cannot accurately be characterized in
terms of wear coefficient alone and is often defined in terms of wear factor and Pv
limit. The wear factor is described as K′ = V/PS = V/Pvt where V is the wear volume,
v the sliding velocity, P the normal load, and t the sliding time. The Pv limit or fac-
tor (which is equal to load times velocity) defines the onset of failure of polymeric
surfaces. Pv limits have to be specified in terms of a limiting load at a given sliding
velocity or limiting velocity at a given load. These values will depend on the test
conditions.

Conclusion

The wear and frictional properties of materials are dependent on tribological

conditions of the tribosystem. Their investigation involves many parameters such
as wear rate, wear mechanisms, transition between initial and steady-­state wear,
and generation and geometry of wear debris. The physical and mechanical prop-
erties of the materials, the environmental and operating conditions, and the
geometry of the wearing bodies are determining factors for these parameters.
Another important tribological attribute that should be reported along with wear
data is the coefficient of friction. The use of different lubricants combined with
the operating conditions (load, velocity) will result in different coefficients of
friction and consequently different lubrication mechanisms of the tribosystem.
The wear mechanisms and wear data are partly governed by the lubrication of the
tribosystem.
By definition, a tribosystem is a dynamic system that can potentially change
over time emphasizing the importance of wear monitoring during testing. It is of
utmost importance to fully define and report the tribological conditions during
testing. The confidence in the quality of the wear data is not only related to the
488 M. LaBerge and J.D. Desjardins

tribological conditions, material properties, material combinations, and experi-

mental method including device used for testing, but also to the number of tests
conducted and number of specimens evaluated, level of statistical significance,
and duration of the test. When these parameters are not reported, it is difficult to
compare wear data obtained with different protocols. This often leads to an incon-
clusive analysis and disagreements. In this respect, the standardization of testing
protocols for wear of biomaterials and devices as well as how they are reported
can significantly improve the wear literature. This literature review clearly indi-
cates that there is no convention for wear-testing systems as well as for reporting
wear data.

Additional Reading and Selected Review Papers

1. Clarke, I.C. and McKellop, H.A. (1986) Wear Testing in Handbook of

Biomaterials Evaluation (A.F. von Recum, Ed.) Macmillan Publishing Co.,
New York. pp. 114–130. A concise review of testing protocols used for wear
testing of cardiovascular, orthopaedic, and dental implants. Introduces some of
the technical parameters involved in the understanding and measurement of wear
performance.
2. Dumbleton, J.H. (1981) Tribology of natural and artificial joints. Elsevier,
New York. An excellent critical review of testing parameters and conditions for
wear and friction of orthopaedic materials.
3. Baykal, D., Siskey, R.S., Haider, H., Saikko, V., Ahlroos, T., Kurtz, S.M. (2014)
Advances in tribological testing of artificial joint biomaterials using multidirec-
tional pin-on-disk testers. Journal of the Mechanical Behavior of Biomedical
Materials, 31, 117–134.
4. Sedel, L. (2000). Evolution of alumina-on-alumina implants: a review. Clinical
orthopaedics and related research, 379, 48–54.
5. Lewis, G. (1997). Polyethylene wear in total hip and knee arthroplasties. Journal
of biomedical materials research, 38(1), 55–75.
6. Bracco, P., & Oral, E. (2011). Vitamin E-stabilized UHMWPE for total joint
implants: a review. Clinical Orthopaedics and Related Research®, 469(8),
2286–2293.
7. Brown, S. S., & Clarke, I. C. (2006). A review of lubrication conditions for
wear simulation in artificial hip replacements. Tribology transactions, 49(1),
72–78.
8. Goel, V. K., Panjabi, M. M., Patwardhan, A. G., Dooris, A. P., & Serhan, H.
(2006). Test protocols for evaluation of spinal implants. The Journal of Bone &
Joint Surgery, 88 (suppl 2), 103–109.
9. Hoeppner, D. W., & Chandrasekaran, V. (1994). Fretting in orthopaedic implants:
a review. Wear, 173(1), 189–197.
8 Wear 489

References

1. Jacobs JJ, Shanbhag A, Giant TT, Black J, Galante JO (1994) Wear debris in total joint
replacements. J Am Acad Orthop Surg 2:212–220
2. Mittlmeier T, Walter A (1987) The influence of prosthesis design on wear and loosening
phenomena. CRC Critical Reviews in Biocompatibility, 3(4): 319–419
3. Dowson D, Wright V (1981) An Introduction to the Bio-mechanics of Joints and Joint
Replacement. Dowson D. and Wright V. Editors. Mechanical Engineering Publications, Ltd,
London, England, pp. 49–60
4. Suh NP (1986) Tribophysics. Prentice-Hall, Inc., Englewood Cliffs, NJ, pp 3–11, 195–222
5. Jahanmir S, Suh NP (1977) Mechanics of subsurface void nucleation in delamination wear.
Wear 44:17–38
6. Rowe CN (1980) Lubricated Wear, in CRC Handbook of Lubrication. Theory and Practice of
Tribology Volume II. Booser, R.E. Editor. Boca Raton, FL:CRC Press; pp. 209–225
7. Waterhouse RB (1992) Fretting wear. Frict Lubr Wear Technol 18:242–256
8. Black J (1988) Orthopaedic biomaterials in research and practice. Churchill Livingstone, NY,
p 255
9. Stachowiak G, Batchelor AW (2014) Engineering tribology, 4th edn. Elsevier, Amsterdam,
852 pp
10. Unsworth A, Dowson D, Wright V (1971) Cracking joints: a bioengineering study of cavita-
tion in the metacarpophalangeal joint. Ann Rheum Dis 30:348–358
11. Williams JA (2005) Wear and wear particles—some fundamentals. Tribol Int 38:863–870
12. Archard JF (1980) Wear theory and mechanisms. In: Peterson MB, Winer WO (eds) Wear
control handbook. ASME, New York, pp 35–80
13. Hornbogen E (1975) The role of fracture toughness in the wear of metals. Wear 33:251–259
14. Zum Gahr K-H (1982) Abrasive Wear of ductile materials. Z Metallkd 73:267–276
15. Hailing JA (1975) A contribution to the theory of mechanical wear. Wear 34:239–249
16. Quinn TFJ (1983) A review of oxidational wear. Parts I and II. Tribol Int 16(257–271):
305–315
17. Quinn TFJ, Rowson D, Sullivan JL (1980) Application of the oxidational theory of mild wear
to the sliding wear of low-alloy steel. Wear 65:1–20
18. Wang A (2001) A unified theory of wear for ultra-high molecular weight polyethylene in
multi-directional sliding. Wear 248:38–47
19. Dowson D (1995) A comparative study of the performance of metallic and ceramic femoral
head components in total replacement hip joints. Wear 90:171–183
20. Dowson D, El-Hady Diab MM, Gillis BJ, Atkinson JR (1985) Influence of counterface
topography on the wear of ultra high molecular weight polyethylene under wet conditions. In:
LeeAm L-H (ed) Polymer wear and its control. Chem. soc. symposium series no. 287,
Chapter 12, pp. 171–187
21. Dumbleton JH (1981) Tribology of natural and artificial joints. Elsevier, New York,
pp 94–103, 110–148, 183–214
22. Gevaert MR, LaBerge M, Gordon JM, DesJardins JD (2005) The quantification of physiolog-
ically relevant cross shear wear phenomena on orthopaedic bearing materials using a novel
wear testing machine. J Tribol 127:740–749
23. Bragdon CR, O'Connor DO, Lowenstein JD, Jasty M, Syniuta WD (1996) The importance of
multi-directional motion on the wear of polyethylene. Proc Inst Mech Eng H 210:157–165
24. Wang A, Stark C, Dumbleton JH (1996) Mechanistic and morphological origins of UHMWPE
wear debris in total joint replacement prostheses. Proc Inst Mech Eng H 210:141–155
25. Sambasivan S, Fischer DA, Shen MC, Hsu SM (2004) Molecular orientation of ultrahigh
molecular weight polyethylene induced by various sliding motions. J Biomed Mater Res B
Appl Biomater 70:278–285
490 M. LaBerge and J.D. Desjardins

26. Bragdon CR, O'Connor DO, Lowenstein JD, Jasty M, Biggs SA, Harris WH (2001) A new
pin-on-disk wear testing method for simulating wear of polyethylene on cobalt-chrome alloy
in total hip arthroplasty. J Arthroplasty 16:658–665
27. Hua Z, Zhang H, Fana Y, Jin Z (2014) Development of a BiotriboPOD testing methodology
for the wear evaluation of orthopaedic biomaterials. RSC Adv 4:19987–19991
28. Joyce TJ, Vandelli C, Cartwright T, Unsworth A (2001) A comparison of the wear of cross-­
linked polyethylene against itself under reciprocating and multi-directional motion with dif-
ferent lubricants. Wear 250:206–211
29. Saikko V (1998) A multidirectional motion pin-on-disk wear test method for prosthetic joint
materials. J Biomed Mater Res 41:58–64
30. Saikko V (2014) In vitro wear simulation on the RandomPOD wear testing system as a
screening method for bearing materials intended for total knee arthroplasty. J Biomech
47(11):2774–2778
31. Toohey KS, Blanchet TA, Heckelman DD (2003) Effect of accelerated aging conditions on
resultant depth-dependent oxidation and wear resistance of UHMWPE joint replacement
bearing materials. Wear 255:1076–1084
32. Turell M, Wang A, Bellare A (2003) Quantification of the effect of cross-path motion on the
wear rate of ultra-high molecular weight polyethylene. Wear 255:1034–1039
33. Yao JQ, Blanchet TA, Murphy DJ, Laurent MP (2003) Effect of fluid absorption on the wear
resistance of UHMWPE orthopedic bearing surfaces. Wear 255:1113–1120
34. McGloughlin TM, Kavanagh AG (2000) Wear of ultra-high molecular weight polyethylene
(UHMWPE) in total knee prostheses: a review of key influences. Proc InstMech Eng H
214:349–359
35. Walker PS, Blunn GW, Lilley PA (1996) Wear testing of materials and surfaces for total knee
replacement. J Biomed Mater Res 33:159–175
36. Dowson DB, Jobbins B (1988) Design and development of a versatile hip joint simulator and
a preliminary assessment of wear and creep in Charnley total replacement hip joints. Eng
Med 17:111–117
37. Kumar P, Oka M, Ikeuchi K, Shimizu K, Yamamuro T, Okumura H, Kotoura Y (1991) Low
wear rate of UHMWPE against zirconia ceramic (Y-PSZ) in comparison to alumina ceramic
and SUS 316L alloy. J Biomed Mater Res 25:813–828
38. McKellop H, Clarke I, Markolf K, Amstutz H (1981) Friction and wear properties of poly-
mer, metal, and ceramic prosthetic joint materials evaluated on a multichannel screening
device. J Biomed Mater Res 15:619–653
39. DesJardins I, Aurora A, Tanner SL, Pace TB, Acampora KB, LaBerge M (2006) Increased
total knee arthroplasty ultra-high molecular weight polyethylene wear using a clinically rel-
evant hyaluronic acid simulator lubricant. Proc Inst Mech Eng H 220:609–623
40. McKellop HA, Clarke IC, Markolf KL, Amstutz HC (1978) Wear characteristics of UHMW
polyethylene: a method for accurately measuring extremely low wear rates. J Biomed Mater
Res 12:895–927
41. Rose RM, Ries MD, Paul IL, Crugnola AM, Ellis E (1984) On the true wear rate of ultrahigh
molecular weight polyethylene in the total knee prosthesis. J Biomed Mater Res 18:
207–224
42. Saikko V (1993) Wear and friction properties of prosthetic joint materials evaluated on a pin
on flat apparatus. Wear 166:169–178
43. Saikko V, Ahlroos T, Calonius O, Keränen J (2001) Wear simulation of total hip prostheses
with polyethylene against CoCr, alumina and diamond-like carbon. Biomaterials 22(12):
1507–1514
44. McKellop HA, Rostlund TV (1990) The wear behavior of ion-implanted Ti-6A1-4V against
UHMW polyethylene. J Biomed Mater Res 24:1413–1425
45. Miller DA, Ainsworth RD, Dumbleton JH, Page D, Miller EH, Shen C (1974) A comparative
evaluation of the wear of ultra-high-molecular weight polyethylene abraded by
Ti-6A1-4V. Wear 28:207–216
8 Wear 491

46. Semlitsch M, Lehmann M, Weber H, Doerre E, Willert HG (1977) New prospects for a pro-
longed functional life-span of artificial hip joints by using the material combination polyeth-
ylene/aluminium oxide ceramic/metal. J Biomed Mater Res 11:537–552
47. Desjardins JD, Burnikel B, LaBerge M (2008) UHMWPE wear against roughened oxidized
zirconium and CoCr femoral knee components during force-controlled simulation. Wear
264:245–256
48. Medley JB, Krygier JJ, Bobyn JD, Chan FW, Tanzer M (1995) Metal-metal bearing surfaces
in the hip: investigation of factors influencing wear. Trans Orthop Res Soc 20(2):765
49. Chan FW, Medley JB, Krygier JJ, Bobyn JD, Podgorsak GK, Tanzer M (1996) Wear perfor-
mance of cobalt–chromium metal–metal bearing surfaces for total hip arthroplasty. In: Trans.
42nd annual meeting of the orthopaedic research society, vol 21, no. 2, p. 464
50. Chan FW, Bobyn JD, Medley JB, Krygier JJ, Tanzer M (1999) Wear and lubrication of metal-­
on-­metal hip implants. Clin Orthop Relat Res 369:10–24
51. Galante J, Rostoker W (1973) Wear in total hip prostheses. Acta Orthop Scand Suppl
145:1–46
52. Streicher RM, Schon R, Semlitsch M (1990) Investigation of the tribological behaviour of
metal-on-metal combinations for artificial hip joints. Biomed Tech 35(5):3–7
53. Clarke IC (1982) Wear-screening and joint simulation studies vs. materials selection and
prosthesis design. CRC Crit Rev Biomed Eng 8:29–91
54. Scholes SC, Unsworth A (2009) Wear studies on the likely performance of CFR-PEEK/
CoCrMo for use as artificial joint bearing materials. J Mater Sci Mater Med 20(1):163–170
55. Scholes SC, Unsworth A (2010) The wear performance of PEEK-OPTIMA based self-­mating
couples. Wear 268:380–387
56. Boutin P (1972) Arthroplastie total de la hanche par prothèse en alumina frittée. Rev Chir
Orthop 58:229
57. Nevelos J, Ingham E, Doyle C, Streicher R, Nevelos A, Walter W, Fisher J (2000)
Microseparation of the centers of alumina-alumina artificial hip joints during simulator test-
ing produces clinically relevant wear rates and patterns. J Arthroplasty 15:793–795
58. Oonishi H, Clarke IC, Good V, Amino H, Ueno M (2004) Alumina hip joints characterized
by run-in wear and steady-state wear to 14 million cycles in hip-simulator model. J Biomed
Mater Res A 70A:523–532
59. Wroblewski BM, Siney PD, Dowson D, Collins SN (1996) Prospective clinical and joint
simulator studies of a new total hip arthroplasty using alumina ceramic heads and cross-­
linked polyethylene cups. J Bone Joint Surg 78(2):280–285, British volume
60. Jin ZM, Dowson D, Fisher J (1993) Wear and friction of medical grade polyurethane sliding
on smooth metal counterfaces. Wear 162–164:627–630
61 Auger DD, Dowson D, Fisher J, Jin MZ (1993) Friction and lubrication in cushion form bear-
ings for artificial hip joints. Proc Inst Mech Eng 207:25–33
62. Auger DD, Dowson D, Fisher J (1995) Cushion form bearings for total knee joint replace-
ment. Part 1: Design, friction, and lubrication. Proc Inst Mech Eng 209:73–81
63. Caravia L, Dowson D, Fisher J (1993) Start up and steady state friction of thin polyurethane
layers. Wear 160:191–197
64. Chow AB, Medley JB, LaBerge M (1994) Mechanical and tribological analyses of elasto-
meric surface layers in load bearing implants. In: Trans of 20th annual meeting of the society
for biomaterials, Boston, MA, p. 434
65. Graham RM, Joseph PF, Dooley RL, LaBerge M (1995) Effect of lubricant viscosity and
bearing surface stiffness on the lubrication mechanism of a point contact as a total hip arthro-
plasty model. Trans Soc Biomater 23:124
66. LaBerge M, Kirkley S, Chow A, Medley JB, Brox WT (1991) Biotribological study of the
contact elastomer-cartilage. In: Proceedings of the combined meeting of orthopaedic research
societies (USA, Japan, Canada), Banff, Alberta, Canada, June 1991
67. Pylios T, Shepherd DET (2008) Wear of medical grade silicone rubber against titanium and
ultrahigh molecular weight polyethylene. J Biomed Mater Res Part B Appl Biomater
84B:520–523
492 M. LaBerge and J.D. Desjardins

68. Teoh SH, Tang ZG, Ramakrishna S (1999) Development of thin elastomeric composite mem-
branes for biomedical applications. J Mater Sci Mater Med 10:343–352
69. Suzuki S, Suzuki SH, Cox CF (1996) Evaluating the antagonistic wear of restorative
­materials. J Am Dent Assoc 127:74–80
70. Deng M, Shalaby SW (1997) Properties of self-reinforced ultra-high-molecular-weight poly-
ethylene composites. Biomaterials 18:645–655
71. Roeder RK, Sproul MM, Turner CH (2003) Hydroxyapatite whiskers provide improved
mechanical properties in reinforced polymer composites. J Biomed Mater Res A
67A:801–812
72. Wang M, Chandrasekaran M, Bonfield W (2002) Friction and wear of hydroxyapatite rein-
forced high density polyethylene against the stainless steel counterface. J Mater Sci Mater
Med 13:607–611
73. Harsha AP, Joyce TJ (2013) Comparative wear tests of ultra-high molecular weight polyeth-
ylene and cross-linked polyethylene. Proc IMechE Part H J Eng Med 227:600–608
74. Kyomoto M, Moro T, Miyaji F, Hashimoto M, Kawaguchi H, Takatori Y, Nakamura K,
Ishihara K (2009) Effects of mobility/immobility of surface modification by
2-­methacryloyloxyethyl phosphorylcholine polymer on the durability of polyethylene for
artificial joints. J Biomed Mater Res A 90A:362–371
75. Oral E, Christensen SD, Malhi AS, Wannomae KK, Muratoglu OK (2006) Wear resistance
and mechanical properties of highly cross-linked, ultrahigh-molecular weight polyethylene
doped with vitamin E. J Arthroplasty 21:580–591
76. Agrawal CM, Micallef DM, Wirth MA, Lankford J, Dearnaley G, McCabe AR (1993) The
effects of diamond-like-carbon coatings on the friction and wear of enhanced UHMWPE-­
metal couples. In: Trans 21st annual meet of the society for biomaterials, Birmingham, AL,
vol 26, p. 10
77. Baykal D, Siskey RS, Haider H, Saikko V, Ahlroos T, Kurtz SM (2014) Advances in tribo-
logical testing of artificial joint biomaterials using multidirectional pin-on-disk testers.
J Mech Behav Biomed Mater 31:117–134
78. Brown KJ, Atkinson JR, Dowson D, Wright V (1976) The wear of ultra high molecular
weight polyethylene and a preliminary study of its relation to the in vivo behaviour of replace-
ment hip joints. Wear 40:255–264
79. Fisher J, Dowson D (1991) Tribology of total artificial joints. Proc Inst Mech Eng
205:73–79
80. Rose RM, Radin EL (1982) Wear of polyethylene in total hip prostheses. Clin Orthop Relat
Res 170:107–115
81. Schmidt MB, Lin M, Greer KW (1995) Wear performance of UHWMPE articulated against
ion implanted CoCr. In: Trans 21st annual meeting of the society for biomaterials, San
Francisco, CA, vol 28, p. 230
82. Streicher RM (1991) Ceramic surfaces as wear partners for polyethylene. In: Bonfiled W,
Hastings GW, Tanner KE (eds) Bioceramics, vol 4. Butterworth-Heinemann Ltd, London,
England, p 9
83. Gavrjushenko NS (1993) Recommendations with respect to the improvement of lubrication
quality of synovial fluid in artificial joints. Proc Inst Mech Eng 207:111–114
84. Simon SR Radin EL (1973) Lubrication and wear of the Charnley, Charnley-Muller, and
McKee-Farrar prostheses with special regard to their clinical behavior. In: Proc. 1st scient.
meet. hip society, Saint Louis, Chapter 5, pp. 33–45
85. Saikko V (1992) Simulator study of friction in total replacement hip joints. Proc Inst Mech
Eng H 206:201–211
86. McKellop H (1981) Wear of artificial joint materials. II. Twelve-channel wear-screening
device: correlation of experimental and clinical results. Eng Med 10(3):123–136
87. Tateishi T, Terui A, Yunoki H (1990) Friction and wear properties for biomaterials for artifi-
cial joint. Bioceramics 2:145–151
8 Wear 493

88. Ruger LM, Love BJ, Drews MJ, Hutton WC, LaBerge M (1996) Effect of antioxidant on the
tribological properties of gamma sterilized ultra high molecular weight polyethylene. In:
Proc. fifth world biomaterials congress, Toronto, Canada, 29 May–2 June
89. Unsworth A, Pearcy MJ, White EFT, White G (1988) Frictional properties of artificial hip
joints. Eng Med 17:101–104
90. Walker PS, Bullough PG (1973) The effects of friction and wear in artificial joints. Orthop
Clin North Am 4:275–293
91. Weightman B, Simon S, Paul LI, Rose R, Radin EL (1972) Lubrication mechanisms of hip
joint replacement prostheses. J Lubr Technol (Trans Am Soc Mech Eng) 94:131
92. Weightman B, Paul IL, Rose RM, Simon SR, Radin EL (1973) A comparative study of total
hip replacement prostheses. J Biomech 6:299
93. Ungethum M, Refior HJ (1974) 1st Aluminiumoxidkeramik als Gleirlagerwekstoff fur
Totalendoprothesen geeignet? Arch Orthop Unfallchir 79:97
94. Eyerer P, Kurth M, McKellop HA, Mittlmeier T (1987) Characterization of UHMWPE Hip
cups run on joint simulators. J Biomed Mater Res 21:275–291
95. Poggie RA, Wert JJ, Mishra AK, Davidson JA (1992) Friction and wear characterization of
UHMWPE in reciprocating sliding contact with Co–Cr, Ti-6AMV, and zirconia implant
­bearing surfaces. In: Denton R, Keshavan MK (eds) Wear and friction of elastomers. ASTM
STP 1145. American Society For Testing and Materials, Philadelphia, pp 65–81
96. Dowson D, Wallbridge NC (1985) Laboratory wear tests and clinical observations of the
penetration of femoral heads into acetabular cups in total replacement hip joints. I: Charnley
prosthesis with polytetrafluoroethylene acetabular cups. Wear 104:203–215
97. Saikko V, Paavolainen P, Kleimola M, Slatis P (1992) A five-station hip joint simulator for
wear rate studies. Proc Inst Mech Eng H 206:195–200
98. Griffith MJ, Seidenstein MK, Williams D, Charnley J (1978) Socket wear in Charnley low
friction arthroplasty of the hip. Clin Orthop Relat Res 137:36–47
99. Deavane PA, Bourne RB, Rorabeck CH, Hardie RM, Home JG (1995) Measurement of poly-
ethylene wear in metal-backed acetabular cups. Clin Orthop Relat Res 319:303–316
100. Livermore J, Ilstrup D, Morrey B (1990) Effect of femoral head size on wear of the polyeth-
ylene acetabular component. J Bone Joint Surg 72A:518–528
101. Isaac GI, Wroblewski PJ, Atkinson JR, Dowson D (1992) Tribological study of retrieved hip
prostheses. Clin Orthop Relat Res 276:115–125
102. Wroblewski BM, McCullagh PJ, Siney PD (1992) Quality of the surface finish of the head of
the femoral component and the wear of the socket in long-term results of the Charnley low-­
friction arthroplasty. Eng Med 20:181–183
103. Oonishi H, Takayama Y (1989) Comparisons of wear of UHMW polyethylene sliding against
metal and alumina in total hip prostheses. Bioceramics 1:272–277
104. Okumara H, Yamamuro P, Kumar T, Nakamura T, Oka M (1989) Socket wear in total hip
prosthesis with alumina ceramic head. Bioceramics 1:284–189
105. Egli A, Weber BG, Sieber H, Semlitsch M, Dörre E (1990) Experience with the pairing of
polyethylene/ceramic materials in hip endoprostheses. In: Willert H-G, Bucchorn GH, Eyerer
P (eds) Ultra-high molecular weight polyethylene as biomaterial in orthopaedic surgery.
Hoffrefe and Huber, Gottingen, pp 154–158
106. Zichner LP, Willert HG (1992) Comparison of alumina-polyethylene and metal-polyethylene
in clinical trials. Clin Orthop Relat Res 282:86–94
107. Atkinson JR, Dowling JM, Cicek RZ (1980) Materials for internal prostheses: the present
position and possible future developments. Biomaterials 1:89–96
Chapter 9
Degradation/resorption in Bioactive
Ceramics in Orthopaedics

H. Oonishi and H. Oomamiuda

9.1   Introduction

Bioceramics have now been widely used as bone replacement materials in orthopaedic
surgery. In particular, calcium phosphate ceramics have been applied as bioactive
ceramics with bone bonding capacities.
Biological responses such as bone bonding and the biodegradation properties of
these materials are very important in clinical applications. Any convincing conclu-
sion has not yet been reached as to whether these materials are biodegradable or not,
although it has been discussed for a long time.
Degradation is an important characteristic for biomaterials, and it is considered
to have a large influence on the bone bonding properties. This degradation charac-
teristic must be considered from the following two view points. These are the solu-
tion mediated dissolution process and the cellmediated process (phagocytosis).
This chapter overviews the literature regarding the biodegradation processes of
bioactive calcium phosphate ceramics from the viewpoint of in vitro physico-­
chemical dissolution processes and in vivo/in vitro biological degradation
processes.

H. Oonishi (*) • H. Oomamiuda

Department of Orthopaedic Surgery, Artificial Joint Section and Biomaterial Research
Laboratory Osaka-Minami National Hospital,
677–2 Kido-Cho, Kawachinagano-shi, Osaka, Japan

© Springer Science+Business Media New York 2016 495

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_25
496 H. Oonishi and H. Oomamiuda

DCPD ; Dicalcium phosphate dihydrate [CaHPO4.2H2O]

DCPA ; Dicalcium phosphate anhydrate [CaHPO4]
OCP ; Octacalcium phosphate [Ca8H2 (PO4)6.5H2O]
TCP ; Tricalcium phosphate [Ca3 (PO4)2]
HAp ; Hydroxyapatite [Ca10 (PO4)6 (OH)2]
TTCP ; Tetracalcium phosphate [Ca4 (PO4)2O]

In this chapter, each calcium phosphate material is abbreviated as follows:

9.2  In Vitro Physico-chemical Dissolution Processes

Dissolution of a solid material continues until an equilibrium condition is reached,

followed by a saturated condition where a solid and a liquid remain in equilibrium.
Solubility is defined as the maximum concentration of solute in the solution under
the equilibrium condition. This concept of solubility is very convenient to know
how much the material has actually dissolved. However, it is not convenient for
showing general solubility of the material since its value depends on the pH of the
solution.
Therefore, a thermodynamic equilibrium constant known as the solubility prod-
uct constant Ksp is used for slightly soluble salts. This solubility product constant is
useful for understanding the dissolution characteristics, because its value does not
change in either acid or basic solutions under the same conditions of temperature,
pressure and ionic strength.
For example, HAp dissolves in water as follows;

{Ca ( PO ) ( OH ) }
10 4 6 2 solid
 10Ca 2 + + 6 PO34− + 2OH −

Therefore, its solubility product constant Ksp is calculated as follows;
10 6 2
K SP ( HA ) = Ca 2 +   PO34−  OH − 
p

where values in brackets represent ionic activities.
There are many reports on solubility product constants of calcium phosphate
compounds obtained by the above equation, and these values are shown in Table 9.1.
Chow [18] calculated, based on these data, the solubility isotherms at 37 °C over
a wide pH range (Figure 9.1). These solubility isotherms were shown as the func-
tion of the concentration of calcium and phosphate ions in a saturated solution of
each calcium phosphate salt. The relative stability of calcium phosphate salts at
various values of pH can be obtained by these solubility isotherms. At a given pH,
a salt with its isotherm lying below that of another salt is less soluble and more
9  Degradation/resorption in Bioactive Ceramics in Orthopaedics 497

Table 9.1  Solubility product constants of calcium phosphate compounds at 25°C

Compound Abbreviation Formula p Ksp Reference
Dicalcium phosphate DCPD CaHPO4. 2H2O 6.59 1,2
dihydrate
6.63 3
Dicalcium phosphate DCPA CaHPO4 6.90 2, 4
anhydrous
Octacalcium phosphate OCP Ca8H2 96.6 5
(PO4)6. 5H2O 98.6 6
α-Tricalcium phosphate α-TCP α-Ca3(PO4)2 25.5 7
β-Tricalcium phosphate β-TCP β-Ca3(PO4)2 28.9 2, 8
29.7 3
Hydroxyapatite HAp Ca10(PO4)6(OH)2 115.5 2, 9, 10
115–117 11
117 12–15
122 16
125 17
Tetracalcium phosphate TTCP Ca4(PO4)2O 38 15
pKsp: the negative iogarithm of Ks.

stable than the other. Therefore, HAp is the most stable and least soluble salt among
these salts in the range of pH below approximately 4.2 where DCPA is the least
soluble. Similarly, TTCP is the least stable and most soluble salt in the range of pH
below 8.5, above which pH DCPD is the most soluble.
Newesely [19] and Monma and Kanazawa [20] reported that α-TCP was con-
verted to HAp by hydration as follows;

3Ca 3 ( PO4 )2 + Ca 2 + + 2OH − → Ca10 ( PO4 )6 ( OH )2

However, this phase diagram can apply only in a thermodynamic equilibrium for the
ternary system of Ca(OH)2–H3PO4–H2O, and it shows only a general tendency
under in vivo conditions where various different ions are involved. Furthermore, this
relation is also considered to be influenced by the experimental conditions and the
characteristics of the material used in the experiments.
Most of the implanted materials used in the investigations were TCP and HAp. It
has been reported in earlier investigations that TCP dissolves more rapidly than
HAp in various solutions [21].
Jarcho [22] compared the relative dissolution rates of dense HAp and TCP. The
dissolution rate of TCP was 12.3 times higher than that of HAp in buffered lactic
acid solution (0.4 M, pH 5.2), and was 22.3 times higher than that of HAp in buff-
ered EDTA solution (0.05 M, pH 8.2). Klein et al. [23] carried out dissolution tests
on HAp and β-TCP with various values of porosity in buffered lactic acid solution.
The dissolution rate of β-TCP was three times higher than that of HAp. And it was
498 H. Oonishi and H. Oomamiuda

Figure 9.1  Solubility phase diagram for the ternary system Ca(OH)2–H3PO4–H2O at 37°C [18].
9  Degradation/resorption in Bioactive Ceramics in Orthopaedics 499

concluded that the degradation rates of these materials were determined by neck
dissolution rate and neck geometry, the latter factor being dependent on the crystal-
lography and stoichiometry of the material and of the sintering conditions. Ducheyne
et al. [15] compared the dissolution rate and the precipitation rate of the following
six calcium phosphates in calcium and phosphate free solution with pH 7.3. The
dissolution rate increased in the following order:

HAp < CDAp < OHAp < β　 TCP < α　 TCP < TTCP

and the precipitation rate increased as follows:

β　 TCP < ( HAP and α　 TCP ) < CDAp

(CDAp; Ca-deficient HAp, OHAp; dehydroxylated HAp)

The precipitate that formed on CDAp was apatitic; on the other hand, the precipi-
tates on HAp and β-TCP had a much lower Ca/P. In TTCP, the precipitate was cal-
cium deficient carbonate containing hydroxyapatite. Niwa et al. [24] evaluated the
concentration of calcium and phosphate ions being slightly dissolved in saline solu-
tion from HAp sintered at the temperature range from 250 to 1250 °C. It was con-
cluded that the sintering temperature was closely related to the crystallinity and the
amount of dissolution of the material. Maximan et al. [25] compared the dissolution
rate of plasma spray coated amorphous HAp and poorly crystallized HAp by expo-
sure to Hank’s physiological solutions (pH 7.2 and 5.2). The poorly crystallized
HAp coating showed faster resorption, greater surface film precipitation and greater
chemical changes than amorphous HAp coating.

9.3  In Vivo/in Vitro Biological Degradation Processes

Biological degradation processes have been investigated either by animal experi-

ments in vivo and clinical applications or by cell cultures in vitro. These results were
obtained by observing the remaining implanted materials and the behavior of each
cell around the materials.

9.3.1  Animal experiments and clinical applications

Most of the implanted calcium phosphate materials that were used in animal experi-
ments and clinical applications were TCP, HAp and calcium phosphate glasses.
Table 9.2 shows major reports of these investigations.
Most of the reports on TCP have concluded that TCP is biodegradable although
there are some differences depending on the characteristics of the materials used.
Bhaskar et al. [26] concluded that this biodegradation of TCP was caused by the
Table 9.2  Biodegradation of calcium phosphate compounds in vivo.
500

Implanted material (sintering temp.,

porosity or density) Implantation Duration Biodegradation Reference
• TCP (plug) Tibia of rats 14 weeks Progressive decrease 26
• TCP (pellet) Tibia of rats 48 days 95% of implant was degradated 27
• β-TCP (porous plug) [36%] Femur of dogs 4 months Considerable resorption 28
• TCP and TTCP (porous cylinder) [45%] Tibia of dogs 10 months TCP; resorbed TTCP; unchanged 29
• TCP (block) Vertebrae cervicales of dogs 22 weeks Partially resorbed (implants were crushed) 30
• HAp and β-TCP (ten types of micro and Tibia of rabbits 9 months HAp: no resorption TCP: more or less 31
macro porous cylinder) degradable
• β-TCP (seven types of micro and macro Tibia of rabbi 12 months Microporous>macroporous porous 32,33
porous cylinder) cylinder) reduced by addition of Mg2+, F-
• HAp and TCP (porous block) [HAp: 1300 Femur of dogs 50 weeks HAp: no resorption TCP: 25–30% (in 22 34
°C – 56%, TCP: 1150 °C – 50%] weeks)
• Coral (porites) – HAp (porous block) Mandible of dogs 1 year 29% of implant 35
• Coral (goniopora) – HAp (porous block) Tibia of dogs 1 year No bioresorption 36
• Coral (porites and goniopora) – HAp Animal studies 12 months Minimal 37,38
(porous block) Clinical application 15 months No degradation
• Coral (goniopora) – HAp and TCP [36%] Animal experiments 1 year Coral-HAp: not apparent 39
Clinical application 4 years TCP: observed in many cases
• HAp (three types of dense block) Dorsal muscle and tibia of 6 months No degradation (no difference in three 40
[900°C – 97%, 1200 °C – 97 and 99.9%] rats types)
• HAp (porous blocks) [1300°C, 56%] Femur of dogs 3.5 years Negligible 41
• HAp (porous blocks) [900 and 1200°C, Tibia of dogs 2 years Slow bioresorption (900°C> 1200 0c) 42
86%]
• HAp (macro and micro porous blocks) Middle ear of rats 1 year Resorption by 15–20 μm during the first yr. 43,44
[macroporosity: 26%, microporosity: 5%]
H. Oonishi and H. Oomamiuda
Implanted material (sintering temp.,
porosity or density) Implantation Duration Biodegradation Reference
• DCPD, DCPA, OCP, α- and β-TCP, HAp, Femur of rabbits 12 weeks (α- and β-TCP) < (DCPD, DCPA, TTCP) < 45
amorphous HAp Bioglass (granule sizes of (amorphous HAp,OCP)
100–300 and 10 μm)
• MBCP (macroporous blocks) Animal experiments 18 weeks Initially fast (~1 month) 46
[40–50%] (60%HAp + 40% β – TCP) Clinical application 16 months
• Plasma sprayed HAp (cylindrical plug) Tibia of rabbits 3 months Loss of coating thickness 47
• Plasma sprayed HAp, TTCP, MWL Femur of rats 4 weeks 60%-HAp and TTCP: distinct bulk 48
(cylindrical rod) (crystallinity of HAp: 10, degradation, 10%-HAp: gradual surface
60, 95%) degradation 95%-HAp and MWL:
negligible
• Plasma sprayed HAp, F Ap, MWL Femur and humerus of 25 weeks HAp: considerable and progressive 49
(cylindrical plug) goats reduction, MWL: considerable reduction in
thickness, FAp: no degradation or
dissolution
• Bioglass (SiO2–P2O5–CaO–Na2O system) Femur of rats 6 weeks Silica-rich layer and Ca-P rich layer 51
formation
• A -W glass-­ceramics (MgO–CaO–SiO2– Tibia of rabbits 25 weeks Ca-P rich layer formation 53
P2O5–CaF2 system)
9  Degradation/resorption in Bioactive Ceramics in Orthopaedics
501
502 H. Oonishi and H. Oomamiuda

Table 9.3  Biodegradation of calcium phosphate compounds in vitro

Substrate (sintering
temp., porosity or
density) Strain Incubation Biodegradation Reference
HAp and β-TCP Human 7 days Human bone cell are 54
(granule, <50 μm) bone cell capable of ingesting
HAP and TCP granules
HAp (dense disc; three Rat bone 8 days Osteoclast-like cells 55
types of surface marrow are capable of
rugosities) [1130°C] cell resorbing HAp
HAp (dense disc; three Rat bone 7 days Osteoclast-like cells 56
types of porosity [dried marrow are capable of
at 200°C, unsintered] cell resorbing HAp
Plasma sprayed HAp Rat bone 18 days 15 & 43%: rapidly 58
(crystallinity; marrow degrated 69%;
15,43,69%) cell degradation rate
was reduced
Plasma sprayed HAp, Rat bone 2 weeks Different bone-bonding 59
FAp, TCP, TTCP, marrow and biodegradation
MWL cell properties

phagocytosis of the mesenchymal cells. Cameron et al. [28] stated that the ingestion
by giant cells did play a significant role in the degradation of TCP although passive
dissolution occurred. Klein et al [32, 33] stated that the micro-pores played an
important role in the biodegradation rate of TCP. The degradation of TCP started
mostly from the medulla by solution mediated disintegration processes, and fine
particles released were phagocytosed and removed by macrophages in the medulla
to the lymph nodes. Renooij et al. [35] reported that HAp was not affected by bio-
degradation processes, while TCP was subject to extensive bioresorption. Resorption
debris from TCP was found in mononuclear phagocytes and multinuclear osteo-
clastlike cells. Although multinuclear cells were occasionally seen near the surface
of HAP, cells carrying HAP debris were never observed. And it was supposed that
TCP was transformed into HAp in a physiologic environment.
Concerning the biodegradation of HAp, there are reports in the case of no degra-
dation, slow or partial degradation and for the degradable case. The differences in
these results are dependent on the experimental conditions such as the characteris-
tics of the materials, animal species, implanted sites and methods of observation.
Holmes et al. [35–39] carried out investigations using HAp which was derived
from marine coral and reported the results as follows. Significant biodegradation
occurred when implanted in load bearing sites such as mandibles, while minimal
biodegradation was observed in cortical bone of radius and no apparent evidence of
biodegradation was observed in cancellous bone of tibia. In clinical applications,
radiographic observations did not show any irrefutable evidence of biodegradation
and history of biopsies showed no conclusive evidence of biodegradation, while
osteoclasts were occasionally seen along the implant surface. In contrast to these
results, degradation of TCP appeared to occur by passive dissolution and osteoclas-
9  Degradation/resorption in Bioactive Ceramics in Orthopaedics 503

tic resorption, and in many cases it was radiographically observed in clinical trials,
especially where the implant was applied in a diaphyseal onlay fashion. Denissen
et al. [40] reportered no degradation of three different dense HAp varying in its
density. Similarly, Hoogendoorn et al. [41] reported through their long-term study
that porous HAp did not undergo biodegradation during 3.5 years of implantation,
while giant multinucleated cells were occasionally seen in pores near the bone and
ceramic surface.
On the other hand, Kurosawa et al. [42] observed the degradation of highly
porous HAp in their experiments, and concluded that this degradation was caused in
two ways; the mechanical collapse of the material and the ingestion of fine particles
released from the HAp surface by multinuclear giant cells. Similarly, Blitterswijk
et al. [43, 44] observed in their implantation experiments with dense and macropo-
rous HAp that the deposition of calcium, partially in the form of calcium phosphate,
was found on the implant surface, and the resorption of the implant occurred as the
result of phagocytosis by mono- and multi-nuclear cells. Oonishi et al. [45] com-
pared bioactivity for bone formation in several kinds of bioceramics. These materi-
als were divided into three groups; bioinert ceramics (alumina), surface bioactive
ceramics (HAp and Bioglass), and resorbable bioactive ceramics (DCPD, DCPA,
OCP, α-TCP, β-TCP, TTCP and amorphous HAp). In resorbable bioactive ceramics,
bioactivity or bioresorbability might increased in the following order:

(α − and β − TCP ) < ( DCPD, DCPA and TTCP )

< ( amorphous HAP and OCP )

Daculsi et al. [46] stated that the bioresorption of macroporous biphasic calcium
phosphate consisting of HAp and β-TCP was conducted by multinucleated cells
(osteoclastlike cells) and was related to the β-TCP content of this material. Bruijin
et al. [48] and Dhert et al. [49] compared the degradation of plasma spray coated
TTCP, MWL (magnesium whitlockite) and three types of HAp with various degrees
of crystallinity. It was revealed that both TTCP and semi-crystalline HAp under-
went distinct bulk degradation and amorphous HAp showed a gradual surface deg-
radation, while the degradation was negligible with the highly crystalline HAp and
MWL. Biodegradation appeared to be related to bone apposition, since more bone
seemed to be present on amorphous HAp and TTCP, as compared to highly crystal-
line HAp and MWL. The degrading surface of TTCP and amorphous HAp coatings
was most likely a dynamic zone in which dissolution and reprecipitation occurs.
This zone was therefore thought to be favourable for rapid bone formation and
bonding. At the interface between bone and MWL, a seam of unmineralized bone-­
like tissue was frequently seen, and a substantial amount of aluminum was detected
in the MWL coating and the unmineralized bone-like tissue, which might cause the
impaired mineralization.
Since the discovery of Bioglass by Hench et al. [50], various kinds of bioactive
glasses and glass ceramics have been developed and applied clinically. Hench et al.
[51] summarized their study on Bioglasses which were based on the SiO2–P2O5–
504 H. Oonishi and H. Oomamiuda

CaO–Na2O system. When a bioactive glass was immersed in an aqueous solution,

three general processes occurred; leaching, dissolution and precipitation. In these
reactions, hydrated silica was formed on the glass surface, resulting in a silica-rich
gel layer, and then a calcia-phosphate-rich layer was formed on or within the gel
layer. This layer was initially amorphous and later crystallized to a hydroxycarbon-
ate apatite structure to which bone could bond. Kokubo et al. [52] developed A-W
glass-ceramics which was based on the MgO–CaO–SiO2-P2O5–CaF2 system. In this
material, oxyfluorapatite [Ca10(PO4)6(O,F2)] and β-wollastonite [CaO · SiO2] both in
the form of rice grain-like particles were dispersed in an MgO-CaO-SiO2 glass
matrix. In their experiments [53] it was shown that a thin layer, rich in Ca and P, was
formed on the surface of this material. This Ca, P-rich layer was identified as a layer
of apatite, and this material was observed to be closely connected to the surrounding
bone through this apatite layer without any distinct boundary. The same type of apa-
tite layer was formed on the surface of this material exposed to the simulated body
fluid, and consisted of carbonate-containing hydroxyapatite of defective structure
and small crystallites. It was concluded that this apatite layer played an essential role
in forming the chemical bond of all bioactive materials which bonded to bone.

9.3.2  Cell cultures

To study biodegradation and interfacial bonding phenomena, in vitro cell culture

systems have been developed. Gregoire et al. [54] investigated the influence of cal-
cium phosphate on human bone cell activities and demonstrated that the isolated
human bone cells were capable of ingesting HAp and β-TCP granules. And the
capacity for ingesting a synthetic mineral component clearly suggested that bone
cells were able to participate in the degradation of calcium phosphates. Gomi et al.
[55] showed that osteoclasts are capable of resorbing sintered HAp in vitro and that
the fusion of osteoclast mononuclear precursors was influenced by substratum
rugosity. Similarly, Ogura et al. [56] demonstrated that osteoclast-like cells were
capable of resorbing unsintered calcium phosphate substrata in vitro. Bruijn et al.
[57–59] carried out a series of cell culture tests on various plasma sprayed calcium
phosphate compounds and reported the results as follows. Rat bone marrow cells
were cultured on plasma sprayed HAp. The cells formed a mineralized extracellular
matrix that exhibited several characteristics of bone tissue. Two distinctly different
interfacial structures were observed on HAp. An electron-dense layer which was
rich in glycosaminoglycans was regularly present. A collagen-free amorphous zone
was frequently seen interposed between the electron-dense layer and HAp. In cell
culture tests on three types of plasma sprayed HAp varying in degree of crystallin-
ity, an electron-dense layer was clearly visible on a stable, nondegrading crystalline
HAp and was frequently observed at the interface of semi-crystalline HAp. An
amorphous zone was regularly seen at degrading surfaces of semi-crystalline and
poorly crystallized HAp. It was concluded that the crystallinity of plasma sprayed
HAp was an important parameter which influenced the establishment of the bony
9  Degradation/resorption in Bioactive Ceramics in Orthopaedics 505

interface and might, as a result, have an effect on the bone formation rate and
­bonding strength between HAp and bone tissue. Similarly, rat bone marrow cells
were cultured on various plasma sprayed calcium phosphate coatings. Mineralized
extracellular matrix was formed on HAp, TCP and TTCP in 2 weeks, and was
formed on FAp (fluorapatite) in 8 weeks. It was only occasionally observed in some
area on MWL, which phenomenon might have been due to aluminium impurities in
the coating. It was concluded that plasma sprayed calcium phosphates would d­ isplay
different bone-bonding and biodegradation properties, depending on their chemical
composition and crystal structures.

9.4  Summary

No convincing conclusion has been reached as to the biodegradation mechanisms of

bioactive ceramics. Many researchers have reported different results, as described
above. These discrepancies are considered to be caused by the fact that materials
used for the experiments were different, and that experimental methods and
­analytical methods were also different. Therefore, when these reported results are
compared, it is important to consider the characteristics of the material used
­(chemical compositions, impurity, crystallinity, dense or porous, micro- or macro-
porous, porosity), experimental methods used (in vivo or in vitro, animal species,
implanted duration, implanted sites, load bearing or not), and analytical methods
used (radiographic, optical microscopic, electron microscopic). Futhermore, a good
understanding of the characteristics of the materials to be used becomes important
when bioactive ceramics are used clinically.

References

1. L.C. Chow and W.E. Brown, J. Dent. Res., 54, 65–16, 1975.

2. F.C.M. Driessens, J.W.E. van Dijk and J.M.P.M. Borggreven, Calcif. Tissue Res., 26, 127–136,
1978.
3. J.L. Meyer, and, E.D. Eanes, Calcif. Tissue Res., 25, 59–68, 1978.
4. H. McDowell, W.E. Brown and J.R. Sutter, Inorg. Chem., 10, 1638–1643, 1971.
5. M.S. Tung, N. Eidelman, B. Sieck and W.E. Brown, J. Res. Nat. Bur. Stand., 93, 613–624,
1988.
6. J.L. Shyu, L. Perez, S.J. Zawacki, J.C. Heughebaert and G.H. Nancollas, J. Dent. Res., 62,
398–400, 1983.
7. B.O. Fowler and S. Kuroda, Calcif. Tissue Int., 38, 197–208, 1986.
8. T.M. Gregory, E.C. Moreno, J.M. Patel and W.E. Brown, J. Res. Nat. Bur. Stand., 78(A), 667–
674, 1974.
9. E.C.Moreno, T.M. Gregory and W.E. Brown, J. Res. Nat. Bur. Stand., 72(A), 773–782, 1968.
10. J.S.  Clark, Can. J. Chem., 33, 1696–1700, 1955.
11. R.M.H. Verbeek, H. Steyaer, H.P. Thun and F. Verbeek, J. Chem. Soc. Faraday Trans. 1, 76,
209–219, 1980.
506 H. Oonishi and H. Oomamiuda

12. R.  Chuong, J. Dent. Res., 52, 911–914, 1973.

13. A.N. Smith, A.M. Posner and J.P. Quirk, J. Coll. Inter. Sci., 54, 176–183, 1976.
14. H.McDowell, T.M. Gregory and W.E. Brown, J. Res. Nat. Bur. Stand., 81(A), 273–281, 1977.
15. P. Ducheyne, S. Radin, L. King, K. Ishikawa and C.S. Kim, Bioceramics vol.4, eds W. Bonfield,
G. W. Hastings and K.E. Tanner, Butterworth-Heinemann, London, UK, pp. 135–144, 1991.
16. M.B. Fawzi, J.L. Fox, M.G. Dedhiya, W.I. Higuchi and J.J. Hefferren, J. Coll. Inter. Sci., 67,
304–311, 1978.
17. M.S. Wu, W.I. Higuchi, J.L. Fox and M. Friedman, J. Dent. Res., 55, 496–505, 1976
18. L.C.  Chow, J. Ceram. Soc. Japan, 99, 954–964, 1991.
19. H.  Newesely, J. Oral Rehab ., 4, 97–104, 1977.
20. H. Monma and T. Kanazawa, Yogyo-Kyokai-Shi, 84, 209–213, 1976.
21. R.W. Mooney and M.A. Aia, Chem. Rev.,61, 433–462, 1961.
22. M.  Jarcho, Cli Orthop. Rel. Res. 157, 259–278, 1981.
23. C.P.A.T. Klein, A. A. Driessen and K. de Groot, Biomaterials, 5, 157–160, 1984.
24. J. Niwa, S. Takahasi and M. Sohmiya, Fineceramics, 2, 25–32, 1981.
25. S.H. Maximan, J.P. Zawadsky and M.G. Dunn, J. Biomed. Mater. Res., 27, 111–117, 1993.
26. S.N. Bhaskar, J.M. Brady, L. Getter, M.F. Grower and T. Driskell, Oral Surg., 32, 336–346,
1971.
27. D.E. Cutright, S.N. Bhaskar, J.M. Brady, L. Getter and W.R. Posey, Oral Surg., 33, 850–856,
1972.
28. H.U. Cameron, I. Macnab and R.M. Pilliar, J. Biomed. Mater. Res., 11, 179–186, 1977.
29. K. Köster, H. Heide and R. König, Langenbecks Arch. Chir., 343, 173–181,
30. T. Shima, J.T. Keller, M.M. Alvira, F.H. Mayfield and S.B. Dunsker, J. Neurosurg., 51, 533–
538, 1979.
31. C.P.A.T. Klein, A.A. Driessen and K. de Groot, J. Biomed. Mater. Res., 17, 769–784, 1983.
32. C.P.A.T. Klein, K. de Groot, A.A. Driessen and H.B.M. van der Lubbe, Biomaterials, 6, 189–
192, 1985.
33. D.C.P.A.T. Klein, K. de Groot, A.A. Driessen and H.B.M. van der Lubbe, Biomaterials, 7,
144–146, 1986.
34. W. Renooij, H.A. Hoogendoorn, W.J. Visser, R.H.F. Lentferink, M.G.J. Schmitz, H.V. Ieperen,
S.J. Oldenburg, W.M. Janssen, L.M.A. Akkermans and P. Wittebol, Cli. Orthop. Rel. Res., 197,
272–285, 1985.
35. R.E.  Holmes, Plast. & Reconstr. Surg., 63, 626–633, 1979.
36. R.E. Holmes, R.W. Bucholz and V. Mooney, J. Bone and Joint Surg., 68-A, 904–911, 1986.
37. R.E. Holmes, V. Mooney, R. Bucholz and A. Tencer, Cli. Orthop. Rel. Res., 188, 252–262,
1984.
38. D.J.  Sartoris, D.H.  Gershuni, W.H. Akeson, R.E.  Holmes and D.  Resnick, Radiology, 159,
133–137, 1986.
39. R.W. Bucholz, A. Carlton and R.E. Holmes, Orthop. Cli. North America, 18, 323–334, 1987.
40. H.W. Denissen, K. de Groot, P.Ch. Makkes, A. van den Hooff and P.J. Klopper, J. Biomed.
Mater. Res., 14, 713–721, 1980.
41. H.A. Hoogendoorn, W. Renooij, L.M.A. Akkermans, W. Visser and P. Wittebol, Cli. Orthop.
Rel. Res., 187, 281–288, 1984.
42. H. Kurosawa, T. Iwata, I. Shibuya, K. Murase and Takeuchi, J. Joint Surg., 8, 1761–1769,
1989.
43. C.A. van Blitterswijk, J.J.  Grote, W.  Kuijpers, C.J.G.  Blok-van Hoek and W.  Th. Daems,
Biomaterials, 6, 243–251, 1985.
44. C.A. van Blitterswijk, J.J. Grote, W. Kuijpers, W. Th. Daems and K. de Groot, Biomaterials,
7, 137–143, 1986.
45. H. Oonishi, S. Kushitani, T. Sugihara, H. Iwaki, H. Ohashi and E. Tsuji, Bioceramics vol. 8,
eds L.L. Hench and J. Wilson, ELSEVIER, New York, 1995.
46. G. Daculsi, N. Passuti, S. Martin, C. Deudon, R.Z. Legeros and S. Raher, J. Biomed. Mater.
Res., 24, 379–396, 1990.
9  Degradation/resorption in Bioactive Ceramics in Orthopaedics 507

47. J.A. Jansen, J.P.C.M. van de Waerden, J.G.C. Wolke and K. de Groot, J. Biomed. Mater. Res.,
25, 973–989, 1991.
48. J.D. de Bruijn, Y.P. Bovell and C.A. van Blitterswijk, Calcium Phosphate Biomaterials: Bone-­
bonding and Biodegradation Properties, ed. J.D. de Bruijn, Offsetdrukkerij Haveka B.V.,
Alblasserdam, pp. 79–92, 1993.
49. W.J.A. Dhert, C.P.A.T. Klein, J.A. Jansen, E.A. van der Velde, R.C. Vriesde, P.M. Rozing and
K. de Groot, J. Biomed. Mater. Res., 27, 127–138, 1993.
50. L.L.  Hench, R.J.  Splinter, W.C. Allen and T.K.  Greenlee, J. Biomed. Mater. Res. Symp., 2,
117–141, 1971.
51. L.L. Hench and B. Andersson, Advanced Series in Ceramic – Vol. 1 An Introduction to
Bioceramics eds L.L. Hench and J. Wilson, World Scientific, Singapore, pp. 41–62, 1993.
52. T. Kokubo, S. Ito, S. Sasaki and T. Yamamuro, J. Mater. Sci., 21, 536–540, 1986.
53. T.  Kokubo, Advanced Series in Ceramics – Vol. 1 An Introduction to Bioceramics, eds
L.L. Hench and J. Wilson, World Scientific, Singapore, pp. 75–88, 1993.
54. M. Gregoire, I. Orly and J. Menanteau, J. Biomed. Mater. Res., 24, 165–177, 1990.
55. K. Gomi, B. Lowenberg, G. Shapiro and J.E. Davies, Biomaterials, 14, 91–96, 1993.
56. M. Ogura, T. Sakae and J.D. Davies, Bioceramics vol. 4, eds W. Bonfield, G.W. Hastings and
K.E. Tanner, Butterworth-Heinemann, London, UK, pp. 121–126, 1991.
57. J.D. de Bruijn, C.P.A.T. Klein, K. de Groot and C.A. van Blitterswijk, J. Biomed. Mater. Res.,
26, 1365–1382, 1992.
58. J.D. de Bruijn, J.S. Flachl, K. de Groot, C.A. van Blitterswijk and J.E. Davies, Calcium
Phosphate Biomaterials: Bone-bonding and Biodegradation Properties, ed. J.D. de Bruijn,
Offsetdrukkerij Haveka B.V., Alblasserdam, pp. 45–62, 1993.
59. J.D. de Bruijn, C.P.A.T. Klein, K. de Groot and C.A. van Blitterswijk, Calcium Phosphate
Biomaterials: Bone-bonding and Biodegradation Properties, ed. J.D. de Bruijn, Offsetdrukkerij
Haveka B.V., Alblasserdam, pp. 63–78, 1993.
Chapter 10
Corrosion of Metallic Implants

M.A. Barbosa

10.1  General Aspects

10.1.1  Incidence of corrosion

The surfaces of passive metals are normally attacked at specific points where the
oxide film has been destroyed and massive quantities of metal ions are released.
Depending on the magnification with which surfaces are observed, various degrees
of localized attack can be detected. Sometimes, however, corrosion may not be eas-
ily distinguishable from mechanical imperfections associated with manufacturing
or handling. Even under the scanning electron microscope (SEM) it is often difficult
to distinguish between mechanical indentations and pitting or crevice attack.
After determining the existence of corrosion, the next step is to assess its magni-
tude. This can be done, by corrosion scores, such as those given in Table 10.1
(Thomas et al., 1988). In this table the ‘no surface degradation’ score is obtained
with a magnification of 60x. If a higher magnification was used some ‘non-degraded’
surfaces might fall in one of the ‘surface degradation’ categories. The borderline
between ‘corroded’ and ‘non-corroded’) surfaces is therefore very much dependent
on magnification, as well as on surface preparation, as explained above. With these
limitations in mind, it is useful to have an idea of the incidence of corrosion, i.e. of
the percentage of implants that suffer some degree of attack. Table 10.2 compares
the data obtained by several authors. Vacuum melting (VM) significantly reduces
the susceptibility of 316L stainless steel to attack, while titanium is practically
immune. Corrosion, apart from affecting the mechanical performance of the
implants, also results in contamination of the tissues with metallic ions.
The detection of ions released from metallic implants is dependent on the tech-
nique used. Very minute amounts of ions can be detected by electrothermal atomic

M.A. Barbosa (*)

INEB-Ma, Rua do Campo Alegre 823, 4150 Porto, Portugal

© Springer Science+Business Media New York 2016 509

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_26
510 M.A. Barbosa

Table 10.1  Grading scale used to evaluate the degree of interface and surface corrosion (Thomas
et al., 1988)
0 = no surface degradation visible at 60x magnification
1 = very mild surface degradation visible at 60x magnification
2 = mild surface degradation visible at 60x magnification
3 = moderate surface degradation visible without magnification
4 = heavy surface degradation visible without magnification
5 = very heavy severe surface degradation visible without magnification

Table 10.2  Incidence of corrosion

Implant/material Corroded implants Ref
38 fixation devices/AISI 316L 95%; crevice corrosion Thomas et al., 88
43 miniplates/AISI 316LVM* 19%; only at the countersinks Torgersen and Gjerdet, 94
19 miniplates/c.p Ti 0% Torgersen and Gjerdet, 94
* VM – Vacuum melted

absorption spectroscopy (ET-AAS), which goes down to concentrations of the

order of ng/g. Electrochemical methods enable the detection of extremely low cor-
rosion current densities, below 1 μA/cm2, corresponding to dissolution in the pas-
sive state. These rates of corrosion do not modify the aspect of the surface and are
not normally considered as surface attack.

10.1.2  Potential-pH (Pourbaix) diagrams

These diagrams indicate the regions of immunity, passivation and corrosion of pure
metallic elements in pure water at 25°C. Fig.10.1 (a, b) gives the potential-pH dia-
grams of Cr and Ti. Cr is the element responsible for the passive behaviour of stain-
less steels and Co-Cr-Mo alloys. In Fig. 10.la passivation by a Cr(OHh film is
assumed. The film is thermodynamically stable over a wide range of pH and poten-
tial values. Below pH 4 the film is unstable and Cr is corroded. Ti is responsible for
the excellent corrosion resistance of Ti-Al-V alloys. In Fig. 10.lb passivation by a
hydrated TiO2 .H2O oxide has been assumed. The passivation region extends to
much higher potentials than in the case of Cr. The passive film is unstable below pH
2.5. In both diagrams the dotted lines give the region of stability of water: below
line a hydrogen is evolved, whereas above line b oxygen is released. Normally, the
corrosion potentials of implant materials do not reach such extreme values.
In spite of their usefulness in predicting the stability of metals and their oxides,
potential-pH diagrams suffer from a number of limitations. They refer to pure metals,
not to alloys, and to pure water, not to environments normally found in practical situa-
tions. For example, the diversity of chemical species, and particularly the presence of
chloride ions in physiological media, is responsible for substantial differences between
practical and predicted behaviour. Localized attack, in the form of crevice, pitting and
corrosion fatigue, is due to the presence of chloride. Furthermore, the kinetics of metal
10  Corrosion of Metallic Implants 511

Fig. 10.1 (a)  (b) Theoretical domains of corrosion, immunity and passivation of titanium, at
25°C, considering Tiltz (Pourbaix, 1971).
512 M.A. Barbosa

dissolution or passivation cannot be assessed by these diagrams, which are purely ther-
modynamic. However, if not misused, potential-pH diagrams can give useful informa-
tion which must be complemented by other type of date, namely of kinetic nature.

10.2  Aspects Related to the Metal Composition

10.2.1  Importance of materials purity in improving

the corrosion resistance

The evolution of stainless steel composition can be used to illustrate the importance
of materials purity in reducing corrosion susceptibility. Chromium and molybde-
num are the key elements in promoting resistance to pitting and crevice attack of
stainless steels, but high chromium and molybdenum concentrations are not suffi-
cient to ensure an adequate corrosion resistance. Low concentrations of impurities,
like carbon, silicon, phosphorous and sulphur, are required. Type 316L and 316LVM
stainless steels are commonly employed to fabricate a variety of fracture fixation
devices. They both have low carbon concentration, below 0.03 wt%, which is indi-
cated by the letter L. VM stands for vacuum-melted, a technique that enables the
production of metals with very low concentrations of impurities.
A retrieval analysis of Kuntscher intramedullary rods (Cook et al. 1990) has
shown that significant surface corrosion, inclusion content and carbon content
occurred on early materials, which had remained in situ for 10 years or longer (maxi-
mum 23 years). Significant relationships were obtained for surface corrosion score
vs. thin globular oxide inclusion content, and for surface corrosion score vs. sulphide
inclusion content. Fig. 10.2 shows the data obtained for the former correlation.

10.2.2.  Type of metallic material and influence of alloying

Due to the presence of a thin oxide film, titanium has a very high corrosion resis-
tance. However, its low resistance under wear conditions may lead to enormously
high titanium concentrations in tissues adjacent to titanium implants (section 9.3.1).
Rapid film formation after surface damage is therefore of critical importance to
guarantee low levels of titanium ions.
The current density (c.d.) required to form a passive film is called the critical c.d.
for passivation, ic. The lower ic the better. Fig. 10.3 shows that Zr, Nb, Ta and Pd
decrease ic, whereas Sn increases it. It has been found (Okazaki et al. 1994) that ic
can be related to the percentages of Pd, Ta, Nb, Zr and Sn by the following
expression
ic(A.m-2 = 10-2{98-89.5[%Pd] - 9.5[%Ta] - 3.4[%Nb]- 0.67[%Zr] + 8[%Sn]}
Fig. 10.2  Relationship between surface corrosion and thin globular oxide inclusion content.
Regression line y= 1.78x+0.52, r= 0.65, n= 18, p <0.05. (Cook et al., 1990.)

Fig. 10.3  Effects of Zr, Nb, Ta, Pd and Sn contents critical current density for passivation in 15%
H2S04 and 5% HCl solutions at 310 K. (Okazaki et al., 1994.)
514 M.A. Barbosa

A new alloy, Ti-15Zr-4Nb-2Ta-0.2Pd, with better corrosion resistance than the

conventional Ti-6AI-4V, was proposed by the same authors.
Replacement of Nb for V, in order to eliminate the possible toxic effects of the
latter, has been carried out (Semlitsch et al., 1992). Ti-6AI-7Nb showed a corrosion
resistance similar to that of Ti-6AI-4V, as concluded from anodic polarization
curves (Fig. 10.4).
Although not as widely used as titanium, tantalum has found a number of appli-
cations, e.g. in vascular clips, as a suture and to fabricate flexible stents to prevent
arterial collapse. The reader is referred to a paper by J. Black (1994), where the
material properties are reviewed, together with the host response and clinical appli-
cations. In terms of corrosion resistance, tantalum is at least equivalent to titanium.
Its oxide, Ta2O5, is very stable over very wide pH and potential ranges, thus explain-
ing the excellent corrosion resistance. According to Black, high cost and difficulties
of fabrication are some of the reasons for its limited usage.
Cr and Mo are the major alloying elements responsible for the corrosion resis-
tance of stainless steels. These alloys are also very sensitive to inclusion content,
which has led to continuous attempts to reduce impurity concentrations. F138 and
F139 are variations of the AISI 316L stainless steel with a lower content of non-­
metallic inclusions. A duplex stainless steel, 25Cr-7Ni-4Mo-0.25N, shows a better
corrosion resistance than conventional austenitic stainless steels (Cigada et al.,
1989). The authors have established the following ranking: 23Cr-4Ni<AISI
316L<ASTM F138<22Cr-SNi-3Mo<27Cr-31Ni-3.5Mo<25Cr-7Ni-4Mo-N.

Fig. 10.4  Current density/potential curves of five different implant materials in Ringer’s solution
bubbled through with nitrogen .■ AISI 316L; ●CoNiCrMo; ▲ Ti6AI-7Nb; ▼ CP-titanium; x
Ti-6AI-7Nb/ODH. (Semlitsch et al., 1992.)
10  Corrosion of Metallic Implants 515

10.2.3  Site for attack

Normally, pitting initiates at non-metallic inclusions. In stainless steels sulphides

are particularly prone to attack. Oxide inclusions may also give origin to attack, but
they are less active than sulphides. Carbides may also nucleate pitting attack and
when they are numerous at grain boundaries they may give rise to intergranular cor-
rosion. However, with surgical grades of stainless steels this type of attack should
not occur. Crevice corrosion is also common with stainless steels and less frequent
with Co-Cr alloys. Sintered beads of Co-Cr-Mo alloy have been studied by scan-
ning electrochemical microscopy (Gilbert et al., 1993a). At any time, some grains
were more active than others, whereas at later stages shifting of the active regions
occurred. Titanium is immune to both types of corrosion under static conditions.
Sliding between titanium and another material (e.g. cement, polyethylene or bone)
may originate severe degradation by corrosive wear.

10.2.4.  Combinations of different materials

The need to combine different materials may sometimes arise. An example is the
use of hard materials for the head of hip joints in combination with a titanium stem.
Titanium has a very high corrosion resistance, but a very poor wear resistance.
Therefore, either surface hardening treatments (e.g. ion implantation of nitrogen or
surface alloying) or a harder material, e.g. a ceramic, are employed for the femoral
head. Ceramics, like alumina or zirconia, do not cause enhanced electrochemical
dissolution of the titanium stem because of their low electronic conductivity.
However, when another metal (e.g. Co-Cr-Mo alloy) is used instead, the possibility
of a galvanic couple between the stem and the head being formed exists. The situa-
tion illustrated by this example can be extended to other couples, including those
involving carbon. Even in the case of hard coatings galvanic couples between the
coating and the substrate may form.
In a first approximation, the safety of couples involving different materials can
be preditected by a number of experimental techniques. Table 10.3 summarizes the
data obtained by several authors. Notice that the couples between stainless steel and
other materials is unsafe. On the contrary, TiAlV/CoCrMo, CoCrMo/C and
TiAIV/C combinations may be considered safe. However, repeated fracture of the
oxide film at the conical taper region between head and stem of Ti6Al4V/CoCr
combinations has been associated with corrosion. Attack also occurred in CoCr/
CoCr combinations and was proportional to the duration of implantation, as seen in
Fig. 10.5 (Gilbert et al., 1993b). A larger percentage (34.5%) of cases of ­corrosion
was found with mixed CoCrffiAlV systems than with CoCr/CoCr systems (7%)
(Cook et al., 1994). Corrosion occurred at the interface between head and neck of
modular components. No correlation between the presence or extent of corrosion
with the time in situ was found. In another study the percentage of corroded tapered
516 M.A. Barbosa

Table 10.3 Predicted Couple Behaviour

behaviour of galvanic couples
TiAlV/C Safe
(Barbosa, 1991)
CoCrMo/C Safe
TiAlV/CoCrMo Safe
316L S.S./C Unsafe
316L S.S./TiAlV Unsafe
316L S.S./CoCrMo Unsafe

Fig. 10.5  Graphs summarizing the percentage of mixed metal components which show signs of
moderate to severe corrosion as a function of duration of implanation. The dotted regression line
was fit to the data for the heads (O) and the solid line was fit to the data for the necks (▲). (Gilbert
et al., 1993b.)

connections between titanium-alloy stems and cobalt-alloy heads was found to be

about 57% (Collier et al., 1991). Titanium-titanium and cobalt-cobalt alloy combi-
nations did not result in interfacial corrosion. 85% of prostheses made of dissimilar
materials exhibited corrosion 24 months or more postoperatively. The data indicate
that a correlation exists between corrosion and time of implantation.
In view of the above clinical data it is advisable to avoid using dissimilar metals
for modular hip prostheses. The occurrence of fretting corrosion at the taper region
is responsible for the release of metallic ions that may have cytotoxic effects.
10  Corrosion of Metallic Implants 517

Laboratory galvanic current measurements can be very useful in the pre-­

screening of materials. For instance, in a study on dental implant bridges it has been
reported that silver-palladium, when brazed to titanium, corrodes in vitro (Ravnholt
and Jensen, 1991), thus eliminating the need to carry out in vivo tests.
As a final note it must be stressed that laboratory static tests should only be used
to eliminate dangerous metal-metal combinations, and not as an acceptance crite-
rion. Fretting may substantially alter the properties of the interface, by continuously
removing the passive oxide film, thereby inducing severe attack of a metal which,
otherwise, would remain unaffected.

10.3  Aspects Related to The Physiological Environment

10.3.1  Contamination of tissues by corrosion products

Contaminations of tissue with metals may have two origins. The first is the
release of ionic species resulting from the process of electrochemical dissolution of
the implant. This is normally associated with static corrosion. Under dynamic con-
ditions, and particularly when fretting occurs, small metallic particles detach from
the surface, and become embedded in the soft tissue around the implant. The fate of
these particles may vary, depending on their size and chemical nature. They may,
for instance, undergo a process of corrosion, with the consequent release of metal
ions. This process may take place both in the extracellular matrix or as a result of
macrophage activity. Table 10.4 gives the concentration of Cr, Ni, Fe and Co in
biological samples. It shows that tissues around implants may be orders of magni-
tude richer in these metallic elements than normal blood or normal bone.
Titanium has a tendency to accumulate in tissues. The concentrations can be very
high, as indicated in Table 10.5. Titanium was not excreted in the urine of hamsters

Table 10.4  Concentration of metallic elements in biological samples (Barbosa, 1992)

Sample Cr Ni Fe Co
Normal blood* 2-6 3-7 200-680 0.1-0.2
ppb ppb ppm ppb
Normal bone* 460 900 91
ppb ppb ppm
Tissue around 316SS 10000 1400 20000
Implants (max)** ppm ppm ppm
Co-Cr-Mo 10200 1500 3650 22000
ppm ppm ppm ppm
* Wet weight; data from Tsalev and Zaprianov, 1983.
** Concentrations in dry tissue; data from Pohlcr, 1983.
ppm: Parts per million.
ppb: Parts per billion
518 M.A. Barbosa

Table 10.5  Concentration of Tissue Concentrations (ppm) Ref.

titanium in tissues
Bone <2100 Ducheyne, 1984
surrounding titanium
implants Soft 2000 Meachim, 1973
tissue
Soft 56–3700 Agins, 1988
tissue

injected with metal salts (Merritt et al., 1992). Small concentrations were found in
the serum, red blood cells and organs. Only 5.5% of the injected titanium was found
in the kidneys, liver, lung and spleen tissues. The authors suggest that titanium accu-
mulates at the injection site due to the high stability of the titanium dioxide that is
formed at physiological conditions. In the same study nearly all the injected vana-
dium was recovered in the urine. This behaviour is similar to that of nickel and
cobalt, and is related to the formation of highly soluble compounds.
High concentrations of metals were found in capsule and fibrous membranes of
loose titanium and Co-Cr stems of total hip prostheses (Dorr et al., 1990). The same
work reports elevated metal ion concentrations in synovial fluid and blood when-
ever cemented and uncemented stems are loose, but no increase when they are fixed.
The average values are given in Table 10.6. The standard deviations (not shown)
were often very large, of the order of magnitude of the averages.
Polyethylene wear debris may artefactually contribute to high ion readings in
periprosthetic tissues, as indicated in Tables 10.7 and 10.8 (Meldrum et al., 1993).
The high concentrations found in UHMWPE are due to the manufacturing pro-
cesses. These tables show that there are statistically significant increases in Co,
Al and Ti in the nonarticulated inserts with respect to bar stock. In retrieved
implants, large increases with respect to bar stock were found for Cr, Mo, Ti and
V. The role of UHMWPE wear debris would be twofold: irritant to tissues and
source of metal ions.
The accumulation of metal ions in periprosthetic tissue is a combination of two
sources: the extracellular matrix and the cells themselves. The ability of fibroblasts
to incorporate metal cations is a linear function of concentration, up to 50% toxicity
concentrations, for Ag+, Au4+, Cd 2+, Cu2+, In3+, Ni2+, Pd 2+ and Zn2+ (Wataha
et al., 1993), as illustrated in Fig. 10.6 for Cu2+, NF+ and Pd2 +. By measuring the
slope of the lines in this figure it is possible to estimate the uptake efficiency
(Table 10.9). The efficiency is highest for In3+ and lowest for Pd2+. Two years after
implantation of femoral components made of Ti-6Al-4V, the titanium and alumin-
ium concentrations measured in the synovial fluid were higher for cemented com-
ponents than for the uncemented (200 μm HA, or porous Ti coatings) components
(Karrholm et al., 1994). Table 10.10 gives the data for the synovial fluid and the
aluminium concentrations in serum and urine. No significant concentrations of
vanadium were found in any of the samples, which was also the case for titanium in
serum and urine. Fast clearance of vanadium from the synovial fluid, due to high
solubility of vanadium complexes, and formation of stable titanium compounds,
10  Corrosion of Metallic Implants 519

Table 10.6  Concentration (μg/l) of metals in tissues and blood retrieved during total arthroplasty
of cementless stems (Dorr et al, 1990)
TiAlV stems CoCr stems
Sample Ti Al V Co Cr Mo Ni
SF 556 654 62 588 385 58 32
SF (control) 13 109 5 5 3 21 5
CAP 1540 2053 288 821 3329 447 5789
CAP (control) 723 951 122 25 133 17 3996
FM 20813 10581 1027 2229 12554 1524 13234
Blood 67 218 23 20 110 1524 29
Blood (control) 17 12.5 5.8 0.1–1.2 2–6 0.5–1.8 2.9–7.0
SF – synovial fluid; CAP – capsule; FM – fibrous membrane.

Table 10.7  Cobalt-chrome alloy ion concentrations in UHMWPE et al. material and manufactured
and retrieved inserts (Meldrum et al., 1993)
Co Cr Mo Ni
Bar stock, n=3 55±5 330±5 5* 650±5
Manufactured inserts, n=9 440±250 520±440 5* 490±600
Retrieved, n=18(all cemented 54±42 1,500±1,400 87±120 1,360±1,300
inserts)
All concentrations are in parts per billion (nanograms/gram).
* This is the minimum detection limit of the spectrometer.

Table 10.8  Titanium alloy ion concentrations in UHMWPE material and manufactured and
retrieved inserts (Meldrum, et al. 1993)
AL Ti Va
Bar Stock n=3 5* 5* 5*
Manufactured inserts, n=9 800±200 2300±980 60±95
Retrieved, n=21 (all metal backed) 5* 6700±4500 220±410
All concentrations are in parts per billion (nanograms/gram).
* This is the minimum detection limit of the spectrophotometer.

e.g. titanium phosphates (Ribeiro et al., 1995), might be reasonable explanations for
these findings.
Experiments with metal salts and with stainless steel and Co–Cr–Mo electrodes
corroded in vivo by applying anodic potentials showed that all the nickel and most
of the cobalt were rapidly excreted (Brown et al., 1988). Acceleration of corrosion
by the use of anodic potentials obeys similar mechanisms both in vivo and in saline
when a potential of 500 mV vs. SCE is applied. This is illustrated by the single
straight line in Fig. 10.7 (weight loss vs. total charge). In particular, this implies that
the valency of the released cations is no different in both media, according to
Faraday’s law.
520 M.A. Barbosa

Fig. 10.6  Plots of the average uptake of metal cation per cell vs. concentration of the metal cation
in the medium for Cu2+, Ni2+, and Pd2+. The least-squaresmethod was used to fit linear curves to the
points. (Wataha et al., 1993.)

Table 10.9  Uptake efficiencies of metal cations by fibroblasts (Wataha et al., 1993)
Uptake efficiency
Metal cation ((fmol/cell )/μM )/h) *
Ag+ 23.8
Au4+ 1.0
Cd2+ 38.0
Cu2+ 0.26
In3+ 45.3
Ni2+ 0.21
Pd2+ 0.11
Zn2+ 0.73
* fmol = femtomoles (10-15 moles).
10  Corrosion of Metallic Implants 521

Table 10.10  Metal concentrations (ng/g) in synovial fluid, serum and urine. Median (range)
(Karrholm et al., 1994)
Cemented HA-coated Porous Controls
Ti/synovial fluid 37 (12-56) 3.5 (0-14) 6.4 (0-7.8) 0 (0-7.5)
Al/synovial fluid 12 (6.7-28) 5.2 (2.6-13) 3.8 (2.9-9.1) 7.3 (1.9-19)
Al/serum 2.1 (0-11) 1.4 (0-5.9) 5.7 (2.1-16) 3.7 (0-17)
Al/urine 6.2 (1.7-17) 4.9 (1.7-7.0) 4.2 (3.7-4.6) 4.6 (2.1-14)

Fig. 10.7  Linear regression analysis of weight loss as a function of total charge for stainless steel
rods at 500 mV (SCE) for 30 min. Symbols: * = in saline, box = in 10% serum, circle = in vivo.
(Brown et al. 1988.)

10.3.2  Problems associated with the chemical analysis

of metallic elements in tissues

Acurate analysis of trace elements in tissues is essential to assess the degree of con-
tamination. This is not an easy task, mainly because we are dealing with normal
levels of the order of μg/litre. Sampling and sample preparation are steps prone to
serious contaminations, if the necessary precautions are not taken. As indicated in a
review by Lugowski et al. (1990), reported ‘normal’ levels of Cr in blood span over
four orders of magnitude. Contamination during sampling can be avoided by using
PTFE or polyethylene materials for blood collection and sample storage. For cut-
ting tissues a blade made of a material free from the elements to be analysed should
be used. For example, in our laboratory we have been using pure titanium blades to
522 M.A. Barbosa

cut soft tissues for Cr and Ni analysis. Contamination during sample preparation
can be minimized by: (i) adopting a very strict protocol of labware cleaning; (ii)
chemical treatment with ultrapure reagents, preferably in a microwave oven to
reduce the time necessary for digesting tissues; (iii) use a laminar flow hood to pre-
pare the samples, in order to avoid airborn contamination.
Lugowski has published a number of excellent works where the reader can find
very detailed information on the above and other aspects. The degree of precision and
accuracy to be expected when adequate experimental methods are used is indicated
in Table 10.11. This table refers to an ‘internal’ lab blood standard and to a Standard
Reference Material (SRM) with vanadium concentration certification. The relative
standard deviation (RSD) ranges from ca. 10% for Ni and Co, to ca. 29% for V.

10.3.3  Corrosion in sweat

The main constituents of sweat are chlorides (0.3–3.0 g/l), urea (0.12–0.57 g/l) and
lactic acid (0.45–4.5 g/l). When metallic objects come in contact with skin corro-
sion may occur, and if the corrosion products are toxic or irritating they may origi-
nate contact dermatitis. The most common example is dermatitis caused by
nickel-containing jewelry. In North America ca. 10% of men and women have a
history of nickel dermatitis (Randin, 1988). Although the degree of sensitization
may not be directly related to the amount of metal ions released from an object, due
to variability of response from person to person, it is generally considered that a
high corrosion resistance gives rise to fewer allergies.
The corrosion resistance of several materials in artificial sweat is given in
Table  10.12 (Randin, 1988). The composition (g/l) of the medium used was: 20
NaCl, 17.5 NH 4CI, 5 urea, 2.5 acetic acid, 15 lactic acid, pH 4.7. The table gives
the corrosion potential, Ecorr, in 02- and N2-saturated medium, the pitting potential,

Table 10.11  Precision of laboratory standard and precision and accuracy of SRM 909 human
serum (Lugowski et al., 1990)
Element X (μg/litre) n SD RSD (%)
Al 1.88 10 0.35 18.8
Co 2.37 9 0.26 10.9
Cr 0.71 8 0.14 19.6
Ni 2.95 8 0.30 10.1
Ti 4.20 7 0.41 9.8
V 0.28 8 0.08 29.4
V in SRM 909 2.73 13 0.14 5.0
certified value =
2.70±0.56 μg/litre
x – concentration; n – number of measurements; SD – Standard deviation; RSD – Relative SD.
10  Corrosion of Metallic Implants 523

Table 10.12  Main Electrochemical Parameters in ISO Sweat (Randin, 1988)

Epit, and the corrosion rate, icorr, measured by the Tafel extrapolation method. icorr, is
given only for those alloys which are in the active state. For the other alloys Epit, is
given. The following materials were found to corrode in the active state: Ni, CuNi25,
NiAl (50:50, 60:40 and 70:30), WC+Ni, white gold, FN42, Nilo Alby K, NiP. Alloys
such as stainless steels, TiC+Mo2C+Ni, NiTi, Hastelloy X, Phydur, PdNi, and SnNi
are in the passive state and may pit under exceptional circumstances. Titanium has
an extremely high Epit and therefore cannot pit under normal use.
524 M.A. Barbosa

10.3.4  Influence of proteins on the corrosion resistance

Albumin has a detrimental effect on the corrosion resistance of cast Co–Cr–Mo

alloy (Tomás et al., 1994). The breakdown potential in 0.15 M NaCI is 0.40 ± 0.02
V vs. SCE, whereas in 0.15 M NaCl+albumin it is 0.25 ± 0.06 V vs. SCE.
The presence of 5% bovine serum in lactated Ringer’s solution (pH=6.5)
increases the corrosion rate of Ti–6Al–4V alloy, as shown in the last two columns
of Table 10.13 (Lewis and Daigle, 1993b). This table gives data obtained by direct
current (d.c.) and alternating current (a.c.) methods. The difference between d.c.
and a.c. corrosion rates found in this system is not unusual. The same table also
shows that decreasing the pH of lactated Ringer’s solution to 1 has a dramatic effect
on corrosion rate.
Table 10.14 summarizes data obtained for Co–Cr, 316L stainless steel and tita-
nium. The type of electrochemical technique used has an important influence on the
results, which might indicate that the electrode potential determines the beneficial
or detrimental effect of proteins on corrosion.

10.3.5  Antibiotic-metal interactions

The interaction between a number of antibiotics (oxytetracycline, tetracycline,

tobramycin, clindamycin, cefamandole, bacitracin and chloramphenicol) and surgi-
cal metallic materials (316L stainless steel, Co-Cr and commercially pure Ti) has
shown that only oxytetracycline exerts an effect on the electrochemical response.
For all the materials this antibiotic shifted the corrosion potential of abraded sur-
faces in the noble direction, as seen in Fig. 10.8.

Table 10.13  Electrochemical characteristics of Ti6Al4V alloy in three biosimulating solutions

(Lewis, 1993b)
βab βcac Rpd Rce idcf iacg
a
Ecorr (m V per
Solution (mv) decade) (M Ω/cm2) (nA/cm2)
Lactated Ringer’s 185 210 301 2.57 1.00 21 54
(pH = 6.5)
Lactated Ringer’s (pH = 6.5) + 5% 336 187 234 1.45 0.70 31 65
bovine serum
Lactated Ringer’s (pH = 1) 147 306 1650 0.33 0.22 340 510
a Corrosion potential.
b Anodic Tafel slope.
C Cathodic Tafel slope.
d Polarization resistance; obtained from d.c. results.
e Polarization resistance; obtained from a.c. results.
f Corrosion current density; obtained from the values for Rp.
g Corrosion current density; obtained from the values for Rc.
Table 10.14  The influence of proteins on the corrosion resistance of metals
Material Effect Remarks
F75 Co-Cr-Mo Increased corrosion Accelerated anodic corrosion method: 10% serum
alloy with porous rate (Hughes et al., 1990)
coating of F75
beads
316L stainless Marginal increase Anodic polarization curves; 10% serum (Chawla
steel in pitting potential et al., 1990)
316L stainless Increased corrosion Polarization resistance method; static conditions
steel rate* 10% serum (Williams et al., 1988)
cp titanium Increased corrosion idem
rate
Ti-6Al-4V Insignificant effect idem
c.p. titanium Dual role Beneficial effect in the absence of breakdown and
detrimental when attack takes place;
potentiodynamic and galvanostatic experiments;
10% serum (Sousa and Barbosa, 1993)
316L stainless Increased pitting Potentiodynamic and galvanostatic experiments;
steel potential 10% serum (Sousa and Barbara, 1991)
F75 Co-Cr-Mo Increased Co and Constant potential (500mV vs. SCE); 10% serum
alloy Cr release (Brown et al., 1988)
316L stainless Decreased weight Constant potential (500mV vs. SCE); 10% serum
steel loss (Brown et al., 1988)
* Under fretting conditions the corrosion rate decreases.

Potential
(mV vs S.C.E.)

−100

−200

−300

−400

−500

−600
0 60 120 180 240 300
Time (minutes)

Fig. 10.8  Potential-time curves for pure titanium in 0.9% saline with and without additions of
oxytetracycline: ○(upper line), as received; ○](lower line), abraded; ◊ 0.01 mg ml-1; □0.1 mg
ml-1; ∆ 1.0 mg ml-l. (von Fraunhofer et al., 1989.)
526 M.A. Barbosa

10.4  Aspects Related to the Oxide and Other Surface Layers

10.4.1  Effect of anodizing and passivation treatments

on the corrosion resistance of titanium

For a detailed description of anodic oxidation of titanium and its alloys the reader
may refer to a review by Aladjem (1973).
The oxide on titanium can grow to thicknesses of the order of 100 nm or more
by applying anodic currents in suitable electrolytes. H3P04 and NaOH baths have
been used for this purpose. The colour of the oxide changes with thickness due to
light interference. A gold colour corresponds to a thickness of the order of
10–25 nm whereas a blue colour is normally associated with thicknesses of
30–60 nm. The corrosion resistance of anodized titanium increases as the oxide
becomes thicker. This is illustrated in Fig. 10.9 (Cigada et al. 1992), which shows
that films formed in H3PO4 are thicker than those formed in air. They are also more
protective, since the passive current density in a buffered physiological solution at
38°C is ca. 10% that measured for specimens oxidized in air. The same figure
shows that anodizing in NaOH is not so effective in reducing the current density as
doing it in H3P04•

Fig. 10.9  Average passivity currents (between 600 and SOO h) and standard deviations in physi-
ological solution of Ti6AI4V specimens, oxidized and anodized in different conditions. (Cigada
et al. 1992.)
10  Corrosion of Metallic Implants 527

The corrosion rate of anodized titanium (solution: 60 ml ethanol, 35 ml water,
10 ml lactic acid, 5 ml phosphoric acid, 5 mg citric acid and 5 mg oxalic acid; 45V,
45s) is much lower than that of passivated titanium (40% volume nitric acid, room
temperature, 30 min.), as indicated in Table 10.15 (Ong et al., 1993). The corrosion
potential of the former is also more noble, as indicated in the same table. The aver-
age thicknesses are given in Table 10.16. The anodized film is ca. 10 times thicker
than the passivated film.
There have recently been reports (Lowenberg et al., 1994l Callen et al.,1995)
indicating that passivation of Ti–6Al–4V in HNO3 increases the release of all three
constituent elements in a culture medium (α-Minimal Essential Medium with 15%
foetal bovine serum and 10% antibiotics). Table 10.17 exemplifies the results
obtained for titanium ions, for three periods of immersion of three days each. The
level of Ti is significantly reduced throughout the 9-day experimental period.

Table 10.15  Corrosion results (Ong et al., 1993)

Treatment Average Ecorr±1 SD (mV) Average Icorr±1 SD (μA/cm2)
Non-passivated -138.4±25.9 0.015±0.0l
Passivated -104.7±22.8 0.003±0.00l
Anodized 34.4±17.4 90.0006±0.000l

Table 10.16  Titanium oxide thickness (nm), relative to tantalum pentoxide (Ong et al., 1993)
Treatment Mean SD Sample size
Non-passivated 3.1 0.6 1.8
Passivated 4.1 1.8 1.8
Anodized 43.6 4.9 1.5

Table 10.17  Trace Levels of Ti, A, and V in culture medium (Callen et al., 1995)
cpTi Wells Ti6Al4V Wells
Time Control
points Not Passivated Passivated Not Passivated Passivated Values
Ti
1st 23.696±12.892 l5.735±3.354 12.599±3.850 23.338±8.497 4.983±0.977
2nd 12.650±5.275 16.640±4.940 l1.050±1.601 24.645±8.419
3rd 6.444±2.495 8.738±2.983 5.513±1.943 10.486±3.674
Al
1st 4.091±0.677 4. 133±0.523 8.933±1.187 16.878±4.574 3.476±0.392
2nd 4.694±1.039 5.523±2.784 5.703±0.707 9.656±2.750
3rd 5.215±1.096 4.l49±0.397 4.516±0.384 6.614±1.407
V
1st 0.508±0.199 0.366±0.167 6.195±2.191 21.104±8.828 0.246±0.082
2nd 0.255±0.018 0.l71±0.05l 2.789±1.129 10.096±5.697
3rd 0 0.330±0.213 0.588±0.334 4.218±2.003
528 M.A. Barbosa

10.4.2  Effect of coatings and surface treatments

on the corrosion resistance of stainless steel and titanium

When metals are used as coatings the possibility of occurrence of galvanic corro-
sion exists, since cracks or pores in the coating enable the corrosive medium to
contact the substrate. Mainly for this reason metallic coatings have not been used in
internal implants. However, surface treatments with inert materials have been
widely applied and are now in clinical practice. The effect of these and other surface
treatments will be addressed in this section.
With the development of ion implantation the plating of practically any element
on any substrate opened new perspectives to surface modification of biomaterials.
Carbon and nitrogen have been the species most widely employed to modify the
corrosion and wear behaviour of stainless steels and titanium alloys. However, the
plating of metallic elements, with a view to modifying either the corrosion perfor-
mance and/or the biological behaviour of metallic implants, is an interesting possi-
bility. This would be particularly valuable in the case of stainless steel substrates.
Very little has been reported in this area. Titanium, niobium and tantalum coatings
on stainless steel act as anodes, therefore indicating that they may retard the transfer
of chromium and nickel into the environment (Gluszek and Masalski, 1992). In the
same medium (Ringer’s solution) the oxide layers formed on titanium, niobium and
tantalum by prolonged (100h) exposure to air are not very stable. Fig. 10.10 shows
this effect (dotted line). The galvanic current first increases, corresponding to modi-
fication/destruction of the original oxide layer, and then decreases, corresponding to
increased stability of the film formed in solution. When freshly ground specimens
are used (solid line) the galvanic current decreases with time, due to film growth,
which follows a logarithmic law [log i ∝ (–log t)].
Laser surface alloying (LSA) of Ti6Al4V with Nb, Mo and Zr, in order to
increase surface hardness, has shown that the latter element is the most promising
(Akgun and Inal, 1994). The hardness increase is almost threefold in comparison to
the substrate and identical to that obtained by laser surface melting (LSM). Since a
nitrogen atmosphere was used in LSA and LSM, TiN formed during melting appears
to be the main reason for the high hardnesses obtained. The hardened zone extends
to a depth of over 0.5 mm. Wear and fretting corrosion could be considerably
reduced with such surface treatments, but no experimental data are yet available.
Radio-frequency (RF) plasma treatments in air (1.0 torr) produced enhanced
ionic release from Co–Cr–Mo and Ti–6Al–4V alloys, without any improvement in
biological behaviour (Kummer et al., 1992). Table 10.18 gives the Cr, Co and Ti
concentrations obtained after 10 days exposure to cell culture fluid (DMEM with
10% FBS). The RF plasma-treated Ti–Al–V alloy shows a 3-fold increase after the
plasma-treatment.
Depassivation of Ti–6Al–4V occurs during planar-planar rubbing against
PMMA in Ringer’s solution (Rabbe et al., 1994). The free corrosion potential drops
to values below -650 mV vs. SCE. This potential is substantially lower than those
obtained for nitrogen ion-implanted and ion-nitrided Ti–6Al–4V, which are of the
10  Corrosion of Metallic Implants 529

Fig. 10.10  Galvanic current density-time relationship for 316L/titanium couple in Ringer’s solu-
tion. (Gluszek and Masalski, 1992.)

Table 10.18  Concentration of Cr, Co and Ti in cell culture fluid after 10 days (Kummer et al.,
1992)
concentration (μ/mL)
Sample Cr Cr Ti
Control 1.1 <2 -
Co–Cr–Mo 96.5 720 –
Co–Cr–Mo/RF 120.5 960 –
Ti–AI–V – – 35.6
Ti–AI–V/RF – – 102.0
RF = Radio-frequency plasma treated.
530 M.A. Barbosa

order of , -100 mV vs. SCE. At high doses (~2xl018 ions/cm2) a TiN layer is formed
on ion-implanted surfaces, whereas TiN and Ti2N form as a result of ion nitriding,
thus increasing the hardness of the alloy surface.
Superalloy MA 956 (Fe–20Cr–4.5Al–0.5Ti–0.5Y2O3, wt%) possesses the inter-
esting ability of developing a fine α-alumina scale on the surface upon isothermal
treatment at 110 °C (Escudero and González-Carrasco, 1994). This layer acts as a
coating, being responsible for an improved corrosion resistance of the alloy, as indi-
cated by the anodic polarization curves given in Fig. 10.11. No pitting corrosion
occurs for potentials up to 700 mV vs. SCE.
Hard ceramic coatings (Al2O3 and SiC) deposited by radio-frequency (RF) sput-
tering on Ti and Co–Cr–Mo alloy resulted in significant corrosion resistance
improvement, as seen in Table 10.19 (Sella et al., 1990). The data in this table were
obtained by applying a constant potential of 1.4 V vs. SCE and measuring the cor-
rosion current density (c.d.) in artificial saliva. SiC coatings deposited on Ti caused
a decrease of c.d. of ca. 300 times. The same coating applied to Co–Cr–Mo was
only effective when an intermediate Ti sublayer was used to avoid cracking. An
Al–Al2O3 cermet sublayer was also very effective in improving the corrosion resis-
tance of Al2O3-coated Co–Cr–Mo alloy; the c.d. decreased 200 times when both
layers were used. The authors indicate that Al2O3 and SiC coatings gave better bio-
compatibility than Ti and that no signs of corrosion were observed on Al2O3-coated
dental implants removed after several years of implantation.
Modification of Ti–6Al–4V alloy surfaces by ion implantation with iridium, at
fluences of 0.74 x 1016 and 1.48x1016 ions/cm2, corresponding to 2.5 and 5.0 at% Ir
peak concentrations, has been reported (Buchanan and Lee, 1990). After pre-­
treatment of the implanted surfaces in 1N H2S04 the surfaces become enriched in Ir

Fig. 10.11  Anodic polarization curves for MA956 in the as-received and oxidized conditions after
nine months of immersion in Hank’s solution. ○ Oxidized; □ as-received. (Escudero and
Gonzalez-Carrasco, 1994.)
10  Corrosion of Metallic Implants 531

Table 10.19  Comparison of the corrosion currents of coated and uncoated metals (Sella et al., 1990)
Corrosion current at E=1.4V/SCE (μA/cm2)
Uncoated metal or alloy
Ni-Cr 6000-8000
Co-Cr-Mo 8000
Ti 260
Experimental coatings
SiC (1 μm) on Ti 0.8
SiC (1 μm) on Co-Cr 10000
Ti (1 μm) on Co-Cr 500
SiC (1 μm) + Ti (1 μm) on Co-Cr 28
Al2o3 (0.5 μm) on Co-Cr 1800
Al2o3 (0.5 μm) + Al-Al2o3 cerment on Co-Cr 40

Fig. 10.12  Corrosion potential vs. time in aerated isotonic saline. (Buchanan and Lee, 1990.)

(the concentrations are over 60% and may approach 100%), as a consequence of
alloy dissolution. The result is a corrosion potential in isotonic saline very close to
that of pure Ir, as depicted in Fig. 10.12. Owing to the very high corrosion resistance
of Ir, its implantation onto titanium is of potential interest, particularly if it becomes
significantly enriched on the surface. Galvanic couples formed between Ir and Ti is
a possibility that justifies further research.
532 M.A. Barbosa

In an attempt to reduce the release of potentially harmful metal ions from

Co-Cr-Mo surgical implants, a thin coating of TiN has been applied via physical
vapour deposition (PVD) (Wisbey et al., 1987). In vitro corrosion performance has
been investigated using electrochemical techniques. The release of Co and Cr ions
is reduced by the presence of the TiN coating. Data concerning this study are shown
in Fig. 10.13.
Thermal heating of titanium at 400 °C or immersion in 30% HN03, followed by
aging in boiling distilled water for times in the range 6–14 h, greatly reduced the
amount of Ti and Al released from Ti–6Al–4V, as shown in Table 10.20 (Browne
and Gregson, 1994). The corrosive medium was bovine serum at 37 °C. The table
also gives the ion release for two other treatments: immersion in 30% HN03, for 10
min., which is the conventional commercial treatment, and immersion for 16h in the
same solution followed by rinsing in distilled water (N). The beneficial effect of the
first two treatments is attributed to formation of rutile, which is more dense and has
a closer packed structure, with fewer paths for ion diffusion, than the oxide formed
upon passivation in nitric acid.
Commercially pure titanium and Ti–6Al–4V implants ion implanted with nitro-
gen heal as well as non-treated samples in cortical bone (Johansson et al., 1993), as
indicated by the existence of no statistically significant differences in total bone-­
metal contact.

Fig. 10.13  Metallic dissolution products released from a polished Co-Cr-Mo alloy after 550 h in
0.17 M NaCl+2.7x10-3 M EDTA solution at 37°C. □ uncoated; ■ TiN coated. (Wisbey et al., 1987.)
10  Corrosion of Metallic Implants 533

Table 10.20  Effect of Total ion release (μg/cm2)

various surface treatments on
Titanium 700 h Aluminium 700 h
the dissolution of titanium
and aluminium from 30% HN03 0.1 0.25
Ti-6Al-4V alloy (Browne and (C)
Gregson, 1994) 30% HN03 0.11 0.06
(N)
400°C (T) 0.03 0.026
Aged 10 h (A) 0.03 0.023
C - Conventional treatment (10 min. immersion).
A - Aging treatment (immersion in destilled water).
N - Immersion in HN03 for 16 h.
T - Thermal heating.

10.4.3  Effect of hydroxyapatite coatings on the corrosion

resistance of titanium and stainless steels

Most of the data available on this topic refer to hydroxyapatite deposited by plasma
spraying. Although compounds may form at the metal/hydroxyapatite interface, partic-
ularly in the case of titanium, their existence has not been unequivocally demonstrated.
Titanium phosphates and phosphides, as well as calcium titanates, may exist, but they
probably form very thin layers. The large surface roughness, caused by grit blasting of
the substrate prior to hydroxyapatite deposition, is another factor that renders identifica-
tion of any interfacial compounds by surface analysis techniques difficult.
Table 10.21 shows that the corrosion resistance of stainless steel increases upon
coating with hydroxyapatite. The presence of calcium phosphate in solution, due to
dissolution of hydroxyapatite, seems to be the cause for these changes. The same
table indicates that calcium phosphate is detrimental to the corrosion resistance of
titanium, both in terms of film breakdown potential and corrosion rate under passive
conditions.

10.4.4  Interaction between metal ions and calcium phosphates

Metallic ions may influence the formation of calcium phosphates in different ways.
Some inhibit (nickel, tin, cobalt, manganese, copper, zinc, gallium, thalium, molyb-
denum, cadmium, antimony, magnesium, and mercury), a few stimulate (iron [fer-
ric] and iridium) whereas others have no effect (cerium, titanium, chromium, iron
[ferrous], iridium, palladium, platinum, silver, gold, aluminum, and lead) (Okamoto
and Hidaka, 1994). Fig. 10.14 gives the induction time for calcium phosphate for-
mation vs. concentration for the above metal ions.
Heat treatment of Ti–6Al–4V at 280 °C for 3 h produced a high accumulation of
Ca deposited next to screws implanted in rats (Hazan et al.,1993), as indicated in
Table 10.22. The oxide was twice as thick as that formed on non-treated screws.
534 M.A. Barbosa

Table 10.21  Effect of hydroxyapatite coatings and calcium phosphate solutions on the corrosion
resistance of titanium and stainless steel
Material Solution Effect Ref
316L ss/HA Saline Increase in breakdown potential Hayashi et al.,
1990
Ti-6Al-4V/HA Saline Decrease in breakdown potential Hayashi et al.,
1990
316L ss Saline+ Increase in breakdown potential Sousa and
Ca phosphate Barbosa, 1991
316L ss Saline+ Decrease in corrosion rate (passive Barbosa, 1991 b
Ca phosphate state)
Ti cp Saline+ Decrease in breakdown potential Sousa and
Ca phosphate Barbosa, 1991
Ti cp Saline+ Increase in corrosion rate (passive Barbosa, 1991 b
Ca phosphate state)

Fig. 10.14  The induction time (min) versus concentration of various metal ions (open circle) and
HEBP: 1-hydroxyethylidene-l, 1-biphosphonate (closed circle). (Okamoto and Hidaka, 1994.)

Table 10.22  Calcium deposition (mg) next to control and heat-treated Ti–6Al–4V implants
(Hazan et al. 1993)
Time after
immersion (days) Control Heat treated
4 – –
5 – –
6 2.0±0.2 4.5±6.5
10 3.l±0.5 7.4±1.1
35 4.0±1.0 9.6±1.0
10  Corrosion of Metallic Implants 535

The presence of Ca on the surface of titanium implants after a period in vivo is

now well established. Ca deposition may be important in influencing protein adsorp-
tion, since it has been suggested that glycosaminoglycans adhere to the surface by a
Ca-O link rather then via a Ti-N bond (Sutherland et al., 1993).
Aluminium induces demineralization of previously formed bone (Frayssinet
et al., 1994), which can be ascribed to formation of stable complex aluminium
phosphate compounds (Ribeiro et al., 1995). Aluminium ions may be produced
either as a result of dissolution of Ti–6Al–4V alloy or of corrosion of alumina coat-
ings. In pH 4 buffer the release of aluminium ions from alumina is much more sig-
nificant than at pH 7 (Frayssinet, 1994). V and Ti retard apatite formation and the
growth of apatite seeds, as illustrated respectively in Figs. 10.15 and 10.16 for V
(Blumenthal and Cosma, 1989). The action of V appears to be related to the forma-
tion of V-PO4 complexes, whereas that of Ti may be due to poisoning of active
growth sites, as in the case of Al.
Hydroxyapatite coatings applied to porous titanium alloys significantly reduced
the titanium and aluminium releases, but had no important effect on vanadium
release, as shown in Fig. 10.17 (Ducheyne and Healy, 1988). No major change was
produced in the ion release kinetics from Co–Cr alloys.

Fig. 10.15  The action of

V ions in affecting direct
HA precipitation, V as
VCl5 in solution at pH 7.4,
0.15 M NaCl, 37°C in a
pH-stat. The quantity of
HA precipitated is
proportional to the extent
of OH uptake. Ca
concentration is 2.79 mM;
P04 concentration is 1.87
mM, A= control (no V);
B=0.50 mM V; C=l.00 mM
V; D=2.00 mM V.
(Blumenthal and Cosma,
1989.)
536 M.A. Barbosa

Fig. 10.16  The action of

V ions on the growth of
HA seeds. V as VCl5 in
solution at pH 7.4, 0.15 M
NaCl, 37°C in a pH-stat.
The amount of HA seeded
growth is proportional to
the OH uptake. Ca
concentration is 1.55 mM;
P04 concentration is 1.07
mM, Seed crystals were
0.15 mg/mL, and the seeds
had a surface area of
110 m2/g. A = control (no
V); B = 1.00 mM V.
(Blumenthal and Cosma,
1989.)

10.4.5  Physico-chemical properties of metal oxides

The corrosion resistance of some metals ultimately depends on the presence of a

thin oxide film formed by the reaction of the metal with the environment. This is the
case of titanium, tantalum, zirconium, molybdenum, aluminium, cobalt, chromium,
etc. Table 10.23 gives the physicochemical properties of the oxides formed on some
metals. A low oxide solubility is important to guarantee a low rate of corrosion,
since any loss in oxide thickness, due to chemical dissolution, will tend to be bal-
anced by oxidation of the metallic substrate. The oxides should also possess low
ionic conductivity.

10.4.6  Passive films on metallic implants

The oxide film on metallic implants is usually very thin (5–10 nm). It is formed as
a result of a spontaneous reaction between the metal and the environment. In spite
of the common use of immersion treatments in nitric acid solutions, usually known
as passivation treatments, they are not necessary to form an oxide. They are often
10  Corrosion of Metallic Implants 537

Fig. 10.17  (a, b, c) The

release of Ti, Al and V
from the Ti alloy after 1, 2
and 4 weeks of immersion;
D.L. indicates the detection
limit of each element.
Error bars represent the
95% confidence interval on
the means. (Ducheyne and
Healy, 1988.)

responsible for an increase in corrosion resistance due to removal of surface con-

taminations or inclusions, as in the case of stainless steels. As indicated in section
9.4.1, there have been reports suggesting that this acid treatment may decrease the
corrosion resistance of titanium.
 538

Table 10.23  Selected physico-chemico properties of metal oxides in water (Tengvall and Lundstrom, 1992)
0.996
Na(FeCN64-
FeCN63-)
Order of Water Dielectric polarization
practical corrosion pKa of Solubility lsoelectric Charge onst for resistance, Rp Corroded in Corrosion
mobility Element product hydrolysis at pH 7 (M) point at pH 7 oxide (KΩcm2) Essential Soft tissue reaction H202 at pH7 product
2 Nb Nb2O5 >20 -10-5 - 280 455 No Inert? No
3 Ta Ta2O5 >20 -10-5 - 12 1430 No Inert? No
4 Au Au2O3 - (pH7) 7x10-2 >10 (calc.) ++ 0.28 No Sequestration Yes AU2O3
Au(OH)3 5x106 E0≃1.04V
7 Ti TiO2 anat. +18 3x106 6.2 -- 48 78 110 714 No No Yes TiO22-
brook. TiO2 TiO2
rutile
14 Ag Ag2O +10 Ag++0.7 104 >1 (Ag+)12 ++ 9 No Sequestrattion Yes AgO-
AgOH
19 Al Al2O3 α 14.6 - 106 10-3 -9 + 5-10 No Sequestrattion ? ?
Al(OH)3
am.
21 Cr Cr2O3 CrO3 -1.8(Cr(OH)2+) 10-11 >10-13 8.4 (Cr3+) + 12 Yes Toxicity Toxicity Yes Yes CrO42-HCrO4-
Cr(OH)3 18.6(CrO2-)
28 Fe Fe2O3 -13.3 (Fe2+) >10-10 10-1 12.4 (Fe2+) + 100 30-38 Stainless steel Yes Sequestration Yes FeO42-?
Fe(OH)2 (pH 9.1) 10-9 8.0 (Fe3+) 316 4.38
Fe(OH)3
29 Ni Ni2+ NiO 12.2 (Ni2+) 10-15 10-11 9.5 + Yes Toxicity Yes NiO42-?
(pH 8.9) NiO2
30 Co C02+ CoO -12.6 10-11 10-12 10.8 + CoCrNi Yes Toxicity Yes CoO2
3.32
- -
40 V V2O5 V2O4 +10.3 (HV2O-) >1 10-4 1-2.5 (V5+) - Yes Toxicity Yes H3V2O7H 2VO4
M.A. Barbosa
10  Corrosion of Metallic Implants 539

Generally, the oxide film grows according to a logarithmic law (log thickness
proportional to log time), reaching a quasi-stationary thickness very rapidly. Under
stationary conditions, film dissolution and film formation rates should be the same.
Normally, film thickness increases slowly with time, after an initial period of rapid
growth. This is illustrated in Fig. 10.18, which depicts film thickening with implan-
tation time (Kasemo and Lausmaa, 1994).

Fig. 10.18  An artist’s attempt to capture some of the complexity involved in the interaction
between a material and living tissue, exemplified here by a titanium implant in bone. Note the wide
range of dimensions and time scales that are relevant. (Kasemo and Lausmaa, 1994.)
540 M.A. Barbosa

The dissolution kinetics of titanium follows a two-phase logarithmic model

(Healy and Ducheyne, 1992, 1993). In the first phase the concentration of OH
groups increases. The second phase coincides with the adsorption of P-containing
species. Fig. 10.19 clearly indicates the presence of a second phase after 400h of
immersion. In the initial phase titanium is released either in the form of Ti(OH)n(4-n)+
or TiO(OH)2. In the second phase adsorption of H2PO4-
Ti(OH)3+(ox) + H2PO4-(aq) → Ti4+(ox) . HPO42-(ad) + H2O(aq)
followed by desorption into the concentration boundary layer
Ti4+(ox) . HPO42-(ad) → Ti HPO42+(aq) + (ox) (charge transfer)

Fig. 10.19  Normalized integral passive dissolution kinetics for titanium thin films immersed in
EDTA/SIE (simulated interstitial electrolyte): (a) real time data empirically fitted with two-phase
logarithmic law relationship; (b) a semilogarithmic plot of the data demonstrating the two-phase
logarithm relationship. The correlation coefficient for the least-squares fit of the linear functions
are given. (Healy and Ducheyne, 1992.)
10  Corrosion of Metallic Implants 541

In the bulk electrolyte the complex ion dissociates

TiHPO42+(aq) + H+ (aq) → Ti 4+(aq) + H2PO4 -(aq)
and forms a more stable complex
Ti4+ (aq) + 4OH- (aq) → Ti(OH) 4 (aq)
In these reactions (ox) represents O24- in TiO2.
This mechanism is consistent with the hypothesis that in the second stage dis-
solution kinetics is dependent on diffusion within the concentration boundary layer.
It is conceivable that in the first stage field assisted dissolution may be the control-
ling step. In this stage formation of Ti(OH)4 or of hydroxy-cations, e.g. Ti(OH)3+,
has different effects on titanium transport. While Ti(OH)4 does not react with
organic molecules, Ti(OH)3+ can form organometallic complexes which may be
transported systemically.
Airborne titanium oxide, TiO2-x, is oxygen defficient but upon immersion in sim-
ulated interstitial electrolyte with EDTA (a metal chelating agent) changes to nearly
stoichiometric TiO2 (Healy and Ducheyne, 1993). TiO2 is also reported to exist on
the surface of a new Ti–15Zr–4Nb–2Ta–0.2Pd alloy (Okazaki et al. , 1994). The
other oxides present were ZrO2, Nb2O5 and Ta2O5.
Sterilization by various methods (conventional steam autoclaving, dry heat ster-
ilization in air at 160–180°C, and packaging and sterilization in sealed glass
ampoules) originates films with the composition TiO2. Their thickness is 2–6 nm,
depending on the method of sterilization. Heat sterilization increased the thickness
of the original oxide by a factor of ca. 2 (Lausmaa and Kasemo, 1990).
Films formed on metallic materials oxidized in pure oxygen at 300° C for 30 min.
have the composition shown in Table 10.24 (Oshida, 1992). Strong oxidative condi-
tions may exist in vivo, for example due to presence of the superoxide anion, O2-,
formed by inflammatory cells. The possibility of O2- originating hydrogen peroxide,
H2O2, has led Tengvall et al., (1989) to suggest that hydrogen peroxide may be of great
importance to the biological behaviour of titanium. Hydrogen peroxide is responsible
for the appearance of an outer layer, formed on top of a TiO2 layer, composed of tita-
nium oxi-hydroxide or hydrates, non-stoichiometric and rich in water (Pan et al., 1994).
The oxide thickness for wet-ground specimens is ca. 3 times that for dry-polished
specimens, as shown in Table 10.25. Hydrogen peroxide reduces the oxide thickness
and results in enhanced dissolution of titanium, according to the same authors. It is not
certain whether titanium acts as a catalyst in the oxidative deterioration of biological

Table 10.24  Type of oxide formed on biomaterials (pure oxygen, 300°C, 30 min.) (Oshida et al.,
1992)
Material Type of oxide
Pure Ti TiO2 (rutile)
Ti-6Al-4V TiO2 with traces of Al2TiO5
Ni-Ti, austenitic and martensitic Mixture of TiO2 and NiTiO3
316L stainless steel spinel-type [(Fe,Ni)O· (Fe,Cr)2O3]a and
corundum-type oxides [(Fe,Cr)2O3]a
a Possible composition.
542 M.A. Barbosa

Table 10.25  Thickness (nm) of titanium oxide films (Pan et al., 1994)
Source Polarized at 0.4 V/SCE
H2O2 in the PBS (mM) Dry-polished Wet-ground 0 1 10
XPS measurements 1.5 4.6 6.3 6.2 5.8
Capacitance measurements 6.7 5.5 6.0
Literature data 1.2-1.6 4-5 6.6

molecules, a property which has been established for other metals, e.g. iron, copper,
cadmium, chromium, lead, mercury, nickel and vanadium (Stohs and Bagchi, 1995).
These metals produce reactive oxygen species, leading to lipid peroxidation, DNA
damage, depletion of sulphydryls, apart from modifying calcium homeostasis. Since
large concentrations of titanium debris may be found around Ti and Ti-alloy implants
(section 9.3.1) the oxidative deterioration of biological molecules induced by the pres-
ence of Ti ions is a process that deserves to be studied.
The oxide formed on titanium upon passivation in HNO3 is composed of regions
of mixed titanium oxides (anatase and rutile), together with areas of amorphous
titanium oxide (Browne and Gregson, 1994). Films formed on anodized titanium
may be one order of magnitude thicker than those formed by passivation (section
9.4.1). The film is predominantly constituted by TiO2, with the presence of carboxyl
groups (Ong, 1993). It appears that upon passivation of cp Ti and Ti-6AI-4V alloy
the film on the former is thinner (3.2 ± 0.8 nm) than that on the latter (8.3 ± 1.2 nm)
(Keller et al., 1994). TiO2 films are generally amorphous, except in the case of thick
films produced by thermal oxidation or anodizing. Table 10.26 summarizes the
characteristics (composition, oxide thickness, surface topography/roughness, and
substrate microstructure/oxide crystallinity) of titanium samples subjected to vari-
ous treatments (Larsson et al., 1994). Electropolished + anodized (1M acetic acid,
room temperature) films are thicker than those formed by electropolishing and on
‘clinical reference’ (machined) surfaces. For 80V the oxide is crystalline.
Table 10.27 summarizes the composition and thickness of oxides formed on a
Co–Cr–Mo alloy exposed to ‘dry air’ and ‘wet steam’ for 1 h (Lewis, 1993a).

10.4.7  Contact angles of oxide-covered surfaces

When a metallic surface covered by an oxide (either formed naturally or by an

appropriate treatment) is placed in contact with the body or a culture medium
adsorption of various species, namely proteins, is the prime event. Contact angle, θ,
measurements can be used to ascertain the afinity of a liquid to a biomaterial sur-
face, in particular when adsorption occurs, which is revealed by a decrease in θ.
Generally speaking, θ is governed by the intermolecular forces between solid and
liquid, and in the case of passive metals by the forces between metal oxide and liq-
uid. is a complex function of surface roughness, oxide crystallinity and composi-
tion, liquid composition and time. Normally, wettability tends to progress from
hydrophobic to hydrophilic. Table 10.28 gives the initial contact angles and the
10  Corrosion of Metallic Implants 543

Table 10.26  Summary of surface characteristics of the four different types of Ti samples (Larsson
et al., 1994)
Substrate
Oxide Surface topography microstructure Oxide
Preparation Composition thickness Surface roughness crystallinity
Clinical reference TiO2+45-80% 4nm Rough, with Rough, Plastically deformed,
C, traces of Ca, with and amorphous metal
S, Si, P, CI and protrusions, ≤ 10 surface Non-­
Na μm Rrms=29±4 nm crystalline oxide
Electropolished TiO2+55-90% 4-5nm Smooth, with Polycrystalline metal
C, traces of Ca, occasional pits, ≤ 1 surface Non-­
S, Si, P, Cl and μm crystalline oxide
Na Rrms=2.7±0.9 nm
Electropolished + TiO2+55-70 21 nm Smooth, with pits Polycrystalline metal
anodized, 10 V %C, traces of and porous regions, surface Non-­
Ca, S, Si, and ~10 μm crystalline oxide
Cl Rrms=1.5±1 nm
Electropolished + TiO2+~34-­40% 180nm Heterogeneous, Polycrystalline metal
anodized, 10 V + C, traces of Ca with smooth or surface. Crystalline
and Cl porous regions, - 10 oxide (anatase)
μm Rrms=16±2 nm

Table 10.27  Composition of oxides formed on a Co-Cr-Mo alloy (Lewis, 1993a)

Medium Composition Thickness Remarks
Dry air Co14Cr5MoO21 4 nm May disaggregate into Co, Cr, Mo, CoO, CrO2
and Mox(OH)y and suboxides of Co, Cr and Mo
Wet steam Co3Cr6MoO20 3 nm May dis aggregate into Co, Cr, Mo, Co(OH2),
CrO3, CoOOH, CoMoO4, Mox(OH)y and
suboxides and hydrated species of Co, Cr and Mo

Table 10.28  Initial contact angles and changes in contact angles as function of time of oxidized
surfaces of biomaterials after mechanical and buff polishing (Oshida et al., 1992)
Mechanical polish-oxidizing Buff polish-oxidizing
θ° (deg) δθ/δt θ° (deg) δθ/δt
Pure Ti 54.24 -0.0046
Ti6A14V 32.08 -0.0010 30.85 -0.0015
NiTi (m) 69.88 -0.0055 68.92 -0.0053
NiTi (a) 71.88 -0.0048
316L s.s. 56.46 -0.0024 55.73 -0.0025
Pure Ni 35.72 -0.0016
Co-Cr alloy 62.04 -0.0023 61.85 -0.0021
α-alumina 60.87 -0.0044
m – martensite; a – austenite.
544 M.A. Barbosa

changes in contact angle as a function of time, dθ/dt, for a number of materials

(Oshida et al., 1992). Pure Ni and Ti-6Al-AV have the lowest initial contact angle,
θ°, with low dθ/dt; 316L stainless steel and Co-Cr alloy have high θ° with low d­ θ/
dt; and pure Ti, Ni-Ti alloys and α-alumina have high θ and high dθ/dt. Shot-peening
and pre-oxidation (300° C, 30 min.) of the above materials reduced the standard
deviation of contact-angle measurements, probably as a result of minimization of
microscopic irregularities (Oshida et al., 1993). In this work pure Ti exhibited the
highest initial contact angle θ°, and also the most noble corrosion potential. Both are
related to the characteristics of the TiO2 oxide that covers the metal surface. Note
that the value of θ for Al2O3 is also high (Table 10.28).
The critical surface tension (CST), which is the highest surface tension of a liquid
that completely wets a given surface, is given in Table 10.29 (Kilpadi and Lemons,
1994) for titanium subjected to various surface treatments. Polar (double-­distilled
water, glycerol and thiodoethanol) and dispersive (diiodomethane, bromonaphtha-
lene, dicyclohexyl and decane) liquids were used in the study. Radio frequency glow

Table 10.29  Critical surface tension of Ti surfaces (Kilpadi and Lemons 1994)
Critical Surface
Specimen Plot Tension, τc (dyn/cm) Comments
I C – No liquids were appropriate
P 46.0±1.08
P 42.5±1.08 Only diiodomethane and bromonaphthalene were
used
II C 31.6±0.48 Water was not used in these analyses, as it did not
fit with Good’s criterion and also did not fall in line
with the other liquids
P 35.4±0.48 Only glycerol and thiodoethanol were used
D 34.9±0.48 Only diiodomethane and bromonaphthalene were
used
III C 40.0±0.41
41.4±0.59
D 40.7±0.59 Only diiodomethane and bromonaphthalene were
used
IV C 41.9±0.79
P 41.5±1.05D
D 42.5±1.05 Only diiodomethane and bromonaphthalene were
used
V C 37.4±0.51
P 31.0±0.30 Water was not included
D 41.8±0.51 Only diiodomethane and bromonaphthalene
were used
I – Non-passivated, TFGD-treated, polished machined flats; II – Non-passivated, unsterilized, pol-
ished machined flats; III – Passivated, dry-heat-sterilized, polished machined flats; IV – Passivated,
dry-heat-sterilized, polished coined flats; V – Passivated, dry-heat-sterilized, unpolished flats; C –
Composite (includes all liquids); P – only polar liquids: D – only dispersive liquids. RFGD – Radio
Frequency Glow Discharge.
10  Corrosion of Metallic Implants 545

discharge (RFGD) treated samples showed the higher CST. Grain size (70 vs. 23
μm) did not affect the CST of polished, passivated, and dry-heat-­sterilized titanium
surfaces.
The equilibrium contact angles of cp Ti and Ti–6Al–4V, both passivated in nitric
acid, were 52±2° and 56±4°, respectively (Keller et al., 1994). Wettability was mea-
sured employing water drops. This similarity in contact angles reflets the similarity
in oxide film composition found in the same work. However, the film on the alloy
surface was significantly thicker (8.3±1.2 nm) than that on the cp Ti (3.2±0.8 nm).

Reference

O.V. Akgun and O.T. Inal, 'Laser surface modification of Ti-6A1-4V alloy', Journal of Materials
Science , 29, 1159-1168 (1994).
A. Aladjem, 'Review Anodic oxidation of titanium and its alloys', Journal of Materials Science, 8,
688-704 (1973).
M.A. Barbosa, 'Corrosion mechanisms of metallic biomaterials', Biomaterials Degradation-­
Fundamental Aspects and Related Clinical Phenomena, European Materials Research Society
Monographs, Vol. 1 (eds. M.A. Barbosa, F. Burny, J. Cordey, E. Dorre, G. Hastings, D. Muster
and P. Tranquilli-Leali), pp. 227-257, Elsevier Science Publishers, Amsterdam (1991 a).
M.A. Barbosa, 'Electrochemical impedance studies on calcium phosphate-metal interfaces',
Bioceramics, Vol. 4 (eds. W. Bonfield, G.W. Hastings and K.E. Tanner), pp. 326-333,
Butterworth-Heinemann (1991 b).
M.A. Barbosa, 'Surface layers and the reactivity of metallic implants', High-Tech Biomaterials,
European Materials Research Society Monographs, Vol. 3 (eds. D. Muster, M.A. Barbosa,
F. Burny, J. Cordey, E. Dorre, G. Hastings, and P. Tranquilli-Leali), pp. 257-283, Elsevier
Science Publishers B.V., Amsterdam (1992).
J. Black, 'Biological Performance of Tantalum', Clinical Materials, 16, 167-173 (1994)
N.C. Blumenthal and V. Cosma, 'Inhibition of apatite formation by titanium and vanadium ions',
Journal of Biomedical Materials Research: Applied Biomaterials , 23, 13-22 (1989).
M. Browne and P.J. Gregson, 'Surface modification of titanium alloy implants', Biomaterials, 15,
894-898 (1994).
S.A. Brown, L.J. Farnsworth, K. Merritt, and T.D. Crowe, ' In vitro and in vivo metal ion release',
Journal of Biomedical Materials Research, 22, 321-338 (1988).
R.A. Buchanan and I.S. Lee, 'Surface modification of biomaterials through noble metal ion implan-
tation', Journal of Biomedical Materials Research, 24,309-318 (1990).
B.W. Callen, B.F. Lowenberg, S. Lugowski, R.N.S. Sodhi, and J.E. Davies, 'Nitric acid passivation
of Ti6A14V reduces thickness of surface oxide layer and increases trace element release',
Journal of Biomedical Materials Research, 29, 279-290 (1995).
S.K. Chawla, S.A. Brown, K. Merritt, and J.H. Payer, 'Serum protein effects on polarization behav-
ior of 316L stainless steel', Corrosion 46, 147-152 (1990).
A. Cigada, G. Rondelli, B. Vicentini, M. Giacomazzi, and A. Roos, 'Duplex stainless steels for
osteosynthesis devices', Journal of Biomedical Materials Research, 23, 1087-1095 (1989).
A. Cigada, M. Cabrini, and P. Pedeferri, 'Increasing of the corrosion resistance of the Ti6A14V
alloy by high thickness anodic oxidation', Journal of Materials Science: Materials in Medicine,
3, 408-412 (1992).
J.P. Collier, V.A. Surprenant, R.E. Jensen, and M.B. Mayor, 'Corrosion at the interface of cobalt-­
alloy heads on titanium-alloy stems', Clinical Orthopaedics, 271, 305-312 (1991).
S.D. Cook, M.R. Brinker, R.C. Anderson, R.J. Tomlinson and J.C. Butler, 'Performance of retrieved
Kuntscher intramedullary rods: improved corrosion resistance with contemporary material
design', Clinical Materials, 5, 53-71 (1990).
546 M.A. Barbosa

S.D. Cook, R.L. Barrack, G.C. Baffes, A.J.T. Clemow, P. Serekian, N. Dong and M.A. Kester,

'Wear and corrosion of modular interfaces in total hip replacements', Clinical Orthopaedics,
298, 80-88 (1994).
S.D. Cook, R.L. Barrack, A.J.T. Clemow, 'Corrosion and wear at the modular interface of unce-
mented femoral stems', The Journal of Bone and Joint Surgery [Br], 76-B, 68-72 (1994).
L.D. Dorr, R. Bloebaum, J. Emmanual and R. Meldrum, 'Histologic, biochemical, and ion analysis
of tissue and fluids retrieved during total hip arthroplasty', Clinical Orthopaedics and Related
Research, nr. 261, 82-95 (1990).
P. Ducheyne and K.E. Healy, 'The effect of plasma-sprayed calcium phosphate ceramic coatings
on the metal ion release from porous titanium and cobalt­chromium alloys', Journal of
Biomedical Materials Research, 22, 1137-1163 (1988).
M.L. Escudero and J.L. Gonzalez-Carrasco, In vitro corrosion behaviour of MA956 superalloy
Biomaterials , 15, 1175-1180 (1994).
P. Frayssinet, F. Tourenne, N. Rouquet, G. Bonel, and P. Conte, Biological effects of aluminium
diffusion from plasma-sprayed alumina coatings, Journal of Materials Science: Materials in
Medicine, 5, 491-494 (1994).
J.L. Gilbert, S.M. Smith and E.P. Lautenschlager, Scanning electrochemical microscopy of metal-
lic biomaterials: reaction rate and ion release imaging modes, Journal of Biomedical Materials
Research, 27, 1357-1366 (1993a).
J.L. Gilbert, C.A. Buckley, and J.J. Jacobs, In vivo corrosion of modular hip pros­thesis compo-
nents in mixed and similar metal combinations. The effect of crevice, stress, motion, and alloy
coupling, Journal of Biomedical Materials Research, 27, 1533-1544 (1993b).
J. Gluszek and J. Masalski, Galvanic coupling of 316L steel with titanium, niobium, and tantalum
in Ringer's solution, British Corrosion Journal, 27, 135-138 (1992).
K. Hayashi, I. Noda, K. Uenoyama, and Y. Sugioka, Breakdown corrosion potential of ceramic
coated metal implants, Journal of Biomedical Materials Research, 24, 1111-1113 (1990).
R. Hazan, R. Brener and U. Oron, Bone growth to metal implants is regulated by their surface
chemical properties, Biomaterials, 14, 570-574 (1993).
K.E. Healy and P. Ducheyne, The mechanisms of passive dissolution of titanium in a model physi-
ological environment, Journal of Biomedical Materials Research, 26, 319-338 (1992).
K.E. Healy, and P. Ducheyne, Passive dissolution kinetics of titanium in vitro, Journal of Materials
Science. Materials in Medicine, 4, 117-126 (1993).
P.J. Hughes, S.A. Brown, J.H. Payer, and K. Merritt, The effect of heat treatments and bead size on
the corrosion of porous F75 in saline and serum, Journal of Biomedical Materials Research,
24, 79-94 (1990).
C.B. Johansson, J. Lausmaa, T. Rostlund, and P. Thomsen, Commercially pure titanium and
Ti6Al4V implants with and without nitrogen-ion implantation: surface characterization and
quantitative studies in rabbit cortical bone, Journal of Materials Science: Materials in Medicine
, 4, 132-141 (1993).
J. Karrholm, W. Frech, K.-G. Nilsson and F. Snorrason, Increased metal release from cemented
femoral components made of titanium alloy, Acta Orthop. Scand., 65, 599-604 (1994).
B. Kasemo and J. Lausmaa, Material-tissue Interfaces: the role of surface properties and processes,
Environmental Health Perspectives, 102, 41-45 (1994).
J.C. Keller, C.M. Stanford, J.P. Wightman, R.A. Draughn and R. Zaharias, Characterizations of
titanium implant surfaces. III, Journal of Biomedical Materials Research, 28, 939-946 (1994).
D.V. Kilpadi and J.E. Lemons, Surface energy characterization of unalloyed titanium
implants,Journal of Biomedical Materials Research , 28, 1419-1425 (1994).
F.J. Kummer, J.L. Ricci, and N.C. Blumenthal, RF plasma treatment of metallic implant surfaces,
Journal of Applied Biomaterials, 3, 39-44 (1992).
C. Larsson, P. Thomsen, J. Lausmaa, M. Rodahl, B. Kasemo and L.E. Ericson, Bone response to
surface modified titanium implants: studies on electropolished implants with different oxide
thickness and morphology, Biomaterials, 15, 1062-1074 (1994).
J. Lausmaa and B. Kasemo, Surface spectroscopic characterization of titanium implant materials,
Applied Surface Science, 44, 133-146 (1990).
10  Corrosion of Metallic Implants 547

G. Lewis, X-ray photoelectron spectroscopy study of surface layers on orthopaedic alloys.

II. Co-Cr-Mo (ASTM F-75) alloy, Journal of Vacuum Science Technology A, 11, 168-174
(1993a).
G. Lewis and K. Daigle, Electrochemical behavior of Ti-6Al-4V alloy in static biosimulating solu-
tions, Journal of Applied Biomaterials, 4, 47-54 (1993b).
B.F. Lowenberg, S. Lugowski, M. Chipman, and J.E. Davies, ASTM-F86 passivation increases
trace element release from Ti6Al4V into culture medium, Journal of Materials Science:
Materials in Medicine, 5, 467-472 (1994).
S. Lugowski, D.C. Smith, and J.C. VanLoon, Critical aspects of trace element analysis of tissue
samples: a review, Clinical Materials, 6, 91-104 (1990).
R.D. Meldrum, R.D. Bloebaum, and L.D. Dorr, Metal ion concentrations in retrieved polyethylene
total hip inserts and implications for artifactually high readings in tissue, Journal of Biomedical
Materials Research, 27, 1349-1355 (1993).
K. Merritt, R.W. Margevicius, and S.A. Brown, Storage and elimination of titanium, aluminium,
and vanadium salts, in vivo, Journal of Biomedical Materials Research, 26, 1503-1515 (1992).
Y. Okamoto and S. Hidaka, Studies on calcium phosphate precipitation: effects of metal ions used
in dental materials, Journal of Biomedicla Materials Research, 28, 1403-1410 (1994).
Y. Okazaki, A. Ito, T. Tateishi, and Y. Ito, Effect of alloying elements on anodic polarization prop-
erties of titanium alloys in acid solution, Materials Transactions, JIM, 35, 58-66 (1994).
J.L. Ong, L.C. Lucas, G.N. Raikar, and J.C. Gregory, Electrochemical corrosion analyses and
characterization of surface-modified titanium , Applied Surface Science, 72, 7-13 (1993).
Y. Oshida, R. Sachdeva, and S. Miyazaki, Changes in contact angles as a function of time on some
pre-oxidized biomaterials, Journal of Materials Science: Materials in Medicine, 3, 306-312
(1992).
Y. Oshida, R. Sachdeva, S. Miyazaki, and J. Daly, Effects of shot-peening on surface contact
angles of biomaterials, Journal of Materials Science: Materials in Medicine, 4, 443-447
(1993).
J. Pan, D. Thierry, and C. Leygraf, Electrochemical and XPS studies of titanium for biomaterial
applications with respect to the effect of hydrogen peroxide, Journal of Biomedical Materials
Research, 28, 113-122 (1994).
O.E.M. Pohler, Degradation of metallic orthopaedic implants, in Biomaterials in reconstructive
surgery (ed. L.R. Rubin), Chap. 15, The CV Mosby Co., 1983, pp. 158-228.
M. Pourbaix, Atlas of electrochemcial equilibria, NACE/CEBELCOR, 1974.
A. Pourbaix, M. Marek and R.F. Hochman, Comportement electrochimique du titane à bas pH et
bas potential d’électrode. Rapports Techniques CEBELCOR, 118, RT 197, (1971).
L.M. Rabbe, J. Rieu, A. Lopez and P. Combrade, Fretting deterioration of orthopaedic implant
materials: search for solutions, Clinical Materials, 15, 221-226 (1994).
J .-P. Randin, Corrosion behavior of nickel-containing alloys in artificial sweat, Journal of
Biomedical Materials Research, 22, 649-666 (1988).
G. Ravnholt and J. Jensen, Corrosion investigation of two materials for implant supraconstructions
coupled to a titanium implant, Scandinavian Journal of Dental Research, 99, 181-186 (1991).
C.C. Ribeiro, M.A. Barbosa, A.A.S.C. Machado, A. Tudor, and M.C. Davies, Modifications in the
molecular structure of hydroxyapatite induced by titanium ions, Journal of Materials Science:
Materials in Medicine, 6, 829-834 (1995).
C. Sella, J.C. Martin, J. Lecoeur, J.P. Bellier, M.F. Harmand, A. Nadji, J.P. Davidas, and A. Le
Chanu, Corrosion protection of metal implants by hard biocompatible ceramic coatings depos-
ited by radio-frequency sputtering, Clinical Materials, 5, 297-307 (1990).
M.F. Semlitsch, H. Weber, R. Streicher and R. Schon, Joint replacement components made of hot-­
forged and surface-treated Ti-6Al-7Nb alloy, Biomaterials, 13, 781-788 (1992).
S.R. Sousa and M.A. Barbosa, Electrochemistry of AISI 316L stainless steel in calcium phosphate
and protein solutions, Journal of Materials Science: Materials in Medicine, 2, 19-26 (1991).
S.R. Sousa and M.A. Barbosa, Corrosion resistance of titanium cp in saline physiological solutions
with calcium phosphate and proteins, Clinical Materials, 14, 287-294 (1993).
548 M.A. Barbosa

S.J. Stohs and D. Bagchi, Oxidative mechanisms in the toxicity of metal ions, Free Radical Biology
and Medicine, 18, 321-336 (1995).
D.S. Sutherland, P.O. Forshaw, G.C. Allen, I.T. Brown and K.R. Williams, Surface analysis of tita-
nium implants, Biomaterials, 14, 893-899 (1993).
P. Tengvall, L. Lundstrom, L. Sjokvist, H. Elwing, and L.M. Bjurstein, Titanium­hydrogen perox-
ide interaction: model studies of the influence of the inflammatory response on titanium
implants, Biomaterials , 10, 166-175 (1989).
P. Tengvall and I. Lundstrom, Physico-chemical considerations of titanium as a biomaterial,
Clinical Materials, 9, 115-134 (1992).
K.A. Thomas, S.D. Cook, A.F. Harding, and R.J. Haddad Jr., Tissue reaction to implant. corrosion
in 38 internal fixation devices, Orthopedics, 11, 441-451 (1988).
H. Tomás, A.P. Freire, and L.M. Abrantes, Cast Co-Cr alloy and pure chromium in proteinaceous
media: an electrochemical characterization, Journal of Materials Science: Materials in
Medicine, 5, 446-451 (1994).
S. Torgersen and N.R. Gjerdet, Retrieval study of stainless steel and titanium mini plates and
screws used in maxillofacial surgery, Journal of Materials Science: Materials in Medicine , 5,
256-262 (1994).
D.L. Tsalev and Z.K. Zaprianov, Atomic absorption spectrometry in occupational and environmen-
tal health practice, Vol. I, CRC Press, Boca Raton, 1983.
J.A. von Fraunhofer, N. Berberich, and D. Seligson, Antibiotic-metal interactions in saline
medium, Biomaterials, 10, 136-138 (1989).
J.C. Wataha, C.T. Hanks, and R.G. Craig, Uptake of metal cations by fibroblasts in vitro, Journal
of Biomedical Materials Research, 27, 227-232 (1993).
A. Wisbey, P.J. Gregson, and M. Tuke Application of PVD TiN coating to Co­Cr-Mo based surgical
implants, Biomaterials, 8, 477-480 (1987).
R.L. Williams, S.A. Brown, and K. Merritt, Electrochemical studies on the influence of proteins on
the corrosion of implant alloys, Biomaterials , 9, 181-186 (1988).
Chapter 11
Carbons

A.D. Haubold, R.B. More, and J.C. Bokros

11.1  Introduction

The biocompatibility of carbon has long been appreciated: ancient man, for exam-
ple, knew that pulverized charcoal could be placed under the skin without any
apparent ill effects (Benson, 1969). The charcoal particles visibly remained indefi-
nitely and thus allowed ancient man the means to decorate himself permanently
with tattoos. However, it was not until the mid-1960s that carbon was first consid-
ered for use as a structural material in implantable prosthetic devices. During this
period, a specific, imperfectly crystalline, man-made, pyrolytic form of carbon was
found to be well suited for application in prosthetic heart valves. Because of the
outstanding clinical success of pyrolytic carbon in long-term structural components
of heart valve prostheses, carbons have assumed a prominent position in our reper-
toire of biomaterials and have sparked investigation of other forms of carbons for
possible in vivo use. A number of these forms are listed in Table 11.1. This chapter
will be devoted to a discussion of the background and historical uses of carbons in
medical devices along with suggestions for future research.

11.1.1  Background

Although only two allotropic crystalline forms of elemental carbon, diamond and
graphite occur in nature, carbon also occurs as a spectrum of imperfect crystalline
forms that range from amorphous through mixed amorphous, graphite-like and
diamond-­like to the perfectly crystalline allotropes. Such imperfect crystalline

A.D. Haubold (*) • R.B. More • J.C. Bokros

Medical Carbon Research Institute,
8200 Cameron Road Suite A-196, Austin, TX 78754–8323, USA

© Springer Science+Business Media New York 2016 549

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_27
550 A.D. Haubold et al.

Table 11.1  Carbon Forms

Pyrolytic Produced at low or high temperature from the thermal pyrolysis of a
Carbon hydrocarbon in a fluidized bed. These materials have a laminar, isotropic,
granular or columnar structure and may be pure carbon or alloyed with
various carbides.
Glassy or Obtained from the thermal pyrolysis (~1000 °C) of selected polymers and
Polymeric may be monolithic, porous or reticulated.
Carbons
Artificial Produced from a variety of starting materials such as petroleum or naturally
Graphites occurring cokes and yield bulk structures of varying grain size, crystallite
orientation, purity, porosity, strength, and particle size.
Carbon Fibers Formed from spun polymeric fibers which are subsequently pyrolysed to
yield structures of unusual strength and stiffness. The properties are a
function of polymer precursor and processing history. More recently, carbon
fibers have been grown from the vapor phase.
Charcoal These are perhaps the oldest and most diverse materials with interesting
adsorptive properties and are produced from many organic material
spanning the range from wood to coconut shells to animal bones.
Vapor Phase Applied, generally at reduced pressures (<1 atm) and often at low
Coatings temperatures to provide a carbonaceous surface coating that ranges from
amorphous to diamond-like with accompanying wide variation in thermal,
mechanical and electrical properties.
Composites Structures have been produced that utilize all of the above materials and
even some other binders. Found in this group of materials are carbon fibers
infiltrated and held together with pyrolytic carbon, silicon carbide, glassy
carbon, PTFE, methyl methacrylate, epoxies, and petroleum pitches as well
as combinations thereof. The structures may contain randomly oriented
chopped fibers or long filaments oriented in random, 2, 3 and n dimensions.

structures are termed turbostatic and give rise to considerable variability in physical
and mechanical properties (Figure 11.1). Indeed, this ability of carbon to assume
either perfectly crystalline or chaotic, turbostatic structures gives rise to confusion
when considering physical and mechanical properties. For this reason, it is best to
consider carbon as a spectrum of materials and to bear in mind that within this spec-
trum, a number of unique combinations of structure and physical and mechanical
properties occur. This is true with respect to biocompatibility: the fact that one type
of pyrolytic carbon has been used successfully in heart valves does not necessarily
imply that other forms of pyrolytic carbons or indeed other forms of carbon in gen-
eral will also prove useful in this or other prosthetic applications. For example,
pyrolytic and glassy carbon can be finished to have identical appearances, yet the
properties of glassy carbon make it unsuitable for use in prosthetic heart valves
(Haubold et al., 1981).
For successful use in implantable prostheses a material (1) must retain its proper-
ties required for device function in the hostile biological environment and (2) must
not provoke adverse effects either locally or systemically. Most of the pure carbons
are relatively inert and unlikely to provoke severe tissue reactions, however, only
certain pyrolytic carbons have sufficient strength, fatigue resistance, wear resistance
 11 Carbons

Figure 11.1  Possible atomic arrangements of crystalline carbon: (a) diamond tetrahydral, cyclohexane chair, crystalline structure, (b) graphite planar hexagonal
layered structure (c) three-­dimensional quasicrystalline turbostatic structure. Biomedical carbons have a turbostatic structure.
551
552 A.D. Haubold et al.

and biodegradation resistance to function as long-term implant structural compo-

nents. In fact, the pyrolytic carbons were deliberately tailored to meet the specific
biocompatibility requirements for heart valve component application (Bokros et al,
1972; Haubold, 1977). Silicon was added in small amounts (5–12 wt%) as an
­alloying element in order to form silicon carbide inclusions to assure adequate wear.
Furthermore, specific processing parameters were identified in order to produce
suitable microstructures and densities, as well as the strength levels required. Glassy
carbon, as mentioned above, could not be prepared with adequate strength and wear
resistance. Thus, biocompatibility cannot be presumed a priori for carbon materi-
als: each particular application has specific demands which require a unique set of
properties. While the spectrum of carbon materials encompasses many such proper-
ties, the particular material must be engineered to uniquely satisfy all of the proper-
ties needed.

11.1.2  Diamond

Diamond, the hardest substance known, has the so called diamond cubic structure
consisting of a network of regular tetrahedral arrays in which each carbon atom is
covalently bonded to four other carbon atoms forming the corners of a regular tetra-
hedron (Figure 11.1). From X-ray diffraction data, there is a single value, 1.54 Å
bond length and a unit cube lattice spacing of 3.56 Å. The entire crystalline array is
a single covalently bonded molecule. Because many covalent bonds must be broken
to break the crystal, a very large amount of energy is required, therefore, the sub-
stance is very hard (Pauling, 1964). There has been an ongoing interest in the use of
diamond or diamond-like coatings (May, 1995). However, suitable manufacturing
processes do not yet exist that allow economical preparation of diamond type mate-
rials in the quantity, quality, shapes and sizes required for durable long term bio-
medical applications. A specific application uniquely requiring diamond or
diamond-like material for clinical success has not been identified as a justification
to compel additional research efforts in preparation techniques.

11.1.3  Graphites

Graphite has a layered hexagonal crystal structure. Each atom forms two single
bonds and one double bond with its three nearest in-plane neighbors to form sheet-­
like layers of flat six atom ring arrays (Figure 11.1). Interatomic bond distances are
1.42 Å within the layer and 3.4 Å between each layer. Within each layer, bonding is
covalent and between the layers, bonding is of the much weaker van der Waals type.
Consequently, the layers are easily separated giving rise to the soft, lubricating
properties of graphite.
Naturally occurring graphite is generally found as a isolated small scales or
imperfect single crystal precipitates in granite. These small crystals such as those
11 Carbons 553

found in Madagascar, Ontario and in New York have dimensions on the order of a

few millimeters. Larger graphite masses have been found in China and Korea.
Natural graphites are thought to be formed by metamorphosis of sedimentary
­carbonaceous materials or by reaction with and precipitation from liquid magma.
The naturally occurring graphite materials are useful primarily as the starting mate-
rial for subsequent fabrication of some ‘artificial graphites’.
Graphite was first synthesized accidentally by heating carborundum to extreme
temperatures. The silicon was driven off leaving a graphite residue. A patent on this
production process was granted over 100 years ago. Since then, petroleum and other
types of cokes coupled with organic binders have become the major raw materials
in the production of a wide variety of carbonaceous conglomerates called artificial
graphite.
The manufacturing processes employed today are as varied as are the properties
of the resultant materials. Bulk graphites are usually molded, extruded, hot isostati-
cally compacted and may even be formed by combinations of these processes. They
may be further reimpregnated with binder subsequent to graphitization (heat-­
treatment step), infiltrated isothermally or in a thermal gradient with methane or
aromatic hydrocarbons. The resultant materials have widely varying properties and
purity which can have a profound influence on the interaction of ‘graphite’ with the
living environment.
Because of low strength and low wear resistance, graphite alone is not suitable as
a structural member of an implant. But, investigations of colloidal graphite coatings
as a blood compatible surface during the mid-1960s led investigators to examine
pyrolytic carbon. Pyrolytic carbon, initially developed and used as a coating for
nuclear fuel pellets, was found to have excellent blood compatibility, strength, wear
resistance and durability for application in long term implants. The historical
account leading to the use of LTI (low temperature-isotropic) pyrolytic carbon has
been described (Bokros et al., 1972, 1975). Graphite is widely used as a substrate
for pyrolytic carbon coatings in heart valve components. In prosthesis applications,
graphite is entirely encased in pyrolytic carbon coatings on the order of one mm
thickness. Here the stronger, more durable coating stabilizes the interior graphite
structures.

11.1.4  Pyrolytic carbons

Pyrolytic carbons of the type developed at General Atomic for use in bioengineer-
ing were an off-spring of research directed at developing carbon materials that
would be suitable for structural applications in the severe environment of high-­
temperature, gas-cooled nuclear reactors. The isotropic carbon forms called high-­
temperature-­isotropic carbon were derived from the gas phase nucleation and
condensation of droplets formed during the pyrolysis of methane at temperatures in
excess of 2000°C where densification can occur by thermally activated processes.
The carbons called low-temperature-pyrolytic carbons are formed by the pyrolysis
554 A.D. Haubold et al.

of other hydrocarbons (such as propane and propylene at lower temperatures in the

range 1300–1500°C. The carbon materials so produced are not uniquely structured
nor are all of them isotropic. Wide and complex variations in properties are possible
ranging from weak to very strong, and in structure ranging from laminar and aniso-
tropic, to isotropic, to columnar and granular, the latter also varying in anisotropy.
A comprehensive review of the deposition and structure of pyrolytic carbon is given
by Bokros (1969).

11.1.5  Glassy carbons

The preparation, structure and properties of glassy, or polymeric carbons has been
described in detail by Jenkins and Kawamura (1976). These carbons are derived
from a polymer by a slow pyrolysis process which results in a vitreous residue free
of macroscopic bubbles.
Fabrication of glassy carbon materials is a relatively straightforward, but time
consuming process. A preformed polymeric precursor such as phenol-­formaldehyde,
polyfurfuryl alcohol, polyvinyl alcohol or oxidized polystyrene is slowly heated in
an inert atmosphere to a high temperature in excess of 2000 °C. Heating times may
be as short as a day or as long as one month. It is not unusual to encounter exother-
mic temperature regions that must be traversed very slowly (i.e., 1 °c temperature
increase per hour) to avoid the nucleation of bubbles.
There is a volumetric shrinkage of about 50% so the resultant structure formed
in this process is a miniature of the precursor preform. The gases generated within
the preformed structure must have time to diffuse out and not nucleate bubbles so
one dimension of glassy carbon structures is limited to about seven millimeters.
Hence materials or objects are limited to thin flat plates or tubes with thin walls.
Massive equiaxed structures are not possible unless they are small with dimensions
compatible with the diffusional requirements.

11.1.6  Carbon fibers

Carbon fibers, thought by many to be a relatively new material, actually have a long
history as evidenced by the issuance of the first patent for incandescent electric
lamp filaments (carbon fibers). The patent was issued to Thomas Edison in 1892.
Hiram Maxim (the inventor of the machine gun, among other things) was issued a
process patent for carbon fibers in 1899. Prior to the 1950s, these fibers had mar-
ginal strength and were used primarily for their electrical properties.
High strength carbon fibers were developed in the 1950s for the aerospace indus-
try and military aircraft. The mechanical properties of rayonbased carbon fibers
were enhanced using stress graphitization. Since that time, a variety of other precur-
sors have been used including polyacrylonitrile (PAN), specific fractions of asphalt
or pitch, lignin, lignosulfonates, hetero and nonheterocyclic aromatic polymers,
11 Carbons 555

linear polymers and even coal. The processes for fiber manufacture are as varied as
the precursors themselves. In the patent literature, hundreds of processes and vari-
ants can be found (Sittig, 1980). Nevertheless, generalizations can be made. The
first step in the process is the selection and treatment of a suitable raw material
which can be carbonized to a high yield. The second step generally consists of a low
temperature (250–500 °C heat treatment or preoxidation followed by high tempera-
ture (up to 2800 °C) carbonization and graphitization steps.
The resultant fibers generally are of three types classified according to their
structure and the degree of crystallite orientation. There are the high modulus (50
million psi or above), high strength fibers which, when incorporated into structures,
give the highest stiffness per unit weight. Fibers with a lower modulus (about 30
million psi) but still of high strength are the second class of generally available
fibers. The lowest modulus (less than 20 million psi) do not have structural
applications.

11.1.7  Vapor phase coatings

Coatings formed at reduced pressure (<1 atm) are generally termed ‘vapor depos-
ited’. These coatings may be formed by physical vapor deposition, chemical vapor
deposition or combinations thereof. Physical vapor deposition such as evaporation
is probably the oldest technique for depositing thin films and involves generating a
vapor from a source material at reduced pressure. The vapor subsequently con-
denses on the object to be coated. This technique suffers from a number of limita-
tions such as only line-of-sight coating is possible. In the case of a carbon source,
which does not evaporate but rather only sublimes at extreme temperatures, the
object to be coated is exposed to direct thermal radiation from the subliming source.
Few materials can withstand the intense thermal radiation for any great length of
time. Hence coating thickness is limited as is the choice of substrate material to be
coated. Much has been written on physical vapor deposition such as the comprehen-
sive text by Maissel and Glang (1970).
More versatile coating techniques are broadly termed chemical vapor deposition
(CVD). Coatings are formed through chemical reactions in the vapor phase or
through the thermal decomposition or reduction of gases generally at reduced pres-
sures. It is interesting to note that the process used to form pyrolytic carbon is a
CVD process but is generally carried out at ambient pressures. The low temperature
CVD processes are usually assisted by means of catalysts, glow discharge plasmas,
ion beams and the like to activate gas reactions. In fact, production of diamond films
is now almost routine through ion beam assisted disassociation of selected hydro-
carbons in the presence of excess hydrogen. On the other hand, amorphous carbon
films with little or no detectable crystallinity can also be produced by CVD. Thus
chemical vapor deposition techniques are extremely versatile and consequently
films and coatings produced by CVD must be carefully characterized and identified.
An extensive and comprehensive review of thin film deposition technologies can be
found in Bunshah (1982).
556 A.D. Haubold et al.

11.1.8  Composites

Even more complex than the materials described above are a family broadly termed
‘composites’. Three-dimensional structures can be formed by combining a filler
material with an appropriate matrix. Herein lies the difficulty with composites; all
too often, the starting materials are not adequately described and the resultant struc-
ture characterized.
Carbon-carbon composites can be produced with a multitude of structures. The
simplest have two-dimensional order and consist of stacked plies of carbon fabric
held together by a carbon matrix. The fabric fibers may be any of those described
previously, prepared from the pyrolysis of polyacrylonitrile and the like. The matrix
could be derived from petroleum pitch or be infiltrated pyrolytic carbon or even sili-
con carbide. The latter are generally referred to as SiC/C composites. From two-­
dimensional, the next progression in structure is three-dimensional on to
n-dimensional. This terminology refers to fiber orientation within the matrix.
Filament wound carbon composites have also been developed. In this case the
desired shape is, as the name implies, wound using carbon fibers onto a suitable
mandrel. The fibers are bonded using an epoxy type or thermoset resin. The bonding
of the matrix to the fibers and direction of fiber orientation in large part determine
the mechanical properties of the composite. The biological properties are generally
governed by the selection of matrix.

11.2  Historical Overview -In Vivo Applications

Encouraged by the success of the LTI form of pyrolytic carbon in the demanding
mechanical heart valve application, other carbon materials and usages were
explored. Different carbon materials have been evaluated because pyrolytic carbon
was patented on the one hand and processing constraints limited its versatility on
the other hand. Shown in Table 11.2 are examples that demonstrate medical and
engineering ingenuity in attempts to expand the use of carbons in the biological
environment. Although the attempts are numerous and varied, only the use of car-
bon as components for artificial heart valves has achieved widespread usage.

11.2.1  Dental

One of the earliest applications of carbon as an implant material was in restorative

dentistry. The first devices were bulky posts fabricated from glassy carbon that were
implanted in the maxilla or mandible to serve as artificial tooth roots. Because of the
inherent lack of strength of glassy carbon, they were bulky and poorly accepted. As
a further complication, the stainless steel post on which a crown was cemented
formed a galvanic couple in vivo leading to complications caused by accelerated
corrosion of the stainless steel.
Table 11.2  Applications of Carbon in Medical Devices
Active Component in Hemodetoxifier Dental Implants
Alveolar Ridge Maintenance Particles • Posts
Carbon Fiber Patch Fabrics • Blades
Cathether Tips • Coating on metallic implants
Coated Components for Membrane Electrodes
Oxygenators • Solid
Coated Emboli Filters • Coated
Coated Mandibular Trays Ex vivo Blood Filters
Coated Tracheal Prosthesis Femoral Stems
Coated TMJ Condyle Prosthesis • Coated
Coated Prosthetic Fabrics and Polymers • Composite
Coated Aneurysm Clips Femoral Condyle Replacements
Coating on Angioplasty Stents Femoral Heads
Coatings on Heart Valve Suture Rings • Coated
Coatings on Indwelling Catheters and Fracture Fixation Devices
Delivery Systems Left Ventricular Apex Inlet Tubes
Coatings on Vortex Blood Pumps Ligaments and Tendons
Coatings in Vascular Grafts Mechanical Heart Valve
Components for Centrifugal Blood Pumps Components
Composites Soft Tissue Replacement Ossicular Replacement Prosthesis
Particles for Filling Periodontal
Defects
Percutaneous Access Devices
Small Joint Replacements
• Hand
• Wrist
• Elbow
• Foot
Tibial Plateau Replacements
Vascular Attachment Prosthesis
Carbon forms
Young’s Flexural Hardness Fracture
Modulus Strength (DPH Density Toughness Wear
(GPa) (MPa) 500 g) (g/cm3) (Mpa √m) Resistance
Diamond
760–104 600– 5700– 2.9–3.5 5–7 Potentially
2000 10400 excellent
Knoop
Pure pyrolytic 28 486 >230 1.5–2.1 1.67 Excellent
carbon
Si-alloyed 31 389 >230 2.0–2.2 1.17 Excellent
pyrolytic
Carbon
Glassy or 21 175 150 <1.54 0.5–0.7 Poor
Polymeric
Carbons
Artificial 4–12 65–300 50–120 1.5–1.8 1.5 Poor
Graphites
Carbon fibers 172–517 896– * 1.6–1.8 * *
2585
Charcoal NA NA NA NA NA NA
Vapor phase 17 1.8
coatings
Composites * * * * * *
* Dependent upon matrix.
558 A.D. Haubold et al.

In another attempt, artificial tooth roots in the form of blades were fabricated
from pyrolytic carbon. Although they were less bulky than the glassy carbon
implants, they were difficult to place. Improper seating of the blade caused micro-
motion after implantation that ultimately caused the prosthesis to fail. The success
rate of 60% after 5 years was judged inadequate. Metal blades coated with carbon
fared a similar fate.
Metallic mandibular reconstruction trays coated with a vapor deposited film of
carbon generally performed well. The application was complicated by the fact that
the trays were custom and many times fashioned directly in the operating room,
making the logistics of coating with carbon unacceptable.

11.2.2  Vascular

Graft prostheses >6mm diameter are generally considered to work well and improve-
ments in performance as a result of modifying the biochemical nature of the graft
lumen with carbon coatings are difficult to quantify and not significant enough to
justify the cost of carbon coating. Improvements in the patency of small diameter
grafts (<4 mm) as a result of carbon coating have been reported but even with the
improvement, these grafts ultimately failed as a result of intimal hyperplasia prolif-
eration at the anastomoses. While the carbon coating may retard clotting of certain
grafts, the coating does little to ameliorate intimal hyperplasia formation. Carbon
vascular attachment prostheses have also been reported to perform poorly, not as a
result of poor biocompatibility, but rather the result of mechanical complications.

11.2.3  Orthopedics

Femoral stems fabricated from carbon composites fitted with a femoral head
­fabricated from pyrolytic carbon have been reportedly used successfully by some
for over 10 years (Chen, 1986). Others have experienced disastrous failures through
a lack of attention to engineering and material property details. Such failures natu-
rally lead to questions on the suitability of ‘carbon’ for use in such medical devices.
Attempts are underway to design and fabricate other joint replacements.
Ligaments and tendons have been fabricated from carbon fibers (Béjui and
Drouin, 1988). These fibers in the initial stage perform well as a scaffolding ­material
that aids in the regeneration of tendons and ligaments in vivo; but in the long term
the fibers fracture and migrate to, for example, the lymph nodes.
Fracture fixation devices that are fabricated from stainless steel are unsuitable for
coating or coupling with carbon because of galvanic effects (Haubold et al., 1986).
Carbon composite devices have been used reportedly with good results but such
usage has not become widespread presumably because of an unaccepted cost/benefit
ratio.
11 Carbons 559

11.2.4  Other

Many applications of devices listed in Table 11.2, while successful, are not in
­widespread use because, even in their non-carbon form, usage is limited. For example,
left ventricular apex tubes (Haubold et al., 1979) are used successfully in the
­construction of a prosthesis to correct idiopathic hypertrophic subaortic stenosis but
fortunately in man, such a medical condition is rare.

11.3  New Directions/Future Trends

LTI pyrolytic carbons, since their introduction in the late 1960s, have become the
material of choice for use in the fabrication of mechanical prosthetic heart valves.
Over 90% of the mechanical valves implanted worldwide utilize such carbon com-
ponents. To date, more than 2 million valves have been implanted which amounts to
an accumulated experience in excess of 12 million patient years. While this material
has proven to be the most successful carbon biomaterial, it can be improved.
Recently, advances in process control methodologies have allowed refinements in
the pyrolytic carbon coating preparation. These improvements allow the elimination
of the potentially thrombogenic silicon carbide from the biomedical coatings.
Furthermore, this pure pyrolytic carbon can be produced with substantially improved
mechanical properties relative to the silicon alloyed material (Emken et al., 1993;
Ely et al., 1994,). Such improvements in the material open possibilities for improve-
ments in heart valve prosthesis design and performance.
Results from investigations on the suitability of other forms of carbon for in vivo
use yielded mixed and often seemingly contradictory results. Some of the confusion
developed because of misunderstanding of carbon structure and misunderstanding of
the relationship of carbon properties to such structures. The use of the generic label
‘carbon’ compounded the problem. A similar situation exists with ‘polyurethanes’.
There are polyether urethanes, polyester urethanes, polyether urethane ureas and even
polyester urethane ureas – all misappropriately called simply ‘­urethanes’. Thus it is
not surprising that the biological responses of ‘carbon’ (Table 11.1) are so varied.
Biocompatibility claims for a particular form of ‘carbon’ based on published
results for a totally different form or structure should be carefully scrutinized. For
example, to claim that carbon fibers have the same biological properties as bulk
pyrolytic carbons or even that all pyrolytic carbons behave similarly is unjustified.
In the case of fibers, geometry plays a significant role. It is well known that bulk
materials may be well tolerated when the same material in particulate form may not.
The lack of characterization and standardization can be devastating.
More and more entrants are anticipated into the field of carbon biomaterials. In the
past, because of technology, patent or cost constraints, there were only several sources
for, for example, pyrolytic carbon. A number of the earlier constraints have now been
removed. Consequently, these materials are being produced in limited but increasing
560 A.D. Haubold et al.

quantities in many countries, using a multiplicity of fabrication techniques. The

­challenge for the future is to be precise in material identification, characterization and
to avoid generalization.

References

Bejui, J.  and Drouin, G. (1988). Carbon Fiber Ligaments. In CRC Critical Reviews in
Biocompatibility, 4(2), 79–108.
Benson, J.  (1969). Pre-Survey on the Biomedical Applications of Carbon, North American
Rockwell Corporation Report R-7855.
Bokros, J.C. (1969). Deposition, Structure, and Properties of Pyrolytic Carbon. In Chemistry and
Physics of Carbon, Vol. 5 (Walker, P.L., ed.). Dekker, New York, pp. 1–118.
Bokros, J.C. LaGrange, L.D. and Schoen, F.J. (1972). Control of Structure of Carbon For Use in
Bioengineering. In Chemistry and Physics of Carbon, Vol. 9 (Walker, P.L., ed.). Dekker,
New York, 103–171.
Bokros, J.C., Akins, R.J., Shim, H.S., Haubold, A.D. and Agarwal, N.K. (1975). Carbon in
Prosthetic Devices. In Petroleum Derived Carbons (Deviney, M.L., and O’Grady, T.M., eds).
American Chemical Society, Washington, DC, pp. 237–265.
Bunshah, R.F. (ed.) (1982). Deposition Technologies for Films and Coatings., Noyes Publications,
Park Ridge.
Chen, Lan-Tian (1986). Carbon-Titanium Combined Joints. In Chinese Journal of Biomedical
Engineering 3, 55–61.
Ely, J., Haubold, A., Bokros, J. and Emken, M. (1994), New Unalloyed Pyrolytic Carbon with
Improved Properties for Implant Applications, XXI Congress European Society for Artificial
Organs, Oct. 20–22, Barcelona Spain. Also US Patent 5 514 410.
Emken, M., Bokros, J., Accuntis, J. and Wilde, D., (1993) Precise Control of Pyrolytic Carbon
coating, Extended Abstracts & Program Proceedings of the 21st Biennial Conference on
Carbon, Buffalo New York, June 13–18, pp. 531–532. Also US Patent 5 284 676.
Haubold, A.D., Shim, H.S., and Bokros, J.C. (1979). Carbon Cardiovascular Devices. In Assisted
Circulation (Unger, F., ed.) Springer Verlag, Berlin, Heidelberg, New York, pp. 520–532.
Haubold, A.D. (1977). Carbon in Prosthetics. In Annals of the New York Academy of Sciences, Vol.
283, The Behavior of blood and its Components at Interfaces, (Vroman, L. and Leonard E.F.,
eds). New York Academy of Sciences, New York.
Haubold, A.D., Shim, H.S., and Bokros, J.C. (1981). Carbon in Medical Devices. In Biocompatibility
of Clinical Implant Materials, Vol II (Williams, D.F., ed.). CRC Press, Boca Raton, pp. 3–42.
Haubold, A.D., Yapp, R.A., and Bokros, J.C. (1986). Carbons for Biomedical Applications. in
Encyclopedia of Materials Science and Engineering (Bever, M.B., ed.) Pergamon Press,
Oxford, New York, Toronto, Sydney, Frankfurt, pp. 514–520.
Jenkins, G.M. and Kawamura, K. (1976). Polymeric Carbons - Carbon Fibre, Glass and Char.
Cambridge University Press, Cambridge. London, New York, Melbourne.
Lewis, J.C. and Redfern, B., and Cowland, F.B. (1963). Vitreous Carbons as Crucible Materials for
Semiconductors. In Solid State Electronics, 6, 251.
Maissel, L.I. and Giang, R. (1970). Handbook of Thin Film Technology, McGrawHill, New York.
May, P.W. (1995), CVD Diamond - a new Technology for the Future?, Endeavor Magazine 19(3),
101–106.
Pauling, L. (1964), College Chemistry, W.H. Freeman and Co., San Francisco.
Pierson, H.O. (1993) Handbook of Carbon, Graphite, Diamond and Fullerenes, Noyes
Publications, Park Ridge, New Jersey.
Sittig, M. (1980). Carbon and Graphite Fibers. Noyes Data Corporation, Park Ridge. Mechanical
Behavior of Diamond and Other Forms of Carbon, Materials Research Society Symposium
Proceedings, Vol. 383, ed. Dory M.D. et al., Materials Research Society, Pittsburgh,
Pennsylvania, 1995.
Part III
Chapter 1
General Concepts of Biocompatibility

D.F. Williams

1.1  Introduction

The host responses to biomaterials are extremely varied, involve a range of different
mechanisms and are controlled by factors that involve characteristics of host, mate-
rial and surgical procedure. These responses themselves constitute a significant
component of the phenomenon of biocompatibility. In this section, the broad con-
cepts of biocompatibility are critically reviewed with particular reference to the role
that the human host response plays in determining the performance of the biomate-
rial and of the device in which it is used. Particular emphasis is given to the influ-
ence of biocompatibility in the clinical applications of devices. It should be
remembered, however, that biocompatibility phenomena are extremely difficult to
interrogate remotely or to study in an active way, so that accurate information of the
details of biomaterial–human tissue interactions is not readily available. As Black
(1) has pointed out with reference to observations on the host response in general,
we are usually limited to detecting events long after they have occurred by examin-
ing end-points, usually histopathologically, after the host is dead. This is largely the
case with experiments on biocompatibility in animals, but is an even more relevant
observation with the human clinical experience. All comments in this section must
therefore be interpreted with this in mind.

D.F. Williams (*)

Department of Clinical Engineering, Royal Liverpool University Hospital,
147, Liverpool L69 3BX, UK

© Springer Science+Business Media New York 2016 563

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_28
564 D.F. Williams

1.2  The Definition of Biocompatibility

Etymologically, the term ‘biocompatibility’ sounds simple to interpret since it implies

compatibility, or harmony, with living systems. This concept, however, is a little too
simple to be useful and the meaning of compatibility has to be explored further.
It is intuitively obvious that a biomaterial or implanted medical device should
cause no harm to the recipient by intent or by accident. This is the underlying prin-
ciple of biological safety but is not the totality of biocompatibility. A material may
well be entirely safe in the human body but unless it actually does something useful,
it is not necessarily appropriate for a medical device. For many years during the
evolution of biomaterials, this was not really taken into account and the ‘require-
ments’ of biomaterials were dominated by the perceived necessity to be safe, which
was interpreted as a requirement that a biomaterial should be totally inert in the
physiological environment and should itself exert no effect on that environment. In
other words, there should be no interaction between biomaterials and the host, in the
latter case implying that the material should be non-toxic, non-irritant, non-­
allergenic, non-carcinogenic, non-thrombogenic and so on.
This concept of biocompatibility, which equates the quality to inertness and bio-
logical indifference, has resulted in the selection of a portfolio of acceptable or
standard biomaterials which have widespread usage. These range from the passivat-
able alloys such as stainless steel and titanium alloys, the noble metals gold and
platinum, to some oxide ceramics such as alumina and zirconia, various forms of
carbon and a range of putatively stable polymeric materials including silicone elas-
tomers (polysiloxanes), polyolefins, fluorocarbon polymers and some polyacry-
lates. of course, if this was all there was to biocompatibility, there would be few
problems other than optimizing inertness and there would be little to write about.
In practice, biocompatibility is far more complex. There are at least four reasons
for this. The first is that inertness in the physiological sense requires a great deal
more than resisting degradation at the atomic or molecular level and the second is
that even if it were so, this goal is extremely difficult to achieve. Indeed it is now
recognized that no material is totally inert in the body. Even those very stable mate-
rials mentioned above will interact to some extent with tissues; titanium, although
one of the most corrosion resistant engineering alloys, corrodes in the body, as
judged by the presence of the metal in the surrounding tissues as well as serum and
urine. With many materials, while the main component itself may be exceptionally
inert, there are often minor components, perhaps impurities or additives which can
be released under some circumstances. The leaching of plasticizers and other addi-
tives from plastics provide good examples of interactions which are not related to
the molecular breakdown of the material but which confer a degree of instability to
the product. Moreover, in the context of interactions which affect the overall perfor-
mance of the material in the physiological environment, it is important to note that
an interfacial reaction involving a physicochemical process such as protein adsorp-
tion will take place with the vast majority of materials, further emphasizing the fact
1  General Concepts of Biocompatibility 565

that inertness is a very relative term and there is indeed no such thing as an inert
biomaterial.
The third reason why biocompatibility cannot be equated with inertness is that
there are several, and indeed an increasing number, of applications which involve
intentionally degradable materials. The two most widely quoted situations here are
absorbable sutures and implantable drug delivery systems but many more circum-
stances where degradable scaffolds and matrices could form an essential component
of a device are envisaged. If biocompatibility is predicated on inertness, then
degradable materials cannot, by definition, be biocompatible. This clearly does not
make sense and suggests that the concept of biocompatibility needs to be altered.
The fourth reason is even more compelling, especially when considering bioma-
terials used in devices for tissue reconstruction. If a device is made from materials
which are inert and which do not interact with the body in any way, then it is unlikely
that it can be truly incorporated into the body. For effective long term performance
in the dynamic tissue environment, it is far more preferable for there to be functional
incorporation, which implies that the device should be stimulating the tissues to be
reactive to it positively rather than negatively. Thus biocompatibility should not be
concerned with avoiding reactions but selecting those which are the most beneficial
to device performance.
On the basis of these ideas, biocompatibility was redefined a few years ago (2),
as ‘the ability to perform with an appropriate host response in a specific situation.’
Clearly this definition encompasses the situation where inertness is still required for
the most appropriate response in some situations is indeed no response. A tradi-
tional bone fracture plate is most effective when it is attached mechanically to the
bone and does not corrode; no response of the tissue to the material is normally
required. Even here, however, we have to concede that a material that could actively
encourage more rapid bone healing might be beneficial so that a specific osteoin-
ductive response would be considered appropriate.
More importantly, the definition allows a material to stimulate or otherwise
favour a specific response, including cell activation, where that response optimizes
the performance of the device. It will be obvious that the required response will vary
with the particular application, which clearly implies that the response, both desired
and actual, will vary with the different types of tissue encountered by biomaterials.
The above definition also stipulates that the biocompatibility of a material has to
be qualified by reference to the specific application. The response to some very
common and popular biomaterials may vary quite considerably and some of the
major problems of implantable medical devices have been caused by a misunder-
standing about transferability of biocompatibility data. To recognize the very effec-
tive performance of a material under one set of conditions but then to assume that
the same material can perform equally well under entirely different circumstances
is inherently dangerous since it takes into account neither the variations one might
expect to see in the host response from site to site nor the fact that what is appropri-
ate for one situation may not be appropriate for another.
566 D.F. Williams

1.3  Components of Biocompatibility

The above definition of biocompatibility helps to explain the subject area but cannot
describe exactly what it is. For this purpose we have to consider the various compo-
nents that are involved in biocompatibility processes. Biocompatibility refers to the
totality of the interfacial reactions between biomaterials and tissues and to their
consequences. These reactions and consequences can be divided into four catego-
ries. These involve different mechanisms and indeed quite separate sectors of sci-
ence but are, nevertheless, inter-related.
The first component is that of the protein adsorption mentioned above. This pro-
cess is initiated as soon as a material comes into contact with tissue fluids such that
relatively quickly the surface of the biomaterial is covered with a layer of protein.
The kinetics and extent of this process will vary from material to material which
will in any case be a dynamic phenomenon with adsorption and desorption pro-
cesses continuously taking place. Under some circumstances, this layer is extremely
important in controlling the development of the host response since cell behavior
near the material may depend on interactions with these proteins. For example,
thrombogenicity is controlled by a number of events including the interaction
between plasma proteins and surfaces, these proteins being able to influence the
attachment of platelets to the surface. In other circumstances, the effects of this
protein layer are far from clear.
The second component of biocompatibility is that of material degradation. It is
emphasized here that degradation is a component of biocompatibility rather than a
separate phenomena. There is still confusion over this since it is often perceived that
degradation, which occurs on the material side of the interface, is the counterpart to
biocompatibility which is equated with the other (tissue) side. This is not correct
since degradation is the counterpart to the local host response, both being contribu-
tory to the biocompatibility of the system.
Degradation phenomena are covered elsewhere in this Handbook and will not be
discussed in detail here. It is necessary to point out, however, that descriptions of
material degradation mechanisms have to take the special, and indeed unique, fea-
tures of the tissue environment into account. Whatever its location, a biomaterial
will continuously encounter an aqueous environment during its use. This is not sim-
ply a saline solution, however, but a complex solution containing a variety of anions
and cations, a variety of large molecules some of which are very reactive chemi-
cally, and a variety of cells which again may be in passive or active states. There are
occasions when a degradation process can be explained, mechanistically and quali-
tatively, by the presence of an electrolyte. This is the situation with most metals
when they suffer from corrosion in physiological environments(3). Even here, how-
ever, it is known that the kinetics of corrosion may be influenced by the organic
species present, especially the proteins, and it is even possible for the corrosion
mechanism to be somewhat different to that found in non-biological situations.
With other groups of materials, however, and especially with polymers, kinetics,
mechanisms and consequences of the degradation are fundamentally related to the
1  General Concepts of Biocompatibility 567

details of the environment. Although hydrolysis remains the substantive mechanism

for degradation of most heterochain polymers, including polyurethanes, polyamides
and polyesters, this hydrolysis may be profoundly influenced by the active species
present in the tissue, especially in the tissue of the inflammatory response to materi-
als. Included here are the influences of lysosomal enzymes (4). Moreover the hydro-
lysis may be supplemented by oxidative degradation, again occurring not only by
virtue of passively dissolved oxygen in body fluids, but (and probably far more
importantly) by active oxidative species such as superoxides, peroxides and free
radicals, generated by activated inflammatory cells such as macrophages. It is thus
possible for homochain polymers not particularly susceptible to hydrolysis and not
normally oxidized at room temperature, to undergo oxidative degradation upon
implantation (5). Polyolefins such as polyethylene and polypropylene come into
this category.
The term ‘biodegradation’ is often used to describe degradation which occurs in
such situations although the circumstances of and requirements for degradation to
be so described have not been entirely clear. A recently agreed definition of biodeg-
radation (6) states quite simply that it is the breakdown of a material mediated by a
biological environment. The interpretation of ‘biological’ is left to the reader.
The purpose of explaining the role of the biological, or physiological environ-
ment in degradation phenomena, was to emphasize the crucial significance of the
interaction between degradation and the host response, for not only can degradation
be influenced by the host response but also it can control that response.
To explain this in a little more detail, let us consider the evolution of the local
host response, which is the third component of biocompatibility, using a model that
involves inflammatory and repair processes (7). Whenever a material is implanted
into the tissues of the body, there has to be a degree of trauma associated with the
insertion process. This will inevitably establish an acute inflammatory response,
which is the body’s natural defence mechanism to any injury. The inflammation is
totally desirable and helpful since it is the precursor to the second phase of the
response, which is that of tissue repair. The response to a surgical incision is acute
inflammation followed by repair, the consequences of which are a zone of fibrous
(collagenous) scar tissue. If a biomaterial is placed within the tissue, this response
will be modified by its presence, but the extent to which that modification occurs
depends on many factors.
Considering first the role of the material, if that material were totally inert chemi-
cally and unable to react at all with the tissues, and if the device were not able to
irritate the tissues in any way, the perturbation to the inflammation/repair sequence
is minimal, and the result will be the formation of a zone of fibrous tissue analogous
to the scar, but oriented in such a way as to envelope the implant. The classical
response to an implant is its encapsulation by soft fibrous tissue. On the other hand,
if the material is able to react with the tissues, chemically, mechanically or any other
way, it will act as a persistent stimulus to inflammation. While there is nothing
inherently harmful about inflammation as a response to injury, persistent inflamma-
tion occurring as a response to a persistent injury is less acceptable. At the very
least, this results in a continued stimulus to fibrosis such that the capsule is far more
568 D.F. Williams

extensive and may intervene between the material and tissue it is meant to be in
contact with (for example bone in the case of joint prostheses) but perhaps more
importantly it can change the immediate tissue environment from one of quiescent
fibrosis to that of active chronic inflammation. This is rarely the appropriate response
and, as noted above, is likely to generate an even more aggressive environment.
In the context of the definition of biocompatibility, therefore, it is important that
the interaction between the material and the tissues is one which leads to an accept-
able balance between inflammation and repair. A few points may serve to explain
this further and qualify appropriateness. First, the nature of the host response and
those features which constitute acceptability will vary very considerably from one
host to another and from one location (or set of circumstances) to another within a
particular host. It is often forgotten that host variables are as important as material
variables in the determination of biocompatibility. This is particularly important
when the wide variety of tissue characteristics is considered. Obviously bone is very
different from nerve tissue or a vascular endothelium and there will be very consid-
erable difference in the details of their responses. Not all tissues of the same variety
will be able to respond in the same way and it should always be remembered that
host variables such as age and overall health status will have a major effect.
Secondly, the importance of time and the sequence of events should never be
underestimated. While the above model describes the sequence from surgical inter-
vention to inflammation to repair, such that the process may undergo rapid resolution,
with a resulting long lasting equilibrium, the inflammation may be restimulated at any
time and rarely can we guarantee long term survival. In this context the third point
becomes important, for any feature of the interaction between material and tissue and
material can be responsible. In many situations it is the chemical reactivity, repre-
sented by degradation processes, which drives the inflammatory response, but it can
equally well be a process by which fragments of the material are extracted by physical
or mechanical means. The release of wear debris from orthopaedic prostheses is a
good example here, since the presence of such particles in sufficient numbers can have
a profound effect on the tissue response, which is mediated by the mechanisms of
inflammation but where the clinical results, manifest by loosening of the prostheses,
may not be seen for quite some time(8). As far as the response to debris is concerned,
in general the effects of released fragments will be quite different to those of bulk
materials, both by virtue of physical factors as well as changed chemical factors.
Thirdly, the identification of these events and their importance leads to various
possibilities for the control of biocompatibility. In the balance of inflammation and
repair we have the possibilities of controlling that balance by aiming to eliminate or
at least minimize those events which are undesirable for one set of conditions or
alternatively enhancing or optimizing those events which are most desirable. This
has led to the emergence of the concept and indeed introduction of bioactive materi-
als which have been defined as biomaterials that are designed to elicit or modulate
biological activity.
As a final point about the local host response, it has to be recognized that there
are significant regional and tissue-specific variations to the phenomena. These
1  General Concepts of Biocompatibility 569

c­ annot be described here, but it is important to mention the particular case with
blood. When a biomaterial comes into contact with blood, there are many different
mechanisms by which the blood can interact with the material, most of which are
preferably avoided. The most important of these are those processes, alluded to in
an earlier paragraph, that are responsible for the clotting of blood. This is a vital
defense mechanism which prevents death by uncontrolled bleeding under everyday
circumstances, but unfortunately in the context of biomaterials, the two processes
which can, either separately or together, initiate the formation of a blood clot, that
is contact phase activation of clotting proteins and platelet activation, are them-
selves initiated by contact with foreign surfaces. Thrombogenicity, defined as the
property of a material which promotes and/or induces the formation of a thrombus,
is clearly an important feature of biocompatibility.
Turning now to the last component of biocompatibility, we have to recognize that
if there is an interfacial reaction, there is no reason why the products of that reaction
and their effects have to be confined to the locality of that interface and the presence
of a benign local response is not necessarily indicative of the absence of any sys-
temic or remote site effects. The possibility of systemic effects arising from the
presence of biomaterials has long been recognized, although extreme difficulties
exist with their identification and interpretation. Indeed, at the present time, there
are few systemic effects that can be readily identified with biomaterials. The trans-
formation of a thrombus into an embolus derived from an intravascular device has
obvious implications and we can imagine and often demonstrate the systemic con-
sequences of using overtly cytotoxic materials. However, the more intriguing specu-
lations refer to the putative implant-related carcinogenicity and even more
speculative implant-related immune responses. At this stage we have to be con-
cerned about such possibilities but have to put the subject into context. While we
cannot deny that there are possible mechanisms for biomaterials to induce tumors,
the evidence that they do so in human clinical experience is very sparse. While it is
possible for some hypersensitivity responses to be seen to implants, the evidence for
any clinically ignificant response from the immune system to biomaterials is even
less available.

1.4  Conclusions

This review attempts to outline the main concept that currently prevail in the
subject area of biocompatibility. Clearly it is a complex subject, about which we
are still relatively ignorant, not least because it involves a juxtaposition between
two quite different sectors of science, the materials sciences and the molecular/
cellular biological sciences. Based on these concepts, however, a better under-
standing is now emerging so that our biomaterials can be chosen, and where
necessary treated, in order to determine that the tissues of the host do indeed
respond appropriately to them.
570 D.F. Williams

Additional Reading

Black, J. (1992) Biological Performance of Materials – Fundamentals of

Biocompatibility, 2nd edn, Marcel Dekker, Inc., New York.
A recently updated introductory text describing the basic principles of the perfor-
mance of biomaterials in biological environments and the relevance of the
biomaterial-­tissue interactions.
Hench, L.L. and Etheridge, E.C. (1982) Biomaterials: An Interface Approach,
Academic Press, New York.
An early text describing concepts of biomaterials and their interactions with tis-
sues, concentrating on the interface and based on the authors’ experiences with
bioceramics.
Williams, D.F. (ed.) (1981) Fundamental Aspects of Biomaterials, Volumes I &
II, CRC Press, Boca Raton.
An edited collection of contributions dealing with the major components of bio-
compatibility mechanisms, including corrosion and degradation phenomena, toxi-
cology and the local tissue response.

References

1. Black, J. (1984) Systemic effects of biomaterials. Biomaterials, 5, 11–18.

2. Williams, D.F. (1987) Definitions in Biomaterials, Elsevier, Amsterdam, pp. 49–59.
3. Williams, D.F. (1985) Physiological and microbiological corrosion. CRC Critical Reviews in
Biocompatibility, 1(1), 1–24.
4. Williams, D.F., Smith, R. and Oliver, C. (1987) The enzymatic degradation of polymers in
vitro. Journal of Biomedical Materials Research, 21, 991–1003.
5. Williams, D.F. and Zhong, S.P. (1991) Are free radicals involved in the biodegradation of
implanted polymers? Advanced Materials, 3, 623–626.
6. Williams, D.F., Black, J. and Doherty, P.J. (1992), in Doherty, P.J., Williams, R.L., Williams,
D.F. et al. (eds.) Biomaterial-Tissue Interfaces, Advances in Biomaterials, Volume 10. Elsevier,
Amsterdam, pp. 525–533.
7. Williams, D.F. (1989) A model for biocompatibility and its evaluation. Journal of Biomedical
Engineering, 11, 185–192.
8. Williams, D.F. (1976) Biomaterials and biocompatibility. Med. Prog. Tech., 4(1/2), 31–42.
Chapter 2
Soft Tissue Response

J.M. Anderson

2.1  Introduction

Soft tissue responses to biomaterials for medical devices are generally viewed from
the inflammation and wound healing perspectives and are usually considered as
parts of the tissue or cellular host responses to injury. Placement of a biomaterial or
medical device in the soft tissue environment involves injection, insertion, or sur-
gical implantation, all of which injure the tissues or organs involved. Early host
responses are dynamic and change with time (Table 2.1). It is important to consider
this time variable in determining the host response or biocompatibility of a
material.

2.2  Types of Response

Four general types of response may occur following the implantation of a biomate-
rial. These are a minimal response, a chemically induced response, a physically
induced response, and cellular/tissue necrosis [1].
A minimal response is generally called fibrous encapsulation and the implant is
encapsulated within fibrous tissue containing mainly collagen with a few fibro-
blasts. At the tissue/implant interface, a one to two cell layer of macrophages and
foreign body giant cells is present which constitutes the foreign body reaction.

J.M. Anderson (*)

Department of Pathology Case Western Reserve University, University Hospitals
of Cleveland, 2074 Abington Road, Cleveland, OH 44106–2622, USA

© Springer Science+Business Media New York 2016 571

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_29
572 J.M. Anderson

Table 2.1  Sequence of Local Injury

Events Following
Acute Inflammation
Implantation
Chronic Inflammation
Granulation Tissue
Foreign Body Reaction
Fibrosis

Chemically induced responses may range from an acute, mild inflammatory

response to a chronic, severe inflammatory response. These responses may be the
result of leaching of biomaterial additives or degradation products.
Physically induced responses are usually the result of the size, shape, porosity,
and other geometric factors of the biomaterial or device. The form and topography
of the surface of the biomaterial may determine the composition of the foreign
body reaction. With biocompatible materials, the composition of the foreign body
reaction and the implant site may be controlled by the surface properties of the
biomaterial, the form of the implant, and the relationship between the surface area
of the biomaterial and the volume of the implant. High surface to volume implants
such as fabrics, porous materials, or particulate, will have higher ratios of macro-
phages and foreign body giant cells at the implant site than will smooth-surface
implants, which will have fibrosis as a significant component of the implant site
[2–5]. These three general types of responses are generally found with biocompat-
ible materials.
The fourth type of response, i.e., cellular necrosis, is a toxic reaction which leads
to cell death. It is generally taken as a sign of the incompatibility of a material and
is generally the response to highly toxic additives, residual monomer, or degrada-
tion products released from the biomaterial [6]. The similarity between chemically
induced responses leading to chronic, severe inflammatory responses and cellular/
tissue necrosis should be considered in determining the biocompatibility of a
biomaterial.
Mechanical factors and edge effects may modify the response to a biomaterial.
Implant motion or micromotion can lead to variations in the fibrous capsule thick-
ness and the composition of the fibrous capsule and the interfacial foreign body
reaction. Edges and sharp changes in surface features may lead to a variation in
fibrous capsule thickness and the presence of variable concentrations of chronic
inflammatory cells, i.e., monocytes and lymphocytes.
Immune and neoplastic responses are specialized responses which are rarely
seen with biomaterials and medical devices. Immune responses are generally cre-
ated by the phagocytosis of particulate by macrophages which biochemically pro-
cess the material and communicate with lymphocytes to produce the immune
response. The metal sensitivity response is a well-known immune response to
metallic corrosion products. Neoplastic, i.e., tumor formation, responses are gener-
ally considered to be an example of solid state tumorigenesis. Solid state tumorigen-
2  Soft Tissue Response 573

Fig. 2.1  The temporal variation in the acute inflammatory response, chronic inflammatory
response, granulation tissue development, and foreign body reaction to implanted biomaterials.
The intensity and time variables are dependent upon the extent of injury created in the implantation
and the size, shape, topography, and chemical and physical properties of the biomaterial.

esis is generally linked to the extent or degree of fibrous capsule formation and the
potential for solid state tumorigenesis is reduced with increasing foreign body
reaction.

2.3  Inflammation

Inflammation is defined as the reaction of vascularized living tissue to local injury

(Table 2.1) [7, 8]. The size, shape, and intended application of a biomaterial or
medical device determine the implantation procedure which in turn determines the
extent or degree of initial injury. The size, shape, and chemical and physical proper-
ties of the biomaterial may be responsible for variations in the intensity and time
duration of the inflammatory and wound healing processes. Figure 2.1 illustrates the
temporal sequence of inflammation and wound healing. The inflammatory response
is a series of complex reactions involving various types of cells whose implant site
concentrations (densities), activities and functions are controlled by various endog-
enous and autogenous mediators [9]. The predominant cell type present in the
inflammatory response varies with the age of the injury, i.e., the time since the
implant was inserted. Neutrophils, which are the characteristic cell type of acute
inflammation, predominate during the first several days following implantation and
are replaced by monocytes as the predominant cell type. Acute inflammation is of
relatively short duration, lasting from minutes to days, depending on the extent of
injury. The main characteristics of acute inflammation are the exudation of fluid and
plasma proteins (edema) and the immigration of leukocytes (predominantly
574 J.M. Anderson

neutrophils). Following localization of leukocytes at the implant site, phagocytosis

and the release of enzymes, reactive oxygen intermediates (ROI), and other agents
occur following activation of neutrophils and macrophages. Agents released from
activated leukocytes, hydrogen ions (acid), enzymes, ROIs and others, may effect
the biodegradation of biomaterials [10, 11]. The major role of the neutrophils in
acute inflammation is to phagocytose and destroy microorganisms and foreign
material.
Acute inflammation is of relatively short duration, lasting from minutes to days,
and is dependent on the extent of injury. As the acute inflammatory response sub-
sides, monocytes and lymphocytes predominate in the implant site and are the char-
acteristic cells of chronic inflammation [7, 8]. Monocytes, migrating from the blood,
in the acute and chronic inflammatory responses differentiate into macrophages
within the tissue in the implant site. These macrophages will fuse or coalesce into
foreign body giant cells (Figure2.1). Macrophages and foreign body giant cells are
prominent at the tissue/implant interface, even with biocompatible materials. In
Figure 2.1, the intensity and time variables are dependent upon the extent of injury
created in the implantation and the size, shape, topography, and chemical and physi-
cal properties of the biomaterial.
In the phagocytosis process, recognition and attachment of neutrophils and
monocytes/macrophages are expedited when the biomaterial is coated by naturally
occurring blood serum factors called opsonins. The two major opsonins are IgG and
the complement-activated fragment C3b. Both of these plasma-derived proteins are
known to adsorb to biomaterials and neutrophils and macrophages have correspond-
ing cell membrane receptors for these opsonization proteins. These receptors may
also play a role in the activation of the attached neutrophils, monocytes, macro-
phages, or foreign body giant cells. Small particles, of the order of 5 µm in largest
dimension, may undergo the phagocytosis or engulfment process by neutrophils,
monocytes/macrophages, or specialized cells in the reticuloendothelial system
(liver, spleen, etc.). Medical devices with surface areas of biomaterial many times
greater than the size of the cell may stimulate frustrated phagocytosis. Frustrated
phagocytosis does not involve engulfment of the biomaterial but rather the extracel-
lular release of leukocyte products in an attempt to degrade or destroy the biomate-
rial [12]. Macrophages and foreign body giant cells adherent to the surface of the
biomaterial may undergo frustrated phagocytosis with the release of hydrogen ion
(acid) enzymes, ROIs, and others. Little is known regarding the extent or time
period of frustrated phagocytosis and its dependence on the chemical and physical
properties of the biomaterial.
The cells and components of vascularized connective tissue (Table 2.2) are
involved in the inflammatory and wound healing responses. Thus, injury to soft tis-
sues involves the specific types of cells which constitute the organ or tissue as well
as the cells and components of vascularized connective tissue. Vascularized connec-
tive tissue can be viewed as the general network which holds together specific cell
types in unique three-dimensional patterns to constitute organs or tissues.
While it is convenient to consider blood-material interactions separately from
tissue-material interactions, it must be emphasized that blood-material interactions
2  Soft Tissue Response 575

Table 2.2  Cells and Intravascular (blood) cells Blood plasma proteins
Components of Vascularized
Erythrocytes (RBC) Coagulative Proteins
Connective Tissue
Neutrophils Complement Proteins
Monocytes Albumin
Eosinophils Fibrinogen
Lymphocytes Gamma-Globulins
Basophils Others
Platelets Extracellular matrix
components
Connective tissue cells Collagens
Mast Cells Elastin
Fibroblasts Proteoglycans
Macrophages Fibronectin
Lymphocytes Laminin

and the inflammatory response are intimately linked and, in fact, early responses to
injury involve mainly blood and blood vessels. Therefore, both cellular and humoral
elements, i.e., plasma proteins, etc., are considered as cells and components of vas-
cularized connective tissue.

2.4  Wound Healing and Fibrosis

The wound healing response is initiated by the action of monocytes and macro-
phages, followed by proliferation of fibroblasts and vascular endothelial cells, i.e.,
capillaries, at the implant site. The proliferation of fibroblasts and the formation of
capillaries constitute granulation tissue. Modified fibroblasts, i.e., myofibroblasts,
which have contractile properties which assist in wound site closure are transiently
present in granulation tissue. As fibroblasts predominate over macrophages in the
healing response, collagens and other extracellular matrix molecules are deposited
in the implant site. The extent of the wound healing response is generally dependent
on the extent or degree of injury or defect created by the implantation procedure.
Wound healing progresses by primary union (or first intention) if the healing is
clean such as a surgical incision in which the wound edges have been approximated
by surgical sutures, clips, or staples. Healing under these conditions occurs with a
minimal loss of tissue. Wound healing by secondary union (or secondary intention)
occurs when there is a large tissue defect that must be filled or there has been an
extensive loss of cells and tissue. In wound healing by second intention, regenera-
tion of specific organ or tissue cells cannot completely reconstitute the original
architecture and more granulation tissue is formed resulting in larger areas of fibro-
sis or scar formation. Thus, the surgical procedure to create the implant site may
influence the extent or degree of the wound healing response.
576 J.M. Anderson

The end-stage healing response to biomaterials and medical devices is generally

fibrous encapsulation by collagenous fibrous tissue. This has been previously
described as the minimal response. In the minimal response, the tissue/implant
interface has a layer of macrophages and foreign body giant cells, i.e., foreign body
reaction, on the surface of the biomaterial and this is surrounded or encapsulated by
a fibrous capsule which is composed of collagen, proteoglycans, and other extracel-
lular matrix molecules. Fibroblasts may be present in the fibrous capsule.

2.5  Repair of Implant Sites

Repair of implant sites involves two distinct processes: regeneration, which is the
replacement of injured tissue by parenchymal cells of the same type, or replacement
by fibrous connective tissue that forms a capsule [7]. These processes are generally
controlled by either (i) the proliferative capacity of the cells in the tissue or organ
receiving the implant and the extent of injury as it relates to tissue destruction or (ii)
persistence of the tissue framework of the implant site. The regenerative capacity of
cells permits their classification into three groups: labile, stable (or expanding), and
permanent (or static) cells. Labile cells continue to proliferate throughout life, sta-
ble cells retain this capacity but do not normally replicate, and permanent cells can-
not reproduce themselves after birth of the host.
Perfect repair, with restitution of normal structure, theoretically occurs only in
tissues consisting of stable and labile cells, whereas all injuries to soft tissues com-
posed of permanent cells may give rise to fibrosis and fibrous capsule formation
with very little restitution of the normal tissue or organ structure. Tissues composed
of permanent cells (e.g., nerve cells, skeletal muscle cells, and cardiac muscle cells)
most commonly undergo an organization of the inflammatory exudate, leading to
fibrosis. Tissues composed of stable cells (e.g., parenchymal cells of the liver, kid-
ney, and pancreas), mesenchymal cells (e.g., fibroblasts, smooth muscle cells,
osteoblasts, and chondroblasts), and vascular endothelial and labile cells (e.g., epi-
thelial cells and lymphoid and hematopoietic cells) may also follow this pathway to
fibrosis or may undergo resolution of the inflammatory exudate, leading to restitu-
tion of the normal tissue structure. The condition of the underlying framework or
supporting stroma of the parenchymal cells following an injury plays an important
role in the restoration of normal tissue structure. Retention of the framework may
lead to restitution of the normal tissue structure, whereas destruction of the frame-
work most commonly leads to fibrosis. It is important to consider the species-­
dependent nature of the regenerative capacity of cells. For example, cells from the
same organ or tissue but from different species may exhibit different regenerative
capacities and/or connective tissue repair. An example of species differences in cell
proliferation and regeneration is the endothelialization process, proliferation of
endothelial cells, on the luminal surface of vascular grafts which does not occur in
humans but does occur in other mammals including other primates.
2  Soft Tissue Response 577

Following injury, cells may undergo adaptations of growth and differentiation.

Important cellular adaptations are atrophy (decrease in cell size or function), hyper-
trophy (increase in cell size), hyperplasia (increase in cell number), and metaplasia
(change in cell type). Hyperplasia of smooth muscle cells at blood vessel/vascular
graft anastomoses may lead to failure of the graft by stenosis or occlusion, i.e., nar-
rowing of the lumen, and thrombosis. Other adaptations include a change in which
cells stop producing one family of proteins and start producing another (phenotypic
change) or begin a marked overproduction of protein. This may be the case in cells
producing various types of collagens and extracellular matrix proteins in chronic
inflammation and fibrosis. Causes of atrophy may include decreased workload
(e.g., stress-shielding by implants), as well as diminished blood supply and inade-
quate nutrition (e.g., fibrous capsules surrounding implants).
Local and system factors may play a role in the wound healing response to bio-
materials or implants. Local factors include the site (tissue or organ) of implanta-
tion, the adequacy of blood supply, and the potential for infection. Systemic factors
may include nutrition, hematologic and immunologic derangements, glucocortical
steroids, and pre-existing diseases such as atherosclerosis, diabetes, and infection.

2.6  Summary

Inflammation, wound healing, foreign body response, and repair of implant sites are
usually considered components of the general soft tissue response to biomaterials or
medical devices. The extent or degree and temporal variations in these responses are
dictated by the inherent biocompatibility characteristics of the biomaterial or medi-
cal device. Factors which may play a role in the soft tissue response include the size,
shape, topography, and chemical and physical properties of the biomaterial. As the
implantation procedure involves injury to vascularized connective tissue, blood
responses and interactions may play a role in the general soft tissue response. The
extent, degree or type of soft tissue response is generally considered to be tissue-­
specific, organ-specific, and species-specific. Thus, a given biomaterial may be con-
sidered to be biocompatible in one shape or form but not in another and in one tissue
but not in another depending on the given application.

Additional Reading

Black, J. (1992) Biological Performance of Materials - Fundamentals of

Biocompatibility, 2nd edn, Marcel Dekker, New York.
This volume is an excellent tutorial text for the engineer/biomaterial scientist/
biologist/and others who have little or no knowledge in the area of biomaterials
and medical devices. The text is divided into four parts: General considerations,
material response: function and degradation of materials in vivo, host response:
578 J.M. Anderson

biological effects of implants, and methods of test for biological performance. The
fourth part, Methods of test for biological performance, is unique to biomaterials
texts and provides the reader with in vitro and in vivo test models and methods as
well as perspectives on the design, selection, standardization, and regulation of
implant materials.
Cohen, I.K., Diegelmann, R.F. and Lindblad, W.J. (eds) (1992) Wound Healing:
Biochemical and Clinical Aspects, W.B. Saunders Co., Philadelphia.
This is an edited volume containing 35 chapters. The volume addresses the fol-
lowing areas: Biological processes involved in wound healing (6 chapters), struc-
tural and regulatory components of wound healing (7 chapters), factors affecting
tissue repair (7 chapters), repair of specific tissues (7 chapters), and clinical man-
agement of healing tissues (7 chapters). This is an excellent volume which provides
an up-to-date and in-depth perspective of various aspects of wound healing. The
references lists provided at the end of each chapter are extensive. The strength of the
volume is its biological perspective and little is provided on biomaterials. The chap-
ter by Frederick Grinnell on cell adhesion does offer a biomaterial perspective.
Gallin, J.A., Goldstein, I.M. and Snyderman, R. (eds) (1992) Inflammation:
Basic Principles and Clinical Correlates, 2nd ed, Raven Press, New York.
This is an edited volume containing 58 chapters by individual authors. The
volume is divided in the following areas: Soluble components of inflammation (10
chapters), cytokines (5 chapters), cellular components of inflammation (21 chap-
ters), responses to inflammation (3 chapters), clinical correlates (13 chapters), and
pharmacologic modulation of inflammation (4 chapters). Each chapter is a critical,
in-depth review of the indicated subject and the references are extensive. This is
an excellent volume for those wanting an in-depth overview of the inflammatory
process and its components. No information is provided on biomaterial/inflamma-
tion interactions.
Greco, R.S. (ed.) (1994) Implantation Biology: The Host Response and
Biomedical Devices, CRC Press, Boca Raton.
This is an edited volume containing 23 chapters. Three chapters deal with bioma-
terials in general, 6 chapters address specific blood and tissue interactions with bio-
materials, 10 chapters address the use of biomaterials in specific surgical disciplines,
and 3 chapters address tissue engineering and genetic manipulation of cells. The
reference list for each chapter is extensive. This is an excellent overview of how
biomaterials interact with the host and the specific use of biomaterials in indicated
applications.
Harker, L.A., Ratner, B.D. and Didisheim, P. (eds) (1993) Cardiovascular
Biomaterials and Biocompatibility: A Guide to the Study of Blood-Tissue-Material
Interactions, Cardiovascular Pathology, 2 (3 Suppl), 1S–224S.
This is the third edition of a standard National Institutes of Health reference
previously entitled Guidelines for Blood-Material Interactions - Report of the
National Heart, Lung, and blood Institute Working Group. The volume contains 20
chapters and 3 appendices. The chapters address the following areas:
Pathophysiologic mechanisms, materials and their physicochemical characteriza-
tion, safety testing of materials and devices, and blood-vessel-material interactions.
2  Soft Tissue Response 579

The appendices are entitled: NIH Primary Reference Materials, International

Standards for Biological Evaluation of Medical Devices, and Blood Analog Fluid
for Medical Device Evaluation. This volume provides an in-depth perspective on
cardiovascular materials and state-of-the-art information is provided regarding bio-
materials. This is an excellent review, however, the editors limited the length and
number of references for each chapter due to space considerations.

References

1. Williams, D.F. and Roaf, R (1973) Implants in Surgery, W.B. Saunders Company Ltd., London,
pp. 233–35.
2. Anderson, J.M. (1993) Mechanisms of inflammation and infection with implanted devices.
Cardiovascular Pathology, 2 (3 Suppl.), M319-M321.
3. Anderson, J.M. (1994) In vivo biocompatibility of implantable delivery systems and biomate-
rials. European Journal of Biopharmaceutics, 40, 1–8.
4. Anderson, J.M. (1994) Inflammation and the foreign body response. Problems in General
Surgery, 11(2), 147–160.
5. Black, J. (1992) The inflammatory process, in Biological Performance of Materials
-Fundamentals of Biocompatibility, 2nd edn, Marcel Dekker, Inc.,New York, pp. 125–147.
6. Marchant, R.E., Anderson, J.M. and Dillingham, E.O. (1986) In vivo biocompatibility studies.
VII. Inflammatory response to polyethylene and to a cytotoxic polyvinylchloride. Journal of
Biomedical Materials Research, 20, 37–50.
7. Cotran, R.Z. et al. (1994) Inflammation and repair, in Pathologic Basis of Disease, 5th edn,
Cotran, R.Z., Kumar, V. and Robbins, S.L. (eds), W.B. Saunders Co., Philadelphia, pp. 51–92.
8. Gallin, J.I., Goldstein, I.M. and Snyderman, R. (eds) (1992) Inflammation. Basic Principles
and Clinical Correlates, 2nd edn, Raven Press, New York.
9. Spector, M., Cease, C. and Tong-Li, X. (1989) The local tissue response to biomaterials. CRC
Critical Reviews in Biocompatibility, 5 (4), 269–295.
10. Weissman, G., Smolen, J.E. and Korchak, H.M. (1980) Release of inflammatory mediators
from stimulated neutrophils. New England Journal of Medicine, 303, 27–34.
11. Henson, P.M. (1980) Mechanisms of exocytosis in phagocytic inflammatory cells. American
Journal of Pathology, 101, 494–511.
12. Henson, P.M. (1971) The immunologic release of constituents from neutrophil leukocytes:
II. Mechanisms of release during phagocytosis, and adherence to nonphagocytosable surfaces.
Journal of Immunology, 107, 1547–57.
Chapter 3
Hard Tissue Response

T.O. Albrektsson

3.1   Introduction

The initial tissue response when a biomaterial is implanted in the body is dependent
on release of specific growth factors. It has been indicated by Frost [1] that the
inevitable bone injury resulting from surgery and the presence of an implant will
release various types of growth factors that will sensitize cells and promote cellular
mitosis. This is a general healing response that will result in growth of all sorts of
local connective tissues, bone as well as various types of soft tissue.
The balance between these tissue varieties is controlled by the action of chemical
mediators which issue ‘instructions’ for the amount of bone and soft tissue to be
formed as an appropriate healing response. This delicate balance can easily be dis-
turbed inadvertently and may cause the undesirable end-result of an interfacial soft
tissue embedment of the implant or, in the case of fracture healing, formation of a
pseudoarthrosis. The discussion in this section will focus on various modes of
implant fixation, such as cementation, ingrowth and osseointegration (Figure 3.1).

3.2  Fixation by Cementation

Bone cement, a two component acrylic, is frequently used for implant fixation in the
cases of hip and knee arthroplasties. Bone cement is toxic with localized as well as
general adverse tissue reactions [2]. Therefore, the good long-term results reported
with cemented arthroplasties seem to be quite puzzling. However, it must be

T.O. Albrektsson (*)

Handicap Research, Institute for Surgical Sciences, University of Gothenburg,
Medicinaregatan 8, Gothenburg S-413 90, Sweden

© Springer Science+Business Media New York 2016 581

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_30
 582

Fig. 3.1  Different classification criteria are used for implants in bone. Here we distinguish between cemented implants, which are dependent upon cure in place
luting or filling agents; cement-free (ingrowth implants) which are dependent on bone tissue growth into surface features for fixation; and osseointegrated
implants, which are dependent upon collagen and cellular processes and their physiochemical interactions with the implant surface for fixation. In each case, a
T.O. Albrektsson

‘ridge’ feature on the implant surface is shown in cross-section; the width of each band of the diagram is several millimeters (cells (osteocytes) are shown larger
than actual size).
3  Hard Tissue Response 583

understood that the strength of the cemented bone interface is not related to the live
state of bone tissue: in reality, the curing bone cement invading the trabecular net-
work results in a substantial cancellous bone interdigitation (Figure 3.1). The inter-
facial bone usually dies from the combined trauma of heat and monomer leakage.
The width of this necrotic zone varies depending on the extent of the surgical trauma
and the type of bone cement used, but is at least a few millimeters [2]. Ironically, a
revitalization of the interfacial bone may prove disastrous, as the bone will then run
an increased risk of resorption. In examining interfacial tissues reactions to retrieved
cemented orthopaedic implants, most studies have examined the tissue after removal
of the implant. This leads to an uncertainty in examining the proportion of true bone
to implant contact.
We have been able to investigate bone implants in situ, at the resolution level of
light microscopy, utilizing the techniques developed by Donath et al. [3]. Computer
assisted calculations, using custom developed software, permit determination of the
proportion of bone to implant contact, the amount of bone in the region adjacent to
the implant and comparisons of the bone density adjacent to and in the immediate
region of the implant (‘outfolded mirror image’) (Figure 3.2).
These data correlate positively to biomechanical tests: the more bone in the inter-
facial region of the implant, the greater the torque necessary to remove the implant
when we apply controlled rotational forces to the implant body. Specimens of 10μm
thickness have been investigated at the light microscopic level of resolution [4].
Other studies using cutting and grinding techniques have investigated the implant in

Fig. 3.2 Computerized
histomorphometric
approach to evaluation of
the bone-implant interface.
The bone-to-implant
contact percentage is the
linear contact area between
bone and implant in the
inside zone; percent bone
ingrowth is the ratio of
bone ‘inside’ to that
‘outside’ in the ‘mirror’
zone.
584 T.O. Albrektsson

situ, but ended up with sections of a final thickness of 40 μm or more leading to

clear overestimates of the true bone to implant contact. In a retrieval study of
cemented resurface arthroplasties, Morberg [2] was able to overcome these method-
ological problems and verify a very poor bone to cement contact, even though he
found bone fragments with disturbed mineralization bordering the cement.

3.3  Fixation by Ingrowth (Cement-Free Implants in Bone)

In the absence of cement, fixation may be obtained by active bone tissue ingrowth
into the implant surface irregularities of medullary stems (Figure 3.2). As pointed
out by Black [5], one disadvantage of cement-free implants is the 3–4 week waiting
time before the bone-implant interface can support significant shear loading; this is
in contrast to a cemented interface which, if successful, has adequate shear strength
within one hour of implantation. Osseointegrated implants have even lower interfa-
cial shear strength at the the same 3–4 week post operative time point [6].
In the clinical experience, however, the results of cement-free joint arthroplasties
utilizing ingrowth fixation have not matched those of cemented devices. The appro-
priate bone ingrowth is often disturbed and incomplete, leading to early failure of
fixation. Cement-free knee arthroplasty components have been observed to migrate
by up to 2 mm in the first post-operative year, while cemented devices of the same
design migrate only half as much or less [7]. The failure rates of cement-free hip
arthroplasties have been so substantial in comparison to cemented devices that the
former mode of implant insertion has been restricted to young patients for whom
cemented devices give poorer results at most clinics [8].
The outcome of cement-free orthopaedic implants depends, as does the outcome
of craniofacial osseointegrated implants, on the more or less simultaneous control
of a number of different factors including biocompatibility, design and surface con-
ditions of the implant, the state of the host bed, the surgical technique and the load-
ing conditions after implant insertion.
Herein is the explanation for the dubious results of many current designs of
cement-free hips and knees: the implant material used, the design and surface finish
have been well adapted to engineering demands, but not so well matched to the
biological needs.
Ideal implant characteristics with respect to bone anchorage are quite different
from those of gliding surfaces. The surgical technique utilized when inserting a
cement-free hip or knee of current designs is, of necessity, traumatizing: resulting in
an impaired bone healing response. Current stem-type cement-free hips migrate
when loaded, as do tibial components of artificial knees, leading to a change in the
chemical mediators and subsequent increase of soft tissue formation. It is, therefore,
not surprising to learn from studies of retrieved cement-free hips that there is, as a
rule, only sparse bony ingrowth into retrieved acetabular cups [9], femoral cups [2,
10] or femoral stems [9, 11, 12]. More modern types of cement free hip prostheses
with titanium meshwork or HA-coated surfaces have had slightly better clinical
3  Hard Tissue Response 585

results than those implanted and retrieved during the 1980s and it may well be pos-
sible that there is more abundant bone ingrowth in some of the more recent designs.
Nevertheless, from multicenter studies it seems quite clear that summed five-year
failure rates for cement-free arthroplasties are greater than those for cemented
devices [8]. This observation illustrates that the anchorage problems associated with
cement free arthroplasties are far from solved.

3.4  Osseointegration

Osseointegration is a term introduced by Brånemark et al. [13] to describe a loaded,

stable implant in direct contact with bone (no apparent intervening soft tissue under
light microscopic examination). Osseointegrated implants differ from ingrown ones
that are dependent upon bone growth into surface macroscopic features or irregu-
larities. By contrast, osseointegration is dependent on tissue ingrowth into minute
surface features, such as the fundamental asperities of a ‘smooth’ surface or, as
postulated for surfaces of various crystalline calcium phosphates (such as calcium
hydroxyapatite) or amorphous, bioactive glasses, on direct chemical bonding
between tissue and implant. Irrespective of the type of interfacial contact – chemical
bonding or mechanical interdigitation – an important difference is that while the
former requires only limited tissue elements, the latter requires complete, mature
bone elements for appropriate function. Theoretically, an ingrown interface may
also be osseointegrated as well but experimental evidence for this, from examina-
tion of thin sections, is virtually nonexistent. Macroscopic features on implant sur-
faces designed for ingrowth fixation include sintered beads, rough plasma sprayed
coatings and sintered meshes, since bone requires features of minimum dimensions
of about 100 μm for successful ingrowth [14, 15].
However, surface irregularities only in the nanometer to micrometer range are
necessary for osseointegration when implant stability is dependent on cellular and/
or collagen ingrowth, rather than bone ingrowth. Since this ingrowth (perhaps better
termed ‘ongrowth’) of tissue elements occurs in a three dimensional manner, the
osseointegrated implant will, from a biomechanical viewpoint, be a directly bone-­
anchored device. In fact, Wennerberg [16] has demonstrated that implant surfaces
with a CLA (center line average) roughness of ~1 μm will experience more rapid
bony incorporation, through osseointegration, than ones with CLAs of 0.1 or 2 μm.
This observation leads to the hypothesis that too smooth surfaces (CLA ~0.1) may
not permit proper collagen attachment while rougher ones (CLA ≥2 μm) may
release too many metal ions that disturb cellular functions necessary for anchorage.
There is no doubt that osseointegration has resulted in a clinical breakthrough in
oral implants. Soft tissue embedded load-carrying devices do not function adequately
in the jaw. In sharp contrast, properly osseointegrated implants do. However, this
does not necessarily imply that every functioning bone implant in other parts of the
body need to be osseointegrated. On the contrary, cemented hip arthroplasties with a
586 T.O. Albrektsson

bone-cement interface consisting of soft tissue or mostly dead bone have demon-
strated significant clinical longevity, in many cases exceeding ten to fifteen years.
Furthermore, the so called osseointegrated interface is still in need of a proper
definition. First described by Brånemark [17] as a bone response that occurred
everywhere around the implant circumference of c.p. titanium screws in placed in
bone, osseointegration is regarded today as a more nonspecific tissue response
resulting in a mix of interfacial soft and hard tissues. In reality, bone anchorage of
foreign bodies is a more general type of tissue response that occurs to c.p. titanium
alone [3]. The only definition of osseointegration that has stood up to a critical
analysis is based on a clinical finding of implant stability: ‘A process whereby clini-
cally asymptomatic rigid fixation of alloplastic materials is achieved and main-
tained, in bone, during functional loading’ [18]. The continued usage of a term such
as osseointegration is motivated by the proven clinical results in the case of cranio-
facial implants and the hope to replicate these findings in the case of orthopaedic
implants in the future. However, from a strict histological point of view osseointe-
gration remains poorly defined.
Osseointegrated implants have resulted in a clinical breakthrough in two differ-
ent clinical applications in the craniofacial skeleton: one of these being oral implants,
irrespective of whether treating total or partial edentulousness [19, 20]; the other
being skin penetrating extra-oral implants. The clinical results of screw-type, c.p.
titanium oral implants in mandibles or maxillas for 5 years or more of follow up
have been in the 90–99% range[21]. The results of skin penetrating implants in the
temporal bone region have been similar, but not in the orbit region, where the host
bed has been irradiated. Now, 20 years since their clinical introduction in 1977,
permanent skin-penetrating, osseointegrated, screw-shaped titanium implants are
regarded as routine clinical treatments for facial disorders or certain types of hear-
ing impairments [22, 23].
Press-fit fixation represents one approach to the osseointegration of orthopedic
implants. The design of press-fit joint replacements is based on three dimensional
geometric data with the intention of fitting the implant as closely as possible to the
host bone. The objective of this design approach is to transfer load across the
implant-bone interface to as wide an area of the bone as possible [24]. In theory,
press-fit fixation may lead to osseointegration of the implanted device. However, as
it is difficult to mimic precisely the resulting intravital loading patterns, osseointe-
gration of initially stable press-fit components is threatened by subsequent bone
remodeling processes. Too stiff a device may cause ‘stress shielding,’ leading to
bone resorption. Conversely, too high local stresses may lead to pressure necrosis
and resorption prior to remodeling. Finally, bone resorption may lead to local inter-
facial movements: these predispose to soft-tissue formation and may cause a subse-
quent failure of fixation.
Polymers are not one hundred percent stable under biological conditions, leading
to highly variable clinical durability. Ultrahigh molecular weight polyethylene
(UHMWPE), although relatively stable, has shown poor outcomes when press-­
fitted in knee replacement arthroplasties [25]. The poor fixation of such devices may
3  Hard Tissue Response 587

relate both to intrinsic properties of the polymers involved as well as to differences

in their elastic moduli from that of bone. For instance, the Young’s modulus of
UHMWPE is approximately 2% of that of cortical bone (0.3 GPa vs. 17 GPa). This
leads to a quite different loading pattern for both the polymer and the surrounding
tissue than that encountered in the case of c.p. titanium, which is approximately 7
times as stiff as cortical bone (127 GPa vs. 17 GPa).
Ceramics are still stiffer, with moduli up to 30 times that of cortical bone. Bulk
ceramics, such as aluminum oxide (alumina) are well tolerated by bone but are
generally insufficiently strong and tough to serve as load bearing implants, espe-
cially in the presence of tensile or bending loads. Calcium phosphates, such as cal-
cium hydroxyapatite (CaHAP), although much weaker still, are an interesting class
of biomaterials due to their assumed capability for ‘bone bonding’. Søballe et al.
[26] has observed that the addition of a CaHAP coating induces proliferating bone
to bridge gaps in the bone-implant interface, in the presence of dynamic loading,
which would be filled with soft tissue around uncoated metallic implants. He also
suggests that such coatings enhance bone growth from osteopenic tissue, utilizing
an experimental animal arthritis model with substantial pre-implantation loss of
bone density.
In experimental studies, c.p. titanium has been demonstrated to induce a stronger
bone response than most other pure metals or alloys, including Ti6Al4V [4, 27].
However, there is substantial evidence that CaHAP leads to a still more rapid heal-
ing response [28]. This may be due to a direct positive influence on interfacial bone
from the calcium phosphate material and/or a relatively rough surface topography
resulting from the manufacturing process in combination with the reduction of
metal ion release by the presence of the CaHAP coating. Long term experimental
and dental clinical data from CaHAP coated implants has been disappointing [29].
However, CaHAP remains a very interesting biomaterial with efforts underway to
explore functional improvement through changes in crystallinity, coating thickness,
method of application, etc.
Retrieval data [30–32] have contributed to our current knowledge of oral
implants. Steflik and co-workers [32] reviewed 51 retrieved oral implants of differ-
ent designs. They claimed that implants inserted 10 years ago or more fail generally
due to loss of bone support and other biological features while more recently placed
oral implants also fail secondary to intrinsic biomaterial failure such as implant
fracture. One hundred stable (uncoated) Brånemark System® ‘Nobelpharma’
implants were retrieved and studied at our laboratories: 33% were removed because
of therapy-resistant pain or progressive bone resorption, 26% were removed after
death from unrelated causes, 24% were fractured implants and 17% were removed
for psychological reasons. The implants had been in situ and functioning for 1–18
years. There was on average 82% bone-to-implant contact and a similar percentage
of bone within the threads of these retrieved implants (evaluated over the three best
consecutive threads on both sides of the implant). Whenever possible, the entire
bone to implant contact percentages were also calculated and found to be ~70%. In
almost every case, there was >60% bone to implant contact. In fact, such extensive
588 T.O. Albrektsson

bone to implant contact percentages may represent a histological correlate to

osseointegration.

3.5  How Bone-Biomaterial Interfaces Fail

The implant-bone interface can fail for various reasons but the most common is so-­
called ‘aseptic loosening’: i.e. loosening not associated with infection. The caus-
ative factors for aseptic loosening may be classified as mechanical or biological. For
example, early post-operative failures of cemented, cement-free (ingrown) as well
as osseointegrated implants may be attributed to lack of initial fixation due to
mechanical failure of the bone-cement (when present) or bone-implant interface.
Such failures may relate to incidents of overload. However, early failures may also
occur subsequent to overheating of the tissues during surgery, leading to bony
necrosis, collapse and subsequent mechanical loosening. Overheating due to a com-
bination of surgical trauma and polymerization heat release may occur at bone-­
cement interfaces; surgical trauma, with associated heating, alone may prevent bone
and other tissue elements from invading cement-free implant-bone interfaces,
whether designed for ingrowth or osseointegration. Biological failure is also possi-
ble in the longer term, associated with generation or release of cytotoxic products,
such as cement monomer or metallic corrosion products or through induction of
specific immune responses [33] (Figure 3.3).
In clinical use, cylindrical shaped oral implants which lack any additional reten-
tion features, such as threads, grooves, etc., are prone to gradual failure due to ongo-
ing bone ‘saucerization’ (gradual loss of bone at the implant-bone-mucosal boundary)
[21]. In the case of orthopaedic implants with medullary structural elements (such as
a femoral stem), Willert and Semlitsch were first to propose that bone loss occurred
secondary to biological response to small particles, such as wear debris. [34]
Macrophages (MP) and foreign body giant cells (GC) ingest these undigestible par-
ticles of metal, polymer or ceramic and release factors which stimulate osteolytic
activity by cells in membranes associated with the implant-bone interface (Figure 3.3)
[35]. With modern histological staining techniques, especially the use of oil red 0
[33], it has now become possible to appreciate the large amounts of small and sub-
micron UHMWPE particles found in the vicinity of loose orthopaedic joint replace-
ment implants. Many investigators believe that biological response to these particles
is the leading biological cause of osteolysis leading to gradual, late implant failure.

3.6  Conclusions

The major advantage of the osseointegrated interface is its remodelling capacity.

Gradual adaptation to load has been verified in retrieval studies; the implants are
known to have almost no bone to implant contact during the first few weeks after
 3  Hard Tissue Response

Fig. 3.3  Theoretical failure modes for an osseointegrated calcium hydroxyapatite-coated femoral stem of a total hip replacement prosthesis [29; by permis-
sion]. Key: GC: multinucleated giant cell; MP: macrophage; Ca: calcium; Ph: phosphorous; OAF: osteoclastic activating factor; IL1: interleukin 1, TNF: tumor
necrosis factor; PGE2: prostaglandin E2.
589
590 T.O. Albrektsson

placement but then demonstrate an increasing amount of interfacial bone tissue. The
main proportion of osseointegrated implant failures occur during the first one or two
years and result from failure to achieve a proper osseointegration. This trend is quite
different from most orthopaedic studies reported in the literature, which show clini-
cal failure rates increasing with time. Cemented arthroplasties of today have resulted
in a significantly better clinical outcome than cement-free ones. It must be pointed
out that the results of orthopaedic implants published in the literature are generally
based on the number of revisions alone. This means that the actual number of suc-
cessful arthroplasties is lower than the figures quoted in the literature. Furthermore,
there is a patient drop-out of more than 20% in most clinical reports. From a bio-
logical viewpoint it is important to strive for improved cement-free implants so that
their clinical results will at least match those of the cemented arthroplasties. New
types of osseointegrated hip and knee constructions have been designed and are
presently in clinical trials. Altering prosthetic design (in comparison with current
practice), and improving surgical instruments and procedures may well overcome
some of the hurdles in the development of new osseointegrated arthroplasty devices.

Additional Reading

Lee, A.J.C. and Ling, R.S.M. (1984): loosening, in Complications of Total Hip
replacement, (ed. R.S.M. Ling), Churchill-Livingstone, Edinburgh, pp. 110–145.
Still useful account of the biological and mechanical features of loosening lead-
ing to clinical failure of the various interfaces formed between rigid biomaterials
and bone. Extensive bibliography.
Manley, M.T. (1993): Calcium phosphate biomaterials: A review of the litera-
ture, in Hydroxylapatite Coatings in Orthopaedic Surgery, (eds G.T.R. Geesink and
M.T. Manley), Raven Press, New York, pp. 1–24.
The title speaks for itself. Additional chapters in the same book provide experi-
mental and clinical results of osseointegration, specifically adhesion fixation.
Spector, M. (1987): Historical review of porous-coated implants. J. Arthroplasty,
2(2), 163–177, 1987.
Historical review of experimental and clinical results of ingrowth fixation, with
extensive bibliography.

References

1. Frost, H.M. (1989) The biology of fracture healing. Clinical Orthopaedics and Related
Research, 248, 283–293.
2. Morberg, P. (1991) On bone tissue reactions to acrylic cement. PhD Dissertation, Biomaterials
Group, University of Göthenberg, Sweden, pp. 1–142.
3  Hard Tissue Response 591

3. Donath, K., Laass, M. and Günzl, H.-J. (1992) The histopathology of different foreign body
reactions in oral soft tissue and bone tissue. Virchows Archiv A Pathologia Anatomica, 420,
131–137.
4. Johansson, C. (1991) On tissue reactions to metal implants. PhD Dissertation, Biomaterials/
Handicap Research, University of Göteborg, Göteborg, Sweden, pp. 1–232.
5. Black, J. (1988) Orthopaedic Biomaterials in Research and Practice. New York, Churchill
Livingstone, pp. 267–284.
6. Steinemann, S.G., Eulenberger, J., Maeusli, P.-A. et al. (1986) Adhesion of bone to titanium.
Adv. in Biomaterials 6, 40–44.
7. Ryd, L. (1986) Micromotion in knee arthroplasty. A Roentgen stereophotogrammetric analysis
of tibial component fixation. Acta Orthop. Scand., Suppl. 220, 1–80.
8. Malchau, H., Herberts, P. and Anhfelt, L. (1993) Prognosis of total hip replacement in Sweden.
Follow-up of 92,675 operations performed in 1978–1990, Acta Orthop Scand., 64, 497–506.
9. Collier, J.P., Mayor, M.B., Chae, J.C. et al. (1988) Macroscopic and microscopic evidence of
prosthetic fixation with porous coated materials. Clin. Orthop. Rel. Res., 235, 173–180.
10. Willems, W.J., Eulderbrink, F., Rozing, P.M. et al. (1988) Histopathologic evaluation in failed
Gerard double cup arthroplasty. Clin. Orthop. Rel. Res., 228, 123–133.
11. Engh, C.A., Bobyn, J.D. and Glassman, A.H. (1987) Porous coated hip replacement. The fac-
tors governing bone ingrowth, stress shielding and clinical results. J. Bone Joint Surg., 69B,
45–55.
12. Cook, S.D., Thomas, A.K. and Haddad, R.J. (1988) Histologic analysis of retrieved human
porous coated joint components. Clin. Orthop. Rel. Res., 234, 90–101.
13. Brånemark, P.-I., Hansson, B.-O., Adell, R. et al. (1977) Osseointegrated implants in the treat-
ment of the edentulous jaw. Scand. J. Plastic Reconst. Surg., Suppl 16, 1–116.
14. Albrektsson, T. (1979) Healing of bone grafts. In vivo studies of tissue reactions at autografting
of bone in the rabbit tibia. PhD Dissertation, Laboratory for Experimental Biology, Göteborg
University, Göteborg, Sweden, pp. 1–90.
15. Pilliar, R.M. (1986) Implant stabilization by tissue ingrowth. In Tissue Integration in Oral and
Maxillofacial Reconstruction, D. van Steenberghe (ed.), Amsterdam, Excerpta Medica,
pp. 60–76.
16. Wennerberg, A. (1995) On surface topography of implants. PhD Dissertation, Biomaterials/
Handicap Research, University of Göteborg, Göteborg, Sweden, pp. 1–202.
17. Brånemark, P.-I. (1985) Introduction to osseointegration, in Tissue Integrated Prostheses (eds
P.-1. Brånemark, G. Zarb and T. Albrektsson), Quintessence Co, Chicago, pp. 11–76.
18. Zarb, G. and Albrektsson, T. (1991) Osseointegration – A requiem for the periodontal liga-
ment? – An editorial. Int. J. Periodontal and Restorative Dent., 11, 88–91.
19. Albrektsson, T., Dahl, E., Enbom, L. et al. (1988) Osseointegrated oral implants. A Swedish
multicenter study of 8139 consecutively inserted Nobelpharma implants. J. Periodont., 59,
287–296.
20. Lekholm, U., vanSteenberghe, D., Herrmann, I. et al. (1994) Osseointegrated implants in par-
tially edentulous jaws: A prospective 5-year multicenter study. Inter. J. Oral & Maxillofacial
Implants, 9, 627–635.
21. Albrektsson, T. (1993) On the long-term maintenance of the osseointegrated response.

Australian Prosthodontic J., 7, 15–24.
22. Tjellström, A. and Granström, G. (1994) Long-term follow-up with the bone anchored hearing aid:
A review of the first 100 patients between 1977 and 1985. Ear Nose and Throat J., 2: 138–140.
23. Jacobsson, M., Tjellström, A., Fine, L., et al. (1992) A retrospective study of osseointegrated
skin-penetrating titanium fixtures used for retaining facial prostheses. Int. J. Oral &
Maxillofacial Implants, 7, 523–528.
24. Poss, R., Robertson, D.D., Walker, P.S. et al. (1988) Anatomic stem design for press-fit and
cemented application. In Non-cemented Total Hip Arthroplasty, R. Fitzgerald, Jr (ed.),
New York, Raven Press, pp. 343–363.
25. Freeman, M.A.R., McLoed, H.C. and Levai, J.P. (1983) Cementless fixation of prosthetic
components in total arthroplasty of the knee and hip. Clin. Orthop. Rel. Res., 176, 88–94.
592 T.O. Albrektsson

26. Søballe, K. (1993) Hydroxyapatite ceramic coating for bone implant fixation. Acta Orthop.
Scand. (Suppl 225), 64, 1–58.
27. Han, C.H., Johansson, C., Wennerberg, A. et al. (1995) A quantitative comparison of commer-
cially pure titanium and titanium-6 aluminum-4 vanadium implants in rabbit bone. Proc. Fifth
Biomaterials Club Meeting, Ischgl, Austria, Albrektsson T. and Tjellstrom, A. (eds), p. 25.
28. Søballe, K., Hansen, E.S., Rasmussen, H.B. et al. (1993) The effect of osteoporosis, bone
deficiency, bone grafting, and micromotion on fixation of porous coated vs. hydroxyapatite-­
coated implants. In Hydroxyapatite Coatings in Orthopedic Surgery, Geesink, R.G.T. and
Manley, M.T. (eds), New York, Raven Press, pp. 107–136.
29. Gottlander, M. (1994) On hard tissue reactions to hydroxyapatite-coated titanium implants.
PhD Dissertation, Biomaterials/Handicap Research, University of Göteborg, Göteborg,
Sweden, pp. 1–202.
30. Lemons, J.E. (1991) Bone-biomaterial interfaces in retrieved implants, in The Bone Biomaterial
Interface (ed. J.E. Davies), University of Toronto Press, Toronto, pp. 419–424.
31. Albrektsson, T., Eriksson, A.R., Friberg, B., et al. (1993) Histologic investigations on 33
retrieved Nobelpharma implants. Clinical Materials 12 (1), 1–9.
32. Steflik, D.E., Parr, G.R., Singh, B.B. et al. (1994) Light microscopic and scanning electron
microscopic analyses of dental implants retrieved from humans. J. Oral Implantol., 20(1),
8–24.
33. Campbell, P. (1995) On aseptic loosening of total hip replacements: The role of UHMWPE
wear particles. PhD Dissertation, Biomaterials/Handicap Research, Göteborg University,
Göteborg, Sweden, pp. 1–225.
34. Willert, H., Semlitsch, M. (1976) Reactions of the articular capsule to joint prostheses. In
Biocompatibility of Implant Materials, D.F. Williams (ed.), London, Sector Publishing,
pp. 157–169.
35. Goldring, S.R., Schiller, A.L., Roelke, M., et al. (1983) The synovial-like membrane at the
bone-cement interface in loose total hip replacements and its proposed role in bone lysis.
J. Bone Joint Surg., 65A, 575–584.
Chapter 4
Immune Response

K. Merritt

4.1 Introduction

There is increasing concern about the role of specific immune response to implanted
materials. This section discusses the general principles governing immune responses
and outlines techniques for their measurement and evaluation. This is a necessarily
brief presentation of the issues, and the reader is encouraged to pursue the topic
through relevant references provided for further study.

4.2 Overview of the Specific Immune Response

The specific immune response is the normal response of vertebrates when a foreign
substance is introduced into the body. This is a desirable protective response which
detoxifies, neutralizes, and helps to eliminate such substances.
However, in some cases, responses to seemingly innocuous substances may
cause harm to the host. Such effects are usually termed allergic or hypersensitivity
responses. The responses have been classified into four types: Type I, Type II, Type
III, Type IV.
These four responses share elements of a common mechanism, triggered by the
presence of a foreign material termed an antigen. The antigen is initially processed
by a cell, usually either a monocyte or macrophage, but occasionally a skin den-
dritic cell also referred to as an antigen processing cell (APC). The APC engulfs the
antigen, processes it (usually by enzymatic digestion or attempted digestion), and
transfers or presents it to another cell, usually a lymphocyte termed a T helper cell.
The T helper cell then presents the processed antigen to another T lymphocyte,

K. Merritt (*)
17704 Stoneridge Dr., Gaithersburg, MD 20878, USA

© Springer Science+Business Media New York 2016 593

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_31
594 K. Merritt

called the T cytotoxic cell, or to a B lymphocyte. The receiving cell, whether T- or

B-type, initiates a response for interaction with the processed antigen, forming a
less biologically active complex. In the former case, the immune response is a Type
IV or cell mediated immunity while in the latter case, the final result is release of
free antibody, which may lead to a humoral Type I, II, or III response. T-cell
responses result in accumulation of T-cells at the site where the foreign material is
present. B-cells remain at remote depots of lymphoid tissue while the antibody cir-
culates and appears at the site of the foreign material.
The main features of the four types of responses are described in the following
table:

Type: Antibody: Cells involved: Mediator: Consequences:

I IgE B-lymphocytes Histamine, Itching, rhinitis, vascular
vasoactive amines collapse
II IgG, IgM B-lymphocytes Histamine, Vascular collapse
vasoactive amines
III IgG, IgM Bclymphocytes Vasoactive amines Pain, swelling, some vascular
plugging and collapse
IV none T-lymphocytes Cytokines Pain, swelling

Both the T- and B-cells are small lymphocytes, which circulate in the blood, and
are found in lymphoid tissue. They arise from a common stem cell and then undergo
processing in the thymus to become T-cells or an unknown site, probably the bone
marrow, to become B-cells. They are difficult to distinguish and a great deal of work
has been done to facilitate identification of these cells in order to elucidate specific
immune responses. Identification of T-cells has been greatly aided by the recogni-
tion that there are unique substances on the cell surface that can be recognized by
the use of monoclonal antibodies generated using murine cells. These antigens are
referred to as cluster differentiation markers (CDs) and are given numbers. There
are a large number of them, more are being identified, and the importance of each is
being evaluated. However, all T-cells express CD3 which is consequently referred
to as a Pan(= all) T-cell marker. CD2 may also be a Pan T-cell marker. Additionally,
the T helper cell expresses CD4 while the T cytotoxic cell expresses CD8. New CDs
continue to be identified and their importance evaluated.
The B-cell retains small amounts of antibody on its surface which can be used for
identification. The B-cell response results in further differentiation to plasma cells
which produce antibody in large quantities. Antibody is a four chain immunoglobu-
lin which has combining sites specific for a single antigen. Antibody is soluble,
circulates in the plasma, and the plasma or serum drawn from a vertebrate, or the
antibody produced in cell culture for monoclonal antibodies, can be stored frozen
and will last virtually forever.
There are five general classes of immunoglobulins produced as antibodies. In order
of concentration in normal human blood from highest to lowest are IgG, IgM, IgA, IgE,
and IgD. IgD is a surface marker on B-cells and will not be discussed further. IgE is the
4 Immune Response 595

antibody associated with Type I responses. IgA is a secretory immunoglobulin, found

in high concentrations in saliva, GI contents, and milk and in the associated organs. IgG
and IgM are present in high concentrations in blood and are excellent antibodies for
immunological testing since they can participate in Type II and III responses.
Type I responses and Type II responses have the same consequences, but the mecha-
nisms are different. Type I responses are best known as hay fever and dust allergies and
are immune responses to these antigens mediated by a skin fixing antibody (IgE).
Type II responses involve the reaction of IgG, rarely IgM, antibody with a cell
surface antigen. The result is lysis of the cell with release of products. If the cell
contains vasoactive amines, then the consequences are signs and symptoms described
here. This is commonly seen in allergies to drugs which bind to blood platelets.
Type III responses are referred to as immune complex reactions and occur when
both antigen and antibody are present in large quantities at the same time. In the
normal immune response, the antigen is processed, the immune response is initi-
ated, and the antigen disappears shortly. However, if the antigen is persistent, then
significant amounts of immune complexes can form, plug small vessels, and result
in tissue or organ damage.
Type IV responses are those most usually associated with chronic presence of a
foreign body, such as an implanted biomaterial. They are typified by the common
contact dermatitis caused by poison ivy.
In each case, an antigen stimulates the immune response and the immune
response in turn reacts specifically with the antigen. T-cells, B-cells and circulating
antibody each recognize only one antigen. For a substance to be antigenic, it must
be foreign to the host, of high molecular weight (> 3000), and processable by an
APC. However, some small substances become antigenic by binding to larger car-
rier molecules, usually proteins, found in the host. Such a small substance is called
a hapten and the immune response is mounted to the hapten-carrier complex.

4.3 Detection of Antibody

To evaluate whether or not a patient has produced antibody against a substance,

such as an implant, a blood sample needs to be drawn, tested and the results com-
pared to those from controls. A pre-implantation sample is an ideal control but usu-
ally this is not available. The choice of appropriate controls is a major problem. The
test procedure itself requires a known positive control (often difficult to obtain for
evaluating responses to biomaterials), and a known negative control (usually saline,
tissue culture media, or bovine or equine serum used for tissue culture). The con-
trols for the patient population under study are generally obtained from normal
individuals without implants and without underlying disease, individuals with the
disease (e.g. arthritis) but without the implant (e.g. total joint replacement), indi-
viduals with the implant and without problems, and individuals with diagnosed
implant failure. The results for all of these need to be analyzed in order to ascertain
596 K. Merritt

Fig. 4.1 Standard

immunoassay for antibody.
An antigen is fixed
(attached) to a solid
support and binds a
specific antibody from
solution. The bound
antibody is detected by
binding a second, labeled
(enzyme, isotope, etc.)
antibody.

whether or not antibody is increased in the patient population and whether or not its
presence can be correlated with failure of the device.
Most current tests are based on immobilization of the antigen to a solid sur-
face such as polystyrene. The general procedure is outlined in Figure 4.1.
Detection of bound antibody involves the use of enzyme (EIA or ELISA assays)
or a radioisotope labeled antibody (RIA). (The tests are simple to perform, but
great care is needed to wash away excess materials and use the appropriate con-
centrations. Individuals wishing to initiate such testing procedures are encour-
aged to attend workshops (often offered by American Society for Microbiology,
American Type Culture Collection, etc.) or obtain training in a clinical immunol-
ogy lab, and also explore detailed manuals on procedures.) Antiserum to human
antibodies, (all classes or individual types) with enzyme labels, can be purchased
from several biological supply houses. Isotope labeled antibody is available to
licensed laboratories.
Problems with interpretation of results:
Positive results
If negative controls are negative and patient samples are positive, then the interpre-
tation is that the patient made antibody to that antigen. However, the question
remains as to whether or not the antigen is the correct one and if contaminating
antigens are contributing to the reaction. This is difficult to determine but inhibition
studies with well characterized antigens are helpful.

In vivo testing
Type I sensitivity is associated with histamine and vasoactive amine release with
vascular responses. Often such sensitivity is determined by ‘patch’ (skin) testing.
The positive reaction will occur in a few minutes as a wheal and erythematous (red-
dening) flare response in the skin. This test is hampered by availability of antigen
for testing but the biomaterial applied directly to skin or a mucosal surface may
stimulate a response. Caution needs to be taken in interpreting these tests, however,
since this is a short term application.
4 Immune Response 597

Negative results
Negative results are the desired response in evaluating biomaterials for clinical use,
but they are also difficult to evaluate. If there are no responses recorded except in the
positive control sample, this is indicative that patient does not produce antibody to
that antigen. Again the question of appropriateness of the antigen and its concentra-
tion on the solid support remains.

In vivo testing
A negative skin test presents the same problem: Was the antigen correct and/or was
it applied to a correct site?

4.4 Detection of Cell Mediated Responses (Type IV)

The procedures for detecting cell mediated responses are much more complicated
and difficult than for antibody determination. Most of the assays require the use of
living cells and thus tests must be done shortly after obtaining the cells. Controls
may have to be done at a different time which complicates the comparisons.
The two most common in vitro test procedures used are lymphocyte proliferation
and cell migration inhibition tests. The basic theory behind both of these is that
T-cells have receptors on their surface which will each respond to a specific antigen.
In the course of the response, soluble substances (cytokines or lymphokines), prin-
cipally blastogenic factor and migration inhibition factor, which act on other cells,
including other T-cells, are produced and released.

Blastogenic factor (lymphocyte transformation factor)

This causes other lymphocytes to transform and divide. If cell counts are done, an
increased number is seen. If a radioactive cell proliferation precursor (such as triti-
ated thymidine) is added to the culture, the isotope is taken up by dividing cells and
the ‘counts’ increase. This test, usually called LTT for lymphocyte transformation
test, requires living cells to produce and respond to the factor. This takes several
days, with 7 days being the general time for response to antigen. Some control
stimulants (mitrogens) such as PHA (phytohemagglutinin) act in 4–5 days.
Interpretation of the tests presents the same problems as with tests for antibody: Are
the appropriate controls included, what was the antigen used, were the cells viable;
if the results are negative, were the culture conditions correct?

Migration inhibition factor

Migration inhibition factor (MIF) is produced by the stimulated T-cell and acts on
cells that are normally motile. The two cell types are the monocyte/macrophage line
and polymorphonuclear leukocytes (polys). Thus the test, usually called LIF test
598 K. Merritt

(leukocyte inhibition test) requires living lymphocytes and living migrating cells
which may be obtained from fresh, whole blood. The results of the LIF test are avail-
able in 18–24 hours. blood does not contain enough monocytes to evaluate inhibition
of their migration and this indicator cell is usually obtained from the peritoneal cav-
ity of other animals, typically the mouse or the guinea pig. It is possible to stimulate
human lymphocytes in culture for 24–48 hours and then harvest the culture fluid and
add it to the macrophages obtained from the animals. Migration (or inhibition of
migration) of cells is observed by placing them into tissue culture medium solidified
with purified agarose and observing them with a microscope in 18–24 hours or by
packing them into capillary tubes and observing migration from the tubes in a few
hours giving the appearance of ‘ice cream cones’. The factor for LIF and the one for
MIF may be slightly different and thus the two separate terms remain.

Choice of test
There is no evidence that one test is better than the other. The choice is usually
based on laboratory preference. This author uses blood cells and migration in aga-
rose since it requires little equipment, is rapid, and small breaks in aseptic procedure
are tolerated. The evidence is that the stimulated T-cells produce a group of cyto-
kines. Thus detecting migration inhibition factor or lymphocyte proliferation
implies the presence of the others and one is not more specific or sensitive.

Direct testing for lymphokines or cytokines

It would seem from the above studies that the ideal test would be for the lymphokines
without the use of a viable indicator cell. These substances are produced in low levels,
and thus the cell based assays such as LTT and LIF or MIF are required. Although ELISA
based or RIA based assays can be used to detect and quantitate cytokines, reagents are not
yet available for these human lymphokines (LIF, MIF, LTT) specifically.

Testing for production of cytokines

There is a current explosion of studies on production of cytokines in response to bio-
materials, especially to particles produced by wear and degradation. Thus it seems
pertinent to discuss these assays briefly. The assays are generally done by inhibition of
ELISA or RIA based assays. This concept is shown in Figure 4.2. Many of these assays
are available as complete ‘kits’ from emerging biotechnology companies. (There are
many companies, each doing only a few of the cytokines; thus no specific reference
will be given. A glance at the methods section in an article reporting on assays will give
their source. In addition, a glance at a biotechnology based journal and Science will
give ads, or an immunologist’s mail box will give you a plethora of choices). These
tests are easy to perform, but have technical difficulties and must be done carefully.
Interpretation of the results involves being sure that the control studies were done suc-
cessfully. Again, as with most studies, a positive test can be evaluated but a negative
test is difficult to interpret until you are confident of the laboratory doing the testing.
4 Immune Response 599

Fig. 4.2 Quantitation of antigen by competitive (inhibition) assay. An antigen is fixed (attached)
and binds a specific antibody from solution. However, additional antigen is provided in solution
and antibody binding to the bound antigen is reduced in direct proportion to the concentration of
free antigen. The bound antibody is then detected by binding a second, labeled (enzyme, isotope,
etc.) antibody.

In vivo studies
The classical test for cell-mediated immunity (CMI) (Type IV) in the early days of
immunology was the skin test. Antigens were applied to the skin or injected under
the skin and swelling was observed in 24–72 hours if there was CMI. This was dif-
ferentiated from Type I, IgE mediated responses, which occur rapidly (in minutes)
and usually disappears in 24 hours. CMI begins in 24 hours, has a swelling with
some resemblance to the wheal but does not show a flare. This author does not advo-
cate skin testing for responses to biomaterials since the testing is difficult to do
correctly, has the potential for producing sensitivity to the test agents and the results
are easily misinterpreted. In general, the use of the actual biomaterial (rather than
extracts, corrosion products or constituents) is contraindicated since mechanical
irritation may be read as a false positive or or the biomaterial may fail to release the
antigen and thus produce a false negative.
Skin testing is an excellent diagnostic procedure in patients with clinical suspicion
of hypersensitivity. However, skin testing with haptens, such as metal ions, involves a
risk of sensitization. For the immune response to be detected, the hapten must bind to
dermal cells or proteins. However, such binding produces a complete antigen which
may stimulate an immune response. Since this immune response takes time to develop,
the skin test will be negative, but future tests may then be positive. Thus repeated test-
ing increases the likelihood of inducing sensitivity and should be avoided. Bulk bio-
materials will probably not give adequate release of soluble materials in the 24–48
period of testing so may produce false negative results.

Histochemical techniques
There are a number of studies now examining tissues removed from sites adjacent
to implants. It is possible, using immunological techniques resembling those out-
lined in Figure 4.1, to identify cell types and perhaps cell products produced at the
600 K. Merritt

site. The major interest is in the detection and typing of lymphocytes by use of
antisera prepared against the CD markers described earlier. The same type of assay
is being initiated for detection of the cytokines in the tissue. The techniques are
simple but not all antisera work and thus a variety of antisera are used. The required
antisera are available commercially.
Interpretation of the results is again a problem. Tests using ‘home made’ monoclo-
nal antibodies are suspect until the antisera is made available to other investigators for
conformation. The use of well characterized antisera from companies which supply to
others is better at this stage. Since the tissue section is examined and scored by an
observer, the ‘data’ from these studies are not really available for analysis by the
reader. Computerized image analysis techniques are still not widely used. Thus, in
evaluating the results, possible bias of the investigator must be taken into account.

4.5 Detection of Immune Responses to Haptens

Detection of immune responses to haptens is the same as that described above, but
there are some special techniques now being used. A hapten-carrier complex can be
prepared in vitro by combination in solution with a large protein such as albumin or
a smaller molecule such as glutathione. These can then be used to coat a solid sub-
strate. Alternatively, the protein carrier can be coated onto a substrate, the hapten
added, and then the assay performed. This probably increases the amount of hapten
that is available for antibody binding and minimizes that which is lost in the tertiary
folding of the protein.

4.6 Human Immune Response to Materials

4.6.1 Latex

The term ‘latex’ actually is confusing since the name is given to some materials
because of the way they are processed and not because of their source. The biomate-
rial latex used to fabricate gloves, condoms, etc. is a rubber (elastomer) extracted
from a plant (Hevea brasiliensis). As such, there is a great deal of antigenic protein
contamination. Allergies to latex are usually of the Type I, IgE mediated response,
with an immediate reaction (in minutes) that can be life threatening. Since latex is
encountered in many household objects such as household gloves, balloons, etc.,
sensitivity to it is a frequent pre-existing condition. Latex material cleaned of protein
seems to be nonallergenic. Other types of immune response to latex have not been as
frequent or of much concern. Latex is not used as a long term implanted material and
thus the long term responses are not noted. The population at greatest risk are the
health care workers with the increased use of examining and surgical gloves.
4 Immune Response 601

4.6.2 Collagen

This is another material that is an extract of material of natural origin with bovine
and orvine skin or tissue the favored source. This is a foreign protein and thus
capable of stimulating a variety of immune responses. Antibodies of IgE, IgM, and
IgG classes have been observed. Cell mediated immune responses have also been
observed. As with latex, the important precaution is to remove as much foreign
material as possible. Since collagen across mammalian species has a similar struc-
ture, it is possible to remove contaminating proteins and leave a relatively nonal-
lergenic material. Chemical treatment and cross linking of the collagen can further
reduce antigenicity. Collagen products need to be carefully evaluated for their abil-
ity to initiate immune responses, but it is possible to produce safe products.

4.6.3 Synthetic polymers

The use of chemically defined synthetic polymers is associated with minimal human
immune responses. These materials are based on carbon, hydrogen, nitrogen and
oxygen which are basic building blocks of the biological system. Thus the genera-
tion of antigenic material would be unlikely. Nevertheless, there are some poly-
meric materials with additional chemical moieties that are of concern.

A. Polysiloxane (silicone elastomer)

There is abundant lay press comment and little scientific material on this topic. It is
apparent that there can be binding of silicone to foreign protein with stimulation of the
immune response. The use of a simple hapten test, as described in section 4.5, has added
greatly to our knowledge of this. It is also apparent that silicone gel is a potent adjuvant
enhancing immune responses to unrelated materials. Whether this is of relevance to the
use of gel filled implants remains unknown. The possible stimulation of related and
unrelated immune responses remain a major concern in the use of these materials.

B. Polyurethanes
This is a complex group of polymers. Their propensity to stimulate an immune
response is very small since there are few molecular groups which would be per-
ceived as foreign by the host, perhaps explaining why clinical immune responses to
polyurethanes have not been reported.

C. Poly(methyl)methacrylate
Acrylics are in widespread use in activities of daily living. As with metals, there are
documented cases of contact dermatitis from the use of these materials, especially
self curing glues containing methacrylate monomer that is very skin sensitizing,
602 K. Merritt

usually stimulating a Type IV response. The use of these materials for implants
generally exposes the patient to the monomer for only a brief period of time as the
bone cement or dental acrylic cures in situ. Acrylics which are fully polymerized
before use will not be associated with an allergic response. Reports of sensitization
responses of patients to acrylics are rare and the health care workers at most risk are
the personnel, such as the surgeon and dental laboratory technician, handling the
monomer frequently.

D. Metals
A number of metallic elements and alloys are used extensively in implants, external
medical devices and are encountered in activities of daily living. Allergy to metals
as a contact dermatitis (Type IV response) is well known in individuals in contact
with metal salts, corroding metals, and jewelry or snaps and fasteners. Reactions
have been seen to metals used in dental, orthopaedic, and general surgery. The con-
tact dermatitis from topical use resolves when the device is removed. The role of the
immune response in reactions to metals implanted into the deep tissue remains con-
troversial. Cell mediated immune responses have been associated, in some studies
but not in others, with pain and swelling at the implant site and loosening of the
device. Antibodies to metals in patients with metallic implants have recently been
reported, but again the consequences of this response remain unknown. Concern
remains about the chronic use of metals that are known human sensitizers, such as
chromium, nickel and cobalt.

4.7 Consequences of an Immune Response

The immune response is apparently intended to neutralize, detoxify, and help elimi-
nate a foreign material. However, sometimes the immune response can inadver-
tently cause harm. This will be discussed in various categories in the next section.
1. Damage to the implant. The inflammation which is part of the initiation of the
immune response is an oxidative response. Materials subject to oxidative attack,
such as polyethylenes and polyurethanes, may be degraded.
2. Damage to adjacent tissues. Products, particularly from Type II and IV responses,
may initiate swelling and other vascular responses at the site. This may resolve
with no further harm, or it may cause tissue necrosis and/or loss of tissue mass
with loosening or movement of the device.
3. Systemic responses. Immune responses of Type I and II generate vasoactive sub-
stances which may circulate and cause vascular collapse. This is seen in response
to latex materials and drugs which bind to platelets, mast cells, or eosinophils,
resulting in an immune response and release of these vasoactive substances.
4. Autoimmune diseases: This is the most controversial area of consequences of
immune response to implants. Autoimmune diseases are technically the result
4 Immune Response 603

of an immune response to host tissue. Autoimmune diseases such as arthritis,

glomerulonephritis, etc., occur in individuals with an unknown cause of
onset although some have an association with a preceding infection (especially
streptococci). Proving cause and effect is an epidemiological problem with sur-
veys of large populations. It is important to refine our immunological testing
techniques to prove cause and effect associated with implants and do thorough
epidemiological studies.
These responses, if present in clinical populations, may arise from several mech-
anisms. The two most likely ones associated with the use of implants are (i) binding
of the material to host tissue making it a foreign substance such as with hapten-
carrier complexes or (ii) altering the host tissue through folding of proteins, degra-
dation of cells or proteins thus making then antigenic for the host. This is the main
issue now with the silicone breast implants. This is difficult to prove and massive
studies as controlled as possible are needed.

4.8 Conclusions

There has been a rapid growth in our knowledge of the immune response and how to
evaluate and quantitate it. As these techniques are applied to the population in contact
with biomaterials, we will learn more about its importance in performance of the
material. We will also learn more about how to process the materials to minimize the
immune response. However, it is important to remember that the immune response is
a protective response and detection of immune responses to products of biomaterials
does not necessarily indicate clinical problems. On the other hand, implants are for-
eign material and will stimulate host responses, some of which may cause harm to the
host or implant.
Thus the important issue is to distinguish between those immune responses
which are normal and help to render antigens less biologically active from those
which are harmful to the host. It is clear that IgE (Type I) responses are harmful.
Detection of a Type I response to products of biomaterials indicates potential
problems in the clinical setting. Responses of the IgG type are generally protec-
tive and may not be predictive of further problems unless there is continual release
of antigenic material leading to a Type III-response. Biomaterial wear and degra-
dation products that bind to platelets or mast cells pose a potential for adverse
Type II responses.
The most commonly observed is the Type IV (cell-mediated) response. This
is a protective response in walling off the stimulating agent and in killing cells
which have the antigen on the surface, thus eliminating the antigen. However,
the tissue reaction accompanying this response may cause harm to the host
through soft and hard tissue necrosis. The difference between protection and
allergy from Type IV responses is still unclear and careful evaluation of patients
is required.
604 K. Merritt

Additional Reading

General Immunology
Golub, E.S. Green, D.R. (1990) Immunology, A Synthesis, 2nd Edition, Sinauer Associates, Inc.,
Sunderland: Good general text.
Roitt, I. (1971) Essential Immunology, Blackwell Scientific Publications, London: Good descrip-
tion of types I–IV reactions.
Annual Review of Immunology, Annual Reviews Inc., Palo Alto, CA: Yearly publication with
timely reviews.
Immunology Today: Elsevier Science Inc. Tarrytown, NY: Monthly: Good review articles.

Antigen Presentation
Celada, A. and Nathan, C. (1994) Macrophage activation revisited. Immunology Today, 15, 100–
102: good review of macrophages.
Chicz, R.M. and Urban, R.G. (1994). Analysis of MHC presented peptides: applications in autoim-
munity and vaccine development. Immunology Today, 15, 155–160: good review on a compli-
cated subject.

CD markers
Kemeny, D.M., Noble, A. Holmes, B.J. et al. (1994) Immune regulation: a new role for the CD8+
T cell. Immunology Today, 15, 107–110: Good description of the function of the CD8+ T-cell
which is a key cell in Type IV responses.
Sclossman, S.F., Boumsell, L., Gilkes, L.W. et al. (1994) CD antigens 1993. Immunology Today,
15, 98–99: good description of recently reported CDs.

Cytokines/interleukins
Miyajima, A., Kitamura, T., Harada, N. et al. (1992) Cytokine receptors and Signal Transduction.
Annual Reviews of Immunology, 10, 295–331: Review of function and methods of
stimulation.
Mizel, S.B. (1989) The interleukins. FASEB J. 3, 2379–2388: good detailed review.

Effects of cytokines/interleukins
Goldring, M.B., and Goldring, S.R. (1990) Skeletal tissue response to cytokines. Clin. Orthop. Rel.
Res., 258, 245–278: review of cytokines and orthopaedics.
Stashenko, P., Obernesser, M.S., and Dewhirst, F.E. (1989) Effect of immune cytokines on bone.
Immuno Invest., 18, 239–249: one of the few reviews focussing on bone.

Immune response to metals/metallic implants

Agrup, G. (1968) Sensitization induced by patch testing, Brit. J. Derm., 80, 631–634: points out
problem of routine skin testing in nonallergic individual.
Benson, M.K.D., Goodwin, P.G. and Brostoff, J. (1975) Metal sensitivity in patients with joint
replacement arthroplasties, Brit. Med. J., 4, 374–375: third of the original 1975 articles point-
ing to a possible problem. Skin test used.
Black J. (1988) Does corrosion matter? J. Bone Jt. Surg., 70B (4), 517–520: discusses issues of
importance of understanding corrosion, minimizing it, and recognizing it can be important for
the patient.
Brown, G.C., Lockshin, M.D., Salvati, E.A. et al., (1977) Sensitivity to metal as a possible cause
of sterile loosening after cobalt-chromium total hip-replacement arthroplasty, J. Bone Joint
Surg., 59A(2), 164–168: Denies existence of metal allergy in orthopaedics. Complete misinter-
pretation of data. Uses negative results of an invalid test to draw conclusions. Limited patient
population deliberately selected to prove lack of allergy.
4 Immune Response 605

Burholm, A.; AI-Tawil, N.A.; Marcusson, J.A. et al. (1990): The lymphocyte response to nickel
salt in patients with orthopedic implants. Acta Orthop. Scand., 61(2): 248–250: Example of use
of LTT test.
Elves, M.W., Wilson, J.N. and Kemp, H.B.S. (1975) Incidence of metal sensitivity in patients with
total joint replacements. Brit. Med. J., 4, 376–378: Second one of the original 1975 articles
pointing to a possible problem. Skin test used.
Evans, E.M., Freeman, M.A.R., Miller, A.J. et al. (1974) Metal Sensitivity as a Cause of Bone
Necrosis and loosening of the Prosthesis in Total Joint Replacement, J. Bone and Joint
Surg., 56B (4), 626–642: One of the original articles pointing to a possible problem. Skin
test used.
Goh, C.L. (1986) Prevalence of contact allergy by sex, race, and age. Contact Dermat., 14, 237–
240: discusses normal population
Grimsdottir, M.R., Gjerdet, N.R. and Hensten-Pettersen, A. (1992) Composition and in vitro cor-
rosion of orthodontic appliances. Am. J. Orthod. Dentofac. Orthop., 101, 525–532: Discusses
sensitivity and stainless steels. Release of nickel related to many metallurgical aspects and not
necessarily to nickel content of the metal.
Lalor, P.A., Revell, P.A., Gray, A.B. et al. (1991) Sensitivity to titanium. J. Bone and Joint Surg.,
73B(1), 25–28: Description of possible titanium sensitivity. Patch test vehicle of unknown
composition, larger cobalt-chromium component than titanium component in device. Of inter-
est and important, but not conclusive.
Menne, T.; and Maibach, H.I. (1989) Systemic contact allergy reactions. Immunol Allergy Clin.
N.A., 9, 507–522: Discusses extension from contact dermatitis to systemic reactions.
Merritt, K. (1984) Role of medical materials, both in implant and surface applications, in immune
response and in resistance to infection. Biomaterials, 5 (1), 47–53.: Review article. Out of date
now but covers literature through 1983.
Merritt, K. (1986) Biochemistry/hypersensitivity/clinical reactions. in: Lang B, Morris, J. and
Rassoog, J. (eds) Proc. International Workshop on Biocompatibility, Toxicity, and
Hypersensitivity to Alloy Systems used in Dentistry. Ann Arbor, U. MI; pp 195–223.: Review
article. Covers the literature through 1984. Good discussion of the problem in the discussion
section of the symposium
Merritt, K. (1986) Chapter 6. Immunological testing of biomaterials, Techniques of Biocompatibility
Testing, D.F. Williams (ed.), Vol. II, CRC Press, Boca Raton: Description of possible test
methods.
Merritt, K.; and Brown, S.A. (1980): Tissue reaction and metal sensitivity: An animal study. Acta
Orthop. Scand. 51 (3), 403–411: Example of use of LIF test.
Rostoker, G., Robin, J., Binet, O. et al. (1987) Dermatitis due to orthopaedic implants. J. Bone
Joint Surg., 69A, 1408–1412: Example of a reaction to implant.
Rudner, E.J., Clendenning, W.E., Epstein, E. et al. (1975) The frequency of contact dermatitis in
North America 1972–1974. Contact Derm. 1, 277–280: Incidence of contact dermatitis.
Shirakawa, T., Kusaka, Y. and Morimoto, K. (1992) Specific IgE antibodies to nickel in workers
with known reactivity to cobalt. Clin. Exp. Allergy, 22 (2), 213–218: Measuring IgE and nickel
cobalt interactions.
Trobelli, L., Virgili, A., Corassa, M. et al. (1992) Systemic contact dermatitis from an orthodontic
appliance. Contact Dermatitis, 27, 259–260: Example of reaction to dental application of
metals.
Yang, J., and Merritt, K. (1994) Detection of antibodies against corrosion products in patients after
Co-Cr total joint replacements. J. Biomed. Mater. Res., 28, 1249–1258: Method for measuring
antibodies to metals

Immune response to latex, collagen, silicones

Belsito, D.V. (1990) Contact urticaria caused by rubber. Analysis of seven cases. Dermatol. Clin.
8, 61–66: Questions whether increased demand for latex may have decreased quality with more
allergens leachable.
606 K. Merritt

Hanke, C.W., Higley, H.R., Jolivette, D.M. et al. (1991) Abscess formation and local necrosis after
treatment with Zyderm or Zyplast collagen implant. J. Amer. Acad. Dermatol. 25, 319–326:
Deals with some adverse responses to collagen materials which may be related to the immune
response. Points to possible problems.
Meade, K.R., Silver, F.H. (1990) Immunogenicity of collagenous implants. Biomaterials, 11, 176–
180: Discusses immunogenicity problem and cross linking. Good place to begin reading.
Naim, J.O., Lanzafame, R.J. and van Oss, C.J. (1993). The adjuvant effect of silicone gel on anti-
body formation in rats. Immunol Inv., 22, 151–161: Shows that the gel is better than Freund’s
adjuvant in stimulating the response to BSA in rats. Caution on use of gel.
Slater, J.E. (1989) Rubber anaphylaxis. New Eng. J. Med. 320, 1126–1130: Good methods. Good
literature review, real cases reacting to anaesthesia mask.
Sussman, G.L., Tarlo, S. and Dolovich, J. (1991). The spectrum of IgE responses to latex. J. Am.
Med. Assoc. 265, 2844–2847: Latex gloves on health workers causing allergic responses in
patients. Can do skin test with latex to check patients or use non-latex gloves.
Warpinski, J.R., Folgert, J., Cohen, M. et al. (1991) R.K. Bush. Allergic reaction to latex: a risk
factor for unsuspected anaphylaxis. Allergy Proc. 12, 95–102: Clinical symptoms of Type I
allergy. Identifies IgE antibodies against latex (gloves, balloons, condoms). IgE against pro-
teins from latex.
Wolk, L.E., Lappe, M., Peterson, R.D. et al. (1993) Immune response to polydimethylsiloxane
(silicone): screening studies in a breast implant population. FASEB J., 7, 1265–1268: Very
important study with a good technique. Hopefully more studies will be done with this tech-
nique. Valid test of antibody to silicone.

Oxidative damage of implants

Carter, W.O., Narayanan, P.K. and Robinson, J.P. (1994). Intracellular hydrogen peroxide and
superoxide anion detection in endothelial cells. J. Leukocyte Biol. 55, 253–258: Good method
for detecting H2O2 release and superoxide production. Example of biologically produced oxi-
dizing species.
Kao, W.J., Zhao, Q.H., Hiltner, A. et al. (1994) Theoretical analysis of in vivo macrophage adhe-
sion and foreign body giant cell formation on polymethylsiloxane, low density polyethylene,
and polyetherurethanes. J. Biomed. Mater. Res., 28 (1), 73–80: Recent article on macrophages
on polymers and references some articles on oxidative events.
Kaplan, S.S., Basford, R.E., Jeong, M.H. et al. (1994) Mechanisms of biomaterial induced super-
oxide release by neutrophils. J. Biomed. Mater. Res., 28, 377–86: Discusses the release of
reactive oxygen species stimulated by biomaterials. Not all materials are activating.

Consequences of immune responses to materials

Angell, M. (1994) Do Breast implants cause systemic disease? Science in the court-room New
Eng. J. Med., 330 (24), 1748–1749: Editorial in response to article by Gabriel et al. (1994)
Reiterates the necessity of doing detailed studies. Indictment of patients, manufacturers and
government jumping to conclusions using inadequate data. Doesn’t say how we get the ade-
quate data though.
Gabriel, S.E., O’Fallon, W.M., Kurland, L.T. et al. (1994) Risk of connective tissue diseases and
other disorders after breast implantation. New Eng. J. Med., 330 (24), (1697–1702): Excellent
study. Shows problems of doing studies on long term consequences. Example of how it ought
to be done.
Chapter 5
Cancer

M. Rock

5.1 Introduction

The widespread use of temporary and permanent implants in the post World War
II era has had a dramatic impact on the practice of medicine and on the life of dis-
abled and ill individuals. Nowhere has this been more obvious than in the frequent
use of implants to stabilize fractures and replace diseased joints which has revolu-
tionized orthopedic practice and afforded millions of patients levels of function
that previously could not be achieved. Although the metal alloys used in these
implants exhibit excellent resistance to corrosion, oxidation of these large compo-
nents ultimately produce free ions, chlorides, oxides, and hydroxides which, in
combination with particulate metal matter released by wear and fretting, are
released into the surrounding environment. Efforts to improve these alloys have
included compositional as well as processing changes. Additionally, modifications
have been made to the plastic articulating components in efforts to produce a much
more consistent ultrahigh molecular weight polyethylene. The perceived need to
improve implant wear and corrosion resistance and alter design has been largely
motivated by the excessive soft tissue staining noted by orthopedic surgeons at the
time of removal or revision of clinically failed joint arthroplasty. The presence of
particulate metal matter, polyethylene, and even fragments of polymethyl methac-
rylate in local tissue has been confirmed histologically and by direct analysis [1–
4]. In spite of all of the modifications made in implant composition, implant
fixation, and articulation, biomaterial degradation and release of these products
persist [4–7].

M. Rock (*)
Department of Orthopaedics, Mayo Clinic,
200 First Street Southwest, Rochester, MN 55905, USA

© Springer Science+Business Media New York 2016 607

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_32
608 M. Rock

5.2 Release and Distribution of Degradation Products

The body’s response to the local presence of debris is dependent on the size, amount,
and composition as well as rate of accumulation. The body attempts to neutralize
these foreign particles by precipitating granulomatous foreign body reactions and/
or removal through local lymphatic channels. If the local accumulation of debris
exceeds the body’s ability to neutralize and/or transport, the debris migrates from
the site to remote areas including the bone-implant interface, possibly contributing
to if not initiating the phenomena of loosening and osteolysis (Figure 5.1).
Of equal or possibly even greater concern is the detection of metal ions, metallic
debris, polyethylene, and even methylmethacrylate in areas remote from the implant
including circulating serum, excreted urine, and regional draining lymph nodes.
Elevated serum levels of metal ions consistent with the composition of the implanted
alloy have been confirmed in animal models [8] and in human patients after total hip
arthroplasty [9]; identifying serum levels of cobalt, chromium, nickel, and titanium
that are two and three fold higher than preoperative determinations. These figures
represent significant elevations both over means for contemporary control groups
and for the individual patients before operation [9]. However, since they are within
the widely accepted normal range for these metallic ions in the unimplanted human
controls, it is assumed that toxic levels of these foreign materials do not materialize.
However, when analyzing the serum to urine concentration in patients subjected to
conventional total hip arthroplasty, it has become apparent that the urinary concen-
tration of chromium in particular does not rise with the same magnitude and time
course as the serum level [9]. This observation parallels that made in the accounts
of industrial overexposure to Cr6+ and suggests that metal ions accumulate in organs
and tissues remote from implantation. Such accumulation is unlike that resulting
from normal systemic circulation. This was previously suggested by Steineman [10]
who calculated the potential release of metallic ions of 0.15 to 0.3 micrograms per
cm2 per day which would translate to between 11 and 22 milligrams per year in
patients with total hip replacement. This incidentally coincides with or exceeds the
total body burden of such metallic ions in a 70 kilogram man [11].

Figure 5.1 Tissue Reaction to Implant Degradation Products.

5 Cancer 609

Evidence for metallic debris accumulating in distant organs has also been con-
firmed by Langkamer et al. [12] who identified wide spread dissemination of par-
ticulate wear debris from hip prosthesis to lymph nodes, liver, and spleen. He
reported increases above normal levels in these organs of 30 fold for aluminum,
chromium, and iron in the lymph node, and 10 fold in the spleen and liver.
These findings suggest that concentrations of metal ions and debris at remote
sites may reach such proportions as to precipitate altered cellular dynamics in
organs principally of the lymphoreticular system. It would only be logical to assume
that local concentrations of such debris at the site of implantation would be even
higher, although attempts at quantifying the effects of local concentrations have
been fraught with inaccuracies mostly due to sampling error and the need to distin-
guish between bioavailable and non bioavailable metal species.
What is potentially more disturbing is that these figures for serum concentrations
and the identification of this debris in remote organs have come primarily from
patients who have received conventional polymethyl methacrylate cemented com-
ponents. With the advent of using uncemented porous coated implants, particularly
in younger patients, these figures would be expected to increase, creating the dis-
tinct possibility of toxic levels in the serum, tissues and organs that will respond
with altered cellular dynamics and function.

5.3 Neoplasia

Perhaps one of the greatest concerns with debris dissemination locally and within
the systemic circulation is the possibility of inducing malignant neoplasia. This is
thought to be possible by one of two mechanisms:
(i) A ‘solid-state’ mechanism has been proposed, whereby a large foreign object
implanted in vivo possibly stimulates mutagenesis of local cells, thereby creat-
ing tumor by its mere presence. Most large foreign objects upon implantation
will initiate a very marked fibrous reaction. The cells within this fibrous reac-
tion ultimately mutate and become cancer growths.
(ii) The other possibility is that particulate metal matter or other debris have an
innate capacity, upon corrosion or dissolution, to induce cancer through a more
traditional chemical route.
Cancer, the end product of carcinogenesis, is the result of transformation of a
normal cells to ones which grow in an uncontrolled or malignant manner. Cancer is
a genetic disease, which may result from expression of genetic pre-dispositions
present from birth or from later insults to cells of many different types. In particular,
the phagocytosis or pinocytosis of foreign matter (in an attempt to neutralize or
eliminate it) may cause or precipitate malignant conversion. Such conversion, if not
lethal to the cell, may then persist through cell duplication, creating first a cluster of
cells with altered DNA and eventually a clinical malignant tumor. Malignancy is
characterized by rapid, uncontrolled growth, invasion in surrounding tissues and
seeding to form tumors (metastases) in other anatomical locations such as the lung.
610 M. Rock

Some of the more common malignant tumors of musculoskeletal origin are osteo-
sarcoma (OS) of bone and malignant fibrous histiocytoma (MFA) of soft tissue.
Osteosarcoma is the most common tumor of bone: it occurs in children, adoles-
cents and, less frequently, in adults. OS may also occur as a consequence of radia-
tion therapy or in Paget’s disease, an ostensibly benign bone embrittling disease of
the elderly. It frequently appears about the knee (distal femur; proximal tibia), and
in the proximal femur and proximal humerus.
MFA is the most common primary malignant tumor of soft tissues and can occur
in bone in adults over the ages of 50–55. The more common soft tissue type usually
involves the large muscular areas of the body, including the thigh, buttock and upper
arm and shoulder.

5.4 Evidence for Carcinogenicity of Implanted Materials

Well-documented cases of carcinoma and sarcoma have developed in refinery work-

ers who inhaled nickel and chromium and in miners who were exposed to iron or
even at local injection sites of iron dextran [13]. Aluminum has been linked to a
high rate of lung and bladder cancer in exposed individuals and titanium has been
associated with experimental induction of lymphoreticular tumors and leukemia.
Although the results have not been universally accepted, many animal experiments
have shown a direct correlation between the initiation of sarcomas and the injection
of particulate metal debris. This appears to be related to the concentration, as well
as the physical nature, of the metal implanted [14]. Metal ions, particularly cobalt,
chromium, and nickel, are known to induce infidelity of DNA synthesis by causing
the pairing of non-complimentary nucleotides and thereby creating a misinterpreta-
tion of the genetic code which may lead to neoplasia.
Furthermore, it must be remembered that particulate metal matter may not be the
only solid-form material that can be, and has been proven to be, carcinogenic in
appropriate environments. In 1954 long before the first total hip arthroplasty was
performed, Laskin [15] observed the carcinogenic capabilities of polymethylmeth-
acrylate after subcutaneous introduction of this material in mice. His conclusions
suggested that similar occurrences of tumor may appears in humans that were being
treated at that time with methylmethacrylate for dental deficiencies and that this
evolution of cancer may take up to 20 years of exposure given the proportional time
exposure before tumors were seen in the mice. A similar conclusion was reached
[16] on the use of polyethylene plastic before it was conventionally used in the
management of arthritic joints. Regardless of form, whether powder or large solid
segments, polyethylene plastic produced sarcomas in 25 percent and 35 percent of
rats, respectively. Their conclusions also suggested a latent period, after exposure,
of 20 years in humans before such an event could be expected to occur.
It is, therefore, with interest that investigators were forewarning the medical
community of the carcinogenic effect of metals and polymers years before the
development and introduction of joint replacement using these very same materials.
5 Cancer 611

In 1961, Sir John Charnley introduced total hip arthroplasty as an alternative in the
management of arthritic hips. No other orthopedic procedure has been adopted with
such enthusiasm. Thirty-five years later we are still witnessing an incremental
increase in the yearly utilization of this operation, attesting to the obvious success
associated with it. According to some investigators, we may be coming into an era
of increased tumor activity in the vicinity of or possibly remote from implantation
sites of these orthopedic appliances.

5.5 Case Reports of Implant Related Tumors

In 1976, Harris et al. [17] were the first to describe an aggressive granulomatous
lesion around a cemented femoral stem in a total hip replacement. This was a condi-
tion of localized tumor-like bone resorption that appeared radiographically as large
lytic defects within the femur, approximating the cement mantle of the implant.
Initially thought to be neoplastic, these lesions were surgically biopsied and found
to be consistent with well-organized connective tissue containing numerous histio-
cytes, monocytes, and fibroblastic reactive zones. Immunohistologic evaluation
revealed multinucleated giant cells and nonspecific esterasepositive monocyte mac-
rophages. These findings suggest a foreign-body type reaction, and with the subse-
quent isolation of polyethylene, polymethyl methacrylate, and metal debris, it was
theorized that these constituents of the construct likely migrated down around the
implant cement mantle in cemented prostheses and implant-bone interface in non-
circumferentially coated ingrowth implants. Such a reaction suggests an excessive
accumulation of debris at the site of articulation that surpasses the body’s ability to
neutralize and/or transport the material resulting in migration of debris to sites
remote from the source. This rapid appearance of bone loss radiographically which
is often associated with a deteriorating clinical course has been termed type-II asep-
tic loosening [17].
In 1978, two years after the recognition of pseudo tumors of bone induced by the
degradation products of total hip arthroplasty, Arden and Bywaters [20] (Table 5.1)
reported a case of a 56-year-old patient who developed a high-grade fibrosarcoma
of soft tissue 2.5 years after receiving a metal-on-metal McKee-Farrar hip prosthe-
sis. The tumor apparently did not have a direct association with the underlying bone
or any components of the total hip arthroplasty. No formal analysis of the tumor for
debris products was performed. This case drew attention to the possibility of tumors
being initiated in the presence of large orthopedic appliances. It was not until 1984
when this concept became fashionable in large part due to three articles that appeared
simultaneously in the Journal of Bone and Joint Surgery recounting two malignant
fibrous histiocytomas and one osteosarcoma at the site of a total hip arthroplasty
[21–23].
This sudden and rather unexpected evolution prompted editorials [24, 25] in the
same journal addressing the issue of sarcoma and total hip arthroplasty and encour-
aged the orthopedic community worldwide to report such cases to a central registry
612 M. Rock

Table 5.1 Malignancies Associated with Joint Replacements (published)

Time
Author Year Implant interval(yrs) Tumor type
Castleman and McNeely [18] 1965 Austin-Moore 1 M.F.H.*
Rushforth [19] 1974 McKee-Farrar 0.5 Osteosarcoma
Arden and Bywaters [20] 1978 McKee-Farrar 2.5 Fibrosarcoma
Bagó-Granell et al. [21] 1984 Charnley-Muller 2 M.F.H.
Penman and Ring [22] 1984 Ring 5 Osteosarcoma
Swann [23] 1984 McKee-Farrar 4 M.F.H.
Weber [26] 1986 Cemented TKA 4.5 Epithelioid sarcoma
Ryu et al. [27] 1987 Uncemented 1.4 M.F.H.
Vitallium**
Vives et al. [28] 1987 Charnley-Muller 2 M.F.H.
Van der List [29] 1988 Charnley-Muller 11 Angiosarcoma
Lamovec et al. [30] 1988 Charnley-Muller 11 Synovial sarcoma
Lamovec et al. [31] 1988 Charnley-Muller 10 Osteosarcoma
Tait et al. [32] 1988 Charnley-Muller 11 M.F.H.
Martin et al. [33] 1988 Charnley-Muller 10 Osteosarcoma
Haag and Adler [34] 1989 Weber-Huggler 10 M.F.H.
Mazabraud et al. [35] 1989 Unknown 9 Epidermoid sarcoma
Brien et al. [36] 1990 Charnley 8 Osteosarcoma
Troop et al. [37] 1990 Charnley-Muller 15 M.F.H.
Kolstad and Högstorp [38] 1990 Freeman TKA 0.25 Metastatic
adenocarcinoma
Jacobs et al. [39] 1992 AML cementless 0.5 M.F.H.
Solomon and Sekel [40] 1992 Charnley-M uller 7 M.F.H.
* M.F.H. = malignant fibrous histiocytoma.
** Trademark, Howmedica, Inc. (Cobalt-Chromium alloy).

to obtain more accurate figures on the incidence of such a problem. These tumors
occurred 2, 4, and 5 years after hip replacement that was performed with various
femoral and acetabular components, some with metal-on-metal articulations and
others with metal on polyethylene. In two of these cases, the proximal femur was
extensively involved with tumor that was in direct contact with the component. The
remaining case was a soft-tissue sarcoma not in direct approximation to the prosthe-
sis. Two of these tumors were malignant fibrous histiocytomas, one of bone and one
of soft tissue. The remaining tumor was osteosarcoma. In this particular case, there
was evidence of gray-brown pigmentation both intra- and extracellularly between
the tumor and femoral component. No formal metal analysis was performed. Three
additional cases emerged prior to 1988 at 15 months, 4.5 years, and 2.0 years after
implantation [26–28].
In 1988, five cases were reported occurring at 10 [29, 30] and 11[29, 31, 32]
years after implantation. The sarcomas included two osteosarcomas, two malignant
fibrous histiocytomas, and one synovial sarcoma. Two of these were soft tissue in a
location with no direct association with the implant, yet in the case reported by Tait
et al. [32] there was evidence of nickel within tumor cells. The remaining three
5 Cancer 613

Table 5.2 Malignancies Associated with Joint Replacements (unpublished)[41]

Author Year Implant Time interval (yrs) Tumor type
Harris 1992 Charnley 1 Chondrosarcoma
Surin 1992 Christiansen 9 Rhabdomyosarcoma
Lightowler 1992 Charnley 10 Osteosarcoma
Rees, Thompson, Burns 1992 Thompson 3 M.F.H.*
Nelson 1992 Muller 9 M.F.H.
Rock 1992 HG ingrowth 0.8 M.F.H.
Rock 1992 PCA TKA 1.2 Osteosarcoma
* M.F.H. = Malignant Fibrous Histiocytoma.

patients all had direct contact with either the cement or implant with the tumor
originating in bone.
In 1990 there were three additional reports in the literature which included an
osteosarcoma developing at the site of a Charnley total hip arthroplasty 8 years [36]
after implantation, malignant fibrous histiocytoma developing 15 years after a
Charnley-Muller total hip arthroplasty [37], and metastatic adenocarcinoma devel-
oping at the site of a Freeman total knee arthroplasty three months after implantation
[38]. In 1992, Jacobs et al. [39] presented a malignant fibrous histiocytoma develop-
ing one half year after implantation of a cementless AML total hip arthroplasty.
In that same journal volume, unpublished but submitted reports of five tumors
occurring around implants were brought to the attention of the orthopedic commu-
nity [41] (Table 5.2). These included malignant fibrous histiocytomas around a
Thompson and a Muller total hip arthroplasties, an osteosarcoma around a Charnley
total hip arthroplasty, a rhabdomyosarcoma of soft tissue in the vicinity of a
Christiansen total hip arthroplasty, and a chondrosarcoma developing in a patient
with Maffucci syndrome having a Charnley total hip arthroplasty. The intervals from
implantation to tumor detection were 9, 3, 10, 9 and 1 years respectively. To this, we
add two previously unreported additional patients, neither of whom had their joint
replacement done at the Mayo Clinic (Table 5.2). The first is that of a 79-year-old
man who nine months previously came to total hip replacement with an uncemented
Harris-Galante component who was found to have a large malignant fibrous histio-
cytoma engulfing the proximal femur and extending to the implant. There was, how-
ever, no evidence of any particulate debris within the tumor cells removed. The
second case was that of a 56-year-old man who developed a soft tissue osteosarcoma
14 months after a left total knee arthroplasty with conventional cemented compo-
nents. The tumor extended down to both the femoral and patellar components.

5.6 Critical Analysis of Tumors

As such, 28 tumors have been reported in direct contact or in close proximity to

joint arthroplasty. The vast majority of these appeared with total hip arthroplasty
[27] with a smaller contribution from total knee arthroplasty [3]. There have been
614 M. Rock

no reported cases of malignant degeneration occurring in the vicinity of total shoul-

der and/or total elbow arthroplasty. Of the reported 26 cases, 8 tumors were of soft
tissue origin, 19 were of primary bone pathology, and 1 metastatic gastric carci-
noma. The histogenesis of the soft tissue tumor included 3 malignant fibrous histio-
cytomas, 1 synovial sarcoma, 1 soft tissue osteogenic sarcoma, 1 fibrosarcoma, 1
epidermoid sarcoma and 1 rhabdomyosarcoma. The histogenesis of the primary
bone tumors included 10 malignant fibrous histiocytomas, 6 osteosarcomas, 1 chon-
drosarcoma, 1 angiosarcoma, 1 fibrosarcoma. Direct contact with the underlying
tumor was noted in 15 of the 19 cases in which sufficient information is known from
which to make such determinations. In three of the cases, particulate metal matter
was determined to be present in the tumor including one case of a soft tissue sar-
coma that appeared on image and exploration to be remote from the implant but had
obvious evidence of nickel present within the tumor cells.
Many of these tumors have not had an appropriate latent interval between
implantation and development to be seriously considered implant induced. Given
that the interval to tumor induction from bone stimulation should be at least as long
as the accepted five year interval from radiation therapy to sarcoma degeneration,
15 of the 28 patients would qualify, all of whom have had tumours around total hip
arthroplasties.
Apart from tumors developing at the site of prosthetic replacement, there have
been ten known malignant tumors that have developed at the site of previous inter-
nal fixation (Table 5.3). To date there have been no malignancies noted around a
titanium implant. The vast majority (> 80%) of malignancies both in the prosthetic
and internal fixation groups have occurred in the vicinity of Vitallium™ (cobalt-
chromium alloy) implants. This is not, however, to exonerate stainless steel because
tumors in the proximity of the implants made of this alloy have been reported in the
animal literature [52] as well as the human experience utilizing stainless steel as
fixation devices for traumatology [42, 44, 50, 51]. It is interesting to note that in
1976 veterinarians were encouraged within their own literature to report similar

Table 5.3 Malignancies Associated with Internal Fixation of Fractures

Author Year Implant Time interval (yrs) Tumor type
McDougall [42] 1956 Stainless steel 30 Ewings
Delgado [43] 1958 Unknown 3 Undifferentiated
Dube and Fisher [44] 1972 Stainless steel 36 Angiosarcoma
Tayton [45] 1980 Vitallium* 7.5 Ewings
McDonald [46] 1981 Vitallium 17 Lymphoma
Dodion et al. [47] 1982 Vitallium 1.2 Lymphoma
Lee et al. [48] 1984 Vitallium 14 M.F.H.*
Hughes et al. [49] 1987 Vitallium 29 M.F.H.
Ward et al. [50] 1990 Stainless steel 9 Osteosarcoma
Khurana et al. [51] 1991 Stainless steel 13 M.F.H.
* Trademark, Howmedica, Inc. (cobalt-chromium alloy).
** M.F.H. = malignant fibrous histiocytoma.
5 Cancer 615

experiences of tumors around implants nearly eight years before such concern was
voiced with the application of these same metallic alloys in humans [52].

5.7 Significance of Clinical Reports

As impressive as these cases may be, they must be put into perspective given the
global use of internal fixation and prosthetic devices. Approximately 300 000 to 350
000 total hip joint replacements are performed worldwide on a yearly basis [53]. It
can be assumed that approximately four million people will have had total hip
arthroplasties performed by the end of 1995. To date, there have been 28 reports of
malignant tumor arising in close proximity to these implants (25 total hip and three
total knee arthroplasties). No direct contact was noted in four. If we assume a mini-
mal latency of five years to suggest association between presence of implant and
tumor, 15 of the 28 could have association. As such, the incidence of sarcomas in
total joint replacement would be approximately 1 in 250 000. There are approxi-
mately 3000 new primary bone tumors and 5000 soft-tissue sarcomas in the United
States per year. This would give an incidence of approximately 1 in 100 000 for the
general population to develop a primary bone sarcoma and 1 in 40 000 to develop a
soft tissue sarcoma a year. This is obviously not stratified for age given that many
primary bone tumors develop in the second and third generation of life, yet it does
afford the opportunity of putting this rather unusual event in perspective.
The prevalence of osteosarcoma among the osseous malignancies in this series is
not entirely unexpected. Of the total osteosarcoma population 15 percent to 20 per-
cent occur after the age of 50 years. Most of these cases are superimposed on Paget’s
disease or in previously irradiated tissue, yet de novo cases of osteosarcoma do
occur in this age group. Malignant fibrous histiocytoma of bone is somewhat less
common. A review of the Mayo Clinic files reveals 71 cases with more than half of
these occurring after age 55. Malignant fibrous histiocytoma of soft tissue is the
most common soft-tissue sarcoma. It is not surprising, therefore, that two of six
soft-tissue tumors in the combined series are of this histogenesis. As such, the
distribution of sarcomas in the combined series could have been predicted from
general population data given the age of the patients and anatomical distribution.
There have been two separate reports that have critically analyzed the cancer
risk after total hip arthroplast [54, 55]. The combined person years of exposure after
operation between the two series was 20 015. The overall cancer incidence among
total hip replacement procedure in both series did not appear to be any different
than what was expected or anticipated. The cancer-observed/expected ratio was
especially low for the first two years following surgery in both series, implying that
patients undergoing this procedure are otherwise generally healthy. In both series,
the observed/expected ratio of developing lymphoma or leukemia was two to three
times higher in patients who had total hip arthroplasty. Additionally, there was a
two fold decrease in breast carcinoma among patients who had total hip
arthroplasty.
616 M. Rock

Of interest, Gillespie et al. [54] suggested a similar decrease in the incidence of

rectal, colon, and lung cancer among total hip arthroplasty patients. The results sug-
gest or are possibly compatible with the hypothesis of chronic stimulation of the
immune system, thereby potentially allowing for malignancies to occur within the
lymphoreticular system. We have already determined a predilection for particulate
metal matter to accumulate in the reticuloendothelial system [12]. This has been
further supported by studies in animals subjected to metal implants, especially those
containing nickel, in which there was an increase in malignancies of the lymphoreti-
nacular systemic [52]. Additionally, due to the added immune surveillance, tumors
of the breast, possibly colon, rectal, and lung may be decreased. A hyper immune
state is not unexpected given the dissemination of debris locally at implantation
sites as well as the well-recognized and documented capacity of this material to gain
access to the systemic and possibly storage sites including the reticuloendothelial
system. This trend obviously needs continued surveillance.
A recent extensive analysis of the cancer risk in a cohort of 39154 patients with
at least one hip replacement operation has been performed by the Swedish
Nationwide In-Patient Registry [56]. Patients were identified by means of a linkage
to the Swedish Cancer Registry. The overall results, although showing a significant
3% increase in cancer, were judged by the authors not to be of clinical significance.
Increases of cancer of kidney, skin and brain in women and of prostate in men were
found, accompanied by a decrease in gastric cancer for women. The study showed
no increase in lymphoreticular cancers as previously reported [54, 55] nor a decrease
in colon, breast or rectal cancers. The authors’ judgement is that the overall cancer
risk associated with total hip replacement arthoplasty is negligible and should not
distract from the obvious benefits of the procedure.
A similar extensive review of the relationship between metallic implants and
cancer in dogs was performed by Li et al. [57]. This case controlled study of 1857
dogs from 22 veterinary medical centers failed to reveal significant association
between stainless steel fracture fixation devices and the development of bone and
soft tissue sarcomas.

5.8 Summary

In summary, careful examination of the scientific and clinical literature suggests

that implant materials commonly used for fixation and joint reconstruction are not
entirely inert. Accumulation of particulate debris is to some extent going to occur in
all patients who have large prosthetic devices. This necessarily includes the distinct
possibility of systemic and remote site exposure to these foreign objects that the
body attempts to neutralize and excrete. Due to the heightened immunologic sur-
veillance and/or possible storage of particulate metal matter in sites remote from the
implantation site, patients with total hip arthroplasty may be at added risk for remote
malignancies, particularly of the lymphoreticular system. The incidence of primary
mesenchymal tumors in close proximity to implants appears to be consistent with
5 Cancer 617

the incidence in the general public. The frequency of occurrence and the associated
individual and group risks of systemic and remote site malignancy remains
unresolved.

Additional Reading

Nyren, O., McLaughlin, J.K. et al. (1995) Cancer risk after hip replacement with
metal implants: A population based study. J. National Can. Inst. 87, 28–33.
An extensive review of risk of cancer in 39 154 total hip replacement patients
who appeared in the Swedish National Cancer Registry between 1965 and 1983. A
review of 60 cancer-specific sites showed an overall, not clinically significant
increase of 3% in incidence, slight increases noted for kidney cancer, prostate can-
cer (in men) and melanoma accompanied by a continuous decline in gastric cancer
for both sexes. This would appear to be the definitive review of the risk for develop-
ing cancers after total hip replacement arthroplasty.
Brand, K.G. and Brand, I. (1980) Risk assessment of carcinogenesis at implanta-
tion sites. Plastic Reconst. Surg. 66, 591–595.
Review of possible foreign body cancer initiation in humans based upon pub-
lished case reports. The authors conclude that, since the clinical use of prosthetic
implants has been popular for more than twenty years and since, extrapolating from
animal experience, at least 25% if not 50% of foreign body tumors should have
appeared by the time of their publication, there is little risk of such non-chemically
mediated tumors occuring in patients.
Gillespie, W.J., Frampton, C.M.A., Henderson, R.J. et al. (1988) The incidence
of cancer following total hip replacement. J. Bone Joint Surg., 70B, 539–542.
A New Zealand study of 1358 patients with total hip arthroplasty, for a total of
14 286 patient years. A significant increase in tumors of the hemopoetic and lym-
phatic systems, accompanied by a significant decrease of cancers of breast (in
women), colon and bowel was observed. The authors suggest that these data are
evidence for increased immune surveillance, allowing or precipitating hemopoetic
and lymphatic tumors but at the same time providing better resistance to the devel-
opment of soft tissue tumors. The first large scale study of this question.
Visuri, T. and Koskenvuo, M. (1991) Cancer risk after McKee-Farrar total hip
replacement. Orthopedics, 14, 137–142.
A study similar to that of Gillespie et al. but on a Finnish patient group (433
patients; 5729 patient years) leading to the same general conclusions. Includes a
historical discussion of the carcinogenic properties of various trace elements.
Jacobs, J.J., Rosenbaum, D.H., Hay, R.M. et al. (1992): Early sarcomatous
degeneration near a cementless hip replacement. A case report and review. J. Bone
Joint Surg., 74B, 740–744.
A review of a patient who developed a malignant fibrous histiocytoma at the site
of a cementless total hip replacement five months after implantation and suc-
cumbed of diffuse metastases, as is typical for such patients, within one year of
618 M. Rock

presentation. Includes an extensive review of world literature on sarcomas in the

vicinity of total hip replacement and suggest the need for an international registry
of such case reports.

References

1. Coleman, R.R., Herrington, S. and Scales, J.T. (1973) Concentration of wear products in hair,
blood, and urine after total hip arthroplasty. Brit. Med. J. 1 1527–1529.
2. Lux, F. and Zeisler, R. (1974) Investigations of the corrosive deposition of components of
metal implants and of the behavior of biologic trace elements in metallosis tissue by means of
instrumental, multi-element activation analysis. J. Radiol. Anal. Chem., 19, 289–297.
3. Rock, M.G. and Hardie, R. (1988): Analysis of local tissue response in 50 revision total hip
arthroplasty patients. Trans. Soc. Biomater., XI, 43.
4. Agins, H.J., Alcock, N.W., Bansal, M. et al. (1988) Metallic wear in failed titanium alloy total
hip replacements. A histological and quantitative analysis. J. Bone Joint Surg. 70A, 347–356.
5. Buchert, B.K., Vaughn, B.K., Mallory, T.H. et al (1986): Excessive metal release due to loos-
ening and spreading of sintered particles on porous coated hip prosthesis. Report of two cases.
J. Bone Joint Surg., 68A, 606–609.
6. Jacobs, J.J., Skipor, A.K., Black, J. et al. (1991), Release in excretion of metal in patients who
have a total hip replacement component made of titanium base alloy. J. Bone Joint Surg. 73A,
1475–1486.
7. Witt, J.D. and Swann, M. (1991) Metal wear and tissue response in failed titanium alloy total
hip replacements. J. Bone Joint Surg. 73B, 559–563.
8. Woodman, J.L., Jacobs, J.J., Galante, J.O., et al. (1984), Metal ion release from titanium-based
prosthetic segmental replacements of long bones in baboons. A long term study. J. Orthop.
Res., 1, 421–430.
9. Bartolozzi, A. and Black, J. (1985) Chromium concentrations in serum, blood clot and urine
from patients following total hip arthroplasty. Biomaterials, 6, 2–8.
10. Steineman, S.G. (1985) Corrosion of Titanium and Titanium Alloys For Surgical Implant. in:
Lutergering, G., Swicker, U., Bunk, W. (eds), Titanium, Science, and Technology, Volume 2,
DG für Metal. e.V. Oberuresel, Berlin, 1373–1379.
11. Harvey, A.M., Johns, R.S., Owens, A.H. et al. (1972) The Principles and Practice of Medicine.
New York: McGraw-Hill, pp. 68–102.
12. Langkamer, V.G., Case, C.P., Heap, P. et al. (1992) Systemic distribution of wear debris after
hip replacement. A cause for concern? J. Bone Joint Surg., 74B, 831–839.
13. Doll, R. (1958) Cancer of lung and the nose. Nickel workers. Brit. J. lndust. Med., 15,
217–223.
14. Swanson, S.A.V., Freeman, M.A.R. and Heath, J.C. (1973) Laboratory tests on total joint
replacement prosthesis. J. Bone Joint Surg., 55B, 759–773.
15. Laskin, D.M. (1954): Experimental production of sarcomas by methylacrylate implant. Proc.
Soc. Exper. Biol. Med., 87, 329–333.
16. Carter, R.L. and Rowe, F.J.C. (1969): Induction of sarcomas in rats by solid and fragmented
polyethylene; Experimental observations and clinical implications. Brit. J. Cancer, 23,
401–407.
17. Harris, W.H., Schiller, A.L., Scholler, J.M. et al. (1976) Extensive localized bone resorption in
the femur following total hip replacement. J. Bone Joint Surg. 58A, 612–618.
18. Castleman, B., McNeely, B.U. (eds) (1965) Case 38–1965, Case records of the Massachusetts
General Hospital. New Engl. J. Med., 273, 494–504.
19. Rushforth, G.F. (1947) Osteosarcoma of the pelvis following radiotherapy for carcinoma of
the cervix. Brit. J. Radiol. 47, 149–152.
5 Cancer 619

20. Arden, G.P. and Bywaters, E.G.L. (1978), in Surgical Management of Juvenile Chronic Poly
Arthritis. Arden, G.P., Ansel, B.M. (eds), Academic Press, London, pp. 269–270.
21. Bágo-Granell, J., Aguirre-Canyadell, M., Nardi, J. et al. (1984) Malignant fibrous histiocy-
toma of bone at the site of a total hip arthroplasty. A case report. J. Bone Joint Surg. 66B,
38–40.
22. Penman, H.G. and Ring, P.A. (1984) Osteosarcoma in association with total hip replacement.
J. Bone Joint Surg. 66B, 632–634.
23. Swann, M. (1984) Malignant soft tissue tumor at the site of a total hip replacement. J. Bone
Joint Surg. 66B, 269–231.
24. Hamblen, D.L. and Carter, R.L. (1984) Sarcoma and Joint Replacement (Editorial). J. Bone
Joint Surg. 66B, 625–627.
25. Apley, A.G. (1989) Malignancy and joint replacement. The tip of an iceberg? (Editorial).
J. Bone Joint Surg. 71B, 1.
26. Weber, P.C. (1986) Epithelioid sarcoma in association with total knee replacement. J. Bone
Joint Surg. 68B, 824–826.
27. Ryu, R.K.N., Bovill, E.G. Jr, Skinner, H.B. et al. (1987) Soft tissue sarcomas associated with
aluminum oxide ceramic total hip arthroplasty. A case report. Clin. Orthop. 216, 207–212.
28. Vives, P., Sevestre, H., Grodet, H. et al. (1987) Histiocytome fibreux malin du fémur après
prosthèses totale de hanche (Malignant fibrous histiocytoma of the femur following total hip
replacement). Rev. Chir. Orthop. 73, 407–409.
29. Van der list, J.J.J. (1988) Malignant epithelioid hemangioendothelioma at the site of a hip
prosthesis. Acta Orthop. Scand. 59, 328–330.
30;31. Lamovec, J., Zidar, A., Cucek-Plenicar, M. et al. (1988): Synovial sarcoma associated
with total hip replacement. A case report. Addendum: Osteosarcoma associated with a
Charnley-Muller hip arthroplasty. J. Bone Joint Surg. 70A, 1558–1560.
32. Tait, N.P., Hacking, P.M. and Malcolm, A.J. (1988) Case report, malignant fibrous histiocy-
toma occurring at the site of a previous total hip replacement. Brit. J. Radiol., 61, 73–76.
33. Martin, A., Bauer, T.W., Manley, M.T. et al. (1988) Osteosarcoma at the site of a total hip
replacement. J. Bone Joint Surg. 70A, 1561–1567.
34. Haag, M. and Adler, C.P. (1989) Malignant fibrous histiocytoma in association with hip
replacement. J. Bone Joint Surg. (Brit), 71B, 701.
35. Mazabraud, A., Florent, J. and Laurent, M. (1989) (A case of epidermoid carcinoma developed
in contact with a hip prosthesis)(Fr.)(au. transl.). Bull. Cancer, 76, 573–581.
36. Brien, W.W., Salvati, E.A., Healey, J.H. et al. (1990) Osteogenic sarcoma arising in the area of
a total hip replacement. A case report. J. Bone Joint Surg., 72A, 1097–1099.
37. Troop, J.K., Mallory, T.H., Fisher, D.A., et al. (1990) Malignant fibrous histiocytoma after
total hip arthroplasty: A case report. Clin. Orthop. Rel. Res.,253, 297–300.
38. Kolstad, K. and Högstorp, H. (1990): Gastric carcinoma metastasis to a knee with a newly
inserted prosthesis: A case report. Acta Orthop. Scand. 61, 369–370.
39. Jacobs, J.J., Rosenbaum, D.H., Hay, R.M. et al. (1992): Early sarcomatous degeneration near
a cementless hip replacement. A case report and review. J. Bone Joint Surg., 74B, 740–744.
40. Solomon, M.I. and Sekel, R. (1992) Total hip arthroplasty complicated by a malignant fibrous
histiocytoma. A case report. J. Arthroplasty, 7, 549–550.
41. Goodfellow, J. (1992) Malignancy and joint replacement (Editorial). J. Bone Joint Surg., 74A,
645.
42. McDougall, A. (1956) Malignant tumor at site of bone plating. J. Bone Joint Surg. 38B,
709–713.
43. Delgado, E.R. (1958) Sarcoma following a surgically treated fractured tibia: A case report.
Clin. Orthop. 12, 315–318.
44. Dube, V.E. and Fisher, D.E. (1972) Hemangioendothelioma of the leg following metallic fixa-
tion of the tibia. Cancer 30, 1260–1266.
45. Tayton, K.J.J. (1980) Ewing’s sarcoma at the site of a metal plate. Cancer 45, 413–415.
46. McDonald, I. (1981) Malignant lymphoma associated with internal fixation of a fractured
tibia. Cancer 48, 1009–1011.
620 M. Rock

47. Dodion, P., Putz, P., Amiri-Lamraski, M.H. et al. (1983) Immunoblastic lymphoma at the site
of an infected vitallium bone plate. Histopathol., 6, 807–813.
48. Lee, Y.S., Pho, R.W.H. and Nather, A. (1984) Malignant fibrous histiocytoma at site of metal
implant. Cancer 54, 2286–2289
49. Hughes, A.W., Sherlock, D.A., Hamblen, D.L. et al. (1987) Sarcoma at the site of a single hip
screw. A case report. J. Bone Joint Surg. 69B, 470–472.
50. Ward, J.J., Thornbury, D.D., Lemons, J.E. et al. (1990) Metal-induced sarcoma: A case report
and literature review. Clin. Orthop. 252, 299–306.
51. Khurana, J.S., Rosenberg, A.E., Kattapuram, S.V. et al. (1991) Malignancy supervening on an
intramedullary nail. Clin. Orthop. 267, 251–254.
52. Sinibaldi, K. (1976): Tumors associated with metallic implants in animals. Clin. Orthop. Rel.
Res., 118, 257–266.
53. Moss, A.J. (1991) Use of selected medical devices in the United States. Advance Data from
Vital and Health Statistics of the National Center for Health Statistics, 191, 1–24.
54. Gillespie, W.J., Frampton, C.M., Henderson, R.J. et al. (1988) The incidence of cancer follow-
ing total hip replacement. J. Bone Joint Surg. 70B, 539–542.
55. Visuri, T. (1992) Cancer risk after McKee-Farrar total hip replacement. Orthopedics, 14, 137–
142, 1992.
56. Nyren, O., McLaughlin, J.K. et al. (1995) Cancer risk after hip replacement with metal
implants: A population based study. J. National Can. lnst. 87, 28–33.
57. Li, X.Q., Stevenson, S., Hom, D.L. et al. (1993): Relationship between metallic implants and
cancer: A case-control study in a canine population. Vet. Comp. Orthop. Traumatol. 6, 70–74.
Chapter 6
Blood–material Interactions

S.R. Hanson

6.1 Introduction

The importance of understanding mechanisms of blood-material interactions is emphasized

by the increasingly widespread use of cardiovascular devices; hence, this field has been the
subject of intense inquiry as described in several excellent reviews [1–4]. Unfortunately, it
is still not possible to simply rank or classify materials with respect to their suitability for
particular blood-contacting applications. Nor is it possible to predict in any general way,
based on the properties of devices and of their blood-contacting surfaces, the behavior of
blood in contact with materials or the propensity of devices to produce clinically adverse
events. Despite many attempts to correlate biologic responses to physicochemical property
measurements, our success in understanding blood-material interactions, and the clinical
application of many blood-contacting devices, has been largely empirical. It is not appropri-
ate to discuss in detail this large and controversial literature, which has been reviewed else-
where [1, 2]. Rather, this section will focus on the available experimental data in humans, or
results which may likely be extrapolated to humans from relevant animal studies, that may
guide in the development of new designs for blood-contacting devices. Cardiovascular
device applications in humans have also been the subject of an excellent review [5].

6.2 Experimental Difficulties

Before summarizing relevant experimental findings, it is appropriate to review

briefly the theoretical and practical limitations to our understanding of blood–
material interactions.

S.R. Hanson (*)

Division of Hematology/Oncology, Emory University, AJ, Atlanta, GA 30322, USA

© Springer Science+Business Media New York 2016 621

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_33
622 S.R. Hanson

There are several factors which have precluded the rational engineering design of
devices based on first principles. While thousands of materials have been put forward as
‘biocompatible’ or non-thrombogenic, based on in vitro studies and animal testing, the
relevance of these tests for outcomes in humans remains uncertain. Device responses in
vivo depend upon actual device configuration and resulting flow geometry as well as
upon intrinsic materials’ properties. In many applications, mechanical and physical
property requirements may dominate materials’ selection. For example, membranes
used in dialyzers and oxygenators must be both solute and gas permeable; chronic vas-
cular grafts and heart valves must be mechanically durable and chemically stable for
years; heart assist devices require flexible pumping chambers. Thus, the use of in vitro
assays or simplified in vivo flow geometries, as in many animal models, has not proven
adequate to predict actual device performance in patients. Furthermore, most animals
and humans, as individuals, differ markedly from one another in both blood chemistry
and in blood response to foreign materials [6]. It is deemed unethical to perform screen-
ing tests in humans, hence relatively few materials have undergone clinical evaluation
and only limited human comparative data are available. In the case of chronic implants,
devices removed at autopsy provide only a single set of observations which cannot be
related to dynamic blood–material interactions prior to explantation.
Another limitation is our recognition that all blood–material interactions of
clinical consequence are preceded by complex interactions between the biomate-
rial surface and circulating blood proteins. Plasma contains more that 100 identi-
fied proteins with specific functions and varying biologic properties [7]. These
proteins interact with surfaces in a complex, interdependent and time-dependent
fashion that remains poorly understood, except in low dilution, simplified model
systems [8]. These reactions may vary from individual to individual depending
upon coagulation status, the use antithrombotic or other drug therapies, or the
administration of contrast media for fluoroscopic imaging. A partial listing of
variables which may affect device outcomes following blood exposure is given
in Table 6.1.
Despite these limitations, the design engineer may be guided by previous success-
ful applications of materials in a variety of device configurations, and by certain
generalizations which have resulted from these studies. Devices which are com-
monly used include catheters, cannulae, guide wires, stents, shunts, vascular grafts,
heart valves, heart and ventricular assist devices, oxygenators, and dialyzers. With
respect to these devices it is important to consider those events which can lead to
serious clinical complications. These complications include: (1) thrombosis, (2)
thromboembolism, (3) consumption (ongoing destruction) or activation of circulat-
ing hemostatic blood elements, and (4) activation of inflammatory and immunologic
pathways. An appreciation for the biologic mechanisms of these events is essential
for understanding the blood-compatibility of devices, and may be briefly described
as follows. Thrombus forms as the localized accumulation of blood elements on,
within, or associated with a device, and thrombus which is actively deposited can
accumulate to the extent of producing device dysfunction or blood vessel occlusion.
Interruption of normal blood flow may produce ischemia (relative lack of oxygen)
and infarction (tissue death due to total oxygen deprivation) in distal circulatory beds
leading to heart attacks and strokes. Thrombus structure may be complex, and is
6 Blood–material Interactions 623

Table 6.1. Variables Device Properties

influencing blood interactions
Size and shape
with cardiovascular devices
Surface composition
Texture or roughness
Mechanical properties
blood Flow Phenomena
Shear forces
Convection and diffusion of reactants, products,
cofactors and inhibitors
Disturbed flow and turbulence
blood Chemistry-related Effects
Coagulation status
Antithrombotic and other therapies
Contrast media
Other Variables
Duration of device blood exposure
Tissue injury
Infection

distinguished from that of whole blood clots which are often formed under static flow
conditions. Thus, clots are relatively homogeneous structures containing red blood
cells and platelets trapped in a mesh of polymerized protein (fibrin), while thrombus
formed under arterial flow conditions and high fluid shear rates (‘white thrombus’)
may be composed primarily of layers of fibrin and platelets (small procoagulant cells
occupying only about 0.3% of the total blood volume). Under conditions of low fluid
shear, as found in veins, thrombus may more closely resemble the structure of whole
blood clots (‘red thrombus’). Thromboembolism is the dislodgement by blood flow
of a thrombus which is then transported downstream, ultimately blocking vessels
which are too small for the thrombus to traverse. Thromboembolism is a common
cause of stroke (cerebrovascular infarction) and peripheral limb ischemia. Often the
balance between dynamic processes of thrombus deposition and its removal by
embolic and lytic mechanisms will produce platelet consumption (ongoing destruc-
tion) and a net reduction in circulating platelet levels. Other clotting factors may be
consumed as well [9]. Finally, certain devices, particularly those having large surface
areas, may activate enzyme systems (e.g., complement) leading to inflammatory or
immunologic responses [10]. With these issues in mind we will now review the per-
formance of various classes of biomaterials in actual device configurations.

6.3 Conventional Polymers

Conventional polymers, such as polyethylene (Intramedic™), poly (vinyl chloride)

(Tygon™), polytetrafluoroethylene (Teflon™), and poly (dimethyl siloxane)
(Silastic™), and certain polyurethanes, have been used for many decades in short-
term applications including catheters, cannulas, arteriovenous shunts for
624 S.R. Hanson

hemodialysis, and tubing components of extracorporeal circuits. When used for

periods of only a few hours, and most often in patients receiving systemic antico-
agulation agents, the performance of such materials has usually been clinically
acceptable. For example, although thrombus on angiographic catheters can be dem-
onstrated in about half of all cases, most thromboembolic or occlusive events are
clinically silent and significant complications occur in less than 1% of procedures
[5]. Even total occlusion of small peripheral veins, by short term catheters used for
venous sampling or drug administration, is usually inconsequential. However, lon-
ger-term indwelling catheters in a variety of configurations and polymer composi-
tions, particularly in infants and children, are now recognized to produce a significant
risk of thrombosis which can ultimately lead to organ or limb damage, and even
death [11]. Comparative, quantitative studies with different polymer formulations
remain to be performed in humans.
Polyurethanes, due in part to their flexibility and toughness, are perhaps the
polymer of choice for ventricular assist devices and blood pumps. Consequently,
they have received considerable interest as blood-contacting materials. In non-
human primates, those polyurethanes, such as Pellethane™, which exhibit the
most hydrophobic surface chemistry produce the least platelet consumption
[12]. In dogs, early platelet interactions with polyurethanes vary considerably
although relationships to polymer surface chemistry remain unclear [13]. Thus
while polyurethanes are chemically versatile and possess many desirable
mechanical properties, it is generally not possible to predict their biologic
responses in humans.

6.4 Hydrophilic Polymers

These materials, which preferentially adsorb or absorb water (hydrogels), were ini-
tially postulated to be blood compatible based on the view that many naturally occur-
ing phospholipids, comprising the cell membranes of blood contacting tissues, are
also hydrophilic. Thus, water, as a biomaterial, was expected to show minimal inter-
action with blood proteins and cells [14, 15]. Interestingly, in animal studies highly
hydrophilic polymers based on acrylic and methacrylic polymers and copolymers, as
well as poly(vinyl alcohol) are all found to consume platelets excessively although
they accumulate little deposited thrombus [12, 16]. The materials have variable
thrombogenicity, but little capacity to retain adherent thrombus, i.e., they shed depos-
ited platelets as microemboli. Thus, while surface-grafted hydrogels (which are
mechanically weak) are currently used to improve catheter lubricity and as reservoirs
for drug delivery, they have not received widespread application for improving
blood-compatibility.
Another hydrophilic polymer that has received considerable attention is
poly(ethylene oxide) [17, 18]. While poly(ethylene oxide) surfaces have been
shown (like hydrogels) to have relatively low interactions with blood proteins and
cells in in vitro studies and in some animal models, the suitability of such poly-
6 Blood–material Interactions 625

mers for actual device applications and long-term implants has not been
established.

6.5 Metals

Metals, as a class, tend to be thrombogenic, and are most commonly applied in situ-
ations requiring considerable mechanical strength, such as in the struts of mechanical
heart valves and as endovascular stents [3, 19] or electrical conductivity, as in pacing
electrodes. Stents are metallic mesh devices placed within blood vessels to preserve
vessel patency after procedures to expand the vessel lumen diameter (e.g., after bal-
loon angioplasty). Metals most commonly employed are stainless steel (316L type)
and tantalum; however, both are thrombogenic [19, 20]. Catheter guide wire throm-
bogenicity, although readily documented, has been less of a clinical problem because
of the usually short period of blood exposure involved in most procedures.
In early canine implant studies, the thrombogenicity of a wide series of metallic
implants was seen to be related to the resting electrical potential of the metal which was
generated upon blood contact [21]. Metals with negative potentials tend to be antithrom-
bogenic, while stainless steel tends to be neutral. Copper, silver, and platinum are posi-
tive and extremely thrombogenic. Indeed, the use of copper coils inserted into canine
arteries continues to be a widely used model for inducing a thrombotic response [22].
Theoretically, reduced thrombogenicity of metallic stents and heart valve com-
ponents might be achieved by thin film polymer coatings, although the clinical
effectiveness of this strategy has not been demonstrated. Thus, chronic systemic
anticoagulation is generally employed in patients with prosthetic heart valves (with
metallic components) and stents.

6.6 Carbons

Cardiac valves with components fabricated from low temperature isotropic carbons
(pyrolytic carbon) are successfully used clinically [23]. These materials are appropriate
for such applications as mechanical valves which require long-term chemical inertness,
smoothness, and wear-resistance. The reasons for the marked improvement in the per-
formance (reduced thrombosis and thromboembolic stroke rates) of these newer vs.
older style heart valves are not entirely understood, but are undoubtedly multifactorial
and related to improved patient management and valve design, as well as to the nature
of the carbon surface. The specific benefits conferred by pyrolytic carbons with respect
to blood cell and protein interactions, resulting in a very low frequency of clinical com-
plications, remain to be defined. The use of carbon coatings has been proposed for other
devices, i.e., vascular grafts, although such devices have not yet been used clinically.
626 S.R. Hanson

6.7 Ultra-thin Film Coatings

Polymeric thin films of widely varying chemical composition may be deposited onto
polymers, metals, and other surfaces using the method of plasma polymerization (also
called ‘glow-discharge’ polymerization) [24]. This method is advantageous since very
thin films (e.g., 100 nm) may selectively modify the surface chemistry of devices, but
not their overall mechanical properties or surface texture. Plasma polymers form
highly cross-linked, covalent, inert barrier layers which may resist the adsorption of
proteins, lipids, and other blood elements. Plasma reacted coatings, based on hydro-
carbon monomers such as methane, may produce durable diamond like coatings.
Plasma polymer coatings have been proposed for vascular grafts and stents, based on
promising animal studies [25], but are not used clinically at the present time.

6.8 Membranes

Blood contacting membranes are used for gas exchange (e.g., blood oxygenators) or
solute exchange (e.g., dialyzers). The large membrane surface area, which may
exceed 2 m2, and the complexity of cardiopulmonary bypass circuits can produce
consumption and dysfunction of circulating blood elements such as platelets, lead-
ing to an increased risk of bleeding as well as thromboembolism [26]. The activa-
tion of inflammatory and immune response pathways (complement system) by
dialysis and oxygenator membranes may also produce significant complications
[27]. Complement activation by dialysis membranes may be related in part to the
availability of surface hydroxyl groups, particularly on cellulosic membranes.
Complement activation may be greatly attenuated by the use of other membrane
materials such as polysulfone and polycarbonate. Complement activation by bioma-
terial membranes has been reviewed [27].

6.9 Biological Surfaces

The use of biological and bioactive molecules as device surface coatings may confer
thromboresistance. Such coating materials include phospholipids and heparin.
Phospholipids such as phosphorylcholine, a normal constituent of cell membranes,
may orient polar head groups towards the aqueous phase and locally organize water
molecules, much like hydrogel surfaces. These surfaces may minimize protein and
complement interactions [28]. In preliminary animal studies, phosphorylcholine
coated stents, guide wires, and vascular grafts have shown improved thromboresis-
tance. This approach is being actively developed for clinical applications.
6 Blood–material Interactions 627

Heparin, a naturally occuring anticoagulant glycosaminoglycan, has been con-

sidered an attractive surface coating based on its ability to directly catalyze the
inactivation of procoagulant enzymes, and thus suppress thrombus development.
Recently, the reduced thrombogenicity and improved biocompatibility of heparin-
ized metallic stents has been demonstrated in animals [29]. In these studies, heparin
was covalently attached to a polymer surface coating. This method has also been
used clinically for the heparin coating of catheters and other devices, although it
remains unclear whether the improved biocompatibility is a function of heparin’s
anticoagulant activity, nonspecific physicochemical properties, or both.
With biomolecule modified surfaces, there may also be important questions
regarding the possible deleterious effects of sterilization procedures required before
implantation.

6.10 Surface Texture

Surface ‘smoothness’ is a generally desirable feature of blood contacting surfaces.

In this context, a smooth surface is one with irregularities with typical dimensions
less than those of cells (< 1 μm). However, in certain applications, device incorpora-
tion by tissue is desirable, or the texturing of polymers may increase mechanical
flexibility and durability. Thus, the sewing ring of mechanical heart valves is typi-
cally composed of poly(ethylene terephthalate) (Dacron™) fabric to permit tissue
in growth and healing, which is associated with a reduction in thromboembolic
events. Vascular grafts used to replace diseased blood vessels are most commonly
fabricated from woven or knitted DacronTM or textured (expanded) polytetrafluo-
roethylene (ePTFE) (Goretex™). In tubular form, these textured polymers remain
flexible and stable for many years following implantation. Smooth-walled vascular
grafts have generally not been considered attractive for long-term applications since
smooth surfaces may not permit tissue ingrowth or flow surface healing. Textured
flow surfaces are initially thrombogenic upon blood contact, although ePTFE
appears less thrombogenic and thromboembolic than fabric Dacron™ prostheses
[30]. These grafts perform acceptably in man when the graft diameter exceeds about
6 mm since the layer of thrombus that forms does not significantly restrict blood
flow. Interestingly, both smoothwalled and textured ventricular assist devices have
also performed successfully in clinical trials [31, 32].
628 S.R. Hanson

6.11 Conclusion

Because the variables affecting cardiovascular device responses are sufficiently

numerous and complex, those properties of blood-contacting surfaces which would
be desirable to minimize adverse reactions cannot, in most instances, be predicted
with confidence. The choice of material is usually constrained by mechanical prop-
erty considerations and by variable requirements for material durability and chemi-
cal stability. Cardiovascular device engineering must therefore be guided by
previous experience in successful clinical applications.

Acknowledgement This work was supported by Research Grant HL 31469 from the Heart, Lung
and blood Institute, the National Institutes of Health.

Additional Reading

Colman, R.W., Hirsch, J., Marder, V.J. and Salzman, E.W. (eds)(1994) Hemostasis
and Thrombosis: Basic Principles and Clinical Practice, 3rd edn, J.B. Lippincott,
Philadelphia.
This book is highly recommended. This state-of-the-art text covers in detail
essentially all important hematological aspects of cardiovascular device blood com-
patibility. In particular, Chapter 76, Interaction of blood with artificial surfaces,
which considers many theoretical, experimental, and animal studies, and Chapter
77, Artificial devices in clinical practice, which describes clinical device thrombo-
embolic complications, are of great practical value.
Harker, L.A., Ratner, B.D. and Didisheim, P. (eds)(1993) Cardiovascular
Biomaterials and Biocompatibility, Cardiovascular Pathology, 2(3) (suppl.),
1S–224S.
In this volume, 20 chapters by expert authors treat all aspects of biomaterials and
blood compatibility including pathologic mechanisms, material characterization,
blood-material interactions and device performance. This volume updates an
excellent earlier book Guidelines for blood-Material Interactions, National
Institutes of Health, Washington, DC, Publication No. 85–2185 (1985).
Szycher, M. (ed.) (1983) Biocompatible Polymers, Metals, and Composites,
Technomic Publishing Co., Lancaster, Pennsylvania.
Many of the same issues of blood–material interactions are broadly covered
while selected polymer and device applications are described in additional detail. of
particular interest are Section I (Fundamental Concepts in blood/Material
Interactions) and Section II (Strategies for Hemocompatibility).
6 Blood–material Interactions 629

References

1. Salzman, E.W., Merrill, E.W. and Kent K.C. (1994) Interaction of blood with artificial surfaces,
in Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 3rd edn, R.W. Colman,
J. Hirsch, V.J. Marder, and Salzman, E.W. (eds), J.B. Lippincott, Philadelphia, pp. 1469–85.
2. Harker, L.A., Ratner, B.D. and Didisheim, P. (eds) (1993) Cardiovascular Biomaterials and
Biocompatibility, Cardiovascular Pathology 2(3)(supplement), 1S–224S.
3. Szycher, M. (ed.) (1983) Biocompatible Polymers, Metals, and Composites, Technomic
Publishing Co., Lancaster, Pennsylvania.
4. Williams, D.F. (ed) (1981) Biocompatibility of Clinical Implant Materials, CRC Press, Boca
Raton, Florida.
5. Clagett, G.P. and Eberhart, R.C. (1994) Artificial Devices in Clinical Practice, in Hemostasis
and Thrombosis: Basic Principles and Clinical Practice, 3rd edn R.W. Colman, J. Hirsch,
V.J. Marder, and Salzman, E.W. (eds), J.B. Lippincott, Philadelphia, pp. 1486–1505.
6. Grabowski, E.F., Didisheim, P., Lewis, J.C. et al. (1977) Platelet adhesion to foreign surfaces
under controlled conditions of whole blood flow: human vs. rabbit, dog, calf, sheep, pig,
macaque, and baboon. Transactions - American Society for Artificial Internal Organs, 23,
141–51.
7. Lentner, C. (ed.) (1984) Geigy Scientific Tables (vol. 3): Physical Chemistry, Composition of
blood, Hematology, Somatometric Data, Ciby-Geigy, Basle.
8. Brash, J.L. and Horbett, T.A. (eds) (1987) Proteins at Interfaces. Physicochemical and
Biochemical Studies, American Chemical Society, Washington, DC.
9. Harker, L.A. and Slichter, S.J. (1972) Platelet and fibrinogen consumption in man. New
England J Med. 287(20), 999–1005.
10. Bennett, B., Booth, N.A., Ogston D. (1987) Potential interactions between complement, coag-
ulation, fibrinolysis, kinin-forming, and other enzyme systems, in: Haemostasis and
Thrombosis (2nd edn), A.L. bloom and D.P. Thomas (eds), Churchill Livingstone, New York,
pp. 267–82.
11. Andrew, M. (1995) Developmental hemostasis: relevance to thromboembolic complications in
pediatric patients. Thrombosis and Hemostasis, 74(1), 415–25.
12. Hanson, S.R., Harker, L.S., Ratner, B.D. et al. (1980) In vivo evaluation of artificial surfaces
using a nonhuman primate model of arterial thrombosis. J Laboratory Clinical Med. 95,
289–304.
13. Silver, J.H., Myers, C.W., Lim, F. et al. (1994) Effect of polyol molecular weight on the physi-
cal properties and haemocompatibility of polyurethanes containing polyethylene oxide macro-
glycols. Biomaterials 15(9), 695–704.
14. Hoffman, A.S. (1974) Principles governing biomolecular interactions at foreign surfaces.
J. Biomedical Materials Res. (Symp.) 5(1), 77–83.
15. Andrade, J.D., Lee, H.B., Jhon, M.S. et al. (1973) Water as a biomaterial. Transactions-
American Society for Artificial Internal Organs 19, 1–7.
16. Strzinar, I. and Sefton, M.V. (1992) Preparation and thrombogenicity of alkylated polyvinyl
alcohol coated tubing. J. Biomedical Materials Research 26, 577–92.
17. Merrill, E.W. (1993) Poly(ethylene oxide) star molecules: synthesis, characterization, and
applications in medicine and biology. J. Biomaterials Science, Polymer Edition 5(1–2), 1–11.
18. Llanos, G.R. and Sefton, M.V. (1993) Does polyethylene oxide possess a low thrombogenic-
ity? J. Biomaterials Science, Polymer Edition, 4(4), 381–400.
19. Sigwart, U., Puel, J., Mirkovitch, V. et al. (1987) Intravascular stents to prevent occlusion and
restenosis after transluminal angioplasty. New England J. Med, 316(12), 701–6.
20. Scott, N.A., Nunes, G.L., King, S.B. et al. (1995) A comparison of the thrombogenicity of
stainless steel and tantalum coronary stents. American Heart J., 129, 866–72.
21. Saywer, P.N., Stanczewski, B., Lucas, T.R., et al. (1978) Physical chemistry of the vascular
interface, in Vascular Grafts, P.N. Sawyer and M.J. Kaplitt (eds), Appleton -Century-Crofts,
New York, pp. 53–75.
630 S.R. Hanson

22. Rapold, H.J., Stassen, T., Van de Werf, F., et al. (1992) Comparative copper coil-induced
thrombogenicity of the internal mammary, left anterior descending coronary, and popliteal
arteries in dogs. Arteriosclerosis and Thrombosis, 12(5), 634–44.
23. Schoen, F.J. (1983) Carbons in heart valve prostheses: Foundations and clinical performance,
in M. Szycher (ed.), Biocompatible Polymers, Metals, and Composites, Technomic Publishing
Co., Lancaster, Pennsylvania, pp. 239–61.
24. Yasuda, H.K. (1985) Plasma Polymerization, Academic Press, Orlando.
25. Yeh, Y.S., Iriyama, T., Matsuzawa, Y., et al. (1988) blood compatibility of surfaces modified
by plasma polymerization. J. Biomedical Materials Research, 22, 795–818.
26. Harker, L.A., Malpass, T.W., Branson H.E., et al. (1980) Mechanism of abnormal bleeding in
patients undergoing cardiopulmonary bypass: acquired transient platelet dysfunction associ-
ated with selective alpha-granule release. blood, 56(5), 824–34.
27. Hakim, R. (1993) Complement activation by biomaterials. Cardiovascular Pathology, 2(3)
(suppl), 187S-198S.
28. Yu, J., Lamba, N.M., Courtney J.M., et al. (1994) Polymeric biomaterials: Influence of phos-
phorylcholine polar groups on protein adsorption and complement activation. International
Journal of Artificial Organs, 17(9), 499–504.
29. Hardhammer, P.A., van Beusekom H.M., Emanuelsson, H.U., et al. (1996) Reduction in
thrombotic events with heparin-coated Palmaz-Schatz stents in normal porcine coronary arter-
ies. Circulation, 93(3), 423–30.
30. Schneider, P.A., Kotze, H.F., Heyns, A. duP., et al. (1989) Thromboembolic potential of syn-
thetic vascular grafts in baboons. J. Vascular Surgery, 10, 75–82.
31. Dasse, K.A., Poirier, V.L., Menconi, M.J., et al. (1990) Characterization of TCPS textured
blood-contacting materials following long-term clinical LVAD support. In: Cardiovascular
Science and Technology: Basic and Applied: II, JC Norman (ed.), Oxymoron Press, Boston,
MA, pp. 218–220.
32. Kormos, R.L., Armitage, J.M., Borovetz, H.S., et al. (1990) Univentricular support with the
Novocor left ventricular assist system as a bridge to cardiac transplantation: An update in
Cardiovascular Science and Technology: Basic and Applied: II, JC Norman (Ed), Oxymoron
Press, Boston, MA, pp. 322–324.
Chapter 7
Soft Tissue Response to Silicones

S.E. Gabriel

7.1 Silicones used in Medicine

Although the term ’silicone’ refers to a group of organic silicone compounds, the
one most commonly used in medicine is composed of a polymer known as dimethy-
polysiloxane (DMPS). In silicone gel the polymer is cross-linked; the more cross-
linking, the more solid is the gel. Liquid silicone consists of glucose-linked DMPS
polymer chains. Silicones first became commercially available in 1943, with the
first subdermal implantation of silicone occurring in the late 1940s [1–3]. Silicones
have since been developed for a wide variety of medical applications, most notably
in joint and breast prostheses.
There is a large body of literature attesting to the chemical and physical inertness
of silicone [4–12]. Recently, there has been increasing interest in the possible
adverse effects of silicones used in implantation. Much of the literature describing
the adverse effects of silicone has been in reference to direct silicone injection.
The following discussion will review the immunologic effects of prostheses used in
breast reconstruction and augmentation.

7.2 Local immunologic reactions to silicone

Immunologic reactions to silicone can be local, regional due to silicone migration,

or systemic. Local cutaneous and subcutaneous reactions to injected silicone or gel
have been reported [13–18]; and it has become apparent that these reactions are not

S.E. Gabriel (*)

Division of Rheumatology Mayo Clinic,
200 First Street Southwest, Rochester, MN 55905, USA

© Springer Science+Business Media New York 2016 631

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_34
632 S.E. Gabriel

due to impurities in the silicone, as was originally suspected. Subcutaneous injec-

tion of silicone liquid in experimental animals provokes an acute inflammatory
response characterized by a primarily polymorphonuclear reaction, followed by a
chronic inflammatory response with lymphocytes, fibroblasts, and plasma cells
[19]. The late response is characterized by a small amount of cellular infiltrate and
an increase in extracellular material. Macrophages with clear vacuoles have been
observed and are suspected to contain silicone. Occasionally, multinucleated for-
eign body giant cells have also been observed [16, 20–23].
In humans, liquid silicone has been injected subcutaneously for cosmetic rea-
sons. Granulomatous reactions have been reported to occur in some instances [14–
17]. Similar reactions have been noted in two case reports following the rupture of
silicone gel-filled breast prostheses [24, 25]. Clinically, these reactions have the
characteristics of an inflammatory response, i.e., redness, swelling, and pain.
Histologic examination shows chronic inflammatory reactions, occasionally with
the presence of refractile material resembling silicone [24].
Migration of silicone has been documented on numerous occasions in the lit-
erature. Following experimental intra-peritoneal injection in mice, silicone was
demonstrated to be present in the liver, spleen, ovaries, and kidneys [26]. Other
investigators have documented the migration of subcutaneously injected silicone
to the lung, associated with an increased incidence of respiratory problems in
experimental animals [27]. Pneumonitis was reported in 3 patients several days
following liquid silicone injection, and silicone was demonstrated in macro-
phages obtained by pulmonary lavage from these patients [27]. The presence of
silicone was confirmed by atomic absorption and infra-red spectrophotometry.
Another case report described a patient with silicone-induced granulomatous
hepatitis; analysis of liver biopsy specimens revealed quantifiable amounts of sili-
cone [28]. Subcutaneous masses or nodules, hepatic granulomas have also been
reported following injections into humans [23, 24, 28]. Regional lymphadenopa-
thy is a frequently reported finding [29–35]. In rare cases, this has progressed to
malignant lymphoma [29, 32, 33]. The relevance of these reports to silicone
breast implants is uncertain.

7.3 Systemic immunologic reactions to silicone

Systemic reactions have been reported following the introduction of silicone into
the body. In one instance, a severe systemic reaction consisting of a febrile illness,
acute renal insufficiency, respiratory compromise, pulmonary infiltration, delirium,
anemia, and thrombocytopenia has been reported following implantation of a sili-
cone gel envelope prosthesis. Improvement followed implant removal. Silicone was
identified by mass spectrophotometry in this case [36]. Another case involved the
injection of a large quantity of free silicone under the breasts by an unauthorized
individual. The patient expired within 10 hours of injection. Silicone was identified
7 Soft Tissue Response to Silicones 633

by absorption spectrophotometry in large quantities in the lung, kidney, liver, brain

and serum [28].
The mechanism underlying the systemic immunologic reactions to silicone has
not been thoroughly investigated. A marked local granulomatous reaction to sili-
cone has been noted in guinea pigs; however, an antibody response to silicone by
Ouchterlony gel diffusion or passive cutaneous anaphylaxis was not demonstrated
[37]. Other investigators have studied macrophage migration inhibition [38]. In
these studies, pigs were sensitized by subcutaneous injection of silicone. Harvested
macrophages demonstrated inhibition of migration in the presence of silicone, sug-
gesting specific antigen recognition. In addition, silicone was demonstrated in the
cytoplasmic bridges joining macrophages and lymphocytes. Alternatively, it has
been suggested that the immune system does not respond with a specific recognition
of silicone but that silicone promotes the immune response to other antigens, i.e.,
acts as an adjuvant. Hypergammaglobulinemia has been noted by some investiga-
tors, and silicon dioxide has been reported to have adjuvant effects [39]. A disorder
which has been termed, ’human adjuvant disease’, was described following injec-
tion of paraffin for breast augmentation mammoplasty [40]. In Japan, in 1973,
Yoshida reviewed seven cases of human adjuvant disease in Japan following injec-
tions of paraffin or silicone for augmentation mammoplasty [41]. The symptoms
included arthritis, arthralgia, lymphadenopathy, hypergammaglobulinemia, elevated
erythrocyte sedimentation rates and positive rheumatoid factor. Removal of the
injected materials resulted in improvement of the condition in some patients [41].
In 1979, Kumagai reported four cases of classical systemic sclerosis following
cosmetic surgery [42]. Five years later, the same investigator described a series of
46 patients with signs and symptoms of connective tissue disease following injec-
tion of either silicone or paraffin [43]. Definite connective tissue diseases, based on
American Rheumatism Association Criteria, were diagnosed in 24 patients. These
conditions included systemic lupus erythematosus, mixed connective tissue disease,
rheumatoid arthritis, Sjögren’s syndrome, and systemic sclerosis. Another group of
22 patients were described as having human adjuvant disease, with signs, symp-
toms, and laboratory abnormalities suggestive but not diagnostic of a connective
tissue disease. In 1984, three patients from Singapore were reported who developed
autoimmune disease following injection augmentation mammoplasty [44]. In the
same year, a 52-year-old woman who developed systemic sclerosis, primary biliary
cirrhosis and Sjögren’s syndrome following silicone/paraffin injection mammo-
plasty was reported [45].
In 1982, the first case series describing autoimmune disorders following aug-
mentation mammoplasty with gel-filled prostheses was reported. This was followed
by other reports involving both gel-filled implants and salinefilled silicone implants
[46–49]. The most frequently reported connective tissue disease associated with
silicone breast implants is systemic sclerosis. Table 7.1 summarizes the clinical and
laboratory characteristics of 19 cases of systemic sclerosis associated with silicone
breast augmentation published in the English language literature. Eleven of these
cases received implantation in the United States [35]. Eleven of the 19 patients were
ANA positive, 15 had Raynaud’s, and 10 had diffuse systemic sclerosis. The inter-
634 S.E. Gabriel

val between augmentation to diagnosis of systemic sclerosis varied from 1 to 25

years, with a mean of 13 years. Fourteen of the 19 patients were exposed to silicone
(11 silicone gel and 3 silicone injection), the remainder being exposed to paraffin
injection. In two cases, histopathologic demonstration of multinucleated giant cells,
vacuoles with refractile droplets, and intracytoplasmic asteroid bodies in lymph
nodes draining the prostheses suggested leakage of silicone from the implants and
its dissemination in lymphoid tissues. The authors used energy-dispersive X-ray
analysis to confirm that the macrophage inclusions in the lymphatic tissue contain
silicone.
Although systemic sclerosis is the most commonly reported disorder occurring
following silicone breast implantation, there have also been reports of systemic
lupus erythematosus [43, 44, 46, 50], Sjögren’s syndrome [45], keratoconjunctivitis
sicca [45], rheumatoid arthritis [4, 34, 43], polymyositis [43], overlap syndromes
(including human adjuvant disease) [34, 43, 46, 50, 51], morphea [35, 43],
Hashimoto’s thyroiditis [43, 52], anticardiolipin antibody syndrome [53], primary
biliary cirrhosis [45] and toxic shock syndrome [54]. Unfortunately, it is impossible
to tell, on the basis of case reports, whether the frequency of these events is greater
than might be expected on the basis of chance alone (Table 7.1).

7.4 Evidence for causation

It has been estimated that two million American women have undergone breast
augmentation or reconstruction since the introduction of the silicone gel-filled elas-
tomer envelope-type prosthesis in the early 1960s [60, 61]. The reported cases of
systemic sclerosis among this population raise the important question of whether
the association between systemic sclerosis and silicone breast implants is a real one.
Unfortunately, almost all of the evidence to date is derived from case reports, which
are the very weakest form of data bearing on the question of causality (Table 7.2).
Indeed, the most important evidence for establishing a cause-effect relationship is
the strength of the research design used to study that relationship [62]. Randomized
control trials provide the strongest evidence but are seldom ethical in studies of
causation because they involve randomly assigning individuals to receive or not to
receive a potentially harmful intervention. In addition, the long latent periods and
large numbers of subjects needed to answer most cause and effect questions in clini-
cal medicine make it impractical to utilize this research design.
Well conducted prospective cohort studies are the next strongest design because
they minimize the effects of selection bias, measurement bias, and known confound-
ers. Such a study would involve following a large population of women, preferably
for one or more decades, looking for the outcomes of interest (e.g., connective tissue
disorders). A relative risk for connective tissue disorders among those women who
elect to have breast implantation compared to those who do not can then be calcu-
lated. Although this is a powerful research design, it is usually impractical because
of the necessary long follow-up period. This problem can be circumvented by a ret-
Table 7.1 Cases of Systemic Sclerosis after Augmentation Mammoplasty Reported in the English-Language Literature 7
Extent of
Age at Age at Interval to Systemic Raynaud Antinuclear
Patient Diagnosis, y Mammoplasty, y Type of Implant Onset, y Sclerosis Phenomenon Systemic Involvement Antibodies* Reference
1 52 50 Silicone bag gel 2 Diffuse No No — (55)
2 41 20 Silicone bag gel 21 Diffuse Yes Pulmonary, + (55)
gastrointestinal,
3 63 53 Silicone bag gel 10 Limited Yes Pulmonary, + (55)
4 37 32 Silicone bag gel 5 Diffuse No Pulmonary, + (55)
5 45 25 Paraffin injection 19 Diffuse Yes No — (42)
6 49 20 Paraffin injection 16 Diffuse Yes Pulmonary, — (42)
gastrointestinal
7 51 25 Paraffin injection 17 Limited Yes No — (43)
Soft Tissue Response to Silicones

8 36 24 Paraffin injection 9 Limited Yes Pulmonary — (43)

9 55 30 Paraffin injection 25 Limited Yes Pulmonary — (44)
10 50 31 Silicone injection 19 Limited Yes Pulmonary, + (45)
gastrointestinal, primary
biliary cirrhosis
11 59 34 Silicone injection 25 Diffuse Yes No + (56)
12 38 26 Silicone bag gel 7 Limited Yes Gastrointestinal + (56)
13 47 32 Silicone bag gel 15 Diffuse Yes Pulmonary + (35)
14 59 50 Silicone gel 9 Diffuse No Pulmonary, + (35)
gastrointestinal
15 44 34 Silicone gel 10 Limited Yes Pulmonary, + (35)
gastrointestinal
16 43 37 Silicone gel 6 Diffuse Yes Gastrointestinal + (35)
17 44 43 Silicone gel 1 Limited Yes Malignant hypertension + (57)
18 44 19 Silicone injection 25 Diffuse Yes Malignant hypertension/ + (58)
renal failure
19 46 34 Silicone gel 12 Limited No No — (59)
635

- negative.
+ positive.
* Determined using immunofluorescence.
636 S.E. Gabriel

Table 7.2 Strength of Strongest Randomized controlled trials

Research Designs Used to
↓ Prospective cohort studies
Determine Causation
↓ Case control studies
↓ Ecological survey
Weakest Case series

Table 7.3 Evidence that an Association is Causal[62]

Characteristic Definition
Temporality Cause precedes effect
Strength Large relative risk
Dose-response Larger exposures to cause associated with higher rates of disease
Reversibility Reduction in exposure associated with lower rates of disease
Consistency Repeatedly observed by different persons, in different places,
circumstances, and times
Biologic plausibility Makes sense, according to biologic knowledge of the time
Specificity One cause leads to one effect
Analogy Cause-and-effect relationship already established for a
similar experience

rospective cohort study, which is similar with the exception that the population and
the exposure (breast implantation) is identified in the past, allowing the patients to be
followed to the present for the outcomes of interest. Although this is a very attractive
research design, it requires that the complete exposed and unexposed populations be
identifiable and that follow-up information be available on all individuals.
Case-control studies retrospectively compare the frequency of breast implanta-
tion in women with and without the outcomes of interest. If, for example, connec-
tive tissue disorders were more likely to occur among women with breast implants,
this would constitute some evidence for causation. Case-control studies typically
require less time and resources than cohort studies. However, they are susceptible to
many more biases than cohort studies [62]. The primary reason not to perform a
case-control study here, however, is that a separate case-control study would be
required for each of the outcomes of interest, i.e., a case-control study of systemic
sclerosis, a case-control study of rheumatoid arthritis, etc. The retrospective cohort
design is much more efficient since it can evaluate multiple outcomes in a single
study, as is the need here.
A set of eight criteria has been proposed as a guide to formulate decisions
regarding cause and effect relationships (Table 7.3). The relationship between
breast implants and connective tissue disorders does fulfill the criterion of tempo-
rality since, at least in the published case reports, the connective tissue disorders
all followed breast implantation. There is no evidence describing the magnitude of
the relative risk in this relationship. There is also no evidence for a dose response
relationship, i.e., that women with bilateral implants perhaps have an increased
likelihood of connective tissue disorders compared to women with unilateral
7 Soft Tissue Response to Silicones 637

implants. The evidence regarding the reversibility of these disorders with removal
of implants is variable. Although there have been some reports of improvement of
connective tissue disorders following removal of the implants, this is not consis-
tent and the number of patients involved is small. The relationship does appear to
be consistent, i.e., it has been observed repeatedly by different persons in different
places, circumstances, and times, however it has not yet been assessed using ade-
quate study designs. Perhaps the most compelling evidence is the biologic plausi-
bility of this relationship due to the hypothesis of silicone acting as an immune
adjuvant. The relationship does not appear specific, as silicone implants have lead
to not just one effect, but several, albeit somewhat related, effects. Finally, a cause
and effect relationship is strengthened if there are examples of well established
causes that are analogous to the one in question. Adjuvant induced arthritis can be
considered analogous [63].
In summary, in spite of the anecdotal evidence, until very recently there was a
lack of evidence to either support or refute a cause-and-effect relationship between
silicone breast implants and connective tissue/ autoimmune disorders.

7.5 Controlled studies examining the relationship

between breast implants and connective tissue disease.

At least seven controlled studies have now been published (Table 7.4), each of
which provided a quantitative assessment of the risk of connective tissue diseases
among women with breast implants [64–70]. The first of these was a case-control
study of augmentation mammoplasty and scleroderma [68]. The aims of this study
were to compare the frequency and temporal relationship of augmentation mam-
moplasty among scleroderma cases and matched controls. Scleroderma patients and
age stratified general practice controls were interviewed using a pretested telephone
questionnaire. Self-reported dates of augmentation mammoplasty were ascertained
as were dates of scleroderma symptoms and diagnoses as relevant. Frequencies of
noriaugmentation mammoplasty silicone exposure between interviewed cases and
controls were expressed in terms of rate ratios and 95% confidence intervals. Rate
ratios were also adjusted for socioeconomic status.
A total of 315 cases and 371 controls were interviewed, of whom 251 and 289,
respectively, were female. The unadjusted rates for augmentation mammoplasty
among interviewed cases and controls were 4/251 (1.59%) and 5/289 (1.73% ),
respectively. The socioeconomic status adjusted rate of augmentation mammo-
plasty in scleroderma patients was 1.54% (95% CI: 0.03–3.04) which is very
similar to the 1.73% rate in interviewed controls. These results indicate that aug-
mentation mammoplasty rates were comparable in cases and controls. In addi-
tion, the rates of exposure to nonmammoplasty silicone mastectomy and breast
lumpectomy were comparable in interviewed cases and controls. This study
failed to demonstrate an association between silicone breast implantation and the
Table 7.4 Summary of controlled studies examining the relationship between breast implants (BI) and connective tissue diseases (CTD)
638

Study Population
Cases Controls Outcome(s)
Reference Study Design (exposed) (unexposed) Examined Main Result Conclusions
68 Case control 315 371 Systemic sclerosis Rates of BI among cases and controls Rates of BI were similar in cases
(SS, scleroderma) were 1.59% and 1.73% and controls.
65 Retrospective 749 1498 Connective Relative risk (cases:controls) of There was no association
cohort tissue and other developing any of these diseases was between BI and the connective
autoimmune diseases 1.06 (95% CI: 0.34–2.97). tissue and other disorders studied.
67 Case control 195 143 Systemic lupus One (0.8%) of the SLE cases and 0 (0%) No association was shown
erythematosus (SLE) of the controls reported having a BI between BI and SLE.
(p=0.57).
64 Case control 349 1456 Rheumatoid Relative risk for a history of BI No increased risk for RA among
arthritis (RA) (cases:controls) was 0.41 (95% CI: women with BI was
0.05–3.13). demonstrated.
60 Multi-center 869 2061 SS (scleroderma) Odds ratio for BI surgery (cases:controls) No significant causal relationship
case control was 1.25 (95% CI: 0.62–2.53). was demonstrated between BI
and the development of SS.
70 Nested case 121700 Connective tissue Five cases with BI were identified among No association was found
control disease 300 patients with RA;0 cases with BI between BI and CTD.
among 123 with SLE, 20 patients with
SS, 3 with Sjögren's syndrome, 13 with
dermato/polymyositis, and 2 with mixed
connective tissue disease.
69 Case control 592 1184 SS (scleroderma) Odds ratio for BI (cases:controls) was No significant association
(preliminary 0.61 (95% CI: 0.14–2.68) between BI and SS was found.
results)
CI = confidence interval
S.E. Gabriel
7 Soft Tissue Response to Silicones 639

subsequent development of scleroderma to a relative risk as low as 4.5 with 95%

statistical power.
In June of 1994, a population-based retrospective cohort study was published
which examined the risk of a variety of connective tissue diseases and other disor-
ders after breast implantation [65]. In this study, all women in Olmsted County,
Minnesota who received a breast implant between 1 January 1964 and 31 December
1991 (the case subjects) were studied. For each case subject, two women of the
same age (within three years) from the same population who had not received a
breast implant and who underwent a medical evaluation within two years of the date
of the implantation in the case subject were selected as control subjects. Each wom-
an’s complete inpatient and outpatient medical records were interviewed for the
occurrence of various connective tissue diseases (i.e., rheumatoid arthritis, systemic
lupus erythematosus, Sjögren’s syndrome, dermatomyositis, polymyositis, sys-
temic sclerosis, ankylosing spondylitis, psoriatic arthritis, polymyalgia rheumatica,
vasculitis, arthritis associated with inflammatory bowel disease, and polychondri-
tis), certain other disorders thought to have an autoimmune pathogenesis (i.e.,
Hashimoto’s thyroiditis), and cancer other than breast cancer. In addition, this study
itemized the results of ten related symptoms and the abnormal results of four related
laboratory tests. A total of 749 women who had received a breast implant were fol-
lowed for a mean of 7.8 years and the corresponding 1498 community controls were
followed for a mean of 8.3 years. The relative risk of developing any one of these
specified connective tissue and other diseases among case subjects compared to
controls was 1.06 (95% CI: 0.34–2.97). This study, therefore, found no association
between breast implants and the connective tissue diseases and other disorders that
were studied [65].
In the summer of 1994, Strom and colleagues published a case-control study
which addressed the risk of systemic lupus erythematosus among women with
breast implants [67]. A total of 219 eligible cases who met the American Rheumatism
Association criteria for systemic lupus erythematosus [71] were identified from the
medical practices of cooperating rheumatologists in the Philadelphia metropolitan
area. One hundred ninety-five (89%) of these were enrolled in the study. Friends of
the cases, matched to the cases on sex and age (±5 years), served as controls. Using
a short telephone interview, cases and controls were contacted and asked to provide
information on any surgery that they had prior to the index date, i.e., the date of
diagnosis of systemic lupus erythematosus, in the cases and the same year for the
age-matched friend controls. Specific questions were asked about plastic surgery in
general and breast implants in particular. One hundred forty-eight (75.9%) of the
195 systemic lupus erythematosus being sought and 111 (77.6%) of the 143 controls
agreed to be reinterviewed for this study. Only 1 (0.8%) of the 133 female systemic
lupus erythematosus cases reported having a breast implant eight years prior to the
diagnosis of systemic lupus erythematosus. This compared to 0 out of the 100
female friend controls (Fisher exact one-tailed p-value = 0.57). These authors con-
cluded, based on this very large case-control study of systemic lupus erythemato-
sus, that no association existed between silicone breast implants and the subsequent
development of systemic lupus erythematosus. However, the modest statistical
640 S.E. Gabriel

power of the study was only able to provide sufficient evidence against a very large
association.
Three additional controlled studies have been presented and are published in
abstract form [64, 66, 70]. As part of a prospective case-control study of the risk of
rheumatoid arthritis, Dugowson et al. recruited 349 women with new-onset rheu-
matoid arthritis and 1456 similarly aged control women. Information about breast
implants was obtained on both cases and controls and age-adjusted risk for a history
of breast implants among cases was compared to that of controls. The relative risk,
i.e., comparing the rate of a history of breast implants among rheumatoid arthritis
cases compared to a similar history among controls, was 0.41 (95% CI: 0.05–3.13)
[64]. These data did not support an increased risk for rheumatoid arthritis among
women with silicone breast implants.
A multi-center, case-controlled study was performed to examine the association
between scleroderma and augmentation mammoplasty [66]. A total of 869 women
with systemic sclerosis recruited from three university-affiliated rheumatology
clinics and 2061 local community controls matched on age in three strata (ages
25–44, 45–64, and ≥65); race and sex were identified by random-digit dialing. Data
on exposure and potential confounding variables were collected from cases and
controls by self-administered questionnaires and telephone interviews, respec-
tively. The frequency of breast implant surgery was compared in both groups and
the odds ratio and 95% confidence intervals for the association of augmentation
mammoplasty with systemic sclerosis, adjusted for age, race, marital status, and
site, was 1.25 (95% CI: 0.62–2.53). These data failed to demonstrate a significant
causal relationship between augmentation mammoplasty and the development of
systemic sclerosis.
Using a nested case-control study, Sanchez-Guerrero et al. examined the asso-
ciation between silicone breast implants and connective tissue diseases among a
cohort of 121 700 registered American nurses followed since 1976 [70]. In 1992, a
questionnaire was sent to nurses who had reported any rheumatic disease from
1976 to 1990 asking about rheumatic symptoms and silicone exposure. The com-
plete medical records were obtained on all participants who confirmed any rheu-
matic or musculoskeletal symptoms. Connective tissue disease cases were classified
according to the American College of Rheumatology or other published criteria.
Ten age-matched controls per case were randomly selected among nurses with no
rheumatic or musculoskeletal complaints. Odds ratios and 95% confidence inter-
vals were used as a measure of association. This study identified 448 cases with
definite connective tissue diseases and 1209 nurses with silicone breast implants.
Five patients had silicone breast implants and any connective tissue disease. The
mean time since implantation was 139 ± 95.81 months among these five patients
and 119.17 ± 76.64 months among all nurses with silicone breast implants. These
five patients were identified among 300 patients with rheumatoid arthritis. No case
with a silicone breast implant was identified among 123 patients with systemic
lupus erythematosus, 20 patients with scleroderma, three with Sjögren’s syndrome,
13 with dermato/polymyositis, and two with mixed connective tissue diseases. In
7 Soft Tissue Response to Silicones 641

conclusion, this study found no association between silicone breast implants and
connective tissue diseases.
Finally, Burns and Schottenfeld are conducting a population-based casecontrol
study examining the relationship between this condition and prior history of breast
implant surgery [69, 72]. The cases and normal population controls are being
assembled from the states of Michigan and Ohio. The cases are being identified
using several sources: a computerized data base of hospital discharge diagnostic
listings during the period 1980–1991; a collaborative network of major medical
centers; a postal survey of certified rheumatologists; and from patient members of
the United Scleroderma Foundation. Although this study is still underway, prelimi-
nary results demonstrate a crude odds ratio of 0.61 (95% CI: 0.14–2.68) for breast
implants among cases compared to controls. These results do not support a causal
relationship between breast implants and systemic sclerosis.
In summary, although numerous anecdotal case reports have suggested an asso-
ciation between silicone breast implants and connective tissue diseases, all seven
controlled epidemiologic studies conducted to date have failed to confirm such an
association. Whether silicone breast implants cause a new and previously unde-
scribed condition is yet to be determined.

Reference

1. Williams, D.F. (1981) Biocompatibility of clinical implant materials. Vol 2. Boca Raton: CRC
Press, pp. 1–272.
2. Blocksma. R. and Braley, S. (1965) The silicone in plastic surgery. Plast. Reconstr. Surg.,
35(4), 366–370.
3. Braley, S.A., Jr (1964) The medical silicones. Trans. Am. Soc. Artif Int. Organs, 10,
240–243.
4. Weiner, S.R. and Paulus, H.E. (1986) Chronic arthropathy occurring after augmentation mam-
maplasty. Plast. Reconstr. Surg., 77(2), 185–192.
5. Ward, T.C. and Perry, J.T. (1981) Dynamic mechanical properties of medical grade silicone
elastomer stored in simulated body fluids. J. Biomed. Mater. Res., 15 511–525.
6. Homsy, C.A. (1970). Bio-compatibility in selection of materials for implantation. J. Biomed.
Mater. Res., 4, 341–356.
7. Rigdon, R.H. and Dricks, A. (1975) Reaction associated with a silicone rubber gel: An experi-
mental study. J. Biomed. Mater. Res., 9, 645–659.
8. Robertson, G. and Braley, S. (1973) Toxicologic studies, quality control, and efficacy of the
silastic mammary prosthesis. Med. Instrum., 7(2), 100–103.
9. Boone, J.L. and Braley, S.A. (1966) Resistance of silicone rubbers to body fluids. Rubber
Chern. Technol. , 39(4), 1293–1297.
10. Boone, J.L. (1966) Silicone rubber insulation for subdermally implanted devices. Med. Res.
Eng., 5, 34–37.
11. Roggendorf, E. (1976) The biostability of silicone rubbers, a polyamide, and a polyester.
J. Biomed Mater. Res., 10, 123–143.
12. Speirs, A.C. and Blocksma, R. (1963) New implantable silicone rubbers. An experimental
evaluation of tissue response. Plast. Reconstr. Surg., 31(2), 166–175.
13. Symmers, W.StC. (1968) Silicone mastitis in ’topless’ waitresses and some other varieties of
foreign-body mastitis. Br. Med 1., 3, 19–22.
642 S.E. Gabriel

14. Ortiz-Monasterio, F. and Trigos, I. (1972) Management of patients with complications from
injections of foreign materials into the breasts. Plast. Reconstr. Surg. , 50(1) 42–47.
15. Parsons, R.W. and Thering, H.R. (1977) Management of the silcone-injected breast. Plast.
Reconstr. Surg., 60(4), 534–538.
16. Wilkie, T.F. (1977) Late development of granuloma after liquid silicone injections. Plast.
Reconstr. Surg., 60(2), 179–188.
17. Pearl, R.M., Laub, D.R., and Kaplan, E.N. (1978) Complications following silicone injections
for augmentation of the contours of the face. Plast Reconstr Surg, 61(6): 888–891.
18. Kozeny, G.A., Barbato, A.L., Bansal, V.K. et al. (1984) Hypercalcemia associated with
silicone-induced granulomas. N. Engl. J. Med., 311(17), 1103–1105.
19. Andrews, J.M. (1966) Cellular behavior to injected silicone fluid: A preliminary report. Plast.
Reconstr. Surg, 38(6), 581–583.
20. Ballantyne, D.L., Rees, T.D. and Seidman, I. (1965) Silicone fluid: Response to massive sub-
cutaneous injections of dimethylpolysiloxane fluid in animals. Plast. Reconstr. Surg., 36(3),
330–338.
21. Ben-Hur, N. and Neuman, Z. (1965) Siliconoma - another cutaneous response to dimethylpo-
lysiloxane. Experimental study in mice. Plast. Reconstr. Surg., 36(6), 629–631.
22. Rees, T.D., Platt, J., and Ballantyne, D.L. (1965) An investigation of cutaneous response to
dimethylpolysiloxane (silicone liquid) in animals and humans - A preliminary report. Plast.
Reconstr. Surg., 35(2), 131–139.
23. Travis, W.D., Balogh, K. and Abraham, J.L. (1985) Silicone granulomas: Report of three cases
and review of the literature. Hum. Pathol., 16(1), 19–27.
24. Mason, J. and Apisarnthanarax, P. (1981) Migratory silicone granuloma. Arch. Dermatol., 117,
366–367.
25. Apesos, J. and Pope, T.L., Jr (1985) Silicone granuloma following closed capsulotomy of
mammary prosthesis. Ann. Plast. Surg., 14(5), 403–406.
26. Rees, T.D., Ballantyne, D.L., Seidman, I. et at. (1967) Visceral response to subcutaneous and
intraperitoneal injections of silicone in mice. Plast. Reconstr. Surg., 39(4), 402–410.
27. Chastre, J., Basset, F., Viau, F. et al. (1983) Acute pneumonitis after subcutaneous injections
of silicone in transsexual men. N. Engl. J. Med. , 308(13), 764–767.
28. Ellenbogen, R. and Rubin, L. (1975) Injectable fluid silicone therapy: Human morbidity and
mortality. J. Am. Med. Assoc. , 234(3), 308–309.
29. Murakata, L.A. and Rangwala, A.F. (1989) Silicone lymphadenopathy with concomitant
malignant lymphoma. J. Rheumatol., 16(11), 1480–1483.
30. Christie, A.J., Weinberger, K.A. and Dietrich, M. (1977) Silicone lymphadenopathy and syno-
vitis. Complications of silicone elastomer finger joint prostheses. J. Am. Med. Assoc., 237(14),
1463–1464.
31. Kircher, T. (1980) Silicone lymphadenopathy. A complication of silicone elastomer finger
joint prostheses. Hum. Pathol., 11(3), 240–244.
32. Benjamin, E. and Ahmed, A. (1982) Silicone lymphadenopathy: A report of two cases, one
with concomitant malignant lymphoma. Diagn. Histopathol., 5, 133–141.
33. Digby, J.M. (1982) Malignant lymphoma with intranodal silicone rubber particles following
metacarpophalangeal joint replacements. Hand, 14(3), 326–328.
34. Endo, L.P., Edwards, N.L., Longley, S. et al. (1987) Silicone and rheumatic diseases. Semin.
Arthritis Rheum., 17(2), 112–118.
35. Varga, J., Schumacher, H.R. and Jimenez, S.A. (1989) Systemic sclerosis after augmentation
mammoplasty with silicone implants. Ann. Intern. Med., 111(5), 377–383.
36. Uretsky, B.F., O’Brien, J.J., Courtiss E.H. et al. (1979) Augmentation mammaplasty associ-
ated with a severe systemic illness. Ann. Plast. Surg., 3(5), 445–447.
37. Nosanchuk, J.S. (1968) Injected dimethylpolysiloxane fluid: A study of antibody and histo-
logic response. Plast. Reconstr. Surg. , 42(6), 562–566.
38. Heggers, J.P., Kossovsky, N. and Parsons, R.W. et al. (1983) Biocompatibility of silicone
implants. Ann. Plast. Surg., 11, 38–45.
7 Soft Tissue Response to Silicones 643

39. Pernis, B. and Paronetto, F. (1962) Adjuvant effect of silica (tridymite) on antibody produc-
tion. Proc. Soc. Exp. Biol. Med., 110, 390–392.
40. Miyoshi, K., Miyamura, T. and Kobayashi, Y. (1964) Hypergammaglobulinemia by prolonged
adjuvanticity in man. Disorders developed after augmentation mammaplasty. Jpn Med. J.,
2122, 9.
41. Yoshida, K. (1973) Post mammaplasty disorder as an adjuvant disease of man. Shikoku Acta
Med., 29, 318.
42. Kumagai, Y., Abe, C. and Shiokawa, Y. (1979) Scleroderma after cosmetic surgery. Four cases
of human adjuvant disease. Arthritis Rheum., 22(5), 532–537.
43. Kumagai, Y., Shiokawa, Y. and Medsger, T.A., Jr (1984) Clinical spectrum of connective tissue
disease after cosmetic surgery. Arthritis Rheum., 27(1), 1–12.
44. Fock, K.M., Feng, P.H., Tey, B.H. (1984). Autoimmune disease developing after augmentation
mammoplasty: Report of 3 cases. J. Rheumatol, 11(1), 98–100.
45. Okano, Y., Nishikai, M. and Sato, A. (1984) Scleroderma, primary biliary cirrhosis, and
Sjogren’s syndrome after cosmetic breast augmentation with silicone injection: A case report
of possible human adjuvant disease. Ann. Rheum. Dis., 43, 520–522.
46. van Nunen, S.A., Gatenby, P.A. and Basten, A. (1982) Post-mammoplasty connective tissue
disease. Arthritis Rheum., 25(6), 694–697.
47. Baldwin, C.M., Jr, and Kaplan, E.N. (1983) Silicone-induced human adjuvant disease? Ann.
Plast. Surg., 10, 270–273.
48. Byron, M.A., Venning, V.A. and Mowat, A.G. (1984) Post-mammoplasty human adjuvant dis-
ease. Br. J. Rheumatol., 23(3), 227–229.
49. Vargas, A. (1979) Shedding of silicone particles from inflated breast implants. Plast. Reconstr.
Surg., 64(2), 252–253.
50. Walsh, F.W., Solomon, D.A., Espinoza, L.R. et al. (1989) Human adjuvant disease. A new
cause of chylous effusions. Arch. Intern. Med. , 149, 1194–1196.
51. Ziegler, V.V., Haustein, U.F., Mehlhorn, J. et al. (1986) Quarzinduzierte sklerodermie
sklerodermie-ahnliches syndrom oder echte progressive sklerodermie? Dermatol. Mon. Schr.,
172(2), 86–90.
52. Taura, N., Usa, T. and Eguchi, K. (1976) A case of human adjuvant disease. J. Jap. Soc. Intern.
Med., 65, 840–841.
53. Alusik, S., Jandova, R. and Gebauerova, M. (1989) Anticardiolipin syndrome in plastic sur-
gery of the breast. Cor. Vasa, 31(2), 139–144.
54. Olesen, L.L., Ejlertsen, T. and Nielsen, J. (1991) Toxic shock syndrome following insertion of
breast prostheses. Br. J. Surg., 78(5), 585–586.
55. Spiera, H. (1988) Scleroderma after silicone augmentation mammoplasty. J. Am. Med. Assoc.,
260(2), 236–238.
56. Brozena, S.J., Fenske, N.A., Cruse, C.W. et al. (1988) Human adjuvant disease following
augmentation mammoplasty. Arch. Dermatol., 124, 1383–1386.
57. Gutierrez, F.J. and Espinoza, L.R. (1990) Progressive systemic sclerosis complicated by severe
hypertension: Reversal after silicone implant removal. Am. J. Med., 89(3), 390–392.
58. Hitoshi, S., Ito, Y., Takehara, K. et al. (1991) A case of malignant hypertension and sclero-
derma after cosmetic surgery. Jpn J. Med., 30(1), 97–100.
59. Sahn, E.E., Garen, P.D., Silver, R.M. et al. (1990) Scleroderma following augmentation mam-
moplasty. Report of a case and review of the literature. Arch. Dermatol., 126, 1198–1202.
60. Deapen, D.M., Pike, M.C., Casagrande, J.T. et al. (1986) The relationship between breast
cancer and augmentation mammaplasty: An epidemiologic study. Plast. Reconstr. Surg.,
77(3), 361–367.
61. May, D.S. and Stroup, N.E. (1991) The incidence of sarcomas of the breast among women in
the United States, 1973–1986. Plast. Reconstr. Surg., 87(1), 193–194.
62. Fletcher, R.H., Fletcher, S.W. and Wagner, E.H. (1988) Cause. In: Collins, N., Eckhart, C.,
Chalew, G.N., eds. Clinical Epidemiology: The Essentials. Baltimore: Williams & Wilkens,
pp. 208–225.
644 S.E. Gabriel

63. Pearson, C.M. (1963) Experimental joint disease. Observations on adjuvant-induced arthritis.
J. Chron. Dis., 16, 863–874.
64. Dugowson, C.E., Daling, J., Koepsell, T.D. et al. (1992) Silicone breast implants and risk for
rheumatoid-arthritis. Arthritis Rheum., 35, S66. (Abstract)
65. Gabriel, S.E., O’Fallon, W.M., Kurland, L.T. et al. (1994) Risk of connectivetissue diseases
and other disorders after breast implantation. N. Engl. J. Med., 330, 1697–1702.
66. Hochberg, M.C., Perlmutter, D.L., White, B. et al. (1994) The association of augmentation
mammoplasty with systemic sclerosis: Results from a multi-center case-control study. Arthritis
Rheum., 37 (Supplement): S369 (Abstract).
67. Strom, B.L., Reidenberg, M.M., Freundlich, B. et al. (1994) Breast silicone implants and risk
of systemic lupus erythematosus. J. Clin. Epidemiol. (in press).
68. Englert, H.J., Brooks, P. (1994) Scleroderma and augmentation mammoplasty - a causal rela-
tionship? Aust. NZ J. Med., 24, 74–80.
69. Burns, C.J., Laing, T.J., Gillespie, B.W. et al. (1996) The epidemiology of scleroderma among
women: Assessment of risk from exposure to silicone and silica. J. Rheumatol., 23(11),
1904–1911.
70. Sanchez-Guerrero, J., Karlson, E.W., Colditz, G.A. et al. (1994) Silicone breast implants (SBI)
and connective tissue disease (CTD). Arthritis Rheum, 37 (Supplement): S282 (Abstract).
71. Tan, E.M., Cohen, AS., Fries, J.F. et al. (1982) The 1982 revised criteria for the classification
of systemic lupus erythematosus. Arthritis Rheum, 25 (11), 1271–1277.
72. Schottenfeld, D., Burns, C.J., Gillespie, B.W. et al. (1995) The design of a population-based
case-control study of systemic sclerosis (scleroderma): Commentary on the University of
Michigan Study. J. Clin. Epidemiol. 48(4), 583–586.
Chapter 8
Vocal Folds

Joel Gaston and Susan L. Thibeault

8.1 Introduction

Vocal folds are two infoldings of complex tissue housed in the larynx whose vibration,
known as phonation, results in voice. To create voice, vocal folds are adducted, effec-
tively closing the glottis or the space between them creating a pressure difference
between the top and bottom of the tissue. This pressure difference, combined with
Bernoulli forces and tissue elasticity, pushes the vocal folds apart and brings them
back together again [1]. This rapid, sustained opening and closing cycle produces
vocal fold vibration at various fundamental frequencies ranging from 0 to 300 Hz for
normal vocal levels [2]. The vocal folds are multilayer structures that consist of
muscle, lamina propria, and epithelium. Fundamental frequency is controlled by
laryngeal muscles which alter vocal fold physical properties, such as length and thick-
ness. The extracellular matrix (ECM) of the lamina propria contributes significantly to
vocal quality. Disruption of vocal fold vibration and/or quality through muscular
dysfunction, airflow disruption, or tissue damage results in disturbance of normal
voice production. Vocal fold scarring, a prevalent vocal fold injury, is characterized by
pathophysiologic changes to the lamina propria ECM, directly causing a marked
decrease in voice quality [3].

J. Gaston
Department of Biomedical Engineering, University of Wisconsin–Madison,
5118 WIMR, 1111 Highland Ave., Madison, WI 53705, USA
S.L. Thibeault (*)
Department of Surgery, University of Wisconsin–Madison,
5107 WIMR, 1111 Highland Ave., Madison, WI 53705, USA
Department of Biomedical Engineering, University of Wisconsin–Madison,
5107 WIMR, 1111 Highland Ave., Madison, WI 53705, USA
Department of Communication Sciences and Disorders, University of Wisconsin–Madison,
5107 WIMR, 1111 Highland Ave., Madison, WI 53705, USA
e-mail: [email protected]

© Springer Science+Business Media New York 2016 645

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1_35
646 J. Gaston and S.L. Thibeault

8.2 Composition

The vocal fold can be divided into three layers, divided primarily by tissue function.
The top layer consists of squamous cell epithelium, which separates the underlying
tissue from the exposed airway. Below that lies the primary vibratory region, the
lamina propria. The cellular and protein composition of this layer determines the
ease of phonation. The lamina propria can be further subdivided into three layers,
superficial, intermediate, and deep lamina propria (Fig. 8.1). The most interior sec-
tion of the vocal fold consists of the thyroarytenoid muscle (muscularis), which
relaxes and shortens the vocal folds. The large majority of biomechanical character-
ization has been completed on the lamina propria for biomaterial applications; sub-
sequently the remainder of this chapter concentrates on the properties of this specific
layer. The most important proteins, glycosaminoglycan (GAG), and proteoglycans
in the lamina propria, as well as their concentrations at different depths, can be
found in Table 8.1.

SUPERFICIAL
INTERMEDIATE
DEEP

EPITHELIUM

MUSCULARIS

100 80 60 40 15 0

Fig. 8.1 Schematic of the divisions of the vocal fold, including lamina propria subdivisions.
Reprinted with permission from [4]
8 Vocal Folds 647

Table 8.1 Lamina propria ECM components and their relative concentrations with respect to
tissue depth
Superficial lamina Intermediate Deep lamina
ECM component propria lamina propria propria
Fibrous proteins Collagen I Low High Highest
Collagen III Low High Highest
Elastin Low Highest High
Fibronectin Highest Low Low
GAG Hyaluronic acid Very low Highest Low
Proteoglycans Decorin High Very low High
Biglycan Intermediate Very low Highest
Versican Very low High High

8.2.1 Collagen

The primary protein component of the vocal fold lamina propria is collagen, which
constitutes roughly 43 % of the total tissue protein by weight [5]. Immunohistological
staining shows that collagen I and collagen III are the primary collagens of the lamina
propria. Both collagen types show the highest staining in the deep lamina propria, close
to the thyroarytenoid muscle, with the intermediate layer having slightly lower staining
intensity. The high fibrillar collagen content imparts strength and resiliency to the vocal
folds. In general, the collagen fibers are spatially oriented from anterior to posterior,
allowing them to support the force applied by the intrinsic laryngeal muscles [6].

8.2.2 Elastin

Elastin plays a crucial role in vocal fold vibration and the mechanical properties of
the lamina propria. Accounting for approximately 8.5 % of the total lamina propria
tissue protein by weight, it is mainly responsible for imparting elastic recoil to the
human vocal folds [7]. Like collagen, higher concentrations of elastin are observed
in the deep and intermediate layers of the lamina propria, with the concentration
decreasing closer to the epithelial layer. Fibrillin, another elastic protein, appears to
have deposition opposite that of elastin, with the highest concentration in the super-
ficial lamina propria and the lowest concentration in the intermediate layer [7].

8.2.3 Hyaluronic Acid

Hyaluronic acid (HA), a linear nonsulfated glycosaminoglycan, is a major compo-

nent of the vocal fold ECM. It is the most prevalent glycosaminoglycan present,
consisting of roughly 6.4 μg of HA for each mg of total protein within the lamina
648 J. Gaston and S.L. Thibeault

propria [7]. The high concentration has a significant effect on vocal fold mechanical
properties. The long chain length and loosely coiled molecular structure of HA
allow it to sequester large volumes of water, resulting in high tissue viscosity. The
high tissue viscosity acts as a shock absorber, resisting tissue compression and dam-
age caused during phonation.

8.2.4 Interstitial Proteins

Several other proteins play prominent roles in the vocal fold lamina propria. These
small proteins contribute to ECM organization and typically are found at different
levels in the lamina propria. Fibronectin plays a key role in both matrix organization
and interaction with resident fibroblasts [8]. It is primarily located in the superficial
layer of the lamina propria and the basement membrane that divides the epithelium
from the superficial layer of the lamina propria. Decorin, a small proteoglycan, is
also predominantly found in the superficial lamina propria [9]. It plays a role in col-
lagen fibril assembly [10] and may contribute to the low density of collagen fibers
in the superficial lamina propria. In addition, the ability of decorin to modulate
stress transmission through the extracellular matrix is crucial for vocal fold phona-
tion. Biglycan, a small proteoglycan with a similar role to decorin, is mainly located
in the deep lamina propria. Within the deep lamina propria, it appears to be most
dense near the intermediate layer [9]. Versican, a large chondroitin sulfate proteo-
glycan, is found primarily in the intermediate lamina propria [9]. It is often found
colocalized with HA, though it is capable of binding with a variety of partners.
Extracellular proteins such as fibrillin, fibronectin, and collagen I are capable of
binding versican, and may affect cell adhesion to these structural proteins.

8.2.5 Voice Disorders

Scarring is the leading cause of poor voice quality after vocal fold surgery. The tis-
sue fibrosis defined by scarring results in disorganized ECM architecture and altered
mechanical properties. Controlled animal studies show that the major fibrous ECM
proteins, collagen and elastin, are altered following scarring of the lamina propria
(Table 8.2). To date, only one ten-patient study has analyzed the histology of scarred
vocal folds [15]. Collagen deposition in the scarred lamina propria was found to be
related to injury depth, with thicker fiber deposition correlated with deeper injury
depth. Elastin was absent or fragmented and disorganized in all but one patient,
consistent with animal studies. HA deposition followed a similar pattern, with HA
being unobservable in all but one patient, where staining was barely visible. Finally,
a quantitative assessment of the collagen, elastin, and HA content could not be per-
formed, as the study did not have a normal control from the patients.
8 Vocal Folds 649

Table 8.2 Review of animal studies detailing vocal fold injury with histological and compositional
changes
Animal Type of Timeline
model injury post-injury Collagen Elastin HA
Rabbit [11] Lamina 60 days Disorganized Short, compact Density similar
propria fibers, less dense fibers. Less to normal
forceps than normal dense than controls
biopsy controls normal controls
Rabbit [13] Lamina 6 months Organized fibers, Fibers Density similar
propria thicker than fragmented and to normal
forceps normal control disorganized control
biopsy
Canine [14] Vocal fold 2 months Disorganized Disorganized Density similar
stripping fibers, thicker fibers, less to normal
than normal dense compared control
control to normal
control
Rat [15] Unilateral 8 weeks Increased density N/A Decreased
vocal fold compared to density
stripping normal controls compared to
normal controls
Rat [15] Unilateral 12 weeks Increased density N/A Decreased
vocal fold compared density
stripping normal to compared to
controls normal controls

8.3 Mechanical Properties

8.3.1 General

The mechanical properties of the vocal fold are critical for normal phonation. Vocal
fold tissue mechanical properties contribute to the minimum amount of air pressure
needed to produce voice, also known as phonation threshold pressure. Phonation
threshold pressure is the minimum subglottal pressure needed to overcome airway
pressure, forcing the vocal folds open to begin oscillation. This measure is often
taken as an objective indication of the energy required by an individual for vocal
function. The combination of large quantities of collagen, elastin, and hyaluronic
acid, and their respective mechanical properties, imparts viscoelastic properties to
the vocal fold lamina propria. In normal conditions, and different disease states, the
amount and organization of each protein affect the overall viscoelasticity of the
tissue. In the past, the Young’s modulus for the human intermediate and deep layers
of the vocal fold has been calculated at 33.1 ± 10.4 kPa and 135 kPa for strains of
15 % and 25 %, respectively [16]. The wide frequency range covered during vocal
fold phonation means direct measurement of the vocal fold viscoelastic properties
has greater clinical application. Presently there are no methods to measure viscoelas-
tic properties in vivo. For in vitro analysis, shear rheology is the favored method for
quantifying the mechanical properties of excised lamina propria vocal fold tissue.
650 J. Gaston and S.L. Thibeault

8.3.2 Rheology

Oscillatory shear deformation, a standard rheological technique, is used to deter-

mine the complex shear modulus of the vocal fold lamina propria [17]. Typically, a
parallel plate rheometer is used to subject the tissue to precisely control sinusoidal
torque, resulting in oscillatory shear stress. During testing, the controlling software
measures the shear amplitudes and phase shift from the shear stress and strain func-
tions. Measurement of these properties allows for the calculation of the elastic shear
modulus, dynamic viscosity, viscous modulus, and damping ratio.

8.3.3 Elastic Shear Modulus

Elastic shear modulus quantifies the elasticity of the tissue under shear stress, and is
proportional to the energy stored elastically due to strain. With rheological testing, the
elastic shear modulus is a function of frequency, and only increases gradually across
most of the frequency range tested (Fig. 8.2). At frequencies greater than 1 Hz, the elas-
tic shear modulus increases with higher frequency. Shear properties are typically only
measured at a frequency range of .01 to 15 Hz, as shear thinning and the small sample
size prevent meaningful results at higher testing frequencies for vocal fold tissue.
Rheology data published indicates that in general, the female vocal fold lamina propria
had a lower elastic shear modulus compared to the male vocal fold lamina propria.

10000

72 y.o.
62 y.o.
1000
60 y.o.
30 y.o.
m (Pa)

28 y.o.
100
50 y.o.
34 y.o.
36 y.o.
10
31 y.o.
59 y.o.

1
0.01 0.1 1 10 100
Frequency (Hz)

Fig. 8.2 Elastic shear modulus of male human vocal fold lamina propria. Reprinted with permis-
sion from [18]
8 Vocal Folds 651

8.3.4 Dynamic Viscosity

Dynamic viscosity characterizes the opposition of the tissue to shear flow, and is
proportional to the energy lost in the viscoelastic tissue. Dynamic viscosity is
related to the viscous shear modulus through the angular frequency. Rheological
testing shows that the dynamic viscosity decreases as a function of frequency, and
as such the vocal folds exhibit shear thinning behavior (Fig. 8.3). This behavior is
fundamental to vocal fold movement—as the fundamental frequency of vibration
increases, the lamina propria will shear thin providing movement of this tissue at
very high frequencies.

8.3.5 Poisson’s Ratio

Only one study has investigated the Poisson ratio of human vocal folds [19]. An
optical method was used, where a marked vocal fold sample was illuminated, cycli-
cally stretched at 1 Hz and 37 °C, and video recorded using a high-speed camera
system. Processing of the recorded images resulted in a Poisson ratio of .392. This
number should be taken with caution; several models consider the human vocal
folds to be transversely isotropic [20, 21], and testing was performed at only one
frequency [19].

1000 72 y.o.
62 y.o.
100 60 y.o.
50 y.o.

10 30 y.o.
h (Pa-s)

28 y.o.
34 y.o.
1
36 y.o.
59 y.o.
0.1
31 y.o.

0.01
0.01 0.1 1 10 100
Frequency (Hz)

Fig. 8.3 Dynamic viscosity of male human vocal fold lamina propria. Reprinted with permission
from [18]
652 J. Gaston and S.L. Thibeault

8.3.6 Vocal Fold Scarring

Numerous studies have shown that scarring of vocal fold tissue results in tissue
mechanics detrimental to normal voice function. Animal models have shown that
vocal fold scarring results in permanently elevated elastic shear modulus and
dynamic viscosity, making the tissue both stiffer and more viscous (Table 8.3).
Rabbit models are particularly useful, as normal vocal fold viscoelastic shear prop-
erties are similar to those of human vocal fold tissue [17]. This can be visualized
further when the viscoelastic moduli are plotted as a function of frequency for rabbit
animal models. Elastic shear modulus of scarred vocal fold tissue is higher than the
normal control tissue for all samples tested (Fig. 8.4a), indicating an increase in
stiffness [22]. The same trend is seen for the dynamic viscosity, with scarred tissue
having a significantly higher modulus than normal controls (Fig. 8.4b). Though no
direct comparison of mechanical properties between normal and scarred human
vocal folds, nearly all mammalian animal models have shown the same patterns
described above [13, 14].

8.4 Biomaterial Injections

Injectable biomaterials have been developed and used to treat vocal fold lamina
propria with abnormal viscoelastic properties due to ECM disruption. The most
important property for consideration for a vocal fold injectable is that the biomate-
rial has biomechanical properties most similar to the lamina propria, so as to allow
for normal tissue movement. When injected into the vocal fold lamina propria, the
viscoelastic properties of injectable biomaterials should be matched as closely as
possible to normal lamina propra. To this end, several natural and synthetic bioma-
terials have been developed to mitigate dysphonia. Upon injection, biomaterials
with significantly higher viscous and elastic moduli make vocal fold oscillation, and
therefore phonation, difficult. Finally, any biomaterial implanted into the vocal fold
must undergo shear thinning in order to match the basic viscoelastic behavior of the
lamina propria.

Table 8.3 Animal studies detailing vocal fold injury and changes in mechanical properties
Animal Timeline
model Type of injury post-injury Elastic shear modulus Dynamic viscosity
Rabbit [12] Lamina propria 60 days Stiffer tissue, higher More viscous, higher
forceps biopsy modulus modulus
Rabbit [13] Lamina propria 6 months Stiffer tissue, higher More viscous, higher
forceps biopsy modulus in eight out modulus in nine out
of ten samples of ten samples
Canine [14] Vocal fold 2 months Stiffer tissue, higher More viscous, higher
stripping modulus modulus
8 Vocal Folds 653

a
10000 Scar One

Scar Two

Scar Three

Scar Four

Scar Five
1000
Scar Six

Norm One
m (Pa)

Norm Two

Norm Three

Norm Four
100
Norm Five

10
0.01 0.1 1 10 100

b
10000 Scar One

Scar Two

Scar Three
1000
Scar Four

Scar Six
100
Scar Five

Norm One
h (Pa-s)

10 Norm Two

Norm Three

Norm Four
1
Norm Five

0.1

0.01
0.01 0.1 1 10 100

Fig. 8.4 (a) Elastic shear modulus (μ) of rabbit vocal fold. (b) Dynamic viscosity (η) of rabbit
vocal fold tissue. Reprinted with permission from [22]

8.4.1 Collagen and Fat Injections

Several different collagen formulations and adipose have been tested as injectables
for vocal fold scarring. A study utilized rheological testing to compare collagen and
adipose tissue to normal vocal fold tissue, with a specific focus on viscosity [17].
Adipose tissue prepared as a fat graft was found to have the closest viscosity to
normal vocal fold tissue, followed by uncrosslinked collagen (Fig. 8.5).
654 J. Gaston and S.L. Thibeault

TABLE II.
Dynamic Viscosity of Implantable Biomaterials and Human Vocal Fold Mucosal Tissues
Measured at 10 Hz and Extrapolated to 100 Hz.
Dynamic Viscosity (Pa-s)

Material Sample Measured at 10 Hz Extrapolated to 100 Hz

Polytetrafluoroethylene (Teflon) 116.144 10.186

Gelatin (Gelfoam) 21.297 2.335
GAX collagen (Phonagel or Zyplast) 12.844 1.480
Noncrosslinked collagen (Zyderm) 8.563 0.980
Human abdominal subcutaneous fat
(70-year-old woman) 3.026 0.296
Vocal fold mucosa (72-year-old man) 2.702 0.281
Vocal fold mucosa (62-year-old man) 0.897 0.099

Fig. 8.5 Comparison of dynamic viscosity of several biomaterial injections for the vocal fold
lamina propria. Reprinted with permission from [17]

8.4.2 Modified Hyaluronic Acid

HA hydrogels show the most promise as injectable biomaterials for the vocal fold
to date. However, natural HA is rapidly degraded upon injection, typically taking
roughly 3–5 days to dissolve [12]. As such several different HA hydrogel formula-
tions with differing cross-links and chemical moieties have been designed and
investigated for vocal fold applications (Table 8.4).
Hylan B, a non-inflammatory divinyl sulfone cross-linked simple HA hydrogel,
has been used in the past as a space-filling injection. It is no longer in the market,
but has since been succeeded by the Restylane line of HA injectables. Multiple
animal studies show that Hylan-B injections into rabbit vocal fold lamina propria
show mechanical results similar to normal vocal fold [24, 27]. Specifically, 6
months after injection no significant difference in dynamic viscosity was observed
between Hylan-B-injected tissue and normal, noninjected tissue. In the same study,
vocal fold tissue injected with Hylan-B was found to have a lower dynamic viscos-
ity than tissue injected with other biomaterials, such as collagen or Teflon. In an
83-person clinical study, Hylan-B had a more favorable effect than collagen on
vocal fold function [28]. Most notable was the increase in vocal fold mucosal wave
production compared to baseline, indicating positive viscoelastic mechanics. This
result could not be directly corroborated, as rheology can only be performed on
ex vivo tissue.
Hydrogel cross-linking is of particular interest in vocal fold applications, as
different cross-linking methods and groups can alter the mechanical properties of
the hydrogel. Several different HA hydrogels have been formulated that utilize a
polyethylene glycol diacrylate (PEGDA) cross-linker, allowing the gelation time
and swelling ration to be easily modified [29]. PEGDA cross-linking is prominent
8 Vocal Folds 655

Table 8.4 Summary of studies utilizing HA derivatives and their outcomes

HA derivative Normal Prophylactic Chronic
Hylan B Fibrotic No inflammation
effects up to 12 months
post-injection
[31]. Hylan B
residence time of
at least 6 months
[23, 31]
Mechanical No change in
effects dynamic viscosity
compared to no
injection [27]
Carbylan-S Fibrotic Decreased fibrosis
effects compared to saline [30]
Mechanical Lower elastic shear
effects modulus and lower
dynamic viscosity [30]
HA-DTPH- Mechanical No significant
PEGDA effects difference compared to
saline [30]
Fibrotic No change in elastic
effects shear modulus. Lower
dynamic viscosity [30]
Extracel Fibrotic Decreased procollagen, Increased collagen
effects fibronectin, and and procollagen
fibromodulin mRNA [23]
[12]
Mechanical Lower elastic shear Lower elastic
effects modulus and lower shear modulus and
dynamic viscosity [12] lower dynamic
viscosity [23]

in two HA hydrogels for vocal fold applications; HA-DTPH and Carbylan-S. Both
HA-DTPH and Carbylan-S are similar in composure, consisting of an HA backbone
with PEGDA cross-links. The Carbylan-S backbone has additional carboxylate
groups on the HA backbone, resulting in a higher elastic shear modulus than
HA-DTPH [25]. To date, both hydrogel formulations have been primarily examined
as a prophylactic treatment to reduce or prevent vocal fold scarring by injection at
the time of injury. Excised tissue from injured rabbit vocal fold injected with
Carbylan-S had a lower elastic shear modulus than injured tissue treated with
saline controls or injured tissue injected with HA-DTPH-PEGDA [30]. Tissue
injected with either Carbylan-S or HA-DTPH-PEGDA was significantly less
viscous than the saline-treated samples as well. Trichrome staining performed on
sections of the excised larynges showed that the average fibrosis levels for animals
injected with the HA-DTPH-PEGDA were moderate and not significantly different
from saline-treated controls.
656 J. Gaston and S.L. Thibeault

The Carbylan-S HA backbone has also been cross-linked using thiolated gelatin
to generate a hydrogel commercially designated as Extracel® [26]. Extracel® has
undergone several naming changes since its first formulation, and can also be found
as Gycosil® or HyStem®. Extracel® has been investigated as both a prophylactic
treatment for vocal fold tissue injury and as a treatment for previously scarred tis-
sue. A rabbit study showed that 6 months after injury and injection, excised vocal
fold tissue injected with Extracel® had a lower elastic shear modulus and lower
viscous modulus compared to saline controls [12]. While histology was not per-
formed, quantitative PCR showed a decrease in mRNA transcript levels for fibro-
nectin, fibromodulin, and procollagen [12]. A similar study was performed on
rabbits where an injury was induced and allowed to scar for 3 months [23], at which
time Extracel® or saline was injected into the vocal fold. Two months following
injection, the tissue injected with Extracel® had a lower elastic shear modulus and
viscous modulus compared to saline injections. Finally, histological staining showed
that the Extracel®-treated vocal folds had higher procollagen and collagen than
saline, a marked contrast from the prophylactic study.

Additional Reading

Miri AK. Mechanical characterization of vocal fold tissue: a review study. J Voice
2014; 28:657–67.
Caton T, Thibeault SL, Klemuk S, Smith ME. In reference to viscoelasticity of
hyaluronan and nonhyaluronan based vocal fold injectables: Implications for muco-
sal versus muscle use—Reply. Laryngoscope 2007, 117(8): 1506–1508.
Chan RW, Titze IR. Dependence of phonation threshold pressure on vocal tract
acoustics and vocal fold tissue mechanics. Journal of the Acoustical Society of
America 2006; 119:2351–62.

References

1. Benninger MS, Alessi D, Archer S, Bastian R, Ford C, Koufman J et al (1996) Vocal fold scar-
ring: current concepts and management. Otolaryngol Head Neck Surg 115:474–482
2. Titze IR (2000) Principles of voice production. National Center for Voice and Speech
3. Smith E, Gray S, Verdolini K, Lemke J (1995) Effects of voice disorders on quality of life.
Otolaryngol Head Neck Surg 113:P121
4. Butler JE, Hammond TH, Gray SD (2001) Gender-related differences of hyaluronic acid dis-
tribution in the human vocal fold. Laryngoscope 111:907–911
5. Hahn MS, Kobler JB, Zeitels SM, Langer R (2006) Quantitative and comparative studies of
the vocal fold extracellular matrix. II: Collagen. Ann Otol Rhinol Laryngol 115:225–232
6. Gray SD, Titze IR, Alipour F, Hammond TH (2000) Biomechanical and histologic observa-
tions of vocal fold fibrous proteins. Ann Otol Rhinol Laryngol 109:77–85
7. Hahn MS, Kobler JB, Starcher BC, Zeitels SM, Langer R (2006) Quantitative and comparative
studies of the vocal fold extracellular matrix. I: Elastic fibers and hyaluronic acid. Ann Otol
Rhinol Laryngol 115:156–164
8 Vocal Folds 657

8. Hansen JK, Thibeault SL (2006) Current understanding and review of the literature: vocal fold
scarring. J Voice 20:110–120
9. Hahn MS, Kobler JB, Zeitels SM, Langer R (2005) Midmembranous vocal fold lamina propria
proteoglycans across selected species. Ann Otol Rhinol Laryngol 114:451–462
10. Kalamajski S, Oldberg A (2010) The role of small leucine-rich proteoglycans in collagen
fibrillogenesis. Matrix Biol 29:248–253
11. Hirano S, Minamiguchi S, Yamashita M, Ohno T, Kanemaru S, Kitamura M (2009) Histologic
characterization of human scarred vocal folds. J Voice 23:399–407
12. Thibeault SL, Klemuk SA, Chen X, Johnson BHQ (2011) In vivo engineering of the vocal fold
ECM with injectable HA hydrogels-late effects on tissue repair and biomechanics in a rabbit
model. J Voice 25:249–253
13. Rousseau B, Hirano S, Chan RW, Welham NV, Thibeault SL, Ford CN et al (2004)
Characterization of chronic vocal fold scarring in a rabbit model. J Voice 18:116–124
14. Rousseau B, Hirano S, Scheidt TD, Welham NV, Thibeault SL, Chan RW et al (2003)
Characterization of vocal fold scarring in a canine model. Laryngoscope 113:620–627
15. Tateya T, Tateya I, Sohn JH, Bless DM (2005) Histologic characterization of rat vocal fold
scarring. Ann Otol Rhinol Laryngol 114:183–191
16. Caton T, Thibeault SL, Klemuk S, Smith ME (2007) In reference to viscoelasticity of hyaluro-
nan and nonhyaluronan based vocal fold injectables: implications for mucosal versus muscle
use—Reply. Laryngoscope 117:1506–1508
17. Chan RW, Titze IR (1998) Viscosities of implantable biomaterials in vocal fold augmentation
surgery. Laryngoscope 108:725–731
18. Chan RW, Titze IR (1999) Viscoelastic shear properties of human vocal fold mucosa: measure-
ment methodology and empirical results. J Acoust Soc Am 106:2008–2021
19. Alipour F, Vigmostad S (2012) Measurement of vocal folds elastic properties for continuum
modeling. J Voice 26:9
20. Alipour F, Berry DA, Titze IR (2000) A finite-element model of vocal-fold vibration. J Acoust
Soc Am 108:3003–3012
21. Rosa MD, Pereira JC, Grellet M, Alwan A (2003) A contribution to simulating a three-
dimensional larynx model using the finite element method. J Acoust Soc Am 114:2893–2905
22. Thibeault SL, Gray SD, Bless DM, Chan RW, Ford CN (2002) Histologic and rheologic char-
acterization of vocal fold scarring. J Voice 16:96–104
23. Thibeault SL, Duflo S (2008) Inflammatory cytokine responses to synthetic extracellular
matrix injection to the vocal fold lamina propria. Ann Otol Rhinol Laryngol 117:221–226
24. Borzacchiello A, Mayol L, Garskog O, Dahlqvist A, Ambrosio L (2005) Evaluation of injec-
tion augmentation treatment of hyaluronic acid based materials on rabbit vocal folds viscoelas-
ticity. J Mater Sci Mater Med 16:553–557
25. Shu XZ, Liu Y, Prestwich GD (2011) Modified macromolecules and methods of making and
using thereof
26. Duflo S, Thibeault SL, Li WH, Shu XZ, Prestwich GD (2006) Vocal fold tissue repair in vivo
using a synthetic extracellular matrix. Tissue Eng 12:2171–2180
27. Dahlqvist A, Garskog O, Laurent C, Hertegard S, Ambrosio L, Borzacchiello A (2004)
Viscoelasticity of rabbit vocal folds after injection augmentation. Laryngoscope 114:138–142
28. Hertegard S, Hallen L, Laurent C, Lindstrom E, Olofsson K, Testad P et al (2002) Cross-linked
hyaluronan used as augmentation substance for treatment of glottal insufficiency: safety
aspects and vocal fold function. Laryngoscope 112:2211–2219
29. Shu XZ, Liu YC, Palumbo FS, Lu Y, Prestwich GD (2004) In situ crosslinkable hyaluronan
hydrogels for tissue engineering. Biomaterials 25:1339–1348
30. Hansen JK, Thibeault SL, Walsh JF, Shu XZ, Prestwich GD (2005) In vivo engineering of the
vocal fold extracellular matrix with injectable hyaluronic acid hydrogels: early effects on tis-
sue repair and biomechanics in a rabbit model. Ann Otol Rhinol Laryngol 114:662–670
31. Hallen L, Johansson C, Laurent C (1999) Cross-linked hyaluronan (hylan B gel): a new inject-
able remedy for treatment of vocal fold insufficiency—an animal study. Acta Otolaryngol
119:107–111
Index

Page numbers in bold refer to figures; those in italics refer to tables

A Al203 whisker-nickel, tensile

Above knee tissues, elastic moduli, 64 properties, 255
Adhesion Amalgamization, 192
bioglass, 157 Amalgams, 191, 191–4, 193, 194
titanium alloys, 157 chemical composition, 191
Age of donors, cortical bone, 5–6 copper content, 192
Alloys, 515–7, 516 corrosion, 194
AgPd, 197, 201 and wear, 193–4
galvanic, 515 high copper type, 194
chemical composition, 202 mercury vapour release, 194
NiCr, 202–13, 212–13 mixed alloys, 192
properties, 202, 203 one-component, 192
Pd, 197, 201 processing, 193
precious metal, 198–200 properties, 191, 191–2, 193
bonding with ceramics, 199 Amino plastics, properties, 237
brazing, 199 Animal cornea, 136
casting, 198 Annulus fibrosus, 46, 47, 50
heat treatment, 198–9 Antibiotics, metal interactions, 524
Alumina, 342–3 Antibodies, 594–5
ceramic hip balls, 340 antiserum, 596
chemical properties, 343 detecting, 595–7, 596
clinical performance, 344 in vivo testing, 596, 597
dental implants, 341 negative results, 597
engineering properties, 342 positive results, 596
implants, 339–40 immunoglobins, 594–5
physical properties, 342 Antigens, 593
standards, 343 immobilization, 596, 606
wear resistance, 343 processing cell, 593
Aluminum/graphite, tensile Antioxidants
strengths, 254 additives, 248
Aluminum/silicon, tensile data, 254 low toxicity, 245
Aluminum/Thornel, mechanical properties, 253 synergism, 248

© Springer Science+Business Media New York 2016 659

W. Murphy et al. (eds.), Handbook of Biomaterial Properties,
DOI 10.1007/978-1-4939-3305-1
660 Index

Aorta ASTM F553, 164

compliance, 87–8 chemical composition, 164, 174
hydrodynamic properties, 87 mechanical properties, 164, 165
morphometric properties, 82, 82 ASTM F799, 161
pressure/cross-sectional area, 87–8 chemical composition, 161
retraction, 87–8, 89 mechanical properties, 162
tapered, 78 ASTM F1058, 164
Apatite parameters chemical composition, 165
dentin, 27 mechanical properties, 165
enamel, 27 Atherosclerosis, 81, 84
Apparent density, 15 Autoclave molding, 221–2
Aramid fibers, 209 Autoimmune diseases, 602–3
disadvantages, 209
Arterial tissues
change with age, 79, 81, 83 B
composition, 81 Basal ganglia, 67
Arterial wall, composition, 79–84, 81, 82, 83, 84 B/5505 boron/epoxy
Arteries aging, 249
coronary, tensile properties, 89 moisture effects, 249
creep, 90–2, 91 Beam theory, 357–60
elastic modulus, 86 Below knee tissues, 64
hypertension, 81, 83 Bending modulus, osteons, 8, 8
mechanical properties, 84–92, 85, 86, 87, Bioactive
89, 90, 91 bonding, 447–8, 448, 449, 458
morphometric properties, 78, 78 ceramics, orthopaedics, 495–5
Reynolds number, 78 compositions, 449–50
stress relaxation, 90–2, 91 glasses, 447–54
tensile properties, 88, 89 bone bonding, 452
Articular cartilage, 37 clinical uses, 453
composition, 38 physical properties, 450, 451, 453
compression, 38–9, 39 Bioactivity
lubrication, 41–2 definition, 447
mechanical properties, 38–42 index, 449
permeability, 41 Bioceramics, 339–51
Poisson’s ratio, 41 ASTM standards, 347
shear properties, 40, 40 biocompatibility, 347–8, 348, 349
tensile properties, 39–40, 40 dental applications, 350
tribologic properties, 41–2 failure loads, 448
wear tests, 42 ISO standards, 347
Artificial saliva, repassivation rates, 193 manufacturers, 350
Aseptic loosening, 588 orthopaedic applications, 348
A-S/3501 graphite/epoxy trade names, 350
aging, 249 Biocompatibility, 563–9
moisture effects, 249 components, 566–9
ASTM F75, 159–61 control of, 568
chemical composition, 160 definition, 564–5
investment casting, 160 local host response, 567–9
mechanical properties, 161 material degradation, 566–7
ASTM F90, 161–2 Biofunctionality, metallic biomaterials,
chemical composition, 161, 162 154–6
mechanical properties, 162 Biological
ASTM F552, 162–3 degradation, 499–5, 500–1, 502
chemical composition, 163 animal experiments, 499–4
mechanical properties, 163 calcium phosphate, 500–1, 502
MP35N, 162 clinical applications, 499–4
Index 661

properties silicone, 640

titanium, 185 systemic lupus erythematosus, 639
Biomaterials, systemic effects, 569 Brittle behaviour, ceramic materials, 353
Biospan, 294 Brittle fiber/Sic, properties, 258
Biospan-D, 294 Brittle ring test, 367
Biospan-S, 294 Bulk modulus, brain tissues, 69
Blastogenic factor, 597
Blood, 115–23
and cardiovascular devices, 623 C
coagulation, proteins, 122 Calcium phosphates, and metallic ions, 533–6,
metals in, 519 534, 535, 536
physiochemical properties, 116 Cancellous bone, 15–20
velocity, 78 compression, 16
vessels, 77–97 fracture toughness, 19
atherosclerosis, 96 mechanical properties, 16, 17
changes with age, 94–5 microstructure, 15
hypertension, 95 Poisson’s ratio, 18
transport properties, 94, 95 stiffness and strength, 16–18
Bone strain, 18
cement, 581 surface-to-volume ratio, 15
formation, interfacial, 450 tissue composition, 16
Bone-biomaterial interfaces, 588, 589 tissue damage, 19–20
Bonelbiomaterial interfaces, 621–7 tissue-level mechanical properties, 20
Bone remodeling, 16 and ultrastructure, 16
Born equation, 379 viscoelastic and fatigue properties,
Boron fibers, 208, 209 18–19
application, 209 Young’s and shear modulus, 16, 18
manufacturing technique, 208 Cancer, 607–17
properties, 208 and metallic implants, 616
Bovine bone, creep, 10 Capillaries, morphometrics, 78
Brain tissues, 67–75 Ca/P molar ratio
bulk modulus, 69 dentin, 27
composition, 68 enamel, 27
creep modulus, 69, 72–3 Carbon fiber
density, 69 precursors, 208
diffusion properties, 74 properties, 208
dimensions, 68 trade names, 208
elastic modulus, 69, 70–71 Carbon prostheses, mechanical
electrical conductivity, 69, 74 properties of, 251
mass, 68 Carbons, 551–62
mechanical stimuli, 68 atomic arrangement, 553
Poisson’s ratio, 69 biocompatibility, 551
shape, 68 charcoal, 552
shear modulus, 69, 70–71 composites, 552, 558
stem, 67 filament wound, 558
thermal properties, 74, 74 dental applications, 558, 560
Breast augmentation, 633 fibers, 552, 556–7
scleroderma, 640 forms, mechanical properties, 559
subsequent diseases, 632–4 glassy, 455, 556
causes, 634–7, 636 in vivo, 558
systemic sclerosis, 633, 635 medical applications, 559
Breast implants orthopaedics, 560
and connective tissue disease, polymeric, 552
637–41, 638 pyrolytic, 552, 554, 555, 562–6
rheumatoid arthritis, 640 low-temperature, 555
662 Index

Carbons (cont.) compliance, 387–392

structures, 551–2 digital imaging correlation, 443–4
vapour phase coatings, 552, 557 double cantilever beam, 396–8, 417–20, 418
vascular applications, 560 double torsion method, 414–7
Carbon/silicon nitride, properties, 257 ductile behaviour, 353
Carbo thane, 308 energy release rate, 385–7
Cardiac valves, 625 engines, 430
Cardiovascular devices, and blood, 623 fracture work, 392–6
Cartilage, 37–42 Griffith’s criterion, 384–5
articular, 37 Griffith’s theory, 381–4
composition, 38, 38 hip joints endoprosthesis
elastic, 38 finite element analysis (FEA), 433–5
mechanical properties, 42 Weibull equation, 435–41
fibrocartilage, 38 microstructure, 354–5
hyaline, 37 porosity, 355–6
water content, 37 SPT diagram, 428–30
Cataractous human lenses stem fixation problem, 442
hardness, 108 strength of solids, 377–81
water content, 105 strength testing
Catheters, 622–626 3 and 4 point bending, 360–65
Causal association beam theory, 357–60
analogy, 636, 646 ceramic machining, 357
biologic plausibility, 636, 646 flexion and flexure testing, 357
consistency, 636, 646 nature of test, 374–7
dose-response, 636, 646 processing of failure data, 369–72
reversibility, 636, 646 ranking, 372–4
specificity, 636, 646 significance of test results, 368–9
strength, 636, 646 tensile testing, 356–7
temporality, 636, 646 test configurations, 365–8
CCR50, 154 test pieces, 357
Cell volume of test piece, 377
adaptation, 576 stress concentrators and cracks,
cultures, 504–5 398–12
testing, 504 stress corrosion, 426–8
mediated stress-volume-integral, 436
immunity, 599 valency type, 356
responses, 597–600 Ceramics, 256–8, 257, 258
responses: choice of test, 598 calcium phosphate, 495
migration inhibition, 597 flexural strength, 258
Cellular necrosis, 572 fracture toughness, 258
Cementation, 581–4, 582, 583 glass, 257
Cement-free implants, 582, 583, 584–5 hip balls, 340
hip prostheses, 584–5 physical properties, 450, 451, 453
knees, 584–5 processing, 347
orthopaedic, 584 Cerebellum, 67
Cement line, cortical bone, 4 Cerebrospinal fluid, 67, 117
Central corneal thickness (CCT), 138 inorganic content, 117–18
Centrifugal organic content, 118–19
casting, 220–21 protein content, 118–20
molding, 226–8 Cerebrum, 67
release agents, 226 C-Flex TPE, 310
Ceramic acetabular cup, 443 Chemical vapour deposition, 557
Ceramic materials Chromium
brittle behaviour, 353 corrosion, 511
Index 663

immunity, 511 particulate, 205

passivation, 511 radiation resistance, 246–7
ChronoFlex, 308 reinforcements, 253
Cluster differentiation markers, 594 Compression
Coatings, 528–33, 529, 530, 531, 532, 533, articular cartilage, 38–9, 39
539, 541 cortical bone, 8
corrosion currents, 531 modulus
HAp, 533, 534, 535 cortical bone, 6–7
hard ceramic, 530 osteons, 8, 8
metallic, 528 Compressive strength
niobium, 528 dentin, 29
tantalum, 528 enamel, 28–9
titanium, 528 Connective tissue disease, and breast implants,
ultra-thin, 625–6 637–41, 638
Cobalt-chromium (CoCr) alloys, 159–66 Contact angles, 542–5, 543
ASTM F75, 159–61 equilibrium, 545
chemical composition, 160 oxidized surfaces, 543
investment casting, 160 Contamination of tissues, 517–21
mechanical properties, 161 Continuous sheets, 228–9
ASTM F90, 161–2 decorative, 228
chemical composition, 161, 162 glass reinforcement, 228
mechanical properties, 162 Coremer, 308
ASTM F552, 162–3 Corethane, 308
chemical composition, 163 Corhesive, 308
mechanical properties, 163 Cornea, 135–6
MP35N, 162 animal, 136
ASTM F553, 164 biomechanics, 140
chemical composition, 164, 174 composition, 137
mechanical properties, 164, 165 diameter, 137, 137
ASTM F799, 161 electric properties of tissue, 141
chemical composition, 161 geometrical and optical properties, 139
mechanical properties, 162 keratometry, 139
ASTM F1058, 164 mechanical properties of tissue, 142–3
chemical composition, 165 refractive index, 140
mechanical properties, 165 stress-strain relationship, 141
Havar thickness, 137–8, 138
chemical composition, 165 ultraviolet and visible radiation
mechanical properties, 166 transmission, 141
UNS R30005, 164–6 Corneal diameter, 137
Collagen Corrosion, 509–45
content, cartilage, 37 contamination of tissues, 517–21
cortical bone, 3 detection of ions, 509
immune response to, 601 duration of implantation, 515, 516
shrinkage temperature, 3 grading scale, 510
Collagen/elastin ratio, 82 incidence of, 510
Collagen fibers, 55 and inertness, metallic biomaterials, 153–4
Combined loading, cortical bone, 9 materials purity, 512
Composites, 205–58 proteins, 524, 525
classifying, 205 Ti-15Zr-4Nb-2Ta-O.2Pd, 514
fabrication, 220–9 and wear, 182, 182–5, 183, 184, 185
fibrous, 205 noble metals, 201
laminar, 205 Cortical bone, 3–12
in medicine, 250–2 age of donors, 5–6
metal matrix, 252–6 collagen, 3
664 Index

Cortical bone (cont.) elastomers, 291

combined loading, 9 ligament, 55
compression, 8 osseointegration, 585
compressive modulus, 6–7 ply, 216
constituents, 3, 4 repassivation, 153–4
creep, 10 solubility, 496
density, 4, 4 tendon, 55
dry testing, 5 thermoplastic resins, 211
elastic moduli, 7 thermoset resins, 211
electromechanical behaviour, 5 tribosystem, 487
fatigue, 10–11 Degradation, 608, 618–19
health of donors, 5–6 Density
mineral, 4 brain tissues, 69
mineralization, 6, 10 cortical bone, 4, 4
physical properties, 4–5 Dental
Poisson’s ratios, 7, 7 alloys
remodelling, 11, 11 AgPd, 197
shear moduli, 7, 7 composition of, 195, 196
small specimens, 7 high gold containing, 195, 200
stiffness, 6–9 low gold containing, 195, 201
strain rate, 6, 8–10 Pd, 197
effects, 10 precious, 198
strength, 9, 9–11 amalgams, 151
tensile modulus, 6 implants, 151, 152, 157
time to failure, 10 restoration, 191–203
viscoelastic damage, 8–6 Dentin, 23–34
Young’s modulus, 11 apatite in, 27
cp-titanium, chemical composition, 168 apatite parameters, 27
Creep Ca/P molar ratio, 27
bovine bone, 10 compressive strength, 29
cortical bone, 10 constituents, 24
modulus, brain tissues, 69, 72–3 critical surface tensions, 33
Critical stress intensity factor, 400 elastic moduli, 28
Critical surface tensions elemental composition, 24
dentin, 33 flexural strength, 29
enamel, 33 fracture toughness, 30
Curing agents mineral density, 24
amine, 213–4 organic density, 24
anhydride, 214 permeability, 31
accelerators, 214 stress, 29
epoxide resins, 213–5 tensile strength, 29
polyamide, 214–5 toughness, 30
Cutting and milling titanium, 173 viscoelastic properties, 28
Cytokines, testing for, 598–9, 599 water density, 24
wetability, 33
work of fracture, 30
D Dentino-enamel junction (DEJ), 23
DAP, properties, 237 Descaling titanium and titanium alloys, 171–2
Debris Diamond, 557
accumulation, 608 Diaphysis, vs. metaphysis, 7, 7, 9–10
metallic, 608, 609 Diencephalon, 67
Definitions Diffusion properties, brain tissues, 74
bioactivity, 447 Diffusion welding, titanium, 174
biocompatibility, 564–5 Digital imaging correlation, 443–4
Index 665

Diluents, 212–3 properties, 301

non reactive, 213 trade names, 301
reactive, 212–3 polyurethane, 301–5
Dimethypolysiloxane, 631 polyurethane-based, 294
Dissolution, in vitro, 496–9 manufacturers, 297–9
Doppler results, 63–4, 64 properties, 297–9
Double cantilever beam, ceramic materials, trade names, 297–9
396–8, 417–20 PVC, 309–10
Double torsion method, 414–7 silicone, 310–1
Drag coefficients, menisci, 47–8 high consistency, 311
Drug delivery, implantable, 565 LIM, 311
Dry testing, cortical bone, 5 silicone adhesives
Ductile behaviour, ceramic, 353 manufacturers, 316
Dynamic viscosity, 651, 661 properties, 316
trade names, 316
sterilization, 335
E methods, 332–3
Ear bones, 6 styrene-based
ECDEL, 293 manufacturers, 306
Elastic properties, 306
behaviour trade names, 306
blood vessels, 77 thermoplastic, 292–10
lymphatic vessels, 77 types of, 295–34
cartilage, 38 water absorption, 333
properties Electrical conductivity, brain tissues, 69, 74
invertebral discs, 48–9, 49 Electromechanical behaviour, cortical bone, 5
menisci, 48–9, 49 Electron beam welding, titanium, 174
Elastic shear modulus of vocal fold, 650, 653 Elemental composition
Elastin/collagen ratio, 82 dentin, 24
Elastomers, 291–37 enamel, 24
ASTM standards, 335, 336 surface enamel, 25–6
biocompatibility, 293, 318–31, 335 whole enamel, 25–6
copolyester Elgiloy, 164
manufacturers, 296 Elliptical cracks, ceramic materials, 398–400
properties, 296 Enamel, 23–34
trade names, 296 apatite in, 27
crosslinked, 310–1 apatite parameters, 27
definition, 291 Ca/P molar ratio, 27
dispersions compressive strength, 28–9
manufacturers, 314–5 critical surface tensions, 33
properties, 314–5 elastic moduli, 28
trade names, 314–5 elemental composition, 24
equivalent, 317 fracture toughness, 30
silicone, 317, 334 hardness of fracture, 31
HC silicone mineral density, 24
manufacturers, 312 organic density, 24
properties, 312 stress, 28
trade names, 312 surface, elemental composition, 25–6
LIM silicone tensile strength, 28
manufacturers, 313 toughness, 30
properties, 313 water density, 24
trade names, 313 wetability, 32
polypropylene-based whole, elemental composition, 26
manufacturers, 301 work of fracture, 30
666 Index

Endoprostheses G
hip, 151 Gauge length, 357
knee, 151 Glass ceramics, 447–54
Environmental stress cracking, 264 clinical uses, 453
Epiligament/epitenon, 55 physical properties, 450, 451, 453
Epoxy resins Glass fibers, 206–7, 208–9
curing agents, 212–5 advantages, 208
description, 240 chemical composition, 206
end uses, 240 electrical properties, 206–7
properties, 244 grades, 209
Equivalence, 317–35 mechanical properties, 206
Extracellular matrix (ECM) of lamina optical properties, 207
propria, 645 physical properties, 206
thermal properties, 206
Glassy carbons, 556
F fabrication, 556
Face-centered cubic (FCC), 162, 167 mechanical properties, 252
Fatigue, 178, 178–81, 179, 180, 181 shrinkage, 556
cortical bone, 10–11 Glycosaminoglycan (GAG), 646
porosity, 11 Graphite-fiber-reinforced
wear, 456, 458, 473, 475 thermoplastics, properties, 235
Young’s modulus, 11 Graphite fibers, 208, 209
Female, fluid volume, 120 Graphites, 554–5
Femoral articulating surface, 45–6 colloidal, 555
Fiberglass, properties, 238 manufacturing processes, 555
Fiber-reinforced plastic pipe, natural, 554–5
220–1 synthesized, 555
Fibrocartilage, 38, 45–51 Griffith’s theory, 381–4
anisotropy, 45
composition, 45–7, 46, 47
mechanical properties, 42 H
structure, 45–7, 46, 47 Hand lay-up, 220
Fibrosis, 575 Hanks’ buffered saline solution, 154
Filament winding, 224 HAp, 496
storage tanks, 225 bioresorption, 503
Fillers degradation, 499–505
brazing temperatures, 199 dissolution rate, 497
chemical composition, 199 highly porous, 503
Finite element analysis (FEA), ceramic from marine coral, 502
materials, 433–5 precipitation, 535
Flexural strength, dentin, 29 seeds, 536
Fluid volume Haptens, immune response to, 600
female, 120 Havar
male, 120 chemical composition, 165
Fractures, internal fixation mechanical properties, 166
malignancies, 614 Haversian
Fracture toughness remodelling, 11
dentin, 30 systems, 4
enamel, 30 HDPE, properties, 230
Fracture toughness, cancellous bone, 19 Health of donors, cortical bone, 5–6
Fracture work, ceramic materials, Heart
392–6 pacemakers, 152
Friction coefficients, implants, 464–71 valves, 551, 552
Index 667

Hip Inflammation, 573, 583–5

replacement, accelerated test data, 251 characteristics, 573, 583
Histochemical techniques, 599–600 chronic, 574
Hooke’s law, 388 Ingrowth behaviour, 156–7
Hot isotactic pressing (HIP), 160 Injection molding, 223–4
Human advantages, 223
adjuvant disease, 633 Injured (T/L) function, 60
calf, stress relaxation, 64 Intervertebral discs, 45
dentition, structure of, 23 adult
Hyaline cartilage, 37 shape, 47
Hyaluronic acid (HA), 647–8 size, 47
Hydraulic annulus fibrosus, 46
conductivity, blood vessels, 77 elastic properties, 48–9, 49
permeability, menisci, 47–8 nucleus pulposus, 46
Hydrocarbon polymers, stability, 248 shape, 46, 47
Hydrophilic polymers, 624 stiffness, 50
Hydrothane, 301 strength, 50
Hydroxyapatite. See HAp viscoelastic behaviour, 49–51, 51
Hydroxy carbonate apatite, 449 Intraocular crystalline lens. See Intraocular
rate of formation, 449 lens (IOL)
Hydroxyl radical, 156 Intraocular lens (IOL), 103. See also Lens
Hypertension, 81, 83 Investment casting, 160
Ion implantation, 154
nitrogen in titanium, 154
I Ischemia, 622, 623
Iliac vessels, 95
Immune
complex reactions, 595 J
responses, 602–3 Joint replacements
implants, 593–603 associated malignancies, 612, 613
Implants
alumina, 339–40
bone, in situ, 583–4 K
calcium deposition, 533, 534 Keratometry, 139
cancer, 607–17 Kevlar 49, 209
carcinogenicity, 610–1 electrical properties, 242
cemented, 582 mechanical properties, 207
cement-free, 582, 583, 584–5 shear properties, 240–1
dental, 151, 152, 157 thermal properties, 207, 242
exposure time, 157 Kraton, 310
friction coefficient, 464–71
hard tissue response, 581–2
immune response, 593–603 L
metallic Lamina propria, 645, 646
corrosion, 509–45 extracellular matrix (ECM), 645, 647
surface roughness, 157 Laminates
oral, 588 angle ply, 219
orthopaedic, 583, 584, 586, 588 properties, 216–20
penetration rates, 486 Large veins, morphometric properties, 78
porous, 157 Latex, immune response to, 600
retrieved, 518, 519 LDPE, properties, 242
site repairs, 576–7 Lead/graphite, properties, 256
tribological conditions, 464–71, 472–85 Length, arteries, 78
668 Index

Lens, 103–11 Malignant fibrous histiocytoma (MFA),

capsule, mechanical properties, 109 610, 615
cataractous, 105 Mass, brain tissues, 68
composition, 104 Matched die molding, 221–4
density, 107 Materials
dimensional variation with age, 107 pre-screening, 517
electrical properties, 110 and tissue interaction, 567
force of contraction, 109 Matrix, materials, 210, 253
hardness of cataractous, 108 Mechanical stimuli, brain tissues, 68
inorganic ion content, 106 Medicaflex, 301
normal Membranes, 626
inorganic ion content, 106 Menisci, 45–51
water content, 105 Mesencephalon, 67
organic content, 106 Metal
physicochemical characteristics, 104–10 cations, uptake efficiencies, 520
refractive index, 107 oxides, physico-chemical properties, 536, 538
shear modulus of, 108 Metallic
spring constants, 110 biomaterials, 151–8, 151–2
tensile modulus, 108 adhesion strength, 152
thickness of capsule, 107 alloys and applications, 152
transmissivity, 108 bioadhesion, 154
viscoelastic properties, 110 biocompatibility, 152, 154
volume, 107 biofunctionality, 154–6
water content, 105 CoCr alloys (see Cobalt-chromium
weight, 107 (CoCr) alloys)
Ligaments, 560 corrosion, 155
Ligaments and tendons (T/L) corrosion: and inertness, 153–4
composition, 56, 56–7 corrosion: measurements, 153
hierarchical structure, 56 densities, 153
injured function, 60 generations of, 151
mechanical behavior, 57 inorganic reaction, 155
mechanical properties of, 59 mechanical properties, 153
normal function, 57–9 Nitinol, 152
stiffness, 58 organic reaction, 155
structural properties, 57 orthopedics, 153
structure, 55–6 oxide layer, 151
viscoelastic behavior, 58 polarization resistance, 153
Linear low density polyethylene, 263 primary corrosion products, 153
LTI carbon, mechanical properties, 252 repassivation, 153
Lubrication, 487 structural applications, 151
Lymph testing, 153
inorganic content, 117–18 thermodynamic stability, 156
organic content, 118–19 titanium, 152
Lymphocyte, 593, 594 Young’s modulus, 151–2, 155
proliferation, 597 implants
transformation factor, 597 and cancer, 616
Lymphokines, testing for, 598 corrosion, 509–45
passive films, 536–42, 539, 540,
541, 542
M ions, and calcium phosphates, 533–6, 534,
Macrophage migration inhibition, 633 535, 536
Magnesium-graphite, mechanical properties, 256 materials
Male, fluid volume, 120 alloys, 512–4, 513, 514
Malignancies, osseous, 615 in biological samples, 517
Index 669

Metals, 625 Osseointegration, 582, 584–7

in blood vessels, 625 calcium phosphates, 587
immune response to, 602 ceramics, 587
thrombogenic, 625 craniofacial skeleton, 586
Metaphysis, vs. diaphysis, 7, 7, 9–10 definition, 585
Microstructure, cancellous bone, 15 interfaces, 586
Mineral oral implants, 585, 586
cortical bone, 4 polymers, 586–7
density press-fit fixation, 586
dentin, 24 surface irregularities, 585
enamel, 24 Ossicles, 6. See also Ear bones
Mineralization, cortical bone, 6, 10 Osteocalcin, 4
Morphometries Osteolysis, 608, 618
pulmonary arteries, 79 Osteonectin, 4
systemic circulation, 78 Osteons
Muscle bending modulus, 8, 8
mechanical properties, 63–5, 64, 65 compression modulus, 8, 8
preservation, 63 secondary, 4, 8, 8
and skin, 63–5 strength, 9
specimen preparation, 63 tension modulus, 8, 8
tissue handling, 63 torsional modulus, 8, 8
Osteosarcoma, 610
Oxide ceramics, 340, 341
N applications, 341
Natural rubber, 310 evaluation, 340
Neoplasia, 609–10 Oxide formation, 541
malignant, 609–10 Co-Cr-Mo alloy, 542, 543
Neutrophils, 573 Oxide layer, metallic biomaterials, 151
Niobium, 152, 153
Nitinol, 152, 186–7
Ni45Ti, properties, 186 P
Noble metals, 194–201, 195–201 PBT, properties, 236
corrosion and wear, 201 PEEK, 272
heat treatments, 199, 200 characteristics, 272
polarization current, 201 chemical structure, 276
polarization resistance, 201 processing conditions, 284
Non-collagenous protein (NCP), 3 properties, 230
Normal (T/L) function, 57–59 trade names, 289
Normal human lens Pellethane, 304
inorganic ion content, 106 PEl, properties, 231
water content, 105 Permeability
Nucleus pulposus, 46 articular cartilage, 41
dentin, 31
PES, 211
O PET, 271–2
Organic density chemical structure, 276
dentin, 24 physical properties, 271
enamel, 24 processing conditions, 284
Orowan–Petch relation, 354 properties, 236
Orowan’s argument, 380 trade names, 289
Orowan value, 400 Petrosals, 6. See also Temporal bones
Orthopaedics, 495–505 Phagocytosis, 573–4
bioactive ceramics, 495–505 frustrated, 574
Orthopedic alloys, 154 Physical properties, cortical bone, 4–5
670 Index

Physiochemical properties chemical structure, 275

blood, 115 processing conditions, 283
cerebrospinal fluid, 115 properties, 271
plasma, 115 sheets, properties of, 231
synovial fluid, 115 trade names, 289
tear fluid, 115 Polyester
whole blood, 115 fiberglass-reinforced, properties, 244
PIE, 211 properties, 238
Plasma Polyester resins, 215–6
concentration, 121–2 accelerators, 216
diffusion coefficient, 122 catalysts, 215–6
extinction coefficient, 122 Polyetheretherketones. See PEEK
half-life, 121–2 Polyetherimides. See PEI
inorganic content, 117–18 Polyether sulfones. See PES
molecular weight, 121–2 Polyethylene, 263–4
organic content, 118–19 chemical structure, 274
proteins plastic, carcinogenicity, 610
content, 119–20 processing conditions, 280
properties, 121–2 trade names, 285
sedimentation constant, 122 wear debris, 518
Ply Polyfluorocarbons, properties, 241
definition, 216 Polyimides, properties, 237
orientated, 216–7 Polymer/carbon, mechanical
rotated, 217 properties, 251
transverse, failure, 219 Polymers
PMMA, properties, 241, 269 manufacturer, 251
Pneumonitis, 632 trade names, 251
Poisson’s ratios Polymethylmethacrylate
articular cartilage, 41 carcinogenicity, 610
brain tissues, 69 immune response to, 601–2
cancellous bone, 18 Polyphenylene sulfides. See PPS
cortical bone, 7, 7 Polypropylene, 264–5
of human vocal folds, 651 chemical structure, 274
Polarization resistance, 182 processing conditions, 280
Polyacetal, 269–70 properties, 230
chemical structure, 275 trade names, 286
processing conditions, 283 Polysiloxane, immune response to, 601
solvents, 270 Polysulfone, 272–3
trade names, 289 characteristics, 273
Polyacrylates, 268–9 chemical structure, 276
chemical structure, 275 processing conditions, 284
processing conditions, 283 properties, 233
trade names, 288 trade names, 289
Polyamide-imide, 210 Polytetrafluoroethylene. See also PTFE
properties, 232 chemical structure, 274
Polyamides, 210, 268 processing conditions, 281
aliphatic, 268 trade names, 287
chemical structure, 275 Polyurethane, 265–6
processing conditions, 282 blood pumps, 624
trade names, 288 chemical structure, 274
Polyarylate, properties, 232 elastomers, 265
Polyblend, polyurethane, 304 immune response to, 601
Polycarbonate, 210, 270–1 polycarbonate-based
chemical resistance, 270–1 manufacturers, 300
Index 671

properties, 300 tensile strength, 11

trade names, 300 ultimate strain, 11
processing conditions, 281 Young’s modulus, 11
ventricular assist devices, 624 Repassivation, definition, 153–4
trade names, 286–7 Residual compression, UHMWPE, 486
Polyvinylchloride. See also PVC Resins
chemical structure, 275 cured polyester, propertie, 239
Porosity, fatigue, 11 epoxy, 212
Potential-pH diagrams, 510–2, 511 description, 240
limitations, 510 end uses, 240
PPS, 211 polyester, 215–6
fabric reinforce, 222 properties, 238
properties, 232 thermoplastic, 210
Presbyopia, 103 thermoset, 210
Preservation vinyl ester, 211–2
muscle, 63 Resin system, choice of, 226, 229
skin, 63 Responses
Press-fit fixation, 586 allergic, 593
knee replacement, 586 cell mediated, 597–600
Prostheses, heart valves, 551 cellular necrosis, 572
Proteins chemically induced, 571
adsorption, 566 hard tissue, 581–91
coagulation, 122 hypersensitivity, 593
complement system, 123 immune, 572, 602–3
PTFE, 266 local host, 567–9
properties, 241 minimal, 571
Pulmonary arteries, morphometrics, 79 neoplastic, 572
Pultrusion, 229 physically induced, 572
PVC, 267 soft tissue, 571–7
Elastichem, 309 silicone, 631–41
elastomers, 309 Retrieved implants, 518, 519
Ellay, 309 oral, 587
Geon, 309 Reynolds number, 78
Multichem, 309 aorta, 78
plasticized, 267 arteries, 78
manufacturers, 302–4 capillaries, 78
properties, 302–4 vena cava, 78
trade names, 302–4 Rheometry, 105
processing conditions, 282 Rheumatoid arthritis, breast implants, 640
Teknor Apex, 310
trade names, 287
unplasticized, 267 S
Santoprene, 293, 295
Sarlink, 308–9
R Secondary, osteons, 4, 8, 8
Reaction bonded silicon nitride, axial Serum
strength, 257 levels, total hip arthroplasty, 608, 618
Refractive index of human lens, 107 metal concentrations, 518, 521
Reinforcement filaments Shape
manufacturing technique, 208 brain tissues, 68
properties, 208 intervertebral discs, 46, 47
Remodelling Shear modulus
cortical bone, 11, 11 brain tissues, 69, 70–71
Haversian, 11 cortical bone, 7, 7
672 Index

Shear modulus (cont.) intervertebral discs, 50

of human lens, 108 osteons, 9
trabecular bone, 16, 18 testing for ceramic material
Shear properties 3 and 4 point bending, 360–5
articular cartilage, 40, 40 beam theory, 357–60
dentin, 29 ceramic machining, 357
Shear stress, 155 flexion and flexure testing, 357
Shrinkage temperature, collagen, 3 nature of test, 374–7
SiC/RBSN, axial strength, 257 processing of failure data, 369–72
Silicones ranking, 372–4
breast implants, 640 significance of test results, 368–9
cosmetic surgery, 632 tensile, 356–7
granulomatous hepatitis, 632 test configurations, 365–8
immunologic reactions, 631–2 test pieces, 357
migration, 632 volume of test piece, 377
soft tissue response, 631–41 Stress
systemic reactions, 632–4 corrosion, ceramic materials, 426–8
Silicon nitride, properties, 257 dentin, 29
Sintering, 515 enamel, 28
Skeletal reconstructions, 151 intensity factor, 381
Skin relaxation
elastic moduli, 63–4 human calf, 64
forehead, indentation, 64 Structure, fibrocartilage, 45–7, 46, 47
mechanical properties, 63–5, 64, 65 Styrene polymer, properties, 236
and muscle, 63–5 Superalloy MA956, 530
preservation, 63 polarization curves, 530
specimen preparation, 63 Surfaces, blood contacting, 627
testing, 599 Surface treatments, 528–33, 529, 530, 531,
tissue handling, 63 532, 533, 539, 541
Slip movement, 353 Surgery, tissue overheating, 588
Soft tissues Sutures
indentation, 64–5 absorbable, 565
stiffness, 64 mechanical propertie, 252
Soldering and brazing titanium, 172–3 Sweat
Solubility corrosion in, 522–3, 523
calcium phosphate, 497 ISO, electrochemical parameters, 523
definition, 496 Synovial fluid, 116
isotherms, 498 cellular composition, 117
product constant, 496 inorganic content, 117–18
Specific immune response, 593–5 metallic concentrations, 521
Specimen preparation organic content, 118–19
muscle, 63 protein content, 119–20
skin, 63 Synthetic polymers, immune response to, 601–2
Spongy bone. See Cancellous bone Systemic
Spray-up, 220 circulation, canine, morphometric
Spring constants of human lens, 110 properties, 80
Sterilization, 541 sclerosis, 633, 635
Stiffness
cortical bone, 6–9
intervertebral discs, 50 T
Strain Tantalum, 155, 156, 514
rate, cortical bone, 8, 10 corrosion resistance, 514
Strength pentoxide, thickness, 527
cortical bone, 9 plates, 156
Index 673

Ta-oxide, 156 dissolution, 533

TCP, 496 electrochemical characteristics, 524
bioresorption, 503 femoral components, 518
degradation, 499–505 laser surface alloying, 528
dissolution rate, 497 passivity currents, 526, 536
T cytoxic cell, 594 and pure titanium, implants, 532
Tear Ti-6AI-4V-50 v/o BORSIC, properties, 255
fluid, 116 Ti-6AI-4V-28 v/o SiC, tensile properties, 255
inorganic content, 117–18 Ti6A14V
organic content, 118–19 ageing, 177
protein content, 119–20 fatigue strength, 178
Tecoflex, polyurethane, 304–5 fretting corrosion, 184
Tecothane, 305 notch factor, 178
Teknor Apex, 310 plasma nitriding, 177
Temporal bones, 6 rotating bending failure, 181
Tendon, 55–61 rotating bending strength, 179
definitions, 55 solution treatment, 177
Tendons, 560 wear properties, 185
Tensile Time to failure, cortical bone, 10
modulus Tissues
annulus fibrosus, 50 density, 15
cortical bone, 6 granulation, 575
menisci, 49 handling
modulus articular cartilage, 39–40, 40 muscle, 63
strength skin, 63
dentin, 29 metals in, 519, 521–2
enamel, 28 vascularized, 574, 575
menisci, 48–9, 49 TiTa30, 152
remodelling, 11 thermal coefficient, 152
Tension, modulus, osteons, 8, 8 Titanium, 156, 167–87
Texin, 305 accumulation, 517, 518
Thermal properties, brain tissues, 74, 74 alloys, 167–87
Thermoforming, 222–3 biological properties, 185
Thermoplastics, 210–1, 261–89 chemical composition, 167, 168
amorphous, 262 corrosion and wear, 182, 192–5, 183,
biocompatibility, 262 184, 185
branching, 262 fatigue, 178, 178–81, 179, 180, 181
mechanical properties, 277–8 fracture toughness, 177
molecular weight, 262 hip endoprostheses, 178
physical properties, 277 mechanical properties, 174, 175, 176
processability, 261 physical properties, 168–9, 169
states of, 262 and alloys, 152
thermal properties, 278–9 anodic, electrochemical data, 183
vulcanizates, 293, 295 anodized, corrosion rate, 527, 537
manufacturers, 295 biological properties, 185
properties, 295 cold deformation, 175
trade names, 297 corrosion, 511
Thornel/Kevlar, properties, 243 formation, 185
Thromboembolism, 622, 623, 626 resistance, 526, 526–7, 527
Thromboresistance, 626 and wear, 182, 192–5, 183, 184, 185
Thrombus, 622, 623 cutting, 172
Ti-6AI-4V deformation temperatures, 170
contact angles, 543 descaling, 171–2
depassivation, 528 dissolution, 539, 540
674 Index

Titanium (cont.) tissue damage, 19–20

hot working, 169–70 tissue-level mechanical properties, 20
immunity, 511 and ultrastructure, 16
implant in bone, 539 viscoelastic and fatigue properties,
international standards, 167 18–19
machining, 172 Young’s and shear modulus, 16, 18
mechanical properties, 174, 174–7, 175, Transfer molding, 222–3
176, 177 Transmissivity of human lens, 108
milling, 172 Tribologic properties
oxide articular cartilage, 41–2
films, 542 Tribosystem, definition, 487
sterilization, 541 Tricalcium phosphate. See TCP
thickness, 527 Tubing, 624
passivation, 511 Tumorigenesis, solid state, 572
corrosion rate, 527, 537 Tumors
physical properties, 168–9, 169 analysis of, 613–5, 614
polarization current, 182 implant-related, 611–3, 614
processing, 169–74, 171, 172, 173 prosthetic replacement, 614
pure titanium implant, 614
potential-time curves, 525
pure
physical properties, 168 U
repassivation time, 182, 183 UHMWPE, 263, 434
sheet working, 170, 170–1 Ultimate
soldering and brazing, 172–3 strain, remodelling, 11
surface characteristics, 543 stress, annulus fibrosus, 50
thermal heating, 532 tensile strength (UTS), 58, 368
welding, 173–4 Ultra high molecular weight polyethylene.
Ti-15Zr-4Nb-2Ta-0.2Pd, 514 See UHMWPE
Torsional modulus, osteons, 8, 8 UNS R30005, 164–6
Total hip Urine, metal concentrations, 521
arthroplasty, 486, 608, 618 Usui process, 227
cancer risk, 616
endoprosthesis, 340
replacements, 154 V
Toughness Vanadium, 152, 155
dentin, 30 Vapor-deposited carbon, mechanical
enamel, 30 properties, 252
Toxicity Vapour phase coatings, 552, 557
additive, 244, 245 Vascular grafts, 627
plasticizers, 245 Vascular wall, layers, 85
polymer stabilizers, 244, 245 Veins
Trabeculae, 15 canine jugular, 92, 93
Trabecular alignment, 18 human saphenous, 92, 93
Trabecular bone. See Cancellous bone mechanical properties, 92–3, 93, 94
compression, 16 tensile properties, 92–3, 93, 94
fracture toughness, 19 Vena cava, morphometric properties, 78
mechanical properties, 16, 17 Venous tissues
microstructure, 15 changes with age, 84, 85
Poisson’s ratio, 18 constituents, 84, 84
stiffness and strength, 16–18 Ventricular system, 67
strain, 18 Vicker’s diamond pyramid, 423, 426
surface-to-volume ratio, 15 Vicker’s indenter, 425
tissue composition, 16 Vinyl ester resins, 211–2
Index 675

Viscoelastic Wear, 455–87

behaviour abrasive, 456
intervertebral discs, 49–51, 51 Rabinowicz model, 457
menisci, 49–51, 51 Zum Gahr model, 457
properties, trabecular bone, 18–19 adhesive, 456
Vitallium, 152, 157, 614 Archard equation, 457
Vitreous body, 125–30 Hornbogen equation, 457
age of protein content, 129 artificial joints, 455
animal, 126 ASTM standards, 462–3
axial length, 129 corrosive, 456
compliance, 130 Quinn model, 458
gel and liquid volume, 130 debris, release of, 568
inorganic ions content, 128 delamination, 456
nature of, 125 fatigue, 456
organic content, 128 Halling model, 458
physical properties, 127 fretting, 456
radiation, 130 linear clinical, 486, 486–7
rheology, 130 rates, 472–85
viscosity, 130 testing
water content, 127 machines, 461
Vitreous humor. See Vitreous body test apparatus, 461
Vocal folds, 645 in vitro, 459–85
animal studies detailing injury on, 649, 652 Weibull equation, ceramic materials, 435–41
biomaterial injections, 652–3 Welding, 173–4
collagen and fat injections, 653–4 diffusion, 174
composition, 646–7 electron beam, 174
collagen, 647 spot, 174
elastin, 647 Welding titanium, 173–4
hyaluronic acid, 647–8 Wetability
interstitial proteins, 648 dentin, 33
voice disorders, 648, 649 enamel, 32
dynamic viscosity, 651, 661 Wet lay-up, 224–6
elastic shear modulus, 650, 653 Whole blood
mechanical properties cellular composition, 117
dynamic viscosity, 651 inorganic content, 117–18
elastic shear modulus, 650 organic content, 118–19
general, 649 Whole enamel, elemental composition,
Poisson’s ratio, 651 25–6
rheology, 650 Work of fracture
vocal fold scarring, 652 dentin, 30
modified hyaluronic acid, 654–6 enamel, 30
Young’s modulus, 649 Wounds, healing, 575
Volume fraction, trabecular bone, 15
von Mises criterion, 353
Y
Young’s modulus
W cortical bone, 7
Wall thickness, 78 fatigue, 11
Water metallic biomaterials, 151–2, 155
content, cartilage, 37 remodelling, 11
density titanium and titanium alloys, 168, 169
dentin, 24 vocal folds, 649
enamel, 24 Young’s modulus
permeability, blood vessels, 77 trabecular bone, 16, 18
676 Index

Z engineering properties, 342

Zinc/graphite, mechanical properties, 256 fracture toughness, 346
Zinc, mechanical properties, 256 mechanical properties, 341, 344
Zirconia, 344–7, 345, 354–5 partially stabilized, 344
biocompatibility, 348 standards, 343
chemical stability, 346 tetragonal polycrystals, 344
clinical performance, 346 wear resistance, 346