02whole v1

t5'3.
'tS
METHODS OF TEST¡NG THE MECHANICAL PROPERTIES OF

ORTHODONTIC WIRES
A report submitted in partial fulfilment of the requirements for the degree

of Master of Dental SurgerY
BY
KATHERINE RUTH ALLEN B.D.S.
Department of Dentistry
Faculty of Dentistry
The University of Adelaide
South Australia
April 1994
TABLE OF CONTENTS
PAGE
NOMENCLATURE VI
LIST OF FIGURES
LIST OF TABLES xil
SUMMARY XVI
SIGNED STATEMENT XVIII
ACKNOWLEDGEMENTS xtx
Chapter 1 ¡rurnooucrtoN 1
Chapter 2 IIERATURE REVIEw 3
MECHANICAL PROPERTIES OF ORTHODONTIC TOOTH 3
MOVEMENT
THE IDEAL ARCHWIRE 3
TERMINOLOGY USED WHEN DISCUSSING THE MECHANICAL 4
PROPERTIES OF WIRES
PRODUCT]ON OF WIRES 11
THE CRYSTALLIZATION OF MOLTEN METALS 11
THE EFFECTS OF COLD WORKING AND ANNEALING ON THE 12
CRYSTAL STRUCTURE
HISTORY OF ORTHODONTIC WIRES 13
THE CHEMICAL COMPOSITION AND MECHANICAL 15
PROPERTIES OF SPECIFIC WIRES
GOLD 15
STAINLESS STEEL 15
M ULTIBRAIDED STAINLESS STEEL 19
ll
CHROMIUM.COBALT 20
NICKEL-TITANIUM 21
BETA-TITANIUM 27
ALPHA-TITANIUM 28
SUMMARY 29
Chapter 3 AtMs oF THE lNVEsrlGATloNS 30
Chapter 4 MATERIALS AND METHoDS 32
CONVENTIONAL TENSILE TESTS 32
BEND TESTS 32
ELECTROPOLISHING 38
CANTILEVER TESTS 40
THREE POINT BEND TESTS 45
USE OF LOAD AND UNLOAD DATA 47
RESONANCE 48
SPEED OF SOUND 50
MICROHARDNESS TEST 52
MACROHARDNESS TEST 54
SURFACE APPEARANCE 54
DETAILED MICROSTRUCTURAL DATA 56
Chapter 5 REsuLTs AND DlscussloN 67
ELASTIC MODULUS VALUES 67
CONVENTIONAL TENSILE TESTS 69
BEND TESTS 75
lll
ARCHFORM 75
CANTILEVER TESTS B2
OTHER TECHNIOUES 87
TESTING OF ANNEALED SPECIMENS 91
OTHER WIRES 94
RESONANCE 9B
VELOCITY OF SOUND 103
SUMMARY OF ELASTIC MODULUS TESTS 108
A COMPARISON OF ELASTIC MODULUS VALUES DERIVED 109
FROM DIFFERENT TECHNIOUES USING LOAD DATA
EFFECT OF DIAMETER CHANGE ON ELASTIC MODULUS 111
VALUE
THE EFFECT OF DATA MOD¡FICATION ON ELASTIC 111
MODULUS VALUES
MICROHARDNESS TESTS 114
MACROHARDNESS TESTS 116
SURFACE APPEARANCE OF WIRES 117
SURFACE EFFECTS OF PULSE STRAIGHTENING 117
SURFACE CHANGES AFTER ELECTROPOLISHING 117
FRACTOGRAPHY 120
DETAILED MICROSTRUCTURAL DATA 127
Chapter 6 coNclusloNs 132
Chapter 7 FUTURE RESEARcH

135
lv
BIBLIOGRAPHY
137
APPENDICES
Vol II
NOMENCLATURE
Unless otherwise Stated the symbols have the following meaning:
A area in mm squared
at alpha-titanium
ct length or distance
b breadth
br beta-titanium
d diameter
d the length of the diagonal of a square in micrometers
D length of the diagonal in millimetres
DPN Diamond Pyramidal Hardness Number
E elastic modulus (or Young's modulus)
ep electropolished
ep 25 electropolished for 25 seconds
F measured load
"fn
frequency of vibration .
GPa Giga Pascal
h height
HV Vickers hardness number
Hz hertz (cycles per second)
I moment of inertia
I second moment of area
I distance
m MASS
m metres
nt nickel-titanium
P applied load
ps pulse straightened
rpm revolutions per minute
se nt superelastic nickel-titanium
SEM scanning electron microscoPe
SS stainless steel
ssp stainless steel premium
sspp stainless steel premium Plus
sspp ps stainless steel premium plus, pulse straightened
SSS stainless steel suPreme
sssp stainless steel special Plus
vl
TEM transmission electron microscope
TMA titanium molybdenum alloY
TTR transition temperature range
v velocity
v extension
YS yield strength
o stress
t strain
t maximum flexibility
p density
pm micrometre
ps microsecond
Dimensions
018 = O.O18" (O.45mm)
Note: Some orthodontic wires are supplied in imperial units but all
scientific studies should use metric.
vl1
LIST OF FIGURES
FIGURE SUBJECT PAGE
1 Stress-strain curve illustrating proportional limit, yield 7
strength, UTS, failure point, elastic modulus.
2 Stress-strain curve illustrating resilience and formability' 1O
3 Stress-strain curve illustrating differences between 26

stainless steel, nickel-titanium and superelastic nickel-
titanium.
4 A universal Instron testing machine with an 33

extensometer in position
5 A wire specimen wrapped around an aluminium casing 34

prior to tensile testing
6 A milled brass wire holder in a stand allowing 36

adjustments in the vertical Plane
7 steel holder glued to the Mettler 61OO pan with anterior 37

section of archwire in a groove in steel support
I Electropolishing unit (Rocky Mountain 70O,USA) 39
9 A cantilever bend test us¡ng a knife edge to support the 42

free end of the specimen
10 A 3-point bend test 46
11 Apparatus for resonance testing 49
12 Apparatus for measuring speed of sound along a wire 51
13 A photomicrograph of microhardness indentations near 53

the edge of the specimen
14 Curing characteristics of G-1 Epoxy used for preparing 58

TEM cross sections
15 Disc cutting and grinding for cross-sectional TEM 58

specimen
vrll
16 Precision dimple grinder and polisher 59
17a Magnified view of the grinding wheel and mount on 61
dimple grinder
17b Dimpled specimen 61
18a lon milling of a cross-sectional specimen 62
18b Effect of OctagunTM voltage on specimen thinning rate. 62
18c Specimen current versus beam angle for argon milling 62

of copper.
19 Preparation of a single wire specimen to minimise the 64

epoxy
20 E.A.Fischione model 12O-23O twin jet electropolishing 66

unit
21 Diagramatic stress-strain curve illustrating tangents to 73

the curve which may be chosen to calculate the elastic
modulus.
22 Diameter changes with continuous electropolishing of 79

sssp 020
23 Diameter changes with cumulative electropolishing of 80

sssp O2O
24 Wave forms as seen on oscilloscope Screen in velocity 1O4

of sound testing
25 Modification of stress-strain data to improve the fit of 112

the straight line for load data
26 Modification of stress-strain data to improve the fit of 1 13

the stra¡ght line for unload data
27 SEM micrograph (75x magnification) of as-received 1 18

sspp 018 wire surface
lx
28 SEM micrograph (75x magnification) of as-received 118
sspp ps O18 wire surface
29 SEM micrograph (75x magnification) of as-received 119

sspp O2O wire surface
30 SEM micrograph (3OOx magnification) of as-received 119

sspp O2O wire surface
31 SEM micrograph (75x magnification) of sspp 020 wire 121

surface after 25" electropolishing
32 SEM micrograph (75x magnification) of sspp 020 wire 121

33 SEM micrograph (75x magnification) of sspp 02O wire 122

surface after 1 50" electropolishing
34 SEM micrograph (3OOx magnification) of sspp O2O wire ' 122

35 Number of bends possible until point of wire fracture 123
36 SEM micrograph showing "tag" formation after 124

deliberate fracture of a wire specimen.
37 SEM micrograph showing variation in the extent of 125

"tag" formation.
38 SEM micrograph showing variation in the extent of 125

"tag" formation.
39 SEM micrograph of a fractured wire surface showing 126

differences in the direction of the crack path.
40 SEM micrograph of a fractured specimen showing 126

distinct steps on the surface.
41 SEM micrograph of a fractured wire illustrating a zone 128

normal or possibly inclined in the opposite direction to
the "tag".
x
42 SEM micrograph of a fractured wire surface where a 128
pronounced crack separates the "tag" and zone normal
to the wire surface.

steplike growth of a central crack.
44 SEM of a "tag" showing the elongated grain structure. 129

fine flat dimples typical of ductile fracture.
46 A printout of the ED analysis of a stainless steel wire. 130
xt
LIST OF TABLES
TABLE SUBJECT PAGE
1 Elastic modulus values for as-received stainless steel 69

wires from tensile tests of 016ss wires (Masson 1969)

wires from tensile tests of ss wires (Twelftree 19741

wires from tensile tests of O1O-O2Oss wires (Sokel
1 984)

wires from tensile and bend tests (Masson 1969)
5 Slopes of stess-strain curves for as-received stainless 76

steel wires in arch form
6 Diameter change of stainless steel wires after 78

continuous electropolishing
7 Diameter change of stainless steel wire after interrupted 80

electropolishing
8 Slopes of stress-stra¡n curves for stainless steel wires 81

tested in arch form after electropolishing

wires calculated from cantilever bend tests with the
support at 18.5mm

wires calculated from cantilever bend tests with the
support at 20mm
11 Elastic modulus values for electropolished stainless 85

steel wires calculated from cantilever bend tests with
the support at 18.5mm
xrr
steel wires calculated from cantilever bend tests with
the support at 2Omm

wires from cantilever tests with the support at l Omm

steel wires from cantilever tests with the support at
l Omm
15 Elastic modulus values for stainless steel wires 90

calculated from three-point bend tests
16 Elastic modulus values calculated for annealed stainless 92

steel wires from bend tests with a support at 2Omm
17 Elastic modulus values calculated for annealed and 93

electropolished stainless steel wires from bend tests
with a support at 2Omm
18 Mean elastic modulus values calculated for annealed 94

and electropolished specimens
19 Elastic modulus values for at and nt wires calculated 95

from cantilever bend tests with the support at 18.5mm
20 Elastic modulus values for at, nt and sent wires 95

calculated from bend tests with the support at 2Omm
21 Elastic modulus values for nt wires calculated from 96

bend tests with the support at 1Omm
22 Elastic modulus values for at, bt and nt wires calculated 96

from bend tests with the support at 2Omm

wires calculated from preliminary resonance tests

steel wires calculated from preliminary resonance tests
xrll
25 Elastic modulus values for nt and sent wires calculated 100
from resonance tests

wires calculated from resonance tests

steel wires calculated from resonance tests
28 Elastic modulus values for stainless steel wire 103

calculated from speed of sound tests using the peaks
as reference points

calculated fróm speed of sound tests using the troughs
as reference points

calculated from speed of sound tests using the point of
deviation from the horizontal as the reference point
31 Elastic modulus values for annealed stainless steel wire 106

calculated from speed of sound tests using the point of
deviation from the horizontal as the reference point
32 Elastic modulus values for new annealed stainless steel 107

wire specimen calculated from speed of sound tests
using the point of deviation from the horizontal as the
reference point
33 A comparison of elastic modulus values derived from 109

different techniques using load data.
34 The effect of w¡re diameter on elastic modulus values 111
35 The effect of data modification on the accuracy of the 114

straight line fit for a sspp ps 018 wire
36 Effect of data modification on elastic modulus values 114

for sspp ps 018
xrv
37 Average values of diagonal measurements of 115
microhardness test indentations
38 Vickers hardness numbers for sspp 018 116
xv
SUMMARY
ln orthodontics a light, continuous force is thought to be most desirable

as it results in maximum tooth movement and minimum patient
discomfort and damage to the supporting tissues. W¡th an increasingly
large selection of wires on the market, selecting the most appropriate
archwire for a given clinical situation is difficult. Choice may be based on
the feel of the wire, clinical impression or from standard mechanical test
data.
of these, the modulus of elasticity is of primary importance since ¡t

determines the stiffness of an archwire, which in turn governs the force
delivered by an appliance. The elastic modulus of a material has been
thought to be a constant value for a particular material. lt was therefore
surprising, to find such a wide range of elastic modulus values quoted in
the literature.
The principal aim of this study was to evaluate available experimental

techniques for the determination of Young's modulus. Tensile, bend,
resonance and speed of sound tests were performed. Elastic modulus
values were calculated from the data acquired, allowing the different
testing methods to be comPared.
Emphasis was placed on stainless steel wires as fundamental questions

remain, despite their long history of use. ln addition some of the newer
archwires such as nickel-titanium, alpha-titanium and beta-titanium were
tested.
It has been proposed that stiffness appeared to be affected when a wire

was electropolished. For this reason, wires were tested in the as-received
condition and after electropolishing, to ascertain whether elastic modulus
was affected.
Elastic modulus values both in theas-received state and after

electropolishing were lower than the textbook quoted values, and were
difficult to reproduce. Electropolishing did not appear to have a
consistent effect on elastic modulus.
xvl
It was proposed that these heavily drawn wires exhibited anisotropic
behaviour and that this may account for the low elastic modulus values.
Wires were annealed to reduce the anisotropic behaviour prior to testing.
This had a varied effect on elastic modulus with values for some wires
increasing while others decreased.
A wire's surface may be more heavily cold worked than the central
region. Wire specimens were embedded in Bakelite and polished to allow
the maximum wire diameter to be microhardness tested. Microhardness
tests did not detect differences between the surface layer and inner core.
lf a surface difference is present it is very small.
The surface appearance of wires was also assessed in the SEM, in the
as-received state and after electropolishing. As-received wires showed a
very elongated grain structure giving a fibrous appearance typical of a
heavily cold worked structure. The striations were removed with
electropolishing, leaving a smooth surface (apart from occasionäl deep
gouges).
To enable any textural differences between the wire surface and inner
core to be assessed, wires were deliberately maltreated to the point of
fracture. Fractured surfaces were then assessed in the SEM'
It is important to remember that it is the microstructure and in turn the

chemical composition and thermomechanical processing during
manufacture, that determines the mechanical properties of wires. For
these reasonS, assessment of the m¡crostructure in the TEM was
attempted. Specimen preparation proved difficult and no suitable foils
were produced. Ability to assess the microstructure would improve our
understanding of these materials and assist with the development of
more advanced materials.
xvlr
SIGNED STATEMENT
This report contains no material which has been accepted for the
award of any other degree or diploma in any university. To the best of
my knowledge and belief, it contains no material previously published or
written by another person, except where due reference is made in the
text of the rePort.
KATHERINE R. ALLEN
lgive consent to this copy of my thesis, when deposited in the University

Library, being available for photocopying and loan.
xvrll
ACKNOWLEDGEMENTS
I wish to express my appreciation to:

Dr. J.V. Bee, Senior Lecturer, Department of Mechanical
Engineering, University of Adelaide, for his guidance, patience and
humour
Dr. M. R. Sims, Professor of orthOdontics, Department of
Dentistry, University of Sydney, for his advice
Mr. l. Brown, Department of Mechanical Engineering, for his
technical assistance and suPPort
and lastly my husband Paul Heijkoop who made this Masters
degree possible. He worked full-time in private practice, cared for our
eleven month old daughter Bridget, ran the house and gave advice to a
computer illiterate wife even in the small hours of the morning.
xlx
CHAPTER 1
INTRODUCTION
orthodontic tooth movement is dependent on remodelling of the

periodontal ligament and alveolar bone. The type of force applied to the
tooth affects the physiological response. Light continuous forces avoid
occluding the blood vessels in the periodontal ligament allowing
continuous and rapid bone removal to occur on the pressure side and
new living tissue to be deposited on the tension side (Sandstedt 19O5'
Storey and Smith 1952, Reitan 1960). This allows maximum tooth
movement to occur with minimum damage to the tooth and investing
tissue (Begg and Kesling 1971).
The physical properties of archwires need to be known to allow a light

continuous force to be applied by the archwire to the teeth. The three
most important characteristics of archwires are thought to be springback,
stiffness and formability. For wires which obey Hooke's law for elastic
bodies, the springback and stiffness can be mathematically determined
(Wilcock 1988).
Orthodontic wires are often classified according to their mechanical

properties. However it is the microstructure which determines these
mechanical properties and this in turn is dependent on the chemical
composition and thermomechanical treatment during manufacture (Bee
1991; Burstone and Goldberg 1980; Twelftree, Cocks and Sims 19771'
Austenitic stainless steels and chromium-cobalt wires have dominated the

market for some years. These wires have similar values for strength,
stiffness and range (Proffit 1986). The similarity of the stress-strain
curves for stainless steel and chromium-cobalt allowed archwire selection
to be made from cross-section alone. An increase in stiffness was
achieved by an increase in cross-section of the wire and this resulted in
the phrase "variable cross-Section orthodontics" (Burstone 1981).
Despite the length of time that stainless steel wires have been used in
orthodontics, fundamental questions remain about their metallurgical and
mechanical properties (Khier, Brantley and Fournelle 1988).
There has been a large increase in the number of archwires available as a
result of the continuing search for improved materials (Andreason and
Morrow 1978). These include the nickel-titanium, beta-titanium and
alpha-titanium wires. The nickel-titanium wires have gained increasing
popularity since their introduction in 1971. The êlastic modulus is about
half that of stainless steel allowing a much lighter force to be produced
compared with an equivalent sized stainless steel wire. This ability to
vary the stiffness by varying the modulus of elasticity is referred to as
"variable modulus orthodontics" (Burstone 1981 )'
"Appropriate use of the available wire types may enhance patient

comfort, reduce chairside time and the duration of treatment" (Kapila and
Sachdeva 1989). Each archwire available has distinct advantages and
clinical limitations. The overall balance of properties and the specific
clinical application should determine which wire is selected (Burstone and
Goldberg 1980; Kusy 1981). There is no single ideal archwire for all
clinical situations.
To allow the most appropriate archwire to be selected and to enable

improved archwire materials to be developed it is important that there is
good understanding of wire structure and the mechanical properties
which it produces.
2
CHAPTER 2
LITERATURE REVIEW
MECHANICAL PRINCIPLES OF ORTHODONTIC TOOTH MOVEMENT
ldeally orthodontic forces should move teeth rapidly with the least
discomfort to patients and minimal damage to the teeth and the investing
tissues (Begg and Kesling 19711. lt is thought that excessive forces
compress the ligament and occlude the blood supply resulting in necrosis
(sandstedt 1904). No tooth movement will occur until the necrosed
tissue has been removed and new living tissue formed. Large orthodontic
forces cause the teeth to become loose and painful and the resultant
orthodontic tooth movement is intermittent and slow. When only light
forces are applied the blood vessels are not occluded and bone on the
pressure side is cont¡nuously and rapidly resorbed and new living tissue
forms on the negative pressure side. This results in continuor-¿s tooth
movement with minimum discomfort and tooth loosening (Sandstedt
1905, Storey and Smith 1952, Reitan 1960, Rygh et al 1986). From
these viewpoints it would appear that the ideal force system should
create a light continuous force (Proffit 1993).
THE IDEAL ARCHWIRE
The ideal properties of a wire will vary depending on its specific

application. Desirable archwire properties include the ability to resist
occlusal forces, provide a light continuous force over a large
displacement, show zero stress relaxation, minimum friction, good
formability, be easily joined, corrosion resistant, biocompatible and
environmentally stable (Kapila and Sachdeva 1989). Some of these
properties are metallurgically opposing and therefore a compromise needs
to be achieved (Wilcock 1988). lt should also be noted that the bracket
design and slot size create the environment in which the wire functions
and limits the properties and effects of the archwire (schudy 1990).
3
Of all the properties of archwires the following are generally considered
to be the three most imPortant.
1. Springback. This is the degree that a wire may be deflected without

permanent deformation. The terms springback, maximum elastic
deflection, flexibility and working range have been used interchangeably.
A high springback is desirable because it enables large activations to be
applied. The working time of the appliance is longer and therefore fewer
archwire changes and adiustments will be needed.
2. Stiffness or load deflection. This determines the magnitude of the

force delivered by the appliance. A wire with low stiffness allows a larger
diameter archwire to be used which will f¡ll the bracket slot to give good
tooth control and yet only produce light physiological forces. However
the archwire must have sufficient stiffness to resist undesirable elastic
deformation and plastic deformation during function'
3. Formability. Most appliances require modification at some stage. The

wire should be highly formable so that it may be easily bent and shaped
without fracture. "Ease of forming is best described in terms of practical
experience with representative formative pliers" (Burstone and Goldberg
1980, Hazel and West 1986, Kapila and Sachdeva 1989)'
For wires which obey Hooke's law for elastic bodies, the springback and
stiffness can be calculated from mathematical formulae. Springback is
proportional to the yield stress divided by the elastic modulus and
stiffness is proportional to the elastic modulus (Wilcock 1988). These
physical terms will be described when discussing the mechanical
properties of wires.
TERM¡NOLOGY USED WHEN D¡SCUSSING THE MECHANICAL PROPERTIES OF WIRES
Orthodontic wires can be classified according to their mechanical

properties. tt should be noted that it is the microstructure which
determines mechanical properties. The' microstructure in turn is
dependent on the chemical composition and thermomechanical treatment
during manufacture (Bee 1991; Burstone and Goldberg 1980; Twelftree,
Cocks and Sims 19771.
4
The following descriptions of terms have been used to describe the
physical properties of a wire (Phillips 1982).
Stress is the force per unit area in a body which resists an external force.
The two forces are equal in magnitude but have opposite directions. So
that stress and external force can be differentiated, external force is
referred to as load.
The type of stress is classified according to ¡ts direction. A tensile stress

is induced by a load which tends to stretch or elongate a body'
Compressive stress occurs when a a load tends to compress or shorten a
body. A stress which resists a twisting or sliding motion is a shear stress.
It should be noted that it is difficult to produce a pure stress. one stress
may be predominant but the other two may be present as well.
Whenever a stress is present there is also deformation or strain which

may be elastic or plastic. Elastic strain is reversible while plastic strain is
permanent. When a wire is stressed past its elastic limit the wire will not
return to its original length when the load is removed. The tft""t at
which this occurs is referred to as the elastic limit of the material' This is
also equal to the yield stress (yield strength).
lf an archwire is loaded in small increments a stress-strain curve can be

plotted. The curve usually starts as a straight line but curves after a
certain stress value is exceeded. The straight line portion of the graph
obeys Hooke's law which states that stress is directly proportional to
strain when the strain is within the elastic limit of the material. The point
at which the curve changes from a straight line is known as the
proportional limit.
The deviation from linearity is often difficult to determine accurately. For

this reason a proof stress value may be more appropriate. Th¡s is defined
as the stress required to produce a specific amount of permanent strain,
usually O.1o/o or O.2o/o. A line with slope equal to Young's modulus (the
elastic modulus) is constructed from the point representing the specified
plastic strain and the magnitude of stress at the point of intersection with
the stress-strain curve is the proof stress.
Elastic limit, proportional limit and proof stress have different definitions
but have nearly the same value and are therefore often used
5
interchangeably. These values represent the StreSS above which
permanent deformation occurs. For good formability a low value is
indicated but to resist deformation such as from masticatory forces a
high value is desirable.
The modulus of elast¡c¡ty (E) is calculated from the stress-strain ratio

determined from any point on the linear section of the graph.
F_ o
€
E elastic modulus
o STTESS
a strain
This value is indicative of the degree of stiffness or rigidity of the wire

and the magnitude of force delivered by the appliance. A smalJ elastic
modulus value enables a low constant force to be applied and allows a
given force to be applied with greater ease and accuracy (Kapila and
Sachdeva 1989). A wire with a higher modulus of elasticity is more
resistant to deformation (Drake et al 1982t..
Figure 1 (Proffit 1993), outlines the elastic-plastic behaviour of metal on

tensile loading. However the diagram is actually in error since the yield
point and yield strength refer to the O.1o/o ptoo't stress. The yield strength
actually lies at or very close to the proportional limit.
Archwire stiffness will vary depending on the chemical composition,

thermomechanical processing and the cross-section. To minimise
confusion between the wide variety of wires on the market a numbering
system indicating the relative stiffness has been suggested. Stainless
steel being the most commonly used alloy was arbitrarily set at 1.O
(Burstone 1981).
Appliance design will also affect the stiffness (Burstone 1982). In the
clinical situation the archwire sections between the brackets act as
beams. A beam is said to be stiff if it requires a large bending moment to
cause it to flex or distort (Hazel and West 1986). With the advent of
lingual orthodontics for the aesthetically conscious patient, the reduced
6
Ultimate Tensile Strength
Yield Stren gt h Failure Point
Yield Point
Proportional Limit
(/)
Ø
q)
t-
E
U) Stif fness ø E
Springiness a 1/E
-------l
l- O.l"l"
Strain
Figure 1: The elastic-plastic behaviour of metal on tensile testing.

Diagram from Proffit 1993.
7
interbracket distance results in an archwire being stiffer. To compensate
for this, wire dimension and composition may be altered (Moran 1987)'
ln orthodontics there are situations where a large strain or deformation

may be needed with a moderate or slight stress. Wires which fulfil these
criteria are said to be flexible. "The maximal flexibility is described as the
strain which occurs when the material is stressed to its proportional limit"
(Phillips 1982l..
According to standard engineering formulae the flexibility of a solid wire

is expected to decrease as its cross section increases. However, lngram,
Gipe and Smith's research suggested that this was not the case for
Nitinol (Unitek) and they stated that "range appeared to be independent
of wire size". As already stated flexibility appears to be affected by
many factors including wire Size, shape and modulus of elasticity
(lngram, Gipe and Smith 1986). Flexibility may be calculated from the
yield strength to modulus of elasticity ratio.
oy
I
E
E : maximum flexibility
oy : yield strength
E : elastic modulus
A high flexibility value allows large activation to be applied resulting in
the appliance having a longer working time with the advantage of fewer
adjustments being required (Burstone and Goldberg 198O; Drake et al
1982; Kapila and Sachdeva 1989).
"Resilience can be defined as the amount of energy absorbed by a

structure when it is stressed not to exceed its proportional l¡mit" (Phillips
1g82). lt is often measured ¡n terms of modulus of resilience and this is
equal to the area under the straight-line portion of the stress-strain curve.
This value represents the work available to move teeth (Kapila and
Sachdeva 1989). A high value should result in increased clinical
efficiency by necessitating fewer archwire changes (Drake et al 1982).
The term resilience is frequently associated with springiness.
8
ln an interview with Kesling, Wilcock stated that the most important
properties of orthodontic wires were flexibility and resiliency (Wilcock
1988). "Resilience and formability are defined as an atea under the
stress-strain curve and a distance along the x-axis respectively (Fig. 2)"
Proffit 1993.
lf the specimen continues to be loaded, a point will be reached where the

wire will fracture. "strength is referred to as the maximum StreSS
required to fracture a structure" (Phillips 1973). Various strengths are
defined according to the type of load. Tensile strength is defined as the
maximum stress it will support in tension, without failure.
Ductility is "the ability of a metal to undergo plastic deformation once the

proportional limit has been exceeded" (Greener, Harcourt and
Lautenschlager 1972). A ductile metal will elongate considerably
before
fracturing, while a brittle material would exhibit minimal elongation. A
degree of ductility is required to enable plastic deformation to occur
during archwire formation. "There are three common methods" for its
measurement: percentage elongation after fracture, reduction in the area
at the fractured ends and the number of turns around a mandrel to
fracture (cold bend test)" (Phillips 1982l..
Whenever one body slides over another body, the force that acts to
oppose the movement is called the frictional force. lt has been shown
that friction increases as wire size is increased (Riley, Garrett and Moon
1979). Different frictional resistances of wires may be explained by
differences in the surface smoothness of a wire. lncreased friction is
generally regarded as unfavourable because larger forces are required to
move the teeth (Garner, Allai and Moore 1986).
"surface hardness is the result of the interaction of numerous properties.

These properties may include material strength, proportional limit and
ductility. The relative hardness of a substance may be based upon the
resistance to indentation" (Phillips 1982).
Several investigators have assessed the mechanical properties of various

orthodontic wires (Twelftree, Cocks and Sims 19771 Drake et al 1982;
lngram, Gipe and Smith 1986; Moran 1987; Johnson and Lee 1989;
Kapilla and Sachdeva 1989). The findings from this research work will
be discussed in a later sect¡on on the chemical composition and
9
-Yield Strength
Prop ortional Limit-
U)
U)
q)
L
U)
Formability
-
Strain
E Resilience and formability a¡e defined as'an area

under the stress-strain curve and a distance along the X-axis re-
spectively, as shown here.
Figure 2: A stress-strain curve illustrating the terms resilience and

formability. Diagram from Proffit 1993.
10
mechanical properties of specific wires. There is however very little in the
literature about the mechanical properties of orthodontic wire in relation
to their chemical composition and microstructure despite the close
relationship.
PRODUCTION OF WIRES
Wires are manufactured from cast metals which are drawn through a
ser¡es of dies of decreasing diameter. This process roughens the outside
surface and results in plastic deformation of the metal which is known as
cold working. The crystallization process of molten metals and the
subsequent effect of cold working will be discussed'
THE CRYSTALLIZATION OF MOLTEN METALS
pure metals are rarely used in dentistry because they tend to be tbo soft.
For this reason two or more metals are often mixed to form alloys, with
superior properties. ln the solid state the atoms composing the alloy are
regularly spaced into a configuration known as a space lattice or crystal.
1. The atomic structures are classified according to the fourteen Bravais

lattices in seven crystal systems.
2. There are several possible atomic arrangements. However, metals and

alloys fall into only three of these: body-centred cubic, face-centred cubic
and close-packed hexagonal crystal structures.
3. The form and dimensions of the space lattice are important in
determining the mechanical properties of the metal.
Defects occur in the crystalline structure. The most important defects are
dislocations and their presence affects the ductility of the metal as will be
discussed later.
As a molten metal solidifies a process known as crystallisation occurs.

Crystal growth is initiated at specific sites called nuclei. The number of
nuclei present is determined to some extent by the amount of impurities
present, seeding and the rate of cooling. Nuclei number will therefore
11
affect the number of crystals and in turn crystal size. In general the
smaller the grain size the better its mechanical properties.
Often, "the crystals grow as dendrites which are described as three-

dimensional branched network structures emanating from a central
nucleus. Crystal growth continues until all the material has solidified and
all the dendritic crystals are in contact" (McCabe 199O). Each crystal is
known as a grain. The area between the two grains in contact ¡s called
the grain boundary. The grain boundaries are a very small region of
transition, approximately two to three atom diameters in extent, and the
structure is usually highly disordered. lf the metal is highly polished and
then etched appropriately the grain structure is visible microscopically.
After crystallization the grains have approximately the same dimensions

in each direction as measured from a central nucleus. They do not form
to any geometric shape and are said to be equiaxed. A change from this
equiaxed structure to a more elongated, fibrous structure can result in
important changes in mechanical properties (McCabe 19901. Such
changes may be brought about by cold working the alloy below the
recrystallization temperature which produces work/strain hardening.
Examples of this occur in wire fabrication and the bending of wires or
clasps.
THE EFFECTS OF COLD WORKING AND ANNEALING ON THE CRYSTAL STRUCTURE
Wires are said to have a wrought structure as a result of the cold

working procedure. This results in wires having different properties to
those of cast metals. They become harder and stronger with a higher
elastic limit value and have decreased ductility.
The drawing of a wire from castings is an example of plastic deformation.

As plastic deformation begins, the dislocations present within the grains
move along a slip plane until another defect (dislocation or foreign atom)
is reached. The resultant interactions reduce the mobility of the
dislocations and higher stresses are needed to continue deformation. This
explains the phenomenon of work hardening. lf dislocations reach a grain
boundary, they "pile up". This is due to the fact that the slip planes in
t2
adjacent grains are not aligned due to the¡r different lattice orientations.
Again higher stresses must be provided to continue plastic deformation.
The effect of crystal size on the mechanical properties is now

comprehendable. As crystal size decreases the concentrat¡on of grain
boundaries increases and consequently the resistance to dislocation
movement increases. Thus materials with a finer grain structure are
harder and have a higher elastic limit (McCabe 1990).
Cold worked metals can be converted back to an equiaxed structure by

heating the structure above the recrystallization temperature. This is
known as a softening heat treatment. The metal reverts to its
undeformed condition, which is softer with a lower elastic limit and has a
higher ductility. ln some cases these changes in mechanical properties
are undesirable and so heat treatment should be avoided. Heat treatment
below the recrystallization temperature is referred to as stress relief
annealing. lt allows the internal stresses to be relieved without affecting
the mechanical properties, since the deformed structure is not- altered
significantly.
HISTORY OF ORTHODONTIC WIRES
precious metal alloys were predominantly used for orthodontic purposes

prior to the 1930's. At this time gold alloys were the only material
available that would tolerate the intraoral conditions. lts physical
properties were only marginal and it was obsolete even before the price
increase in the 1970's (Proffit 1986).
Historically stainless steel (Wilcock) and chromium-cobalt (Elgiloy) have

been the dominant orthodontic wires because of their desirable physical
characteristics. Austenitic stainless steel was introduced in 1929.
Stainless steel w¡res possess excellent formability, have good corrosion
resistance, high values of stiffness and resilience, are environmentally
stable and are of moderate cost (Drake 1982; Burstone and Goldberg
1980; Hazel, Sokel and West 1984). Chromium-cobalt (Elgiloy) wires
have similar strength, stiffness and range properties (Proffit 1 986).
Despite the long history of use of stainless steel wires in orthodontics
T3
fundamental questions still remain regarding their metallurgical and
mechanical properties (Khier, Brantley and Fournelle 1988).
As a result of a continuing search for improved performance there has

been a large increase in the number of archwire materials available
(Andreason and Morrow 1978). These include the nickel-titanium, beta
titanium and (near) alpha titanium wires. Since the introduction of nickel-
titanium wires in 1g71 they have gained increasing popularity. Some
possess unique properties of superelasticity and shape memory which
will be described later (Hurst et al 1990). Beta titanium has a unique
balance of low stiffness, high springback, formability and weldability
(Burstone and Goldberg 1980). A near alpha titanium is also on the
market. properties of titanium wires will be discussed in detail in sections
5,6 and 7.
The increase in new alloys and wire fabrication techniques has made
archwire selection much more difficult. ln the past owing to the similarity
of stainless steel and chromium-cobalt Stress-strain curves, archwire
selection was made almost exclusively on cross-section alone. lncreases
in wire stiffness were achieved by progressively increasing the cross-
section of the wire. This resulted in the phrase "variable cross sect¡on
orthodontics". Now the archwire stiffness can be varied by varying the
modulus of elasticity and this is referred to as "variable modulus
orthodontics" (Burstone 1981; Combe et al 1985). ln the latter technique
the cross sectional shape, alloy content and heat treatment of the
archwire needs to be assessed (Johnson and Lee 1989).
As a result of these advances "the play between the wire and the
attachment is not dictated by the stiffness required but is under full
control of the operator" (Burstone 1981). The low elastic modulus newer
alloy archwires allow light rectangular wires to be placed even during
early stages of treatment. This results in better tooth control and
decreases the number of archwire changes.
Each archwire on the market has distinct advantages and clinical

limitations. The overall balance of properties and the specific clinical
application should determine which wire is selected (Burstone and
Goldberg 1980; Kusy 1981). "Appropriate use of all the available wire
types may enhance patient comfort and reduce chairside time and the
duration of treatment (Kapila and Sachdeva 1989)."
t4
THE CHEMICAL COMPOSITION AND MECHANICAL PROPERTIES OF SPECIFIC W¡RES
1. cor-o
pure gold was too soft for most dental purposes so alloys similar to that
of Type lV castings were used for wires, A typical alloy contained 60%"
gold, 15% silver, 15o/o copper and about 1Oo/o platinum or palladium' The
platinum or palladium content raises the melting point and
recrystallization temperature of the wires making them more suitable for

soldering. Gold alloy wires have a lower modulus of elasticity than
stainless steel and therefore apply lower forces (McCabe 1990)' Gold
tolerates the intraoral conditions well. ln general though their physical
properties are only marginal and with the advent of other alloys they
have become obsolescent. Only the Crozat appliance is still occasionally
made from gold (Proffit 1993)'
2. srnlruless steel
There are three major forms of stainless steel - ferritic, martensitic and
austenitic and there is wide variation in their respective compositions and
properties (Phillips 19821. There are also grades which contain two co-
existing phases (ferrite-martensite, or ferrite-austenite) which are termed
dual-phase or duplex stainless steels.
The typical stainless steel orthodontic wire is referred to as 18/8

austenitic stainless steel. The term 18/8 refers to the composition - 18%
chromium and 8o/o nickel. The high chromium content results in good
resistance to corrosion (Proffit 1993). The addition of nickel makes the
austenite-martensite transformation sluggish and lowers the martensitic
transformation Start temperature below room temperature So that no
hardening heat treatments are possible (Greener, Harcourt and
Lautenschlager 1972). lt should be noted that some individuals are
sensitive to nickel and chromium (Kapila and Sachdeva 1989)'
lnvestigations have shown that nominally austenitic sta¡nless steel wires

may not be entirely austenitic. Whether a stainless steel wire is a single
phase austenitic (face centred cubic) structure or a duplex austenitic and
martensitic (body centred cubic) structure is strongly dependent on the
15
alloy composition and wire manufacturing processes (Khier, Brantley and
Fournelle 1988). The austenitic stainless steel structure is metastable and
can decompose to the martensitic phase under certain conditions such as
proprietary cold working and intermediate heat treatment associated with
wire manufacture (Khier, Brantley and Fournelle 1988; Phillips 1982)' ln
fact, austenite can be metastable with respect to deformation and it is
this which produces martensite in different orthodontic wires (Singh
1991).
By varying the amount of cold working and annealing during manufacture

the properties of stainless steel wire can be altered. lt may be softened
by annealing and hardened by cold working (Phillips 1982l'. A range of
stainless steel archwire materials in partially annealed states are
available. They are categorised according to their ductility for example
soft, half-hard and hard. Fully annealed stainless steel is quite soft and
highly formable. Ligature wire which is used to tie an archwire into the
bracket is an example of "dead soft" wire. As the amount of annealing is
increased formability is progressively enhanced at the expense of
decreased yield strength. The steel wires with the highest yield strength
are brittle and break when bent sharPly (Proffit 1993).
ln Adelaide the commonly used stainless steel wires are produced by

A.J. Wilcock at Whittlesea, Victoria, Australia. They ate available in
several different grades - Special, Special plus, Premium, Premium plus
and Supreme. These wires are made from an 18/8 austenitic stainless
steel. Their properties are varied by the thermomechanical treatment
during manufacture.
Khier, Brantley and Fournelle (1988) investigated the structure and

mechanical properties of as-received and heat-treated stainless steel
orthodontic wires. Two different proprietary brands of stainless steel
were selected (Tru-Chrome and Permachrome). Selected specimens were
heat treated in air at 7OO", 900" and 1 100' for a period of ten minutes
followed by air cooling to room temperature. X-ray diffraction patterns
showed that both brands of the as-received wires consisted of a two-
phase structure. This two-phase structure was converted to austenite
with heat treatment for Tru-Chrome but persisted in Permachrome. The
modulus of elasticity in tension was significantly lower in the latter wires
when compared with the annealed austenitic sta¡nless steels. This can be
16
attr¡buted to phase transformation and the development of considerable
residual Stress in the heavily cold-worked structure. This is in agreement
with findings from Goldberg, Vanderby and Burstone 1977, Asgharnia
and Brantley 1986, PhilliPs 1982.
When a stainless steel wire is ptastically deformed dislocations pile up

and accumulate at various microstructural obstacles. When the load is
removed the dislocation lines will not move appreciably because the
structure is stable (Orowan 1959). lf the direction of loading is reversed
some dislocation lines can move a considerable distance at a low shear
stress because the barriers behind are not as strong and closely spaced.
As a consequence initial yield will occur with a smaller load. lf a load is
applied in the same direction as the original load there is increased
resistance to plastic deformation. This phenomena is referred to as the
Bauschinger effect.
The reduction in proof Stress and Young's modulus and increased

elongation noted when the direction of deformation is reversed mpy have
the following clinical effects. These include:
i. The amount of stored energy is reduced and therefore there is less

energy to move teeth.
ii. The maximum force that can be applied before the wire is plastically
deformed is reduced.
iii. The lower stress:strain ratio reduces the rate of change of force
magnitude, that is the force level remains more constant over a given
distance.
iv. The wire has increased ductility and is therefore more formable'
To gain the best elastic properties from the wire it should be activated in
the same direction as the original curvature in the wire although this is
not always possible (Gullotta, West and Hazel 1987)'
Stress relief anneals may be carried out after bending to reduce internal
stresses and enhance the elastic properties. Stress relief anneals involve
heating the wire to 450"C for about ten minutes. This can only be done
to stabilised stainless steel wires where small quantities of titanium (or
niobium) have been added. Titanium is added to the alloy to prec¡pitate
t7
as a carbide in preference to chromium, so that chromium is retained in
solid solution (Phillips 1982).
A study assessing the effects of stress relief generally showed an

increased resilience which was reflected in increased stiffness and range.
This results in a higher initial force magnitude for a given activation level
(Lane and Nikolai 198O). The use of smaller diameter wires enables lower
forces to be applied but this results in poorer bracket fit and
consequently less control of tooth movements. The force produced is
higher than nickel-titanium and beta-titanium but dissipates faster than
either (Kapila and Sachdeva 1989). lt exhibits low friction - less than
Nitinol and beta-titanium (Garner, Allai and Moore 1986).
In Stage I of the Begg technique, traditionally Wilcock O.016" stainless

steel archwire is used. Recently, as a result of more flexible archwires
becoming available, archwires such as O.O16" Japanese NiT¡, O.O16"
Nitinol and O.O1O" Supreme Wilcock wire are being used in Stage l. This
has been shown to considerably decrease the time for initial alignment
(Lew 1988).
Supreme grade stainless steel wire (ultra high tensile) has a flexibility
similar to that of beta-titanium but has nearly three times the resilience. lt
is not as flexible as nickel-titanium but the nickel-titanium flexibility is
probably not fully utilised orthodontically. lt has good formability and the
resiliency of nickel-titanium can be achieved by engineering the yield
stress and diameter (Wilcock 1988).
Stainless steel wires can be joined by soldering or welding but the

technique is demanding (Kapila and Sachdeva 1989). Care must be taken
to avoid overheating as it may cause recrystallization and reduce the
springiness of the wire. ln unstabilised wires it may cause chromium to
react with carbon to form carbides. This phenomena is referred to as
weld decay and results in a loss of corrosion resistance at the soldered
joint and the introduction of a degree of brittleness. To decrease the
potential for this silver solders are used because they contain silver and
copper with small quantities of other elements which lower the fusion
temperature.
Welding isaccomplished by passing an electric current through two

pieces of wire held together between two electrodes. Sufficient current
18
needs to be used to melt the wires at the point of contact (McCabe
1990).
The corrosion of stainless steel orthodontic wires has been investigated

in the "as received" condition, after forming and after heat treatment in a
hot air oven at 4OO"C for ten minutes. Corrosion rate varied depending
on the strain history. Heat treatment permitted a degree of stress relief
without carbide formation and reduced the corrosion rate of the formed
wires in relation to the "as-received" wires (Toms 1988).
3. MULTIBRAIoED srAlNLEss srEEL
These wires consist of specified numbers of thin stainless steel wire

sections coiled around each other to provide a round or rectangular cross
section. These wires have a high springback, low stiffness, high stored
energy but poor formability. They are biocompatible and environmentally
stable and may be joined by soldering or welding. Little is knowñ about
their friction (Kusy and Dilley 1984).
ldeally the initial archwire should have a great range to enable malaligned
teeth to be accommodated, have low stiffness so that the forces applied
are gentle and have high strength so that it is not easily deformed with
masticatory forces. Historically, looped archwires were formed to
increase the flexibility and reduce the force (Begg and Kesling 197U.
Disadvantages included the increased time required to bend the archwire,
they were mechanically disadvantageous (resulting in bite closure and
excessive incisor proclination) and were difficult to keep clean. An
alternative was the use of multistranded stainless steel wires which have
the flexibility of small diameter wires but the strength of many strands.
Titanium alloys are a third alternative but are more expensive (Kusy and
Stevens 1987).
It has been noted that multistranded stainless steel wires have a similar
springback to nitinol but a larger springback when compared to solid
stainless steel wires or beta-titanium wires. The springback of
multistranded stainless steel and titanium wires is relatively independent
of wire size and therefore does not obey the engineering principal that
19
spr¡ngback decreases with increasing thickness (lngram, Gipe and Smith
1986).
Four triple-stranded stainless steel wire products were tested in one

investigation: Twist (American Orthodontics), Wildcat (GAC
tnternational), spiral (ormco) and Hi-T TwistFlex (unitek). The four
different wire diameters tested were 0.0150", 0.0175", 0.0195" and
O.0215". Force storage and delivery properties were determined and
compared with reported values for titanium alloy archwires' Except for
Spiral (Ormco) wire all of the O.O15" multistranded wires demonstrated a
greater working range than Nitinol or TMA. All of the 0.O15" triple
stranded wires were unmatched for application of light forces over a long
distance. Excluding Hi-T TwistFlex (Unitek) the titanium wires are
stronger. The elastic property ratios of triple stranded wires varied widely
and compared favourably with some of the titanium wires offering a
cheaper viable alternative (Kusy and Stevens 1987).
A study comparing 43 patients with either a 0.O15" multistranded

stainless steel or a O.O14" superelastic nickel-titanium initial aligning
archwire showed no statistically significant difference in the amount of
incisor alignment over the mean time interval of five weeks. These results
support the conclusions Kapilla and Sachdeva (1 989) that a
of
multistranded stainless steel wire may be a less expensive alternative to
the nickel-titanium wires for initial alignment and levelling (Jones,
Staniford and Chan 199O).
4. cxRorvllurv¡-cosRLr
Chromium-cobalt alloys were originally developed for use as watch

springs but their properties are also excellent for orthodontic purposes
(Phillips 1982). These alloys contain approximatelY 7Oo/" cobalt and 30%
chromium however some cobalt can be replaced by nickel. The carbon
content is never above O.15o/". The distribution and size of the carbide
precipitates within the matrix may determine the hardness and brittleness
of the material. Alloys with lower carbon content are softer (Greener,
Harcourt and Lautenschlager 1972l..
20
These wires are unique in that they are supplied in a softened state' This
is achieved by heat treating the wires between 1 1OO' and 12OO'C and
then quenching them. As a result the wires have excellent ductility and
are easily formed.
They can be hardened after bending by "age-hardening" at temperatures

between 2600 and 6SOoC (Phillips 1982). This results in the formation of
precipitates which produce the required springback and resilience
properties (Burstone and Goldberg 198O).
Hardness, yield and tensile strength and the modulus of elasticity are
similar to those of austenitic stainless steel (Hazel, Sokel and West 19841
proffit 1gg3). The high modulus of elasticity enables them to deliver
twice the force of beta-titanium and four times that of Nitinol.
They are difficult to join by soldering and care must be taken to avoid
overheating as this can result in annealing with loss of yield strength
(McCabe lggo). Tarnish and corrosion resistance is excellent -(Phillips
1g82). Their advantages over stainless steel include greater resistance to
fatigue and distortion and longer function as a resilient spring. Frictional
forces are comparable to stainless steel (Kapila and Sachdeva 1989)'
Elgiloy (Rocky Mountain Co.) is an example of a chromium-cobalt

orthodontic wire. lts nominal composition is cobalt 4Oo/o, chromium 2Oo/o,
iron 160/o, nickel 15o/o, molybdenum 7o/o and manganese 2% (Elgiloy,
Rocky Mountain Orthodontics Brochure, Denver, Colorado, 1977!.. These
wires are available in several cross-sectional shapes and dimensions.
There are four grades of Elgiloy - red, green, yellow and blue. Each grade
represents different degrees of cold work.
5. ru¡cret-rltRrulun¡
Nickel-titanium was the first of the titanium alloy wires and was marketed
as "Nitinolrnr" (Unitek Corp.) in the late 1970's. "NitinolrM" was initially
developed for the space programme but has proved very useful in clinical
orthodontics. lts name was derived from its metal composition Ni - nickel
and Ti - titanium and the place of origin NoL - Naval ordinance
Laboratory (Proffit 1993).
2l
These alloys contain almost equal amounts of nickel and titanium with
small quant¡ties of other metals. At high temperatures (880'C), the
crystal structure is austenitic (body centred cubic) beta phase, and at
lower temperatures a martensitic transformation to a (close packed
hexagonal) alpha phase can be produced (Buehler and Cross 1969). The
change in crystal structure alters the mechanical properties. The material
is ductile and may be plastically deformed in the alpha phase (Andreason,
Wass and Chan 1985). ln the beta phase it is difficult to induce
deformation (Miura et al 1986).
The temperature range over which the change in crystalline structure

occurS is referred to as the transition temperature range (TTR)' The
specific TTR is dependent on the chemical composition of the alloy and
its processing history. TTR can be changed by altering the proportion of
nickel to titanium or by substituting cobalt for nickel (Buehler and Cross
1 969).
Nickel-titanium has outstanding elasticity and can almost be bent back on

itself without being plastically deformed. This characteristic offers real
advantages in enabling clinicians to engage an archwire into malposed
teeth (Andreason and Morrow 1978). Studies have shown that nickel-
titanium wires have superior springback characteristics to stainless steel
and beta-titanium (Hudgins, Bagby and Erikson 199O), "Nitinol exhibited
5o/o-55o/o greater working range than single or multistranded stainless
steel wires, depending on the deflection, interbracket distance and wire
(Andreason and Barret 1970)."
Hudgins, Bagby and Erikson's study aimed to quantify the amount of

permanent deformation after deflection of nickel-titanium wire over
different time intervals of one, fourteen and twenty eight days. The
majority of the wires exhibited more deformation at the later time
intervals, especially twenty eight days, suggesting that deformation may
be a function of time (Hudgins, Bagby and Erikson 1990)'
The elastic modulus of nickel-titanium is about half that of stainless steel.

Clinically this means that nickel-titanium wires will produce a lighter force
than an equivalent sized stainless steel wire, The low elastic modulus but
moderately high strength results in nickel-titanium having high resilience.
The stored energy of nitinol is much larger than stainless steel
(Andreason and Morrow 1978).
22
The outstanding elasticity and low elastic modulus allow large dimension
wires (which nearly fill the bracket), to be used without the risk of
deforming during bracket engagement or exerting too great a force at an
early treatment stage. This decreases the number of archwire changes
and keeps chairside time to a minimum. Treatment time to derotate and
level teeth is shortened and patient discomfort lessened (Andreason and
Morrow 19781 Lew 1988; Chen 1990). These properties make nickel-
titanium favourable as a starting and intermediate wire (Kusy 1981)'
The nickel-titanium wires initially available for orthodontic use were in the
stabilised martensitic form at room temperature. These wires possess
excellent springback, but shape memory and superelasticity are poor due
to manufacture by a work hardening process (Miura et al 1986).
"Titanal" (Lancer Pacific), is a newer martensitic nickel-titanium alloy. lt
has very similar strength and spring characteristics to the early nickel-
titanium wires but has the advantage of being highly formable (Proffit
1993).
ln the late 1980's nickel-titanium wires with a primarily austenitic grain

structure became available. Like the original nickel-titanium wires they
exhibited low stiffness and high springback. ln addition they
demonstrated superelasticity and/or shape memory (chen 199o)' A three
point bending test was designed to differentiate between wires with and
without superelasticitY.
Shape memory is described as the ability of a wire to return to its

previously manufactured shape when it is heated through its transitional
temperature range (Andreason, Wass and Chan 1985). This change can
be metallurgically explained by a stress-induced martensitic change. This
is a partial transition in structure from an austenitic to martensitic form.
The shape memory effect is brought about by a high temperature heat

treatment which alters the crystal structure of the alloy. The nitinol wire
must be held in the desired shape at a temperature between 23O"-260"C
for 1O minutes (Andreason 1980). This allows the archw¡re to be formed
in the laboratory ahead of time, saving chairside time and making it more
accurate (Miura, Mogi and Okamoto 1990). Providing the wire is
deformed at temperatures beneath the TTR, the original shape can be
recovered by heating it through its unique temperature trans¡tion range.
For clinical purposes the TTR should be sufficiently high to avoid a return
23
to the original shape at room or oral temperature (Andreason and Morrow
1978; Hurst et al 199O).
ln contrast to other metals where deformation is predominantly induced

by lattice slip, with nickel-titanium it is obtained by martensitic
transformation. To obtain maximum shape recovery the amount of plastic
deformation should be limited to 7-8o/o of the original length. This enables
the process to be repeated many times. The change from the distorted to
the original form involves a transformation of nickel-titanium from the
martensitic to the austen¡tic phase (Andreasen and Brady 1972l..
The shape memory effect has been quantitatively analysed. Maximum

shape memory was determined by calculation of the amount of recovery
that occurred. The mean percentage recovery ranged from 89 to 94o/" for
N¡-Ti, Nitinol, Orthonol, Titanol, Sentinol Light and Sentinol Medium.
Sentinol Heavy wire showed a mean recovery of approximately 41o/o
(Hurst et al 199O). This is in contrast to earlier findings by Buehler and
Cross, and Andreason and Brady, where it was suggested that shape
recovery was usually 1OOY" (Buehler and Cross 1969; Andreason and
Brady 1972l,.
This temperature transition or shape memory potential of nickel-titanium

wires is not fully understood or utilised in orthodontics at this stage but it
is an area that warrants further investigation (Andreason, Wass and Chan
1985; McCabe 199O).
"Nitinol" has a lower elastic modulus and can be deformed almost five
times more than stainless steel but the stress-strain curve is similar to
that of StainleSS steel and chromium-cobalt. "Japanese NiTi" wire
produces a totally different cUrve. When Stretched Up to 2o/o, sLress and
strain are almost proportional, but additional strain is accumulated at
constant stress. lnit¡ally the curve appears to be similar to that shown
when plastic deformation occurs. However on unloading the specimen
the stress and strain values are almost proportional followed by a period
where the stress remained constant despite a decrease in the strain. This
phenomena is known as superelasticity (Miura et al 1986). Clinically this
means that the archwire can be deflected a small or large distance and
yet exert the same force (Chen 1990).
24
Superelastic nickel-titanium wires are unusual in that the unloading curve
differs from its loading curve (F¡g. 31. This may be interpreted as the
force needed to activate the wire is not the same as the force that the
wire delivers (Proffit 1993).
Several factors affect the force level of superelastic nickel-titanium wires.

These include the heat treatment conditions, the processing method, and
most importantly, the constituent ratio of the alloy. Heat treatment and
reduction of the nickel content reduce the forces produced by the wire.
Wire size and the amount of deflection do not affect the superelastic
force (Miura, Mogi and Okamoto 1990).
The bending properties of superelastic (Nitinol SE, Sentinol and NiTi) and
non-superelastic (Nitinol, Titanol and Orthonol) wires have been assessed
in the as-receíved condition and after heat treatment at 5OO and 6OO"C.
Permanent deformation for the superelastic wires after unloading from
80'activation were 1O-15" and about twice this for the non-superelastic
wires. A superelastic region of nearly constant bending moment was
noted in the round wires but was less evident in rectangular wires. The
slopes of the non-linear bending plots were considerably less for the
superelastic wires. Heat treatment only produced small changes in the
bending plots for the non-superelastic but considerable response for the
superelastic wires. The maximum amount of activation and deactivation
decreased with heat treatment. Heat treatment at 6OOoC for 1O minutes
resulted in complete loss of the superelastic properties (Khier, Brantley
and Fournelle 1991).
Nickel-titanium wires have limited ductility and are therefore not easy to
bend without fracturing (McCabe 1990). They cannot be bent with sharp
cornered instruments and bending adversely affects springback. ln
addition they cannot be soldered or successfully welded without being
annealed (Andreason and Morrow 1978). These properties limit the
usefulness of the wire (Kusy 1981).
These wires have more surface roughness than stainless steel which may
be significant to corrosion and friction (Garner, Allai and Moore 1986).
The numerous surface irregularities could predispose the wire to
corrosive attack in the mouth (Harris, Newman and Nicholson 1988). The
corrosion resistance findings are inconsistent. According to some studies,
their frictional resistance is higher than stainless steel but |ower than
25
E 2000
E
Stainless Steel
Ë 500
(') 1
c(l, Nitinol
1
E
o
NiT¡
=o) 50
c
tc
(l)
@
0 20 40 60 80
Deflection (degrees)
Figure 3: Stress-stra¡n curves for a range of orthodontic wires: stainless

steel, Nitinol (martensit¡c NiTi) and NiTi (austen¡t¡c N¡T¡). (Proffit 1993)
26
beta-titanium (Garner, Allai and Moore 1986; Kapila and Sachdeva
1989).
Nickel-titanium wires are costly and as a result some clinicians are

reusing them. This necessitates the wires being sterilised and being
subjected to repeated mechanical stress. No appreciable loss in physical
properties has been noted after repeated cycles of sterilising/disinfection
(Buckthal and Kusy 1988; Mayhew and Kusy 1988). The loss of elastic
deformability may be the limiting factor in the number of times that the
wire may be reused (Harris, Newman and Nicholson 1988).
6. BETA-TtrANtuM
Unalloyed titanium has two allotropic forms - a low temperature

hexagonal close packed, alpha, and a high temperature (above 88O"C)
body centred cubic, beta, structure. Additions of certain elements such
aS iron, molybdenum, vanadium and chromium stabilise the beta phase
and when in sufficient amounts allows the beta to be present at room
temperature. Other additions such as tin and zirconium suppress the
formation of an embrittling omega phase in beta (Burstone and Goldberg
1980; Wilson and Goldberg 1987).
Beta titanium dental alloys consist of titanium with additions of

molybdenum which stabilises the body centred cubic phase even when
cooled to room temperature' Beta-titanium is marketed as TMA (titanium
molybdenum alloy) (Ormco/Sybron) (Wilcock 1 988). Unlike nickel-
titanium, beta-titanium was developed primarily for orthodontic use
(Proffit 1993).
The extent of deformation and the number of passes used to draw the
wire affects the properties. With increased deformation the tensile
strength increases but the elastic modulus remains unchanged. Lower
reductions per pass improve elongation and yield strength (Shastry,
Elinson and Goldberg 1982).
These wires offer a desirable combination of strength (greater than

stainless steel) and springiness and are reasonably ductile and therefore
formable (Kusy 1981). Formability is similar to stainless steel although it
cannot be bent over as sharp a radius (Wilson and Goldberg 1987).
27
The modulus of elasticity value is less than half that of stainless steel
but is twice that of nitinol. This enables these wires to impart a low force
per unit of activation. Springback is superior to stainless steel permitting
beta-titanium wires to be deflected over long distances without
permanent deformation. This allows the wire to have a greater range of
action (Burstone and Goldberg 1980). The flexibility of beta titanium
wires is much less than nickel-titanium wires (Kusy 1981).
Beta-titanium can be joined by welding alone without reinforcement

solder. The welding must be done within a narrow optimal voltage setting
to avoid affecting the wire properties. Overheating could result in the
wire becoming brittle (McCabe 199O). Corrosion resistance is similar to
stainless steel and chromium-cobalt wires. Friction levels are higher than
stainless steel, chromium-cobalt and Nitinol (Garner, Allai and Moore
1 986).
The properties of beta-titanium are intermediate between stainless steel

and martensitic nickel-titanium (Proffit 1993). These characteristics make
it suitable for an intermediate and finishing wire (Kusy 1981)'
Some of the titanium alloys are amenable to thermomechanical

processing to produce alpha plus beta structures. The presence of the
alpha phase improves the tensile ductility in comparison to when only the
beta phase is present but the fracture toughness is inferior. The proof
stresses for both structures are similar (Cornet, Mille and Muster 1982l-.
7. ALPHA-TITANIUM
This new alloy wire is titanium-based and contains aluminium and

vanadium. ln equilibrium at room temperature it has a close-packed
hexagonal crystal structure termed alpha-titanium, lt is slow to form from
the high temperature body-centred cubic phase. There is some retention
of the beta phase at room temperature and due to this it is usually
referred to as a near-alpha alloy. The fewer slip planes in the hexagonal
lattice render it less ductile (Wilcock 1988).
Alpha-titanium wire has similar tensile properties (proof stress and elastic
modulus) to gold. The elastic deflection or springback is similar to high
tensile stainless steel. This allows the wire to be activated more than
28
stainless steel wires, without problems of permanent deformation or an
excessive force being produced. lts properties lie between the low moduli
nickel-titanium wires and beta titanium and the high moduli stainless steel
and chromium-cobalt.
The formability of the wire is affected by the surface finish. Poor surface
finish or surface cracks in the wire decrease the formability. Alpha-
titanium wire is corrosion resistant and much harder than beta-titanium
alloy wires (Hazel, Sokel and West 1984).
SUMMARY
As stated in the literature there is a lack of knowledge about the

metallurgícal and mechanical properties of stainless steel orthodontic
wires despite their long history of use. The use of wires is determined by
their mechanical properties. Particularly important is the elastic modulus
as this determines the stiffness and resilience of a wire.
The elastic modulus values measured for orthodontic wires have

generally been lower than the standard values quoted in metallurgical
references for bulk materials. ln addition there is large disagreement in
absolute values between authors.
A study is therefore warranted on the determination, accuracy and

reproducibility of Young's modulus in commonly used mechanical tests.
Results from these studies should then be compared with values obtained
using more sophisticated mechanical and physical test methods.
29
CHAPTER 3
AIMS OF THE INVESTIGATION
In the literature survey, it was noted that there was disagreement in
elastic modulus values for orthodontic wires. Most of the elastic modulus
values were determined from tensile and bend tests and these values did
not agree with bulk metallurgical values. For these reasons it was
decided to look at different methods of measuring the elastic modulus
and to check the reproducibility of the results. Current orthodontic wires,
including stainless steel, nickel-titanium, alpha-titanium and beta-titanium,
were assessed.
Knowledge of mechanical properties of as-received wires would

complement observations by Singh 1991 on microstructure. Some wires
were also assessed after being electropolished. This was prompted by a
comment made by an experienced clinician that wire stiffness appeared
to change after electropolishing.
1. Mechanical property data was acquired using
a. Simple tensile test using a 1O tonne Instron universal testing machine
b. Bend tests
Three different bend tests were used in order to compare the load-
deflection behaviour and the elastic modulus values obtained. The tests
are as l¡sted below:
(1) wires in arch form
(2) a conventional cantilever bend geometry
(3) three point bend sYstem.
The properties of the wires were determined in both the as-received

condition and after manipulative treatments such as electropolishing
and/or heat treatment.
c. Resonance tests.
d. Measurement of speed of sound along a length of wire.
30
e. Microhardness / Macrohardness tests
2. Etfect of electropolishing on surface structure

The surface structure of wires in the as-received state and after different
times of electropolishing was assessed in the SEM.
3. Fracture behaviour
Orthodontic wires are heavily cold worked with most deformation in the
surface layers. As-received wires were deliberately cold-worked to the
point of fracture and then assessed in the scanning electron microscope
to assess the extent of the heavily distorted surface layer and the
distribution of deformation through the wire cross-section.
Electropolished specimens were also assessed, to determine if the heavily

cold worked surface layer was removed.
4. Another objective was to study the detailed microstructural data in

transverse and longitudinal sections of specified wires using transmission
electron microscopy. Fine scale features such as grain size, dislocation
density, and second phase size, shape and distribution were to be
assessed.
No-one other than Singh 1991 has looked at the m¡crostructure of

orthodontic wires. Comparison of mechanical property data results
between different studies and especially different sized wires are
therefore probably meaningless.
The more a wire is drawn, the more cold worked ¡t is, the finer the
effective grain size and consequently the higher the strength. To
understand any differences in wire properties we must know if the
structure is different.
3T
CHAPTER 4
MATERIALS AND METHOD
The materials studied included orthodontic wires of (1) stainless steel

(special plus, premium, premium plus, supreme and pulse straightened
grades), (2) nickel titanium and (3) superelastic nickel-titaniumr.
1. a. Conventional tensile tests using a universal lnstron testing

instrument (model 8501 ), 1OOkN load cell, in conjunction with an
extensometer, with a cross head speed of 0.5mm per minute to comply
with the Australian standard 1964-1977 (Fig. 41. The stainless steel
wires assessed were as listed below2.
i. sssp o16, 01 8, O20, O22

ii. sspp o12, O14, 016, 018
iii sss 010
Approximately 24Omm lengths of wire were attached to

aluminium
casings before entering the lnstron grips which were spaced
approximately 2OOmm apart (F¡g. 5). This overcame the problem of the
wires failing prematurely at the point of grip. The wires were loaded until
fracture.
b. Bend tests. Three different bend tests were compared to obtain

information on the load deflection behaviour : wire in arch form, a
cantilever bend geometry and a three point bend system.
Testing apparatus
The testing apparatus consisted of:
A milled brass wire holder held in a stand which allowed adjustments to

be made in the vertical plane, using a micrometer drive (calibrated in
lThe material/condition abbreviated identifications are given in the forward p' (vi).
2Although Sl units should be used in scientific studies the nomenclature 018 = 0.018"
(0.45mm) is used for identification since most orthodontic wires are supplied in imperial
units and will be familiar to practitioners.
32
I
Figure 4: A photograph of a universal Instron testing instrument with an

extensometer in Position.
33
Figure 5: A photograph illustrating a stainless steel wire wrapped around
an aluminium casing.
34
ímperial units) (Fig. 6). Slots in the wire holder enabled the wires to be
positively located. When the wires were tested in archform the distance
between the distal ends when placed in the slots was 7Omm.
A milled steel holder was glued to the scale (Mettler 6100) pan. The
anterior section of the archwire was placed into a groove in the steel
holder. In the cantilever tests the wire was either supported in a slot (Fig.
7l or knife edge while the three point bend test specimens were
supported on two knife edges 25mm apart.
Vertical movement of the brass holder would then apply a load on the
supported anter¡or segment of the wire which could be recorded on the
Mettler scale.
Tests performed
The wires were loaded in increments of ten thousandths of an ínch

(O.25mm) and the load reading recorded in grammes to two. decimal
places. The wire was then unloaded in ten thousandths of an inch
increments and readings recorded.
Some archwires were then electropolished where indicated and the test
repeated.
(1.) Arch wire preoaration
Plain archwires without circles were formed with anchor bends of varying
sizes, from 140mm lengths of ss wire of nominated diameter. 1O
millimetre lengths of steel tube were glued to each end of the wire wíth
Selleys five minute Araldite (epoxy adhesive). This allowed the wire to be
reproducibly located in the apparatus used to apply a load in increments.
35
Figure 6: A milled brass wire holder in a stand with a micrometer drive
which allowed vert¡cal adjustments
36
,+
A bend test of a wire specimen in archform in a slot support.

Figure 7:
Support glued to the scale pan to allow load/unload values to be
recorded.
37
Archwires tested
¡.) Archwires with 45" anchor bends were formed from three different
spools of Australian Wilcock stainless steel wire and then loaded and
unloaded as described above.
sspp 016 5 specimens

ssp O2O 5 specimens
i¡.) Six archwires were formed from sssp O2O. Archwires 1-3 had 35'
anchor bends while archwires 4-6 had 25" anchor bends. This allows the
effect of different anchorage bends to be assessed.
sssp O2O 6 specimens

The wires in groups ¡.) and ii.) were electropolished later and then
retested.
i¡i.) Eight archwires were formed from three different spools of

stainless steel Wilcock wire of varying diameter and grades including the
recently released pulse straightened wires. 45" anchor bends were placed
in the archwires excluding sssp O2O (12-13) where 55" anchor bends
were formed. This enabled the effects of different sized anchorage bends
to be assessed.

sspp ps O1 I 2 specimens
sssp O2O 4 specimens
Electrooolishing
An electropolishing unit (Rocky Mountain 7OO, USA) containing 85o/"

orthophosphoric acid was used to electropolish the archwires at the fixed
potential of 13V D.C., for varying lengths of time as listed. The specimen
was attached with a crocodile clip placed at the opposite end to the steel
tube. Care was taken to ensure that the specimen was fully immersed to
avoid the archwire overheat¡ng and annealing. The specimen was also
gently agitated during electropolishing (F¡g. 8).
38
Figure 8: An electropolishing unit. Specimens were attached to a
crocodile clip, fully immersed in orthophosphoric acid and then a current
passed through the wire.
39
Ten sections of sssp O2O (O.50mm) were cut. Six specimens were
electropolished continuously for 25,50,75, 1O0, 125 or 150 seconds
respectively. Four wires were electropolished for twenty-five second
intervals until a total of either 25,50,75 or lOO seconds was reached'
This was to assess ¡f the intermittent electropolishing affected the wire
properties differently to continuous electropolishing. Wire diameter was
recorded at three different locations before and after electropolishing;
near the point of attachment, midwire and furthest from the attachment.
Wires tested in the as-received state were then electropolished for

varying lengths of time as stated below prior to loading and unloading.
i.) sspp 016 3 specimens ep30"

sspp O16 2 specimens ep45"
sspp 018 2 specimens ep3O"

sspp 018 1 specimen ep45"
ssp O2O 3 specimens ep30"

ssp O2O 2 specimens ep45"
ii. ) sssp O2O 3 specimens ep25" plus ep5O"

sssp O2O 3 specimens ep5O" plus ep50"
Archwire diameters were recorded before and after each electropolishing

with a micrometer (Mitutoyo Japan).
(2.) Cantilever tests
Many of the bend tests reported in the literature are conventional

cant¡lever tests. The American Specification, No 32 specifies that a one
inch length of wire should be used for cantilever tests. Drake et al, 1982
stated that one inch lengths were used ¡n the cantilever tests. For these
reasons a similar length was used in this study.
40
Thirty five millimetre lengths of wire were cut to take into consideration
the lOmm lengths of steel tube glued to one end.
F
:¡:il;i
(diagram from Ware 1967)
i.) The following wires were assessed in the as-received state. A steel
milled holder with a slot was glued to the scale pan and used to support
the free end of the archwire at a distance of 18mm. Some wires were
also electropolished as listed below and retested.
at O18 x O25
nt 016
sspp 016 ep30" or ep45"
sspp 018 ep30" or ep45"
sspp ps O18
sssp o20 ep25" plus ep50"
ep50" plus ep50"
ii.) A knife edged milled steel holder was glued to the scale pan to
support .the free end of the wire (F¡g. 9). A distance of 2Omm was left
between the steel tube and the knife edge support' The wire types and
electropolishing times were the same as for (2) ¡.).
¡ii.) Six lengths of sspp ps O18 wire specimens were prepared. Each
specimen was tested twice, using a distance of 10mm and 2Omm
between the steel tubing and knife edged support. This was to assess
whether distance affected the results.
Two specimens were tested by loading and unloading, prior to

electropolishing.
4l
'1
Figure 9: A cantilever bend test with a knife edged support glued to a
scale pan to record load/unload values.
42
One of these specimens was then electropolished for periods of twenty
five seconds until a total electropolishing time of one hundred seconds
was achieved. After each period of electropolishing the wire was loaded
and unloaded at lOmm and 20mm.
The remaining four specimens were electropolished for 25, 50, 75 and
1OO seconds respectively and then tested. This was to establish if the
results were altered when the electropolishing was interrupted.
iv.) Annealed cantilever specimens were assessed using a distance of

2Omm between the steel tubing and knife edged support. The wires
assessed were as listed.
at 018 x O25
bt O17 x O25
nt 016
SSS 010
SSPP 012
SSPP 014
sspp 016
sspp 018
sssp o20
One specimen each, excluding sssp 020 where two Speclmens were
tested.
v.) Four specimens of annealed sssp O2O were prepared. These

specimens were tested using a distance of 2Omm between the knife
edged support and steel tub¡ng.
One wire was electropolished for intervals of 25 seconds and retested

after each period of electropolishing, until a total of 1OO seconds
electropolishing time was reached.
The remaining four wires were electropolished for 25,50, 75 and 1OO
seconds respectively and then tested.
43
Plastic deformation of the annealed wires probably occurred with loading.
Fewer unload values were consequently recorded when compared with
the load values.
vi.) After assessment of the initial elastic modulus values calculated

from the cantilever tests some tests were repeated.
It was noted in previous tests that the load and unload values took a long
time to equilibrate possibly due to stress relaxation. To standardise, thirty
seconds was allowed to elapse between altering the extension and
recording the load/unload value. lnstantaneous values are probably more
accurate but are too difficult to measure.
(¡.) Six new sssp O2O specimens were prepared and tested on a knife
edge support at a distance of 20mm. These wires were not annealed or
electropolished. Three specimens were tested without rernoving the
curve present as a result of being on a spool. The other three specimens
were straightened with fingers prior to loading.
The initial reason for repeating tests on sssp 020 was that the elastic
modulus values calculated from the load data, were significantly higher
when tested in the annealed state compared to the as-received state.
(ii.) Six annealed sssp O2O specimens were prepared and tested on a
knife edged support at 2omm. Two specimens were tested after
annealing but no electropolishing.
One specimen was electropolished for 25 second intervals and retested
until a total electropolishing time of 100 seconds was achieved'
The remaining four wires were electropolished 25, 50, 75 and 1OO
seconds respectively and tested.
The reason for repeating these tests was that the elastic modulus value
increased when the electropolishing time was 50 seconds compared with
a 25 second electropolishing time. The general trend noted was a
decreasing value for elastic modulus as the electropolishing time
increased. The initial increase noted may be explained by hydride
formation and precipitation hardening'
44
vii.) To assess for potential effect of wire diameter on elastic modulus the
following wires were assessed in the as-received state on a knife edge
support at 2Omm. The specimens were 35mm long and had l omm
lengths of steel tubing glued to one end.
nt o14 I 016 / 018

sent 016/018x025 lO22xO28
The nickel-titanium wires were retested at a distance of 1Omm.
(3.) Three ooint bend tests
Two milled steel knife edge supports were glued to a scale pan with the
edges parallel to each other and spaced 20 mm (Fig' 1O)'
(diagram from Ware 1967)
Six 25mm lengths of sspp ps 018 wire were cut. Two specimens were
tested in the as-received state while the remaining four were
electropolished 25,50,75 and 1OO seconds respectively. The diameters
were recorded before and after electropolishing. A milled steel circular
plunger was attached to the stand and centrally placed on the wire
specimen. lt was then loaded in increments of ten thousandths of an inch
(O.25mm). Load readings were recorded in grammes to two decimal
places. The wire was then unloaded in ten thousandths of an inch and an
unload value was recorded.
45
rt
f--
,l
I
*r t t
Figure 1O: A three point bend test with two knife edge supports glued to
the scale pan and a centrally placed circular plunger to load and unload
the wire.
46
Use of the load and unload data
The thousandths of an inch measurements were converted to a metric

unit and a graph with a stra¡ght line of best fit was drawn to show the
load and unload separately, in relation to the extension. Where the
accuracy of this line of best fit was poor the values at the start or
completion of loading and unloading were not used.
The equation of best straight Iine fit y = mx + c was determined. A

value of 1 was used for x to determine the y value. ln this case x and y
values refer to the x and y axis respectively.
Elastic modulus for cantilever tests were calculated from the follow¡ng
formula (Beam Theory Methodology)
E - I'a
-3
3vI
where F: load/unload value (y axis)

a: distance between steel tube and support
slot knife
a 4
I L
/ :
extension (x axis)
Calculation for 1 (second moment of area) varies depending
on whether the wire is round or rectangular in cross section.
: *do
I for circular wire d: diameter
64
I tor rectangular wire : bh3

ó: breadth
t2
å: height
47
To calculate values of elastic modulus for three point bend tests the
following formula was used
Fa'
E
48yI
The formulas used for calculation of the elastic modulus are from Blake,
1982.
c. Resonance
The following wires were tested in the as-received state
sss Ol O
sspp 012,014,016,018
sspp ps 016,018
sssp O2O
1 15mm lengths of wire were cut and one end was clamped inside a
milled brass wire holder. A 1g plastic weight was attached l OOmm from
the support. The distance was reduced to 50mm with the more flexible
sss O1O wire.
The wire was manually displaced to oscillate the wire specimen. The
frequency of a Strobotac type 1531 General Radio Company, Concord,
Massachusetts, USA was adjusted until the wire appeared stationary
(subjective approximately SHz) (Fig. 11). This was best done in a
darkened room. The modulus of elasticity can be calculated from:
I I-trEr
.f^
2n \ mt'
Where "fn
: frequency of vibration
E : elastic modulus
I : moment of inertia
m : mass
I : distance
48
Figure 11 Equipment used to determine the natural frequency of
vibration of wire spec¡mens.
49
The formula appeared very sensitive to diameter changes. A Mitutoyo O-
25mm O.OOOI mm accuracy m¡crometer was used to measure w¡re
diameter. The diameter used was an average of three different
measurements.
The sspp 016 and sssp O2O were tested after electropolishing. Three
specimens of each were electropolished for 25" and another three
specimens of each were electropolished for 1O0".
d. Soeed of sound
The speed of vibration through a wire is thought to be the most accurate
method of elastic modulus calculation. A Systron Donnor Corporation
Datapulse 1 16 Pulse Generator was used to generate a voltage which
was amplified in a SOW amplifier prior to being passed through a l2Omm
length of as-received sssp O2O wire. A piezoelectric crystal
(cyanacrylated to transducers) detects the vibration, which is then
passed through a Charge Amplifier Type 2635 Bruel and Kjaer and finally
to a Hewlett Packard 54601A Oscilloscope (Fig. 121. The vibration was
then detected by a second piezoelectric crystal spaced at either 40,60,
80, lOO or 12Omm and as before passed to the oscilloscope. Two wave
forms appear and cursor placement (at the peak of the wave crest, depth
of the trough or at the first sign of deviation) allows the time elapsed
between the vibration detected at the first and second crystal to be
calculated in microseconds:
Where v : velocity :
distance x time
E : elastic modulus
p : density ( SMgm-3 for stainless steel 304 )
Different amounts of tension were placed on the wire. Tension did not
appear to affect the consistency of the results. Varying the distance
between the piezoelectric crystals may have an effect. This may be due
to proximity of the crystals to the grips holding the wire specimen.
50
Q,Ê ç
Figure 12 Equipment used to calculate the speed of sound through a
wire.
51
The formula used to calculate the velocity was taken from Ashby and
Jones 198O.
A sssp O2O wire specimen was annealed and then tested under similar
conditions. This was to remove the potential anisotropic effect possible in
such heavily cold drawn wires (Goldberg et al 1977l.'
e. Microhardness test uring the Leitz Wetzlar Germany 4262 miniload

hardness tester. A length of as-received sspp 018 wire was shaped to
enhance mechanical retention in the Bakelite (coloured phenolic resin)' A
Buehler Simplimet ll was used to embed the specimen in Bakelite. The
specimen was polished to a mirror finish, placed on the hardness tester,
viewed at 4Ox magnification and then tested using a "confocal" small
pyramidal diamond indentor. A 30 second time period was allowed to
elapse from the time of release of the (camera) trigger used for applying
the load, before removing the load. Approximately 5Oo/o of this time was
the actual loading time. This method of loading was continued across the
width of the wire and repeated with a 25 pond and then a 50 pond
weight. A pond (P) is equivalent to a 1 gram force.
Due to the Bakelite mounting the specimen needed to be gold / palladium

coated prior to viewing in the Philips 505 SEM. Photomicrographs were
recorded at 1.42F.3 to assess for changes in hardness. A collage was
then made. The length of the two diagonals were calculated from the
photomicrograph. When assessing the photom¡crographs ¡t must be
remembered that the specimen is thinner near the edges so greater
deformation would be anticipated. These values would be less reliable
(Fig. 131.
From this information the Vickers microhardness number could be
calculated from tables or the following formula.
1854 x P
HV
d2
Where P: applied load in grams

d:the distance of the diagonal of the square in micrometers
52
Figure 13: Photomicrograph of microhardness test (magnification 142Oxl
which illustrates the greater deformation near the edge of the specimen.
53
Macrohardness test using the Vickers Pyramid Hardness Tester.
The standard Vickers indentor is a diamond in the form of a square based
pyramid. VHN is independent of applied load but for ease and greater
accuracy of measurement it was recommended to choose a load which
gives an impression that fills two thirds of the field of view. A 2O-3Okg
load was recommended for steels.
The specimen used for microhardness test was polished sufficiently to

remove the gold palladium coat¡ng. A 20kg load was initially attempted
but it was not possible to indent the wire only and consequently a
calculated hardness number would be meaningless. Even when a Skg
load was used it was very difficult to indent the wire only. A microscope
was used to measure the lengths of the diagonals of the indentation.
From these measurements the hardness number was calculated by either
of the following formulae.
P
DPN
A
t.854P
or DPN
D,
Where DPN : Diamond Pyramid Hardness Number

P : the applied load in kilograms
A : the surface area of the pyramidal impression in
millimetres squared
D : distance of the diagonal in millimetres
The formulas used to calculate microhardness and macrohardness tests

are from the Miniload Hardness Tester Manual, 1968.
2. Surface appearance.
a. A sample of as received sspp 018 and sspp ps 018 were prepared to
enable surface differences as a consequence of pulse straightening to be
assessed in the SEM. Pulse stra¡ghtened wires feel smoother to the
fingers.
54
b. Wire surfaces were examined for changes after electropolishing. Seven
lOmm lengths of wire, were cut from sssp O2O wire' Six were
electropolished for 25,50, 25 + 50, 50 + 50, 25 + 50 + 50, 50 + 5O
+ 50 seconds respectively. The wire sections were then attached to
stubs with conducting "silver dag" and assessed in the scanning electron
microscope (Philips 505 SEM) at magn¡fications of 75 and 3OO' The
seventh specimen was not electropolished to act aS a control to
compare surface changes.
c. Fractography. Wires were deliberately maltreated by bending in pliers

at a single point many times until point of fracture. This is not indicative
of clinical practice but the aim was to assess the fractured surface to see
if there were textural differences between the surface layer and central
core. The range of wires chosen included sspp 01 2, 014, 016, 018, sspp
ps 016 and 018, nt 016, se nt 01 6, O2O,018 x O25, 0215 x 028. The
first bend was 9Oo and subsequent bends were 18O". The number of
bends required to fracture the wire were recorded.
Seven 25mm specimens of sspp 018 were cut. Four specimens were
electropolished for 25, 50, 75 and 10O seconds respectively. All
specimens were then work hardened until point of fracture. The fractured
specimens were attached to stubs with silver dag, assessed in the Philips
5O5 SEM at varying magnifications and photographed.
Another Seven specimens of the wire from the same spool were
prepared. Prior to manipulation these wires were heat treated at 3sO"C
for ten minutes and then allowed to bench cool. A high temperature
(14OO'C) chamber furnace was used and the temperature was checked
using a "Carbolite" thermocouple and ten minutes were allowed for
equilibration to occur. The reason for the heat treatment was to remove
the potential effect of hydrogen embrittlement. Steels, especially high
strength steels are prone to hydrogen embrittlement'
The composition of Some specimens were assessed using a Tracor

Northern TN 55OO Model EDS system. Aluminium was present in some
areas assessed despite being well away from the stub. An additional two
fractured specimens of sspp 018 were prepared, one from the same
55
spool and the other from a new spool. These were then attached to a
graphite covered aluminium stub prior to ED analysis.
3. Detailed microstructural data.
Specimen preparat¡on determines the quality of TEM micrographs'

Mechanical grinding is the fastest technique but may cause unacceptable
damage and is generally not feasible for final thinning. Chemical and ion
or fast atom thinning methods thin slowly and unevenly resulting in only
small electron transparent areas but produce minimal damage. Dimple
grinding produces an exceptionally smooth surface and reduces specimen
thickness w¡th m¡nimal damage. lt is precise enough to produce a final
thickness of less than 3pm in the centre and leave a thick supporting rim.
A short period of chemical/particle beam thinning to produce electron
transparent areas is then required.
Alldiagrams are copíed from the "TEM specimen preparation of

semiconductors, ceramics and metals by ion milling" handbook (Gatan
1992).
Four different wire types were initially selected:

1. sspp ps O18
2. ss O175 multibraided
3. ss O21 x 025 braided
4. nt 016
Since foil preparation of multibraided stainless steel wires proved

particu¡arly difficult due to the small cross section of the individual wires
this was discontinued.
56
The final wires that were prepared for foil specimens in cross section
included
1. sspp 018
2. sspp ps O18
3. at O18 x O253
4. nt 016
As many short lengths of wire as possible were embedded in Gl Epoxy /

Hardener (Gatan) (10 parts resin to 1 part hardener) withín tub¡ng of
approximately 3mm external diameter to approximate the size of grids
used in TEM holders. The wires and steel tub¡ng were first cleaned with
acetone to remove any contaminants. Thick walled tubing was used to
minimise the amount of epoxy present. Wire samples from spools tended
to be curved, so to assist placement in the tubes the wires were first
straightened. The epoxy was syringed into the empty tub¡ng and the
wires were individually coated with epoxy before being placed ¡nto the
tubes to minimise the chances of air entrapment.
The samples were then placed in an oven for five minutes at 1OO"C for
the epoxy to be cured (Fig. 14). Curing was indicated by a colour change
of the epoxy from amber to Pink.
The tubing was cut in cross section to approximately O.5mm thickness,

using a carborundum disc at slow speed and an oil water coolant to
avoid overheating. The specimens were held in a milled steel holder and
manually thinned on silicon carbide 12OO grit paper with water to
approximately 0.1 - O.O7mm (Fig. 15).
A dimple grinder Gatan Model 656 (Fig. 16) was used to thin the
sections after centralising the specimens on the platform with cross hairs
under magnification. A 2Og weight was applied during grinding and the
slowest speed was chosen (1OO revolutions in 1 minute and 38
seconds).
A 15mm spherical bronze wheel 656-0106 with Gatan Desc diamond

cutting compound 2-4pm and water lubricant were used. To avoid
contamination a wheel is used with only one polishing compound. To
minimise surface scratch¡ng the specimens were polished with Gatan
57
r60
40 G-l EPOXY
O
Il rzo
õ
q
o)
roo
t
o)
t--
gso
f
O
60
40 ll
50 r00
Curing Time (minules)
Figure 14: Curing character¡st¡cs of G - 1 epoxy used for prepar¡ng TEM

cross sect¡ons.
D¡sk Cutl¡nq
70-100pm
Oisk grinding
Figure 15: This diagram illustrates cutting the tub¡ng in cross section of
SOOpm and then disc thickness after manual gr¡nd¡ng to 70 - 1OOpm.
58
ønleMþht
bad scåle
0 - sogm
III
*
fÍi/otsd
plâtlom oro
a
o 116,l
ra¡s€ loer€r mlcroíreler sp€omn

cam dr¡ve platform
wtÞel o€r nragætc
lumtaue
Figure 16: Precision dimple grinder and polisher
59
Alumina Pol suspension using a felt wheel. After dimple grinding the
centre of the specimen should be approximately .0o5mm (F¡9. 17).
Specimen thinning from both sides was attempted to start with. Due to
the thinness of the foil it proved difficult to invert the specimen without
disrupting the epoxy matrix. Sticky wax (a low melting point
thermoplastic polymer) was used at first to firmly attach the specimens
to the stub. To form a strong and uniformly thin bond the wax must be
heated to around 130'. The effect of heating the sticky wax to remove
the specimen after dimpling, was sufficient to soften the epoxy and
dislodge the wires. Superglue was then used to the adhere the specimens
to the stubs and 75% acetone was used to dissolve the superglue.
Specimen removal was still difficult due to close adaptation of the
specimen to the stub and limited exposure of the superglue to acetone.
An alternative method was devised where the specimen was
mechanically held with a mounting plate fixed with screws.
lnitially specimens were thinned in the Gatan dual ion mill model 60O at
room temperature using a grinding angle of 15", gun current O.5mA DC
and gun voltage 3kV (Fig. 18). The time required to perforate the
specimen was much greater than that stated in the manual where 30
minutes was suggested.
The specimens disintegrated and were lost before perforating due to

preferential loss of the epoxy. The performance of the G1 Epoxy I
Hardener (Gatan) was questioned as its shelf life was approaching twelve
months. ln addition, although storage at room temperature was stated to
be adequate, in Australia where high temperatures are experienced,
refrigerated storage may be more appropriate. New epoxy was ordered
and specimens were reprepared. However the problem was not
overcome.
To avoid loss of the specimens the foils were placed between two nickel
lOO slotted grids placed at right angles to each other. The grids were 2o/o
nitrocellulose coated followed by carbon coat¡ng. Only one gun was used
with the same current and voltage but at a lower grinding angle of 8". An
earlier attempt at ion milling between copper grids was unsuccessful due
to the copper being milled at a faster rate than the specimens.
60
a b
1smm dia.
lOmmdia. .5Ém
Figure 17: ThiS diagram illustrates: (a) a standard grinding wheel and
mount and (bl the thickness of the specimen centrally after dimple
grinding.
61
lon Beam
a
lon Beam
25
b
e Maler¡al: Copp€r
3zo
È
Gun Current: 0-5mA
Beam Angle: 20degrees
f Gas: Argon
o
-c
Ø
3ls
-9
g
(I)
(õ lO
(r
o,
.E
c
c
7s
l-
0
2_ 6 I 10 t2
Gun Voltage (kV)
Effect of OctogunrM Voltage on Specimen Thinning Rate
c 00
gl¡orcas. l.rú
lldal gün úed)
a
o-
E
(ú
C
o)
þN BEAM
()f
c(l) e
E
'õ
q)
o 9ECM€N
U)
40
0 t0 20 30 40
Beam Angle (degrees)
Specimen Current vs Beam Angle for Argon Milling

of Copper
Figure 18: These diagrams illustrate the thinning action of the Gatan dual
ion mill. The effect of OctagunTM voltage, spec¡men current and beam
angle on thinning rate are ¡llustrated.
62
The specimens were removed every two hours for viewing in the
scanning electron microscope to assess the thinned areas' After
approximately ten hours of ion milling the specimens appeared to have
sufficiently thinned areas. They were then transferred to folding copper
grids (because the nickel grids were not able to be attached) which had
also been coated in with 2o/o nitrocellulose and carbon. This proved
difficult owing to the fact that the Specimens were now magnetic'
Floating the specimens on a water bath minimised handling of the
specimens but small fragments did appear to break off from the fo¡ls.
The specimens were examined in the JEOL l OOS but no electron

transparent areas were found. The goal is to examine these specimens in
the transmission electron microscope (JEOL 2OOOFX) to assess the
crystalline structure and diffraction patterns.
Gatan was contacted regarding the problems being experienced. Their

suggestions included using a tube of the Same material as the wire which
is feasible with the stainless steel wires but not for the alpha-titanium or
nickel-titanium wires. The importance of minimum epoxy and maximum
wire density was re-emphasised. Stainless steel rods were milled with an
O.O16" and O.018" central hole and a slot of 0.O18" x O.025" to allow
new specimens to be made with minimum epoxy present (Fig' 19)' ln
addition the rectangular specimens were cured under pressure using a
small G clamp.
The wire surface was also roughened with 12OO grit paper before
embedding to enhance mechanical locking of the epoxy to the specimen.
Desc. Cubic Boron Nitride (4-6pm and O-2¡rm) were recommended as the
polishing paste and a low angle ion mill was also recommended'
Longitudinal
Two wire types were selected:
1. ntO18
2. sspp ps O18
The specimens were attached to double sided cellotape and mechanically
thinned on 12OO grit paper to O.1mm. This was a very labour intensive
procedure.
63
Figure 19: Rods of stainless steel were milled with a central hole the
same size as the wire specimen to minimise the epoxy present.
64
A m¡croprocessor-controlled table top machine called Abramin was
utilised for automatic grinding, lapping and polishing of all materials. Wire
specimens were attached to acrylic blocks with sticky wax and then
placed in a specimen holder. The specimen holder and polishing disc
rotate at 1Sorpm. The holder is eccentrically positioned relative to the
polishing disc to ensure a fast and good polish without creating comet
tails. A lOON grinding pressure was used on 22OP and lOOOP grit paper
with a water lubricant.
For final polishing and thinning to perforation an E. A. Fischione model

12O-23O twin-jet electropolishing machine was used (F¡g. 20)' A 3mm
length of the thinned specimen was cut and placed in the lollipop holder
ensuring that there was contact with the platinum ring on either side.
The polishing solution consisted of 1oo/o petchloric acid, 20% glycerol
and 7Oo/o ethanol. This was cooled to approximately -2O"C with liquid
nitrogen. Once the solution was cooled the speed was adjusted so that
the jets would just meet. A 25V D.C. voltage was selected. When the
specimen perforated it was removed from the lollipop, rinsed in ethanol
and placed on a copper grid and viewed in the JEOL 2OOOFX.
Despite careful preparation no suitable thin foils could be produced'
65
Figure 20: Twin-jet electropolishing machine used to perforate prethinned
specimens.
66
CHAPTER 5
RESULTS AND DISCUSSION
Tests are performed on resil¡ent orthodontic wires to determine their

mechanical properties. There are three main methods of testing the
mechanical properties of an archwire. They include using a tensile testing
machine, utilising a cantilever beam set up and measuring the resilience
of a wire.
The Australian Standard (AS 1964) requires three tests; tensile strength'
wrapping ability and resistance to bending. The force required to fracture
the wire being held between the two grips is used to calculate the tensile
strength (standards Association of Australia 19771. The American
Specification, No. 32, tests similar wire properties; flexure yield strength,
modulus of stiffness and resistance to bending. Flexure yield strength is
derived from cantilever beam tests and from this the modulus of stiffness
is mathematically catculated (Council on Dental Materials and Devices
19771.
Stainless steel wires have dominated in orthodontics because of their

desirable physical characteristics. They possess a good balance of
environmental Stab¡l¡ty, stiffness, resilience, formability and low coSt.
Despite their long history of use fundamental questions st¡ll remain
regarding their metallurgical and mechanical properties (Khier, Brantley
and Fournelle 1988). For these reasons emphasis was placed on
assessing stainless steel orthodontic wires in this research project'
Elastic modulus values
Stiffness, springback and formability are generally considered to be the

three most important properties of an archwire. The stiffness of a wire
determines the magnitude of the force delivered by the appliance.
Although light physiological forces are desirable for tooth movement, the
archwire must have sufficient stiffness to resist undesirable plastic
deformation during function.
Goldberg and Burstone 1979 (as stated in Sokel 1984) considered that
wires with a low load deflection rate have distinct clinical advantages.
67
The desirable characteristics include the ability to apply lower forces; a
more constant force over a longer period of time; greater ease and
accuracy in applying a given force; the ability to use larger act¡vat¡ons
and the assoc¡ated increased "working time" of the appl¡ance.
Stiffness may be mathematically calculated for wires which obey Hooke's

law for elastic bodies. lt is proportional to the modulus of elasticity which
may be calculated from the stress-strain ratio determined at any point on
a linear section of the graPh.
ln this research project elastic modulus values were calculated from data
acquired from a simple tensile test, bend tests, natural frequency of
osc¡llation and measuring the velocity of a vibration along a wire.
properties of orthodontic wires were assessed in the as-received state
and after manipulative treatments such as electropolishing and heat
treatment.
Modulus of elasticity for a pafticular material has traditionally been

thought to be an invariable property (whittle 1993). The value for
young's modulus is determined by the atomic number of the metal and is
said to be only marginally affected by heat treatment or cold work¡ng
(lngerstev, 1966). Forsyth and Stubbington (1975) (as stated in Sokel
1g84) drew attention to the importance of the crystal orientation to the
value of the modulus with respect to the titanium alloys."
Masson 1969, assessed 016" orthodontic wires and noted that Young's
Modulus of Elasticity for stainless steel orthodontic wire in the cold-
worked or heat treated condition is not a constant value. The accepted
value for the modulus of elasticity of stainless steel wires was
approximately 30 x t 06 lOs/¡n2 (ZOlCea) but Masson claimed that we
have been mistaken in making this assumption. Later in 1977, Goldberg
et al showed a substantial decrease in the value of modulus in stainless
steel with heavY cold working.
It ¡s of interest to note that values of 155 GPa, which are twenty percent
below the published values of 193-2OO GPa, have been calculated for
016 unspecified stainless steel wire (Yoshikawa et al 1981). At the other
extreme values as high as 248 GPa have been reported by Goldberg'
68
Accurate measurement of the modulus of elasticity for thin wires poses
some difficulties (Goldberg et al 1983a/b).
A. Conventional tensile tests
Masson 1969 stated that the elastic modutus value for stainless steel
varies depending on which wire is being assessed. The variation is most
likely due to the amount of cold work and stress relief heat treatment
that the wire has undergone during manufacture, The values for the
Modulus of Elasticity (lbs/sq.in.) in tensile testing arc listed below
together with the metric equivalents3. Values increased through the
range of .016 Wilcock wire from "Regular" up to "special Plus" as listed
in Table 1.
Table 1: Elastic modulus values for as-received stainless steel wires from
tensile tests of O.O16" stainless stee¡ (ss) wires (Masson 1969)'
wrRE 016 SS ELASTIC MODULUS

(106 lbs/¡n2) (GPa)
Regular 22.8 ¡ O.9 157 t6

Regular Plus 23.2 * O.7 160 + 5
Special 23.2 * O.9 160 * 6
Special Plus 24.6 * O.8 170 Í 5
It was noted that an .O16 Dentaurum Remanit "Super Spring Hard" is a

precipitation-hardening stainless steel that has been heat treated and its
E was the highest of all the wires tested at 26.1 t O.8 x 106 lbs/in2
(180 + 5 GPa).
Twelftree 1g74, noted that elastic modulus values calculated from tensile
testing varied from 150 to 174GPa depending on the grade of stainless
steel (Table 2). This finding was in agreement with Masson's 1969'
3Ahhough orthodontic wire is supplied in imperial units and many early results
(particularly American) also used imperial units, scientific results are now invariably
presented in S.l, units. lt has been decided in this thesis therefore to present where
necessary both original published results and their metric equivalents.
69
tensile tests of stainless steel wires (Twelftree 19741'
WIRE STAINLESS STEEL ETASTIC MODULUS

lGPa)
S cial 150 i 5
S al lus 164 t4
mrum 17o^ f 6
premium plus 174f6
sokel (1984) compared the tensile properties of different wilcock

stainless steel wires and noted the large variation in values for Young's
modulus of elasticity (Table 3). Values for Young's Modulus showed a
considerable scatter with a range of 154 - 199 GPa and a mean value of
174 GPa was noted.
70
Table 3: Elastic modulus values for as-received O.O1O"-O.O2O" stainless
steel wires from tensile tests (Sokel 1984)'
WIRE TYPE ETASTIC MODULUS
O1O unspecified (probably supreme) 172

179
012 Special 168

168
012 Special Plus 173

184
014 Regular 156

181

164
016 Regular 188

199
193
018 Regular 163

163
189
020 Premium 166

154
No obvious correlation was noted between the elastic modulus value and
either wire diameter or wire grade.
Sokel ment¡oned the difficulties in the tensile testing of orthodontic wires

due to the the¡r high tensile strength and elast¡c modulus and low
ductility. For these reasons experimental apparatus must be extremely
accurate s¡nce very small percentages of elongation are being measured'
Secondly any defect in the wire such as a kink would greatly affect the
results. Wire spec¡mens and the grips were closely assessed for defects
before testing. The third difficulty is in accurately read¡ng the results.
ldeally the stress-stra¡n sect¡on is straight in the elastic region while an
ever increasing curvature is noted in the plastic region.
'tt
Many wires used are in the cold drawn condition and consequently the
Stress-Strain curve was found to be curved almost from the start of
loading. Young's Modulus is determined by measurement of the slope of
the ¡n¡t¡al portion of the curve. Many different tangents could be chosen
and therefore values obta¡ned were subject to a large var¡ation (Fig. 211.
As a consequence a range of elastic modulus values may be more
appropr¡ate than a single elastic modulus value.
Despite the difficulties to measure elastic modulus values the results are
comparable to the 175 GPa quoted by Yoshikawa et al (1981)' The
values are well below the figure of 193 GPa quoted by Goldberg and
Burstone (1g79). lt was proposed that a different experimental method is
required to measure Young's modulus with greater accuracy.
Singh lggl noted after preliminary tensile tests on alpha-titanium

straightened and combination wires, consistently low elastic modulus
values despite repeated test¡ng and the use of an extensometer. These
low values were found to be similar to those obtained by Hazel, Sokel
and West , 1984.lt was interesting to note that values for yield strength,
tensile strength and O.2o/o proof stress were consistent w¡th published
values for alpha-titanium (Singh 1991).
Tensile testing is unable to test the formability of wires. A wire may look
promising according to the quantitative tensile tests but may be clinically
unsuitable due to poor formability. Despite this limitation of tensile
testing and the large variation in elastic modulus values, tensile testing is
still important because of the comparable data available.
It has been suggested therefore that the use of tensile data to determine
flexural data for fine orthodontic wire is inappropriate. Orthodontists
primarily stress wire in bending in the clinical situation (Goldberg 1983b;
Nikolai et al 1g88). For these reasons alternative methods of test¡ng the
mechanical properties of wires were explored such as bend tests'
Not much information is available as to the effect of bend versus tensile

testing on the elastic modulus values. The imperial values from Table 4
are taken from the thesis of Masson 1969. Again metric values have
been calculated to aid comparison with other data'
72
U)
U)
11,
E.
t--
(/)
D
B
A-
c./
STRAIN
Figure 21.. Diagrammatic representation of a stress-strain curve

illustrating two tangents A - B and C - D which could be chosen to
determine the elastic modulus.
73
tensile and bend tests (Masson 1969).
MATERIAL TENSILE
ELASTIC MODULUS
(1 06 lbs/in2) (GPa)
Wilcock Special Plus 016 24.6 + 0.8 170 + 5
Unisil (ss wire from Unitek) 22.8 + O.8 157 ¡. 5
016
MATERIAL BEND
ETASTIC MODULUS
(106 lbs/in2) (GPa)
Wilcock Soecial Plus 016 24.3 *. 1.5 168 t 10

Unisil (ss wire from Unitek) 24.8 + 1.2 171 + I
016
It was noted that the value for Young's modulus of Elasticity for Wilcock
"Special plus" wires in tension was very similar to the value obtained in
bend tests. This was not the case for "unisil" wires where values
obtained from tensile tests were lower than those from bend tests'
Masson 1969 believed that the Wilcock stainless steel Special Plus wire
values were similar in tensile and bend tests because of the stress-relief
heat treatment. Stress-relief heat treatment is discussed later in Section
3(2). Unisil was in the cold-worked cond¡t¡on and therefore contained
many internal stresses which are unpredictable.
lf tensile testing is to be used, it would be preferable to load and unload

the same specimen prior to fracture several times, as the initial load
stress-strain curve is unreliable due to kinks in the wire. The initial slopes
of any load (excluding the first curve) and unload stress-strain curves
should be the same and may be used to mathematically calculate the
elastic modulus. Sokel 1984 cyclically loaded and unloaded specimens to
measure the elastic modulus of wires but elastic modulus values with
large variations were still produced.
74
ln the present study, some preliminary tests were performed, but due to
major problems with extensometers and computer software, the results
obtained were unreliable and not reproducible. lt was decided therefore
to concentrate on other testing techniques which more closely followed
clinical orthodontic application or corresponded to other accepted dental
material standards.
B. Bend tests
Wires were ¡n¡tially tested in archform to assess them in a similar

configuration to that used in clinical practice and provide comparitive
data to results from the more conventional test methods. Due to the
complexity of calculating the elastic modulus for this geometry, elastic
modulus values could not be calculated and cantilever and three-point
bend tests were used in preference in the later sections. sections 1-5
will refer to tests on stainless steel wires while Section 6 will be "other
wires".
1. Preliminary experiments in arch form.

(1 ) as-received sPecimens
Bend tests were performed on wires, as-received in archform. 45o anchor

bends were placed unless otherwise stated. The stress-strain curves for
each test are in the Appendix: graphs 1 - 27. Straight lines of best fit
were computer chosen for the data plotted. Equations for these lines are
included in the form y = c I mx (m : slopel.
Many of the stress-strain curves were curved despite the specimen being
tested within the elastic limit. Several different tangents could be
selected as discussed by sokel 1984 and illustrated in Figure 21. As a
consequence the slopes and equations for the lines of best fit would
vary.
Elastic modulus values were not calculated due to the complexity of

mathematics required but the wires may be compared by the slopes (rn)
of the stress-strain curves which are listed in Table 5.
75
The data presented is in its raw state and the number of decimal places
is not a reflection of the accuracy.
Table 5: Slopes (m) of stress-strain curves for as-received stainless steel

wires in arch form
MATERIAL ANCHOR m m
BEND (LOADt (UNLOAD}
ssoo 01 6 030 .o28

034 030
.031 o29
.051 o44
.o44 .038
ssoo 01 8 055 .049
.050 046
061 .053
194 .1 89
.189 177
018 .194 208
214 231
sssp 020 25" .o87 079
25" .o74 .068
25" .099 .096
35" .041 039
350 038 045
350 038 038
183 .177
232 .229
55" .349 364
55" .194 .1 89
sso 020 .069 .o63

.061 .055
065 059
108 .093
.069 .o64
The slopes from unload data were generally 5% lower than those for load
data. This is evident from the lower m value. This was surprising given
that the wire was not plastically deformed and may be as a consequence
of the experimental set-up, in that frictional forces between the wire and
support slot may differ on loading and unloading.
The range of values for some wire types (sspp O16, sspp ps O18 and ssp
O2O) were narrow. However for sssp O2O and sspp O18 wires large
variations were noted. Values ranged from 0.038 to 0.349 (although the
bend angles may have some, probably small, effect) and O.O5O to O.189
76
respect¡vely, with factors of more than 9 and almost 4 difference which
could not be explained.
(2) after electropolishing

An experienced clinician commented that wire stiffness appeared to
change after electropolishing (personal communication with M.R.Sims).
Electropolishing removes the outer surface of the wire which may affect
the surface appearance and diameter of the wire.
Sokel, 1984 ment¡oned the possibility of an oxygen embrittled surface

layer on near-alpha titanium wires.
Wilcock wire is deliberately oxidised by heat treatment during

manufacture. On occasions oxygen is forced through the oven to ensure
adequate oxidation. The oxidation cleans the wire and leaves it with a
roughened surface which ensures that lubricant adheres to the wire. The
lubricant protects the dies as the wire passes through during the drawing
process (Masson 1969). A similar oxygen embrittled surface may be
present on stainless steel wires. Removal, such as by electropolishing'
may alter the wires ProPerties.
Orthodontic wires are heavily cold worked and when magnified the wire
surface shows very elongated grains giving a fibrous appearance' The
wire may appear homogeneous but the surface is theoretically more
severely cold worked than the central area during wire drawing (Dieter
1961 as stated in Twelftree 1974). This heavily cold-worked surface area
may represent the tag often noted after wire fracture. The loss of this
heavily cold-worked surface layer may result in altered wire properties.
Analysis of wires in cross section may reveal differences in the surface
and central areas of the wire.
(¡) Surface changes were assessed in the PhiliPs 505 SEM.

Photomicrographs are included later in the section F on "surface
appearance".
(iil Diameter changes with electropolishing

To catculate the elastic modulus of a wire, the diameter is required. Some
formulae, such as that used for the calculation of elastic modulus from
resonance tests, were shown to be extremely sensitive to diameter
77
changes. An average of three values measuring wire diameter were used
when calculating the elastic modulus.
A Mitutoyo micrometer with ratchet was used to measure the wire

diameters. This removes potential operator error induced by pressure
variation used to t¡ghten the jaws of the micrometer. An alternative
method may be measurement from sEM photomicrographs.
The results for changes in wire diameter are given in Table 6.
Table 6: Diameter change of stainless steel wires after continuous

electropolishing (eP).
SPECIMEN EP NEAR MIDWIRE FURTHEST MEAN

(secs) ATTACHMENT (mm) FROM OF
(mm) ATTACHMENT MIDWIRE/

(mm) FURTHEST
(mm)
1 25 0.50 0.s0 0.50 o.50
2 50 0.45 o.47 o.48 0.48 t .Ol
3 75 0.48* 0.45 o.47 0.46 + .Ol

4 100 0.45 o.42 0.45 O.44 ¡ .O2
5 125 0.33 0.40 o.44 O.42 + .O2
6 150 o.20 0.40 o.44 O.42 + .O2
* may be explained by variation in the start diameter of this particular

wrre
Closest to the attachment preferential polishing appears to occur. When

catculating the mean an average of the diameters midwire and furthest
were used. A maximum error of .O2mm was noted.
The data listed in table form above is also shown below graphicallY (Fig.
22t.
78
Figure 22: Diameter change with electropolishing sssp O2O'
DIAMETER CHANGE WITH ELECTROPOLISHING SSSP O2O
o.5
o.45
o.4
o.35
DIAMETER (mM)
o.3
o.25
o.2
o.15
o 25 50 75 100 125 150
CONTINUOUS ELECTROPOLISH¡NG TIMI

(secs)
near attachment A mid wire X furthest from +mean

attachment
Minimal w¡re diameter changes are noted after twenty-five seconds

electropolishing. After lengthy periods of electropolishing the diameter
red uces significantlY.
It was noted that diameter changes were not even along the length of
the wire. The effect on diameter was reduced with increasing distance
from the clip although the diameter immediately under the po¡nt of
attachment was minimal. Variation in diameter increased as the
electropolishing t¡me lncreased. lt is noted that the electropolishing times
are much longer than that used clinically. Lengthy electropolishing times
were deliberately chosen to highlight any potential changes.
When calculating the elastic modulus, variation in wire diameter after

electropolishing and in the as-received state needs to be considered
although var¡ation of the tatter is probably not as marked. The results
above suggest a range (mean with a standard deviation) of elastic
modulus values may therefore be more appropriate than a single value.
The results for changes in wire diameter due to interrupted

electropolishing are given in Table 7.
79
Table 7: Diameter change of sta¡nless steel wire after interrupted
electropolishing (ep).
SPEC¡MEN EP NEAR MIDWIRE FURTHEST MEAN
(secs) ATTACHMENT (mm) FROM OF
(mm) ATTACHMENT MIDWIRE/

(mm) FURTHEST
(mm)
1 25 0.48 0.49 0.49 o.49
2 25+25 0.45 o.47 o.47 o.47
3 25+25+ 0.43 o.47 o.47 o.47

25
4 25+25+ 0.39 0.46 0.46 o.46
25+25
Mean diameter values were calculated again from midwire and po¡nts
furthest from the attachment diameters. Data in table form was then
plotted graphicallY (Fig. 231.
Figure 23: Diameter change with electropolishing sssp O2O.
DIAMETER CHANGE WITH ELECTROPOLISHING SSSP O2O
o.5
A
o.45
E
DIAMETER (mm)
o.4
o.35
o 25 25+25 25+25+25 25+25
+25+25
CUMULATIVE ELECTROPOLISHING TIME (secs)
E near attachment A mid wire X furthest from +mean

attachment
80
Bend tests were performed on wires after electropolishing to determine
whether the elastic modulus values were altered. Wires in archform with
45o anchor bends, unless otherwise stated, were tested after varying
times of electropolishing. The data was plotted and from the stress-strain
curves a line of best fit was chosen. The equation for the straight line y
: C * mx where m : slope allows the data to be compared. The slopes
are listed in Table 8. The slopes are listed for each of the stress-strain
curves which are included in the Appendix numbers 28 - 46. Slopes for
load graphs were cons¡stently h¡gher than for unload graphs.
lnstabi¡it¡es were often noted in these stress-strain curves for load and
unload data making selection of the l¡ne of best fit difficult. This was also
noted in stress-strain curves in both the as-received and electropolished
condition.
Table 8: Slopes (m) of stress-strain curves for stainless steel wires tested
in arch form after electropolishing (ep).
Anchor EP m m
Bend secs (LOAD) (UNLOAD)
sspp 016 30 030 o27
30 .053 046
30 .030 o26
45 .o27 o25
45 o23 o22
o18 30 .o42 o38
30 .051 o47
45 067 058
sssp 02O 35" 25+50 040 .038
350 25+50 040 .037
350 25+50 o26 o24
25" 50+50 .057 .054
25" 50+50 .053 .051
25" 50+50 089 085
sso O2O 30 .073 .066
30 o62 058
30 080 .o71
45 060 055
45 .066 o60
81
Although these tests provided comparative data, without actual elastic
modulus values few conclusions can be drawn.
ln the literature many of the bend tests are conventional cantilever tests.
Cantilever bend tests, using short lengths of wire, mimics the clinical
situation where the archwire is more comparable to a series of short
beams between brackets. ln addition simple mathematical formula are
available to calculate the elastic modulus value. For these reasons it was
decided not to pursue archform bend tests but rather to perform
cant¡lever bend tests.
2. Cantilever tests
(1) as-received
From the experimental data stress-strain curves were plotted and a line of
best fit was computer chosen to calculate the slope. The curves and lines
of best fit included in the Appendix were those initially drawn. As stated
earlier, the initial portion of the stress-strain curves were often curved.
fgnoring ze(o, the straightest portion of the stress-strain curve was
identified and the appropriate data replotted and a line of best fit chosen.
By modifying the graphs in this manner the slopes and elastic modulus
values increased. This was repeated for all stress-strain graphs where
appropriate. The elastic modulus values were then calculated using the
beam theory methodology described in the materials and methods
sect¡on.
lnitial Stress-strain curves are included in the Appendix section

(i) graPhs 47 - 50
(ii) graphs 51 - 62'
The results giving the elastic modulus values calculated in each test from
the loading and unloading curves are given in Table 9 and 1O.
82
(il Table 9: Elastic modulus values (E) for as-received stainless steel wires
calculated from cant¡lever bend tests with the support at 18.Smm.
WIRE TYPE E E
(GPal (LOAD) (GPa) UNLOAD)
ss 016 80.48 61.89

018 97.15 97.70
18 148.26 1 19.81
sssp 020 62.18 45.41
The elastic modutus values exhibit a very large scatter and are much
lower than the published data which suggest values of 193 to 2OO GPa'
These exper¡mental values were therefore dismissed. lt was thought that
the slot support may result in friction between the wire and support,
sufficient to invalidate the results. A knife edge was used as an
alternative supPort.
(ii) Table 1O: Elastic modulus values (E) for as-received stainless steel
wires calculated from cantilever bend tests with the support at 20mm'
WIRE TYPE E E
(GPa) (LOADI (GPa) (UNLOAD)
016 201.89 171.18
ssoo 01 8 196.24 172.65
sspo ps 01 B 193.01 169.49

190.24 168.89
186.02 167.1
sssp 020 't28.75 113.28

1 19.46 101.16
122.70 100.39
149.58 106.96
+ *
155.81 129.39
127.82* gg.06*
+
151.75 104.17*
* curvature resulting from wire being spooled removed prior to testing
83
lgnoring differences in wire types (ie premium plus and special plus) the
elastic modulus values decreased as wire diameter increased from O16 to
O18 to O2O. As discussed in section A on tensile test¡ng Masson 1969
and Twelftree 1974 stated that elastic modulus varied depending on the
grade of stainless steel wire. The values for the 016 and O18 wires are
closer, but still low, when compared to the published values for elastic
modulus. Elastic modulus values for the 020 wires were much lower'
Elastic modulus values for the O16 and 018 wire groups were reasonably
consistent with mean values of 2O2GPa and 191 t 4GPa respectively.
Values for the O2O group were scattered and inconsistent and there
appeared to be two groups with mean values o'f 126 + SGPa and 153
+
15GPa. Two groups of O2O were tested where in one group the
curvature resulting from the wire being spooled was removed prior to
testing but this did not explain the d¡fferences as the two groups cut
across this category. Another explanation may be the presence of large
variation along the wire but th¡s would seem unlikely.
lnitial elastic modulus values showed that pulse straightening appeared to

have little effect.
(21 Effect of electroPolishing
As previously mentioned a clinician suggested that electropolishing may

affect the elastic modulus value. Additional tests to compare elastic
modulus of as-received wires with electropolished wires were performed.
lnitial stress-strain curves for the bold data are included in the Appendix
(i) graphs 63 - 68
(ii) graphs 69 - 84.
The results are presented in Tables 11 and 12.
(i) Slot support at 18.5mm
selected wire samples were assessed after electropolishing using a

cant¡lever bend test w¡th a slot support at 18.5mm and the results are
given in Table 1 1. Values for the wires tested in the as-received state (in
non-bold type) are included for comparison.
84
Table 11: Elastic modulus values (E) for electropolished ("p) sta¡nless
steel wires calculated from cantilever bend tests w¡th the support at
18.5mm.
WIRE TYPE EP E E
(secs) (GPa) (LOAD) (GPa) (UNLOAD)
016 80.48 61.89

30 76.12 58.O0
45 a2.92 71.61
018 97.15 97.70
30 68.19 59.87
45 60.66 45.63
ssso 020 62.18 45.41
25+50 23.53 11.43
50+50 92.51 79.65
Again the figures were very low and of limited value'
(ii) Knife-edge support at 2omm
Cantilever bend tests were performed on wires after electropolishing

using a knife edge support at 2Omm. These elastic modulus values are
¡isted in Table 12. As-received values are included again (in non-bold
type) for comparison.
85
Table 12: Elastic modulus values (E) for electropolished ("p) sta¡nless
steel wires calculated from cantilever bend tests with the support at
2Omm.
WIRE TYPE EP E E
(secs) (GPa) (LOAD) (GPa) (UNLOAD)

016 201.89 "171.18
30 220.59 189.OO
45 212.46 184.68
018 196.24 172.65

30 193.58 167.13
45 179.34 155.69
018 193.01 169.49

190.24 168.89
186.02 167.15
25 17A.73 160.62
152.34 136.27
25+25 147.95 129.41
50 165.25 150.62
25 + 25 +25 167.55 157.52
75 1A2.92 169.20
25+25+25 152.20 136.18
+25
100 152.55 141.O2
sssp o20 128.75 113.28

119.46 101 .16
122.70 100.39
149.58 106.96
*
155.81 * 129.39
127.82
+
gg.06*
+ +
151.75 104.17
25 160.19 127.50
50 162.24 136.33
25+50 176.39 141.33
50+50 176.57 145.37
*curvature resulting from wire being spooled removed prior to testing
86
No particular pattern could be given to the effect of electropolishing on
the elastic modulus values. The elastic modulus values for the
electropolished o18 wires appeared reduced compared with the as-
received specimens but not proportionately to the electropolishing time.
This observation may be explained by the high surface stresses being
removed with electropolishing. In contrast elastic modulus values
increased after electropolishing when compared with the as-received
values for both 016 and o2o wires so the theory of removing the high
stress surface does not hold making the effect of electropolishing on
elastic modulus values confusing. However it is likely that errors in
measurement are much greater than any effect if it exists.
3. Other techniques
To compare the effect of changing experimental parameters and of
different methods of testing on elastic modulus values samples were
tested on a knife edge support at lOmm or using a three point bend test.
lnitial stress-strain curves for bold values are included in the Appendix
(1)(i) graPhs 85 - 86
(ii) graphs 87 - 94
(2)(i) graPhs 95 - 100
(11 Cantilever tests on a knife edge support at lOmm
It would seem appropriate to determine the mechanical properties of

orthodontic wires using very short lengths of wire in a cantilever bending
test as this mimics the clinical situation (Brantley and Myers 1979). lt has
been shown that varying lengths of the same mater¡al between half, one
and two inch specimens, gave elastic modulus values which differed by
more than lOOo/o (Brantley 1978 as stated in lngram et al 1986)'
Reducing the distance of the support to lOmm would have the effect of
reducing specimen length.
87
il as-received
These results are given in Table 13. Results for the same wire type
tested on a cantilever support but at 2Omm are included (in non-bold
type) for comparison.
Table 13: Elastic modulus values (E) for as-received sta¡nless steel wires
from cantilever tests with the support at 1Omm.
WIRE TYPE E E
(GPa) (LOAD) (GPa) (UNLOADI
sspp ps 01 B 172.82 124.51

17A.78 107.38
20mm 193.01 169.49

190.24 168.89
186.02 167.15
Reducing the distance of the knife-edge support from 20mm to l Omm

reduced the elastic modulus values'
iil electropolished
The modulus values obtained after electropolishing are given in Table 14.
Again, results of electropolished wires tested on a knife edged support at
2Omm are also included (in nin-bold type) for comparison.
88
Table 14: Elastic modulus values (E) for electropolished ("p) stainless
steel wires from cantilever tests with the support at lOmm'
WIRE TYPE EP E E
(secs) (GPa) (LOAD} (GPa)

(UNLOAD)
018 172.82 124.51

178.78 107.38
25 140.4 95.96
157.85 101.29
20mm 178.73 160.62

152.34 136.27
25+25 127.38 98.28

20mm 147.95 129.41
50 158.68 1 15,57
2Omm 165.25 150.62
25+25+25 130.73 95.42

20mm 167.55 157.52
75 170.17 130.40
20mm 182.92 169.20
25+25+25+25 't23.15 92.57

20mm 152.20 1 36.1 8
100 116.23 94.97
20mm 152.55 141.O2
The elastic modulus values calculated from tests at l Omm are lower than
the values calculated with the support at 20mm with the possible
exception of one spec¡men electropolished 25 seconds. The elastic
modulus values calculated at 2Omm were already less than the publ¡shed
89
elastic modulus values. Varying the position of the support appears to
have an effect on the elastic modulus value. This is in agreement with
Brantley 1978 as stated in lngram et al 1986.
(2) Three-point bend tests
Twenty-five millimetre lengths of wire were supported by two knife

edges placed 2Omm apart and then loaded centrally. These results are
given in Table 15.
Table 15: Elastic modulus values for stainless steel wires after
electropolishing (ep), calculated from three-point bend tests.
E E
WIRE TYPE EP
(secsl (GPa) (LOAD) (GPa) (UNLOAD)
sspp ps 01 I o 147.97 95.59
o 149.62 96.21
25 136.85 76.65
25+25 149.96 85.98
25+25+25 159.65 113.80
25+25+25+2b 153.67 153.66
All values are much lower than those quoted in the published data.
Elastic modulus values after electropolishing were above and below those
of wires tested in the as-received state. This is not surprising given the
difficulty of achieving a consistent result even when retesting a wire'
As previously reported Goldberg et al 1977 showed a substantial

decrease in the value of modulus in stainless steel with heavy cold
working. As stainless steel orthodontic wires are progressively drawn to
smaller diameters they ate increasingly cold worked. Goldberg et al
1983b mentioned the anisotropic tendency of heavily drawn wires.
Forsyth and Stubbington (1975) Sokel 1984, have also drawn attention
to the importance of the crystal orientation to the value of the modulus
with respect to the titanium alloys.
90
It was thought that the reduced elastic modulus values noted in both
tensile and cantilever tests may be associated with the heavy cold-
working procedures. Cold-worked metals can be converted back to an
equiaxed Structure by heating the structure above the recrystallization
temperature. This is referred to as annealing the wire'
Masson (1969) noted that Wilcock wires are heat treated after each
drawing but the actual temperature of treatment is not available. All that
Wilcock revealed was that the process is " more than a stress-relieving
process" although the heat treatment temperature after the final draw
could not be very high since the cold worked properties need to be
maintained. A stress-relief process at low temperature only releases some
of the residual strain induced by plastic deformation. There does not
seem to be a consensus on the optimal temperature and length of time
required for heat treatment (Funk, 1951; Backofen and Gales, 1951;
Kemler, 1956; Mutchler, 1961; Howe et al., 1968 recommend different
temperatures and times). The approximate temperature lies between 371
and 484"C which is below the recrystallisation temperature'
4. Testing of annealed specimens
Specimens of orthodontic wire were annealed to minimise the anisotropic

behaviour prior to cant¡lever bend tests. The aim of the annealing was to
achieve elastic modulus values closer to the published data. Goldberg et
al 1g77 had noted that the elastic modulus value of cold-drawn sta¡nless
steel wire was considerably less than that normally quoted but when
recrystallized the wire possessed a modulus closer to textbook values
(being 177 to 2O3 GPa) as stated by Braden et al 1979 in a review of
dental materials.
A knife edge support was used at a distance of 2Omm
lnitial stress-strain curves for bold values are included in the Appendix
(1) graPhs 1O1 - 1O9
(2) graphs 110 - 125
Less data was recorded and therefore plotted for the unload curves for
annealed specimens due to the reduced springback'
9l
(1) annealed
The modulus values for annealed samples are given in Table 16' Where
available the results from wires tested in the as-received state are
included (in non-bold type) for comparison.
Table 16: Elastic modulus values (E) calculated for annealed stainless
steel wires from bend tests with a support at 2Omm'
WIRE TYPE E E
(GPa) (LOAD) (GPa) (UNLOAD)

sss 01O 205.33 95.56
012 150.90 81.89
014 186.88 31.59
o1 172.60 170.63
as-received 201.89 171.18
018 121.46 119.73
ssso O2O 168.59 46.66
166.51 67.O2
140.78 70.67
166.72 2a
119.46 101 .16
122.70 100.39
149.58 106.96
155.81
- 129.39-
127.82- 98.06*
151 .75- 104.17-
*curvature resulting from wire being spooled removed pr¡or to testing
Elastic modulus values were calculated from the load data. The sss O1O
specimen elastic modulus value was close to the accepted value of
between 193 to 2OO GPa.
Elastic modulus values for as-received and annealed specimens of sspp

016, sspp 018 and sssp O2O tested on a knife edge support at 2Omm
were compared. Values for annealed sspp 016 and O18 wires were lower
than when tested in the as-received state. sssp 02O elastic modulus
values were higher when compared to as-received values. This result was
unexpected. O2O special plus wires are drawn less than 018 or 016 wires
and are therefore less cold-worked. lt would be anticipated that annealing
92
would have a greater impact on the elastic modulus value of the more
cold-worked wires but the reverse was true.
The elastic modulus values calculated from the unload data were all
extremely low and less than values calculated from the load data. This
may be accounted for by the reduced springback of annealed wires.
(2) Annealed and electroPolished
The effect of electropolishing on annealed samples was also assessed

and results given in Table 17.
Table 17 Elastic modulus values (E) calculated for annealed and

electropolished (ep) stainless steel wires from bend tests with a support
at 20mm.
WIRE TYPE EP
E E
(secs)
(GPa) (LOAD) (GPal (UNLOAD)
ssso 020 annealed / not ep 168.59 46.66
166.51 67.02
140.78 70.67
166.72 28.85
25" 144.O8 54.94

158.82 74.13
153.14 9.O2
116.15r 19.O3r
not annealed I eP25 1 60.1 I 127.50
25" +25" 188.62 29.34

143.08 61.60
50" 170.15 81.37
138.22¡ 3.63r
not annealed / eP50 162.28 136.33
25" +25" +25" 146.72 129.57

145.62 55.36
75" 128.08 126.73
123.4Or 121.97t
25" +25" +25" +25" 157.16 10.69

147.65 74.66
1 00" 150.35 153.30
158.14r 44.35t
r denotes retest of wire immediately above
93
There is insufficient data to determine an effect of interrupting the
electropolishing versus continuous electropolishing. Given that no
differences were noted when diameter changes were compared after
continuous versus ¡nterrupted electropolishing, ¡t is unlikely to be
significant. For these reasons it was decided that results of wires
electropolished for the same total time could be grouped together to
provide the average values given in Table 18.
Table 18: Mean elastic modulus values (E) calculated for annealed and
electropolished (ep) sta¡nless steel specimens'
WIRE TYPE TOTAL EP TIME E mean

(secsl (GPa) (LOAD)
annealed sssp 020 0 161
25 143
50 160
75 135
100 153
There is no systematic change in Young's modulus with increasing

electropolishing time.
A comparison can be made for as-received wires after ep25" and 50"
(16OGPa and 162GPa respectively). Again, annealing of the wires did not
have the anticipated effect of increasing the elastic modulus values.
It was noted that the elastic modulus values for the retested wires were
lower excluding the wire electropolished 1OO seconds'
5. Other wires
These wire specimens were all tested in the as-received state. Wires
have been grouped together according to the tests performed. lnitial
stress-strain curves are included in the Appendix
(1) graPhs 126 - 127-
(2) graphs 128 - 132
(3) graphs 133 - 138
(4) graphs 139 - 141
94
For preformed nickel-titanium archwires, the straightest section of the
archwires were used.
(11 Cantilever bend tests using a slot support at 18.Smm
These results are given in Table 19.
Table 19: Elastic modulus values (E) for alpha-titanium (at) and nickel-
titanium (nt) wires calculated from cantilever bend tests with the support
at 18.5mm.
WIRE TYPE E E
(GPa) (LOAD} (GPa) (UNLOADI

* 19.11
at 018 xO25 38.40
nt 016 23.75 18.91
*measurement given as height x breadth
Very low elastic modulus results were noted previously when' a slot
support was used, and again l¡ttle reliability could be placed in these
results.
(21 Cantilever bend tests with a knife edge support at 2Omm
These test results are given in Table 20.
Table Elastic modulus values (E) for alpha-titanium (at), nickel-

20
titanium (ntl and superelastic nickel-titanium (se nt) wires calculated from
bend tests with the support at 2Omm.
WIRE TYPE E E
(GPa) (LOAD} (GPa) (UNLOAD}

+
at 018 x 025 54.49 41.88
nt 016 52.52 45.50
se nt 016 35.00 23.82
sent018x028 15.88 30.16

se nt 022 x 025 22.56 19.61
*measurement given
as height x breadth
95
(3) Cantilever bend test us¡ng a knife edge support at lOmm
These test results are given in Table 21
Table 21: Elastic modulus values (E) for nickel-titanium (nt) wires
calculated from bend tests with the support at lOmm'
WIRE TYPE E E
(GPa) (LOAD) (GPa) (UNLOAD}
nt 014 28.17 21.49

30.50r 25.76¡
nt 016 57.11 38.95

52.84r 41.83r
nt 018 66.50 44.98

70.89r 55
r denotes retest of wire immediately above
(41 Gantilever bend test using a knife edge support at 2Omm
These results for the annealed specimens are given in f able 22'
Table 22= Elastic modulus values (E) for alpha-titanium (at), beta-titanium
(bt) and nickel-titanium (nt) wires calculated from bend tests with the
support at 2Omm.
WIRE TYPE E E
(GPa) (LOADI lGPa) (UNLOAD)
at018x025 58.63 41.81
bt 017 x 025 263.1 I 102.44
nt 016 74.48 61.09
Data calculated from tests (2), (31 and (4) will be discussed according to
wire type.
96
1. alpha-titanium
Two specimen of 018 x O25 were tested resulting in elastic modulus
values of 55 and 59GPa. These values are lower than the range 78 - 99
GPa as quoted by sokel 1984.
2. beta-titanium
One beta-titanium O17 x O25 wire was tested. The elastic modulus value
of 263Gpa was very high. A wide range of elastic modulus values are
quoted for beta-titanium wires ranging from 55 - 11OGPa (Goldberg and
Burstone 1979) and 65GPa (Burstone 198O).
3. nickel-titanium
preliminary tests using a cant¡lever bend test with a knife edge support at
1Omm, indicated that the elastic modulus value increased from 28GPa to
67GPa as the wire diameter was increased from 014 to 018.
The cantilever test was repeated for nt 016 using a knife edge support at
2Omm. The elastic modulus value was 53GPa which was lower than the
57Gpa recorded at 1Omm. This is in contrastto trend noted in section 3.
(1) where the elastic modulus values calculated with the support at
2Omm were higher than those calculated with the support at lOmm'
Burstone and Goldberg 1980 quoted an elastic modulus value of 33GPa

for nickel-titanium wires.
No elastic modulus values are quoted in the literature for the superelastic
nickel-titanium wires. lnitial tests showed a range of 16 - 35GPa for the
wires tested. When discussing elastic modulus values for these wires
with one company it was stated that an actual value is not quoted but
the shape of the stress-stra¡n curve is what is relied on for quality
control.
There was considerable scatter of the data. only limited data was
plotted, to enable the straightest section of the stress-strain curve to be
drawn. Despite this, a large range of elastic modulus values were
calculated.
97
Mechanical tests still appear unsatisfactory so alternative physical
methods of assessing the elastic modulus were investigated. These
included resonance and the velocity of sound.
G. Resonance
A wire specimen with a 1 gramme plastic weight attached to one end

was manually displaced to oscillate the wire specimen. The frequency of
a strobe light was adjusted until the wire specimen appeared stationary'
A frequency was recorded in rpm. The elastic modulus value was
calculated using the formula described in the materials and method
section.
1. Preliminary
(1) Wire size
Several different diameter stainless steel wires were tested. The results
are listed in Table 23 according to wire diameter.
Table 23: Elastic modulus values (E) for as-received stainless steel wires
calculated from preliminary resonance tests.
WIRE DIAMETER FREO FREO LENGTH MASS I VALUE E VALUE
(metres) (rpm) {Hz) (metres) (ko) (GPal

TYPE
sss 010 0.000252 262 4.366667 0.05 0.001 1 .98E-16 158
122 2.033333 0.1 0.001 3.98E-16 't37

sspp 0.0003
o'12
sspp 0.000358 180 3 0.1 0.001 8.O6E-16 147
014
sspp ps 0.000412 245 4.O83333 0.1 0.001 1.41E-15 155
016
sspp o.000412 260 4.333333 0.1 0.001 1.41E-l5 175
016
sspp ps 0.00045 285 4.75 0.1 0.001 2.01E-l5 148
018
sspp 0.0004s2 300 5 0.1 0.001 2.05E-15 161
018
sssp 0.000509 350 s.833333 0.1 0.001 3.29E-15 136
o20
98
The elastic modulus values for the sta¡nless steel premium plus wire
ranging from 012 to O18 were inconsistent. The sspp pulse straightened
016 and 018 wires elastic modulus values were lower when compared
with the same dimension non-pulse straightened stainless steel'
(21 Electropolishing effects
The sspp 016 and sspp O2O wires were electropolished for either twenty-
five or one hundred seconds prior to testing. These results are given in
Table 24. Results for as-received wires are also included (in non-bold
type) for comParison.
steel wires calculated from preliminary resonance tests.
W¡RE DIAMETER FREO FREO LENGTH MASS I VALUE E VALUE
(metres) (roml (Hz) (metres) (ks) (GPa)

TYPE
016 0.000412 260 4.333333 0.1 0.001 1.41E-15 175
sspp 016 0.000404 250 4.166667 0.1 0.001 1.31E-15 175
ep25"
sspp 01 6 0.000395 235 3.916667 0.1 0.001 1.19E-15 169
eol 00"
sssp 02O o.000509 350 5.833333 0.1 o.001 3.29E-15 136
sssp 020 o.o00506 345 5.75 0.1 o.001 3.22F-15 135
ep25"
sssp 020 0.000495 325 5.416667 0.1 0.001 2.95E-15 131
ep1 00"
A twenty-five second period of electropolishing did not appear to affect

the elastic modulus. A period of one hundred seconds electropolishing
appeared to slightly reduce the elastic modulus values'
(31 Other wires
A limited range of nickel-titanium and superelastic nickel-titanium wires

were assessed. lt was difficult to oscillate these wires for a sufficient
length of time to allow the strobe frequency to be adjusted until only one
99
image was seen. Although reducing the distance at which the weight
was suspended assisted experimentation, tests were only pursued on
stainless steel wires. The results are given in Table 25'
Table 25: Elastic modulus values (E) for nickel-titanium (nt) and super
elastic nickel-titanium (sent) wires calculated from resonance tests'
(rom) (Hz) (metres) lko) (GPa)

TYPE
nt 016 0 780 13 0.025 0.001 1.23E-15 28.2
sent 01 6 o.oo0407 690 11.5 0.025 0.001 1.35E-15 20.2
sent 020 0.000511 430 7.166666 0.05 0.001 3.35E-15 25.2
Results for nt 016 and nt se 016 were significantly lower when

compared with values for the same wire types from cantilever tests in
section 5. No comparable data was available for nt se O2O.
2. Detailed tests
Three specimens of each wire type were prepared and then tested. The
wires were tested in the as-received state. Six additional specimens of
sspp O1G and sssp O2O were prepared. Half were electropolished twenty
five seconds and the remainder one hundred seconds.
(11Wire size
Results from tests on as-received stainless steel w¡res are listed in Table
26 according to diameter.
Table 26: Elastic modulus values (E) for as-received stainless steel wires
calculated from resonance tests.
100
WIRE DIAMETER FREO FREO LENGTH MASS I VALUE E
TYPE (metres) (rpm) (Hz) (metres) (ks) VALUE

(GPa)
sss 010
wire 1 0.000254 250 4.166667 0.05 o.001 2.048-16 140
wi¡e 2 0.000254 255 4.25 0.05 0.001 2.03E-16 146
wire 3 0.000253 255 4.25 0.05 0.001 2.O1E-16 148
012
wire 1 0.000301 145 2.416667 0.1 0.001 4.O5E-16 190
wi¡e 2 0.000303 150 2.5 0.1 0.001 4.14816 199
wire 3 0.000303 160 2.666667 0.1 o.001 4.12E-16 227
o'14
wire 1 0.000358 180 3 0.1 0.001 8.04E-16 147
wi¡e 2 0.000358 195 3.25 0.1 o,001 8.06E-16 172
wire 3 0.000358 230 3.833333 0.1 0.001 8.06E-16 240
ssoo 01 6
wire 1 0.000411 245 4.083333 0.1 0.001 1.4E-15 156
wire 2 0.000412 250 4.1 66667 0.1 0.001 1.41E-15 162
wire 3 0.000411 250 4.166667 0.1 0.001 1.48-15 163
sspp ps
016
wire 1 0.000411 245 4.083333 0.1 0.001 1.4E-15 156
wi¡e 2 0.000412 240 4 0.1 0.001 1.41E-15 149
wire 3 0.00041 245 4.083333 0.1 0.001 1.39E-15 158
18
wire 1 0.00045 305 5.083333 0.1 0.001 2.01E-15 169
wire 2 0.00045 300 5 0.1 0.001 2.O2E-15 163
wire 3 0.00045 290 4.833333 0.1 0.001 2.01E-l5 153
sspp ps
018
wire 1 0.00045 290 4.833333 0.1 0.001 2.01E-l5 153
wi¡e 2 0.000449 285 4.75 0.1 0.001 1 .99E-15 149
wire 3 0.000448 280 4.666667 0.1 0.001 1.98E-15 145
o20
wire 1 0.000508 385 6.416667 0.1 0.001 3.26E-15 166
wire 2 360 6 0.1 0.001 3.27E-15 145
wire 3 o.000 I 335 5.s83333 0.1 0.001 3.29E-15 125
101
The elastic modulus values for the sspp O12 wires are close to the
published data. All othervalues are lowexcluding sspp 014 wire 3 which
is too high. lt is interesting to note the large range of values for both the
sspp 014 and sspp O2O wires. No particular pattern was observed'
(21 Electropolishing effects

The effects of electropolishing on modulus values are listed in Table 27
according to wire type and electropolishing time.
steel wires calculated from resonance tests.
TYPE (metres) (rom) (Hzl (metres) (kql (GPa)
O16 as-received E values were 1 Pa 162GPa and 163GPa
sspp 01 6
eo25"
wire 1 0.00041 255 4.25 0.1 o.001 1.38E-15 172
wire 2 0.00041 250 4.166667 0.1 o.001 1.39E-15 165
wire 3 0.000409 240 4 o.1 0.oo1 1,38E-15 153
sspp 01 6
ep1 00"
wire 1 0.000404 235 3.916667 0.1 o.001 1.31E-15 154
wire 2 0.000406 255 4.25 0.1 0.001 1.33E-15 179
wire 3 0.00040s 235 3.916667 0.1 0.001 1.32E-15 153
020 as-received E values were 166GPA 14 Pa and 125GPa
sssp 020
ep25"
wire 1 0.000505 345 5.75 0.1 0.001 3.2E-15 136
wire 2 0.000s08 360 6 0.1 0.001 3.28E-15 145
wire 3 0.000508 360 6 0.1 0.o01 3.28E-15 145
sssp 020
ep1 00"
wire 1 0.000s06 355 5.916667 0.1 o.001 3.21E-15 143
wi¡e 2 o.000505 340 5.666667 0.1 0.001 3.2E-15 132
wire 3 0.000505 360 6. 0r1 o.001 3.19E-1 5 148
toz
Despite using an alternative technique it was still proving difficult to
produce a consistent elastic modulus value. A large range of elastic
modulus values were calculated. ln addition the values were still
significantly lower than the textbook figures often quoted' For these
reasons it was decided to pursue a further technique which measures the
veloc¡ty of sound.
D. Velocity of sound
1. Preliminary experiments
A vibration was passed through as-received sssp O2O wire with attached
piezoelectric crystals spaced at either 40,60, 80, lOO or 12Omm' Two
wave forms appear on the oscilloscope screen which represent the
initiating and "received/dampened" vibration. The distance between the
two wave forms was measured by cursors to calculate the time elapsed.
The reference po¡nts used were varied: peak to peak, trough to trough
and the points of first deviation from the horizontal. The diagram of the
wave patterns observed on the oscilloscope screen illustrates the
different reference points used (Fig. 241.
(1) The distance between the peaks was used as a reference. These
results are given in Table 28. lt was noted that it was not easy to
determine accurately the crest of the peak.
Table 28: Elastic modulus values (E) for stainless steel wire calculated
from speed of sound tests using the peaks as reference points.
DISTANCE TIME VELOCITY (vELOClrY)2 E VALUE
(m) (us) (GPa)
0.04 2.528-05 1587.302 251 9526 20.16
0.06 2.788-05 2158.273 4658144 37.27
0.08 2.64E-05 3030.303 91 82736 73.46
o.1 3.08E-05 3246.753 10541407 84.33
o.12 3.52E-Os 3409.091 11 621 901 92.98
103
PEAK
DEVIATION (1)
*
TROUGH
+
(21
(11 wave detected by first piezoelectric crystal

(2) wave detected by second piezoelectric crystal
Figure 24: Dtagrammat¡c representation of wave pattern on oscilloscope

screen after a single vibration passed along the wire length. The different
reference points used to measure the time elapsed are also illustrated.
104
The elastic modulus values were all very low. lt was noted that as the
distance between the piezoelectric crystals is increased the elastic
modulus values increased'
l2l A different reference point (the base of the troughs) was chosen to
see ¡f more accurate values could be calculated. These results are given
in Table 29.
Table 29: Elastic modulus values (E) for stainless steel wire calculated
from speed of sound tests using the troughs as reference points'

(GPa)
{us)
0.04 0 5 2666.667 7111111 s6.89
0.06 1 3125 9765625 78.13
0.08 2.258-05 3555.556 12641975 101.14
0.1 2.64E-05 3787.879 14348026 114.78
o.12 2.948-05 4081.633 1 6659725 133.28
The elastic modulus values were higher than when the peaks of the wave
form were used as a reference, but still lower than the published data.
Again it is noted that the values for elastic modulus increased as the
distance between the piezoelectric crystals increased.
(31 An alternative reference point was tried, inwhich it was attempted to

determine the first deviation from linearity in the crystal output. The
results of the point of first deviation from the horizontal as a reference
are given in Table 30.
105
Tabte 3O: Elastic modulus values (E) for stainless steel wire calculated
from speed of sound tests using the point of deviation from the
horizontal as the reference point.
(ps) (GPa)
(ml
0.04 0.000009 4444.444 1 9753086 158.02
0.06 0.000014 4285.714 18367347 146,94
0.o8 1.88E-O5 4255.319 1810774',|' 144.86
o.1 2.18E-05 4587.156 21042000 168.34
o.'t2 2.62E-05 4580.153 20977798 167.82
The elastic modulus values were much better but still low. As the elastic
modulus values were most promising using this reference point ¡t
continued to be used.
(4) Annealed
To minimise the anisotropic behaviour of heavily cold drawn orthodontic
wires ¡t was decided to test an annealed section of sssp o2o. The
variation in modulus with distance between the piezoelectric crystals is
given in Table 31.
Table 31: Elastic modulus values (E) for annealed stainless steel wire
calculated from speed of sound tests using the point of deviation from
the horizontal as the reference point.
(m) (ps) (GPa)
0.04 6.4E-06 6250 39062s00 312.5
0.06 1.26E-05 4761.905 22675737 181.41
0.08 0.000016 s000 25000000 200

0.09 1.74E-O5 5172.414 26753864 214.O3
The elastic modulus values were much higher than when tested in the as-
received state and most were close to the published data range. The
value of 2OO is within this range.
106
The high elastic modulus value of 312.5GPa may indicate that the
distance between the piezoelectric crystals is too short.
The results of repeating these tests using a new sect¡on of annealed wire
are given in Table 32.
Table 32: Elastic modulus values (E) for new annealed stainless steel wire
specimen calculated from speed of sound tests using the point of
deviation from the horizontal as the reference point'
TIME VELOCITY (VELOCITY}2 E VALUE

DISTANCE
(rrs) (GPa)
(m)
0.04 0.000006 6666.667 44444444 355.56
0.06 o.000009 6666.667 44444444 355.56
o.08 1.36E-O5 5882.353 34602076 276.82
o.09 1.78E-05 s056.18 25564954 204.52
0.1 1.96E-05 5102.041 26030820 208.25
0.11 2.04E-05 5392.1 57 29075356 232.60
o.12 2.16E-05 5555.s56 308641 98 246.91
Unfortunatety when the tests were repeated the majority of elastic

modulus values were too high. Where the crystals were separated 90 to
lOOmm the values were more in keeping with anticipated results'
Early tests indicated that ¡t was not easyto achieve consistent results.
The use of a square wave form which would give a well defined
reference point may be advantageous.
The potential for variation in wire tension to affect the elastic modulus
value needs to be investigated. lf tension does affect the elastic modulus
values a weight may need to be hung from the wire to eliminate this
variable.
ro7
Summarv of the elastic modulus tests
A comparison of elastic modulus values derived from different techniques

using load data
Table 33 has been compiled to allow easy comparison of elastic modulus

values for different wire types and test methods. Where more than one
wire specimen was tested, mean elastic modulus values with a standard
deviation are quoted . All values are calculated from load data only'
108
Table 33: A comparison of elastic modulus values derived from different techniques using load data.
Wre Type Slot Knife Knife Knife edge 3-point Resonance Speed Speed
Support edge edge support bend of of
18.Smm Support Support (annealed) test GPa sound sound
20mm 1Omm 20mm (annealed)
GPa GPa GPa 0.08
GPa GPa GPa deviatíon
GPa
at018x025 38 55 59
o25 263
nt 014 29 r.2
nt 016 24 53 75 28
55r3
nt 018 69r3
s 35 20
25
sent018x028 16
2 o28 23
sss 010 205 148 r.7
sspp 01 2 151 188r38
sspp 014 187 177 r.58
sspp 0l 6 81 202 173 164r8
EP 25sec 166 r 10
EP 30sec 76 221
EP 45sec 83 213
EP 100sec 164 + 13
sspp ps 01 6 155r4
109
Table 33 continued
Wre Type Slot Knife Knife Knife edge 3-poínt Resonance Speed Speed
Support edge edge support bend of of
18.Smm Support Support (annealed) test GPa sound sound
20mm 1Omm 20mm (annealed)
GPa GPa GPa 0,08
GPa GPa GPa deviation
GPa
ssop 01 I 97 196 122 152 r.7

EP 30sec 68 194
EP 61 179
018 148 190r4 176t4 149 r 1 149r3
EP 179 140 137
+25 148 127 150
+25 168 131 160
+25 152 123 154
EP 25sec 152 158
EP 50sec 165 159
EP 75sec 183 170
EP 100sec 153 116
ssso 020 62 137r15 161 r 13 143 t 17 145 239 r 38
EP 25sec 160 140 r 6
+50 24 176
EP 50sec 162
+50 93 133
EP 25sec 143 r 19
+25 166t23
+25 147 r.1
+25 153r4
EP 50sec 154 r 16
EP 7Ssec 155r4
EP 100sec 154r4 139 r I
110
Effect of diameter change on elastic modulus values
A comment was made earlier about the effect of wire diameter on elastic
modulus values. To assess this, the wire diameter value was increased
by 2o/o and the elastic modulus values were recalculated without
changing any other parameters. sspp ps 018 wire specimens were
chosen and the wire diameter was increased from 0.45 to O.46mm and
the results are as listed in Table 34.
Table 34: The effect of wire diameter on elast¡c modulus values
Test Diameter of 0.¿15mm Diameter of O.46mm

GPa GPa
Cantilever test 1Omm 173 135
Cantilever test 20mm 190 167
Three-point bend test 148 130
Resonance test 153 135
The elastic modulus values decreased significantly when the diameter

was marginally increased. The cant¡lever bend test at lOmm appeared to
be most sensitive to diameter change with a 22o/o reduction noted
compared with a 12o/o change noted with the other test methods. This
emphasises the importance of accurately measuring wire diameter'
The effect of data modification on elastic modulus values
The difficulty in selecting the most appropr¡ate tangent to the stress-

strain curve to calculate elastic modulus has been discussed. The effect
of ignoring zero and limiting the data chosen on the accuracy of the line
of best fit, and consequently the elastic modulus values, were assessed.
Load and unload data for an as-received sspp ps 018 wire tested on a
knife edge support at lOmm was modified as can be seen in Tables 35
and 36.
Figures 25 and 26 illustrate the effects of modifying the data.
111
a
y = O.I ló32 + O.9Z]7OX R^2 = O.991
2
z
a
o
o
o
Er(ension (mm)
b
I=3.4912c-3+ 1.0628x R^2= rl
z
d
o
o
(,
Ertcnsion (mm)
Figure 25: Modification of load data. The first graph shows the slope of
the l¡ne of best fit when all data were ploffed. ln the second graph data is
limited to the first six values.
Lt2
a
2-4{v10rc-2 < O6?485x l{^2 =O-9tì(r
z
z
ó
o
d
ð
o
o
(, l
Extcnsion (mm)
2
b t= -o.rs222 + 0.77212¡ R^2 =o.s(
2
€G
o
cé
o
(, l
Extcnsion (mm)
12
c . 0.8ó951r R^2 = 0.99)
0.245?0
08
z
d
o o.6
a
.J
o.4
o2.
0-o
o4 o6 o_a to r: l4 t6
Ertension (mm)
Figure 26: Modification of untoad data. ln the first graph all data are
plotted. ln the second graph, only eight data po¡nts are plotted and zero
po¡nt is ignored. ln the last graph the slope is calculated from five data
po¡nts only.
Lt3
Table 35: The effect of data modification on the accuracy of the stra¡ght
line fit for a stainless steel premium plus, pulse straightened O'O18" wire'
R^2 (11 R^2 (21 R^2 (3)
LOAD o.991 1.000
UNLOAD 0.986 0.995 o.999
R^2 indicates the accuracy of the straight line fit to the stress-strain data
plotted. For column 1 all data were plotted. When load data was limited
to six points, a perfect straight line fit was achieved. For the unload data,
a nearly perfect fit of 0.999 was noted when five data points were
plotted and zero ignored.
Table 36: Effect of data modification on elastic modulus values for a
stainless steel premium plus, pulse straightened 0.O18" wire.
E E E
GPa (1) GPa tzt GPa (3)
LOAD 172 177
UNLOAD 108 103 103
Despite modifying the data, only relatively small changes in elastic

modulus values of less than 5% were noted.
E. Microhardness and Macrohardness tests

1. Microhardness
After embedding the sspp 018 wire specimen in Bakelite, it was polished
to a mirror finish, exposing the maximum wire diameter for
microhardness testing. This enabled the microhardness of the wire
surface to be compared with the inner core.
The surface layer of the wire is more heavily cold drawn and therefore
harder. Removal of the surface layer, such as with electropolishing, may
make the wire softer. lf the wire was homogeneous surface layer removal
would not be significant.
l14
Both the diagonal distances of the indentation square were measured
from photomicrographs (magnification 142Oxl in millimetres. An average
of these two distances are listed in Table 37 below according to the
loads used. Values are listed as measured, from one surface across to
the other.
Table 37: Average values of diagonal measurements of microhardness

indentations, in a stainless steel premium plus 0.018" wire.
BASE LOAD 25P 50P
mm mm mm
6 7.25 10.25
5.25 6.5 10
5 6.5 10
5.5 7 9.5
5 6.75 9.75
5.25 7 9.25
5 6.75 9.75
5.5 7 9.75
5.5 6.75 9.75

5.5 7 9.5
5.25 7 9.75
5.75 7 9.5
5.25 7.25 10
5.25 7 10
5.5 7 10
5 7 10
5.75 7.75 10.25
5.75
5.75
lf there is a surface difference, due to variation in wire texture (ie a more

heavily cold drawn surface), it must be very small or thin as excluding
the larger values at the edge due to specimen geometry, there were only
minor differences in diagonal dimensions across the wire.
115
The measurements were adjusted to allow for the magnification (1'42 x
163¡ used. An average of all the values (excluding the edges) were used
to calculate the Vickers hardness numbers given in Table 38. This varied
between fifteen and seventeen measurements for each'
Table 38: Vickers hardness numbers for a stainless steel prem¡um plus
0.O18" wire.
LOAD ¿ (pm) HV HV
(mathematically (referenced from
calculated*) table* *)
base load 6.88 588 584
25P 8.75 605 599
50P 12.18 623 623
*Vickers formula as stated in the materials and methods section.

* *tables for the determination of Vickersharte with the LEITZ-MIniload
Hardness-Tester.
As the load was increased, from base load to 25P to 5OP, HV increased.
lf these wires do have a more heavily cold worked surface layer, it would
be expected that as the load was increased, HV would decrease. The
rationale for this being that a lighter load would only penetrate the outer
hard surface while a heavier load would penetrate the softer core. The
value of the latter would be an average of the hard outer surface and the
softer core lowering HV. This was not noted.
2. Macrohardness
It proved difficult to position the diamond indentor on the wire specimen

alone. As a consequence bakelite hardness may be being tested rather
than the sta¡n¡ess steel wire. lnit¡ally a 2Okg load was used but only half
the indentation was located across the wire. A 5kg load was used to
indent the specimen at three locations.
The hardness values calculated were 501,753, 516. The middle value of
753 is so different to the other values and should be rejected as a
consequence. The macrohardness values were lower than the
116
microhardness values. This is to be expected ¡f the macrohardness values
are an average of the hard surface layer and softer core.
As a consequence of the curved surface of wires, a small indentor is

probably a more accurate method of calculating the hardness as it would
allow a more planar wire surface to be examined.
F. Surface appearance
a. Surface effects of pulse straightening.
ln premium plus O18 wire, closely spaced striations (S), parallel to the
long axis of the wire, are evident on the surface (Fig. 271. The elongated
grains give a fibrous appearance. which is typical of heavily cold worked
wires. The low temperature stress relieving heat treatment after the final
draw is below the recrystallization temperature so this does not affect
the fibrous appearance. The surface roughness is possibly a function of
die wear.
Differences in the number of striations, height of the ridges and the

presence of porosity and pits, may help to explain the different
mechanical properties (Slngh 1991).
fn the pulse-straightened condition (Fig. 281 there are no significant

differences in the surface of the wire compared w¡th the untreated
premium plus.
b. Surface changes after electropolishing
The surface structure of as-received 020 special plus wire is shown in

Figure 29. Many evenly spaced striations (S) are present throughout the
longitudinal section. The striations are parallel to the wire axis. lndividual
grain boundaries cannot be distinguished because of the large amount of
distortion during the drawing process.
Non metallic inclusions which may impair the mechanical properties of

wires if present would have been drawn to such an extent that they
cannot be detected (Honeycombe 1968 as in Twelftree 19741.
rt7
-vsC¡ --?-. -j- :=- r-É _ _ <
70pm
Figure 27: Scanning electron micrograph of a

longitudinal section of
stainless steel premium plus o.o18" as-received viewed at a
magn¡fication of 75x showing closely spaced striations (S).
, Fl /. r*'
l:r'
I f.À¡' ,a
\ .-
I
t
I t
:lio'i
)(t
t
fD- 'If
70pm
Figure 28: A scanning electron micrograph of a longitudinal section of as-

received stainless steel premium plus pulse straightened o.o18", at a
magnification of 75x.
118
69pm
Figure 29: A scann¡ng electron m¡crograph of a longitudinal section of as-

received stainless steer special plus 0.020" at a magnification of 75x
showing many evenly spaced striations (S)'
0.[nn26.1 kU 3.øøEe 6608/øE KÊ
Figure 30: A scanning electron micrograph of a longitudinal section of as-

received stainless steel special plus 0.020" wire at a magnification of
3OOx showing the striations (S).
This is the same specimen as Figure 29 but at increased magnification.
119
Singh 1991 noted that premium plus O16 ss wire showed less striations
than special plus 02O ss.
At higher magnifications the striations (S) seem to be evenly spaced

although some (wider) gouges are present (Fig. 3Ol. However in this
material hardly any pits are visible.
After electropolishing for 25 seconds these wires have a very different

appearance to that in the as-received state, the surface striations are
only just visible (Fig.31l. This has the potent¡al to reduce friction.
Studies on the effect of electropolishing on friction would be interesting.
After further electropolishing of 25 seconds, no surface striations are

evident (Fig. 321. The surface appears slightly irregular and gives the
impression of very shallow depressions being present, although an
(unidentified) intermittent larger defect can be seen in the centre of the
wire surface.
A long electropolishing time of 150 seconds was used to exaggerate the

effects of electropolishing. This is not a clinically realistic electropolishing
time. No surface striations are present and the surface appears very
smooth excluding a deep gouge (G) possibly placed at the time of
drawing (F¡g. 331. lf the gouge was placed at the time of drawing the
wire appears to be rotated as it is drawn through the die as the scratch is
not parallel to the edges.
At higher magnification it can be seen that the original surface striations

have been removed leaving a very smooth surface excluding the deep
gouges (G) noted above which were not removed by electropolishing
(F¡9. 341.
c. Fractography
Wire surfaces were assessed in the SEM after being deliberately

maltreated to the point of fracture. Examination of the fractured surfaces
enabled the wire cross-section to be examined to locate variation in
texture between the surface and inner core.
r20
Figure 31: A scanning electron micrograph of a longitudinal section of
stainless steel special plus O.O2O" wire which has been electropolished
tor 25" and viewed at a magnification of 75x'
72¡tm

stainless steel special plus O.O2O" wire which has been electropolished
for 5O" and viewed at a magnification of 75x'
l2l
sta¡nless steel special plus O.02O" which has been electropolished for
15O' and viewed at a magnification of 75x. A deep gouge (G) was
noted.

stainless steel special plus 0.020" which has been electropolished for
150" and viewed a magnification of 300x. A deep gouge (G) was noted.
This is the same specimen as Figure 33 but at increased magnification'
t22
The number of bends required before the point of fracture were recorded
and this data is graphically illustrated (Fig. 35).
Figure 35: Number of bends required before point of fracture.

BEND UNTIL FRACTURE TEST
30
25
20
Emean
hlgh
NUMBER OF BENDS 15
_to*
10
o
l)12 o14 016 016 o18 018 016 016 o20 018 x O215x
.tPP t3PP rsPp ppps ttPp ppps nl ntsc nt!c 026 024
ntsc ntsê
WIRE TYPE
The nickel-titanium wires could w¡thstand many more bends prior to

fracture than stainless steel wires.
After fracture the wires were assessed in the SEM. Photomicrographs

were recorded and ED analysis performed to allow the chemical structure
to be assessed.
The extreme loading condit¡ons applied by reverse bending produced

generally consistent fractures, typified by the formation of "tags" (Fig.
36). The only minor var¡ations noted were in the extent of the "tags"
(Figs. 37 and 38).
At higher magnifications, some differences in fracture behaviour can be

seen. ln Figure 39, the fracture surface appears relatively uniform with
only a gradual change in direction of the crack growth path from almost
normal to the surface (at A) to parallel to the axis (at B) where the fibres
appear to be parallel to wire axis probably due to elongated grains
produced in wire drawing. Distinct steps are also visible on the fracture
surface (at C) which are often clearly associated with similar features at
123
Figure 36: A of a deliberately fractured
scanning electron micrograph
stainless steel premium plus O.O18" wire surface showing "tag"
formation.
t24
Figures g7 and 38: Scanning electron micrographs of the fractured
surfaces of deliberately maltreated stainless steel premium plus O'O18"
wires showing variation in the extent of "tag" formation.
r25
._:_-
Figure 39: A scanning electron micrograph of a fractured wire surface at

higher magnif¡cation. The direction of the crack path appears normal to
the surface at (A) while at (B) it appears parallel to the axis'
BlmmZSBl.r.r t!'JIE:' f.1 Uril ,.1 '' r;

- - -
Figure 40: A scanning electron micrograph of a fractured wire surface.
Distinct steps (C) are visible on the surface'
t26
the wire surface (Fig. a0) and represent ¡ncremental changes in the plane
of the crack. However in another sample (Fig.41), there is a clearly
different zone (at D) normal to the wire surface, or even inclined in the
opposite direction to the "tag". The remaining surface area of the
fracture surface is similar to that described above. A similar fracture can
be seen in Figure 42, but here the two zones are separated by a
pronounced crack (at E). Another photomicrograph (Fig. 43), shows that
these central cracks can grow in a steplike manner to produce the step
features noted in all the "tag" samples.
At even higher magnifications, the fracture surfaces again appeared

largely similar. ln the "tag" areas, fractures were generally associated
with the heavily elongated grains (at F) produced in the manufacturing
process for the wire (Fig. 441. However other areas exhibited regions of
fine flat dimples (at H), typical of ductile fracture {F¡9. a5l. Both these
features are characteristic of ductile failure in drawn wires, for which the
formability has been almost exhausted during manufacture. This type of
fracture is known as fibrous fracture.
The scanning electron micrographs included are typical of many fractured

samples viewed with l¡ttle difference being noted between samples'
However few conclusions can be drawn, other than that the failures are
typical of the material.
Microanalysis of the sample in Figure 42 showed that it contained iron,

chromium and nickel (Fig. 46), as expected with no major impurities
which might influence properties or fracture.
G. Detailed microstructural data
Most authors support the concept that advances in testing and

developing orthodontic wires would be gained ¡f microstructure were
correlated with mechanical properties (Williams and VON Fraunhoffer
1971 as in Twelftree 1974l..
TEM studies would reveal detail of the wire structure by allowing the fine
granular structure to be assessed. Different phases (matrix or
tn
I
r-
P] I = a"tF I r1r'
rn rn - r"-t;'-'
Figure 41: A scanning electron m¡crograph of a fractured wire surface

(D).
illustrating a zone normal, or even inclined in the opposite direction
I
a
/
i
.4
@
?
\-
i-
{
{l
o.1nm2B.6 kU 1.58E2 BBBI J L_ì!ì Þ_-,
-
Figure 42: A scann¡ng electron micrograph -of a fractured wire surface
where a pronounced Crack (E) separating a "tag" and zone normal to the
wire surface, or even inclined away from the "tag"'
128
¿-
\
ãÞ
---T'r-')
f
t\
\-i
- - .1--t
-
l-tà'2
fr
Ê1mm1'f-l'
-)
Figure 43: A scann¡ng electron m¡crograph of a fractured wire surface
showing the steplike growth of a central crack'
ð
t
I E um3E I t' : þi iF ; s:1 Llri ; ¡1 -

- - -
Figure 44 A scann¡ng electron micrograph of a "tag" at higher

magnification showing the elongated grains (F).
129
a
! 1.
.a
\ I
\,
a'-
æ1
:fl
\
\r.
\
t-
),i { I .1+. rl'
Figure 45: A scanning electron micrograph of a fractured wire surface at

high magnif¡cation showing fine flat dimples (H) typical of ductile
f racture.
rl E H I'l =. F Ur'i c,+ F¡¡elEiae Pr,i I rf¡s 5ur= :'1r-¡¡l :<:1-l'ìlìY-E=: I-..
Cut'sol^: Ø.@@ØKeV = Ø RÖI (3:i cl €rqlo: l¿.ØØØ
-' - Fq,qcTi iÊE StlFFFllE ': '-trr TT; :Þ1TTË
t :
il
i!
;l
:!
.'..:-...i
:l
ø taulo VF5 = 4O96 LØ.?4ø

lElA
Figure 46: A printout of the EDS analysis of sta¡nless steel wire showing
the presence of iron, chromium and nickel'
130
prec¡p¡tate), ¡mpurities and defects or heavily deformed regions may be
revealed
The preparation of thin 3mm foils for TEM assessment is not easy due to
the small diameter of the wires. One way of overcoming this was to
prepare the fo¡ls from wires of larger diameter. Although the structure
could be assessed the effects of cold drawing would not be present
(Singh 1991).
Several different techniques were pursued to produce suitable TEM foils

of the wires in cross section as discussed in the materials and methods
section, but unfortunately none were successful.
131
CHAPTER 6
coNcLUsroNs
A knowledge of the mechanical properties of orthodontic wires is

important as it enables the most appropriate archwire for a given clinical
situation to be selected and assists with the development of newer
improved wires. A variety of different properties were tested in the as-
received state and after manipulative treatment.
Stainless steel wires were predominantly tested, as despite their long

history of use, fundamental questions remain about their mechanical
properties.
The stiffness of an archwire is considered to be one of its most important

propert¡es. Stiffness determines the force delivered by an appliance. ln
general a low, continuous force is thought to be most desirable, to
achieve maximum tooth movement with minimum patient discomfort.
The stiffness of a wire is proportional to the elastic modulus. Despite the

use of an extensometer, elastic modulus results from tensile tests by
previous authors have been lower than textbook quoted values. An
alternative testing technique was used to assess the elastic modulus for
this reason and the fact that orthodontists primarily stress wire in
bending not tensile direction.
Bend tests were first performed on wire specimens shaped into an

archform to mimic the clinical situation more closely. Only the slopes of
the stress-strain curves could be compared as actual elastic modulus
values were not calculated due to the complexity of the mathematics.
The range of values for the slopes varied for different diameter wires:
some exhibiting almost constant values while others showed large
variations up to 9 times.
To enable elastic modulus values to be calculated, short lengths of wire

were tested in a cantilever or 3-point bend test. Bend tests using a
slotted support produced very low elastic modulus values which had to
be discarded. The low values may be explained by friction between the
archw¡re and support. A knife edge support was used as an alternative.
r32
Although elastic modulus values were increased they were still low when
compared to the textbook quoted values.
Less conventional test methods of resonance and the speed of sound

were also used to determine the elastic modulus. Despite using these
alternative techniques it was stilt proving difficult to produce a consistent
elastic modulus value. lt appears that determining the mechanical
properties of fine orthodontic wires is not easy.
The influence of wire diameter on elastic modulus values was assessed.

Diameter is critical when calculating elastic modulus for mechanical and
resonance tests where d is used to the fourth power. A 2o/o error in
diameter changed the elastic modulus values between 12 and 22o/". The
speed of sound test method avoids this source of error as wire diameter
is not used to calculate the elastic modulus value.
A comment had been made by an experienced clinician that

electropolishing appeared to affect the stiffness of an archwire. Wires
were tested in the as-received state and after electropolishing. No
particular pattern could be established between electropolishing and
elastic modulus value.
Orthodontic wire surfaces were assessed in the SEM in the as-received

state and after electropolishing. Wires in the as-received state had a
fibrous appearance due to the elongated crystals which is typical of a
heavily cold worked wire. As the wires were electropolished for longer
periods of time, the wire surface appeared smoother and smoother.
Although surface striations were removed, deep gouges probably
produced during the drawing process, remained.
It was proposed that orthodontic wires may exhibit anisotropic

behaviour. Wires were annealed to reduce the anisotropic effect and then
tested in the hope that elastic modulus values closer to textbook quoted
values may be achieved. Elastic modulus values were calculated from the
load data as the unload values were limited and extremely low. This may
be explained by the reduced springback of annealed wires. The effect of
annealing on elastic modulus values was not consistent with values for
t33
O16 and O18 sta¡nless steel wires decreasing while for O2O wires
increasing.
Wires were deliberately maltreated to the point of fracture and then the
fractured surfaces were assessed in the SEM. This was to assess for
textural differences between the wire surface and inner core. The wires
exhibited a ductile fibrous fracture but no significant textural differences
were noted.
Microhardness tests were also performed across the diameter of an

embedded and polished wire. lf there are differences between the surface
layers and inner core they must be small as no variation was noted when
HV was calculated.
Although wires may be selected according to their mechanical properties,

it is the microstructure and in turn the chemical composition and
thermomechanical treatment during manufacture that determines these
properties. TEM studies are the only method of examining in detail the
internal structure. Foil preparation was difficult and time consuming and
despite many different techniques being trialled, no successful foils were
produced. Microstructural assessment would improve our understanding
of mechanical properties and assist in the development of new improved
materials.
134
CHAPTER 7
FUTURE RESEARCH
1. Further tensite tests on orthodontic wires should be performed using

accurate calibrated load cells w¡th extensometers and a large number of
accurate determinations of wire diameter to provide a good average'
2. Most tests reported in the literature use stra¡ght lengths of wire when
determining the elastic modulus. Brantley (1978) (in lngram et al, 1986)
used beam theory technique, to show results for the measurement of
varying lengths of the same material which gave elastic modulus values
that differed by more than l}Oo/o between half, one and two inch
specimens. This should be investigated in more detail.
3. The formula being used to calculate elastic modulus values from

tensile and bend tests may not be appropriate for orthodontic wires due
to their small diameters. This should be investigated.
4. Refining the velocity of sound technique by using a Square wave form'

may assist with accurate identification of the same reference point on the
two wave forms.
5. An article in "Materials Forum" (Bell et al 1993), discussed a new

method of determining hardness and elastic modulus values of thin films
and cross-section using an ultramicroindentat¡on system. This alternative
method of test¡ng the micromechanical properties would provide
additional data for comParison.
6. lf stiffness changes are noted after electropolishing, it may be the

result of precipitate formation (eg hydride formation) or the removal of an
oxygen embrittled or work hardened surface layer. This would need to be
investigated with quantitative and qualitative element identification in the
TEM.
7. The wire surface appeared smoother after electropolishing. This has

the potential effect of reducing friction. Frictional studies should be done.
135
to provide additional
B. Testing should be extended to include other tests
mechanical data such as proport¡onal limit, elastic limit, yield point'
tensile strength, proof stress, ductility, resilience and toughness'
9. Alpha-titanium has been reported to become brittle after being placed

in the oral environment. Such wires should be tested before and after
exposure to an oral environment to assess for changes in the mechanical
properties.
10. TEM on wires to determine the microstructure in cross-section.

Findings from these tests may explain some of the tensile/bend test
findings.
t36
BIBLIOGRAPHY
Andresen,G.F. (1980) A clinical trial of alignment of teeth using a o.o19

inch thermal nitinol wire with a transition temperature range between 31
degrees C and 45 degrees C. Am. J. Orthod . 78 : 528 - 537 '
Andreasen,G.F. and Barret,R.D. (197O) An evaluation of cobalt

substituted nitinol wire in orthodontics' Am. J' Orthod. 63 : 462 - 47O'
Andreasen,G.F. and Brady,P.R. (1972}A use hypothesis for S5-nitinol

wire for orthodontics. Angle Orthod . 42 : 172 - 177.
Andreasen,G.F. and Morrow, R.E. (1978) Laboratory and clinical analysis

of nitinol wire. Am. J. Orthod. 73 : 142 - 152.
Andreaseî,G., Wass,K. and Chan,K'C. (1985) A review of superelastic

and thermodynamic nitinol wire. Ouintessence lnternational 9 : 623 -
626.
Asgharnia,M.K. and Brantley,W.A. (1986) Comparison of bending and

tension tests for orthodontic wires. Am. J, Orthod. 89 : 228 - 236.
Ashby,M. and Jones,D. Engineering Materials An introduction to their

properties and applications. 1st edition (198O) Pergamon Press.
Bee,J.V. (1991) New instruments complete world-class facility at the

University of Adelaide. Lumen. The University of Adelaide News
Magazine. 20 : 1
Begg,P.R. and Kesling,P.C. Begg orthodontic theory and technique. 2nd

ed. (1971) W.B.Saunders Co., Philadelphia, London'
Bell,T.J., Field,J.s. and swain M.V. (1993) Precision hardness and

elastic modulus measurements of thin films and cross-sections. Materials
Forum. 17: 127 - 138.
t3l
Blake,A. Practical Stress Analysis in Engineering Design. (1982). Marcel
Dekker, Inc. New York and Basel.
Braden,M., (Ed) Brown,D. , Combe,E.C., Lloyd,C.H., Main, Miller,M',

Tranter,T.C., von Fraunhofer,A.J., Waters,N.E., Williams,D.F. et al.
Review of dental materials : 1977 Literature Review, Part 1. J Dent., 7 :
283 - 300.
Brantley,W.A. and Myers,C.L. (1979) Measurement of bending

deformation for small diameter orthodontic wires. J. Dent. Res., 58 :
1696 - 1700.
Buckthal,J.E. and Kusy,R.P. (1988) Effects of cold disinfectants on the

mechanical properties and the surface topography of nickel - titanium
arch wires. Am. J. Orthod. Dentofac. Orthop. 94: 117 - 122
Buehler,W.J., and Cross,W.B. (1969) 55 nitinol - Unique wire alloy with

a memory. Wire J. 1 : 41 - 49.
Burstone,C.J. (1981) Variable-modulus orthodontics. Am. J. Orthod. 80 :
1 -16.
Burstone,C.J. and Goldberg,A.J. (1980) Beta titanium : A new

orthodontic alloy. Am. J. Orthod .77 : 121 - 132.
Burstone,C.J., Oin,B. and Morton,J.Y. (1985) Cinese NiTi wire:a new

orthodontic alloy. Am. J. Orthod.87 : 445 452.
Chen,R. (199O) Advanced Chinese NiTi alloy wire and clinical

observations. The Angle orthodontist 62 : 59 - 66.
Coombe,E.C., Bradeñ,M., Sheriff,M., Brown,D., Strang,R., Causton,B.E.,

van-Noort,R., Fletcher,A.M., Waters, N.E., Lloyd,C. H., et al. ( 1 985)
Dental materials: lgS2literature review. Part 1. J. Dent. 13 : 1 - 40.
Cornet,A., Mille,P. and Muster,D. (1982) lmprovement of mechan¡cal

properties of Ti 6 Al 4 V by thermomechanical treatments for
endoprosthesis applications. J. Dent. Assoc. S.Afr. 37 :786.
138
Council on dental materials and devices (1977) American Dental
Association Specification No, 32 for orthodontic wires not containing
precious metals. J. Am. Dent. Assoc.,95: 1169 - 1171.
Dieter,G.E., JI. (1961) Mechanical Metallurgy. McGraw-Hill Book

Company, New York.
Drake,S.R., Wayne,D.M., Powers,J.M. and Asgar,K. (1982) Mechanical

properties of orthodontic wires in tension, bending, and torsion. AJO 82
: 206 - 21O.
Elgiloy, Rocky Mountain Orthodontics Brochure, Denver, Colo. , 1977.
Garner,L.D., Allai,W.W. and Moore,B.K. (1986) A comparison of

frictional forces during simulated canine retraction of a continuous
edgewise archwire. Am. J. Orthod. Dentofac. orthop. 90 : 199 - 2O3
Gatan 1992: Workshop on "TEM specimen preparation of

sem¡conductors, ceramics and metals by ion milling", organised and
sponsored by University of Adelaide, CEMMSA and Thompson Scientific
lnstrument Pty Ltd, Australia, offered by Gatan lnc, USA.
Goldberg,A.J., Burstone,C.J. and Koenig,H.G. (1 983a) Plastic

deformation of orthodontic wires. J. Dent. Res., 62: 1016 - 1O2O
Goldberg,A.J., Morton,J. and Burstone,C.J. (1983b) The flexure modulus

of elasticity of orthodontic wires. J.Dent.Res., 62 : 856 - 858.
Goldberg,A.J., Vanderby,R. and Burstone,C.J. (1977) Reduction in the

modulus of elasticity in orthodontic wires. J. Dent. Res. 56 = 1227 -
1231.
Greener,E.H., Harcourt,J.K. and Lautenschlager,E.P. Materials science in

denrisitry. lst ed. 11972l'The Williams and wilkins co.
Gullotta,A., West,V.C. and Hazel,R.J. (1987) The Bauschinger effect in

stainless steel orthodontic wires. Aust. Ortho. J. 1O: 11O - 113.
139
Harris,E.F., Newman,S.M. and Nicholsen,J.A. (1988) Nitinol archwire in
a simulated oral environment : changes in mechanical properties. Am. J.
Orthod. Dentofac. Orthop. 93 : 508 - 513.
Hazet,R.J., Soket,B.N. and West,V.C. (1984) Near - alpha titanium - a

potentially new orthodontic archwire material. Aust. Orthod. J. I : 90 -
91.
Hazel,R.J. and West,V.C. (1986) The importance of mechanical

behaviour in orthodontic archwires. Aust- Orthod. J. 9 :317 - 32O.
Hudgins,J.J., Bagby,M.D. and Erikson,L.c. (199O) The effect of long-

term deflection on permanent deformation of Nickel-Titanium archwires.
Angle Orthod. 60 : 283 - 288.
Hurst,C.L., Duncanson,M.G., Nanda,R.S. and Angolkar,P.V. (199O) An

evaluation of the shape memory phenomenon of nickel-titanium
orthodontic wires. Am. J. Orthod. Dentofacial. Orthop. 98 :72 - 76.
Ingerslev,C.H., et al. (1966) lnfluence of heat treatment on the physical

properties of bent orthodontic wire. Angle Orthod. 36: 237 - 247.
lngram,S.B.Jr., Gipe,D.P. and Smith,R.J. (1986) Comparitive range of

orthodontic wires. Am. J. Orthod. Dentofac. Orthop.9O: 296 -3O7.
Johnson,E. and Lee,R,S. (1989) Relative stiffness of orthodontic wires.

J. Clin. Orthod. 23 :353 - 363.
Jones,M.L., Staniford,H. and Chan,C. (199O) Comparison of superelastic

NiTi and multistranded stainless steel wires in initial alignment.
J.Clin.Orthod. 24: 61 1 - 613.
Kapilla,S. and Sachdeva,R. (1989) Mechanical properties and clinical

applications of orthodontic wires. Am. J. Orthod. Dentofac. Orthop. 96
: 1OO -109.
140
Khier,S.E., Brantley,W.A. and Fournelle,R.A. (1988) Structure and
mechanical properties of as- received and heat -treated stainless steel
orthodontic wires. Am. J. Orthod. Dentofac. Orthop. 93 : 206 - 212.
Khier,S.E., Brantley,W.A. and Fournelle,R.A. (1 991 ) Bending properties

of superelastic and non superelastic nickel-titanium. Am. J. Orthod.
Dentofacial Orthop. 99 : 310 - 318.
Kusy,R.P. (1981) Comparison of nickel - titanium and beta - titanium wire

sizes to conventional orthodontic archwire materials. Am. J. Orthod.79 :
625 - 629.
Kusy,R.P. and Dilley,G.P. (1984) Elastic modulus of triple - stranded

stainless steel archwire via three- and four-point bending. Dent. Res. 63
: 1232 -1240.
Kusy,R.P. and Stevens,L.E. (1987) Triple - stranded stainless steel wires

- evaluation of mechanical properties and comparison with titanium alloy
alternatives. Angle Orthod. 57 : 18 - 32.
Lane and Nikolai,R.J. (1980) Effects of stress relief on the mechanical

propert¡es of orthodontic wire loops. Angle Orthod. 50 : 139 - 145.
Lew,K. (1988) A comparison of archwires used for initial alignment in

Begg treatment. Aust. Orthod. J. 10 : 180 - 182.
Mason,R.J. (1969) An evaluation of .O16 inch orthodontic light wires

Thesis submitted in partial requirement for the degree of MDSc, Dept. of
Preventive Dentistry, University of Sydney.
Mayhew,M.J. and Kusy,R.P. (1988) Effects of sterilisation on the

mechanical properties and surface topography of nickel - titanium
archwires. Am. J. Orthod. Dentofac. Orthop. 93 : 232 - 236.
McCabe, J.F. Applied dental materials. 7th Edit¡on. Blackwell Scientific

Publications, London, Edinburgh, Boston, Melbourne, Paris, Berlin, Vienna
1 990.
t4t
Miniload Hardnes Tester manual. Ernst Leitz, GMBH and Wetzlar. (1968)
Miura,F., Mogi,M., Ohura,Y. and Hamanaka,H. (1986) The superelastic
property of the Japanese NiTi alloy wire for the use in orthodontics. Am.
J. Orthod. Dentofac. Orthop. 90 : 1-1O.
Miura,F., Mogi,M. and Okamato,Y. (1990) New application of

superelastic N¡Ti rectangular wire. J.Clin.Orthod. 24;544 - 548
Moran,K.l. (1987) Relative wire stiffness due to lingual versus labial

interbracket distance. Am. J. Orthod. Dentofacial Orthop. 92 : 24 - 32.
Nikolai,R.J., Anderson,W.T. and Messersmith,M.L. (1 988) Structural

reponses of orthodontic wires in flexure from a proposed alternative to
the existing specification test. Am. J. Orthod. Dentofac. Orthop. 93 :
496 - 504.
Orowan,E. (1959) Causes and effects of internal stresses. Internal

stresses and fatigue in metals , eds., Rossweiler,G.M.
Phillips,R.W. Skinner's science of dental materials. 8th ed. (1982) W.

B.Saunders Co.
Proffit,W.R. Contempory orthodontics. 2nd ed. (1993) The C.V.Mosby

Co.
Reitan,K. (1960) Tissue behaviour during orthodontic tooth movement

AJO 46: 881 - 898.
Riley,J.L., Garrett,S.G. and Moon,P.C. (1979) Frictional forces of ligated

plastic and metal edgewise brackets. J. Dent. Res. 58 : 421.
Rygh,P., Bowling,K., Hovlandsdal,L. and Williams,S. (1986) Activation of

the vascular system: A main mediator of periodontal fibre remodeling in
orthodontic tooth movement. Am. J. Orthod. 89: 453 - 468.
Sandstedt,G.: Eine Betrage zur Theorie der Zanregulierung, Nordisk

Tendlakare, Tidsskrift No. 4, 1904; Nos. 1 and 2, 1905 from "Begg
Ortho theory and text".
t42
Schudy,G.F. (1990) Bracket design and wire flexibility. JCO 24: 106 -
114.
Shastry,C., Elinson,A. and Goldberg,J. (1982) lnfluence of drawing

parameters on beta-titanium wires. J. Dent. Res. 61 : 305.
Singh,J., (1991) The mechanical properties and microstructure of

orthodontic wires. First year report submitted in partial fulfilment of the
requirement of Master of Dental Surgery.
Sokel B.N. (1984) Tensile properties of orthodontic wires. Thesis

submitted in partial fulfilment of the degree of MDSc in the Unversity of
Melbourne, Dept of Restorative Dentistry, University of Melbourne.
Standards association of Australia (19771 Australian standard

specification for orthodontic wires (Resilient) (Excluding precious metal
wires). Stand. Assoc. Aust,, North Sydney.
Storey,E. and Smith,R. (1952) Force in orthodontics and its relation to

tooth movement. Aust. J. Dent. 56 : 1 1 - 13.
Thomson,W. VibrationTheory and Applications. (2nd edition 1969)

Ruskin House, London.
Toms,A.P. (1988) The corrosion of orthodontic wire. Eur. J. Orthod' 1O:

87 - 97.
Twelftree,C.C. 119741 Elastic properties of orthodontic wire. Project

report submitted in partial fulfilment for the degree of MDS, Dept. of
Dental Health, University of Adelaide.
Twelftree,C.C., Cocks,G.J. and Sims,M.R. 119771 Tensile properties of

orthodontic wires. Am. J. Orthod .72 : 682 - 687.
ware, A.L. (1967) A study of the physical properties of resilient

orthodontic wires. Aust. Ortho. Bull. 5: 3 - 10.
1.43
Wilcock,A.J. (1989) Applied materials engineering for orthodontic wires.
Aust. Orthod. J. 11 : 22 - 29.
Wilcock,A.J. Jr. (1988) JCO lnterview on orthodontic wires. JCO 22:

484 -489.
Wilson,A.J. and Goldberg,A.J. (1987) Alternative beta-titanium alloys for

orthodontic wires. Dent. Mater. 3 : 337 - 341.
Yoshikawa,D.K., Burstone,C.J., Goldberg,A.J. and Morton,J. (1 981 )

Flexure modulus of orthodont¡c stainless steel wires. J. Dent. Res. 60 :
139 - 145.
r44

02whole v1

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

02whole v1

Uploaded by

Copyright:

Available Formats

t5'3.

METHODS OF TEST¡NG THE MECHANICAL PROPERTIES OF

A report submitted in partial fulfilment of the requirements for the degree

KATHERINE RUTH ALLEN B.D.S.

The University of Adelaide

LIST OF TABLES xil

SIGNED STATEMENT XVIII

Chapter 2 IIERATURE REVIEw 3

MECHANICAL PROPERTIES OF ORTHODONTIC TOOTH 3

THE IDEAL ARCHWIRE 3

TERMINOLOGY USED WHEN DISCUSSING THE MECHANICAL 4

THE CRYSTALLIZATION OF MOLTEN METALS 11

THE EFFECTS OF COLD WORKING AND ANNEALING ON THE 12

HISTORY OF ORTHODONTIC WIRES 13

THE CHEMICAL COMPOSITION AND MECHANICAL 15

PROPERTIES OF SPECIFIC WIRES

M ULTIBRAIDED STAINLESS STEEL 19

Chapter 3 AtMs oF THE lNVEsrlGATloNS 30

Chapter 4 MATERIALS AND METHoDS 32

CONVENTIONAL TENSILE TESTS 32

THREE POINT BEND TESTS 45

USE OF LOAD AND UNLOAD DATA 47

DETAILED MICROSTRUCTURAL DATA 56

Chapter 5 REsuLTs AND DlscussloN 67

ELASTIC MODULUS VALUES 67

CONVENTIONAL TENSILE TESTS 69

TESTING OF ANNEALED SPECIMENS 91

VELOCITY OF SOUND 103

SUMMARY OF ELASTIC MODULUS TESTS 108

A COMPARISON OF ELASTIC MODULUS VALUES DERIVED 109

FROM DIFFERENT TECHNIOUES USING LOAD DATA

EFFECT OF DIAMETER CHANGE ON ELASTIC MODULUS 111

THE EFFECT OF DATA MOD¡FICATION ON ELASTIC 111

MICROHARDNESS TESTS 114

MACROHARDNESS TESTS 116

SURFACE APPEARANCE OF WIRES 117

SURFACE EFFECTS OF PULSE STRAIGHTENING 117

SURFACE CHANGES AFTER ELECTROPOLISHING 117

DETAILED MICROSTRUCTURAL DATA 127

Chapter 6 coNclusloNs 132

Chapter 7 FUTURE RESEARcH

Unless otherwise Stated the symbols have the following meaning:

018 = O.O18" (O.45mm)

FIGURE SUBJECT PAGE

1 Stress-strain curve illustrating proportional limit, yield 7

strength, UTS, failure point, elastic modulus.

2 Stress-strain curve illustrating resilience and formability' 1O

3 Stress-strain curve illustrating differences between 26

4 A universal Instron testing machine with an 33

5 A wire specimen wrapped around an aluminium casing 34

6 A milled brass wire holder in a stand allowing 36

7 steel holder glued to the Mettler 61OO pan with anterior 37

I Electropolishing unit (Rocky Mountain 70O,USA) 39

9 A cantilever bend test us¡ng a knife edge to support the 42

10 A 3-point bend test 46

11 Apparatus for resonance testing 49

12 Apparatus for measuring speed of sound along a wire 51

13 A photomicrograph of microhardness indentations near 53

14 Curing characteristics of G-1 Epoxy used for preparing 58

15 Disc cutting and grinding for cross-sectional TEM 58

17a Magnified view of the grinding wheel and mount on 61

17b Dimpled specimen 61