na~~orces between bracket and are

Dieter Drescher, DMD, Christoph Bourauel, Dipl. Phys., and Hans-Albert Schumacher,
Bonn, West Germany

Guiding a tooth along an arch wire results in a counteracting frictional force. Clinically, a
mesiodistally applied force must exceed the frictional force to produce a tooth movement. A
friction-testing assembly simulating three-dimensional tooth rotations was constructed to study
factors affecting friction magnitude. Five wire alloys (standard stainless steel, Hi-T stainless
steel, Elgiloy blue, nitinol, and TMA) in five wire sizes (0.016, 0.016 x 0.022, 0.017 x 0.025,
0.618, and 0.018 x 0.025 inch) were examined with respect to three bracket widths (2.2, 3.3,
and 4.2 mm) at four levels of retarding force (0, 1, 2, and 3 N). The following factors affected
friction in decreasing order: retarding force (biologic resistance), surface roughness of wire, wire
size (vertical dimension), bracket width, and elastic properties of wire. The study recommends the
application of 0.016 x 0.022 inch stainless steel wire combined with a medium (3.3 mm) or wi
(4.2 mm) bracket for an arch-guided mechanism with an 0.016 inch slot. The effective force of
this arrangement has to increase twofold to overcome the friction. For TMA wire, however, the
effective force must increase sixfold, resulting in a hazardous overload of the anchorage units.
(AN J ORTHOD DENTOFAC ORTHOP 1989;96:397-404.)

iction is defined as a force tangential to
the common boundary of two bodies in contact that
resists the motion or tendency to motion of one relative
to the other.’ Since this force operates in the opposite
direction to the mobile body, it is important that fric-
tional forces should be eliminated or minimized when
orthodontic tooth movement is being planned. Mesio-
distal tooth movement can be accomplished by free
body movement or by guidance of a tooth along an arch
wire. The major advantage of the former mechanism
(e.g., sectional retraction arch wire) is the lack of fric-
tional forces between bracket and arch wire. Unfortu-
nately, such a mechanism is associated with undesired
tooth rotations in the sagittal and transverse planes,
resulting in an increase in the leveling requirements.
By contrast, the latter technique of guiding a tooth along
an arch wire decreases adverse rotating movements but
leads to friction, which results in a delay in tooth move-
ment, an increase in anchorage requirements, or both.
Orthodontic literature has a dearth of information
regarding this essential topic. Andreasen and Quevedo2
pulled wires of diverse sizes through fixed preangulated
brackets and found that increased tipping and a greater
wire cross section led to an increase in frictional forces.
The influence of bracket width and salivary lubrication
was negligible. Using the same system, Frank3 and
Frank and Nikolai4 agreed in their conclusions regard-
ing the wire dimension but found friction to increase
From the Department of Orthodontics, University of Bonn.
Wl/lQ423
with an increase in bracket width. To obtain the best
combination for canine retraction, they recommend the
use of edgewise wires with a relatively high stiffness
and narrow brackets. Riley et a1.j studied friction with
respect to diverse material composition (i.e., steel ver-
sus plastic) of ligation auxiliaries. They concluded that
steel ligatures generated greater frictional forces than
plastic modules and that moistening caused an insig-
nificant increase in friction in the case of steel ligature
and was irrelevant for the plastic modules. Using An-
dreasen’s apparatus, Peterson et al.” found nitinol wire
to exhibit less friction than stainless steel. Moreover,
the effect of wire size was negligible in the case of
nitinol. Garner et al.? examined diverse brackets and
arch wire materials on a simulated canine-retraction
assembly. Their results indicated that friction increased
in stainless steel, nitinol, and beta titanium (TMA), in
that order. Friction also increased with wire size.
Guiding a tooth along an arch wire can be divided
in four consecutive phases (Fig. 1):
Phase 1. Before application of force in the me-
siodistal direction and on completion of the leveling
stage, the arch wire lies in the slot, with no conflict
(Fig. 1, A).
Phase 2. Concomitant with force application in the
mesiodistal direction, the tooth tips and rotates since
the point of force application lies above the center of
resistance (Fig. 1, B and C).
Phase 3. Continuous force application sets an elas-
tic deformity in the arch wire (Fig. 1, D). The load
at the contact points between wire and bracket in-
 Am. J. Orthod. Lkntofac. Orthou.
Drescher, Bourauel, and Schumacher

i
Fig. 1. Three phases of mesiodistai tooth movement. In Phase 1, before application of mesiodistal
force and after leveling of the dental arch, the arch wire sits in the bracket’s slot with no conflict (A).
In phase 2, With initial mesiodistal force application, the tooth rotates in both the sagittal plane (tip)
(5) and the occlusal plane (C). Thus friction is generated at the contact points between bracket and
arch wire and between ligature and arch wire. In phase 3, with continuous force application, an elastic
deformity is set in the arch wire. The load, and hence the friction at the contact points, is increased,
resulting in antitip and antirotation counterforces (arrows). A force equilibrium is achieved, with the
tooth gliding along the arch in tipping upright oscillations (D).

creases as well as the friction. Thus a portion of
the mesiodistal force is lost. This elastic deformity
concurrently produces antitip and antirotational move-
ments of the tooth.
Phase 4, In an unbalanced situation, a permanent
deformity of the arch wire can be developed. Obviously,
the latter situation should be avoided.
Arch-guided tooth movement consists of repeated
movements of tipping and uprighting (phases 1 to 3).
In clinical situations, however, additional factors might
be involved; for example, masticatory impediment can
break this cycle by causing a permanent set in the wire
(phase 4), It has been suggested that saliva may reduce
friction by acting as a lubricant film. However, a pre-
liminary study has shown no difference between dry
models and wet models. This supports findings by
Andresean and Quevedo’ and Riley et a1.5 On the
other hand, the adhesive characteristic of saliva and
the accumulation of materia alba along the wire in
a nonhygienic mouth condition might hinder tooth
movement.
The objectives of the present study were to deter-
mine, on a dry model, changes in magnitude of friction
with respect to arch wire material, arch wire size,
bracket width, and biologic resistance. Inferences from
the study should assist the clinician in making the right
decision in selection of the proper mechanical system.
Other factors, such as masticatory impediment and sa-
liva, will be discussed in a forthcoming study.
MATERIALS AND METHODS
The effect of diverse wire material on friction was
studied on the following orthodontic wires: Unitek stan-
dard stainless steel (UNI), Unitek Hi-T stainless steel
(HIT), Rocky Mountain Elgiloy blue (ELI%), Unitek
nitinol (NIT), and Ormco TMA (TM-A). The influence
of wire size on friction was studied on the following
wire dimensions for each of the above-mentioned wire
materials: 0.016, 0.016 x 0.022, 0.017 x 0.025,
0.018, and 0.018 X 0.025 inch. (The latter dimension
was not available for nitinol and TM-4 at tbe time of
the investigation.) The effect of bracket width on fric-
tion was studied on 2.2 mm, 3.3 mm, and 4.2 mm
standard edgewise brackets (0.0185 inch slot, with no
torque and no angulation). The effect of biologic re-
sistance on friction was studied on four levels of re-
tarding force: 0 newton, 1 newton (102 gm), 2 newton
(204 gm), and 3 newton (306 gm).
 Frictional forces between bracket and arch wire 3

Fig. 2. Friction measurements were performed with a friction-testing assembly mounted on a universal
material-testing machine. The friction-testing assembly consisted of a simulated tooth (brass bar) with
an 0.018 inch bracket attached to it and a spring ligature providing constant ligation force of 1 N. The
simulated tooth (brass bar) and its supporting structure (elastic rubber foam) were mounted in an
aluminum housing (A). The aluminum housing, supported by a carriage, traveled linearly along two
steel shafts by means of ball bearings. The arch wire was mounted in a wire-holding device. The
applied force pulled the tooth crown (brass bar) in one direction, and the biologic resistance (retarding
force) pushed the roots (carriage) in the counter direction (B). Since the friction-testing assembly was
mounted on a material-testing machine designed for vertical-force measurement only, a second carriage
was employed to balance the weight of the first carriage. The applied force was measured with a
strain-gauge sensor during downward movements of the base table of the testing machine, during
which the arch wire was drawn through the bracket (C).

For simulation of an arch wire-guided tooth move-
ment, a full three-dimensional freedom of mobility of
the examined body (tooth) is required. This spatial mo-
bility allows for a complete simulation of the oscillated
cycle of tipping and uprighting tooth movements. More-
over, it was found that the tipping/uprighting cycle is
accomplished with low tooth velocity of 2 mm/ set and
is discontinued at higher velocities. Under low-velocity
conditions, both types of static friction* and sliding
(kinetic) frictioni8,’ occurred alternately but almost co-
incidentally. Moreover, according to Bilge,” the co-
efficient of friction for the steel/steel combination (in
“Static tictim = The smallest force needed to start motion of solid surfaces
at rest with respect to each other. Mathematically, f, 5 k,N where )L. =
coefficient of static friction and N = load.
Xinetic friction (4) = The force needed to resist the sliding motion of one
solid object over another at a constant speed. f, = p+N where bk = coefficient
of kinetic friction and N = load.
dry conditions) is almost identical (km = 0.15). For
this reason, the present study did not attempt to analyze
static and kinetic friction separately.
The friction-testing assembly consisted of two parts
(Fig. 2). The first part comprised a small carriage sup-
ported by ball bearings, which moved linearly along
two steel shafts. The simulated tooth was embedded in
an elastic rubber foam mediator, and both parts were
placed within an aluminum housing that was linke
the carriage. This construction permitted the desired
spatial rotations of the simulated tooth (Fig. 2, A
and B, and Fig. 3). A bracket and a spring-ligature unit
(Fig. 2, A) were attached to the tooth (brass bar). The
spring-ligature unit provided a light constant force of
1 N to engage the arch wire in the bracket’s slot. The
orthodontic force (mesiodistal force) was applied to the
bracket wing. The second part of the assembly-the
arch wire-holding device (Fig. 2, B)-comprised two
40 Drescher, Bourauel, trrzd Schumacher
Fig. 3. The simulated tooth it1 the course
set in the arch wire.
edgewise tubes that received the ends of the arch wire
and a coil spring to hold the arch wire under tension.
During measurement, the arch wire-holding device
is moved, causing the arch wire to slide through the
bracket slot. Such a design is equal to drawing a bracket
along a fixed arch wire. The orthodontic force devel-
oped during this process was measured by means of a
force sensor linked to the tooth by a rigid wire (Fig.
2, C). The retarding force (i.e., the counterforce to the
orthodontic force that, in an in vivo condition, is gen-
erated by the tooth-supporting structures) was applied
here to the carriage in the opposite direction (Fig.
2, B). Clinically, the retarding force is determined by
intrinsic factors, such as bone density and age, root
factors (number, configuration, surface area), and oc-
clusion. The greater the retarding force, the more tilt
the simulated tooth will undergo. This will increase the
elastic deformity of the wire and, consequently, the
load (friction) between bracket and wire will also
increase.
The friction-testing assembly was mounted on a
universal materiaf-testing machine (Zwick 1445, Ulm,
Einsingen, West Germany) consisting of a vertical slid-
ing table and a strain-gauge force sensor (Fig. 2, C).
Because of the vertical operation of the testing machine,
an additional carriage was incorporated into the friction-
testing assembly to compensate for the weight of the
first carriage. B;oth carriages were linked together by
of sliding movement. Note the S-shaped elastic deformity
two pulleys. The system was baianced by means of
auxiliary weights.
Movements of the sliding table of the testing ma-
chine, with the first carriage of the friction-testing as-
sembly attached to it: and the data obtained from the
strain gauge of the testing machine were computer con-
trolled. A single data collection was initiated after
3 mm idle movement downward of the sliding table.
This precaution assured that the simulated tooth oscil-
lated in the tippingiuprighting cycle. Because of
surface-texture irregularities of orthodontic wires, mul-
tiple measurements were conducted along the 10 mm
length of wires. A unit sampling for a wire consisted
of five measurements with no wire as well as with wire
loaded with retarding forces of 0, 1, 2, and 3 N. Each
unit sampling was repeated five times. With a sampling
frequency of 1000 Hz and with data acquisition of 5
seconds, 125,000 measurements were collected for
each bracket-wire combination. The low standard error
of the obtained frictional forces indicates high accuracy
of the measuring assembly.
The measurements recorded by the strain gauge
(Fig. 2, C) consisted of the retarding force, as well
as the frictional forces generated between the ball
bearings of the carriages and the shafts, the pulleys and
the ropes, and the bracket and the arch wire. The ef-
fective frictional force (bracket/ wire) was calculated
by subtracting the assembly friction (carriages/ shafts
 Frictional forces between bracket and arch wire

r
Friction vs Retarding Force
I 52 $ Friction vs Bracket Width

retarding forces [N]: bracket width [mm]:

wire: I wire:
HIT UN1 ELB NIT TMA ; I HIT ‘L’NI ELB KIT TMA

Fig. 4. A direct relationship exists between friction and the re- Fig. 6. Narrow brackets (2.2 mm) increased friction most sig-
tarding force for the different wire alloys. That is, friction in- nificantly in all examined alloys. However, effects of medium
creased linearly with an increase in the retarding force. (3.3 mm) and wide (4.2 mm) brackets depended on wire materia!
(Fig. 7).

AFriction vs Wire Dimension (HIT)

74 wire dimension [inch]: r
L3
f?d 0.016

L
retarding force [N] wlre:
Fig. 5. The influence of wire dimensions on friction is determined
by the vertical dimension of the wire. The histogram shows such Fig. 7. The bar diagram friction versus wire material as calcu-
an effect for a Hi-T wire in a narrow bracket at different levels lated for all measured data showed friction to be weakly related
of retarding forces. Note that the differences between a 0.016 to the elastic properties of the wires. Hence TMA is stiffer than
inch round wire and a 0.016 x 0.022 inch rectangular wire are nitinol but was demonstrated to have greater friction.
negligible.

effective force is a practical indicator for evaluation of

and pulleys /ropes), which was obtained from the zero
measurements, and the given retarding force from the
recorded measurement.
Frictional forces were evaluated in relation to the
following:
Retarding force. Fig. 4 shows a direct relationship
between frictional force and the retarding force for each
wire material. Thus friction increases with an increase
in retarding force (biologic resistance). There is almost
a linear increase in friction for each wire material. In
the case of TMA wire, however, a substantially steeper
slope can be observed. The ratio of applied force to
friction involvement. For example, to move the tooth
along a Hi-T wire a mesiodistal force of 2.3 N (234.6
gm) is required. This orthodontic force is composed of
a frictional force of 1.3 N (132.6 gm) (Fig. 4, second
bar from left) and a retarding force of 1 N (102 gm).
The ratio of 2.3 between the orthodontic force and the
retarding force means that the orthodontic force must
be increased by 2.3 to achieve mesiodistd tooth move-
ment. For the various bracket-wire combinations, the
ratio of applied force to effective force Auctuated be-
tween 2.0 (HIT) and 6.1 (TMA).
Wire dimension. Fig. 5 depicts the frictional forces
developed in Hi-T wires of different sizes drawn along
a narrow bracket. It is apparent that friction depends
 Fig. 8. Scanning electron micrographs
a direct relationship between surface
stainless steel (UNI)---smooth surface
coarse surface texture (C), and beta titanium
primarily on the vertical dimension of the wire. In gen-
eral, a 0.016 inch round wire and a 0.016 x 0.022
inch rectangular wire generated the lowest frictional
force, while the 0.018 X 0.025 inch rectangular wire
generated the greatest frictional force.
Bracket width. Fig. 6 demonstrates the relationship
between frictional force and bracket width for each wire
material. In contrast to the findings reported by Frank,’
this study found that, for each wire material, narrow
brackets (2.2 mm) always generated the greatest fric-
tional force. However, friction for medium (3.3 mm)
and wide (4.2 mm) brackets showed the same distri-
bution as for wire material (compare Figs. 6 and 7).
Wire material. Fig. 7 shows an overall analysis of
the collected data. Mean values were calculated for all
measurements except the 0.018 X 0.025 inch rectan-
gular wires. which were not available for all wire ma-
(original magnification x 5000) of the examined alloys depicted
texture and friction, which increased in the following order:
(A), Elgiloy blue (ELB)-flai relief (B), nitinol (NIT)-more
(TMA)-very crude surface tex?ure (D).
terials. Accordingiy, frictional forces increased in the
following order: Hi-T (HIT), standard stainless steel
(UNI), Elgiloy blue (ELB). nitinol (NIT), and beta
titanium (TMA). TMA wires always showed signifi-
cantly higher frictional force values.
DISCUSSION
The direct relationship found between the retarding
force (i.e., biologic resistance) and the frictional force
(Fig. 4) is due to augmentation in the load at the contact
points between the bracket and the arch wire as well
as between the bracket and the Ligature wire. This direct
relationship emphasizes the need for a passive leveling
arch before application of mesiodistal forces.
When selecting the proper arch wire size for a me-
siodistal tooth movement, the clinician should bear in
mind that friction depends primariiy on the vertical
Volume 96
Number 5
dimension of the wire (Fig. 5). This finding can be
further substantiated by the physical definition of fric-
tion (F = ~.LN), where the coefficient of friction (k)
and the vertical force (iV) (not a horizontally applied
farce) determined the magnitude of friction. A 0.016
inch Hi-T round wire and a 0.016 X 0.022 inch Hi-T
rectangular wire showed virtually the same amount of
friction (Fig. 5). Thus, for mesiodistal tooth movement,
the rectangular wire is preferred because of its addi-
tional feature of buccolingual root control.
Unlike other investigations,2-4 this study found nar-
row brackets to intensify friction by enhancing tipping
movements ( Fig. 6). The conflict is due to the degree
of freedom given to the simulated tooth in this setup
as compared with the spatially restrained tooth move-
ment designed in previous studies.2-4 This implies a
preference for the use of medium or wide brackets
in arch-guided tooth movement, particularly in cases
in which excessive mesiodistal tooth translation is
required.
A multifactorial analysis (Fig. 7) of all four ex-
amined factors (retarding force, wire size, bracket
width, and wire material) disclosed that wire material
was the decisive factor in affecting frictional involve-
ment. The increase in friction in Hi-T (HIT), standard
stainless steel (UNI), elgiloy blue (ELB), nitinol (NIT),
and beta titanium (TMA) (in that order) is in full agree-
ment with the observations of Garner et al.’ Further-
more, surface-structure analysis by means of scanning
electron microscopy (Fig. 8, A to 0) revealed distin-
guishing surface texture characteristics between diverse
wire materials. Stainless steel (UNI) and Elgiloy blue
(ELB) demonstrated a smooth surface texture, whereas
the nickel-titanium alloy (NIT) and especially the
titanium-molybdenum alloy (TMA) showed an exten-
sive surface roughness. Although no quantitative data
are currently available, it is apparent that surface texture
is a substantial factor in dictating friction magnitude in
edgewise mechanics. Elastic properties of orthodontic
wires are one expression of material characteristics. The
relationship between friction and the elastic properties
showed only a weak increase in friction with decreased
stiffness (Fig. 7). This relationship was based on the
assumption that tipping in mesiodistal tooth movement
decreased with increased stiffness (Fig. 7). However,
elastic properties per se cannot explain the frictional
behavior of TMA wire, since this wire is stiffer than
nitinol but exhibited significantly greater frictional
forces than nitinol. It is more plausible that the elastic
properties are secondary in affecting friction. This can
be elucidated by the force diagram analysis of the sim-
ulated tooth; that is, in an equilibrated condition (slow
and constant velocity), the tooth oscillated between the
Frictional forces between bracket and arch wire
tipping moment caused by the applied mesiodistal force
and the counter antitip moment elicited by the spring-
back force of the wire. Thus the force resulting at the
contact points between the bracket and the wire is al-
most independent of the elastic properties of the wire.
Wires with equal size and surface texture but with di-
verse elastic properties will generate nearly the same
frictional forces for a given bracket width. The only
difference between two such wires would be in the
equilibrated angle formed between the bracket slot and
the arch wire. In the case of TMA, high vaiues of
frictional force are the result of surface roughness and
a high coefficient of friction.
CONCLUSlON
The present study found the following factors to
affect friction in tooth-guided arch wire mechanics (in
decreasing order): retarding force (biologic resistance),
surface roughness of wire, wire size (vertical dimen-
sion), bracket width, and elastic properties of wire.
Clinically, for guiding a tooth along a continuous arch
wire in an 0.018 inch bracket, the study recommends
(1) the use of medium (3.3 mm) or wide (4.2 mm)
brackets, particularly in cases of major mesiodistal
tooth displacement; (2) the use of the 0.016 X
0.022 inch rectangular arch wire as a working arch to
provide better spatial root control; (3) the preclusion of
wires with rough surface texture (TMA, nitinol); and
(4) the reinforcement of the anchorage units (e.g., ex-
traoral) in maximal anchorage problems or, alterna-
tively, the application of nonguided arch wire mechan-
ics (e.g., sectional arches).
The clinical evaluation should take into account that
the effective force has to increase by twofold (stainless
steel) to sixfold (TMA) to overcome bracket-to-wire
friction.
We wish to express our gratitude to Dr. Alexander D.
Vardimon for his constructive criticism and to Dr. Lawrence
R. Voss as well as to Mrs. Rocio Avalos for revising the
manuscript.
REFERENCES
1. Morris W, ed. The American Heritage dictionary. Boston: Amer-
ican Heritage and Houghton Mifflin, 1969:577.
2. Andreasen GF, Quevedo FR. Evaluation of friction forces in the
0.022 x 0.028 edgewise bracket in vitro. J Biomech 1970;3:
151-60.
3. Frank CA. An evaluation of the frictional force generated be-
tween orthodontic bracket and arch wire [Master’s thesis].
St. Louis, Missouri: St. Louis University, 1979.
4. Frank CA, Nikolai RJ. A comparative study of frictional resis-
tances between orthodontic bracket and arch wire. AM J ORTHOD
1980;78:593-609.
5. Riley JL; Garrett SG, Moon PC. Frictional forces of ligated
 Am. .3. Orthod. Deniofac. Orthop.
4 Drescher, Bourauel, and Schumacher
November I989

plastic and metal edgewise brackets [Master’s thesis]. Richmond, New Encyclopedia Britannica, Macropedia. Chicago: Encyclo-
Virginia: Virginia Commonwealth University, Medical College pedia Britannica, Inc., 1981;705-9.
of Virginia, 1979. 10. B&e A. Formeln und Tabellen zur Mechanik und Festigkeits-
6. Peterson L, Spencer R, Andreasen GF. A comparison of friction lebre. BraunschweigiWiesbaden: Vieweg & Sohn. 1983:7.
resistance for Nitinol and stainless steel wire in edgewise brack-
ets. Quintessence Int 1982;5:563-5. Reprint requests to:
7. Garner JL, Allai WW, Moore BK. A comparison of frictional Dr. Dieter Drescher
forces during simulated canine retraction of a continuous edge- Department of Orthodontics
wise arch wire. Aiv~J ORTHOD DENTOFAC ORTHOP 1986;90:199- University of Bonn, School of Dentistry
203. Welschnonnenstr. 17
8. Sears FW, Zemansky MW, Young HD. College physics. 53 Bonn 1 West Germany
Menlo Park: Addison-Wesley. 1977:25-33.
9. Rabinowicz E. Tribological phenomena. Volume 18. In: The

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS

Bound volumes of the AMERICAN JOURNAL OF ORTHGDONTICS AND DENTO-
FACIAL ORTHOPEDICS are available to subscribers (only) for the 1989 issues from
the Publisher, at a cost of $42.00 ($52.00 international) for Vol. 95 (January-June)
and Vol. 96 (July-December). Shipping charges are included. Each bound volume
contains a subject and author index and all advertising is removed. Copies are
shipped within 60 days after publication of the last issue in the volume. The binding
is durable buckram with the journal name, volume number, and year stamped in
gold on the spine. Payment must accompany all orders. Contact The C.V. Mosby
Company, Circulation Department, 11830 Westline Industrial Drive, St. Louis, iM0
63146-3318, USA; telephone (800)325-4177, ext. 351.
Subscriptions must be in force to qualify. Bound volumes are not av~~lab
in place of a regular Journal subscription.