Centers of rotation with transverse forces: An

experimental study
Hans N~igerl,* Charles d. Burstone, DDS, MS, b Benedict Becker," and
Dietmar Kubein-Messenburg d

Grttingen, W. Germany, and Farmington, Conn.

A new noninvasive method for simultaneously applying a known force system and measuring tooth
movement was developed and tested. The instruments proved to be highly accurate and reliable.
Using a model of an upper canine with a simulated periodontal ligament, we located the center of
rotation and the center of resistance and found that the product of two distances--a, from the center
of resistance to the line of application of force (F) and, b, from the center of resistance to the center
of rotation (CJ--yields a constant that represents the parallel transverse forces in the same plane.
Knowing this constant enables the clinician to determine the appropriate force position for a given
center of rotation. We also found that the occlusoapical position of the center of resistance varies
somewhat, depending on the transverse direction of loading around the long axis. Differences in
values of the Constant were found also to depend on the direction of loading in a given plane. Thus
a transverse force exerted in a mesiodistal direction at the level of an orthodontic bracket through
the long axis produces different effects from those produced when an identical force is applied in a
labiolingual direction. (AMJ ORTHOD DENTOFACORTHOP 1991;99:337-45).

T h e basis of orthodontic therapy is the ability to control the force system applied to the teeth. Yet
the relationship between the point at which a given force
is exerted on a tooth and the center of rotation for the
movement that such force produces is not completely
understood.
Previously, the problem has been attacked by analysis of both analytical and numerical models, t4 but,
effective as these models may be, it is necessary to test
them experimentally with suitable instrumentation in
laboratory models and, ideally, in vivo in human subjects. Previous experimental techniques have been limited by the invasive nature of mechanical force gauges,
the complexity of techniques such as holography, and
the inability to vary the position of the force on the
tooth by small incrementsJ s
The purpose of the present study was to develop a
reliable and relatively noninvasive technique for simultaneously measuring the pressure that exerted on a
tooth by the application of a specific force and the
resulting tooth movement.
=Dr. rer. nat., IV. Phys. Institute of the University of Grttingen, West Germany.
bProfessor, Health Center, University of Connecticut, Farmington, Connecticut.
elY. Phys. Inst. der Univ. Grttingen, West Germany.
aProfessor Dr.med., Kieferorthop. Abtlg. Zentrum Zahnmedizin, University
of G6rtingen, West Germany.
811120918

Burstone developed a two-dimensional analytical

model by assuming linear-elastic PDL characteristics,
a two-dimensional parabolic root, and a center of rotation that coincided with the location of zero stress in
the PDLJ
Nikolai,: in an improved theoretical model, considered the effects of root taper, both mesiodistally and
labiolingually. Although his analysis is basically twodimensional, it does include three-dimensional aspects of the root and hence offers a simplified threedimensional model.
We have developed a more general theory of tooth
movement, also based on linear-elastic assumptions,
that is three-dimensional and that describes motion in
the case of irregularly shaped roots. Because our theory
goes beyond the special cases previously described, it
is referred to as a general [heory Of tooth movement.
Burstone ~'6 stated in his simple formula:
y x (M/F) = 0.068 h 2 where
y = Distance from center of resistance to center of rotation
M / F = Distance from center of resistance to point of force application
h = Root length
Thus, in this special case of a two-dimensional parabolic root, 0.068 is a constant for a given root length.
Using the newer general theory of tooth movement,
337

338

Am. J. Orthod. Dentofac. Orthop.

April 1991

Niigerl et al.

i F

or
Torqu e~lexible

shaft

1- crown extension
| Force ,,

b = cr2
B
Fig. 1. A, A general theory of tooth movement states that the
product of the distances a and b equals a constant (~2). B, o~
is a constant, provided that the single forces are parallel and
act in the.same plane. Any
individual forces shown in B
produces the same o-2.

ofthe

we also found a constant for the general case, which

is referred to as 0.2. The present theory states in part
that a x b = 0.2, where a is the distance from the point
of application of force to the center of resistance and
b is the distance from the center of resistance to the
center of rotation (Fig. 1, A). It is beyond the scope
of this article to derive the formulas of the general
theory; rather, this study was designed to test the hypothesis that, for parallel forces acting in the same
plane, 0.2 is a constant that represents the distribution
of the restraining forces in the periodontal ligament. In
this study, transverse forces were applied through the
long axis of the tooth, in a special case of the general
theory that shows the constancy of 0.'~ for loading in
all directions and with forces away from the long
axis.
Experimental determination of 0.2 subsequently allows simple calculation of either the point at which
force must be applied for a specified center of rotation,
or, for a given point of application, the center of rotation
that will result. In this study, forces were applied transversely (at 90 to the long axis of the tooth) in various
planes through the long axis to identify the position(s)
of the center of resistance and the value of 0.2.

METHOD
An extracted upper right canine was used for the
experimental model. A mold was formed that allowed
placement ofa 1 mm thickness of soft plastic to simulate
the periodontal ligament (Moloplast B E = 7 N / m m 2,
lvoclar Dental GmbH, Ellwangen, West Germany).
The model was constructed of relatively rigid acrylic
(SR Kunststoff" Evoclar E = 2500 N / m m 2, Ivoclar

!i
I

root

load

.J/

Fig. 2. Loading method. The range of clinically relevant centerof-rotation locations obtained with the inherent sense of the
couple (torque) opposite the moment of the force about thecenter of resistance. The magnitude of the torque was varied
with a constant force.

Dental GmbH) in comparison with the low stiffness of

the periodontal ligament material.
A rectangular brass extension was attached to the
coronal portion of the tooth, where a torque and a force
were applied simultaneously to the tooth (Fig. 2). The
force was applied by a weight from a pulley, and the
torque was varied by a torque motor that delivered
torque to the tooth by means of a flexible shaft. Both
force and torque acted in the same plane. The shaft was
relatively noninvasive, since its spring constant was so
low as not to influence the movement of the tooth
(2.5 10 -6 N/mm).
Resolution of the torque was within I% or better.
The varying torque and the constant force were equivalent to a single transverse force acting at 90 angles
to the initial orientation of the long axis of the tooth
(Fig. 1, B). The position of the equivalent force was
moved occlusoapically in a given plane; in addition,
four different planes of loading were selected through
the long axis of the tooth, with two directions of loading
for each plane. Overall displacement occurred in the
following directions: labial, mesiolabial, mesial, mesiopalatal, palatal, distopalatal, distal, and distoiabial.
Three sensors for measuring two transverse displacements and one occlusoapical displacement were
used for full definition of the tooth movement and for
locating a uniplanar center of rotation. The center of
rotation is defined as that point around which all points
of the tooth will rotate; it need not be on the long axis

Volume 99
Number 4

Centers of rotation with transverse forces

339

I
|
1

,, it

.....

-1- ......

"x

Fig. 3. Method of displacement measurement. A minimum of three sensors was required to define
displacement in one plane.

of the tooth. At least three points are required for mathematical determination of the center of rotation in one
plane, since loading could also produce intrusion or
extrusion of the tooth. The sensors were inductive in
nature and loaded with a very low force (0.05 N, Feinpriif GmbH, G6ttingen, Germany), at the initial loading
position. The effect of the small load was corrected in
the computer program; hence the inductive transducers
were noninvasive (Fig. 3). Displacement resolution of
the sensors was 0.02 ~m.
The computer integrated and controlled the application of the forces and torques and the measurement
of displacement (Fig. 4). The applied torque was proportional to the current and was regulated by the computer. The sensors measured the displacement and, after
A - D conversion, the displacement figures were fed into
the computer, which calculated the locations of the centers of rotation.
The equipment tested 160 different equivalent force
positions by varying the magnitude of torque with respect to a constant force in each plane and for each
direction of loading. Readings were made 1 second after
application of the force to allow the tooth to reach a
state of equilibrium. Five separate readings were made
for each loading condition. Means and standard deviations were calculated for the location of the center of
resistance and tr z. The center of resistance was defined
as that point through which a single force produced no
XZ-plane rotation as measured directly by the transducers.
The overall equipment, including the tooth model,
torque motor and weight, displacement sensors, and
computer, is shown in Fig. 5.

AT
computer

A/D I

C 0 nve ~er

" I

converter

amplifier
displacement
-~sensor 1 0 ~

current
I
I
controlL_=,..
[+_0__400 reAl r
I 24V- /

-.gPIsensor 2 - ' ~ -,~ mot or

.~1~ mfml;t M
- ~ sensor 3

Fig. 4. Tooth modelmloading and measurement are controlled

by computer.

RESULTS
Loading of the tooth through its long axis demonstrated that a b = tr'- for any given plane. A representative graph is shown in Fig. 6, in which the dis-

340

Am. J.

Niigerl et al.

Orthod. Dentofac. Orthop.

April 1991

Fig. 5. Overall view of equipmentshowing tooth model, displacementsensors, torque motor, and pulley
with weight.

Table I. Position o f center o f resistance (CR), 0-2 values, and force direction
Force direction
Labial
Distolabial
Distal
Distopalatal
Palatal
Mesiopalatal
Mesial
Mesiolabial

CR (ram)

CelL

(cm2)

oe l L 2

6.9
5.4
6.4
6.6
6. I
6.9
7.0
8.3

0.35
0.27
0.32
0.33
0.31
0.35
0.35
0.42

1.029
1.009
0.869
1.028
i .237
0.885
0.909
0.985

0.257
0.252
0.217
0.257
0.309
0.22 l
0.227
0.246

Center of resistance is measured from a uniform alveolar crest. Average standard deviation for CR was 0.25 mm. Average standard deviation
for ~ was 0.033 cmL

tance between the center of rotation and the center of

resistance is plotted against the reciprocal o f the distance from the center o f resistance to the location o f
the single force. Note that all data points fall very close
to the regression line. The slope of the line is cr2. Coefficients o f correlation in all planes approached 1.0;
hence the theory was supported experimentally for this
special case.
Data for eight loading conditions through the long
axis o f the tooth are given in Table I. A statistically
significant difference in the positions o f the centers o f
resistance associated with the different planes was
found. The difference between the maximum and minimum values was significant at p < 0.001. This vari-

ation is a striking observation and will be discussed

further. The average position o f the center o f resistance
was 6.7 m m from the alveolar crest, which was 34%
o f the root measured from the alveolar crest (Table I).
Although center-of-rotation and center-of-resistance locations were determined with respect to a constant point
on the brass bar, the data are given with respect to a
uniform alveolar crest around the tooth on the model
because alveolar crest measurements are more clinically
relevant. Table I also gives both cr2 and "/ values for
the tooth. The 3, value is calculated by dividing 0-2 by
the square o f the root length (L). By using % one can
determine force positions or positions o f the center of
rotation as percentages of root length.

Volume 99
Number 4

Centers of rotation with transverse forces

341

'b

8- err/
Force F = 1N d i s t a l

6,,o
Jo

4-

.t~ o*
s"

2
ot./-'
i

I
-8

-6

-4

I
-2

...-'

2t

4:

6:

: a '--~
cm-1

.=.-"
.#"
#,

//

-- 4

a,b

= a= = 0 . 9

cm 2

/"

--6

.
.s

Fig. 6. Representative plot of center-of-rotation position versus reciprocal of force position. The constancy of slope (o=) is demonstrated.

In Fig. 7 the "7 values are plotted around the long

axis of the tooth with polar coordinates. It readily can
be seen that the ",/values are not equal for loadings in
different planes. Higher values were obtained for loading in a labiopalatal direction than in a mesiodistal
direction.
When the magnitude of force was varied with the
location of the force at the center of resistance in a
given plane, displacement was proportional to load in
the range of forces that were applied. This effect demonstrated the linear response of the PDL model and also
the accuracy of measurement within 3 I.tm of displacement (Fig. 8). For the same range of forces it was found
that the location of the center of resistance was independent of the magnitude of force (Fig. 9). These resuits correlate with the biologic characteristics of teeth
in vivo, where it has been shown that a relatively linear
relationship exists between force and tooth displacement at low force levels.
DISCUSSION

The equipment was accurate and gave highly reproducible data, both in determining the centers of ro-

Fig. 7. Gamma values are dependent on the direction of loading.

Mesiatiy and distally directed loads produced the smallest
magnitudes. Note that ~/distributions around the long axis approach an ellipse (dotted line).

342

Am. J. Orthod. Dentofac. Orthop.

April 1991

Niigerl et al.

Az__.
/Jm

-i

Tf"

"

elasticityconstant
D = 5,4kN,Zcm

___e

0.5

1.0

1.5

F/N

Fig. 8. Representative plot of force magnitude versus displacement during translation. Force-deflection
rate (D) is a constant.

uCR

cm

1.0'

t--j

r--rl--~--------

0.5
0

0.5

1.0

1.5

F/N

Fig. 9. Representative plot of force magnitude versus center-of-resistance position. No significant

difference occurs with increasing magnitudes of force. Dashed fine represents mean position of center
of resistance.

ration of the tooth and in its ability to vary by small

increments the point of loading along the long axis of
the tooth. The simplicity of design suggests that this
equipment is suitable for laboratory studies and, possibly, in vivo studies. Currently, pilot studies that use
palatolabial loads on the maxillary central incisor are
being carried out in the mouth.
One of the observations from the study was the
variation in the position of the center of resistance when
the plane of loading was altered. Although it is true
that a free body has a specific center of mass, regardless
of the direction of loading, this is not necessarily the

case with a restrained body such as a tooth. Orthodontists may have assumed that there is only one center of
resistance for a given tooth and its periodontal support.
This study shows that this may not be the case, and
one may anticipate that future clinical studies will show
that the center of resistance will vary with the direction
in which the tooth is to be moved. In the present study,
the location of the center of resistance varied from 27%
of the root length to 42%, as measured from a constantlevel alveolar crest. It has been previously reported that,
with a root of parabolic shape, one might predict a
center of resistance at 33% of the root length. 8 It is

Volume 9 9
Number 4

Centers of rotation with transverse forces

343

Table II. Distal canine movement: Position of force for various centers of rotation

Lcticeter l
rotation

bIL

Apex
Midpoint of CR to A p e x
2 Root lengths apical to CR
Translation (infinity)
Incisal edge
Bracket (13 m m from CR)

0.217
0.217
0.217
*
0.217
0.217

0.680
0.330
2.00
*
0.845
0,645

aI L

c IL

0.32
0.66
0. i 1
0.00
- 0.26 t
--0.34

0.00
0.34
--0.2 i 2
- 0.32
- 0.58
--0.66

a(mm )
6,4
13.2
2.2
0.0
- 5,2 t
--6.8

c(mm)
0.0
6.8
- 4.22
-6.4
-!1.6
13.2

Minus signs refer to distances or distances/root-length ratios apical to the center of resistance I or alveolar crest:. See Fig. I.
0.2

"Y-VE
Ca = Center of resistance.
L = Root length.
a = Distance from point of application of force to center of resistance.
b = Distance from center of resistance to center of rotation.
c = Distance from alveolar crest to point o f application of force.
*Undefined.

interesting to note that in this study the average position

of the center of resistance was found at 34% of the root
length from the alveolar crest.
The purpose of the study was not only to develop
equipment for measuring the location of the center of
rotation with a known force system but also to test the
theory that a b = cr2 such that tr 2 is a constant for
individual parallel forces applied at different points in
a single plane. This has been previously shown in the
special cases described by both Burstone 6 and Nikolai 2
in their mathematical models. In the general theory, tr 2
is not just a constant; it has an intrinsic meaning. It
represents the distribution of forces that support the
tooth in the periodontal ligament during loading. It is
beyond the scope of this article to discuss the derivation
of the formulas and to discuss in greater detail the
meaning of trS; nevertheless, it can be stated that the
greater o.2, the more resistant the tooth is to tipping or
rotation. This function can be influenced by anatomic
considerations such as the morphology of the root (root
diameter and length) and the constitutive behavior of
the periodontal ligament when properties vary from one
part of the ligament to the next. Gamma values are
useful clinically because they allow for calculations of
a (the distance from the center of resistance to the point
of force application), b (the distance from the center of
resistance to the center of rotation) in root-length units.
Gamma also allows convenient comparisons among
similarly shaped roots with similar periodontal support.
The formula can be used to determine the required
position of force for any desired center of rotation. The
change in position of the force for different centers of
rotation can be readily observed in Table II. For sim-

ilarly shaped teeth, the a l L column gives the position

of the force measured in root-length units from the
center of resistance and the c / L column gives the position of the force in root-length units measured from
the alveolar crest. This information is useful clinically:
for example, with distal movement, a force 32% of the
root length apical to the alveolar crest produced translation; a force 58% of the root length apical to the
alveolar crest produced root movement (rotation around
the incisal edge); moving the force further apically to
66% of the root length caused the tooth to rotate around
the bracket; and a force at the alveolar crest produced
tipping about the apex. The model in our study had a
uniform alveolar crest, unlike teeth in the human
mouth, where alveolar crest height varies circumferentially. A varying alveolar crest height could alter the
results somewhat.
In an earlier analytical study with a two-dimensional
model, the position of the force for a center of rotation
at the apex was found to lie at 16% of the root length
occlusal to the center of resistance. ~ This contrasts
with the 32% found in this study (Table II). The
general theory tells us that a two-dimensional model
will tend to underestimate ~'- or ",/values, so it appears
that the force positions for commonly used centers of
rotations may be further away from the center of resistance than previously anticipated. It should be noted,
also, that previous three-dimensional holographic studies showed a greater occlusoapical spread between the
positions of the force for commonly used centers of
rotation.
The (r2 or ",/values determine the sensitivity or precision required in the location of a single force for a

344 N?igerl et al.

desired center of rotation. When either of these values
is increased, the orthodontist is not required to use as
much precision to achieve that force position. In most
previous analytical studies, inadequate modeling gave
an apparent 0.2 value that was too small; however, the
correct O"2 values will not be known until relevant clinical studies are carried out. In this study, errors in 0.2
may have occurred because the applied forces may have
been slightly off the long axis of the tooth, producing
some rotation around the long axis. This rotation, because of the nature of the equipment, tends to increase
the value of 0.:. Nevertheless, these tr 2 values and the
derived 3' values may give a good prediction of the
clinical situation. The actual 0.2 values, however, must
await in vivo clinical studies.
The effect of different 3' values on the position of
the center of rotation can be found from data on the
canine that was studied. If one compares distal movement witl~ palatal movement when a single force is
directed through the long axis at the level of the bracket,
the centers of rotation can be calculated with the gamma
values. A higher gamma value of 0.309 was found in
a palatal direction in comparison to 0.217 in a distal
direction (significance: p < 0.001). The position of the
center of rotation with a palatally directed force lies
51% of the root length apical to the center of resistance,
and, with a distally directed force, it lies 33% apical
to the center of resistance. Thus, the angle of tip of a
tooth was less for lingual movement than for distal
movement during displacement of the same crown. Because there is a difference in the position of the center
of resistance with the force in the palatal direction versus distal direction, a better way to visualize where the
tooth will move is to measure the center of rotation
from a constant reference such as the alveolar crest,
which was assumed to be uniform circumferentially in
this study. Considering the effects of varying the location of the center of resistance, as well as the influence
of the 3' values, we found that the center of rotation in
the canine we studied was 16.5 mm below the alveolar
crest for palatal movement and 13.7 mm below the
alveolar crest for distal movement when a single force
was applied through the long axis of the tooth at the
level of the bracket.
Now that the constancy of 0.2 has been demonstrated, clinical studies should aim at determining 0.5
for different teeth (single- and multirooted) under different conditions of the periodontal ligament. Clinical
studies are difficult to carry out in all loading conditions;
therefore the theory helps, because a limited number
of loading conditions can establish 0.2 in a given plane.
Thus the general theory offers a rational framework for
future research that uses analytical, numerical, and ex-

Am. J. Orthod. Dentofa. Orthop.

April 1991

perimental methods. It should not be unreasonable to

suggest that similarly shaped teeth with sim'ilar coronal
root diameters, when corrected for root length, may
have similar 3' values or clinically useful ranges of 3'
values.
SUMMARY AND CONCLUSIONS

A new experimental method has been developed for

varying the torque while applying a constant force to
a tooth in a given plane; this situation is equivalent to
applying a single force at different occlusoapical locations. The method has proved to be both reliable and
accurate.
The equipment was tested on a simulated model
with an actual human upper right canine. Several significant findings were observed:
I. Although the center of resistance averaged about one
third the distance from the alveolar crest to the apex,
measured from the alveolar crest, the center of resistance was found in different occlusoapical positions, depending on the direction of the force. Thus
the location of the center of resistance cannot be
considered to be constant, independent of the
direction of loading, f o r ' a tooth with a given
support.
2. Two important parameters 0.2 and 3', which measure
the resistance of the tooth to tipping, were found to
be constants for loading in one plane of space, independent of the position of occlusoapical force.
3. Using experimental 0.2 or y values, one can calculate
the location of the center of rotation of the tooth for
a given force position or, conversely, when a center
of rotation is desired, the position of the force (or
the equivalent moment/force ratio at the bracket)
can be calculated.
4. Because the y values differed as the load was
changed from one plane to another through the long
axis of the tooth, it was shown that different
centers of rotation would be produced for a given
force location if the direction of loading was
changed.
5. The center of rotation of the canine was located more
apically, with forces directed labiopalatally than mesiodistally or palatolabially forces with forces occlusal to the center of resistance.
6. The experimental methods that have been developed, together with the general theory of tooth
movement, give a rational framework for future
clinical studies. In particular, the determination of
o-~ or 3' offers a promising approach for determining what forces should be used to control tooth
movement.

Volume 99
Number 4

Centers of rotation with transverse forces

REFERENCES
I. Burstone CJ. The biomechanics of tooth movement. In: Kraus
BS, Reidel RS, eds. Vistas in orthodontics. Philadelphia: Lea &
Febiger, 1962;197-213.
2. Nikolai RJ. Periodontal ligament reaction and displacements of a
maxillary central incisor subjected to transverse crown loading.
J Biomech 1974;7:93-9.
3. Nikolai RJ. An optimum orthodontic force theory as applied to
canine retraction. At4 J ORTHOD 1975;68:290-302.
4. Tanne K, Koenig HA, Burstone CJ. Moment to force ratios and
!he center of rotation. Ar,t J ORTtlOD DE.,'q'rOFACORTItOP 1988;
94:426-31.
5. Mi.ihlemann HR, Zander HA. Tooth mobility. 11I. The mechanism
of tooth mobility. J Pcriodont 1954;25:128-37.
6. Christiansen RL, Burstone CJ. Centers of rotation within the periodontal space. AM J OR'IqtOD 1969;55:351-69.

345

7. Burstone CJ, Pryputniewicz RJ, Bowley WW. Holographic measurement of tooth mobility in three-dimensions. J Periodont Res
1978; 13:283-94.
8. Burstone CJ, Pryputniewicz RJ. Holographic determination of
centers of rotation produced by orthodontic forces. Ar,I J ORTtIOD
1980;77:396-409.
9~ Tanne K. Stress induced in the periodontal tissue at the initial
phase of the application of various types of orthodontic force:
three-dimensional analysis by means of the finite element method.
J Osaka Univ Dent Soc 1983;28:209-61.

Reprint requests to:

Dr. Charles Burstone
University of Connecticut Health Center
School of Dentistry
Farmington, CT 06032

AAO MEETING CALENDAR

1991--Seattle, Wash., May 11 to 15, Seattle Convention Center

1992--St. Louis, Mo., May 10 to 13, St. Louis Convention Center
1993--Toronto, Canada, May 16 to 19, Metropolitan Toronto Convention Center
1994--Orlando, Fla., May 1 to 4, Orange County Convention and Civic Center
1995--San Francisco, Calif., May 7 to 10, Moscone Convention Center
1996-- Denver, Colo., May 12-15, Colorado Convention Center