ORIGINAL ARTICLE

An evaluation of the shape of some popular nickel titanium

alloy preformed arch wires

Stanley Braun, DDS, MME,a William P. Hnat, PhD,b Reena Leschinsky, DDS,c and Harry L. Legan, DDSd
Nashville, Tenn, Louisville, Ky, and Chicago, Ill

The mathematical Beta function is shown to be an accurate planar representation of the natural human
arch form defined by the spacial coordinates of the labial and buccal dental/bracket interfacing surfaces
in the maxillary and mandibular arches. Graphic planar representations of the corresponding bracket
base spacial coordinates of 33 popular preformed nickel titanium arch wires and bracket assemblies
were superimposed on each of the relevant maxillary and mandibular natural forms. The arch forms of
the preformed nickel titanium arch wires and bracket assemblies did not emulate the natural human arch
form. The average mandibular natural human arch form first molar/canine width ratio is 2.38/1; the same
preformed arch wire/bracket ratio is 1.87/1. These ratios for the maxillary arch are 1.92/1 and 1.54/1,
respectively. The average canine width exceeded the natural canine width by 5.95 mm in the mandibular
arch and 8.23 mm in the maxillary arch. The corresponding mandibular first molar and maxillary first
molar widths exceeded the natural human first molar arch width by 0.84 mm and 2.68 mm, respectively.
These findings have implications with respect to posttreatment stability and facial esthetics. “Round
tripping” teeth resulting from subsequent change to stainless steel arch wires to restore a more natural
human arch form and size may result in deleterious tissue effects. (Am J Orthod Dentofacial Orthop
1999;116:1-12)

W ith the advent of nickel titanium highly
elastic preformed arch wires, the clinician is often able
to introduce larger cross-section wire in the early stages
of “leveling and alignment.” This type of arch wire has
become popular for it is believed to provide “greater
efficiency” during this stage of treatment.1 Because of
the metallurgic composition of these arch wires, the
clinician is unable to adjust their form or size to a given
patient. The principal clinical factors that affect a
patient’s dental arch form and dimensions are: dental
perimeter (sum of mesiodistal widths), cross-arch
width, and arch depth. By using nickel titanium pre-
formed arch wires, the clinician fails to recognize the
uniqueness of arch form and size required for a partic-
ular patient. This may be a significant factor in con-
tributing to posttreatment problems.2-4 Boone5 has sug-
gested that the individuality of a patient’s arch form and
dimensions must be recognized and respected if a suc-
aClinical Professor of Orthodontics, Vanderbilt University Medical Center, and
Marquette University.
bAssociate Professor of Mechanical Engineering, Speed Scientific School, Uni-
versity of Louisville.
cResident in Orthodontics, Vanderbilt University Medical Center.
dProfessor and Chairman, Department of Orthodontics, Vanderbilt University
cessful treatment outcome is to be achieved. Many
believe that arch form and size are unique for each
individual and are principally controlled by the form of
the basal bones initially and by the balance of energy
imparted to the teeth in all planes of space.6 Arch form
and size should be recognized as part of a morpholog-
ic human pattern.7-9
Recently planar projections (occlusal views) of
untreated human dental arches in Angle Class I occlu-
sion have been described by the mathematical Beta
function with an average correlation coefficient of 0.98
for the mandible and 0.97 for the maxilla.10 This earli-
er study revealed a variation in only dental arch size
related to the Angle classification of occlusion. With
these considerations in mind, an investigation was
undertaken to evaluate the dental arch form and size
determined by several popular brands of nickel titani-
um preformed arch wires and related bracket systems
compared with the untreated Angle Class I human arch
form and size described mathematically by the Beta
function. The Class I arch form was chosen because the
preponderance of malocclusions are corrected to this
Angle classification.

Medical Center. MATERIAL AND METHODS

Reprint requests to: Dr. Stanley Braun, 7940 Dean Road, Indianapolis, IN 46240.
Copyright © 1999 by the American Association of Orthodontists. The Beta function determined in the previously
0889-5406/99/$8.00 + 0 8/1/91942 referenced study10 is based on a highly accurate planar
1
2 Braun et al
Fig 1. A, Brown and Sharp Micro Val coordinate mea-
suring machine. B, Frictionless air-bearing touch probe
recording bracket position coordinates.
projection of the coordinates of the cusp tips of the
buccal teeth, and of the center point of each of the
incisal edges of the anterior teeth bilaterally for the
maxillary and mandibular dental arches. In order to
compare the arch form and dimensions to that delin-
eated by an arch wire and related bracket system, the
appropriate coordinates of the labial and buccal dental
surfaces that interface with the bracket base surfaces
must be described. Fifteen sets of casts exhibiting
American Journal of Orthodontics and Dentofacial Orthopedics
July 1999
Fig 2. Preformed arch wire/bracket assembly holding fix-
ture.
Table I. Bracket height relationships
Maxillary height Mandibular height
Tooth (mm) (mm)
Central incisor 4.0 3.5
Lateral incisor 3.5 3.5
Canine 4.0 4.0
Premolars 4.0 4.0
Molars 3.5 3.5
Class I occlusion that had been used in the earlier ref-
erenced study10 were used to establish the human arch
form and sizes for comparison purposes. Casts
exhibiting subdivision malocclusions, fractured, or
ectopically erupted teeth, deciduous teeth, or arch
length deficiencies in excess of 5 mm, were excluded.
All casts were of fully developed adult dentitions
including second molars.
A bracket height gauge was used to suitably mark
the labial and buccal surfaces of the teeth in a com-
monly used dimensionally coordinated plan as seen in
Table I. The dental center points of these markings
were also defined and their spacial coordinates identi-
fied using a Brown and Sharp Micro Val measuring
machine (North Kingston, RI) seen in Fig 1. A fric-
tionless air bearing probe records the coordinates of a
point in space to 0.001 mm in each of the three orthog-
onal axes. Linear accuracy is 0.006 mm, and repro-
ducibility is 0.004 mm. The casts were each secured to
a fixed plane, and the touch trigger probe used to iden-
tify the markings corresponding to the dental center
points at the appropriate heights, while recording the
American Journal of Orthodontics and Dentofacial Orthopedics Braun et al 3
Volume 116, Number 1

Table II. Correlation of each bracket coordinates relative to the Beta function
Class I: Mandibular arch
Measured values Curve fit values Correlation coefficient
Cast Depth (mm) Width (mm) Depth (mm) Width (mm) r

1a 41.742 60.543 43.632 61.892 0.98

2a 43.835 62.023 46.415 63.070 0.98
3a 48.304 58.212 51.139 60.712 0.93
4a 45.214 58.466 48.655 59.664 0.97
6a 48.630 58.097 51.520 58.950 0.98
7a 46.642 43.133 51.078 64.272 0.96
8a 42.784 59.131 46.127 61.630 0.97
9a 51.624 65.889 54.746 66.243 0.98
10a 48.374 60.356 52.568 62.997 0.94
1la 41.257 59.739 45.015 60.500 0.98
12a 48.947 59.653 51.227 65.598 0.91
13a 46.941 60.956 50.332 63.353 0.98
14a 41.787 58.211 42.807 58.500 0.99
15a 48.502 60.175 52.326 61.316 0.98
Average 46.042 58.899 49.113 62.050 0.97
Standard deviation 3.279 4.969 3.687 2.347 0.02
Maximum 51.624 65.889 54.746 66.243 0.99
Minimum 41.257 43.133 42.807 58.500 0.91
Class I: Maxillary arch

1b 38.321 56.553 39.378 58.732 0.97

2b 41.180 61.295 42.450 61.750 0.99
3b 42.621 54.777 45.244 55.807 0.98
4b 39.684 54.636 43.280 55.934 0.98
6b 44.967 53.637 46.556 55.093 0.94
7b 40.563 59.365 43.152 59.929 0.97
8b 38.858 57.001 41.019 59.123 0.98
9b 47.591 64.739 48.100 64.850 0.99
10b 45.628 52.850 47.797 58.982 0.82
1lb 36.793 55.085 38.250 55.120 0.97
12b 44.828 63.450 46.663 65.301 0.98
13b 42.168 57.498 45.371 58.132 0.98
14b 37.714 56.185 38.350 56.250 0.98
15b 42.778 56.243 44.790 58.293 0.97
Average 41.692 57.380 43.600 58.807 0.96
Standard deviation 3.257 3.594 3.346 3.289 0.04
Maximum 47.591 64.739 48.100 65.301 0.99
Minimum 36.793 52.850 38.250 55.093 0.82

corresponding X, Y, and Z coordinates automatically to
a computer file. Fourteen points were recorded for each
maxillary and mandibular cast. A planar projection of
each dental arch was subsequently obtained by reduc-
ing the Z coordinate of each landmark to zero. All mea-
surements were made by the same individual. A repro-
ducibility study using one cast (14 points, 42 coordi-
nates) was conducted to evaluate the variation in the
measurements. On the basis of the actual measure-
ments in each of the three coordinate axes, it revealed
the accuracy to be equivalent to the study cited earli-
er.10 This outcome was less than 0.5%.
The Beta function was subsequently tested as to fit
of these planar coordinates for each of the function
points for each of the 15 maxillary and 15 mandibular
casts. Using the average correlation coefficient as a
measure of fit, each maxillary and mandibular arch
was measured and found to be 0.96 and 0.97, respec-
tively (see Table II). Thus, the mathematical Beta
function is an accurate representation of the arch
forms of the maxillary and mandibular dentitions at
the interfaces with any preformed arch wire and
bracket system.
The ‘A’ Company (San Diego, Calif), Ormco
4 Braun et al American Journal of Orthodontics and Dentofacial Orthopedics
July 1999

Table III. Canine width and first molar width comparisons between arch wire/bracket assemblies and the Beta function
Mandibular arch: Beta function (width = 58.81 mm, depth = 43.60 mm)

Canine width (mm) First molar width (mm)

Wire no. Wire Beta Difference Wire Beta Difference Wire ratio Beta ratio Ratio difference

1 26.35 21.20 5.15 51.07 47.21 3.86 1.94 2.23 0.29

3 27.52 19.71 7.81 47.22 46.89 0.33 1.72 2.38 0.66
9 26.86 20.28 6.58 47.26 47.35 -0.09 1.76 2.33 0.58
10 26.46 19.71 6.75 45.48 47.78 -2.30 1.72 2.42 0.71
12 25.42 21.07 4.35 49.56 47.90 1.66 1.95 2.27 0.32
13 26.15 21.07 5.08 50.84 47.26 3.58 1.94 2.24 0.30
15 25.74 21.15 4.59 50.25 47.11 3.14 1.95 2.23 0.28
17 24.98 20.32 4.66 49.18 47.57 1.61 1.97 2.34 0.37
20 25.72 20.30 5.42 47.85 48.58 -0.73 1.86 2.39 0.53
22 25.20 19.91 5.29 46.98 48.43 -1.45 1.86 2.43 0.57
24 25.55 19.06 6.49 47.10 48.55 -1.45 1.84 2.55 0.70
26 25.85 19.60 6.25 49.09 47.41 1.68 1.90 2.42 0.52
28 26.66 20.72 5.94 49.25 48.67 0.58 1.85 2.35 0.50
30 26.84 19.16 7.68 49.38 47.50 1.88 1.84 2.48 0.64
31 25.82 18.29 7.53 47.79 46.95 0.84 1.85 2.57 0.72
33 25.74 20.17 5.57 48.89 48.52 0.37 1.90 2.41 0.51
Average 26.05 20.11 5.95 48.57 47.73 0.84 1.87 2.38 0.51
Maxillary arch: Beta function (width = 62.05 mm, depth = 49.11 mm)

2 35.04 25.92 9.12 55.55 50.23 5.32 1.59 1.94 0.35

4 34.71 25.66 9.05 52.28 50.21 2.07 1.51 1.96 0.45
6 34.22 26.41 7.81 49.39 50.38 -0.99 1.44 1.91 0.46
7 35.40 26.21 9.19 52.34 49.18 3.16 1.48 1.88 0.40
8 34.17 26.36 7.81 49.67 50.27 -0.60 1.45 1.91 0.45
11 34.57 26.27 8.30 53.86 50.26 3.60 1.56 1.91 0.36
14 35.07 26.65 8.42 55.27 49.69 5.58 1.58 1.86 0.29
16 35.02 27.71 7.31 52.62 49.65 2.97 1.50 1.79 0.29
18 34.78 25.89 8.89 55.18 49.56 5.62 1.59 1.91 0.33
19 33.22 25.58 7.64 52.89 50.79 2.10 1.59 1.99 0.39
21 34.08 27.22 6.86 52.17 51.02 1.15 1.53 1.87 0.34
23 34.77 25.82 8.95 51.63 51.18 0.45 1.48 1.98 0.50
25 35.08 25.70 9.38 51.68 49.72 1.96 1.47 1.93 0.46
27 32.20 26.90 5.30 54.91 51.45 3.46 1.71 1.91 0.21
29 34.99 24.48 10.51 53.50 49.26 4.24 1.53 2.01 0.48
32 36.57 26.18 10.39 53.78 51.27 2.51 1.47 1.96 0.49
34 32.68 27.76 4.92 53.84 50.91 2.93 1.65 1.83 0.19
Average 34.50 26.28 8.23 52.97 50.30 2.68 1.54 1.92 0.38

(Glendora, Calif), and Unitek (Monrovia, Calif),
donated several of their more popular nickel titanium
preformed arch wires and related bracket systems for
this study. The appropriate brackets were permanent-
ly affixed to the preformed arch wires using instant
glue (Borden Co., Columbus, Ohio). This procedure
eliminates any relative motion at the bracket/arch
wire interfaces. Each bracket was located along the
arch wire in the appropriate anatomic position as
determined by teeth ideally arranged in a typodont in
Class I occlusion. Fourteen brackets, from second
molar to second molar, were positioned and attached
in this way to each maxillary and mandibular pre-
formed arch wire tested. Points on each of the brack-
et mesh surfaces were identified that corresponded to
the points previously marked on the dental casts.
Next, the arch wires and appropriate bracket systems
were placed in the Brown and Sharp Micro Val mea-
suring machine and the X, Y, and Z coordinates for
each bracket/dental interface points were recorded as
described previously. A specially designed aluminum
fixture, seen in Fig 2, 6 inches square by 1/2 inch
deep, supported the arch wire/bracket assemblies
during measurement. A thin layer of wax, 4 inches
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 116, Number 1
Fig 3. Graphic representation of selected Ormco preformed arch wires/bracket assemblies contrasted
to the Beta function.
square by 1/8 inch deep, secured the test arch
wire/bracket assembly to the fixture without deform-
ing its initial shape. The arch wire/bracket assemblies
were easily changed by heating the wax layer. After
the Z coordinates were zeroed, a planar projection
(occlusal view) of the resulting arch form was
obtained for comparison to the natural human arch
form described previously.
Braun et al 5
RESULTS
The planar projections of each of the 33 nickel
titanium arch wire/bracket systems were superim-
posed on X and Y axes of symmetry passing through
the contacts of the central incisors for each of the rel-
evant maxillary and mandibular planar projections of
the natural human arch form. These superpositions
are seen in Figs 3 through 10. This allows for differ-
6 Braun et al
Fig 4. Graphic representation of selected ‘A’ Company preformed arch wires/bracket assemblies con-
trasted to the Beta function.
ences in form and size to be readily observed. Table III
lists the canine and first molar width differences
between each of the 16 mandibular arch wire/bracket
assemblies and 17 maxillary arch wire/bracket assem-
blies as compared to the natural human arch form. The
average canine width exceeded the natural canine
American Journal of Orthodontics and Dentofacial Orthopedics
July 1999
width by 5.95 mm in the mandibular arch and 8.23 mm
in the maxillary arch. The corresponding mandibular
first molar and maxillary first molar widths exceeded
the natural human arch width by 0.84 mm and 2.68
mm, respectively. First molar/canine width ratios were
also calculated to quantitate the arch form of the arch
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 116, Number 1
Fig 5. Graphic representation of selected Unitek preformed arch wires/bracket assemblies contrasted
to the Beta function.
wire/bracket systems to the natural human arch form.
The average mandibular natural human arch form first
molar/canine width ratio is 2.38/1, whereas the same
arch wire/bracket ratio was found to be 1.87/1. These
ratios for the maxillary arch wire/bracket systems were
1.92/1 and 1.54/1, respectively.
Braun et al 7
DISCUSSION AND CONCLUSIONS
All arch widths, measured at the canines and first
molars, determined by the preformed nickel titanium
arch wire/bracket systems were greater than those of
the natural human arch form. This has two important
implications: (1) if these increased widths are main-
8 Braun et al
Fig 6. Graphic representation of selected Ormco preformed arch wires/bracket assemblies contrasted
to the Beta function.
tained, posttreatment stability may be compromised2-4;
and (2) because a particular patient’s arch length is det-
termined by the sum of the mesiodistal widths of all of
the teeth present in the arch, an increased width will
result in a decreased arch depth. This will subsequent-
ly impact on lip support, and facial esthetics may be
affected.11
American Journal of Orthodontics and Dentofacial Orthopedics
July 1999
Because the natural human arch form is
described by the Beta function (r = 0.96 for the
maxilla, and r = 0.97 for the mandible), Figs 3A
through 10B illustrate that this form is not general-
ly emulated by the popular preformed nickel titani-
um arch wire/bracket systems tested. The pre-
formed nickel titanium arch wire/bracket systems
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 116, Number 1
Fig 7. Graphic representation of selected Ormco preformed arch wires/bracket assemblies contrasted
to the Beta function.
exhibit an average canine width increase of 5.95
mm and a lesser (0.84 mm) increase in first molar
width, when compared with the natural human arch
form. The larger arch width increases are confined
to the anterior portion of the arch wires. This is
reflected in the wire ratio (molar width divided by
canine width) being significantly lower than the
Braun et al 9
same ratio for the natural untreated arch form (see
Table III). This was found to be the case in the
maxillary and mandibular arches. If the clinician
believes that the arch form and size can subse-
quently be altered to a predetermined treatment
goal for a specific patient by changing to a stainless
steel arch wire, he or she should consider the
10 Braun et al
Fig 8. Graphic representation of selected ‘A’ Company preformed arch wires/bracket assemblies con-
trasted to the Beta function.
potential deleterious effects of “round tripping”
teeth.12-14 On the other hand, if the arch form and
size determined by the preformed nickel titanium
arch wire/bracket system is maintained, posttreat-
ment instability may ensue.2-4
Manufacturers should consider a redesign of cur-
American Journal of Orthodontics and Dentofacial Orthopedics
July 1999
rent preformed arch wires to approximate the Beta
function that more closely emulates the human arch
form in Class I occlusion. Thus a more natural arch
form could be readily designed to accommodate vari-
ous arch widths, ensuring the uniqueness of each
patient’s dental arch size.
American Journal of Orthodontics and Dentofacial Orthopedics
Volume 116, Number 1
Fig 9. Graphic representation of selected Unitek preformed arch wires/bracket assemblies contrasted
to the Beta function.
REFERENCES
1. Roberin M, Laumon B, Marten J, Brumner F. Dimensions and form of dental arches
in subjects with malocclusions. Am J Orthod Dentofacial Orthop 1993;104:67-72.
2. Little RM, Wallen TR, Riedel RA. Stability and relapse of mandibular anterior align-
ment: first premolar extraction cases treated by traditional edgewise orthodontics. Am
J Orthod 1981;80:349-65.
3. Felton MJ, Sinclair PM, Jones DL, Alexander RG. A computerized analysis of the
shape and stability of mandibular arch form. Am J Orthod Dentofacial Orthop
1987;92:478-83.
Braun et al 11
4. Burstone CJ. Retention and stability in orthodontics. Philadelphia: WB Saunders;
1993.
5. Boone GN. Arch wire designed for individual patients. Am J Orthod 1963;33:178-85.
6. Braun S, Bluestein M. A new concept: impulse in orthodontics. Am J Orthod Dento-
facial Orthop 1997;111:437-40.
7. Brader AC. Dental arch form related with intraoral forces: PR = C. Am J Orthod
1972;61:541-60.
8. Weinstein S. Minimal forces in tooth movement. Am J Orthod 1967;53:881-903.
9. White LW. Individualized ideal arches. J Clin Orthod 1978;11:779-87.
12 Braun et al American Journal of Orthodontics and Dentofacial Orthopedics
July 1999

Fig 10. Graphic representations of selected Ormco preformed arch wires/bracket assemblies contrast-
ed to the Beta function.

10. Braun S, Hnat WP, Fender DE, Legan HL. The form of the human dental arch. Angle
Orthod 1998;68:29-36.
11. Proffit WR. Contemporary orthodontics, 2nd edition. St. Louis: Mosby–Year Book; 1993.
12. Wehrbein H, Fuhrmann RAW, Diedrich PR. Periodontal conditions after facial root tipping
and mpalatal root torque of incisors. Am J Orthod Dentofacial Orthop 1994;106:455-62.
13. Reitan K. Some factors determining the evaluation of forces in orthodontics. Am J
Orthod 1957;43:32-45.
14. Reitan K. Biomechanical principles and reactions. In: Graber TM, Swain BF, edi-
tors. Orthodontics: current principles and techniques. St. Louis: CV Mosby;
1985.