ORIGINAL ARTICLE

The wire material and cross-section effect on double delta closing

loops regarding load and spring rate magnitude: An in vitro study

Marcelo do Amaral Ferreira, CD, EOa

Curitiba, Paraná-Brasil

The mechanical behavior of orthodontics closing loops, with three different wire materials (stainless steel,
cobalt-chromium and titanium-molybdenum) and with different cross-sections and a double delta design,
was studied in tension tests. The springs were stress-relieved, except the titanium-molybdenum wires.
There were 72 sample springs, divided into 33 stainless steel, 26 cobalt-chromium and 13 titanium-
molybdenum, activated at 0.5 mm intervals, from neutral position to 3.0 mm. It was hypothesized that
loads, after spring activation, and spring rate, are dependent on cross-section, wire material, and
activation. The analysis of variance and the Tukey-Kramer test were applied to verify the differences
between all coupled averages of the loads. Regression analysis was also used to verify if closing loops
behavior was in accordance with Hooke’s law and to obtain the spring rate. The results show that the loads
are dependent on activation, cross-section, and wire material. Titanium-molybdenum 0.017 × 0.025 inch
(Ormco) springs showed the smallest loads and the best spring rate. (β = 84.9 g/mm) (Am J Orthod
Dentofacial Orthop 1999;115:275-82)

T he segmented arch technique has been
recognized for achieving a more efficient force system,
increasing the distance between points of force appli-
cation, which lowers the load-deflection rate.1-5 In
addition, the segmented arches might include prefabri-
cated and precalibrated springs,5,6 which provide more
precision concerning the force magnitude and
moments. In this way, a more efficient and controlled
dental movement can be achieved.
The sectional springs can be used to retract canine
teeth5-7 as well as to take part in a segmented arch5 to
retract the anterior segment, the posterior segment, or
both at the same time, depending on the treatment plan
(alpha and beta moments and loop position will deter-
mine the type of tooth movement).
Studies have been made with the aim of learning the
force levels,7-9 and, currently, different spring designs
are used that take into consideration several aspects
such as vertical forces, horizontal forces and the alpha
and beta moments.5,6
With the creation of new alloys in orthodontics,
such as nickel-titanium and beta-titanium,3,10,11 there
This study was supported by CDTN-CNEN and Pontificia Universidade Católi-
ca de Minas Gerais, Belo Horizonte-Brasil. An abbreviated form of a disserta-
tion submitted to the Centro de Odontologia e Pesquisa-Pontifícia Universidade
Católica de Minas Gerais. Belo Horizonte-Brasil.
aIn private practice.
Reprint requests to: Marcelo do Amaral Ferreira, 1183 Pref. Omar Sabbag Ave.,
Jardim Botânico-Zip Code: 80.210.000, Curitiba, Paraná-Brasil; E-mail,
marcelo.ferreira@avalon.sul.com.br
Copyright © 1999 by the American Association of Orthodontists.
0889-5406/99/$8.00 + 0 8/1/91525
was the need to compare them to the alloys already in
use by checking their mechanical properties. Spring
rate (load-deflection rate) is a very important mechan-
ical property when it comes to the study of springs
because it allows the orthodontist to choose the ones
that deliver low loads per millimeter of deactivation.
Springs with a low load-deflection rate deliver more
constant forces over the periodontal membrane during
unloading.2,5,6
The purpose of this study is to determine the resul-
tant loads after successive activations and the spring
rate of double delta closing loops of different wire
materials and cross-sections through tension tests, and
to check whether there was any statistical difference
among the spring groups studied. It was hypothesized
that resultant loads after spring activation and spring
rates are dependent on cross-section, wire material, and
activation.
MATERIAL AND METHODS
Material
The sample consisted of 72 springs with a loop
design of double delta (Fig 1) as used by Ricketts9 in
continuous arches. The springs were fabricated with
three different wire materials (stainless steel, cobalt-
chromium and titanium-molybdenum) and different
cross-sections as shown in Table I. The springs were
built by the same operator, over a template, and their
angles were rounded to reduce residual stress and to
avoid fracture. Titanium-molybdenum alloys cannot
stand bends over sharp radius. Bends were made on
275
276 Ferreira American Journal of Orthodontics and Dentofacial Orthopedics
March 1999

Fig 1. Standard form. (Arrows indicate rounded areas.)

Fig 3. Transducer and microvoltimeter set.

Fig 2. Spring attached on linear displacement transducer.

Fig 4. Transducer calibration set.

Table I. Wire material, manufacturer, cross-section, and
samples
Manu- Cross-section Modulus of 450oC for 3 minutes. Foil was used to keep the springs
Wire material facturer (inch) Sample elasticity from getting in touch with the leaden solution. The
temperature was accurately checked through a thermo-
Stainless steel Morest 0.018 × 0.025 7
0.019 × 0.025 7 couple (high temperature thermometer adapted to a
3M/Unitek 0.017 × 0.022 6 23-26 × 106psi microvoltimeter), type J, for iron-constant at CDTN-
0.018 × 0.025 7 CNEN (Nuclear Technology Developing Center -
0.019 × 0.025 6 Nuclear Energy National Committee) laboratories.
Titanium- Ormco 0.017 × 0.025 6 8.7 × 106psi
molybdenum 0.019 × 0.025 7 9.9 × 106psi Methods
(β-titanium)
Cobalt- RMO 0.016 × 0.016 17 26 × 106psi The tests were performed on a linear displacement
chromium 0.016 × 0.022 9 transducer where the ends of the spring were attached
(Blue Elgiloy) (Fig 2). This apparatus was manufactured at CDTN-
Morest Corp, Cable Strand, Long Beach, Calif. CNEN, with the support of the Centro de Odontologia
3M/Unitek Corp, Monrovia, Calif. e Pesquisa - Pontifícia Universidade Católica de Minas
Ormco Corp, Glendora, Calif. Gerais, Belo Horizonte, Brazil. In this study only hori-
RMDl Inc, Denver, Colo.
zontal forces were tested. The transducer was charac-
terized by having 5 g of sensibility and scale rate of 1
each end of the springs in order to make them fit into a kg. It was connected to a microvoltimeter of u = 2.5 v
transducer. The samples in stainless steel and in cobalt- of supply source (Fig 3). Each spring was activated by
chromium were submitted to heat treatment for stress- the same operator. Each 0.5 mm interval of activation
relief and improvement of their elastic properties. The was verified through a tensor micrometer in millime-
heat treatment was carried out in a leaden solution at ters. The resultant millivoltage after each activation
American Journal of Orthodontics and Dentofacial Orthopedics Ferreira 277
Volume 115, Number 3

Table II. Transducer calibration

Dead Load Unload Load Unload Load Unload Average Delta
load (g) (MILI V) (MILI V) (MILI V) (MILI V) (MILI V) (MILI V) (MILI V) (V)

0 −3131 −3125 −3129 −3124 −3126 −3124 −3127 (A) 0=A-A

200 −4229 −4222 −4225 −4219 −4223 −4218 −4223 (B) −1096 U=B-A
400 −5323 −5314 −5317 −5312 −5316 −5312 −5316 (C) −2189 V=C-A
600 −5871 −5862 −5865 −5860 −5864 −5860 −5864 (D) −2737 X=D-A
800 −7520 −7520 −7513 −7515 −7511 −7512 −7514 (E) −4387 Y=E-A
1000 −8615 −8615 −8614 −8614 −8613 −8613 −8614 (F) −5487 Z=F-A

Table III. Descriptive statistics of stainless steel groups (SS)

Activation Average Coefficient of
Spring groups (mm) load (g) SD variance (%) Median Minimum Maximum

SS 0.019 × 0.025 inch 0.5 138.5 24.7 17.9 143 96 166

(3M/Unitek) 1.0 266.8 20.2 7.6 268.5 235 290
1.5 403.6 19.9 4.9 404.5 380 433
2.0 544.1 20.4 3.7 541 520 576
2.5 690.0 25.3 3.7 685 660 730
3.0 838.0 29.5 3.5 834 805 884
SS 0.019 × 0.025 inch 0.5 132.4 25.3 19.1 132 101 175
(Morest) 1.0 253.7 32.6 12.8 251 207 295
1.5 379.5 41.4 10.9 368 327 431
2.0 506.1 43.34 8.6 495 452 557
2.5 643.8 69.7 9.4 621 575 738
3.0 781.2 69.8 8.9 753 708 894
SS 0.018 × 0.025 inch 0.5 121.0 10.4 8.6 124 109 138
(3M/Unitek) 1.0 238.2 11.3 4.7 237 228 260
1.5 357.5 19.9 5.6 355 333 394
2.0 497.2 52.3 10.5 482 441 583
2.5 624.8 61.3 9.8 608 550 725
3.0 751.4 68.7 9.1 738 664 861
SS 0.018 × 0.025 inch 0.5 146.5 16.0 10.9 145 118 165
(Morest) 1.0 263.7 17.7 6.7 261 240 294
1.5 377.8 21.1 5.6 374 356 416
2.0 498.0 26.0 5.2 496 464 543
2.5 620.5 33.9 5.5 619 571 675
3.0 746.7 41.5 5.6 750 685 811
SS 0.017 × 0.022 inch 0.5 119.0 6.6 5.6 118.5 113 131
(3M/Unitek) 1.0 204.0 17.3 8.5 205 179 229
1.5 295.1 20.8 7.0 294 272 324
2.0 395.3 17.2 4.3 392.5 374 417
2.5 492.1 32.3 6.6 498 445 521
3.0 591.8 36.4 6.2 599 536 623

was checked when deactivation started. Checking Statistical Analysis

whether the values for each activation interval were the
same was essential. If there were not repeated values,
the spring would not be following Hooke’s law and it
was consequently discarded. The experiment was car-
ried out three times. Afterward, those millivoltage val-
ues were transformed into load units (g). The transduc-
er was previously gauged through a connected scale
where dead weights of 200, 400, 600, 800, and 1000 g
were being placed (Fig 4). It was established that 1 g =
−5.5 mv (Table II). The minus sign is a microvoltime-
ter referential.
The variables were statistically analyzed by means
of an analysis of variance (ANOVA) through the F
Snedecor test for a hierarchy experiment.12 The Tukey-
Kramer test was also used to verify the difference
between all coupled averages of the achieved loads; the
regression analysis, using the Student t test, was
applied to verify if the spring’s behavior was following
Hooke’s law. The degree of linear regression was des-
ignated by the purpose coefficient R2. The test f of
Snedecor was also used to test the hypothesis that the
independent variables (wire material, cross-section,
278 Ferreira American Journal of Orthodontics and Dentofacial Orthopedics
March 1999

Table IV. Descriptive statistics of cobalt-chromium groups

Average Coefficient of
Spring group Activation load (g) SD variance (%) Median Minimum Maximum

Co-Cr 0.016 × 0.022 0.5 112.6 9.4 8.4 111 102 127
inch (RMO) 1.0 203.1 15.5 7.6 200 186 229
1.5 301.3 26.3 8.7 304 267 339
2.0 398.5 38.1 9.5 395 353 454
2.5 501.5 50.3 10.0 494 440 573
3.0 605.1 61.6 10.1 602 531 692
Co-cr 0.016 ×0.016 0.5 75.8 16.2 21.3 77 50 104
inch (RMO) 1.0 145.1 21.7 14.9 144 112 179
1.5 219.6 26.9 12.2 221 171 259
2.0 297.8 32.4 10.8 303 228 348
2.5 376.9 38.5 10.2 382 288 436
3.0 456.1 47.6 10.4 458 352 528

Table V. Descriptive statistics of titanium-molybdenum groups (TMA)

Average Coefficient of
Spring group Activation load (g) SD variance (%) Median Minimum Maximum

TMA 0.019 × 0.025 0.5 97.2 22.4 23.0 86 75 134

inch (ORMCO) 1.0 156.1 30.6 19.6 142 129 207
1.5 215.2 31.8 14.7 202 181 267
2.0 277.5 35.0 12.6 263 235 333
2.5 342.7 39.2 11.4 332 291 404
3.0 408.8 42.8 10.4 398 350 475
TMA 0.017 × 0.025 0.5 48.6 23.1 47.6 38 29 85
inch (ORMCO) 1.0 86.6 24.6 28.3 77.5 63 122
1.5 125.6 24.6 19.6 116.5 101 162
2.0 167.8 25.2 15.0 157.5 140 204
2.5 211.0 25.0 11.8 218 183 249
3.0 257.0 24.8 9.6 250 227 294

and activation) were statistically relevant, explaining
by these means the variance of the stored loads of the
springs. After the regression adjustment, the confi-
dence interval was built for the spring rate on the con-
fidence level of 95%.
The difference between stainless steel and cobalt-
chromium was 124 g, whereas the difference between
stainless steel and titanium-molybdenum reached 231
g. Finally, the difference between cobalt-chromium and
titanium-molybdenum was 107 g (Table VI).

RESULTS Wire Material and Cross-section—Effect on Load

The results reached in several activations in the dif-
ferent spring groups are summarized in Tables III, IV,
and V and Fig 5.
Wire Material—Effect on Load
There was a significant difference among the three
types of researched alloys. The stainless steel springs
showed the highest total average load, which means the
arithmetic average of all reported activations for each
spring group (cross-section and wire material) (431 g),
followed by the cobalt-chromium springs (308 g), and
finally by the titanium-molybdenum springs (200 g).
When the three different types of researched alloys were
compared, they showed critical significant differences.
• These findings showed that the total average load on
stainless steel 0.019 × 0.025 inch (3M/Unitek) was the
highest and significantly different from the average of the
other ones.
• The second highest total average load was observed on
stainless steel 0.019 × 0.025 inch (Morest), stainless steel
0.018 × 0.025 (Morest), stainless steel 0.018 × 0.025 inch
(3M/Unitek), which did not show any significant differ-
ence among them.
• The third highest total average load was observed on
cobalt-chromium springs 0.016 × 0.022 inch (RMO) and
stainless steel 0.017 × 0.022 inch (3M/Unitek), which did
not show significant difference between them.
• The fourth highest total average load was observed on
cobalt-chromium 0.016 × 0.016 inch (RMO) and titani-
American Journal of Orthodontics and Dentofacial Orthopedics Ferreira 279
Volume 115, Number 3

Fig 5. Elastic behavior of spring groups.

Table VI. Wire material, arithmetic average of total load, Table VII. Wire material and cross-section (spring groups)
difference between averages, and critical difference (5%) Wire material and Arithmetic
Arithmetic Difference cross-section (inch) average of total
average between Critical (spring groups) load (g)
Wire of total averages difference
material load (g) Comparison (g) (5%) Stainless steel (3M/Unitek) (A1) 480.1
0.019 × 0.025 (A1)
Stainless 431 A×C 124 8.62* Stainless steel (Morest) (A2) 449.5
steel (A) 0.019 × 0.025 (A2)
Cobalt- 308 A×T 231 10.76* Stainless steel (3M/Unitek) (A3) 442.2
chromium (C) 0.018 × 0.025 (A3)
TMA (T) 200 C×T 107 11.17* Stainless steel (Morest) (A4) 431.7
0.018 × 0.025 (A4)
*Statistical significant difference by Tukey-Kramer test (5%).
Stainless steel (3M/Unitek) (A5) 349.5
0.017 × 0.022 (A5)
Cobalt-chromium (RMO) (C1) 353.7
um-molybdenum 0.019 × 0.025 inch (Ormco), which did 0.016 × 0.022 (C1)
not show significant difference. Cobalt-chromium (RMO) (C2) 261.9
• The smallest total average load was shown for the titani- 0.016 × 0.016 (C2)
um-molybdenum 0.017 × 0.025 inch (Ormco). Tables VII TMA (Ormco) (T1) 249.6
and VIII show the results. 0.019 × 0.025 (T1)
TMA (Ormco) (T2) 149.4
Activation—Effect on Load 0.017 × 0.025 (T2)

Total average loads related to showing activations

are significantly different (Tables IX and X).
Estimated Coefficient of Spring Rate, Standard
Error, Superior and Inferior Limits
Based on the square minimum method, it was pos-
sible to adjust the regression equation to the nine dif-
ferent spring groups. Afterward, the confidence interval
was constructed for the spring rate (β) on the confi-
dence level of 95%.
The elastic behavior of the studied springs is shown
graphically in Fig 5. The modulus of elasticity (E) is
given by the steepness of the elastic curves. It repre-
sents the measure of the stiffness of a wire.2,13
280 Ferreira American Journal of Orthodontics and Dentofacial Orthopedics
March 1999

Table VIII. Comparison, difference between arithmetic Table IX. Activation and total arithmetic average (g)
average, and critical difference (5%) Activation (mm) Total arithmetic average (g)
Difference between Critical
Comparison arithmetic average (g) difference (5%) 3.0 (D) 585.2
2.5 (E) 484.7
A1 × A2 30.69 24.26* 2.0 (F) 385.3
A1 × A4 37.96 24.26* 1.5 (G) 287.5
A1 × A3 48.46 24.26* 1.0 (H) 194.7
A1 × C1 126.47 22.98* 0.5 (I) 105.8
A1 × A5 130.61 25.17*
A1 × C2 218.26 20.70*
A1 × T1 230.55 24.26* Table X. Comparison and average difference (5%)
A1 × T2 330.72 25.17*
A2 × A4 7.26 23.30 Comparison Average differences (5%)
A2 × A3 17.76 23.30
A2 × C1 95.78 21.97* D×E 100.47*
A2 × A5 99.92 24.26* D×F 199.89*
A2 × T1 199.86 23.30* D×G 279.69*
A2 × T2 300.03 24.26* D×H 390.49*
A3 × C1 78.02 21.97* D×I 479.40*
A3 × A5 82.15 24.26* E×F 99.42*
A3 × C2 169.81 19.58* E×G 197.22*
A3 × T1 182.10 23.30* E×H 290.01*
A3 × T2 282.27 24.26* E×I 378.93*
A4 × A3 10.50 23.30 F×G 97.81*
A4 × C1 88.52 21.97* F×H 190.60*
A4 × A5 92.65 24.26* F×I 279.51*
A4 × C2 180.31 19.58* G×H 92.79*
A4 × T1 192.60 23.30* G×I 181.71*
A4 × T2 292.77 24.26* H×I 88.92*
C1 × A5 4.14 22.98 *Statistical significant difference by Tukey-Kramer test (5%).
C1 × C2 91.79 17.97*
C1 × T1 104.08 21.97*
C1 × T2 204.25 22.98*
properties,2,6 and also by making use of nomograms14
A5 × C2 87.65 20.70*
A5 × T1 99.94 24.26* and equations comparing elastic properties rate.15,16
A5 × T2 200.11 25.17* The present study proposes to compare orthodontic
C2 × T1 12.29 19.58 retraction springs with the same design (double delta
C2 × T2 112.46 20.70* loop), different cross-sections, and different wire mate-
T1 × T2 100.17 24.26*
rials, to determine the spring rate and the achieved
*Statistical significant difference by Tukey-Kramer test (5%). average load in different activations.
It is advisable that cobalt-chromium wires undergo
heating treatment at 482oC from 7 to 12 minutes in a
The stainless steel group of springs 0.019 × 0.025 heat-treating unit, RMO adapting cables, a dental fur-
inch (3M/Unitek) has shown a higher estimated value nace, a match or blush flame, until the alloy reaches an
(β = 275.9 g/mm). That means that about 275.9 g/mm orange color.17 In this work, an uncommon but effi-
are stored on each millimeter of activation or converse- cient type of heating treatment was used. Cobalt-
ly 275.9 g/mm are released for every millimeter of chromium and stainless steel springs have reached sat-
deactivation. The smallest spring rate (β = 84.9 g/mm) isfactory elastic behavior (Fig 5).
was found in the titanium-molybdenum 0.017 × 0.025 Although the present study has not considered the
inch (Ormco) spring group. An intermediate value was vertical forces and the effects of the alpha-beta
found in the cobalt-chromium group of springs. The moments, it is possible to have an idea of the influence
results are summarized in Table XI. of the cross-section and the wire material over the
achieved loads in the horizontal forces.
DISCUSSION The cross-section, wire material, and activation are
Because new alternatives in alloys are available in very important variables related to the resultant average
the market, it is important to consider a precise criteri- loads after the activations. Even though this investiga-
on to choose such alloys for clinical use. Some authors tion only considers springs having the same design, this
have studied these alloys through their mechanical is an essential factor in relation to spring choice.2,5,6,18
American Journal of Orthodontics and Dentofacial Orthopedics Ferreira 281
Volume 115, Number 3

Table XI. Spring rate, standard error, inferior limit, superior limit, and confidence interval (95%)
Spring groups Spring rate Standard Inferior Superior
(inch) (g/mm) error limit (g) limit (g) R2

Stainless steel (3M/Unitek) 275.9 1.8 272.1 279.7 0.9984

0.019 × 0.025
Stainless steel (Morest) 257.5 3.5 250.3 264.6 0.9964
0.019 × 0.025
Stainless steel (3M/Unitek) 248.6 3.4 241.7 255.5 0.9964
0.018 × 0.025
Stainless steel (Morest) 249.5 2.1 245.1 253.9 0.9964
0.018 × 0.025
Stainless steel (3M/Unitek) 197.6 1.9 193.5 201.7 0.9964
0.017 × 0.022
Cobalt-Chromium (RMO) 201.1 2.5 195.9 206.3 0.9964
0.016 × 0.022
Cobalt-Chromium (RMO) 150.5 1.9 146.6 154.3 0.9964
0.016 × 0.016
Titanium-molybdenum (Ormco) 138.3 2.6 133.0 143.6 0.9964
0.019 × 0.025
Titanium-molybdenum (Ormco) 84.9 1.8 81.1 88.6 0.9964
0.017 × 0.025

According to Burstone et al,2 several factors might
influence the spring design, such as the mechanical
properties of the spring alloy, cross-section, and linear
configuration.
The optimal force levels have been extensively stud-
ied in order to obtain an efficient dental movement that
is fast and with a minimum of tissue damage and dis-
comfort. Smith and Storey7 found optimal forces vary-
ing from 150 to 200 g during the translation of canines,
while Reitan8 says that the amount of force applied
depends on the kind of movement required; in this way,
a force of 250 g would be required for bodily movement,
but for extrusion movements 25 g would be enough.
Neto et al19 tested stainless steel 0.019 × 0.025 inch
springs (American Orthodontics) with a tear drop loop
design and found that 80% of the springs has a load of
1.000 g when submitted to activation of 1.25 mm. The
spring rate was 800 g/mm. In the current study, the
stainless steel springs with the same cross-sections
(0.019 × 0.025 inch), showed smaller loads when acti-
vated at 1.5 mm, 379.5 g (Morest) and 403 g
(3M/Unitek), and a spring rate of 257.5 g/mm and
275.9 g/mm, respectively. The loop design certainly
has some influence in the resultant loads.
Gjessing6 developed a spring in stainless steel
0.016 × 0.022 inch (RMO) with a designed ovoid dou-
ble helix loop showing 160 g when the double helix are
separated 1 mm. The spring rate was 45 g/mm. The
present study did not test the behavior of the stainless
steel springs with cross-section studied by Gjessing.6
However, cobalt-chromium with 0.016 × 0.022 inch
(RMO) were tested. According to Thurow,13 cobalt-
chromium and stainless steel have similar stiffness. In
the present study, it was found that cobalt-chromium
with 0.016 × 0.022 inch (RMO) presented 203 g when
activated 1.0 mm and a spring rate of 201.1 g/mm.
Comparing these different wire material springs, it can
be concluded that Gjessing’s spring design would rep-
resent a better choice for releasing a lower rate decay.
Leão18 studied the behavior of several spring
designs in tension tests, all in stainless steel material
with 0.019 × 0.026 inch and 0.019 × 0.025 inch cross-
sections, with and without heat treatment. In his study,
it was found that at 1.0 mm activation, the heat treated
0.019 × 0.025 inch (3/M Unitek) with the T designed
springs presented 405 g, with reversal loop 606 g, with
helicoidal loop 893.6 g, and with tear drop loop 963.1
g. The performances of the 0.019 × 0.025 inch
(3M/Unitek) and 0.019 × 0.025 inch (Morest), report-
ed in the present investigation, show a smaller average
load (266 g and 253 g, respectively) when submitted to
activation at 1.0 mm.
Ricketts9 advises that sectional springs cobalt-
chromium 0.016 × 0.016 inch (RMO) should produce
100 to 150 g when activated 2 to 3 mm.
Bench et al20 verifying cobalt-chromium 0.016 ×
0.016 inch (RMO) springs on a double vertical helical
closing loop design, 60 mm long, found that 2 to 3 mm
of activation are necessary to get a load of 100 to 150 g.
The spring rate was 75 g/mm.
In the current study, the cobalt-chromium 0.016 ×
0.016 inch (RMO) springs had average loads of 297.8
g and 456.1 g when activated at 2 to 3 mm, respective-
ly. The spring rate was 150.5 g/mm; therefore, the
282 Ferreira
design tested by Bench et al20 would be chosen because
of its superior spring rate.
Burstone2,6,10 carefully investigated titanium-
molybdenum (Ormco) springs. In 1982, the author
tried a composite TMA 0.018-0.017 × 0.025 inch
(Ormco) retraction spring with a T loop design (the
round cross-section is attached to an active member
and the rectangular to the anchorage member) achiev-
ing an excellent spring rate of 33 g/mm. By activating
the spring at 6 mm, just 201 g was stored. Later, in
1995, the author investigated titanium-molybdenum
0.017 × 0.025 inch (Ormco) springs, 17 mm long, with
a T loop design. The T centered loop springs stored
173.1 g when activated 3.0 mm and a spring rate of
approximately 55 g/mm. In the present study, the tita-
nium-molybdenum 0.017 × 0.025 inch (Ormco)
springs stored 257 g when activated 3.0 mm and a
spring rate of 84.9 g/mm. Comparing such outcomes
with Burstone’s (composite TMA 0.018-0.017 ×
0.025 inch and 0.017 × 0.025 inch T centered loop), it
is important to consider the loop design and cross-sec-
tion wire.
The results of this study comparing the three differ-
ent alloys are not surprising. The titanium-molybde-
num alloys were expected to present the smallest
spring rate; however, the statistical analysis has shown
that the cobalt-chromium spring group 0.016 × 0.016
inch did not show significant difference in relation to
the titanium-molybdenum spring group 0.019 × 0.025
inch, and the same applies to stainless steel 0.017 ×
0.022 inch and 0.016 × 0.022 inch groups. In this way,
the orthodontist might have another option if the first
one is not available.
CONCLUSIONS
This study tested the wire material and cross-sec-
tion effect of orthodontic spring retraction with double
delta design in relation to the average load and spring
rate achieved after several activations. The following
conclusions were reached:
1. The titanium-molybdenum 0.017 × 0.025 inch (Ormco)
springs displayed the lowest average load for each test-
ed 0.5 mm of activation and also the lowest spring rate.
American Journal of Orthodontics and Dentofacial Orthopedics
March 1999
2. The stainless steel 0.019 × 0.025 inch (3M/Unitek)
springs showed a higher average load for each tested
0.5 mm activation and the highest spring rate as well.
3. The stainless steel 0.017 × 0.022 inch (3M/Unitek) and
cobalt-chromium 0.016 × 0.022 inch (RMO) springs did
not show any significant difference between each other.
4. The cobalt-chromium 0.016 × 0.016 inch (RMO) and
titanium-molybdenum 0.019 × 0.025 inch (Ormco)
springs did not show any significant difference between
each other.
5. The stainless steel 0.018 × 0.025 inch (3M/Unitek),
0.018 × 0.025 (Morest) and 0.019 × 0.025 inch (Mor-
est) springs did not show any significant difference be-
tween each other.
6. The spring rate is dependent on wire material, cross-
section, and spring design.
REFERENCES
1. Braun S, Marcotte MR. Rationale of the segmented approach to orthodontic treatment.
Am J Orthod Dentofacial Orthop 1995;108:1-8.
2. Burstone CJ, Baldwin JJ, Lawless DT. The application of continuous forces to ortho-
dontics. Angle Orthod 1961;31:1-14.
3. Burstone CJ, Goldberg AJ. Beta titanium: a new orthodontic alloy. Am J Orthod
1980;77:121-32.
4. Burstone CJ. The segmented arch approach to space closure. Am J Orthod
1982;82:361-78.
5. Burstone CJ, Steenbergen E, Hanley K. Modern edgewise mechanics and the
segmented arch technique. Farmington: University of Connecticut Health Center,
1995.
6. Gjessing P. Biomechanical design and clinical evaluation of a new canine retraction
spring. Am J Orthod 1985;87:353-62.
7. Smith R, Storey E. The importance of force in orthodontics. Aust J Dent 1952;56:291-
304.
8. Reitan K. Some factors determining the evaluation of forces in orthodontics. Am J
Orthod 1957;43:32-45.
9. Ricketts RM. Bioprogressive therapy as an answer to orthodontic needs: part II. Am J
Orthod 1976;70:241-68.
10. Goldberg J, Burstone CJ. An evaluation of beta titanium alloys for use in orthodontic
appliance. J Dent Res 1979;58:593-600.
11. Goldberg J, Burstone CJ. Status report on beta titanium orthodontic wires. J Am Dent
Assoc 1982;105:684-5.
12. Kirk, RE. Experimental design: procedures for the behavioral sciences. 2nd edition.
Pacific Grove, CA: Books:Cole Publishing Company, 1990. p. 456-66
13. Thurow RC. Edgewise orthodontics. 3rd edition. St Louis: Mosby; 1972. p. 51-5 and
213-23.
14. Kusy RP. On the use of nomograms to determine elastic property of orthodontic arch
wires. Am J Orthod 1983;83:374-81.
15. Kusy RP. Comparison of nickel-titanium and beta titanium wire sizes to conventional
arch wire materials. Am J Orthod 1981;79:625-9.
16. Kusy RP. Comparison of the elastic properties of nickel-titanium and beta titanium
arch wires. Am J Orthod 1982;82:199-205.
17. Orthodontic wires. Denver: Rocky Mountain Dental Products Company; 1990.
18. Leão LAT. Avaliação de molas de fechamento de espaços em ortodontia:ensaios em
laboratório.Tése. Rio de Janeiro: UFRJ, 1992. (Tese de mestrado).
19. Neto SP, Mucha JN, Chevitarese O. Mola de fechamento de espaços em ortodontia em
forma de “lágrima”: desempenho em tração. RBO 1984;5:10-4.
20. Bench RW, Gugino CF, Hilgers JJ. Bioprogressive therapy. JCO 1978;12:123-39.