titanium-niobium

M DalstraG DenesB Melsen

Authors’ affiliations:Michel Dalstra, Gabriella Denes, Birte Melsen,

Department of Orthodontics, Royal Dental

College, University of Aarhus, Denmark

Correspondence to:Assoc. Prof. Michel DalstraDepartment of Orthodontics

Royal Dental College, University of Aarhus

Vennelyst Boulevard 9

8000 Aarhus C

Denmark

Tel: +45 8 942 4037

Fax: +45 8 619 6029

E-mail: [email protected]

Dates:

Accepted 20 September 1999

To cite this article:

Clin. Orthod. Res. 3, 2000; 6–14

Dalstra M, Denes G, Melsen B:

Titanium-niobium, a new finishing wire alloy

Copyright © Munksgaard 2000

ISSN 1397-5927

Titanium-niobium, a newfinishing wire alloy

Abstract: The mechanical properties of the newly introduced

titanium-niobium finishing wires were investigated. Both in

bending and torsional loading mode, the stiffness, yield point,

post-yield behavior, and springback of titanium-niobium wires

were experimentally determined and compared to those of

equally sized stainless steel wires. The experimentally obtained

values were also validated with theoretical values from

engineering formulas of cantilever deformations. The ratios for

these parameters for the two materials proved to be different

in bending and torsion. The stiffness of titanium-niobium in

bending is roughly half of that of stainless steel, whereas in

torsion it is roughly one-third. These characteristics enable the

clinician to use titanium-niobium for creative bends without the

excessive force levels of steel wires. The springback of

titanium-niobium in bending is 14% lower than that of steel,

whereas in torsion it is about the same or even slightly higher

than that of steel, thus making it possible to utilize the wire for

even major third-order corrections. Finally, the weldability of

titanium-niobium wires was found to be good, so it is possible

to weld wires of different dimensions together for the

generation of differentiated force systems.

Key words: force system; orthodontic wires; titanium-niobium;

stainless steel; weldability

Introduction

Since the introduction of the first new orthodontic alloy

in the early 1970s (1, 2), a wide range of new alloy wires

have become available to the clinician (3). The selection of

a wire is no longer a trivial routine, but has to be based on

a clear definition of the specific task to be performed by

the wire. New wires are often presented by the manufac-

tures with only the sales pitch of indications of possible

clinical use. For the clinician to judge a new wire ratio-

nally, information on its physical properties must be

Dalstra et al. Titanium-niobium, a new finishing wire alloy

provided. Burstone (4) suggested that the use of stainlesssteel (SS) as a standard would facilitate the comparison ofstiffness between different wires. The selection of a wire isnot only based on stiffness, but also information onspringback, formability, joinability, and temperature-de-pendency variables. It is furthermore important to realizethat wires may react differently in bending and torsion,and that the mode of loading may affect these properties(5).

Recently, a new ‘finishing wire’ made from a nickel-freetitanium-niobium alloy (TiNb) was introduced (TitaniumNiobium/FATM, Sybron Dental Specialities Inc., Orange,CA, USA). According to the manufacturer’s productinformation, TiNb is soft and easy to form, yet it has thesame working range as SS. Its stiffness is 20% lower thanTMA and 70% lower than SS. The purpose of this articleis to present the material properties of the TiNb wire. Thespecific objectives were: 1) to evaluate the load-deflectionbehavior of TiNb wires compared to SS in both bendingand torsional mode, 2) to assess its springback, 3) to relatethese variables to the mode of activation, 4) to evaluateintra-batch variation, and finally 5) to analyze weldingproperties.

Material and method

The wires evaluated were TiNb and their properties werecompared to conventional SS arch wires, both with a

nominal cross section of 0.017×0.025%%. All productswere manufactured by Sybron Dental Specialities Inc.(Orange, CA, USA) and were taken from the samebatches to exclude inter-batch variations.

All wires were tested in a PC-controlled orthodonticwire-testing device, also known as the force system iden-tification (FSI) system, developed at the Department ofOrthodontics of the Royal Dental College at the Univer-sity of Aarhus (6). In this device, an orthodontic wire canbe fixed between two wire holders, in which mechanicalsensors are placed. The moments and forces generatedbetween the wire and the holders are transformed intoelectrical impulses by means of specially developed straingauges. In the FSI system, six step motors control thetranslation and the rotation of the two holders. Theprecision of the machine has been analyzed and describedby Menghi et al. (7). The displacements of the holders arecomputer controlled, and input for these are supplied bythe user. Data of the moments and forces, together withthe positions of the holders, are stored in the computerfor further statistical analysis.

The bending tests were performed as cantilever bend-ing, which meant that only one end of the wire was fixedin one of the holders of the FSI system. The other endwas kept in one-point contact with a special support pinattached to the other holder (Fig. 1). In the case oftorsion, the wire was clamped tightly in both holders,thus eliminating the influence of play between bracket andwire.

Two different modes for bending and torsion wereemployed in this study. For the first bending mode (A),five wires with an effective length of 12.5 mm (10 mmfrom the holder to the support pin and 2.5 mm into theholder before being fixed by a fixating screw) were bentfrom 0 to 60° and back. In the second mode (B), anotherfive wires were bent from 0 to 30° and back, and then 30°in the opposite direction and back again. For the secondpart of this test, moments and forces were zeroed again at0°. For both bending modes, the forces and moments atthe sensors were determined in increments of 3°.

For torsional testing, a similar approach was employed,but here the final angles were 90 and 45°, respectively, andthe wires had an effective length of 12 mm (7 mm betweenthe two holders and two times 2.5 mm into the holdersbefore being fixed by the fixating screws).

Based on the measured data on the bending and tor-sional moments versus the bending and torsional angles,respectively, the following parameters were determined: 1)

Fig. 1. Experimental set-up for the cantilever bending mode. It can beseen that the wire is already bent about 30°.

Clin Orthod Res 3, 2000/6–142 7


the stiffness, as the slope of the curve calculated at 3°deformation; 2) the yield point (My and uy), as the pointat which a permanent deformation of 1° was reached; 3)the maximal moment (either the ultimate moment (Mult)or the moment at the end of the deformation range); and4) the springback. In case of the testing modes B, wherethe wire is first bent or twisted in one then in theopposite direction, a relative measure of the hysteresiswas considered by looking at the stiffness, maximal mo-ment, and the springback in the second loading leg as apercentage of the respective parameters in the first load-ing leg.

In order to test the validity of the measurements andto calculate the Young’s and shear moduli of the wirematerial, the following theoretical considerations weremade. For cantilever bending and torsion, the respectivestiffness of a wire are given by:�M

u

�bending

=2·E·Ix

Land

�Mu

�torsion

=G·Ip

L

where M is the applied moment, u the angular deforma-tion, E the material’s Young’s modulus, Ix the momentof inertia, L the wire’s length, G the material’s shearmodulus, and Ip the apparent polar moment of inertia.For a wire with a rectangular cross section with a longside of 2a and a short side of 2b, Ix and Ip are given by:

Ix=43

ab3 and Ip=ab3 �163

−3.36ba�

1−b4

12a4

��Assuming a theoretical Young’s modulus and shear

modulus for SS of 200 and 75 GPa, respectively, andsubstituting the known dimensions of the wires (a=0.317 mm, b=0.216, Lbending=12.5 mm, and Ltorsion=12.0 mm), we find the following stiffness for bending andtorsion:�M

u

�bending

=237.9cNmm

deg and�M

u

�torsion

=107.9cNmm

deg

Comparing these respective stiffness values with theexperimentally measured stiffness of the TiNb and SSwires, it is possible to extract the actual values for theYoung’s and shear moduli of the two materials.

In order to test the welding stability of TiNb, speci-mens were prepared such that the two straight pieces ofwire were welded perpendicular to one another. A spotwelding was performed on a Rocky Mountain model506A. The heat selector regulating the power input wasset in position 1 and the voltage regulating the amount

of stored energy was set at position 5. The clampingpressure was high. The flat sides of the electrodes wereused when bringing the perpendicularly situated arch-wires in contact. The specimens were then either bent(both legs 90° each) or twisted (one leg two times aroundits own longitudinal axis) with a pair of pliers to seewhether or not fracture would occur at the weld.

The statistical analyses was performed using SPSS(SPSS Inc., Chicago, IL, USA). Both for bending andtorsion, stiffness and yield point data were pooled forboth testing modes (A and B). Post-yield data (maximalmoment and springback) and the respective reductions instiffness, maximal moment, and springback were onlyconsidered for their respective loading modes (A or B).An independent samples t-test was used to analyze thedifferences between TiNb and SS, and statistical signifi-cance was assumed for pB0.05.

Results

The results of the bending tests based on the averages ofthe five respective specimens are illustrated in the graphsin Figs. 2 and 3. The values of the relevant parameterstaken from these curves are given in Table 1. The bend-ing stiffness of TiNb is 48% that of SS. Yield occursslightly earlier for the TiNb wires (13 vs. 17°) and theyield moment is 36% that of SS. The ultimate bendingmoment is reached on average at 31°, while for SS wiresthis first occurs at 35°. This means that both the elasticand plastic working ranges of TiNb in bending areslightly smaller than for SS. The ultimate moment forthe TiNb wires is about 38% of that of the SS wires andthe springback for the TiNb wires is significantly lowerthan for the SS wires (25 vs. 29°). Similar findings areobserved for the first part of the curves for bendingmode B. Looking at the stiffness, maximal moment, andthe springback in the second part of the curves as apercentage of the respective values in the first part of thecurves, then the percentages are lower for TiNb than SS(stiffness: 64 vs. 74%; maximal moment: 59 vs. 68%;springback: 63 vs. 73%). It must be noted, however, thatthe reduction in stiffness is not statistically different.This means that in bending, TiNb is more sensitive toprevious deformations than SS.

The results of the torsion tests based on the averagesof the five respective specimens are shown in the graphs

2Clin Orthod Res 3, 2000/6–148


Fig. 2. Moment versus angle for bending mode A (60°) for both SS and TiNb wires, based on the averages of five specimens. Error bars representthe SDs for the five specimens.

in Figs. 4 and 5. The relevant parameters corresponding tothe graphs are given in Table 2. The torsional stiffness ofTiNb is 36% that of SS. In contrast to bending, yieldoccurs both for the TiNb and SS wires at around 25°.

This finding implies that the elastic working range is thesame for both metals, although the yield moment of TiNbis only 38% of that of SS. At 90°, the maximal momentfor the TiNb wires lies at 33% of that of SS wires. Both

Fig. 3. Moment versus angle for bending mode B (30 and −30°) for both SS and TiNb wires, based on the averages of five specimens. Error barsrepresent the SDs for the five specimens.



Table 1. Average values and SDs of the relevant parame-ters from the bending tests

Bending

Mode TiNb SS

Stiffness (cNmm/deg) A+B 115.995.3 243.296.3

Young’s modulus (GPa) A+B 97.494.4 204.595.3

A+B 1 4239327My (cNmm) 3 9609769

uy (deg) A+B 13.392.9 17.393.3

Mult (cNmm) A 2 177958 5 770957

A 31.291.6uult (deg) 34.891.6

A 24.791.0 28.590.3Springback (deg)

B 64.494.4 73.698.0Reduction in stiffness (%)

Reduction in maximal moment (%) B 59.393.5 67.992.9

Reduction in springback (%) B 63.194.3 73.493.7

All parameters are significantly different (pB0.05) between TiNb and SS, exceptthe reduction in stiffness.

cally significant. Similar findings are observed for thefirst part of the curves for torsion mode B. An exami-nation of the stiffness in the second part of the curvesas a percentage of the respective values in the first partof the curves reveals this percentage as nearly equal forTiNb and SS (84 vs. 86%). The percent values for themaximal moment and the springback, however, are con-siderably lower for TiNb than for SS (maximal moment:60 vs. 87%; springback: 64 vs. 86%). This finding sug-gests that in torsion too TiNb is more sensitive to pre-vious deformations than SS.

Comparing the theoretically determined bending andtorsional stiffness of SS (238 and 108 cNmm/deg) withthe values in Tables 1 and 2, it can be seen that theseare indeed within the ballpark of experimental values.The mean and SDs of these Young’s and shear moduliare given in Tables 1 and 2. Additionally, the weldabil-ity of TiNb was found to be good. The applied defor-mations to the welding specimens did not cause theweld to give away (Fig. 6).

Discussion

Orthodontic wires can be classified into three categories:1) wires characterized by a large plastic range, similar

for TiNb and for SS, the ultimate moment was notreached at 90° torsion. TiNb has a springback of 56° asopposed to a value of 54° for SS. Although this differ-ence between both materials is not large, it is statisti-

Fig. 4. Moment versus angle for torsion mode A (90°) for both SS and TiNb wires, based on the averages of five specimens. Error bars representthe SDs for the five specimens.



Fig. 5. Moment versus angle for torsion mode B (45 and −45°) for both SS and TiNb wires, based on the averages of five specimens. Error barsrepresent the SDs for the five specimens.

activation and deactivation curves, and relatively lowspringback, (e.g. SS); 2) nickel-titanium wires without aclearly defined yield point, different activation and deacti-vation curves, and medium-range springback (e.g.,Nitinol); and finally 3) superelastic, temperature-sensitive

wires. The new TiNb wire belongs to the first category. It

can be easily bent and therefore utilized to generate a

predetermined force-driven force system.

The TiNb wire was introduced as a finishing wire.

According to the manufacturer’s specifications, the stiff-

ness of a TiNb wire is 20% less than TMA and 70% less

than SS. Interestingly, the actual loading mode to which

this stiffness applies (bending, torsion) is not further

specified. The results of the present study demonstrate

that the latter claim does hold true for the torsion stiff-

ness, while in bending the stiffness of TiNb is almost 50%

of that of SS. Although TMA wires were not tested in

this study, the reduced stiffness of TiNb in comparison to

TMA seems, at least in bending, unlikely. Burstone (4)

reports a relative stiffness of TMA of 42% of that of SS,

and with the present results this would imply that the

bending stiffness of TiNb is actually 14% higher than

TMA, rather than 20% lower.

Within the parameters of this study, the mean stiffness

value for SS in bending (243.2 cNmm/deg) was slightly

higher than the theoretically calculated one for a 12.5 mm

long 0.017×0.025%% SS wire (237.9 cNmm/deg). These

differences will have to be attributed to the slight geomet-

rical variations within the specimens tested. The speci-

men’s length is the obvious culprit here and the difference

Table 2. Average values and SDs of the relevant parame-ters from the torsion tests

Torsion

SSTiNbMode

Stiffness (cNmm/deg) A+B 39.793.3 109.197.9

27.692.3 75.995.5A+BShear modulus (GPa)

My (cNmm) A+B 9819317 2 5659172

24.792.825.697.2A+Buy (deg)

1 8189160A 5 550995M90 (cNmm)

Springback (deg) 54.491.1A 56.390.5

Reduction in stiffness (%) B 84.493.1 86.492.0

Reduction in maximal moment (%) 60.294.1B 86.595.0

BReduction in springback (%) 64.293.6 85.894.1

All parameters are significantly different (pB0.05) between TiNb and SS, exceptthe yield angle and the reduction in stiffness.



Fig. 6. Examples of an unloaded (left), a bent (middle), and a twisted welding specimen (right).

between 237.9 and 243.2 cNmm/deg would imply a differ-

ence in wire length of less than 0.3 mm. The yield

moment in bending for SS found in this study (39609769 cNmm) is somewhat lower than the value (43509200

gmm) reported by Burstone and Goldberg (8). This might

be due to the strong dependence in the determination of

the yield point on the initial slope of the loading curve.

Slight variations here already cause substantial changes in

the exact location of the yield point. The ultimate bending

moment shows a better comparison: 5770957 vs.

5750925 gmm. The accompanying yield angle found in

the present study (17.3°) compares well to the 18.0°

reported by Burstone and Goldberg, while the angle for

the ultimate bending moment is a little higher in our

study (34.8 vs. 25.5°).

With a bending stiffness corresponding to 48% of that

of SS and a springback 14% lower than that of steel, the

clinician can easily make creative bends and avoid the

excessive force levels of a steel wire. In the B mode, the

TiNb wire loses relatively more stiffness, maximal mo-

ment, and springback in the second loading leg, indicating

that the TiNb wire behaves more plastically than a corre-

sponding SS wire.

The mean value of the stiffness of SS wires in torsion

(109.1 cNmm/deg) comes very close to the theoretically

calculated one for a 12 mm long 0.017×0.025%% SS wire

(107.9 cNmm/deg). As discussed for the bending behav-

ior, any disparities here will have to be attributed to

geometrical variations in the specimens. In comparing the

results of torsion of the steel wires to those reported by

Larson et al. (9), it can be seen that the measured shear

modulus in the present study (75.9 GPa) is higher than

theirs (67.9 GPa). However, they admit that their value is

actually 10% under the theoretical value of around 75

GPa due to clamping errors.

It is important to note that the stiffness of TiNb intorsion is only 36% of steel and that the maximal momentwas at the same level of that exhibited for bending. Yet,the springback of TiNb in torsional mode is slightlyhigher than SS. This characteristic makes it possible toutilize the TiNb wire for even major third-order correc-tions. When the torsion was tested in the B mode, asignificant decrease was observed in both the springbackand in the maximal moment. This property could beuseful in the event the clinician would like to reduce theactive forces during a third-order correction.

Like beta-titanium (10, 11), the weldability of TiNb isquite good. It takes considerable deformation before theweld gives away (Fig. 6). It would be possible, therefore,to join different dimensions of this wire to create adifferential force system.

Abstrakt

Es wurden die mechanischen Eigenschaften der neu vorgestellten Tita-nium-Niobium-Finishing-Drahte untersucht. Fur Biegung und Torsionwurden experimentell Elastizitat, Elastizitatsgrenze, Verhalten jenseitsder Elastizitatsgrenze und Ruckstelikraft bestimmt und mitStahidrahten gleicher Dimension verglichen. Die experimentell ermit-telten Daten wurden weiterhin mit theoretischen, aus mathematisch-physikalischen Formeln ermittelbaren Werten verglichen. Die Wertedieser Parameter war fur beide Materialien bei Biegung und Torsionunterschiedlich. Die Biegelastizitat von Titanium-Nioblum ist ungefahrhalb so hoch wie die von Stahl, die Torsionselastizitat hingegen istungefahr ein Drittel so hoch. Diese Eigenschaften ermoglichen es demKliniker, Titanium-Nioblum fur Detailbiegungen zu verwenden, ohnedie sehr hohen Kraftniveaus von Stahidrahten zu erreichen. Die Ruck-stelikraft von Titanium-Niobium ist bei Biegungen 14% niedriger alsdie von Stahi, wahrend bei Torsionen die Ruckstelikraft genau sogroß oder sogar leicht großer ist als bei Stahidrahten. Diese Eigen-schaft ermoglicht es, diesen neuen Draht auch fur großere Torque-Ko-rrekturen einzusetzen. Schlußendlich erwies sich auch dieSchweillbarkeit als gut, so daß es moglich ist, unterschiedliche Draht-starken zur Erzeugung differenzierter Kraftsysteme zusammenzufugen.



Resumen

Las propiedades mecanicas del alambre de titanio niobio recientementeintroducido pare terminaciones en tratamientos, fueron investigadas.Tanto en el doblez como en el modo de carga torsional, la rigidez, elpunto de elasticidad, el comportamiento de elasticidad posterior, y elefecto de resorte a su punto original de los alambres de titanio-nobio,fueron determinados experimentalmente y comparados con los alam-bres de acero inoxidable de igual tamano. Los valves obtenidos experi-mentalmente fueron tambien validados con valves teoricos deformulas de ingenierıa de deformaciones de cantilever. Las propor-ciones de estos parametros pare estos dos materiales probaron serdiferentes en doblez y torsion. La rigidez del titanio-niobio en eldoblez es casi la mitad del de acero inoxidable, mientras que entorsion es mas o menos un tercio. Estas caracterısticas permiten alclınico utilizar el titanio-niobio pare dobleces creativos sin niveles defuerza excesiva de los alambres de acero. El efecto de resorte a supunto original del titanio-niobio en los dobleces es 14 porciento masbajo que el de acero, mientras que en torsion es mas o menos o aunun poco mas alto que el de acero, haciendo posible que se utilice elalambre haste pare correcciones mayores del tercer orden. Finalmente,la habilidad pare soldadura de los alambres de titanio-niobio fue cla-sificada como buena. Por ende, es posible solder alambres de dimen-siones diferentes pare la generacion de sistemas de fuerza diferenciada.



Structured Abstract

Authors – Dalstra M, Denes G, Melsen B.Objectives – To compare the mechanical properties in bending andtorsion of the newly introduced titanium-niobium wires to stainlesssteel wires and to assess the weldability of titanium-niobium.Design – Experimental force system identification testing machine.Setting and Sample Studied – Laboratories of the Department ofOrthodontics, Royal Dental College, Aarhus University, Denmark.The materials tested were 0.017×0.025%% titanium-niobium andstainless steel wires.Experimental Variable – Titanium-niobium ver-sus stainless steel.Outcome Measures – Stiffness, yield point, post-yield behavior, andspringback both in bending and torsion.Results – Titanium-niobium has a bending stiffness of 48% and atorsional stiffness of 36% of that of stainless steel. The springbackof titanium-niobium is 14% less than that of stainless steel inbending, whereas it is comparable or even slightly higher than thatof stainless steel in torsion. Intra-batch variations of these proper-ties were found to be very small. The weldability of titanium-nio-bium is very good.Conclusion – With a considerably lower stiffness, yet a comparablespringback of stainless steel, titanium-niobium can be used formoderate predetermined force-driven force systems.

Clinical Orthodontics and Research 3, 2000; 6–14Copyright © Munksgaard 2000, ISSN 1397-5927

References

1. Andreasen GF, Hilleman TB. An evaluation of 55 cobalt substi-tuted Nitinol wire for use in orthodontics. JADA 1971;82:1373–5.

2. Andreasen GF, Morrow RE. Laboratory and clinical analyses ofNitinol wire. Am J Orthod Dentofac Orthop 1978;73:142–51.

3. Kusy RP. A review of contemporary archwires: their propertiesand characteristics. Angle Orthod 1997;67:197–208.

4. Burstone CJ. Variable-modulus orthodontics. Am J Orthod Dento-fac Orthop 1981;80:1–16.

5. Lopez I, Goldberg J, Burstone CJ. Bending characteristics ofNitinol wire. Am J Orthod Dentofac Orthop 1979;75:569–71.

6. Leth KB, Melsen HM, Frederiksen MAA, Godtfredsen E, An-dersen KL, Melsen B. Force system identification. In: 68th EOSCongress, Venice, Italy. 1992.

7. Menghi C, Planert J, Melsen B. 3-D experimental identification offorce systems from orthodontic loops activated for first ordercorrections. Angle Orthod 1999;69:49–57.

8. Burstone CJ, Goldberg AJ. Maximum forces and deflections fromorthodontic appliances. Am J Orthod Dentofac Orthop 1983;84:95–103.

9. Larson BE, Kusy RP, Whitley JQ. Torsional elastic property mea-surements of selected orthodontic archwires. Clin Mater1987;2:165–79.

10. Burstone CJ. Welding of TMA wire – clinical applications. J ClinOrthod 1987;21:609–15.

11. Nelson KR, Burstone CJ, Goldberg AJ. Optimal welding of betatitanium orthodontic wires. Am J Orthod Dentofac Orthop1987;92:213–9.


titanium-niobium

Documents

Transcript of titanium-niobium