J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/jmbbm

Research paper

The influence of heat treatment and role of boron on slidingwear behaviour of β-type Ti–35Nb–7.2Zr–5.7Ta alloy in drycondition and in simulated body fluids

P. Majumdar∗, S.B. Singh, M. Chakraborty1

Department of Metallurgical and Materials Engineering, Indian Institute of Technology, Kharagpur, India

A R T I C L E I N F O

Article history:

Received 10 May 2010

Received in revised form

15 October 2010

Accepted 23 October 2010

Published online 30 October 2010

Keywords:

Titanium alloys

Scanning electron microscopy

Hardness

Wear

A B S T R A C T

The wear behaviour of heat-treated Ti–35Nb–7.2Zr–5.7Ta (TNZT) and Ti–35Nb–7.2Zr–5.7Ta-

0.5B (TNZTB) alloys (all compositions are in wt%) was investigated in dry condition and

in simulated body fluids. It has been found that there is no straightforward relationship

between the wear rate and the microstructure. The hardness has no appreciable effect on

the wear behaviour of these alloys. The presence of boron in the TNZT alloy deteriorates

its wear properties. The wear rate of TNZT and TNZTB alloys in various media increases in

the following sequence: dry condition < Hank’s solution < bovine serum.c⃝ 2010 Elsevier Ltd. All rights reserved.

d

1. Introduction

Friction wear of implant materials is a major clinical prob-lem (Güleryüz and Cimenoglu, 2004; Hao et al., 2002; Niinomiet al., 1999). Sliding and rubbing action at the contact surfacesof biomedical devices during their service in the body leads tolocalized stresses at the contact regions and may cause sub-stantial damage on their surfaces. Moreover, the combinedeffect of corrosion in the human body and corrosion-assistedwear may cause wear to progress very rapidly on the im-plant (Güleryüz and Cimenoglu, 2004). In addition, the wearof prosthetic components generates debris and releases par-ticles, which play a major role in implant loosening (Niinomiet al., 1999). The presence of corrosion and wear products in

∗ Corresponding author. Tel.: +91 3222 283290; fax: +91 3222 282280.E-mail address: m.pallab@gmail.com (P. Majumdar).

1 Present address: Indian Institute of Technology, Bhubaneswar, Ind

1751-6161/$ - see front matter c⃝ 2010 Elsevier Ltd. All rights reservedoi:10.1016/j.jmbbm.2010.10.007

ia.

the tissues surrounding the implant may result in a series ofevents leading to bone loss and ultimately to the failure of theimplant (Dearnley et al., 2004; Geetha et al., 2004; Gonzalez-Mora et al., 2003; Kim et al., 2006; Long and Rack, 1998; Ni-inomi et al., 1999).

Among the conventional surgical implant materials, ti-tanium and several of its alloys are traditionally used asbiomaterials because of their excellent combination of corro-sion resistance, biocompatibility and mechanical properties(Kobayashi et al., 1998; Niinomi et al., 1999; Williams, 2001).They are generally preferred to stainless steels and Co–Cr al-loys because of their high strength to density ratio, superiorbiocompatibility and corrosion resistance, good mechanicalproperties and low elastic modulus (Hao et al., 2002; Velten

.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 285

et al., 2003). Research work on titanium alloys for biomedicalapplications is currently focused on the development of lowmodulus β titanium alloys. Promises of properties better thanthose of the existing titanium alloys (Ti–6Al–4V, Ti–6Al–7Nb,Ti–5Al–2.5Fe etc.) for implants and scope of wide use particu-larly for load bearing applications have drawn the attentionof many researchers in recent years (Akahori et al., 2005;Banerjee et al., 2005; Kuroda et al., 1998, 2005; Li et al., 2004a;Niinomi, 2003; Niinomi et al., 2005; Sagaguchi et al., 2005a;Takahashi et al., 2000; Zorn et al., 2005). However, poor tribo-logical properties of titanium and its alloys reduce the servicelife of the implants significantly. This can be attributed to lowresistance to plastic shearing and low work hardening thatinhibit their applications as articulating components (Jianget al., 2000; Molinari et al., 1997). The wear resistance of α + β

titanium alloys is found to be inferior when compared withCo–Cr alloys (Long and Rack, 1998; Manivasagam et al., 2003;Williams, 2001). For example, Ti–6Al–4V alloy shows morethan 15% higher wear than Co–Cr–Mo alloys (Hallab and Ja-cobs, 2003). On the other hand, Niinomi (1998) has observedthat the wear properties of β titanium alloys are better thanα + β type Ti–6Al–4V and Ti–6Al–7Nb alloys.

Surface modification, composition adjustment, heat treat-ment, reinforcing the matrix with some hard precipitates etc.are some of the basic approaches that are adopted to im-prove the wear properties of Ti and its alloys (Dearnley et al.,2004; Jiang et al., 2000; Li et al., 2004b; Long and Rack, 1998;Manivasagam et al., 2003; Niinomi, 1998; Niinomi et al., 1999).Reinforcing the matrix of titanium alloys (including the newgeneration β-Ti alloys) with hard ceramic phases such asborides offers the possibility of substantially enhancing thewear resistance of these composites (Samuel et al., 2008).Among many possible reinforcing compounds for titanium-matrix composites, TiB is the most promising one due to itshigh elastic modulus, comparable density and excellent in-terfacial bonding with a titanium matrix (Feng et al., 2004;Geng et al., 2003; Lu et al., 2001; Yang et al., 2006). The com-posite can be produced by simple and economic solidificationtechnique. Small additions of boron to titanium alloys leadto the formation of a strong and stiff titanium-boride (TiB)phase that precipitates in situ during solidification (Chen andBoehlert, 2008; Zhang et al., 1999).

The present work was therefore undertaken to explorethe possibility of improving the wear resistance of a lowmodulus β titanium alloy by the addition of boron thatleads to the formation of hard reinforcing particles of TiB inthe β matrix. The composite was produced by solidificationprocessing where in situ formation of TiB particles takes placeduring solidification. The effect of boron on the mechanicalproperties of this titanium alloy has been reported elsewhere(Majumdar, 2009; Majumdar et al., 2010).

Laboratory wear tests were conducted to assess and com-pare the wear properties of boron free and boron containingβ titanium alloys. This test does not exactly reproduce the ac-tual conditions in the human joint (Dumbleton, 1981). Never-theless, they are useful for predicting the wear behaviour andperformance of the material in vivo (Amaral et al., 2007; Chibaet al., 2007; Choubey et al., 2004; Dumbleton, 1981; Gispertet al., 2006, 2007; Leng et al., 2003; López et al., 2003; Thomannand Uggowitzer, 2000; Zander et al., 2003; Zheng et al., 2007).

Such tests are also valuable for comparison of different mate-rials, heat treatment and surface treatment procedures.

Wear testing equipment with different types of config-urations are used to study the wear behaviour of materi-als for orthopaedic applications. These equipment includepin-on-disc, disc-on-plate, ring-on-conforming block, ball-on-flat configurations (Dumbleton, 1981; McKellop and Ros-tlund, 2001). All these testing methods employ a simplegeometry which does not exactly reproduce in vivo condi-tions in the human joint which are rather complicated andnot completely understood (Dumbleton, 1981). However, theyare widely used for first-level evaluation of the material andcharacterization of the wear behaviour. For pin-on-disc test-ing machine either plastic or metal pins are used. This set uphas an advantage of simplicity of surface pressure calculationand a modest load gives high pressure at the contact surfaceas may be found in some prostheses and loading situations.The pin-on-disc wear testingmachine is therefore extensivelyused for evaluating the wear characteristics of implant mate-rials (Gispert et al., 2006; Li et al., 2004b; Ng et al., 2008; Ocelíket al., 2005; Samuel et al., 2008; Shi et al., 2000; Thomann andUggowitzer, 2000).

The conditions for laboratory wear tests of implant mate-rials have been spelt out by Dumbleton (1981). Both dry andwet wear tests have been described. Though dry wear testsdo not replicate the actual conditions in the human body,they are performed to highlight the effect of the lubricants.A number of lubricating liquids has been listed for wet tests;these include distilled water, saline, Ringers or Hank’s so-lutions, human or animal serum, animal synovial fluid etc.(Dumbleton, 1981; Long and Rack, 1998). A large body of liter-ature exists where one or more of these lubricants have beenused for evaluating the wear characteristics of implant ma-terials (Amaral et al., 2007; Chiba et al., 2007; Gispert et al.,2006, 2007; Leng et al., 2003; Long and Rack, 1998; López et al.,2003; Thomann and Uggowitzer, 2000; Zander et al., 2003;Zheng et al., 2007). For example, Hank’s solution was used fortests reported by Amaral et al. (2007), Chiba et al. (2007), Gis-pert et al. (2006, 2007), Leng et al. (2003), López et al. (2003),Thomann and Uggowitzer (2000), Zander et al. (2003) andZheng et al. (2007). Similarly, bovine serum and a mixture ofHank’s solution and bovine serum have been used as lubri-cants by Amaral et al. (2007) and Gispert et al. (2006, 2007).In the present work, the role of heat treatment and the effectof boron on wear behavior of Ti–35Nb–7.2Zr–5.7Ta alloy in drycondition, in Hank’s solution and in bovine serum were stud-ied.

2. Experimental

Charge for melting was prepared by weighing the elementalpieces of Ti, Nb, Zr, Ta and B according to the desired composi-tion of Ti–35Nb–7.2Zr–5.7Ta (TNZT) and Ti–35Nb–7.2Zr–5.7Ta-0.5B (TNZTB) alloys (compositions in wt%). TNZT and TNZTBalloys were prepared by arc melting of the respective chargeswith a non-consumable tungsten electrode in a vacuum arcmelting unit supplied by Vacuum Techniques Pvt Ltd, Bangalore.The melting chamber was first evacuated to less than 5×10−4

Pa and then flushed with high purity argon and evacuated

286 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

again; the chamber was finally back-filled with high purity ar-gon. Before melting the alloys, a piece of pure titanium wasmelted as getter. The ingots were turned over and remeltedat least six times in order to attain homogeneity in compo-sition and microstructure. Three or four small buttons wereremelted together and cast into ingot (approx. 40 g) in thesame melting unit.

Ikeda et al. (2002) have made an approximate estimationof the β transus temperature of Ti–29Nb–13Ta–4.6Zr alloy (inwt%) and found it to be about 727 ◦C. In the present investiga-tion, the composition of the TNZT alloy (Ti–35Nb–7.2Zr–5.7Tain wt%) is different from that of the Ti–29Nb–13Ta–4.6Zr al-loy in the sense that the amount of β stabilizer Nb is higher(35 wt%) in TNZT. Thus, it is expected that the β transus tem-perature of the TNZT alloy is not higher than 727 ◦C. The heattreatment schedule of the TNZT alloy was thus designed onthe basis of this temperature. Boron has a negligibly smallsolubility in Ti (Banerjee et al., 2003; Molchanova, 1965) andtherefore, it was assumed that the addition of 0.5 wt% boronto the TNZT alloy would not affect the transformation tem-perature appreciably. Hence, the same heat treatment sched-ule was also applied to the TNZTB alloy.

The average composition of the alloys was determinedusing scanning electron microscopy–energy dispersive spec-troscopy (SEM-EDS) on a scanning electronmicroscope (Model:JSM-5800, JEOL, Japan), attached with Energy Dispersive X-raysystem (Model: ISIS 300 Oxford Instruments Limited, UK). The op-erating voltage of the SEM was kept at 20 kV. For chemicalanalysis of the alloys, a number of regions were examinedfor each alloy in order to have an average composition. In thecase of boron containing alloy, the composition was scaled toTi+Nb+Ta+Zr = 100 wt% as suggested by Samuel et al. (2008),because the boron content of the alloy cannot be quantifiedby EDS. In addition, interstitial gas analyses were carried outusing Laboratory Equipment Corporation (LECO) analyzers inWah Chang lab, USA in order to measure the oxygen, nitrogen,hydrogen and carbon content of the TNZT alloy.

The alloys were hot rolled after melting and casting.Around 40% reduction was given to the ingots by hot rollingat 850 ◦C (above the β transus temperature) followed by aircooling to room temperature. In all the cases, deformationwas carried out in a number of passes giving about 5%reduction in each pass.

The hot rolled TNZT samples were solution treated at800 ◦C for 1 h followed by water quenching to room tem-perature. These water quenched samples were divided intotwo batches and aged in two separate salt baths, the com-position of the bath was: 40 wt% NaNO2 + 60 wt% KNO3.Aging was carried out at 300 and 400 ◦C for 2, 4 and 10 h.One set of samples from each salt bath was directly waterquenched to room temperature. Theses samples are referredto as single-stage aged samples. Another set of samplesfrom each salt bath was up-quenched to 580 ◦C in an-other salt bath and aged at that temperature for 8 h fol-lowed by water quenching. The composition of this bath was:20 wt% NaNO2+30 wt% KNO3+50 wt% CaCl2. Theses samplesare identified as duplex aged samples.

In the case of TNZTB samples, the same treatment asdescribed above for the TNZT alloy was followed but theholding time at 300 and 400 ◦C was 10 h only. This is because

Fig. 1 – Schematic diagram illustrating thethermomechanical treatment of TNZT and TNZTB alloys(RT = room temperature; AC = air cooled; WQ = waterquenched).

TNZT samples showed higher hardness when the period ofaging was longer (10 h) and it is usually expected that thehigher hardness would lead to better wear resistance of thematerial (Archard, 1953). The heat treatment cycle of TNZTand TNZTB alloys is shown schematically in Fig. 1 and thenomenclature of samples used in this work for these heattreatment cycles is also shown in the same figure.

Optical microscopy and electron microscopy were usedfor microstructural analysis of the heat treated samplesand examination of wear surfaces and wear debris. Opticalmicroscopy and image analysis were carried out on a LeicaDMLMmicroscope and a Leica DFC320 image analyzer (Model:Q550IW) interfaced with Leica Q-win V3 image analysissoftware. For this purpose, the polished samples were etchedwith Kroll’s reagent (10 vol% HF and 5 vol% HNO3 in water).For the measurement of grain sizes of different heat treatedsamples, a large number of images were recorded and theprocedure prescribed in ASTM E-112-96 was followed. Theaverage grain size was determined from several micrographsat different magnifications. For electron microscopy analysis,an FEI Nova Nanolab-200 scanning electron microscope(SEM) with FEG-SEM column operating at 20 kV was usedin backscattered electron (BSE) imaging mode. In addition,another SEM supplied by JEOL, Japan (Model: JSM-5800)operating at 20 kV voltage and high resolution transmissionelectron microscope (HRTEM) supplied by FEI-Philips (model:EM 400) operating at 200 kV were used. Hardness of thepre-polished samples was measured at 10 kg load using aVickers hardness tester (Diamond Vickers Hardness tester LV 700,supplied by Leco Corporation, USA); an average of ten readingswas taken.

The sliding wear test was started at ambient temperature(30± 1 ◦C) using a pin-on-disc wear testing machine suppliedby Ducom, Bangalore, India. The disc was made of hardenedhigh chromium steel having a hardness of Rc 69. Wearsamples in the form of cylindrical pins (8 mm diameterand 25 mm long) were machined from rolled bars, whichwere then heat treated as described above. Prior to the weartests, the pin and the disc were polished metallographicallyto a roughness (Ra) level of 0.15–0.17 µm and 0.4–0.55 µm,

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 287

respectively. Tests were carried out in dry condition as wellas in wet conditions. Wet wear tests were carried out inHank’s solution and in bovine serum. Freshly prepared Hank’ssolution was used for each experiment. The solution hadthe following chemical composition dissolved in one liter ofdistilled water: 8 g NaCl, 1 g glucose, 0.4 g KCl, 0.35 g NaHCO3,0.14 g CaCl2, 0.1 g MgCl2.6H2O, 0.06 g Na2HPO4.2H2O, 0.06 gKH2PO4 and 0.06 g MgSO4.7H2O (López et al., 2003). The pH ofthe solution was maintained at 7.4.

Loads acting on human joints vary considerably from jointto joint. For a particular joint, it varies with time duringthe loading cycle (e.g., from “heel strike” to “toe off” duringwalking) and with loading rate (e.g., with pace (slow, normalor fast) of walking,) (Dumbleton, 1981; Park, 1984). It has beenreported that the stresses in the living joints are of the orderof 1 MPa (Dumbleton, 1981). Consistent with this report, anormal load of 50 N (equivalent to a nominal contact pressureof 1 MPa) was applied to the specimen. A linear sliding speedof 1 m/s was maintained. Tests were continued up to a slidingdistance of 2000 m; this was sufficient for attainment of asteady state situation in all the cases.

The temperature at the contact surface was measured by athermocouple which was inserted in the pin-sample a smalldistance away from the contact surface. The output of thethermocouple was constantly recorded with a DAS attachedto a computer. During the wear tests, the temperature atthe articulating surfaces increases rapidly at first and then itattains a steady value with only a small degree of fluctuation(±2–3 ◦C for dry wear and ∼1 ◦C for wet wear tests). Thewear rate was estimated from the rate of height loss or heightreduction of the pin in the steady temperature range. Thus,the measurement of the height loss of the pin is not affectedby thermal expansion due to gross temperature changesduring the test and is entirely due to wear. Three sampleswere tested for each set of parameters. The wear rate wascalculated from the slope of the volume loss of the pin vs.sliding distance plot assuming a simple linear relationshipbetween the two and the data are presented as specific wearrate in the unit of m2/N as suggested by Majumdar et al.(2008). The slope was estimated at the 99% confidence level.

3. Results and discussion

3.1. Microstructure

The composition of the boron free and boron containingTNZT alloy, obtained by SEM-EDS studies and LECO analysis,is given in Table 1. The microstructure of the heat treatedTNZT and TNZTB alloys is briefly described here; the detailsare given elsewhere (Majumdar, 2009).

The optical micrographs of all the heat treated TNZT andTNZTB alloys showed equiaxed β grains (Fig. 2). The single-stage aged samples exhibited smaller grains than the duplexaged samples. The mean grain diameter of the TNZT alloyvaried between 43 and 64 µm and that of the TNZTB alloyvaried in the range of 19–32 µm. Although the TNZT alloyappeared to be a single phase material consisting of β grainsonly under optical microscope, a more detailed TEM studyshowed the presence of the ω phase (Fig. 3) or the α phase

Table 1 – The chemical composition (wt%) of the TNZTand TNZTB alloys.

Alloy Ti Nb Zr Ta O C N H

TNZT 51.3 35.5 7.3 5.9 0.13 0.0093 0.0074 0.011TNZTB 52.72 34.86 7.12 5.3 – – – –

Note: Interstitial contents were not measured for boron containingalloys.

(Fig. 4) within the β matrix. A very small amount of the ω or α

phase was present on a very fine scale in the aged samplesand therefore, theses could not be observed in the opticalmicrostructure.

The microstructure of TNZT alloy water quenched from asolution treatment temperature of 800 ◦C consisted of the β

phase. The martensite was not found in the microstructuredue to the presence of a higher amount of Nb (35 wt%) inthe alloy. As a result of the presence of a higher amountof β stabilizing Nb in the alloy, the Ms temperature of thealloy is lower than room temperature and therefore, theβ phase is retained after water quenching from the singlephase β region; however, a very small amount of additionalmetastable phases (mainly the ω phase) may be formed (Leeet al., 2002; Tang et al., 2000).

Precipitation of ω was found in the β matrix during single-stage aging at both the aging temperatures (300 ◦C and400 ◦C). Since this ω phase formed during isothermal treat-ment at the aging temperature, it is called isothermal ω (ωiso)

(Tang et al., 2000). Aging of metastable β Ti-alloys at loweraging temperatures (i.e. between α nose and Ms temperature)induces the precipitation of ω phase (Kuroda et al., 1998).Bagarjatskij et al. (1961) have found that the metastable ω

phase appears when quenched β titanium alloys are annealedat temperatures not higher than 450–550 ◦C. Qazi et al. (2005)have reported that aging of Ti–35Nb–7Zr–5Ta–0.06O and 0.46O(wt%) at 427 ◦C for 8 h resulted in ω precipitates in the matrix.Results of the present work are consistent with these reports.

Formation of ωiso involves diffusion and hence, it requirestime and temperature sufficient for solute redistribution inthe matrix (Tang et al., 2000). Therefore, it is expected thatthe ω phase would grow with increasing the aging time from2 to 10 h and the amount of ωiso phase would increase.

In case of duplex aged samples, the microstructure con-sisted of α precipitates in the β matrix. Irrespective of thefirst-stage aging temperature and time, second stage aging(at 580 ◦C for 8 h) led to the disappearance of ω precipitatesand the formation of the fine secondary α in the β matrix(Fig. 4). The ω particles can act as potent nucleation sites forα precipitates (Sagaguchi et al., 2005b). Thus, by isothermallyaging the as-quenched alloy in the ω + β field followed by up-quenching it in the α + β field and holding at that tempera-ture for a certain period of time a fine dispersion of α phasecan be produced (Fig. 4). The ω precipitates are formed duringthe first stage of aging and during the second stage of aging,the tendency of α phase to nucleate on ω is utilized. Ohmoriet al. (2001) have also observed that isothermal holding of theTi–9.87V–1.78Fe–3.20Al (wt%) alloy at 300 ◦C for 6 ks followedby up-quenching to 400 ◦C for 0.3 ks resulted in uniformly dis-persed α laths nucleated on ω phase particles.

288 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

Fig. 2 – Optical micrograph of (a) TNZT and (b) TNZTB alloys; samples single-stage aged at 300 ◦C for 10 h.

Fig. 3 – TEM micrographs and diffraction pattern of TNZT alloy after single-stage aging at 300 ◦C for 10 h (a) dark fieldimage (b) corresponding selected area diffraction (SAD) pattern and (c) schematic diagram of SAD pattern.

Fig. 4 – TEM micrographs and diffraction pattern of TNZT alloy subjected to duplex aging (300 ◦C for 10 h and then at580 ◦C for 8 h); (a) dark field image (b) corresponding selected area diffraction (SAD) pattern and (c) schematic diagram ofSAD pattern.

Addition of boron to the TNZT alloy resulted in in situthe formation of dispersed precipitates of TiB particles inthe β matrix (Fig. 5) and it had no other appreciable effecton the microstructure. The other features present in theTNZTB alloy were similar to those in the TNZT alloy. Themajority of the boride particles show an acicular (needle like)morphology with length varying in the range of 5–10 µm withsome particles being even larger than that. The solubility of

boron in titanium is negligible (Banerjee et al., 2003) and evenat 700 ◦C it is only 0.05 wt% (Molchanova, 1965). Hence itis expected that the entire amount of boron in TNZTB alloyprecipitates out as TiB particles. Similar in situ formation ofTiB precipitates in the matrix of boron containing titaniumalloys has been reported by many researchers (Chen andBoehlert, 2008; Hill et al., 2005; Soboyejo et al., 2004;Tamirisakandala et al., 2005).

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 289

Fig. 5 – (a) Backscatter SEM Micrographs and (b) dark field TEM of TNZTB alloy single-stage aged at 300 ◦C for 10 h.

Fig. 6 – Vickers hardness of the aged TNZT and TNZTB alloys; (a) solution treatment and water quenching (the pointcorresponds to 0 h), and single stage aging at 300 or 400 ◦C for different times and (b) duplex aging of the single stagesample at 580 ◦C for 8 h.

3.2. Hardness

Single-stage aging at 300 or 400 ◦C for 2 h of the waterquenched sample did not change the hardness of the TNZTalloy appreciably (Fig. 6(a)). The hardness increased with in-creasing single-stage aging time for both the aging tempera-tures. However, for any particular aging time, aging at 300 ◦Cresulted in higher hardness than aging at 400 ◦C.

After duplex aging of the single-stage aged sample, thehardness decreased (Fig. 6(b)). This decrease was high in caseof the sample aged for 4 h at 300 or 400 ◦C in the first stage.However, the sample that was aged for 10 h at 300 or 400 ◦C inthe first stage showed a small decrease in hardness after du-plex aging. In general, the hardness of the single-stage agedsamples varied in the range of 181–205 HV10 and that of du-plex aged samples in the range of 160–179 HV10.

The presence of the ω phase in the matrix has a morepronounced hardening effect than the presence of secondaryα precipitates (Banerjee et al., 2005). During single-stageaging, the amount of the precipitated ω phase increases withaging time and hence, the hardness of the samples increases.The increase in hardness during aging at 400 ◦C is smaller

than that at 300 ◦C presumably because relatively coarserprecipitates form during aging at higher temperatures. Duringduplex aging, the ω precipitates are replaced by α precipitatesand hence, the hardness decreases.

Addition of boron to the TNZT alloy led to a small increasein the hardness for all the heat treatment conditions (Fig. 6).Like the TNZT alloy, the hardness of the single-stage agedTNZTB samples was higher than the corresponding duplexaged samples. Compared with the TNZTB samples single-stage aged at higher temperature (400 ◦C) the samples single-stage aged at lower temperature (300 ◦C) produced higherhardness (Fig. 6(a)) and same trend was observed after duplexaging (Fig. 6(b)).

The small increase in hardness after the addition of boronis due to the grain refinement effect of the β grains wheregrain size of the TNZT alloy decreased from the range of43–64 µm to the range of 19–32 µm by addition of boronand due to the formation of hard TiB precipitates (hardness27.5 GPa measured by nanoindentation technique (Cao et al.,2006)) in the matrix. The hardness of this boride phase issubstantially higher than the other phases present (β, α or ω)in the TNZTB alloy.

290 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

Fig. 7 – Wear rate of the (a) solution treated and water quenched (the point corresponds to 0 h), and single-stage aged; and(b) duplex aged TNZT and TNZTB alloys in dry condition.

3.3. Wear characterization

3.3.1. Dry condition

In this condition, the TNZT sample subjected to single-stage aging at 300 ◦C for 2 h showed better wear resistancethan the sample water quenched from a solution treatmenttemperature of 800 ◦C (Fig. 7(a)). However, the wear loss of thesamples aged at 400 ◦C for 2 or 4 h was found to be higherthan that of the water quenched sample.

The wear rate of the TNZT samples single-stage aged at300 ◦C increased with increasing the aging time from 2 to 4 h(Fig. 7(a)). No significant change in the wear rate was observedwith further increase in the aging time to 10 h. Similarly,when single-stage aging was done at 400 ◦C, the wear rateof the samples increased with aging time up to 4 h and thendecreased with further aging up to 10 h. In case of TNZTsamples single-stage aged for 2 or 4 h, aging at 400 ◦C resultedin a higher wear rate than aging at 300 ◦C. However, the wearresistance of the samples single-stage aged for 10 h at boththe aging temperatures was comparable.

Single-stage aged samples and duplex aged samplesshowed an almost similar trend of wear behaviour with re-spect to the first stage aging time (Fig. 7(b)). The duplex agedsamples which were aged for 4 h at both the temperatures(300 or 400 ◦C) in the first stage exhibited higher wear lossthan the duplex aged samples aged for 2 or 10 h at the cor-responding temperatures in the first stage. For the samplesaged at 300 ◦C for 2 or 4 h in the first stage, a small increase inwear rate was observed after second stage aging. An improve-ment in the wear resistance was observed when the samplesaged at 400 ◦C for 2 or 4 h in the first stage were duplex aged.However, there was no appreciable change in the wear ratewhen the sample aged for 10 h at 300 or 400 ◦C in the firststage was subjected to second stage aging (duplex aging).

In case of dry wear, the addition of boron to the TNZTalloy led to a substantial increase in the wear rate of all theheat treated samples (Fig. 7(a) and (b)) and this was morepronounced for the water quenched sample. In case of TNZTBsamples single-stage aged for 10 h, aging at 300 ◦C resulted ina higher wear rate than aging at 400 ◦C. However, after duplexaging, the sample that was aged at 400 ◦C in the first stage

showed a higher wear loss than the sample that was aged at300 ◦C.

In dry condition, the coefficient of friction of single-stageaged as well as duplex aged TNZT samples followed analmost similar trend as the wear rate with a sample havinga higher wear rate showed a higher coefficient of friction. Thecoefficient of friction (measured in the steady temperatureregion) of the heat treated TNZT alloy (solution treated andwater quenched, single-stage aged and duplex aged) variedfrom 0.38 to 0.51 (Fig. 8(a) and (b)) and that of similarlyheat treated TNZTB samples varied from 0.47 to 0.51. Thetemperature at the contact surface of TNZT samples as wellas TNZTB samples varied from 65 to 75 ◦C during dry weartests.

3.3.2. Wet conditions

Wear in Hank’s solutionCompared with the wear rate in dry condition, the wear

rate of all the heat treated TNZT samples increased substan-tially in Hank’s solution (Fig. 9(a) and (b)). In this medium, thewear rate of the single-stage aged (aged at 300 or 400 ◦C) sam-ples was higher than the solution treated andwater quenchedsample. With increasing the single-stage aging time from 2to 4 h at 300 ◦C, no substantial change in wear rate was ob-served. However, longer aging of 10 h led to slight increase inwear loss. On the other hand, the wear rate did not changesubstantially with aging time at 400 ◦C. The wear resistanceof the TNZT samples single-stage aged at 300 ◦C was foundto be higher than that of samples single-stage aged at 400 ◦C(Fig. 9(a)).

After second stage aging of the single-stage aged samples,no appreciable change in wear rate was observed when testedin Hank’s solution (Fig. 9(b)) and the wear rate of all the duplexaged samples varied within a narrow range.

TNZTB samples after solution treatment and waterquenching as well as after single-stage aging at 300 or 400 ◦Cfor 10 h showed a higher wear rate than the similarly heattreated TNZT samples in Hank’s solution. This increase wassignificant for single-stage aged samples. However, after du-plex aging the wear rate of the TNZTB samples was compa-rable with that of similarly heat treated TNZT samples andthe wear resistance of these duplex aged TNZTB samples was

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 291

Fig. 8 – Coefficient of friction of the (a) solution treated and water quenched (the point corresponds to 0 h), and single-stageaged; and (b) duplex aged TNZT and TNZTB alloys in dry condition.

Fig. 9 – Wear rate of the (a) solution treated and water quenched (the point corresponds to 0 h), and single-stage aged; and(b) duplex aged TNZT and TNZTB alloys in Hank’s solution.

Fig. 10 – Coefficient of friction of the (a) solution treated and water quenched (the point corresponds to 0 h), andsingle-stage aged; and (b) duplex aged TNZT and TNZTB alloys in Hank’s solution.

substantially higher than the corresponding single-stage agedTNZTB samples.

The coefficient of friction of the heat treated (solutiontreated and water quenched, single aged and duplex aged)TNZT and TNZTB alloys was higher in Hank’s solution thanin dry condition. In this medium, the coefficient of friction of

the TNZT alloy varied from 0.50 to 0.59 (Fig. 10(a) and (b)) andthat of TNZTB alloy was found to be in the range of 0.49–0.60(Fig. 10(a) and (b)). The temperature at the contact surface ofboth the alloys (TNZT and TNZTB) in Hank’s solution variedfrom 40 to 45 ◦C which was significantly lower than that indry condition.

292 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

Fig. 11 – Wear rate of the (a) solution treated and water quenched (the point corresponds to 0 h), and single-stage aged; and(b) duplex aged TNZT and TNZTB alloys in bovine serum.

Wear in bovine serumThe wear loss of TNZT alloy was more severe in bovine

serum (Fig. 11) than in Hank’s solution (Fig. 9). In thismedium,the wear rate of the solution treated and water quenchedsample was lower than the sample single-stage aged at 300 ◦Cfor 2 h. However, it was higher than the sample single-stage aged at 400 ◦C for 2 h. The wear rate of the samplessingle-stage aged at 300 ◦C did not change appreciably withincreasing the aging time from 2 to 4 h. But a further increasein aging time to 10 h led to an increase in thewear rate. In caseof TNZT sample single aged at 400 ◦C, the wear loss increasedwith increasing aging time from 2 to 4 h. However, a longeraging time of 10 h at this temperature did not appreciablychange the wear rate. Compared with the aging temperatureof 400 ◦C, the lower aging temperature of 300 ◦C led to a higherwear loss of the alloy in bovine serum (Fig. 11(a)).

Compared with the samples single-staged aged at 300 ◦Cfor 2 or 4 h, the corresponding duplex aged samples showed adecrease in wear rate (Fig. 11(b)). However, duplex aging hadno significant influence on the wear behaviour of the samplewhich was single-stage aged for 10 h at this temperature. Asignificant increase in wear loss was observed after duplexaging of the sample which was aged at 400 ◦C for 2 h in thefirst stage. Compared with the duplex aged sample that wassingle-stage aged at 400 ◦C for 2 h, the duplex aged samplethat was aged for 4 h at the same temperature in the firststage showed a decrease in the wear rate and it was lowerthan the corresponding single-stage aged sample. However,the wear rate of the duplex aged sample which was aged at400 ◦C for 10 h in the first stage was higher than that of theduplex aged sample which was aged at the same temperaturefor 4 h in the first stage (Fig. 11(b)) and it was higher than thecorresponding single-stage aged sample (Fig. 11(a)).

Like the TNZT alloy, the wear loss of the TNZTB alloywas also substantially higher in bovine serum (Fig. 11) thanin Hank’s solution (Fig. 9). In this simulated body fluid, thesolution treated and water quenched TNZTB sample pre-sented a higher wear rate and single-stage aged TNZTB sam-ples showed a significantly higher wear rate than similarlyheat treated TNZT samples (Fig. 11(a)). On duplex aging ofthe single-stage aged TNZTB samples, a substantial decreasein wear rate was observed (Fig. 11(a) and (b)) and the wear

rate was comparable with that of similarly heat treated TNZTsamples (Fig. 11(b)).

In bovine serum, the coefficient of friction and thetemperature at the contact surface of the heat treated TNZTalloy (solution treated and water quenched, single-stage agedand duplex aged) varied from 0.42 to 0.59 (Fig. 12(a) and (b))and 40 ◦C–45 ◦C, respectively. The coefficient of friction of thesolution treated and water quenched as well as aged (single-stage aged or duplex aged) TNZTB samples was in the rangeof 0.57–0.65 (Fig. 12(a) and (b)). Compared with the TNZTalloy, the TNZTB alloy showed an increase in the coefficientof friction in this medium. The temperature at the contactsurface of the TNZTB alloy was similar to that of the TNZTalloy during wear testing in this medium.

In general, there is no clear relationship between the hard-ness andwear resistance of the heat treated TNZT and TNZTBalloys (although the number of TNZTB alloy samples testedis limited). A similar observation has been reported for aTi–13Zr–13Nb (wt%) alloy (Majumdar et al., 2008). It can there-fore be concluded that hardness has a relatively small in-fluence on the wear properties of TNZT and TNZTB alloyswhich is governed mainly by several mechanisms that op-erate simultaneously (Stachowiak and Batchelor, 1993). How-ever, a direct relationship exists between the hardness andmicrostructure obtained after different heat treatment cycles.Thus it can be concluded that there is no straightforward re-lationship between wear rate and microstructure.

In dry condition, a substantial increase in wear rate afteraddition of boron to TNZT alloy may be related to the pull-outof the TiB particles from the matrix of the TNZTB alloy duringrubbing of the samples against the high chromium steel discof high hardness. This leads to a three-body abrasive wearsituation that increases the wear loss of the materials.

In Hank’s solution, simultaneous corrosion of the samplesespecially for mixed phase alloy samples accelerates the wearloss (Manivasagam et al., 2003). The present study indicatesthat the electrochemical reaction on the contact surface ofthe titanium alloy in the presence of Hank’s solution has asubstantial influence on the wear rate of the material.

Higher wear rate in bovine serum than in Hank’s solutionin the present investigation is consistent with the results ofAmaral et al. (2007) and is possibly due to incomplete fluid

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 293

Fig. 12 – Coefficient of friction of the (a) solution treated and water quenched (the point corresponds to 0 h), andsingle-stage aged; and (b) duplex aged TNZT and TNZTB alloys in bovine serum.

Fig. 13 – SEM micrograph of the wear track of the duplex aged TNZT alloy that was aged first at 300 ◦C for 10 h and thenaged at 580 ◦C for 8 h; (a) dry condition, (b) Hank’s solution and (c) bovine serum.

lubrication. Wear behaviour in bovine serum depends on thematerial characteristics of the tribological pair with solid-lubricant interfacial energy playing a major role. It seemsthat the interfacial energy is influenced by the presence,size and distribution of the precipitated phase in the matrixof the TNZT alloy. In bovine serum, the difference in thewear resistance of the single-stage and duplex aged samplesmay be associated with the difference in wettability of thesamples. In this body fluid, the wear behaviour of thematerialis mainly influenced by the effect of the protein layer onthe contact surfaces, which is affected by the presence ofdifferent phases in the matrix (Gispert et al., 2006, 2007).In both the lubricating media (Hank’s solution and bovine

serum) the wear loss of the material increases when TiBparticles are present along with the ω phase in the β matrix.

3.3.3. SEM observationsSEM examination of the wear surface of single-stage and du-plex aged TNZT samples tested in dry condition revealedthat the worn surface was smeared and that there werescratches parallel with the sliding direction (Fig. 13(a)). InHank’s solution, the wear tracks were less smeared butscratches were more prominent (Fig. 13(b)). Similar observa-tion was reported by Li et al. (2004b) for Ti–Nb–Ta–Zr alloyarticulated against stainless steel in 0.9% NaCl solution. TheSEM micrograph of the wear surface of the TNZT alloy after

294 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

Fig. 14 – SEM micrograph of the wear debris of duplex aged TNZT alloy that was aged first at 300 ◦C for 10 h and then agedat 580 ◦C for 8 h; (a) dry condition, (b) Hank’s solution and (c) bovine serum (the magnification in figures are not same; thishas been done to show a number of particles).

Fig. 15 – SEM micrograph of the wear track of duplex aged TNZTB alloy that was aged first at 300 ◦C for 10 h and then agedat 580 ◦C for 8 h; (a) dry condition, (b) Hank’s solution and (c) bovine serum.

testing in bovine serum showed a heavily smeared andgrooved surface in the direction of sliding (Fig. 13(c)).

The morphological analysis using SEM of the wear de-bris obtained after wear testing in all the three media (air,

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 295

Fig. 16 – SEM micrograph of the wear debris of duplex aged TNZTB alloy that was aged first at 300 ◦C for 10 h and thenaged at 580 ◦C for 8 h; (a) dry condition, (b) Hank’s solution and (c) bovine serum (the magnification in figures are not same;this has been done to show a number of particles).

Hank’s solution and bovine serum) showed irregular shapedparticles with very sharp edges (Fig. 14). Compared with thefragmented particles produced by dry wear, the particles pro-duced by wear in wet conditions were larger.

The SEM photograph of the wear surface of the TNZTBalloy tested in dry condition showed zones of micro-fragmentation indicating brittle detachment (Fig. 15(a)). InHank’s solution and in bovine serum, surface deformationand scratches parallel with the direction of wear were ob-served on the wear surfaces (Fig. 15(b) and (c)). The wear de-bris of the TNZTB sample tested in dry condition consistedmainly of irregular shaped particles (Fig. 16(a)) whereas thesamples tested in Hank’s solution and bovine serum con-sisted mainly of elongated particles (Fig. 16(b) and (c)). Likethe TNZT alloy, the TNZTB alloy also showed relatively largefragmented particles in both the wet media than in dry con-dition.

Oxidative wear is the main damage process that occursduring dry sliding of titanium alloys (Long and Rack, 1998).During wear testing, titanium oxides (mainly TiO2 and TiO)form at the contact surface of the titanium alloys (Long andRack, 1998). The complex oxide film, which forms on titaniumsurfaces in air, although chemically protective, is unlikely toplay a direct protective role in the abrasion process becausethis film is very thin (2–5 nm) and is thus some three tofour orders of magnitude smaller than the typical abrasivescratch penetration (Mercer and Hutchings, 1998). However,the formation of an oxide layer is important indirectly, sinceit facilitates the removal of metallic debris from the abrasionprocess by reducing capping and debris adhesion (Mercer andHutchings, 1998).

The repassivation capacity of the surface layer plays a ma-jor role during the wear process in Hank’s solution. Comparedwith wear in dry condition, wear in Hank’s solution exhibitsless smearing because of rapid formation of a passive layeron the contact surface. In the case of bovine serum, heavy de-formation on the wear surface is due to large contact stressdeveloped as a result of localized contacts between the op-posing surfaces interleaved with zones where the fluid pre-vented direct contact. This results from a deficiency in fullfluid lubrication in the entire contact region (Amaral et al.,2007). Primarily the abrasive wear mechanism is operative inHank’s solution as well as in bovine serum, and it is more se-vere in bovine serum. The larger contact stresses developedin bovine serum than in Hank’s solution lead to increase inwear in the former medium.

4. Conclusions

In the present work, the wear behaviour of Ti–35Nb–7.2Zr–5.7Ta (TNZT) and Ti–35Nb–7.2Zr–5.7Ta–0.5B (TNZTB) alloys(all compositions are in wt%) subjected to different heattreatment conditions has been investigated in dry conditionand in simulated body fluid. The following conclusions can bedrawn from this study:

(1) There seems to be no straightforward relationship be-tween wear rate and microstructure.

(2) Hardness has a relatively small influence on the wearproperties of TNZT and TNZTB alloys which are governedmainly by several wear mechanisms that operate simulta-neously.

296 J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7

(3) The wear rate of TNZT and TNZTB alloys in variousmedia increases in the following sequence: dry condi-tion < Hank’s solution < bovine serum.

(4) Boron deteriorates the wear properties of the TNZT alloyin all the media (air, Hank’s solution and bovine serum)tested in the present investigation.

R E F E R E N C E S

Akahori, T., Niinomi, M., Fukui, H., Ogawa, M., Toda, H., 2005.Improvement in fatigue characteristics of newly developedbeta type titanium alloy for biomedical applications bythermo-mechanical treatments. Mater. Sci. Eng. C 25, 248–254.

Amaral, M., Abreu, C.S., Oliveira, F.J., Gomes, J.R., Silva, R.J., 2007.Biotribological performance of NCD coated Si3N4-bioglasscomposites. Diam. Relat. Mater. 16, 790–795.

Archard, J.F., 1953. Contact and rubbing of flat surfaces. J. Appl.Phys. 24, 981–988.

Bagarjatskij, J.A., Nosova, G.I., Tagunova, T.V., 1961. On thenature of the omega phase in quenched titanium alloys. ActaCrystallogr. 14, 1087–1088.

Banerjee, R., Collins, P.C., Genç, A., Fraser, H.L., 2003. Direct laserdeposition of in situ Ti–6Al–4V-TiB composites. Mater. Sci. Eng.A 358, 343–349.

Banerjee, R., Nag, S., Fraser, H.L., 2005. A novel combinationapproach to the development of beta titanium alloys fororthopaedic implants. Mater. Sci. Eng. C 25, 282–289.

Cao, G., Geng, L., Naka, M., 2006. Elastic properties of titaniummonoboride measured by nanoindentation. J. Am. Ceram. Soc.89, 3836–3838.

Chen, W., Boehlert, C.J., 2008. The elevated-temperature fatiguebehavior of boron-modified Ti–6Al–4V (wt%) castings. Mater.Sci. Eng. A 494, 132–138.

Chiba, M., Kumagai, K., Nomura, N., Miyakawa, S., 2007. Pin-on-disk wear behavior in a like-on-like conguration in a biologicalenvironment of high carbon cast and low carbon forgedCo–29Cr–6Mo alloys. Acta Mater. 55, 1309–1318.

Choubey, A., Basu, B., Balasubramaniam, R., 2004. Tribologicalbehavior of Ti-based alloys in simulated body fluid solution atfretting contacts. Mater. Sci. Eng. A 379, 234–239.

Dearnley, P.A., Dahm, K.L., Cimenoglu, H., 2004. The corrosion-wear behaviour of thermally oxidized CP–Ti and Ti–6Al–4V.Wear 256, 469–479.

Dumbleton, J.H., 1981. Tribology of Natural and Artificial Joints,first ed. In: Tribology Series, vol. 3. Elsevier, New York.

Feng, H., Zhou, Y., Jia, D., Meng, Q., 2004. Microstructure andmechanical properties of in situ TiB reinforced titaniummatrixcomposites based on Ti–Fe–Mo–B prepared by spark plasmasintering. Compos. Sci. Technol. 6, 2495–2500.

Geetha, M., Mudali, U.K., Gogia, A.K., Asokamani, R., Raj, B.D.,2004. Influence of microstructure and alloying elements oncorrosion behavior of Ti–13Nb–13Zr alloy. Corros. Sci. 46,877–892.

Geng, K., Lu, W., Yang, Z., Zhang, D., 2003. In situ preparationof titanium matrix composites reinforced by TiB and Nd2O3.Mater. Lett. 57, 4054–4057.

Gispert, M.P., Serro, A.P., Colaco, R., Botelho do Rego, A.M.,Alves, E., da Silva, R.C., Brogueira, P., Pires, E., Saramago, B.,2007. Tribological behavior of Cl-implanted Tin coatings forbiomedical applications. Wear 262, 1337–1345.

Gispert, M.P., Serro, A.P., Colaco, R., Saramago, B., 2006. Frictionand wear mechanisms in hip prosthesis: comparison of jointmaterials behavior in several lubricants. Wear 260, 149–158.

Gonzalez-Mora, V., Hoffmann, M., Chiesa, R., Cigada, A.,Hampshire, J., Stroosnijder, M.F., 2003. Screening wear testerversus HIP joint simulator: results of UHMWPE sliding againstdifferent CoCrMo counterfaces. Surf. Eng. 19, 45–50.

Güleryüz, H., Cimenoglu, H., 2004. Effect of thermal oxidation oncorrosion and corrosion-wear behaviour of a Ti–6Al–4V alloy.Biomaterials 25, 3325–3333.

Hallab, N.J., Jacobs, J.J., 2003. Orthopedic implant frettingcorrosion. Corros. Rev. 21, 183–213.

Hao, Y.L., Niinomi, M., Kuroda, D., Fukunaga, F., Zhou, Y.L.,Yang, R., Suzuki, A., 2002. Young’s modulus and mechanicalproperties of Ti–29Nb–13Ta–4.6Zr in relation to α′′ martensite.Metall. Mater. Trans. A 33, 3137–3144.

Hill, D., Banerjee, R., Huber, D., Tiley, J., Fraser, H.L., 2005.Formation of equiaxed alpha in TiB reinforced Ti alloycomposites. Scr. Mater. 52, 387–392.

Ikeda, M., Komatsu, S.Y., Sowa, I., Niinomi, M., 2002. Agingbehavior of the Ti–29Nb–13Ta–4.6Zr new beta alloy for medicalimplants. Metall. Mater. Trans. A 33, 487–493.

Jiang, P., He, X.L., Li, X.X., Yu, L.G., Wang, H.M., 2000. Wearresistance of a laser surface alloyed Ti–6Al–4V alloy. Surf. Coat.Technol. 130, 24–28.

Kim, H.S., Kim, W.Y., Lim, S.H., 2006. Microstructure andelastic modulus of Ti–Nb–Si ternary alloys for biomedicalapplications. Scr. Mater. 54, 887–891.

Kobayashi, E., Doi, H., Yoneyama, T., Hamanaka, H., Gimson, I.R.,Best, S.M., Shelton, J.C., Bonfield, W., 1998. Influence of agingheat treatment on mechanical properties of biomedical Ti–Zrbased ternary alloys containing niobium. J. Mater. Sci. Mater.Med. 9, 625–630.

Kuroda, D., Kawasaki, H., Yamamoto, A., Hiromoto, S., Hanawa,T., 2005. Mechanical properties and microstructures of newTi–Fe–Ta and Ti–Fe–Ta–Zr system alloys. Mater. Sci. Eng. C 25,312–320.

Kuroda, D., Niinomi, M., Morinaga, M., Kato, Y., Yashiro, T., 1998.Design and mechanical properties of new β type titaniumalloys for implant materials. Mater. Sci. Eng. A 243, 244–249.

Lee, C.M., Ho, W.F., Ju, C.P., Lin, J.H.C., 2002. Structure andproperties of titanium-25 niobium-x iron. J. Mater. Sci. Mater.Med. 13, 695–700.

Leng, Y.X., Chen, J.Y., Yang, P., Sun, H., Huang, N., 2003. Structureand properties of passivating titanium oxide films fabricatedby DC plasma oxidation. Surf. Coat. Technol. 166, 176–182.

Li, S.J., Niinomi, M., Akahori, T., Kasuga, T., Yang, R., Hao,Y.L., 2004a. Fatigue characteristics of bioactive glass–ceramic-coated Ti–29Nb–13Ta–4.6Zr for biomedical applications. Bio-materials 25, 3369–3378.

Li, S.J., Yang, R., Li, S., Hao, Y.L., Cui, Y.Y., Niinomi, M., Guo,Z.X., 2004b. Wear characteristics of Ti–Nb–Ta–Zr and Ti–6Al–4Valloys for biomedical applications. Wear 257, 869–876.

Long, M., Rack, H.J., 1998. Titanium alloys in total jointreplacement—a materials science perspective. Biomaterials19, 1621–1629.

López, M.F., Jiménez, J.A., Gutiérrez, A., 2003. Corroson study ofsurface-modified vanadium-free titanium alloys. Electrochim.Acta 48, 1395–1401.

Lu, W., Zhang, D., Zhang, X., Wu, R., Sakata, T., Mori, H., 2001.Microstructural characterization of TiB in situ synthesizedtitanium matrix composites prepared by common castingtechnique. J. Alloys Compd. 327, 240–244.

Majumdar, P., 2009. Processing and characterization of lowmodulus biocompatible titanium based alloys. Ph.D. Thesis.Indian Institute of Technology, Kharagpur, India.

Majumdar, P., Singh, S.B., Chakraborty, M., 2008. Wear responseof heat-treated Ti–13Zr–13Nb alloy in dry condition andsimulated body fluid. Wear 264, 1015–1025.

Majumdar, P., Singh, S.B., Chakraborty, M., 2010. Fatigue behaviourof in situ TiB reinforced β-titanium alloy composite. Mater. Lett.64, 2748–2751.

J O U R N A L O F T H E M E C H A N I C A L B E H AV I O R O F B I O M E D I C A L M A T E R I A L S 4 ( 2 0 1 1 ) 2 8 4 – 2 9 7 297

Manivasagam, G., Mudali, U.K., Asokamani, R., Raj, B., 2003.Corrosion and microstructural aspects of titanium and itsalloys as orthopaedic devices. Corros. Rev. 21, 125–159.

McKellop, H.A., Rostlund, T., 2001. Wear of titanium 6–4 alloy inlaboratory tests and in retrieved human joint replacements.In: Brunette, D.M., Tengvall, P., Texfor, M., Thomsen, P. (Eds.),Titanium in Medicine. Springer, New York, pp. 747–770.

Mercer, A.P., Hutchings, I.M., 1998. The influence of atmosphericcomposition on the abrasive wear of titanium and Ti–6Al–4V.Wear 124, 165–176.

Molchanova, E.K., 1965. Ti–B phase diagram. In: Glazunov, S.G.(Ed.), Phase Diagrams of Titanium Alloys, Israel, Program forScientific Translation, Jerusalem, pp. 102– 104.

Molinari, A., Straffelini, G., Tesi, B., Bacci, T., 1997. Dry sliding wearmechanisms of the Ti6Al4V alloy. Wear 208, 105–112.

Ng, K.W., Man, H.C., Yue, T.M., 2008. Corrosion andwear propertiesof laser surface modified NiTi with Mo and ZrO2. Appl. Surf.Sci. 254, 6725–6730.

Niinomi, M., 1998. Mechanical properties of biomedical titaniumalloys. Mater. Sci. Eng. A 243, 231–236.

Niinomi, M., 2003. Fatigue performance and cyto-toxicity of lowrigidity titanium alloy. Ti–29Nb–13Ta–4.6Zr. Biomaterials 24,2673–2683.

Niinomi, M., Akahori, T., Takeuchi, T., Katsura, S., Fukui, H.,Toda, H., 2005. Mechanical properties and cyto-toxicity of newbeta type titanium alloy with low melting points for dentalapplications. Mater. Sci. Eng. C 25, 417–425.

Niinomi, M., Kuroda, D., Fukunaga, K.I., Morinaga, M., Kato, Y.,Yashiro, T., Suzuki, A., 1999. Corrosion wear fracture of newβ type biomedical titanium alloys. Mater. Sci. Eng. A 263,193–199.

Ocelík, V., Matthews, D., De Hosson, J.Th.M., 2005. Sliding wearresistance of metal matrix composite layers prepared by highpower laser. Surf. Coat. Technol. 197, 303–315.

Ohmori, Y., Ogo, T., Nakai, K., Kobayashi, S., 2001. Effects ofω-phase precipitation on β → α, α′′ transformations in ametastable β titanium alloy. Mater. Sci. Eng. A 312, 182–188.

Park, J.B., 1984. Biomaterials Science and Engineering, third ed.Plenum Press, New York.

Qazi, J.I., Marquardt, B., Allard, L.F., Rack, H.J., 2005. Phasetransformation in Ti–35–7Zr–5Ta (0.06–0.68)O alloys. Mater. Sci.Eng. C 25, 389–397.

Sagaguchi, N., Niinomi, M., Akahori, T., Takeda, J., Toda, H.,2005a. Effect of Ta content on mechanical properties ofTi–30Nb–xTa–5Zr. Mater. Sci. Eng. C 25, 370–376.

Sagaguchi, N., Niinomi, M., Akahori, T., Takeda, J., Toda, H., 2005b.Relationship between tensile behaviour and microstructure inTi–Nb–Ta–Zr system alloys. Mater. Sci. Eng. C 25, 363–369.

Samuel, S., Nag, S., Scharf, T.W., Banerjee, R., 2008. Wearresistance of laser-deposited boride reinforced Ti–Nb–Zr–Taalloy composites for orthopedic implants. Mater. Sci. Eng. C28, 414–420.

Shi, W., Dong, H., Bell, T., 2000. Tribological behaviour andmicroscopic wear mechanisms of UHMWPE sliding againstthermal oxidation-treated Ti6Al4V. Mater. Sci. Eng. A 291,27–36.

Soboyejo, W.O., Shen, W., Srivatsan, T.S., 2004. An investigation offatigue crack nucleation and growth in a Ti–6Al–4V/TiB in situcomposite. Mech. Mater. 36, 141–159.

Stachowiak, G.W., Batchelor, A.W., 1993. Engineering Tribology,third ed. In: Tribology Series, Elsevier, New York.

Takahashi, M., Kobayashi, E., Doi, H., Yoneyama, T., Hamanaka, H.,2000. Phase stability and mechanical properties of biomedicalβ type titanium–zirconium based alloys containing niobium.J. Japan Inst. Met. 64, 1120–1126.

Tamirisakandala, S., Bhat, R.B., Tiley, J.S., Miracle, D.B., 2005.Processing, microstructure, and properties of β titanium alloysmodified with boron. J. Mater. Eng. Perform. 14, 741–746.

Tang, X., Ahmed, T., Rack, H.J., 2000. Phase transformationsin Ti–Nb–Ta and Ti–Nb–Ta–Zr alloys. J. Mater. Sci. 35,1805–1811.

Thomann, U.I., Uggowitzer, P.J., 2000. Wear-corrosion behaviorof biocompatible austenitic stainless steels. Wear 239,48–58.

Velten, D., Schenk-Meuser, K., Biehl, V., Duschner, H., Breme, J.,2003. Characterization of thermal and anodic oxide layers onβ- and on near-β-titanium alloys for biomedical application. Z.Met. kd. 6, 667–675.

Williams, D.F., 2001. Titanium for medical applications.In: Brunette, D.M., Tengvall, P., Texfor, M., Thomsen, P.(Eds.), Titanium in Medicine. Springer, New York, pp. 13–24.

Yang, Z., Lu, W., Qin, J., Zhang, D., 2006. Microstructure and tensileproperties of in situ synthesized (TiC + TiB + Nd2O3)/Ti-alloycomposites at elevated temperature. Mater. Sci. Eng. A 425,185–191.

Zander, D., Olson, D.L., Eliezer, D., 2003. Mutual effects ofhydrogenation and deformation in Ti–Nb alloys. Metall. Mater.Trans. A 34, 2199–2206.

Zhang, X., Lu, W., Zhang, D., Wu, R., 1999. In situ technique forsynthesizing (TiB1TiC)/Ti composites. Scr. Mater. 41, 39–46.

Zheng, L.Y., Zhao, L.X., Zhang, J.J., 2007. Tribological study ofthree-dimensionally braided carbon Fiber–Nylon 6 compositesagainst 316L stainless steel. Curr. Appl. Phys. 7, 120–124.

Zorn, G., Lesman, A., Gotman, I., 2005. Oxide formation on lowmodulus Ti45Nb alloy by anodic versus thermal oxidation.Surf. Coat. Technol. 201, 612–618.