- Email: [email protected]
Effect of Sn addition on the microstructure and superelasticity in Ti–Nb–Mo–Sn Alloys
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jmbbm
Research Paper
Effect of Sn addition on the microstructure and superelasticity in Ti–Nb–Mo–Sn Alloys D.C. Zhanga,b, S. Yanga,b, M. Weia,b, Y.F. Maoa,b, C.G. Tana,b, J.G. Lina,b,n a
Key Laboratory of Low Di-mensional Materials and Application Technology of Ministry of Education, Xiangtan University, Xiangtan, Hunan 411105, China b Faculty of Material and Optical-Electronic Physics, Xiangtan University, Xiangtan, Hunan 411105, China
ar t ic l e in f o
abs tra ct
Article history:
Ti–7.5 Nb–4Mo–xSn (x¼0–4 at%) alloys were developed as the biomedical materials. The effect of
Received 23 November 2011
the Sn content on the microstructure and superelasticity of the alloys was investigated. It is
Received in revised form
found that Sn is a strong stabilizer of the b phase, which is effective in suppressing the
16 April 2012
formation of a00 and o phases in the alloys. Moreover, the Sn addition has a significant impact
Accepted 21 April 2012
on the mechanical properties of the alloys. With the increase of Sn addition, the yield stress of
Available online 4 May 2012
the alloys increase, but their elastic modulus, the fracture strength and the ductility decrease,
Keywords:
and the deformation mode of the alloys changes from (322) twining to a00 transformation and
Titanium alloys
then to slip. The Ti–7.5 Nb–4Mo–1Sn and Ti–7.5 Nb–4Mo–3Sn alloys exhibit a good superelasticity
Microstructure
with a high sSIM due to the relatively high athermal o phases containing or the solution
Mechanical properties
hardening at room temperature. Under the maximum strain of 5%, Ti–7.5 Nb–4Mo–3Sn (at%)
Deformation mode
alloy exhibits higher super elastic stability than that of Ti–7.5 Nb–4Mo–1Sn alloy.
Superelasticity
1.
Introduction
Nickel–titanium alloys have been found to be the most useful of all the titanium-based shape memory alloys (SMAs) as biomedical materials due to their large recoverable strain value, low elastic modulus, and high corrosion resistance (Barras and Myers, 2000; Frick et al., 2006). However, it has been found that pure nickel may exhibit hypersensitivity and carcinogenicity to human body (Wang et al., 1996). There is a prohibition tendency for nickel–titanium alloy system in the biomedical field in European countries in recent years due to the problem of human nickel allergy. Therefore it is n Corresponding author at: Xiangtan University, Faculty of Material and Optical-Electronic Physics, Xiangtan, Hunan 411105, China. Tel./fax: þ86 731 58298119. E-mail addresses: [email protected], [email protected] (J.G. Lin).
1751-6161/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jmbbm.2012.04.017
& 2012 Elsevier Ltd. All rights reserved.
preferable to develop absolutely safe nickel-free titaniumbased shape memory alloys for biomedical applications. Recently, some Ni-free Ti alloys such as Ti–Nb (Souza et al., 2010; Brailovski et al., 2011; Laheurte et al., 2010; Sun et al., 2011; Chai et al., 2008; Wang et al., 2009; Miura et al., 2011; Al-Zain et al., 2010; Nozoe et al., 2007), Ti–Mo (Al-Zain et al., 2010; Lin et al., 2007; Oliveira et al., 2007; Oliveira et al., 2009a,b; Sutou et al., 2006; Zhang et al., 2005;) and Ti–Zr (Brailovski et al., 2011; Wang et al., 2009) have been extensively investigated as SMA due to their low elastic modulus, high biocompatibility and non-toxicity. It is well known that the SME of Ti–Nb alloys is attributed to the reversible martensitic transformation between a00 martensite and parent phase (Ozaki et al., 2004), and the martensitic transformation temperatures (Ms) and the shape memory behavior of Ti–Nb alloys are strongly dependent on their compositions. It has been reported that Ms of Ti–Nb alloys decreases from the melting point of Ti to 90 1C with the increase of the Nb content (Kim et al., 2004). A SME of 3% has
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
been obtained in Ti–(22–29 at%) Nb alloys. Moreover, for Ti–Nb alloys, their superelasticity (SE) strain can be improved up to 3.3% by cyclic loading–unloading training but their SME are not prone to be improved due to their low martensitic transformation stress (sSIM is about 200 MPa) (Kim et al., 2006). By contrast, Ti–Mo based alloys exhibit a relatively small superelasticity recovered strain (about 2.5%), but a high sSIM (about 450 MPa) (Zhang et al., 2005). Thus, in order to enhance the SE, and its stability, Miyazaki et al. added Mo to Ti–Nb alloys to develop the Ti–15Nb–4Mo alloy, which sSIM and a total recovery strain reached about 420 MPa and 3.5%, respectively (Al-Zain et al., 2010). In addition, Sn addition has an important influence on the SME and SE behavior of Ti–Nb or Ti–Mo alloys. It has been reported that Ms of Ti–Nb–Sn alloys decreases rapidly with increasing Sn content and a large SE strain was obtained in Ti–16Nb–4.9Sn alloy (Al-Zain et al., 2010). Moreover, Ozaki et al. (Ozaki et al. 2004) found that the addition of 2.5 at% Sn to Ti22 at% Nb could suppress the a00 martensitic transformation. While, Yang et al. ( Hu et al., 2008) found that Sn is a stronger b stabilizer than Nb in the Ti–Nb–Zr–Sn system from first-principles calculations, and a certain amount of Sn addition could reduce the elastic modulus of the alloy by suppressing o phase formation. Based on the previous work, we designed a quaternary Ti-based alloy, Ti–7.5Nb–4Mo–1Sn (in atom percent), according to the d-electron orbit theory, and the alloy exhibits stable SME and SE (Zhang et al., 2011). But how the Sn content affects the microstructure and the mechanical properties of the quaternary alloy is not reported yet. Therefore, in the present work, the Ti–7.5Nb–4Mo–xSn (x¼ 0–4) alloys were fabricated by an arc melting method, and the microstructure and mechanical properties of the alloy were investigated, emphasizing on clarifying the effect of the Sn content on the SME and SE of the alloy.
2.
157
Fig. 1 – Schematic geometry of the tensile test specimen.
Experimental
The ingot of the Ti alloy was fabricated by an arc melting method using pure Ti, Nb, Mo and Sn as raw materials. To ensure the composition homogeneity, the ingot was flipped and remelted five times. The nominal composition of the ingot is Ti–7.5Nb–4Mo–xSn (x¼0–4), and the ingots were hot-pressed by 50% at the temperature of about 1073 K. The samples for X-ray diffraction (XRD) measurement and tensile tests were cut from the ingot by an electro-discharge machine, and vacuum sealed in quartz tubes with Ar atmosphere. The samples were solutiontreated at 1273 K for 1.8 ks followed by quenching them into ice water. After that, the specimens were acid etched to remove the oxidized skin. The phase structures present in the microstructure were identified at room temperature by XRD using a Rigaku D/Max 2500PC diffractometer operated at 50 kV and 100 mA with a Cu Ka radiation (l¼1.5406 nm). To evaluate the SE of the alloy, tensile tests were carried out on an Instron 5569 universal testing machine, using the tensile specimens with a gage section of 1 mm 2.5 mm 8 mm, and the geometry of the samples is schematically shown in Fig. 1. The transmission electron microscopy (TEM) observations were conducted using a JEM-2100 operated at 160 kV. Samples for TEM were prepared by an electro-polishing in a solution of 200 vol% methanol, 20 vol% glycol and 10 vol% HNO3 at 15 V–20 V under 301C. Microhardness was obtained using a Vickers hardness tester (HVS-30Z)
Fig. 2 – XRD spectra Ti–7.5Nb–4Mo–xSn alloys.
of
the
solution-treated
with a load of 10 kgf. Elastic moduli were determined by a Tribo Indenter from monotonic load tests to a depth of 7000 nm using a Berkovich tip with a measured radius of 5000 nm. The data were analyzed using the Oliver and Pharr method (Pharr, 1998).
3.
Results and discussion
3.1.
Microstructure
Fig. 2 shows the XRD spectra of solution-treated Ti–7.5Nb–4Mo– xSn (x¼0–4) alloys. The alloy without Sn content (Ti–7.5Nb–4Mo)
158
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
mainly contains b and a00 phases (Fig. 2a), but a small amount of the athermal o phases are present in the alloy (Fig. 2 b). With Sn content increasing, the volume fraction of b phase increases, which is proved by the increasing intensity of the peaks of b phase in the Sn containing alloys. As the Sn content increases to more than 3%, the peaks from a00 and o phases disappear on the XRD pattern, but only the peaks from b phase can be observed in the alloys. More carful observations of the microstructures of the alloys were conducted on a TEM. Fig. 3 presents the TEM dark-field images of Ti–7.5Nb–4Mo–xSn (x¼ 0–2) alloys with the corresponding selected area diffraction patterns with zone axis [113]b in the inserts. The dark-field images were formed using the spot of the athermal o phase in the diffraction patterns marked by a circle in the diffraction patterns. It can be seen that the volume fraction of the athermal o phase decreases with the Sn addition increasing. The result is in coincident with that of XRD, which indicates that Sn is a strong b stabilizer, which is effective in suppressing the athermal o transformation. In the meantime, the presence of Sn decreases the start temperature of the matensitic phase transformation (Ms) of the alloys. With further increasing of Sn content, a00 transformation is suppressed due to the solid-solution
hardening effect of Sn. The findings are supported by some previous work (Ishiyam et al., 1991; Ozaki et al., 2004; Hu et al., 2008). Ishiyam et al. (Ishiyam et al., 1991) found that the martensitic transformation is difficult to occur in the alloys containing a large amount of Sn and A1. Kim et al. (Ozaki et al., 2004) found that Ms decreases rapidly with increasing Sn content in Ti–Nb–Sn alloys. Yang et al. (Hu et al., 2008) investigated the effects of the alloying elements, Nb, Zr and Sn on the stability of Ti alloys from the first principles calculations, and their results demonstrated that Sn is a stronger b stabilizer than Nb theoretically, and a certain amount of Sn addition could reduce the elastic modulus of the alloy by suppressing o phase formation in the Ti–Nb–Zr–Sn alloys. The optical microstructure of Ti–7.5Nb–4Mo–xSn alloys is illustrated in Fig. 4. It can be seen that all the alloys exhibit nearly equiaxed grain microstructure, and the grain size of the alloys decreases obviously with the Sn content increasing. Some fine acicular martensitic a00 phase can be observed in Ti–7.5Nb–4Mo alloy and Ti–7.5Nb–4Mo–1Sn alloy. While, as Sn content increases to 2% or more, the microstructure of the alloys only contains b phase. The result is in agreement with that of XRD analysis.
Fig. 3 – Dark-field images and the selected area diffraction patterns for (a) Ti–7.5Nb–4Mo, (b) Ti–7.5Nb–4Mo–1Sn, (c) Ti–7.5Nb–4Mo–2Sn alloys.
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
159
Fig. 4 – Optical micrographs of Ti–7.5Nb–4Mo–xSn alloys. (a) Ti–7.5Nb–4Mo alloy, (b) Ti–7.5Nb–4Mo–1Sn alloy, (c) Ti–7.5Nb–4Mo–2Sn alloy, (d) Ti–7.5Nb–4Mo–3Sn alloy, (e) Ti–7.5Nb–4Mo–4Sn alloy.
3.2.
Mechanical properties
To evaluate the effect of Sn content on the modulus of the alloys, nano-indenter tests were carried out on the alloys with different Sn contents. The elastic modulus of the alloys as a function of Sn content is presented in Fig. 5. It can be seen that the Sn-free alloy, Ti–7.5Nb–4Mo, yields a relative high elastic modulus, which is about 120 GPa. With the increase of Sn content, the modulus of the alloys sharply declines. As the Sn content is about 2%, the alloys exhibit the lowest modulus, which is 87 GP. The result is coincident with the change trend of the microstructure with the Sn content. As mentioned above, the microstructure of the Sn-free alloy contains bþoþa00 phases. The Sn addition suppresses the formation of oþa00 . As the Sn content increases from 1% to 4%, the phase constitution of the alloys is from bþo to b. While, it is well documented the modulus of the phases ranges in a diminishing sequence from o, a, a00 and b (Lee et al., 2002). Especially for the o phase, it has a significant influence on the elastic modulus of Ti-based alloy (Souza
Fig. 5 – Elastic modulus Ti–7.5Nb–4Mo–xSn alloys.
and
microhardness
of
et al., 2010). Therefore, the alloys containing 2%–4% Sn exhibit a relatively low modulus due to their single b structure. In addition, Fig. 5 also displays the change of the
160
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
microhardness of the alloys with the Sn content. It can be seen that the microhardness of the alloys sharply declines as the Sn content increases from 0 to 2% due to the content reduction of o and a00 phases, because o and a00 phases are harder than b phase. However, as the Sn content further increase to 3% and 4%, the microhardness of the alloys increases, which implies that the Sn addition has a strong solid-solution hardening effect. Fig. 6 shows stress–strain curves of solution-treated alloys obtained by tensile tests at room temperature. Ti–7.5Nb–4Mo– xSn (x¼ 1–3) alloys exhibit two-stage yielding as shown by arrows. It is supposed that the Ti–7.5Nb–4Mo–xSn (x¼1–3) alloys exhibit the SME or SE at room temperature. However, the Ti–7.5Nb–4Mo–xSn (x¼ 0, 4) alloys exhibit a single-stage yielding. To clearly illustrate the effect of the Sn content on the first yield stress (sSIM, defined by the 0.2% offset yield stress) indicated by the black-headed, the variation of sSIM as a function of the Sn content is present in Fig. 7. It can be seen that sSIM decreases with the increase of Sn content, and then it increases as the Sn content further increases, and the
Fig. 6 – Stress–strain curves obtained for Ti–7.5Nb–4Mo–xSn alloys.
Fig. 7 – rSIM of Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloy.
Ti–7.5Nb–4Mo–2Sn alloy exhibits the lowest sSIM. Moreover, XRD analysis on the alloys with the different Sn contents before and after tensile test clearly indicates that the peak of b phase at 38.641for all the alloys after the deformation splits into two peaks from b phase and a00 phase, respectively (see Fig. 8). By comparison, more a00 martensites are present in the deformed Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloys than the deformed Ti–7.5Nb–4Mo and Ti–7.5Nb–4Mo–4Sn alloys. Therefore, the martensitic transformation induced by stress is a dominant deformation mode in Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloys, which provides an evidence that the deformation induced a00 transformation tends to be suppressed by adding Sn in the alloys. For comparison, the deformation microstructures of the Ti–7.5Nb–4Mo and Ti–7.5Nb–4Mo–4Sn alloys were observed by using TEM. Fig. 9 illustrates TEM microstructures of the Ti–7.5Nb–4Mo and Ti–7.5Nb–4Mo–4Sn alloys after the deformation. A number of the deformation twins can be found in the deformed Ti–7.5Nb–4Mo alloy (see Fig. 9a). By contrast, many dislocations can be observed, but the deformation twins are hard to be found in the deformed Ti–7.5Nb–4Mo–4Sn alloy (see Fig. 9b). The occurrence of (322) deformation twins in the Ti–7.5Nb–4Mo alloy may be attributed to its relatively high athermal o phase content within b matrix. Oka et al. and Hanada et al. [Oka and Taniguchi, 1978; Hanada et al., 1986] explained that shuffling of one-half of atoms in the alloys with a large amount of athermal o phase into the direction different from that of twinning shear, which is needed to create (332) twinning, is ready to occur in an unstable b with o phase precipitation in it. While, the Ti–7.5Nb–4Mo–4Sn alloy deforms in the slip mode instead of the martensitic transformation, which may be attributed to its relatively high sSIM for the martensitic transformation due to the solid-solution hardening effect of the high Sn content. As a result, the deformation mode of the Ti–Nb–Mo–Sn alloys is governed by its chemical composition. With the increase of the Sn additions, the deformation mode of the alloy changes from twinning to a00 transformation, and then to slip. The result is in good agreement with that reported in the previous work [Ishiyam et al., 1991; Ohyama and Nishimura, 1995; Hanada and Izumi, 1982], in which, it was found that (332) deformation twins and a00 transformation are the dominant deformation modes in the alloys with the highly unstable b phases, while the slip is the dominant deformation mode in the alloys with the stable b phases. Further more, the yielding stresses sS for the plastic deformation, indicated by gray-headed arrows in Fig. 6, the ultimate tensile strength sult and the fracture strain d of the alloys are measured from the stress–strain curves of Ti–7.5Nb–4Mo–xSn (x¼0–4) alloys, which are plotted against Sn content in Fig. 10. It is found that sS increases with Sn content increasing. On the other hand, d decreases with Sn content increasing from 47% for the Ti–7.5Nb–4Mo to 26% for the Ti–7.5Nb–4Mo–4Sn. Irrespective of the solid-solution strengthening effect of the Sn addition, the Ti–Nb–Mo–Sn alloys exhibit a good ductility. Fig. 11 displays the SEM fractographs of the Ti–7.5Nb–4Mo–xSn (x¼0–4) alloy. The fracture surfaces of the alloys are covered with dimples, suggesting that the fracture of the alloys is in a ductile fracture mode. It also can be seen that the dimple’s size and depth reduce with Sn
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
161
Fig. 8 – XRD profiles of the Ti–7.5Nb–4Mo–xSn alloys before and after deformation (a) x ¼0, (b) x ¼1, (c) x ¼2, (d) x¼ 3, (e) x ¼4.
content increasing, which is coincident with the change trend of the ductility of the alloys with the Sn content.
3.3.
Superelasticity
Fig. 12 shows stress–strain curves of the Ti– 7.5Nb–4Mo–2Sn alloy. Each stress–strain curve was obtained by a loading and unloading cycle at room temperature. The similar measurement was repeated by increasing the maximum strain upon loading in the same sample.
In order to characterize the SE of the alloy, tow types of strains and a recovery rate are defined as follows: (1) the permanently remained strain eresidual after unloading (see Fig. 12); (2) the recovered strain erecoverable upon unloading (see Fig. 12); (3) the strain recovery rate, which can be described as Z ¼ erecoverable =ðerecoverable þ eresidual Þ
ð1Þ
It can be seen that almost complete superelastic recovery occurs in the first four cycles, while as the applied strain emax
162
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
Fig. 10 – Effect of Sn content on yielding stresses for plastic deformation rS, the ultimate tensile strength rult and the failure strain d in the Ti–7.5Nb–4Mo–xSn alloys.
Fig. 9 – TEM images showing the deformation microstructures of the deformed alloys (a) deformation twins in Ti–7.5Nb–4Mo alloy, (b) dislocations in Ti–7.5Nb–4Mo–4Sn alloy.
increases to 5%, the permanently remained strain eresidual is found, and it increases with the increase in applied strain. Similar cyclic tensile tests were also carried out on the Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–3Sn alloy. In order to clarify the effect of Sn content on the SE of the alloys, eresidual, erecoverable and Z of the Ti–7.5Nb–4Mo–xSn (x¼1–3) alloys are plotted as a function of the applied strain in Figs. 13 and 14, respectively. In general, for the alloys with different Sn content, erecoverable increases fast with applied strain increasing up to 4.0%, and then it increases slowly with the applied strain further increasing. In contrast, eresidual increases slowly with the applied strain up to 4.0%, and then it increases fast with the applied strain further increasing; In addition, Z remains unchanged (over 85%) as the applied strain changes from 1% to 4%, and then it decreases gradually with the applied strain further increasing. Moreover, it is clear that eresidual, erecoverable and
Z of Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloy are roughly identical as the applied strain increases from 1 to 4%, but the three parameters show a little difference with the applied strain further increasing. Compared with Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–2Sn alloy, Ti–7.5Nb–4Mo–3Sn alloy exhibits relatively high erecoverable, low eresidual and high Z, implying that it has better SE. In order to investigate the stability of SE of the alloys at room temperature, the cyclic tensile tests were carried out on the alloys with the Sn content of 2% and 3% at room temperature. To do it, the samples of the two alloys were elongated up to the strains of 5% followed by unloading at each cycle. The cyclic stress–strain curves for the Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–3Sn alloys are shown in Fig. 15(a) and (b), respectively. For the Ti–7.5Nb–4Mo–3Sn alloy, the first curve exhibits incomplete superelastic behavior with a plastic strain of 0.4%. After the fifth cycle, almost complete superelasitc behavior is observed with a narrow stress hysteresis. The yielding stress indicated by a blackheaded arrow, which is the stress for inducing martensitic transformation (sSIM), decreases with increasing cyclic number. However, they are constant after the fifth cycle irrespective of increasing cyclic number. Thus, the alloy exhibits good superelastic stability at room temperature. In comparison, the corresponding cycling curves of the Ti–7.5Nb–4Mo–1Sn alloy exhibits wider stress hysteresis and larger plastic strain at the same conditions, implying that the Ti–7.5Nb–4Mo–3Sn alloy has a better SE behavior than the Ti–7.5Nb–4Mo–1Sn alloy. Moreover, it is clear that sSIM of the Ti–7.5Nb–4Mo–3Sn alloy is higher than that of the Ti–7.5Nb–4Mo–1Sn alloy. It may be associated that the former has a lower Ms due to its higher Sn content, and thus, it is required a higher stress for inducing martensitic transformation in the former alloy. Taking it into the consideration, we carried out the cyclic tensile tests on the Ti–7.5Nb–4Mo–1Sn alloy with a relatively low maximum strain of 4.5%, the results are shown in Fig. 15(c). It can be seen that the Ti–7.5Nb–4Mo–1Sn alloy also exhibit a good superelastic stability at room temperature under the maximum strain of 4.5%.
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
163
Fig. 11 – SEM fractographs of Ti–7.5Nb–4Mo–xSn alloys. (a) Ti–7.5Nb–4Mo alloy, (b) Ti–7.5Nb–4Mo–1Sn alloy, (c) Ti–7.5Nb–4Mo–2Sn alloy, (d) Ti–7.5Nb–4Mo–3Sn alloy, (e) Ti–7.5Nb–4Mo–4Sn alloy.
4.
Conclusions
Ti–7.5Nb–4Mo–xSn (x¼1–4) alloys were hot-pressed with thickness reduction of 50% at the temperature of about 1073 K and then solution-treated at 1273 K for 1.8 ks followed by quenching them into ice water. The microstructural observations and the tensile tests at room temperature draw the following conclusions:
Fig. 12 – Stress–strain curves obtained by cyclic loadingunloading tensile tests for Ti–7.5Nb–4Mo–2Sn alloys.
1. The Sn addition has a significant influence on the microstructure of the Ti–Nb–Mo–Sn alloys. For the Sn free alloy, Ti–7.5Nb–4Mo, the microstructure of the alloy consists of bþoþa00 . The Sn addition is effective in suppressing the formation of a00 and o phase. As Sn content reaches 2% or more, the alloys exhibit a single b Structure. 2. Mechanical properties and the deformation mode of the alloys are dependent on the Sn content. With the increase
164
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
Fig. 13 – Permanently remained strain eresidual and recovered strain erecoverable plotted against tensile strain for Ti–7.5Nb–4Mo–xSn alloys.
Fig.14 – Strain recovery rate g plotted against tensile strain for Ti–7.5Nb–4Mo–xSn (x ¼1–3) alloys.
of Sn content, the deformation mode of the alloys changes from twining to a00 transformation and then to slip, and the yielding stress increases, but the elastic modulus and fracture strength and strain decrease. In addition, Ti–7.5Nb–4Mo–xSn (x¼ 2–4) exhibit low elastic modulus with the value of about 85 GP. 3. The superelasticity is found in Ti–7.5Nb–4Mo–xSn (x¼ 1, 2, 3) alloys, in which, Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–3Sn alloys exhibit a good superelasticity with relatively high sSIM at room temperature due to the athermal o phases containing or the solid-solution hardening effect of Sn. The maximum superelastic recovery strain of the two alloys is about 5.5%, and their strain recovery rateZis about 50% at the applied strain of 11%. 4. Ti–7.5Nb–4Mo–3Sn and Ti–7.5Nb–4Mo–1Sn alloys exhibit the stable SE at room temperature, but, under the
Fig. 15 – Stress–strain curves for Ti–7.5Nb–4Mo–1Sn alloy obtained by repeated loading to the maximum strain of (a) 5% and (c) 4.5% followed by unloading. (b) Stress–strain curves for Ti–7.5Nb–4Mo–3Sn alloy obtained by repeated loading to the maximum strain of 5% followed by unloading at room temperature.
maximum strain of 5%, the former alloy exhibits higher superelastic stability.
Acknowledgments This work was partially supported by Projects (10972190, 09A089, CX2011B253) supported by National Natural Science Foundation, Scientific Research Fund of Hunan Provincial Education Department and the Innovation Found Project for Graduate Student of Hunan Provice, China.
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
r e f e r e n c e s
Al-Zain, Y., Kim, H.Y., Hosoda, H., Nam, T.H., Miyazaki, S., 2010. Shape memory properties of Ti–Nb–Mo biomedical alloys. Acta Materialia 58, 4212–4223. Barras, C.D.J., Myers, K.A., 2000. Nitinol—Its Use in Vascular Surgery and Other Applications. European Journal of Vascular and Endovascular Surgery 19, 564–569. Brailovski, V., Prokoshkin, S., Gauthier, M., Inaekyan, K., Dubinskiy, S., Petrzhik, M., Filonov, M., 2011. Bulk and porous metastable beta Ti–Nb–Zr(Ta) alloys for biomedical applications. Materials Science and Engineering C 31, 643–657. Chai, Y.W., Kim, H.Y., Hosoda, H., Miyazaki, S., 2008. Interfacial defects in Ti–Nb shape memory alloys. Acta Materialia 56, 3088–3097. Frick, C.P., Lang, T.W., Spark, K., Gall, K., 2006. Stress-induced martensitic transformations and shape memory at nanometer scales. Acta Materialia 54, 2223–2234. Hanada, S., Izumi, O., 1982. Deformation and Fracture of Metastable Beta Titanium Alloys (Ti–15Mo–5Zr and Ti–15Mo–5Zr–3Al). Transactions of the Japan. Institute of Metal 23, 85–94. Hanada, S., Yoshio, T., Izumi, O., 1986. Plastic deformation mode of retained b phase in b-eutectiod Ti–Fe alloys. Journal of Materials Science 21, 866–870. Hu, Q.M., Li, S.J., Hao, Y.L., Yang, R., Johansson, B., Vitos, L., 2008. Phase stability and elastic modulus of Ti alloys containing Nb, Zr, and/or Sn from first-principles calculations. Applied Physics Letters 93, 121902. Ishiyam, S., Hanada, S., Izumi, O., 1991. Effect Phase of Zr, Sn and Al Additions on Deformation Mode and Beta Phase Stability of Metastable Beta Ti Alloys. ISIJ International 31, 807–813. Kim, H.Y., Ikehara, Y., Kim, J.I., Hosoda, H., Miyazaki, S., 2006. Martensitic transformation, shape memory effect and superelasticity of Ti–Nb binary alloys. Acta Materialia 54, 2419–2429. Kim, H.Y., Satoru, H., Kim, J.I., Hosoda, H., Miyazaki, S., 2004. Mechanical properties and shape memory behavior of Ti–Nb alloys. Materials Transactions 45, 2443–2448. Laheurte, P., Prima, F., Eberhardt, A., Gloriant, T., Warye, M., Patoor, E., 2010. Mechanical properties of low modulus b titanium alloys designed from the electronic approach. Journal of the Mechanical Behavior of Biomedical Materials 3, 565–573. Lee, C.M., Ju, C.P., Chern Lin, J.H., 2002. Structure-property relationship of cast Ti–Nb alloys. Journal of Oral Rehabilitation 29, 314–322. Lin, D.J., Chuang, C.C., Chern Lin, J.H., Lee, J.W., Ju, C.P., Yin, H.S., 2007. Bone formation at the surface of low modulus Ti–7.5Mo implants in rabbit femur. Biomaterials 28, 2582–2589. Miura, K., Yamada, N., Hanada, S., Jung, T.K., Itoi, E., 2011. The bone tissue compatibility of a new Ti–Nb–Sn alloy with a low Young’s modulus. Acta Biomaterialia 7, 2320–2326. Nozoe, F., Matsumoto, H., Jung, T.K., Watanabe, S., Saburi, T., Hanada, S., 2007. Effect of Low Temperature Aging on
165
Superelastic Behavior in Biocompatible b TiNbSn Alloy. Materials Transactions 48, 3007–3013. Oka, M., Taniguchi, Y., 1978. Stressed-induced products in metastable beta Ti–Mo and Ti–V alloys. Journal of the Japan Institute of Metals 42, 814–820. Oliveira, N.T.C., Aleixo, G., Caram, R., Guastaldi, A.C., 2007. Development of Ti–Mo alloys for biomedical applications: Microstructure and electrochemical characterization. Materials Science and Engineering A 452, 727–731. Oliveira, N.T.C., Guastaldi, A.C., Piazza, S., Sunseri, C., 2009a. Photo-electrochemical investigation of anodic oxide films on cast Ti–Mo alloys. I. Anodic behaviour and effect of alloy composition. Electrochimica Acta 54, 1395–1402. Oliveira, N.T.C., Guastaldi, A.C., 2009b. Electrochemical stability and corrosion resistance of Ti–Mo alloys for biomedical applications. Acta Biomaterialia 5, 399–405. Ohyama, H., Nishimura, T., 1995. Effects of Al and Sn on Deformation Mode in Ti–V Based b Titanium Alloy System. In: Proceedings of 8th World Conference on Titanium, Birmingham, UK. pp. 2273–2280. Ozaki, T., Matsumoto, H., Watanabe, S., Hanada, S., 2004. Beta Ti Alloys with Low Young’s Modulus. Materials Transactions 45, 2776–2779. Pharr, G.M., 1998. Measurement of Mechanical Properties by Ultra-low Load Indentation. Materials Science and Engineering A 253, 151–159. Souza, S.A., Manicardi, R.B., Ferrandini, P.L., Afonso, C.R.M., Ramirez, A.J., Caram, R., 2010. Effect of the addition of Ta on microstructure and properties of Ti–Nb alloys. Journal of Alloys and Compounds 504, 330–340. Sun, F., Hao, Y.L., Nowak, S., Gloriant, T., Laheurte, P., Prima, F., 2011. A thermo-mechanical treatment to improve the superelastic performances of biomedical Ti–26Nb and Ti–20Nb–6Zr (at%) alloys. Journal of the Mechanical Behavior of Biomedical Materials. http://dxdoi.org/10.1016/j.jmbbm.2011.06.003. Sutou, Y., Yamauchi, K., Takagi, T., Maeshima, T., Nishida, M., 2006. Mechanical properties of Ti–6 at% Mo–4 at% Sn alloy wires and their application to medical guidewire. Materials Science & Engineering, A: Structural Materials: Properties, Microstructure and Processing 438–440, 1097–1100. Wang, K., Yin, X.M., Chao, D.T., Milliman, C.L., Korsmeyer, S.J., 1996. BID: a novel BH3 domain-only death agonist. Genes and Development 10, 2859–2869. Wang, X.J., Li, Y.C., Xiong, J.Y., Hodgson, P.D., Wen, C., 2009. Porous TiNbZr alloy scaffolds for biomedical applications. Acta Biomaterialia 5, 3616–3624. Zhang, D.C., Lin, J.G., Jiang, W.J., Ma, M., Peng, Z.G., 2011. Shape memory and superelastic behavior of Ti–7.5Nb–4Mo–1Sn alloy. Materials and Design 32, 4614–4617. Zhang, L.C., Zhou, T., Alpay, S.P., Aindow, M., 2005. Origin of pseudoelastic behavior in Ti–Mo-based alloys. Applied Physics Letters 87, 241909.
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jmbbm
Research Paper
Effect of Sn addition on the microstructure and superelasticity in Ti–Nb–Mo–Sn Alloys D.C. Zhanga,b, S. Yanga,b, M. Weia,b, Y.F. Maoa,b, C.G. Tana,b, J.G. Lina,b,n a
Key Laboratory of Low Di-mensional Materials and Application Technology of Ministry of Education, Xiangtan University, Xiangtan, Hunan 411105, China b Faculty of Material and Optical-Electronic Physics, Xiangtan University, Xiangtan, Hunan 411105, China
ar t ic l e in f o
abs tra ct
Article history:
Ti–7.5 Nb–4Mo–xSn (x¼0–4 at%) alloys were developed as the biomedical materials. The effect of
Received 23 November 2011
the Sn content on the microstructure and superelasticity of the alloys was investigated. It is
Received in revised form
found that Sn is a strong stabilizer of the b phase, which is effective in suppressing the
16 April 2012
formation of a00 and o phases in the alloys. Moreover, the Sn addition has a significant impact
Accepted 21 April 2012
on the mechanical properties of the alloys. With the increase of Sn addition, the yield stress of
Available online 4 May 2012
the alloys increase, but their elastic modulus, the fracture strength and the ductility decrease,
Keywords:
and the deformation mode of the alloys changes from (322) twining to a00 transformation and
Titanium alloys
then to slip. The Ti–7.5 Nb–4Mo–1Sn and Ti–7.5 Nb–4Mo–3Sn alloys exhibit a good superelasticity
Microstructure
with a high sSIM due to the relatively high athermal o phases containing or the solution
Mechanical properties
hardening at room temperature. Under the maximum strain of 5%, Ti–7.5 Nb–4Mo–3Sn (at%)
Deformation mode
alloy exhibits higher super elastic stability than that of Ti–7.5 Nb–4Mo–1Sn alloy.
Superelasticity
1.
Introduction
Nickel–titanium alloys have been found to be the most useful of all the titanium-based shape memory alloys (SMAs) as biomedical materials due to their large recoverable strain value, low elastic modulus, and high corrosion resistance (Barras and Myers, 2000; Frick et al., 2006). However, it has been found that pure nickel may exhibit hypersensitivity and carcinogenicity to human body (Wang et al., 1996). There is a prohibition tendency for nickel–titanium alloy system in the biomedical field in European countries in recent years due to the problem of human nickel allergy. Therefore it is n Corresponding author at: Xiangtan University, Faculty of Material and Optical-Electronic Physics, Xiangtan, Hunan 411105, China. Tel./fax: þ86 731 58298119. E-mail addresses: [email protected], [email protected] (J.G. Lin).
1751-6161/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jmbbm.2012.04.017
& 2012 Elsevier Ltd. All rights reserved.
preferable to develop absolutely safe nickel-free titaniumbased shape memory alloys for biomedical applications. Recently, some Ni-free Ti alloys such as Ti–Nb (Souza et al., 2010; Brailovski et al., 2011; Laheurte et al., 2010; Sun et al., 2011; Chai et al., 2008; Wang et al., 2009; Miura et al., 2011; Al-Zain et al., 2010; Nozoe et al., 2007), Ti–Mo (Al-Zain et al., 2010; Lin et al., 2007; Oliveira et al., 2007; Oliveira et al., 2009a,b; Sutou et al., 2006; Zhang et al., 2005;) and Ti–Zr (Brailovski et al., 2011; Wang et al., 2009) have been extensively investigated as SMA due to their low elastic modulus, high biocompatibility and non-toxicity. It is well known that the SME of Ti–Nb alloys is attributed to the reversible martensitic transformation between a00 martensite and parent phase (Ozaki et al., 2004), and the martensitic transformation temperatures (Ms) and the shape memory behavior of Ti–Nb alloys are strongly dependent on their compositions. It has been reported that Ms of Ti–Nb alloys decreases from the melting point of Ti to 90 1C with the increase of the Nb content (Kim et al., 2004). A SME of 3% has
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
been obtained in Ti–(22–29 at%) Nb alloys. Moreover, for Ti–Nb alloys, their superelasticity (SE) strain can be improved up to 3.3% by cyclic loading–unloading training but their SME are not prone to be improved due to their low martensitic transformation stress (sSIM is about 200 MPa) (Kim et al., 2006). By contrast, Ti–Mo based alloys exhibit a relatively small superelasticity recovered strain (about 2.5%), but a high sSIM (about 450 MPa) (Zhang et al., 2005). Thus, in order to enhance the SE, and its stability, Miyazaki et al. added Mo to Ti–Nb alloys to develop the Ti–15Nb–4Mo alloy, which sSIM and a total recovery strain reached about 420 MPa and 3.5%, respectively (Al-Zain et al., 2010). In addition, Sn addition has an important influence on the SME and SE behavior of Ti–Nb or Ti–Mo alloys. It has been reported that Ms of Ti–Nb–Sn alloys decreases rapidly with increasing Sn content and a large SE strain was obtained in Ti–16Nb–4.9Sn alloy (Al-Zain et al., 2010). Moreover, Ozaki et al. (Ozaki et al. 2004) found that the addition of 2.5 at% Sn to Ti22 at% Nb could suppress the a00 martensitic transformation. While, Yang et al. ( Hu et al., 2008) found that Sn is a stronger b stabilizer than Nb in the Ti–Nb–Zr–Sn system from first-principles calculations, and a certain amount of Sn addition could reduce the elastic modulus of the alloy by suppressing o phase formation. Based on the previous work, we designed a quaternary Ti-based alloy, Ti–7.5Nb–4Mo–1Sn (in atom percent), according to the d-electron orbit theory, and the alloy exhibits stable SME and SE (Zhang et al., 2011). But how the Sn content affects the microstructure and the mechanical properties of the quaternary alloy is not reported yet. Therefore, in the present work, the Ti–7.5Nb–4Mo–xSn (x¼ 0–4) alloys were fabricated by an arc melting method, and the microstructure and mechanical properties of the alloy were investigated, emphasizing on clarifying the effect of the Sn content on the SME and SE of the alloy.
2.
157
Fig. 1 – Schematic geometry of the tensile test specimen.
Experimental
The ingot of the Ti alloy was fabricated by an arc melting method using pure Ti, Nb, Mo and Sn as raw materials. To ensure the composition homogeneity, the ingot was flipped and remelted five times. The nominal composition of the ingot is Ti–7.5Nb–4Mo–xSn (x¼0–4), and the ingots were hot-pressed by 50% at the temperature of about 1073 K. The samples for X-ray diffraction (XRD) measurement and tensile tests were cut from the ingot by an electro-discharge machine, and vacuum sealed in quartz tubes with Ar atmosphere. The samples were solutiontreated at 1273 K for 1.8 ks followed by quenching them into ice water. After that, the specimens were acid etched to remove the oxidized skin. The phase structures present in the microstructure were identified at room temperature by XRD using a Rigaku D/Max 2500PC diffractometer operated at 50 kV and 100 mA with a Cu Ka radiation (l¼1.5406 nm). To evaluate the SE of the alloy, tensile tests were carried out on an Instron 5569 universal testing machine, using the tensile specimens with a gage section of 1 mm 2.5 mm 8 mm, and the geometry of the samples is schematically shown in Fig. 1. The transmission electron microscopy (TEM) observations were conducted using a JEM-2100 operated at 160 kV. Samples for TEM were prepared by an electro-polishing in a solution of 200 vol% methanol, 20 vol% glycol and 10 vol% HNO3 at 15 V–20 V under 301C. Microhardness was obtained using a Vickers hardness tester (HVS-30Z)
Fig. 2 – XRD spectra Ti–7.5Nb–4Mo–xSn alloys.
of
the
solution-treated
with a load of 10 kgf. Elastic moduli were determined by a Tribo Indenter from monotonic load tests to a depth of 7000 nm using a Berkovich tip with a measured radius of 5000 nm. The data were analyzed using the Oliver and Pharr method (Pharr, 1998).
3.
Results and discussion
3.1.
Microstructure
Fig. 2 shows the XRD spectra of solution-treated Ti–7.5Nb–4Mo– xSn (x¼0–4) alloys. The alloy without Sn content (Ti–7.5Nb–4Mo)
158
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
mainly contains b and a00 phases (Fig. 2a), but a small amount of the athermal o phases are present in the alloy (Fig. 2 b). With Sn content increasing, the volume fraction of b phase increases, which is proved by the increasing intensity of the peaks of b phase in the Sn containing alloys. As the Sn content increases to more than 3%, the peaks from a00 and o phases disappear on the XRD pattern, but only the peaks from b phase can be observed in the alloys. More carful observations of the microstructures of the alloys were conducted on a TEM. Fig. 3 presents the TEM dark-field images of Ti–7.5Nb–4Mo–xSn (x¼ 0–2) alloys with the corresponding selected area diffraction patterns with zone axis [113]b in the inserts. The dark-field images were formed using the spot of the athermal o phase in the diffraction patterns marked by a circle in the diffraction patterns. It can be seen that the volume fraction of the athermal o phase decreases with the Sn addition increasing. The result is in coincident with that of XRD, which indicates that Sn is a strong b stabilizer, which is effective in suppressing the athermal o transformation. In the meantime, the presence of Sn decreases the start temperature of the matensitic phase transformation (Ms) of the alloys. With further increasing of Sn content, a00 transformation is suppressed due to the solid-solution
hardening effect of Sn. The findings are supported by some previous work (Ishiyam et al., 1991; Ozaki et al., 2004; Hu et al., 2008). Ishiyam et al. (Ishiyam et al., 1991) found that the martensitic transformation is difficult to occur in the alloys containing a large amount of Sn and A1. Kim et al. (Ozaki et al., 2004) found that Ms decreases rapidly with increasing Sn content in Ti–Nb–Sn alloys. Yang et al. (Hu et al., 2008) investigated the effects of the alloying elements, Nb, Zr and Sn on the stability of Ti alloys from the first principles calculations, and their results demonstrated that Sn is a stronger b stabilizer than Nb theoretically, and a certain amount of Sn addition could reduce the elastic modulus of the alloy by suppressing o phase formation in the Ti–Nb–Zr–Sn alloys. The optical microstructure of Ti–7.5Nb–4Mo–xSn alloys is illustrated in Fig. 4. It can be seen that all the alloys exhibit nearly equiaxed grain microstructure, and the grain size of the alloys decreases obviously with the Sn content increasing. Some fine acicular martensitic a00 phase can be observed in Ti–7.5Nb–4Mo alloy and Ti–7.5Nb–4Mo–1Sn alloy. While, as Sn content increases to 2% or more, the microstructure of the alloys only contains b phase. The result is in agreement with that of XRD analysis.
Fig. 3 – Dark-field images and the selected area diffraction patterns for (a) Ti–7.5Nb–4Mo, (b) Ti–7.5Nb–4Mo–1Sn, (c) Ti–7.5Nb–4Mo–2Sn alloys.
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
159
Fig. 4 – Optical micrographs of Ti–7.5Nb–4Mo–xSn alloys. (a) Ti–7.5Nb–4Mo alloy, (b) Ti–7.5Nb–4Mo–1Sn alloy, (c) Ti–7.5Nb–4Mo–2Sn alloy, (d) Ti–7.5Nb–4Mo–3Sn alloy, (e) Ti–7.5Nb–4Mo–4Sn alloy.
3.2.
Mechanical properties
To evaluate the effect of Sn content on the modulus of the alloys, nano-indenter tests were carried out on the alloys with different Sn contents. The elastic modulus of the alloys as a function of Sn content is presented in Fig. 5. It can be seen that the Sn-free alloy, Ti–7.5Nb–4Mo, yields a relative high elastic modulus, which is about 120 GPa. With the increase of Sn content, the modulus of the alloys sharply declines. As the Sn content is about 2%, the alloys exhibit the lowest modulus, which is 87 GP. The result is coincident with the change trend of the microstructure with the Sn content. As mentioned above, the microstructure of the Sn-free alloy contains bþoþa00 phases. The Sn addition suppresses the formation of oþa00 . As the Sn content increases from 1% to 4%, the phase constitution of the alloys is from bþo to b. While, it is well documented the modulus of the phases ranges in a diminishing sequence from o, a, a00 and b (Lee et al., 2002). Especially for the o phase, it has a significant influence on the elastic modulus of Ti-based alloy (Souza
Fig. 5 – Elastic modulus Ti–7.5Nb–4Mo–xSn alloys.
and
microhardness
of
et al., 2010). Therefore, the alloys containing 2%–4% Sn exhibit a relatively low modulus due to their single b structure. In addition, Fig. 5 also displays the change of the
160
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
microhardness of the alloys with the Sn content. It can be seen that the microhardness of the alloys sharply declines as the Sn content increases from 0 to 2% due to the content reduction of o and a00 phases, because o and a00 phases are harder than b phase. However, as the Sn content further increase to 3% and 4%, the microhardness of the alloys increases, which implies that the Sn addition has a strong solid-solution hardening effect. Fig. 6 shows stress–strain curves of solution-treated alloys obtained by tensile tests at room temperature. Ti–7.5Nb–4Mo– xSn (x¼ 1–3) alloys exhibit two-stage yielding as shown by arrows. It is supposed that the Ti–7.5Nb–4Mo–xSn (x¼1–3) alloys exhibit the SME or SE at room temperature. However, the Ti–7.5Nb–4Mo–xSn (x¼ 0, 4) alloys exhibit a single-stage yielding. To clearly illustrate the effect of the Sn content on the first yield stress (sSIM, defined by the 0.2% offset yield stress) indicated by the black-headed, the variation of sSIM as a function of the Sn content is present in Fig. 7. It can be seen that sSIM decreases with the increase of Sn content, and then it increases as the Sn content further increases, and the
Fig. 6 – Stress–strain curves obtained for Ti–7.5Nb–4Mo–xSn alloys.
Fig. 7 – rSIM of Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloy.
Ti–7.5Nb–4Mo–2Sn alloy exhibits the lowest sSIM. Moreover, XRD analysis on the alloys with the different Sn contents before and after tensile test clearly indicates that the peak of b phase at 38.641for all the alloys after the deformation splits into two peaks from b phase and a00 phase, respectively (see Fig. 8). By comparison, more a00 martensites are present in the deformed Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloys than the deformed Ti–7.5Nb–4Mo and Ti–7.5Nb–4Mo–4Sn alloys. Therefore, the martensitic transformation induced by stress is a dominant deformation mode in Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloys, which provides an evidence that the deformation induced a00 transformation tends to be suppressed by adding Sn in the alloys. For comparison, the deformation microstructures of the Ti–7.5Nb–4Mo and Ti–7.5Nb–4Mo–4Sn alloys were observed by using TEM. Fig. 9 illustrates TEM microstructures of the Ti–7.5Nb–4Mo and Ti–7.5Nb–4Mo–4Sn alloys after the deformation. A number of the deformation twins can be found in the deformed Ti–7.5Nb–4Mo alloy (see Fig. 9a). By contrast, many dislocations can be observed, but the deformation twins are hard to be found in the deformed Ti–7.5Nb–4Mo–4Sn alloy (see Fig. 9b). The occurrence of (322) deformation twins in the Ti–7.5Nb–4Mo alloy may be attributed to its relatively high athermal o phase content within b matrix. Oka et al. and Hanada et al. [Oka and Taniguchi, 1978; Hanada et al., 1986] explained that shuffling of one-half of atoms in the alloys with a large amount of athermal o phase into the direction different from that of twinning shear, which is needed to create (332) twinning, is ready to occur in an unstable b with o phase precipitation in it. While, the Ti–7.5Nb–4Mo–4Sn alloy deforms in the slip mode instead of the martensitic transformation, which may be attributed to its relatively high sSIM for the martensitic transformation due to the solid-solution hardening effect of the high Sn content. As a result, the deformation mode of the Ti–Nb–Mo–Sn alloys is governed by its chemical composition. With the increase of the Sn additions, the deformation mode of the alloy changes from twinning to a00 transformation, and then to slip. The result is in good agreement with that reported in the previous work [Ishiyam et al., 1991; Ohyama and Nishimura, 1995; Hanada and Izumi, 1982], in which, it was found that (332) deformation twins and a00 transformation are the dominant deformation modes in the alloys with the highly unstable b phases, while the slip is the dominant deformation mode in the alloys with the stable b phases. Further more, the yielding stresses sS for the plastic deformation, indicated by gray-headed arrows in Fig. 6, the ultimate tensile strength sult and the fracture strain d of the alloys are measured from the stress–strain curves of Ti–7.5Nb–4Mo–xSn (x¼0–4) alloys, which are plotted against Sn content in Fig. 10. It is found that sS increases with Sn content increasing. On the other hand, d decreases with Sn content increasing from 47% for the Ti–7.5Nb–4Mo to 26% for the Ti–7.5Nb–4Mo–4Sn. Irrespective of the solid-solution strengthening effect of the Sn addition, the Ti–Nb–Mo–Sn alloys exhibit a good ductility. Fig. 11 displays the SEM fractographs of the Ti–7.5Nb–4Mo–xSn (x¼0–4) alloy. The fracture surfaces of the alloys are covered with dimples, suggesting that the fracture of the alloys is in a ductile fracture mode. It also can be seen that the dimple’s size and depth reduce with Sn
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
161
Fig. 8 – XRD profiles of the Ti–7.5Nb–4Mo–xSn alloys before and after deformation (a) x ¼0, (b) x ¼1, (c) x ¼2, (d) x¼ 3, (e) x ¼4.
content increasing, which is coincident with the change trend of the ductility of the alloys with the Sn content.
3.3.
Superelasticity
Fig. 12 shows stress–strain curves of the Ti– 7.5Nb–4Mo–2Sn alloy. Each stress–strain curve was obtained by a loading and unloading cycle at room temperature. The similar measurement was repeated by increasing the maximum strain upon loading in the same sample.
In order to characterize the SE of the alloy, tow types of strains and a recovery rate are defined as follows: (1) the permanently remained strain eresidual after unloading (see Fig. 12); (2) the recovered strain erecoverable upon unloading (see Fig. 12); (3) the strain recovery rate, which can be described as Z ¼ erecoverable =ðerecoverable þ eresidual Þ
ð1Þ
It can be seen that almost complete superelastic recovery occurs in the first four cycles, while as the applied strain emax
162
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
Fig. 10 – Effect of Sn content on yielding stresses for plastic deformation rS, the ultimate tensile strength rult and the failure strain d in the Ti–7.5Nb–4Mo–xSn alloys.
Fig. 9 – TEM images showing the deformation microstructures of the deformed alloys (a) deformation twins in Ti–7.5Nb–4Mo alloy, (b) dislocations in Ti–7.5Nb–4Mo–4Sn alloy.
increases to 5%, the permanently remained strain eresidual is found, and it increases with the increase in applied strain. Similar cyclic tensile tests were also carried out on the Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–3Sn alloy. In order to clarify the effect of Sn content on the SE of the alloys, eresidual, erecoverable and Z of the Ti–7.5Nb–4Mo–xSn (x¼1–3) alloys are plotted as a function of the applied strain in Figs. 13 and 14, respectively. In general, for the alloys with different Sn content, erecoverable increases fast with applied strain increasing up to 4.0%, and then it increases slowly with the applied strain further increasing. In contrast, eresidual increases slowly with the applied strain up to 4.0%, and then it increases fast with the applied strain further increasing; In addition, Z remains unchanged (over 85%) as the applied strain changes from 1% to 4%, and then it decreases gradually with the applied strain further increasing. Moreover, it is clear that eresidual, erecoverable and
Z of Ti–7.5Nb–4Mo–xSn (x¼ 1–3) alloy are roughly identical as the applied strain increases from 1 to 4%, but the three parameters show a little difference with the applied strain further increasing. Compared with Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–2Sn alloy, Ti–7.5Nb–4Mo–3Sn alloy exhibits relatively high erecoverable, low eresidual and high Z, implying that it has better SE. In order to investigate the stability of SE of the alloys at room temperature, the cyclic tensile tests were carried out on the alloys with the Sn content of 2% and 3% at room temperature. To do it, the samples of the two alloys were elongated up to the strains of 5% followed by unloading at each cycle. The cyclic stress–strain curves for the Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–3Sn alloys are shown in Fig. 15(a) and (b), respectively. For the Ti–7.5Nb–4Mo–3Sn alloy, the first curve exhibits incomplete superelastic behavior with a plastic strain of 0.4%. After the fifth cycle, almost complete superelasitc behavior is observed with a narrow stress hysteresis. The yielding stress indicated by a blackheaded arrow, which is the stress for inducing martensitic transformation (sSIM), decreases with increasing cyclic number. However, they are constant after the fifth cycle irrespective of increasing cyclic number. Thus, the alloy exhibits good superelastic stability at room temperature. In comparison, the corresponding cycling curves of the Ti–7.5Nb–4Mo–1Sn alloy exhibits wider stress hysteresis and larger plastic strain at the same conditions, implying that the Ti–7.5Nb–4Mo–3Sn alloy has a better SE behavior than the Ti–7.5Nb–4Mo–1Sn alloy. Moreover, it is clear that sSIM of the Ti–7.5Nb–4Mo–3Sn alloy is higher than that of the Ti–7.5Nb–4Mo–1Sn alloy. It may be associated that the former has a lower Ms due to its higher Sn content, and thus, it is required a higher stress for inducing martensitic transformation in the former alloy. Taking it into the consideration, we carried out the cyclic tensile tests on the Ti–7.5Nb–4Mo–1Sn alloy with a relatively low maximum strain of 4.5%, the results are shown in Fig. 15(c). It can be seen that the Ti–7.5Nb–4Mo–1Sn alloy also exhibit a good superelastic stability at room temperature under the maximum strain of 4.5%.
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
163
Fig. 11 – SEM fractographs of Ti–7.5Nb–4Mo–xSn alloys. (a) Ti–7.5Nb–4Mo alloy, (b) Ti–7.5Nb–4Mo–1Sn alloy, (c) Ti–7.5Nb–4Mo–2Sn alloy, (d) Ti–7.5Nb–4Mo–3Sn alloy, (e) Ti–7.5Nb–4Mo–4Sn alloy.
4.
Conclusions
Ti–7.5Nb–4Mo–xSn (x¼1–4) alloys were hot-pressed with thickness reduction of 50% at the temperature of about 1073 K and then solution-treated at 1273 K for 1.8 ks followed by quenching them into ice water. The microstructural observations and the tensile tests at room temperature draw the following conclusions:
Fig. 12 – Stress–strain curves obtained by cyclic loadingunloading tensile tests for Ti–7.5Nb–4Mo–2Sn alloys.
1. The Sn addition has a significant influence on the microstructure of the Ti–Nb–Mo–Sn alloys. For the Sn free alloy, Ti–7.5Nb–4Mo, the microstructure of the alloy consists of bþoþa00 . The Sn addition is effective in suppressing the formation of a00 and o phase. As Sn content reaches 2% or more, the alloys exhibit a single b Structure. 2. Mechanical properties and the deformation mode of the alloys are dependent on the Sn content. With the increase
164
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
Fig. 13 – Permanently remained strain eresidual and recovered strain erecoverable plotted against tensile strain for Ti–7.5Nb–4Mo–xSn alloys.
Fig.14 – Strain recovery rate g plotted against tensile strain for Ti–7.5Nb–4Mo–xSn (x ¼1–3) alloys.
of Sn content, the deformation mode of the alloys changes from twining to a00 transformation and then to slip, and the yielding stress increases, but the elastic modulus and fracture strength and strain decrease. In addition, Ti–7.5Nb–4Mo–xSn (x¼ 2–4) exhibit low elastic modulus with the value of about 85 GP. 3. The superelasticity is found in Ti–7.5Nb–4Mo–xSn (x¼ 1, 2, 3) alloys, in which, Ti–7.5Nb–4Mo–1Sn and Ti–7.5Nb–4Mo–3Sn alloys exhibit a good superelasticity with relatively high sSIM at room temperature due to the athermal o phases containing or the solid-solution hardening effect of Sn. The maximum superelastic recovery strain of the two alloys is about 5.5%, and their strain recovery rateZis about 50% at the applied strain of 11%. 4. Ti–7.5Nb–4Mo–3Sn and Ti–7.5Nb–4Mo–1Sn alloys exhibit the stable SE at room temperature, but, under the
Fig. 15 – Stress–strain curves for Ti–7.5Nb–4Mo–1Sn alloy obtained by repeated loading to the maximum strain of (a) 5% and (c) 4.5% followed by unloading. (b) Stress–strain curves for Ti–7.5Nb–4Mo–3Sn alloy obtained by repeated loading to the maximum strain of 5% followed by unloading at room temperature.
maximum strain of 5%, the former alloy exhibits higher superelastic stability.
Acknowledgments This work was partially supported by Projects (10972190, 09A089, CX2011B253) supported by National Natural Science Foundation, Scientific Research Fund of Hunan Provincial Education Department and the Innovation Found Project for Graduate Student of Hunan Provice, China.
journal of the mechanical behavior of biomedical materials 13 (2012) 156 –165
r e f e r e n c e s
Al-Zain, Y., Kim, H.Y., Hosoda, H., Nam, T.H., Miyazaki, S., 2010. Shape memory properties of Ti–Nb–Mo biomedical alloys. Acta Materialia 58, 4212–4223. Barras, C.D.J., Myers, K.A., 2000. Nitinol—Its Use in Vascular Surgery and Other Applications. European Journal of Vascular and Endovascular Surgery 19, 564–569. Brailovski, V., Prokoshkin, S., Gauthier, M., Inaekyan, K., Dubinskiy, S., Petrzhik, M., Filonov, M., 2011. Bulk and porous metastable beta Ti–Nb–Zr(Ta) alloys for biomedical applications. Materials Science and Engineering C 31, 643–657. Chai, Y.W., Kim, H.Y., Hosoda, H., Miyazaki, S., 2008. Interfacial defects in Ti–Nb shape memory alloys. Acta Materialia 56, 3088–3097. Frick, C.P., Lang, T.W., Spark, K., Gall, K., 2006. Stress-induced martensitic transformations and shape memory at nanometer scales. Acta Materialia 54, 2223–2234. Hanada, S., Izumi, O., 1982. Deformation and Fracture of Metastable Beta Titanium Alloys (Ti–15Mo–5Zr and Ti–15Mo–5Zr–3Al). Transactions of the Japan. Institute of Metal 23, 85–94. Hanada, S., Yoshio, T., Izumi, O., 1986. Plastic deformation mode of retained b phase in b-eutectiod Ti–Fe alloys. Journal of Materials Science 21, 866–870. Hu, Q.M., Li, S.J., Hao, Y.L., Yang, R., Johansson, B., Vitos, L., 2008. Phase stability and elastic modulus of Ti alloys containing Nb, Zr, and/or Sn from first-principles calculations. Applied Physics Letters 93, 121902. Ishiyam, S., Hanada, S., Izumi, O., 1991. Effect Phase of Zr, Sn and Al Additions on Deformation Mode and Beta Phase Stability of Metastable Beta Ti Alloys. ISIJ International 31, 807–813. Kim, H.Y., Ikehara, Y., Kim, J.I., Hosoda, H., Miyazaki, S., 2006. Martensitic transformation, shape memory effect and superelasticity of Ti–Nb binary alloys. Acta Materialia 54, 2419–2429. Kim, H.Y., Satoru, H., Kim, J.I., Hosoda, H., Miyazaki, S., 2004. Mechanical properties and shape memory behavior of Ti–Nb alloys. Materials Transactions 45, 2443–2448. Laheurte, P., Prima, F., Eberhardt, A., Gloriant, T., Warye, M., Patoor, E., 2010. Mechanical properties of low modulus b titanium alloys designed from the electronic approach. Journal of the Mechanical Behavior of Biomedical Materials 3, 565–573. Lee, C.M., Ju, C.P., Chern Lin, J.H., 2002. Structure-property relationship of cast Ti–Nb alloys. Journal of Oral Rehabilitation 29, 314–322. Lin, D.J., Chuang, C.C., Chern Lin, J.H., Lee, J.W., Ju, C.P., Yin, H.S., 2007. Bone formation at the surface of low modulus Ti–7.5Mo implants in rabbit femur. Biomaterials 28, 2582–2589. Miura, K., Yamada, N., Hanada, S., Jung, T.K., Itoi, E., 2011. The bone tissue compatibility of a new Ti–Nb–Sn alloy with a low Young’s modulus. Acta Biomaterialia 7, 2320–2326. Nozoe, F., Matsumoto, H., Jung, T.K., Watanabe, S., Saburi, T., Hanada, S., 2007. Effect of Low Temperature Aging on
165
Superelastic Behavior in Biocompatible b TiNbSn Alloy. Materials Transactions 48, 3007–3013. Oka, M., Taniguchi, Y., 1978. Stressed-induced products in metastable beta Ti–Mo and Ti–V alloys. Journal of the Japan Institute of Metals 42, 814–820. Oliveira, N.T.C., Aleixo, G., Caram, R., Guastaldi, A.C., 2007. Development of Ti–Mo alloys for biomedical applications: Microstructure and electrochemical characterization. Materials Science and Engineering A 452, 727–731. Oliveira, N.T.C., Guastaldi, A.C., Piazza, S., Sunseri, C., 2009a. Photo-electrochemical investigation of anodic oxide films on cast Ti–Mo alloys. I. Anodic behaviour and effect of alloy composition. Electrochimica Acta 54, 1395–1402. Oliveira, N.T.C., Guastaldi, A.C., 2009b. Electrochemical stability and corrosion resistance of Ti–Mo alloys for biomedical applications. Acta Biomaterialia 5, 399–405. Ohyama, H., Nishimura, T., 1995. Effects of Al and Sn on Deformation Mode in Ti–V Based b Titanium Alloy System. In: Proceedings of 8th World Conference on Titanium, Birmingham, UK. pp. 2273–2280. Ozaki, T., Matsumoto, H., Watanabe, S., Hanada, S., 2004. Beta Ti Alloys with Low Young’s Modulus. Materials Transactions 45, 2776–2779. Pharr, G.M., 1998. Measurement of Mechanical Properties by Ultra-low Load Indentation. Materials Science and Engineering A 253, 151–159. Souza, S.A., Manicardi, R.B., Ferrandini, P.L., Afonso, C.R.M., Ramirez, A.J., Caram, R., 2010. Effect of the addition of Ta on microstructure and properties of Ti–Nb alloys. Journal of Alloys and Compounds 504, 330–340. Sun, F., Hao, Y.L., Nowak, S., Gloriant, T., Laheurte, P., Prima, F., 2011. A thermo-mechanical treatment to improve the superelastic performances of biomedical Ti–26Nb and Ti–20Nb–6Zr (at%) alloys. Journal of the Mechanical Behavior of Biomedical Materials. http://dxdoi.org/10.1016/j.jmbbm.2011.06.003. Sutou, Y., Yamauchi, K., Takagi, T., Maeshima, T., Nishida, M., 2006. Mechanical properties of Ti–6 at% Mo–4 at% Sn alloy wires and their application to medical guidewire. Materials Science & Engineering, A: Structural Materials: Properties, Microstructure and Processing 438–440, 1097–1100. Wang, K., Yin, X.M., Chao, D.T., Milliman, C.L., Korsmeyer, S.J., 1996. BID: a novel BH3 domain-only death agonist. Genes and Development 10, 2859–2869. Wang, X.J., Li, Y.C., Xiong, J.Y., Hodgson, P.D., Wen, C., 2009. Porous TiNbZr alloy scaffolds for biomedical applications. Acta Biomaterialia 5, 3616–3624. Zhang, D.C., Lin, J.G., Jiang, W.J., Ma, M., Peng, Z.G., 2011. Shape memory and superelastic behavior of Ti–7.5Nb–4Mo–1Sn alloy. Materials and Design 32, 4614–4617. Zhang, L.C., Zhou, T., Alpay, S.P., Aindow, M., 2005. Origin of pseudoelastic behavior in Ti–Mo-based alloys. Applied Physics Letters 87, 241909.