NiTi alloys (nitinol) show the shape memory effect, superelastic behavior, high strength, good corrosion resistance and biocompatibility. These characteristics make them of interest for medical applications, such as dental medicine or manufacture of stents, i.e. tubular implants serving to restore blood vessels. In manufacture of stents, nitinol experience various heat treating procedures which may significantly, and sometimes negatively, affect its mechanical properties. For this reason, this work is aimed in determination of mechanical properties of nitinol short-time heat treated at around 500°C. The temperature of 500°C was selected, because in manufacture of stents, shape setting is a step generally performed at about 500°C.
INTRODUCTION
Nitinol, i.e. nearly equi-atomic Ni-Ti alloy, became of interest in production of various medical devices, such as stents, due to shape memory behavior, superelasticity, biocompatibility, corrosion resistance and good mechanical properties. Stents are often manufactured from nitinol wires and, during processing of these wires, nitinol experiences various heat treatment and forming procedures to achieve shape, mechanical properties and transformation behavior suitable for the final application. Final steps in production of superelastic nitinol wires are often cold drawing to a desired diameter followed by straight annealing. Straight annealing consists of heating a pre-loaded (20-100 MPa) cold drawn wire at an appropriate temperature (450-700°C). It ensures an optimum straight shape and desired functional properties of a wire. A very important step in a following fabrication of stents from the straight annealed superelastic wire is the shape setting. It involves a short (several minutes) heat treatment of the wire which is wound in a desired pattern on a mandrel. The shape setting treatment is generally carried out at moderate temperatures (around 500°C) and its purpose is to induce relaxation of a material for achievement of a desired stable shape of an implant.Moderate temperatures and short times are used to prevent the permanent deformation of implants and to maintain their superelastic behavior.
In general, nitinol shape memory alloys exhibit three phases, the high-temperature B2 austenite phase (structure of CsCl), low-temperature B19´ martensite phase (monoclinic structure) and intermediatetemperature R-phase (rhombohedral structure) [1-11]. Transformations of these phases are of a great importance, because they determine the superelastic and shape memory characteristics of nitinol, as well as its mechanical and functional properties and performance. These transformations can proceed by variousways, B2↔B19´, B2↔R, R↔B19´, depending on thermal and mechanical history of alloys. The direct transformation of austenite B2 to martensite B19´upon cooling generally occurs when an alloy is in a solution annealed state, i.e. annealed at a high temperature and water quenched. Upon subsequent ageing, the solid solution decomposes to form Ti3Ni4 precipitates. All stages of precipitation strongly influence both phase transformations and mechanical characteristics of a material, mainly yield strength and tensile strength.Strength may be increased by an elastic lattice stress introduced by coherent and semi-coherent Ti3Ni4 precipitates
To our best knowledge, relatively little information is available on changes mechanical properties of straight annealed nitinol wires due to a short-term heat treatment at moderate temperatures. For this reason, our study is concerned with the short-time annealing of a nitinol wire commonly employed in stent fabrication at 410-540°C. Influence of these heat treatments on mechanical properties is the main objective of our study.
EXPERIMENT
Nitinol wire having a thickness of 0.2 mm and chemical composition of 50.9 at. % Ni was used in our experiment. The wire was provided by an industrial supplier and it was produced using a standard procedure, including cold drawing (46 % deformation) and subsequent straight annealing. The wire was annealed at temperatures of 410, 435, 460, 485, 500, 515 and 540°C. Annealing periods in our experiment were short, i.e. 2, 4, 8 and 16 minutes. The moderate temperatures and short times are generally used in the shape setting of stents. Annealing was carried out in a resistance furnace under a flow of argon. In each heat treatment regime, approximately 20 cm long pieces of the wire were placed to the pre-heated furnace and their temperature was monitored by a thermocouple positioned near them. After annealing, samples were withdrawn from the furnace and immediately quenched into water at 20°C.
Phase composition of samples was determined by using x-ray diffraction (XRD) (X´Pert Philips, 30 mA, 40 kV, X-ray radiation Cu Kα). Structure of wires was examined by optical microscopy (OM) and scanning electron microscopy (SEM, Hitachi S4700, acceleration voltage of 15 kV), ground and polished samples were etched by a solution of 3 ml HF, 5 ml HNO3, 20 ml CH3COOH and 72 ml H2O for this purpose. Tensile tests were conducted on an Instron 3343 tensile machine at a strain rate of 8.3⋅10-4 s-1 and at a temperature of 37°C. All samples showed well developed upper plateau on the stress-strain diagrams, suggesting that matrix of the alloy was dominated by austenite B2 phase at this temperature. During all tensile tests, tensile loading increased up to the fracture to determine both the onset of stress-induced B2→B19´ transformation (onset of the upper plateau) and tensile strength.
RESULTS AND DISCUSSION
Structures of the as-received wire and wire heat treated at 540°C/8 min are presented in Fig.1. The asreceived structure (Fig.1a) consists of fully recrystallized B2 austenitic grains with no preferential orientation relative to the cold drawing direction. It suggests that a temperature of the straight annealing was sufficient to induce recrystallization of the cold drawn wire. Average grain size for the as-received wire computed from 100 grains is 18 μm. The measurement of grain size after various heat treatment regimes imply that the short-time heat treatments at 410-540°C applied on the wire do not induce any significant changes of the grain size. Although certain small variations of the values are observed, they can not be attributed to the heat treatment and recrystallization, because they show no dependence on heat treatment regime. In addition to austenite grains, there are also non-metallic inclusions in the wire structure (marked by arrows in Fig.1a). These inclusions are based on titanium carbides and result from a contamination of the melt by a graphite crucible during melting process.
High resolution SEM image of the wire annealed at 540°C/8min in Fig.1b reveals the presence of fine Precipitates in austenitic matrix. The precipitates of about 50 nm in size appear as light particles and are relatively homogeneously distributed in the observed area. According to XRD, these precipitates correspond to rhombohedral Ti3Ni4 phase. As will be demonstrated later, precipitation process considerably affects mechanical properties of the alloy.
Fig.1. a) Optical micrograph of the longitudinally sectioned as-received wire showing recrystallized B2 grains. Arrays mark inclusions resulting from the melting process. b) SEM micrograph of the wire annealed at 540°C/8 min showing the presence of Ti3Ni4 precipitates (light).
Fig.2 shows a typical stress-strain diagram of the NiTi wire. This diagram consists of four distinct regions: elastic deformation of B2 phase (I), stress-induced B2→B19´ transformation manifesting itself as a plateau (II), elastic deformation of B19´ phase (III) and plastic deformation of B19´ phase (IV) occurring by dislocation . Stress-strain diagrams for the wire heat-treated by the applied regimes are similar to that in Fig.2, because they exhibit identical stages of deformation.
Therefore, we were able to determine the influence of heat treatment of the wire on the tensile strength (UTS) and on the stress at the onset of B2→B19´ transformation (σB2→B19´), see Fig.2. Hereafter, the
latter is denoted as transformation stress (TS).
Fig.2. Stress-strain diagram of the as-received wire.
Values of UTS and TS for the as-received wire are 1653 MPa and 620 MPa, respectively. Despite the short heat treatment times, there is a strong dependence of both mechanical characteristics on heat treatment, as is presented in Table 1, where both UTS and TS are listed as functions of annealing time at various temperatures. It is observed in this Table that lower temperatures of 410-460°C induce a slight strengthening of the wire. Particularly, heat treatments at 410-435°C/4-16 min are efficient in this sense, because they induce an increase of UTS from the original 1653 to about 1700 MPa. After annealing at 460°C, strengthening is less pronounced but still evident. Strengthening may be a consequence of Ti3Ni4 precipitation from a slightly supersaturated NiTi matrix. Based on the previously reported data [12], precipitation process achieves a maximum rate at about 450°C. In our case, however, the strengthening effect at temperatures close to 450°C is relatively weak (50 MPa), indicating that a driving force of precipitation, i.e. nickel supersaturation, is low in our experiment. Probably, a large part of the excess nickel in the matrix was already consumed for the precipitation during straight annealing performed after cold drawing of the wire.
Table 1. Ultimate tensile strength (UTS) and transformation stress (TS) of heat treated wire.
| heat treatment | UTS [MPa] | TS [MPa] | ||||||
|---|---|---|---|---|---|---|---|---|
| 2 min | 4 min | 8 min | 16 min | 2 min | 4 min | 8 min | 16 min | |
| 410°C | 1685 | 1695 | 1699 | 1710 | 630 | 630 | 629 | 601 |
| 435°C | 1679 | 1688 | 1699 | 1696 | 620 | 614 | 603 | 576 |
| 460°C | 1670 | 1678 | 1677 | 1673 | 615 | 608 | 590 | 574 |
| 485°C | 1582 | 1634 | 1563 | 1574 | 590 | 595 | 570 | 550 |
| 500°C | 1595 | 1567 | 1530 | 1449 | 594 | 570 | 550 | 502 |
| 515°C | 1488 | 1522 | 1411 | 1400 | 577 | 580 | 520 | 500 |
| 540°C | 1300 | 1259 | 1246 | - | 560 | 560 | - | - |
| as-received | 1653 | 620 | ||||||
In contrast, higher temperatures of 485-540°C lead to a reduction of strength. At 485 and 500°C, 2 minute annealing results in a strength decrease by about 50 MPa. Longer annealing at 485°C have a negligible effect on the strength, while annealing at 500°C leads to an additional strength drop to about 1450 MPa. Temperatures above 500°C significantly accelerate the processes occurring in the wire and, as a result, a strength reduction after a heat treatment at 540°C/2 min is more than 300 MPa. At these temperatures, precipitate nucleation rate decreases, due to a low driving force. On the other hand, the growth rate of existing precipitates increases which is a consequence of an accelerated diffusion [12]. Therefore, a lower number of coarser precipitates is present in the matrix and this coarsening may be reason for the observed strength reduction. The precipitate coarsening is illustrated in Fig.1b which presents a sample annealed at 540°C/8 h containing precipitates of about 50 nm in size. It was shown in several papers that such large precipitates are not very effective for strengthening [1]. Beside the precipitate coarsening, another source of the strength reduction due to annealing at higher temperatures may be a loss of precipitate/matrix interface coherency [6].
TS values behave in a way slightly different from the tensile strength, see Table 1, because all annealing temperatures result in TS reduction, except for a very short annealing at 410°C. As in the previous case, the rate of TS reduction increases with increasing temperature. It is well known that the precipitation of Ni-rich Ti3Ni4 particles from the NiTi matrix leads to a nickel depletion in the matrix, resulting in an increase of transformation temperatures and in a proportional decrease of the transformation stress, according to the Clausius-Clapeyron relationship [13]. Moreover, coherent and semicoherent precipitates formed at the lower temperature region (410-460°C) may induce a lattice stress which contributes to this TS decrease [6].
CONCLUSIONS
It is shown in the present work that the short-time (2-16 min) heat treatments of the straight annealed nitinol wire at moderate temperatures strongly influence its tensile strength and transformation stress. Annealing temperatures between 410-460°C improve the strength to some extent. At higher temperatures above 485°C, the wire behaves in an opposite manner, i.e. strength reduces. The observed development of mechanical properties and transformation behavior is related to precipitation processes occurring in the wire during heat treatment.