NiTiHf-based shape memory alloys

Abstract

NiTiHf-based shape memory alloys have been receiving considerable attention for high temperature, high strength and two-way shape memory applications since they could have transformation temperatures above 100°C, shape memory effect under high stress (above 500 MPa) and superelasticity above 100°C. Moreover, their shape memory properties can be tailored by microstructural engineering. However, NiTiHf-based alloys have some drawbacks such as low ductility and high slope in stress induced martensite transformation region. In order to overcome these limitations, studies have been focused on microstructural engineering by aging, alloying and processing. It has been revealed that microstructural control is crucial to govern the shape memory properties (e.g. transformation temperatures, matrix strength, shape recovery strain, twinning type, etc.) of NiTiHf-based alloys. A summary of the most recent improvements on selected NiTiHf-based systems is presented to point out their significant shape memory properties, effects of alloying, aging and microstructure of transforming phases and precipitates.

Keywords

NiTiHf High temperature shape memory alloys Microstructure control Precipitation hardening Superelasticity Work output

Introduction

Shape memory alloys (SMAs) are a unique class of smart materials with the ability of changing their shapes depending on the applied temperature, stress and in special case of ferromagnetic alloys, magnetic field. Shape memory alloys can produce very high actuation strains, stresses and work outputs as they undergo reversible martensitic phase transformation.¹ In addition to their remarkable properties in actuation, vibration damping, noise reduction and sensing, they are compact, robust, lightweight, frictionless, quiet, environment-friendly (no hydraulic liquids), easy to inspect and have low aftermarket costs for inspection and maintenance.^2–4 Shape memory alloys are playing a growing role in supplying key actuation forces and sealing functions in oil and gas, automotive, aerospace and biomedical industries.^2–4 The ability to remain elastic under very large deformation makes SMAs potential candidates for superelastic devices for civil structures.^5,6 Moreover, their superelasticity, good corrosion resistance, biological and magnetic resonance compatibility and high bending resistance resulted in their employment as the biomedical devices in the orthodontic, orthopaedic, vascular, neurosurgical fields.^7,8

Among the various SMA systems, NiTi alloys have good dimensional stability, shape memory properties, ductility and workability. Currently, NiTi alloys are the most commercially viable SMAs and practically being used in various medical and engineering applications where the operating temperature is below 100°C.⁹ It has been found that the transformation temperatures (TTs) of NiTi can be adjusted by tailoring the stoichiometry or formation of precipitates.^10,11 However, the TTs of binary NiTi cannot be increased above 120°C. The development of a shape memory material with properties similar to those of near equiatomic NiTi, but with higher strength and TTs, especially above 100°C, is urgently needed for a broad range of applications in the aerospace, automotive and oil and gas industries to serve as compact actuators for flow and clearance controls, actuation tubes for rotors, moving or morphing surfaces as well as inlet/exhaust configurations, linear actuators and sealants.^2–4

Ternary element addition to NiTi alloys is the most promising method to obtain commercially available high temperature shape memory alloys (HTSMAs) in the near future.¹² Ternary element addition should not only increase the TTs, but also help to maintain the good mechanical and shape memory properties of NiTi alloys. It has been found that the addition of Hf, Zr, Pd, Pt and Au elements to NiTi increases its TTs.^2,9 Among those elements, Pd, Pt and Au are very expensive and will limit the use of their respective ternary alloys to some critical applications only (i.e. aerospace), while Zr is associated with high oxygen affinity.^2,9,13 Among the potential HTSMAs, due to its low cost, medium ductility and high work output NiTiHf seems to be the most encouraging HTSMA for a wide range of applications in the critical 100–300°C temperature range.¹²

The TTs of NiTiHf alloys do not increase much up to 10 at- Hf content, however, at chemical concentrations higher than 10 at-, they tend to increase linearly up to 525°C for 30 at- Hf when Hf is added at the expense of Ti.^14–16 Transformation temperatures of NiTiHf alloys are not notably affected by a change in Ni composition as long as the alloys are Ni-lean, but dropped steeply when Ni content is increased beyond the equiatomic (50 at-) composition, consistent with the behaviour of NiTi alloys.^15–17

The main disadvantages of Ni-lean NiTiHf alloys are their large hysteresis (>50°C), poor ductility at room temperature, lack of cyclic stability due to the high stress for the reorientation of martensite and detwinning, the low strength for slip and poor formability.^2,18 It should be noted that Wojcik¹⁹ studied the possibility of the commercialisation of the NiTiHf (Hf content less than 10 at-) alloys and showed that hot rolling can be successfully utilised to produce thin sheets. Another drawback of the alloy is the absence of stress plateau during phase transformation that results in the lack of superelasticity. This behaviour has been attributed to the simultaneous occurrence of stress induced martensite (SIM) and dislocation slip.^20,21 To increase the strength for slip, NiTiHf alloys were severely deformed that resulted in increased recoverable transformation strain, decreased irrecoverable strain levels and thermal hysteresis under constant stress experiments, as well as improved cyclic stability.¹⁸ However, no superelasticity was observed due to large hysteresis and low material strength.

Precipitation strengthening has been used to improve the mechanical properties of NiTiHf alloys as a successful method. Meng et al.^22,23 revealed that it is possible to form precipitates in Ni-rich NiTiHf alloys and TTs can be increased drastically to temperatures above 100°C. They have also reported that coherent precipitates increase the matrix strength and enhance the thermal stability.²² If the chemical composition is slightly Ni-rich with high Hf content (15 to 20 at- Hf), fine nanometer size precipitates which are face centred orthorhombic structure, simply referred to as the H-phase,^24,25 are formed upon aging treatments. The formation of fine precipitates provides high resistance to dislocation motion resulting in exceptional strength and stability limiting residual strain during transformation under isothermal and isobaric conditions.^26–28

Quaternary alloying and precipitation strengthening have also been used to improve the overall behaviour of NiTiHf polycrystalline and single crystal alloys. The shape memory properties of heat treated Ni_45·3Ti_29·7Hf₂₀Pd₅ (at-) alloys in single crystalline and polycrystalline forms have been reported.^29–33 The replacement of 5 Pd with Ni of Ni_50·3Ti_29·7Hf₂₀ alloy resulted in a very high strength alloy that has high damping capacity of 35 J cm⁻³ in polycrystalline form and 44 J cm⁻³ in [111] oriented single crystals.^32–33 Transformation strain of 2 was observed in aged [111] oriented Ni_45·3Ti_29·7Hf₂₀Pd₅ single crystals under a compressive biasing stress of 1500 MPa.³¹ Moreover, perfect superelastic behaviour with recoverable strain of 4·2 was observed in the solutionized condition even when compressive stress levels as high as 2·5 GPa were applied.³² However, it is also known that Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys are brittle, since they generally fail after limited plastic deformation in compression and during phase transformation in tension in superelasticity experiments.³⁴

It has been considered that low workability is one of the main problems with NiTiHf alloys for practical use. Kim et al.³⁵ reported that an addition of Nb to NiTiHf alloys caused the formation of a soft Nb-rich β phase and improved the cold workability, although the TTs and plastic strain in thermal cycling experiments under stress were decreased. Cu has been another alloying element to NiTiHf systems where, in general, it improved the glass forming ability and thermal stability of NiTiHf alloys while decreasing their TTs.^36,37 NiTiHfCu alloys have also demonstrated two-way shape memory effect.³⁸ It has recently reported that Ni_45·3Ti_29·7Hf₂₀Cu₅ alloys have the capability to recover compressive strains of 2 above 100°C and two-way shape memory strain of 0·8 above 80°C.³⁹

Hsieh and Wu⁴⁰ investigated the TTs and hardness values of Ti_50·5−xNi_49·5Zr_x/2Hf_x/2 (x = 0–20 at-)⁴⁰ and revealed that TTs can be increased from 50 to 323°C with increased Zr and Hf contents. Their shape memory responses under stress (e.g. constant stress thermal cycling, superelasticity) have not been reported yet.

This article reviews the effects of alloying, aging and processing on the shape memory properties and microstructure of NiTiHf-based alloys. Special attention is given to recently developed Ni-rich NiTiHf-based alloys.

Transformation temperatures of NiTiHf-based shape memory alloys

Many studies have been conducted in order to gain the fundamental understanding on how to change the TTs of SMAs.⁴¹ It is known that chemical composition alteration is very effective to change the properties such as TTs, transformation strain and matrix strength of SMAs. Figure 1a shows the effects of Ni content on the M_p (martensite peak temperature) of Ni_xTi_90−xHf₁₀.¹⁶ It is clear that M_p is insensitive up to 50 at- Ni and then suddenly decreases to below 0°C with increased Ni content.

M_p temperature as a function a Ni and b Hf contents in NiTiHf alloys^14–16,42 (chemical compositions are in at-)

Figure 1b shows the change in M_p as a function of Hf.^14–16,42 It is clear that M_p does not change up to 3 of Hf and then increases after 5. Up to 10 Hf, the increase of M_p is about 5°C/at- Hf. As the Hf increases beyond 10, there is an abrupt increase of M_p by almost 20°C/at- Hf in NiTiHf alloys and M_p reaches up to 400°C for 25 Hf.

Figure 2a shows the differential scanning calorimetry responses of the Ni_50·3Ti_29·7Hf₂₀ alloys after heat treatment at selected temperatures from 300 to 900°C for 3 h.^{27

27} Initially, TTs slightly decreased compared to the as extruded (extruded at 900°C) material when aged at 300 and 400°C. Then, TTs increased with heat treatment temperature up to 700°C and then TTs decreased. The maximum A_f (austenite finish temperature) was revealed to be 210°C in Ni_50·3Ti_29·7Hf₂₀ alloys aged for 3 h at 600°C. Figure 2b shows the change in TTs for Ni_45·3Ti_29·7Hf₂₀Pd₅ polycrystalline specimens aged for 3 h at temperatures between 400 and 900°C.³³ The trend in TTs with heat treatment temperature was similar to that of Fig. 2a. The maximum A_f was about 150°C in Ni_45·3Ti_29·7Hf₂₀Pd₅ after aging at 600°C for 3 h. The main reason for the TTs change with aging in the both alloys could be attributed to the change in the chemical composition of matrix³³ due to the formation of precipitates that will be discussed in details in the microstructure part.

Transformation temperatures of a Ni_50·3Ti_29·7Hf₂₀ and b Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys after heat treatment of 3 h at selected temperatures^27,33

Zarinejad et al.⁴¹ revealed a practical relationship between the chemical composition and TTs by considering the number (e_v/a) and concentration (c_v) of valence electrons in NiTi-based alloys. The number of d and s electrons is accepted as the number of valence electrons for an atom in transition metals while the number of valence electrons is considered to be p and s electrons for an atom in non-transition metals.⁴³ The number of valence electrons of alloys can be calculated with the following equation⁴⁴ (1) where f_A, f_B and f_C are atomic fractions of A, B and C elements, are the related valence electrons for the elements in an alloy system.

The following equation can be used to determine the average concentration of valence electrons⁴⁴ (2) where Z_A, Z_B and Z_C are the atomic numbers of elements A, B and C, respectively.

Figure 3 shows the relationships between the M_s (martensite start temperature) (or M_p) and e_v/a and c_v in NiTiHf-based SMAs.^{30,35,37,40,41,44,45} It is clear that the TTs do not have a clear trend with e_v/a while they generally decrease with increasing c_v. It is commonly agreed that higher electron concentration results in higher bulk (resistance to volume change) and shear (resistance to shape change) moduli.^44,46 Thus, the concentration of the electrons may affect the strength of atomic bonds in metallic materials. In general, as the concentration of valence electrons increases, the resistance to shear also increases. Thus, further energy provided by undercooling is necessary for the martensitic transformation resulting in decreased TTs.

M_s as a function of a e_v/a and b c_v in NiTiHf-based shape memory alloys^{30,35,37,40,41,44,45}

As stated above, even though there are some guidelines in predicting the TTs of NiTiHf-based alloys, the relationship between the nominal chemical composition and TTs is not completely established since there are many other factors that may alter TTs such as precipitation and grain size effects.^37,43–45 For instance, if the precipitates are fine and interparticle distances are small, nucleation of martensite could be more difficult and require additional undercooling, resulting in decreased TTs. Transformation temperatures are also sensitive to local chemical composition changes due to formation of precipitates. Moreover, it is known that internal stress fields increase the TTs in SMAs.^37,43–45

Crystal structure and microstructure of NiTiHf-based alloys

The shape memory and superelastic properties of NiTi-based alloys are significantly influenced by the microstructure, such as the precipitation size, interparticle distance and martensite morphology. In order to obtain good shape memory and superelastic properties, it is important to strengthen the matrix to prevent the introduction of dislocations during the martensitic transformation. One of the well-known procedures to improve the strength of the matrix is the precipitation hardening. Aging of NiTiHf alloys produces several precipitates that affect the martensite morphology that will be discussed in details.

Crystal structure of NiTiHf alloys

In general, the crystal structures of austenite and martensite phases in NiTiHf alloys are cubic (B2) and monoclinic (B19′), respectively, which are similar to those in NiTi binary alloys. Zarinejad et al.⁴⁷ investigated the effect of Hf on the lattice parameters of the B19′ martensite in NiTiHf alloys. The lattice parameters a, b, c and β angle of the martensite are plotted in Fig. 4 as a function of Hf content for Ni_(100−x)/2Ti_(100−x)/2Hf_x, Ni_50−xTi₅₀Hf_x and Ni₅₀Ti_50−xHf_x (x = 5–20 at-) alloys. The addition of Hf increased all the lattice parameters for the Ni_(100−x)/2Ti_(100−x)/2Hf_x and Ni_50−xTi₅₀Hf_x alloys. On the other hand, when Ni is constant, the increase in Hf in the Ni₅₀Ti_50−xHf_x alloy increased a, c and β but decreased b. Potapov et al.⁴⁵ also observed a similar dependence of lattice parameters on the Hf content for Ni_49·8Ti_50·2−xHf_x (x = 8–25 at-) alloys where the increase in Hf while Ni was kept constant to 49·8 slightly decreased the lattice parameter b, while it increased a, c and β of B19′ martensite. It was also reported that the addition of Hf increased the lattice parameter of B2 austenite.⁴⁵ The volume change during transformation was smaller than 0·5 which was similar to that in NiTi binary alloys (∼0·3 or less).^48,49 It should be noted that in some studies, NiTiHf alloys with more than 15 at- Hf in Ni_48·5(Ti_51·5−xHf_x)⁵⁰ and between 20 and 30 at- Hf of Ni₅₀(Ti_50−xHf_x)⁵¹ were reported to have orthorhombic B19 martensite.

Lattice parameters a a, b b, c c and d β of B19′ martensite as a function of Hf in NiTiHf alloys⁴⁷

Precipitation characteristics and their effects on martensite morphology

Ni-lean NiTiHf-based alloys

König et al.⁵² fabricated NiTiHf thin films with a wide composition range by magnetron sputtering method and investigated their TTs, precipitate structure and thermal cycling properties. Multilayer thin films (individual layers ∼15 nm thick) were sputtered from elemental targets and annealed at 550°C for 1 h in order to transform their multilayer structure into alloys. Figure 5 depicts the composition regions in which different precipitates are formed.⁵² The relative intensity of one characteristic X-ray diffraction peak belonging to the phase of interest was plotted colour-coded within a section of the NiTiHf ternary phase diagram. Four different precipitates, i.e. HfNi(Ti), Ti₂Ni(Hf), Hf₂Ni(Ti) and Laves phase, were confirmed in Ni-lean composition regions. They concluded that the observation of reversible phase transformation was limited by the formation of Ti₂Ni(Hf), HfNi(Ti) and/or Hf₂Ni(Ti) precipitates. These precipitates restricted the transforming region to compositions with Ni contents above ∼40 at- and Hf contents below ∼30 at-.

Composition regions in which different precipitate phases exist. The relative intensity of an X-ray diffraction peak for each phase is plotted colour-coded within a section of the ternary Ni–Ti–Hf diagram for a HfNi(Ti), b Ti₂Ni(Hf), c Hf₂Ni(Ti), and d Laves phase (colour code: red = high; green = medium; blue = low intensity)⁵² Figure 5 will be reproduced to be mono on the printed version

The Ti₂Ni(Hf) precipitates have also been observed by many other researchers in Ni-lean NiTiHf alloys.^{17,36,53–55} It has been reported that the volume fraction of the Ti₂Ni(Hf) precipitates decreased with increasing the Ni content, although the Ti₂Ni(Hf) precipitates were still observed in slightly Ni-rich compositions.^17,23 Fine Ti₂Ni(Hf) precipitates strengthen the matrix and improve shape memory and superelastic properties of NiTiHf-based alloys.^36,53 The effects of aging temperature and time on the formation of Ti₂Ni(Hf) precipitates were investigated by Meng et al. in Ni₄₉Ti₃₆Hf₁₅⁵³ and Ni₄₄Ti₃₆Hf₁₅Cu₅³⁶ alloys. The size of the precipitates increased with increasing aging temperature and time. Figure 6a and b shows the bright-field transmission electron microscopy (TEM) images of the Ni₄₄Ti₃₆Hf₁₅Cu₅ ribbons annealed at 500 and 700°C for 1 h, respectively.³⁶ According to the selected area diffraction (SAD) pattern (Fig. 6c) taken from the specimen annealed at 500°C, the precipitate was confirmed to be Ti₂Ni(Hf). The diameter of the precipitates was estimated to be 20–40 nm when the annealing temperature was 500°C. After annealing at 700°C, the size of the precipitates increased to ∼150 nm. The fine precipitates formed in the ribbon after annealing at 500°C strengthened the matrix and prohibited plastic deformation, which resulted in a perfect superelastic shape recovery after deformation to 3·5 strain. On the other hand, the ribbon annealed at 700°C showed an incomplete shape recovery due to the lower strength of the matrix with large precipitates.

a typical bright-field image of martensite in Ni₄₄Ti₃₆Hf₁₅Cu₅ ribbon annealed at 500°C for 1 h and SAD pattern taken from region W, electron beam//[1 0]_M,T; b typical martensite structure in the ribbon annealed at 700°C for 1 h and the SAD pattern taken from region D, electron beam//[2 1]_M1,M2//[ 1]_M3; c SAD pattern obtained from Ti₂Ni(Hf) type precipitates formed in ribbon annealed at 500°C for 1 h, electron beam//[110]_Ti2Ni(Hf)³⁶

It is important to note that the size of the Ti₂Ni(Hf) precipitates are very effective to control the martensite morphology. It was found that (001)_B19′ compound twins were dominant when the material contained homogeneously distributed Ti₂Ni(Hf) precipitates with 20–40 nm in diameter (Fig. 6a). Similar martensite morphology has been observed in a Ti-rich NiTi thin film with a homogeneous distribution of fine Ti₂Ni precipitates.⁵⁶ When the annealing temperature was 700°C, {011}_B19′ type I twins became dominant and the martensite variants showed mainly spear-like and mosaic-like morphologies as shown in Fig. 6b. Martensite domains with (001)_B19′ compound twins were also observed around the coarse Ti₂Ni(Hf) precipitates. The spear-like and mosaic-like morphologies have been reported as typical morphologies of the martensite in Hf-added NiTi alloys.^57,58

Ni-rich NiTiHf-based alloys

Meng et al.^23,59 have reported that Ni₄(Ti, Hf)₃ precipitates were formed in Ni-rich NiTiHf alloys similar to the Ni₄Ti₃ precipitation in NiTi binary alloys. However, recently, it has been reported that a new precipitate which has a more complicated structure than that of Ni₄(Ti,Hf)₃ forms in Ni-rich NiTiHf alloys^24,25,60 and improves their shape memory and superelastic properties due to precipitation strengthening.^26,27,33 Initially, Han et al.⁶¹ reported a precipitate with a face-centred orthorhombic lattice with a space group of F 2/d 2/d 2/d in an aged Ni_48·5Ti_36·5Hf₁₅. There are six different variants in this orthorhombic precipitate with habit planes of (100)_P//{001}_B2 and long axes of [001]_P//< 10>_B2. However, they did not provide an atomic structure model for the observed precipitate.

Recently, Yang et al.²⁵ proposed an atomic structure model which contains of 192 atoms in an orthorhombic unit cell for the observed precipitate in Ni-rich NiTiHf alloys. The orthorhombic precipitate phase was named as ‘H-phase’ and Fig. 7a shows the unit cell of this precipitate.²⁵ In order to refine the structure model, ab initio density functional theory calculations have also been performed to relax the structure model.^24,25 Selected area diffraction patterns obtained from a single large H-phase precipitate in a Ni₅₂Ti₂₈Hf₂₀ alloy are shown in Fig. 7b–d.²⁵ All the SAD patterns revealed the orientation dependence between the precipitate and austenite B2 phase (the diffraction spots are indexed according to the austenite phase). There were additional reflections at 1/3 positions along <110>_B2^* in reciprocal space as shown by arrows, which was a characteristic of the H-phase. The composition of the proposed H-phase was Ni₅₀Ti_16·7Hf_33·3, whereas it has been indicated by energy dispersive spectroscopy analysis that the Ni content of the H-phase precipitate was always slightly richer than that of the nominal composition of Ni-rich NiTiHf alloys in contrast to the proposed Ni content of 50 at- .^24,25,62 Therefore the formation of H-phase precipitates depleted Ni from the matrix and increased TTs as shown in Fig. 2. Yang et al.²⁵ observed anti-site defects within the precipitate which may slightly change the composition of the precipitate, and proposed that the H-phase did not have a unique composition. The effects of the alloy composition on the H-phase precipitation were investigated by Santamarta et al.²⁴ They concluded that the H-phase precipitates grew faster in alloys with higher Ni content since the precipitates were richer in Ni content compared to the nominal composition of the alloys. Similarly, for a fixed Ni content, the growth of the H-phase became faster when the Hf content was increased.

a unit cell of unrelaxed orthorhombic model of the H-phase.²⁵ SAD patterns of the b [001]_B2, c [1 1]_B2, d [1 0]_B2 and e [110]_B2 zone axes obtained from a single large particle in a Ni₅₂Ti₂₈Hf₂₀ alloy.²⁴ The small arrows and circles mark the additional reflections arising from the precipitate

The control of the size and interparticle distance of H-phase precipitates is important to obtain good shape memory and superelastic responses. It has been reported that the aging temperature and time significantly affected the size and interparticle distance of the precipitates formed in Ni-rich NiTiHf-based alloys.^23,24,33 Figure 8a–c illustrates the representative microstructure of Ni_50·3Ti_29·7Hf₂₀ alloys in as extruded and aged conditions.²⁷ The bright-field image of the as extruded Ni_50·3Ti_29·7Hf₂₀ alloy is shown in Fig. 8a. Precipitate formation was not confirmed in the as extruded condition. Figure 8b and c shows TEM micrographs of the extruded Ni_50·3Ti_29·7Hf₂₀ alloy aged at 550 and 650°C for 3 h, respectively. Fine and coherent H-phase precipitates were formed in the 550°C aged specimen. When the aging temperature increased from 550 to 650°C, the precipitate size increased from about 20 to 40–60 nm. The interparticle distance also increased after aging at 650°C for 3 h compared with the 550°C for 3 h case. Figure 8d and e shows the H-phase precipitates and B19′ martensite in slightly (Ni+Pd)-rich Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys in aged conditions.³³ The size of the spindle-shaped H-phase was increased when the aging temperature was increased from 550 to 650°C. The fine and coherent H-phase precipitates in the Ni_50·3Ti_29·7Hf₂₀ and Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys aged at 550°C improved the shape memory and superelastic properties due to precipitation strengthening. However, the alloys aged at 650°C exhibited relatively poor shape memory and superelastic properties due to large precipitate sizes.

Bright-field images of the Ni_50·3Ti_29·7Hf₂₀ alloy a extruded at 900°C, b aged at 550°C for 3 h and c aged at 650°C for 3 h.²⁷ Bright-field images of the Ni_45·3Ti_29·7Hf₂₀Pd₅ alloy aged at d 550°C and e 650°C for 3 h.³³ Inset in d is the enlargement of area D. The SAD patterns shown in d and e were taken from the area D and E, respectively

The martensite morphology in Ni-rich NiTiHf-based alloys is affected by the size and interparticle distance of H-phase precipitates. The martensite variants in the as extruded Ni_50·3Ti_29·7Hf₂₀ alloy show spear-like morphology and high density of twins can be seen inside the martensite plates (Fig. 8a). Han et al.^57,58 have reported two types of martensite morphologies; spear-like and mosaic-like in NiTiHf alloys and they also revealed that each martensite lath is consisted of (001)_B19′ compound twins. If the precipitates were small and interparticle distance was short, the growing martensite plates can absorb all the precipitates during growth as it can be seen in the 550°C aged Ni_50·3Ti_29·7Hf₂₀ alloys (Fig. 8b). The large martensite plates were related by the {011}_B19′ type I twinning mode, which was confirmed by the SAD pattern shown in Fig. 8b taken at the interface of the plates. It should be noted that no internal twins were observed in the large martensite plates in the 550°C aged specimen. On the other hand, when the precipitates were big and interparticle distance was large, martensite plates can be formed between the precipitates and the thickness of the plates was controlled by the interparticle distance of the precipitates (Figs. 8c–e). In Ni_45·3Ti_29·7Hf₂₀Pd₅, the SAD patterns were taken from the area D for the 550°C aged specimen (Fig. 8d) and from the area E for the 650°C aged specimen (Fig. 8e). It was revealed that the main twinning mode observed in the martensite was (001)_B19′ compound twin in both aging conditions. It was suggested that the internal twinning type was not affected by the size of the H-phase precipitates if the martensite plates are formed between the precipitates.

Addition of Nb and Pd to NiTiHf alloys

In NiTiHf-based alloys, the lattice invariant shear (LIS) of the martensitic transformation depends on the alloy composition. The (001)_B19′ compound twins have been frequently observed in martensite plates and considered as the LIS in NiTiHf alloys.^57,58 However, recently, it was found that the <011>_B19′ type II twin was the LIS in a (Ni+Pd)-rich Ni_45·3Ti_39·7Hf₁₀Pd₅ alloy which was homogenised at 900°C followed by furnace cooling.³⁰ The Ni_45·3Ti_39·7Hf₁₀Pd₅ alloy exhibited less hardening during transformation compared to a Ni_45·3Ti_29·7Hf₂₀Pd₅ alloy which has (001)_B19′ compound twins.

Figure 9a shows a bright-field TEM image for the Ni_45·3Ti_39·7Hf₁₀Pd₅ alloy³⁰ which consisted of two phases, B2 austenite and B19′ martensite at room temperature. In the SAD pattern taken from the austenite phase (Fig. 9b), there were diffuse streaks along the <110>_B2^* directions in reciprocal space. The diffuse streaks could be attributed to the formation of very small precipitates during the slow furnace cooling process from the homogenisation temperature. Sandu et al.⁶³ also observed similar diffuse streaks in an aged Ni-rich NiTiZr alloy. The SAD pattern taken from the martensite phase (Fig. 9c) indicated that the internal twins formed in the martensite variants were the <011>_B19′ type II twins. Compared to the (001)_B19′ compound twin, lower density of twins is found when the LIS is the <011>_B19′ type II twin. It is noted that the LIS in NiTi binary alloys is known as the <011>_B19′ type II twin and the (001)_B19′ compound twin has been observed in NiTi alloys as a deformation twin.⁶⁴ The (001)_B19′ compound twin has been also found in nanocrystalline NiTi alloys⁶⁵ and in aged Ni-rich NiTi alloys with fine Ni₄Ti₃ precipitates.⁶⁶ These results suggested that the LIS in NiTiHf-based alloys depends on the alloy composition and the size and interparticle distance of precipitates.

a bright-field image of Ni_45·3Ti_39·7Hf₁₀Pd₅ alloy homogenised at 900°C followed by furnace cooling, b SAD pattern taken from B2 austenite phase and c SAD pattern taken from martensite phase indicating B19′ monoclinic structure³⁰

Kim et al.³⁵ reported that addition of Nb to NiTiHf alloys causes the formation of a soft Nb-rich β phase and improves the cold workability. The stability of shape memory properties is improved by the precipitation of the β phase, although the shape recovery strain decreases by the addition of Nb. Figure 10 shows the back-scattered scanning electron images of (Ni_49·5Ti_35·5Hf₁₅)Nb alloys.³⁵ In Fig. 10a, the Ti₂Ni type precipitate can be seen in the Ni_49·5Ti_35·5Hf₁₅ ternary alloy with a slightly dark contrast. The β phase, which appears white on the images, was observed even after 1 Nb addition (Fig. 10b), indicating that the solubility limit of Nb in the matrix was less than 1. The amount of the β phase increased with increasing Nb content. When 15 Nb was added, it exhibited a fully lamellar microstructure as shown in Fig. 10c, which is a characteristic of eutectic solidification. This fine lamellar structure strengthened the matrix and prohibited plastic deformation during transformation.

Back-scattered scanning electron images of a Ni_49·5Ti_35·5Hf₁₅, b (Ni_49·5Ti_35·5Hf₁₅)Nb₁ and c (Ni_49·5Ti_35·5Hf₁₅)Nb₁₅ alloys³⁵

Morphologies of reoriented martensite and stress induced martensite

Acar et al.⁶⁷ have reported the morphology of the reoriented martensite in a Ni_45·3Ti_34·7Hf₁₅Pd₅ alloy. Transmission electron microscopy observation was carried on a homogenised sample after 8 compressive deformation at 15°C (below martensite finish temperature, M_f). Figure 11a and b shows the TEM micrographs obtained from the as homogenised and deformed samples, respectively. There are fine twins in the martensite plates in the as homogenised sample (Fig. 11a). In the deformed sample (Fig. 11b), thicker martensite plates were formed by the reorientation of martensite variants as compared to the as homogenised sample. The thick martensite plates are considered to be favourable martensite variants under stress. The inset in Fig. 11b is an SAD pattern taken from the interface between the martensite plates A and B. It was revealed that the fine twins in the martensite plates are (001)_B19′ compound twins and the boundary between the plates A and B is close to the {111}_B19′ type I twin plane (so called {111}_B19′-type boundary).³⁶ It is considered that the {111}_B19′-type boundary can move under stress without significant detwinning of fine (001)_B19′ compound twins in martensite plates.

Bright-field image (TEM) of a as homogenised Ni_45·3Ti_34·7Hf₁₅Pd₅ and b 8 deformed alloy with corresponding SAD pattern.⁶⁷ Bright-field image of Ni₄₉Ti₃₆Hf₁₅ c deformed to 8 at 250°C and d deformed to 16 at 250°C.²¹ The SAD pattern shown in c was taken from area II

Dalle et al.⁶⁸ investigated the morphology of reoriented martensite of annealed (800°C for 1 h) Ni_49·8Ti_42·2Hf₈ after 10 tensile deformation. They observed finer (001)_B19′ compound twins in the deformed material compared to the as annealed material with self-accommodated martensite. They suggested that the detwinning of the (001)_B19′ compound twins is difficult and proposed that, instead of the detwinning, a supplementary (001)_B19′ mechanical twinning could take place during deformation by a mechanism of the repetition of the dislocation slip on the (001)_B19′ plane.

Meng et al.^20,21 investigated the morphologies of the SIM in Ni₄₉Ti₃₆Hf₁₅ which were solution treated at 1000°C for 1 h and deformed in tension at 250°C. Figure 11c shows the typical morphology of the preferentially oriented SIM variants and the SAD pattern taken from the area II for the 8 deformed Ni₄₉Ti₃₆Hf₁₅.²¹ (001)_B19′ compound twins were mainly observed in the SIM plates. The SAD pattern revealed that the SIM plates were twin-related with {011}_B19′ type I mode, which was similar to the thermally transformed martensite.^57,58 The preferentially oriented SIM variants were disappeared and several martensite variants were intersected into each other after deformation. Figure 11d shows the variant-crashed/variant-intersected morphology after deformation of 16. The interfaces of the martensite variants are blurred in the variant-crashed/variant-intersected morphology. They noted that the stress induced martensitic transformation and dislocation slip occurred simultaneously during loading and suggested that the introduction of dislocations increases the martensite variants with the variant-crashed/variant-intersected morphology.

Mechanical behaviour of NiTiHf-based shape memory alloys

The relatively high degree of brittleness or poor cyclic stability in NiTiHf alloys are the main obstacles for their commercial high temperature applications. It has been observed that ductility of NiTiHf alloys could be improved by deformation at higher temperatures in Ni-lean NiTiHf alloys. Ni₄₉Ti₃₆Hf₁₅ alloys failed after 7 of bending deformation at room temperature while they did not fracture until 30 tensile strain at 260°C.^42,69

Material properties of NiTiHf-based alloys can be controlled by aging at different temperatures and time as illustrated in Fig. 12. Meng et al.⁵³ illustrated that yield strength of Ni₄₉Ti₃₆Hf₁₅ can be adjusted by aging at 700°C while ductility was constant as shown in Fig. 12a. The strength of matrix was improved after 20 h but further increase in aging time decreased the strength of alloy which can be related to the size, interparticle distance and volume fraction of Ti₂Ni(Hf) precipitates.

a effect of aging time on yield strength and elongation of Ni₄₉Ti₃₆Hf₁₅⁵³ and b hardness values of Ni_50·3Ti_29·7Hf₂₀ and Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys as a function of aging temperature

Figure 12b shows the hardness (HV) of Ni_50·3Ti_29·7Hf₂₀ and Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys as a function of aging temperature for 3 h aging. The increase in the hardness in the both alloys can be attributed to formation of nanosize coherent precipitates that minimises the dislocation motion. The decrease in hardness at high aging temperatures in the both alloy systems can be linked to formation of semi-/non-coherent precipitates and larger interparticle distance due to over-aging and thus the lack of precipitation strengthening which was also demonstrated in Fig. 8. It is also clear that Ni_45·3Ti_29·7Hf₂₀Pd₅ is harder in nature when is compared to Ni_50·3Ti_29·7Hf₂₀.

Initially, SMA properties of Ni-lean NiTiHf alloys were mainly investigated due to low TTs of Ni-rich NiTiHf alloys.^23,59 Shape memory effect with 3 recoverable strain or 80 recovery of 6 applied strain is observed in compression and bending while 80 recovery of 2·5 applied tensile strain is observed in Ni₄₉Ti₃₆Hf₁₅.^20,69 The poor shape memory effect is attributed to high stress (∼500 MPa) for martensite reorientation and high slope in SIM transformation region (no plateau region observed) confirmed by tensile experiments.^69,70 Although no superelasticity is observed in Ni-lean NiTiHf alloys,^21,23 0·88 strain for two-way shape memory effect has been observed.⁷¹ Unstable cyclic behaviour is a major problem in Ni-lean NiTiHf alloys where it has been observed that TTs were decreased by 40°C during stress free thermal cycling of Ni₄₉Ti₄₁Hf₁₀ after 20 cycles.¹⁴

In Ni-rich NiTiHf alloys, almost perfect dimensional stability with 3 strain under a compressive stress of 500 MPa was observed as illustrated in Fig. 13a.²⁶ It can be seen from Fig. 13b that the stress–strain curve of solutionised Ni-lean Ni₄₉Ti₃₆Hf₁₅ at temperature above A_f showed high slope in SIM transformation region and deformation was not fully recovered upon unloading which was similar to that in cold worked TiNi,⁷² TiPd⁷³ and as extruded or overaged Ni-rich NiTiHf²⁷ alloys. Aging can improve the shape memory and material properties of Ni-rich NiTiHf alloys. Figure 13c shows the superelasticity responses of as extruded and aged Ni_50·3Ti_29·7Hf₂₀ alloys.²⁷ Perfect superelastic behaviour with 4 recoverable strain was revealed at 240°C after aging at 550°C for 3 h in Ni_50·3Ti_29·7Hf₂₀. The improvement in superelastic behaviour with aging can be attributed to the presence of coherent and fine H-phase precipitates (as discussed in the microstructure part and shown in Fig. 8), which strengthen the matrix. Poor superelastic response after aging at 650°C for 3 h can be attributed to loss of the coherency of the coarsened precipitates. It is worth to note that beside the fully recoverable strain, Ni-rich NiTiHf exhibited high yield strength at high temperature and the Clausius–Clapeyron (C–C) slopes were between 7 and 13 MPa °C⁻¹. It should also be noted that almost fully recoverable strain with small amount of plastic deformation under 1000 MPa with low plastic and perfect superelastic behaviour was obtained in Ni-rich Ni_50·3Ti_29·7Hf₂₀ along the [111] orientation.²⁸

a thermal cycling under stress response of Ni-rich Ni_50·3Ti_29·7Hf₂₀,²⁶ b superelastic behaviour of hot rolled Ni-lean Ni₄₉Ti₃₆Hf₁₅²⁰ and c superelastic behaviour of Ni-rich Ni_50·3Ti_29·7Hf₂₀²⁷

Another method to improve the shape memory properties of NiTiHf has been the quaternary alloying. Recently, several studies^29,31–33 have revealed the effects of Pd addition on the mechanical properties of Ni_50·3Ti_29·7Hf₂₀. It was shown that perfect superelastic curves (with negligible plastic strain) at stress levels as high as 2 GPa were possible for aged polycrystalline Ni_45·3Ti_29·7Hf₂₀Pd₅³³ for a temperature window of 50–130°C as illustrated in Fig. 14a. However, full strain recovery was not observed in over aged materials due to the formation of large precipitates as shown in Fig. 8 and high degree of hardening was observed during transformation. When the Hf content was decreased to 10 in Ni_45·3Ti_39·7Hf₁₀Pd₅, high slope in SIM transformation region was not observed that can be attributed to the formation of different type and density of twinning as discussed in the microstructure section (see Fig. 9). It was mentioned before that the <011>_B19′ type II twin was the LIS in a (Ni+Pd)-rich Ni_45·3Ti_39·7Hf₁₀Pd₅ alloy³⁰ in contrast to the (001)_B19′ compound twin in Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys.³³ Type II twins could be held responsible for the lack of hardening in the transformation region of Ni_45·3Ti_39·7Hf₁₀Pd₅ alloys containing 10 Hf compared to Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys that has (001)_B19′ compound twins. The growth of martensite variants with thin (001)_B19′ compound twins is more difficult in contrast to <011>_B19′ type II twins. Thus, the required energy to complete the SIM transformation increased and resulted in high slope in SIM transformation region.

Stress–strain responses of a Ni_45·3Ti_29·7Hf₂₀Pd₅³³ and b Ni_45·3Ti_39·7Hf₁₀Pd₅³⁰ at temperatures above A_f

Another alloying addition to NiTiHf alloys has been Cu where it generally improved the two-way shape memory effect and thermal stability of NiTiHf alloys while decreased their TTs.^36–38 It has been reported that Ni_45·3Ti_29·7Hf₂₀Cu₅ can recover compressive strains of ∼2·2 under 700 MPa at temperature above 100°C and can produce 0·8 two-way shape memory strain at temperature above 80°C.³⁹ Perfect superelasticity response was not observed in Ni_45·3Ti_29·7Hf₂₀Cu₅ due to its high C–C slope (about 14–25 MPa °C⁻¹) and work hardening coefficient in addition to low yield stress for plastic deformation.³⁹

It can be seen from Fig. 15 that the Ni-rich Ni_50·3Ti_29·7Hf₂₀ alloys had higher TTs and strength than the Ni-lean Ni_49·5Ti_35·5Hf₁₅ alloys. This can be attributed to the difference in precipitates types where H-phase particles were observed in Ni-rich and Ti₂Ni(Hf) precipitates were formed in Ni-lean NiTiHf after aging. NiTiHfNb has higher transformation strain but lower strength than Ni-rich NiTiHf. Also, as the Nb content was increased the TTs were decreased and the shape memory properties of (Ni_49·5Ti_30·5Hf₁₅)Nb₅ alloys became more stable as the plastic strain was decreased due to the fine lamellar structure that strengthen the matrix as shown in Fig. 10. Moreover, the shape recovery ratio was increased as Nb content increased. Addition of Pd to Ni_50·3Ti_29·7Hf₂₀ decreased the TTs and transformation strain while it improved the strength of the alloy.

Thermal cycling under constant compressive stress of 500 MPa results for NiTiHf-based SMAs^27,33,35

Cycling instability is a major concern of the NiTiHf alloys for high temperature applications. Figure 16 shows the thermal cycling under 200 MPa experiments of Ni_49·8Ti_42·2Hf₈ in homogenised and equal channel angular extruded at 650°C conditions.¹⁸ It is clear that severe plastic deformation improved the thermal cyclic stability and decreased the thermal hysteresis of Ni-lean Ni_49·8Ti_42·2Hf₈ alloys. Moreover recoverable strain increases and irrecoverable strain decreases. Formation of precipitates influences the cyclic degradation resistance in NiTiHf²⁷ where small coherent precipitates generally improve the thermal cyclic stability while larger precipitates do not affect the stability.

Strain–temperature response of Ni_49·8Ti_42·2Hf₈ under 200 MPa a homogenised and b equal channel angular extruded at 650°C using route 2C under 200 MPa¹⁸

Work output, damping capacity and potential applications of NiTiHf-based alloys

Shape memory alloy based actuators can be employed as light weight and energy efficient alternatives of hydraulic or pneumatic systems² in automotive, aerospace and down-hole energy exploration industries. Alloys with higher work output values can be used to decrease the required weight or size of actuators.

The maximum work output levels of various NiTiHf-based SMAs as a function of their average operating temperature range are shown in Fig. 17a. Work output can be calculated as the mathematical multiplication of reversible transformation strain and applied stress in constant-stress thermal cycling experiments. NiTi alloys have work output densities of about 12–18 J cm⁻³,⁷⁴ while NiTiPd and NiTiPt alloys have work output capabilities of 6–9 and 13 J cm⁻³ respectively⁷⁵ at temperatures above 150°C. The work output of Ni-rich NiTiHf polycrystalline alloys was found to be 18–20 J cm⁻³.²⁷ Ni_45·3Ti_29·7Hf₂₀Cu₅ alloys can generate work outputs of around 14–15 J cm⁻³ while NiTiHfNb alloys have work output levels of 17–18 J cm⁻³ above 100°C and 150°C, respectively.^35,39 On the other hand, Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys can generate higher work outputs of 32–35 J cm⁻³ (up to 120°C) compared to other NiTiHf-based SMAs, while upper temperature capability is somewhat limited compared to the above mentioned NiTiHf-based alloys.³³

Comparisons of a work outputs and b damping capacities for typical NiTi-based SMAs

Figure 17b shows the damping capacities/absorbed energies of NiTi-based alloys as a function of transformation stress. Damping capacities can be calculated from the area between the loading and unloading curves in a superelastic cycle and can be explained as the ability to repeatedly dissipate unwanted energy from a system. SMA based mechanisms could be employed under high number of cycles in real practical applications. Thus, stability of a superelastic curve is essential for damping applications. A HTSMA that could absorb large energy will be very appealing for high temperature damping devices. In addition to the high work output, NiTiHf-based SMAs have high damping capacities. They could be employed in aircraft engines as a damper for acoustic energy and construction for countering seismic movements in impact damping devices.

The damping capacity of Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys is 30–34 J cm⁻³ stemming from its outstanding mechanical hysteresis (around 900 MPa) and good superelastic strain of 4.³³ In related systems, the damping capacity is 16, 18–20, 38, and 54 J cm⁻³ for NiTi, NiTiHf, NiTiNb and NbTi/NiTi nanocomposites, respectively.^76–78 The Ni_45·3Ti_29·7Hf₂₀Pd₅ alloy has similar damping capability to NiTiNb alloys that are often used in coupling applications. However, it should be noted that Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys have the ability to operate at much higher stresses (∼2 GPa) than the NiTiNb systems. Damping capacities of NiTiHfCu and NiTiHfNb were not compared since full recoverable superelastic cycles have not been reported in literature.

Conclusions

From the present review of NiTiHf-based alloys, it is clear that NiTiHf-based alloys are attractive candidates for high temperature, high strength and damping applications. Their TTs and strength can be adjusted by heat treatments. They could show perfect superelasticity above 100°C and shape memory effect under high stress levels. However, some of their drawbacks such as low ductility, high slope in SIM transformation region and low cyclic stability are still remained to be improved.

It has been shown that microstructural control by composition alteration and aging is essential in tailoring shape memory and mechanical properties (e.g. TTs, strain, hysteresis and strength) in NiTiHf-based alloys. Based on the composition and precipitation characteristics (e.g. precipitate size and interparticle distance), the main microstructural features such as twin type, martensite morphology can be adjusted that would affect the shape memory and mechanical properties. In Ni-lean NiTiHf-based alloys, the size of the Ti₂Ni(Hf) precipitates are effective to control the martensite morphology. (001)_B19′ compound twins are dominant when Ti₂Ni(Hf) precipitates are small (about 20–40 nm) and homogeneously distributed while {011}_B19′ type I twins become dominant with increasing the size of Ti₂Ni(Hf) precipitates. The martensite morphology in Ni-rich NiTiHf-based alloys is affected by the size and interparticle distance of H-phase precipitates. When the precipitates are small and interparticle distance is short, martensite plates can absorb the precipitates during their growth and they are mainly twin-related with the {011}_B19′ type I mode. On the other hand, when the precipitates are big and interparticle distance is large, martensite plates can be formed between the precipitates. The thickness of the plates is governed by the interparticle distance of the precipitates. The formation of fine H-phase precipitates significantly improves the shape memory and superelastic properties due to precipitation strengthening.

Pd addition decreases the TTs of NiTiHf alloys while the matrix strength was increased by solid solution strengthening. Ni_45·3Ti_29·7Hf₂₀Pd₅ alloys can show perfect superelastic response under extremely high compressive stress levels of 2 GPa with negligible plastic deformation. NiTiHf(Pd) alloys have high work outputs and damping capacities reaching up to 30–35 J cm⁻³ owing to their good strain, high strength and large mechanical hysteresis. Nb addition to NiTiHf alloys improves the cold workability and the stability of shape memory properties while decreases the shape recovery strain by the precipitation of the β phase.

Detailed studies are needed to gain the fundamental understanding on processing–composition–microstructure–property relationships and reveal the true potential of NiTiHf-based alloys. Currently, they are the most promising alloys for high temperature and strength applications.

Footnotes

Acknowledgement

This work was supported by the NASA EPSCOR program under grant no. NNX11AQ31A.

References

Leo

: ‘Engineering analysis of smart material systems’; 2007, Hoboken, NJ, John Wiley & Sons.

and Padula

: ‘NiTi-based high-temperature shape-memory alloys: properties, prospects, and potential applications’, in ‘Advanced structural materials: properties, design optimization, and applications’, (ed. Soboyejo

W O

and Srivatsan

T S

); 2007, New York, Taylor & Francis Group.

LeBlanc

: ‘Part I – “Smart metals” providing actuation, sealing, and completion functions downhole’, Offshore, 2001, 61 (12), 58.

LeBlanc

: ‘Part II – “Smart metals” providing actuation, sealing, and completion functions downhole’, Offshore, 2002, 62 (1), 54.

Song

, Ma

and Li

H.-N

: ‘Applications of shape memory alloys in civil structures’, Eng. Struct. , 2006, 28, (9), 1266–1274.

Van Humbeeck

: ‘Non-medical applications of shape memory alloys’, Mater. Sci. Eng. A , 1999, A273, 134–148.

Petrini

and Migliavacca

: ‘Biomedical applications of shape memory alloys’, J. Metall., 2011, 2011, 501483.

El Feninat

, Laroche

, Fiset

and Mantovani

: ‘Shape memory materials for biomedical applications’, Adv. Eng. Mater. , 2002, 4, (3), 91–104.

Firstov

, van Humbeeck

and Koval

: ‘High temperature shape memory alloys problems and prospects’, J. Intell. Mater. Syst. Struct. , 2006, 17, (12), 1041–1047.

10.

Karaca

, Kaya

, Tobe

, Basaran

, Nagasako

, Kainuma

, Chumlyakov

: ‘Shape memory behavior of high strength Ni₅₄Ti₄₆ alloys’, Mater. Sci. Eng. A , 2013, A580, 66–70.

11.

Otsuka

and Ren

: ‘Physical metallurgy of Ti-Ni-based shape memory alloys’, Prog. Mater. Sci. , 2005, 50, (5), 511–678.

12.

, Karaman

and Noebe

: ‘High temperature shape memory alloys’, Int. Mater. Rev. , 2010, 55, (5), 257–315.

13.

Firstov

, Van Humbeeck

and Koval

: ‘Comparison of high temperature shape memory behaviour for ZrCu-based, Ti-Ni-Zr and Ti-Ni-Hf alloys’, Scr. Mater. , 2004, 50, 243–248.

14.

Besseghini

, Villa

and Tuissi

: ‘Ni-Ti-Hf shape memory alloy: effect of aging and thermal cycling’, Mater. Sci. Eng. A , 1999, A273–A275, 390–394.

15.

Angst

D. R

, Thoma

P. E

and Kao

M. Y

: ‘The effect of Hafnium content on the transformation temperatures of Ni₄₉Ti_51-xHf_x shape memory alloy’, J. Phys. IV, Colloq. , 1995, C8, 747–752.

16.

Abu Judom

, Thoma

, Kao

and Angst DR : ‘High transformation temperature shape memory alloy’, US patent 5,114,504, 1992.

17.

Thoma

and Boehm

: ‘Effect of composition on the amount of second phase and transformation temperatures of Ni_xTi_90-xHf₁₀ shape memory alloys’, Mater. Sci. Eng. A , 1999, A273, 385–389.

18.

Kockar

, Karaman

, Kim

and Chumlyakov

: ‘A method to enhance cyclic reversibility of NiTiHf high temperature shape memory alloys’, Scr. Mater. , 2006, 54, (12), 2203–2208.

19.

Wojcik

C. C

: ‘Properties and heat treatment of high transition temperature Ni-Ti-Hf alloys’, J. Mater. Eng. Perform. , 2009, 18, (5), 511–516.

20.

Meng

, Cai

, Zheng

, Tong

, Zhao

and Zhou

: ‘Stress-induced martensitic transformation behavior of a Ti-Ni-Hf high temperature shape memory alloy’, Mater. Lett. , 2002, 55, (1–2), 111–115.

21.

Meng

, Cai

, Wang

, Zheng

, Zhao

and Zhou

: ‘Microstructure of stress-induced martensite in a Ti-Ni-Hf high temperature shape memory alloy’, Scr. Mater. , 2001, 45, (10), 1177–1182.

22.

Meng

, Cai

, Fu

, Li

, Zhang

and Zhao

: ‘Shape-memory behaviors in an aged Ni-rich TiNiHf high temperature shape-memory alloy’, Intermetallics , 2008, 16, (5), 698–705.

23.

Meng

, Cai

, Chen

and Zhao

: ‘Effect of aging on martensitic transformation and microstructure in Ni-rich TiNiHf shape memory alloy’, Scr. Mater. , 2006, 54, 1599–1604.

24.

Santamarta

, Arróyave

, Pons

, Evirgen

, Karaman

, Karaca

and Noebe

: ‘TEM study of structural and microstructural characteristics of a precipitate phase in Ni-rich Ni–Ti–Hf and Ni–Ti–Zr shape memory alloys’, Acta Mater. , 2013, 61, (16), 6191–6206.

25.

Yang

, Coughlin DR , Phillips

, Yang

, Devaraj

, Kovarik

, Noebe

and Mills

: ‘Structure analysis of a precipitate phase in an Ni-rich high-temperature NiTiHf shape memory alloy’, Acta Mater. , 2013, 61, (9), 3335–3346.

26.

Bigelow

, Garg

, , Gaydosh

and Noebe

: ‘Load-biased shape-memory and superelastic properties of a precipitation strengthened high-temperature Ni_50·3Ti_29·7Hf₂₀ alloy’, Scr. Mater. , 2011, 64, (8), 725–728.

27.

Karaca

, , Ded

, Tobe

, Basaran

, Maier

, Noebe

and Chumlyakov

: ‘Effects of nanoprecipitation on the shape memory and material properties of an Ni-rich NiTiHf high temperature shape memory alloy’, Acta Mater. , 2013, 61, (19), 7422–7431.

28.

Karaca

, , Basaran

, Bigelow

, Noebe

and Chumlyakov

: ‘Compressive response of nickel-rich NiTiHf high-temperature shape memory single crystals along the [1 1 1] orientation’, Scr. Mater. , 2011, 65, (7), 577–580.

29.

Acar

, Karaca

, Basaran

, Yang

, Mills

, Noebe

and Chumlyakov

: ‘Role of aging time on the microstructure and shape memory properties of NiTiHfPd single crystals’, Mater. Sci. Eng. A , 2013, A573, 161–165.

30.

Acar

, Karaca

, Tobe

, Noebe

and Chumlyakov

: ‘Characterization of the shape memory properties of a Ni_45·3Ti_39·7Hf₁₀Pd₅ alloy’, J. Alloys Compd , 2013, 578, 297–302.

31.

Karaca

, Acar

, Basaran

, Noebe

, Bigelow

, Garg

, Yang

, Mills

and Chumlyakov

: ‘Effects of aging on [111] oriented NiTiHfPd single crystals under compression’, Scr. Mater. , 2012, 67, (7–8), 728–731.

32.

Karaca

, Acar

, Basaran

, Noebe

and Chumlyakov

: ‘Superelastic response and damping capacity of ultrahigh-strength [111]-oriented NiTiHfPd single crystals’, Scr. Mater. , 2012, 67, (5), 447–450.

33.

Karaca

, Acar

, Ded

, Basaran

, Tobe

, Noebe

, Bigelow

and Chumlyakov

: ‘Shape memory behavior of high strength NiTiHfPd polycrystalline alloys’, Acta Mater. , 2013, 61, (13), 5036–5049.

34.

Ded

: ‘Characterization of Ni-rich NiTiHf-based shape memory alloys’, ‘Mechanical engineering 2010’, University of Kentucky, Lexington, KY, USA, 2010.

35.

Kim

, Jinguu

, Nam

T.-H

and Miyazaki

: ‘Cold workability and shape memory properties of novel Ti–Ni–Hf–Nb high-temperature shape memory alloys’, Scr. Mater. , 2011, 65, (9), 846–849.

36.

Meng

, Cai

, Fu

, Zhang

and Zhao

: ‘Martensite structure in Ti–Ni–Hf–Cu quaternary alloy ribbons containing (Ti,Hf)₂Ni precipitates’, Acta Mater. , 2010, 58, (10), 3751–3763.

37.

Meng

, Cai

, Lau

, Zhao

, Zhou

: ‘Phase transformation and microstructure of quaternary TiNiHfCu high temperature shape memory alloys’, Intermetallics , 2005, 13, (2), 197–201.

38.

Liang

, Chen

, Shen

, Zhang

, Li

and Wang

: ‘Thermal cycling stability and two-way shape memory effect of Ni-Cu-Ti-Hf alloys’, Solid State Commun. , 2001, 119, (6), 381–385.

39.

Karaca

, Acar

, Ded

, Saghaian

, Basaran

, Tobe

, Kok

, Maier

, Noebe

and Chumlyakov

: ‘Compression response of Ni_45·3Ti_29·7Hf₂₀Cu₅ high temperature shape memory alloys’, submitted to Mater. Sci. Eng. A, 2014.

40.

Hsieh

and Wu

: ‘Martensitic transformation of quaternary Ti_50·5-XNi_49·5Zr_X/2Hf_X/2 (X = 0–20 at.) shape memory alloys’, Mater. Charact. , 2000, 45, (2), 143–152.

41.

Zarinejad

, Liu

and Tong

: ‘Transformation temperature changes due to second phase precipitation in NiTi-based shape memory alloys’, Intermetallics , 2009, 17, (11), 914–919.

42.

Wang

, Zheng

, Cai

and Zhao

: ‘The tensile behavior of Ti₃₆Ni₄₉Hf₁₅ high temperature shape memory alloy’, Scr. Mater. , 1999, 40, (12), 1327–1331.

43.

Zarinejad

and Liu

: ‘Dependence of transformation temperatures of shape memory alloys on the number and concentration of valence electrons’, in ‘Shape memory alloys: manufacture, properties and applications’, (ed. Chen

H R

), 339–360; 2010, Hauppauge, NY Science Publishers.

44.

Zarinejad

and Liu

: ‘Dependence of transformation temperatures of NiTi-based shape-memory alloys on the number and concentration of valence electrons’, Adv. Funct. Mater. , 2008, 18, (18), 2789–2794.

45.

Potapov

P. L

, Shelyakov

, Gulyaev

, Svistunov

, Matveeva

and Hodgson

: ‘Effect of Hf on the structure of Ni-Ti martensitic alloys’, Mater. Lett. , 1997, 32, (4), 247–250.

46.

Gilman

: ‘Electronic basis of the strength of materials’; 2003, Cambridge, Cambridge University Press.

47.

Zarinejad

, Liu

and White

: ‘The crystal chemistry of martensite in NiTiHf shape memory alloys’, Intermetallics , 2008, 16, (7), 876–883.

48.

Otsuka

, Sawamura

, Shimizu

and Wayman

: ‘Characteristics of the martensitic transformation in TiNi and the memory effect’, Metall. Trans. , 1971, 2, (9), 2583–2588.

49.

Dilibal

, Sehitoglu

, Hamilton

, Maier

and Chumlyakov

: ‘On the volume change in Co–Ni–Al during pseudoelasticity’, Mater. Sci. Eng. A , 2011, A528, (6), 2875–2881.

50.

Proc. SPIE , 1995, 2442, 171–178.

51.

K-H Wu, Z. Pu : ‘Martensite transformation of (Hf_xTi_50−x)Ni50 shape memory alloys’, in ‘ICOMAT 95: International Conference on “Martensitic transformation”’, Lausanne, Switzerland, August 1995, 759.

52.

König

, Zarnetta

, Savan

, Brunken

and Ludwig

: ‘Phase transformation, structural and functional fatigue properties of Ti–Ni–Hf shape memory thin films’, Acta Mater. , 2011, 59, (8), 3267–3275.

53.

Meng

, Zheng

, Wang

and Zhao

: ‘Effect of aging on the phase transformation and mechanical behavior of Ti₃₆Ni₄₉Hf₁₅ high temperature shape memory alloy’, Scr. Mater. , 2000, 42, (4), 341–348.

54.

Tong

, Chen

, Tian

, Li

and Zheng

: ‘Microstructure and martensitic transformation of Ti₄₉Ni_51−xHf_x high temperature shape memory alloys’, Mater. Lett. , 2009, 63, (21), 1869–1871.

55.

Javadi

M. M

, Belbasi

, Salehi

M. T

and Afshar

M. R

: ‘Effect of aging on the microstructure and shape memory effect of a hot-rolled NiTiHf alloy’, J. Mater. Eng. Perform. , 2011, 20, (4–5), 618–622.

56.

Zhang

J. X

, Sato

and Ishida

: ‘Structure of martensite in sputter-deposited Ti-Ni thin films containing homogeneously distributed Ti₂Ni precipitates’, Philos. Mag. A , 2002, 82A, (7), 1433–1449.

57.

Han

, Zou

, Wang

, Zhang

and Yang

: ‘Structure and substructure of martensite in a Ti_36·5Ni_48·5Hf₁₅ high temperature shape memory alloy’, Acta Mater. , 1996, 44, (9), 3711–3721.

58.

Han

, Chung

, Wang

, Zhang

, Lai

JKL

and Yang

: ‘Martensitic transformation in Ti_36·5Ni_48·5Hf₁₅ high temperature shape memory alloy’, Mater. Trans. JIM , 1997, 38, (10), 842–851.

59.

Meng

X. L

, Cai

, Zheng

and Zhao

: ‘Phase transformation and precipitation in aged Ti–Ni–Hf high-temperature shape memory alloys’, Mater. Sci. Eng. A , 2006, A438, 666–670.

60.

Coughlin

, Phillips

, Bigelow

, Garg

, Noebe

and Mills

: ‘Characterization of the microstructure and mechanical properties of a 50·3 Ni–29·7Ti–20Hf shape memory alloy’, Scr. Mater. , 2012, 67, (1) 112–115.

61.

Han

X. D

, Wang

, Zhang

and Yang

: ‘A new precipitate phase in a TiNiHf high temperature shape memory alloy’, Acta Mater. , 1998, 46, (1), 273–281.

62.

Evirgen

, Basner

, Karaman

, Noebe

, Pons

and Santamarta

: ‘Effect of aging on the martensitic transformation characteristics of a Ni-rich NiTiHf high temperature shape memory alloy’, Funct. Mater. Lett. , 2012, 5, (4), 1250038.

63.

Sandu

, Tsuchiya

, Todaka

M. Tabuchi, S. YamamotoY

and Umemoto

: ‘Microstructural evolution during isothermal aging in Ni-rich Ti-Zr-Ni shape memory alloys’, Mater. Trans. , 2007, 48, (3), 432–438.

64.

Nishida

, Ohgi

, Itai

, Chiba

and Yamauchi

: ‘Electron microscopy studies of twin morphologies in B19′ martensite in the Ti-Ni shape memory alloy’, Acta Metall. Mater. , 1995, 43, (3), 1219–1227.

65.

Waitz

, Kazykhanov

and Karnthaler

H. P

: ‘Martensitic phase transformations in nanocrystalline NiTi studied by TEM’, Acta Mater. , 2004, 52, (1), 137–147.

66.

Nishida

, Wayman

C. M

and Chiba

: ‘Electron microscopy studies of the martensitic transformation in an aged Ti-51at Ni shape memory alloy’, Metallography , 1988, 21, (3), 275–291.

67.

Acar

, Tobe

, Kaya

, Karaca

and Chumlyakov

: ‘Compressive response of Ni_45·3Ti_34·7Hf₁₅Pd₅ and Ni_45·3Ti_29·7Hf₂₀Pd₅ shape memory alloys’, submitted to Journal of Materials Science.

68.

Dalle

, Perrin

, Vermaut

, Masse

and Portier

: ‘Interface mobility in Ni_49·8Ti_42·2Hf₈ shape memory alloy’, Acta Mater. , 2002, 50, (14), 3557–3565.

69.

Meng

, Zheng

, Wang

and Zhao

: ‘Shape memory properties of the Ti₃₆Ni₄₉Hf₁₅ high temperature shape memory alloy’, Mater. Lett. , 2000, 45, (2), 128–132.

70.

Olier

, Brachet

, Bechade

, Foucher

and Guenin

: ‘Investigation of transformation temperatures, microstructure and shape memory properties of NiTi, NiTiZr and NiTiHf alloys’, J. Phys. IV , 1995, C8, 741–746.

71.

Meng

, Zheng

, Cai

and Zhao

: ‘Two-way shape memory effect of a TiNiHf high temperature shape memory alloy’, J. Alloys Compd , 2004, 372, (1–2), 180–186.

72.

Lin

and Wu

: ‘The tensile behavior of a cold-rolled and reverse-transformed equiatomic TiNi alloy’, Acta Metall. Mater. , 1994, 42, (5), 1623–1630.

73.

Otsuka

, Oda

, Ueno

, Piao

, Ueki

and Horikawa

: ‘The shape memory effect in a Ti50Pd50 alloy’, Scr. Metall. Mater. , 1993, 29, (10), 1355–1358.

74.

Grummon

D. S

: ‘Thin-film shape-memory materials for high-temperature applications’, J. Miner. Met. Mater. Soc. , 2003, 55, (12), 24–32.

75.

Bigelow

G. S

, Padula

II , Garg

, Gaydosh

and Noebe

: ‘Characterization of ternary NiTiPd high-temperature shape-memory alloys under load-biased thermal cycling’, Metall. Mater. Trans. A , 2010, 41A, (12), 3065–3079.

76.

Tanaka

, Himuro

, Kainuma

, Sutou

, Omori

and Ishida

: ‘Ferrous polycrystalline shape-memory alloy showing huge superelasticity’, Science , 2010, 327, (5972), 1488–1490.

77.

Hao

, Cui

, Wang

, Jiang

, Yu

, Jiang

, Brown

and Ren

: ‘The ultrahigh mechanical energy-absorption capability evidenced in a high-strength NbTi/NiTi nanocomposite’, Appl. Phys. Lett. , 2011, 99, (2), 024102.

78.

Takagi

, Sutou

, Kainuma

, Yamauchi

and Ishida

: ‘Effect of prestrain on martensitic transformation in a Ti_46·4Ni_47·6Nb_6·0 superelastic alloy and its application to medical stents’, J. Biomed. Mater. Res. Part B: Appl. Biomater. , 2006, 76B, (1), 179–183.