R-phase formation in Ti39Ni45Cu16 shape memory thin films and bulk alloys discovered by combinatorial methods

The combinatorial fabrication and high-throughput characterization of a Ti–Ni–Cu shape memory thin film composition spread led to the discovery of the shape memory alloy Ti₃₉Ni₄₅Cu₁₆, which exhibits a single B2 → R-phase transformation above 25 °C with a thermal hysteresis width <1 K. Here we show that the thin film results correctly predict the phase transformation behavior of bulk material upon cooling from the high temperature phase. For both thin film and bulk, a two-step B2 → R-phase → B19′ transformation was found. The B2 → R-phase transformation can be exploited independently, due to a significant temperature separation of the two transformation steps. The shape memory effect in both thin film and bulk samples is limited due to the two-phase microstructure. Transmission electron microscopy investigations revealed the existence of Ti(Ni,Cu)₂ precipitates within the TiNi(Cu) matrix, which are concluded to be responsible for the R-phase formation and separation of the transformation steps.     

Textures in cold-rolled and annealed Ti50Ni50 shape memory alloy

In the 50% cold-rolled and 600 °C annealed Ti₅₀Ni₅₀ alloy, this study observes (1 1 0)[1 1&#x0304; 0]_P α-fiber II in RD as the main texture in the B2 phase by ODF test. Cyclic tensile tests and DMA results indicate that (1 1 1&#x0304;)[1 1&#x0304; 1]_M is the main texture in B19^′ martensite.     

High hydrogen permeability in the Nb-rich Nb–Ti–Ni alloy

The microstructure and the hydrogen permeability of the Nb-rich Nb–Ti–Ni alloy, i.e., the Nb₅₆Ti₂₃Ni₂₁ alloy were investigated and compared with those of the Nb₄₀Ti₃₀Ni₃₀ alloy. The Nb₅₆Ti₂₃Ni₂₁ alloy consisted of a combination of the primary phase bcc- (Nb, Ti) solid solution with the eutectic phase {bcc- (Nb, Ti) + B2-TiNi}. The volume fraction of the former and the latter phases were 62 and 38 vol.%, respectively. The Nb₅₆Ti₂₃Ni₂₁ alloy showed the higher Φ value of 3.47 × 10⁻⁸ (mol H₂ m⁻¹ s⁻¹ Pa^−0.5) at 673 K, which is 1.8 times higher than that of the Nb₄₀Ti₃₀Ni₃₀ alloy, which has been reported to be highest in the Nb–Ti–Ni system. The present work demonstrated that the Nb-rich Nb–Ti–Ni alloys consisting of only the primary phase bcc- (Nb, Ti) and the eutectic phase {bcc- (Nb, Ti) + B2-TiNi} are promising for the hydrogen permeation membrane.     

Effect of precipitation on the shape memory effect of Ti50Ni25Cu25 melt-spun ribbon

The present research aims to provide accurate understanding of the relation between precipitation (volume fraction, morphology, type) and shape memory effect of Ti₅₀Ni₂₅Cu₂₅ melt-spun ribbon. Rapid thermal annealing was used to control the microstructural development while the shape memory effect of the ribbon was determined under constraint thermal cycling. The results show that the precipitation process takes the following sequence: B11 TiCu → B11 TiCu + Ti₂(Ni, Cu) → Ti₂(Ni, Cu) with increasing annealing temperature or duration. The shape memory effect is found to depend on both the volume fraction and the distribution of the precipitates. The former affects the shape recovery strain through reduction of the transformation volume participating the shape recovery. The latter affects the shape recovery strain through strengthening the matrix thus reducing the martensite strain which is more predominant under low constraint stresses. Precipitation strengthening, on the other hand, reduces the tendency of dislocation generation/movement, thus reducing the irreversible strain and improving shape recovery strain. This understanding provides guidelines on the optimization of the shape memory properties of the Ti₅₀Ni₂₅Cu₂₅ melt-spun ribbon via post-processing annealing.     

The effect of annealing on the microstructure and tensile properties of a β/γ′ Ni–Al–Fe alloy

The evolution of the microstructure of a (β/γ ′) Ni–32 at.% Al–5 at.% Fe alloy during annealing has been studied by electron microscopy and X-ray diffraction. Annealing at 800°C and 1100°C causes a reverse martensitic transformation, L1₀→B2 (β), and a B2→L1₂ (γ ′) phase transformation. The lower annealing temperature leads to a higher volume fraction of the γ ′-phase but a smaller size of the γ ′-particles. The kinetic laws of the coarsening and of the increase in the volume fraction of the γ ′-phase are discussed. The orientation relationships between the β and γ ′ phases appeared to be mainly of Nishiyama–Wassermann and Bain types after 800°C annealing, while Kurdjumov–Sachs and Bain orientation relationships were predominant in the alloys annealed at 1100°C. A strong correlation between the volume fraction of the γ ′-phase and the tensile characteristics of the alloy has been established.     

Formation,thermal stability and corrosion behavior of glassy Ti45Zr5Cu45Ni5 alloy

Ti-rich Ti₄₅Zr₅Cu₄₅Ni₅ bulk metallic glass with critical diameter reaching 3 mm and supercooled liquid region of 42 K was prepared by copper mold casting. The glass transition temperature and onset temperature of crystallization are 673 K and 715 K, respectively. The glassy Ti–Zr–Cu–Ni alloy is passive in 3 mass% NaCl, 1 N HCl and 1 N H₂SO₄ solutions, although pitting corrosion occurred by anodic polarization at higher potential in the Cl⁻-containing solutions. Corrosion rates of the glassy Ti₄₅Zr₅Cu₄₅Ni₅ alloy are of the order of 10⁻³ mm/year in the NaCl and H₂SO₄ solutions and about 1 × 10⁻² mm/year in the HCl acid.     

Ordered ω phases in a 4Zr–4Nb-containing TiAl-based alloy

A considerable amount of B2 phase with a cellular morphology is retained in a 4Zr–4Nb-containing TiAl-based alloy. Heterogeneous precipitation of ordered ω from B2 is found to occur readily after HIPping: B2 → ω with the B8₂-structure in cell regions and B2 → ω with the D8₈-structure in cell-wall regions. Congregated D8₈-ω domains and particles form as a network surrounding the well-developed B8₂-ω cells. The heterogeneous formation of different ω variants is caused by a heterogeneous distribution of Zr + Nb elements across B2, which plays an important role in stabilizing vacancies and promotes the formation of D8₈-ω. Fine D8₈-ω particles are also observed to precipitate from the B8₂-ω cell matrix after ageing at 700 °C for 1000 h, showing a transformation path of β → B2 → B8₂-ω → D8₈-ω for the aged cells. The heterogeneous formation of a D8₈-ω network and B8₂-ω cells is found to be detrimental to ductility and fatigue strength. A very brittle fine-grained TiAl alloy is produced as a result.     

Energy landscape for martensitic phase transformation in shape memory NiTi

First-principles calculations are presented for parent B2 phase and martensitic B19 and B19′ phases in NiTi. The results indicate that both B19 and B19′ are energetically more stable than the parent B2 phase. By means of ab initio density functional theory, the complete distortion–shuffle energy landscape associated with B2 → B19 transformation in NiTi is then determined. In addition to accounting for the Bain-type deformation through the Cauchy–Born rule, the study explicitly accounts for the shuffle displacements experienced by the internal ions in NiTi. The energy landscape allows the energy barrier associated with the B2 → B19 transformation pathway to be identified. The results indicate that a barrier of 0.48 mRyd atom^?1 (relative to the B2 phase) must be overcome to transform the parent B2 NiTi to orthorhombic B19 martensite.     

Shape memory behaviour of Ti51.5Ni(48.5−x)Cux (x = 23.4–37.3) thin films with submicron grain sizes

The shape memory behaviours of Ti_51.4Ni_25.2Cu_23.4, Ti_51.3Ni_21.1Cu_27.6, Ti_51.2Ni_15.7Cu_33.1 and Ti_51.4Ni_11.3Cu_37.3 thin films annealed at 773, 873 and 973 K for 1 h were investigated. The Ti_51.3Ni_21.1Cu_27.6 film annealed at 773 K, the Ti_51.2Ni_15.7Cu_33.1 film annealed at 873 K, and the Ti_51.4Ni_11.3Cu_37.3 films annealed at 873 and 973 K showed a perfect shape memory effect at a stress as high as 1 GPa. This improvement in shape memory behaviour was attributed to their fine grain sizes less than 500 nm. Whereas the Ti_51.2Ni_15.7Cu_33.1 and Ti_51.4Ni_11.3Cu_37.3 films annealed at 873 K or higher showed a martensitic transformation start temperature above room temperature, these films annealed at 773 K were in the parent phase at room temperature owing to their very fine grain sizes. The effects of Cu content and annealing temperature on the shape memory behaviour of the Ti_51.5Ni_48.5?xCu_x (x > 27) films with submicron grain sizes were discussed in comparison with those of the Ti_51.5Ni_48.5?xCu_x (x < 24) films on the basis of their microstructures.     

Phase equilibria in the Ni-rich portion of the Ni–Ga binary system

The phase equilibria in the composition range from 0 to 60 at% Ga of the Ni–Ga system were determined by electron probe microanalysis (EPMA) using diffusion couples, differential scanning calorimetry (DSC) and X-ray diffraction (XRD). It was found that while the phase equilibria between the α′ (L1₂: Ni₃Ga) and α (Ni-solid solution) or β (B2: NiGa) phases are basically in agreement with the diagram evaluated by Lee and Nash, those between γ (B8₁: Ni₁₃Ga₇), δ (Cmmm: Ni₅Ga₃) and ɛ (C2/m: Ni₁₃Ga₉) are topologically different from that diagram. Three eutectoid reactions (γ → δ + ɛ, β → γ + ɛ, β → α′ + γ) and one peritectoid reaction (α′ + γ → δ) were confirmed and the temperatures and concentrations of those invariant reactions were determined.     

Microstructures of crystallized Ti51.5Ni48.5−xCux (x = 23.4–37.3) thin films

The microstructures of Ti_51.4Ni_25.2Cu_23.4, Ti_51.3Ni_21.1Cu_27.6, Ti_51.2Ni_15.7Cu_33.1, and Ti_51.4Ni_11.3Cu_37.3 thin films annealed at 773, 873 and 973 K for 1 h were investigated. The Ti_51.2Ni_15.7Cu_33.1 and Ti_51.4Ni_11.3Cu_37.3 films showed very small grain sizes (120 and 50 nm, respectively) compared with the Ti_51.4Ni_25.2Cu_23.4 and Ti_51.3Ni_21.1Cu_27.6 films (1 and 0.3 μm, respectively). They had no precipitates within the B2 grains. On the other hand, the Ti_51.4Ni_25.2Cu_23.4 films annealed at 773 and 873 K showed GP zones and Ti₂Cu precipitates, respectively, and the Ti_51.3Ni_21.1Cu_27.6 film annealed at 773 K showed TiCu precipitates in the grain interiors. The formation of precipitates in the grain interior was discussed in terms of the lattice mismatch between the precipitates and the matrix. The difference in grain size was attributed to different crystallization processes.     

Infrared brazing of Ti50Ni50 shape memory alloy using two Ag–Cu–Ti active braze alloys

Microstructural evolution, shape memory effect and shear strength of infrared brazed Ti₅₀Ni₅₀ shape memory alloy using Cusil-ABA^® and Ticusil^® active braze alloys are investigated. The Ag–Cu eutectic braze alloy can readily wet Ti₅₀Ni₅₀ substrate by minor titanium additions. The brazed Ti₅₀Ni₅₀/Cusil-ABA^®/Ti₅₀Ni₅₀ joint is mainly comprised of Cu-rich, Ag-rich and CuNiTi phases. On the other hand, the brazed Ti₅₀Ni₅₀/Ticusil^®/Ti₅₀Ni₅₀ joint consists of Ag-rich, Cu-rich and TiCu₂ phases. Because the chemical composition of Ticusil braze alloy is located inside the huge miscibility gap, the molten braze tends to be separated into two liquids during brazing. One is rich in Ag, and the other is rich in both Cu and Ti. The Ag-rich liquid does not react with Ti₅₀Ni₅₀ substrate. In contrast, the copper content is depleted from the matrix of brazed joint due to the formation of interfacial TiCu₂ phase. The TiCu₂ phase is less detrimental to the shape memory effect than CuNiTi phase during the shape recovery bending test. Shear strength of brazed joints exceeds 200 MPa for both braze alloys if the brazing time exceeds 180 s. However, thick interfacial CuNiTi and TiCu₂ layers can deteriorate the shear strength.     

Low-frequency internal friction of hydrogen-free and hydrogen-doped NiTi alloys

The internal friction (IF) and Young’s modulus of the Ni_50.8Ti_49.2 shape memory alloy have been measured as a function of temperature (130 K < T < 335 K) by a dynamic mechanical analyser at various strain amplitudes and frequencies. Besides the one associated with the austenite/martensite transformation, several other IF peaks have been observed both in the hydrogen-free and in the hydrogen-doped states of the material. Some of these peaks are non-thermally activated processes caused by stress-assisted hysteretic motions of twin boundaries and dislocations; some others represent thermally activated relaxations caused by reorientation of hydrogen elastic dipoles or by stress-induced motions of twin boundaries interacting with hydrogen. The present low-frequency measurements provide new information concerning the amplitude and frequency dependences of the damping processes, thus throwing new light on their structural mechanisms.     

Phase transformation behavior of Ti49Ni49.5Fe1V0.5 and Ti48Ni48.5Fe1V2.5 alloys after different heat treatments简

The effects of heat treatments on the phase transformation behavior of Ti49 Ni49.5 Fe1 V0.5 and Ti48 Ni48.5 Fe1 V2.5 alloys were investigated. The results indicate that the alloys subjected to different heat treatments have B2 structure at room temperature. All the specimens exhibit a twostage B2→R→B190martensitic transformation on cooling, but a B190→B2 one-stage reverse martensitic transformation on heating except aged A1 alloy, which undergoes an abnormal two-stage transformation upon heating. The phase transformation temperatures are affected by heat treatments and V content, which can be attributed to the variation of the second-phase particles content in the matrix.     

Oxidation behavior of nano-scaled and micron-scaled TiC powders under air

The oxidation behavior of several batches of TiC nanopowders and micropowders under air has been studied. Using TGA, the oxidation of nanopowders is completed faster and at a lower temperature compared to micropowders. This is related to the higher specific surface area of the nanopowders. Using an amount of powder of about 41.6 mg through DTA, three different exothermic peaks were observed for both kinds of powders. Correlated to the in-situ temperature XRD diffractograms, these peaks were attributed to the following oxidation reactions: i) TiC → TiC_xO_1 − x up to TiO, ii) → Ti₃O₅, and iii) → TiO₂. The same successive oxide phases for nanopowders and micropowders were observed, meaning that the oxidation mechanism is similar in both cases but with a shift of temperature. In this article, the role of the oxygen trapped in the powder and the effect of powder quantity on the oxidation kinetics and on the occurrence of one or three oxidation peaks are discussed.     

Shape memory behavior of high strength NiTiHfPd polycrystalline alloys

Systematic characterization of the shape memory properties of a quaternary Ni_45.3–Ti_29.7–Hf₂₀–Pd₅ (at.%) polycrystalline alloy was performed in compression after selected aging treatments. Precipitation characteristics were revealed by transmission electron microscopy. The effects of aging temperature and time on transformation temperatures, recoverable and residual strains, and temperature and stress hystereses were determined by differential scanning calorimetry, constant-load thermal cycling experiments and isothermal strain cycling (superelasticity) tests. The crystal structure and lattice parameters of the transforming phases were determined from X-ray diffraction analysis. It was revealed that precipitation hardening significantly improved the shape memory properties of the NiTiHfPd alloy. Under optimum aging conditions, shape memory strains of up to 4% under 1 GPa were possible, and superelasticity experiments resulted in full strain recovery without any plastic deformation, even at stress levels as high as 2 GPa. The NiTiHfPd polycrystalline alloy exhibited very high damping capacity/absorbed energy (30–34 J cm^?3) and work output (30–35 J cm^?3), which were attributed to the ability to operate at high stress levels without significant plastic deformation and to a high mechanical hysteresis (>900 MPa) at temperatures ranging from 20 °C to 80 °C.     

Effect of nano-scaled precipitates on shape memory behavior of Ti-50.9at.%Ni alloy

In order to clarify the effect of low temperature aging (<600 K) on the shape memory behavior of a Ti-50.9at.%Ni alloy, the deformation behavior and microstructures were investigated using tensile tests, X-ray diffractometry and transmission electron microscopy (TEM). The specimens were solution-treated at 1073 K for 3.6 ks followed by aging at 373, 473 and 573 K for various times ranging between 1.2 and 10,800 ks. TEM observation revealed that after aging at low temperature (<600 K) the size of Ti₃Ni₄ precipitates hardly increases, but their density increases with increasing aging time in the early stage of formation of Ti₃Ni₄ precipitates. The critical stress for slip deformation abruptly increases with increasing density of fine Ti₃Ni₄ precipitates (<10 nm) in the specimens aged at the low temperature region. This is attributed to the increase in the number of pinning points to hinder the movement of dislocations. With increasing aging time and temperature the size of Ti₃Ni₄ precipitates increases, resulting in decreasing the critical stress for slip.     

Microstructure and hydrogen permeation of cold rolled and annealed Nb40Ti30Ni30 alloy

The ultimate rolling reduction rate (r_u) of the as-cast Nb–TiNi alloys was evaluated by measuring the thickness changes in cold rolled samples. Furthermore, the effects of cold rolling and subsequent anneal on hardness, microstructure, hydrogen permeability (Φ) and hydrogen flux J of the as-cast Nb₄₀Ti₃₀Ni₃₀ alloy (mol%) were examined by a Vickers hardness tester, a scanning electron microscope (SEM), and a mass flow meter, respectively. The value of Φ for the Nb₄₀Ti₃₀Ni₃₀ alloy was reduced with increasing rolling reduction rate and attained to one third of the original one by 50% rolling reduction, but recovered to the original one by anneal at 1373 K for 605 ks. Hydrogen flux J varied inversely proportional to the membrane thickness at 673 K. J of 12 ccH₂/cm²/min was attained for the sample with the thickness of 120 μm. The present work has demonstrated that rolling and the subsequent anneal are effective and useful for the preparation of the hydrogen permeation Nb–TiNi alloy membrane.     

Combined effects of work hardening and precipitation strengthening on the cyclic stability of TiNiPdCu-based high-temperature shape memory alloys

The combined effects of work hardening and precipitation strengthening were employed to improve the cyclic stability of TiNiPdCu-based high-temperature shape memory alloys. Annealing after cold deformation resulted in the formation of nano-scale TiPdCu and Ti₂Pd precipitates, stable at high temperatures in Ti₅₀Ni_25?xPd₂₅Cu_x alloys. The nano-scale precipitates were also observed to retard recovery/recrystallization processes at higher temperatures. It was found that the combined effects of work hardening and precipitation strengthening remarkably enhanced the high-temperature stability of the Ti₅₀Ni₂₀Pd₂₅Cu₅ alloy and increased its maximum working temperature range while keeping the transformation temperatures and recovery strains at sufficiently high levels. Precipitation strengthening helped to greatly improve the high-temperature cyclic stability of the alloy. Creep tests at 673 K under 500 MPa confirmed that the better high-temperature cyclic stability of the precipitate-containing alloy was mainly due to its higher creep resistance.     

Mechanical spectroscopy and twin boundary properties in a Ni50.8Ti49.2 alloy

The internal friction (IF), Young's modulus (E) and heat flow (DSC) have been measured in a Ni_50.8Ti_49.2 alloy as a function of temperature after both ageing and solubilization followed by water quenching. The IF of aged material measured at kilohertz frequencies, besides hardly detectable peaks associated with the A↔R and R↔M transitions, displays two additional non-thermally activated peaks P_TWM and P_TWR, which appear to be due to stress-induced hysteretic motions of twin boundaries within the martensite and the R-phase, respectively. In the solubilized state a large peak corresponding to the A↔M transition is observed together with a thermally activated relaxation P_d occurring at around 100 K (for f=1.2 kHz). The activation energy and limit time of this peak are 0.11 eV and 10⁻¹⁰ s, respectively. Peak P_TWM was not observed in the solubilized state. Peak P_d, which is found to increase with the number of thermal cycles through the transition region undergone by the sample, is attributed to stress-induced motions of dislocations around their equilibrium positions within the B19′ structure of the martensite. The absence of a twin boundary peak in the martensite formed from the solubilized austenite, as well as the large difference in the strength of the transformation relaxations in the aged and solubilized states of the material are accounted for in terms of the different nature of the predominant twin boundaries in the two different states of the material.