The atom positions in the structure of the R-phase in a TiNi(Fe) intermetallic compound have been refined using the multi-slice least-squares method based on electron diffraction data obtained from 50 nm sized regions in a CM30 FEG TEM instrument. The refinement accounted for dynamic scattering including the specimen thickness and crystal misorientation as refining parameters. Compared with the original positions in the (111) planes of the parent B2 phase, each third Ti and Ni plane in the R-phase separates into three layers with different z-coordinates. This agrees with previous experiments in which it was concluded that the R-structure belongs to the space group P3 rather than P\\bar31m as determined earlier by convergent beam electron diffraction (CBED). However, our data reveal a centre of symmetry in the R-phase structure leading to the P\\bar3 space group (R-factor of 5.5%). Finally, it is suggested that the fine scale antiphase-like domains observed in the R-phase appear due to mirror planes over which the above mentioned atomic shifts in the structure are reversed. This phenomenon is expected to cause an artificial symmetry increase in the space group determination when using CBED with a probe size close to the size of the domains.

Titanium nickel (TiNi) and the pseudo binary containing a small amount of iron has the R phase as an intermediate phase. To examine theoretically the stability of the R phase, the electronic structures of TiNi and TiNi_8⁄9Fe_1⁄9 were calculated for four structures of B2, P3, P\\bar31m and P3lm. The total energies predict that the P3 structure is most stable among the four structures and in the P3 structure the added iron atoms prefer the 1c site. It is also revealed that the approach between a nickel (or iron) atom at the 1c site and a titanium atom at the 1c site in the P3 structure plays an important role in stabilizing the P3 structure. These features are clearly reflected on the band energies and the local density-of-states of the constituent atoms.

Ni–Ti Shape Memory Alloys (SMA) are of great technological interest because they have the best shape memory behaviour of all SMA. Moreover, Ni–Ti thin films are considered to be one of the most promising solutions for the development of new microactuators. In order to use Ni–Ti in the thin film-state, it is therefore important to know if the properties of the martensitic phase transformation responsible for the shape memory effect are different in Ni–Ti thin films and in bulk materials. For that purpose, ⁶¹Ni Nuclear Magnetic Resonance (NMR) measurements have been performed at very high magnetic field (14 T) in Ni–Ti bulk alloys as well as in thin films. The process of nucleation and growth of the R-phase and of the martensite in the different samples have been studied by means of a careful analysis of the ⁶¹Ni NMR spectra recorded at different temperatures from T>R_s to T<M_f. The complex structure of the NMR spectra during phase transformation has been interpreted as a sum of contributions arising from the coexisting crystalline phases. The spectral deconvolution is very difficult to achieve because the NMR martensitic line of complex shape is partly superimposed on the austenitic line. However, additional NMR measurements in a ⁶¹Ni enriched Ni–Ti bulk alloy allowed us to identify the microscopic interactions responsible for the martensitic line shape. With the help of this analysis it was then possible to determine the volume fraction of the different phases present at each temperature in the Ni–Ti bulk alloys and thin films. The NMR results showed that in Ni–Ti bulk alloys, where the R-phase transformation does not take place or is only partial, the martensitic transformation is not complete even at temperatures well below M_f as determined by calorimetry. On the other hand, in thin films showing a well distinct two-step transformation via the R-phase no remaining austenite could be detected below the R_f temperature.

It has been observed that deformation stabilises martensite in a number of shape memory alloys, as evidenced by the increase of the critical temperature for the reverse transformation of the deformed martensite as compared to undeformed martensite. Some hypotheses have been proposed in the literature to explain this phenomenon, including the pinning effect of deformation-induced defects and the release of internal elastic energy stored in thermal martensite. This study continues the experimental work by studying the effect of ferroelastic cycling via martensite reorientation on the transformation behaviour of a binary near-equiatomic NiTi, with the aim to provide further experimental evidence for the clarification of the mechanisms responsible for this effect. It was observed that the critical temperature and the heat of the reverse transformation of oriented martensite increased progressively with deformation cycles, although the limits of the deformation cycles remained unchanged.

Hydrogenated TiNi alloy were investigated by X-ray diffraction and the Rietveld refinements from the view of transformation behavior. Small expansion of lattice constants indicated that hydrogen atoms located at interstitial sites in parent and martensite phases were small amount. The refinements also indicated that the amount of hydride TiNi(H) was approximately 0.1%, which was small. These refinements and optical microscopic observations indicated most of hydrogen atoms were located around grain boundaries or martensite variant boundaries. DSC measurements showed that hydrogenated TiNi displayed multiple-stage transformation behavior and thus hydrogen atoms around grain boundaries may cause mutiple-stage behavior of the transformation.

Phase transformation behavior, the shape memory characteristics and the superelasticity of Ti–Ni–Cu–Mo alloys have been investigated by means of electrical resistivity measurements, X-ray diffraction, thermal cycling tests under constant load and tensile tests. A fully annealed Ti–44.7Ni–5Cu–0.3Mo alloy transformed in two-stage, i.e., the B2-B19-B19^′ on cooling and the B19^′-B19-B2 on heating. Fully annealed Ti–39.7Ni–10Cu–0.3Mo, Ti–34.7Ni–15Cu–0.3Mo and Ti–29.7Ni–20Cu–0.3Mo alloys transformed in one-stage on, i.e., from the B2 to the B19 on cooling and from the B19 to B2 on heating. The maximum recoverable elongation deceased from 6.0 to 2.4% with increasing Cu-content from 5 to 20 at%. Transformation hysteresis associated with the B2-B19 transformation decreased from 11 to 5 K with increasing Cu-content from 5 to 20 at%. Substitution of Mo for Ni in Ti–Ni–Cu alloys improved the superelasticity.

We proposed a new model, “series-parallel combined model”, for the fatigue of Ti–Ni–Cu shape memory alloy subjected to superelastic cyclic deformation in order to clarify a fatigue of shape memory alloy due to the cyclic phase transformation. Our model is based on the detailed observation of the stress-strain behavior during the one cycle superelastic deformation, considering together the peculiar fatigue caused by Lüders deformation type of cyclic phase transformation. The model was used to predict the fatigue crack origin in the fracture surface. The predictions showed that fatigue life was governed by the failure of the earliest transformed martensite phase. For the purpose of proving the credibility of the model, the fatigue tests were carried out by using our original machine, which was made attentively so as not to change the given nominal strain amplitude and generate the bending deformation of the specimen due to the irrecoverable strain. The fatigue life curve showed the peculiar strain amplitude dependence, and had the peculiar strain amplitude region where the decrease of fatigue life with increasing the given strain amplitude did not occur under either of the conditions able to generate the phase transformation in the parent phase. The fatigue origins of the entire specimens exist in the central region of the fracture surface. The multi-fatigue cracks were observed in the region of origins and ran axially. The transverse cracks among the multi-fatigue cracks propagate from the central region into the surrounding regions. The prediction by our new model was found to agree well with experimental results and the detailed fracture surface observation by electron microscopy.

The effects of cyclic deformation and copper content on the thermo-mechanical characteristics in Ti–Ni–Cu alloys were investigated. Thermo-mechanical cyclic tests were conducted for various strains at a fixed heating temperature. Specimens were Ti–45Ni–5Cu, Ti–40Ni–10Cu and Ti–37Ni–13Cu (at%), annealed at 673 K for 3.6 ks after cold drawing with 30% reduction. The results show that the change of functions such as residual strain and the strain energy is significant in early cycles, but it becomes insignificant after 100 cycles. Also, the change of functions with number of cycles shows the dependence on copper content. In order to clarify the effects of cyclic deformation and copper content on the degradation of functions, the volume fraction of slip-deformed martensite is evaluated by a two-phase model consisting of the parent phase and the martensitic phase connected in series. The volume fraction of slip-deformed martensite represents the variation of the residual strain and the degraded recovery strain energy with number of cycles and copper content. Based on these results, it is concluded that the volume fraction of slip-deformed martensite is a measure which represents the effects of cyclic deformation and copper content on the degradation of functions.

Our new model, “series combined model”, for the superelastic deformation in Ti–Ni–Cu shape memory alloy was proposed considering the mechanical deformation phenomenon in a Lüders deformation type of phase transformation in order to clarify the progressing behavior of the phase transformation and then the suitable physical formula for it in future. By assuming the competitive phenomenon between the phase transformation and the plastic deformation, this model was applied for transformation behavior progressing together with the local plastic deformation. In the application of the model, some significant transformation points in stress-strain curve were determined originally to be well fit for the superelastic deformation of shape memory alloy. The original constitutive equation, which based on this model, was used to calculate the martensite volume fraction at every point in one cycle of superelastic deformation by making the inverse analysis for the stress-strain data obtained experimentally under the various temperatures. The validity of this model and constitutive equation for superelastic deformation was confirmed by comparison between the experimental result of a half cycle and the prediction for the stress-strain curve based on the inverse analysis result in a full cycle. The prediction for temperature dependence of the volume fraction and the magnitude of plastic strain in the plastic deformed martensite phase could explain enough the temperature dependence of the irrecoverable strains obtained experimentally. The progressing rate of the martensite transformation as well as the reverse transformation, which was calculated by making the inverse analysis results, decreased depending on the increase of temperature.

The effect of copper content on superelasticity characteristics in Ti–Ni and Ti–Ni–Cu alloy wires was investigated by performing isothermal cyclic tensile tests at a temperature of A_f+25 K. Specimens were Ti–50Ni, Ti–45Ni–5Cu, Ti–40Ni–10Cu and Ti–37Ni–13Cu(at%), annealed at 673 K for 3.6 ks after cold drawing with 30% reduction. The results show that the changes in residual strain, the critical stress for inducing martensite and the strain energy in all alloys are significant in early cycles, but become insignificant after 10 cycles. The degradation of residual strain and the strain energy during loading increases with decreasing copper content. However, changes in the critical stress for inducing martensite and the dissipated strain energy in Ti–Ni–Cu alloys are insensitive to copper content. Furthermore, in order to clarify the effect of copper content on the degradation of materials functions, the volume fraction of martensite subjected to slip deformation is evaluated by a two-phase model consisting of the parent phase and the martensitic phase connected in series. The volume fraction for a residual strain becomes larger as copper content increases, and it is directly related to the critical stress for inducing martensite and the dissipated strain energy with number of cycles. Based on these results it can be stated that the volume fraction of martensite subjected to slip deformation is a measure which represents the effects of cyclic deformation and copper content on the degradation of materials functions.

Relation between tensile deformation behavior and microstructure in variously aged Ti–49.7 at%Ni–1.3 at%Co shape memory alloys has been investigated. Stress-strain curves for the alloys aged at 623 K and 723 K for 1.8 ks were of a work-hardening type, but those for the alloys aged at the two temperatures over 28.8 ks were nearly of a constant-stress type, as for binary Ti–51 at%Ni aged alloys. Lenticular precipitates were observed in both the 623 K and 723 K aged alloys, and the distance among the precipitates increased with increasing aging temperature and period, as well as the size of the precipitates. The lenticular precipitates were identified to be of the composition of Ti₃(Ni, Co)₄ from an EDX analysis. Based on these observations, the constant-stress type stress-strain behavior for the alloys aged over 28.8 ks was attributed to some composition change accompanied with the aging progress by which Ti:(Ni+Co) composition ratio in the matrix of the Ti–Ni–Co alloys approached 1:1, as in the equi-atomic Ti–Ni binary alloys.

The Co-free materials with high erosion resistance are anticipated for parts of equipment in nuclear power plants. The erosion resistance of Ti–Ni shape memory alloys (SMAs) against the impact of hot water jet onto the specimen surface was investigated experimentally and by finite element method (FEM). The results are compared with that of an existing Co-based alloy (Stellite). Both of the erosion damaged cross-sectional area and the maximum damaged depth increased linearly with the elongated exposure time. The damaged areas of SMAs were extremely small compared to those of Stellite. However, no significant difference in the maximum damaged depth was found between the two materials. From the FEM results, larger values of the maximum shear stress and the mean stress were found to distribute in the specimens in testing. It is estimated that the essential erosion damage mechanism of Ti–Ni SMAs is cavitation. In Stellite a combination of the shear stress and the cavitation controls erosion. The erosion resistance of the Ti–Ni SMAs is superior to that of Stellite. It is suggested that the Ti–Ni SMAs will be one of the promising alternative materials for Stellite.

In shape memory Ni₂MnGa based alloys, the structural transition temperature (T_t) increases with increasing valence electron concentration per atom (e⁄a). That is, when Ni or Mn atoms in Ni₂MnGa are replaced with a fourth element (X atom), the T_t increases with increasing atomic number of the X atom. To examine the experimental results, the electronic structures of these alloys were calculated for the cubic and monoclinic structures. The difference ΔE of total energies between the two structures was also calculated as a function of e⁄a. It was found that the features of T_t (e⁄a) are similar to those of ΔE (e⁄a). Their features are different in the lower and higher ranges than the boundary e⁄a=7.625. The two features are characterized by two cases: a case that X atoms occupies Ni sites and the other case that X atoms occupy Mn sites. The characteristics mainly come from the difference of the density of states of X atoms at Ni and Mn sites.

The effect of heat treatment and aging on hardness was investigated using near-stoichiometric and off-stoichiometric Ni₂MnGa ferromagnetic shape memory alloys. The alloys fabricated by Ar arc-melting method were homogenized at 1273 K, and systematic heat treatments from 773 to 1173 K. Moreover, in order to clarify the aging behavior, some of the alloys were homogenized at 1373 K followed by aging at temperatures of room temperature (RT), 373 K and 473 K. The effects of heat treatment and aging were evaluated through micro-Vickers hardness measurements. It was revealed that hardness increases with increasing heat treatment temperature. It was also found that the hardness values are also higher with deviating from the stoichiometric composition, and that the hardness is higher at the Mn-rich side than at the Mn-poor side of stoichiometry. Moreover, aging behavior was observed for both stoichiometric and off-stoichiometric alloys even at RT. The low temperature aging must be caused by excess vacancies introduced during heat treatment. It is concluded that hardness of Ni₂MnGa is sensitive to the chemical composition, heat treatment and aging conditions.

The structural changes and magnetic anomalies accompanying martensitic transformations in Ni–Mn–Ga alloys are briefly discussed. The role of lattice tetragonality of martensite in the reduction of magnetic field needed for the observation of large magnetostrain effect is theoretically analyzed, considering the compensation of the magnetic anisotropy. The possibility of the field reduction is based on the previously observed lattice parameter dependence on the temperature and proper fit of the alloy specimen shape. The model shows that a significant reduction of the magnetic field needed for the giant MSE can take place in martensites with 0.98<c⁄a<1.04.

Ni₂MnGa alloy is an intelligent material with ferromagnetic and shape memory properties. The application of the alloy films to micro-actuators has been proposed. The Ni-rich Ni₂MnGa alloy films with a thickness of nearly 5 \\micron were deposited on Al₂O₃ substrates by a radio-frequency magnetron sputtering apparatus using a Ni₅₂Mn₂₄Ga₂₄ target. They were heat-treated at 1073 K for 36 ks for homogenization and ordering. The martensitic transformation temperatures of the heat-treated films were higher than room temperature. To investigate the effect of aging time on shape memory properties, the heat-treated films were aged at 673 K for various times between 0.9 and 57.6 ks in a constrained condition. The constraint-aged films showed the two-way shape memory effect by thermal cycling. Fine precipitates with a crystal structure of L1₂ were observed in the constraint-films aged for a long period. As for their two-way shape memory properties, a range of transformation temperature narrowed and the amount of macroscopic shape change increased with increasing aging time.

Ni_2.17Mn_0.83Ga transforms between a cubic structure and a coexistent state of an orthorhombic and a monoclinic structure. The electronic structures of Ni_2.17Mn_0.83Ga and Ni₂MnGa were calculated for thin films to investigate the influence of excess nickel and the effect of the surface on the phase stability. The total energy and band energy were also calculated to examine the role of the each constituent atoms in the structural stability. It is theoretically predicted that the Ni_2.17Mn_0.83Ga film preserve the shape memory property like the bulk states and that the excess nickel atoms contribute largely to the stability of the monoclinic structure.

Optical emission spectroscopy can be used to monitor the composition of Ni₂MnGa thin films during sputtering. By choosing peaks of Ni:341.5 nm, Mn:403.1 nm and Ga:417.2 nm, the Ar pressure is found to affect the spectrum intensities of Ni, Mn and Ga atoms, as well as the intensity ratios of I_Mn⁄I_Ni and I_Ga⁄I_Ni. However, the r.f. power has no obvious effect on them. This may be due to the ferromagnetic characteristic of Ni, or that different metals have different energy distributions of sputtered atoms, or that they need various p·d values to be thermalized. Here, p is the Ar pressure and d is the target and substrate distance. The intensity ratios of these peaks are found to be proportional to the composition ratios (mol ratio) of thin films with the relations: C_Mn⁄C_Ni=0.0151(I_Mn⁄I_Ni)+0.392 and C_Ga⁄C_Ni=0.0720(I_Ga⁄I_Ni)+0.273. Hence, the composition of sputtered thin films can be predicted by monitoring the intensity of light emission from the sputtering plasma.

An equivalence principle for mechanical and magnetoelastic stresses is used for the quantitative theoretical description of giant magnetostrain effect in ferromagnetic martensite. Field-induced strains are computed in the framework of phenomenological magnetoelastic model for two different orientations of magnetic field with respect to the crystal axes. Good agreement between the theoretical and experimental field dependencies of the strains in Ni–Mn–Ga alloys is achieved.

Fe–Ga alloys, in which the α-Fe structure is maintained, are rich sources of high strength, low cost magnetostrictive alloys for transducer and vibration reduction applications. Although the magnetostriction of Fe itself is very low, when a relatively small fraction of the Fe atoms are replaced by Ga, the magnetostriction, (3/2)λ₁₀₀, increases greatly. Until recently, the highest magnetostriction was found with the replacement of Fe by Al (Alfenol). In this paper, we present our measurements of magnetostriction on Fe_1−xGa_x, 0.13≤x≤0.24, (Galfenol). With the substitution of 19% Ga for Fe in Fe_1−xGa_x, a 12-fold increase in magnetostriction to ∼ 400 ppm occurs, even though Ga is non-magnetic. In these alloys, the saturation magnetizations remain high, M_s≅1.7 T, and the Curie temperatures are far above room temperature, T_C≅700°C. In most alloys studied, the magnetostrictions and magnetizations are fully saturated in fields less than 24 kA/m, even under compressive stresses >100 MPa. For x=0.24 (near Fe₃Ga), an anomalous increase in magnetostriction with temperature occurs with a peak magnetostriction above room temperature. Small additions of Ni and Mo to the binary Fe–Ga alloys decrease the room temperature value of λ₁₀₀.

Effects of magnetic field and hydrostatic pressure on martensitic transformation have been examined by using Fe–Pt, Cu–Al–Ni, Ni₂MnGa and Fe–Pd shape memory alloys and Fe–Ni alloys. Following results are obtained; (i) in Fe–Pt, Cu–Al–Ni and Fe–Ni alloys, the experimental magnetic field and/or hydrostatic pressure dependence of martensitic transformation start temperature, M_s, is in good agreement with the dependence calculated by the equation previously proposed by our group to evaluate the relation between M_s and magnetic field and hydrostatic pressure. (ii) Giant magnetostriction has been observed in the martensite state of Ni₂MnGa and Fe–31.2 at%Pd ferromagnetic shape memory alloy single crystals. The values (contraction of 3.8% for Ni₂MnGa and expansion of about 3% for Fe–31.2 at%Pd) are nearly the same values as can be expected from the perfect conversion of variants, i.e., variants are converted to preferable variants whose magnetocrystalline anisotropy energy is minimum among them under the magnetic field.

We fabricated bottom spin valves (SV) films using Mn–17 mol%Ir–2 mol%Pt antiferromagnetic material. A bottom SV composing of Ta/Ni–20 mol%Fe seed layer shows an improved exchange field (H_ex). The high H_ex of about 17.4 kA/m was obtained in Si(100)/Ta 3 nm/NiFe 3 nm/Mn–17 mol%Ir–2 mol%Pt 7 nm/Co–10 mol%Fe 1 nm/NiFe 5 nm/Ta exchange biasing layer. The giant magnetoresistance (GMR) and the thermal stability of bottom SVs were evaluated by comparing with a top SV. Bottom SV showed the higher GMR ratio of about 5.2% than a top SV with same thick ferromagnetic layer. It seems that a large short circuit effect of conductance in a free layer of a bottom spin valve. In order to improve thermal stability of a bottom SV, we inserted the synthetic antiferromagnetically coupled pinned layers (Co–Fe/Ru/Co–Fe) between the Mn–17 mol%Ir–2 mol%Pt and Cu layers. Thus, the enhanced thermal stability of H_ex can be obtained in bottom synthetic SVs.

Effect of aging on successive martensitic transformation in a Ti–47 at%Pd shape memory alloy has been studied by means of differential scanning calorimetry (DSC) and transmission electron microscopy (TEM). After short time aging successive transformation takes place in Ti–47 at%Pd alloy irrespective of aging temperature. On the other hand, after relatively prolonged period only the specimen aged above 1073 K shows successive transformation. The change of transformation behavior with aging condition is discussed on the basis of equilibrium between TiPd matrix and Ti₂Pd precipitate. The homogeneity range of TiPd compound is also estimated from the transformation behavior and TEM observations.

The combination of 9R martensite plate variants in Ti_50.0Pd_43.0Fe_7.0 shape memory alloy has been investigated by conventional transmission electron microscopy (CTEM). Three fundamental combinations of plate variants are identified in the plate group. These are designated as A:B, A:C and A:D types, which correspond to ⟨\\bar591⟩_9R Type II, {11\\bar4}_9R Type I and {105}_9R compound twins, respectively. They show the same morphological characteristics of 9R and 18R martensite in Cu-base shape memory alloys, i.e., wedge, spear and fork (or kink) types. The Type II and compound twins are new findings. Irrational nature of the A:B interface is also studied in the edge-on state by high resolution electron microscopy (HREM). The boundary is gradually and randomly curved with strain contrast.

A systematic study on martensitic transformation in Ti-rich Ti–Pd alloys has been carried out using differential scanning calorimetry (DSC) and transmission electron microscopy (TEM). The alloys quenched from the single region of B2 parent phase show a successive transformation during DSC measurement. On the other hand, the furnace-cooled alloys show a single transformation. The successive transformation behavior is closely related to the formation of fine Ti₂Pd precipitates with C11_b-type structure during transformation cycle. The first peak on DSC heating curve is attributable to the reverse martensitic transformation of the TiPd matrix, while the second one is due to the reverse martensitic transformation in local regions around the Ti₂Pd precipitates where Pd concentration is higher than that in matrix. Morphological characteristics of the precipitate are also discussed.

This paper presents experimental results relating the initial austenite grain size to bulk hardness, compressive yield stress (σ_0.2%) and volume fraction of stress-induced ε martensite. It is shown that the bulk hardness obeys quite closely the Hall-Petch equation while the yield stress, σ_0.2%, decreases with decrease of grain size, indicating that the induction of ε martensite mechanically is easier for the materials with finer grains. This fact is corroborated by the observed increase of the volume fraction of stress-induced ε martensite with decrease of grain size. Inversely, after shape recovery heating, the volume fraction of residual stress-induced ε martensite increases as the grain size increases, e.g. the increase in grain size hinders the reversible martensitic transformation.

Nitrogen-microalloying and partial substitution of Cr for Mn have been employed to enhance the shape memory effect and corrosion resistance of Fe–Mn–Si based alloys. Typically, the tested alloys with nominal composition Fe–25Mn–6Si–5Cr–(0.12–0.14)N in mass% exhibit perfect shape recovery for a 3% pre-strain after only one cycle of thermomechanical training. The related mechanism has been discussed, taking account of the effect of nitrogen on the stacking fault energy (SFE) or the stacking fault probability (P_sf) of the alloy and the strengthening of the austenite matrix. Thermodynamic calculation and P_sf measurement showed that the SFE increases with increasing N-content in the concentration range investigated, e.g. less than 0.3 mass%. Thus, the critical stress for the formation of stress-induced martensite increases with N-content. It is believed that the interstitial strengthening of the matrix by nitrogen predominantly contributes to the improvement of shape memory effect. Besides, nitrogen-microalloying remarkably improves the corrosion resistance of the alloys in aqueous solutions containing NaOH and NaCl, but not in HCl solution as indicated by the long-term immersion tests.

Experimental mesoscopic observations of the parent phase in Cu-based shape memory alloys, the associated martensitic transformation and the corresponding hysteresis cycle, are carried out. The coexistence effects between martensite and parent phase and the evolution of the transformation temperature (Ms), due to yearly effect on the parent phase, are determined. Several years of continuous measurements, mainly using Cu–Zn–Al alloys, establish: (A) The evolution of the transformation temperature with time and temperature in parent phase (amplitude close to 15% of the ambient temperature change). (B) The changes related with the coexistence among the phases (close to 5 K). (C) The existence of remnant after quench effects are also partially visualized via neutron diffraction. A model based on the experimental results describes the hysteresis cycle and the internal loops. The diffusion effects can be included in the model and the time evolution of Ms and of the hysteretic behavior can be well established. The simulated results allow predicting the damping performance of the material for long-term actions.

We present a study demonstrating the capability for controlled shape memory thin film growth using molecular beam epitaxy. Here, NiTiCu alloy films were grown which are known to exhibit the martensitic transformation well above room temperature. Remarkably, the microstructure of these films was found to be very different compared to conventionally sputtered polycrystalline films: here, the crystallites are highly oriented within ±3^° along the film plane normal. Moreover, a splitting of the martensite orientation is detected indicating the selection of two specific sets of martensite variants. Mechanical stress measurements reveal a high ratio of recoverable stress even for films below 500 nm thickness. These results open up the possibility to specifically modify the microstructure and crystallographic orientation of shape memory thin films and thus suggest promising characteristics, especially in regard to their superelastic behavior.

TiO₂ films were formed on aluminum plates by a dip-coating in an advanced sol-gel precursor solution, and the effect of sintering conditions on the photocatalytic properties of the oxide were investigated by UV adsorption. The samples prepared in this way exhibit photocatalytic activity in a range of suitable sintering conditions compatible with the region of crystallization of anatase characterized by X-ray diffraction. The surface morphology and adhesion between the TiO₂ nano-film and aluminum plate were confirmed by energy filtering transmission electron microscopy, and the surface morphology was found to affect the photocatalitic activity. The TiO₂ film was 30 nm thick and was confirmed to be finely crystallines with a mean diameter of 12 nm.

The effect of postannealing on surface potential stability was investigated for silicon dioxide (SiO₂) electret thin films with a thickness of 2 to 5 \\micron. The SiO₂ films were prepared on Al-coated and uncoated Si substrates by r.f. magnetron sputtering using a fused quartz target. Subsequent to the sputter deposition, the SiO₂ films were postannealed in the deposition chamber in order to improve stability for use in a highly humid atmosphere. The obtained surface potential stability was dependent on not only the postannealing conditions but also the deposition conditions. The surface potential of SiO₂ films postannealed in an oxidizing atmosphere at 275 to 350°C for 10 to 60 min was found to be highly stable when tested at a relative humidity of 90% and a temperature of 60°C. In addition, the postannealed SiO₂ films were stable for use in air for a long term at room temperature.

The uniaxial stress field in shape memory alloy (SMA) films patterned into thin strips increases the transformation induced deflection of SMA/Si cantilever bimorphs in comparison to cantilevers with planar films. In the single phase temperature ranges T>A_f, M_s and T<A_s, M_f (A_f-austenite finish, A_s-austenite start, M_f-martensite finish and M_s-martensite start temperatures), where the deflection is controlled by the thermoelastic stress, the change reflects the difference between the uniaxial and biaxial stress states. In the temperature regimes A_s<T<A_f, M_f<T<M_s, the martensitic microstructures created by uni- vs. biaxial stress fields are responsible for the difference of the cantilever deflection.

A mono-layer sheet of 24 mm×24 mm and a multi-layer sheet of 12 mm×12 mm were fabricated as follows. Semiconducting BaTiO₃ particles were packed between two electrodes of aluminum foil. As the particles and electrodes were enclosed in a plastic bag, evacuated and heat-sealed, they were fixed and compressed by atmospheric pressure. One layer is formed by the packed particles in the mono-layer sheet, and two to three layers are formed in the multi-layer sheet. The particles are semiconducting BaTiO₃ for the mono-layer sheet and composite particles of semiconducting BaTiO₃ and In for the multi-layer sheet. Indium particles always exist between the semiconducting BaTiO₃ particles in the multi-layer sheet and lower the contact resistance between the semiconducting particles. The properties of the sheets are investigated and the following results are obtained. (1) Both the mono-layer sheet and the multi-layer sheet are flexible and show the PTC property. (2) The performance of the multi-layer sheets is almost the same with that of the mono-layer sheet. (3) The apparent resistance is higher than that of the sintered disk, because of the imperfect contact in the sheet. (4) The thickness is about 1.1 mm and it is thinner than commercial thin PTC plates.

Composite materials containing thin Shape Memory Alloy (SMA) wires show great promise as materials able to adapt their shape, thermal behaviour or vibrational properties during service. Tools for designing such materials are however far from being available. The work presented here reports the main achievements of a concerted European effort towards the establishment of a fundamental understanding for manufacturing and design of SMA composites. The following major steps are examined: selection and characterisation of the material constituents, development of manufacturing processes for the production of composites with pre-strained SMA wires, analysis and modelling of the action of the SMA wires in the composite, the contribution of the SMA-resin interface, analysis and modelling of the functional, thermomechanical, impact and durability properties of SMA composites and the development of a simple, large-scale, aerodynamic model. It is argued that the achievements of this research have brought the knowledge on SMA composites to a substantially higher level enabling reliable manufacturing and design, and the emergence of new industrial applications.

This paper reports the fabrication method and mechanical properties of a shape memory alloy (SMA) fiber/plaster smart composite, which can be used in architectural and civil engineering applications. Fe–Mn–Si–Cr SMA fibers are subjected to pretensile strain at room temperature, and are embedded into plaster matrix. The Fe–Mn–Si–Cr SMA fiber/plaster composites are then heated up to 250°C (above A_s) to generate a compressive residual stress in the matrix. Three-point bending test is performed for the mechanical property characterization. Fiber pull-out test is also conducted to evaluate the bonding strength at interface between SMA fiber and plaster matrix. Finite element analyses are carried out to have some further insights on the experimental results. It is found that the bending strength of the composites enhances with increasing level of pretensile strain. By using the inexpensive Fe–Mn–Si–Cr SMA fibers for the reinforcement of the SMA composite, one can obtain materials for practical engineering applications at low cost.

Material design has recently become one of the key topics in the development of smart adaptive composites. In particular, different material constituents of the hybrid polymer composites with embedded Shape Memory Alloy elements (SMA composites) have to be combined and positioned in such a way that predetermined functional properties are obtained. Due to the complexity arising from the inherently non-linear and hysteretic thermomechanical response of SMA elements, modelling of the functional behaviour of SMA composites has become an indispensable part of the SMA composite technology. In this paper, design of SMA polymer composites using a recently developed SMA composite model is demonstrated. The simulations carried out in a preliminary stage of the smart composite design help to find optimal material parameters of the SMA wires (Young’s modulus and coefficient of thermal expansion of austenite and martensite, transformation temperature, strain, hysteresis, entropy, etc.) and of the polymer matrix (longitudinal Young’s modulus, coefficient of thermal expansion), as well as optimal composite fabrication parameters (layout of the composite, volume fraction of wires, prestrain given to the SMA wires when hybridising it with the matrix).

SiC/SiC composites with a thermal coating have been developed as smart composites for high temperature applications, where the joint techniques are indispensable for constructing large scale structures due to the fabrication and economical reasons. Since one of the important issues is the establishment of a design database for the joint strength, the testing method has to be carefully selected based on the theoretical background. In this study, the stress distribution of a SiC/SiC composite specimen containing a butt joint consisting of reaction-formed silicon carbide tested by the asymmetrical four-point bending test was precisely analyzed by the finite element method as a means to evaluate the applicability of analytical results. In the case without the effect of the thermal residual stresses, the shear stress distribution at the interface between the base and the joint almost agreed with the analytical theory. For the case with the residual stress, however, the shear stress near the surface was very large and the possibility of an initial crack induced by the residual stress was considered. Moreover, it was found that the residual shear stress distribution near the surface was significantly affected by the joint thickness.

The superelastic deformation of an adaptive layer composite containing a shape memory alloy as an active component is considered in this paper. It is shown that the intrinsic instability of superelastic deformation can be suppressed in such kind of composite. The stability analysis of superelastic deformation allows one to formulate design principles of adaptive composites with controlled stress-strain hysteresis. The analysis of composite with sufficiently different elastic moduli of passive and active layers is a necessary step for using adaptive composites in applications which require a large reversible superelastic deformation.

This paper describes the manufacture and testing of polymer matrix Terfenol-D particulate composites fabricated with a [112] preferred crystal orientation. The results demonstrate that crystal orientation permits higher saturation strain than previously obtained in particulate composites. Oriented particle composites were fabricated using needle shaped Terfenol-D particulate made from commercially available [112] textured polycrystals. A magnetic field applied during manufacture oriented the particles along their longest dimension. Three materials of 22, 36, and 49% particulate volume fraction, were produced and tested to obtain the magnetostriction at constant mechanical loads from 0.5 to 16 MPa. Results demonstrate that crystallographically oriented particle composites saturate at 1475 microstrain whereas the previous non-oriented particle composite saturates at 1150 microstrain. The authors attribute the increased magnetostriction to a predominance of particle orientation along the [112] crystal direction. Modulus measurements indicate that the composites can be described using an upper bound 1–3 composite model. Upper bound rule of mixtures models show that the 36% and 49% oriented particle composites exhibit 94% and 90% respectively of the predicted upper bound magnetostriction. Further rule of mixtures models are used to illustrate how applied stress increases the effective anisotropy in composites as a function of particulate volume fraction. The results demonstrate that by preferential orientation of particles, Terfenol-D particulate composite materials may be fabricated with magnetostriction properties close to that of bulk Terfenol-D.

In this paper the design and the manufacture of a 3-dof (degrees of freedom) robot driven by shape memory alloys (SMA) is presented. This robot has a parallel structure including a fixed plate and a moving plate. The plates are linked together by 3 SMA wires and a mechanical spring is located in the central part. Possible applications are the control devices to orient a mirror, a sample under a microscope or to orient the head of a micro snake like robot. The paper explains the kinematic model, the mechanical design and the control system of the robot. The feedback signals of the closed loop control system are the displacements of the SMA joints located on the moving plate, measured by three conductive potentiometers. The control system is P.C. based. The SMA actuators are driven by Nitinol wires of a diameter of 0.15 mm. The robot takes up a cylinder with a diameter of 100 mm and a height of 180 mm. A prototype of the robot has been manufactured and some experimental tests were carried out. These tests are carried out both using a simple test bed made by a SMA wire and a pulley, and using the prototype itself. The step response of a single SMA wire and the trajectory control to describe a circle in the prototype are also shown as validation tests of the robot. The results of the experimental validation show the feasibility of this design, but particular attention has to be paid to the machining and to the assembly.

Two types of multilayer actuators based on a shape memory alloy (SMA) film as an active component are explored theoretically. One of them uses the bending of an actuator due to the movement of an austenite/martensite interface of a SMA film parallel to a film plane. Another type of actuation uses the combination of passive layers with different coefficients of thermal expansion for engineering curvature. It is shown that in both cases the actuating deformation of the multilayer actuators can be optimized by the combination of the layers with different elastic properties, misfits and thickness.

A three-dimensional (3D) model has been developed for simulation of the physical properties of an electrically driven shape memory alloy (SMA) microactuator used for control of a microvalve. The mechanical behavior is decribed by a two phase macromodel taking into account material and geometrical nonlinearity. The simulation makes use of mechanical, electrical and thermal finite element programs, which are coupled by a program for management of the simulation sequence as well as the exchange of data between the different finite element programs. Mechanical tests confirm the simulated forces and displacements. From the pressure-dependent gas flow in the valve an effective heat transfer coefficient is determined to simulate the convective heat exchange. The simulated heat transfer times are in quantitative agreement with experimentally determined time constants of the microvalve.

Motivated by many experimental efforts to develop suitable shape-memory micropumps, we propose a multiscale framework to study the behavior of pressurized films. We use recoverable deflection as a measure to design large stroke micropumps and develop a model to estimate it. We show that the recoverable deflection of a polycrystalline shape-memory film depends on the transformation strain of the underlying martensitic transformation, the texture and especially on the size effects. We find that flat grains are preferable to long grains in columnar films concerning the purpose of large recoverable strain. We also show that common sputtering texture is not ideal for recoverable deflection in both Ti–Ni and Cu-based shape-memory films. It turns out that {100} Cu-based films may have better behavior than Ti–Ni films. We conclude with comparison with experiment.

Porous titanium-nickel shape memory alloys (TiNi SMA) can be fabricated using special engineering technique. This material is a whole porous material with interconnected pores. This porous TiNi SMA retains the unique properties of solid TiNi SMA. Its porosity and pore size can be controlled. Its application to orthopaedic field is very expected especially in bone substitute and bone implant and so on. The purpose of this study was to evaluate bone tissue response and histocompatibility of porous TiNi SMA in vivo. Thirty block implants (5 mm×5 mm×7 mm) of porous TiNi SMA were prepared. Analysis of pore structure of the implant was performed using Hg-porosimetry and scanning electron microscope. Fifteen New Zealand white rabbits were used. Sterile porous TiNi SMA implant was implanted in the defects of proximal tibia metaphysis. Limbs of five rabbits were harvested respectively at 2, 4 and 6 weeks post implantation. Each specimen was embedded in PMMA. Embedded specimen was sectioned into 300 \\micron thickness with isomet-diamond saw. Quantitative histomorphometric analysis was performed within the each implant. The pore sizes of porous TiNi SMA were 323±89 \\micron. Porosity was 55.3±6.7%. No apparent adverse reactions such as inflammation and foreign body reaction were noted on or around all implanted porous TiNi SMA blocks. Bone ingrowth was found in the pore space of all implanted blocks. The percent bone ingrowth into the pore space of porous TiNi SMA increased over time. At six week post-implantation, bone ingrowth into pore in TiNi SMA block was very excellent (at 6 week, 78.3±9.7%). This percent bone ingrowth was much higher than that of other porous materials. This in vivo response of porous TiNi SMA observed in this study opens to the possibility that porous TiNi SMA could be used as an ideal bone substitute.

Titanium–Nickel shape memory alloy (TiNi SMA) has great potential as a biomaterial in orthopaedic applications due to its unique thermal shape memory effects, superelasticity and high damping properties. We designed and manufactured bone fixaters using newly developed TiNi SMA wire (Af, 35±2°C). Two bone fixater designs (single and double ring) were prepared for the treatment of bone fracture in twenty patients (6 distal femur, 5 distal fibular, 4 distal tibia, 2 metacarpal bone, 2 periprosthetic fracture and 1 subtrochanter of femur). Serial radiographs, complete blood count (CBC) and urine analysis were performed postoperatively. Radiological union was achieved without complications in approximate eight weeks after operation. There were no abnormal findings on follow-up CBC or urine analysis. On a subjective level, use and application of the TiNi SMA fixater was not as demanding as conventional fixation methods, such as cerclage or the Dall-Miles technique. The efficacy of SMA bone fixater in this study is very excellent as demonstrated in this clinical study. It gives the new armament to orthopedic surgeon.

An application of shape memory alloys (SMAs) for artificial anal sphincters is presented. The artificial anal sphincter consists of two all-round shape memory alloy (ARSMA) plates as the main functional parts, and heaters attached on SMA plates for generating thermal cycles required for phase transformation accompanied shape changes of the plates. The SMA artificial sphincter could be fitted around intestines, performing an occlusion function at body temperature and a release function upon heating. For reducing the potential of infection, a transcutaneous energy transmission (TET) system is incorporated into the artificial anal sphincter, facilitating the complete implantation of the device. Investigation on the thermomechanical responses of the artificial sphincter has been conducted, with both the in vitro and in vivo experiments, showing great potential of practical uses. The relation between the values of applied power and the response times, and the thermocompatibility of the device are discussed.

Steady-state and laser-induced transient surface plasmon bands of copper nanoparticle composites, fabricated by ion implantation, were studied by optical measurements. Negative ion implantation has been applied to generate the Cu nanoparticles with a narrow distribution in amorphous SiO₂, MgO2.4(Al₂O₃) and LiNbO₃ with various refractive indices. The Cu nanoparticles were embedded within a depth of 100 nm by implantation of 60 keV Cu⁻. The surface plasmon band in steady-state absorption spectra resulted from formation of nanoparticles in the various substrates and shifted to red with increasing refractive index of the matrix. Transient absorption was measured with the technique of pump-probe femtosecond spectroscopy. The transient bleaching band also shifted in parallel with the steady-state plasmon resonance. The bleaching recovered in several picoseconds due to energy transfer from the excited electron system to the phonon system via the electron-phonon interaction. The electron-phonon coupling constant, g, of Cu nanoparticles in amorphous SiO₂ was obtained to be a value of 2.4×10¹⁶ W/m³K.

To realize channel cross-connecting in optical communications systems, a high speed optical matrix switch was fabricated using z-cut LiNbO₃. Four 2×2 directional couplers were integrated on one substrate for construction of a 4×4 switch. Single-mode optical waveguides were formed by Ti-diffusion at a wet O₂ atmosphere. Ti-diffusion profile, refractive index variation and waveguide morphology were analyzed by SIMS, Prism coupler and SEM, respectively. The optical properties of the fabricated switch was measured in terms of insertion loss, cross-talk, spectral flatness and switching speed.

Characteristics of thin-film NTC infrared sensors fabricated by micromachining technology were studied as a function of the thickness of membrane. The overall-structure of thermal sensor has a form of Au/Ti/NTC/SiO_x/(100)Si. NTC film of Mn_1.5CoNi_0.5O₄ with 0.5 \\micron in thickness was deposited on SiO_x layer (1.2 \\micron) by PLD (pulsed laser deposition) and annealed at 600–800°C in air for 1 h. Au (200 nm)/Ti (100 nm) electrode was coated on NTC film by dc sputtering. By the results of microstructure, X-ray and NTC analysis, post-annealed NTC films at 700°C for 1 h showed the best characteristics as NTC thermal sensing film. In order to reduce the thermal mass and thermal time constant of sensor, the sensing element was built-up on a thin membrane with the thickness of 20–65 \\micron. Sensors with thin sensing membrane showed the good detecting characteristics.

The activated carbon sphere containing zinc oxide was prepared by carbonizing a zinc ion-exchange resin at different temperatures in nitrogen gas. Zinc oxide of hexagonal type was detected in all carbon samples, the amount of which decreased with an increase in the carbonization temperature. However, the specific surface areas of carbon samples increased with increasing temperature of the resin. The antibacterial activity on their carbon samples was studied without the presence of light. The antibacterial activity on carbon samples containing zinc oxide increased with the amount of zinc oxide in the carbon samples. The antibacterial activity for Staphylococcus aureus was stronger than that for Escherichia coli. By an oxygen electrode analysis, it is shown that hydrogen peroxide was generated on the carbon samples. The concentration of hydrogen peroxide increased with increasing carbonization temperature of the resin. The antibacterial activity is found to be caused by the generation of hydrogen peroxide from zinc oxide dispersed in activated carbon sphere.

Regenerated cellulose and chitosan membranes were studied for the pervaporation separation of an aqueous solution of urine component (ammonia, uric acid or creatinine). The permeation rate of water increased with increase of the temperature of feed solution induced into the upstream side of membrane module. Uric acid, creatine and creatinine were not found in the permeate through the all membranes investigated. Selective permeation of water and ammonia depends on membrane. The removal of ammonia through the chitosan membrane was from 57% to 59%. Adsorption of ammonia from the downstream vapor by silica gels was carried out. And desorption of ammonia from the adsorbents by heating under the reduced pressure to regenerate the capacity of adsorption was also confirmed. In the case of new pervaporation system, the combination of pervaporation and adsorption/desorption process, ammonia was almost completely removed, and finally the pure condensed water was obtained in the cold trap.

X-ray diffraction measurements on the Cr–H system were made using synchrotron radiation at high hydrogen pressures and high temperatures, and the phase diagram was determined up to p(H₂)=5.5 GPa and T\\lesssim1400 K. Three solid phases were found to exist; a bcc phase (α) of low hydrogen concentrations, x=[H]⁄[Cr]\\lesssim0.03 existing at low hydrogen pressures (\\lesssim4.4 GPa), and two high-pressure phases, an hcp (ε) phase at lower temperatures and an fcc (γ) phase at higher temperatures, both having high hydrogen concentrations x∼1. A drastic reduction of the melting point is caused by dissolution of hydrogen. A gradual lattice contraction observed in the fcc phase indicates the formation of superabundant Cr-atom vacancies (vacancy-hydrogen clusters). Thermal desorption measurements after recovery from high p(H₂), T treatments revealed several desorption stages including those due to the release from vacancy-hydrogen clusters and from hydrogen-gas bubbles, and allowed determination of relevant trapping energies.

Structural changes and the hydrogen content in pseudobinary Ti₃Al–Zr₃Al, i.e., Ti_1−x−yAl_yZr_x alloys (y=0.25, 0.30 and 0.35) by hydrogenation were examined with a powder X-ray diffractometer (XRD) and a hydrogen analyzer. Single-phases with the D0₁₉ and the B2 structures were formed in the homogenized alloys, while an fcc phase coexisted with the B2 phase. These D0₁₉, B2 and fcc phases changed into an amorphous phase by hydrogenation below 573 K. This is the first report of hydrogen-induced amorphization (HIA) by hydrogenation of a single-phase bcc type (B2) alloy. Furthermore, crystallization behavior of the hydrogen-induced amorphous Ti–Al–Zr alloys was investigated by a differential scanning calorimeter (DSC) and XRD.

The pressure dependence of structural changes of C15 Laves phase TbFe₂ heated in a hydrogen atmosphere was investigated using a pressure differential scanning calorimeter (PDSC), a powder X-ray diffractometer (XRD), a differential scanning calorimeter (DSC) and a hydrogen analyzer. Hydrogen absorption in the crystalline state, hydrogen-induced amorphization (HIA), precipitation of TbH₂ and decomposition of the remaining amorphous alloy into α-Fe and TbH₂ occurred exothermally with increasing temperature above 0.5 MPa H₂, independent of the heating rate. HIA and precipitation of TbH₂ occurred simultaneously in 0.2 MPa H₂ (0.33 K/s) and in 0.1 MPa H₂ (0.17 K/s). On the contrary, hydrogen absorbed crystalline c-TbFe₂H_3.5 decomposed directly into α-Fe and TbH₂ in 0.1 MPa H₂ (0.33 K/s). That is, no amorphous phase was formed at the lower hydrogen pressure and at a high heating rate. The reason why HIA occurs above the critical hydrogen pressure and below the critical heating rate was discussed on the basis of the hydrogen content absorbed in the crystalline state of TbFe₂.

The hydrogenation and dehydrogenation behavior of LaNi₅, LaNi_4.75Co_0.25 and LaNi₃Co₂ was studied by the pressure differential scanning calorimetry (PDSC) at the hydrogen pressure range of 1 to 5 MPa in the temperature range from 323 to 473 K with the heating and cooling rates of 2 to 30 K min⁻¹. In the heating run of the hydride of LaNi₅, two endothermic peaks were observed. One was the peak for the transformation from the γ phase (full hydride LaNi₅H₆) to the β phase (hydride LaNi₅H₃). The other was the peak for the transformation from the β phase to the α phase (solid solution). In the cooling run, one exothermic peak for the transformation from the α phase to the γ phase was observed. These endothermic and exothermic peaks shifted to higher temperatures with the increase in hydrogen pressure. In the heating and cooling runs of the LaNi_4.75Co_0.25–H₂ system the PDSC curves similar to those of the LaNi₅–H₂ system were observed. However, in the heating run of the hydride of LaNi₃Co₂ only one endothermic peak was observed. Using Ozawa’s method, the activation energies for dehydrogenation of the hydrides were estimated. The activation energy for the γ-β transformation was higher than that for the β-α transformation. Substitution of cobalt for a part of nickel in LaNi₅ increased the activation energies for the phase transformations.

Pure Mg₂NiH₄ was first prepared by Hydriding Combustion Synthesis (HCS), under which pressure and temperature were sensitively controlled. The purpose of this paper was to study “Pressure-Composition-Temperature (PCT)” properties of the HCSed product, in which hydrogen storage capacity, hydriding reaction rate, and PCT curves were mainly examined, in comparison to the commercially available product, i.e., Ingot Metallurgy product. The results showed that the HCSed Mg₂NiH₄ has 3.6 mass%, same as theoretical value, in hydrogen storage capacity in the first cycle just after synthesized without activation treatment, and very large reaction rate; only five minutes for full charge. Plateau pressure of the HCSed Mg₂NiH₄ in PCT curves also became relatively lower than the conventional one. One of the most interesting results is the high activity of our product: it stored hydrogen even at room temperature. These results appealed a possibility of new productive process of hydrogen storage alloy from the viewpoint of product property.

Hydrogenation of MgNi₂ by atomic hydrogen was examined at elevated temperatures. Mg₂Ni powder was compacted into a disk and heated in vacuum at 773 K for outgassing. During this heat treatment, Mg₂Ni was transformed into MgNi₂ by evaporation of Mg. This specimen was hydrogenated at 573, 673 and 773 K by hydrogen gas of 30 Pa and atoms produced by rf-discharge. No significant hydrogenation took place by hydrogen gas. On the other hand, the specimen was hydrogenated by the exposure to atoms at 773 K up to [H]⁄[M]=0.14 within 25 ks. New peaks appeared in the diffraction pattern of Cu Kα X-rays at 2θ=34.0, 42.0, 48.9 and 86.6^°. This observation indicated that the hydride formed was in a cubic structure.

Hydrogen behavior in V–Zr–Ni–Ti alloys with different amount of Laves phase has been examined by the tritium radioluminography and the PCT measurement. Tritium radioluminography has shown that the local hydrogen concentration in the Laves phase is higher than that in the BCC phase in the range of lower hydrogen content of ppm order, that is, hydrogen preferentially dissolves into the Laves phase. PCT measurement has shown that the maximum hydrogen concentration decreases as the ratio of the Laves phase to the BCC phase increases in the range of higher hydrogen content of percent order. This means that the BCC phase mainly absorbes hydrogen. It has been concluded that the Laves phase acts as the penetration path for hydrogen in the early stage of hydrogenation, and that the BCC phase is the main hydrogen storaging phase. An extraordinary phenomenon has been observed in the time dependence of the local hydrogen concentration in V_54.5Zr_18.25Ti_11.25Ni₁₆ alloy. The local hydrogen concentration in the BCC rich region decreases for a few days after hydrogen charging and then changes to increase, while the hydrogen concentration in the Laves phase rich region gradually decreases. This phenomenon has been thought to be related to the microstructure in the specimen.