Search

Internal friction of a Mn-11Cu (at.%) alloy was studied in the main [001], [011] and [111] crystallographic directions by applying tensile stress. Two internal friction peaks were observed in three directions. One is twin boundary-related relaxation peak (in the temperature range of 200–300 K), which is highest in the [001] direction and lowest in the [111] direction. Other is associated with reverse martensitic transformation (~430 K), and its orientation dependence is small. The orientation dependence of the twin boundary-related internal friction peak is discussed based on the Schmid factor for the movement of {011} twin boundaries.

We were successful in controlling the arrangement of Ti3Ni4 in a single-crystalline Ti-50.8Ni (at.%) alloy. Only three variants of Ti3Ni4 form when tensile stress is applied in the [111] direction during aging heat-treatment while four variants form when the stress is absent. The lack of one variant of the precipitate causes different arrangement of the R-phase variants. This results in obvious decrease in the internal friction in the R-phase state. An extremely high internal friction of 0.25 was also found in the R-phase state formed in the specimen including four variants of Ti3Ni4.

In this paper, we report how the charged hydrogen in a Ti-50.8Ni (at. pct) alloy acts during stress-induced martensitic transformation and aging. Hydrogen is preferentially trapped in the B19′ martensite phase near the surface of the specimen. The hydride formed by hydrogenation dissociates during the stress-induced martensitic transformation. The hydrogen diffuses into the interior of the specimen through aging in air, resulting in suppression on the stress-induced martensitic transformation.

In this paper, we report that the superelasticity and the elastocaloric effect of Ti-50.8Ni (at.%) alloy are drastically improved by a combination of precipitation and grain refinement strengthening. The process is as follows: first, aging at 773 K after solution treatment to form coherent fine precipitate of Ti3Ni4; second, cold working followed by annealing at 673 K to form nanocrystalline grain containing Ti3Ni4. The decay of the superelasticity and the elastocaloric effect by 100 stress cycles is

In the region of a martensitic transformation, conventional elastocaloric effect causes cooling when the forward parent to martensite transition is induced by loading under adiabatic conditions and by heating when the load is removed and the reverse transition occurs. An inverse elastocaloric effect might be expected if the reverse transition could be induced upon loading. However, this behavior is quite uncommon and inverse elastocaloric effect rarely occurs. Herein, we have successfully manufactured a Ti-50.8 at.% Ni crystal in a high proportion of favorable $R$ phase (martensite) variant containing single-variant ${\mathrm{Ti}}_{3}{\mathrm{Ni}}_{4}$ nanoprecipitates aligned along the ${\ensuremath{\langle}111\ensuremath{\rangle}}_{B2}$ direction of the $B2$ parent phase, by means of a compressive stress associated aging along the ${\ensuremath{\langle}111\ensuremath{\rangle}}_{B2}$ direction. The material displays inverse elastocaloric effect induced by compressive stress along the ${[111]}_{B2}$ direction in the temperature range of stability of the $R$ phase. This inverse elastocaloric effect is caused by the stress-induced reverse transformation of the high proportion of favorable $R$-phase variant that occurs when the external compressive stress is applied along the direction of the internal tensile stress induced by the aligned ${\mathrm{Ti}}_{3}{\mathrm{Ni}}_{4}$ nanoprecipitates. The mechanism giving rise to this inverse elastocaloric effect is explained by the transformation strain in the ${[111]}_{B2}$ and ${[1\overline{1}0]}_{B2}$ directions for different $R$-phase variants based on crystallography. Results indicate that the stress-associated aging process is an effective way of patterning materials to display inverse elastocaloric effect through controlling the alignment of the ${\mathrm{Ti}}_{3}{\mathrm{Ni}}_{4}$ nanoprecipitate in the Ni-rich Ti-Ni shape memory alloy.

In this paper, we propose a novel approach to fabricate fluidic chips. The method utilizes molecular cross-diffusion, induced by photopolymerization under ultraviolet (UV) irradiation in a channel pattern, to form the channel structures. During channel structure formation, the photopolymer layer still contains many uncured molecules. Subsequently, a top substrate is attached to the channel structure under adequate pressure, and the entire chip is homogenously irradiated by UV light. Immediately thereafter, a sufficiently sealed fluidic chip is formed. Using this fabrication process, the channel pattern of a chip can be designed quickly by a computer as binary images, and practical chips can be produced on demand at a benchtop, instead of awaiting production in specialized factories.

Significant grain growth assisted by a martensitic transformation near or below room temperature was found in a nanocrystalline Ti-50.8Ni (at.%) alloy. Significant grain growth was observed when thermally- or stress-induced B2 ↔ R ↔ B19′ (B2: parent phase, R and B19′: two types of martensite phases) martensitic transformation is repeated while not when B2 ↔ R transformation is repeated. This implies that martensitic transformation to and from the B19′ martensite phase assist the grain growth. Transmission electron microscope observations indicate that migration of grain boundaries could be the main mechanism for the grain growth.

The size effect on the mechanical behaviors of single crystalline Fe-31.2Pd (at.%) micropillars with four different pillar diameters (approximately 2 μm, 1 μm, 500 nm and 400 nm) were studied by compressing in the [001] direction. Both the Young's modulus and yield stress increase with the decrease of pillar diameter. The main defects in the plastically deformed pillars are the {111} deformation twins, which could be covered by high density of dislocations after further plastic deformation in a 400 nm pillar. A repeatable elastic-like strain ~4% was observed for more than 40,000 cycles.

Depending on Pd content, disordered Fe-Pd alloys take various crystal structures: body-centered cubic (BCC), body-centered tetragonal (BCT), face-centered tetragonal (FCT) and face-centered cubic (FCC) structures. We have measured specific heat of Fe-Pd alloys to understand the lattice stability and electronic state of the alloys. At the phase boundary between FCT and FCC phases, electronic specific heat coefficient shows a local maximum and the Debye temperature shows a local minimum. A local maximum of electronic specific heat coefficient also appears in the BCC region with Pd content of near 10 at.%.

Elastocaloric effects in shape memory alloys are usually caused by latent heat associated with the stress-induced martensitic transformation. We report here that the rubber-like behavior of the R-phase in nanocrystalline Ti-50.8Ni (at.%) wires, realized by an aging heat treatment under tensile stress, also contributes a significant elastocaloric effect. The efficiency of the elastocaloric effect caused by the rubber-like behavior is higher than that caused by the stress-induced B2 ↔ B19′ transformations because of the small energy dissipation. The usage of the rubber-like behavior is a method to extend the temperature window of the elastocaloric effect in shape memory alloys.

We have investigated the influence of thermal cycles on martensitic transformation of a Co2Cr(Ga,Si) ferromagnetic Heusler alloy. The as-quenched specimen exhibits successive L21(L)–D022–L21(H) martensitic transformation in the cooling process, which is known as reentrant martensitic transformation. However, heating to 800 K (527 °C) for reverse D022–L21 transformation with a rate of 10 K/min (10 °C/min) stabilizes the parent phase, meaning that the martensitic transformation is suppressed by the thermal cycles. We found precipitate after thermal cycles, and it will be the reason for the stabilization of parent phase.

We have detected an inverse elastocaloric effect in a Ti-51Ni (at%) alloy which contains aligned coherent particles of Ti3Ni4 formed by aging under tensile stress. In this alloy, the temperature of the specimen decreases when stress is applied due to a stress-induced reverse martensitic transformation (from the R-phase to the B2-phase) The temperature decrease caused by this effect is about 1.1 K at 300 K. The internal stress field around the Ti3Ni4 precipitate is responsible for this effect. A further increase in the tensile stress causes a temperature rise due to a conventional elastocaloric effect.

We have determined the easy axis of magnetization of Fe3C (cementite) using a powder specimen prepared by an electrolytic extraction method. The easy axis is determined to be the c-axis (the shortest axis) of the orthorhombic structure with the space group Pnma. The result is different from the one previously reported by neutron diffraction (b-axis). The magnetic energy difference between the easy axis and the average of other two axes is approximately 400 kJ/m3 at 5 K.

Iron-rich Fe–Al binary phase diagram in the composition range of 23–26 at% Al shows a complex structure partly due to magnetic properties of the disordered A2-phase, the ordered B2-phase and the ordered DO3-phase. In this study, the disorder-order transformations between the constituent phases have been investigated through magnetization measurements of Fe-xAl alloys (x = 22.0, 24.0, 24.3, 24.7, 26.5 at%). It is found that spontaneous magnetization of the B2-phase is larger than that of the DO3-phase and smaller than that of the A2-phase. The former relation disagree with previous reports but was further confirmed as the decrease of the B2-DO3 transformation temperature under magnetic field (−1.58 K/T). It is also found that the Curie temperature of the B2-phase (836 K for Fe-24.7 at%Al alloy) coincides with that of the A2-phase; furthermore, these temperatures agree with the phase boundary between B2-region and (A2+B2)-region.

We have examined the transformation strain, Δe, in the [0 0 1] direction of an Fe-31.2Pd (at.%) shape memory alloy under compressive stress applied in the same direction. When the stress is absent, the alloy exhibits a cubic to tetragonal martensitic transformation at 230 K with |Δe| of 1.4%. As the stress increases, the transformation temperature increases linearly and |Δe| decreases linearly and vanishes at 40 MPa (280 K). This point is the critical point of this transformation at which the first-order nature disappears, and the critical exponent β is evaluated to be 0.47 ± 0.04.

Fe 3 Pt with a degree of order of S = 0.75 shows a second-order-like martensitic transformation from a cubic structure to a tetragonal one at approximately 90 K. We found by neutron diffraction that this alloy exhibits a large elastic strain of more than 6% by the application of compressive stress in the [001] direction. A large elastic anisotropy caused by lattice softening in elastic constant C’ , which occurs on approaching the transformation temperature, was found to be responsible for the large elastic strain of Fe 3 Pt.

In this Rapid Communication we report a phenomenon in a disordered interacting electron system. The measurements of structural, magnetic, and transport properties of $\mathrm{FeA}{\mathrm{l}}_{2\ensuremath{-}x}\mathrm{G}{\mathrm{a}}_{x}$ ($0\ensuremath{\le}x\ensuremath{\le}0.5$) show that antiferromagnetic transition in these intermetallic compounds occurs concomitantly with a disorder-induced weak localization of electrons; the temperatures ${T}_{N}$ and ${T}_{m}$, at which antiferromagnetic transition and the weak localization respectively occur, closely track each other as the Ga concentration is varied. The antiferromagnetic transition is confirmed from the magnetic and specific heat measurements, and the occurrence of weak localization is confirmed from the temperature variation of resistivity and magnetoresistance measurements. With increasing Ga concentration, substitutional disorder in the system increases, and the consequent disorder-enhanced magnetic exchange interaction and disorder-induced fluctuations simultaneously drive antiferromagnetic transition and weak localization, respectively, to higher temperatures.

We have investigated the relation between the softening of elastic constants and martensitic transformation in Fe3Pt, which exhibits various kinds of martensitic transformation depending on its long-range order parameter S. The martensite phases of the examined alloys are BCT (S = 0.57), FCT1 (S = 0.75, c/a < 1) and FCT2 (S = 0.88, c/a > 1). The elastic constants C′ and C44 of these alloys decrease almost linearly with decreasing temperature. Although the temperature coefficient of C′ decreases as S increases, C′ at the transformation temperature is the smallest in the alloy with S = 0.75, which transforms to FCT1. This result implies that softening is most strongly related to the formation of the FCT1 martensite with tetragonality c/a < 1 among the three martensites.

Diffuse satellites appearing in electron diffraction pattern of shape memory Ti–(50 − x )Pd– x Fe (14, 16, 18, 19 and 20, in at.%) alloys have been investigated. The satellites appear in each alloy below T min , where its electrical resistivity shows a local minimum. The positions of satellites are g B2 + 〈 ζ ζ ¯ 0〉 * , where g B2 is a reciprocal lattice vector of the B2-phase. The value of ζ is smaller than 1/5 at T min for all the alloys; it increases with decreasing temperature and decreases with increasing iron content. The value of ζ at T min agrees with the length of the nesting vector previously calculated by the present authors. This result implies that Fermi surface nesting is the origin of diffuse satellites in Ti–(50 − x )Pd– x Fe alloys.

Magnetic behavior of equiatomic FeAl alloy is still not satisfactorily understood. In this work, we studied the magnetic properties of disordered FeAl alloy both experimentally and using first-principles theories and revisited the alloy system in perspective of the inhomogeneity present in the system. After obtaining magnetic exchange interactions from first-principles theories, we carried out Monte-Carlo simulations on special quasi-random structures (SQS) and compared the results with experimental measurements. We tried to understand the plethora of often differing results and explain them in terms of possible inhomogeneities in the system.

Submicron-size of colloidal spheres itself and their self-assembled structures are useful basis for obtaining various functional structures. Up to now, we have succeeded to fabricate several unique structures that are expected to be applied to various components of a specific optical device, or the substrate for material- and/or bio-science research purpose. Facile fabrication procedures and obtained structures for an inverse opal structure, a binary colloidal structure, an alternating photonic crystal and glass structure, and an anisotropic hemispherical structure are introduced. Their optical functions are also described.

An Fe–55Pd (at.%) alloy exhibits a disorder–order transformation from the A1-type disordered cubic phase to the L10-type ordered tetragonal phase. We have investigated the influence of the magnetocrystalline anisotropy energy on the selective formation of a variant of the ordered phase during heat treatment. We found that the strength of the magnetic field required to produce a single variant decreases from approximately 4 T to approximately 0.4 T when the ordering heat-treatment temperature is decreased from 673 to 603 K. The magnetocrystalline anisotropy energy was evaluated at these temperatures as a function of magnetic field. The nucleation ratio between different variants was evaluated through the magnetocrystalline anisotropy. It is revealed that a single variant state is obtained during the growth process when the nucleation ratio of the dominant variant exceeds approximately 1.1.

We have studied the structure change of an ordered Fe3Pt (degree of order ∼0.75) under a compressive stress applied in the [001] direction by neutron diffraction. In the absence of the stress, the alloy exhibits a weak first order martensitic transformation at 90 K from the L12-type cubic structure to the L60-type tetragonal structure. Under the compressive stress of 100 MPa, the first order nature of the thermally-induced martensitic transformation was undetectable in the temperature range of between 70 K and 270 K. The first order nature of the stress-induced martensitic transformation was also undetectable in the stress range of between 6 MPa and 300 MPa when tested at 120 K and higher temperatures. Under these conditions, the lattice parameters change continuously both in the cooling process and in the stress-applying process. Despite the disappearance of the first order nature of martensitic transformation, a significant stress-induced softening of lattice, which is regarded as a precursor phenomenon of martensitic transformation, was observed between 120 K and 265 K but not at 93 K and 295 K.We have studied the structure change of an ordered Fe3Pt (degree of order ∼0.75) under a compressive stress applied in the [001] direction by neutron diffraction. In the absence of the stress, the alloy exhibits a weak first order martensitic transformation at 90 K from the L12-type cubic structure to the L60-type tetragonal structure. Under the compressive stress of 100 MPa, the first order nature of the thermally-induced martensitic transformation was undetectable in the temperature range of between 70 K and 270 K. The first order nature of the stress-induced martensitic transformation was also undetectable in the stress range of between 6 MPa and 300 MPa when tested at 120 K and higher temperatures. Under these conditions, the lattice parameters change continuously both in the cooling process and in the stress-applying process. Despite the disappearance of the first order nature of martensitic transformation, a significant stress-induced softening of lattice, which is regarded as a precursor phenomenon...

Magnetocrystalline anisotropy and magnetic field-induced strain were investigated for three tetragonal martensite phases formed in Fe 3 Pt ferromagnetic shape memory alloys. The phases examined are the body-centered tetragonal (bct, c / a S = 0.57, the face-centered tetragonal (fct, c / a S = 0.75 and the fct ( c / a > 1) martensite with S = 0.88. It was found that the c -axis is the easy axis of magnetization for the two fct martensites and the hard axis of magnetization for the bct martensite. The uniaxial magnetocrystalline anisotropy constant, K u , is nearly proportional to |1 − c / a | for the two fct martensites, but the relation is not satisfied for the bct phase. When a magnetic field is applied in the [0 0 1] direction, a large magnetic field-induced strain appears in the two fct martensites but not in the bct martensite. The strain in the field direction is a contraction in the fct martensite with c / a 1, whereas it is an expansion in the fct martensite with c / a > 1. These behaviors are explained by evaluating the magnetic shear stress, which is given by K u divided by the twinning shear.

We have investigated the relation between the negative temperature coefficient (NTC) in electrical resistivity and the formation of the athermal ω phase in Ti–xNb (26 ≤ x ≤ 29 at.%) alloys. The alloys with 26 ≤ x ≤ 28 exhibit NTC in electrical resistivity and a temperature hysteresis appears in the temperature region of NTC. In these alloys, athermal ω phase forms in the temperature range of NTC in resistivity. The alloy with x = 29 also shows NTC but no detectable temperature hysteresis appears. In this alloy, the athermal ω phase does not form, but diffuse satellites appear at g β + 1 / 2 ζ ζ ¯ 0 ∗ , where gβ is a reciprocal lattice vector of the β phase in the temperature range of NTC. These results suggest that NTC in the electrical resistivity in Ti–xNb alloys is related not only to the formation of the athermal ω phase but also to the appearance of diffuse satellites at g β + 1 / 2 ζ ζ ¯ 0 ∗ .

Time dependence of the first order magnetostructural transition in Pd doped B2-type FeRh (Fe0.45Rh0.45Pd0.1 alloy) has been investigated. This alloy exhibits a first order ferro–antiferro magnetostructural transition at 170 K in the cooling process under a low magnetic field (0.1 T). The transition also proceeds by holding at temperatures near 170 K. In addition, we also found that the transition temperature decreases with increasing magnetic field, and the transition is suppressed under a magnetic field of 5 T or higher. Removal of the magnetic field at 4.2 K induces the magnetostructural transition, but a part of the ferromagnetic phase remains at 4.2 K. In the subsequent heating process, the remaining ferromagnetic phase (high temperature phase) transforms to the antiferromagnetic phase (low temperature phase). Such an abnormal transition implies that the first order magnetostructural transition in Fe0.45Rh0.45Pd0.1 alloy is caused by a thermal activation process.

Recently, we reported single variant formation in an Fe–55at.%Pd is certainly realized from a disordered fcc-phase to an ordered L1 0 -phase by heat-treatment under magnetic field. In the present study, we have investigated microstructure evolution during the process of the single variant formation by an X-ray diffraction and an electron microscopy observation. As a result, followings are obtained: size of the ordered particles at the early stage of ordering is about 2 nm and the nucleation ratio of preferable variant, whose easy axis lies in the field direction, is higher than that of other variants. Each of the ordered preferable variant grows by consuming the order variants and finally come together to become a single variant. Based on the observation, a model is proposed for the single variant formation of the ordered L1 0 -phase under magnetic field.

Martensitic transformation (MT) start temperature of the Fe–33Ni nanoparticles prepared by a Sol–gel and Reduction method was found to be higher than that of bulk specimens with the same composition, which is in contrast to the suppressing MT behaviors of nanoparticles in the previous works. Pre-existing bcc phase and polycrystalline structure were confirmed to assist MT in the Fe–33Ni nanoparticles. The mechanism associated with the promotion effect of the pre-existing bcc phase and the grain boundary on MT is discussed.

Displacement of atoms causing diffuse satellites in a Ti–32Pd–18Fe alloy (in at%) has been investigated by transmission electron microscopy (TEM). In this alloy, diffuse satellites appear at incommensurate positions given by gB2+q when the zone axis is [111] and [001], but they are missing when [110]. This results imply that the satellites appear because of a transverse wave with a propagation vector of q∼1/5〈ζζ¯0〉⁎ and displacement in 〈110〉 direction. Possibly, this transverse wave occurs as a result of phonon softening of the TA2 phonon branch, which is a common feature for B2-type Ti-based shape memory alloys.

The B2–B19′ transformation in Ti–51.2Ni (at.%) alloy exhibits a clear time dependence. The transformation starts after an incubation time when the specimen is held above M s , and proceeds while holding in the temperature range between 160 and 100 K. The nose temperature is expected to be near 130 K. The reverse transformation also shows a clear time dependence.

We have investigated the martensitic transformation in B2-type Fe 50 Rh 40 Pd 10 alloy. The B2-phase (cubic structure) of this alloy exhibits a magnetic transition on cooling from a ferromagnetic phase to an antiferromagnetic phase at 270 K. It then shows a martensitic transformation at 175 K. The martensite phase has a tetragonal structure with a tetragonality c/a = 1.17. We confirmed that the specimen deformed in the martensite phase show an imperfect shape memory effect in association with the reverse transformation. The reverse transformation (from the antiferromagnetic tetragonal phase to the ferromagnetic B2-phase) is expected to be controlled by a magnetic field.

Some shape memory alloys exhibit so-called temperature memory effect (TME) associated with their thermoelastic martensitic transformations. In this paper, we report that the same TME also appears in association with a first order magnetic transition between an antiferromagnetic phase and a ferromagnetic phase in FeRh. Although the TME in shape memory alloys appears only in the heating process, the TME in FeRh appears both in the cooling and heating processes. Presumably, the arrangement of antiferromagnetic domains and ferromagnetic domains is responsible for the TME in FeRh.

We have investigated superelastic behavior of an Fe-30.8Pd (at.%) alloy. This alloy transforms from a FCC parent phase to a FCT martensite phase at 255 K with a latent heat of 38 J/mol. When a compressive stress is applied in the [001] direction above the transformation temperature, the specimen exhibits a large elastic-like strain of more than 3% under relatively small stress of 100 MPa. The large elastic-like strain mainly comes from the elastic strain of the parent and martensite phases. The transformation strain also contributes to the strain, but it decreases linearly with increasing temperature and stress. The transformation strain || is extrapolated to be about 1.4 % under zero stress, and 0 % under 100 MPa. Probably a critical point, at which first order nature of the transformation disappears, exists for the FCC to FCT transformation in Fe-Pd alloys.

Transmission electron microscope (TEM) and scanning transmission electron microscope (STEM) observations were carried out to investigate microstructure formation and variant selection process in L10-type ordered CoPt alloy upon a two-step ordering heat-treatment. The first step corresponds to nucleation process carried out under a magnetic field of 10T and the second step represents growth process without magnetic field. After the first step of ordering, ordered domains of about 5nm in size were observed and fraction of the preferred variant with the c-axis parallel to applied magnetic field was slightly higher than that of the other two variants. Formation of tweed microstructure along f011gL10 was confirmed at the initial stage of ordering. This structure is considered to be derived from the periodic alignment of interface between two ordered variants with twin relation. At the early stage of the second step of ordering, numerous micro-twins were formed through tweed microstructure and the volume fraction of the preferred variant was increased accompanying with modulation of twins, while that of other two variants was decreased. After the second step of ordering, the twins were vanished and single variant was obtained. [doi:10.2320/matertrans.MAW201313]

Fe68.8Pd31.2 exhibits an anomalously large magnetostriction of ~400 ppm at room temperature as well as linear, isotropic and hysteresis free magnetization behavior. This near perfectly reversible magnetic response is attributable to the presence of a large number of premartensitic magnetoelastic twin clusters present in the system made possible through the elastic softening that occurs near a martensitic transformation temperature of 252 K. It is proposed that the twin clusters in the material reduce both internal elastic and magnetic energy, causing the elastic and magnetic behavior of the material to be intimately linked. In such a framework, the anisotropy energy becomes extremely low causing the material to bear no crystalline dependence on magnetization and application of a magnetic field causes simultaneous magnetic and twin domain movement which relaxes the system.

π-Conjugated polymers consisting of 9,10-disubstituted 9,10-dihydrophenenthrene, with the substituents octyl, 2-ethylhexyl, -OSiBu3, etc., units are prepared by organometallic polycondensation. Homopolymers (PH2Ph(9,10-R)) have a π-conjugation system similar to that of polymers of 9,9-dialkylfluorene and show UV-Vis peaks at ≈380 nm. In addition to the peak at ≈380 nm, some homopolymers give rise to a peak at a longer wavelength, suggesting molecular assembly of the polymers. X-ray diffraction data support the molecular assembly. The homopolymers show photoluminescence (PL) with PL peaks at ≈430 nm, and PL spectrum of the polymer film is essentially unchanged after heating the polymer film at 150 °C in air. The homopolymers undergo electrochemical p-doping at about 1.5 V versus Ag+/Ag.

The martensitic transformation in a nearly ordered Fe3Pt is weak first order, and the martensite phase shows a large magnetic field-induced strain of several percent. We have investigated the origin of this transformation from its electronic structure. A characteristic feature in the electronic structure is the existence of a relatively high peak in the density of states of the minority spin band just below the Fermi energy. This peak splits into two peaks by tetragonal distortion, and one of them shifts to lower energy by the distortion, suggesting that the band Jahn-Teller effect is the main cause for the transformation.

We have investigated electron diffraction patterns of a Ti-44Ni-6Fe alloy exhibitng a negative temperature dependence in electrical resistivity below Tmin = 210 K. The electron diffraction patterns taken near Tmin show diffuse satellites at gB2 + * when the zone axis is [111] and [001]. For both the beam directions, the value ζ is slightly smaller than 1/3. On the other hand, the satellites are missing when the zone axis is [110]. This means that the incommensurate phase has a modulated structure with the propagation vector * (ζ~1/3) and the displacement direction is one of which is vertical to the propagation vector. This modulation is obviously the consequence of the phonon softening of TA2-branch with the propagation vector near * (ζ~1/3). In addition to the satellite at gB2 + * (ζ~1/3), satellites appear at gB2+* with ζ = 1/2 when the zone axis is [001] and rod-like steaks appear in * direction when the zone axis is [110]. However, these satellites and rod-like streaks do not show clear temperature dependence, suggesting they are not directly related to the phonon softening of TA2-branch.

We have investigated athermal and isothermal martensitic transformations (typical displacive transformations) in Fe–Ni, Fe–Ni–Cr, and Ni-Co-Mn-In alloys under magnetic fields and hydrostatic pressures in order to understand the time-dependent nature of martensitic transformation, that is, the kinetics of martensitic transformation. We have confirmed that the two transformation processes are closely related to each other, that is, the athermal process changes to the isothermal process and the isothermal process changes to the athermal one under a hydrostatic pressure or a magnetic field. These findings can be explained by the phenomenological theory, which gives a unified explanation for the two transformation processes previously proposed by our group.