1.Thermally Assited Solid State Amorphization
遷移金属(T=Nb,Ta)粉末ならびにAl金属粉末の混合体をロッドミルを用いてメカニカル・アロイングすることにより得られる中間生成結晶相を600〜650Kに加熱すると発熱をともなってアモルファス相が形成されることを見い出した。このプロセスを「Thermally Assisted Solid State Amorphization(TASSA)」と呼称することにした。中間生成結晶相は、X線回折によりAlが過剰に固溶したbccNb(Ta)固溶体であると判断されるが,全相学的には多層組織を有している。TASSA反応の熱力学的ならびに速度論的検討を行ったが,詳細な構造研究は今後の課題として残された。
2.Ta金属のメカニカル・ミリングによるナノ結晶粒化
bccTa結晶粉末をロッド・ミルを用いて約200時間メカニカル・ミリングすると,平均結晶粒径20nmのナノ結晶組織が形成される。中性子非弾性散乱により原子振動エネルギー・スペクトルを測定した結果,ナノ結晶粒化は主として剪断力の働きにより生起していることが分った。
3.Mn-AlならびにFe-N系非平衡磁性合金の作成
MnならびにAl金属粉末の混合体をメカニカル・アロイングするとα-Mn型結晶構造を有する強磁性非平衡合金が得られる。この合金からAlをアルカリ抽出すると磁化は室温で36emu/gに達する。
メカニカル・ミリングによるナノ結晶の形成にともない歪み誘起マルチサイト変態fccγ'.FeN→bct α'・Fe(N)が生起することを見い出した。さらにミリングを続行して得られるアモルファスFe(N)を約650Kで焼鈍するとかなり多量のα"-Fe_<16>N_2非平衡強磁性相が折出することが分かった。

In toughness of steels at the ductile-brittle fracture transition region, both the brittle and ductile natures are involved. The origin of the temperature dependence of toughness must be thus examined from both mechanisms. The ductile nature is represented by the resistance against the stable ductile crack propagation, i.e. R curve. By using steels with various microstructures and compositions, precise measurements of R curves were conducted. It was revealed that the temperature dependence of R curve is small and that factors such as grain size and toughness of ferrite matrix have almost no effects. However, carbon con-tent and particularly grain boundary carbides which act as obstacles of slip propagation were found to have remarkable effects.
Accordingly, the origin of the temperature dependence of toughness is at the initiation stage of brittle fracture, while the processes leading to the initiation constitute the toughness value. By referring to the analyses by means of fracture mechanics calculation, it was shown that the initiation is controlled by the local tensile stress in the crack front area at a low toughness level. On the other hand, at the toughness level where the crack tip blunting is substantial and the stable crack advances, the brittle fracture initiation is controlled by local strain. It implies that the plastic deformation is the key process for the determination of toughness at the ductile-brittle transition region.
Transmission electron microscopy of the area near the brittle fracture ini-tiation site revealed complicated dislocation cell configurations and formation of micro-voids at cell walls even at a very low toughness level. It suggests that deformation structure rather than the cracking of secondary particles induces the brittle fracture initiation.

鉄鋼の延性・脆性破壊遷移を支配する材料因子の作用機構解明

The factor that plays the essential role in hydrogen-related failure has been examined for Inconel 625 and iron by means of tensile testing with interposed unloading and reloading with/without hydrogen charging. Aging at 30 degrees C or annealing at 200 degrees C was conducted during the unloaded stage in order to diffuse out hydrogen or to anneal out strain-induced defects. Hydrogen thermal desorption analysis was used to evaluate strain-induced defects that act as trapping sites of hydrogen. Fracture strain decreased in the initially hydrogencharged specimens even though hydrogen was absent at the late stage of straining. Annealing at 200 degrees C at the unloaded stage almost completely recovered the decrease in fracture strain. Enhancement of strain-induced defects by hydrogen and their involvement in degradation were revealed by means of hydrogen thermal desorption analysis. The results provide direct evidence of the primary role of vacancies rather than hydrogen itself in hydrogen degradation, and agree well with the hydrogen-enhanced strain-induced vacancy model with respect to the mechanism of hydrogen-related failure. (C) 2008 Published by Elsevier Ltd on behalf of Acta Materialia Inc.

The factor that plays the essential role in hydrogen-related failure has been examined for Inconel 625 and iron by means of tensile testing with interposed unloading and reloading with/without hydrogen charging. Aging at 30 degrees C or annealing at 200 degrees C was conducted during the unloaded stage in order to diffuse out hydrogen or to anneal out strain-induced defects. Hydrogen thermal desorption analysis was used to evaluate strain-induced defects that act as trapping sites of hydrogen. Fracture strain decreased in the initially hydrogencharged specimens even though hydrogen was absent at the late stage of straining. Annealing at 200 degrees C at the unloaded stage almost completely recovered the decrease in fracture strain. Enhancement of strain-induced defects by hydrogen and their involvement in degradation were revealed by means of hydrogen thermal desorption analysis. The results provide direct evidence of the primary role of vacancies rather than hydrogen itself in hydrogen degradation, and agree well with the hydrogen-enhanced strain-induced vacancy model with respect to the mechanism of hydrogen-related failure. (C) 2008 Published by Elsevier Ltd on behalf of Acta Materialia Inc.

The features and mechanisms of hydrogen embrittlement of metals that undergo hydride formation and martensitic transformation during service are reviewed. Metastable austenitic stainless steels, Nb and V as well as Ti and its alloys are the metals examined. The embrittlement of austenitic stainless steels is susceptible to the stability of austenite, and hydrogen promotes the martensitic transformation forming hydride as a precursor. However, martensite per se is not always the origin of the embrittlement, but.lattice defects associated with dislocation dynamics in the crack front is noted as a factor. In Nb and V, stress-induced hydride formation successively taking place at the crack front supplies a crack path with a reduced resistance for growth. In Ti, the primary role of hydrides is not always the case. In superelastic and shape memory Ni-Ti alloys, the embrittlement appears prominent in the martensitic transformation, and some dynamic process invoking lattice defects is suggested to play a role. In an appendix, the assessment of the susceptibility in terms of the critical hydrogen concentration is critically discussed, and some mechanical response to hydrogen is proposed as new criteria for the susceptibility of materials to hydrogen embrittlement.

Following the preceding article, a theory that considers the primary role of strain-induced vacancies in hydrogenrelated failure is presented. Detection of lattice defects the creation and agglomeration of which are enhanced by hydrogen is described. As a support of the theory, correlations between the mount of strain-induced vacancies and the susceptibility to hydrogen-related failure are shown with steels of different microstructures. Interaction of fatigue and delayed fracture is shown as another evidence of the theory. The role of hydrogen concentration in the failure is discussed in regard to a time-dependent alteration of defects during delayed fracture test.

Major models so far proposed on the mechanism of hydrogen embrittlement are critically reviewed. The theories are classified into two streams, brittle or ductile fracture in nature. The theories for brittle fracture are triggering by internal pressure, surface energy reduction and lattice decohesion. The ductile fracture models presented here are hydrogen-assisted cracking, hydrogen-enhanced localized plasticity, adsorption-induced dislocation emission and autocatalytic effect on plastic instability. Interconnections of hydrogen embrittlement and stress corrosion cracking are also presented. Particular attention is paid on experimental conditions and assumptions on which the theories are based.

Characteristics of hydrogen-related failure are reviewed from mostly experimentally observed features. Comparison with the failure under normal atmospheres is made so as to extract the way by which hydrogen plays a role in the fracture process. Fractographic features associated with plasticity and the promotion of the crack initiation and growth analyzed by means of an R-curve method are presented. Enhanced plastic instability is noted as a precursor of the failure. Effects of hydrogen on stress relaxation and creep are presented as the dislocation dynamics that operate on delayed fracture. Effects of environmental variation of applied stress and external hydrogen potential are also presented.

Positron lifetime spectroscopy was applied to examine the synergistic effect of hydrogen and plastic straining on the vacancy generation in iron. Hydrogen enhanced the increase in mean positron lifetime, T., by plastic straining and elevated the recovery temperature of T. on isochronal annealing. Multi-component analyses of positron lifetime spectra showed that the presence of hydrogen enhances the generation of vacancies, rather than of dislocations. These results are consistent with previous interpretations on thermal desorption analysis of hydrogen in deformed steels. (c) 2006 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Environmental factors that affect the entry of hydrogen into materials are briefly summarized with respect to electrochemical reactions on electrode. Effects of promoters, particularly of H 2S, are reviewed concerned with their mechanism. The effects of localized corrosion, such as crevice corrosion and pitting corrosion, and atmospheric corrosion under wet/dry cycles on hydrogen entry are reviewed. Further, interactions between electrochemical reactions on the surface of electrode and plastic deformation, which appear on polarization during stressing, and vice versa, are noticed as phenomena that suggest some surface effects on internal dislocation dynamics.

Basic electrode reactions for the entry of hydrogen into metals under aqueous environment have been reviewed. Some analytical models that determine the parameters in the kinetics of the entry of hydrogen, including the hydrogen surface coverage and absorbed hydrogen concentration, are presented. On corrosion processes, the significance of surface film for the entry of hydrogen is emphasized from both the structure of the electric double layer at the electrode/electrolyte interface and under-potential deposition of hydrogen.

Adsorption behavior of gaseous hydrogen on metal surface has been briefly reviewed, focusing on fundamentals for hydrogen entry that causes degradation of mechanical properties. The measurements for the adsorbed sites and states by means of thermal desorption spectroscopy and the work function are presented. Studies on the isotherm and adsorption kinetics are also presented. The effects of alloying elements and impurities on hydrogen coverage are summarized as well as the effects of gas components.

Analytical methods for hydrogen states in materials are reviewed. Principles and characteristics of visualization techniques, such as tritium radioautography, hydrogen microprint technique and scanning secondary ion mass spectroscopy, and thermal desorption analysis are briefly described. Examples of applications to practical cases are presented putting emphasis on the limitations and uncertainties involved in experiments and theories.

The effect of grain size on the susceptibility of martensitic steels to hydrogen-related failure has been examined with respect to the role of the hydrogen concentration and lattice defects acting as hydrogen trap sites. Plastic straining increased the hydrogen absorption capacity, but annealing at 250 degrees C after straining eliminated the increase, implying the point-defect-like nature of the increased number of trapping sites for hydrogen. Refinement of the prior austenite grain size, by means of repeated induction heating and quenching, increased the hydrogen absorption capacity in as-heat-treated samples as a consequence of the increased boundary area. Grain refinement reduced the susceptibility, as evaluated by means of a slow strain-rate tensile test of hydrogen-precharged specimens. This reduction of susceptibility is analysed to be associated with a smaller strain-induced increase in hydrogen absorption capacity, i.e. in the defect density. Fractographic features showed a decrease in ductile crack growth resistance in the presence of hydrogen, being less pronounced with grain refinement. The present results support the notion that the susceptibility of steels to hydrogen-related failure originates in the response of microstructures to plastic straining, primarily through the creation of vacancies.

The equilibrium distribution of hydrogen at trapping sites is shown putting emphasis on the effect of the binding energy. Trapping effects on hydrogen diffusivity derived from both the solution of diffusion equation and statistical thermodynamics are reviewed. Analytical and experimental estimations of the kinetics of trapping are reviewed. Dislocation transport of hydrogen by moving dislocations is critically examined.

Succeeding the hydrogen states in solid solution, interactions of hydrogen with various defects such as dislocations, vacancies, precipitates, grain boundaries and voids have been reviewed. The binding energies of hydrogen with defects are collected with emphasis on experimental methods and assumptions in the theoretical analyses. The origins of the interactions are briefly described in terms of electron theories on the states of hydrogen.

The alloying process by mechanical milling and subsequent annealing of a 1:2 mole ratio Ti/2Si system has been analyzed from a crystallographic viewpoint. An intermediate phase appeared after prolonged milling or prior to the formation of TiSi2 on annealing. The observed XRD pattern of the intermediate phase was coincident with that of a structure constructed by introducing a stacking fault on the close-packed plane of the hexagonal Ti lattice with partial replacement of Ti by Si atoms. The structure of TiSi2 is formed from the intermediate phase by an interchange of Ti with Si atoms on the close-packed plane, being associated with structural relaxation. (C) 2003 Elsevier B.V. All rights reserved.

The susceptibility of a Ni-Ti shape memory alloy towards hydrogen embrittlement has been examined with respect to the constituent phase of the alloy. Fracture always occurred in the presence of martensite. A steep decrease in fracture stress took place initially with increasing hydrogen content irrespective of the initial phases. The effect of the constituent phase appeared at high hydrogen contents. The work-hardened martensite phase showed the highest susceptibility succeeded by the parent phase that fractured in the course of martensite transformation. The critical stress for the martensite transformation increased at high hydrogen contents. The embrittlement is likely to be enhanced by the interactions between hydrogen and dynamic processes in materials. (C) 2004 Elsevier B.V. All rights reserved.

Hydrogen-related degradation of the mechanical properties of a Ni-Ti superelastic alloy has been examined by means of delayed fracture tests in acidic and neutral fluoride solutions and hydrogen thermal desorption analysis. Delayed fracture took place in both solutions; the time to fracture was shorter in the acidic solutions than in the neutral solutions with the same fluoride concentration. The time to fracture was reduced in both solutions when applied stress exceeded the critical stress for martensite transformation. In the acidic solutions, Ni-Ti superelastic alloy underwent general corrosion and absorbed substantial amounts of hydrogen. Fractographic features suggested that the delayed fracture in the acidic solutions was attributable to hydrogen embrittlement, whereas in the neutral solutions, a different fracture mode appeared associated with localized corrosion only in the vicinity of the fracture initiation area. In the neutral solutions, the amount of absorbed hydrogen was much less than that in the acidic solutions, and the delayed fracture was likely to be induced by active path corrosion accompanying hydrogen absorption. The results of the present study imply that the hydrogen-related degradation of performance of Ni-Ti superelastic alloys occurs in the presence of fluoride. (C) 2004 Wiley Periodicals, Inc.

Fundamentals of hydrogen behaviors in materials are reviewed with respect to degradation of mechanical properties. The present article is the first part that deals with the states of hydrogen in solid solution. The temperature dependence of the solubility, permeability and technical problems particularly of electrochemical permeation measurement have been explained. The interstitial site of hydrogen and the symmetry of lattice strain have been examined.

Main features of hydrogen effects on mechanical behavior of steels have been reviewed. While the effects on macroscopic tensile properties are not definite, the increase in dislocation mobility by hydrogen appears as an enhanced stress relaxation the extent of which correlating with the susceptibility to failure. Hydrogen enhances the creation of vacancies during plastic deformation. The extent is susceptible to microstructures of steels and correlates with the susceptibility to failure. Flow localization in the presence of hydrogen is substantial, leading to shear instability. Characteristic fractographic features in hydrogen-related failure, such as striations, have been shown to originate in deformation bands in which the density of defects is high. Prominent localization of void nucleation takes place at the advancing crack tip. Amorphization has been observed in front of the crack and just below the fracture surface in a hydrogen-charged steel in accord with flow localization and associated vacancy creation. Hydrogen remarkably reduces the fatigue resistance of steels, and, conversely, prior fatigue treatment increases the susceptibility to delayed fracture. The mechanism has been ascribed to the creation of vacancies during fatigue and interactions between vacancies and hydrogen. Further studies for assessing the susceptibility to hydrogen-related failure have been suggested to be based on the accumulation of damage rather than hydrogen content.

The fracture of commercial pure titanium in acid and neutral fluoride solutions has been examined by a sustained tensile-loading test and hydrogen thermal desorption analysis. It was found that the fracture of titanium occurred in neutral 2.0% NaF solution as well as in 2.0% acidulated phosphate fluoride (APF) solution. The time to fracture decreased with increasing applied stress in both 2.0% APF and 2.0% NaF solutions. In the case of the same applied stress, the time to fracture in the 2.0% APF solution was shorter than that in the 2.0% NaF solution. General corrosion was exhibited on the side surface of the tested specimens. The formation of sodium titanium fluoride was observed on the surface of the immersed specimens in the 2.0% APF solution. Hydrogen desorption of the tested specimen in the 2.0% APF solution was observed with a peak at approximately 600°C. The amount of absorbed hydrogen was &gt
300 mass ppm in the 2.0% APF solution under an applied stress for 24 h. The results of the present study imply that applying stress to titanium by immersing in fluoride solutions leads to the degradation of its mechanical properties. © 2003 Wiley Periodicals, Inc.

Amorphization, following a decrease in the dislocation density and fragmentation of the matrix, has been revealed in a thin layer just below the tensile-fractured surface of a hydrogen-charged ferritic steel. The creation of a high density of vacancies associated with hydrogen is discussed as the origin. (C) 2003 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

The distribution of hydrogen trapped at defects has been visualized by means of a hydrogen microprint technique (HMT). A low carbon steel having a ferrite/pearlite structure with second phases along grain boundaries was subjected to notch-tensile and three-point bending tests with/without hydrogen pre-charging. HMT was applied to tested specimens re-charged with hydrogen. Accumulation of hydrogen was confirmed in strain-concentrated areas as well as along grain boundaries and in pearlite. A noteworthy finding was a substantial increase in defects acting as traps of diffusive hydrogen when specimens were deformed in the presence of hydrogen. The increased defects were annealed out at temperatures as low as 150 degreesC, indicating their point defect nature. Another finding was the appearance of a zone denuded of silver precipitation in highly strained areas such as near grain boundaries and the fracture surface, where tritium autoradiography revealed the evolution of defects having high binding energy with hydrogen. (C) 2003 Elsevier B.V. All rights reserved.

Fatigue damage involving creation of point defects has been revealed by means of hydrogen thermal desorption analysis, utilizing hydrogen as a probe of defects. Two types of high-strength martensitic steels were subjected to a rotational bending test for various cycles. The amount of absorbed hydrogen decreased or was almost constant in the early stage of the fatigue process and then began to increase after a substantial number of fatigue cycles. The increase in the final stage was due to the creation of point defects, presumably vacancies, while the decrease in the early stage was ascribed to changes in microstructural constitutions. Hydrogen-precharging drastically reduced the fatigue life, but the effect of hydrogen on the fatigue limit was not significant. The fracture surface of hydrogen-precharged specimens was smooth without cleavage-like crystallographic features. The defect density in fractured specimens in the presence of hydrogen was higher than in specimens fatigued for a similar number of cycles without hydrogen. Involvement of hydrogen in fatigue damage in the crack nucleation and growth stages is discussed. (C) 2002 Elsevier Science B.V. All rights reserved.

Hydrogen embrittlement of Ni-Ti superelastic alloy in a fluoride solution (0.2% APF) has been investigated by means of a tensile test (after immersion) and hydrogen thermal desorption analysis. Upon immersion, the tensile strength of the alloy decreased to the critical stress level of martensite transformation. Hydrogen desorption of the immersed specimens appeared with a peak at around 500°C. The amount of absorbed hydrogen in the alloy ranged from 100 to 1000 mass ppm when immersed in the fluoride solution for 2 to 24 h. The immersion in the fluoride solution led to the degradation of mechanical properties due to hydrogen embrittlement. The results of the present study imply that one reason that Ti and its alloys fracture in the oral cavity is the fact that hydrogen is absorbed in a fluoride solution, such as prophylactic agents. © 2003 Wiley Periodicals, Inc.

Interaction between fatigue damage and hydrogen is the concern of the present study. The susceptibility of a high-strength martensitic steel to delayed fracture has been examined using as-heat-treated and pre-fatigued specimens. The pre-fatigued specimens showed a tendency to fail earlier, but annealing the pre-fatigued specimens at 200 degreesC recovered nearly all of the delayed fracture life. Production of point defects during fatigue was detected by means of hydrogen thermal desorption analysis (TDA), using hydrogen as a probe of defects. Hydrogen absorption capacity increased in fatigued specimens, but it was reduced to the level of the as-heat-treated specimens when fatigued specimens were annealed at 200 degreesC, implying that increased hydrogen-trapping defects, presumably vacancies, were produced during fatigue. Hydrogen TDA peak profiles showed alterations that imply agglomeration of vacancies associated with an increase in fatigue cycles. The involvement of vacancies created during fatigue in the enhanced delayed fracture is consistent with a model that proposes strain-induced vacancies and their agglomeration are the primary mechanism of hydrogen-related failure. (C) 2002 Elsevier Science B.V. All rights reserved.

The hydrogen thermal desorption of a martensitic steel has been simulated assuming lattice hydrogen diffusion under a local equilibrium with reversibly trapped hydrogen as the rate-determining process, The calculated desorption curves reproduced the observed shift of the peak temperature associated with the specimen thickness and the heating rate. The calculation method involves a combination of a defect density and a hydrogen/defect binding energy as parameters, The dependence of the peak temperature on the defect density and the binding energy has been quantitatively shown, Assignment of the lattice defect relevant to the desorption curves is discussed. A calculation that took into account the increase in the defect density yielded results consistent with the observed change in the desorption curves associated with plastic strain,

The susceptibility of Ni-Ti superelastic alloys to hydrogen embrittlement has been examined by means of a delayed-fracture test and hydrogen thermal desorption analysis. It was found that the time to fracture was drastically reduced when the applied stress exceeded the critical stress for martensite transformation. In the applied stress range lower than the critical stress, the time to fracture lessened in the order of instability of the alloys to undergo reversible martensite transformation. Hydrogen thermal desorption of specimens subjected to delayed-fracture test is classified into two types according to the applied stress level. The amount of desorbed hydrogen was markedly increased when the applied stress was higher than the critical stress. It was concluded that Ni-Ti superelastic alloys transformed to martensite are sensitive to environmental conditions accompanying accelerated hydrogen embrittlement. (C) 2002 Elsevier Science B.V. All rights reserved.

The mechanism of intergranular-mode fracture in hydrogen-related failure of high-strength martensitic steels has been investigated. Pronounced degradation of tensile properties appeared with increasing manganese content in a slow-elongation-rate test under concurrent hydrogen charging. The fracture mode was intergranular with tear traces along martensite lath boundaries. The tear traces disappeared and the average surface roughness decreased with increasing manganese content. Thermal desorption analysis of hydrogen charged to deformed specimens. It was revealed that the density of point defects increased owing to straining and was more noticeable in steels with a higher manganese content. In common with transgranular-mode fracture, the primary function of hydrogen in intergranular-mode fracture is thought to be one of stabilizing and increasing the density of strain-induced vacancies that lead to the formation of microcracks or microvoids in the vicinity of boundaries. The constraint of plastic deformation at grain boundaries due to boundary phases is likely to determine the susceptibility to hydrogen-related failure induced by strain concentration.

A nickel-titanium superelastic alloy is susceptible to environmental embrittlement in a corrosive atmosphere. Because a delayed fracture of the alloy is associated with hydrogen absorption and subsequent formation of brittle hydride phases, the diffusion rate of hydrogen is thought to be one of the factors determining its service life. The Ni-Ti alloys subjected to hydrogen charging of I or 10 A/m(2) for 24 or 120 hours, respectively, were arranged using an electrochemical system. Both the hardness numbers in the cross-sectional area of the alloy and the amount of evolved hydrogen were determined. The fracture surface of the alloys, under tension, was observed using a scanning electron microscope (SEM). Theoretical distributions of the hydrogen concentration were computed for an infinite cylinder model using the differential equation of diffusion. The diffusion constant of hydrogen through the alloy is estimated to be 9 x 10(-15) m(2)/s, assuming that the hardness is proportional to the concentration of hydride and/or hydrogen. Experimental results of the hardness measurements and fractography support the estimated diffusion constant. The process of fracture formation in a biological corrosive environment was discussed. It was concluded that galvanic currents and fretting corrosion of the alloy might be effective factors in fracture formation during function.

The susceptibility to hydrogen embrittlement (HE) of martensitic steels has been examined by means of a delayed-fracture test and hydrogen thermal desorption analysis. The intensity of a desorption-rate peak around 50 degreesC to 200 degreesC increased when the specimen was preloaded and more remarkably so when it was loaded under the presence of hydrogen. The increment appeared initially at the low-temperature region in the original peak. As hydrogen entry proceeded, the increment then appeared at the high-temperature region, while that in the low-temperature region was reduced. The alteration occurred earlier in steels tempered at lower temperatures, with a higher embrittlement susceptibility. A defect acting as the trap of the desorption in the high-temperature region was assigned to large vacancy clusters that have higher binding energies with hydrogen. Deformation-induced generation of vacancies and their clustering have been considered to be promoted by hydrogen and to play a primary role on the HE susceptibility of high-strength steel.

Various models so far proposed for the mechanism of hydrogen embrittlement (HE) of steels are critically reviewed with respect to the manifestation of hydrogen in the fracture process. Recent studies that elucidate the hydrogen states and their relevance to HE are discussed. Particular attention is paid to the role of deformation-induced defects that interact with hydrogen. A model is proposed in which increased vacancy density and agglomeration lead to the promotion of failure. The model ascribes HE to the context of ductile fracture in which vacancies play the primary role.

The effect of hydrogen on the shear localization and associated crack nucleation has been investigated by means of a three point bending test of hydrogen-charged steels. The ductile crack growth resistance in terms of the slope of R-curve was lowered under the presence of hydrogen, the decrease being more pronounced in the steel with more abundant slip constraint phases along grain boundaries. Enlargement of size and reduction in depth/width ratio of primary dimples, occasionally associated with quasi-cleavage, were observed on the fracture surface of the hydrogen-charged steels. By means of a FEM calculation, the increase of the nucleation void volume fraction localized at the crack tip with strain localization as well was shown to take place in the hydrogen-charged steel in consistent with enhanced shear instablity. It was discussed that the evolution of vacancy-type defects, rather than void nucleation at second phase particles, in the course of plastic straining was enhanced under the presence of hydrogen, reducing the ductile crack growth resistance.

Defects induced by plastic deformation have been revealed by means of room temperature desorption and thermal desorption spectroscopy of tritium with regards to the ductile crack growth resistance and brittle fracture initiation in steels in the ductile-to-brittle transition region. Tritium, as a probe for detecting defects, was introduced into non-deformed or deformed samples. Three steels, the microstructures of which are characterized by the constrains factor for slip extension, were employed. The residual tritium in a specimen after three days at room temperature increased to the extent corresponding to the constraint factors when plastic deformation was applied. The thermally desorbed tritium, with a peak desorption rate around 150 degrees C, also increased according to the constraint factors. Referring to the previous analysis of the R-curves, the constraint for the extension of slip across grain boundaries is shown to control both the ductile crack growth resistance and the brittle fracture initiation through the deformation microstructures that evolve vacancy clusters or microvoids. (C) 2000 Acta Metallurgica Inc. Published by Elsevier Science Ltd All rights reserved.

The effect of HNO3 surface treatment on the early stages of pitting of Type 430 stainless steel has been studied by AC impedance method. At a given low potential. the Warburg impedance coefficient sigma, which is calculated from impedance plots, is a maximum after wet polishing, and decreases with increasing HNO3 concentration in the solution of surface treatment. It has been suggested that the surface becomes more passive and the most susceptible sites for pit nucleation are eliminated after HNO3 treatments, and that increasing the HNO3 concentration in treatment solution results in increasing the resistance of the steel to metastable pitting. The critical potential (E-m) at which metastable pit or pits start to grow on the surface of the steel depends on HNO3 concentration in treatment solution. The higher the HNO3 concentration in treatment solution, the higher the E-m values, showing that the nucleated pit or pits proceeding to metastable growth stage becomes more difficult. (C) 2000 Elsevier Science Ltd. All rights reserved.

Defects induced by plastic deformation have been revealed by means of room temperature desorption and thermal desorption spectroscopy of tritium with regards to the ductile crack growth resistance and brittle fracture initiation in steels in the ductile-to-brittle transition region. Tritium, as a probe for detecting defects, was introduced into non-deformed or deformed samples. Three steels, the microstructures of which are characterized by the constrains factor for slip extension, were employed. The residual tritium in a specimen after three days at room temperature increased to the extent corresponding to the constraint factors when plastic deformation was applied. The thermally desorbed tritium, with a peak desorption rate around 150 degrees C, also increased according to the constraint factors. Referring to the previous analysis of the R-curves, the constraint for the extension of slip across grain boundaries is shown to control both the ductile crack growth resistance and the brittle fracture initiation through the deformation microstructures that evolve vacancy clusters or microvoids. (C) 2000 Acta Metallurgica Inc. Published by Elsevier Science Ltd All rights reserved.

Thermal desorption spectroscopy (TDS) has been used to reveal the nature of defects acting as trapping sites of hydrogen. Hydrogen was charged to ferritic and eutectoid steels deformed to various degrees and then given annealing treatment. Desorption with a single peak appeared between roam temperature and 600 K from ferritic steels. Under constant hydrogen charging conditions, the amount of desorption increased with strain. However, when the deformed samples were subjected to annealing at temperatures as low as 500 K, the increase of desorbed hydrogen no longer appeared. Vacancy clusters, which themselves annihilate in the course of TDS measurement, are the probable source of hydrogen desorption. When heavy deformation was given to ferritic steels, a two-step decrease of hydrogen desorption took place with increasing annealing temperature, corresponding to annihilation of vacancy clusters and decrease of dislocation density, respectively. The desorption with a single peak has two origins, one due to the annihilation of the trapping sites themselves and the other to desorption from stable sites. For heavily deformed eutectoid steel, an additional desorption peak centered at around 640 K appeared. The peak likely results from deformation-induced defects within the cementite phase or supersaturated carbon in ferrite. Two types of desorption, one due to the annihilation of trapping sites in the course of measurement and the other due to desorption from stable sites, should be discriminated. TDS using hydrogen as a tracer can be applied as a tool to investigate the various defects induced by plastic deformation. (C) 1999 Elsevier Science S.A. All rights reserved.

Synthesis of TIC and Ti/TiC composite has been conducted by mechanical alloying and following Spark-Plasma-Sintering starting from elemental Ti powder with graphite or heptane. From graphite, formation of TiC takes place rapidly after a definite period of milling while it proceeds gradually from heptane. The apparent compositional range of Ti1-xCx is extended to a C/Ti ratio as low as 30/70 when graphite is used, while a mixture of Ti and TiC is obtained at a C/Ti ratio of 32/68 with heptane. Sintering can proceed at substantially lower temperatures compared with the conventional process, the sintering temperature decreasing with lower C/Ti ratios. The lattice parameter of TiC and the density of the compacts decrease with lower C/Ti ratios, but are larger than previously reported ones through conventional methods. The mechanism which causes the differences between graphite and heptane is discussed with respect to the reaction process during milling.

Microstructural evolution of commercial Ti and Zr powders upon mechanical milling was characterized by means of X-ray diffraction, scanning electron microscopy and transmission electron microscopy. Evident structural inhomogeneity was observed upon milling the two commercial powders. A partial amorphous state was attained in Zr after 100 h of milling. The possible origins for the amorphous phase formation were discussed.

In order to clarify the differences in the anodic surface oxidation mechanisms of PAN-based and pitch-based carbon fibres, the fibres were oxidized in an electrolyte and characterized using the coulostatic method, X-ray photoelectron spectroscopy, laser Raman spectroscopy, and X-ray diffraction. The interfacial bonding strength to an epoxy resin was evaluated based on the interlaminar shear strength (ILSS). The results showed a good correlation between the differential double layer capacities, which were measured with the coulostatic method, and the ILSS values of PAN-based high tensile strength carbon fibres (PAN-HTCFs), PAN-based high modulus carbon fibres (PAN-HMCFs), and pitch-based high modulus carbon fibres (pitch-HMCFs). Their morphologies for the anodic oxidation were as follows: PAN-HTCFs are anodized homogeneously; pitch-HMCFs are selectively oxidized and promote crevice etching; and PAN-HMCFs resist crevice etching due to the many defects in the hexagonal network. (C) 1999 Kluwer Academic Publishers.

The effect of Ca addition on fracture toughness has been examined by means of the analysis of fracture mechanics and tensile tests together with micro-structural observation. The Ca addition results in the spheroidization and reduction of the number of non-metallic inclusions, while the matrix structure does not show significant changes. Fracture mechanics test in terms of J-integral using a three point bending rest shows the increase in the slope of the R-curve, indicating enhanced increase in the stable crack growth resistance. The dimple morphology on the ductile fracture surface is classified into two types, i.e. primary and secondary dimples. By Ca addition, the depth of the primary dimple and local shear area on the fracture surface increase. It implies that Ca addition causes increased local plastic deformation in the course of stable crack extension, resulting in the improvement of fracture toughness. On the other hand effect of Ca addition does not appear in tensile test. It suggests that the effect of micro-structural factors on mechanical properties depends on the testing methods.

NbC nanorods and nanoparticles have been synthesized using a vapor-solid reaction path starting with carbon nanotube precursors. Their structures are studied using XRD, TEM, and HRSEM. Their superconducting properties of nanoparticles were also characterized using a SQUID magnetometer. For reactions at lower temperatures, NbC nanorods (10 similar to 20 nm) which replicate the precursor carbon nanotubes are observed. For reactions at higher temperature, coarsened NbC nanoparticles (100 similar to 500 nm) are observed which have spherical or faceted morphologies. A morphological Rayleigh instability can be applied as initiating the transition from NbC nanorod to nanoparticle morphologies. NbC nanoparticles with a little larger lattice parameter show a higher T-c than stoichiometric bulk NbC.

Synthesis of TiC and Ti/TiC composite has been conducted by mechanical alloying and following Spark-Plasma-Sintering starting from elemental Ti powder with graphite or heptane. From graphite, formation of TiC takes place rapidly after a definite period of milling while it proceeds gradually from heptane. The apparent compositional range of Ti1-xCx is extended to a C/Ti ratio as low as 30/70 when graphite is used, while a mixture of Ti and TiC is obtained at a C/Ti ratio of 32/68 with heptane. Sintering can proceed at substantially lower temperatures compared with the conventional process, the sintering temperature decreasing with lower C/Ti ratios. The lattice parameter of TiC and the density of the compacts decrease with lower C/Ti ratios, but are larger than previously reported ones through conventional methods. The mechanism which causes the differences between graphite and heptane is discussed with respect to the reaction process during milling.

Intergranular proeutectoid ferrite nucleation in a V-N steel has been examined at isothermal transformation at 600 °C with respect to the austenitizing temperatures. While substantial grain growth takes place, the number of intergranular ferrite grains per unit grain boundary area decreases with increasing austenitizing temperatures. Non-stationary increasing rate with time is observed. Overall transformation during isothermal transformation delays after austenitizing at higher temperatures and follows the Johnson-Mehl-Avrami formula for the case of fast grain boundary nucleation. It is discussed that the decrease in the intergranular ferrite nucleation rate associated with higher austenitizing temperatures is partially due to grain boundary structures, not merely to the decrease in the nucleation sites associated with grain growth. Use of high austenitizing temperature is requisite for the intragranular ferrite precipitation.

Pitch-based high modulus carbon fibers were oxidized in air and characterized using the coulostatic method in order to compare them with anodized fibers. The interfacial shear strength (IFSS) of these fibers to an epoxy resin was evaluated by a micro debonding test. The results showed a good correlation between the differential double layer capacities, Cd, which were measured by the coulostatic method, and the IFSS values for air-oxidized and anodized fibers. In order to investigate the variation of the air-oxidation and anodization in the tow, the coefficients of variation for the Cd and polarization resistance (Rp) by the mono-filament coulostatic method were compared. As a result, the variation of oxidation in the tow of air-oxidized fiber seemed to be slightly larger than that of anodized fiber. (C) 1999 Elsevier Science Ltd. All rights reserved.

Intergranular proeutectoid ferrite nucleation in a V-N steel has been examined at isothermal transformation at 600 degrees C with respect to the austenitizing temperatures. While substantial grain growth takes place, the number of intergranular ferrite grains per unit grain boundary area decreases with increasing austenitizing temperatures. Non-stationary increasing rate with time is observed. Overall transformation during isothermal transformation delays after austenitizing at higher temperatures and follows the Johnson-Mehl-Avrami formula for the case of fast grain boundary nucleation. It is discussed that the decrease in the intergranular ferrite nucleation rate associated with higher austenitizing temperatures is partially due to grain boundary structures, not merely to the decrease in the nucleation sites associated with grain growth. Use of high austenitizing temperature is requisite for the intragranular ferrite precipitation.

Toughness and related characteristics of microstructures are examined on heat affected zone (HAZ) of weldments of Ti-killed steel given single- or double-heat cycle treatment, Brit-tie fracture initiation and stable crack propagation behavior are examined from fracture toughness rest, observation of tensile deformation microstructures and local strain distributions, It is revealed that resistance for the stable crack propagation of Ti-killed steel is hardly affected by the simulated thermal cycle conditions. However, the ductile-brittle fracture transition temperature increases in double-hear cycled specimens compared with single-heat cycled specimens. It is also revealed that higher strain concentration takes place at the prior austenite grain boundary in a double-heat cycled specimen. It is discussed that strain localization leads to the brittle crack nucleation.

The mechanism of retarded growth rate of intergranular proeutectoid ferrite in a V-C-N steel has been investigated by examining the effect of holding in austenite region during cooling from the solutioning prior to the isothermal transformation at 600 degrees C. The growth rare increases and the transformation start time decreases by increasing the holding temperature, while a C-curve behaviors appears. The growth rare can be ascribed to the supersaturation of C at the transformation interface, THERMO-CALC calculation indicates that substantial precipitation of V(C, N) takes place in the V-C-N steel and the fraction of C in the precipitate decreases with the increasing holding temperature. Analysis of the transformation kinetics in terms of the Johnson-Mehl-Avrami equation shows a two-step process, suggesting decrease in the ferrite nucleation rate by the holding. It is concluded that absorption of C into the precipitate during the incubation period decreases the supersaturation of C at the transformation interface.

(A) quaternary Ti(34)Z(11)Cu(47)Ni(8) alloy is subjected to mechanical alloying on a low energy ball mill. The structure evolution of the as-milled powers upon heating is investigated by means of thermal analysis and X-ray diffraction. An unknown metastable phase is formed by heat treating the 25 h-milled sample which retains the crystalline state. Amorphization is clearly confirmed in the 100 h-milled Ti34Zr11Cu47Ni8 alloy. Thermal crystallization of the as-milled amorphous phase proceeds via two separate steps corresponding respectively to crystallization of NiTi and Cu51Zr11 intermetallic compounds, respectively. The influences of chemical composition and milling time on amorphization and thermal crystallization of the Ti34Zr11Cu47Ni8 alloy are discussed.

Growth kinetics of intergranular ferrite in a V-N steel has been investigated from the measurement of isothermal transformation kinetics. When compared with a v-C steel, the start of intergranular ferrite transformation starts earlier on isothermal transformation in the V-N steel, whereas irs growth is retarded. The growth rates of the intergranular ferrite are under the parabolic rate rule for both steels with a smaller rate constant for the V-N steel. When the rate constants are ascribed to the supersaturation of carbon at the ferrite/austenite interface, the carbon concentrations are estimated to be 0.29 and 0.35 mass% for the V-N and V-C steels, respectively. Johnson-Mehi-Avrami plot of the transformation kinetics indicates that the ferrite nucleation rate is larger for the V-N steel. II is discussed that the precipitation of intragranular ferrite in the V-N steel is facilitated owing to the retardation of the intergranular ferrite growth in the steel.

Pure elemental Ti and Zr powders are respectively subjected to mechanical milling. It is discovered that after milling for 100 h, amorphization occurs in Zr while only nanocrystallization is achieved in Ti. Further thermal analysis and electron microscope observations verify amorphization in the 100 h-milled Zr sample. The crystallization peak temperature of the 100 h-milled Zr sample is determined to be 875 K by using a differential scanning calorimeter at a heating rate of 40 K/min. The above experimental results are discussed.

Pure elemental powders of Ti, Zr, Cu and Ti corresponding to the compositions of Ti34Zr11Cu47Ni8 and Ti37Zr17Cu42Ni4 (at%) were subjected to mechanical alloying on a low energy milling machine in the present work. The premixed powders of the above two compositions underwent the following transformations during mechanical alloying up to 100 h: mu m-sized (Ti, Zr, Cu, Ni)--&gt;nm-sized (Ti, Zr, Cu, Ni)--&gt;amorphous+ a small amount of nm-sized crystallites. The 100-h milled Ti34Zr11Cu47Ni8 sample was found to exhibit an excellent thermal stability against crystallization with the crystallization temperature of about 200 K higher than that of the as-quenched amorphous counterpart. Contrary to the previous rapid quenching result, amorphous phase is easier to occur in Ti37Zr17Cu42Ni4 than in Ti34Zr11Cu47Ni8 by MA, indicating that the glass forming ability by rapid quenching is different from that by MA.

Hydrogen absorption characteristics of pure iron and eutectoid steel fabricated by cold drawing have been evaluated by thermal desorption spectrometry (TDS). The pure iron specimens with ferrite (alpha) structure and the eutectoid steel specimens with alpha and Fe3C structures, were produced under various degrees of reduction and also under various heat treatment conditions. These amount of hydrogen absorbed in specimens dipped in 20 mass%NH4SCN solution were measured using TDS. TDS analysis showed that hydrogen evolution rate of pure iron has only one peak, while that of eutectoid steel has two peaks. From transmission electron microscopy and TDS analysis, the lower temperature peak was attributed to hydrogen released from the trapping sites such as point defects, clustered defects, and dislocations in alpha. The higher temperature peak was found to correspond to hydrogen released from the trapping sites such as defects in Fe3C, and/or such as the disordered interface between alpha and Fe3C.

Hydrogen absorption characteristics of pure iron and eutectoid steel fabricated by cold drawing have been evaluated by thermal desorption spectrometry (TDS). The pure iron specimens with ferrite (α) structure and the eutectoid steel specimens with α and Fe3C structures, were produced under various degrees of reduction and also under various heat treatment conditions. These amount of hydrogen absorbed in specimens dipped in 20 mass%NH4SCN solution were measured using TDS. TDS analysis showed that hydrogen evolution rate of pure iron has only one peak, while that of eutectoid steel has two peaks. From transmission electron microscopy and TDS analysis, the lower temperature peak was attributed to hydrogen released from the trapping sites such as point defects, clustered defects, and dislocations in α. The higher temperature peak was found to correspond to hydrogen released from the trapping sites such as defects in Fe3C, and/or such as the disordered interface between α and Fe3C.

Microstructure and tensile properties at elevated temperatures of fine composites prepared by mechanical alloying of Ti and SiC powders have been investigated. Mixtures of elemental Ti and SiC powders in the composition of Ti-24 mass%SiC are milled under an Ar atmosphere using an attritor ball mill. After milling for 36 ks, formation of TiC and Ti5Si3 is observed and further milling adding n-heptane for 3.6 ks makes the powder particles finer. Milled powders are consolidated by hot pressing at 1773 K for 10.8 ks under 70 MPa and hot isostatic pressing at 1773 K for 10.8 ks under 196 MPa. The compact consists of mostly equi-axed TiC and Ti5Si3 grains of an average diameter of 0.86 μm partially mixed with plate-like phase of Ti3SiC2. The compact shows elongation of 130% at tensile test at 1773 K with an initial strain rate of 4.0 × 10-4 s-1 with the strain rate sensitivity exponent of 0.38. Most grains remain equi-axed after tensile test, while random orientations of plate-like phases of Ti3SiC2 turn parallel along the tensile axis. The large elongation observed above 1773 K is likely due to the superplasticity, while rotation of the plate-like phase accompanies deformation of the surrounding matrix.

Local stress and strain fields in front of a crack have been computed by FEM analysis using a Gurson model which takes into account void nucleation and growth within the material. The original Gurson model has been revised to fit the actual material by introducing experimentally obtained deformation and fractographic data of the material. Calculation has been applied for the ductile-to-brittle transition regions up to the ductile crack growth stage in two steels with different transition behaviors. Higher magnitudes of local stress and strain as well as larger extensions have been shown for a steel which has less nucleated void volume fraction, resulting in higher resistance against the ductile crack growth. Local stress increases with lowering test temperature whereas local strain decreases. Insignificant temperature dependence of R-curves has been understood from the calculated J-integral values because of the cancellation of the stress increase and strain decrease. The calculated R-curves have been shown to be consistent with the experimentally observed ones.

Microstructures controlling the ductile crack growth resistance in the ductile-brittle fracture transition region have been investigated with three low carbon low alloy steels, which showed characteristic differences in the R curves. The crack growth resistance is related to both the primary dimple morphology and the total length of local shear zones appearing on the fracture surface; the latter contribution predominates over the former. The heterogeneity of the microstructures, which constrains slip propagation at the grain boundaries, supplies sites for easy void nucleation and induces local shear and the resulting surface roughness.

Microstructure and tensile properties at elevated temperatures of fine composites prepared by mechanical alloying of Ti and SiC powders have been investigated. Mixtures of elemental Ti and SiC powders in the composition of Ti-24 mass%SiC are milled under an Ar atmosphere using an attritor ball mill. After milling for 36 ks, formation of TiC and Ti5Si3 is observed and further milling adding n-heptane for 3.6 ks makes the powder particles finer. Milled powders are consolidated by hot pressing at 1773 K for 10.8 ks under 70 MPa and hot isostatic pressing at 1773 K for 10.8 ks under 196 MPa. The compact consists of mostly equi-axed TIC and Ti5Si3 grains of an average diameter of 0.86 mu m partially mixed with plate-like phase of T3SiC2. The compact shows elongation of 130% at tensile test at 1773 K with an initial strain rate of 4.0 x 10(-4) s(-1) with the strain rate sensitivity exponent of 0.38. Most grains remain equi-axed after tensile test, while random orientations of plate-like phases of Ti3SiC2 turn parallel along the tensile axis. The large elongation observed above 1773 K is likely due to the superplasticity, while rotation of the plate-like phase accompanies deformation of the surrounding matrix.

Role of microstructures on the improved fracture toughness has been analyzed for the heat affected zone of weldment in Ti-killed steel. A microstructural features of Ti-killed steel is decrease of coarse bainitic ferrite, alpha(B) degrees, and increase of the mixed area of alpha(B) degrees and quasipolygonal ferrite, a(q), besides the presence of intragranular ferrite. The brittle fracture initiation is not from large particles or inclusions, but at the intersections of small alpha(B) degrees grains with different orientations in a mixed area of alpha(B) degrees and alpha(q) in the proximity of boundary between coarse alpha(B) degrees. Ductile crack growth resistance represented by the slop of R-curve is also higher in Ti-killed steel compared with Al-killed steel. Together with observation of deformed microstructures, it is discussed that the decrease of coarse alpha(B) degrees grains in Ti-killed steel suppresses brittle fracture initiation by reducing plastic constraint and the increase in the mixed area of alpha(B) degrees and alpha(q) enhances ductile crack growth resistance by accommodating large strain.

In 0.5 M NaCl solution, the early stages of pitting corrosion of Type 301 stainless steel which is wet-ground on silicon carbide papers ranged from 240 grit to 400, 800, 1000 and 1500 grits have been studied by AC impedance method. It is found that the Warburg impedance coefficient (sigma), which is calculated from Nyquist impedance plots, decreases with increasing grit number of silicon carbide paper for final surface grinding. This is because that the number of metastable pits on a smooth surface is less than that on a rougher surface. The change in E-m (the potential at which metastable pit or pits start to grow on the steel) relates with the change in grit number of silicon carbide paper. The higher the number of the silicon carbide paper, the higher the E-m value. This fact suggests that metastable pit or pits starting to grow on the smoother surfaces is more difficult than that on rougher surfaces. (C) 1997 Elsevier Science Ltd.

Structural changes in amorphous states during prolonged milling after the completion of amorphization have been investigated through the crystallization behaviors with Ni-Ti alloys of compositions corresponding to the intermetallic compounds. In 33Ni-67Ti, the crystallization to NiTi2 takes place in two steps, a higher crystallization temperature dominating after prolonged milling. In 50Ni-50Ti, structural relaxation leading to the crystallization to NiTi2 and Ni3Ti takes place. In 30Ni-70Ti, which is close to the border of the amorphization range, the initially formed amorphous phase turns into a supersaturated solid solution of Ti in Ni. (C) 1997 Elsevier Science S.A.

The early stages of pitting corrosion of Type 430 stainless steel in 0.5 M NaCl solution containing 0, 0.01, 0.05 and 0.1 M Na2SO4 have been studied by an AC impedance method. It is found that the Warburg impedance coefficient (sigma), which is calculated from Nyquist impedance plots, decreases with increasing Na2SO4 in the solution at a given potential in passive region. This is because that increasing SO42- concentration in NaCl solution decreases the total number of surface sites available for metastable pits. The change in E-m (the potential at which metastable pit or pits start to grow on the steel) is a function of SO42- concentration in NaCl solution. The values of E-m in millivolts are linear with logarithm of SO42- concentration in 0.5 M NaCl solution, following the equation, E-m=59 log(SO42-)-114. This fact suggests that increasing SO42- concentration in NaCl solution causes the shift of not only pitting potential but also E-m to more positive potentials. (C) 1997 Elsevier Science Ltd.

The load drop or pop-in phenomenon at the fracture toughness test of the heat affected zone (HAZ) of a low carbon steel has been investigated. Microstructural features such as the size of the bainitic ferrite were varied with the thermal cycles simulating a double pass welding. It is revealed that the initiation of the load drop is not dependent on test temperatures, whereas the arrest of the load drop decreases with the decrease of the test temperature, thus increasing the amount of the load drop. Fractographic examination has revealed that the amount of the load drop scales with the crack length formed by the local brittle fracture. Also, the size and density of cleavage facets increases with the load drop. It is discussed that the temperature dependent amount of a load drop is determined by the ductile linking of cleavage facets produced at bainitic ferrite grains.

The effect of chloride concentration on early stages of pitting for Type 304 stainless steel has been studied by using an AC impedance method. The Warburg impedance coefficient, which is calculated from Bode plots, increases with increasing chloride concentration at low potentials in the passive region when the diffusion process begins to occur at the surface. For a pit which is nucleated under a given potential, there exists a minimum chloride concentration above which the pit on the surface of the steel can be activated into metastable propagation, and below which it cannot. It is also found the effect of chloride concentration is reflected qualitatively in the potential (E(m)) at which the metastable pit or pits start to grow on the surface of Type 304 stainless steel. The values of E(m) in millivolts linearly increases with the logarithm of chloride concentration of solution, following the equation, E(m) = -37 log(C1(-))-212. (C) 1997 Elsevier Science Ltd.

As a new type of material, the nanocrystalline material has drawn world-wide attention in recent years. As a challenge subject, much work has been done concerning the grain size dependence of mechanical properties, e.g., on the Hall-Petch relationship of these new materials. The currently-known Hall-Petch relations in the nanocrystalline material were examined in this work with an emphasis on crystallized nanocrystalline materials. Nanocrystalline materials usually show decreased microhardness compared with the prediction from the Hall-Petch relationship and even a negative Hall-Petch slope. The possible mechanisms for above phenomena were analyzed in relation to the quantitative measurements on the microstructures of crystallized nanocrystalline materials by means of positron annihilation spectroscopy and X-ray diffraction.

Composites of gamma-TiAl intermetallic compound and a complex carbide (Ti2AlC) have been formed by mechanical alloying of Ti-Al-heptane system followed by hot consolidation. The amount of the carbide can be controlled by the milling parameters. The composites have fine microstructures of an average grain size of a few hundreds nm. Mechanical properties at elevated temperatures have been investigated by tensile tests. At 1073K, different deformation behaviors depending on the carbon content are distinct, while the total elongation is still small. The total elongation of specimen tested at 1273K and at a strain rate of 1x10(-3) s(-1) exceeds 300%.

In situ synthesis of fine grain composites of TiN and Ti5Si3 has been conducted by the reaction milling of Ti and Si3N4 powder mixtures followed by hot-consolidation. Formation of TiSi2 could be suppressed by increasing the Ti content in the starting powder mixtures to more than 75 wt.%. Full densification was attained at sintering temperature of 1300 degrees C using Hot-Isostatic-Pressing. The mean grain size in the compacts was hundreds nm when HIP treated at 1300 degrees C. Tensile strength at high temperatures was improved by the elimination of TiSi2. Even in this case, substantial plastic deformation appeared at testing temperatures over 1300 degrees C, and the elongation at 1500 degrees C exceeded 100%. The strain rate exponent of the flow stress at 1400 degrees C and at strain of 0.1 was 0.47. It indicates intrinsic superplasticity in this type of ceramics composites.

The trapping sites of hydrogen detected by Thermal Desorption Spectrometry (TDS) have been identified by Secondary Ion Mass Spectrometry (SIMS). High-strength steel specimens were given applied loads in D2O and 20% NH4SCN solution at 323 K. In SIMS analysis, deuterium ions can be detected with greater sensitivity than hydrogen ions and measurement can be started in a matter of minutes. TDS analysis shows that hydrogen thermal desorption rate has two peaks, corresponding to trap activation energies of 20.1 similar to 22.5 kJ/mol and 82.2 similar to 87.4 kJ/mol for the lower and higher temperature peaks, respectively. These Values are close to 26.8 kJ/mol reported in the literature as the trap activation energy for desorption from dislocations and greater than or equal to 72.3 kJ/mol from the interfaces of inclusions and precipitates. By applying SIMS image-analysis to specimens cooled after the respective peak temperatures, we could identify the trapping sites corresponding to the lower temperature peak as sites within the matrix such as defects including dislocations, and those corresponding to the higher temperature peaks as inclusion interfaces, precipitate interfaces, and segregation bands of phosphorus. These SIMS results confirm the location of the trapping sites to be the same as those estimated from trap activation energy by TDS.

A new method for improving delayed fracture strength by the precipitation of ferrite along prior austenite grain boundaries has been demonstrated with high strength steels of 1300 MPa in tensile strength. Intergranular fracture which is characteristic to the delayed fracture of tempered martensitic steels is suppressed and is changed to quasi cleavage by the intergranular ferrite precipitation. Hydrogen thermal desorption analysis has revealed that the amount of diffusive hydrogen increases under loading. The mean hydrogen occlusion rate rather than the hydrogen absorption under no loading or the amounts of hydrogen at the time of failure corresponds to the delayed fracture strength levels. The mechanism of the effect of intergranular ferrite on the hydrogen embrittlement based on the accumulation of lattice defects has been discussed.

The changes of electrochemical parameters on the passivated electrode of type 304 stainless steel in a 0.5 M NaCl0 solution at different low potentials ranging From -250 to -100 mV (SCE) in the passive region are studied using the AC impedance method. It is found that the electrode reaction is under metal charge-transfer control at potentials below 200 mV, and the charge-transfer resistance decreases with increasing potential. It is possible that pit nucleation occurs on the passivated stainless steel at this stage. When the potential is above -200 mV, because of the diffusion process appearing in Bode impedance plots, both the metal charge-transfer and diffusion processes occur on the electrode. The metal charge-transfer resistance decreases and the diffusion coefficient increases with increasing potential. The nucleated pits on passivated stainless steel proceed to the stage of metastable growth at the potentials above -200 mV. This potential of -200 mV can be estimated as the level at which metastable pits start to grow on passivated type 304 stainless steel. Copyright (C) 1996 Elsevier Science Ltd

Charpy toughness of the heat-affected zone (HAZ) of weldment of a low carbon steel has been investigated by means of an instrumented Charpy test and fractographic analysis. Microstructures were varied with thermal cycles simulating double-pass welding. The ductile-brittle transition temperature is the most deteriorated at an intermediate second-cycle heating temperature. The origin of the difference in the transition temperatures has been analyzed to exist in the brittle fracture initiation stage. Fractographic examination correlating with microstructural features has revealed that the brittle fracture initiation site is associated with the intersection of bainitic ferrite areas with different orientations rather than the martensite-austenite constituents. The role of the constraint of plastic deformation on the brittle fracture initiation is discussed.

The effect of chromium concentration in the surface film formed by surface treatments on the corrosion resistance of Type 430 stainless steel was studied. It was found that the chromium content depends on the surface treatment method. HNO3 treatments result in a significant enrichment of chromium in the surface film on the Type 430 stainless steel, and the chromium enrichment increases with increasing HNO3 concentration. The pitting potential (E(p)), corrosion potential (E(c)) and polarization resistance (R(p)) are changed by the chromium enrichment (S-Cr) in the surface film formed by treatment. The higher the S-Cr value, the higher the E(p), E(c) and R(p) values. Copyright (C) 1996 Elsevier Science Ltd.

The trapping sites of hydrogen detected by Thermal Desorption Spectrometry (TDS) have been identified by Secondary Ion Mass Spectrometry (SIMS). High-strength steel specimens were given applied loads in D2O and 20% NH4SCN solution at 323 K. In SIMS analysis, deuterium ions can be detected with greater sensitivity than hydrogen ions and measurement can be started in a matter of minutes. TDS analysis shows that hydrogen thermal desorption rate has two peaks, corresponding to trap activation energies of 20.7∼22.5 kj/mol and 82.2∼87.4 kj/ mol for the lower and higher temperature peaks, respectively. These values are close to 26.8 kj/mol reported in the literature as the trap activation energy for desorption from dislocations and ≧72.3 kj/mol from the interfaces of inclusions and precipitates. By applying SIMS image-analysis to specimens cooled after the respective peak temperatures, we could identify the trapping sites corresponding to the lower temperature peak as sites within the matrix such as defects including dislocations, and those corresponding to the higher temperature peaks as inclusion interfaces, precipitate interfaces, and segregation bands of phosphorus. These SIMS results confirm the location of the trapping sites to be the same as those estimated from trap activation energy by TDS.

The reaction process during milling of Zr and Ti with n-heptane has been investigated. Formation of a metastable intermediate phase characterized by fee-like structure has been revealed at a certain milling time. It disappears upon heating with reversible appearance of the original hcp structure. The amounts of H and C in the metal powders increase with milling time to a constant level at which stabilization of the intermediate phase takes place. The intermediate phase is presumably the stacking fault of hcp structure stabilized by H and C atoms as the precursor of the stable carbides. Dissociation of H from heptane precedes that of C. Site occupation is competitive between H and C, and stabilization of the intermediate phase during milling takes place by the increase of C content excluding H to sites with lower binding energy with H.

The fracture process at ductile-brittle fracture transition region of low carbon steels has been compared for J-integral tests and instrumented V-notch Charpy tests. Three steels oi ferrite-pearlite structure, A and B, and of mixed structure with bainite, C, were employed. Only differences of the compositions are respective Ni to Mn ratios, and characteristic differences in the ductile-brittle transition behaviors such as the transition temperature, R-curve and brittle fracture initiation stage have been observed. From the analysis of the instrumented test a Charpy energy is decomposed into the energies dissipated in involved fracture processes. On contrary to steel A and B, decreases of the energy transition temperature and of temperature dependence of stretched zone width at Charpy test compared with at the fracture toughness test are observed with steel C. These differences are originated in Region II where blunting of the notch tip proceeds associated with the extension of the stretched zone, Referring to the stress analyses, the estimated local stress at the brittle fracture initiation site of Charpy tests is about a half of that oi J-integral tests. The inconsistency of the transition temperature takes place for steel C which shows pop-in al the J-integral test. It is discussed that the inconsistency of the transition temperature is caused by the suppression of pop-in at Charpy test because of its low local stress.

Structural relaxation in amorphous Ni50Ti50 alloy prepared by mechanical alloying has been investigated by differential scanning calorimetry and X-ray diffraction analysis. An exothermic DSC peak originating from structural changes in the amorphous state was observed with alloys milled after completion of amorphization. The peak was characterized by the increase in the intensity with milling time and reduced heating rate, and also with accompanying formation of phases such as NiTi2 and Ni3Ti with compositions different from the original one after crystallization. The kinetics of the reaction taking place at the peak was expressed by the Johnson-Mehl-Avrami equation with the Avrami exponent of 0.57. The origin of the peak was discussed with respect to the free energy diagram of Ni-Ti system. Structural relaxation was considered to take place at the peak to stabilize the configuration in the as-milled alloy, and to enhance phase separation or compositional fluctuation during milling.

The fracture process of low carbon steels have been investigated by means of a fracture mechanics test and microscopic observations. Three steels with characteristic differences in the ductile-brittle transition behaviors such as the transition temperature, R-curve and brittle fracture initiation stage are employed. A new microstructural parameter, the constraint factor, is defined to take into account the constraint of plastic deformation by grain boundary carbides. The constraint factor is successfully correlated with the resistance against the stable crack growth. The crack tip field is analyzed in terms of the elastic-plastic analyses in terms of J/delta(y) and COD, and the elastic-plastic analyses are found to be applicable at the ductile-brittle transition region. The brittle fracture initiation is controlled either by the local stress or by the local strain for respective steels. The constraint factor is also correlated with the location of the brittle fracture initiation sites. The estimated local stress at the brittle fracture initiation site shows a unique temperature dependence irrespective of the microstructure of steels. It is suggested that the nucleation of an incipient crack at the brittle initiation site is microstructure dependent through a local deformation behaviors.

Recent results on reaction milling have been reviewed focusing on metal-hydrocarbon and metal-ceramics systems. An intermediate metastable phase during the reaction was first revealed with Ti-heptane system. It was shown that stacking faults in Ti introduced by milling is stabilized by C and H which are decomposition products of heptane. Formation of a metastable Ni3C which has a positive heat of formation was revealed at the milling of Ni with SiC. An incubation period for the reaction was observed and related to a threshold value of the stored energy during milling as the driving force of the reaction. The reaction milling of Ti-Al-heptane and Ti-Si3N4, Ti-SiC systems followed by hot consolidation could produce fine composites of TiAl-Ti4Al2C2, TiN-Ti silicides and TiC-Ti silicides. The phases present in the compacts were altered with the process parameters. A kinematical model for the preference of the phases was presented. Improved sinterbility and superplastic-like ductility at elevated temperatures have been demonstrated.

The fracture process that determines the Charpy energy at the ductile-brittle transition region was investigated by means of the instrumented Charpy test and fractographic analysis with a low carbon low alloy steel subjected to different control-rolling conditions. The decomposition of a Charpy energy into the energies dissipated in the course of the notch-tip blunting, stable crack growth, and brittle crack propagation is unique irrespective of the testing temperatures and specimen series. Toughness level can be divided into four regions according to the predominating fracture process. The;temperature dependence of toughness and effects of the anisotropy of a specimen originates in the brittle fracture initiation stage rather than the resistance against the notch-tip blunting or stable crack growth. From fractographic examination referring to the stress analyses, it is discussed that the brittle fracture initiation is controlled by the local deformation microstructures in the plastic zone together with the stress field ahead of the notch or the stable crack front.

Structural relaxation in amorphous Ni50Ti50 alloy prepared by mechanical alloying has been investigated by differential scanning calorimetry and X-ray diffraction analysis. An exothermic DSC peak originating from structural changes in the amorphous state was observed with alloys milled after completion of amorphization. The peak was characterized by the increase in the intensity with milling time and reduced heating rate, and also with accompanying formation of phases such as NiTi2 and Ni3Ti with compositions different from the original one after crystallization. The kinetics of the reaction taking place at the peak was expressed by the Johnson-Mehl-Avrami equation with the Avrami exponent of 0.57. The origin of the peak was discussed with respect to the free energy diagram of Ni-Ti system. Structural relaxation was considered to take place at the peak to stabilize the configuration in the as-milled alloy, and to enhance phase separation or compositional fluctuation during milling.

Deformation microstructures of a low carbon steel with ferrite-pearlite structures have been examined with tritium autoradiography and thermal desorption spectroscopy. The amount of non-diffusible tritium at room temperature increases with strain. When deformed at room temperature which is in the ductile fracture region, the distribution of tritium is fairly uniform within a grain, while it is inhomogeneous when deformed at -80 degrees C which is in the brittle fracture region. In the latter case, the accumulated sites are distributed regularly in arrays and localized as bands. Comparison with slip band traces indicated that the distribution is likely along slip bands. The amount of non-diffusible tritium which desorbs at 180 degrees C is remarkably large when deformed at -80 degrees C. Desorption at 180 degrees C disappears when the deformed sample is annealed at 500 degrees C, suggesting that the defects as the accumulation sites are microdefects presumably point defects or their clusters.

Chemical short-range order (CSRO) in amorphous CuxTi100-x samples prepared by mechanical alloying of elemental metal powders and mechanical grinding of intermetallic compounds has been examined by EXAFS. Curve fitting for the Fourier filtered first shell of the Fourier transform of EXAFS was conducted to determine the Cu coordination numbers. The Cowley CSRO parameter deduced from the coordination numbers shows a monotonic decrease with increasing Cu concentration. It is positive in the range of x &lt; 68, and negative for x &gt; 68. This behavior is quite different from the results of rapidly quenched Cu-Ti amorphous alloys whose CSRO parameter is negative for all compositions. The ratio of coordination numbers of Cu and Ti atoms, N-Cu/N-Ti, increases linearly for x &lt; 57 and saturates to 2.1 for x &gt; 57. The origin of CSRO in mechanical alloyed samples is probably due to inter-diffusion of Ti into Cu, while CSRO of samples derived from the liquid state is fixed during rapid quenching.

Mechanical alloying of elemental Ti and Si3N4 powders was conducted by high energy ball milling and solid state reactions were observed to take place during milling. An incubation period for the reaction to begin existed which was related to the critical stored energy during milling. Hot compaction of milled powders could be efficiently conducted at lower temperatures than for conventional sintering of Si3N4. The phases formed after sintering varied with the milling time and sintering temperature and a kinematical model has been proposed for the mechanism of phase formation.
The grain size of samples sintered at 1300-degrees-C was about 100 nm and grain growth took place at increasing sintering temperatures preferentially at Ti5Si3 grains. The Vickers hardness decreased monotonically with increasing test temperature, but the bending strength showed an inverse temperature dependence. Substantial plastic deformation took place at over 900-degrees-C and superplastic behaviour was observed.

The fracture process that determines the temperature dependence of toughness at the ductile-brittle transition was investigated for low carbon steels. It was revealed that the resistance against the stretch zone at the crack tip and the stable crack extensions is temperature independent while it depends on the carbon content. Applicability of elastic-plastic analyses of the crack tip fields was examined in terms of the relationship between J integral value normalized by the flow stress and crack opening displacement. Temperature dependence of toughness is determined primarily by the brittle fracture initiation. Locations of the initiation sites were revealed to coincide with the maximum stress triaxiality rather than the maximum tensile stress. It was suggested that the brittle fracture is caused by the deformation-induced initiation and triaxial stress assisted growth of an incipient crack.

Deformation microstructures of a low carbon steel with ferrite-pearlite structures have been examined with tritium autoradiography and thermal desorption spectroscopy. The amount of non-diffusible tritium at room temperature increases with strain. When deformed at room temperature which is in the ductile fracture region, the distribution of tritium is fairly uniform within a grain, while it is inhomogeneous when deformed at -80°C which is in the brittle fracture region. In the latter case, the accumulated sites are distributed regularly in arrays and localized as bands. Comparison with slip band traces indicated that the distribution is likely along slip bands. The amount of non-diffusible tritium which desorbs at 180°C is remarkably large when deformed at -80°C. Desorption at 180°C disappears when the deformed sample is annealed at 500°C, suggesting that the defects as the accumulation sites are microdefects presumably point defects or their clusters.

Synthesis of Sm2Fe17Nx was conducted by mechanical alloying mixtures of intermetallic compounds Sm2Fe17 and SmFe3. Effects of the mixing ratio and nitriding conditions on the magnetic properties were examined. Milled powders were subjected to a stabilizing treatment followed by annealing and nitriding. The nitriding temperature, holding time and N2 pressure were systematically varied. Milling of mixtures of intermetallic compounds gave an amorphous Sm-Fe alloy and alpha-Fe as in the case of milling of elemental metal powders. Optimum magnetic properties were obtained at a composition Sm15Fe85 at which the precipitation of alpha-Fe, SmO1-xNx and SmFe3 was minimum after nitriding. The increase in the coercivity was parallel to the nitrogen content which increased drastically and then gradually as the nitriding time was extended. It was noted that lattice constants of nitride were constant irrespective of nitrogen content, suggesting that the nitriding conditions affect mainly through the amount of nitride. The optimum nitriding temperature for coercivity was 450-degrees-C. The precipitation of alpha-Fe increased at higher temperatures, while the absorption of nitrogen decreased at lower temperatures. The nitrogen pressure was insensitive to the absorbed nitrogen content. The existence of alpha-Fe together with SmO1-xNx and Sm2O3 may affect the magnetization curve.

Crystallization kinetics of amorphous Cu50Ti50 alloy prepared by mechanical grinding of the intermetallic compound powder with a high energy ball mill was studied. Kinetics at the isothermal annealing was analyzed in terms of the Johnson-Mehl-Avrami equation, giving a velocity exponent of 0.85 and an activation energy of 183 kJ/mol. Both figures are less than the corresponding values of material prepared by a rapid cooling process. Kinetics at the athermal annealing was analyzed using the Kissinger method, and an activation energy of 154 kJ/mol was obtained in accord with analysis with the JMA equation. The amorphous state and crystallization process are discussed. In an amorphous material prepared by mechanical grinding, the crystalline fraction remains and is likely to act as nuclei of crystallization. Crystallization is likely to proceed two-dimensionally on these nuclei. The number and average size of nuclei are estimated to be 1024m3 and of 103 unit cells of Cu50Ti50, respectively. The activation energy of diffusion estimated by assuming two-dimensional crystal growth kinetics is 214 kJ/mol and the diffusion of Ti atoms is likely to be rate-controlling. Atom migration necessary for the crystallization is considered to take place in a short range on the order of 5-10 nm.