日本金属学会誌
Online ISSN : 1880-6880
Print ISSN : 0021-4876
ISSN-L : 0021-4876
55 巻, 7 号
選択された号の論文の13件中1~13を表示しています
  • 岡崎 義光, 市川 冽, 松尾 宗次
    1991 年 55 巻 7 号 p. 729-738
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    An investigation was made on the effct of stirring on the microstructures, distribution of solute elements, carbide formation, phosphorus segregation and distribution coefficients between solid and liquid phases in the the Ni-base Superalloy TMP-3 (Ni-11.15%Cr-6.74%Co-3.12%Mo-3.98%Al-2.74%Ti-3.93%Nb-3.25%W) by rheocasting with the ratation of a stirrer at a speed in the range of 20 to 70 rev/s. The main results obtained are summarized as follows:
    (1) When the TMP-3 superalloy was sufficiently agitated at 70 s−1 from the start of solidification, the dendrite structure was broken up and equiaxed grain structure with the average primary solid particle size of 65±15 μm was formed.
    (2) The distributions of Ni, Cr, Co, Al, Ti, Mo, Nb, W, P and C in the superalloy TMP-3 ingots rheocast at 70 s−1 and stationarily cast were measured by using a computer-aided microanalyzer (CMA). The concentrations of main eight metallic elements in the center of equiaxed grains in the rheocast alloy were almost equal to those in the center of dendrites in the stationary cast alloy.
    (3) It was observed from the measurement on the distribution of Ti and Al that the size of γ′ phases in the rheocast alloy tend to decrease as compared with that in the stationarily cast one.
    (4) It was observed that granular MC carbides containing mainly Nb, Ti and Mo precipitate uniformly along the grain boundary in the rheocast alloy, although the plate-like MC carbides precipitate in the grain boundary of the stationary cast alloy.
    (5) The distribution of P measured by CMA showed that P segregates considerably in the grain boundary, but the degree of phosphorous segregation in the rheocast alloy tend to decrease as compared with that in the stationary cast one.
    (6) The solid-liquid distribution coefficients estimated from the concentration of main eight elements in the center of primaly crystals were almost consistent with the equilibrium distribution coefficients obtained by thermodynamic calculation including an additional consideration on solute interactions.
  • 佐藤 知敏, 酒井 洋典
    1991 年 55 巻 7 号 p. 739-747
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    To clarify the mechanism of normal grain growth and the preferred order of migration among grain boundaries, computer simulation of two-dimensional normal grain growth was carried out for the grain structures generated by the constant nucleation rate and growth model.
    Five kinds of processes which are different in the preferred order of grain boundary migration were investigated, and grain structures after each simulation time were compared in terms of the grain size distribution, distribution of number of sides per grain, average morphological characteristic values, average grain diameter, etc. The five kinds of processes are: (1) large grains grow against the neighboring smaller ones in order of largeness, (2) small grains shrink against the neighboring larger ones in order of smallness, (3) grains, if smaller than the neighboring ones, shrink, and if larger, then they grow, and both happen simultaneously in random order. (4) grains grow only against the neighboring smaller ones in random order, and (5) grains shrink only against the neighboring larger ones in random order. The results are summarized as follows:
    (1) Within the present simulation times, the average morphological characteristic values were almost the same in all of these five processes.
    (2) The average grain area increased slightly in the first few simulations, then remarkably, and as more simulations were made, they increased slowly again.
    (3) The logarithm of the average grain diameter decreased slightly at first for the logarithm of simulation time, but then it increased at a constant rate. The reciprocal grain growth exponent was 0.42 for the arithmetic mean diameter and 0.53 for the geometric mean diameter.
    (4) In every simulation, the distributions of grain diameters were approximately log-normal in shape, while the distributions of numbers of sides per grain were normal in shape.
    (5) The distributions of relative grain diameters and number of sides per grain was time-invariant.
    (6) In the processes (3) and (4), the frequency of number of sides per grain was maximum at 5 sides in the first several simulation times, but later it became maximum at 6 as the theory predicts.
  • 原勢 二郎, 清水 亮, 中島 正三郎
    1991 年 55 巻 7 号 p. 748-755
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    (1) Two texture groups A and B are evolved from the primary texture composed mainly of {334} ⟨6 10 13⟩, {111}⟨012⟩ and {434}⟨5 12 4⟩ oriented grains in primary recrystallized Fe-3%Si containing MnS and AlN.
    The group A is composed mainly of Goss oriented grains rotated around the ND axis about 15° and 20°, while the group B has orientations centering around {335}⟨110⟩±10° rotated around the ND axis. Grain diameter of the group A is larger than that of the group B although the group A has an intensity level of about 30% of that of the group B. The intensity level of the groups A and B before grain growth were about between 15 and 20% and between 9 and 30% of the intensity of the main texture component before grain growth, respectively.
    (2) The mechanism of the evolution of A group is considered to be due to the higher frequency of adjacent grains in the Σ9 coincidence orientation relationship and their higher mobility under precipitates rather than the size effect as their grain size is smaller than the average grain size before grain growth. The mechanism of the evolution of the group B is considered to be due to the higher frequency of adjacent grains in the Σ9, Σ19a and Σ29b coincidence orientation relationships and their higher mobility under precipitates rather than the size effect as their grain size is smaller than the average grain size before grain growth.
    It is concluded from the above results that even under a weaker intensity level of inhibitors, the texture evolved by grain growth is more affected by the presence of coincidence boundaries rather than the effect of grain size before grain growth.
  • 増野 茂美, 堀谷 貴雄, 鈴木 洋夫, 岸 輝雄
    1991 年 55 巻 7 号 p. 756-764
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    Microfracture processes of Ti-6Al-4v alloys with various acicular α structures were investigated by acoustic emission (AE) source location, source characterization and fractography.
    The specimen with a coarse acicular α structure showed higher fracture toughness than with a fine acicular α structure.
    AE source characterization showed that two types of microcracks were generated before the final unstable fracture. One was a smaller microcrack with a higher propagation velocity, and the other was a larger microcrack with a lower propagation velocity.
    In the fracture process, the microcracks of the former type were generated around the fatigue pre-crack front in the early stage of loading. The main crack extension occurred when the microcracks of the former type coalesced. The latter corresponded to the coalescence of the former.
    In the specimen with a coarse acicular α structure, microcracks distributed widely out of the fatigue pre-crack plane and the main crack extension was hindered or distorted by phase boundaries.
    It was concluded that both the distribution of microcracks and the resistance of acicular α structure to the coalescence of microcracks had a significant influence on the fracture toughness of this alloy.
  • 青木 茂
    1991 年 55 巻 7 号 p. 765-772
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    Silicon mesa stripes and columns with a few 100 nm in size are fabricated on (100) Si wafer surfaces by a dry etching technique, and then thermally oxidized. Shapes of oxidized regions, stress distribution and dislocation generation in the regions are analyzed by means of transmission electron microscopy (TEM). In the oxidation of Si wafers having uneven surfaces, oxidation rate reduction due to the compressive stress within SiO2 occurs, and dislocations sometimes are observed with unusual structures. Such stress distribution in a sub-micron region can be analyzed by thickness measurement (measurement of thickness fringes) of the TEM sample, because the Ar ion etching rate in the sample preparation is anomalous in the stressed reagion.
  • 吉武 道子, 吉原 一紘
    1991 年 55 巻 7 号 p. 773-778
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    The diffusion behavior of titanium (Ti) in copper (Cu) films was studied. Cu films were deposited on Ti substrate by the r. f. magnetron sputtering method.
    When the deposited Cu film was heated above 630 K in a vacuum, the substrate element Ti diffused through the Cu film very rapidly and concentrated on the surface of the Cu film. The diffusion coefficient was obtained by measuring the time required to diffuse. The pre-exponential factor of the diffusion coefficient was 3.5×10−5 m2 s1, and the activation energy was 127 kJmol−1. The obtained activation energy is about 65% of that of Ti in the bulk Cu.
    Above 800 K, after Ti diffused rapidly onto the surface of the Cu film, intermetallic compounds were formed at the interface between the Cu film and the Ti substrate. The intermetallic compounds formed were γ-TiCu, Ti3Cu4 and TiCu3. It has been reported that Ti2Cu, γ-TiCu, Ti3Cu4, Ti2Cu3, TiCu2 and Ti2Cu7 are formed from ‘bulk cu’-‘bulk Ti’ couple. In this work, from the ‘film Cu’-‘bulk Ti’ couple, we did not observe the formation of Ti2Cu, Ti2Cu3, TiCu2 and Ti2Cu7. We observed the formation of TiCu3 which is not formed from the ‘bulk Cu’-‘bulk Ti’ couple.
  • 田辺 晃生, 千葉 幸文, 朝木 善次郎
    1991 年 55 巻 7 号 p. 779-785
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    Oxidation of dense plates of mixed Fe-Co sulfides whose compositions were CoS-10 mass%FeS and CoS-50 mass%FeS was conducted at 973, 1023 and 1073 K in a 20 vol%O2-N2 gas mixture. Mass change of the samples and SO2 concentration of the off gas were continuously measured during the oxidation. Examination of the oxidized samples was carried out by using SEM, EPMA and XRD.
    Two types of reactions proceeded during the oxidation; the one is the reaction of forming a dense oxide layer without the evolution of SO2 gas resulting in a mass gain of the sample, and the other is the reaction of forming a porous oxide with the evolution of SO2 gas resulting in a mass loss. The dense oxide formed on the sample surface was CoxFe3−xO4 and the porous oxide formed in the interior of the sample was composed of Fe3O4 and CoO. The concentration profiles of Fe and Co in CoxFe3−xO4 layer were proved to be uniform. The layer thickness grew in accordance with the parabolic rate law, with an activation energy of 145 kJ·mol−1.
  • 田辺 晃生, 西浦 正孝, 朝木 善次郎
    1991 年 55 巻 7 号 p. 786-795
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    Oxidation of dense plates of Cu5Fe1−xS4−y in which the (1−x)-value varied from 1.01 to 0.13 was carried out at 973, 1023 and 1073 K in the O2-N2 gas flow. Mass change of the samples and the concentration of evolved SO2 gas were continuously measured during the oxidation.
    The mass change of the sample Cu5Fe1.01S3.77 exhibited that the reaction proceeded through following 3 steps. In the early stage of oxidation, the sample mass increased without the evolution of SO2 gas; a dense layer composed of Fe2O3 and Fe3O4 was formed on the sample surface. The oxidation rates in the 1st and 2nd steps of oxidation were thought to be controlled by the diffusion of Fe in the sulfide and in the oxide, with activation energies of 145 and 160 kJ/mol, respectively. In the 3rd step, SO2 gas evolved and the sample mass decreased to form porous oxides of Fe2O3, Cu2O and CuO. Estimation of the sulfide composition at the maximum mass revealed that the composition of each sample was around the tie line between Cu2S and Cu5FeS4.
  • 下尾 聰夫, 木矢 拓見, 岡村 清人
    1991 年 55 巻 7 号 p. 796-804
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    In relation to the joining of silicon nitride ceramics to heat-resistant alloys, the mechanism of reaction of Si3N4 with Ni has been studied. The powdered mixture of Si3N4-Ni has been heated at temperatures from 973 to 1573 K in Ar or N2 gas stream. The rate of reaction was determined by thermogravimetric analysis. The reaction product was investigated by X-ray diffraction and SEM observation.
    Si3N4 reacts with Ni to form Ni-Si solid solution at the lower temperatures. At the temperatures higher than 1273 K in Ar gas stream and than 1323 K in N2 gas stream, the nickel silicide, Ni31Si12 is formed. The reaction rate is independent on the atmosphere. The rate increases with an increase in the amount of Ni in the powdered mixture and with a decrease in the particle size of Ni.
    Jander equation was applicable to the rate of the reaction at temperatures 1273 K-1423 K. The apparent activation energy is 329 kJ/mol. The reaction of Si3N4 with Ni is controlled by the solid state diffusion through the product layer formed around Ni particles.
  • 若松 良徳, 柴下 泉, 下崎 敏唯, 大西 正己
    1991 年 55 巻 7 号 p. 805-812
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    Iron and chromium specimens were annealed in the molten lead saturated with Cr at 1073 to 1373 K for various periods of time, in order to investigate the diffusion behavior in the chromizing process using the molten Pb as a medium for transfer of the Cr atoms to a Fe surface.
    A diffusion layer composed of an α phase was formed on the surface of the Fe specimen. The region having the composition range of γ phase could not be detected in the diffusion layer, although its appearance was predicted from the Fe-Cr phase diagram. Surface concentraion of the α phase layer ranges from 55 to 70 mass%Cr and Cr concentration in the cross section decreases exponentially as the distance from the surface increases. The thickness of the α phase layer increases with annealing time according to a parabolic rate law and temperature dependence of the rate constants can be expressed approximately by an Arrhenius equation.
    A very thin layer of α phase having the surface concentration of 60-75 mass%Cr was formed on the Cr specimen. This result suggests that Fe as well as Cr atoms dissolve in the molten Pb and diffuse in the Cr specimen. It appears that interdiffusion between solid Fe and Cr occurs through molten Pb in the chromizing process using the molten Pb as a transfer medium.
  • 市川 洌, 岡崎 義光, 木下 好司, 石塚 哲
    1991 年 55 巻 7 号 p. 813-819
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    An investigation was made on the microstructure refinement of Cu-37%Al and Al-17%Si alloys by the vacuum stirring synthesis. In this method the molten alloys were allowed to flow down onto a polygonal rotor that rotated at a high speed up to 167 s−1 about the horizontal shaft in a heat-controlled vessel, the rotor spattered the molten alloys, many initial grains formed in the spattered droplets were refined with repeating a collision against the rotor surface and the inner wall of the vessel by the stirring action of the rotor rotating at high speed, and finally the alloys with slurrying of gathered semi-solidified droplets were poured down into a metallic mold through a taphole at the bottom of the vessel. Considerable refinement of microstructure was accomplished by the present stirring synthesis process as compared with the microstructures formed by conventional casting processes. The average grain size was about 20 μm in the Cu-37%Al alloy made by the stirring synthesis at a rotation speed of 167 s−1, although the grain size was 300 to 1000 μm in the alloy conventionally cast without rotation of rotor. The average silicon grain size was about 16 μm in the Al-17%Si alloy made by the stirring synthesis at a rotation speed of 167 s−1, and it was 89 μm in the alloy rheocast at a rotation speed of 70 s−1.
  • 森 信幸, 浮田 求, 大城 桂作
    1991 年 55 巻 7 号 p. 820-829
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    The irfluences of surface morphology and porosity on the wettability of sintered Al2O3 powder or fiber substrates by molten Al, Ag, Pb-Sn alloy and cast iron were investigated by the sessile drop method under a vacuum of 2×10−3 Pa. The apparent contact angles (θ) of molten aluminium on the sintered fiber and lustrous sintered powder substrates were larger than that (θ1) on the transparent Al2O3 substrate with a smooth surface. The measurered θ values were almost equal to those calculated by Cassie’s equation from θ1 and the surface porosity (fvs) which was evaluated from the bulk porosity (fv) of the substrate by a simple surface structure model. θ on the lusterless sintered powder substrate was much larger than that estimated from the above simple structure model, because the surface of the substrate was very rough. The surface porosity (fvs) of the lusterless substrate was correctly estimated by using an image processing system. θ of molten Ag and Pb-Sn alloy to the sintered powder substrates with a lustrous or lusterless surface were almost equal to the angles estimated by Cassie’s equation from fvs. θ of molten cast iron to the sintered substrates was also estimated from fvs of the original unreacted substrate during the first 30 s of wetting time, though the molten iron reacted with Al2O3 occurred and θ changed with time.
  • 中島 正三郎, 高嶋 邦秀, 原勢 二郎
    1991 年 55 巻 7 号 p. 830-837
    発行日: 1991年
    公開日: 2008/04/04
    ジャーナル フリー
    A study has been made on the effect of the increase of Si content on the primarily recrystallized structure, texture and the dispersion of precipitates in the grain-oriented electrical steel produced by the single-stage cold rolling process by use of MnS and AlN as inhibtors.
    (1) The size of primary grains increases. This is considered to be attributed to the larger size and coaser dispersion of precipitates in the specimen with higher Si content.
    (2) The amount of nuclei with Goss texture decreases. This is considered to be attributed to the less amount of free C during cold rolling, which is cosidered to decrease the amount of transition bands in the specimen with higher Si content.
    (3) The size of precipitates increases and their dispersion gets coaser. This is considered to be attributed to the less uniformity of structure and content of elements in the specimen with higher Si content.
    (4) From the three points mentioned above, it is considered that higher content of Si makes secondary recrystallization more difficult.
feedback
Top