Below is a translation of the English text from second Jung paper: Damage Mechanism and Deterioration of Hydrogen Storage Ability on Pd at Hydrogen Absorption-Desorption Multi-Cycles
https://www.jstage.jst.go.jp/article/jsms1963/50/9/50_9_999/_article The rest of the paper, including the abstract and captions are in English. The download is free, so you should download it to see what's what. Again, this was done by translate.google.come and by me. The Google translation program seems to be improving almost day by day. I submitted some of the paragraphs from this paper to Google last week, and again today, and they came out better today! (It is possible the translation improved somewhat because I have gotten better at using the program. The Google program as it applies to Japanese does call for some skill. You have to prepare the text in certain ways to avoid translation errors. I do not know about other source languages.) Here is the text: Damage Mechanism and Deterioration of Hydrogen Storage Ability on Pd at Hydrogen Absorption-Desorption Multi-Cycles Young-guan JUNG * and Yuzuru SAKAT ** † Received Aug. 18, 2000 * Student member Yokohama National University Graduate Student, Yokohama National Univ., Hodogaya-ku, Yokoham, 240-8501, Yokohama-ku Hodogaya-ku Yokohama-shi 240-8501 * * Full member Major Subject of Yokohama National University Faculty of Science and Engineering Yokohama National University Joban Tai, Hodogaya-ku Yokohama-shi 240-8501, Dept. of Materials Sci. Faculty of Eng., Yokohama National Univ., Hodogaya-ku, Yokohama, 240-85011000 [The following abstract and keywords are in English in the original] Damage mechanism and hydrogen storage ability variation of palladium (Pd) were investigated on hydrogen absorption-desorption multi-cycles. In order to study this problem, palladium plates and round bars with mechanical working or annealing have been used. Specimens were hydrogenated cyclically by the electrochemical method, and hydrogen absorption ratio (H/Pd) as well as deformation of specimens were measured at each hydrogenation cycle. As results, it was shown that damage mechanism of Pd specimens was occurred differently with their geometries and mechanical or heat treatment. In thin plate Pd specimens, the thickness increased in increasing hydrogenation cycles whereas length and width decreased, and grains were greatly deformed however damage of micro and macro structures were scarcely observed, and also hydrogen storage ability scarcely changed. On the other hand, in round Pd specimens, both length and diameter increased with increasing hydrogenation cycles, and significant damage of structures were widely observed which resulted in decrease of the ratio H/Pd to -4.2% at the final hydrogenation stage. The damage of structures and deformation were by far intensive in the specimens with mechanical working than with heat treatment however the hydrogen storage ability were not so different within the two type specimens. Key words: Palladium, Hydrogen absorption, Electrochemical method, Hydrogen cycles, Damage, Deterioration [The following text was translated] Introduction In recent years, development of an internal combustion engine that uses clean hydrogen energy and an eco-car using a hydrogen electrode battery has been actively developed as a measure against environmental problems such as air pollution caused by fossil fuel and global warming caused by CO2. Along with this, research and development is also actively being conducted on hydrogen storage materials for hydrogen energy systems along with this. Now then, it is known that when hydrogen is absorbed and released into a metal material, it causes a large expansion / shrinkage deformation. 1) As a result, dislocations and defects occur in the material, hydrogen precipitation occurs in the material, and plastic strain There are accumulated, micronized is promoted micro-cracks occur. 1) ~ 3) As a measure for preventing the material strength deterioration and hydrogen storage performance deterioration due to pulverization and embrittlement, alloying materials. Despite studies on structural element substitution, structural structure alteration, amorphization, etc., materials with superlative lifetime and absorption capacity yet to be developed. 4) ~ 9) Degradation of the strength of the material and degradation of hydrogen storage capacity are caused by plastic deformation of the material due to hydrogen absorption, generation of cracks, progress of pulverization and the like, but unresolved questions remain regarding the basic process of deterioration. This is an obstacle to improving the durability of hydrogen storage material. In this paper, we observed the occurrence of breakdown in the inside bulk and on the surface of the material due to hydrogenation cycle, large deformation and fine fragmentation of crystal grain, hydrogen storage performance of the overall resulting material is considered the degradation process. 2 Test sample and experiment method We used 1 mm thick Pd plate material made by rolling, and 10 mm diameter round bar by drawing process (both manufactured by Nilaco, purity 99.9 mass%), and those were annealed at 600°C for 2 hours. As shown in Fig. 1, the length, size-plate test piece having a width 20 × 10 mm, was used to prepare a round bar-type specimen of length 10 mm. Figures 2 (a) and (b) show structural photographs of the rod cross-section E of the round bar test piece, and the drawn surface D. The crystal grain is horizontally elongated on the D side under the influence of processing. The internal structure contains relatively many large voids and cracks. In the rolled processed flat test piece, described in the previous paper 10), there were not as many voids and cracks in the internal structure. It seems that damage of internal structure is more severe in drawing process than in rolled processing. A structural photo of the flat plate (foil) test piece is shown in the previous paper 10). Figures 3 (a) and 3 (b) show structural photographs of cross section E and the drawn surface D of a round bar test piece subjected to heat treatment. As with the flat plate heat treatment material of the previous report 10), the crystal grains have an aspect ratio of 1:1, but a few large defects thought to be due to drawing are left after the heat treatment, and annealing twin crystals also appeared. However, micro level defects, voids and dislocations received from processing were almost all removed by the heat treatment. Table I show results of the JIS H 0501 linear analysis testing method to determine crystal grain size, and the areal analysis method showing a crystal grain size, calculated by comparing the round specimens (the data for the flat plate test piece came from our previous paper). In this research, we mainly use the areal analysis method to obtain the change in crystal grain size, and we also use the linear method to obtain the direction of deformation of crystal grains by processing and hydrogen absorption / desorption. In the crystal grain size calculation by the areal method for the drawn material, the appearance of crystal grain flattening due to drawing was observed. The grains in the E plane were 43 μm, and 76 μm in the D plane, or about 1.8 times larger in comparison. Also, according to the linear method, Gw and GT in the cross section of the E plane are almost the same values of 39 μm and 37 μm, but on the D plane there is a difference of about 3.5 times from GT (42 μm) and GL (149 μm). In other words, it can be seen that the crystal grain shape of the D plane extends to about 1:3.5 aspect ratio, whereas in the heat treated material, it is almost the same as the D face (57 μm), E face (56 μm) according to the areal method, And the crystal grain shape in the two surfaces measured with the linear method are also almost the aspect ratio as with the areal method. Hydrogen absorption was carried out by the same electrochemical method as in the previous report 10) The electrolysis temperature was 25°C, the atmospheric pressure was applied, the electrolysis time was 30 hours, and absorption and desorption cycles was repeated 20 time with the flat plate test piece, and 10 times with the round bar test piece. There were only 10 cycles with the round bar test piece because, as will be described below, the bar-shaped sample rapidly breaks down, and the original shape is lost after about 10 cycles. After the end of electrolysis, the absorption ratio and dimensional change were measured with the methods described in the previous study 10). 3 Experimental results 3.1 Deformation Behavior of Flat Plate Pd Test Specimen Figure 4 (a), (b) and (c) show the deformation behaviors in the length, width and thickness direction of flat plate specimens (the rolled and heat treated materials), after up to 20 hydrogenation cycles. During absorption, open voids indicate values at the time of release of hydrogen. In the thickness direction, as shown in Fig. 4 (c), both the rolled material and the heat treated material increase in a quadratic curve manner, in the length direction and the width direction, it shows a linear decrease until 20 cycles. Except that the shape of the plate thickness cross section is deformed like a pillow as shown in Figs. 5 and 6, so the dimensions of the central flat part are used. The deformation ratio in the thickness direction in the 20th hydrogenation cycle is 2.8 times and 2.3 times in the rolled worked material and the heat treated material respectively, and in the rolled material, it expands to nearly 3 times the original thickness. The difference between the expansion coefficients of the rolled material and the heat treated material, which were almost the same in the first hydrogenation cycle, also increases as the number of cycles increases. In the final cycle, the deformation in the longitudinal direction is 0.72 times the size for the rolled workpiece and 0.77 times for the heat treated material, and in the width direction, 0.70 times for the rolled material and 0.75 times for the heat treated material. Figures 5 (a) and (b) are photographs showing changes in the shape of the flat test piece (R) before the hydrogenation cycle and the rolled material (L) and the heat treated material (C) after 20 cycles. The difference in deformation between the rolled material and the heat treated material appears as the thickness increases, the width and the length decrease in the test piece. In terms of the thickness direction, the plastic deformation of the rolled material is higher than that of the heat treated material. It seems that hydrogen precipitation occurs in the internal damage (dislocations, micro cracks, voids, etc.) caused by the rolling process, and this deformation is accelerated by pressure. This tendency also appears in the round bar test piece, as described below. Figure 6 is a schematic diagram showing the outline change of a flat plate specimen after up to 20 cycles of hydrogenation. The swelling preferentially expands in the plate thickness direction by hydrogen absorption and shrinks in the length and width direction. Hydrogen release relieves most of the strain in the sample, but part of the strain (the result of a single cycle Experiment #10) leaves about 5% volume strain. With the next hydrogenation cycle, the deformation occurs to the same extent as in the previous cycle, and this accumulates. From the results in Figure 4, it can be said that in the case of plate samples up to about 20 cycles, the amount of plastic deformation for each cycle occurs to about the same extent and it is cumulative. We think that with the flat plate test piece, the plate expands in the thickness direction and the length and the width dimensions contract, because the test piece is a thin plate shape. As a result of lattice expansion due to hydrogen absorption, a strong stress field develops inside the material, but in the thin plate shape, the stress field is in a biaxially constrained state (a plane stress field) in the plane and the constraint degree in the plate thickness direction is low. For this reason, a large expansion deformation in the thickness direction occurs preferentially, so that it we think that the length and the width direction are subject to compression. 3.2 Grain deformation behavior of flat specimens Changes in crystal grains of rolled and heat treated materials after 20 hydrogenation cycles are shown in Figs. 7 and 8. The aspect ratio of crystal grains varies greatly, due to hydrogen absorption. In the rolled material, the crystal grains that were laterally in a long shape before the hydrogenation cycle exhibited an extremely large deformation that became a vertically long shape after 20 cycles. Also, after 20 cycles from almost the same aspect ratio before the test, the heat treated material changed from being almost the same aspect ratio before the test, also deforming largely into vertically long crystal grains. Table II shows the crystal grain sizes at the A and B cross sections of the rolled material and the heat treated material after 20 cycles using the areal method and the linear method. In the case of the rolled material, the Table I data came from our previous report 10). In comparison, the crystal grain size by the areal method does not change very much between the A side and the B side before the test after 20 cycles, whereas with the linear method, the grain size ratio before the test (GL: Gw : GT) is 5: 2: 1 after 20 cycles, the value of 2: 3: 4 was shown and the transversely elongated crystal grains underwent great deformation in a vertically elongated shape as though it was rotated by 90°. In the case of the heat treated material, the crystal grain size by the areal method before the test and after 20 cycles is as small as about 20% on the A side and about 70% on the B side as measured with the linear method. The crystal grain diameter ratio (GL: Gw: GT) before the test was 1: 1: 1, and after 20 cycles it underwent a large deformation of 1: 1: 3, and the crystal grain was transformed into a vertically long shape. Figure 9 shows a schematic of the process of structure deformation by the hydrogenation cycle of the cross section B and C of the flat plate specimen. In 20 rolled strips, after 20 cycles, the crystal grains in the central part of the plate changed from the initial lateral long shape to the vertically elongated shape It seems that the extremely large deformation of the crystal grains which crushes the sample from the width direction and expand even in the thickness direction shows the specificity of the deformation behavior of the material due to hydrogen absorption. As a result of the internal stress generated by hydrogen absorption, the crystal grain undergoes plastic deformation, which is repeated for a plurality of cycles, eventually resulting in large deformation such that the grain shape is completely changed, but macroscopic damage of the structure (voids and cracks) is not observed much. As can be seen from the drawing (Fig. 2), large damage is caused to the internal structure of the material subjected to machining, so that the grain size is largely changed, but in the deformation process by hydrogenation cycle not much macroscopic damage was caused, and large deformation of the crystal grain shape progresses relatively smoothly. Figure 10 shows the relationship between the hydrogenation cycle and the hydrogen absorption ratio in flat plate specimens (rolled and heat treated materials). In the first cycle, the absorption ratio of the heat treated material is somewhat higher, and as the cycles are repeated, the absorption ratio is slight, but it decreases by about 1.5% after 20 cycles. For rolled material as well it decreases to 1.4% as the number of cycles increases. The hydrogen absorption ratio for the first hydrogenation cycle is 0.70 for the rolled material, and 0.71 for the heat treated material. In the final cycle, the ratios are 0.69 and 0.70 respectively, and the deterioration of the hydrogen storage capacity is quite small compared to [considering] the extent of material deformation. 3.3 Deformation Behavior of Round Bar Pd Test Specimen Figures 11 (a) and (b) show deformation behavior of the round bar specimen in the axial direction and diametrical direction from the hydrogenation cycle. In the round bar specimen, large cracks formed on the surface after several cycles, and the breakage of the material rapidly progressed, so the size measurement was limited to 10 hydrogenation cycles. After the first cycle, the expansion coefficient and the residual expansion ratio in the axial direction were 1.053 and 1.003 for the drawn material and 1.056 and 1.005 for the heat treated material. The heat treated material showed a slight deformation, whereas in the diametrical direction, 1.053 and 1.012 for the drawn workpiece and 1.052 and 1.010 for the heat-treated material, so that the expansion of the workpiece was be large even though it was negligible [sic]. In the round bar specimen, there is not much difference in the swelling behavior for each direction observed in the flat specimen, which showed an expansion of approximately 5.2 to 5.6% in one cycle in the diametral direction and the axial direction. However, the expansion deformation of the sample is not uniform, and it shows a gentle barrel-shaped deformation with the maximum at the center (see Figs. 12 and 13). The above expansion ratio uses the maximum dimension at the central portion of the sample. Until the tenth hydrogenation cycle, the coefficient of expansion in the axial direction is slightly larger than that of the drawn material in the heat treated material, and reversed in the diametrical direction. This seems to be caused by the propagation of cracks generated on the surface of the specimen. Figure 12 shows an overview of the drawn material (L) and the heat treated material (C) before the experiment (R) and after 10 cycles. In the processed material, longitudinal cracks appeared on D surface after 3 cycles, and after 5 cycles it penetrated in the vertical direction, forming deep cracks like D cracks. The stress near the surface is alleviated by this surface cracking, and the diameter of the test piece increases almost linearly. Here, the sample diameter after the occurrence of the crack was used by averaging the values at several places of the central section cross section. On the other hand, the coefficient of expansion in the axial direction gradually increases up to 3 cycles in the workpiece, then it becomes nearly constant up to 8 cycles, and rapidly increases from cycle 9 on. This is considered to be due to temporary suppression of deformation in the axial direction as a result of preferential expansion in the radial direction due to occurrence and propagation of the surface crack described above. As described above, macroscopic cracks tend to occur in the material of the worked material, and the deformation behavior becomes nonuniform. On the other hand, in the heat treated material, no occurrence of surface cracks was observed up to 5 times in the hydrogenation cycle, but in the 6th cycle longitudinal cracks occurred on the D side, and destruction occurred in the vertical direction at the 10th cycle. Figure 13 shows a schematic outlining the change of a round bar test piece, after up to 10 hydrogenation cycles. Depending on the hydrogenation cycle, the round bar specimen swells axially and diametrically but has a barrel-like shape with its central part expanding more than the end part. In the thin plate specimen, the deformation behavior was direction-dependent from the two-axis constrained state of the plane stress type, but in the round bar specimen, the interior of the material was in a triaxial stress state and the expansion behavior also had a direction dependency. It shows a relatively low expansion and exhibits almost the same expansion behavior in the axis and the radial direction. 3.4 Deformation Behavior of Grains of Round Bar Test Specimen Figures 14 and 15 are microstructure photographs of drawn and heat-treated specimens after ten hydrogenation cycles. The internal structure is a crystal structure where neither drawn material nor heat treated material is damaged, macroscopic void, macroscopic crack, etc. In the former, the crystal grain shape on the D plane is laterally long (the work material in Fig. 2) before the hydrogenation cycle However, after 10 cycles, the ratio of length to width and height is the same. Table III is the grain size of the round bar specimen after ten hydrogenation cycles, determined by the areal method and the linear method. In the E side of the drawn material, according to the areal method, before the hydrogenation cycle The crystal grains have been refined down to nearly half from 43 μm of 10 μm to 27 μm and on the D side are refined to about 1/3 from 76 μm to 22 μm because this is due to expansion by hydrogenation cycle · As a result of the quartzing method for the heat treated material, the result of the areal method for the heat treated material is as follows: 56 μm to 35 μm for E surface, 57 μm for D surface to 34 μm for each fine , And the pulverization is progressing. According to the result of the linear method, in the case of drawn material, the grain size ratio GL: GT is 1: 1 after 10 cycles from 3.5: 1 before the test, GL, GW, GT value is 1 / 2 to 1 / 7. Considering the results of the linear method and the areal method together, it can be said that the crystal grains of the worked material were divided into about one crystal grain by the hydrogenation cycle. In the case of the heat-treated material, the size of the workpiece is similarly reduced to about 4. The reason why the workpiece is miniaturized intensely is probably because the crystal grain was horizontally long and was likely to be divided, It was not observed at all in the sample. Figure 16 shows a schematic diagram of crystal grain change in the E cross section by the hydrogenation cycle. In the structure in the cross section before the test, crystal grains of approximately the same size are distributed, but considerable damage (voids and cracks) due to drawing exists The expansion of the crystal lattice by hydrogen absorption causes a strong tensile stress in the three axial directions inside the specimen and the internal strain accumulates due to the increase in the number of cycles. Through the triaxial stress field In addition, voids grow to extremely large cracks in the vicinity of the surface, on the other hand, macro cracks occur in large quantities in the vicinity of the surface, stress relaxation occurs, miniaturization does not progress much and the grain size is kept It tends to be lean. Figure 17 shows the relationship of hydrogen absorption ratio to hydrogenation cycle in a round bar specimen The absorption ratio slightly decreases as the number of cycles increases After approximately 10 cycles of hydrogenation cycle about 4.4% in drawn material, heat treated The absorption ratio of the first cycle of the hydrogenation cycle was 0.71 for the drawn material and 0.72 for the heat treated material and it was almost the same value as in the case of the flat plate test piece Although the reduction ratio of the hydrogen absorption ratio of the round bar specimen to the number of cycles appears somewhat higher than that of the flat specimen, this is due to the large hydrogen induced cracking to increase the surface area, the hydrogen absorption capacity Will be reduced. 4 Discussion Figure 18 shows the surface photographs of the flat plate test pieces (heat treated material) for hydrogenation cycles 5, 10, 15 and 20 times. Until cycle 10, the occurrence of microscopic cracks and voids increases as the number of cycles increases, and the unevenness of the surface is intensified, but after 10 cycles the surface gradually forms a blister state, and it forms cracks, voids and other defects, and becomes covered with fine particles. Figures 19 (a) and 19 (b) show the surface photographs after 20 cycles of flat plate test pieces (processed and heat treated materials, respectively). Macroscopic cracks covered with blisters are developed in a mesh pattern on the surface portion. Also, the surface of the processed material has stronger irregularities, and the degree of damage is also greater. Figures 20 (a) and (b) are photographs of the surface after 10 cycles of the round bar test pieces (processed and heat treated materials, respectively), but the vicinity of the surface is a severe breakage, so the structure is extremely uneven. In this case as well, the damage of the surface of workpieces is more severe. A lot of large cracks occurred in the round bar test piece after 10 cycles of hydrogenation, although the number of cycles was smaller than that of the flat test piece after 20 cycles. >From the above results, it can be seen that damage to the surface texture is much more severe in the round bar test piece than in the flat test piece, and the processed material tends to be more damaged than the heat treated material. Figure 21 is a photograph of the internal structure of the flat specimen after 20 cycles of the hydrogenation cycle. Voids are developed from inside the crystal grain and grain boundary, and the tendency is more marked in the processed material than in the heat treated material. However, as mentioned above, macroscopic defects (cracks and voids) are relatively small in the internal structure of the flat test piece, and the maximum defect observed is shown in Fig. 21. The size is several tens of microns. In the flat plate specimen, almost no crystal grain fragmentation was observed, and macroscopic damage of the internal structure was also mild. Figures 22 (a) and (b) are internal structural photographs of the round bar test pieces (drawn and heat treated materials respectively) after 10 cycles. Both large voids are observed, and the voids are connected to form macroscopic cracks, causing grain boundary cracking. Internal structure damage is much more severe than for the flat plate specimens. Hydrogen absorbed in defects (voids and cracks) in the vicinity of the surface of the test piece is released as a foamy gas due to a change in chemical potential after electrolysis is stopped, when the hydrogen weight after hydrogen absorption is obtained. For this reason, after the sample is left standing for a sufficiently long time, we measured the weight of the Pd sample to obtain the H/Pd value. 10) Figure 23 (a) shows the relationship between the duration of the time the sample was left standing, and the change in the H/Pd value in the flat plate test piece (processed material and heat treated material). The difference between the first cycle and the 20th cycle is not so large for any material. That is, it seems that the amount of hydrogen accumulated in the damaged structure does not change so much from the 1st cycle to the 20th cycle. This seems to indicate that the damage due to hydrogenation cycle does not increase much despite the increase in the number of cycles. On the other hand, Figure 23 (b) shows the relationship between the duration the sample is left standing and the hydrogen absorption ratio in the round bar test piece (processed material and heat treated material). The amount of released hydrogen is clearly larger in the 10th cycle. That is, the amount of hydrogen, which is more likely to be released as gas in the tenth cycle, is increased, but on the other hand, the amount of solid solution hydrogen that is stable is smaller. This means that gaseous hydrogen accumulates mainly in the various defects and damaged parts generated in the vicinity of the sample surface, and that proportion increases as the number of cycles increases. Since this gaseous hydrogen is mostly spontaneously released without lodging in the material under normal temperature and pressure, it does not contribute to the hydrogen absorption performance of the material. The decrease in the hydrogen absorption ratio due to the increase in the number of cycles shown in Figure 17 is due to the increase in the structure damage near the material surface, and it would appear the solid solution hydrogen that can be stably stored in the crystal lattice for the increase in the unstable gaseous hydrogen rather decreases. 5 Conclusion A hydrogenation cycle experiment was conducted by the electrolysis method using palladium rolled material, drawn material, and heat treated material. The accumulation of damage in the inside and the surface of the material due to an increase in the number of hydrogenation cycles, deformation behaviors of crystal grains and the mechanism of fine fragmentation were investigated. As a result, the following conclusions were obtained. (1) Differences in the process of deformation, damage, and deterioration of materials due to hydrogenation cycle were observed due to the geometrical shape of the material. As a result of preferential expansion of the flat plate type specimen in the plate thickness direction, the plate width and the length direction gradually shrink. As a result, the laterally elongated crystal grains formed by the rolling process undergo deformation that is large enough to be vertically elongated in the plate thickness direction, but the crystal grains do not become much more finely divided. Also, the internal structure of the sample does not undergo much macroscopic damage (macroscopic voids or cracks). On the other hand, in the round bar test piece, almost the same expansion occurs in the diametrical direction and the axial direction, and when the hydrogenation cycle is repeated several times, macroscopic cracks are generated on the surface, the sample rapidly breaks down, Macroscopic damage occurs internally, and the fine fragmentation of crystal grains rapidly progresses. (2) Damage (cracks and voids) in the processed material is more severe than that of the heat-treated material in any of the test pieces, because gaseous hydrogen penetrates the micro voids and micro cracks that are generated in the material by the processing process. (3) The hydrogen absorption performance gradually decreases due to damage to the specimen, by about 1.5% in the flat plate specimen after 20 cycles, and by 4.4% in the bar test piece after 10 cycles. But the decrease in absorption is relatively small compared with the severe shape deformation and destruction. References 1) Hideki Hagi, Ankuto Hayashi, Journal of the Japan Institute of Metals, 49, 327 (1985). 2) Yoshihiro Ogino, Mitsuru Nagahama, Atsushi Yamada, Journal of the Japan Institute of Metals, 52, 395 (1988). 3) Toshiki Kabutomori, Keizo Ohnishi, Mataria, 36, 298 (1997). 4) Yuzo Kamakura, Yoshiaki Shida, Koichi Kamiya, Makaria, 36, 104 (1997). 5) Hiroyuki Miyamura, Tatsuo Sakai, Nobuhiro Kuriyama, Hideaki Nakada, Susumi Uehara, BOUNDARY, 34 (1994). 6) Hiroyuki Miyamura, Tetsuo Sakai, Makaria, 34, 189 (1995). 7) Takemori Toshiki, Ohnishi Keizo, Makaria, 36, 298 (1997). 8) Hikyaku Hiki, Moshi Mohri, Miwako Shionoya, Etsuo Akiba, TOYOTA Technical Review, 48, 123 (1998). 9) Yasutaka Yasuda, Resources and materials, 114, 455 (1998). 10) Jung Yong-guan, Suehiro Hideki, Sakai Yuzuru, Material, 49, 1242 (2000).
