日本金属学会誌
Online ISSN : 1880-6880
Print ISSN : 0021-4876
ISSN-L : 0021-4876
33 巻, 11 号
選択された号の論文の22件中1~22を表示しています
  • 椙山 正孝, 福迫 達一, 佐藤 彰
    1969 年 33 巻 11 号 p. 1209-1212
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The recrystallization texture of annealed pure aluminium sheets made from direct-rolled plates at a low directrolling speed is {110}⟨001⟩ and the formability of the annealed sheets with {110}⟨001⟩ is better than that of those with {112}⟨111⟩ and {112}⟨123⟩. Conditions for the formation of the recrystallization texture {110}⟨001⟩ have been studied and the following results are obtained. (1) In order to form the recrystallization texture {110}⟨001⟩, the rolling texture {110}⟨112⟩ must be strong in the cold-rolled sheets before annealing. And conditions for the formation of the rolling texture {110}⟨112⟩ and the recrystallization texture {110}⟨001⟩ are as follows: (a) The cold-rolling direction must be parallel to the growth direction of columnar grains. (b) The reduction degree of cold-rolling should not be so high as to develop the rolling texture {112}⟨111⟩ and {112}⟨123⟩. (c) Direct-rolled plates contain the solid solution element of aluminium. (2) Soaking of direct-rolled plates promotes the rolling texture {112}⟨111⟩, {112}⟨123⟩ and the cube texture {100}⟨001⟩ as the recrystallization texture.
  • 前田 春興, 中桐 明和, 中井 正明
    1969 年 33 巻 11 号 p. 1213-1218
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The static tention pieces were, prepared to the length of 1/2 of the ultrasonic wave length (21.6 kc in longitudinal wave). The value of the tensile stress became remarkably smaller when the ultrasonic wave of resonance frequency was superimposed on a specimen continuously or intermittently.
    At the moment when ultrasonic power was turned on, the transient phenomena of deformation at the node and the loop were observed by using a dualbeam syncloscope and a semiconductor strain gage. The node of the specimen was deformed with a few cycles of ultrasonic vibration, but the loop was not deformed.
    After superimposition of the ultrasonic wave, the deformation behaviors of the node and the loop were observed with a foil strain gage. In order to obtain the same degree of deformation at the node and the loop by stationary tensile deformation, the amount of deformation should be twice or three times greater than that given by the superimposed ultrasonic vibration.
    The increment of the temperature of the specimen with the superimposed ultrasonic vibration was proportional to its amplitude, and the temperature of the node was higher than that of the loop.
  • 長崎 久弥, 春山 志郎, 金子 正己, 伊藤 勲
    1969 年 33 巻 11 号 p. 1218-1224
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The dezincification of α-brasses (2∼30 wt%Zn) in vacuum at high temperature was investigated by a weight loss measurement. The weight loss was proportional to the square root of time for brasses of low zinc contents, as was expected from the ordinary diffusion mechanism. For the high zinc content, however, the weight loss was found to be proportional to time. For the specimens with intermediate zinc contents (20∼26 wt%Zn), the weight loss-time relation was as follows:
    (This article is not displayable. Please see full text pdf.)
    When a specimen showed such an unusual time law, voids and micro-cracks were found in the region near the surface by optical microscopic observation. By using an electron probe micro-analyzer, the zinc content was found to be abnormally depressed. In this region.
  • 長崎 久弥, 春山 志郎, 伊藤 勲, 鈴木 秀生
    1969 年 33 巻 11 号 p. 1224-1227
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    It has been reported in the preceding paper that the formation of voids and cracks plays an important role in the mechanism of dezincification in α-brasses. In this work, the dezincification in α-Ni-Zn alloys was investigated with reference to the relation between the rate law and voids formation, using the alloy specimens containing 2, 15 and 20 wt%Zn. The experiments were carried out at 800°, 900° and 1000°C in vacuum (c.a. 10−5 mmHg).
    The results showed that the weight loss was proportional to the square root of time for 2 and 15 wt%Zn alloys and to time for 30 wt%Zn alloy. The cross sectional concentration profiles in the specimens were then traced by EPMA. In 30 wt%Zn alloy dezincified at any given temperature, the zinc content was abnormally depressed in the outer layer of the dezincified zone and lowered only at grain boundary in the inner layer. In the outer layer, a number of voids and cracks was observed by an optical microscope.
  • 井藤 忠男, 松浦 圭助, 渡辺 勝也
    1969 年 33 巻 11 号 p. 1227-1232
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The precipitation phenomena of an Al-6 wt%Mg alloy containing 0.5 wt%Ag were investigated by means of a transmission electron microscope and hardness measurements to clarify the effect of different heat treatments on the morphology of precipitation.
    The structure of the specimen water-quenched and aged below 200°C showed a fine and homogeneous distribution of small precipitates. On the other hand, by ageing above 200°C, the distribution of precipitates was inhomogeneous : the “precipitate-free zones” along grain boundaries were recognized besides the distribution of fine precipitates. When the direct-quenched specimens were aged above 200°C, the precipitates were mainly formed on dislocation lines at an early stage and only a coarse distribution of long precipitates was recognized for prolonged ageing, while by ageing below 200°C, their distribution was fine and homogeneous in the same way as in the water-quenched and aged specimen. The effect of the two-step ageing on the precipitation phenomena was also studied.
    The above results suggest that the clusters form on ageing below a critical temperature near 200°C and they act as the nucleation sites of precipitates, and that the quenched-in vacancies accelerate the nucleation of precipitates indirectly through the formation of clusters.
  • 井藤 忠男, 古谷 忠, 松浦 圭助, 渡辺 勝也
    1969 年 33 巻 11 号 p. 1232-1238
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The present work was carried out so as to determine the phase diagram of Al side in the Al-Mg system with 0.5 wt%Ag by using the micrographic method, electrical resistivity measurements and differential thermal analysis.
    The solid solubility curve of the α-phase in this alloy shifted to the side of appreciably higher temperatures compared with that in the binary Al-Mg alloy, indicating that the solubility of Mg in Al was decreased by the addition of 0.5 wt%Ag.
    The results obtained by electron probe microanalysis and electron diffraction analysis showed that equilibrium precipitates in the alloys with 0.5 wt%Ag were mainly a ternary compound, Mg32 (Al, Ag)49-T phase.
    Furthermore, the ageing behaviour of an Al-6 wt%Mg-0.5 wt%Ag alloy was studied by hardness measurements and electron microscopic observation.
    The precipitates in an Al-6 wt%Mg alloy with 0.5 wt%Ag were finer and more uniformly dispersed than those in a binary Al-6 wt%Mg alloy, thus leading to a remarkable increase in age-hardening. It was considered that the acceleration of precipitation by the addition of silver was mainly caused by the effect of the increased supersaturation for precipitation of the T phase. However, seems it more reasonable to conclude that the increased supersaturation contributed greatly to the stability of G.P. zones which could act as nuclei of precipitates.
  • 長谷川 正, 佐藤 洋, 辛島 誠一
    1969 年 33 巻 11 号 p. 1239-1243
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    Dislocation arrangements in copper single crystals crept at high temperatures were investigated by means of transmission electron microscopy. Creep tests were performed in argon atmosphere at 0.65Tm=610°C and 0.75Tm=745°C (Tm: the melting temperature in Kelvin). Thin foil specimens with the primary slip plane (111) parallel to the plane of the foils were prepared from the deformed crystals and examined by a 500 kV electron microscope.
    The main results obtained are as follows: (1) The deformation substructure at 610°C was similar to that at 745°C, except that spacings between the sub-boundaries at 610°C were smaller than those at 745°C. (2) From the results obtained by the Berg-Barrett method, etch-pit observation and transmission electron microscopy, the mobile dislocation densities in the steady-state creep were estimated to be ∼5×106/cm2 and ∼3×106/cm2 at 610°C and 745°C, respectively. (3) Dislocations bowing out from the sub-boundaries and dislocations having dipoles were observed in the vicinity of the sub-boundaries. They may be the dislocations emitted from and absorbed into the sub-boundaries, respectively. (4) In the tensile test inside the electron microscope of the thin foil specimens prepared from the deformed crystals, it was found that the dislocations bowing out from the sub-boundaries formed during creep were emitted from the sub-boundaries, and also that the glide motion of the dislocations was stopped by the sub-boundaries. It was concluded from these observations that the sub-boundaries act as sources and sinks of the mobile dislocations during high-temperature creep deformation. The mean free path of the mobile dislocations is thus limited by the sub-boundaries, that is, a few μ∼10μ and 10μ∼ a few 10μ for the primary edge and screw dislocations, respectively. (5) The fact that the dislocation density in the sub-boundaries increases and the mobile dislocation density remains constant during steady-state creep, shows that some kinds of dislocation multiplication should occur in high-temperature creep deformation.
  • 三浦 維四, 浜中 人士
    1969 年 33 巻 11 号 p. 1244-1248
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The authors have studied the method of preparation of Al-Si-Al3Ni ternary eutectic composites by unidirectional solidification, and their micro-structures have already been reported. In this paper, the tensile behaviour of Al-Si-Al3Ni eutectic composites were studied over a range from room temperature to 500°C and as a function of solidification rate, and results were compared with those of as-cast alloys. Al-Si-Al3Ni eutectic composites are reinforced with 12.9 vol%Si and 8.1 vol%Al3Ni fibers, although such a composite reinforced with fibers of different properties was scarecely been reported. The main results are as follows.
    (1) Tensile properties of controlled Al-Si-Al3Ni eutectic composites were much improved at room temperature and elevated temperatures as compared with those of as-cast alloys. (2) The tensile behaviour of the eutectic composite showed an interesting change with the solidification rate, associated with changes of their microstructures. (3) Fracture of the eutectic composites occurred by the failure of Al3Ni fibers after the Si fiber had ruptured. (4) Analytical models for the tensile strength and elastic modulus of the composite reinforced with two kinds of fibers were derived, and these appear to provide reasonable approximations for Al-Si-Al3Ni eutectic composites.
  • 山内 睦文, 佐野 幸吉
    1969 年 33 巻 11 号 p. 1249-1255
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    Deoxidation of molten metals by hydrogen is not only used frequently as a deoxidation method in the melting operation on a laboratory scale, but also as an indirect industrial application in the refining of blister copper by ammonia gas. As an example of gas-molten copper reactions, the authors studied the velocity of the deoxidation by hydrogen diluted with argon under various experimental conditions by the use of the resistance furnace. From the experimental results on the diffusion control of hydrogen molecules in the gaseous film and of oxygen atoms in the stagnant molten copper, it was found that the velocity of the deoxidation could not be explained by a simple mechanism. However, it became evident the resistance of diffusion of hydrogen molecules in the gaseous film to the velocity of the deoxidation was very small and molten copper was stirred even in case the resistance furnace was used.
    It is assumed that the deoxidation of molten copper by hydrogen occurs in a complicated reaction mechanism, because of the following three reasons in addition to the facts described above: (1) the rate of increase of the deoxidation velocity decreases with rising PH2, (2) the reaction plane exists in addition to the free surface, (3) the apparent activation energy of the deoxidation reaction is 12.1 kcal/mol which is thought to be too large as a value of mass transfer.
  • 山内 睦文, 佐野 幸吉
    1969 年 33 巻 11 号 p. 1255-1261
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The previous paper dealt with the velocity of the deoxidation reaction of molten copper by hydrogen gas by the use of the resistance furnace. But the results obtained could not be explained by such a simple mechanisms as the diffusion control of hydrogen molecules in the gaseous film and of oxygen atoms in the stagnant molten copper. In the present work, in order to make clear the deoxidation mechanism, the effect of stirring on the velocity of the deoxidation reaction was studied when molten copper was inductively stirred by the use of a high-frequency furnace. From the results obtained, it becomes clear that the faster velocity of the deoxidation by the use of the high-frequency furnace than by the use of the resistance furnace is due to the acceleration of both the velocity of the deoxidation reactions occuring at the free surface and that on the crucible wall contacted with molten copper.
    The authors proposed a deoxidation model by which the experimental results reported in the previous and present papers were satisfactorily explained.
  • 山内 睦文, 佐野 幸吉
    1969 年 33 巻 11 号 p. 1262-1267
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    As described in the previous paper, the surface of molten copper is fluid from the onset of the deoxidation reaction and dispersion of bubbles is observed below a certain oxygen content in case a crucible of opaque quartz and the resistance furnace are used. Following the assumption that these bubbles are generated and grow up at a singular point of the crucible wall contacted with molten copper and surface from the points, the conditions of generation of bubbles, i.e. the overall velocity of the deoxidation reaction would be affected by the surface conditions of the crucibles. Therefore, the velocity of the deoxidation reaction of molten copper was investigated by the use of various materials of crucibles, mainly in relation to generation, growth, and surfacing of bubbles on or from the crucible walls. From the results obtained it becomes clear that the effect of materials of crucibles on the velocity of the deoxidation reaction is remarkable and by the use of the more porous materials of crucibles, the overall velocity of the deoxidation reaction becomes faster and therefore, the deoxidation reaction occurring on the crucible wall contacted with molten copper is obviously related to the deoxidation reaction of molten copper. When crucibles of fine materials are used at PH2≥0.2 atom the velocity of the deoxidation reaction occurring on the crucible wall depends on the process of generation of bubbles. When the porous materials of crucibles are used, the propotion of the deoxidation reaction on the crucible wall to the overall deoxidation reaction increases and the velocity of it depends on the process of transfer of H through the gas-liquid interface at the free surface.
    As described above, the propotion of the deoxidation reaction occurring on the crucible wall to the overall deoxidation reaction is considerably large and amounts to 20∼30% in case of the fine materials of crucibles and to 50∼60% the porous materials of crucibles.
  • 作井 誠太, 佐藤 公子
    1969 年 33 巻 11 号 p. 1268-1275
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The effects of strain rate and temperature on the flow curves of Al single crystals were investigated systematically by using many single crystals with same orientaion. These effects on each of the three stages of the flow curves were discussed separately.
    During the differential test on the effect of strain rate, the specimen prestrained by impact loading showed a remarkable yield drop during the subsequent static loading in the same way as reported by Cottrell during the differential test on the effect of temperature. Temperature seemed to give a more profound effect on the flow curves and the work softening process than the strain rate.
  • 大庭 幸夫, 関野 泰広
    1969 年 33 巻 11 号 p. 1275-1280
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    Effects of temperature and strain-rate on upper and lower yield points of pure iron in which dislocations were locked in different degree were examined under a condition in which any clear Lüders bands were not observed. The experimental results were compared with the calculated data which were obtained from Hahn’s equation.
    It has been shown that even if the dislocations are locked very strongly, the strain-rate dependences of upper and lower yield points obtained from Hahn’s equation are not consistent with the experimental results. As a main cause of this discrepancy, it is considered that the degree of nonuniform deformation in the specimens increases with increase in the crosshead speed of the tensile test machine rather than variation of the density of the mobile dislocation existing before testing as pointed out by Arsenaut.
  • 栂野 秀夫, 佐々木 英次, 神田 幸雄
    1969 年 33 巻 11 号 p. 1280-1285
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The anodic polarization behavior of commercially pure titanium in hydrochloric acid was investigated and the availability of anodic protection was examined. Acid concentrations ranging from 10 to 30% were used in the temperature range 35° to 90°C. In this electrode system, the potentiostatic current-time curves in a passive range generally showed the following characteristic change; the current decreased at first with rapidity and then gradually to a minimum value, following which it began to increase to attain a maximum and after showing some decrease, reached a steady state. The shapes and relative positions of these curves were affected by set potentials, acid concentrations, temperature, etc., and the effect of set potentials was most significant.
    In hydrochloric acid, the anodic oxide film of titanium is unstable as stated above, and furthermore titanium exposed to the vapor of hydrochloric acid cannot be protected by anodic polarization. From the above considerations, the applications of anodic protection to titanium in hydrochloric acid should encounter a number of questions, contraly to the views of other researchers.
  • 内田 弘, 門 智, 山田 光二, 加藤 忠一
    1969 年 33 巻 11 号 p. 1286-1290
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The thickness of the hydrated chromium oxide films on chromium plated type Tin Free Steel was directly measured by ellipsometer and the film growth was observed by an electron microscope.
    The results obtained are as follows:
    (1) The optical constant of the hydrated chromium oxide film is estimated as 1.55-0.02i.
    (2) The rate of film growth is fast at the initial stage of electrolytic treatment and then it becomes slow gradually.
    In this case, a simple functional time dependence such as a logarithmic or parabolic corelationship may not be observed.
    (3) The chromium content in the film is linearlly related to the film thcickness. From this fact, it may be assumed that the hydrated chromium oxide film consists of same components in the range of 0 to 600 Å.
  • 内田 弘, 門 智, 山田 光二, 加藤 忠一
    1969 年 33 巻 11 号 p. 1290-1295
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The hydrated chromium oxide film on the chromium-plated type Tin Free Steel was investigated by means of a micro-thermobalance, a T.O.F. mass-spectrometer, thermal decomposition gas-chromatograpy and an infrared-spectrometer.
    The results obtained are as follows:
    (1) The water in the film vaporizes up to 240°C and the amount of H2O was estimated as 24 per cent.
    (2) The infrared absorption spectra of the film are found at 3400 cm−1 due to OH stretching vibration and 1600 cm−1 due to HOH bending vibration. These absorptions increase with the film thickness and decrease by heating.
    (3) From the surface observation by means of a scanning-type electron microscope, it is found that some cracks are formed in the thick film by heating.
    From the above results, it may be concluded that the behavior of the water in the film is analogous to that in the passivated film on steel or stainless steel, and the dehydration of the film affects the paintability and corrosion of the substrate.
  • 内田 弘, 門 智, 山田 光二, 小笠原 正
    1969 年 33 巻 11 号 p. 1295-1299
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    According to the determination of the thickness and weight of the hydrated chromium oxide film and electro-microbalance, respectivelly, the density of the film is estimated as 2.32 g/cm3. By using the chromium content (44.8%) and the amount of water (24%) in the film, the chemical formula for the film is considered to be 2CrOOH·3H2O.
    Five peaks are recognized at 3400, 1600, 1425, 910 and 530 cm−1 in the infrared absorption spectra of the film.
    From the above results, it may be assumed that the hydrated chromium oxide film on T.F.S. forms macro-molecules bonded with the OH bridge.
  • 増本 量, 菊地 紀雄, 沢谷 昭八
    1969 年 33 巻 11 号 p. 1299-1303
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    Temperature dependences of thermal expansion and Young’s modulus of single crystals of 63%Fe-32%Ni-5%Co alloy were measured over a temperature range −30° to 600°C. The single crystal bars with several crystal directions were prepared, with a length of about 12 cm and a diameter of about 3 mm. The principal elastic parameters and Young’s modulus for the three principal axes were calculated from the measured values of thermal expansion and Young’s modulus of the single crystals. The results showed that the temperature dependences of Young’s modulus for the three principal orientations were similar to each other, and that the Young’s moduli at 20°C in the ⟨100⟩, ⟨110⟩ and ⟨111⟩ directions were 5.80×105, 12.53×105 and 20.42×105 kg/cm2, respectively. Also, the crystal anisotropy of Young’s modulus, E⟨111⟩E⟨100⟩ of the alloy was estimated to be 3.52 at 20°C.
  • 五弓 勇雄, 木原 諄二, 木村 敏郎
    1969 年 33 巻 11 号 p. 1303-1308
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    In the previous reports, it has been concluded that the hot-rolling simulation by counter-blow forging in the plane strain condition may be a satisfactory operation. It has been also known that the (222) component of both 80% cold-rolled and annealed 0.05% carbon-steel sheets water-quenched just after hot working is the greatest when the hot-working temperature is about 800°∼860°C and the hot worked specimens have a very fine grain of 8 μ.
    In the present study, investigations were made so as to make clear the effect of the grain size of hot worked specimen on the texture of cold-rolled and annealed low carbon steel sheets and the effects of the casting conditions and the Al or N content on the relation between the textures and the grain size.
    The specimens prepared were 0.07% carbon rimmed steel with normal rimming action (R-1), 0.07% carbon rimmed steel without rimming action by Ar gas seal (R-3), IN treated Al killed steel with normal Al content (K-A) and IN treated Al killed steel with Al and N contents 3 times larger than normal Al killed steel (K-B).
    The experimental procedure is the same as in the previous reports.
    The main results are as follows :
    (1) In the cold rolled sheets, there are a small amount of the near (111)[11\bar2] component, whereas there is a large amount of the component in the annealed sheets. A large amount of the near (111)[1\bar10] component is present in the cold rolled sheet.
    (2) The amounts of components other than (222) are not affected by hotworking conditions.
    (3) The dependence of the (222) component in the cold-rolled sheets on the grain size after hot working is not so apparent, but that of the annealed sheets is very large. The relation between the amount of the (222) component and the grain size after hot working is uniquely petermined in the cases of R-1, R-2 and one-step rapidly heated annealed K-A. In the case of K-B, however, this relation does not hold when K-B sheet is worked above A3 point.
    (4) The reason why the (222) component in the annealed sheets depends on the grain size may be made clear by considering the grain-size dependence of probability or rotation about the near (111) axis normal to the plane occurring in the recrystallization.
  • 美馬 源次郎, 山口 正治
    1969 年 33 巻 11 号 p. 1308-1313
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    The effect of vanadium addition on the miscibility gap in the iron-chromium binary system has been investigated by means of differential thermal analysis and Moessbauer effect measurements. The miscibility gap obtained experimentally in the iron-chromium-vanadium ternary system are qualitatively in agreement with that predicted from thermo-dynamical data of three binary systems, iron-chromium, iron-vanadium and chromium-vanadium system.
    The ternary tie-lines has been determined by means of the Moessbauer measurements of the binary and ternary alloys aged at 480°C. It has been found that when the ternary tie-lines are extrapolated the lines cut the iron-chromium binary line at the point at which the chromium content is nearly 90 at%.
  • 岩間 義郎, 稲垣 松美, 宮本 毅信
    1969 年 33 巻 11 号 p. 1314-1319
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
    Alnico 8-type alloys containing more than 5%Ti has been found to have very high coercive forces, particularly when the isothermal magnetic treatment is applied. In order to explore its cause, in this work the transformations, microstructures and magnetic properties were investigated on the specimens of Alnico 8-type alloys with various titanium contents. The critical temperatures of the transformations in solid state were determined as a function of titanium content. It may be reasonably concluded that a mechanism of magnetic hardening in the alloys is similar to that in Alnico 5-type alloys containing no titanium.
    It was shown that the coercive force of the alloys depends largely upon the temperature of the isothermal treatment, and slightly upon the treating time. For each specimen, the optimal temperature of the treatment was found to be about 50°C below the critical temperature of the immiscibility gap reaction, and it is lowered with increasing titanium content. In addition the magnetic anisotropy induced by the isothermal treatment in a magnetic field seems to decrease in magnitude with increasing titanium content. Furthermore, the electron-microscopic observation has revealed that the precipitate particles developed by the optimal treatment are likely to increase in both size and dimension ratio with increasing titanium content.
  • 柴田 俊夫, 八戸 光夫, 竹山 太郎
    1969 年 33 巻 11 号 p. 1320-1321
    発行日: 1969年
    公開日: 2008/04/04
    ジャーナル フリー
feedback
Top