溶接学会論文集
Print ISSN : 0288-4771
検索
OR
閲覧
検索
4 巻 , 3 号
選択された号の論文の24件中1~24を表示しています
    • |<
    • <
    • 1
    • >
    • >|
  • 野村 博一, 杉谷 祐司, 村山 雅智
    4 巻 (1986) 3 号 p. 502-507
    公開日: 2009/06/12
    ジャーナル フリー
    The high speed rotating arc process originally developed for NGW-MAG welding has significant features for bead formation, such as decentralization of penetration and formation of concave bead profile. The examination carried out for horizontal fillet welding by this process has proved that the high speed rotating arc process is also suitable for horizontal fillet welding. Foremost, rotating action of the arc prevents hanging of weld bead and forms flat bead surface under high speed welding condition. Another advantage is the function of automatic seam tracking by arc sensor control. Real time and precision guidance can be achieved by detecting arc voltage waveform which changes with the arc rotation.
    抄録全体を表示
  • 野村 博一, 杉谷 祐司, 玉置 尚弘
    4 巻 (1986) 3 号 p. 508-514
    公開日: 2009/06/12
    ジャーナル フリー
    The automatic seam tracking system by arc sensor method developed by NKK has an original feature, that a weaving arc in a groove performs groove surface tracking under constant arc length control (AVC or ACC) and the turning point of weaving is controlled by the condition when torch height reaches to constant preset level. In this system, weaving width well corresponds to groove width independently to variation of deposited amount of metal. Namely, variation value of weaving width is same as that of groove width. Based on this relationship, welding speed control system has been developed. Where, only the variation value of weaving width is used as detected parameter, and optimum value of welding speed to keep bead height constant is calculated and output at every cycle of weaving. The pulsed MAG arc welding system for multi-pass welding of V shaped butt joint has been developed. Main function of this system is adaptive microcomputer control with above mentioned arc sensor method, such as seam tracking, weaving width control and bead height control for every pass of welding.
    In addition, following functions are also composed: built-up sequence control where number of passes is automatically decided from weaving width, simplified operating system by man-machine dialogic style with liquid crystal display and a key board, and use of a couple of optical fiber cable for transmission of multiple control signals. This system has been already applied to the fabrication of steel box columns.
    抄録全体を表示
  • 大橋 修
    4 巻 (1986) 3 号 p. 514-520
    公開日: 2009/06/12
    ジャーナル フリー
    The article gives the results of studying the effect of welding atmosphere such as vacuum at 4 and 4×10-3 Pa, and Ar, H2 and air at 1×105 Pa, on composition and pressure of the residual gas in voids present in diffusion-welded joints. The residual gas analysis was carried out using a mass spectrometer system evacuated to less than 4×10-6 Pa. Measurements were made both prior to and after breaking the diffusion-welded joints inside the apparatus. The materials examined were aluminum, iron, copper, titanium, SUS304 stainless steel and Ti-6AL-4V alloy, and the results obtained are as follows:
    The inert gas amount in the voids is constant for a given inert gas pressure in the welding atmosphere. For gases which react with the base metal to form compounds we do not detect their presence in voids. In the case of gases that dissolve in the base metal but do not form compounds, the gas pressure approaches a value which is in equilibrium with the gas content of the metal.
    抄録全体を表示
  • 桑名 武, 粉川 博之, 辻井 浩
    4 巻 (1986) 3 号 p. 520-526
    公開日: 2009/06/12
    ジャーナル フリー
    The effect of the nitrogen absorption on the transmission electron microscopic structure of SUS304L austenitic stainless steel weld metal was investigated by varying the nitrogen partial pressure of the Ar-N2 gas mixture welding atmosphere in an atmosphere-controlling chamber. An electron diffraction study suggested that the primary solidification mode of the weld metal alters from ferritic mode to austenitic one with the nitrogen partial pressure and the nitrogen content of weld metal. The austenite frequently adopts a near Kurdjumov-Sachs orientation relationship with the adjacent primary ferrite, while the eutectic ferrite does not adopt the K-S one with the adjacent austenite. The amount of the eutectic ferrite phase increases with the nitrogen partial pressure and the nitrogen content. In the weld metal with no ferrite, precipitation on dislocations and grain boundaries was observed.
    抄録全体を表示
  • 中尾 嘉邦, 西本 和俊, 辰巳 雅彦, 野井 伸悟, 塚原 宏, 原 泰弘
    4 巻 (1986) 3 号 p. 527-533
    公開日: 2009/06/12
    ジャーナル フリー
    The low interstitial ferritic stainless steel which has been successfully developed recently, can be uti-lized for an extensive range of applications. The use of the clad steel using the low interstitial 30Cr-2Mo steel as a cladding Plate contributes to not only reducing the material cost, but also providing the excellent corrosion resistance for devices of the chemical plants.
    This study was performed for the purpose of developing the optimum welding procedure. As the first phase of the study, embrittlement in the overlay weld was investigated by means of the side-bend test and its mechanism was discussed in this report.
    The main results obtained were as follows;
    (1) The bending ductility in the overlay weld was dependent on C pick-up from the base metal and grain coarsening during welding and 475-C embrittlement during PWHT. In these factors, C pickup from the base metal was most detrimental, that is, crack-crack free transition temperature of the weld in the side-bend test increased by 60K with increase of C content in the first layer from 55 pp-to 135 ppm.
    (2) Appearances of fractured surfaces of side-bend tested specimens for overlay welds showed typical cleavage fractured surfaces. Most initiation sites were observed in the vicinity of the weld interface of the first layer in the overlay weld when the penetration ratio of the first layer being below 35%.
    抄録全体を表示
  • 上田 幸雄, 金 裕哲, 柄谷 和輝, 山北 晃久
    4 巻 (1986) 3 号 p. 533-539
    公開日: 2009/06/12
    ジャーナル フリー
    Repair welding produces welding deformation, residual stresses, etc. which may newly produce repair welding cracks. Therefore, it is important to clarify the mechanical characteristics of repair welds in thick plates.
    In this paper, three-dimensional unsteady thermal conduction analysis and three-dimensional thermal elasto-plastic analysis are performed. Based on the results, study is made on the characteristics of distributions of three-dimensional welding residual stresses and plastic strains, and their production mechanisms. The main conclusions can be summarized as follows :
    1) Residual stresses and plastic strains in the base plate are largely affected by compressive hydrostatic stresses and compressive plastic strains produced in the heating stage, respectively. This is remarkable in the base plate near HAZ.
    2) Irrespective of elastic or plastic, severity of mechanical restraints in respective directions can be known from the magnitude of produced stress components, {σx, σy, σz)T.
    3) Since plastic strains are assumed to be incompressive, the sign of plastic strain produced in the direction where maximum stress is produced (mechanical restraint is the severest) is opposite to that in the direction where minimum stress is produced (mechanical restraint is the weakest). Accordingly, the signs of stress and plastic strain may disagree.
    4) For evalution of mechanical damage of the material suffered from welding, it is necessary to pay attention to the whole process of a series of elasto-plastic behaviors, not only to the magnitude of residual plastic strains. In this paper, plastic work is proposed as a parameter to evalute the severity of mechanical condition including transient state in the plastic region. The plastic work becomes maximum in the base plate near HAZ.
    抄録全体を表示
  • 4 巻 (1986) 3 号 p. 540-569
    公開日: 2009/06/12
    ジャーナル フリー
  • 4 巻 (1986) 3 号 p. 543
    公開日: 2009/12/10
    ジャーナル フリー
  • 4 巻 (1986) 3 号 p. 547
    公開日: 2009/12/10
    ジャーナル フリー
  • 4 巻 (1986) 3 号 p. 569
    公開日: 2009/06/12
    ジャーナル フリー
  • 種子 彰, 荒木田 史穂, 高木 健治
    4 巻 (1986) 3 号 p. 570-576
    公開日: 2009/12/10
    ジャーナル フリー
    The arc rotation velocity in the magnetically impelled arc butt welding of a steel pipe was measured using the probe method and a high speed camera and the condition of arc plasma inside the pipe was photographed. The following is the conclusion thus obtained.
    (1) The arc rotation velocity and arc angle were measured by the probe at each position on the pipe circumference immediately after an arc is ignited and the relationship between these and the method of supplying power was analyzed. With the steel pipe (60.5 mm in OD, 3.8 mm in wall thickness) used for the experiment, it was necessary to supply power at least two or more points to prevent ex- cessive local fusion or arc extinction.
    (2) The arc rotation velocity was measured using a high speed camera, and the arc rotation velocity was classified into three regions of the low velocity region, high velocity region and speed fluctuation region.
    (3) It was surmised that both the time th and tf that divide the above three regions are physical property values, and the time th is the time required for the area 1.0 mm from the pipe end exceeds the Curie temperature and the time tf is the time required for the entire surface of the pipe end to reach the melting temperature.
    (4) Changes in the arc rotation velocity in the high velocity region and the velocity fluctuation region were analyzed by using the equation of motion to surmise the virtual mass, saturation magnetic flux density, and drive force.
    抄録全体を表示
  • 種子 彰, 荒木田 史穂, 高木 健治
    4 巻 (1986) 3 号 p. 577-581
    公開日: 2009/06/12
    ジャーナル フリー
    The heating and cooling curve of Magnetically Impelled Arc Butt Welding of steel pipe were calculated by one dimensional heat flow and compared with the measured data. The swelling shape of welded joint was also calculated with results of temperature distribution of pipe. The measured heating and cooling curve corresponded to the calculated one at η=0.72, k=5.64 mm2/s. The time that the temperature at l mm position from the pipe end was heated to 1200°C corresponded to the arc time of the optimum welding conditions.
    抄録全体を表示
  • 仲田 周次, 藤本 公三, 玉井 尚治
    4 巻 (1986) 3 号 p. 581-586
    公開日: 2009/06/12
    ジャーナル フリー
    In the high-current density resistance spot welding process, the fusion layer at the interface of the joints is much thinner than one of the conventional spot welding process, the strength of the joints is sufficiently high, and its indentation is very small, and so the contact resistance at the interface is used efficiently to form nugget and the thermal loss is obstructed due to the pulsed high current density with very short duration.
    As this joining process utilizes the contact resistance at the interface efficiently, the current path is main factor to be determined the joining area. The current path is mainly under the control of the electrode diameter, sheet thickness, electrode force, and current value. For obtaining the joints with sufficiently high strength, the current density in the current path among sheets must be larger than the certain value (2.0 kA/mm2). Furthermore, if the current density at the current path between electrode and sheet is larger than the certain value (2.6 kA/mm2), the electrode sticks on the sheet.
    Thus, for obtaining the satisfactory joints with sufficiently high strength and without the electrode sticking on the sheet surface, the proper joining conditions exist under certain range of sheet thickness and electrode tip diameter for joining.
    抄録全体を表示
  • 西口 公之, 高橋 康夫, 高橋 邦夫
    4 巻 (1986) 3 号 p. 586-592
    公開日: 2009/06/12
    ジャーナル フリー
    The process of solid state diffusion bonding is analized, noting the transient behaviour of interfacediffusion, which is an atom-flow along a bonded interface-boundary. A non-linear partial differential equation for the interface-diffusion is led by proposing a two-dimensional bonding model. Numerical calculations are performed with Euler method. It is indicated that an enhancement of the bonding proscess is caused by the transient behaviour when the bonding pressure is frequently removed, although it is negligible in a conventional diffusion bonding. So, "pulse-pressure bonding", in which the pressure is repeatedly applied at intervals of τoff (pressure-off interval), is proposed according to the calculated results. Influence of τoff on the enhancement-effect is numerically and experimentally examined. It is shown that τoff is a very important factor as a period of "relaxation" in order to attain an effect of the transient behaviour. And also, it is suggested that the enhancement-effect becomes striking with reducing the pressure-on interval τon, when τoff is suitablely determined.
    It is stated in conclusion that the bonding process enhanced by the transient behaviour is successful even if τoff is less than τon, that is, the pulse-pressure method can be useful to enhance the bonding process under a high vacuum condition.
    抄録全体を表示
  • 西田 隆法, 稲垣 道夫, 宮本 栄
    4 巻 (1986) 3 号 p. 592-596
    公開日: 2009/06/12
    ジャーナル フリー
    The production of fumes during brazing is a well-known phenomenon. One of the most important problems associate with such fumes is cadmium due to this element's hazardous nature.
    In the present work, the behavior of cadmium fumes during brazing was closely examined for BAg-1 and BAg-3, the typical silver filler metals containing cadmium.
    The conclusions drawn from the experiments conducted are as follows:
    1) The amount of fumes produced by either BAg-1 or BAg-3 was found to increase as the heating temperature increased when the optimum brazing temperature range was exceeded. This phenomenon can well be explained by the secondary electron images obtained by EPMA.
    2) For the both types of filler metals, the cadmium contents of the fumes produced remained as low as less than 0.80% and 0.70% for BAg-1 and BAg-3, respectively, when brazing was conducted at a heating temperature within the optimum temperature range. When this range was exceeded, however, the cadmium content of the fumes increased as the temperature increased.
    3) For this experimental device, so long as the heating temperature was kept within the optimum brazing temperature range, Cd concentration of fumes in the air can be controlled less than the WES (9007) control limit. When this range was exceeded, however, the fume concentration surpassed the WES limit by an amount which increased in proportion to the temperature employed.
    抄録全体を表示
  • 奈賀 正明, 田中 扶, 岡本 郁男
    4 巻 (1986) 3 号 p. 597-603
    公開日: 2009/06/12
    ジャーナル フリー
    Si3N4/Si3N4 joint was made using Cu66Ti34, Cu50Ti50 and Cu43Ti57 amorphous filler metals, where pressureless sintered Si3N4 was used. The joining strength of Si3N4 joint was measured by fracture shear testing, and the joining mechanism was investigated by microstructure observation with SEM, and elements analyses at the joint interface with EDX and EPMA. The results obtained are summarized as follows.
    (1) The molten Cu-Ti alloys containing 20 at% Ti or more exhibit the equilibrium contact angle of about 8°, and are applicable to the filler metal for joining of Si3N4.
    (2) The joining strength of Si3N4 using Cu66Ti34 filler is higher than that of Si3N4 using other Cu-Ti fillers at any brazing temperature except for brazing temperature of 1273 K. Thus, the joining strength of Si3N4 with Cu66Ti34 filler shows the maximum value of 313.8 MPa at brazing temperature of 1323 K. At the brazing temperature of 1373 K, Si3N4 joint with Cu50Ti50 filler shows the maximum value of 176.5 MPa at brazing time of 1.8 ks, and gradually decreases with longer brazing lime, and to 19.6 MPa at brazing time of 7.2 ks. The elevated temperature fracture shear strength of Si3N4 joint brazed at 1373 K for 1.8 ks using Cu50Ti50 filler increases to 199.1 MPa at 373 K, and gradually decreases to 105.9 MPa at 973 K.
    (3) During brazing Ti in molten Cu50Ti50 filler reacts with Si3N4 by the following reactions. Si3N4(s) +4Ti(1)=4TiN(s)+3Si(s), and Si3N4(s)+9Ti(1)=4TiN(s)Ti5Si3(s). TiN at the interface between Si3N4 and the filler, Ti5Si3, and Cu-Si phase at the central part of the filler are formed. The activation energy for growth of TiN at the joining interface is 206.3 kJ/mol, and the growth of TiN is dominated by the diffusion of N in TiN. The silicon content of the Cu-Ti phase increases about 7 at%Si to about 20 at%Si with increasing the brazing temperature.
    抄録全体を表示
  • 弘田 実保, 町田 一道, 奥田 滝夫
    4 巻 (1986) 3 号 p. 603-608
    公開日: 2009/06/12
    ジャーナル フリー
    Gold wire bonding has been employed in mass production process of semiconductors because the ball bonding system using gold wire can be very effectively utilized on the aspect of productivity. There is, however, an increasing demand for development of a new technology to replace gold wire with other materials in the light of further enhancing the reliability of devices and saving the material cost. In this background, the authors developed the copper wire bonding technology for plastic molded semiconductor, fundamentally on the basis of employing the hardware system for conventional gold wire bonding, Primarily, the ball formatio technology and ball bonding technology were major subjects of development. These two problems were successfully solved by the development of power source for S.P. discharge, use of hot capillary tip, control of physical properties of aluminum pad, and other techniques.
    Furthermore, as a result of steady and versatile fundamental researches and applied studies such as material development, process engineering development, quality assurance and hardware development, the practicability of copper wire bonding has been confirmed. This paper mainly refers to the studies of its process.
    抄録全体を表示
  • 二俣 正美, 棟 徹夫, 竹内 貞雄, 中西 喜美雄
    4 巻 (1986) 3 号 p. 609-615
    公開日: 2009/06/12
    ジャーナル フリー
    The new measuring apparatus of plasma jet force using the load-cell on the parallel beams structure has been developed and applied to the measurement of argon gas plasma jet force.
    The load-cell of parallel beams structure is easily manufactured and has excellent sensitivity of measurement. The experimental results of the plasma jet force measurement are described and discussed.
    It has been indicated that the plasma jet force depends on arc current, arc voltage, working gas flow rate, gas supplying method, and stand-off condition. The plasma jet force is closely related with the input power and is proportion to the square root of input power.
    抄録全体を表示
  • 三宅 洋, 芦田 栄次, 小倉 慧, 二瓶 正恭, 小山 高一, 伊藤 嘉敏
    4 巻 (1986) 3 号 p. 615-621
    公開日: 2009/06/12
    ジャーナル フリー
    In a nuclear power plant, large numbers of pipe are set in very complex and narrow space. Because welding in such places is difficult, remote control pipe welding equipment, in such all welding sequence are controlled.by mini-computer, has been developed. In this report authors investigated the monitoring equipment for TIG welding.
    1) For optical system reduction contrast ratio in welding part, small monitoring equipment produced with combination of pinhole, screened filter, selective filter and lens has been developed.
    2) Informations can be obtained the following phenomena from monitoring equipment. (1) arc states (2) states of molten pool, (3) bead appearance, (4) groove shape, (5) weaving width, (6) molten states of filler wire.
    3) The monitoring equipment developed has been applied to welding procedure in a nuclear power plant.
    抄録全体を表示
  • 松本 俊美, 佐藤 宏, 幡谷 文男, 小倉 慧, 唐津 義憲
    4 巻 (1986) 3 号 p. 621-627
    公開日: 2009/06/12
    ジャーナル フリー
    Superconducting magnets in fusion reactor system produce enormous electromagnetic force which generates very high stress working on the containers of these magnets at cryogenic temperature. The candidate alloys for the structural material of containers are non magnetic stainless steels with improved yield strength and toughness at cryogenic temperature. Containers are produced as welded structures, so that these properties on the structural material are also indispensable for the weld metal.
    The chemical compositions of stainless steel weld metal are usually controlled to include a small quantity of δ-ferrite for the prevention of hot cracking in welds. The δ-ferrite is, however, ferromagnetic and causes the reduction in cryogenic toughness of weld metal. On the other hand, the weld metal of 19Cr-13Ni-5Mn type is known as fully austenitic one with very low susceptability on hot cracking in welds. The improvement of weld metal of this type resulted in the development of the fully austenitic weld metal of l7.5Cr-l6.5Ni-6.5Mn type with improvde cryogenic strengths. The effects of carbon and molybdenum contents on the tensile properties and Charpy impact properties were examined on this type of weld metal.
    Conclusions are as follows. It is possible to assure the yield strength of the weld metal of 2.3% molybdenum content higher than 1000 MPa when carbon content is higher than 0.07% and carbon plus nitrogen content is higher than 0.13%. While the weld metal decreases in the elongation and reduction in area as the addition of carbon, the weld metal which contains 0.13% of carbon and 2.3% of molybdenum indicates about 40% of elongation and about 300% of reduction in area at 4.2 K. The Charpy absorved energy and lateral expansion of the weld metal at 4.2 K are equal to those at 77 K. The weld metal decreases in the absorved energy and lateral expansion as the addition of carbon content, but the weld metal which contains 0.13% of carbon and 2.3% of molybdenum indicates about 60J of absorved energy and 0.8 mm of lateral expansion at 4.2 K. The relation between yield strength and absorbed energy at 4.2 K on the weld metal coincides with the lower limit line of absorbed energy of the stainless steels SUS304LN and SUS316LN of high yield strength.
    抄録全体を表示
  • 菅 泰雄
    4 巻 (1986) 3 号 p. 627-634
    公開日: 2009/06/12
    ジャーナル フリー
    The nucleation and the growth of gas bubbles at the solidification front of fused metal and the growth of blowholes in weld metal were discussed, in case the concentration distribution of hydrogen exists in front of the solidification front.
    Main results obtained are summarized as follows:
    (1) When a radius of gas bubble R is larger than critical radius Rc, the gas bubble is able to exist or grow in the fused metal. However, the gas bubble is not able to grow over the maximum radius Rmax. When R is smaller than Rc, the gas bubble becomes smaller and finally disappears.
    (2) When the thickness of highly concentrated area of hydrogen in front of solidification front is larger enough than the diameter of nucleus of gas bubble, the critical radius of the gas bubble Rc is given by following formula:
    Rc=2r(Co2/k2K2-pd-Pa)
    (3) The growth rate of gas bubble VR increases and critical radius Rc decreases with an increasing concentration of hydrogen Co* and with a decreasing solidification speed Vs.
    (4) The growth rate of gas bubble VR decreases with an increasing ambient pressure Pa in relatively larger radius of gas bubble. Nevertheless it increases with an increasing ambient pressure in smaller radius of gas bubble. Consequently the critical radius of gas bubble Rc decreases with an increasing ambient pressure Pa.
    (5) The contour of blowhole changes from "closed type", "cylindrical type" to "bubbling type", as Co* increases and Vs decreases.
    (6) As ambient pressure Pa increases, a smaller nuclear of gas bubble may grow more easily to be a blowhole. Therefore the occurrence of the blowhole becomes easier with an increasing ambient pressure. However, diameter of blowhole has a tendency to decrease with an increas of Pa.
    抄録全体を表示
  • 向井 喜彦, 西村 新, 金 應俊
    4 巻 (1986) 3 号 p. 634-639
    公開日: 2009/06/12
    ジャーナル フリー
    An experimental investigation was conducted to clarify the behavior of residual stress redistribution caused by fatigue crack propagation itself initially through compressive residual stress field. Test specimens were prepared from HT80 steel plate and two parallel welds comprising one pass along each side of the specimen were applied by electron beam welding to induce centrally located compressive residual stress field.
    The results obtained from fatigue crack propagation tests were reviewed comparing with the results from sawcut simulating a propagating crack. While the crack propagated in the initial compressive residual stress field, the initial distribution of residual stress was maintained until the crack tip reached tensile field in the vicinity of the weld bead. However, as the crack propagated beyond compressive field and crack tip entered into the tensile field, accompaning the residual crack opening just behind of the crack tip, progressive redistribution of residual stress was occurred.
    抄録全体を表示
  • 阿倍 博司, 片岡 成典, 佐藤 次彦
    4 巻 (1986) 3 号 p. 639-646
    公開日: 2009/06/12
    ジャーナル フリー
    This study intends to make clear the effects of the properties of parent metal, specimen width, sheet thickness and nugget diameter on the initiation life and the propagation life of fatigue crack in singlespot-welded joint under tensile-shear load, and using these results to establish an empirical formula according to which the ΔL-N curve can be drawn.
    The empirical formula obtained is expressed in the following equation.
    ΔL=C(1+N/T2)m/(N/T1)m (kN)
    here, m=0.25, C=0.1×W0.22×t0.91×d0.61
    Ti=107×1/1-Y' T2=107×1-Y/3×t/0.8
    Y=1.2×E1./100
    ΔL: Load range (kN), N: Number of cycles (cycle), W: Specimen width (mm), t: Sheet thickness (mm), d: Nugget diameter (mm), E1.: Elongation of parent metal (%)
    It was confirmed that accuracy of the empirical formula is at considerable high level, when this formula is applied for data of hitherto published many reports and obtained results are compared with their experimental ones.
    抄録全体を表示
  • 佐藤 善美, 伊藤 吉保, 志田 朝彦, 皆川 貞利
    4 巻 (1986) 3 号 p. 646-652
    公開日: 2009/06/12
    ジャーナル フリー
    Some effects of welding conditions on the strength of spot welded joints were studied. Spot welds under various welding conditions were performed on SPC-1 (cold rolled carbon steel sheets). Static tensile and bending fatigue tests were carried out, and the following results were obtained:
    (1) A dominant facto: on bending fatigue strength was different from that of static tensile fracture strength. The static tensile strength increased with nugget diameter. The bending fatigue strength changed with HAZ angle (θ), which is the angle between the heat-affected zone and the plate surface.
    (2) It was found that the fatigue crack propagated along the parallel direction to heat-affected zone.
    (3) The welding condition of R.W.M.A. class A gave the maximum static tensile strength but the minimum for bending fatigue strength.
    抄録全体を表示
    • |<
    • <
    • 1
    • >
    • >|
feedback
Top