A comparative study has been made of characteristics of gas-tungsten-arc (GTA) cathodes, consisting of tungsten mixed with a small quantity of lanthanum-, yttrium-, cerium-, thorium-, zirconuim- and magnesium-oxides respectively and pure tungsten. Arc starting performance, arc pressure, change in shape due to heavy loading are compared in DCEN polarity, and it is shown the superiority of La-oxide, Y-oxide and Ce-oxide electrodes in above characteristics.
A comparative study has been made of characteristics of Gas-Tungsten-Arc cathodes, consisting of tungsten mixed with a small quantity of lanthanum, yttrium, cerium, thorium oxides respectively. Through observation of oxide contained in electrode after arcing, the change in electrode properties are discussed. The oxide changes its shape and distribution during arcing by the melting of oxide itself, and consequently disappears from the surface layer, which is the reason why arc starting loses the normal behavior. The introducing oxygen into shielding gas promotes intensely the electrode consumption and resulted in forming the "rim" at the periphery just behind the cathode area. It gives serious problems in durability of electrode and the stability of arc.
The inverter controlled CO2 gas shielded arc welder was developed, which contained micro-computer for controlling the current and voltage waveforms. The values of current and voltage can be shifted to any level ai any time during short circuit duration and/or arc duration. For the purpose of controlling the waveforms, a time period which contains a short circuit duration and an arc duration is divided into 6 stages, and also 14 factors of the waveforms which concern with current, voltage and time were selected. The waveform factors concern with the behavior of droplet at the primary stage in short circuit duration and at the later stage in arc duration had remarkable effect on sticking spatter in CO2 gas shielded arc welding. Using the new waveform control method, spatter in CO2 gas shielded arc welding decreased. Compared with the conventional method with thyristor controlled power sourse, sticking spatter on the nozzle and on the base metal by new method decreased to 1/5 and 1/50 respectively.
For the purpose of studying the welding operativeness of new waveform controlled CO2 gas shielded arc welder, some welding tests concerning with arc voltage, welding speed and wire brand were carried out. And relationships between bead shape and spattering were investigated. The new waveform controlled welder gives the decrease of sticking spatter on the nozzle and base metal compared with the conventional welders. However the spatter increases when the arc voltage is apart from the most suitable value at which spattering is the least, and when the welding speed is high. The selection of wire brand also affects the amount of spatter. The spatter increases remarkably when the current level at the later stage of arc duration is high. On the other hand, the current level at the later stage of arc duration decides the penetration depth. The penetration depth is deep when the current level is high. Therefore, there is a negative correlationship between bead shape and spatter.
Liquid phase diffusion welding of copper to austenitic stainless steel (SUS304) was studied. The surface of copper and SUS304 was cleaned by Ar ion beam bombardment and thin alloyed layer (Cu-30 wt.%Ti) was deposited on the bonding surface by sputtering. In this report, the effect of alloyed layer thickness on mechanical properties of liquid phase diffusion welded joints was investigated. The result indicated that: Mechanical properties of welded joints depended on alloyed layer thickness. Mechanical properties of welded joints equal to those of copper were obtained by the following condition: Alloyed layer thickness of 0.7μm, surface roughness of 0.7μm under, welding temperature 1213 K, pressure of 2.5 MPa and welding time of 3.6 ks.
A new chromium (Cr) metallization process was developed for brazing or soldering of SiC ceramic to metal. This report describes the characteristics of Cr metallization, bond strength and helium leak rate of the silver brazed layer. SiC ceramics used in the experiments was hotpress-sintered material containing a small amount of BeO. It has a high thermal conductivity (270 W/m.K) and also a high electrical resistivity (1014Ω.m). Silver brazing of SiC ceramics to metal was carried out at 1073K using eutectic alloy of 72wt% Ag-28 wt% Cu. The results of this study are as follows, (1) Cr metallization was achieved by printing Cr paste on the surface of SiC ceramics, and then heating the ceramics at a temperature between 1173K and 1273K for 1.8ks in Ar atmosphere. (2) The thickness of Cr metallized layer did not depend on the thickness of printed Cr paste under the same heating condition. (3) The bending strength of the SiC/SiC and SiC/WC-Co hard alloy, brazed layer were 320MPa and 360MPa respectively at 723K and fractured through SiC. The fracture remained in the brazed layer at temperatures above 773K. (4) The helium leak rate of the silver brazed layer (combination: SiC/A1203, SiC/kovar, SiC/WC-Co) was less than 1.2×10-10 atm.cc/s even after the thermal cycle testing.
In the previous report, development of Cr metallization process has been described. Silver brazing process of SiC ceramics was conducted at 1073K by using this process. Bond strength and Helium leak rate of the bonded layer was evaluated. In the present report, the reaction products in Cr metallized layer were investigated, and the mechanism of Cr metallization was clarified. The results of this study are as follows, (1) The reaction products of the Cr metallized layer were identified from the SiC side as three-layer compounds of a complex compound (Cr·Si·C), Cr silicide (Cr3Si, CrSi2) and Cr carbide (Cr7C3, Cr3C2). (2) The form of these reaction products was rectangle crystals, and the crystals were piled up like brick blocks. The grain sizes were from 1 to 1.5μm. (3) Activiation energy (Q) of the growth of Cr metallized layer was 102.9 kJ/mol. This is almost equal to the activiation energy of diffusion of C and Si into Cr. Therefore, the growth rate of Cr metallized layer is considered to be determined by diffusion of C and Si into Cr.
The bonds in mass-produced electric devices are recently required high reliability in precision and in quality. For the purposes, new bonding process for dissimilar metals by resistance heating is successfully developed. To control the material deformation, the optimum electrode constitution of materials, sizes and their arrangement is established as described in Report-1 and Report-2. To get high quality at bonds, the vapor shield process available to clean the material interface while bonding is developed as described in Report-3 and Report-4. Moreover, inprocess control system for getting high quality bonds stably is newly developed and automatic bonder for mass-production line of contact-bars is successfully manufactured as described in Report-5. In this study, main properties of contact-bars bonded by the new process which is operated in massproduction line are examined in comparison with the conventional one by brazing. The main results obtained are as follows; 1) Bond strength increases of 40% as compared with the conventional one. 2) Temperature rise at bond by arcing heat input to the contact-top is sharply suppressed. 3) Bond strength at high temperature of 973 K is three times as high as the conventional one. 4) Hardness of carrier bar required spring property is kept at higher value. 5) Noticeable longer life of electromagnetic switchgears is confirmed by JEM-standard test.
Utilization of conventional arc-welding robot by actual robot teaching (by a human operator) is not suitable for constructing a flexible production system, because applicability of the teaching data is limited to only the target robot. As a solution to this problem, the authors have proposed a compact computer aided welding system which consists of a personal computer and a conventional teaching-playback arc welding robot, and indicated the concept of off-line programming by a simple robot language. In the system, wled information processing is divided into four levels; task level, tool level, robot level and machine code level. At each level, necessary information is obtained from external files. In task level, source program of the weld is parsed and weld parts are extracted from the work model. In tool level, the motion of weld torch is generated on the work model at each weld line and between weld lines referring to both of work and torch models. In robot level, each angle of robot arms, which realizes required torch angle, is claculated considering the movable range of the robot. In machine code level, angle data of robot arms and weld conditions are combined together and converted into robot codes, which are transmitted to the robot controller through a communication line. Realization of torch motion in robot level is discussed for 5-axis articulated type robot. Possibility of required torch angle by the robot is checked by calculation of intersection between arm plane, in which robot's arms move, and movable range of torch rotation. Collision check and avoidance algorithm is shown, and its property was confirmed by an example program of the weld.
The three pixel vector method is a newly developed method of digital length measurement, whose principle was already reported. Some experiments for particle analysis are done to check its practical applicability and to compare its properties with other conventional methods. A test for this method is also done to check computer time required to process an image which has many particles, with along other methods. The practical usefulness of the or three pixel vector method is proved by these results.
Mild steel was welded in pressurized Ar, Ar-20%CO2 and CO2 atmospheres in the pressure range from 0.1-2.0 MPa abs. using a solid wire of 1.6 mm dia.. The effects of ambient pressure on the melting rate of electrode wire, the geometry of weld metal and the behaviour of alloying element during the arc welding process were studied. The results obtained are as follows: 1) In Ar atmosphere, the melting rate of an electrode wire decreases with the increase of pressure. On the contrary, in CO2 atmosphere, it increases with the increase of pressure. 2) In each atmosphere, the penetration width decreases, and penetration depth increases with the increase of ambient pressure. 3) Except for manganese, the loss of alloying elements during the arc welding process is little in pressurized Ar atmosphere. 4) In pressurized Ar-20%CO2 and CO2 atmospheres, Si, Mn, Ti loss are observed during the arc welding process. 5) The decrease rate of alloying elements depend on the CO2 partial pressures at a same ambient pressure. 6) In pressurized Ar-20%CO2 and CO2 atmospheres, the oxygen content of weld metal increases with increasing the pressure. 7) The oxygen content of weld metal made in Ar-20%CO2 atmosphere is lowre than made in pure CO2 atmosphere at a same CO2 partial pressure.
Thermal and mechanical testing approach is applied to evaluating the susceptibility to cold cracking of quenching crack type in welds of medium/high carbon low alloy steels. The simulation test was done with notched specimen during cooling of simulated weld thermal cycle by means of high frequency induction heating. The test shows the availability to get the correlation between temperature and fracture stress by applying tensile deformation from just above Ms temperature under different cross-head speeds corresponding to contraction rate in actual welding. The measured fracture stress decreases with increasing peak temperature, and that measured under the condition of peak temperature resulting in grain boundary liquation is nearly the same as that in the RRC test which was discussed in the previous paper. Grain boundary liquation promotes intergranular fracture. Moreover, intergranular facets with definite grain boundary liquation are more brittle, and are more similar to that in the RRC test than that without grain boundary liquation. Therefore, crack susceptibility of these materials can be evaluated with the simulation test using the specimen heated up to the liquating temperature of grain boundary.
The joint strength of brazed SUS 304 stainless steel has been investigated in relations to silver brazing filler metal composition and test temperatures. The strength was found to depend on the microstructure of silver filler metals. Joints brazed by β-phase type filler metals (with much β-phase) showed high tensile strength at low temperature (-196°C), the a-phase type filler metals, mainly consisted of solid solutions and their eutectic, exhibited medium strength, whereas the AgZn phase type filler metals with AgZn phase has low strength. The a-phase type filler metals showed the highest strength among these three types at higher temperatures. Above all the addition of nickel improved the high temperature strength. The relation between the joint strength at 400°C and zinc equivalence of silver filler metals showed aproximately negative linear relation, increase in zinc equivalence decreased the strength at 400°C. Nickel bearing copper base metal brazed with nickel free filler metals showed high strength at 400°C. The results of creep rupture test at 300°C were similar to the results of tensile test at high temperature. Joints brazed with high zinc equivalence (β-phase type) has inferior creep rupture resistance and a-phase type and nickel bearing filler metals offered longer rupture time.
An auther's previous paper had reported on the effect of heat-treating on the low temperature toughness of Type 316L weld metal. In this paper, the same study on Type 308L weld metal was carried out whose the heat-treating condition was at temperature from 923 to 1523K for time from 2s to 24h. In heat-treating the weld metal with δ ferrite above 2%, the absorbed energy decreased with increasing the amount of prior δ ferrite and particulary in the case of heat-treated temperature of 923K, degradation of absorbed energy was remarkable. This tendency was opposite to the result of Type 316L weld metal. This aws suggested that the addition of Mo was shifted Time-Temperature-Sensitization curve of the weld metal to the left side. In the other hand, the absorbed energy of fully austenitic weld metal was constant independent of heat-treating time and temperature. In the Type 316L weld metal reported previously, the absorbed energy tended to rise slightly by heat-treating for time up to 120s. In this Type 308L weld metal, however, such a tendency was not observed. This was considered to be attributed to somewhat good coherency of δ/γ intreface as compared to the case of Type 316L weld metal, in particular, in case of eutectic δ ferrite weld metal. The decomposition of δ ferrite in this Type 308L was somewhat slower than that in the Type 316L. This was considered to be due to good coherency of δ/γ interface.
The effect of molybdenum on notch toughness of austenitic stainless steel weld metal was investigated in detail. The weld metals where molybdenum content varied from 0.1 to 2.9% and δ ferrite content 0 to 10.8%, were prepared by submarged arc welding. The absorbed energy of fully austenitic weld metal was constant independent of molybdenum content. However, in the case of eutectic δ ferrite weld metal (EWM), the absorbed energy decreased monotonously with the molybdenum content and in particular, this tendency was remarkable in the weld metal of which the concentration of molybdenum within δ ferrite grain was more than 5%. In the weld metal of primary δ ferrite (PWM), the absorbed energy was constant up to 2.3% Mo and began to drop in the specimen above 2.9% Mo where the concentration of molybdenum within δ ferrite grain became more than 5%. These results revealed that if the concentration of molybdenum within δ ferrite grain exceeded 5%, the notch toughness Called drastically. This was considered to be attributed to following causes: the increase of molybdenum resulted in degradation of the coherency of δ/γ interface and moreovre, in the eutectic δ ferrite of EWM where the concentration of molybdenum was especially high, the formation of carbide of M23C6 was observed which affected harmfully the low temperature toughness, while in the primary δ ferrite of PWM the carbide was not observed at all. The multiple regression analysis revealed that each element of C, Cr and Mo had inherently a reducing action in toughness and particularlly, this action of molybdenum was stronger in EWM than in PWM.
This paper deals with the effect of mechanical and physical properties, and specimen size on residual stress in bonded dissimilar materials using Boundary Element Method (BEM). The sign of residual stress changes when the ratio of linear expansion (αB/αA) becomes greater than the unit value. The value of residual stress is chiefly influenced by the ratios of Young's modulus (EB/EA) and αB/αA, but is little influenced by Poisson's ratio (vB, vA). When the height of specimen is greater than the width of specimen, the change of height does not affect the residual stress in the neibourhood of bond line. When the width is greater than the height, the residual stress is not changed by the width.
The objective values of the yield strength and fracture toughness are respectively more than 1200 MPa and 200 MPa√m at 4K on the structural material for the superconductnig magnet container in the Fusion Reactor. It should be desireable to assure the above value for the properties of the weld metal used for the containers. However, it is very difficult to assure the both properties for the same weld metal, especially made by electron beam welding which has the very good cost efficency. It should be reasonable to use a little bit undermatched weld metal in order to assure the fracture tonghness. So, it is understood to be very important to know the applicability and permitted limit based on the design condition of loading and deformation for the superconducting magnet container. From this view point, the initial yielding behavior of the undermatched joint is investigated in this paper. Results are as follows. (1) On the nominal stress vs. nominal strain relationship at the central section of the undermatched metal of the joint, the stress increases higher than the yield stress of undermatched metal itself then gradually comes up to a constant value, after the whole area of the undermatched metal starts to generate plastic deformation as the stress reaches the yield stress of the metal. (2) The increase in the stress is caused by the restraint effect of the base metal on the plasticity, and the stress becames larger with the larger strength ratio of the undermatched metal to the basemetal and with the smaller relative thickness of the joint. (3) The strain does not distribute uniformly at the central section of the undermatched metal along the radial direction during the initial yielding process of the joint. (4) In the case that the yield strength of the joints is considered as the proof stress desided for a constant offset strain, the relationship between the dimensionless yield strength of the joint to the yield stress of the base metal and the relative yield strength of the undermatched metal to the base metal cannot be not established con-sistently. (5) The assumption that the base metal is rigid, causes the excessive evaluation of the joint yield strength of theoretical analysis. (6) As the quantitative evaluation method, the stress vs. strain relationship and then the algebraic expression has been formulated for the initial yielding behavior of the joint.
It is very important to analyze accurately and practically the stress concentration factor due to structural discontinuity, Kts, in evaluation of fatigue strength at weld toes. In this paper, a practical method for analysis of Kts was proposed with the aid of finite element method using plate elements taking weld bead effect into account. As a result, concerning Kts of the plate with rib under uniform tension and bending, it was found that the calculated values using the proposed model of fine mesh agree with those using brick elements with an accuracy of ±10 percent. For five types of welded joints, the good agreement was obtained between calculated stress distribution and measured one near the weld toe.
This paper deals with a proposal of equation to estimate the cooling time, t8/5, from 800°C to 500°C in various grooves and joints. It is shown that t8/5 calculated by an instantaneous line heat source becomes nearly equal to that calculated by a moving point heat source. The estimated equation for t8/5 in grooves and joints is obtained from the concept of thermal diffusion length and that in bead-on-plate. Experiments for examining the proposed equation were done under various welding conditions. It is concluded from the examination that the estimated equation is used without serious error.
It is generally impossible to non-destructively inspect spot welds of thin materials by the conventional normal beam technique of ultrasonic testing, because surface and flaw echoes appear so near together on the cathode-ray tube that these echoes can not be distinguished from each other. In this paper, the nugget formation in resistance spot welding of two-plate lap joints of 0.5 mm thick mild steel sheet is examined micrographically, and an improved technique of ultrasonic testing with an immersion-type focusing probe is tried to evaluate the welds. As a result of this study, the welding variables to produce sound fusion or solid-state welds without expulsion during welding process are determined. In the present ultrasonic technique under such a water distance, that the focus of ultrasonic beam locates beneath the weldment, the echoes from back reflection within the weld region and from faying surface reflection within the nonwelded region converge by multiple reflection, and the echo peak appears with an observable delay behind the front surface echo on the A-scan display. By this process, the back and interface echoes can be detected on the scanning inspection too. The scanning graph of the echo height of back and interface reflections versus probe location for each weld mostly records a composite pattern of protrusion and depression from the level of interface echo due to convergence of the back echo and scattering of wave by the indentation of surface respectively. The C-scan display drawn from these scanning graphs indicates shape of the weld region and data corresponding to the indentation.