The world's first 25 kW laser welding machines at its 24-inch pipe mill has been installed. The technology for the high-power laser pipe welding process that assures welds with the same properties as those of the base metal while maintaining high productivity has been started to develop. Key technologies for this process such as high power laser welding, high-accuracy seam tracking and nondestructive inspection have been successfully established. The results obtained are as follows. (1)The productivity of this process is much higher than conventional fusion weldings, e.g., SAW for less than steel plate 16 mm thick. (2)High energy density of laser beam helps to produce finely distributed oxides. (3)Fine oxide particles with a size of 0.1-0.2 μm were found in weld metal. And they also prevent the heat-affected zones from deteriorating low temperature toughness and so on, a problem that occurs in conventional fusion weldings. (4)The performance of the welds was as sound as the base metal by using the heat treatment process after high-power laser welding. (5)The characteristics of low temperature toughness and SSCC etc. of the laser welds are almost at the same level as the base metal.
The mechanism of porosity formation in CO2, laser welding of thin steel sheets was investigated. Laser melt-runs were made in ultra-low-carbon steel sheets of 0.7 mm in thickness under a wide range of shielding gas conditions. In addition to the normal melt-run condition with Ar coaxial shielding gas and Air atmosphere at the backside, the nitrogen gas content was varied in coaxial shielding gas as well as in backside atmosphere. The number of pores in weld metal was measured by radiographic tests, and the nitrogen content in weld metal was analyzed. The composition of gas in the pores was also quantitatively analyzed. These experimental results lead to an understanding that nitrogen in the backside atmosphere dissolves into molten weld metal and during solidification the dissolved nitrogen forms porosity due to an abrupt decrease in the nitrogen solubility. It is considered that nitrogen molecules are dissociated in the laser induced plasma at the backside, enhancing the dissolution of nitrogen into weld metal.
Joining processes for Bi-2223 multi-filament tapes in Bi-family superconductor tapes were investigated, in order to develop a technique that can obtain the joint with superior mechanical and superconductive characteristics. The superconductive characteristics of the diffusion-bonded joints were better than those obtained by soldering, but did not reach the level of the base metal. Critical current of the lap joints was a third of the base superconductor tape. The electric properties of the lap joints were superior to the scarf joints, and the application of the superconductor materials as interlayer to bonding zone did not improve the electric properties of the scarf joint. Although the lap joint had a silver layer with a normal conductive state at the bonding interface, the silver layer was so thin that it had only slight influence on the electric property. The diffusion-bonded joints were broken at the base metal. Though the electric property of the base materials was depressed during the diffusion bonding process by heating and pressing superconductor tape, it had a recovery with bonding temperatures from 700 to 800°C.
A free failing experiment has been conducted as a simulation of a thermal spray process to investigate the effect of substrate temperature on wetting at splat / substrate interface. A gold-coated substrate was prepared by ion sputtering to change the wetting conditions. The experiments were conducted both in atmospheric pressure nitrogen and under low-pressure condition. The grain size in the cross-section microstructure of the splat obtained under low-pressure condition was quite small even on the room temperature substrate and the bottom surface microstructure was quite dense. It was found that the ambient gas has a physical restraint effect on the wetting by introducing the pores at interface. The grain size in the cross-section microstructure of the splat on gold-coated substrate was also small and the bottom surface microstructure was quite dense. It was recognized that both materials factor and temperature factor affect the wetting at interface with substrate temperature increasing. The transition phenomenon in the wetting with substrate temperature increasing was indicated to be due to materials factor induced wetting, such as a surface roughening by an oxidation of the substrate.
The reactive thermal spraying is expected as the reasonable process of fabricating the sublimable particle dispersed coatings. The purpose of the present study is to develop the coating by the reactive spraying which is superior in the wear-resistance to Cu-alloy coating. For this purpose, both Cu-alloy/Ti pre-mixed and mechanically alloyed composite powder were sprayed to fabricate the titanium nitride dispersed coating by the reactive spraying and the wear characteristics of the coatings were evaluated. The results obtained are summarized as follows. 1) It was found that the titanium nitride dispersed coating can be made by the reactive spraying of pre-mixed powder. The amount of nitride phase could be controlled by the deposition rate. 2) The wear-resistance of the nitride dispersed coating was superior to that of Cu-alloy coating and improved with the dispersed particle content. 3) The titanium nitride phase was formed by the reactive plasma spraying of MA powder. The finer and more homogeneously dispersed composite coatings were established by using the MA powder. 4) The coating made from MA powder was superior in wear-resistance to that made from the pre-mixed powder.
In the arc welding, spatters attached around the bead of base metal. The adhered spatters lower the surface quality of the welding product. Therefore, it is very important to remove the spatters from the welding product. In general, wire brushes of bevel type or cup type have been used to remove the spatters. However, the adherent force of spatters on base metal is not constant. Then, the removal force, which is needed to remove the adhered spatters, is not also constant. Moreover, spatters are scattered on the base metal. Therefore, in order to remove the spatters by the brush, it is very important to clear the relationship between brushing conditions and the removal behavior of spatters. In this study, the removal forces of the adhered spatters are measured by using the removal tool to which strain gages attached. The shapes of the adhered spatters are also measured. Moreover, the force that is applied to the spatter in the brushing has been measured by using a model spatter. The shape of the model spatter is very similar to that of the adhered spatter. A steel plate with scale is used as a base metal. In brushing the metal, the pushing force of the brush is 29.4 N and the holding angle, that is the angle between the brush and base metal, is 15, 20, 25, or 30 degrees. As the result, it is indicated that the applied force to the model spatter in the brushing increases with the height of model spatter proportionally. The holding angle of the wire brush has little influence on the applied force to the model spatter. When the temperature of base metal is less than about 450 K, most of the removal shear forces of spatters are less than about 50 N. In this case, all of the attached spatters can be removed by the brushing. However, when the temperature of the base metal become high, the removal shear force become large and it become hard to remove all of the adhered spatter.
We tried to braze commercially pure titanium (CPTi) at the temperature below its transformation temperature of 882°C using Ag-based filler metal in argon atmosphere. The effect of brazing temperature, isothermally holding time and brazing pressure on the strength and the microstructure of brazement was examined. In addition, ultrasonic vibration was applied during brazing in order to improve the performance of the brazement. The following results were obtained in this study. By using BAg-7 filler metal, CPTi could be successfully brazed below its transformation temperature, and the brazement had the strength equivalent to the base metal. The strength of the brazement increased with brazing temperature and isothermally holding time at that temperature. This was because brittle intermetallic compounds disappeared from the brazed region. The disappearance of the intermetallic compounds accelerated with the increase of brazing pressure because extra molten filler metal could be more effectively expelled from the brazed region. Application of ultrasonic vibration during brazing was capable of increasing the strength of the brazement at lower temperature and lower brazing pressure for shorter holding time than when no ultrasonic vibration was applied.
As is generally known, arc sensors do not require special sensor elements, and feature a number of practical advantages such as welding current/voltage-based detection of deviation in torch aiming from the groove center. However, conventional arc sensors suffer from erroneous detection in cases where gap width within the groove is uneven, making it impossible to detect deviation in torch aiming. This paper describes the application of a high frequency oscillation arc (developed by the authors) to the development of a new type of arc sensor capable of detecting torch aiming deviation regardless of the gap width within the groove, as well as allowing detection of gap width itself. First, based on quantitative analysis, consideration was given to the arc characteristics of the high frequency oscillation arc in comparison with conventional sin curve oscillation. Then, arc sensor principles for simultaneous detection of torch aiming deviation and gap width were elucidated. The results obtained are as follows. 1) The welding current and voltage characteristics of the high frequency oscillation arc feature the coexistence of a dynamic state in which self-regulation of the arc does not function, and a static state governed by self-regulation effects. 2) Frequency characteristics have been revealed for welding current/voltage amplitude with respect to the high frequency oscillation arc for welding in V-type joint grooves. The low frequency range of welding current amplitude tends to decline steadily with increases in the maximum frequency, and no critical frequency phenomenon is displayed in sin curve oscillation. Welding voltage amplitude is virtually constant with respect to frequency, and there is no frequency dependence. 3) Quantitative analysis illustrates arc sensor principles for the simultaneous detection of deviation in torch aiming and gap width using the high frequency oscillation arc. 4) In feedback control of an actual welding robot system equipped with the arc sensor, the gap width was successfully detected during seam tracking with a maximum gap of 3.5 mm.
Robotic welding is applied in many manufacturing industries due to higher welding efficiency and improvement in the working environment. In ships and bridges, however, the latter stages of the manufacturing process often involve large three-dimensional structures featuring tight spaces and poor groove precision, and the use of robotic welding is difficult. The main issues to be addressed in these cases are automation technology for groove sensing, as well as superior handling. In order to solve these issues, A compact, lightweight and high frequency oscillation arc method has been developed. Initial report on this subject confirmed the operations of an oscillation mechanism (maximum oscillation of 40 Hz) internalized in the torch and weighing less than 1 kg, and clarified the bead formation phenomenon in horizontal fillet welds. Second report discussed a new arc sensor that utilizes arc characteristics unique to this welding method so as to enable simultaneous detection of deviation of torch aiming and groove gap. In continuation, this paper here focused on development of an automatic control system with welding speed as a control parameter, based on the arc sensor. Using this newly developed automatic welding control system, stable weld quality was obtained for gap variation up to 3 mm. The method under discussion was applied together with this automatic control system for the further development of self-propelled automatic welding robot that is both compact and lightweight, and which features superior handling.
The factors affecting HAZ crack susceptibility were investigated for castalloy 718. The materials used were laboratory-cast alloy 718 plates with three levels of grain size. The alloys were heat-treated at 1368 K×5.4 ks or 1378 K×10.8 ks before welding. Welding was conducted by a CO2, laser apparatus with a variation of laser power and welding speed. Crack susceptibility in HAZ was evaluated by a bead-on-plate test. The penetrated shape in the laser weld was classified into types of a wine cup-like Type W and nail head-like Type N. Type W and Type N were observed at the condition of lower and higher welding speed, respectively. Welding defects such as porosity, underfill and microcracking were seen in the laser welds as well. Most of the HAZ cracking occurred on grain boundaries in HAZ adjacent neck region of Type N weld metal. The fractogra-phic investigation clearly suggested that the HAZ cracking is produced by the presence of a liquid film along grain boundaries. According to the crack test, it wasfound that the HAZ cracking susceptibility decreases with decreasing the grain size or homogenization heat treatment, and with increasing a radius of curvature in the neck zone. These results suggested that the HAZ cracking susceptibility can be improved by the control of penetrated shape to be Type W, a decrease in grain size and a appropriate heat treatment.
In order to study the mechanism of liquation cracking in cast alloy 718, a mathematical model was developed to calculate grain boundary liquation in weld thermal cycle. According to microscopic observation of specimen heated at high tempera-tures, the laves cluster has started to melt at about 1383 K. Thus, it was considered on modeling that the grain boundaries were liquated by the melting of laves cluster which were located at the center of the grain boundaries. In this model, two factors were considered for solute diffusion in solid and liquid phase, namely, (1) high diffusivity at the grain boundary region, (2) the balance between the grain boundary energy and solid/liquid interfacial energy. Calculation has shown that the experimental values for the amount of grain boundary liquation fairly agree with the calculated results where the grain boundary diffusivity is to be the ten times value of the bulk solid diffusivity. The calculation has also indicated that the liquid strongly infiltrates along grain boundary with increasing the heating temperature, and with decreasing the heating rate. In addition, the amount of grain boundary liquation increases, as the grain size and laves cluster increases. On the basis of these calculations, it has been elucidated that the liquation cracking in cast alloy 718 strongly depended upon the factors, such as weld thermal cycle, grain size and the amount of laves cluster, etc..
The authors have pointed out in preceding papers that four types of temper embrittlement arise in the temperature range below 900 K when the heat affected zone of 2 1/4Cr-lMo steel is reheated. The causes of these temper embrittlements were examined in the present paper. The first type of embrittlement which arises at 775 K was compared with "the low temperture temper embrittlement" which has been known to arise at 575 K and to be caused by the fine precipitates of cementite. Experimental results on a series of specimens which were tempered at each temperature from 575 K to 775 K showed that both types of temper embrittlement are identical. The second type of embrittlement, which arises at 825 K in the short term range and produces cleavage fracture by the impact test, was found to be caused by the hardening of ferrite which was induced by the fine precipitates of Mo2C type carbide. The thirdd type of embrittlement arises at temperatures around 825 K and produces intergranular fracture by the impact test. A new technique for analyzing the grain boundary revealed that the segregation of phosphorus caused this embrittlement. The fourth type of embrittlement, which arises at 875 K in the long term range and produces cleavage fracture, was found to be caused by the large sized particles of Mo2C type and Cr, C, type carbide. In the impact test, fracture occurred preferentially along the planes in which carbide particles of large size exist. The proper terms of the first to the fourth types of embrittlement were proposed by the authors : the low-temperature-, the short-term-, the intergranular- and the long-term temper embrittlement.
Crack propagates in Al2O3 when Al2O3 is bonded to SUS304 by Ag-Cu-Ti brazing. This crack is due to residual stress in Al2O3. Bonding Al2O3 to SUS304 by pulse current heated bonding after brazing Ag, Cu thin plate to Al2O3, SUS304 was studied for the purpose of prevention of this crack. Effect of residual stress relief was evaluated with thermo-elasto-plastic analysis by FEM and joint properties were investigated by bonding test. Main results obtained are as follows : (1) It was confirmed with FEM that residual stress in Al2O3 could be reduced by proposed bonding method. Insert of soft metal (Ag, Cu) and thermal load reduction by pulse current heated bonding are effective for residual stress relief in Al2O3. (2) Crack in Al2O3 doesn't generate and excellent joint strength is obtained by proposed bonding method. Joint strength obtained by optimum bonding parameters is 119-122 MPa (average value) at 4 point bending test.
The failure cases in plants by microbially influenced corrosion (MIC) of stainless steels were often observed around the welded joints. It is significant to understand the underlying reasons that influence the susceptibility of weld regions of stainless steels to MIC. Experiments were carried out to investigate the effect of the shape of the weld bead on the adhesion of bacteria and the occurrence of pitting corrosion. Coupon exposure studies were conducted using Type 304 stainless steel including the weld bead with TIG using type 308 filler. Coupons were immersed in the bacterial culture medium and then the bacteria (Methylobacterium sp. or Bacillus sp.) were inoculated. After a fixed period of incubation, the bacteria adhered on the surface of the specimen were observed by epifluorescence microscope. In the case of reciprocal shaking condition, the adhesion area of bacteria at HAZ and the toe region were more than those at base metal and top of weld region. In the case of non-shaking condition, the bacteria were adhered uniformly at base metal, HAZ, toe and top of weld metal. After 60 days of incubation, the surface of the specimen was observed by SEM. Pitting corrosion was observed at the toe region. The austenite phase was preferentially attacked and δ-ferrite phase was retained like a skeleton. Bacterial adhesion correlated with the occurrence of corrosion. Therefore the weld bead shape is considered to have significant influence on bacterial adhesion and in turn, MIC occurrence in stainless steel. (250 words)
The information of the creep behavior of the thick welded joint is very important to secure the safety of the elevated temperature vessels like the nuclear reactors. The creep behavior of the thick welded joint is very complex, thence it is difficult to practice the experiment or the theoretical analysis. A simple accurate model for theoretical analysis was developed in the first study. The simple model is constructed of several one-dimensional finite elements which can analyze three-dimensional creep behavior under a assumption. The model is easy to treat, and needs only a little labor and computation time to simulate the creep curve and local strain of the thick welded joint. In this second study, the capability of the model is expanded to estimate the creep life of the thick welded joint. New model can easily estimate the time of the rupture of the thick welded joint. It is verified by comparing the result with the experimental one that the model can accurately predict the creep life. The histories of the local strains to the rupture time may be observed in the simulation by using the model. The information will be useful to improve the creep characteristics of the joints.
The fatigue strength of welded joints has been investigated in synthetic sea water. The welded beam which contains yield strength magnitude tensile strength has been fatigued at R=0.1. The small welded specimens of which residual stress is not high enough to be yield strength has been tested at σmax=σy condition. The results of both tests coincided well for each other. The automatic detection of fatigue crack initiation life has been developed by using eddy current detector. This method was effective even in synthetic sea water. It has been revealed that the fatigue crack initiation life shares 10% in fracture life in synthetic sea water, and 20% in air.
Based on the interface element proposed for crack propagation problems, a finite element method (FEM) using a temperature dependent interface element is developed. The proposed method is applied to the analysis of the formation and the propagation of hot cracking in welding. In particular, the hot cracking extending from the starting edge of the narrow trapezoidal plate under bead welding is examined using the proposed method. In case of the trapezoidal specimen with welding from the narrower side, the crack length becomes longer when the welding speed and the heat input is large. This agrees with the phenomenon observed in experiments. In case of the specimen with welding from the wider side, the welding speed and the heat input influence in the opposite manner. Further, the effect of the parameters involved in the proposed method on the computed phenomena is closely examined. Through this study, it is found that the hot cracking simulated by the proposed method is mainly influenced by the BTR (Solidification Brittleness Temperature Range) and the scale parameter r0. This agrees with the concept proposed by Matsuda through high temperature ductility curve.
The fatigue strength of box welds was tripled by using low transformation temperature welding wire and postweld heat treatment (PWHT). The improved fatigue strength was obtained due to the stress ratio effect resulting from compressive residual stress. The low transformation temperature welding material which contains 10 mass % Cr and 10 mass % Ni expands during the final cooling process of welding. This expansion of weld metal induced the compressive residual stress around the welded part. The PWHT at 720°C for 1 hour and subsequent cooling in air induced the transformation from austenite to martensite at the same time for three passes. That is, the volume expansion on three passes by PWHT became large compare with that on final pass in as-welded condition, and the compressive residual stress could be induced efficiently.
The effect of Mg on the sintering behavior of the aluminum alloy powder particles has been examined by using the deoxidization reaction of the Al2O3 surface film by Mg. When sintering an aluminum alloy green compact with relative density of 88%, the relative density of sintered material with and without Mg was 97% and 94%, respectively, that is, Mg was remarkably effective on sintering between aluminum alloy powder particles. It is considered that the solid phase diffusion of the aluminum atoms occurs at the interface between the primary particles during heating due to the deoxidization of the Al2O3 surface film by Mg. When the content of Mg increased, the pores located at the triangle junction of the primary particle boundary decreased and the mechanical properties, such as the tensile strength and elongation, were remarkably improved. The area of the dimple patterns at the fractured surface of the tensile specimen also increased.
Direct nitriding process for in-situ formation of AlN has been developed by using the reaction between solid-state metallic Al and nitrogen gas after the deoxidizing reaction of the Al2O3 surface film by Mg contAlned in aluminum alloy powder particles. In-situ reacted AlN grows on the particle surface and has a layer structure with a width of 20-30 nm. A good bonding is observed at the interface between in-situ reacted AlN and the aluminum matrix of the primary particle. The Direct nitrided and sintered Al-AlN composite material shows superior mechanical properties, such as ultimate tensile strength and elongation, to the Al-AlN composite material produced by the conventional premixing process. This is because the hardening effect due to the fine AlN dispersions in the matrix and the good bonding at the interface between in-situ reacted AlN and the aluminum matrix.