In automatic butt welding of relatively thin plates, it is important to control welding conditions in order to obtain a sound full penetration weld. Especially, it is difficult to obtain a full penetration weld, because the optimum welding conditions change with groove conditions, such as gap of groove, shape of groove and so on. Recently, it was reported that there is an intimate relationship between the oscillation of the molten pool and penetration. The frequency of a molten pool decreases drastically with transformation from partial penetration to full penetration. Accordingly, estimation of penetration mode by detecting frequency of the molten pool is attempted. In this study, the principle of detection of penetration, detection method of molten pool oscillation and construction of system for estimating peculiar frequency of the molten pool oscillation are investigated. As the results of welding experiments, it is shown that the system constructed is effective in detecting molten pool oscillation and estimating penetration mode during TIG arc welding of thin plates.
The Ni-50Cr coatings are prepared under different plasma spray conditions which are arranged according to the orthogonal experimental design method using three typical parameters of plasma arc current, Ar flow and spray distance in order to obtain the quantitative regression correlations between spray parameter and coating mechanical properties, and examine the relationship between the mechanical properties and microstructure features. The mechanical properties of sprayed Ni-Cr coatings are characterized by hardness, abrasive wear weight loss and erosion rates. The microstructure of the coatings is examined by optical and electron scanning microscopes. The porosity in the coating is quantitatively estimated from by the increase in Cr content of the coating after the impregnation of Cr2O3 into pores of the coating. The effects of spray parameters on the mechanical properties of the coating are discussed according to regression formulas systematically. The results evidently show that three typical parameters have significant inter-effects on the microstructure and properties of the Ni-Cr coatings. There are evident correlations among the hardness and abrasive wear resistance, and porosity for plasma sprayed Ni-Cr coatings. With the decrease in the porosity of the coating, the hardness and abrasive wear resistance of the Ni-Cr coating are increased. On the other hand, there are no correlations between erosion resistance and hardness, and porosity as well. It has been found that the abrasive wear occurs mainly by micro-cutting of abrasives while the erosion occurs mainly by the successive separation of flattened particles exposed to the surface which are subjected to the direct impact of erosive particles. Therefore, the abrasive wear resistance is primarily dependent on the density and hardness of the coating, while the erosion wear is mainly dependent on the condition of the cohesion between flattened particles. The results also revealed that the deposition of partially melted particles into the coating deteriorates the erosion resistance of the coating, especially at low impact angle.
A numerical analysis in the inert-gas nozzle is performed with a novel injection concept using swirling flow in the laser nozzle, to control the mass transfer near the laser-irradiated area and outlet of the inert gas nozzle. Consequently, the following results were found. 1) By changing swirl strength, it is easy to control the flow pattern as well as the direction of the flow. 2) Swirl number has a critical importance in controlling the flow pattern near the outlet of the tornado nozzle. 3) Negative pressure near the laser-irradiated area can be easily obtained ; accordingly plume is sucked in the inert gas-flow, and thereafter pushed out to the atmosphere together with the gas. 4) Spatter going up towards the laser-mirror in the nozzle, is prevented through the going down inert gas flow, and pushed out to the atmosphere. Those findings mentioned above are very useful to control the flow pattern to remove the plume and spatters.
Narrow gap welding (NGW) joints offers many advantages over conventional welding methods, such as good mechanical properties of joints, high welding efficiency and low residual stress. As the groove gap width becomes narrower, the arc heat input can be reduced and the merits in narrow gap welding increases more. Generally, GMA welding method has been never applied to less than 5 mm groove gap, because it is guessed that it is arc instability and lack of fusion at the groove bottom area occur. In this paper, first of all, arc behavior under narrower gap joints is discussed, and it was concluded that the arc in MIG arc welding irregularly perturbates up-to-downwards along the groove wall under less than 5 mm gap, but CO2 arc was stable under narrower gap. Next, penetrations at the groove bottom area in CO2 arc welding were discussed. Characteristics of bead formation phenomena in CO2 buried arc welding of bead-on-plate were analyzed. From the results, the relationship between hydrostatic potential of molten metal and arc force corresponding with welding current was estimated. Furthermore, the width of gauging region of penetration by arc force was measured and the relationship between the melting width at groove bottom and welding conditions (welding current and welding speed) can be suggested. With these results, numerical simulation model was proposed and the optimum welding conditions to melt the groove bottom area sufficiently and to minimize heat input were searched by numerical simulation. And then narrow gap welding with 5 mm groove gap was carried out using these simulated welding conditions. In the experimental results, the weld bead was obtained without lack of fusion at groove bottom, but the convex surface bead was formed which is disagreeable in multi-pass welding. The new welding process was proposed from numerical simulations in order to prevent this convex bead and to obtain sufficient melting at bottom area. In the new process, the wire extension can be controlled by welding current waveform and then arc regularly oscillated up-to-downwards along the groove wall. In this arc oscillation, arc heating distribution along groove wall led to both sufficient penetration at groove bottom and concave surface bead shape.
Three kinds of Fe-8P-2C alloy powders containing 10 mass%Cr (10Cr), 20 mass%Cr (20Cr) and 10 mass%Cr-10 mass%Mo (10Mo), respectively, were thermally sprayed by a high velocity oxy-fuel (HVOF) process. The as-sprayed coatings of the 10Cr and 20Cr alloys are a mixture of an amorphous phase and a small quantity of crystalline phases, while as-sprayed coating of the 10Mo alloy is composed of only an amorphous phase. The anodic polarization curves of the coatings of the 10Cr and 20Cr alloys exhibited an activation-passivation transition in 1NHCl solution. On the other hand, the coating of the 10Mo allo, showed no active state and the lowest current density among the alloy coatings in the anodic polarization curve. The passive state of the coating of the 10Mo alloy was stable compared with the other coatings. The coating of the 10Mo alloy als, underwent no corrosion in 1NHCl solution at 0.5 and 1.0 V (vs. SCE), whereas Inconel 600 alloy plate was attacked remarkably.
Two types of intergranular fractures (type I and II) were observed in the specimens after the creep-rupture test on the synthetic HAZ of Cr-Mo steel. The types I and II fractures, respectively, may be induced by the weld-thermal-cycle, and by some mechanism of secondary segregation. The authors assumed that the type I fracture will be brought about by the segregation induced by the α/γ transformation ("the transformation-induced segregation") during welding. The possibility of arising the transformation-induced segregation was examined by the experiments on the synthetic HAZ and also by the calculations. The grain boundaries of prior-austenite was analyzed by using two methods ; "the EDX analysis in a micro-region" and "the grain boundary etching method". The experimental results informed that phosphorus was concentrated in the grain boundary at the time when the α/γ transformation occurred. The greater the heating rate, the larger the quantity of phosphorus in grain boundary. The segregation of phosphorus will be induced by a great heating rate such as 1000 K/s. The mechanism of this type of segregation could be explained from the view point of the distribution of phosphorus in transformed phase (austenite) and original one (ferrite). Phosphorus is accumulated at the α/γ interface as the distribution coefficient ko (=[P]γ/[P]α) is smaller than the unit.
Charpy U-notch impact test was conducted on brazed joints of type 316 stainless steel. Two kinds of newly developed Ni-Cu based filler metals, two kinds of standard nickel filler metals, and a standard silver filler metal were tested. For each of the filler metals, brazing time was varied from 600 s to 7200 s under a constant brazing temperature. Each of the joints were made to have a constant joint clearance of 50 micrometers. The joints with the nickel filler metals showed brittle fracture and their absorbed impact energy were lower than 2 J for the brazing times up to 3600 s. Absorbed energy of the silver filler brazed joints were about 20 J although their fracture surface showed microscopic ductile fracture. The joint with Ni-Cu-Mn filler metal, one of the developed filler metals, showed 120 J of absorbed impact energy for the brazing at 1333 K for 3600 s. By micro-hardness measurement across the joint, the hardness of this joint was found to be practically identical with that of base metal. This implies that the excellent impact strength of this joint should be attributed to the uniformity in mechanical properties across the brazement to the base metal.
The effect of Ni content on weld hot cracking susceptibility of Fe-20%Cr-Ni alloys, which solidify as austenitic single phase, was investigated. In addition, the effect and effect mechanism of alloying elements on weld hot cracking susceptibility of such alloys was studied in the varied ranges of Ni constitution. A theory that can explain general effect and effect mechanism of alloying elements on weld hot cracking susceptibility has not completely been established in the different ranges of major constituents. However, there are many reports concerning the improvement of hot cracking susceptibility in only limited ranges of major constituents. As an experiment, a Trans-Varestraint test and hot tensile test were conducted. The influence of alloying elements on solidification temperature range and liquid fraction during solidification was also investigated using Thermo-Calc. Solidification sequence of Fe-20%Cr-Ni alloys was also investigated by freezing the solidification microstructure with liquidtin quenching during gas tungsten arc welding. As a result, the maximum hot cracking susceptibility of Fe-20%Cr-Ni alloys was observed in the Ni content range from 30% to 60%. Therefore, dependence of hot cracking susceptibility on Ni content can be explained qualitatively by the effect of increase in solidification temperature range around 45%Ni.
We investigated the influences of bonding parameters on joint strengths of diffusion bonds between swaged tungsten and titanium (case A), and between rolled tungsten and titanium (case B). In case of A, joints bonded at temperatures of 1073 K and 1473 K for a bonding time of 1.8 ks showed the tensile strengths of 120 MPa. The joint strength obtained at a bonding temperature of 1073 K increased with increase of bonding times, and for a bonding time of 30 ks reached about 240 MPa, which was 88% of the tensile strength of the tungsten base metal. In the case of B, joint strength obtained at a bonding temperature of 1173 K was 16 MPa, when fracture occurred along grain boundaries. In the both case, the voids were formed in titanium base metal away from the W/Ti interface. It is therefore seemed that the existence of voids give no effect on the joint strength.
Precipitation behavior of sigma (σ) phase was investigated for duplex stainless steel and its weld metals by reheating. The material used was SUS329J1 base metal, and the samples were single-pass welded using SMAW 329J4L (DP3) electrode and SMAW 329J3L (DP8) electrode. The fine secondary austenite (γ*) was formed at primary austenite (γ) /ferrite (α) phase boundaries and grew into ferrite, and then σ phase began to precipitate at γ*/α phase boundaries and grew into ferrite for SUS329J1 base metal. In case of weld metals, γ* was formed both at γ/α phase boundaries and in the interior of ferrite matrix, and thereafter precipitation of a phase at γ*(γ)/α phase boundaries occurred. Time-Temperature-Precipitation (T-T-P) diagrams of sigma (σ) phase were constructed in order to investigate beginning of precipitation of sigma (σ) phase for SUS329J1 base metal and 329J4L (DP3), 329J3L (DP8) weld metals. As a result, the nose position in TTP diagram of σ phase was located at 1123 K for 0.06 ks for 329J4L (DP3) weld metal, while its position was located at 1123 K for 0.3 ks in case of 329J3L (DP8) weld metal which was lower chromium content.
The formation mechanism of vermicular ferrite and lacy ferrite observed in austenitic stainless steel weld metal solidified in the FA mode was investigated. The formation of vermicular ferrite or lacy ferrite is not decided by chemical compositions but it is decided by both the crystallographic orientation relationship between ferrite and austenite established at the stage of ferrite nucleation and the relationship between the welding heat source direction and the preferential growth directions of ferrite and austenite respectively. On the basis of the results, the effect of ferrite morphology on a cryogenic impact toughness and a pitting corrosion resistance of weld metal was investigated. In the case of almost the same amount of ferrite, as the proportion of lacy ferrite increases, the cryogenic impact toughness and the pitting corrosion resistance increase. Consequently, even though the identical chemical composition of stainless steel, it is possible to improve a cryogenic toughness and pitting corrosion resistance by controlling the ferrite morphology.
YAG laser welding was performed on A5052-O and -H34 sheets (thickness: 2 mm) in several different welding speeds with the constant out put power of 2000 W. The authors investigated the distribution of Vickers hardness in both weld metal and heat-affected zone (HAZ), and the influence of heat treatment on Vickers hardness in the weld metal from the view point of micro segregation of magnesium in dendrites. In the HAZ of A5052-H34, the soft region where the hardness is almost the same as A5052-O base metal is developed just near the fusion boundary. Then the hardness increases with the increase in the distance from the fusion boundary toward the base metal and finally reaches the value equal to A5052-H34 base metal. This is caused by the change of the maximum temperatures in the HAZ due to welding. In the weld metal of both A5052-O and -H34 the hardness increases and dendrite arm spacing decreases as welding speed increases. When the heat treatment is performed on the weld metal at 500°C for 24 h, the hardness decreases down to the value of A5052-O base metal. According to the simulation on the diffusion of magnesium in aluminum during the heat treatment, the micro segregation in dendrites tends to decrease with the increase in both the heat temperature and the holding time. This phenomenon was confirmed by EPMA analysis. These results lead to the conclusion that the higher hardness in the weld metal than that of A5052-O base metal is caused by the micro segregation of magnesium in dendrites.
The friction welding for copper and aluminium is useful to minimize the diffusion thickness at the interface between copper and aluminium. However, it is generally difficult to maintain the joint strength. Because brittle intermetallic compounds are necessarily formed at the friction weld interface between copper and aluminium. In this study, stress singularities at free-edge and intersection of copper/aluminium interface were investigated using finite element analysis. It was informed that the no free-edge stress singularities existed below an apex angle of 59°and between angles of 98°and 129°Also, intersection stress singularities exist over all apex angles except an apex angle of 90°.However, in case of apex angles between 98°and 129°, compressive stress concentration existed near the intersection of copper/aluminium interface. In this experiment, copper and aluminium friction welded joints were fabricated as to minimize (>1μm) the welding condition of the diffusion thickness at the interface between copper and aluminium. It was confirmed that the good tensile strength could be observed in all specimens tested. However, the Charpy impact energy became comparatively low in case of an apex angle of 90°and 135°, and showed a remarkable high impact energy can be obtained in case of no free-edge stress singularities.
The relaxation behavior of welding residual stresses by both reversed and repeated loadings was experimentally investigated using austenitic stainless steel JIS SUS316 and weldable high strength steel of JIS SM490A. The weld bead was deposited by electron beam welding on the plates of both materials along their longitudinal center lines. After welding, small size longitudinal welded joint specimens of SUS316 and medium size ones of SM490A with the center hole notch were cut out from the welded plates. Eighteen small size welded joint specimens of SUS316 were subjected to reversed loading and 16 medium size ones of SM490A were subjected to repeated loading with the parameter of maximum stressand number of cycles imposed. Then the residual stress distributions of the plates were measured by the cut-release method. The result showed that welding residual stresses at the notch root were considerably relieved by the first one cycle of both repeated and reversed loadings and then gradually relieved with the following loading cycles. Also found was that welding residual stress in as-welded condition was more effectively relieved by repeated loading than reversed one. In addition to the relaxation test, comparative fatigue tests were carried out of center hole-notched longitudinal welded joint and base metal specimens. The fatigue test result led to the conclusion that the fatigue strength of the longitudinal welded joint of SM490A by repeated loading decreased approximately 15 MPa than that of the base metal in the life regime up to 8×106 cycles. On the contrary, the fatigue strength of longitudinal welded joint of SUS316 by reversed loading only decreased in the high cycle regime over than 105 cycles.
Surface treatment by laser irradiation leads to softening in a surface region for Ni-base superalloys. This paper investigates the utility of surface softening to reduce a stress concentration at the weld toe of Ni-base alloy joints. Three-dimensional FE-analysis is conducted to obtain the stress-strain fields of the welded joints. It is revealed that surface softening is effective to reduce the stress concentration at the weld toe. The relaxation of stress concentration is almost proportional to the degree of softening. The demerit of surface softening is found in an elevation of strain level at the weld toe. However, the strain concentration can be marginal in the applied stress level below 0.5σYB(σYB: yield stress of base metal), at which initial yielding occurs at the weld toe. It is expected from the results that surface softening will improve a high-cycle fatigue property of welded joints.
Investigating the general conditions which the specimen to evaluate the susceptibility of reheat cracking must have, a series of processes from welding to PWHT (Post Weld Heat Treatment) were modeled by 3D thermal elastic-plastic-creep analysis based on FEM for the reheat cracking test specimen used in this study. Noting the components of the welding residual stress owing to the test bead itself, the component in the direction of weld line, σx, was remarkably large and another two components perpendicular to weld line, σy, and thickness direction, σz, were small. Performing the third layer of restraint beads, σy, on the position where the reheat cracking was anticipated was large because the bending moment in the cross section became large due to the shrinkage of the following restraint beads. After the fourth layer restraint beads, σy, was not changed. This is because the bending stiffness in the cross section of the specimen became large. Performing the PWHT without whole restraint beads, σy was largely relaxed without accumulation of creep strain because the bending moment in the cross section was released. So, the evaluation of the reheat cracking susceptibility using the specimen should be performed after the whole restraint beads was laminated. When the susceptibility of the reheat cracking caused by PWHT is evaluated by the specimen used in this study, the test bead welding should be done remained the stress concentration in coarse grained HAZ but the restraint bead welding which is performed from back side should be full penetration. The plate thickness of restraint beads side should be at least four layers, which become to be full penetration.
In order to clarify the reliability of Au wire bonds to Al pads, failure causes such as void formation and corrosion behavior of Au-Al intermetallics were investigated in accelerated tests. The effects of annealing environments, temperatures and bonding conditions on bond reliability were also examined. The molding resin has a great influence on the corrosion behavior, GC-MASS (gas chromatography) analyses of outgasses produced from the resin revealed that a major Br-containing species was methyl bromide (CH3, Br). The corrosion reaction between the intermetallics and bromides produced minute Au precipitates and amorphous Al oxides. This corrosion behavior caused significant fluctuations in electrical resistivity. Non-uniform formation of intermetallics at the bonds accelerated the corrosion behavior. It was confirmed that initial conditions at the bond interface affected bond failure. Activation processes of void-induced and corrosion-induced failures were compared in the respect of bond lifetime. The void formation was controlled by an interdiffusion process, whereas the corrosion behavior was controlled by a decomposition process of the resin. Au/Al bond failure could be governed by the intermetallic corrosion at higher temperatures and by void formation at lower temperatures.
The effect of Mg on the deoxidization of the Al2O3 surface film has been examined by means of the investigation on the chemical bonding changes at the topmost-surface of the aluminum alloy powder particle at the elevated temperature by using SR-XPS analysis. It was clarified that the diffusion of Mg solved in the matrix to the A12O3 surface film started in the aluminum alloy powder with Mg at 573 K, and the deoxidization reaction of the Al2O3 surface film by Mg occurred over 670 K and it completed around 774 K. SEM observation showed that the holes caused by the deoxidization of the Al2O3 film existed at the particle surface. On the other hand, the aluminum alloy powder without Mg did not show such a deoxidization during heating to 825 K and its surface was covered with the Al2O3 film.