In the cutting, required heat input density for melting balances with the heat flux from heat source on the cutting front surface. If the required heat input density and the heat flux values were given, the cutting front shape and kerf shape can be estimated by calculation, because the both values were decided by the numerical simulations described in the Report 1 and 2 of this series. So, the authors tried to make a new algorithm to estimate the cutting front shape and kerf one by using the previous described information for the required heat input density to melt the cutting front and the heat flux value from the arc to the cutting front surface. First, heat source model was examined. Then, the heat source could be supposed to be exponentially decreasing type. Secondary, it was fixed that the kerf shape can be calculated by the heat balance at the maximum kerf width point of the solidification line on the cutting front. Finally, it was clarified that not only kerf shapes but also gouging depth in cutting can be estimated by the new simulation algorithm.
Radial distribution pattern of heat source can be supposed to be influenced not only on the kerf shape but also on the gouging characteristics, that is based on the previous our reports that are Report 1 and 2 of this series. And then, it was clarified that the heat source can be assumed to be exponentially decreasing type along the flow axis. Moreover, the authors have made the new simulation algorithm to estimate the kerf shape and gouging depth at various cutting conditions in the Report 3 of this series. So, the relationships between the heat distribution pattern and the cutting characteristics were examined in the present study with the simulation program that was described in the Report 3. Gaussian type, a truncated corn type and a near square type were selected for the radial heat distribution model of heat source in the calculation. The simulated results were verified with some experimental results with various sorts of working gas ; argon, argon-hydrogen, nitrogen and oxygen that are used to change the radial heat distribution pattern of heat source. As a result, it was clarified that the radial heat distribution pattern remarkably influences the kerf shape and the relationship between cutting speed and gouging depth. The bevel angel of kerf was improved by using a heat source of near square distribution pattern. And then, in this case of heat source, the gouging depth (cutting depth) and cutting speed was in almost inverse proportion to each other. Moreover, it was made clear that the differences of cutting characteristics among various sorts of gas can be explained by the radial heat distribution pattern.
This paper describes a guide line to make low bevel angle cuts in plasma arc cutting by changing of only nozzle conditions ; nozzle shape, electrode tip position in the nozzle and torch height (standoff distance). The idea was deduced from the view point of that the kerf shape is influenced by the radial heat distribution pattern that was established in our another report. First, a new and easy evaluation technique is proposed to be estimated the radial heat distribution pattern with information of cutting speed and top kerf width. Then, the adaptability is examined with some experiments. Secondary, actual radial distribution patterns are experimentally observed at various nozzle conditions. As a result, it is concluded that high throat length, small nozzle convergent angle, long distance between electrode tip to nozzle inlet and low torch height conditions are useful to realize more square type of radial heat distribution pattern with plasma arc. Finally, the relationships between torch conditions and bevel angles are examined. Then, it is indicated that more square radial heat distribution condition is effective to obtain lower bevel angle cuts.
Laser induced plasma plume was characterized by a spectral analysis in terms of temperature and electron density in lap welding of mild steel sheet of 0.8 mm thickness by using CO2 laser in a power range from 1.4 to 2.5 kW. Spatially and temporally averaged temperature of the plasma plume was determined as a function of focal position, F, location of focal point with respect to the work surface in Ar shielding ; it was found that the temperature ranged from 7500 K to 8500 K, and increased with increasing the absolute value of F in penetration welding. Abel transformation analysis indicated that the local temperature and electron density were higher in the area near the laser beam axis than the surrounding area due to laser beam heating ; the maximum temperature and electron density were 9500 K and 1017/cm3, respectively. According to inverse Bremsstrahlung theory, this electron density corresponds to the absorption coefficient of approximately 0.2 cm-1. Although this value proved little effect on bead formation, it was high enough to heat the plasma plume depending on laser focusing conditions, indicating possibilities to monitor the laser welding by detecting plasma temperature or brightness for establishing intellectual laser welding.
This paper describes the effect of evaporation on laser welding, especially on the keyhole formation. In order to investigate the evaporation, a numerical model of evaporation at the molten pool surfase of laser welding is developed. The effect of evaporation on laser welding is numerically analysed by the model and the numerical result has been compared with the experiment. This paper has been summarized as follows. (1) It is made clear by the numerical results that the molten pool geometry in laser welding has been affected remarkably by the evaporation process. (2) Numerical results show that the evaporation on molten pool surface produces the recoil pressure and a keyhole is formed by the process. (3) A calculated critical condition of keyhole has been compared with experiments and a satisfactory result has been obtained.
The solid phase joining of ductile cast iron (FCD 600) and carbon steel (SS 400) was carried out using thermal spray coating of nickel based self-fluxing alloy under atmospheric conditions for basic data. The size of joining specimen is 50 mm in length and 10 mm in diameter. After the each edged surface of FCD600 and SS400 specimens were blasted by used steel grid material sized No. 100, those were sprayed with self-fluxing alloy (S.F.A.) at the each surface of both specimens coated about 100 μm in thickness. The effects of joining conditions on the structure and strength of joints were studied. The joints with excellent metallographic structures were obtaind at the joining temperature near the melting point of the spray material. The results are as follows : (1) The maximum tensile strength at the joint part was more than 500 MPa under conditions of 50 sec raising time and 40 sec holding time at 1223 K temperature. (2) The Vickers hardness distribution near the joining zone of FCD 600 was between Hv250 and Hv340. That was a little bit higher than base metal. (3) The microstructure near the joining zone of FCD 600 did not indicate transformation to white pig iron because of its lower cooling rate. The structure around graphite was ferrite and the base metal was pearlite structure. (4) By the EPMA line analysis across the joining zone of FCD 600, Cr, Ni and B were not observed because the inserted material S.F.A. was a very little diffused and the most of it was discharged out of the jointed part.
Previous report has clarified that in DC-TIG arc welding of SUS 304 stainless steel the penetration ratio (penetration depth/bead width=D/W) can be improved by using flux cored stainless steel wire for CO2 gas welding. In this report an effect of flux cored wire on penetration is further examined, and at the same time the mechanism of the penetration forming is investigated paying attention on changes of arc phenomena. The report shows that shrinked anode is formed directly under electrode by using flux cored wire and weld with large D/W is obtained without influence of the penetration caused by quite small quantites of elements in the base metal. Meanwhile, the idea that the primary factor of the large D/ W is due to electro-magnetic convection likely generated by electric current concentrating in shrinked anode is made clear in this welding method. For stabilized welding with large D/W penetration, a method to supply intermittently flux cored wire immediately before starting of step-movement welding is proposed.
The butt welding of 0.3 mm thick very thin Al sheets by the ordinary inverter controlled TIG welder at ultra high speed(12, 000 mm/min)have been debeloped by means of the improvements of the nozzle. The welding procedures are as follows; (1)Fine 2%Th-W electrodes of 1.0 or 1.6 mm diameter were employed. The polarity is DCEN. No filler metal rod was used. (2) The Al sheets were clamped between two 30 mm thick Cu plates except some part of the sheets, thus avoiding the heat flow to the preceeding parts which cause the deflection of the sheets. (3) As the welding speed becomes higher and higher, the arc is dragged backward more and more. At last the arc goes out of the atmosphere of the protective Ar gas. To prevent this phenomenon, the size of the cross section of the nozzle was expanded, as the shape of a ellipse of which the long axis lies on the welding line. (4) On the other hand, some sheets of wire cloths were inserted vertically to the gas flow as the "damper"in the gas flow hole of the nozzle, to chenge the type of gas flow from the turbulent to the laminar one. (5) However the empty gap without the wire cloth is set in the surrounding narrow area around the electrode. Thas the strong gas blow from this gap squeezes and focuses the arc stably on the welding spot. (6) By the nozzle with the combination of the shape of its cross section and the variety of the wire cloths, the butt welding of 0.3 mm thick Al sheets at 12, 000 mm/min speed became possible and that of 18, 000mm/min became hopeful.
Soldering of surface-modified AIN to copper was investigated using Sn-38 mass%Pb solder. AlN substrate was dually metallized by Ti and Cu using ion plating technique as follows ; the 10μm thick Ti film was first plated on the AIN substrate, heat-treated in vacuum on various conditions and then the 5-10μm thick Cu film was plated on the heat-treated Ti film again. The thermal stress and heat conduction analyses suggested that the soldered joints could possess the superior properties when the thickness of soldering layer was kept less than 0.4 mm. Defects such as micro voids and cracks were occurred in Ti film after heat treatment, and the fraction of defects increased with increasing the treating temperature and time. The tensile strength of Ti+Cu dual metallized AlN to copper joint was approx. 20 MPa, and slightly decreased as the treating time of Ti film was increased. The Ti+Cu dual metallized AIN to copper soldered joint indicated the superior reliability for thermal fatigue and heat conduction.
In a newly developed ultrasonic testing of thin plates with a focused probe, the flaw echo or back wall echo appears with an observable time delay behind surface echo under the condition of shorter water path than the focal length of probe. The sound field of the probe is an important factor for this method, which is determined by the details of probe such as test frequency, diameter of active element and radius of acoustic lens. In this study, the relation between the sound field and the details of focused probe was investigated experimentally and theoretically. These results were used for the discussion of the effect of sound field on the flaw echo or back wall echo of thin plate in the ultrasonic testing based upon the multiple reflections. The experiments and numerical calculations showed that a focused probe with steep increase of sound pressure at the focal point was useful for the ultrasonic testing of thin plates based on the multiple reflections. The steep increase of sound pressure was evaluated by the calculation of sound pressure of probe with Hirose's model modified for pulse wave. Application of the developed ultrasonic testing with a focused probe to the thin plates was limited by the critical thickness, which was minimum detectable thickness, and depended upon the relation between water path and sound pressure of probe. These results were confirmed by the ultrasonic testing of thin plates and spot welds.
The sensitivity of reheat cracking (SR cracking) of 720 N/mm2 class high strength steel was discussed from the viewpoint of the chemical composition. The content of each impurity or alloying elements was shifted in a certain range from the original chemical composition of this steel (2.8%Ni-1.5%Cr-0.5%Mo, 0.10%C, 0.005%P, 0.001%S, 0.006%N), and the change of the cracking sensitivity was examined. The following results were obtained by the implant type cracking test. (1) This steel is originally less sensitive to reheat cracking ; the low contents of phosphorus and sulfur contribute to minimize the cracking sensitivity. (2) The harmful effects of phosphorus and sulfur appear when their contents exceed certain critical values. The critical phosphorus content (Pcrit) is 0.011%. The critical sulfur content (Scrit) is about 0.006%. Reducing the nitrogen content increases the Pcrit and Scrit. (3) The critical carbon content (Ccrit) is 0.15%. (4) Small quantities of vanadium and niobium increase the cracking sensitivity. The critical con-tents, Vcrit and Nbcrit are 0.02% and 0.03%, respectively. (5) Those results agree well with the results of fundamental researches on Cr-Mo type heat-resisting steels. The Y-type joint cracking test was also carried out on all the specimens mentioned above, but no cracking was detected.
In this paper, we investigated the effect of welding speed and external restraint on the grain growth behavior in the heat-affected zone of austenitic stainless steel, and related it to the growth behavior of the crystals in the weld metal which epitaxially grows from the bond. The following results were obtained in this study. The restraint during welding makes the heat-affected zone hardened due to the plastic thermal strain originated during heating, and results in the considerable grain-coarsening in the heat-affected zone after being heated over 1400 K-1500 K. Even thougth the grains in the heat-affected zone are coarsened by the restraint, there is almost no difference between the columnar width of the weld metal and that in the case of non-restraint welding. The bond region where there is no epitaxial relation between the weld metal and the heat-affected zone often appears when the grain-coarsening occurs in the heat-affected zone subject to large plastic strain during welding.
The purpose of this paper is to clarify the weld joint performance, especially the impact characteristics of the weld metal produced by electroslag welding (ESW) and submerged arc welding (SAW) by directing our attention to the high heat input welding processes which are usually applied for the manufacture of four side thick plate box column. Two types of SM490A 40 mm thickness have been used in this test. The test results are summarized in the following. (1) As for the impact characteristics of high heat input weld metal, the absorption energy values (vE value) of weld metal made by ESW are generally lower than those by SAW. (2) As for the vE values of various locations in high heat input weld metal by ESW, the vE value in the center of weld metal (Core) is considerably lower than those in peripheral locations of weld metal (Rim). (3) Macrostructure of ESW weld metal suggested that fine crystal grains were generated in the core with coarse crystal grains in the rim. (4) In the weld metal by ESW the impact value of fine grain core area is low and those in the peripheral coarse grain rim area are high on the contrary. This result disagrees with the conventional theory about the general relation between impact value and grain size. (5) It is found that the vE value of weld metal by ESW shows a tendency to decrease according to the increase of welding heat input (Q) and becomes low at the Q of 30.0 kJ/mm (up to 80.0 kJ/mm). Furthermore, the decrease of vE value related to the increase of welding heat input has not be caused by the change of chemistry such as C, Si, Mn, P and S in the weld metal.
The purpose of this paper is to investigate the impact characteristics of the weld metal produced using various welding parameters by directing our attention to the electroslag welding (ESW). This is one of high heat input welding processes which are usually applied for the manufacture of four side thick plate box column for general multistoried buildings, ets. In particular, this paper clarified the impact characteristics of ESW weld metal in its both center [Core (C) ] and peripheral locations [Rim (R) ]. The test results obtained are summarized in the following. (1) The impact test value (vE) of weld metal in its core (C) is lower than those in the rim (R) locations. (2) As for the change of vE value according to the change in welding heat input (at six levels of Q=10.1 to 126.7 kJ/mm), the vE value of both C and R locations generally decreases according to the increase of heat input. However, vE. value in the core (C) is always lower than those in R locations. (3) Regardless of the changes in steel types (such as TMCP and SM490A), flux types (three molten types) and amount of flux supplies (0.5 N or 0.7 N), the vE value of C location is always low without any significant change. Futhermore, the vE value decreases according to the change in shielding gas types from Argon gas (Ar) to air and then to oxygen (O2) in this order. (4) The formation of grain boundary ferrite in the coarse prior austenite (γ) boundaries and that of fine acicular ferrite inside the grains are mainly noticed in R locations as the result of the observation of microstructure. However, the formation of grain boundary ferrite in the fine prior γ boundaries and that of coarse ferrite inside the grains are densely noticed in C locations. (5) As the result of the observation of impact fractures with a scanning electron microscope (SEM), it is found that the facet size of weld metal in C locations is actually large which matches the low vE value. Moreover, selective fractures propagation along the grain boundary ferrite in the core of the impact fracture is found as the result of the observation on both two faces of structure at the same time. This indicates that the low vE value in C locations has been caused by the existence of both dense grain boundary ferrite and coarse ferrite inside the grains.
Investigations for the improvement of impact characteristics of Electroslag (ESW) weld metal was conducted using different types of welding wire. The test was carried out by using six types of welding wire and three levels of welding heat input such as Q=45.6, 79.0 and 126.7 kJ/mm. The test results obtained are summarised shown as the following. (1) In the case of all three heat input levels, the vE (absorbed energy) value in core (C) location of ESW weld metal was generally lower than those in rim (R) locations. Furthermore, regardless of the types of welding wires, the same tendency was noticed. (2) vE value was improved at C part of the weld metal produced by the wire containing Ni (wire F containig 4.18% Ni) and standard welding heat input of Q=45.6 kJ/mm. (3) In the previous paper it was comfirmed that the difference of vE values between C and R parts was significantly caused by the difference of ferrite structure types between them. In the C location the formation of extremely dense and massive polygonal ferrite (PF) was certified, and in the R location the formation of fine acicular ferrite (AF) was noticed inside the large austenite (γ) grains. However, cooling rate of weld metal of C and R locations for -τ1073(800)to773k(500c) in Q=45.6 kJ/mm was not so much to cause the change of structure. (4) The chemical analytical results of C, Mo and Ni in welding wire, base metal and weld metals showed that the dilution from the base metal into weld metal was comparatively stable (about 50-60%). (5) Improvement of the vE values at C part were expected thanks to formation of the coarse former austenite (γ) grains by the addition of Ni element.
62 level block fatigue loadings approximating to the Rayleigh distribution of peak were applied on center cracked transverse butt welded joint specimens in which tensile residual stress was always induced at the crack tips. The test environment was periodically changed between synthetic sea water and ambient air. The fatigue crack propagation rate, da/dn, obtained in this environment became low compared with da/dn in synthetic sea water by increasing the period in the ambient air. The decrease of da/dn occurred from the crack closure due to oxide debris in spite of the high stress ratio condition around the crack tips.
A study was made on the relation between the microstructure and the mechanical properties of the HAZ which was produced by a thermal simulation test for a high strength steel. It was found that the HAZ toughness deterioration of the high strength steel is closely related to the amount and existing form of martensitic islands in the matrix. Blocky martensitic islands surrounded by ferrite grains did not deteriorate toughness, but lath shape martensitic islands formed in a large matrix packet of the same crystal orientation were quite detrimental to toughness depending on the amount of martensite. The latter structure was formed in the unaltered coarse-grained zone (UACGHAZ), in the intercritically reheated coarse-grained zone (ICCGHAZ), and in the sub-critically reheated coarse-grained zone (SCCGHAZ). But among these structures, the structure in the ICCGHAZ had relatively short lath martensitic islands with softer lath matrix. This structure was especially apt to create micro-cracks. The dependence of strength and toughness on the martensite morphology was considered to result from the degree of restriction against the plastic deformation of matrix due to the microscopic heterogeneity in strength.
To visualize the fracture initiation characteristics in the multipass weld heat affected zone (HAZ), the yield strength distribution map and the CTOD toughness distribution map were drawn up graphically by a computer, based on a series of HAZ simulation test results. The comparison between these computationally estimated property distribution maps and the real HAZ CTOD test result revealed that: The fracture of the HAZ CTOD specimens was initiated from a peripheral part of a local hard zone (LHZ), which does not always correspond to the most brittle point on the crack front line, and The critical CTOD value derived by the full-thickness specimen was adequately represented by the local critical CTOD level at its fracture initiation point. The result of this study suggested that the fracture initiation is prevailed not only by the local brittleness but also by the heterogeneity in strength in the HAZ.
In this report, the problems are investigated for improving the performance of substrates, such as a large current capacity, a high dielectric breakdown voltage, a low thermal conductivity and a high reliability. The large current capacity and the high dielectric breakdown voltage are determined respectively by the size of conductor and insulator. It was clarified by theoretical consideration that increase of conductor thickness lower the reliability and increase of insulator thickness lower thermal conductivity. The design acceptable region of performance of the substrates has been generally bounded by a dielectric breakdown voltage, a current capacity, a thermal conductivity and a thermal stress. The design acceptable region become more narrow as the performance of the substrates improved. As the reslult, a layered structure of copper and molybdenum has been suggested as the conductor and it was shown by computer simulation that molybdenum had good properties to reduce the stress of ceramics, for molybdenum properties of a low thermal expansion and a high young's modulus are suitable to restrict a expansion of copper. Furthermore, the layered conductor is effective to improve current capacity and thermal conductivity of the substrates system.
New assembly process of the substrates system with high performance has been establised by adopting the same bonding condition of both Cu sheet-Mo sheet by vaccum diffusion bonding process and Cu sheet-alumina by the adjusting Ti content in active brazing metal. The substrates system has been consisted by the conductors with three layered structure of Cu and Mo sheets, and 96% alumina as insulator, as already reported. The bond with high reliability of Cu and Mo sheets has made by vacuum diffusion bonding process under the condition of 1, 300K for 2, 400s under the pressure of 10-2 Pa. Furthermore, in order to assemble the substrates system with high performance by one processing stage, the bonding condition of Cu sheet and alumina requires to coincide with that of Cu and Mo sheets. It was shown experimen-tally that the bond of Cu sheet-alumina with high reliabilty was obtained by selecting nearly 30 wt% Ti powder mixed with Cu powder as active brazing metal. As the results, the assemly process of the both the conductor and the insulator-Mo sheet, Cu sheet and alumina-has been able to be achieved with bonding of one processing stage by adopting the bonding condition of 1, 300 K for 2, 400 s. In addition, it was also confirmed that voids or other defects are not observed in the bond interface by the inspection of SEM etc.