In the present work, bead-on-plate welds of mild steel and aluminum alloy are made with an automatic TIG arc welding machine, the traveller and filler rod feeder of which have been made to be driven by individual stepping motors manually and automatically controlled, and welding thermal cycles, weld bead shape and penetration are examined. Detailed discussions cover the heat flow in filler rod as well as in parent plate and the cross-sectional configuration of weld bead in connection with the process variables such as welding heat and ratio of feed speed to travel speed, i.e., length of rod deposited per unit weld length. The main results are summarized as follows:
The traveller and feeder, individually, are smoothly driven at accurate speeds with acceptable step response. By increasing the welding heat, both the penetration depth and the cross-sectional area of molten base metal are increased linearly, and the bead width is increased by increments reduced with welding heat. For a given welding heat, increasing the ratio of feed speed to travel speed decreases the cross-sectional area of molten base metal as well as the high temperature region in parent plate during welding, and increases the ratio of bead width to penetration depth. The temperature distribution along the filler rod during welding can be approximately derived from the quasi-steady-state one dimensional heat conduction theory, and a fraction of the welding heat which is transferred into the filler rod is 2 to 5 percent and increases with an increase in feed speed of the filler rod.

抄録全体を表示

New equations giving constriction resistance R_C as a function of area S_M and number density n of true contact spots have been obtained from a model experiment, and applied to the analysis of the electric resistance of the bond interface ΔR for the diffusion-welded joint of titanium. The new equations obtained are described as
R_C= ρ/√n ⋅S ( √π/2 √S/√S_M -1.21) ……(1)
for (S_M/S)<1/4, and
R_C= ρ/√πn⋅S {√ S/S_M tan^-1(√ S/S_M -1)-0.842(1-√ S_M/S )} ……(2)
for (S_M/S)> 1/4, where ρ is the resistivity of base metal and S the apparent contact area. It can be seen from these equations that parameter A=(ΔR⋅S/ρ) depends only on (S_M/S) and n, if ΔR=R_C. Parameter A measured at temperatures from 77 K to room temperature for joints of titanium, however, decreased with a rise in the temperature of measurement, and its temperature dependence became less pronounced with the increase in welding temperature and time. On the other hand, area S_M and density n were estimated from fractured surfaces of joints on the assumption that spots where grooves caused by machining of the faying surface were annihilated corresponded to true contact spots. Parameter A estimated from S_M and n thus obtained using eqs. (1) and (2) was significantly smaller than that measured at 77 K and rather in good agreement with that measured at room temperature. This result indicates that not all the true contact spots observed on fractured surfaces corresponded to completely bonded spots having electrical properties identical with those of the base metal. The dependence of A on the temperature of measurement can be accounted for by a model that the bond interface consists of three characteristic spots: unbonded spot, completely bonded spot and incompletely bonded spot containing inclusions such as oxide films.

抄録全体を表示

Liquidus surface of copper phosphorus brazing filler metals with silver and tin addition was investigated. Spreadability and erosion of Cu-Ag-Sn-P quaternary filler metals on copper base metal were also studied. Primary phase could be predicted by using the phosphorus equivalent. The filler metals with the phosphorus equivalent more than 8.38 crystallizes Cu₃P primary phase, whereas the filler metals with the value less than 8.38 crystallizes copper solid solution. The primary phase is also predicted by the electron concentration of filler metal, and also the ratio of electron concentration of tin plus phosphorus content to that of copper plus silver content. The erosion depth of copper base metal increased almost linearly with tin, silver and phosphorus content.

抄録全体を表示

The microstructure and mechanical properties of commercial pure titanium (CPTi) and Ti-6Al-4V alloy joints brazed with newly developed Ti-base amorphous filler metals have been investigated by microscopy, electron probe microanalysis, tensile test, fatigue test and salt immersion test.
The developed amorphous filler metals were three kinds of Ti-37.5Zr-15Cu-10Ni, Ti-35Zr-15Cu-15Ni and Ti-25Zr-50Cu. Their melting points were approximately 100°C lower than those of the conventional Ti-base brazing filler metals. With these filler metals, CPTi and Ti-6Al-4V alloy could be brazed at temperatures below α-β transformation (880°C) and β transus temperature (995°C), respectively.
In brazing at temperature below α-β transformation or β transus temperature of base metal used, the original fine structure of the base metal was nearly preserved and a brazed zone was recognized to exist obviously. However, at temperature above these temperatures, the grains of the base metal were coarsened and in the brazed zone the fine acicular phases were observed.
The tensile properties of the joints brazed at temperatures below these transformation temperatures, was compared with that of the base metal and then in brazing for a suitable brazing time, the joint was fractured in the base metal.
From fatigue test results of the Ti-6Al-4V alloy brazed joint, in the joint brazed at temperature below β transus, the fatigue life indicated a fairly good value but all joints failed in the brazed zone. When brazed at 1000°C, joints showed low fatigue life considerably.
The corrosion behavior of the brazed joints was quite equivalent to that of the base metal.

抄録全体を表示

It is required, as an indispensable condition, to secure the specified hermeticity as well as to restrain temperature rise at the minimum and prevent deterioration of an element inside the package from thermal effect during hermetic sealing process of a ceramic package for which high reliability is demanded. For this reason, Thermal Video System, which is able to detect JR thermal image in real-time, was used in this experiment to observe progressive change in surface temperature distribution of a ceramic package during Micro Parallel Seam joining process. Also, the relationships between various seam joining conditions and surface temperatures obtained were studied, and some conditions were established to restrain package temperature rise while maintaining her meticity of 10^-8 atm cm³/s which is defined as the criteria for acceptance/reject of seam joining quality.
The results obtained are as follows.
(1) Package surface temperature distributions were first shown so that two high temperature areas (hot spots) appeared in the vicinity of left and right roller electrodes immediately upon start of current flow, and temperature continued to rise toward the center of lid thereafter as isothermal ring expanded from those hot spots which were to be the center. Such status continued for approximately 70% of total joining time and finally the center of lid reached the highest temperature.
(2) In order to restrain package temperature rise;
(a) Roller electrode with a greater taper angle should be used.
(b) As for seam joining conditions, electrode force and current should be selected as small as possible and joining speed faster within the range that specified herrneticity can be maintained.
(c) Lid with plating tends to cause temperature rise, especially when plating material, of which melting point is lower than that of lid material itself, used.
(3) Also, voltage between roller electrodes, current, resistance and contact width between the lid and roller electrode were discussed in relation to temperature rise of ceramic package.

抄録全体を表示

Characteristics of underwater gas cutting in the flat position, compared with gas cutting in air, were investigated from the viewpoint of cutting ability. This ability was estimated from the maximum cut thickness (Tm) of wedge-shaped specimen, and the relationship between Tm and velocity of cutting oxygen jet was inquired.
The results obtained are summarized as follows;
(1) When the proper cutting conditions employed in air were applied to underwater gas cutting, Tm was remarkably smaller than that in air, especially by using a large diameter-tip. And also, unstable phenomena were observed on cutting process and cut surface.
(2) By the increases in cutting oxygen pressure and preheating gas consumption, Tm was increased in the similar tendency as that in air. However, Tm was still smaller than that in air.
(3) It was known that Tm was influenced by the decrease of velocity of cutting oxygen jet in relation to tip diameter, cutting oxygen pressure and kerf width.

抄録全体を表示

It was previously reported that the maximum cut thickness (Tm) obtained in underwater gas cutting in the flat position was remarkably smaller than that in air, especially by using a large diameter-tip. It was considered that this -was caused by the difference of the state of oxygen jets in the kerf in underwater or in·air cutting process. In this report, underwater gas cuttings by such three cutting positions as the vertical up, the vertical down and the horizontal position were carried out to make clear the phenomena in underwater gas cutting, from the view point of the maximum cut thickness.
The results obtained are summarized as follows;
(1) The value of Tm obtained in underwater gas cutting was affected with the cutting position. The Tin obtained in the vertical down position by using the large diameter-tip was smaller than those of the vertical up and horizontal position.
(2) It was cleared. that the state of oxygen jets was influenced by the direction, the width, and the shape of the kerf, and this caused the difference of the Tm in various cutting positions.
(3) Unstable phenomena were observed in oxygen flow in the kerf model of acrylic resin in the vertical down position. In actual cutting, the unstable exclusion of slag was observed and the cutting surface was disturbed.

抄録全体を表示

In the digital particle analysis by image processing, errors generate in the measurement of the perimeter in case a figure whose size is relatively small in the frame of the whole image. These errors are caused by the size and the orientation of the figure, and are unavoidable in some cases with the usual method. The new method, called the "Three Pixels Vector Method" has been developed. The above errors can be reduced remarkably by this new method. The principle and the ability of.this method, together with the other usual methods, are described in this report.

抄録全体を表示

This investigation was made to establish a new testing method for cold cracking of weld metal of high strength steels such as HT60, HT80 and HY130. This test was named the LB-TRC (Longitudinal Bead-TRC) test.
The effect of loading temperature on the lower critical stress was studied and moreover the lower critical stress obtained by the LB-TRC test was compared with that obtained by the conventional TRC test. On the other hand, fractographic investigation of this test was undertaken.
Consequently, it was shown that the lower critical stress and the fracture mode in the LB-TRC tested specimen were approximately equal to those obtained in the conventional TRC test and that the LB-TRC test was useful to evaluate cold cracking susceptibility of weld metal.

抄録全体を表示

In this study, the effects of hardness and diffusible hydrogen content on the lower critical stresses o f weld metals (LCS) of high strength steels in the strength range 490 to 1950 MPa were evaluated utilizing the LB-TRC test which was newly developed in order to understand the cold cracking characteristics of weld metal.
Test beads were laid utilizing SMA welding and GTA welding in which argon and argon-(0.35, 1.0, 2.0, 3.2 Vol.%) hydrogen mixed gases were used as shielding gases. The diffusible hydrogen content range was 0.1 to 9 ml/100 g. Moreover, yield strength of root-pass deposited metal ((σ_r)_y) and fracture stress of LB-TRC tested specimen which had evolved diffusible hydrogen for 10 days at room temperature (σ^*_F) were measured in order to evaluate the cold cracking susceptibility. σ^*_F was regarded as the hydrogen-free LCS, so the cold cracking susceptibility of weld metal was evaluated using two parameters such as σ_cr/(σ_r)_y and (σ^*_F-σ_cr)/σ^*_F.
Consequently, the LCS of weld metals of high strength steels decreased rapidly with an increase in the diffusible hydrogen content in the range 0 to 3 ml/100 g. Moreover, cold cracking susceptibility in weld metal increased with an increase in strength level and/or hardness and then especially, weld metals of HY130 and HY150 were most susceptible to cold cracking rather than that of HY180.

抄録全体を表示

In this study, the effects of remaining hydrogen content on the lower critical stress of weld metal (LCS) in high strength steels were evaluated, because it was reasonably considered that remaining hydrogen content related with LCS more closely than initial hydrogen content at high temperature measured by JIS or IIW method.
The remaining hydrogen contents at 100°C on bond line ((H_R)₁₀₀) of HT60, HT80, HY130 and HY 180 weldments were measured in the LB-TRC test specimen using GTA welding with Ar-H₂ mixed shielding gases by means of gas chromatograph, of which the measurement accuracy was 0.01 ml. Those were measured under four kinds of thermal cycles and then the relations between (H_R)₁₀₀ and the thermal factor ((ΣDΔt)₁₀₀) which related with thermal cycle were deduced. Consequently, the experimental equations between (ΣDΔt)100 and (HR)₁₀₀/H₀ of all materials used were deduced as follows.
(H_R)₁₀₀/H₀=exp {-A(ΣDΔt)₁₀₀}
where A; 83 (HT60), 69 (HT80), 46 (HY130) and 41 (HY180)
H₀; initial hydrogen content
Moreover, the relations between the LCS obtained by the LB-TRC test and (H_R)₁₀₀ were evaluated, but the LCS were not related to (H_R)₁₀₀. Therefore, the remaining hydrogen contents of weld metal ((H_RW)₁₀₀) were calculated using the analytical solutions of one-dimensional diffusion equation. Consequently, the lower critical stress of weld metal strongly depended on (H_RW)₁₀₀ than (H_R)₁₀₀.

抄録全体を表示

This study was done to develop the possibility of sound welding of higher grade HY-type of steel having a low lower critical stress, utilizing transformation expansion of weldment including weld metal and the heat-affected zone.
The restraint stresses in root pass welded joint of various strength steels such as SM41, HT60, HT80, HY130 and HY150 were quantitatively evaluated utilizing the RRC test under the restraint intensity of 9.8 and 29.4 kN/mm.mm by GTA welding with filler wire. On the other hand, transformation expansion of welded joint was measured by dilatometric technique. Then the relationship between the restraint stress and the transformation expansion was quantitatively investigated.
Consequently, the final value of the restraint stress reduced with an increase in strength of base steel, namely in the order of SM41, HT60, HT80, HY130 and HY150 under the same restraint intensity. For example, the final restraint stress in SM41, HY130 and HY150 was about 490, 216 and 147 MPa, respectively, in the restraint intensity of 9.8 kN/mm·mm. Reducing the restraint stress was attributed to transformation expansion of welded joint. The increase of transformation expansion was much effective for reducing the final restraint stress in HY130 and HY150 weldments in comparison with the other weldments. The transformation temperatures of HY130 and HY150 weldments were lower than those of other weldments, so that not only large transformation expansion but also low transformation temperature were very useful to reduce the final restraint stress. In addition, these results suggested that the crackfree weldments of HY130 and HY150 steels may be completed by developing the higher alloy filler wire having high expansion coefficient and low transformation temperature during cooling after welding.

抄録全体を表示

An authors' previous paper which reported on the influence of delta-ferrite on low temperature toughness of Type 316L austenitic stainless steel weld metal, revealed that the toughness of as-welded specimen was dependent not only on the delta-ferrite content but also on the solidification mode.
In this paper, various factors being affected on the toughness such as concentration difference, coherency, etc., were investigated in detail by using a transmission electron microscopy (TEM), an analytical electron microscopy (STEM/EDX) and a scanning electron microscopy equipped with an energy dispersive X-ray spectrometer (SEM/EDX).
From an electron diffraction analysis, a crystallographic orientation of primary delta-ferrite was coherent with that of its adjacent austenite matrix but in the eutectic delta-ferrite, the relation was incoherent.
The amount of molybdenum contained in the eutectic delta-ferrite was 1.8 times as that in the primary delta-ferrite and moreover, a cell or a cellular dendrite boundary adjacent to the eutectic delta-ferrite, was enriched in molybdenum slightly. Probably, in these molybdenum-rich sites, a lattice strain was considered to be higher than that in the primary delta-ferrite.
From the SEM/EDX analysis on the fractured surface of the Charpy tested eutectic delta-ferrite specimen, the chromium and molybdenum rich zones were measured, where these compositions were equivalent to that of delta-ferrite or cellular dendrite. However, in the primary delta-ferrite specimen, such a zone could not be detected.
From results above, when the eutectic delta-ferrite specimen was Charpy-tested, a crack initiation occurred at the delta-ferrite/austenite boundary or within the delta-ferrite and subsequently, the crack propagated preferentially along the cell or the cellular dendrite boundary or the delta-ferrite.
In the specimen heat-treated at 923 K for 100 hrs, the concentration of the principal elements such as chromium, molybdenum and nickel within the prior delta-ferrite fluctuated widely, independent of its morphology, in which furthermore, the presence of carbide M₂₃C₆ and σ phase were confirmed.

抄録全体を表示

Type 316L weld metals contained various delta ferrites ranging from 0% to 9.5%, which were deposited by one pass submerged arc welding, were heat treated at 923 K to 1523 K for 2 s to 86.4 Ks (24 h). From the results of absorbed energy at 77 K and analysis of precipitates by electron and X-ray diffraction, the influence of reheating on low temperature toughness was investigated.
In the weld metal of eutectic delta ferrite (delta=2.7%) which had minimum absorbed energy, the heat treatment for 2 s at the temperature above 1123 K enhanced its absorbed energy considerably as compared to as-welded specimen and in particular the tendency was remarkable at the temperatures of 1423 K and 1523 K.
The absorbed energies of weld metals which contained some delta ferrites decreased with the increase in the heat treating time and this decreasing rate was pronounced at more delta ferrite content and high heat treating temperature. However, in the case for short heat treating time within 120 s, the absorbed energy had a tendency to increase, compared with as-welded one, while in the fully austenitic weld metal, the decrease in absorbed energy could be scarcely detected independent of the heat treating time.
Retained delta ferrite contents after the heat treatment decreased faster at eutectic delta ferrite than at primary delta ferrite. This was considered to be due to the precipitation of M₂₃C₆ and sigma phase facilitated in the case of eutectic delta ferrite because of poor coherency of delta/gamma interface.

抄録全体を表示

Generally, fatigue crack propagation rate is well put in order by stress intensity range ΔK_eff by considering crack closure period in crack opening process. But, in weld residual stress field, it seems that crack propagation behavior can not be exactly expressed in ΔK_eff. This would be caused that ΔK_eff was calculated by detecting crack full opening point which was able to be well measured experimentally. On the other hand, partially crack opening phenomena, which was presumed theoretically by considering the residual stress distribution in the case of crack laying over the tension and compression stress transition region, has not reflected in ΔK calculation process by this time.
In this paper, as mentioned above consideration, calculation method of stress intensity factor in partially crack opening condition was proposed and its model was based on superpose principle of elastic stress. As a result of applying our calculation of stress intensity range ΔK_act to evaluate the fatigue crack propagation rate in welded plate, it was made clear that the crack propagation behavior could be put in order well by ΔK_act than by ΔK_eff.
Namely, ΔK_act concept would not be only modification of KΔ_eff, because of applying new calculation technique for partially crack opening model.

抄録全体を表示

Initial residual stress due to welding was produced by thermo-elasto-plastic analysis based on F.E.M. calculation, thereafter cyclic-elasto-plastic analysis for fatigue crack propagation was carried out. A crack was propagated by 0.2 mm per one load cycle from compressive residual stress field to tensile stress field (stress ratio was 0.0). Crack opening and closing behavior was investigated in relation to re-dis-tribution of residual stress and change of plastic region.
Main results obtained are summarized as follows:
1. When the crack tip was in the stationary compressive residual stress field, as crack planes were kept in contact, the re-distribution of residual stress did not occur so much at unloading state. The crack opening was occurred at the crack root then moved to the crack tip, and this was also recognized in base metal analysis.
2. As the crack approached the initial tensile residual stress field, the residual stress re-distributed and the crack-opened-region remained near the crack tip even at unloading. The crack opening points were moved both to crack tip side and crack root side, and the crack tip opening occurred earlier than the full crack opening.
3. It was clarified that crack opening stress level would be able to be determined by strain gages glued just behind the crack tip, and using superposition model, the stress level could be estimated from the compressive residual stress distribution on the crack plane.

抄録全体を表示

In order to investigate the effect of welding residual stress field on the fatigue crack propagation direction, the plate specimen which had crossed beads and asymmetric residual stress field, and another specimen which had one bead parallel to the load axis and located in the center of the specimen width were prepared.
The material was high tensile strength steel and the welding was done by electron beam welding. Fatigue crack propagation tests were conducted under constant amplitude loading and stress ratio was nearly equal to zero.
As the results, it was clarified that the direction of maximum tensile principal stress on a certain point beyond the crack tip was rotated to the direction of load axis, the crack tip coming near and loading being applied. And also it was recognized that the fatigue crack even though on the crossed beads specimen was propagated straightly perpendicular to the load axis, and that the fatigue crack propagation rate was plotted on same scatter band obtained in base metal against to ΔK_eff.

抄録全体を表示