Gold-aluminum wire bonds in LSIs is drastically degraded by corrosion of Au-Al intermetallic compound. This corrosion is due to bromine gas that is emitted when an encapsulant resin is exposed to a high temperature. In this study, temperature dependence on this bond degradation was investigated. Degradation starting times of Au-Al compounds as a function of annealing temperatures ranging from 453 to 553 K were studied. Increase of annealing temperature minimizes the degradation time. Activation energy of this degradation behavior was calculated, 110 kJ/mol, which was similar to that of Au-Al intermetallic layer growth. Intermetallic compounds formed within gold-aluminum bond by high temperature annealing were corroded by bromine gas at room temperature. The corroded compound was found to be Au4Al.
Tungsten has been used as one of the target materials of X-ray tube. So far, the tungsten is brazed copper for the cooling structure. But the defects such as void often occur at the brazed area. In this study, diffusion bonding process was tried to make the joint of tungsten to copper. Especially the effect of oxygen content in the copper was investigated using Auger spectroscopy, X-ray diffraction and EPMA. Copper materials used were tough pitch copper (O content : 0.0266mass%) and oxygen free high conductivity copper (O content : 0.0001mass%). The tensile strength of the joints of tungsten to copper depends on the oxygen content in the copper. The joints of oxygen free high conductivity copper to tungsten fractured at the base metal. On the other hands, the oxide layer of W18O49 was formed at the interface of joints of tough pitch copper to tungsten. The tensile strength of the joints decreased extremely due to the formation of the oxide.
In the present work, a three-dimensional numerical model for circumferential GTA (Gas Tungsten Arc) welding of pipes with single-Vee groove is developed. It consists of a heat conduction model and a molten pool balance model. In the heat conduction model, the transient temperature distribution on pipe is numerically analyzed to estimate the molten pool size by using a finite difference method. In the molten pool model, using the molten pool size derived from the heat conduction model, the theoretical configuration of molten pool is calculated in consideration of the balance of gravity, surface tension and arc pressure. By using the mathematical model, the influence of various welding parameters, such as welding positions and joint geometries, and the influence of arc pressure on the welding process have been discussed. As a result, it is made clear that the numerical model, proposed in the present work, can be available for the estimation of welding process parameters in groove welding of pipes.
This paper discusses the detection process on the defect in the TAB bonds by infrared thermal imaging with laser irradiation. We have previously discussed about the detection process for QFP gullwing type lead. But TAB bonds have three different factors which differentiate the detection process from that of the QFP gullwing type lead, i.e., (1) large valid heat capacity (Si-chip) under the bond interface, (2) lead thickness and (3) two different structures (face-up type, face-down type). This paper discusses about the influence of these three factors using thermal analysis. (1) The influence of the TAB bonds large heat capacity on the temperature distribution is that the thermal flow through the bond interface is increased. (2) The influence of the lead thickness is that thermal flow reflection from the defective part on the bond interface is attenuated by the lead thickness value. (3) We can detect the defect of face-up type structure by using the same detection process as the QFP gullwing type and the face-down type structure by changing the position of the laser irradiation and observation from the lead surface above the bond part to an observable lead surface. Then a detection process of internal micro defect in TAB bonds is theoretically established by infrared thermal imaging with laser irradiation.
Microstructural analyses of the bonding interface in R-SUS304ULC/Ta/Zr explosive bonded joints were conducted in order to clarify the embrittlement behavior of the joint. SEM and TEM observations revealed that the wavy interface was formed at the R-SUS304ULC/Ta and Ta/Zr interfaces, and that several products existed in the vortex zone (fused zone). The intermetallic compounds such as TaFe2, TaNi2 and TaCr2 were identified at the R-SUS304ULC/Ta interface, however, ω-Zr as well as Ta was precipitated in α-Zr matrix at the Ta/Zr interface. The hardened area was located in the vortex zone at R-SUS304ULC/Ta and Ta/Zr interfaces, and the severe embrittlement was occurred in the vortex zone at the Ta/Zr interface where the ω-Zr and Ta were precipitated.
The weld metal of stainless steel SUS310S was investigated for the purpose of making clear the direction of grain boundary migration. Using a vertex dynamics model, we calculated the transient process of the grain growth in the base metal and grain boundary migration in the weld metal. The results was compared with the observation of the grain boundary migration in the weld metal. The results are obtained as follows. (1) A frictional coefficient for grain boundary migration is assumed to be proportional to the third power of the grain boundary length. The grain growth rate calculated using this coefficient agrees with the quarter power of growth time shown in our experiment. (2) The distribution of numbers of edges surrounding a grain is apt to be universal and independent of the initial conditions and the vertex dynamics model. (3) The moving direction of the grain boundary triple junctions calculated in the weld metal has good agreement with the experiment results.
The grain boundary migration in the surface of the weld of an austenitic stainless steel was observed. In the cross section shape of the grain boundary groove, the thermal grooving theory in the surface based on the surface flow was compared to the experimental results. Furthermore, the relationship between the angle which is the plane of the solidification grain boundary to the surface plane and the amount of the grain boundary migration were discussed. The obtained results are as follows. The relationship between the cross section shape of the grain boundary groove which migrated in the surface of the weld metal and the direction of the grain boundary migration is good agreement with those of thermal grooving theory. That is, the plane of the grain boundary migrates to approach to be a right angle to the surface plane. The amount of the grain boundary migration increases with increase of the angle of the solidification grain boundary to a right angle of the surface. This reason is why the plane of the grain boundary which has higher angle to a right angle of the surface is able to migrate with less friction force than that which has lower.
Ni-base superalloy Inconel 706 which was developed from Inconel 718 has an economical advantage to save the major element Ni content less than Inconel 718. This alloy has such high strength, ductility and resistance against high temperature oxidation as equivalent to those of Inconel 718, but have a slight higher susceptibility to hot cracking. However the hot cracking susceptibility of Inconel 706 has not been investigated in detail. The Trans-Varestraint test was conducted in order to examine the solidification cracking susceptibility of Inconel 706 as a function of Ni content in base metal. Three kinds of Ni-base alloys with three different levels of Ni content were used. Typical dendrite structure was observed in the fractured surface of the specimen of Trans-Varestraint tested. Quantitative analysis of crack revealed that the solidification crack length and the temperature range in which hot cracking occurred (brittle temperature range, BTR) decreased with the increase of Ni content. Calculation of liquidus and solidus temperature of alloys by thermo-calc data indicated that the solidification temperature range also decreases with increase of Ni content. On the basis of these results, it was deduced that the improvement of the solidification cracking susceptibility with Ni content was attributed to the decrease of the solidification temperature range of alloys.
Low cycle fatigue tests were performed at room temperature and at 673 K for 1%Cr-Mo steel coated with self fusing alloy by gas spraying. The fatigue properties of self fusing alloy-sprayed steel and the mechanism of fatigue-fracture were discussed comparing to those of ceramic-sprayed steel. Following observations and conclusions were obtained. (1) The fatigue life of the self fusing-sprayed steel at 673 K was shorter than that at RT, while the fatigue lives at RT and 673 K almost coincided with those of base metal, respectively. (2) The crack initiation in the self fusing layer did not occurred in the early stage of fatigue lives, which was observed in the ceramic coated layer. (3) Fusing heat treatment did not improve the fatigue life of the self fusing alloy-sprayed steel remarkably. (4) The fatigue-fracture process of the ceramic-sprayed steel is considered that in the early stage of fatigue life, a fatigue crack initiates at the surface of the ceramic layer and propagates quickly to the substrate metal. On the other hand, the above phenomena is not observed in the self fusing alloy-sprayed steel, whose behavior is more like that of the base metal.
MAG welding was applied to the butt welded joint for a 500 MPa class TMCP steel. Fatigue tests were performed under a constant-amplitude load with the stress ratio of 0.1 in air and in synthetic seawater. In particular, corrosion fatigue tests were carried out under a natural corrosion condition with the cycling rate of 0.17 Hz at 25°C. The fatigue crack initiation life, Nc, was evaluated from a crack length measured by a DC electric potential difference method. In this paper, the S/N relations, macroscopic stress concentration factors and crack initiation behaviors at a toe of weld are discussed. The results obtained are as follows: (1) It was found that there exists a linear relation between ΔSn and Nc when tested in the seawater, where the ΔSn is a nominal stress range. The Nc-value tested under a higher applied stress range in the air agrees well with that observed in the seawater. It may be, therefore, quite reasonable to apply the ΔSn/Nc relation obtained in the seawater as the most safety crack initiation life in the air. (2) There is an applied stress range in which no fatigue failure arises even after 1×107 cycles in spite of the existence of a crack at the toe of weld when tested in the air. On the other hand, the crack at the toe of weld grows in the seawater due to corrosion dissolution of the crack tip. (3) The macro-stress concentration factors at the toe of weld, Kt2, obtained from strain gages were smaller than the localized stress concentration factor, Kt, , estimated from two-dimensional cross section at the crack initiation point. The scattering of the Kt2 was also small. The ΔSt in the ΔSt/Nc relation evaluated at the toe of weld, where bending stress components caused by the angular deformation as well as tensile stress components of the axial force exist, can be estimated as ΔSt=Kt2*ΔSn. (4) The ratio of crack initiation life to failure life, Nc/Nf, was about 0.84 in the air, and about 0.91 in the seawater. This difference may be caused by an environmental effect on crack growth rates. (5) The fatigue crack always initiates at the toes of weld, and propagates in a direction perpendicular to the loading direction.
The influence of the strengthening mechanism on the non-propagating behavior of fatigue cracks was discussed for low carbon steels which were strengthened by silicon or copper solid solution and copper precipitation. The notched round bar specimens with different stress concentration factors were fatigue-tested under uni-axial loading at stress ratio R=-1 (tension-compression). Fractographic observation of non-propagating cracks at stress levels below the fatigue limit were performed, and the correlation between the stage of non-propagating crack and the strengthening mechanism was investigated. Non-propagating cracks could be classified as Stage I or Stage II cracks, depending on the strengthening mechanism and notch root radius. In the case of smooth specimens, the fatigue limit was governed by the resistance of the microstructure to Stage I crack initiation or Stage II crack initiation. Therefore both strengthening mechanisms which prevent the dislocation movement in the ferrite phase are effective in improving fatigue strength. On the other hand, the fatigue limit for notched round bar specimens depends on the balance of microstructural resistance against to Stage I crack initiation, Stage II crack initiation and Stage II crack propagation. Therefore the effective strengthening mechanism varied with the stress concentra-tion factor.
Two kinds of thermal cycle tests with short (10 min) and long (23h) holding times were carried out to investigate thermal cycling damage characteristics and thermal barrier property of thermal barrier coatings (TBCs). Surface cracks vertical to the coating surface were nucleated due to thermal stress during both the thermal cycle tests. Holding time had no significant influence on the surface cracking behavior. On the other hand, holding time had a significant influence on the thermal barrier property, which was degraded with increasing total holding time. The degradation of thermal barrier property was caused by the decrease in porosity and hence by the increase in density of ceramic layer.
The beam-to-column connections of architectural steel structures have conventionally been welded by a semiautomatic multi-layer welding method using a CO2 arc welding process. So, there are the limit of its efficiency and the problem that the welding quality often depends on the welder's workmanship. By using the electrogas arc welding process which was generally considered difficult to apply to the beam-to-columr connections of architectural steel structures because of the shortage problem in weld toughness due to its conventional high heat input. With the architectural steel plate for large heat input welding newly developed, the reporters, et al. have developed an one pass upward electrogas arc welding process which can produce welds in their deposition rate of two times of the conventional without any program in their toughness and with extremely less weld defects. Moreover, this electrogas arc welding process can be applied to new seismic details such as a non scallop details, etc.. It is the most conspicuous characteristic for this welding process that stable penetration can be secured even for thick steel plates by moving the welding torch as trapezoid within the groove formed by water-cooled copper shoe. The heat input in this welding process is approximately 7 times as high as that in the conventional process but the charpy impact values in both weld metal and heat-affected zones are not less than 150 J (at 0 degree) in case of the architectural steel plate for large heat input welding newly developed. So, this new welding process with new developed steel can keep the same degree of toughness in weld zones as those in the conventional steel welded by the conventional CO2 arc welding process.