The microstructure and bond strength of the friction-welded joint of Al (commercially pure aluminum) to OFC (oxygen free copper) have been investigated in order to understand the formation of intermetallic compounds and its effects on the mechanical properties of the joint. A mixing layer of Al and Cu which showed a fine stratified microstructure was formed in Al adjacent to the weld interface by mechanical picking-up of Cu into Al. In the mixing layer, intermetallic compounds CuAl2, CuAl and Cu9Al4 were detected by X-ray diffraction analyses and TEM. As forge pressure P2 was increased, the width of the mixing layer was decreased and the stratified microstructure became considerably isolated flaky one. The tensile strength of the joint was increased as P2 was increased, and at P2 above 120 MPa the joints were fractured in the heat affected zone of Al. On the other hand, the joints welded at P2 below 80 MPa were fractured mainly in the mixing layer. EDX analyses of the area around cracks suggested that the crack on the tensile test initiated at CuAl in the mixing layer, since the composition around the initiation site of the crack corresponded approximately to that of CuAl. From these results, it can be concluded that the intermetallic compounds, especially CuAl have harmful effects on the mechanical properties of the joint. The increase in the joint strength with rise in P2 can be explained as a consequence of the decrease in the thickness of the mixing layer which includes a number of fine intermetallic compound layers. The formation of mixing layer having the fine stratified microstructure is probably due to the mechanical mixing in the solid state, since the average temperature of frictionn interface estimated from the thermoelectric power between Cu and Al was below the lowest eutectic temperature of the Cu-Al system.
The microstructure and bond strength of the friction-welded joint of Al (commercially pure aluminum) to Cu-W (copper-tungsten sintered) alloy of various W contents have been investigated with particular reference to effects of W particles on the formation of intermetallic compounds at the weld interface and on resulting mechanical properties of the joint. A lot of mixing layer of Cu, Al and W was formed by picking-up of Cu and W into Al in the joint of Al to Cu W alloy of W content less than 30 mass%, whereas no mixing layer was found for W contents more than 50 mass%. Since the mixing layer included a number of brittle intermetallic compound layers, tensile strength for joints of Al to Cu-W of W contents below 30 mass% was considerably lower than those of W contents above 50 mass%. The highest temperature of friction interface during the friction-welding, estimated from thermoelectric power between Cu and Al, was about 800 K both for the joint of Al to Cu-W including 70 mass% W and for the joint of Al to oxygen free copper (including no W particle). Microhardness of the Cu-W alloy with 50 mass% W was significantly higher than those of W contents below 30 mass% at temperatures ranging from room temperature to 800 K. This suggests that the flow stress of Cu-W alloys containing W above 50 mass% was significantly higher than that of W contents below 30 mass%. This difference in the flow stress of the Cu-W alloy prevented the formation of the mixing layer in the joint of Al to Cu-W alloy of W contents above 50 mass%, since it needed plastic flow and rupture of Cu-W alloy to pick up Cu and W from Cu-W alloy into Al.
The formation of intermetallic compounds and its effects on the bond strength have been investigated for the friction-welded joint of Ti (commercially pure titanium) to Cu-W alloy (copper-70% tungsten sintered alloy) by being compared with those of Ti to OFC (oxygen free copper). In the joint of Ti to OFC, a mixing layer of Cu and Ti which included brittle intermetallic compounds TiCu4, and Ti3Cu4, was formed. As forge pressure P2 was increased, the thickness of the mixing layer was decreased, since the mixing layer was excluded toward the collar during forge process. The tensile strength of the joint was increased with the rise in P2 ; joints welded at P2 above 250 MPa were fractured in the weld heat affected zone of OFC about 3 mm away from the weld interface, while joints welded at P2 below 200 MPa were fractured mainly in the mixing layer. A similar mixing layer was also observed in the Cu-W/Ti joint. In this case, however, the mixing layer was hardly excluded during the forge process, and consequently the joint was fractured in the mixing layer even at P2 above 250 MPa. A marker experiment using alumina powder was carried out in order to investigate the plastic flow caused by friction welding near the weld interface ; i.e., alumina powder was packed into a hole drilled in one of the specimens, and distribution of alumina powder resulting from the plastic flow during friction welding was observed. The result suggested that plastic flow in the radial direction of the weld interface for Cu-W/Ti joint was insufficient to exclude the mixing layer even at P2=340 MPa. The temperature of the weld interface measured with two-color radiation pyrometer was increased up to 1333 K. At this temperature, the Cu matrix of the Cu-W alloy adjacent to the weld interface may be regarded as melted, since spot analyses with EDX indicated that the Ti content of the Cu matrix within 4 μm from the weld interface was greater than the solidus composition of Cu-Ti system at this temperature. It seem that the melting of the Cu matrix facilitated the formation of the mixing layer and the intermetallic compounds.
High performance metal matrix composite coatings could be fabricated by using the highly mixed metal-ceramic composite powders. In this study, composite powders both of Cu-Al2O3 system and of Cu-AlN system were produced by mechanical alloying (MA) method. The composite coating were made by plasma spraying of these pre-composite powders. To compare with these composite coatings, conventional coatings were made by using the conventional Cu and Al2O3 powder. The quality of the coatings obtained were evaluated by abrasion were test and the electric properties were also investigated. The main results of this study are summarized as follows, (1) In the case of spraying of MA powders, the separation of the ceramics from the metal occured due to the relative difference of specific gravity between them. Nonetheless, the sprayed coating made with MA pre-composite powders had finer microstructure and higher were resistance than made with conventional powders. (2) The electric resistivity values of the coatings made with MA powders did not scatter largely compared with those of the conventional coatings because of the homogeneous structures. Therefore, it was found that highquality composite coatings with fine and homogeneous structure can be obtained using MA powders.
Metallization and bonding of AIN ceramics were made using active brazing metal (Ti-AgCu). Active brazing metal used in metallization and bonding was in paste form prepared by adding Ti powder and organic solvent in to eutectic silver brazing metal powder. Characterization and bonding mechanism of the bonded layer, and relation between the thichness of the reacted layer and bonding strength were studied. The results of this study are as follows : (1) The microstructure of the bonded layer between AIN ceramics and Ti-AgCu brazing metal is found to be TiN layer of about 0.2 to 0.8 μm in thickness at the AIN interface, a composite layer of TiC, AICu as a middle layer and Ag-Cu alloy layer at the surface side. (2) Thickness of TiN layer did not change by the amount of Ti addition in silver brazing metal but the thicness of TiCu and AlCu layers are increased with increase of Ti addition in silverbrazing metal. Activation energy of growth of TiN layer is found to be 20 kJ/mol. (3) When the amount of Ti addition is more than 15 mass%, the bonding strength is decresed and the bonded layer is broken. Because the thickness of brittle TiCu layer increases with the increase of the Ti cotent in silver brazing metal.
The penetration in GTA welding is affected by the fluid flow in molten pool, and the fluid flow is profoundly influenced by the minor elements in base metal such as sulfur, oxygen and chlorine. The purpose of this work is to make clear the mechanism of the minor element effect on the fluid flow in metal pool. As described in the previous paper, the anode size on arc weld pool is sensitive to the minor element in base metal. This means that the electro-magnetic force is one of the important factors for the minor element effect in arc weld pool. In the present work, a defocused electron beam has been used as a heat source to discuss the influence of the element on the surface tensional convection. The results of the present work are summarized as follows, (1) The fluid flow pattern in the pool by a defocused electron beam is dependent on the evaporation process in molten pool. (2) Under the condition of high evaporation rate, the metal flow on the pool surface, induced by the frictional force of evaporation jet, is outward from the center of the molten pool. (3) Under the condition of low evaporation rate, where the frictional force due to the jet is negligible, the fluid flow is inward along the pool surface toward the center. (4) The sulfur in base metal depresses the metal vaporization from molten pool. (5) The ratio of sulfur content in the pool decreases during evaporating, i.e., the sulfur element evaporates predominantly from the pool surface. (6) The above result suggests that a surface tensional gradient on molten pool surface due to the evaporation of surface active element (sulfur) is important as the mechanism of the inward flow on pool surface.
In the object modeling for a product that consists of many parts, it is required to describe not only a shape of each of the parts but also relations among them. In this paper, we propose a new product model for plate structures. Each of the parts is described as a predicate of prolog, which predicate consists of part's name and boundary-represented shape of the plate. Connectivity between two parts is also described by a predicate as a binary relation. Some fundamental concepts of local structure, that are important in design and manufacturing process, and their prolog implementation are shown.
This study was made on the oxygen and nitrogen contents and notch toughness of the multilayered weld metal produced by submerged arc welding. The results obtained are summarized as follows. (1) The oxygen content of multilayered weld metal increased as the number of layered weld metal increased. (2) The nitrogen content of multilayered weld metal is almost constant irrespective of the number of layered weld metal. (3) The silicon content of multilayered weld metal increased as the number of layered weld metal increased, and the oxygen content of multilayered weld metal shows the similar tendency. (4) The manganese content of multilayered weld matal increased by MnO included flux, and decreased by flux without MnO as the number of layered weld metal increased. (5) The carbon content of multilayered weld metal decreased as the number of layered weld metal increased. (6) The impact value of multilayered weld metal is roughly constant, but the value of last layered weld metal decreased slightly. (7) The oxygen and nitrogen contents of weld metal does not change remarkablly irrespective of heat treatment. (8) The oxygen content of weld metal increased as the welding current increased, but the nitrogen content in weld metal decreased as the welding current increased, and the impact value of weld metal generally decreased as the welding current increased.
The contact angle of molten Cu-Ti alloys containing Ti content up to 60 at% or more on SiC was investigated by a sessile drop technique at 1373 K in vacuum. The contact angle of molten Cu-Ti alloys reaches the equilibrium value at 3.6 ks at 1373 K. The equilibrium contact angle of Cu-Ti alloys on SiC reduces drastically with increasing Ti content. Cu -Ti alloys containing Ti content of 30 at% or more shows the equilibrium contact angle of 7 degree or below. The addition of Ti to copper depresses the reaction of Cu with SiC forming TiC and Ti3SiC2 carbides at SiC/Cu-Ti interface. This accounts for the superior wetting of the Cu-Ti alloys on SiC. The Cu-Ti alloys are applicable to the filler for joining SiC.
The effect of δ ferrite morphology and grain boundary property on the low temperature toughness of austenitic stainless steel SUS304L weld metal has been studied. The types of δ ferrite were classified into three categories : vermicular, lacy, and globular. When Charpy test were performed at low temperature, brittle fracture easily occurred in a vermicular type δ ferrite because the longitudinal direction of this δ perallel to δ, that is a cleavage plain. On the other hand, brittle fracture scarcely occurred in both a globular type and a lacy type δ ferrite, because the the stress concentration must be low at globular type δ and the Kurdjumov-Sachs (K-S) relationship holds between lacy type δ and γ phase. Grain boundary holding the K-S are coherence to grains and hardly cause the decohesion. The present data indicate that changing the δ ferrite morphology from a vermicular type to lacy type prevents the cleavage fracture and increases the fracture toughness at low temperature.
A type 304 austenitic stainless steel was heat-treated at temperatures of 800-1300 K, and the heat-treated specimens were observed by optical and transmission electron microscopy to examine the effect of crystallographic character of grain boundary on carbide precipitation and corrosion.At sensitizing temperatures, grain boundaries with carbide precipitation or corroded boundaries increased with holding time. Kikuchi line analyses showed that grain boundary carbide precipitation and corrosion sensitively depended on the grain boundary structure. A more ordered boundary needed longer time for carbide precipitation or corrosion than a less ordered boundary. This fact suggests that weld decay in austenitic stainless steel can be prevented by increasing the percentage of ordered boundaries in the material.
The valence Auger spectra have been studied in order to understand interatomic chemical bonding across the interface between metals and ceramics. SiO2 was adopted as the ceramic and Ti was used as the metal. Ti/Si interface was also analized to comprehend the characteristic features of Si-O, Ti-O and Ti-Si bond in spectra. Ti was deposited on SiO2 and Si substrates under high vacuum environment about 5×10-4 Pa. Ti/SiO2 was annealed at 473K for 1800s. Ti/Si was annealed at 823K for 3600 s. Some of the spectral feature observed on the Ti/Si interface were applicable to understand the chemical bonding at the Ti/SiO2 interface. The Ti-O bond was detectable at the interface by the intensity ratio of Ti-LM23M45/Ti-LM23M23. The Si-O bond was also detectable from the energy of Si -LVV peaks. It was found to be difficult to comfirm the existence of Ti-Si bonds at the Ti/SiO2 interface. A damage induced by Ar-and electron-bombardment was observed. It didn't disturb the detection of Ti-O and Si-O bonds.
The valence electronic structure of the Ti/SiO2 interface was discussed by analyzing their Auger spectra. The change in the intensity ratio of SiO2-L23V1D+Si-L3M23M23/SiO2-L23V1V1 with Ti contents were helpfull to confirm the interfacial Ti-Si bonds under the condition that the electronic dose were kept constant at each analyze-point through the interface. This concept was also applied to annealed interface of Cr/SiO2 and Cu/SiO2. M-O or Si-O bonds were detected at the interface. The possibility of the existence of M-Si bonds were suggested from the Auger intensity ratio. The transition elements which exhibit small number of 3d electron, such as III a-VI a metals, tend to lose the metallic property in electronic structure.
This study developed the new Transient Liquid Insert Metal Diffusion Bonding (TLIM-bonding) technique using alloying powder whose chemical compositions were similar to those of base metal. The base metals used were commercial Ni-base superalloys such as MM007, MarM247 and Inconel 713C alloys. They were bonded with both a Ni-15%Cr-4%B amorphous insert metal and IN100 alloying powder sheet with the thickness of about 250 μm. In the case of a MM007 alloy, the times required for completion of isothermal solidification and sequent homogenizing process were about 0.6 ks and 86.4 ks at 1423 K, respectively using this technique even with the bonding clearance of about 100 μm. The former time was about 1/80 and the latter time was about 1/30 compare to the times obtained by a conventional TLIM bonding. The short period during homogenizing process may be due to element diffusion from alloying powder to base metal by three-dimensional diffusion process. The tensile strength of MM007, MarM247 and Inconel 713C joints at 1255 K were almost equal to those of each base metal, respectively. From the experimental results, it should be useful to insert alloying powder to a wide bonding clearance joint during TLIM-bonding in fabrication
This paper was concerned with mechanism of isothermal solidification process on Transient Liquid Insert Metal Diffusion Bonding of Ni-base superalloys using Ni-base alloying powder. MM007 and IN100 alloys were used for base metals. IN 100 powder sheets were inserted between specimens with MBF-80 insert metal and then specimens were heated up at bonding temperatures of 1398 K. Content of liquid phase among alloying powders decreased with the lapse of bonding time. On the contrary, content of borides on dendrite boundary in alloying powders increased with the lapse of bonding time. Thermal analysis of mixed alloys with IN 100 and MBF-80 showed the possibility of crystalizing of borides in alloying powders during isothermal solidification process. It was believed concerning with the mechanism of isothermal solidification process that boron in liquid phase was consumed due to boride crystalizing in alloying powder and subsequently solid phase growth occurred in liquid phase.
This paper deals with the testing method for estimating the adhesive strength of sprayed coatings by the Implant type specimen loaded torsion because sprayed coatings usually are loaded in shear stress. The shearing strength obtained from this testing method was analysed based on the concept of energy release rate. To obtain the shearing strength of sprayed coatings, we made a new testing machine which can detect as a small stress as a strength of sprayed coatings. Energy release rate was calculated using numerical analysis. The result of the strength obtained from this method was reproducible to plus or minus 10% of a stress. From numerical analysis, the conditions for specimen size, in which the energy release rate caused by the residual stress of each specimen should be equal, was clarified. The energy release rate was useful to estimate the adhesive strength of sprayed coatings.
In the study of the functionally gradient material (FGM) systems as for the high temperature structural materials, designing of these and evaluation of their thermal shock properties under the temperature gradient condition would be the most essential problems to be investigated. In this study, ZrO2/NiCrAlY functionally gradient plasma sprayed coatings were produced on the stainless steel substrates. Simple heating-cooling thermal shock testing apparatus was made, and repeated thermal shock resistance properties of the coating specimens under the temperature gradient conditions were investigated using this apparatus. Conventional thermal shock test regulated by JIS was also conducted. Moderate coating configuration for minimizing the thermal stress were estimated by the simple analytical model. The results obtained are summarized as follows, (1) Feasibility of the functionally gradient coating were verified by the temperature gradient thermal shock testing conducted using the simple apparatus made. (2) The effect of coating configuration on the repeated thermal shock resistance properties could be evaluated by the temperature gradient test condition. (3) From the results of simple numerical analysis, the moderate coating configuration profile of the functionally gradient coatings was given as the compositional distribution parameter P of less than 1.
In a Power Fusion Reactor, a huge megnetic field is necessary for plasma confinement. This field also interacts with the high current to produce large forces tending to deform the superconducting cables ; this electro-magnetic force demands very high values of the cryogenic yield strength and fracture toughness in the supporting structural materials. It is said to be difficult to develop welds satisfing the same criteria as for the structural material. It is useful to investigate the usability of welded joint with a lower yield strength than the base metal, but associated with a higher fracture toughness. From this viewpoint, the influence of the undermatched joint on the deformation and the stressstrain property of the coil case was investigated by using a finite element method. It is concluded that (1) there is only a slight difference in the overall deformation of the coil case whether the undermatched joint is included in the calculation or not, (2) the maximum strain and the muximum plastic strain at the undermatched welds can be controlled to be less than those in the base metal with a judicious choice of weld location and yield strength level of welds.
In this study, we have been investigated on joining of titanium-matrix composites reinforced with SiC-CVD fiber (FRM) to FRM by tensile tests of the joints using butt joint, with a doubler and overlap length of 20 mm. The joinability was compared with solid state diffusion bonding and diffusion brazing using commercially pure copper and Ti-37.5Zr-15Ni-10Cu amorphous filler. As for mechanical properties of the joint, even the joint made at 1143 K for 5 min under the joining pressure of 12 MPa, which is slightly higher than that required in solid state diffusion bonding of the same material, was fractured in the base metal with tensile strength of 1400 MPa, resulting a joint efficiency of about 96%. The fiber damage decreases the strength of the joint as the joinging pressure increases. The diffusion brazability using commercially pure copper of 5 μm thickness at 1153 K was obtained good joint efficiency for 3 min under the joining pressure of 2 MPa, lower than that of diffusion bonding. In the diffusion brazability using Ti-base filler metal, the strength of the joint was equal to the strength of the base metal. Butt joint, with a doubler and overlap length of 10 and 15 mm, failed in shear at a tensile stress in the base material, resulting in a shear strength of about 125 MPa. The poor results were attributed to composites material of lower interlaminar shear strength.
A method of thermal history simulation has been developed to obtain a better understanding of mechanical properties of the grain growth region (heat-affected zone-HAZ) of gold bonding wires. The simulator is degigned to simulate the HAZ microstructures over a several inches length of the wire, which is necessary for mechanical testing. It has been found that grain size of the coarse grained HAZ cal be satisfactorily reproduced on the simulated thermal history condition. The breaking load obtained in this condition found to exhibit good agreement with the pull load of actual ball bonded wires. Using the simulator, changes in mechanical properties with varying simulated thermal cycles are briefly studied. With increase of grain size, strength of simulated HAZ decreased. On the other hand, elongation first increased with decreasing the strength but decreased sharply to a very low level. The intensive decrease of the elongation occurs because deformation does not occur uniformely but concentrated on heavily coarsened grains.