In order to clarify spalling mechanism of oxide scale on NiAl, microstructures in the vicinity of NiAl–X(Hf, Zr)⁄oxide scale interfaces were observed by using TEM. The oxidation tests were done at 1373 K in air for up to 360 ks. In NiAl, many voids were formed at the substrate⁄Al2O3 interface, leading to spalling of the alumina scale. On the other hand, addition of Hf or Zr to NiAl led to suppression of the void formation, resulting in prevention of spalling of alumina scale. In addition, a Ni–AlO layer tended to be formed at the interface in NiAl. This demonstrates appreciable increase in oxygen partial pressure and nickel activity at the interface. In fact, GDS measurements indicated depression of the Al concentration in the substrate close to an alumina scale. Consequently, it was speculated that the addition of Hf or Zr led to increase in diffusivity of Al in the substrate and suppression of void formation at NiAl⁄alumina scale, resulting in prevention of spalling of an alumina scale.
Cu-5Si-xTi filler alloys for joining Si3N4 were developed, and the joining ability of the alloys was investigated by measuring contact angles on Si3N4 in vacuum and strength of Si3N4 joint brazed with the filler alloys. The addition of Ti content of 3 at% or more decreased the contact angle on Si3N4. The alloys containing Ti content of 5at% showed the lowest contact angle and good wettability. Si3N4 was brazed to Si3N4 with the filler alloys containing Ti content of 3 at% or more, and the strength of the Si3N4 joint increased with the Ti content in the filler alloys. The increase in the amount of TiN and Ti5Si3 which was formed by the reaction of Ti in the filler with Si3N4 raised the strength of the Si3N4 joint.
It is well-known that liquid zinc-induced cracking occurs at the heat affected zone when the welded steel structures are immersed in the liquid zinc. The composition, the microstructure and the hardness of steel are related to this crack. Moreover, this crack is caused by bending work, but there has been no systematic research paper published on this to date. In this research, the influences of the bending work level and the degree of cleanness for steel surface with liquid zinc induced cracking were examined. The following results were obtained. The liquid zinc-induced cracking was caused by only bending work when the residual stresses induced by bending work were over 680MPa. When the residual stresses were under 680MPa, the cracking was caused at the yield stress of steel. The liquid zinc-induced cracking initiated at the grain boundaries of the steel surface opened by bending work and propagated along the grain boundaries of ferrite or the interface of ferrite and pearlite. The liquid zinc-induced cracking from bending work was able to be prevented when the surface of steel was ground by 100μm.
The keyhole behaviour and bubble formation were investigated in the full penetration laser welding of 11 and 15 mm thick plate with various kinds of atmosphere at the bottom side. Prevention of some defects was also attempted. Back surface atmosphere was controlled using a shielding box fixed on the sample back surface to avoid contamination from the air and also the effect of gas flow on the keyhole behaviour. A lot of bubbles were formed during welding with the back surface atmosphere of nitrogen and air. This was deduced to be caused by supersaturation of nitrogen, which was dissolved from the backside molten pool. The critical nitrogen concentration in the weld metal to form the porosity was independent of the back surface atmosphere. Oxygen enhanced the porosity formation, since the nitrogen concentration increased in the presence of oxygen. The keyhole was significantly perturbed and hot cracking occurred in the inert gas back shielding. It was shown that the keyhole stability determined by solubility of the shielding gas in the molten steel. Aluminum coating on the sample back surface was effective in preventing the porosity due to denitrification of the molten pool and hot cracking.
The effect of heat input on joint strength in the brake type friction welding has become clear, but the effect in the inertia type friction welding is not clear even at present. Then in this report, inertia type friction welding of 6061 aluminum alloy was carried out in order to examine the effect of unit heat input, burn-off length and burr shape on the joint strength. The heat input (mechanical work) was calculated by burn-off speed and welding pressure in the final stage. The joint strength was examined by tensile test. The final stage (welding time) of inertia type friction welding corresponding to the upset stage of brake type friction welding was searched using the relationship between heat input and joint strength. Using the most proper-final stage obtained, the relationships between unit final deformation heat input and joint strength, final burn-off length and joint strength, and burr shape and final deformation heat input were examined. Also the minimum unit final deformation heat input and minimum final burn-off length required for making a sound joint were also examined. The results showed that within this experiment the most proper-final stage (welding time) was 0.1s, and the limit unit final deformation heat input and limit final burn-off length were 19J/s and 0.3mm, respectively.