Large-scale duplex ESSO tests were conducted to investigate the effect of distance of loading points on long brittle crack propagation/arrest behavior. The test results showed that the greater load drop was found to be easier to arrest a long brittle crack in the tests of a 5m loading-points distance, compared to the tests of a 10m loading-points distance. Traditional static equivalent spring model could roughly describe the differences in load drop phenomena, due to differences in loading-points distance. Dynamic FEM analyses showed that specimen length was needed to be expanded to about 20m in order to prevent a load drop in large-scale duplex ESSO test, when the brittle crack velocity was about 500∼800m/s. Also, if the distance of loading points was over 10m or the specimen length of large-scale duplex ESSO test was over 6.8m, the dynamic stress intensity factor of the long brittle crack were found to be almost equivalent to the actual conditions in which the effects of load drop and stress wave were small. In the large-scale duplex ESSO test to evaluate a long brittle crack propagation/arrest behavior, it was considered that setting the appropriate test conditions was important, so as not to cause significant load drop and unloading stress wave.
The welding defects in friction stir welds (FSW) such as kissing bond and wormhole were inspected by non-destructive techniques based on X-ray inspection and ultrasonic testing with phased-array probe. Although the X-ray inspection could detect wormhole defects, kissing bonds were recognized only by the ultrasonic testing. The sizes of kissing bonds were estimated based on the maximum echo height of the ultrasonic testing, while the estimated sizes were smaller than the actual ones observed on the fatigue fracture surfaces. Fatigue tests were performed using welds with defects, revealing that the fatigue strengths were significantly reduced due to the early fatigue crack initiation from defects. The prediction of fatigue life of weld was conducted based on the Paris law of fatigue crack propagation rate. The predicted lives of the welds with defects correlated with the actual ones. The ultrasonic testing led to the linear relationship between the estimated and actual sizes of kissing bond, thus it would be possible to predict the fatigue life of the weld with defects by non-destructive inspection.
Creep tests of 2.25Cr-1Mo-V steel weld metal, which was made with tandem submerged arc process, were conducted. Temperature and stress levels for the creep tests were chosen so as to lead to short-time rupture of about 1,000 hours. Three creep prediction methods, such as Monkman-Grant law, Modified θ Projection and Ω method, were applied to each obtained creep curve. All three methods were proved to be practical for weld metal as well as base material. On the other hand, in the aspect of the perdition from early-stage, the Monkman-Grant law was proved to be most effective in this research. The equation, obtained in this study, was applicable also to 2.25Cr-1Mo weld metal containing various amounts of V, Nb, B, and W.
Mixing behaviors of the nitrogen into the arc plasma in GTA welding is studied by numerical analysis. The whole region consisted of a tungsten cathode, an arc, a shield nozzle and a water cooled cupper anode is treated in the unified two-dimensional numerical model. Simulation results show that the argon arc has higher shielding gas concentration compared with the helium arc. Additionally, it is assumed that the shielding gas concentration in the arc plasma is affected by the shielding gas density and the diffusion coefficient. Lower gas density reduces the shielded area and higher diffusion coefficient increases the nitrogen mixture in plasma region of high temperature range. The reduction of shielding gas concentration in helium arcs, especially, is caused by the higher diffusion coefficient of the helium plasma.
Recently, manufacturing of flat panel displays has been required to achieve high cost-performance and to concern about environmental issues. For these reasons, application of the lift-off process in place of the etching process to electrode pattern formation in manufacturing of plasma display panels was considered. To resolve problems in conventional lift-off process caused by inappropriate resist profiles, the inversely-tapered resist profile with interstice was proposed, and its fundamental feasibility was experimentally proved. However, even if this resist profile is employed, the problems in the lift-off process still occur when its dimensions are inappropriately given. Therefore, proper design of that resist profile is important. Generally, design of the profile is achieved by iteration of two processes: one is calculation of thickness distribution of the formed electrode pattern, and the other is optimization of dimensions of the profile. It is required that the calculation process is carried out with small computational load and supplies useful information for the optimization process. For this reason, a calculation method which meets the requirements and a design method based on this new method were proposed. However, this calculation method considers only two-dimensional behavior of depositing material in a cross sectional plane. This paper describes enhancement of the calculation method by taking three-dimensional behavior of depositing material into consideration, and reports better consistency between calculation results by the proposed method and experimental results.
The coaxial multi-layer solid wire for 9% Ni steel have been developed to perform stable GMA welding in pure Argon shielding gas. This coaxial multi-layer solid wire has a double structure consisting of a different composition in its inner and outer parts. The average composition of this wire was the same level as the composition of the solid welding wire of 9% Ni steel. To stabilize the welding in pure Argon shielding gas, potassium compounds were used for this coaxial multi-layer solid wire. The potassium compounds were added between inner part (center wire) and outer parts (hoop). First, the wire melting behavior was observed with high-speed video camera system. Then the effect of potassium was confirmed experimentally. Next, the relation between the amount of potassium and the length of column of liquid molten metal (CLM), which causes arc instability in pure Argon shielding gas, was analyzed. The coaxial multi-layer solid wire which contained potassium more than 0.001 wt % made the CLM shorten. As a result, the stable welding in pure Argon shielding gas was achieved. Lastly, the welding was carried out at V-groove to assess the ductility and strength of joints. The coaxial multi-layer solid wire which has appropriate composition and constant current characteristics power source were used. An excellent weld joint was obtained in pure Ar shielding gas. The Charpy absorbed energy was 89 J at 77 K (-196°C), 0.2% proof stress was 710 MPa and tensile strength was 855 MPa.
The effects of welding conditions on the residual stress and distortion behaviors of the full and partial-penetration weld joints with thin steel plates in the use of the electron beam welding process have been studied. Residual stresses and angular distortion were measured, and the experimental results have been verified with the aid of the thermal elastic-plastic analysis by the finite element method. The experimental results of residual stress and angular distortion have been clarified to be similar to the analysis results. The residual stresses (σT and σL) in the transverse and longitudinal directions with respect to the weld line have been confirmed to be shifted to the compression side in the vicinity of the weld toe when the beam power was higher and the beam diameter was smaller. The magnitude of residual stresses at the weld toe has been found to be related to the penetration depth. It has also been revealed that angular distortion tends to increase as the beam power becomes higher.
Dissimilar metals of 6022 aluminum alloy and galvannealed steel were joined successfully by laser brazing with new zinc brazing filler metal which was added Si.Si addition in the filler metal has great effects in preventing the growth of the IMC layer such as Fe2Al5 and minimizing its thickness. The peel strength of the joint with ZnSi filler metal increased by 3.5 times from that of joint with ZnAl filler metal. Fracture occurred at aluminum HAZ for joint with ZnSi filler metal and at brazing interface for joint with ZnAl filler metal. From TEM observation and diffraction analysis of brazing interface with ZnSi filler metal, Fe3Al2Si3 that had the same lattice spacing as α-Fe was observed. Thus, Fe3Al2Si3 that formed on the surface of steel prevented the formation of brittle intermetallic compound like Fe2Al5.