In arc welding, heat source property of arc plasma depends on temperature of arc plasma. Temperature distribution of arc plasma is affected by metal vapor which evaporate from molten metal. In this paper, dynamical temperature distribution during TIG welding is obtained through optical measurement with a high speed video camera. Furthermore, spectral images of metal vapor are taken simultaneously, therefore distributions of metal vapor are also obtained. Temperature of arc plasma decreases with mixing metal vapor into the arc plasma. Manganese vapor, namely the metal which has low boiling point, is distributed widely, and concentration is high. On the other hand, iron vapor, namely the metal which has high boiling point, is distributed only near the center of the arc, and concentration is low.
In present study, a method for measurement and visualization of temperature of plasma and concentration of metal vapor in arc welding with consumable electrode is conducted. Monochromatic images by three high-speed video cameras are utilized in order to clarify behavior of plasma states. The plasmas photographed by the high-speed video cameras during arc welding. Consequently, visualizations of plasma temperature and iron vapor concentration succeed during welding. The dual structure consisting of a low temperature region by metal vapor plasma near the arc axis and a high temperature region by shielding gas plasma at surrounding of the arc exists because of metal vapor which is vaporized at the tip surface of a consumable electrode and is transported by the plasma jet from the consumable electrode to the weld pool.
Ag-based filler metals containing Ni element have been recommended for brazing of cemented carbide because the Ni element is considered to improve the wetting characteristics to the cemented carbide. However, the reason for the improvement is still unknown. The present study was aimed at clarifying the reason why the Ni element in Ag-based filler metals makes the wetting characteristics improved and the effects of the Ni element on the bending strength of the brazed joints. Ag-Cu-Zn ternary alloys with and without Ni element were produced to investigate the effects of the Ni element. Furthermore, Co element was added into the filler metals to prevent the Co element in the cemented carbide from diffusing into the brazed layer, and the effects of the Co element on the bending strength of the brazed joints were investigated. The obtained results in this study are as follows. By the Ni element addition into the Ag-based filler metal, α-Cu solid solution phases were formed in the brazed layer along the interface; consequently, the distribution of Ag and Co elements across the interface between the cemented carbide and the α-Cu solid solution phases continuously changed. It seemed that the continuous distribution of Ag and Co elements across the interface caused the wetting characteristics to improve. The Ni element addition expanded the Co-depleted zone of the cemented carbide around the interface and decreased the bending strength of the joint brazed at 750°C. However, when the brazing temperature was set at 850°C, much amounts of Ni element diffused into the Co-depleted zone. As a result, the deterioration of the bending strength of the brazed joint was small. The Co element addition made the Co-depleted zone narrow, resulting in the increase of the bending strength of the brazed joints.
On TIG-MIG hybrid welding process, MIG arc is stable even though pure Ar shielded condition. Therefore, this process has a possibility as new type of welding process with high quality and high efficiency. However, the mechanism of this process is not clear due to complex interaction between TIG arc and MIG arc. In this study, we simulated arc phenomena in TIG-MIG hybrid welding process by using three-dimensional numerical model. Consequently, it was found that the direct path of electrical current into TIG electrode from MIG electrode is provided by the arc plasma distributed between TIG and MIG electrodes. Furthermore, the predicted penetration depth by simulation result which includes the influence of convective flow in the weld-pool is about in agreement with the experimental result, and there is remarkable difference in thermal efficiency and the detail of MIG arc current on base metal surface by whether the model includes the convictive flow in weld pool.
The objective of this study is to ensure the safety of nuclear reactors. A number of accidents occurred from the welded zones of the reactor vessel or the coolant pipes have been reported at sites around the world. One of the main causes is welding residual stress. It is therefore very important to know the welding residual stress in order to maintain the safety of the plants. Welded joints in nuclear reactors have complex shapes, and the welding residual stresses have complex three-dimensional distributions. In this study, the mock-up is idealized and manufactured for the welded joint at the pipe penetration part of actual reactor vessel. The inherent strain method is applied to accurately measure the welding residual stress of the joint. The inherent strain method is an analytical method which solves an inverse problem using a least squares method. The least squares method is the maximum likelihood estimation method, if the errors of data satisfy a normal distribution, and the fitting model has no approximation error. But these assumptions are hardly satisfied. The robust estimation method intends to improve the estimation accuracy, by weighting to the data, even if the data and the model are incomplete. In this study, Biweight method and adaptable Biweight method from the robust estimation methods are applied to the inherent strain method. In these methods, all measured data are used with the same weight 1 in the first analysis. The calculation is repeated, adjusting the weights based on the residuals of last calculation. The theory of the inherent strain method incorporated with such robust estimate method is developed. The welding residual stress of the above joint is analyzed by Biweight method and adaptable Biweight method. As a result, high-accurate estimation is achieved.
Hybrid welding process is effective technique to solve the problem of the welding execution by combining the merit of various heat sources. However, the hybrid welding process is complicated by interaction of the heat sources and selection of the suitable welding condition is difficult. The numerical analysis enables to investigate an influence of each parameter on hybrid arc phenomena theoretically. In this study, we simulated the influence of the hybridization condition such as TIG current and torch angles on the TIG-MIG hybrid arc phenomena. The main results are as follows: (1) Repulsion between TIG and MIG arc depends on the distance between the tip of both electrodes. (2) More than certain degree of TIG and MIG torches that is cross each other causes formation of direct current path between both electrodes. (3) The direct current does not increase so much with TIG current and reaches upper limit in certain value. When the TIG current is over the limit value, all increase current flow forward the base metal, and contributes to heat input.
In order to verify whether the initiation of the fatigue cracks at the inside of the U-rib of the steel plate deck (the invisible part) could be detected and the propagation of them toward the thickness direction of the deck plate could be monitored by FSM or not, a series of experiments and 3D static electric field analyses were carried out. The initiation of these fatigue cracks could be detected and the propagation of them toward the thickness direction of the deck plate could be monitored with high accuracy by the potential difference change (FC value) of the pairs sandwiching the fillet welds of the left and right sides of the U-rib. The fatigue cracks could be detected extremely earlier by FSM than the visual inspection by which the crack penetrated the deck plate. The features of the potential difference change with the initiation/propagation of the crack could be simulated by the 3D static electric field analysis. The results indicated that the initiation/propagation of the invisible cracks at the left or the right sides of U-rib could be detected and monitored by noting the behavior of FC value.
Buried pipelines may be deformed due to earthquakes and also corrode despite corrosion control measures such as protective coatings and cathodic protection. In such cases, it is necessary to ensure the integrity of the corroded pipelines against earthquakes. This study developed a method to evaluate the earthquake resistance of corroded pipelines subjected to seismic motions. Pipes were subjected to artificial local metal loss and to axial cyclic loading tests to clarify the cyclic deformation behavior until buckling occurs under seismic motion. As the cyclic loading progressed, displacement shifted to the compression side due to the formation of a bulge. Finally, the pipe buckled after several cycles. To evaluate the earthquake resistance of different pipelines, with varying degrees of local metal loss, a finite-element analysis method was developed that simulates the cyclic deformation behavior. A combination of kinematic and isotropic hardening components was used to model the material properties. These components were obtained from small specimen tests that consisted of a monotonic tensile test and a low cycle fatigue test under specific strain amplitude. This method enabled the successful prediction of the cyclic deformation behavior, including the number of cycles required for the buckling of pipes with varying degrees of metal loss.
Pure Al baseplates integrated with AlN substrates for power modules were tried to produce by joining-casting process, and applicability of the directional solidification technique using chills was investigated. Although no fetal casting defects were observed in obtained integrated baseplates, string-like surface defects, called 'wrinkles', were observed on them. On the basis of these generation status and the calculated solidification pattern, the wrinkles were considered to be appeared by the deformation of the front of the initially solidified shells, as in the case of low melting point metals in past studies. In the calculated surface temperature gradient distribution of the castings, a temperature gradient area smaller than that of surroundings ('minimal zone') appeared at a position similar to that of the wrinkle. These results suggest that the generation of wrinkles in the integrated baseplates can be predicted by checking the existence of minimal zones. In fact, the availability of this criterion for investigation of casting conditions was confirmed. Depending on optimization of casting design, it was found that the integrated baseplates which have no wrinkles were able to produce by joining-casting using double-layerd molds. In obtained Al-AlN interface, there were remarkably less joining defects. It was also found that the interface also have practically enough thermal-cycle reliability.
The measuring methods of the residual stress are classified into destructive one and nondestructive one. The inherent strain method (ISM) is destructive one. The neutron diffraction method (NDM) is nondestructive one. But the measurable depth is limited within about 20 mm and the method cannot measure the weld zone, without destruction of the object. So, in this study, the hybrid measuring method has been developed, by combining the ISM and the NDM. The theory of the hybrid method is the same as the ISM. In the analysis, the strains measured by the NDM without destruction are used. This hybrid measuring method is a true nondestructive measuring method for a thick welded joint. The applicability of the hybrid method has been verified by simulation, using a butt welded joint of thick pipes. In the simulation, the reliable order of the strains measured by the present NDM is very important, and was considered as 10 micro. The measurable regions by the present NDM were assumed. Under the above conditions, the data (the residual elastic strains assumed to be measured by the NDM) were made, and used in the ISM. As a result of such simulation, it has been cleared that the estimated residual stress has very high accuracy, if enough data are used. The required number of data is less than the ISM.