One of the important problem in welding engineering is to develop a mathematical method for the determination of optimal process parameters. From this viewpoint, an algorithm for the determination of optimal heat input condition of arc welding has been discussed in the previous paper, where an analytical model of heat conduction has been used to evaluate the temperatuer field during welding and the nonlinear programming technique has been applied to estimate the optimal heat input sequence for a required temperature field. In this paper, the optimal heat input sequence derived from the above method has been discussed experimentally and as a result of the comparison between experimental and required temperature fields, it is made clear that the algorithm for optimal heat input sequence is reliable. In the second part of this paper, a computerized optimal control system for thin plate GTAW has been developed on the basis of the algorithm, where the arc efficiency is continuously identified by the temperature detected by IR sensor. Under the system, a real-time and adaptive optimization of process parameters is possible by taking account of the actual arc efficiency. From the satisfactory result of experiment on the system, it is concluded that the algorithm described in the previous paper is a useful guide for optimal heat input control of arc welding.
In this paper, a feasibility study has been performed to determine if a computerized method by the numerical model of process could be used to estimate the optimal process parameter in arc welding. In the first part of this paper, a method of optimization of process parameters in arc welding has been discussed. The method is based on the numerical calculation of weld pool by a finite difference model, and a nonlinear optimization technique is applied to estimate the optimal process parameter in arc welding. In the second part of this paper, a system of on-line optimization has been developed on the basis of the above method, where the temperature in welding detected by infrared sensors is used to identify the numerical moedl. The developed system for TIG arc welding made it possible to optimize the process parameter in real time. Accordingly, it is concluded that the method based on the numerical simulation of welding process is a useful guide to the optimal process parameter in arc welding.
In deep penetration electron beam welds, some characteristic defects such as porosity and solidification crack have been occurred depending on the occurrence of the locally delayed solidification pattern. In this report, some dominant factors affected on the delayed solidification have been investigated. The results obtained were as follows. (1) The delayed solidification was resulted from remelting or secondary melting effect of the solidified wall by the stagnated molten metal in a beam cavity. (2) The amount of the stagnated molten metal depended on how easily the molten metal flowed out of the beam cavity. (3) Strong evaporation near the root promoted the smooth flow of molten metal, reslting in suppressing the delayed solidification. (4) Gases involved in the base metal and evaporation from the front wall of the cavity perturbed the molten metal flow, resulting in promoting the delayed solidification.
Aluminum specimens consisted of couples of tubes and cores were welded by electromangetic welding at elevated temperatures up to 500°C. An equipment with capacitor of 470, UF and nominal voltage o f 10 kV was used. To discuss the results of welding test, the collision velocity and the collision acceleration were evaluated by the deformation analysis on specimen tubes at elevated temperatures. Several flow stresses of aluminum required for the analysis were obtained by the deformation tests on specimen tubes and the hardness tests on deformed ones. The following results were obtained. (1) The electromagnetic energy required for welding at 500°C amounts to about half of that at the room temperature. (2) Althouhg the strain rate dependance of flow stress of aluminum decreases with a raising temperature, the flow stress itself decreases with the temperature, because the strain hardening is remarkably reduced with the temperature. (3) The favorable effect of heating the specimen is caused by two factors, i) the increases of collision velocity and collision acceleration of tube, ii) the decreases of the critical collision velocity and the collision acceleration required for welding. The fractions of their contributions are 30%(i) to 70% (ii).
A reactive bniding method has been investigated to metalize SiC substrates with Cu foils. The bonding was carried out in argon atmosphere by using Cu-Mn liquid insert alloys. Bonding strength was measured by peel test, and the bodning interface was analyzed with EPMA. Cu-Mn insert alloys containing Mn more than 35wt% showed an excellent bondability by heating just above the melting point of insert fiols for a short time. Interfacial analysis revealed that SiC is decomposed by Cu-Mn liquid, and the decomppsed element of C remains at SiC/Cu-Mn interface, but Si dissolves into Cu-Mn liquid. Cu bonded to SiC kept a high bonding strength in reheating treatment below 600°C, but decreased the strength above 600°C. Thermal cycle test (-50°C-150°C) showed that no crack was observed in SiC substrates bonded with Cu foiis less than 30 μm in thickness even after 1000 cycles. SiC surface from which Cu was removed by etching had almost same electric resistivity as original surface.
Austenitic stainless steel pipes, type 316 and 304, were joined by liquid phase diffusion bonding process. Six specially designed interlayer alloys, three Fe-base and three Ni-base, which were produced by rapid quenching from the liquid state, were used in this experiment. Austenitic stainless steel pipes bonded by Fe-base interlayer alloys have proper mechanical properties same as the base metal although intergranular corrosion was observed after Strauss tests because of the diffusion of boron near bonding interface. Welded joints by Ni-base interlayer alloys have good resistance to intergranular corrosion even after Strauss test. Intergranular cracks were observed in the straight seam weld metal of type 304 pipe heated up to the sensitization temperature after Strauss test, but were prevented by solution treatment before joining. Lowering the carbon content in austenitic stainless steel was also effective to prevent the intergranular cracks.
It is the ultimate purpose of this investigation to elucidate the fundamental phenomena in cooling and solidification process and to establish reasonably the methods of estimating hot cracking sensitivity and preventing Al alloy weld from hot cracking. In this report, temperature measurement was carried out by CA thermocouple in coiling and solidification process on TIG arc spot welds of commercial Al alloy 2024 and 5083 and various analyses were performed. On the basis of the measured value, the authors made Continuous Cooling Solidification Process diagram (CCSP diagram) showing the relation among passed time from arc stop, phase change and the position of welds. By using the CCSP diagram, the authors could investigate the solidification process of the welds. With regard to fraction solid cooling rate (cooling rate in solid-liquid coexisting zone), it was found that the cooling rate at every definite fraction solid was more meaningful than one in a whole solidification time. Increasing behavior of fraction solid were very different between 2024 and 5083 Al alloy at the latter stage of solidification process in the molten pool.
Recently, it has been reported that sensitization in welded joints could be promoted remarkably in the range of lower temperature, in which region, sentization would be hardly indcued for solution heat treated materials. This phenomenon, known as LTS (Low Temperature Sensitization), is becoming a serious problem for operating all sort of reactors and pressure vessels in the high temperature condition. This would be caused by fine carbide nucleation which was induced in weld thermal cycle duration. In this paper, effects of weld thermal cycle on the susceptibility to LTS were discussed by unstable heat conduction analysis. The following properties became evident after examining the experimental results. The peak temperature in weld thermal cycles which could cause LTS in the post weld heat duration, was decreased gradually and LTS region was extended with the increase of weld heat input. And it was also recognized that the susceptibility to LTS in Type 304L was much lower than that in Type 304. According to welding method to imporve the resistance to LTS, application of water quenching welding method was very effective, because of preventing of the carbide nucleation by rapid cooling in weld thermal cycle duration.
The relation between the parameter of "the concentration of dissolved sulfur, (S)" and the reheat cracking sensitivity was examined on 1%Cr-0.5%Mo steel by the following experiment and metallurgical considerations. (1) The reheat cracking sensitivity of each steel specimen was compared by the critical restraint stress to produce cracking, σAW-crit obtained by the modified implant test. Seventeen steels were used for the test; they contained 0.015 to 0.09%S and 0.1 to 1.3%Mn. (2) The sulfide was extracted by the electrolysis, and the (S) was calculated by reducing the weight of sulfur in the extracted sulfide from that in the bulk steel. (3) The critical restraint stress, σAW-crit was plotted against the (S); there is a critical value of (S), (S) crit, from which σAW-crit begins to decrease with increasing the (S) value. (4) A nomograph was proposed to estimate the content of dissolved sulfur by the sulfur and manganese contents in steel. This informs that manganese reduces (S) remarkably when manganese exceeds the critical value. (5) Sulfide will render much smaller effect on the reheat cracking than dissolved sulfur did.
In order to improve the mechanical properties of accelerated cooled TMCP type high strength steel UOE pipe flash butt weldment, reheating and quenching were proposed as a PWHT. Laboratory scale examination was performed, and the following conclusion is obtained. (1) Significant deterioration in toughness and tensile strength was found in both of TMCP steel weld-line. (2) After PWHT (reheating and quenching), the tensile strength of flash butt welded joints was restored at the base metals, and no softened region was observed in the welded joints. (3) The microstructure of flash butt welded joints after PWHT consisted of fine grained ferrite and bainite. (4) The simulated reheat transition zone showed the good mechanical properties. The tests results demonstrate that the proposed process is a promising measure to solve the problems of as-welded joints.
This paper describes an extensive study on the effects of C, Nb and Ni on the toughness of heat affected zone of welded joints in low Ni (1.5-3.5%) steels for LPG storage tanks and 9%Ni steel for LNG storage tanks. The toughness of heat affected zone of welded joints is mainly studied with use of the specimens orenared by thermal cycle simulation. The decrease in C content to 0.01 % improves the toughness of fusion line of welded joint of the 2.5% Ni steel, while it deteriorates the toughness of 9%Ni steel. In both steels, the fraction of island martensite in the welded joint decreases, resulting in the improvement of toughness. However, coarsening of microstructure, which lowers toughness, is more extensive for the 9%Ni steel. The difference in microstructure coarsening of both steels seems to be a cause for the difference in the toughness. The addition of 0.03%Nb improves the toughness of the 0.01-0.05%C-2.5%Ni steel through the refining of microstructure, but does not improve the toughness of the 0.01-0.06%C-9%Ni steel. The increase in Ni content from 1.5% to 2.5% improves the toughness of the 0.01%C steel, while a further increase of Ni content to 3.5% lowers toughness. This comes from the intensive formation of island martensite. These results are applied to the production of 0.01%C-2.5%Ni and 0.06%C-9%Ni steels with good toughness of fusion line of welded joints for LPG and LNG storage tanks, respectively.
The final purpose of our study is to discuss the selection of surface modification processes and/or materials from a viewpoint of blast erosion properties. In this study, an experimental investigation was carried out on the blast erosion resistance of overlay weld metals. Seven different weld metals of Ni-Cr steel, Co-base alloy and Ni-base alloy etc., whihc were produced by GMAW, SMAW and PTA, were examined in comparison with SUS304 steel and SCS5 cast steel. Experimental results showed that the better correlation between volume erosion rate and Vickers hardness after erosion could be obtained except the weld metals by PTA. The higher the Vickers hardness after erosion, the lower was the volume erosion rate. However, in the case of the weld metals by PTA, the volume erosion rates were higher than the rate of other materials for the large hard phase precipitated. The volume erosion rate of NbC dispersed weld metal showed good agreement with that of SUS304 matrix.
The increase of chromium and molybdenum content as alloying elements is effective to improve the corrosion resistance of duplex stainless steels. However, the increase of these elements have the tendency to induce the poor hot workability and inferior weldability of the steels. In the weldment the more those elements increase, the more often the embrittlement in HAZ takes place. The mechanism of the HAZ embrittlement in 28Cr-3.8Mo-Ni-N duplex stainless steels was investigated. It was clarified that there are two reasons for the HAZ embrittlement from the investigation on the effects of nickel and nitrogen contents. One was the sigma phase precipitation in case of the lower ferrite ratio. Another was the nitride precipitation in case of the higher ferrite ratio. In the proper range of nickel and nitrogen content, the HAZ showed the high toughness. The sigma phase precipitation was accelerated by the decrease of the ferrite ratio and the increase of the nickel content. The nitride precipitation was accelerated by the increase of the ferrite ratio and the decrease of the nickel content. The acceleration of the sigma phase precipitation in the increase of nickel content can be explained by as follows. One is that the increase of nickel enriches the ferrite phase in the chromium and molybdenum concentration. The other is that the nickel makes the ferrite phase unstable and increases the transformation driving force.
The effect of the transformation superplasticity (that is a martensite transformation) on the stress releasement has been examined. In the first study, using the new apparatus (a restraint thermal cycle test) by which the displacement of only the transformed part of the specimen can be measured, the proportion between transformation superplasticity and transformation expansion on the amount of the stress releasement was examined. The new composite specimen of which the test 9%Ni steel was joined by friction welding to SUS304 stainless steel on the both ends was also used. The results are shown as follows; 1) The expansion by only transformation superplasticity was separated from whole expansion during transformation using the new apparatus and specimen. 2) When the much amount of thermal stress is generated in spite of a small transformed area, the stress releasement during transformation chiefly occurs by transformation superplasticity. 3) Regard as stress releasement in the midst of cooling just after transformation start, it mainly occurs by superplasticity because of its high thermal stress. But, in the latter half of the transformation, the stress releases mainly by transformation expansion. 4) The proportion of the transformation expansion and superplasticity was measured.
This study was performed to obtain the optimum condition of TIG treatment, to confirm how there is improving effect on fatigue test based on maximum stress, and to analyze the improving factors quantitatively, in improving the fatigue strength of fillet welded joint by TIG arc remelting of weld toe region. Fourteen S-N fatigue tests (JSME-5002: Standard Method of Statistical Fatigue Testing) based on maximum stress fixed to yield stress (S=σmax-σmin, σmax=σY) were performed on non-load-carrying cruciform fillet welded joints of 569 MPa tensile strength class high strength steel (SM58Q). The joints were made by shielded metal arc welding, and the toe region of fillet weld was treated by TIG arc remelting. It was found that the fatigue limit of TIG arc treated test specimen obtained by staircase method judged at 2×106 cycles was about 2.7 times as high as one of as welded test specimen, and that weaving TIG arc treatment executed with relatively high heat input was more effective than other treatments, judging from various statistical analyses of fatigue strength on fatigue test based on maximum stress.
A series of studies has been performed with an objective of applying "Photo-acoustic spectroscopy (PAS)" as NDT to thermal sprayed specimens. In this report, the authors have developed a prototype instrument of PAS in order to apply to solid specimens with and without thermal spraying coatings, and basically studied PA characteristics of mild steel and copper spray deposited specimens. The main result obtained are as follows; The prototype instrument was able to apply to solid specimen with and without thermal spraying coating by using double beam method. As for PA characteristics of mild steel, it increased with decreasing of light wavelength and increasing of specimen volume, and was almost constant for specimens of which thickness was over than 2 mm. As for PA signal intensity of mild steel with copper spray coating, it decreased with increasing of light wavelength.
The feasibility of a rapid plasma ion nitriding process assisted by hollow cathode discharge has been examined for austenitic stainless steel (SUS304), ferritic stainless steel (SUS340) and Ni base heat resistant alloys (Inconel 625 and Udimet 500). Hollow cathode discharge was easily realized by arranging two specimens on a cathode plate, so that their treating faces were set parallel at an optimum distance, which was different depending on the pressure of N2+H2 mixed gas as 3 mm for 800 and 1330Pa, 5 mm for 270Pa or 15 mm for 80Pa. At these optimum conditions, the nitriding speed was increased to about 2 times as fast as that for a conventional plasma ion nitriding process at the same nitriding temperature and time without any decrease in the hardness of nitrided layer. This increase in nitriding speed was considered to be due to the high ionization rate of hollow cathode discharge in comparison with conventional glow discharge.
Pressureless sintered SiC ceramics joints and hot press sintered SiC ceramics joints were made by using active brazing metal (Ti-AgCu). Then the bonding strengths, the microstructures and reaction products of bonding layer were compared and discussed. The results of this study are as follows: (1) The four point bending strengths of pressureless sintered SiC ceramics joints were found to be equivalent to the bending strength of the SiC ceramics substrate (400 MPa). However, the four point bending strength of hot press sintered SiC ceramics joints were less than 100 MPa. Those strengths were extremely lower than that of the SiC ceramics substrate. Furthermore, the fracture position was at the interface between SiC ceramics and Ti-AgCu brazing metal. (2) A TiC layer (0.2 pm thick) was observed in the bonding layer between pressureless sintered SiC ceramics and Ti-AgCu brazing metal. However, in the case of hot press sintered SiC ceramics, Ag2Si, Ti5Si4 and TiC were observed on the SiC side fractured surface. This was supposed to be the cause for the low bonding strengths of hot press sintered SiC ceramics joints.
A bonding method in which the barrier layer was formed on the bonding surface of hot press sintered SiC ceramics (BeO or AlN as sintering additives) prior to active brazing metal (Ti-AgCu) was investigated in order to improve the bonding strength of SiC ceramics joints. In this report the material for the barrier layer and its effect were investigated. The results of this study are as follows: (1) The layer of reaction products between Cr powder and SiC ceramics as a barrier layer is effective to improve the bonding strength. (2) The four point bending strength of the SiC joints by Ti-AgCu brazing metal with the Cr barrier laver of about 1.4μm in thickness on SiC was improved by the factor 3 to 5, as compared with the joints without Cr barrier layer. Then fracture position was in the SiC substrate. It is considered that Ag2Si, Ti5Si4, etc. which were assumed to be the cause for low bonding strength in the previous report were not formed in the bonding layer.
Fiber reinforced metal composites were fabricated by diffusion welding. The matrices used were commercially pure titanium (CPTi) and two kinds of titanium alloys (Ti-6Al-4V and Ti-21V-4Al), and the filaments were two kinds of SiC-CVD fiber. The structure and the distribution of elements in reaction zone between matrix and filament both in as-fabricated state and in heat-treated state were investigated by microscopy, SEM/EDX microanalyses and transmission electron microscopy (TEM). From the results obtained, the kinetics and mechanism of growth of the reaction zone were discussed. It was found that the reaction zone was composed of TiC and Ti5Si3 and the growth rate of the zone in Ti-21V-4Al alloy matrix was lower than those in CPTi and Ti-6Al-4V alloy matrices. The apparent activation energy for the growth of the reaction zone, derived from the relation between heating temperature and reaction rate constants, revealed that the value of the activation energy was different between the heating temperatures above and below the α-β transformation (1153K) in case of CPTi or β transus temperature (1268K) in case of Ti-6Al-4V alloy matrices because of rapid diffusion rate and low solubility of carbon in beta phase. In the composite of the Ti-21V-4Al alloy matrix, the growth rate of raection zone was very low compared with that in the Ti-6Al-4V alloy, suggesting that large amounts of alloying elements of V and Al reduced the growth rate of reaction zone. This result indicates that the Ti-21V-4Al alloy is suited to the matrix of composites reinforced with SiC filament.