This study was carried out to achieve the granulation or grain refinement of solidification structure of fully austenitic stainless steel weld metal by pulsed TIG welding. Welding variables examined in this study were the pulse frequency and the ratio of base current time (tb) to peak current time (tp) in a rectangular pulsed current wave. The granulation or grain refinement of the solidification structure depended on the pulse frequency with the most effective refinement achieved at a pulse frequency of nearly 2.5 Hz and al. tb/tp of 9 to 10. The welding conditions in this case were a welding speed of 6 cm/min, average welding current of 50 A, and base current of 10 A. The pulse frequency-dependent granulation or grain refinement could be explained in terms of the difference in the degree of constitutional supercooling.
In order to investigate the weld cold cracking of medium, high carbon low alloy steels which is considered to be caused by quench cracking, the fundamental behaviors of this type of cracking were studied by means of the RRC test. Namely, HAZ cracking in JIS SNCM439 (0.4C-1.8Ni-0.8Cr-0.3Mo steel) and JIS SK5 (0.8C plain carbon steel) was studied under the conditions of restraint intensity of about 9.8 to 29.4 kN/mm·mm without preheating utilizing GTA welding with austenitic filler wire to eliminate the effect of diffusible hydrogen. Main conclusions obtained are as follows: (1) General behavior of restraint stress after welding of these steels nearly agrees with that of weldable high strength steels in which martensitic transformation reduces restraint stress. (2) The last stage which gives the maximum or saturated restraint stress after welding is not seen even in the low restraint intensity of about 10 kN/mm.mm, because the cracking develops largely before the last stage. (3) Crack initiation estimated with AE technique occurs earlier with increasing the restraint intensity. The highest temperature of crack initiation, however, was about 35- K and 393 K in SNCM439 and SK5, respectively, meaning fairly lower than their Ms temperatures. (4) The stress at the crack initiation was about 400 MPa and 270 MPa in SNCM439 and SK5, respectively irrespective of the restraint intensity. (5) Crack initiation part has an intimate correlation with liquated grain-boundary in HAZ near fusion boundary, and gives very brittle intergranular fracture surface.
Effect of welding tensile residual stresses on fatigue crack growth retardation following single tensile overload in low ΔK region was studied experimentally. Material used was steel of 800 MPa in tensile strength. Specimens were center-cracked-plates and consisted of two kinds; specimens without weld, specimens with longitudinal weld. Major results obtained are summarized as follows: (1) Tensile residual stresses decreased the level of crack growth retardation. (2) Compressive overload succeeded to tensile overload decreased the level of crack growth retardation. Especially the effect was more remarkable in specimens with longitudinal weld. (3) The retardation couldn't be explained from crack closure behavior.
In previous work it has been shown that diffraction effects in the relatively coarse-grained metal of the weld is a prime cause of the mottling. However, the mottling is most commonly observed in thin welds when one is using low voltage X-rays, with which differential absorption will be greater. There appears to be a possibility that a part of the mottling might be due to small differences in X-ray absorption in individual grains in the metal. The extent of the contribution of the differential absorption to the mottled appearance of radiographs has, therefore, been studied using austenitic stainless steel welds. Experimental results show that the extent of the contribution from the differential absorption is unimportant in so far as the austenitic stainless steel weld is concerned.
Dissimilar joint of nonferrous metals like as contact-bars of electromagnetic swtichgears are required both high quality at bonds and non-deformation of works. For the purposes, vapor-shielded pressure bonding process and inprocess control system applicable to the process are developed. From the view point of controlling work deformation, heating characteristics and temperature distribution while bonding is suggested to be controlled by computer simulation shown in the Report 1. In this study, the actual proof for the computer simulation, clarification of heating mechanism at each material system and selection of electrode constitution to prevent the work deformation are experimentally examined for the joint couple of disk-shape Ag and Cu. Results obtained are summarized as follows; (1) Electrodes Mo-Mo give relatively flat temperature distribution all over the joint materials. (2) Electrodes CuCr-CuCr can give the peak temperature at the bonding interface, so that the material (Ag) deformation is effectively controlled. But, twice current is necessary to bond in comparison with electrodes Mo-Mo. (3) The computer simulation applied in the Report 1 is proved to be proper for estimating heating characteristics and temperature distribution in the bonding process, that is, both results of the experiments and the simulation agree with each other. (4) Arrangement of electrodes such as CuCr to Ag side and Mo to Cu side can achieve a deformationfree bonding under relatively lower current condition and lower average temperature in Ag materials. (5) In the above-mentioned constitution of electrodes, the optimum current range becomes to be enlarged. Since the allowable current limit against deformation extends to higher current condition.
Bonding of Si3N4 to refractory metals has been investigated using liquid insert metals of copper base alloys. The bonding has been carried out in an electric furnace in vacuum of 6 mPa. Strength of joints was evaluated by tensile-shear tests made at temperature from room temperature to 1073 K. Copper-Haynes No. 188 alloys which showed good wettability with Si3N4 were selected as an insert metal. Strength obtained for joints of Si3N4 to Si3N4, tungsten, molybdenum, niobium and tantalum were about 160 MPa, 80 MPa, 100 MPa, 60 MPa and 55 MPa, respectively. Strength of joints of Si3N4 to tungsten and molybdenum increased with the rise of test temperature from room temperature to 473 K and reached 90 MPa and 100 MPa, respectively. In the joint of Si3N4 to Si3N4, voids were observed' in the insert layer except for the insert metal of copper with 20-30%Haynes No. 188 alloys. Liquidus and solidus temperature of insert metals decreased with increasing copper content. The improved fluidity of the insert metal caused by the decrease in the melting temperature seemed to have the effect of preventing the void formation in the insert layer, resulting in improvement of the strength of the joint.
The roles of elements in the insert metal of 80%copper-20%Haynes No. 188 for bonding of Si3N4 to tungsten and molybdenum have been investigated. EDX analysis of the interlayer of joints revealed that chromium was concentrated in the region adjacent to Si3N4 and copper was concentrated in the center region of the insert layer. CrSi2, Cr2N and CrN were found in the region adjacent to Si3N4. As the results of bonding of Si3N4 to tungsten using insert metals of copper base ternary and binary alloys containing iron, nickel, chromium or cobalt, good strength of joints were obtained only when chromium was contained in the insert metal. These results suggest that chromium is the main element to contribute to bonding of the insert metal to Si3N4 by forming chromium silicide or nitride at the interface and copper which enriched in the center region had the function to relieve the thermal stress by forming buffer layer in the insert layer.
Resistance pressure welding of SUS304 stainless steel using insert material of Ni-base BNi-2 brazing alloy powder was studied. The effect of the welding conditions upon mechanical and metallurgical properties of the joint was investigated. The results are summarized as followed: (1) The tensile strength of the joint is higher than 600 MPa (equal to that of the base metal), and the shrinkage of weldment for 10 mm is less than 0.1 mm. These properties are obtained by using the following conditions: weld current 157 A/mm2, weld time 0.2 sec and pressure 49 MPa. (2) Temperatures at the joint surface and at a distance of 1 mm from the joint interface are different by more than 300°C on the welding conditions of weld current 254 A/mm2, weld time 0.04 sec and pressure 49 MPa. (3) In the case of the same deformation, metallurgically better welding results are obtained by using the welding conditions of higher weld current and shorter weld time. For example, on the welding conditions of 254 A/mm2 and 0.08 sec, Si content in the joint layer equals to that in the base metal, but on 58 A/mm2 and 5.0 sec, Si element gathers like a film at the center of the joint layer.
In the spot welding process using high-current density (nearly 2.0 kA/mm2) and short current duration (4.0 ms ), the indentation on sheet is smaller than one in the conventional spot welding process (heat conduction type). But, in case of welding using R type or CF type electrode tips, its indentation is clear in visual inspection at welds showing the tear rupture at the tensile shear testing. In this report, it is cleared that the indentation is formed mainly by following three factors; the softening in sheet surface caused by heating by the contact resistance between electrode and sheet, the current concentration in the electrode edge, and electrode force. Furthermore, it is shown experimentally that the joint of thin steel sheets or Al alloy plates with enough tensile shear strength and extremely small indentation (about 0.01 mm) are obtained easily by using the combination of R type and F type electrodes.
This report is conserned with the effect of chemical composition of base material, especially Si and Mn, on flash weldability. Si and Mn influence the fracture or crack of welded part in practical production line, for example, wheel-rim forming line in the automobile companies and pickling line or cold-rolling line in the steel industries. The reason why Si and Mn influence the fracture or crack is that fracture is caused by Si-Mn-Al oxide in the weld line which hasn't been removed out from weld line to outside at the upset stage. Though the fracture or crack ratio is proportional to the contents of Si and Mn, the influence of Si is more effective than that of Mn. In addition, even if the Mn content is high, fracture or crack can be decreased by the selection of optimum welding condition. But in the case of Si, that improvement is almost impossible. The reason why the influence of Si is different from that of Mn is as follows: It's observed by plasma arc welding that amount of oxide and condition of oxide formation on molten pool in the high-Si steel are different from those in the high-Mn steel. Namely, the amount of oxide on molten layer during flash welding is much in the high-Si steel than that in the high-Mn steel. In addition, the fluidity of oxide in the high-Mn steel is better than that in the high-Si steel. Therefore, Si influences the fracture or crack more significantly than Mn.
First the welding current and voltage waveform was analysed by the waveform analyser containing a microcomputer. Then the correlation between waveform data and arc stability was im estigated. When welding current is low, the standard deviation of an average current in a short circuit duration and the average interval between short circuits, etc. correlate-well with arc stability. But when the current is increased, these correlations become poor. From the results of multiple regression analysis about waveform factors and arc stability, four factors which correlate well with arc stability were selected. A general index for arc stability which uses four factors was established. In the welding tests under various conditions, this index was shown to be capable of differentiating between stable and unstable arc.
Liquidus plane and some cross sectional phase diagrams of Cu-Ag-P and Cu-Sn-P ternary alloys at copper rich corner were established using differential thermal analysis and microscopic observations. Primary phase planes of copper solid solution (Cu) and Cu3P phase were divided by a straight line which started from Cu-P system binary eutectic composition. In respect of the primary phase, the addition of tin was similar to the phosphorus addition, because the primary phase changed from Cu phase to Cu3P phase at lower phosphorus. The phosphorus equivalent of tin was determined to be 0.238 in case of mass % expression. The phosphorus equivalent' of silver was only 0.066. The spreadability of Cu-Sn-P filler metals were inferior to that of Cu-Ag-P, which is explained by the thermodynam c calculation that tin oxide is not so easily reduced by phosphorus. The activation energy of erosion of copper base metal was coincided well with that of diffusion of tin in solid copper.
The method for predicting the toughness of multi-pass weld HAZ by simple thermal cycle simulation was investigated on the basis of the distribution of toughness which had been discussed in the previous report. The main results obtained were as follows. (1) It was confirmed that the toughness of single-pass weld HAZ could be predicted by single thermal cycle simulated specimen with the same cooling rate. It was difficult to predict the toughness of multi-pass weld HAZ by single thermal cycle simulated specimens. (2) Double thermal cycle simulation test (Peak temperature: 1350°C-650°C) was effective for the prediction of the toughness of multi-pass weld HAZ. That is, tempering thermal cycle (Peak temperature: 650°C) was essential for the simulation test. (3) Average toughness calculated by using value of the toughness and area fraction of each microstructure distributed in HAZ almost agreed with the measured toughness of multi-pass weld HAZ.
In this study, the effects of applied stress, diffusible hydrogen content, hardnesses of HAZ and weld metal on the crack initiation site were studied utilizing the small-sized TRC test for HT60, HT80, HY130 and 2 1/4Cr-1Mo steels in order to reveal the criterion under which cold cracking in root pass welding occurred alternatively in HAZ or weld metal. The vickers hardness ranges in HAZ and weld metal were 290-390 and 360-410, respectively. The diffusible hydrogen contents range was 2-7 ml/100 g. Consequently, in the diffusible hydrogen content range 2 to 4 ml/ 100 g, the cold cracking in weld metal occurred under the higher hardness of weld metal, the higher applied stress and the higher diffusible hydrogen content in the weld metal hardness range about 315 to 370 and occurred only in the weld metal in more than 370. Then, "critical stress vs. time diagram" which represented the change in the critical stress in weld metal and HAZ vs. time laspe was proposed as the criterion considering the distribution of diffusible hydrogen in weld zone. The tendencies of crack initiation in root pass welding as mentioned above could be well explained by combining the progress of applied stress and "critical stress vs. time diagram".
The effect of preheating temperature, chemical compositions and hardness of weld metal on delayed cracking susceptibility of type 405 stainless clad steel welding was investigated by the slit type cracking test using nine different covered electrodes. The main results were summarized as follows: 1) For the prevention of the weld cracking, preheating was more effective than lowering of weld metal hardness, In the same preheating condition, the weld metal was less susceptible to cracks if its hardness was lower. 2) As preheating temperature rised, the diffusible hydrogen content in weld metal decreased remarkably at the temperature range in which cracks were liable to occur. 3) The Vickers hardness of weld metal (Hv) was satisfactorily given as below by Ferrite index (Fiax) calculated from the chemical compositions. Hv=804-47 Fidx (8<Fidx<13.5) where Fidx=[Cr+Mo+1.5Si+2Al+4Nb+9Ti]-[Ni+30(C+N)+0.5Mn] (wt%) As preheating temperature rised, cracks generated in lower Fidx.
In underwater welding by the use of commercial covered austenitic stainless steel electrode, the mechanical properties are not so good and the hot crack occurs in the weld metal. This study is investigated the mechanical and metallurgical properties with trial covered stainless steel electrode (RMl). The following results are obtained: (1) The compositional analysis of weld metal with RMl rod are the same as Schaeffler's diagram (penetration 40%) using trial covered stainless steel electrode. (2) The quantity of diffusible hydrogen collected in mercury reaches to a constant volume in about 500 hr, it is about 8 ml/100 g Fe in case of RMl rod. The hydrogen embrittlement and the delayed cracking are hard to occur because of the lower quantity of diffusible hydrogen, so the RMl rod is adapted for the underwater welding rod. (3) The weld metal seldom has blowholes, and no cracks are observed to the weld metal in X-ray inspection. (4) The vicker's hardness in heat-affected zone is about Hv250 and that of weld metal is about Hv220 when used RMl rod, it does not have quench hardenability in underwater welding. (5) In notched tensile test, the tensile strength of weld metal is about 660 MPa and that of bond about 645 MPa. (6) In the impact test, the RMl rod is good for notch toughness because of the lower transition temperature.
Fatigue crack propagation tests were carried out on HT80 steel components on which residual stress fields were induced by electron beam welding, and the influence of residual stress on the fatigue crack opening behavior was investigated. An array of strain gauges attached along the section adjacent and parallel to the section including crack plane was used to examine the variation of stress distribution along the gauges attached section during uploading. In case of notched status, the variation of stress distribution both in welded and unwelded components subjected tensile load showed same aspects regardless the distribution of residual stress. However, in case that fatigue crack propagated, the variation of stress distribution depended on the distribution of residual stress. Based on the above results, it is clarified that residual stress directly affects the fatigue crack opening behavior.
The effective stress intensity factor range based on the measurements of crack opening loads gives the same crack propagation behavior for various stress ratios. Many authors have reported that the crack opening ratio U depends only on the stress ratio R. It is observed in experiment, however, that U depends also on stress intensity factor range ΔK. In this study, fatigue crack propagation tests were carried out on a high strength steel with yield strength of 450 MPa class under various stress ratios. And a crack opening model, that contains the both effects of R and ΔK was investigated based on the cyclic elasto-plastic finite element analysis (FEA) and the fracture mechanics approach with the small scale yielding concept. The FEA can simulate well the plasticity-induced closure even under the plane strain condition. Generally, it is difficult to apply the FEA for oxide- and roughness- induced closures that play an important role in the near threshold region. On the other hand, in the small scale yielding model with the assumption of the existence of the residual deformation on the wake of a crack, the crack opening stress intensity factor can be calculated as the value at which the fatigue crack opening displacement becomes zero. The resultant formula of U contains two parameters, one is for the plastic deformation at a crack tip and the other is for the oxide and the roughness in a fatigue crack. This model can explain successfully the experimental results. As a result, the fatigue crack propagation behavior under various stress ratios is described by the equation with five parameters which can be obtained with only one specimen. The predicted crack propagation rates showed good agreement with empirical results. Therefore, the formula derived in this study may be applied well to the evaluation of fatigue crack extension.
The fatigue crack propagation rate and the fatigue threshold were measured on transverse butt welded joints. The center cracked type specimens were made from as-welded joints and heat treated joints. The heat treatment were performed at 630°C and 720°C. The residual stresses were tensile around the middle of specimen and compressive near the edges. The tensile residual stresses were about 200 MPa, 20 MPa and 5 MPa for as-welded and heat treated specimens at 630°C and 720°C, respectively. The fatigue crack propagation properties obtained at various stress ratios on as-welded and heat treated specimens at 630°C were similar to each other and inferior to those obtained at the stress ratio of 0 on the base metal specimens. The fatigue threshold for as-welded specimens was about a quarter of that for base metal specimens. These properties obtained at the stress ratio of 0 on the heat treated specimen at 720°C were similar to those for the base metal specimens. While, the properties obtained avoiding crack closure at the very high stress ratio on the heat treated specimen at 720°C coincided with those for as-welded specimens. The coincidence of these properties means that the critical stress ratio conditions at which crack tips are fully open during all range of loading are made around crack tips by welding tensile residual stresses of 200 and 20 MPa.
The occurrence of a short-arrested brittle crack (Pop-in) in the COD test for welds is usually regarded as a critical event because significance of Pop-in has not yet been made clear. It is, however, considered to be possible to accept the occurrence of Pop-in which doesn't make significant influence on the safety of actual structures. Based on this conception, authors have investigated the dynamic behaviours of Pop-in in order to establish the assessment method for its significance from the viewpoint of brittle fracture and the followings are considered to be the key points on predicting the behaviours of Pop-in occurred in an actual structures from the results of 3-point bending COD test. 1) Influence of applied load drop on the behaviours of Pop-in crack initiated in a small sized specimen as a 3-point bending one. 2) Influence of applied loading type on propagation and arrest of the small crack in Pop-in About the latter, it was shown by numerical analyses performed in the previous report that 3-point bending loading was able to be severer than tensile one for the small crack propagation and arrest under some limited conditions. In this report, more detailed investigations are carried out experimentally and analytically in order to clarify the conditions under which the crack arrested as a Pop-in in a 3-point bending specimen is sure to be arrested even in an infinite body under tensile one. About the former, the existence of Pop-in of which crack was arrested without the influence of applied load drop was already shown together with the verification by dynamic FEM code. The results are discussed again with some additional investigations because they have much importance to show the theoretical background of the assessment method proposed in this report. Finally, assessment procedure for the significance of Pop-in is proposed from the results of these investigations.