Based upon the idea that hydrogen cold cracking in a steel weld occurs only when the local hydrogen concentration at the instant of crack initiation exceeds a critical value which depends on HAZ ductility and local stress, a new cracking parameter PH was proposed in the author's previous report to predict the critical preheating temperature to avoid root cracking in the slit-type oblique-Y (Tekken) cracking test. In the present paper, an improved and generalized formula of cracking parameter is proposed with which the values of critical preheating temperature and critical restraint stress can be predicted for root cracking in any groove shape and eccentrically located weld. The predicted values are shown to, agree satisfactorily with experimental ones. Since the new parameter is based on the stress concentration ratio and the accumulated local hydrogen at the highly stressed zone close to the root of a weld, the proposed prediction procedure seems to be effective also in most of the cold cracking problems in steel welding and will be described in the following reports. Moreover, a new formula is proposed to. predict the critical restraint stress for root cracking in the oblique-Y groove Tekken test. It differs considerably from other formulae already proposed by other researchers on Implant test data. This may suggest that the Implant test data are not necessarily satisfactory for prediction of welding conditions in actual welding fabrication.
This paper deals with the stability of the welding arc in a constant feeding speed system. The stability of the arc depends on the welder characteristic. The behavior of the arc welding system is described by two first-order nonlinear differential equations. The phase-plane analysis is used for the present investigation. The stability of the arc is discussed by studying the solution in the transient state, which, with the lapse of time, ultimately yields the steady solution. From the present analysis, it is to be noted that though the welding system may produce a stable arc when the welder has a drooping characteristic or a constant voltage, the steady arc might become unstable and change into an oscillating arc when the welder has a rising characteristic. Experiments are also carried out by making use of welders which contain transistors and integrated circuits, and the arc phenomena are observed by means of the high speed motion pictures. The validity of the theoretical results is confirmed by the experimental investigations.
A new apparatus for determing the diffusible hydrogen in weld metals was developed, (Vacuum-extraction Method). This method is based on the principle that the diffusible hydrogen evolved from weld metals in a closed evacuated vessel is concentrated, by means of a Toepler pump, into the measuring unit and its gas volume is accurately measured. Hydrogen evolution from JIS. type specimens was calculated by using the solution of one-dimensional diffusion model for the movement of hydrogen in uniform metals. Experimental results of hydrogen evolution by using the Vaccum-extraction method agree approximetely with the calculated values. An apparent diffusion coefficient of hydrogen in the welded mild steel (SS41) at a room temperature of 20°C, D=1.06×10-6 cm2/sec was obtained from an analysis of the evolution curve of hydrogen from the weld metal.
In order to clarify the creep properties of EB welds on Hastelloy X which is one of the candidate alloys for components of VHTR, creep tests on EB weld metal and welded joint were carried out. The results were discussed in comparison with those of base metal and TIG welds. Further, EB welds were evaluated from the standpoint of high temperature structural design. The results obtained are summarized as follows. 1) Both creep rupture strengths of EB weld metal and EB welded joint are almost equal to that of base metal, but those of TIG welds are lower than base metal. As for the secondary creep rate, EB weld metal is higher and TIG weld metal is lower than base metal. As for the time to onset of tertiary creep, no remarkable difference among base metal, EB weld metal and TIG weld metal is observed. 2) In case of EB weld metal, although anisotropy is slightly observed, the ductility is same or more as compared with base metal. In case of TIG weld metal, on the contrary, anisotropy is not observed and the ductility is essentially low. 3) Such rupture morphology of EB weld metal as appears to have resulted from interconnection of voids which occurred at grain boundary is similar to base metal. In case of TIG weld metal, however, many cracks with sharp tips are observed at grain boundary, and the rupture appears to have occurred in brittle by propagation and connection of the cracks. 4) It can be said from the standpoint of high temperature structural design that EB welding is very suitable to welding for structure. where creep effects are significant, because both of the creep ductility and the rupture strength are almost equal to those of base metal.
The local non-wetted Area occurred on the surface of copper specimen dipped in molten eutectic solder bath was measured as functions of the surface contamination of specimen, the immersion time, depth, rate etc. of specimen and the activity of used flux, by using a commercial type meniscograph tester and a profile projector. The results obtained were as follows; 1) The oxide film on specimen greatly influenced the occrrence of non-wetted parts. Before dipping, when the specimen was heated at 150°C within 10 min in air, the ratio of non-wetted area to the soldered area of specimen was below a few percent. 2) The ratio of non-wetted area was depended on the immersion depth of specimen. When the depth equals a equilibrium height of negative meniscus by non-wetting liquid, or is below, the ratio was fairly small. This case was named as I type wetting process. However, when the depth is over the equilirbium height, the ratio increased with increasing of the depth, because the flux painted on specimen was removed by the hydrostatic pressure of molten solder. This case was named as II type wettig process. 3) The equilibrium height of negative meniscus was 3.6 mm for copper plate of infinite width and eutectic solder at 230°C. From this value, the radius of curvature at the height was calculated as 1.8 mm. Moreover, the equilirbium height of negative meniscus for a lead wire of 0.8 mm dia. was measured as 1.6 mm. In this case, the radii of curvature on two liquid surfaces which intersect perpendicularly at the height of negative meniscus were calculated as 1.8 mm and 3.2 mm respectively. 4) For a lead wire of 0.8 mm dia., the immersion depth of 2-2.5 mm standeadized by JIS C 5033 belongs to II type wetting process in our experiment.
Fracture toughness tests of cross-bond notch type specimens with controlled fractions of weld metal in crack front have been carried out. The processes of fracture are observed in detail by using fractographic technique. The processes of fracture of cross-bond type specimens are as follows. Cleavage fracture firstly initiates in the weld metal region, even though its region is small and away from the center of thickness. Then it propagates in the heat affected zone and the base metal region. It seems that, in the case of cross-bond type specimens, stretched zone width can be a proper parameter of fracture toughness, which depends not on stress-strain conditions but on materials. It is necessary to take note of stretched zone width in the lower toughness region, in estimating the macroscopic and average value of fracture toughness of a cross-bond type specimen. It is satisfactory to consider that the critical COD values of cross-bond type specimens used here, even though the weld metal region is away from the center of thickness, are roughly equal to those of specimens of which crack front are completely in weld metal.
This paper describes the correlation between notch toughness and microsturcture of weld metal produced by the large current MIG arc welding. The single bead weldings were made in the inert gas (Ar-60% He and Ar-2-70% CO2) shielding atomosphere by using the titanium and boron bearing consumable wires of 4 mm diameter. The microstructure of weld metal varied significantly depending on the combined effects of titanium, boron and oxygen. When the consumable wire containing 0.11%Ti and 0.0067%B was used, the most appropriate range of oxygen content of weld metal was determined to be 210-300 ppm. In those weld metals, the coarse proeutectoid ferrite was prevented to form, resulting into the very fine acicular ferrite structure and ductile to brittle transition temperatures determined were in the range of -90--106°C. As a result of phase analysis, a large proportion of titanium was evidenced to exist in these weld metals in the form of nonmetallic inclusions which were detected as the acid-insoluble titanium. The majority.of these .inclusions is thought to be oxides. As the oxygen contents of weld metal increased, boron as oxides also increased. Notch toughness of weld metal depended upon the weld heat inputs. In the range of 0.010-0.037% Ti and 0-0.0021%B, ductile to brittle transition temperatures determined on the weld metal were considerably improved with increasing titanium and born contents.
The high frequency current which is superimposed on the AC TIG arc welding current to stabilize the arc gives sometimes rise to radio interferences. However, the perfect means to solve the troubles have not yet been found. A new method to reduce the radio interferences is described in the paper. This technique makes skilful use of the known fact that the restriking voltage on each half-cycle decreases remarkably owing to the high temperature in arc column and electrodes after establishing a welding arc. This method shows one of the possibilities that the radio interferences which are.caused by the AC TIG arc welding process with high frequency current are almost neglected.