About 40 commercial mild steel electrodes of different types (all except three passing the mechanical test specified by AWS or JIS) were used for welding slit-type crack test specimens prepared of several grades of rimmed steels and the crack development in them has been observed with the folio wing results: 1) Apparently there is no relationship between the mechanical properties of all deposit metals and crack tendency. 2) In the room-tempreture welding of steel plates with ordinary composition, with some exceptions there is a very low incidence of crack development whatever electrode may be used. 3) Considerable variance in crack tendency is observed depending upon the grade of electrode, with high-S (for instance, 0.046%) steel plates to be welded at room temperature. 4) Crack tendency of steel plates with usual components when they are welded at low temperature, say 0°C, follows the same pattern as. that of high-S steel but is not so marked as the latter. 5) Deoxidation of. deposit metal seems to have something to do with crack tendency. Effect of deoxidation is great when high-S steels are welded or welding is done at low temperature. Every type of electrode with high Si/Mn deposit metal, excepting low hydrogen type, possesses high crack tendency. Low hydrogen type is most stable to cracks, followed by the ilmenite type, containing a conside-rable amount of Mn and low Si/Mn, with fairly good stability.
Welded tear test piece of Kahn type was proposed to investigate the effect of manual arc welding on the notch sensitivity of structural mild steel plates, 11 charges of rimmed and killed ones produced by four representative steel plants in Japan. Snigle bead was deposited on one side of plate surfaces, and then, machined flush to the surface. The saw cut was located alongside the deposited bead to evaluate the notch sensitivity of heat-affected zones of the base material. For comparison, tests were also carried out with standard Kahn tear test pieces of un-welded plates. Table 5 and Fig. 14 show the transition temperature of steel plates thus determined before and after manual arc welding. Fig. 14 also demonstrates the relations between Mn/C ratio and transition temperatures, oblique zones in the figure showing the ones of American semi-killed and killed steels on Kahn's data. Welded plates generally showed higher transition temperatures than unwelded plates, ranging from 5 to 11°C in 13 mm thick and 4 to 13°C in 23 mm thick ones of rimmed steels, from 1 to 13°C in killed steels, 23mm thick. But these adverse effects. were comparatively slight, so it could be concluded that the notch sensitivity of mild steel plates was but slightly increased as the result of manual arc welding so far as fracture transition was concerned. Moreover, the amount of rise in transition temperature after welding was not related at all to the notch sensitivity of unwelded plate studied in this report.
By using the mother plates containing Radioactive Isotope S the-authors investigated the phenomenon of diffusion of sulphur from the mother plates to the weld-metals under multi-pass electric arc-welding. As the result of this experiment it was revealed that ; 1) The influence of sulphur of the parent plate extends clearly to the fifth pass of weld metals in the multi-pass welding. 2) Although the absolute mass of sulphur absorbed by slag from the weld metal increases with sulphur content of parent-metal, the Desulphurization Ratio" decreases according to the increases of sulphur content of parent plates or weld metal on which the next welding pass may be run in multi-pass welding. 3) It was found by counting test that there is a slightly more sulphur in the heat-affected zone under the weld metal than the other parts of parent-plate, and, it is guessed that this phenomenon will be caused by the thermal diffusion of sulphur which existed, at first, in the outside of the heataffected zone.