With the aid of high speed motion pictures taken by the American investigators, it has now become clear that in inert gas shielded metal arc welding, the molten metal of electrode tip transfers to the base metal in a fine spray form. The same phenomenon is very likely to occur in the submerged arc welding which utilizes heavy current as well. This is to supplement to our earlier works and to show what is happening in manual open arc welding at high current densities. Fig. 1 to 4 and Fig. 6 to 8 show the case of 4mm diameter bare steel electrode at 250A and 150 or 140A DCSP respectively, the speed of pictures being 720 exposures per second. At normal welding current i. e. 140 or 150A, molten metal accumulates at the electrode tip when the arc length is held long enough to limit the mechanical contact between electrode and base metal, and it finally drops off the tip due to gravity. At higher current densities, the pinch effect-hydrostatic squeezing action generated in a liquid conductor due to electromagnetic force-become more predominant than the restraining force due to surface tension, and beyond a certain critical current, electrode metal is torn off quickly and is transfered to the base metal in fine sprays. In one of our pictures such pinching off took place 21 times in 0.4 second at 250A DCSP. These photos were taken in prewar days, and many other phenomena that were found in ordinary welding conditions were already reported in the Journal of Welding Society in Japan.
Numerous reports are available on the study of the characteristics of MIG arc. But they are some time not systematic, sometimes conflict with one another in certain points. To attain as rational selection of welding conditions as possible, the authors tried to collect the most systematic data on the state of burn-off with aluminum alloy filler wires of varied diameters. In the present instalment, a welding source having a specified current-voltage characteristic curve was adopted to exclude the necessity of considering the influence of the characteristic of the source. The tested filler wires were the following six kinds : 43 S, 0.8mmφ, 3/64"φ. 1/16" φ, and 3/32"φ ; 2 S 1/16"φ ; and NP 5/6 1/16"φ. The results are summarized with some comments : 1) Under less than 18V, 43 S and 2 S wires had the metal globules causing a short-circuit between electrodes ; under 18-19 V, such a short-circuit sometimes happened, sometimes not ; under more then 20V, it scarcely did. (The corresponding voltages for NP 5/6 are respectively 19, 19-20, and 21V.) A sharp rise of the burn-off rate of the filler wire with the drop of arc voltage might be related with this fact. 2) In the range 0.8mm-3/32"φ, the volume burn-off rate of 43 S wires is practically independent of diameter change. With same diameter (1/16"φ), 43 S and 2 S have same value of the rate ; NP 5/6 has a somewhat larger value. 3) There is a linear relation between the minimum current ensuring the favourable transfer of globu-les and the sectional area of filler wire, so far as same material (43 S) is concerned. With same diameter of wire, NP 5/6 gives a larger value than 43 S, and 2 S is rather small. 4) The size of globules under the current which is sufficienty larger than the minimum is generally constant and, for the measured range, approximately proportional to the sectiodal area of filler wire.
In this paper, thermal conduction of spot welding is calculated as one dimension phenomina. There are four kinds of resistances which generate heat on spot welding. These are 1 contact resistance between two sheets, 2 specific resistance of sheets, 3 contact resistance between electrode tip and sheet, 4 specific resistance of electrode tip. Therefore, we must consider four sources of heat, but electrode has so large heat capacity that we can suppose contact plane of electrode and sheet is always at 0°C during welding. So the author solved differential equations of following four cases. 1 one substance (sheet only), neglecting a (thermal coefficient of specific resistance), 2 one substance, considering α 3 two substances (electrode and sheet), neglecting a 4 two substances, considering a but special case. And by putting numercal values in each equation, temperature-time curves as follows are obtained.
Impact tests were conducted at various temperatures on many impact specimens with V notch taken from one-pass weld metal obtained by one-pass welding using the various kinds of one-pass fillet welding elecctrodes. Tests results reported in this paper indicate that transition temperatures of one-pass weld specimens are clase to room temperature and higher than that of base metal.
The C. S. Arc Welding Process does not produce so much quantity of slag as the submerged arc welding process or manual arc welding process with coated electrode, because this process needs no flux. But since the filler wire for this process contains much quantities of deoxidizers, the quantity of deoxi-dation-product is comparatively large. And the product, rised up from the molten weld metal, remains partly on the surface of bead. In case of the multi-pass-welding of thick plates, it is desirable to remove completely this small amount of slag. This slag can be removed by a chipping-hammer, and it is not easy to remove the slag by wire-brushing only. If it will be able to controll the composition of deoxidation-product, the covering and the removal-facility of slag may be improved. In this process, it is unable to controll the properties of weld metal by flux as in cases of the submerged arc welding process and manual arc welding process. Authors intended to controll the weld metal and slag obtained by the C. S. Arc Welding Process by means of adjustment of supplied gas. And they researched experimentally on the effects of oxygen mixed in the supplied gas. According to the results of experiments, they recognized the following facts ; (a) In case of using filler wires having high silicon content, with the increasing of oxygen content of the supplied gas the deoxidation-product changes from silica or silicate saturated with SiO2 to unsaturated molten solution, FeO-MnO-SiO2. (b) The addition of a suitable smount of oxygen to the supplied gas gives good effects on the stability of arc deposition, the facility of removal of thin slag and the improvement of appearance of bead. (c) It is also able to controll the mechanical properties of weld steels by the addition.
The rapid heating maximum temperature of the heat-affected zone due to welding decreases with the increase of the distance from the fusion line. But the cooling times, from A3 transformation point to 500°C, in the cooling process from those various maximum temperatures at the heat-affected zone do not differ so much. However, the cooling time changes extensively by thickness and width of a plate, welding current, proceeding rate of bead and other welding conditions. Structures and hardnesses at the neighbourhood of the fusion line in the heat-affected zone of a welded joint are similar to those of the specimen, cooled with the same cooling time as those of that part of affected zone, which can be known by the continuous cooling transformation diagram in case of the rapid heating maximum temperature 1300°C. Accordingly, if we obtain such continuous cooling transformation diagram, the maximum hardness and the corresponding 'structure which appear in the heat-affected zone can be supposed by the diagram. The absorbed energy by impact bend and bend angle of the specimen, used to obtain the continuous cooling transformation diagram in case of the rapid heating maximum temperature 1300°C, show extremely small values when the cooling time is shorter than the critical cooling time Cf. Therefore, it is considered that if the welding is performed so as to cool the neighbourhood of the fusion line in the heat-affected zone with the longer time than the Cf', the resistances of the obtained joint for shock and deformation are large and the joint will be secure. Further, it is desirable to make steels which have necessary mechanical properties and show the critical cooling time C1' being as small as possible.
It is well known that the strength of the brazed joint depends on the methods of the test, the imension of the test piece and so on. But the interpretation of these phenomena has never been clear. The author get a qualitative formula showing the relationship between the strength and the dimensions of the test piece in the brazed joint by applying the results of Bridgman's analysis. By using this formula the discrepancy between many investigator's results is explained clearly.
In the previous paper the author reported that the toughness of high tensile steel through the standard V-notch Charpy test is recovered by the tensile prestrains at room temperature which are at the slight necking point (the 1st recovery point) and the maximum load-point (the 2nd recovery point). In one part of this study the 1st recovery of toughness is interpreted from the stress-strain relation of a torsional test. And another part gives the examples of the recovery of the toughness by prestrain of the practical cold-worked steels though it is difficult to judge whether it is the 1st or the 2nd recovery. The results obtained are summerized as follows: 1. The 1st recovery at the slight necking point is correlated to the increasing steepness of the transition curve. 2. The slight necking point almost coincides with the decreasing point of the rate of strain hardening. 3. The increasing steepness is caused by the decreasing of the rate of strain hardening. Therefore the 1st recovery can be interpreted by the decreasing of the rate of strain hardening. 4. The cold-bent semi-killed steel in pipe type and the cold-drawn killed steel, also, show the recovery of the toughness by prestrain through the notch impact test.