The most widely held view on the flux action with soldering and brazing is presented by three principal functions which are described below. 1. To dissolve or releasen any oxides present on the base metal and molten filler alloy surfaces which are to be brazed. 2. To prevent the oxidation of these surfaces during the brazing process. 3. To lower the surface tension of the molten filler alloy, namely, the molten flux should be readily displaced by the molten filler alloy, thus promoting the wetting of the base metal by the molten filler alloy. The last of the above mentioned functions belongs to the physical meaning of the flux action and, accordingly the investigation of the viscosity of flux saturated with metal oxides is one of the most important for gaining a thorough knowledge of the brazing phenomena. In this paper, the relation between the viscosity of the molten flux and the interfacial tension between the filler alloy and the molten flux can not be reported but the paper describes the variation of viscosity with various fluxes which were heated to either the lower or the higher brazing temperature in a porcelain or graphite crusible. The results obtained from the investigation are as follows : 1. By small additions of powder ammonium chloride or granular stannous chloride, or both of them to the molten zinc chloride, the viscosity can considerably be decreased. In the binary molten flux system, the minimum value (about 1.5 poise at 450°C) of the viscosity was recognized when using the 75% ZnCl2-25% NH4Cl, the 70% ZnCl2-30%SnC12.2 H2O On the other hand, the maximum value (about 2 poise at 300°C) was found when using the 50% ZnCl2-50% SnCl2⋅2H2O. 2. The viscosity of either the eutectic composition or compound in the ZnCl2-NH4Cl system seems to indicate the minimum value, as well as the fluidity of the molten metal and alloy in cast processing ; for instance, if the relation between fluidity and viscosity is shown reversibly under physical considerations, the antimony-lead alloy or the antimony-cadmium alloyhas the same phenomena. 3. Boric acid (about 80 poise at 900°C) has a higher viscosity than borax (about 10 poise at 900°C) and the viscosity decreases as the borax in the mixtures is increased. But it should be noted that the transition point (at about 780°C) was observed when producing the phase change which would be probably considered as the chemical reaction : 2H3BO3=B2O3+3H2O in mixed system. Further research is needed to understand this "transition point."
It is unavoidable even at present the gas blemishes on cut sides remain in cutting process by the automatical flame-cutting machine. The object of this experiment was firstly in determining the fatigue strength of flame-cut specimens containing an intentional gas blemish (notch) (t : 1.8 mm. ρ : 0.7 mm), and moreover, in comparing the fatigue strength thus obtained with that of machined specimens; and secondly in examining several methods for the repair of notch. The plate used was SM 50 A. The experiment revealed the following ; (1) The fatigue strength of the flame-cut specimens containing an intentional notch (stress concentration factor 3) was 72 percent of that of the evenly flame-cut specimens. But the abovementioned value was greater than that of the machined specimens containing an identical, but machined notoch. (2) The location of crack-initiation was limited to the melted edges in the evenly flame-cut specimens. But the crack initiated from the whole surface of the notch root in the notched specimens (MN and GN testing series). (3) All the repair methods tried in this research succeeded in giving the same fatigue strength as that of non-notched specimens.
High temperature tensile and creep rupture properties of six low alloy high strength steels were investigated for base metals and welded joints which were made with covered electrodes. Test steels were all 30 mm thick, four HT 60 (three normalized and one quenched & tempered) and two HT 70 quenched & tempered steels which were produced experimentally for nuclear pressure vessels. The results are that: (1) The high temperature tensile and rupture strengths of welded joints were generally inferior to those of base metals. (2) The tensile strength of both base metals and welded joints began to decrease at 300°C for quenched & tempered HT 60 and HT 70 steels, and at 350°C for normalized HT 60 steels. (3) The rupture of tensile specimens of welded joints was generally located in base metal at lower temperatures, and in weld metal at higher temperatures. (4) Generally, higher creep rupture strengths were observed for steels with higher tensile strengths at high temperatures. (5) As far as the present test steels are concerned, normalized steels seem to have better high temperature strength than quenched & tempered steels.
It is a well known fact that ferritic stainless steels are more useful than austenitic stainless steels as the chemical plant materials used in sulphuric atmosphere. We have made high temperatuse creep rupture tests on the mother plates, deposited metals and their welded joints of type 446 (25 Cr stainless steel) and Sicromal-12 (25Cr+ 1.5Si+1.6 Al stainless steel) that is known so far as the highest class material for ferritic steels. The result can be summarised as follows : (1)High temperature creep rupture strengths on the mother plate and deposited metal of type 446 are more or less superior than Sicromal-12. (2)Creep rupture strengths of the welded joints show no difference in comparison with the respective mother plate. (3)We consider that a-phase-that is specific characteristic on these materials-is an important factor to have an effect on the creep rupture strength, especially as Sicromal-12 remarkably forms more a-phase than type 446.
A method of estimating the crack-sensitivity of single welds using the testing plate of tapered width with slit gap was proposed. It is important in estimating the crack-sensitivity to determine the variation of that property depending on the severity of the testing conditions. A series of crack tests under several degrees of mechanical severities can be made by this method using only one testing plate. The method proposed shows many advantages over the traditional ones: Better reproducibility, less expensiveness in the preparation, capability of re-use. etc. Some testing results using this plate are cited.