It is necessary to determine experimentally the standard welding parameters before beginning the welding. However, the social loss of such experiments, which may be repeated in each work and in each case, is not small. If the relationship among the welding parameters may be computerized and the standard welding parameters may be available by computer, the material, labor and time losses will be saved. This study is to make an algorithm for computerizing welding parameters. At first step the algorithm for submerged arc welding proposed by Richter et al. was modified and experimentally studied. In conclusion, the algorithm for submerged arc I-butt one pass welding is proposed under the condition that a calculated penetration depth (PC) is from 0 to 1 mm less than the plate thickness.
When structures are constructed by welding, they contain, consequently, residual stress and deformation. These may cause weld cracks and influence performance of the structures. In order to avoid weld cracks, it is necessary to have information about mechanical behavior of welded joints by welding especially, thermal stress. When connection is simple such as of one dimensional members or of members under one dimensional constraint state, the magnitude of residual stress can be fairl yaccurately estimated by the conventional concept of restraint intensity. However, for shorter slit weld such as that in the y-slit type restraint weld cracking test specimen, experimental facts show that weld crack is more observed in the middle portion of a slit where restraint intensity is lower than in the vicinity of the slit ends. This fact can not be explained in terms of the conventional restraint intensity and welding residual stress is not necessary to follow directly the restraint intensity. In this paper, an investigation is carried out into the mechanism of production of thermal stress and deformation, and applicability of the concept of restraint intensity for slit welds, based on the thermal elastic-plastic analysis by the finite element method. Based on the results obtained in this investigation, the following is important and useful information. (1) It may be concluded that there are three kinds of deformation which cause thermal stress; (a) the shrinkage of the weld metal, (b) locally confined thermal deformation of the base plate, and (c) over-all deformation of the base plate. In case of short slit, the over-all deformation is the main cause of thermal stress. In case of long slit, this does not play an important role and then the conventional concept of restraint intensity is applicable to estimate the residual stress. (2) To seek appropriate welding condition for prevention of weld cracks in slit weld, the test result by the Tekken type specimen may provide conservative one, since the specimen produces the severest condition among those specimen which have the same restraint intensity.
Many researches have been conducted on the influence of the surface finish of the weld toes on fatigue strength but little is known about fatigue strength of welded joints subjected to high mean stresses encountered in actual service of machines. In this paper, the effects of finish of the weld toes on bending fatigue strength for minimum stress Smin=0 and for high mean stresses in the neighbourhood of yield point Sy are investigated. Four kinds of finish, as welded, ground, peened and ground+peened were given to the weld toes of SM41 and HT60 respectively. The results are summarized as follows. (1) The fatigue strength (Stress amplitude) Sa of as welded for mean stress Sm at 5×106 cycles can be written as follow. SM41; Sa=10.3-0.125Sm HT60; Sa=14.3-0.158Sm. (2) The fatigue strength increases approximately in proportion to the reduction of mean stress concentration factor Kt by grinding for Smin=0 and for high mean stresses, if the finish is fine. However, for poor finish with remaining defects its effect is little in high cycle region with Smin=0. (3) The effect of peening is caused mainly by compressive residual stress induced by peening and little by work hardening in the case of both HT60 and SM41. Therefore significant increases in fatigue strength can be obtained by peening for Smin=0, but its effects vanishes in low cycle region with Smin=0 as well as entire region with high mean stresses because of the release of compressive residual stress induced by peening. (4) The fatigue strength of ground+peened can be estimated by superposing the effects of ground and peened.
Oxy-arc cutting of mild steel by stick hollow electrode is being conducted as a established in practical application, because it is fast and easy to handle. But the chief drawback to the use of conventional oxy-arc cutting is electric shock hazard and high degree of skill is required to keep steady arc in unstable cutting condition and poor visibility. To solve the problems, the simple automatic oxy-arc cutting method that is called Gravity cutting method was developed in our programe. Experiment was carried out at down hand position using the said hollow electrode and the same power source as used for underwater Gravity welding and the following conclusions were obtained. (1) Once set up, the actual cutting is made automaticaly without diver and welder. (2) The equipment is simple, cheap at price and easy to operation. (3) Actual cutting speed was high, and therefor in case of plate thickness in 100 mm and 40 mm, cutting speed was 25 cm/min and 60 cm/min respectively.
According to the studies on the ductile strength of cruciform welded joints, it is known that the fracture strength of the joints with comparatively small amount of gaps, does not decrease in comparison with that of the joints with no gaps, because of the increase of penetration, which is roughly proportional to the increase of amount of gap. However, such joints are frequently subjected to fatigue failure, therefore the effect of root gaps on low cycle fatigue strength of cruciform joints is investigated. The main conclusions obtained are as follows. (1) At the roots of the joints, initial microcracks are observed occasionally. The average length and the occurrence percentage of these cracks become minimum when amount of initial gaps before welding is about 2 mm. (2) As for the fillet welded joints used in this study, the penetrations are approximately equal to the amount of initial gaps before welding. (3) It is shown that the crack initiation lives can be neglected, therefore, the fatigue lives of the joints are almost regarded as the crack propagation lives. (4) It is confirmed that the well known relationship between the crack propagation rate, dl/dN and the stress intensity factor range, Δk; dl/dN=c(Δk)m are applied well to the fatigue cracks advanced toward the right angle to the loading direction. (5) The fatigue strength of the joints with actual gaps below 3 mm after welding does not decrease in comparison with that of the joints with no gaps.
An analysis on spot welds using capacitor discharge resistance welding machine is made by employing the method ofstatistical analysis. Studies are made on the partial correlation among a tensile shear strength, a nugget dimension and a contact resistance. Materials used are mild steel and aluminium alloy. The thicknesses of sheet selected are 0.6 mm and 0.8 min, respectively. Fifty specimens of the spot welds are prepared under the same conditions of welding. The number of specimens is determined from the results of the statistical analysis. It becomes evident from this analysis that there is little partial correlation between the contact resistance and the tensile shear strength or the nugget dimension within 5% level of significance under the suitable welding conditions. That is, the strength of spot welds in the test of tensile shear type and the nugget dimension are independent of the contact resistance under the suitable welding condition.