As a method to estimate the fatigue life of unnotched ductile specimens under combined bending and torsion, an ellipse quadrant expressed by the following equation was proposed. (σ/σN)2+(τ/τN)2=1 where semi-axes σN and τN are bending and torsional stresses under which failure would occur at N cycles. It was assumed that any combined stress (σ, τ) satisfying the above equation would induce failure at the same N cycles. Fatigue tests of unnotched steel specimens under monofrequency in-phase combined bending and torsion were carried out, and a good correspondence was found between the experimental fatigue lives and those calculated by the ellipse equation.
Structural materials which face a thermonuclear plasma are subjected to thermal shock and cyclic loading under service conditions, since tokamak devices are usually operated as pulsed-type fusion reactors. In order to investigate the thermal fatigue behavior of the structural materials for pulsed-type fusion reactors, a new testing apparatus was developed. The testing apparatus was composed of a closed loop fatigue testing unit, an ultra-high vacuum system and a quick heating system. As the first testing example, thermal fatigue tests were conducted on molybdenum at cyclic temperatures between 500 and 1600°C under complete restraint conditions. Molybdenum materials are to be used as the first wall of the vacuum chamber, fixed limiters and magnetic limiters of JT-60 (JAERI tokama device). The features of the new testing apparatus and the preliminary result of the thermal fatigue behavior of molybdenum materials are described. The thermal fatigue tests were successfully performed as expectedly. It is noted that molybdenum materials showed a remarkable buldging at the center of all the specimens manufactured from three forming methods, namely sintering, electron-beam melting and vacuum-arc melting.
The EMPA quantitative analysis has been considered unapplicable to porous matter. This investigation emphasizes the possibility of chemical analysis for the porous catalysts of two series of silica-alumina and one series of silica-magnesia having micro pores less than 100Å. The density ρ in the mass absorption coefficient (μ/ρ) used in the conventional ZAF correction calculation is that of nonporous matter. In the modified ZAF method proposed here, the apparent density ρa, which is given by ρ(1-ε) with the void fraction ε, can be used for the porous sample analysed, while the value of ρ should be used only for the non-porous reference substance. This modified ZAF method appeared reasonable for three series of samples of porous catalysts with different void fractions.
The effect of plastic strain history on the stress-strain hysteresis curve in the initial stage of cyclic loading was analysed for commercially pure aluminum subjected to various strain controlled push-pull cyclic tests: constant strain cycling with or without a mean strain, and constant strain cycling after various prestraining. Each test was repeated until a stable stress-strain hysteresis loop was obtained. The stress magnitude for each half cycle was plotted against the cumulative plastic strain and compared with the monotonic tension curve. It was found that in the initial stage of cycling the stress-strain curves for different prestrains were not similar nor had such shapes that coincide with each other by a suitable translation. The stress-strain curve for each half cycle was expressed in terms of two quantities: (1) the value of permanent softening, which is the stress difference between the monotonic tension curve and the cyclic curve at the parallel point on the stress-cumulative plastic strain plot and (2) the transient hardening curve. The former depends strongly on the plastic strain range and cycle number but weakly on the prestain, while the latter depends only on the plastic strain occurred in the just preceding half cycle. A new parameter which is related to the monotonic tension curve was introduced to identify the stable behaviour under cyclic loading.
Investigation of the behaviour of plastic deformation in materials is very important in the fields of plastic working and strength evaluation of materials. Metallic materials are generally polycrystalline aggregate, and it is seldom possible to deduce the deformation behaviour of an aggregate directly from the behaviour of single crystals because of the interaction of each grain with the neighboring grains and grain boundaries. In the present paper, the concept of G. I. Taylor that the plastic strain in each grain of a polycrystal is the same as that in a bulk specimen was assumed, and the deformation behaviour in polycrystals was studied in relation to the fundamental equations for X-ray stress measurement The lattice strains of aluminum and low carbon steel measured by X-ray diffraction technique agreed with the analytical ones at a state of uniaxial tension in the initial stage of plastic deformation, indicating that the stress on each grain depends upon its orientation of crystal in plastic deformation. However, it is considered that the effect of anisotropy of work-hardening on each slip system may influence the measured values as the applied strain increases. Examination of the unloaded conditions after plastic deformation made it clear that the stress was unloaded uniformly in each grain. Therefore, it is considered that the uniform local stress model can be applied to both the loading and unloading processes in the case elastic deformation. Since the present experimental results can be explained by accepting this analysis, it is concluded that the behaviour of these materials is approximated by Taylor's model in the initial stage of plastic deformation.
On the assumption that craze behavior under static tension can be expressed by a simple rate theory, the deformation curves by crazing were discussed for creep and constant strain-rate tensions in poly (methyl methacrylate) wetted by kerosene as a crazing agent, by applying the Johnston-Gilman theory of dislocations. An increase in theoretical creep strain εc by crazing was approximated by the following equation used commonly. εc=ε0exp(ασ)tn where σ is the applied stress, t is the time, and ε0, α and n are constants, respectively. This may provide a physical foundation for the widely-used equation derived from the rate theory. The yield points by crazing, tested at two strain rates, agreed well with the theoretical calculations. This agreement between the theory and the experimental results indicates the applicability of the present method to the deformation kinetics by crazing.
The yield behavior of polycarbonate and polyvinyl chloride was investigated by several static load tests: tension test and torsion tests of hollow cylinders having various ratios of the inside to the outside diameter. Particular attention was given to the effect of the macroscopic stress distribution in the specimen on the yield behavior. Close observation in tension test and torsion tests of thin-walled cylinders revealed that there are two processes in yield of plastics: the initiation process of yield accompanying load drop and the progressive process of yield under constant load. The observed yield behavior suggests that the relation between stress and strain of plastics can be represented by two characteristic curves: the first goes up to the point where yield starts, and the second goes up to the failure after the progressive process of yield appears. From the yield behavior mentioned above, two kinds of yield stresses were found to exist: the stress at the yield point and the stress at the progressive process of yield. By applying the above interpretation to the torsion tests of hollow cylinders, it is possible to determine the relation between the torsional moment and shear strain quantitatively. The calculated results were confirmed by the experiments.
The yield curves of two types of copper powder preforms with different loading paths, i. e., the one compacted hydrostatically and the other compacted in a closed die, were determined by uniaxial, triaxial and hydrostatic compression experiments and were drawn in Rendulic stress plane. For the both type preforms (hydrostatic and closed die preforms), the yield curves were similar to elliptic curves or lemniscates. The yield curve for the hydrostatic preform had a major axis in σm-axis (the axis of equal triaxial stresses σ1=σ2=σ3) direction, while the Curve for the closed die preform had a major axis in the pre-loading direction. For the same initial density of the both type preforms, their yield curves approximately coincided to each other if their major axes were superimposed by rotating about the origin in Rendulic plane. Therefore, the copper powder compacts (preforms) were found to obey to the kinematic hardening rule. With the increase in initial density of the hydrostatic or closed die preform, the yield curve in Rendulic plane expanded without rotation of the respective major axis; the yield curves of the both type preforms with different initial densities were approximately similar, having their center at the origin in Rendulic plane.
The transition behaviours of fracture have been examined on lamellar pearlite steel containing proeutectoid cementite. The transition temperature was determined from the tensile properties-temperature curve. Micro-cracks of the fractured specimen were measured on the longitudinal section, and then the relation between the initiation and propagation of micro-cracks and testing temperature was discussed. In the temperature range below the ductile transition temperature TD, three types of micro-cracks were observed (types a, b and c). Type a was the micro-crack which was initiated by cracking of proeutectoid cementite and propagated by subsequent separation of pearlite. Type b was the micro-crack grown from the cracking initiated at the pearlite colony boundary. Type c was the micro-crack initiated by the cracking of lamellar cementite which was connected to proeutectoid cementite. Above TD, dotted micro-cracks in lamellar cementite (type d) and transgranular cracks of pearlite colony (type e) were observed in addition to micro-cracks of types a, b and c.
In order to clarify the reason of low notch sensitivity in the strength of cast iron, tensile tests and stress analyses were performed on the circumferentially notched bars of flake and nodular cast iron. A fracture criterion was induced from the non-elastic stress distributions obtained by FEM. The experiments showed that the rupture load of flake cast iron was varied only by ±5% at room temperature by the notches with the elastic stress concentration factor up to 2.6. Accordingly the flake cast iron showed very low notch sensitivity. The stress analyses revealed that such low sensitivity in flake cast iron was caused by the release of stress concentration due to the non-elastic behaviour and the existence of the stress boundary layer δ of 2-3mm in depth which was proposed by the author for edge-notched plates in the previous paper. At liquid nitrogen temperature the iron showed hight notch sensitivity because of its highly elastic behaviour and the decrease in δ. For the nodular cast iron, the rupture load of notched bars was increased by 15-25%. This increase in strength was attributed to the constraint of deformation and fracture at the notch root. According to the metallography, in the case of blunt notch, fracture initiated at the center of notch section and grew outwards in ductile manner. In the case of sharp notch, however, fracture initiated at the notch root, and the criterion with δ was applicable.
The purpose of this investigation was to examine the S-N diagram and fatigue process of mortar specimens under repeated impact bending load. The results obtained were summarized as follows. (1) The S-N diagram of mortar specimens under repeated impact bending load consists of two straight lines. When the ratio of impact level S is larger than that of singular impact level S0, impact fracture behavior is the main behavior on the S-N diagram. When S is smaller than S0, fatigue fracture behavior is the main. Here, S0 is about 55 percent of impact level. (2) As the number of impact cycles increases, the impact flexural strain increases. Therefore, the fatigue process under repeated impact bending load can be estimated by the measurement of impact flexural strain. (3) The flexural strength of mortar specimens unbroken by repeated impact bending load of four hundred cycles is larger than the unloading flexural strength in spite of large residual strain.
In order to clarify high-temperature, high-cycle fatigue properties of Cr-Mo alloy steel plates used for pressure vessels, rotating bending fatigue tests were carried out on three kinds of steels, JIS SCMV 2 (1 Cr-0.5 Mo), SCMV 3 (1.25 Cr-0.5 Mo) and SCMV 4 (2.25 Cr-1 Mo), at room temperature, 400, 500 and 600°C with a frequency of 7500 cycles/min, and the fatigue strength at 108 cycles was obtained. The conclusions obtained were as follows: (1) No specimen failed in the vicinity of 108 cycles and the endurance limit was found at room temperature. The number of cycles to failure increased up to 108 cycles with decreasing stress amplitude at elevated temperatures. (2) For the steel of SCMV 2, the fatigue strength at 400°C was larger than that at room temperature, but it decreased with increasing test temperature above 400°C. For the steels of SCMV 3 and 4, the fatigue strength decreased with increasing temperature. (3) For the steels of SCMV 2 and 3, the ratio of fatigue strength at 108 cycles to 0.2% proof stress was larger than unity at 400 and 500°C. (4) The coaxing effect was observed a little at room temperature but not at elevated temperatures. (5) The above phenomena observed at elevated temperatures can be explained qualitatively by the strengthening mechanism due to chemical interactions between interstitial atoms (C or N) and substitutional atoms (Cr or Mo).
The polarization behavior of a welded joint of SUS 304 stainless steel was investigated in boiling 42% MgCl2 solution, and the effects of the variations of the residual stress and metallurgical factors coused by welding and machining after welding on the polarization behavior were discussed. Furthermore, the polarization behavior was studied under the applied stress of 25kgf/mm2, and the relation between this behavior and the susceptibility to stress corrosion cracking was discussed. The results obtained were as follows: In the polarization behavior of the weld metal zone, a wide unsteady passive region existed. However, the unsteady passive region was narrow for the base metal zone and was not existent for the heat affected zone. It was clarified that this difference in polarization behavior was due to the differences of alloyed Cr and Ni concentration, metallurgical structure and residual stress values. In the case of unloaded situations, it was deduced that the heat affected zone was most sensitive to stress corrosion cracking, while the weld metal zone was immune to stress corrosion cracking. In the case of loaded situations, all the polarization behaviors changed to more sensitive behaviors to stress corrosion cracking, It was to more sensitive behaviors to stress corrosion cracking. It was concluded from these behaviors that the base metal zone was most sensitive to stress corrosion cracking.