The ultrasonic wave reflection method has been used to determine the elastic constants, Young's modulus, shear modulus, and Poisson's ratio of various steels and aluminium alloys in the temperature range from room temperature to 800°C. It was found that the both elastic moduli tend to decrease with the rise of temperature. Poisson's ratio, however, shows no significant difference among the materials with temperature and there is a slight increase with rise of temperature. The data were successfully represented by the empirical formulae in the form of exponential function of temperature. They are very convenient expressions for the analyses of thermal stress problems.
Rotating bending fatigue tests were carried out on a martensitic stainless steel (Type 403) at room temperature, 200, 300, 400, 450 and 500°C with a frequency of 7500 r. p. m. and the fatigue strength at 108 cycles was obtained by the staircase method. The results obtained are summarized as follows: (1) The endurance limits were found, and the fatigue strength at 107 and 108 cycles were identical at 200 and 300°C and were nearly equal at room temperature and 400°C. At 450 and 500°C, the number of cycles to failure continued to increase with decreasing stress amplitude up to 108 cycles. (2) The ratio of fatigue strength at 108 cycles to 0.2% proof stress was 0.64∼0.69 at R. T.∼500°C. (3) The coaxing effect was slightly observed in the temperature range from room temperature to 400°C. (4) The room temperature hardness at the surface of the specimens endured 108 cycles at high temperatures from 200 to 400°C was larger than that of a virgin specimen. (5) The strengthening of the material is considered to be caused by dynamic aging during the fatigue test at 200∼400°C. This phenomenon can be pointed out as a reason why endurance limits were found and the coaxing effect was observed in this temperature range.
Long-time stress relaxation tests up to one year have been carried out at 500°, 550°, and 600°C on 12Cr-Mo-W-V steel and at 600°, 650°, and 700°C on 19-9 DL steel for high temperature bolts of a steam turbine for total strains ranging from 0.10% to 0.25%. The stress relaxation curves of both steels showed an inflection of convex nature in a certain time range on the longer time and higher temperature side. The inflection of the curves for 12Cr-Mo-W-V steel at 550° and 600°C might be due to the decrease of nitrogen content in solid solution, which was mainly caused by the formation of aluminium nitride. For 19-9 DL steel the inflection was considered to be caused by the decrease of the alloying elements in solid solution, such as tungsten, molybdenum and chromium, together with the coarsening of carbide precipitates. The comparison of the present relaxation strength with the published data was made. The influence of the total strain on the residual stress and the correlation of the observed stress relaxation curves with the estimated ones obtained from the creep data applying the creep constitutive equations were discussed.
The changes in microscopic structures and mechanical properties occurring during the repetition of heating and cooling were investigated on cast iron. The repeating temperature ranges were as follows; (1) from 800°C to 600°C, (2) from 650°C to 350°C, (3) from 350°C to 150°C and (4) from 800°C, 650°C, 600°C, 350°C and 150°C to the room temperature, respectively. In the case (1), the tensile strength and the brinell hardness decreased gradually with increasing number of repetition, and showed about 40% reduction from the initial values after 15 times of repetition. These values, however, did not changed on further repetition. The similar results were also obtained in the repeated annealing (800°C). In these studies, the dissolution of pearlite and the increase in the percentages of graphite and ferrite areas were detected. In the case (2) and in the repeated annealing (650°C), the mechanical properties remained unchanged, even though the repetition was continued up to 15 times. After 20 times of repetition, however, 20% reduction was detected in the tensile strength and brinell hardness. The repeating tests in the ranges of temperatures below 350°C showed no effect on the mechanical properties. It was also revealed that the mass effect decreased with increasing number of repetition like the behaviors of strength or hardness. These results seemed to suggest that the changes in mechanical properties and mass effect were closely associated with the repetition of heating over the eutectoid temperature of this material (700°C).
The creep crack growth behavior in a bar with a circumferential hyperbolic crack was analysed by means of a method proposed in previous papers. This method combines the stress and strain analysis with a generalized creep damage hypothesis applied to the point located ahead of the current crack by a characteristic length ρs. The effect of several parameters on the creep crack growth behavior was discussed. In parallel with this analysis creep rupture tests of deeply precracked round bars, which simulate the bars with a hyperbolic crack, were carried out on S15CK low carbon steel at 450°C. The results obtained are summarized as follows: (1) The analytical predictions as to the creep rupture lives of deeply cracked bars agreed well with the experimental data. (2) It was reconfirmed by the experimental data that notch-weakening prevails under torsional creep in the absence of metallurgical strengthening mechanisms. (3) It was found that rupture life of a precracked specimen can be successfully correlated with the initial value of modified J-integral, which is a measure of stress and strain fields in the vicinity of a precrack under the steady-state creep conditions.
Steam turbine blade steel of 12 Cr-Mo-W-V type has been rupture-tested at 500°-650°C for periods up to 40000hr. The data obtained showed significant differences in the long time properties among various production heats. Metallurgical examination after exposure at 600°C showed that the microstructures and solute contents of alloying elements, such as tungsten and molybdenum, are similar for all the heats and have no correlation with their rupture strength. Low aluminum content heats (≤0.007%) showed high strength in the long time tests. It appears that the high aluminum content (≥0.02%) reduces the nitrogen content in solution due to precipitation of AlN during prolonged rupture testing and leads to deterioration in rupture strength. The rupture ductility variation is suggested to arise from the differences in pre-existed austenite grain, aluminum content, and copper content. A heat with fine grains of pre-existed austenite (G: 5.0), low aluminum content (0.005%), and low copper content (0.03%) showed high ductility over a wide testing range.
Creep and stress rupture tests on three heats of Hastelloy X differing in the manufacturing process were carried out at 800°C, 900°C and 1000°C. Interpretation of the observed creep properties was made, and a method for predicting necessary design data from the experimentally obtained results was discussed. The results are as follows. (1) It was difficult to separate the primary, secondary and tertiary creep stages in the creep curve of Hastelloy X of the present tests. However, those were made distinguishable by plotting the results in a double-logarithmic coordinates. From these creep rate curves, the primary and secondary creep rates and the times to the initiation of secondary and tertiary creeps were derived. (2) It is considered that the same stress and temperature dependences between the primary and secondary creep rates exist in the creep behaviour of Hastelloy X of the present tests. (3) All the creep data, except the isochronous stress-strain curve, required for the design such as stress vs. rupture time, stress vs. secondary creep rate and stress vs. time to initiation of tertiary creep could be arranged through the Larson-Miller parameter. On the other hand, the isochronous stress-strain curve was figured out by estimating creep curves. The constitutive equations of creep for a heat of Hastelloy X proposed in this paper and the isochronous stress-strain curves derived from these constitutive equations were consistent with the experimental data obtained for the corresponding material.
Isothermal low-cycle fatigue tests including both load controlled and strain controlled tests were conducted on Hastelloy X, one of the materials for VHTR (Very High Temperature Gas Cooled Reactor). The effects of temperature, atmosphere, strain rate and hold time were studied. Special attention was given to the creep-fatigue criteria. The results are summarized as follows. (1) In the simulated VHTR He gas, the fatigue life increases as compared with that in air for the test without any strain hold time, but remains almost unchanged for the test with a long strain hold time. (2) In the temperature range from 800 to 1000°C, the Coffin-Manson relationship holds between plastic strain range and fatigue life, but it does not agree well with the result estimated from the tensile test. (3) The relationships between plastic strain range and fatigue life obtained from the strain controlled test under different strain rates and strain hold times are almost equal. The fatigue life under the strain controlled test can be estimated using these relationships. (4) The result of the load controlled test agrees well with that obtained by applying the linear damage rule on the assumption that the total damage consists of the two kinds of damages caused by cyclic plastic strain and by accumulation of creep strain on one side. The both damages can be calculated by applying the life fraction rule to the relationship between plastic strain range and fatigue life and to that between stress and creep rupture ductility, respectively.
In order to establish a proper method for testing the hardness of plastic pipes, various kinds of experiments were perfomed on plastic pipes being used most commonly. From the results, it was possible to establish the most proper method. The main results obtained are as follows: (1) Pipes should be tested after 72 hours or more since they were molded. (2) In the case that the measurements are to be made at 5 spots on a pipe, the length of the samples should be more than 70mm for types A, C, D and H, and more than 20mm (except when pipes of the standard length are used) for type I. The thickness of the samples should be at least 4.0mm for PVC, 3.0mm for MA, 4.0mm for P. P, and 3.5mm for ABS. (3) The degree of final roughness of the surface of the samples should be 3.2S, which can be achieved by using Emery Sandpaper #800∼#600.
Thin plate specimens of annealed 7-3 brass were subjected to gaseous ammonia under the loading conditions of constant displacement of tension, and constant deflection of bending and equi-biaxial bending. The followings are the summary of the results obtained (1) The time to crack nucleation, defined as the time when the applied stress fell down 3% of the initial value, was found to become shorter in the order of equi-biaxial bending, tension, and bending tests. The ratio of crack growth period to total life was 0.35 in the case of tension and 0.20∼0.24 in the case of bending. In the bending test under the stress level of 4kg/mm2, the specimen did not fracture although cracks were nucleated. (2) The surface stress measured by X-rays was found to decrease markedly with the nucleation of plural cracks. At the time ratio t/tf of 0.7, the mean span between adjacent cracks became nearly equal to the mean crack length and the stress on the specimen surface vanished. X-ray measurement of the surface stress was proposed in the present study to be a possible non-destructive method for monitoring the progress of damage due to stress corrosion cracking.
Center-notched plates of five kinds of high polymers available commercially, i.e., polymethylmethacrylate, polycarbonate, nylon 6, polypropylene and polyethylene were stretched under monotonic loading at room temperature. The plastic zone size and the opening displacement at the notch tip were measured experimentally during loading. The measured values of the plastic zone size and the opening displacement were found to be predictable using a modified Bilby-Cottrell-Swinden model where the cohesive stress distribution within the plastic zone was assumed to be parabolic. The X-ray diffraction method was found to show a clear difference in structure of the plastic zone ahead of the notch tip and the cold-drawn zone of the stretched smooth specimen. This method was confirmed to yield useful information in explaining the relation between the critical crack tip opening displacement and the tensile fracture ductility.
It is well known that aluminum single crystal having <100> tensile orientation shows initial rapid hardening but after about several per cent elongation, its tensile stress-strain curve becomes very flat until failure. The present authors observed and reported previously that the clustered slip accompanied by prominent cross slips propagates in this flat region of the curve. Similar stress-strain curves were reported in the <100> oriented gold and silver single crystals deformed at room and high temperatures, respectively. However, the detailed deformation behavior of copper single crystal having <100> tensile orientation still remains unclarified. In the present study, the <100> oriented copper single crystal was tested in tension at 200, 20 and -196°C in order to clarify the mechanism of deformation. The effects of tensile orientation, stacking fault energy and deformation temperature on the occurrance of prominent cross slips were discussed. The prominent cross slips have occurred in the <100> oriented copper single crystal and they were observed more frequently with increasing the deformation temperature. However, it was found that the number of prominent cross slips in the copper crystal was much fewer than that in the aluminum crystal, and the flat region of the stress-strain curve was not observed. The difference in deformation behavior between copper and aluminum crystals is considered to be due to the difference of stacking fault energies in these two metals.
The effect of grain size on fatigue damage in pure aluminium was studied in fully reversed plane bending by using the X-ray microbeam technique and optical microscopy. The results are summarized as follows: (1) The fatigue strength was found to be affected by the grain size in aluminium, although the substructure was well developed. This is interpreted on the basis that the crack initiation is affected by the grain size and the crack propagation takes place mostly along the grain boundary at higher stress amplitudes, but is retarded in the vicinity of the grain boundary at lower stress amplitudes. The grain size dependence of fatigue strength also seems to be associated with the purity of material. (2) The Petch's type relation did not exist between the fatigue strength and the grain size. The fatigue strength plotted against the inverse square root of grain size did not fall on a straight line but on a downwards bent line. This appears to have resulted from the facts that the stress amplitude at the transition from intergranular to transgranular cracking is also affected by the grain size as is the fatigue strength, and that the substructure develops more in the specimens having smaller grain sizes. (3) The subgrain size was found to be inversely proportional to the stress amplitude. But the subgrain size was also affected by the grain size and became smaller in the specimens with grains less than 50μm. On the other hand, both (Db)max and (Db)mim increased with an increase of stress amplitude and, furthermore, the former increased a little with increasing grain size.