Compressor discs of jet engines and turbine discs of gas turbine engines are usually subjected to a combined stress σcomb of centrifugal stress σt and thermal stress σth in the tangential direction during flight or start-stop time. However, there has been few studies on the life prediction for such components made of Al-alloy (2014-T6) at the condition of about 120°C and σcomb of 21 to 22 kg/mm2 during rotation. In this report, the life was estimated from the creep rupture data at 212, 400 and 600°F using the prediction method proposed by the authors.
The dynamics of melt spinning of glass was analyzed by solving a set of simultaneous differential equations consisting of the equations of heat, force and material balance. The steady state solutions corresponding to many different spinning conditions of glass fiber showed a good agreement with the experimentally measured values of temperature and cross-sectional area. Trouton viscosity β of glass fiber derived from the experimental results was expressed as a function of absolute temperature T by the following equation. β=2×10-3·e7200/T
Thermal behaviors of glass fiber have been studied by measuring their linear thermal expansion. The low transition temperature (T1) and the high transition temperature (T2) decreased with increasing spinning speed and with decreasing spinning temperature. These results suggest that the low transition temperature (T1), which changes with thermal history, is associated with alkali diffusion in glass fiber, while the high transition temperature (T2) is related to the degree of orientation of glass structure. The time effect on the recovery was measured at various temperatures in the transition region, and the isotherms were found superposible closely each other by horizontal shifts. The temperature dependence of the logarithmic shift factor (log aT) was in accord with the W. L. F. equation in the temperature range above 500°C, but diverged from the equation at lower temperatures.
Several fcc metals and alloys with low stacking fault energy have been rolled to various reductions. Detailed metallographic examination of the deformation structure has been carried out. The results for single crystals and polycrystals of α-brass are presented as the typical ones. The (211)  crystals are deformed with peculiar deformation behavior such as shear banding after onset of mechanical twinning, while the (101)  crystals are deformed by normal slip. It is shown that a shear band is a plate-like region which is nearly perpendicular to the side surface, and inclined 30-40 deg to the rolling plane of the crystal. In polycrystals, nearly the same type of shear bands as in the single crystal is formed in the grains deformed by mechanical twinning. The shear bands contribute to the plastic deformation of the specimen, and could be approximated to a shearing system. The occurrence of shear bands may be explained on the basis of plastic instability or localized deformation. However, the activation of shear bands should be considered as one of the deformation modes inherent in rolling, and differentiated from the inhomogeneous deformation such as deformation banding.
In order to investigate the effects of loading history and variation of temperature on the deformation behavior of softened celluloid by heating, the proportional loading tests were made at a constant temperature on specimens made of a celluloid plate and a thin walled cylindrical celluloid tube for various periods of heating time. The effect of heating time on the mechanical property of the material was examined, and the Poisson's ratio and the stress-strain relations under tensile and torsional loadings were discussed. The conclusions obtained are summarized as follows: (1) The mechanical property of celluloid is not affected by heating time. (2) The residual stress appearing in the thin walled cylindrical tube specimen is small and an isotropy in a plane is maintained. (3) The Poisson's ratio is 0.47 and the incompressible condition is satisfied still for the inelastic deformation within the range of 6 percent error. (4) The deformation resistance of torsional loading is larger than that of tensile loading. This may be attributed to the effect of a first invariant of stress tensor.
The crack initiation behavior in delayed failure was studied on various specimen configurations, using the apparent stress intensity factor, Ka, which is obtained by substituting notch length for crack length in the equation of stress intensity factor K. When Ka is constant, the time to crack initiation is independent of notch length a0 and loading methods, but is dependent on specimen thickness B and notch root radius ρ. When the crack initiation time is constant, the next experimental equation holds: Ka=KI(B0/B)δ(ρ/ρ0)λ, ρ≥ρ0 where KI is the stress intensity factor under plane strain state, B0 is the critical specimen thickness over which crack initiation time is constant, ρ0 is the critical notch root radius under which Ka at crack initiation is constant, and δ and λ are constant. ρ0 varies with crack initiation time and seems to depend not only on mechanical factors but also on chemical factors, such as the formation of protective film on the notch root surface.
In order to elucidate the effect of non-uniform stress distribution on the notch strength of cast iron, some plate specimens with edge notches or a hole were examined at room temperature and the liquid nitrogen temperature. Non-elastic stress analysis was performed by FEM to discuss the fracture criterion. The notched specimens with the elastic stress concentration factor of 1.6∼5.7 ruptured at the load 5∼40% smaller than the nominal stress criterion, but 8∼200% larger than the maximum stress criterion. The maximum stress at the notch root was estimated as 15∼80% higher than the tensile strength of the material. These results were explained by introducing the“boundary layer depth”δ of 2∼4mm, which was equivalent to the size of 3∼5 of graphite eutectic cells. The effect of the stress distribution decreased at the liquid nitrogen temperature where the cast iron showed almost an elastic and brittle behaviour. At that temperature, the value of δ was about 0.7mm, which was nearly the size of an eutectic cell.
Attempts were made to produce microduplex structures in 32%Cr-8%Ni steel with ferrite matrix and stable austenite phase by thermomechanical treatments. The specimens having 0.8 to 26.9μ grain size of austenite were obtained and their fatigue behaviors were studied. The main results obtained are as follows; (1) The microduplex structures having small austenite grain sizes up to 0.8μ were produced simply by quenching from the ferrite region, cold rolling and subsequent annealing at the two phase region. (2) Various appearances of fatigue cracks and various types of propagation were observed by changing the grain size and stress amplitude. The appearance of fatigue cracks was different at different locations of a specimen. Those were considered to be caused by the variation in crystallographic orientation, morphology and length of free path of ferrite and austenite. (3) Although the appearance of cracks depended on the grain size, it was observed by using an electron microscope (replica) that the initiation process of the cracks had similar features in both large and small grained specimens. In fine grained specimens fine slip band cracks initiated close to the ferrite-austenite interphase or inclusions during the initiation process of ferrite-austenite interphase or inclusion cracks, while in large grained specimens slip band cracks initiated within ferrite and austenite phases. (4) The fatigue strength followed the Petch-type relation. This may be caused by the fact that, as mentioned above, cracks initiate at the slip band irrespective of the grain size but the length of slip band cracks depends on the grain size.
Fatigue behaviors of structural adhesives composed of epoxide resin and aliphatic α, ω-dicarboxylic acids HOOC-(CH2)p-COOH (p=2, 4, 8 or 10) were investigated by analyzing the results of fatigue test and bulk properties of the systems. The fatigue strength and maximum value of damping increased with increasing p. The glass transition temperature, tensile strength, shear adhesive strength by tension, shear modulus and crosslinking density decreased with increasing p. On the basis of the above observations, improvement of fatigue strength was explained by considering low restriction of segmental motions, high relaxational ability, transformation of mechanical stress to heat and flexible structure of the cured system. A good correlation of fatigue strength with crosslinking density or area of tan δ curve was found to exist.
In the field of rock mechanics, it is well recognized that the elastic anisotropy is one of the important mechanical properties of mineral, rock, and rock mass. But, because of troublesomeness of the measurement, anisotropic elastic constants of these materials are not measured so frequently. Particularly the constants which give the relationship between shear stress and strain are rarely measured. The reason is that the preparation of torsion test pieces is relatively difficult, although the torsion test is usualy thought to be essential for obtaining these anisotropic constants. In this paper, it is shown that the all 21 elastic constants can be calculated using the transformation law of the co-ordinate of elastic constants, from the test results of uni-axial compression in which the axes of test pieces are varied to the principal axes of anisotropy. As an example, the experimental results and the calculated compliances of calcite (trigonal) are shown. The preciseness of the measurement is also discussed.
Generally, the fluctuation of arc resistance of electrical insulating materials is remarkably large even if the tests are carried out under the same condition with the same testing aparatus. In this paper, the cause of the fluctuation of arc resistance was discussed from the results of arc resistance tests on a large number of specimens. The experimental method was nearly same as the ASTM D495-71 and various plastics was used as the test materials. For each material, 80-100 tests were carried out. The main results obtained are as follows: (1) The frequency curve of arc resistance generally has plus skewness and can be represented approximately by a logarithmic normal distribution. Furthermore, this distribution curve agrees well with the result of calculation obtained by considering a stochastic process. Thus, it is considered that the phenomenon of arc deterioration is a statistical one. (2) The loading with fillers, the temperature rise of circumference and the increase of test voltage equally decrease the fluctuation of arc resistance. (3) No significant difference is observed on the fluctuation of arc resistance between the test with rod electrode and that with rectangular plate electrode in ASTM method. (4) It is convenient to elucidate the fluctuation of arc resistance when the probability of arc deterioration p is introduced. This value of p can be obtained from the slope of log r-t curves, r being the non-deterioration probability. Moreover, p is small at the beginning but reaches nearly constant after a while. (5) The non-deterioration probability can estimated when the log r-t curve is availabe. Even if the recurrent speed of discharge is different in each test as in the ASTM method, the non-deterioration probability r can be derived by using a number of the log r-t curves corresponding to each of the test steps.