In order to examine fatigue life prediction methods at high temperatures where creep damage need not be taken into account, fatigue tests were carried out on plane bending specimens of alloy steels (SCM 435, 21/4Cr-1Mo) under superposed and combined superposed stress waves at room temperature and 500°C. The experimental data were compared with the fatigue lives predicted by using the cycle counting methods (range pair, range pair mean and zero-cross range pair mean methods), the modified Goodman's equation and the modified Miner's rule. The main results were as follows. (1) The fatigue life prediction method which is being used for the data at room temperature is also applicable to predict the life at high temperatures. The range pair mean method is especially better than other cycle counting methods. The zero-cross range pair mean method gives the estimated lives on the safe side of the experimental lives. (2) The scatter bands of N/Nes (experimental life/estimated life) becomes narrower when the following equation is used instead of the modified Goodman's equation for predicting the effect of mean stress on fatigue life. σt=σa/(1-σm/kσB) σt; stress amplitude at zero mean stress (kg/mm2) σB; tensile strength (kg/mm2) σm; mean stress (kg/mm2) σa; stress amplitude (kg/mm2) k; modified coefficient of σB
The apparent compressive stress σ of powdery layer in a steel mortar was found to be directly proportional to the compressive strain φ as follows: φ=Δl/(Δl∞-Δl) σ/φ=α σ=P/F where Δl is the displacement of powdery layer under compressive load P, Δl∞ that in the case P=∞, F the cross sectional area of plunger, and α the modules of compressibility of powdery layer.
It has been known that the creep strength of polyvinyl chloride (PVC) in conc. nitric acid is remarkably lower than that in air. This behavior has been explained from the viewpoint of the plasticizing effect of nitric acid in PVC. However, the diffusion process of nitric acid in PVC has not yet been made clear. In this paper the diffusion of conc. nitric acid in rigid polyvinyl chloride was studied experimentally by using the film-adhesion method which is similar to the film-roll method. The tests were carried out under unloaded and loaded conditions at temperatures 25 to 55°C. The results obtained are as follows: (1) Under the unloaded condition, the diffusion process is Fickian and the surface concentration of nitric acid in PVC does not change with immersion time. The diffusion coefficients of nitric acid are in the range of 10-12-10-11cm2/sec. (2) In the case of tensile stress levels less than 75kg/cm2, the diffusion process is almost the same as that in the case of the unloaded condition at 40°C. At a stress level of 95kg/cm2, the diffusion process is non-Fickian where the surface concentration increases with immersion time.
From the experimental results, it has been found that the molten bonding pellet, which is obtained by drying slowly a green pellet in a vinyl chloride bag, shows the strength as much as one thousand times of the strength of dry pellet. When the molten bonding is not formed enough, the strength of this pellet can be expressed by the same equation (3) as that for dry pellet. When the molten bonding proceeds, the strength of the pellet can be expressed by the equation (4) independent of particle diameter. This phenomenon could be explained as follows: Na and K constituents in glass pellet dissolve first into water contained in a pellet and congeal to solid state again as the pellet is dried slowly in a vinyl chloride bag. Consequently, the volume of pellet shrinks like in the sintering process as it dries.
A two-dimensional model of granular materials was made by using polyurethane rods, which are utilized as a photoelastic material. Simple shear tests on the model beds were carried out in order to observe microscopic internal mechanisms. The rotation angle (α) of particles, the contact angle (θ) of contact points, interparticle forces (f) and their inclination angles (δ) were detected at a constant interval of shear strain (γ). The interaction between adjacent particles was calculated from the observed results, and the changes of contact direction (Δθ/Δγ), sliding movements (Δψ/Δγ) and rolling movements (Δρ/Δγ) at contact points were obtained. Their statistical values (the average and standard deviation) were investigated. It was found that these values depended on the contact angle as well as on the interparticle force. The interaction between particles became remarkable with increasing shear strain. Therefore, the number of contact points in a special direction increased remarkably during the early stage of shear, and then, it decreased slightly during the steady state of shear. The stress-strain relation was calculated from the microscopic behavior, and was in good agreement with the experimental one.
This paper presents a method of predicting the nonlinear stress-strain curves up to the tensile failure for plain woven glass fiber cloth reinforced polyester laminates, when the load direction does not coincide with the fiber direction. The application of the laminate plate theory to the small stress increments of the stress-strain curve was done in the proposed analytical procedure. Comparisons were made between the analytical prediction and the experimental results for plain woven and nonwoven glass fiber reinforced laminates. As the results, the calculated stress-strain curves were found to be in considerably good agreement with the experimental results and the nonwoven laminate had higher strength and rigidity than the plain woven laminate for those with low fiber orientation angles, but had lower strength and rigidity for those with high fiber orientation angles.
Flexural properties of moldings made by RIM (Reaction Injection Molding), which are structural foams consisting of high density skin and low density core, were investigated by three-point bending tests. The following two breaking modes were observed in bending tests of the moldings, depending upon the density ratio of skin layer and core layer. (1) The other side of the skin layer on which load was applied was broken by tensile stress. (2) The same side of the skin layer on which load was applied was broken by compressive stress, causing Wrinkling buckling. Then, the conventional composite beam theory was applied to the former breaking mode and Hoff's buckling to the latter, and equations were derived to predict the flexural properties of the structural foams which involve buckling from the flexural properties of solid construction. In addition, it has been shown that there exists the best density distribution that maximizes the flexural strength of the moldings made by RIM with a given overall density. The results obtained here should be useful to optimum structural designing of moldings made by RIM.
In this paper, fracture criterion of gray cast iron under non-uniform stress was discussed on the basis of bending and eccentric tension tests performed on the rectangular specimens at room and liquid nitrogen temperatures. The experimental results were analyzed with a concept of“over stressed depth δ”, which was proposed by the author as a fracture criterion of notched cast iron in the previous paper. The values of δ, obtained from the calculated stress distribution at fracture and the tensile strength of the material, were almost constant and 2-5mm for both bending and eccentric tension at room temperature, coinciding with those obtained in notched plates and bars. At liquid nitrogen temperature, δ decreased to less than 1mm, also coinciding with that in notched tension. Thus the applicability of constant δ as the fracture criterion under the stress gradient conditions was confirmed. The value of δ at liquid nitrogen temperature was almost equivalent to the size of a graphite eutectic cell which is considered to be a microstructural unit of cast iron, and at room temperature, increased by 3-5 times. From the fact that δ depends not only on material but on specimen size or stress gradient and especially on temperature, δ was infered to be a parameter which represents the propagating condition of a crack, initiated at highly stressed surface or notch root before final fracture. By the concept of over stressed depth, mechanical characteristics of gray iron such as high bending strength, low notch sensitivity and their dependences on temperature and specimen size can be explained rationally with a single parameter δ.
Blast-plastic analysis was carried out for the corner region of a slit under compressive stress. From this analysis the ratio of work hardening coefficient H' to elastic modulus E play an important role in transition from the elastic stress field to the plastic stress field. The aspect ratio d/a of a slit under the elast-plast condition affects little the stress concentration at the slit corner. But stress apart from the corner decreases rapidly when the aspect ratio decreases.
An attempt was made to correlate bending and direct stress fatigue strengths by using the cyclic stress-strain relation of the material. The cyclic stress-strain relation was shown to be representable with the ultimate tensile strength: σB as follows σ=1.73σBεp0.16 By using this relation, the nominal stress σn in bending was calculated and related with the true stress σ at the specimen surface. Then, by replacing the nominal stress by the surface stress, S-N curves obtained from the bending and direct stress fatigue tests on S15 CK steel were compared each other and shown to be in good agreement. Therefore, the bending fatigue strength was found to be correlatable with the reversed direct stress fatigue strength through σn-σ diagrams.
The elastic-plastic deformation of a square plate with a center crack under biaxial stresses was analysed by the finite element method of plane stress. The loading conditions analysed were equi-biaxial, uniaxial and shear. The values of the crack-tip opening displacement and J integral increased as the stress parallel to the crack changed from tension to compression, when compared at the same perpendicular stress. The increment under shear loading was considerably large, while the difference between the equi-biaxial and uniaxial cases was relatively small. The interrelation among the near-tip stress-strain distribution, the crack-tip opening displacement and J integral was confirmed to be characterized by the HRR type singurality field. A simple method to estimate J integral for a center-cracked plate under biaxial stresses was proposed based on the finite element analysis. The fatigue crack growth data obtained with a cruciform specimen under biaxial cyclic stresses were analysed in terms of J integral range by using the proposed method. The relation between crack growth rate and J integral range was found to be a unique power relation independent of biaxiality and plasticity, although a considerable acceleration was detected under shear loading when the rate was correlated to the stress intensity range.
Sharply notched specimens of structural low-carbon steel were fatigued under several ratios of maximum to minimum stresses. The growth behavior of a fatigue crack near the notch root was analysed based on the crack closure measurement. A fatigue crack first decelerated with increasing crack length, and then accelerated or became non-propagating depending on the range and ratio of the applied stress. The effective range of the stress intensity factor was found to be the single-valued parameter in describing the growth rate of small cracks near the notch root, and their relation agreed well with that obtained for long cracks. By considering the increase in crack closure stress with crack length, a quantitative method was proposed for predicting the non-propagating crack length and the fatigue strength reduction of notched specimens as a function of the applied stress and the notch geometry.
Fatigue crack growth tests in a wide range of fatigue crack propagation rates covering the threshold condition were conducted at various stress ratios using a quenched-tempered and high strength steel SCM435 with changing heat treatment. The crack closure behavior due to both the plastic deformation at the crack tip and the formation of oxide deposits on crack faces at an ultralow, near-threshold propagation rate were confirmed experimentally by the proposed Top-on ultrasonic method. The influences of stress ratio, strength level and microstructure on fatigue crack growth characteristics were discussed on the basis of the crack closure concept. It was shown that the fatigue crack propagation rate, da/dN, can be related uniquely to the effective stress intensity factor range, ΔKeff, and the threshold level defined at da/dN=10-8mm/cycle in terms of effective stress intensity factor range, ΔKeff, th, are nearly constant regardless of stress ratio and strength level. However, the relationship between da/dN and ΔKeff at such a high stress ratio as R=0.8 where the plasticity induced crack closure and the oxide induced crack closure don't exist, didn't agree with the above relationships at R=0.05, 0.15, 0.4 and 0.6, and ΔKeff, th was Large.
The stress corrosion cracking (SCC) susceptibility of SUS304 stainless steel cathodically precharged with hydrogen was investigated in boiling 35% MgCl2 solution at 398K. The specimen preparation was carried out by precharging at 200A/m2 for 7.2ks in the SCC test solution after 0∼35% prestrained at 195K, and successively SCC tested by the constant loading method. The mechanical properties of specimens precharged with hydrogen are closely related to the hydrogen-induced phase transformation. By precharging with hydrogen, especially the lattice of α'-martensite in high prestrain is expanded and activated. In this case, the crack initiation during SCC of these specimens is controlled, consequently the SCC susceptibility shows a marked decrease. On the other hand, the SCC fracture surface morphology of specimens precharged with hydrogen depends on the distribution and form of strain-induced martensite (ε, α'), which contributes considerably to the susceptible path for cracking. It is considered from these results that the role of hydrogen dissolved in specimen during SCC has the effect of embrittlement and active dissolution to ε-and α'-martensite, respectively.
The tensile, creep and fatigue tests were carried out on stress aged SUS316 stainless steel in order to clarify the effect of creep damage on fatigue strength. The interaction between creep damage and high temperature strength, especially fatigue life, was discussed. The precipitates of M23C6, χ-phase and Laves-phase were observed, and σ-phase precipitated mainly along grain boundaries in the long-term stress aged material. In the long-term stress aged material the creep cavities were found to exist along the grain boundary particles and non-metallic inclusions in grain. The fatigue life of the stress aged material decreased with increasing ageing time. On the other hand, the fatigue life of the thermal aged material increased by ageing. It was considered that the increase of fatigue life of the thermal aged material was related to structural change, especially the precipitation of carbides, etc., and that the decrease of fatigue life of the stress aged material was related to creep cavities which accelerate fatigue-crack initiation and propagation. The creep damage of the stress aged material accompanies structural change as well as formation of creep cavities. Therefore, the fatigue life of the stress aged material is affected by these two factors; deceleration of fatigue crack propagation due to the precipitation and acceleration of fatigue crack initiation and propagation due to the creep cavities.
In order to develop a non-destructive method to detect the creep damage and estimate the residual life for the machine parts used at high temperatures, the changes in material properties caused by long-term heating and by creep were observed with the test methods of X-ray diffraction, eddy current and hardness. The material tested was the tempered Cr-Mo-V steel. The results obtained are summarized as follows: (1) Hardness test was found the most suitable method to detect the material property change for Cr-Mo-V steel. (2) The change in hardness during long-term heating was able to be expressed by a simple curve in the plotting against Larson-Miller parameter. Softening was caused by the heating above a certain temperature but below the temper temperature. (3) The softening was accelerated by the stress larger than 11kg/mm2. The change in hardness during creep under the stress of σ (kg/mm2) was correlated with those of long-term heating by the plotting against the parameter of logT+log(20+logt)+0.00217(σ-11), where T is the test temperature (K) and t is the loading time or heating time (hour). (4) A method to detect the creep damage and to estimate the residual life was deduced from the above mentioned results.
In order to investigate the basic thermal fatigue (TF) life properties of materials used for high-temperature components in a power plant, in-phase and out-of-phase TF tests as well as isothermal low-cycle fatigue tests at maximum temperature, Tmax, were carried out on low-alloy steels, austenitic steel and superalloys. Also, based on these test results, discussion was done on TF life prediction from other mechanical properties. Ter results obtained were summarized as follows: (1) The dependence of TF life on strain range could be classified into four categories, and two of them (i. e., I-and O-types) showed shorter in-phase and out-of-phase TF lives at low strain ranges. The I-type was due to creep-fatigue interaction in the in-phase TF, while the O-type was caused by the difference in crack initiation life between both phases. (2) The TF lives of low-alloy and stainless steels with low-strength and high-ductility, except a case of the I-type, were almost the same, i. e., approximately 500 and 5000 cycles at the total strain range, Δε, of 1.5 and 0.5% respectively. In general the TF lives of superalloys with high-strength and low-ductility were shorter than those of the above steels, but approached to them at low strain ranges. (3) Good or safe prediction of TF life, except in-phase in the I-type, could be done by applying the Δε-based relations, particularly the Δε vs. life relation in isothermal fatigue at Tmax. (4) High-temperature low-cycle fatigue failure seemed to have the same stress-dependence as that appearing remarkably in the high-strength/low-ductility superalloys.
Acoustic Emission (abbreviated as AE) is the transient elastic wave emission due to microfracturing in a solid. In the previous paper, theoretical representation of AE wave motions in concrete was reported by applying the theory of dislocation and elastodynamics. Once sources are mathematically described by the dislocation model, AE waveforms generated by various types of source mechanisms are characterized with emphasis on features of the radiation pattern, which is the spatial distribution of amplitude and polarization of first motions. In this paper, fundamental studies in regard to radiation patterns of AE wave motions are reported. Such patterns are verified in the experiments. In the source location experiments, the types of sources can be classified into two groups from AE characteristics based on the radiation pattern. They are associated with tensile cracks and others. From the relation between the radiation pattern and the dislocation model, in addition, the orientations of dislocation models may be determined. The improved source location technique useful for such an attempt is developed and applied in the experiments. The results show that the determination of source kinematics by the aid of the radiation pattern is promising in concrete.