When Ti-6Al-4V alloy was polished by using diamond paste or #400 emery paper, fatigue cracks initiated at the surface and its fatigue strength scattered widely. It was considered that the scatter of fatigue strength depended on the local surface morphology. On the other hand, when the fatigue crack initiated at the subsurface with shot-peening, the scatter of fatigue strength became narrowed remarkably. The fatigue strength of the shot-peened specimen, however, decreased again in the long-life region, and no fatigue limit was recognized even more than 107 cycles. This phenomenon might depend on the residual stress distribution.
The internal stress of thermosetting resin generated during curing process was analyzed by using a viscoelastic model. The changes in viscoelastic properties in the curing process were measured experimentally by a rotating rehometer. The viscoelastic constitutive relations in the curing process were formulated by the three elements spring dashpot model. The deformation of two layered resin and metal in the curing process was investigated experimentally. The results were compared with F.E.M. analysis on the basis of the viscoelastic relations and shrinkage model of the resin. The viscoelastic analysis gave more realistic results than the elastic analysis.
Carbon steel (S45C) for machine construction and alloy tool steel (SKS3) were coated with chromium or chromium carbide by using electroplating or powder pack-method (Chromizing) and then titanium carbide. It is possible thermodynamically that titanium carbide (TiC) forms on the chromium carbide (Cr7C3) layer. The total thickness of this dual layer was approximatelly constant, i.e. 10-12μm. The hardness distribution profile of the dual layer changed gradually from the surface. When the chemical potential of carbon in the substrate became higher or the chromium carbide- or electroplated chromium layer became thinner, a thicker double carbide layer consisting of chromium- and titanium carbide was achieved.
In manufacturing high strength bolts, the zinc phosphate coating is generally perfomed in order to increase the effect of lubrication. However, a phospharized layer is often formed on the bolt surface by diffusion of phosphate during subsequent heat treatment. In this study, the effect of this layer on delayed fracture was investigated. The results were as follows: (1) In high strength bolts, the phospharized layer at a thickness of about 40μm including the 5-10μm ferrite layer was formed on the surface. (2) After the ferrite layer was corroded, the grain boundary of the phosphorized layer having a rich phosphate concentration was corroded preferentially. The corrosion pits became coupled and coalesced, growing to an initial crack. (3) As the result of acceleration of corrosion rate and diffusivity of hydrogen with the existence of phospharized layer, the freqency of occurrence of initial cracks was increased, giving a detrimental effect on the characteristics of delayed fracture.
The main objective of this study was to clarify the effects of microstructural change on minimum creep rate and rupture life of Hastelloy X, and was also to develop a creep life prediction method from the microstructural observations. For this purpose, long-time aging at 750 to 900°C up to 104hr was used for making different degraded microstructural materials. The accomplishments of this study were as follows; (1) During the aging, carbides of M12C, M6C, and M23C6, and an intermetallic compound of μ phase were precipitated at the grain boundaries and in the grains. Intergranular precipitates strengthened the creep resistance, but intragranular ones weakened, which was not commonly expected. (2) The minimum creep rate (εm) of the aged Hastelloy X at 850°C was explained as a function of grain boundary coverage ratio of precipitates (ρ), volume fraction of intragranular precipitates (V), aging temperature (Ta), and applied stress (σ). The equation was as followed. εm=[ε*0+A1(1-ρ)(V-V0)2exp(-QD/κTa)]σn (3) Creep rupture life was also explained from the minimum creep rate by using the Monkman-Grant equation for the aged degraded Hastelloy X.
In our previous paper, an analytical crack growth law to predict crack growth rate during the retardation period was established through the detailed evaluation of fatigue crack growth behavior under two-step varying load conditions with the stress ratio R(=σmin/σmax)=0. The main purpose of this study is to develop the above crack growth law to make applicable to various types of two-step varying load conditions. For this purpose, the effects of the tensile mean stress and the stress ratio of the lower stress level on the crack growth behavior during the retardation period induced by load change were investigated by carrying out fatigue crack growth tests on aluminum alloy (A2017-T3) compact tension specimens under four types of two-step varying load conditions with the lower stress level of positive stress ratio. Detailed observations for the crack length grown by the application of lower stress cycles were made by an optical microscope and/or on SEM micro-fractographs. The results indicated that crack growth behavior in the deceleration stage during the retardation process was governed only by the lower stress intensity range, ΔKL, and that the crack growth ratio in the acceleration stage was presented as a function of the maximum values of the higher and the lower stress intensity ranges together with the value of ΔKL. A new crack growth law to estimate the fatigue crack growth rate during the whole retardation process including the deceleration and acceleration stages was derived by using the empirically determined parameters.
The authors have been carrying out a series of fatigue crack growth tests under two-step varying load conditions to reveal the fundamental characteristics of fatigue crack growth behavior under varying load conditions, and have established an empirical law governing the crack growth rate during the retardation period observed after the load change from high to low load amplitudes. This report describes the results of fatigue crack growth tests carried out on Al alloy CT specimen to investigate precisely the effects of pre-straining and aging treatment on the crack growth behavior during the retardation period under two-step varying load conditions. The results indicated that the crack growth rate under constant amplitude load conditions was not affected by the pre-straining and the aging treatment, but that the crack growth rate during the retardation period was remarkably influenced by these treatments. By taking the above mentioned crack growth behaviors into account, the analytical crack growth law to estimate the crack growth rate in the retardation period, derived empirically in our previous studies, was modified to be available for a wide range of material conditions. Furthermore, a reasonable crack growth model was also proposed.
From the fractographic observation on silicon nitride specimens failed under rolling contact fatigue, it has been found that the crack initiation in silicon nitride specimens subjected to rolling contact fatigue is induced by cyclic subsurface shearing stress, as is known in steel bearings. Though these subsurface cracks are produced in the direction parallel to the cyclic shearing stress, cracks grow in the direction near to the plane of the maximum tensile stress if mode II loading is applied to them. The difference between the crack growth in simple mode II loading and the crack growth in rolling contact fatigue is, we suppose, whether or not there is a superimposed compression stress. Based on this hypothesis, we developed an apparatus to obtain intrinsic mode II fatigue crack growth characteristics, as a simplified model of determining the subsurface crack growth in rolling contact fatigue. Some preliminary results on da/dN-ΔKII relations have been obtained by using this apparatus on a steel and aluminum alloys.
In this study, crack propagation is regarded as a successive random walk process. Two Markovian models of crack growth described by the Fokker-Planck equation are introduced through Paris-Erdogan's law, and the crack length distribution at any fatigue cycle and the life distribution at any crack length are set up analytically with some approximations. The first model is expressed by the Fokker-Planck equation with constant coefficients, which are calculatable from the crack propagation data, and another model is introduced theoretically from the distribution of the coefficient of Paris-Erdogan's equation directly. These Fokker-Planck equations are analysed by the Lax-Wendroff scheme, one of the finite-difference methods. As a result, the crack-length distribution in random stress sequences is evaluated successfully.
Fatigue crack growth under thermal fluctuation is theoretically investigated by the use of a Markov approximation method, under the condition that the temporary variation of the inner surface temperature of the plate can be modeled as a narrow-band stationary process. First, under an assumption that the peak stress causes the crack growth, a crack growth equation is formulated based upon the Paris' law and it is extended to a random differential equation, in which the random variation of crack propagation resistances are also taken into account. Next, a residual life distribution as well as a probability distribution function of the crack growth processes is derived. Finally, some numerical examples are shown to examine the quantitative behavior of the residual life distribution in the highreliability region. The result indicates that the residual life in such a region is of order of 108 cycles.
In the fracture toughness tests of perfectly brittle and crack-rate sensitive ceramics (R∝am*, R; the crack resistance force, a; the crack velocity), it has been analytically as well as experimentally clarified that an initial crack begins to extend before a maximum load appears in load-deflection curve. This crack-extension phenomenon was named PSP (psuedo-stable phenomenon) in the previous paper. The amount of the crack-extension due to the PSP depends on the crack-rate sensitivity parameter m*, the non-dimensional compliance of testing machine EBΛ* and the initial crack length x0. Nevertheless, the fracture toughness value, KIC has been currently determined by the maximum load and the precracked crack-length measured before testing. In the testing regulation for high performance ceramics in JIS, the KIC is also evaluated in the same way. For a single edge cracked 3 point bending specimen of JIS, the amount of crack-extension is numerically calculated as a function of m*, EBΛ* and x0, and it is studied whether or not the KIC is evaluated exactly. The results obtained show that the KIC is reasonably evaluated when both m* and EBΛ* are small enough, but the KIC is considerably underestimated when these values are large.
Monte Carlo simulation method was used to evaluate the theoretical distributions of the estimates of Weibull (shape) and scale parameters for various sample sizes. It was found that the theoretical distributions of the estimates of Weibull and scale parameters are well approximated by log-normal distributions, and the scatter and bias of each distribution decrease with the increase of sample size. The comparison of the distribution of Weibull parameter estimated by experiments with the theoretical distribution indicates that the scatter of the experimentally obtained Weibull parameter results from the statistical nature of the estimates depending on the sample size and the difference of test conditions is not a major factor for the scatter. And it was also revealed that the distribution of scale parameter obtained by experiments does not agree with the theoretical one obtained by Monte Carlo simulation. This means that the scale parameter or the intrinsic strength of sintered silicon nitride and sintered silicon carbide depends on the fabrication and testing techniques.
For evaluating the thermal shock resistance of ceramics precisely by quenching test, one must know the heat transfer coefficient h between quenching liquid and surface of specimen, when a critical temperature difference Δθc is measured. However, few studies have been devoted to quenching liquids applicable to thermal shock test of ceramics, and the heat transfer coefficient is estimated from the Δθc value when needed. In general, h in a quenching liquid exhibits a maximum value in the temperature region where the heat transfer mode changes from bubble boiling to transition stage. Therefore, it may be meaningless to measure the Δθc value for some ceramics, which are fractured by thermal stress near the region of maximum h. In this study, the temperature changes at two different positions in the specimen were measured by using various quenching liquids and the results were analyzed. The useful so-called heat transfer coefficient for thermal shock test was obtained from the temperature difference. On the basis of this method, the mixture of mineral oil and butanol was proved to be a suitable quenching liquid for thermal shock test, in which h did not show maximum in the usual quenching temperature difference range of 150 to 450°C of ceramics.
High strength concrete with the compressive strength in excess of 80MPa has recently been a matter of concern in its application to the column or core wall of high-rise reinforced concrete building. High strength concrete can potentially generate high hydration temperature at early ages due to its relatively high cement conctent. It has been pointed out that a high temperature profile at early ages has a detrimental effect on the strength of high strength concrete, reducing long-term strength development and increasing thermal cracking risk. In this study, a temperature matched curing system (TMC) which can simulate in-situ temperature profiles of the column was developed in order to investigate the influence of high hydration temperature at early ages on the strength development of high strength concretes with and without silica fume. It was found that under sealed conditions, a high temperature of 70°C at early ages in high strength concretes with silica fume led to a little higher 7-day strength but no increase in strength beyond 28 days, whilst it had no significant detrimental effect on the strength development in those without silica fume. The evaporable and non-evaporable water content data showed that both the hydration of cement and the pozzolanic reaction of silica fume almost completed after 3 days TMC due to the shortage of water in concrete when high strength concretes with silica fume were subjected to a high temperature of 70°C at early ages.
In order to clarify the method of expressing tree form, the effect of wind on tree form and the branching system from the dynamic point of view, Strahler's system was applied to Masaki (Euonymus japonicus), Tobera (Pittosporum tobira), Enoki (Celtis sinensis var. japonica), Mochinoki (Ilex integra), Hisakaki (Eurya japonica) which grow at the tip of Miura peninsula in Kanagawa Prefecture. The results obtained can be summarized as follows: The number of branches exponentially decreased with an increase of the order of branch. The ratio of decrease of branches was constant for all the species under weak winds (point B). However, those of Masaki, Tobera and Enoki increased, while that of Hisakaki remained almost unchanged and that of Mochinoki decreased under strong winds (point A). The width and lengths of branches exponentially increased with an increase of the order. Winds checked both thickening and elongation growth for all the species except Mochinoki. Enoki likely accommodated strong winds with an decrease of aspect ratio of branches, while Masaki and Tobera did it with restraint of elongation of higher order branches to the growth check of wind. The mechanical safety allowance S of each branch was proposed. For all the species, it increased with a decrease of the order of branches. The branching system changed to a safer side toward winds.
Thermal power plants are now operating on a DSS or WSS basis in response to electricity supply and demand. This tends to increase the pump start/stop frequency and operation time in the low flow rate range so that the crack is possibly initiated. Therefore, the boiler feed water pumps can be maintained in stable operation if more accurate inspection techniques for maintenance and control are employed. In this study, three sophisticated nondestructive inspection techniques that are sensitive to fatigue damage have been utilized. They are: 1) X-ray diffraction method to detect damaged microstructure before crack initiation, 2) short crack detection by image processing of surface structure and 3) potential drop method to detect surface crack configuration. Those techniques are packaged as a system with a specially designed mechanical scanner for the pump shaft. The system has been applied to more than 200 actual pump shafts in a factory during in-service inspection.