In a solid with many cracks, the stress field is affected by the mutual interface of cracks. In the present paper, to investigate the interaction of many cracks, the torsional stress state of a cylindrical rod with three parallel ring-shaped edge cracks is considered on the basis of the theory of elasticity. The radius of the central crack in different from those of the others. Numerical results are illustrated for the distributions of displacements and shear stresses and for the variations of the stress intensity factors with distance between cracks and with the sizes of the cracks. The stress intensity factors in the deep cracks are larger than those in shallow ones. The maximum stress intensity factor in a cylindrical rod with three cracks increases with increasing distance between cracks, and approaches that in the case of a single crack.

View full abstract

Static and fatigue tests of thin-walled, [±45] filament-wound tubular specimens of a carbon fiber reinforced plastic, T300/827, under combined axial tensile and torsional loading have been conducted, and the influences of water absorption on the failure behavior have been investigated. The tests were conducted under a load-controlled condition, keeping a constant value of the combined stress ratio, α=γ/σ. The static strength of dry specimens, which had been held in air, showed good agreement with the Tsai-Hill failure criterion. The static strength of wet specimens, which had been preconditioned in water for two months, atα= 0 - 1 decreased by water absorption, whereas atα= 2 - ∞, where compressive strength in the fiber direction dominates over the failure, the decrease in the strength of wet specimens from dry ones was small. Fatigue strength of dry specimens atα= 0 - 1 was smaller than those of α= 2 - ∞. This indicates that a combined loading mode influences the fatigue strength. The decrease in fatigue strength of wet specimens from air data at α= 0 - 0.5 was higher than those at α= 2 - ∞.

View full abstract

Tensile fatigue properties of three types of composite laminates were examined. Stacking sequences of laminates were [0/90/+45/-45]_s, [+45/0/-45/90]_s and [0/90₂]_s. Modulus reduction during the experiments was measured and delamination growth was investigated. Interlaminar stresses along the free edge were calculated by the finite-element method and the stress distributions were compared with experimental results. Test results revealed that the unnotched and the notched specimens had almost the same fatigue strength properties. Modulus reduction with respect to the initial value for [+45/0/-45/90]_s laminates was about l0%, whereas [0/90/+45/-45]_s and [0/902]_s laminates showed no change from the initial modulus. This modulus reduction for [+45/0/-45/90]_s laminates was observed at the first l0%20% of failure life and it was mainly due to free edge delamination. Interlaminar stresses were predominant in free edge delamination and fairly enabled to infer delamination onset interfaces.

View full abstract

Tensile and fatigue characteristics of GFRTP were examined in the longitudinal (L) and transverse (T) directions. Young's modulus E and tensile strengthσ_B were higher in the L-direction than those in the T-direction. As a result, fatigue strength σ_a showed a higher value in the L-direction. The relation σ_a/σ_B and number of cycles to failure were coincident in both directions.

View full abstract

The effect of shock pressure on the stress-strain behavior of a low-carbon steel of four different grain sizes, 9, 16, 33 and 80μm, was investigated. The specimens were explosively shock-1oaded in water. The shock pressure effect was compared with the effect of pressurizing or temper rolling on the stress-strain behavior. Sensibility to twinning induced by shock loading was also examined. Results obtained in this investigation are summarized as follows. ( 1 ) The yield stress of all kinds of specimens began to decrease at the shock pressure of 300 MPa and continued to decrease to the minimum with increasing pressure, owing to generation of free dis1ocations in the specimens. For the specimen of smaller grain size, the amount of decrease in the yield stress became larger and the shock pressure (p_s)_m which minimized the stress became higher. ( 2 ) Yield ratio of the specimens of four different grain sizes of 9, 16, 33 and 80 μm, which had the original ratios of 0.80, 0.76, 0.65 and 0.52, respectively, could all be reduced to about 0.4 after shock loading. ( 3 ) The shock pressure (p_s)_t at which the mechanical twins began to form in the specimen increased linearly with increasing d^-1/2, which is a reciprocal of the square root of grain size d. Twin density, defined as the ratio of number of crystal grains with twins to total number of grains, was larger in the specimen with larger grain size and the density increased monotonously with increasing shock pressure. ( 4 ) For the specimen with a given grain diameter, the shock pressure (p_s)_m at which the yield stress became the minimum was identical in value to the pressure (p_s)_t at which the mechanical twins appeared in the specimen. This indicates that the macroscopic plastic strain begins to occur in the specimen at the pressure of (p_s)_m. ( 5 ) The decrease in the yield stress due to shock loading was considerably larger than the decrease in the cases of pressurizing and temper rolling.

View full abstract

The interface crack initiated between the die pad (lead frame) and the resin encapsulant in the LSI package sometimes leads to package cracking and becomes critical to package fracture design. On the other hand, the property of thermal diffusion has become a main concern for recent package design of power semiconductor devices. To simulate these problems, we developed the boundary element heat conduction and thermoelastic analysis code, which can deal with thermoelastic contact of the interface crack. The stress intensity factors for an interface crack are calculated by the extrapolation method which is proposed by one of the authors. Using the above technique, the interface crack initiated in a plastic-molded LSI package is analyzed under conditions of uniform temperature loading and heat generation in the Si-chip. Calculated interfacial stress intensity factor Ki shows a decreasing trend against the interface crack length in all the above cases. Taking into account the contact of the crack surface, the results of the disturbance in temperature distribution are outstanding. However, the normalized stress intensity factor K_i/|ΔT_max| Shows almost identical behavior against with the crack length and is less dependent on the contact region length and the amount of heat generation.

View full abstract

The present study provides a formulation of the creep constitutive equation which is descriptive of anisotropic creep hardening behavior. The basic structure of the model is based on the general framework of irreversible thermodynamics; we define a set of energy-conjugate state variables using the Helmholtz free energy function and the dissipation energy function. The prototypal model of the unified type accounts for combined isotropic and kinematic hardening, and it is applied to a description of anisotropic creep behavior. In order to enhance the predictive accurcy of prototypal evolution equations of the Bailey-Orowan type, we furthermore develop an auxiliary creep hardening rule in general form ; the concept of material memory is introduced where isotropic hardening is suppressed and kinematic hardening is promoted. It is shown that the auxiliary creep hardening rule is applicable to any model of the pure kinematic type. The proposed creep model, which is a combination of the prototypal model and the auxiliary rule, can well reproduce creep softening behavior under complete stress reversals, as in the ORNL creep model.

View full abstract

The aim of this work is to show that creep voids or cavities are observed at the secondary stage of the creep curve. Using commercially pure aluminum, creep damage tests were carried out at the temperatures of 200°C and 250°C. Creep tests were interrupted at several points in the creep curve, and the number of creep voids, creep void area per a unit area, and change in density in specimens were measured to evaluate creep damage. It was found that creep voids were formed in the middle of secondary creep when the fracture mode was intergraunlar, and that they developed rapidly with increasing creep strain. The creep damage parameterω introduced by Kachanov and Rabatonov agreed with the dimensionless value of change in density. There was no significant difference between secondary and tertiary creep in view of the creep damage measured.

View full abstract

In continuum mechanics it is not clear what kind of principle is equivalent to the principle of material frame indifference (PMFI). In the present paper, it is first shown that not only mechanical conservation laws but also rules of frame transformation for constitutive quantities and source quantities are simultaneously derived by changing frame on the assumption that the forms of the first and second laws of thermodynamics are invariant for the changes of frame. Next, considering whether the reverse proposition is true or false, an explanation is given for the relation between the PMFI and the invarinance of the laws of thermodynamics for the changes of frame. Then the essential standpoint of PMFI is definitely placed, making the relation between the PMFI and the frame-invariance of Clausius-Duhem inequality clear. Moreover it is indicated that the method which derives the conservation laws using the changes of frame is equivalent to both the principle of virtual power and the method by Green-Rivlin.

View full abstract

In order to examine the effects of plastic deformation on the stress fields near the bonding edge, an elastic-plastic finite-element analysis was performed on a bimaterial plate. The materials above and below the interface were assumed to be different from each other in Young's modulus and/or yield stress. The plate was subjected to a prescribed uniform axial displacement along the upper boundary. Computation has been performed for various combinations of the material constants, and the following results were obtained. ( 1 )Development of the plastic region is affected not only by the yield stress but also by the elastic modulus. ( 2 )Stress concentration of tensile stress at the edge disappears as yield region extends. ( 3 ) Stress concentration of shear stress at the edge is expressed by |γ|=Kr^-λ under general yielding, and the value of λ takes a constant (λ_p) for a wide range of the deformation stage. ( 4 )The value of λ_p is affected by the ratio of Young's moduli when the ratio of the yield stresses is small.

View full abstract

This paper deals with applications of the stress and displacement field of twist disclination dipole in the infinite isotropic medium to an analysis of the St. Venant torsion problem by means of indirect boundary element method. The disclination dipole has a singularity at its center so that we introduce the method of fictitious boundary in order to avoid the reduction of accuracy of the analysis which is caused by the singularity. Consequently, it is found that we can obtain sufficient accuracy of the analysis if we choose 3 or more for the coefficient of the distance of the fictitious boundary, and it is ascertained that the fields of the twist disclination dipole can be applied to the analysis of the St. Venant torsion problem.

View full abstract

The residual stress in silicon substrates after thermal oxidation was discussed experimentally. Microscopic Raman spectroscopy was used for the stress measurement. It was confirmed that the peak position of the Raman spectra shifted linearly with existing stress. The Ar⁺ laser beam of 1μm diameter was radiated into the silicon substrate and the back scattered light was detected by a photon multiplier. It was found that tensile stress occurred near the silicon surface after the plain oxidation. The stress increased with lowering oxidation temperature and thickening oxide film. However, complicated stress change occurred in the silicon substrate after the local thermal oxidation. The stress change was explained considering the curvature change of the Si/SiO₂ interface during the local thermal oxidation.

View full abstract

Cavitation erosion of a SiC-whisker-reinforced aluminum composite (FRM) was studied by vibratory erosion tests in ion-exchanged water, and SEM observations of the eroded surfaces and erosion particles. The FRM shows fairly good erosion resistance. The FRM made by a high-pressure infiltration process is superior to that by a powder metallurgy process. The erosion resistance of the former shows good correlation with H²_B/E (H_B : Brinell hardness, E : elastic modulus). The FRM in which the volume fraction of the whiskers is more than 30% shows almost the same resistance as high-carbon steel and stainless steel. These behaviors are because the whiskers prevent the plastic deformation of the matrix metal and the growth of eroded cracks, resulting in decrease of the size of erosion particles.

View full abstract

This paper formulates a method for creating an adaptive truss structure composed of variable stiffness members. In the method, an optimization law is applied to make a decision on the basic structural composition. The optimization law is called the adaptive region maximum law, and if maximizes the function region of the adaptive structure. The formulation is presented to make a decision on the adaptive truss structure. Moreover, its efficiency is explained with three examples of statical determinate and indeterminate truss structures.

View full abstract

Examinations of the statistical characteristics of corrosion damage generated in heat exchangers with mild steel tubes and the effects of operating conditions on them were investigated. As a result, types of setting and cooling water paths indicate close correlation to the corrosion damages generated from the cooling water side. Heat exchangers, which have vertical settings and which are cooled by water from the shell side, show inferior characters in corrosion damage to those which have horizontal settings and are cooled from the tube side. Further, they show remarkable scatters in renewed life distribution.

View full abstract

It is shown that Weibull statistics of strength may be extended to consider the random in-homogeneity of defects in volume (or area or length) of brittle materials. A new statistical distribution function is obtained, which exhibited good statistical characteristics, and is in better agreement with the experimental data for several kinds of carbon materials.

View full abstract

Some ceramic materials after proof testing have lower strength than the proof stress because of stress corrosion cracking behaviour. Therefore, estimation of the strength degradation during proof testing is quite important for proof testing on ceramics components. In this paper, we first made clear the meaning of the minimum strength of the specimens after testing, then a Statistical technique was proposed to estimate the actual minimum strength on a Weibull plot. Proof tests were also carried out on alumina material which has stress corrosion cracking behaviour by moisture. Four-point flexure strength was measured in water before and after proof testing. The results showed that the strength degradation during proof testing could be represented by a two-parameter Weibull distribution, and that the minimum strength after proof testing was estimated on the Weibull plot.

View full abstract

An efficient formulation for a 4-node shell element which includes the effect of large rotation increments is presented. The formulation is based on the MITC element proposed by Bathe et al. A complete and symmetric tangent stiffness matrix is obtained by considering up to the quadratic term of Taylor expansion of a finite rotation tensor. Several numerical examples are demonstrated to show the superior convergence by the present formulation compared with the conventional MITC formulation which assumes infinitesimal rotation increments. It is also shown in sensitivity analysis that accurate gradients are always obtained by the complete tangent stiffness, although erroneous gradients can be obtained by the conventional ones.

View full abstract

Microcrack growth behavior under axial and torsional loading at 550°C was observed. Cracking behavior was correlated with the results of microdamage observations using several strain parameters. From the microdamage observation, micro-cracking of oxide film at the crack tip in the strain localized region seems to be a dominant factor of micromechanism of fatigue crack growth. The maximum principal strain, which is the linear summation of the shear strain and the normal strain perpendicular to the crack path, has been identified as an important parameter in multiaxial fatigue at high temperature. A good correlation was obtained between the principal strain range and the crack growth rate.

View full abstract

Stresses around two coplanar Griffith cracks in an elastic layer sandwiched between two dissimilar elastic half-planes are determined. The self-equilibrated system of pressure is applied to the surfaces of the cracks. Application of the Fourier transform technique reduces the problem to that of solving dual integral equations. To so1ve the equations, the differences of the crack surface displacements are expanded in a series of functions which are automatically zero outside of the crack. The unknown coefficients in the series are solved by the Schmidt method. The stress intensity factors are calculated numerically for the layer of the epoxy resin sandwiched between the upper half-plane of the ceramic and the lower half-plane of the steel.

View full abstract

Rotating bending fatigue tests were carried out smooth specimens of ferrite-base and mainly pearlite-base spheroidal graphite cast iron (FDI and PDI, respectively) at room temperature. The statistical properties of initiation and propagation lives of surface cracks and fatigue life were discussed in detail from results of successive observations of specimen surface. The distrubution of the initiation life N_i, the propagation process life N_p1, N_p2 and the fracture life N_f, determined in this study, were well represented by a three-parameter Weibull distribution. The coefficient of variationηof those distributions decreased with increasing crack length 2_α during the fatigue process. The corrrelation coefficient in order Z of N_i, N_p for N_f increased with increasing 2_α, and in the lower stress level the Z reached about 1.0 in the early stage of fatigue (2_α=700μm).

View full abstract

To examine the effect of a small defect on the high cycle fatigue strength of spheroidal graphite ductile iron, rotating bending fatigue tests were performed. To simulate surface defects small artificial holes were introduced on specimen surface. A geometrical parameter, the square root of the projected area (√(area)) proposed by Murakami, was applied to estimate quantitatively the effect of the defect. From the relationship between the fatigue limit and √(area), critical values of √(area) were about 50μm and 100μm for the ferritic and the pearlitic material, respectively. The length of the nonpropagating crack observed at the fatigue limit was about twice as large as the size of spheroidal graphite. The concept an equivalent half-ellipse in calculating √(area) gave a reasonable prediction of the fatigue strength of plain specimens. agreement with the experimental results.

View full abstract

The authors have proposed an evaluation method for thermal fatigue strength, by which the thermal fatigue strength of Al-CFRP bonded joints at low temperature can be well explained by the maximum amplitude of thermal stress in the adhesive layer. In this paper, the stress distributions along the adhesive layer of an Al-CFRP bonded joint have been analyzed by a finite-element method code. Then an experimental method of mechanical fatigue in low-temperature environment has been developed to try to estimate the thermal fatigue strength of an Al-CFRP bonded joint effectively. It has been found that the strength of the thermal fatigue experiments coincides with that of proposed mechanical fatigue experiments, whose environment temperatures are fixed as the average temperatures of the thermal fatigue experiments.

View full abstract

In order to make quantitative assessment of the subcritical crack growth caused by stress corrosion cracking (SCC) of Ni-base alloy 600 in BWR and/or PWR water environment, cracking behavior was characterized by a slow strain rate test (SSRT). The effects of both sensitizing heat treatment and the crevice on critical cracking potential have been made clear. The sensitizing heat treatment of 650°C for 30 min lowered the critical potential by about 200 mV. In addition, the crevice which was introduced on the specimen surface by a grain boundary etching technique further dropped the critical potential by about 200 mV.

View full abstract

In X-ray stress measurement, the stress value is obtained from the product of the stress constant K and the slope M of the sin²ψdiagram which represents the lattice strain. Structural carbon steels of JIS-type S45C were quenched and tempered at various temperatures, and their 95% confidence limits of the stress constant were determined by measuring lattice strains of the specimens subjected to various applied stresses in order to determine the stress of steels accurately. The stress constant increased only slightly with increasing tempering temperature below approximately 773 K (500°C). At higher temperature, it increased rapidly. This remarkable increase in the stress constant is closely related to the microstructure change of the specimen ; above 773 K, the tempered martensite recrys-tallized to form the mixure structure of the ferrite matrix, and the spheroidized cementites which were growing to form laminar cementite plates at higher temperature.

View full abstract

General equations are derived for calculating the resolved shear stresses (RSS) on any slip system during arbitrary homogeneous slip deformation of single crystals in tensile tests and compression tests. RSS can be evaluated by the equations (ERSS) from shape and orientation (or initial shape, initial orientation and the change in shape) of the crystal specimens and the applied load. Much misunderstanding and confusion in the literature about the conventional ERSS is elucidated:e. g., ( 1 ) the conventional ERSS for tensile tests is the same as the ERSS for ideal tensile tests, wherein both stress and slip deformation are homogeneous; ( 2 ) the conventional ERSS for compression tests coincides with the equation to calculate volume average of RSS for frictionless compression tests, in which the load is perpendicular to the end-faces of the cylindrical specimen and the stress is usually inhomogeneous; ( 3 ) we can suppose ideal tests in compression, for which the same ERSS holds as in tension and, when single slip occurs, the loading axis rotates in the direction opposite to that in tension.

View full abstract