A bimaterial problem of strips which are bonded at their ends is solved. The bonded part is one part and debondings occur on both sides of the interface. The debonding lengths are changed, and concentrated loads are applied at the tips of each strip. The complex variable method and a rational mapping function are used for the analysis. Stress distributions are shown for two different locations of the interface. Stress intensity of debonding is defined and the values are obtained for various debonding lengths. Debonding and the effect of material constants on debonding are also investigated.

View full abstract

Stress fields in transistor structures are analyzed with consideration of the internal stresses of thin films. Internal stresses of amorphous silicon and tungsten silicide films are measured by detecting changes in the surface curvature of film-covered substrates as a function of temperature. Internal stresses of both films change upon annealing due to phase transitions, and reach about 1000 MPa. The stress predicted for transistor structures without considering the internal stress of the films differs markedly from results obtained using microscopic Raman spectroscopy. On the other hand, the stress predicted with consideration of film internal stress agrees very well with measured data. Stress design is performed for an actual transistor structure by adjusting the annealing temperature depending on the internal stress of an amorphous silicon thin film to eliminate the generation of dislocations. It is confirmed that stress design is effective in eliminating dislocations in transistor structures, thus improving device reliability.

View full abstract

A solution is presented for the stress-wave response of a partially transparent finite elastic circular plate with a penny-shaped crack subjected to impulsive electromagnetic radiation. The radiation is assumed to occur at a constant rate for the duration of the pulse, to be deposited with a radial Gaussian distribution and to diminish exponentially with distance from the exposed surface of the plate. The development of the analysis is based on the equations of uncoupled dynamic thermoelasticity with heat conduction neglected. The numerical procedure employs explicit finite difference approximations with second-order accuracy based on the integration of the governing equations along the bicharacteristics. Numerical calculations are carried out for the dynamic behavior of the thermal stresses and the stress intensity factors, and the results are shown in figures.

View full abstract

A finite element formulation using the penalty function method to analyze exactly the junction of plate and shell built-up structures is suggested for an isoparametric shell element. The connectivity condition at the junction is added to the potential energy functional by the penalty parameter and the interpolating function of displacements. This formulation yields an integral-type stiffness matrix of the special junction elements, which can directly evaluate the surface tractions at the junction. The suggested technique is applied to the stress analyses of isotropic and laminated plates with several types of stiffeners, and the validity of the technique is discussed.

View full abstract

We discuss the stress concentration problem of an elastic solid with an elastic circular-cylindrical inclusion under tension. A method of solution is developed for the above problem using fundamental solutions of axisymmetric problems of elasticity. The fundamental solutions are defined as solutions for the problem of an elastic solid subjected to axisymmetric body forces acting along a circle. Through numerical calculations, the influence of the length of the elastic circular-cylindrical inclusion on the stress distribution around the inclusion and on the central section is investigated. The influence of the share modulus of elasticity on the stress distribution is also shown.

View full abstract

We describe the complementary relationship between stress and displacement for nonlocal elasticity. This relationship is derived from the definition of nonlocal stress and the equilibrium equation. The stress solutions in an infinite plane subjected to a uniform compressed load are given. It is assumed that the nonlocal effects of the matrix and the inclusion are the same, and these effects cross the boundary between the matrix and inclusion. The stress distribution around a hole is drawn graphically using this solution. One of the effects of nonlocal elasticity is the development of a stress relaxation zone around a hole.

View full abstract

Thermal residual stress distributions on the interface and in the vicinity of the intersections of the interface and the free surfaces of bonded dissimilar materials are calculated using the boundary element method. Thermoelastic constant stress terms are calculated using Airy's stress function. The thermal residual stresses, when the thermoelastic constant stress terms are subtracted, show free-edge stress singularity. The values of the order of thermal residual stress singularity and stress distribution functions agree well with the theoretical ones calculated based on Airy's stress function. It is shown that the thermal stress singularity disappears for certain ranges of wedge angles of a pair of materials, as predicted.

View full abstract

The discretization of a whole domain is a prerequisite process for Finite Element Method (FEM). Since manual mesh generation is tedious, mesh generation schemes have been studied in recent years. In iterative analyses, the geometry of the domain is drastically changed by each iteration. It is known that mesh regeneration eliminates mesh distortion caused by the geometry change, which greatly affects the convergence of the whole analysis. If error control in the analysis is additionally considered in the mesh regeneration, such a process should be adaptive ; adaptive remeshing is required for a reliable solution. There are many mesh generators available, however, most of them require interaction with the user, and thus cannot be used for adaptive remeshing. In this paper, a sophisticated mesh generation scheme for the adaptive remeshing in the 2D FEM is discussed in detail. After comparison with other schemes, it is concluded that our scheme is superior to the others in terms of flexibility and time complexity.

View full abstract

This paper is concerned with a numerical technique based on the method of characteristics for three-dimensional dynamic thermo-elastic/viscoplastic problems. A constitutive model covering a wide range of strain rates and a wide range of temperatures, proposed by Tanimura, is used. As a numerical example, the three-dimensional stress wave propagation in a thermo-elastic/viscoplastic bar of square cross section subjected to both an impact loading and a thermal shock is presented. The stability and convergence of these numerical solutions are examined by checking the error in the total energy of the system.

View full abstract

An adaptive finite element technique for thermal stress analysis of built-up structures has been developed. A finite element formulation for a triangular membrane element and a new plate bending element, used for modelling such structures under both mechanical and thermal loadings, is presented. The associated finite element matrices have been derived in closed form. The performance of the new plate bending element is evaluated for a plate with temperature gradient through its thickness by comparing the predicted solution with the exact solution. The effectiveness of the adaptive meshing technique combined with the finite element method is evaluated by thermal stress analysis of a built-up structure with intersecting panels. The application demonstrates that the adaptive meshing technique can provide an accurate solution with fewer elements and shorter computational time than the standard finite element procedure.

View full abstract

In our previous paper, an approximate solution to the main crack-microdefect interaction model was derived using a main crack stress field, and the effect of a microdefect ahead of a main crack on the strength of solids was discussed in terms of the model. In the present paper, in order to investigate more precisely the above effect, the main crack-microdefect interaction model is formulated more specifically based on the method of continuously distributed theory of dislocations, and the distribution functions for both a main crack and a microdefect are obtained. As a result, we obtain stress intensity factors K both at the tip of a main crack and a microdefect in the closed form. Using these K values, we elucidate the effective range of the above model and the crack shielding effect by a microdefect. Furthermore, the crack length dependence of the fracture strength and fracture toughness of engineering ceramics can be explained well theoretically using the present model.

View full abstract

The macroscopic plastic deformation of polycrystalline metals, as revealed by features such as yield surface under multiaxial stress or the normality rule, is related to the microscopic slips in grains. In the present paper, a rate-type constitutive equation and a constant stress model of polycrystals are adopted and the equal-strain-rate curves under biaxial stress, which are equivalent to the yield curves in plasticity, are studied. The rate-sensitivity exponent in the rate-type constitutive equation is closely related to the number of active slip systems on the yield curves. Hence, it is possible to examine the effect of the number of active slip systems on the yield curve. The yield curves of fcc single crystals as well as fcc polycrystalline metals are calculated. The shape of the obtained yield curve is dependent on the number of active slip systems as well as the distribution of strain in polycrystals. The direction of strain rate vectors is also discussed. A new model of polycrystals called the "constant maximum shear strain rate model" is proposed.

View full abstract

In order to study the effects of dissolved hydrogen atoms and hydrogen damage on elongation of mild steel in relation to the susceptibility of iron and steel to hydrogen embrittlement, tensile tests were carried out immediately after cathodic hydrogen charging and after degassing of hydrogen at room temperature. Because cathodic hydrogen charging caused internal damage, such as microcracks (blisters) and plastic deformation owing to hydrogen precipitation, the influence of the dissolved hydrogen on the elongation was distinguished from that of hydrogen damage. The dissolved hydrogen and hydrogen damage were found to cause reversible and irreversible reductions in elongation, respectively.

View full abstract

In this study, a new method for evaluating the adhesive strength of the interface between resin liner film and cement mortar has been developed. The peel test for vinyl ester resin liner films on cement mortar surfaces was carried out. The critical energy release rate (G_c value) was obtained from the peel load and crack length measured during the peel test. G_c values were not affected by the crack length or the thickness of resin liner film. Therefore it was established that G_c values could be used to evaluate the adhesive strength of resin liner film. Furthermore, the effect of primer coats on the adhesive strength of resin liner film was revealed by obtaining the G_c values.

View full abstract

Both mullite/SiC (0.27μm and l.20μm) and mullite/TiC composite ceramics were prepared by hotpressing at 1650°C under 35 MPa for 4h. Room-temperature bending fracture stress, Young's modulus, Vicker's hardness and fracture toughness were investigated as functions of SiC and TiC volume fraction (0-20%). Grain growth of mullite was prevented by the existence of dispersed particles (SiC, TiC) in the matrix. As a result, bending fracture stress of both mullite/SiC and mullite/TiC composite ceramics was improved. In the case of the mullite/SiC system, bending fracture stress inceased with increasing SiC content and showed a maximum value of 604 MPa at 20vol%, which was about 80% higher than that of monolithic mullite. On the other hand, fracture toughness of mullite/TiC ceramic composite was observed to incease from 2.65 to 3.9 MPa√(m) with the addition of 20vol% TiC. Correspondingly, the bending-fracture stress increased from 330 to 410MPa. The strengthening mechanism of thermal treatment in air was also investigated for mullite/SiC composite ceramics and it was concluded to be useful for increasing bending fracture stress. Detailed reseach on the microstructure showed that the Hall-Petch relationship was satisfied for grain size and bending fracture stress.

View full abstract

In this paper, we present a probabilistic theory of propagation of ultrasonic waves in porous ceramics, and propose a new method of ultrasonic inspection for nondestructive pore characterization. The idea is based upon the arrival probability of ultrasonic rays. When incident rays impinge on a pore, they travel around the pore surface and increase the propagation time. We studied this process probabilistically, and found that the propagated waveforms can be expressed as Gaussian functions. The Gaussian waveform is determined by the porosity, pore size and pore shape. This new finding led to the following important laws. (1) The delay time of an ultrasonic wave passing through porous ceramics is proportional to the porosity. (2) The pulse width of the wave increases with increasing mean pore size. (3) The amplitude of the wave decreases with mean pore size. (4) The delay time and pulse width of the wave increase as the mean pore perimeter increases. Formulae for these relationships between ultrasonic and pore characteristics were derived, and an ultrasonic method for evaluating porosity and pore size was proposed.

View full abstract

We present a numerical analysis and results using the traction method for optimizing domains in terms of which linear elastic problems are defined. In this paper we consider the application of the traction method, which was proposed as a solution to domain optimization problems in elliptic boundary value problems. The minimization of the mean compliance is considered. Using the Lagrange multiplier method, we obtain the shape gradient functions for these domain optimization problems from the optimality criteria. In this process we consider variations in the surface force acting on the boundary and variations in the stiffness function and the body force distributed in the domain. We obtain solutions for an infinite plate with a hole and a rectangular plate clamped at both ends.

View full abstract

We report topological optimization of elastic frames by means of the boundary cycle used in algebraic topology and a genetic algorithm. In this study the optimum frame is defined as that in which the deformation at a point is minimal for a given weight limit. Members that have a tip which is not connected to other members, and increase the weight without making any mechanical contribution are neglected. The optimum topology is identified efficiently using a boundary cycle which yields a one-dimensional simplicial complex with no tips, satisfying the topological condition of no idle tip. The boundary cycle is derived from a chain and boundary operator, which plays the important role of decoding the genotype into the phenotype in the genetic algorithm, and is included in the string used in the genetic algorithm to represent the frame topology. The numerical examples are concerned with minimization of the deformation of two-dimensional frames subject to bending, and three-dimensional frames subject to torsion or expansion analyzed by the finite element method.

View full abstract