Some kinds of functionally gradient materials have been produced in our laboratory by hot isostatic processing (HIP), using powders of copper and molybdenum themselves and their mixtures for the purpose of thermal stress relaxation. Then, both elements have remarkably difference in thermal conductivity. This paper refers to the mechanical and fatigue properties of products using powders of copper and molybdenum themselves and their mixtures. The functional gradient of one of the above products had been observed by an optical microscope and micro-Vickers hardness distribution. In addition, the static tensile and fatigue properties of the mixture exhibited a lower value as compared with those based on the rule of mixture. It can be seen under observation by a scanning electron microscope (SEM) that the specimens of copper and molybdenum show typical ductile fracture surfaces, while those of mixtures exhibit brittle ones.

View full abstract

Fatigue threshold for SUS304 stainless steel containing 0.2 weight percent dispersed lead particles was about 40% higher for the host steel at test temperatures above 200°C. This was explained by considering that dispersed lead particles could control or arrest the fatigue crack growth. In this study, using an AFM/STM hybrid system which enables one to obtain topographic and current images at the same time, fatigue surfaces for SUS304 steel with lead particles were examined to clarify the controlling mechanisms in the fatigue crack growth. Topographic and current images showed that the oxide layer spalled at the fatigue crack tip was repaired by the molten lead at test temperatures higher than 200°C. Accordingly, we concluded that dispersed lead particles controlled the fatigue crack growth at elevated temperatures.

View full abstract

To determine some of the microstructural parameters governing the low cycle fatigue strength and the crack propagation behavior of ferritic ductile cast irons, an experimental analysis of fatigue damage has been undertaken using four materials with different spheroidal graphite morphologies. It is shown that the frequency of the large graphite grain in the specimen surface layer strongly controlled the fatigue strength and the crack propagation behavior. In detail, the fatigue strength is mainly affected by the volumetric distribution of the spheroidal graphite, while the spheroidal graphite distribution in the surface layer played an important role during the microcrack propagation stage.

View full abstract

Low cycle fatigue tests were performed for 316 stainless steel (316SS), alumina-dispersion-strengthened copper (DS Cu) base metal and 316SS-DS Cu diffusion-bonded joints. Plastic strain distribution of joint specimens during fatigue was measured to estimate deformation of the joint specimens. Furthermore the microstructure of the diffusion zone of the joints and fracture surface of the fatigue specimens were examined with scanning and transmission electron microscopes. Intermetallic compounds and recrystallization of DS Cu were observed near the interface in DS Cu. The fatigue life of the joints was lower than that of DS Cu. The joint specimens fractured in the DS Cu near the interface in the case of low strain range fatigue tests. For high strain range fatigue tests, however, the fracture point was in the DS Cu 6∼7 mm from the interface, at which the maximum plastic strain occurred.

View full abstract

The influence of pre-crack length on an R-curve for ceramics with bridging is investigated by numerical simulation. Generally a pre-crack is introduced into the specimen for measurement of the R-curve. However, it is considered that the R-curve is influenced by the pre-crack length because the bridging stress distribution depends on the pre-crack length. In this study, R-curves are obtained by numerical simulation. In the numerical simulation, we establish an algorithm to evaluate the R-curve based on bridging stress. This algorithm can be used to estimate the characteristics of bridging through a certain R-curve obtained experimentally. Based on the results, it is shown that the R-curve for ceramics with bridging depends on the pre-crack length and the specimen geometry.

View full abstract

In this paper we describe an experimental study of the CTOD under impact loading in an attempt to clarify the dynamic fracture mechanism of mild steel. The drop test apparatus and air-gun exprimental apparatus, which allowed measurement of the displacement rate of the load point, were developed by the authors, and were used in an impact three-point bend test to investigate the influence of displacement rate on CTOD under various loading speeds. Three-dimensional FEM calculations were performed to compare the experimental results with the simulation results. Additionally, CTOD-Δa curves and values of critical CTOD were determined through three-point bend tests under three different displacement rates. Finally, influences of displacement rate on CTOD, critical CTOD and fractured surface pattern of specimens were discussed based on the experimental results, calculated results and observation of the fractured surface. It was found that the CTOD values and the critical CTOD values decrease with increasing displacement rate.

View full abstract

In this study, we attempt to develop a more efficient method of thermal fatigue life evaluation for solder joints than the thermal cycle test. The mechanical fatigue tests for solder joints of surface mount devices are performed under several temperature and cyclic frequency conditions. In addition, thermal cycle tests are carried out in order to compare the results with the mechanical fatigue test. On the other hand, the strain range of solder joint in the mechanical fatigue and thermal cycles tests is evaluated by using the elasto-plastic finite element method, considering the effects of temperature and strain rate on the stress-strain curve of solder. From the experimental and analytical investigations, it is found that the relation between the elasto-plastic strain range and the fatigue life of the mechanical fatigue test agrees well with that of the thermal cycle test. Furthermore, an evaluation method of thermal fatigue life employing the mechanical fatigue test is proposed on basis of the investigations mentioned above.

View full abstract

The use of the metal matrix system in advanced composite has attracted much attention for applications in high-temperature environments. The behavior of the metal matrix composite is very complex as the metal matrix undergoes plastic deformation at elevated temperatures. Moreover, the residual stress and interfacial property degradation affect the behavior of the composite. An elasto-plastic analysis is made to study the mechanical behavior of a metal matrix fiber composite with damage on interface and residual stress. A homogenization theory with multiscale asymptotic expansion is used to describe a homogenized constitutive equation of the composite. The interface which consists of separate and contacted zones is modeled as a spring layer. A boundary integral equation method is applied to obtain perturbed displacement increments for homogenization. The elasto-plastic behavior of the metal matrix composite at elevated temperatures is evaluated from the homogenized equation with the perturbed displacement increment. The numerical results are compared with the experimental results of the metal matrix composite.

View full abstract

Monotonic compressive loading tests and relaxation tests are conducted on high-density polyethylene (HD-PE) at the strain rates of 1.1×10^-1 to 1.1×10^-5s^-1 and temperatures of 10, 25 and 40°C. The observed stress-strain responses are similar to those for steels at elevated temperature. The material constants of the viscoplasticity model based on overstress is determined for HD-PE at every temperature and the model is shown to successfully simulate such rate-dependent phenomena numerically. The examination of the stress-strain curves suggests that the strain-rate and temperature effects on the deformation curves can be expressed using a unified parameter. The viscoplastic behavior observed in the strain-rate cycling test at 25°C is reproduced in the similar test at 40°C designed for this parameter, and the phenomenon is verified in the numerical experiment by the overstress model.

View full abstract

The dynamic stress-strain curves for Pb-Sn alloys have been obtained by a longitudinal impact test of tapered bars with rectangular cross section. The history of stress was estimated using an elastic input bar (a stress bar), while the corresponding plastic strain was measured directly using a strain gauge on the test bar. The relationship between the dynamic overstress and the plastic strain rate was derived from these data of the history of stress and strain. Thus the constitutive equation of viscoplastic type was described by a power-law relationship between the foregoing overstress and strain rate. The problems corresponding to the longitudinal impact tests were analyzed numerically using the estimated constitutive equation. The numerical results were in good agreement with the experimental results.

View full abstract

In the present paper, the experimental estimation of dynamic plastic bending moments for metal strips is investigated. The three-point bending test under impact and static loads is applied to aluminum alloy (JIS A 6063 S) and mild steel (JIS SS 400). It is confirmed that the dynamic bending deformations in the three-point bending test can be modeled as a plastic hinge, the experimental results show that the absorbed energies of the specimens are proportional to the bend angles. The ratio of the absorbed energy to the bend angle is approximately equal to the plastic bending moment. In the case of aluminum alloy, the dynamic bending moments for the different average bend angular rates coincide with the static bending moments. On the other hand, in the case of mild steel, the dynamic bending moments are proportional to the average bend angular rates. These results show the strain rate effect for mild steel and aluminum alloy which are known well. As a result, we confirm that the present method based on the plastic hinge model and the absorbed energy is efficient for determining the dynamic plastic bending moment.

View full abstract

This paper presents the analytical solutions for an anisotropic medium containing an elliptical cavity or inclusion subjected to uniform out-of-plane loads at infinity, concentrated force, screw dislocation or dipole force at an arbitrary point. An elliptical inclusion can be assumed to be rigid or elastic. For the case of elastic inclusion, some sliding between the matrix and the inclusion at the elliptical boundary is assumed.

View full abstract

An ultrasonic reflectivity measurement technique is applied to anisotropy detection within a small area of material. The ultrasonic reflectivity is microscopically measured using an ultrasonic microspectrometer with spherical-planar-pair lenses. The acoustic reflection coefficients of small area on a water-cubic (111) interface were measured for an incident wave at an angle θ to the normal and φ to some direction lying in the (111) surface. The reflection coefficients measured as a function of θ and φ were in good agreement with the theoretical calculation. This reveals that the microscopically measured reflection coefficient is reasonable and can be used as a promising detector of anisotropy in a small area. Furthermore, a mean reflection coefficient is defined for quantitative evaluation of elastic anisoropy in a small area of polycrystalline materials. It has been confirmed that, by performing numerical and experimental examinations with texture-controlled steels, the reflection coefficient measured in polycrystalline materials corresponds to the mean reflection coefficient.

View full abstract

In this study, the nonlinear thermal bending problem is developed for a nonhomogeneous beam. Assuming that the beam is composed of the medium with nonhomogeneous thermal and mechanical material properties in the transverse direction, the transient heat conduction problem for such a nonhomogeneous beam is analyzed making use of the methods of finite sine transform and Laplace transform, and the analytical results are obtained under the condition that heat transfer between the surrounding media occurs at the upper and lower surfaces of the beam. Thereafter, the associated nonlinear thermal deformation fields are analyzed theoretically making use of the von Karman deformation theory under the assumption of the Bernoulli-Euler hypothesis. As for the nonlinear response of the beam, the stationary principle of total potential energy is employed, and the numerical simulations are carried out to elucidate the nonlinear thermal bending behaviors of the beam under the clamped edge condition, making use of the calculation program for the nonlinear search procedure. The results are shown in the figures, and the influences of geometrical nonlinearty, the nonhomogeneous material properties, and condition of heat transfer at surfaces of beam are discussed briefly.

View full abstract

This paper describes theoretical treatment of problems of thermal stress and thermal bending due to partially distributed heat supply for a multilayered anisotropic laminated plate. As an analytical model, we consider a laminated rectangular plate consisting of a diagonal stack of layers having orthotropic material properties, i. e. an angle-ply laminate. Introducing the Rayleigh-Ritz method based on the variational principle, a three-dimensional temperature solution for a transient state is given. Moreover, for the thermoelastic problems, the components of displacements are analyzed using the Rayleigh-Ritz method, and the characteristic behaviors of thermal stress and thermal deformation for a simply supported plate are determined. As an example, calculations are carried out for a 2-layered angle-ply laminate, and the results are examined precisely.

View full abstract

The tensile creep tests were performed on sintered silicon nitride ceramics at 1300°C, using the tensile creep specimen with the four projections which work as targets for the laser beam. Then the creep displacement was measured by means of the laser-beam-type displacement measuring system. The creep curve of the ceramics exhibited the steady creep period following the transient crrep period. The relationship between minimum creep strain rate ε_c and nominal stress σ was expressed by the straight line on double logarithmic scale. The creep deformation occurring in the specimen with the projections was calculated through the finite element method, where the measured ε_c-σ relationship was introduced as the basic material constitutive equation. The relationship between ε_c and σ calculated from the displacement between the two projections through FEM analysis was in good agreement with the above material constitutive equation. It was suggested from this and the FEM calculation on the progress of the creep deformation that the ε_c-σ relationship obtained from he displacement measured between the two projections expressed the correct creep property of the ceramics.

View full abstract

Residual stress distribution of Si₃N₄/SUS304 joints with an insert layer of copper was analyzed by two- and three-dimensional thermo-elastic-plastic finite-element methods (FEM). Analytical results were compared with experimental results of an X-ray diffraction method. It is found that the 3D-FEM results show good agreement with the experimental results, although they differ from the 2D-FEM results. The superposition principle (mean value) of 2D-FEM results was applied to estimate the 3D-FEM results. The superposition results (mean value) show good agreement with the 3D-FEM results, except for the results near the interface. Stress singularity near the interface at the ceramic side was also studied on the basis of 3D-FEM results, 2D-FEM results and superposition results. The stress singularity of 3D-FEM results differs from that of 2D-FEM results at the edge line, although it shows a good agreement with that of 2D-FEM (plane strain) results at the center line.

View full abstract

In this paper, Genetic Algorithm is applied to the discrete optimization of a structure composed of composite material. The finite-element method is used for analysis of the structure. To improve the reliability and the efficiency of the simple GA, automatic linking of design variables, mutation with high probability, and reduction of the population size are proposed. The numerical example presented here will show that the efficiency of GA can be improved without sacrificing the reliability of GA by introducing these procedures.

View full abstract

The plane elasticity problem of an infinite anisotropic plate containing a circular isotropic inclusion is considered and Green's functions are derived for a point force and/or a dislocation acting at a point in the interior of the inclusion. The complex potential approaches of Muskhelishvili and Lekhnitskii are used to express the elastic fields in the isotropic material and the anisotropic material, respectively. The general solutions are obtained in a closed form.

View full abstract

The authors already discussed the crack tip singularities of various kinds of open crack meeting at the interfaces of different materials for plane problems. Based on these results, this paper is concerned with stress singularities near the tip of an arbitrary open crack yielding by two inclined slip planes. Using the eigenfunction-expansion method by Williams, the crack tip stress singularities for mode I are computed for plane problems. The theoretical process of the analysis and some numerical results are given.

View full abstract

Adhesively bonded joints will fail under comparatively low stress by addition of unexpected peel loads to shear loads. Thus, it is essential to evaluate the strength of adhesively bonded joints under the mixed mode conditions consisting of peel loads and shear loads. However, evaluation methods for thin adherends under mixed mode conditions have not been establised. In this study, a biaxial loading apparatus and multilayer specimens are developed to evaluate static strength and fatigue strength of adhesive joints. This method is applied in adhesive tests of metallic coatings. As a result of the evaluation tests, the static fracture standard can be described by a quarter-ellipse which is subjected to pure shear strength and pure peel strength. The fatigue strength can be predicted if the static fracture standards and the voids caused by peel loads are considered. The adhesion strength of galvanized steel under the mixed mode can be evaluated in liquid nitrogen.

View full abstract

The making-up process of the premium connections for oil or gas wells is generally controlled by the recommended make-up torque. In order to determine the recommended torque, an analysis method for the make-up torque with short computing time and good accuracy is required. In this paper, so as to obtain the contact load at the metal-sealed portion of the connection, the connection model of combined cylindrical shells was analyzed, considering the shrink-fit values at the threaded portion and the metal-sealed one of the connection. Then, the calculation method of the make-up torque just before the contact of the torque shoulder was obtained. As the torques calculated by the calculation method agreed well with those by FEM and actual tests, this method can be applied to determine the make-up torque of premium connections.

View full abstract

For a horizontal well or an extended reach well, which is promising for well construction in . future, a well casing is subjected to local wear of wall thickness on its inside surface due to rotational contact with tool joint or drill pipe during drilling. Collapse strength of worn casing is analyzed as an unstable problem by the elastic-plastic finite-element method (FEM) and the decrease ratio of collapse strength due to casing wear is formulated by nonlinear programming using parameters of the wear shape. The elastic-plastic FEM is fully applicable as values calculated by the FEM program agree well with those of experiments. It was found that the decrease ratio of collapse strength of a worn casing is linearly proportional to the decrease ratio of wall thickness in both modes of plastic collapse and elastic collapse, and that the apical angle of the wear shape has a slight influence on collapse strength.

View full abstract

Generally, measurements of the stress-strain curve are required in order to determine the yield strength. However, it is difficult to measure dynamic stresses induced in materials due to the complicated propagation of the stress waves. In contrast to this, measuring strain data is relatively easy. When the materials are subjected to impact force, elastic waves propagate first. After the strain becomes larger than yield strain, plastic waves occur. Therefore the strain rate changes discontinuously at this point, although the strain increases smoothly. By measuring the strain waves induced in a test piece rod subjected to longitudinal impact force as a step waveform, the strain rate profiles are obtained based on numerical differentiation. For the previous reason, the dynamic yield point can be estimated by change of the strain rate. The dynamic yield stresses are obtained by multiplying those values by Young's modulus measured from the static compress or tensile tests. This paper confirmed the possibility of the present method based on numerical simulations and discussed practicality by measuring the dynamic yield strength of materials under various strain rates.

View full abstract

This paper deals with optimization of truss topology using boundary cycle in algebraic topology. Elimination of unnecessary members from the ground structure, one of the popular means to optimize truss topology, is employed. The elimination has a disadvantage that unstable structures possibly appear in the process of optimization. Boundary homomorphism, which makes the boundary cycle from chain, is equipped with an important feature to represent equilibrium of internal forces in members. Design variables derived by the boundary cycle can always satisfy this equilibrium and avoid a category of unstable structures without imposing any constraint. Dimensions of members are dependent on each other and related mutually through the design variables. Numerical examples of minimizing the total weight of a plane truss are attempted under the condition that the distribution of strain energy density is uniform and equal to a certain value. The truss is fixed to a rigid wall and supports a vertical load acting at a point distant from the wall. The validity of this formulation is verified by the numerical examples.

View full abstract

Recently, various electronic printed circuit boards on which IC and LSI are mounted are required in personal computers (PCs), because of the widening applications of PCs, for example, in the area of education, business, art, and multimedia. To meet this ever-increasing demand, an efficient and convenient method of thermal stress analysis for the printed circuit boards with electronic devices is necessary. Almost all of the thermal stress analysis methods in conventional use employ computational techniques such as the finite-element method or boundary-element method. These methods are not adequate for the design of the electronic printed circuit boards because of their complexity. In this paper, we proposed an effective and convenient method for thermal stress analysis of the printed circuit board by formulating theoretical equations using the beam theory. The method is adopted to analyze thermal stress in a wire lead frame of an electronic device mounted on a printed circuit board, and the efficiency is discussed comparing results obtained by the method and finite-element analysis.

View full abstract

Since the computing environment with many workstations on networks has recently become available, these workstations can be regarded as a virtual parallel computer to perform finite element analyses. There are some problems such as the limitation of communication capacity for large-scale computing using a cluster of workstations on a network. The parallel performance of finite element algorithms, such as the Gaussian elimination method, the conjugate gradient method and the domain decomposition method, depends strongly on the parallel parameters. The purpose of this paper is to evaluate the parallel computing time of the algorithms and optimize the parallel parameters. The parallel computing systems are developed based on these techniques and applied to a large-scale problem with 350000 degrees of freedom using twenty workstations.

View full abstract