The fracture parameter of an interface crack has yet to be established even for an elastic crack. The CED (crack energy density) was proposed as a crack parameter which enables us to be free from the restrictions on the constitutive equation, and it has been shown that it is applicable, in a unified way, to various kinds of crack behavior in a homogeneous material. Therefore, the CED may also be applicable to an interface crack. In this paper, the fundamentals of the CED for an interface crack are studied. The concrete definition of the CED for an interface crack is given first, and it is shown that the CED can be divided into mode I and mode II contributions that can be expressed by domain integrals without any restrictions on the constitutive equation. Subsequently, the relationships to the conventional crack parameters of the CED are derived. Moreover, the substantial mode I and mode II contributions of the energy release rate are introduced, and their correspondence to the mode I and mode II contributions of the CED is discussed.

View full abstract

The fracture parameter of an interface crack has yet to be established even for an elastic crack. The CED (crack energy density) was proposed as a crack parameter which enables us to be free from the restrictions on the constitutive equation, and it has been shown that it is applicable, in a unified way, to various kinds of crack behavior in a homogeneous material. Therefore, the CED may also be applicable to an interface crack. In the previous paper, the fundamentals of the CED for an interface crack were studied theoretically. In this paper, the elastic finite element analyses of a bimaterial specimen with a center crack in the interface are carried out under tension, and it is shown that the CED and its mode I and mode II contributions can be evaluated from the corresponding domain integrals virtually independent of the radius of curvature ρ at the notch (crack) tip over the some range of ρ and, therefore, these quantities are promising as crack parameters of an interface crack. Moreover, the substantial mode I and mode II contributions of the energy release rate introduced in the previous paper are evaluated and the relationships of the mode I and mode II contributions of the CED to them are discussed.

View full abstract

The behavior of a surface crack crossing the fusion line is studied experimentally and numerically. By changing the aspect ratio of the surface crack, three kinds of surface-cracked specimens are tested. The distributions of Δa, CTOD and CTOA values along the crack front are measured. The experimental results show that the distribution patterns of these parameters along the crack front across the fusion line are not largely different from those of homogeneous specimens. Significant differences in CTOA values between the base metal and weld metal are recognized. By referring to the basic material test, three-dimensional FEM analyses are carried out. The effect of the fusion line on the J-integral distribution and CTOD are obtained and are compared with experimental results. The crack tip fields are compared with HRR solutions. It is found that the existence of the fusion line has no significant effect on the crack tip field.

View full abstract

This paper deals with the two-dimensional rolling-sliding contact problem with frictional heat generation on an elastic half space containing multiple surface cracks. Rolling-sliding contact is simulated as an arbitrarily distributed contact load with normal and shear components, moving with constant velocity over the surface of the half-space. The frictional heat generation on the contact region is estimated by the use of sliding velocity, frictional coefficient and contact pressure. Numerical results of stress intensity factors are obtained for the case of two parallel cracks due to Hertzian and parabolic distributed loadings. The interference effects on the stress intensity factors with the distance between two cracks, and the effects of the slide/roll ratio, frictional coefficient and crack angle on the results, are considered.

View full abstract

Characteristics of crack initiation and propagation in rotating bending fatigue tests were investigated in a Ti-15V-3Cr-3Sn-3Al alloy. Successive observations were made through the plastic replica method to clarify the physical meaning of fatigue limit, and the crack propagation behavior was studied using a small crack growth law. The main results are summarized as follows. ( 1 )When σ > σ_ω, microcracks appear at β single phase. ( 2 ) Crack initiation lives are smaller than 10% of the lives for the final fracture. Therefore, the total fatigue life is mainly determined by the crack propagation life. ( 3 ) When σ=σ_ω, no microcracks are observed on the specimen after 10⁷ cycles under a metallurgical microscope. Therefore, the fatigue limit in this material is determined by the limiting condition for the initiation of a microcrack. ( 4 ) The crack propagation rate is controlled by the small crack growth law; dl/dN=Cσⁿl(σ : stress amplitude, l : crack length, C and n : constants) .

View full abstract

The authors have developed an experimental and computational hybrid measuring system for the purpose of nondestructive detection of cracks embedded in structural members. The system consists of an infrared thermal video system (TVS-5000), by which the temperature distribution of a body surface can be measured and processed to express its thermal image, and an engineering work station (EWS), by which image processing of the thermal image can be carried out. Various methods for heating or cooling the specimen have been examined. Further more, an image processing expert system has been developed to obtain definite pictures of the defects.

View full abstract

In this paper, rotating bending fatigue tests were carried out to investigate the fatigue strength of plain and notched specimens of spheroidal graphite cast iron. The fatigue limit of the plain specimen depends on the propagating limit of micro cracks which are nucleated from the graphite or the nodule. The maximum length of the micro non-propagating cracks is nearly equal to 200 μm. Therefore, the fatigue strength σ_ω1 of the notched specimen is insensitive to the notch root radius. The fatigue strength σ_ω2 increases as the notch root radius decreases, because the micro non-propagating crack is filled with graphite and oxide. Moreover, it is found that the fatigue strength of an arbitrary notched specimen can be estimated on the basis of the linear notch mechanics.

View full abstract

In order to establish a quantitative evaluation method for the fatigue limit of materials with defects, a 4-parameter method, based on tensile strength σ_B, Vicker's hardness HV, threshold stress intensity range ΔK_th and maximum defect size a_max, was proposed and applied to three kinds of ductile cast iron and those austempered materials. As the result, the fatigue limit values estimated by the 4-parameter method agreed well with the experimental values.

View full abstract

The strength and the fracture morphology of cemented carbides in cyclic compression have been investigated through experiments and a finite-element analysis. Flanged cylindrical specimens with filleted shoulder of WC- (5 to 25 wt.% ) Co cemented carbides were considered. The number of cycles to crack initiation and to failure in compression fatigue of all alloy compositions, except for low-cobalt-content alloys such as 5 wt. % Co alloy, was well arranged by using the load ratio, which is a ratio of the applied maximum compressive stress to the static compressive strength. The low-cobalt-content alloy was split in two from the end surface, and the higher-cobalt content alloys, such as 15 to 25 wt% Co alloys, failed after stable crack growth around the root area. The difference in fracture morphology between alloy composition was discussed through the finite-element stress analysis.

View full abstract

Three-point bending and fracture toughness tests of alumina specimens with various grain sizes were carried out to discuss the effects of grain size on bending strength and fracture toughness. The bending strength increased with reducing grain size. The valid fracture toughness is independent of grain size. On the other hand, the apparent fracture toughness with slow crack growth increased with increasing grain size due to the R-curve behavior.

View full abstract

A simple void growth model in ductile fracture is proposed. The model is that an elliptic void grows under the plane strain condition in an infinite material, which has a rate-dependent constitutive equation of the exponential type. Assuming that the void maintains elliptic shape during the growth process, a simple analytical relationship is derived through the use of kinematical relationships of the elements. On this basis, the void growth process is simulated by simple calculations. The applied parameters are the triaxiality of uniform remote stress, the initial shape of the void (aspect ratio of the ellipse) and the exponent of the constitutive equation. It is shown that there is a very strong dependence of void elongation on the exponent of the constitutive equation, and that the initial circular void elongates in the perpendicular direction to the loading axis when the exponent is large enough. In general the growth rate and shape are determined by the combinations of stress triaxiality, exponent of constitutive equation and initial shape of void. These relationships are quantitatively obtained by the analysis.

View full abstract

Growth behavior of a 3-dimensional curved surface crack (arrow-headed crack) under lubricated rolling-sliding contact loading is analyzed by considering the stress intensity factors of mode I, II and III. K_σmax and K_τmax criteria in mixed mode fatigue fractures are used as the conditions for fatigue crack growth. The analytical results support the hypothesis that an arrow-headed crack in the pitting phenomenon first propagates in the shear mode, and tensile mode growth follows due to the hydraulic pressure effect on crack faces by the oil penetrated into the crack. It is also shown that the pitting shape at the contact surface is influenced by the magnitude of the frictional force and the ratio of the crack size to the Hertzian contact width. It is suggested that the application of fracture mechanics analysis to an arrow-headed crack under contact loading is promising for the prediction of the operating conditions in the experiment.

View full abstract

Oxidation behavior of silicon nitride ceramics (SSN-1, SSN-2, SSN-3) was investigated in air environment at constant high temperature and under a thermal cycle condition. Flexural strengths were measured for specimens before and after oxidation at room and high temperature. ( 1 ) The formation of pits was observed for SSN-1 exposed at high temperature. The large pits resulted in strength degradation. ( 2 ) High-temperature strength of SSN-2 after oxidation did not change due to the disappearance of cracks by melting of the oxidized layer. ( 3 ) Strength degradation of SSN-3 was 40 % after thermal cycle oxidation. The strength did not reduce beyond the accumulated exposure time of 20 hr.

View full abstract

This paper gives a finite-element formulation for structural analysis of hyperelasticity using rotationless strain, which can predict accurately the stress-strain curve in simple shear problems compared to the previous Green's hyperelasticity model. This study employs the displacement-based finite-element method due to its conceptual simplicity, even though recent finite-element codes use the displacement-pressure method. Numerical examples of plane block subjected to shear deformation show the performance of the present formulation as well as the fundamental features of mechanical responses involving combined loading conditions and the effects of the steel layers included in isotropic rubberlike materials.

View full abstract

The fast breeder reactor (FBR) uses liquid sodium as a primary coolant, and is characterized by operation at high temperature and under atmospheric pressure. Therefore, the main vessels are essentially required to cope with thermal stresses in the structural design. To reduce such stresses, it is necessary to make the vessel walls as thin as possible. This makes it indispensable to study the stability of reactor structures against buckling under earthquake loadings. However, it generally takes an enormous amount of computer time to execute dynamic buckling analysis by the finite element method. The purpose of the present paper is to describe the development of an elastic-plastic large-displacement analysis computer program for shells of revolution with high efficiency as a first step, which will be finished at final step of elastic-plastic dynamic buckling analysis including fluid-structure interaction.

View full abstract

Stress-relaxation behavior is studied in polypropylene samples subjected to different cyclic preloadings and to simple tension. The relaxation tests are performed under different sets of strain amplitude, number of cycles, and strain rate using a closed-loop, electrohydraulic, servocontrolled testing machine. The calculated stress-relaxation curves are determined from the constitutive equation based on an overstress theory, in which an equilibrium stress and a viscosity function are treated. The calculated results agree well with the experimental ones. It is concluded that the overstress theory explains the nonlinear viscoelastic-plastic behavior of polypropylene.

View full abstract

By our finite-difference method (FDM) based on an exact theory, impact problems of a 2-dimensional elastic-viscoplastic solid under step stress are analyzed. In our FDM code, the inner region is discretized by Lax-Wendroff difference schemes, and the boundary region is discretized by Butler difference schemes. Particle velocity and stress of 2-dimensional solid obtained by our FDM Code are compared with those obtained by finite-element-method (FEM) code DYNA2D. In an elastic problem, the particle velocity and stress obtained by FEM code fluctuate in spite of artificial viscosty more than those obtained by our FDM code. But in elastic-viscoplastic problems, the two codes are in a good agreement.

View full abstract

A simple instrumentation system is developed to determine the elastic moduli of engineering ceramics and their composites from the sound produced by the impact of a sphere on a specimen which is suspended by thin metallic wires or ceramic threads in a furnace. The natural frequencies of the flexural and torsional vibration modes are measured from the impact noise using a fast Fourier transform analyzer. Elastic moduli are then computed from these frequencies using Timoshenko's beam theory and also Saint-Venant's torsion theory. The validity of the measurement method is demonstrated by room temperature tests on ceramic and metallic specimens of known elastic moduli. Sialon and Al-Al₂O₃ composites are tested at elevated temperatures up to 800°C. It is shown that Young's and the shear moduli of the materials tested decrease with an increase in temperature, while Poisson's ratio remains almost constant.

View full abstract

As athletic skills improve and athletic equipment, such as baseball bats, tennis rackets, and golf clubs, is exposed to high-speed loading, the dynamic strength as well as static strength should be examined. In this article, baseball bats made of aluminum alloy which are supported elastically at their center of percussion are impacted by actual balls at various speeds and the induced strain waves and the dynamic deformations are observed. In order to conduct unified high-speed impact tests of baseball bats and other equipment, and to develop some new industrial materials to be used for athletic equipment. a high-speed gas gun to launch baseballs and many kinds of projectiles is constructed.

View full abstract

In the previous report, impact phenomena of baseball bats made of aluminum alloy and balls were examined. In order to clarify questions pertaining to the application of some new industrial materials to athletic equipment, the dynamic deformation and failure of cylindrical pipes made of carbon-fiber-reinforced plastics (CFRP) which are impacted by balls are studied experimentally in this paper. It is pointed out that the delamination of CFRP pipes precedes the destruction of carbon fibers and the strength of CFRP pipes is reduced abruptly even if the impact velocity is relatively low.

View full abstract

We make clear the properties of two methods for analyzing penetration into geological materials. One method is based on spherical cavity expantion theory (SCET method). The other method is based on computer simulation of 3-dimensional FEM impact code DYNA 3 D (DYNA 3 D method). Under the condition of a low-velocity region (below 16 m/s), we show the projectile accelerations of the SCET method to be good agreement with those of our drop experiment with sand target. Under the condition of a high-velocity region (over 212 m/s), we show that the projectile accelerations of the DYNA 3 D method are smaller than those of the other method (experiment or SCET method) at the early stage of penetration. The reason is that the resistance of the target is smaller for the small-hole target model of DYNA 3 D calculation at the early stage.

View full abstract

In this paper, the characteristics of the stress field for a V-notch, with its tip on the bimaterial interface are studied as a plane problem. The eigenequation, which gives the order of stress singularities, is given in an explicit closed form. Based on the eigenequation, the eigenvalues are given for various notches and material combinations.

View full abstract

This paper deals with application of the stress and displacement field of a wedge disclination dipole in an infinite isotropic medium to analysis of a two-dimensional elastic problem by means of the indirect boundary element method. The disclination dipole has a singularity at its center, and hence we introduce the fictitious boundary method in order to avoid the reduction in the accuracy of the analysis which is caused by the singularity. Consequently, it is found that we can obtain sufficient accuracy of 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 wedge disclination dipole can be applied to the analysis of the two-dimensional elastic problem.

View full abstract

It is well known that the stresses become singular at the edge point of the interface between dissimilar materials. The singularity has been reported by Bogy for the isothermal elastostatic case with stress free edges and also with the edge subjected to surface tractions. However the singularity for the thermal stress or the bonding residual stress has not yet been clarified. In this study, we fouud that the logarithmic singularity appears in the thermal or residual stresses at the edge of the interface similar to the isothermal problems subjected to discontinuous surface tractions at the edge point. By the boundary element analysis of residual stress for various combinations of dissimilar materials, it is verified that the logarithmic singularity exists for any combination of materials.

View full abstract

Stress singularities in different-material wedges with interlayer were studied by the finite-element method with unique bond elements, plane elements, which represented the interfacial stress. The results were as follows. Stress singularities in different-material wedges were successfully analyzed by the plane element method. As the effect of the thickness of the soft interlayer, the stress intensity coefficient, decreased with decreasing interlayer thickness. This was caused by restraining of the deformation of the interlayer near the surface by the other hard material.

View full abstract

The stresses of band adhesive butt joints, in which the interfaces were partially bonded, were analyzed using a two-dimensional theory of elasticity in order to establish the fracture criteria when the joints of dissimilar adherends were subjected to external bending moments. In the analysis, the dissimilar adherends and the adhesive were replaced with finite strips when the interfaces were bonded by an adhesive at two regions. In the numerical calculations, the effects of the ratios of Young's modulus among adherends and an adhesive, the thickness of the adhesive, the bonding area and the position on the stress distributions at the interfaces were demonstrated. As a result, it was seen that band adhesive joints were available when the bonding area and positions were determined taking into account the external load distributions. Experiments were performed on the strains of adherends and the joint strengths. Analytical results were consistent with the experimental ones.

View full abstract

An analysis method to evaluate the sensitivities in postbuckling problems is developed. In recent years, sensitivity analysis using the perturbation method has been applied to nonlinear static/dynamic problems. In postbuckling analysis, however, the load values at limit or bifurcation points are also unknown variables of which the gradients cannot be evaluated by the conventional sensitivity analysis. The new extended formulation can calculate the gradients of the load values as well as the displacement gradients under a constant 'arc-length' condition, which is generally used as a path-following method to find the equilibrium state beyond the limit or bifurcation points. A couple of buckling analyses of sheli structures are demonstrated to examine the validity of this method, and satisfactory results are obtained.

View full abstract

Three-dimensional elastic-plastic analyses were carried out for two small specimens using finite deformation theory. For a 1/4-inch compact tension specimen of 1 mm thickness, the J-integral along the crack front is higher at the midsection of the specimen than at the surface of the specimen. The J-integral at the midsection of the specimen coincides with the J-value estimated by Merkle-Corten's conventional equation using 0.8 B₀ (B₀=specimen thickness) as an effective thickness. For a 2 by 2 by 11-mm three-point-bend specimen, the J-integral distribution along the crack front is similar to that of the 1/4 CT specimen. The J-integral at the midsection of the specimen can be estimated from the load-versus-displacement curve within 10% error by Rice's conventional equation using 0.85 B₀ as the effective thickness. The three-point-bend specimen is mere appropriate for the fracture toughness test of a small specimen than is the 1/4 CT specimen.

View full abstract

There are various compression test methods for a unidirectional composite material, but they use different specimen lengths without cosidering the effect of specimen length in detail. In this paper, the effects of specimen length and stress concentration on the compressive strength of unidirectionally carbon-fiber-reinforced plastics were investigated. Compression tests for the CFRP were carried out for a wide range of specimen lengths, three different end tab configurations and two different end tab materials under the condition of constant specimen thickness. This was performed using the Celanese compression test method. The apparent compressive strength measured by the Celanese test method is lower than the true compressive strength because of the stress concentrations near the tab tip of the specimen. However, the apparent compressive strength of the specimen with a circular profile (R25) between the chucking part and the testing part (CFRP) which has slight stress concentration at the tab tip is higher than those of the Celanese specimen configurations. The true compressive strength was obtained using the specimen with the end tab of stainless steel and lengths of 3.2 mm and 6.4 mm.

View full abstract

The uniaxial creep deformation of shape memory polymer of polyurethane series at temperatures higher than the glass transition temperature was investigated experimentally. The main results are summarized as follows. ( 1 ) The time-independent strain is recoverable and does not vary under cyclic loading. The time-independent strain is represented by the sum of linear and nonlinear terms of stress. ( 2 ) The time-dependent strain is resolved to be a recoverable component and an irrecoverable component. The recoverable strain does not vary under cyclic loading. The irrecoverable strain appears only during the first loading process. ( 3 ) A creep limit exists for the time-dependent strain. The time-dependent strain is represented by a power function of time.

View full abstract

The θ projection method, originally developed to describe the creep curve even in the tertiary creep section, was applied to creep curves observed in pure aluminium. Since this test material has wide expansion of the secondary stage of creep, the original θ projection method cannot necessarily be applied to the whole creep curve. Then the authors modified the θ projection method, replacing a Blackburn-type equation of creep to a Bailey-type equation for the primary creep section. Four parameters derived from the modified equation were expressed as functions of testing temperature, applied stress and grain size of the test material. As a consequence, the proposed equation can determine the whole creep curve except the rupture point, and the minimum creep rate. Using four paramaters extrapolated on the test condition of creep, unknown creep curve can be easily obtained.

View full abstract

Combined tension-torsion creep of nickel-base single-crystal thin-walled tubular specimens, in which [001] is oriented in the axial direction, is analyzed using the 3D finite-element method. A crystal creep equation of the power-law type for steady-state creep is implemented in the analysis. Features in distributions of stress and strain rate in the specimens of long and short gate-length types are discussed so as to give foundations to the 2D models of simplified analysis formulated in the 1st report. Comparisons between the 3D finite-element and 2D simplified analyses are made with respect to overall tensile and torsional strain rates of the specimens. Effects of the deviation of [001] from the axial direction are also discussed. Consequently the finding on active slip systems obtained by use of the 2D models in the 1st report is ascertained on the basis of the 3D finite-element analysis.

View full abstract

A computational system using the general-purpose code code ABAQUS for laser transformation hardening was developed through finite-element method analysis. In this system the temperature history of the laser-irradiated area was calculated and then compared to a TTA (time-temperature-austenitization) diagram. The elements for which temperature histories were sufficient to produce austenitization and martensitization were taken as the hardened zone. Circular beams with three typical kinds of laser beam profile and rectangular beams with three various intensity distributions were considered in the model because the laser beam profile has a critical influence on laser processing characteristics. When circular beams were used, a ring-mode beam was found to be better than a multi-mode or single-mode beam from the standpoint of obtaining the widest hardened zone. In obtaining the widest and deepest hardened zone a rectangular beam having a peak in the rear part was found more effective than a homogeneous beam. On the basis of these results, it is suggested that the temperature history of the laser-irradiated area can be controlled by the variation of intensity distribution.

View full abstract

This paper is concerned with the boundary integral equation formulation and its numerical implementation for steady-state vibration analysis of spatial frame structures in the context of small elastic deformation and small strain theory. First, the integral equation formulation and its method of solution are presented in detail for a single straight bar subjected to distributed axial load, bending and torsional loads. Then, these are applied to the analysis of three-dimensional frame structures by assembling the element stiffness matrix of each element into the global stiffness matrix of the whole structure to be analyzed. A new computer code is developed in this study, and is employed to compute the steady-state responses of some frame structures, where by the usefulness of the proposed analysis method is demonstrated.

View full abstract

The static strength of a weld by lap resistance welding and TIG welding on TiNi shape memory alloy and the cyclic behavior of recovery strain under loading-unloading and heating-cooling processes have been studied. The influence of the welding schedule on the static strength of a welded wire has been described. For cyclic characteristics, base on the experimental results of the cyclic loading-unloading and heating-cooling processes, the cycedlic behavior of recovery strain for the wire and welded wire has been investigated.

View full abstract

The formation process of a gradation in composition under a centrifugal force in functionally graded material (FGM) fabrication has been studied. The motion of ceramic particle in the viscous liquid, which corresponds to liquid metal, was analyzed by numerical calculation. The gradations of ceramic particle distributions in viscous liquid are visualized with the aid of computer graphics. Thses results are in good agreement with results of experiments which use the corundum/plaster model, including the size effect. Thus, the method of simulation presented in this study can properly estimate the gradation in composition. From the results of simulation and experiment using the model, it is also found that precise control of composition can be achieved by simultaneously using particles of different grain sizes.

View full abstract

In this paper, the optimum directions of bone tunnels at the isometric points on tibia and femur for anterior cruciate ligament reconstruction are determined. The crossed four-bar linkage model is applied for knee kinematics. Then the directions of bone tunnels at the isometric points are mathematically described as the anterior angle in the sagittal plane and the medial angle in the coronal plane. Next the following three variables are considered: ( 1 )the bending angle φ between the center line of the fiber part of the graft and the bone tunnel, ( 2 ) the bending angle φ between the center line and the width direction of the fiber part at the exit of bone tunnel, ( 3 ) the strain ε of the graft. In the actual reconstructive procedure, the directions of bone tunnels are limited by surgical accessibility factors such as the design of the drill guide and the soft tissue around the knee. Hence the optimum directions of bone tunnels are determined to minimize the three variables stated previously for knee flection within the area in which bone tunnels can be made. Next, in order to examine the influence of the assumptions for modelling and the individual differences, an experiment using fresh cadaver knees is carried out. The theoretical bending angle and the experimental bending angle are found to be in good agreement. Therefore, in the actual reconstructive procedure, it should be effective to give the bone tunnels the optimum directions determined in this studdy.

View full abstract

Conventional photoelasticity and holographic interferometry could be successfully combined into holographic photoelasticity, which is able to give engineers isopachics, a family of the contour lines of the sum of principal stresses as well as isochromatics both solely and simultaneously. As an extension of this advanced technique developed for determining principal stresses independently to the photoelastic coating method, holographic coating method may be naturally inspired. Apart from holography, it is also possible to be used as an alternative method for yielding isopachics, which is composed of moire techniques and superposition of interference fringes produced by light waves from front and rear surfaces in transparent material adhered to reflective part of the structure. The light intensity equation governing the light fields in these newly developed methods was calculated analytically. A comparison of the experimental and theoretical fringe patterns was made. Improved simulation patterns generated by the light intensity equation and elastic stress distribution demonstrate the characteristics of these methods and some advantages in moire coating rather than holographic coating method.

View full abstract