Location of defects in solid can be presumed in terms of arrival time differences of AE-waves to each sensor. A small amount of the arrival time error, however, might influence on the AE source location. Present paper investigates the influence of errors including both the arrival time and sensor location on the AE source. Superiority of the neural network technique over the least square method was discussed on accuracy of location obtained by the arrival time differences. Applicability of source location using simulated AE source to real structure of pipe was also discussed. It was concluded that the neural network technique should provide available means to the AE source location because of less influence characteristics from the noise.

View full abstract

In a previous study, for the case of elastic deformation field, the authors have developed a hybrid experimental-numerical method based on a variational principle minimizing experimental measurement errors. In this stydy, for the case of elastic-plastic deformation field, first, an incremental variational principle is derived to minimize the errors associated with a full-field displacement measurement such as moire interferometry. Next, on the basis of this variational principle, a new hybrid moire-interferometry and finite element method is developed for the case of elastic-plastic deformation field. The present hybrid method demonstrates automatic detection and elimination of the measurement errors and smooth visualization of stress and strain contours. Furthermore, the present hybrid method automatically achieves the path independence of the T^* integral, restoring the path-dependence caused by the measurement errors. This paper provides the foundation of an intelligent measurement and visualization of displacement and stress fields in actual structural components.

View full abstract

The hierarchical domain decomposition method (HDDM) is a solver for large scale algebraic equations in the finite element method. It is suitable for various kinds of parallel computers such as massively parallel processors (MPP) and workstation/PC clusters. In this study, the original HDDM is first modified to improve computation speed, and then applied to static elastic-plastic finite element analyses. Some key techniques employed in the static nonlinear finite element analysis are presented. As illustrative examples, the parallel elastic-plastic analyses are performed on a nuclear structure with 1.3 millions degrees of freedom and a cube with ten millions degrees of freedom. It is shown that both the conjugate gradient method used in the HDDM and the Newton Raphson method are successfully converged, respectively, in the two examples.

View full abstract

In this paper, an inverse analysis method to identify the nonlinear material behavior of solid, using a nonlinear sensitivity analysis, is presented. The proposed methodology can be applied to estimate the stress-strain relationships of various types of materials. The nonlinear sensitivity analysis methodology, which does not require repetitive analyses for general nonlinear problems to calculate the sensitivity of the solution to the variation of material constants, stress-strain relationship, etc., forms a basis of the material behavior identification procedure, by combining with a nonlinear programming algorithm. The identification problem is defined to be a minimization problem, by choosing an appropriate objective function, which represents the deviation of result of finite element analysis from experimentally measured displacement field and applied loads. The procedure is constructed to be quite general, so that it can be applied to a wide range of material behavior identification problems, such as stress-strain curve, changing material properties, etc. In the previous authors' article, the identification problem of stress-strain curve was successfully completed and a possibility of simplifying experimental procedure in obtaining stress-strain curve was suggested. In this paper, the proposed methodology is applied to internally changing material property determination problems.

View full abstract

As computers can be connected to each other through the internet, we are able to use a remote computer even if it is located overseas in the same way as using a computer which is located nearby. This paper describes a parallel finite element system using the domain decomposition method, which is implemented on a computer network made from remote computers at the Internet. The above system was successfully applied to the 3 D structural analyses with a high parallel efficiency.

View full abstract

This paper shows an optimal layout algorithm of a damping material which is affixed on a vibrating surface of a structural component. This method is based on the generalized layout optimization method⁽¹⁾ where a prescribed amount of material is distributed in a design space so that an objective function may be minimum. The complex elastic moduli are employed to express the visco-elastic behaviours of a damping material under periodic loads. Two different objective functions are employed in numerical examples. One is dynamic mean compliance, which is useful in reducing the displacement at the loading points. The other is square product of the nodal displacement vector, which is effective to reduce the overall dynamic response of the structure. It is shown that both of the results show good convergence and give reasonable layout.

View full abstract

Cyclic stress-strain responses of annealed copper were numerically calculated in the minimum cross section of cylindrical notched components under axial load-controlled condition using elastic-plastic finite element method. The constitutive equation was developed for analyzing cyclic deformation by modifying the Johnston-Gilman's equation, which is derived from the concept of dislocation dynamics. The cyclic load-diametral displacement response of the notched component was in quite good agreement with the test result. The stress-strain response at a given position in the minimum cross section was calculated, which exhibited characteristic cyclic hardening. Then residual stress, mean stress, mean strain, and cyclic hardening behavior were successfully evaluated in the minimum cross section.

View full abstract

The near-tip stress fields around an interface edge of butt-jointed plates subjected uniform tension are analyzed by finite element method. Based on the result the dependence of the singular stress fields around the interface edge on the material combination is discussed. The relation between the exponent of the stress singularity and its intensity is investigated. The uniqueness of the intensity of singular stress fields for Dundurs's parameters is also studied. The equations which is useful for materials selection of joint concerning with the stress singularity are newly developed.

View full abstract

The constitutive equation for the TRIP steel has been improved through precise uniaxial tension tests under several environmental temperatures. To elucidate highly nonuniform deformation behavior dependent on martensitic transformation, computational simulations with the proposed constitutive equation are carried out for TRIP steel bars with a ringed notch under Mode I loading. The distribution of volume fraction of martensite phase is compared with that obtained by micro hardness measurements around the notch tip in the cross-sectional plane. Fairly good correspondence between these two results verify the applicability of proposed constitutive model to highly inhomogeneous deformation behavior.

View full abstract

The purpose of this study is to clarify the dynamic behavior of stress waves in strip plates with fillets experimentally. The reflection coefficient, with corresponds to the changes in the duration of the incident pulse and the fillet shape, has been obtained by the strain gauge method. The relationship between the duration of incident stress waves to the fillet and the dynamic stress concentration factor is obtained by a dynamic photoelastic method. An interaction of the incident wave and the reflected wave takes place at the fillet. Depending on the fillet shape, the dynamic stress concentration factor becomes 10-70% higher than the static stress concentration factor for incident waves. The relationship between the duration of stress pulse and the dynamic stress concentration factor has also been obtained.

View full abstract

In order to clarify mechanisms of surface damage and to establish methods for evaluating strength degradation in ceramics subjected to impact of solid particles, testings of silicon nitride sphere impact to silicon nitride specimens were carried out. The residual strengths of specimens were measured by flexural tests, in which the damaged surfaces were the tensile sides. Fractures were classified to several types depending on their own impact damages. Influences of specimen thickness on the impact damage and the residual strength were also examined. From the comparison of test results of indentation and impact, the possibility was shown that the residual strength due to sphere impact would be evaluated using a method based on the elastic indentation theory.

View full abstract

In the 1st report, it was shown that it was possible to apply an elastic indentation theory to the evaluation of surface damage of ceramics due to sphere impact, although this was limited to low velocities in which elastic damage dominated. In this research, further investigation on the applicability of the theory to the residual strength evaluation was conducted. In the case that elastic damage or cone crack was related in the final fracture, the residual strength of ceramic specimens damaged by sphere impact was estimated using the elastic indentation theory and conventional fracture mechanics. Good agreement between estimated curves and test results were obtained.

View full abstract

In order to elucidate mechanisms of surface damage and to establish methods for evaluating strength degradation in ceramics subjected to dynamic contact with solid particles, silicon carbide sphere impact tests were carried out for silicon carbide specimens. A gas-gun-type apparatus was used for these tests. The residual strengths of the surface damaged specimens were measured by flexural tests. Damage analysis and residual strength evaluation were conducted. For the case of elastic damage (cone crack), the residual strength could be evaluated using indentation analysis, up to low speed range, in which the specimen was not ruptured by impact.

View full abstract

Shape memory TiNi fiber reinforced/epoxy matrix composite is fabricated to demonstrate the suppression effect of crack-tip stress concentration and the control of fracture toughness K value of the composite. The decrease of fracture toughness was attributed to the compressive stress field in the matrix which was induced when the prestrained TiNi fibers contract to the initial length upon heating above austenitic finish temperature of TiNi fiber (>A_f). The dependencies of K values on the prestrain value of TiNi fibers as well as crack-tip compressive stress domain size are discussed based on the analytical equivalent inclusion model for composite.

View full abstract

In this study, the problem of jointed dissimilar materials subjected to a suddenly applied dislocation or force which travels constant velocity on the interface is analyzed. The transformed solutions in the Laplace and Fourier domains include the function of Stoneley Wave, and then the inverse transforms are carried out using the Cagniard-deHoop method. The exact transient closedform solutions in the time domain are obtained. The features of Stoneley Wave and the typical stress concentrations near the dislocation tip are showed by graphical representation.

View full abstract

By modeling a damage portion of a laminated composite, strategic way of numerical analysis how to identify the characteristics of damage from bending deflection has been shown. The model is developed by extending the concept of equivalent inclusion. The damage is represented effectively in the formulation by equivalent additional force in view of inhomogeneity of bending stiffness due to existing damage. Deflection of the laminated composite plate with damage can be reproduced by the additional force based on the classical plate theory. Therefore the identification of damage is numerically performed by determining the additional force from given bending deflection of the plate, inversely. When the additional force is assumed as a point force, position and intensity of the force correspond to location and level of the damage, respectively. Identification process of the damage characterization is illustrated with representative numerical results obtained by finite element analysis for cross-play laminated composites.

View full abstract

Metal matrix composites have damage for matrix cracking, fiber break and /or interfacial debonding by thermal and mechanical loads. This paper is focused on interaction between fiber break and interfacial debonding of the composite. The interface between fiber and matrix is subjected to frictional force under compressive residual stress. It gradually slips from the edge of the broken fiber during loading. A homogenization theory with multi-scale asymptotic expansion is used to solve the problem for the composite in the analysis. The displacement is expanded to macro-and microscopic fields for homogenization. And governing equations are reconstituted for each level. To solve the perturbed displacement which indicates the microscopically disturbed term, three dimensional boundary element analysis is made for a unit cell. In numerical calculation, the interfacial slip and nonlinear properties of the composite are illustrated during longitudinal loading.

View full abstract

The multiple strata plasticity model, which was presented in the previous paper by introducing the tensor Q expressing the activating state of slip systems in the polycrystalline metal, has been developed with a new internal state factor B₅. This factor B₅ denotes the ratio of the grains having the activating slip systems greater than five to the total grains. The variation features of B₅ and Q can be computationally evaluated with a virtual polycrystalline specimen. The microstructure state such as dislocation density and so on can be reflected in the internal structure variablle A in this model. The relation between B₅ and A has been examined on the tensile flow curve of pure copper. As a result, a simple formula connecting the A with B₅ was found. Moreover, it has been clarified that this formula is applicable for describing the plastic deformations under another proportional loadings.

View full abstract

A series of deformation tests was conducted on Ni-based superalloy IN 738 LC at 850°C to clarify inelastic behavior of the material. The inelastic behavior was summarized as follows. The material shows time/rate dependent deformation, no cyclic hardening (or softening) behavior and an eminent shift of stress-strain hysteresis loops with cycle for tensile strain-hold waveforms, which are rarely observed on other materials. The conventional unified constitutive equation (Chaboche model) was applied to the test results to check applicability of the constitutive model for IN 738 LC. It was found that Chaboche model was capable of describing the inelastic behavior well except the condition of tensile strainhold waveforms. For tensile strain-hold waveforms, it showed no shift of hysteresis loops with cycle, and the expression capability of stress-strain response was insufficient. Further, the relationship between the shift of hysteresis loops and kinematic rules of a back stress was discussed.

View full abstract

A new unified constitutive model was proposed in order to describe inelastic deformation behavior of Ni-based superalloy IN 738 LC precisely. In the constitutive model, an internal state variable corresponding to the origin of dynamic recovery of a back stress was introduced into an evolution law of the back stress. Due to this modification, the constitutive model could express a shift of stress-strain hysteresis loops for tensile strain-hold waveforms, which was specific inelastic behavior of the material and was hardly described by conventional unified constitutive models. To check the validity of the constitutive model, numerical simulations were done for various inelastic behavior of IN 738 LC at 850°C. The simulations agreed with experimental results adequately, including the description of the shift of hysteresis loops.

View full abstract

An electrical resistance method is applied for identifications of size and position of delamination cracks of components of CFRP in the present study. The method does not cause strength reduction of CFRP structures, because the method adopts reinforcement fiber itself as sensors. Due to orthotropic electrical residence of laminates of CFRP, electric potential fields of CFRP laminates are different from the fields of conventional methods for isotropic metallic materials. Authors have investigated the effect of orthotropic electrical resistance on the method by using FEM. Based on the results. identification of delamination crack size and position was impossible in the case of only an entire electric resistance change of a structure is available. In the present study, multiple points for measurement of electric potential are adopted for identifications of the delamination crack size and position. The response surface methodology is adopted to estimate delamination size and position with electric potential change data. Using FEM analysis, identification of the delamination crack size and position is conducted. As a result, a 5-segment-method is numerically shown to be very effective for identification of delamination crack size and position.

View full abstract

The penetration of a viscous lubricant into a two dimensional surface crack under a rolling contact stress was calculated. The finite element method was applied to solve the Reynolds equation coupled with the elastic deformation of the crack. The lubricant begins to flow into the crack when the contact pressure is passing the crack mouth. The lubricant enters to the crack tip and increases the stress intensity factor, when a nondimensional parameter β=ηνE²/(ap³₀) is small. The crack does not open when β is large, because the contact pressure has passed before the lubricant flows into crack. The critical values of β were calculated. The penetration of the lubricant becomes easy when a frictional stress acts as the crack opening.

View full abstract

Fatigue crack growth properties for SUS 316 stainless steels containing nitrogen in the mass. percent from 0.001 to 0.66 were investigated in air at room temperature. The near-threshold fatigue crack growth properties for SUS 316 steel containing 0.001, 0.02 and 0.07 mass. % N were coincident each other. On the other hand, SUS 316 steels containing 0.24, 0.25, 0.40 and 0.66 mass. % N showed the remarkable decrease in fatigue crack growth rate and 50% increase in fatigue threshold. This behavior was related to the slip deformation at the fatigue crack tip. The slip deformation was localized for low nitrogen steels, while it was uniformed for high nitrogen steels.

View full abstract

Low cycle fatigue properties under pure fatigue and creep/fatigue conditions were evaluated using a powder metallurgy superalloy, AF 115, containing a variety of defect-types and defect-sizes. The relationship between life, defect size, and loading condition were analyzed on the basis of crack growth rules and strain energy parameters. Following results were obtained ; (1) Inclusions, micro porosities and prior particle boundaries were observed at the origin of fatigue specimens. (2) By treating these defects as either cracks or notches, the lower or upper boundary of fatigue life can be obtained from the elastic-plastic strain energy parameter, ⊿W_f, and the fatigue crack growth rule. (3) The introduction of a tension hold period to the stress cycle or a decrease in the strain rate in the strain cycle significantly affects the fatigue life. The linear summation rule for partitioned lives, based on the elastic-plastic and creep strain energy parameters, ⊿W_f and ⊿W_c, and fatigue and creep crack growth rules, is suitable for life estimation under creep/fatigue conditions.

View full abstract

In order to make clear the effect of the crack coalescence for crack growth rate, low cycle fatigue tests were carried out using two kinds of spheroidal graphite cast iron (SGI). The crack propagation behaviors from the initiation crack to the fracture were investigated to observe surface and fracture section. When a main crack was connected with preexisting cracks starting from microshrinkages on neighboring surface, crack growth rate was accelerated. After connection, the rate was suppressed. So, the crack growth curve in SGI showed the state of steps. As a whole, its crack growth rate was the same as average rate in SGI. That is, the crack coalescence was not the main factor for accelerating crack growth rate. In addition, it was clear that the large variation of the crack growth curves in SGI was caused by the crack coalescence and the difference of the crack initiation length.

View full abstract

The fatigue strength of Ni-Ti-Nb shape memory alloy was investigated focusing on the influence of surface roughness. The specimens had the surface roughness between 20 and 50 μm and a notch with depth of 250 μm, and the cyclic axial stresses of different stress ratios were applied at 293 K. When the surface roughness was below 50 μm, the alternating fatigue limit leveled off at 285 MPa. However, when the notch depth was 250 μm, it decreased to 145 MPa. The microscopic observation of the fracture suface revealed that the crack started at the Ti base precipitates formed in the subsurface of the specimens with roughness below 50 μm. In contrast the crack began at the surface for notch specimens. Fracture mechanics analysis shows that Ti-rich precipitates affect the fatigue limit of the 50 μm sample, which is four to five times thicker than that of the precipitates. These results indicate that the influence of surface roughness on the fatigue strength of Ni-Ti-Nb shape memory alloy can be disregarded surface roughness less than 50 μm.

View full abstract

It is well known that the fatigue life of carbon steel in oxygenated high-temperature pure water is much less than the fatigue life of carbon steel at room temperature in air when the applied strain rate is very low. The fatigue life of carbon steel was examined in order to develop an evaluation method. Cracks were found to initiate immediately after the fatigue test started in high-temperature pure water. The fatigue life was regarded to be the crack propagation life. The crack propagation rate was found to be constant against the fatigue life ratio, N/N_f. A time-domain evaluation method for fatigue crack growth rate was applied to evaluate the corrosion fatigue life. A time-based nominal crack growth rate was assumed as the reciprocal of time to failure by fatigue. The relationship between time-based nominal crack growth rates in high-temperature water and room temperature air was linear on a diagram of logarithms even when strain amplitudes and rates were changed. An environmental effect index, which was defined as the slope of this line, seemed to be influenced only by the environmental conditions. The newly developed evaluation method could successfully estimate the corrosion fatigue life of carbon steel in a high-temperature pure water environment.

View full abstract

The influence of dual phase microstructure on the fatigue mechanism of TiAl intermetallic alloy, coated with CoCrAlY alloy by the plasma spraying, was studied in vacuum at 1073 K from view point of mesocopic size which is several μm-1 mm size range. The crack growth was observed and photographed under high resolution conditions using elevated-temperature loading cyclic stage for the scanning electron microscope. The results revealed strong influences of coating layer and dual phase microstructure on the fatigue mechanism in the alloy. The main crack initiated from pores in coating layer. The main crack grew into TiAl substrate and meso-cracks initiated in TiAl substrate affected the crack growth behavior. Finally, the main crack linked with Meso-cracks. The fatigue property was dependent upon the mechanism retarding the main crack growth due to dual phase microstructure.

View full abstract

The purpose of this study is to investigate the relationship between nonlinear behavior of fatigue life curves and fatigue life for carbon-fiber-reinforced PEEK and PA 66 under two-step loading above T_g (Glass transition temperature) and below T_g. As a results, the fatigue life curves in the case of PEEK/Below T_g, PA 66/Above T_g and PA 66/Below T_y were found linear. Moreover, the fatigue lives were calculated by linear cumulative damage law under two step loading. On the other hand, the fatigue life curves in the case of PEEK/Above T_y was found nonlinear. Moreover, the fatigue lives were not calculated by linear cumulative damage law under two-step loading. It was found that these factors occurred because of 'Transition phenomenon' which fatigue fracture mechanisms transferred in the range between two fatigue life curves.

View full abstract

The response of thin piezothermoelastic composite plate subjected to stationary thermal and electric fields is investigated. Solutions are obtained using the thin plate theory and Rayleigh-Ritz method. As an analytical model, we consider a simply supported antisymmetric angle-ply laminate piezothermoelastic plate. The plate is exposed to an environment with a temperature rise on the upper surface only. To reduce the deflection produced by the thermal loading. electric potential is applied to the piezoelectric layer in the composite. Numerical results show that the ply angles of laminate configuration and the number of layers give influence of the response of the thermal deflection and the applied voltage.

View full abstract

In a previous work, a new method was proposed to estimate 3-D welding residual stresses nondestructively. Firstly, 3-D eigenstrain distributions are estimated by the inverse analysis from surface strain changes due to removal of reinforcement of the weld. Then, residual stresses are easily determined by an FEM analysis using estimated eigenstrains as boundary conditions. Under an existence of possible experimental errors. however, a solution of the eigenstrain distribution becomes unstable due to the ill-posedness of the problem, which is not acceptable as an actual evaluation method. To overcome above ill-posedness, a regularization method based on the singular value decomposition was incorporated. It is confirmed that 3-D residual stress distributions can be obtained in good accuracy. Then, a new optimizing method determining strain measuring locations under reasonably small number, was proposed based on the experimental design theory using the Doptimality criterion and its utility is shown.

View full abstract

Mode I Double Cantilever Beam (DCB) and mode II End Notched Flexure (ENF) tests for Interlaminar fracture toughness were carried out on unidirectional carbon fiber reinforced plastics (CFRP) using a long and a short strain gages glued on the specimen surface. For the DCB test a new analysis method has been developed to calculate the fracture toughness, G_IC, based on the strain output. The results were compared with the G_IC calculated by an usual analysis method based on a load-displacement curve. For the ENF test, a new control method to stabilize the fracture propagation has been developed. Here the test is conducted so as to increase at a constant rate the difference of the strain of the long and the short gages. This method makes it possible to obtain a continuous R-curve of the mode II interlaminar fracture toughness, G_IC. Further an analysis method to calculate crack length and G_IC, based on the strain data is also established. Because the strain output is significantly sensitive to the crack propagation, fracture initiation is detected more precisely and accurate determination of crack initiation becomes possible by these method.

View full abstract

A new global search method for general contact problems is developed and implemented in a Lagrangian program. The proposed algorithm improves the grid (bucket) method by utilizing both multiple grids and an overflow linked list. Without these improvements the search efficiency of the original method deteriorates under such configulations that contacting nodes are distributed sparsely or packed closely in the computational domain. A numerical simulation of a debris impacting double shields is conducted to verify the utility and the efficiency of the present method. We also discuss antitangle forces which is required to continue the computation long enough to form a debris cloud after the first shield is perforated.

View full abstract

In this paper, we discuss the topology and shape optimization scheme of continuum structures by using genetic algorithm (GA). The profiles of the structures are defined by the cross-section between the 3 D spline function surface and the 2 D plane. The genetic representation of the structure is defined by the series of the coordinates of the control points. Then, the genetic operations are applied for determining the profile of the structure satisfying the design objectives. The present scheme is applied to minimum weight design of two dimensional elastic problems in order to confirm its validity.

View full abstract

This paper describes the microstructural effect of Diamond-Like Carbon (DLC) thin film on micro-tribological properties. DLC thin films with density of 1.9-2.4 g/cm³ were deposited on Silicon (111) surface by Plasma-Enhanced CVD method of the hot cathode PIG discharge type. The effect of deposition conditions on the microstructure of DLC thin film was evaluated by Raman spectroscopy. Increasing the bias voltage and the discharge current shifted the peak of the G-band shift toward high frequency and enlarged I (D)/I (G) ratio. Pin-on-Plate friction experiments were carried out using a diamond pin in high vacuum. The friction coefficients of DLC thin films indicated constant values during the friction. The friction coefficients increased with shifting G-band toward low frequency in the range above 1540 cm^-1, but those did not vary with increasing I (D)/I (G). The friction coefficients were directly proportional to density of the film. Molecular orbital calculations revealed that the adhesion energy on the sp³ bonding atom cluster was higher than that on the sp² bonding atom cluster.

View full abstract