The temperature dependence of the static fatigue behaviours of SiC was examined. The experimental result of the fatigue strength which can be represented using Larson-Miller plots indicates that the thermally activated processes such as oxidation and viscous flow of grain boundaries are dominant in the elevated temperature static fatigue. However, the temperature dependence of the static fatigue strength becomes unclear with the decrease of the temperature because the influence of the scatter of the defect sizes becomes realatively stronger. The static fatigue crack growth shows complicated behaviour involving thc crack velocity reduction and crack arrest. So that the static fatigue limit below which thc crack does not show the extension possibly exists. The new fatigue design rules for structural ceramics based on the defect size and temperature dependence of static fatigue strength is proposed.

View full abstract

The rolls of matrix microstrucure and graphite at fatigue crack tip in retarding the initiation and growth of fatigue crack was examined with the objective of achieving maximum resistance to fatigue in toughened austempered ductile cast iron, ADI. Especially, composite microstructures, ferritebainite structure has been developed as typical examples of toughened ADI by alloying and heat treatment. It is found that the fatigue threshold, ΔK_th in toughened ADI with composite microstructures can be increased by 50 % to 15 MPa √(m). From fractography it could be interpreted that the above increasing in fatigue threshold should be caused by the crack closure induced by zigzag crack path in composite microstructures.

View full abstract

Fatigue crack tip opening stress σ_op and closing stress σ_ct can be easily determined with the relationship between the following strain functions f, h and the cyclic stress σ, f=(ε_y-β)/(ε_x+α), h=ε_y+λε_x, where ε_y and ε_x are the strains measured in the vicinity of the crack center (or the crack face), and α, β and λ are constants. With regard to small-size specimens having a center crack or notch, the accuracy of this measurement method is confirmed. If the constants α, β and λ are chosen to proper values, the crack tip opening stress σ_op and closing stress σ_ot can be judged from the inflection points of the σ-h diagram and σ-f diagram. These stresses coincide with the opening stress σ_op and the closing stress σ_ct determined by the unloading elastic compliance method.

View full abstract

In the present study, the influences of microstructures of material on the characteristics of small-crack growth were investigated and the minimum crack length to which the small-crack growth law (dl/dN ∝ l) is applicable was evaluated using a ferrite-pearlite steel, a ferrite-martensite steel and a quenched and tempered carbon steel. The crack growth is influenced markedly by grain boundary, pearlite structure and martensite structure at the initial stage of crack growth, and the influences decrease with the increase in crack length. Even in the process in which crack growth is influenced by the microstructures, however, the mean value of crack growth rate is estimated by extrapolating the crack growth rate in the region where the crack growth is mainly controlled by the mechanical parameter: that is, the small-crack growth law is applicable in most regions of the crack growth process.

View full abstract

In order to study the probabilistic characteristics of the initiation and propagation of cracks, rotating bending fatigue tests of heat-treated 0.45 % C steel smooth specimens were carried out in three stress ranges (1.1σ_w < σ_a < 1.6σ_w, σ_w : fatigue limit). The 16 specimens were tested at each stress range. The following results were obtained through the successive observations of cracks which led to final fracture. The propagation life of a crack from initiation up to fracture, N_i-f, and the total number of cycles to fracture, N_f, are expressed by the three-parameter Weibull distribution. A slight dependency of the coefficient of variation, CV, in N_i-f on the stress range is recognized. Although the initiation life of a crack, N_i, is also expressed by the three-parameter Weibull distribution, the value of CV increases with the decrease in stress range.

View full abstract

The conditions for corrosion fatigue crack initiation were examined on 12 CrMo martensitic stainless steel in dilute sodium chloride solution at 80°C under either galvanostatic or potentiostatic control. The effects of mean stress, NaCl concentration and corrosion current on crack initiation were investigated. The results confirmed that each corrosion fatigue crack had initiated at a corrosion pit. It was found that the critical condition for crack initiation from the corrosion pits could be evaluated using fracture mechanics by treating the pit as a sharp surface crack.

View full abstract

For spheroidal graphite iron, FCD 80, and the same material with plasmasprayed Alumina coating, the axial-load fatigue tests were carried out at elevated temperature of 350-550°C. The fatigue limit of FCD 80 at elevated temperatures was improved by the plasma-sprayed Alumina coating. Through fractography using SEM, it was found that the crack initiated at an inner porosity of micro casting defect in Alumina-coated FCD 80 and that the fish eye was formed on the fracture surface. On the other side, the fatigue crack originated from the surface porosity in FCD 80. Thus, it could be concluded that the improvement in the fatigue limit of FCD 80 at elevated temperatures would be caused by the prevention of propagation of "metallurgical crack" due to which had been formed from a surface porosity.

View full abstract

The influences of hardness of bolts on the fatigue strength of bolted-joints tightened by the plastic-controlled tightening method were examined. The fatigue strength of bolted-joints was improved by 20% through hardening bolts from H_υ=300 to 400. Hardening bolts induced the higher prestress and restrainted the separation of bolted-joints. So, it decreased the additional stress of bolt under applied cyclic loadings. Also, hardening bolts improved the fatigue strength of bolts them-selves. From the microscopic observation, it was found that bolts of higher hardness could considerably restrict the initiation of fatigue crack at the root of thread groove. For the bolt hardened after roll-formed, decarbonizing the root of thread groove during heat treatment will deteriorate the fatigue strength of bolt.

View full abstract

The impact fracture toughness K_Id of alumina-powder/aluminium composite and alumina-fiber/aluminium composite was measured at temperatures up to 200°C by using a one-point bend impact tester. A fatigue precrack was introduce d in a specimen. Young's modulus, Poisson's ratio and yield stress were measured at testing temperatures, and the dynamic stress intensity factor was calculated by applying the plastic zone correction. The results are as follows: (1) as the loading rate increases, the fracture toughness of both alumina-fiber/aluminium and alumina-powder/aluminium increased. This tendency is stronger at 200°C than at room temperature; (2) as the testing temperature increases, the K_Ic and K_Id values of both alumina-fiber/aluminium and alumina-powder/aluminium decreased. This tendency is stronger for K_Ic than for K_Id; (3) the above experimental tendencies can be qualitatively explained by taking the creep deformation of the matrix into consideration.

View full abstract

The effect of the direction of the crystal axis (C-axis) on fracture toughness of single-crystal ice specimens was investigated by the three-point bending test at -10°C. The loading rate K^^.₁ was about 200 kPa√(m)/s. Four types of single-crystal ice specimens were used in which the C-axis direction of the crystal was placed at an angle θ equalling 0°, 45°(I), 45°(II) and 90°, respectively, to the notch plane. These specimens have a sharp edge notch made by pressing the edge of a razor blade on the bottom of the saw-cut slot. In advance of the test using single-crystal ice specimens, the effect of pressing the razor blade on fracture toughness was examined using columnar grained ice specimens under the loading rate K^^.₁ ranging from 1.0 to 1.0×10⁴ kPa√(m)/s at -10°C. The effects of loading rate K^^.₁ on fracture toughness of both blade-pressed specimens and blade-molded specimens under K^^.₁ > 30 kPa√(m)/s were very small and the difference botween them was approximately 10 kPa√(m). Fracture toughness values of θ=0°and θ=90°specimens were nearly equal, K_C]_50%=110.5 kPa√(m), and those of θ=45°(I) and θ=45°(II) specimens were nearly equal too, K_C]_50%=99.2kPa√(m). The fracture toughness of columnar grained ice specimens estimated from that of single-crystal ice specimens was higher than the experimental values.

View full abstract

The temper embrittlement in low-alloy steels, one of the typical material degradations, occurs in service and it is important to detect this embrittlement by means of a nondestructive technique. In order to develop the nondestructive evaluation method of embrittlement for the Cr-Mo-V rotors, a study was conducted using 1ab-charged steels of P dope and Sn dope. P-doped steels showed evident embrittlement by the segregation of P at grain boundaries. On the other hand, Sn-doped steels showed little embrittlement when compared with P-doped steels. Various testing methods were investigated and compared in terms of their sensitivity and the factors detected, using these steels. The chemical etching test was found to be a very good measure of the embrittlement. That is, it was found that there is a good correlation between the width W of the etched grain boundary measured by the penetration of replicas and FATT. The correlation between W and FATT was also recognized for the samples taken from long-term serviced rotots. The relation had the same slope as that found for the lab-charged steels. From this, the multiple regression was conducted in order to express FATT using W and the variables which are known or nondestructively measurable. As a result, a regression equation which can estimate the actual FATT with the scatter of ±20°C was obtained.

View full abstract

This paper describes an application of statically indeterminate fracture mechanics, SIFM, to, structural integrity evaluation tests under pressurized thermal shock, PTS, which were carried out by United Kingdom Atomic Energy Authority, UKAEA, and Oak Ridge National Laboratory, ORNL, as national projects in the U.K. and U.S.A., and are objects of international round-robin analyses in the project for Fracture Analyses of Large-Scale International Reference Experiments, Project FALSIRE. It was clarified that SIFM can predict ductile crack growth behaviors under PTS with good accuracy in an engineering sense.

View full abstract

In some creep buckling experiments conducted by other authors, axially compressive cylindrical shells with a large ratio of radius to thickness were observed to buckle with circumferential waves. In the previous paper, we demonstrated that the circumferential waves observed in the creep buckling experiments are due to the bifurcation. In the present paper, we performed the analysis of this creep buckling phenomena. We paid special attention to the effect of end boundary conditions on the critical time of creep buckling and on the bifurcation buckling mode. We found that the bifurcation buckling of a cylindrical shell with both ends simply supported occurs earlier than that of a cylindrical shell with both ends clamped. It was also found that circular cylindrical shells exhibit three types of bifurcation buckling modes, depending on the dimensions of cylindrical shell, applied load levels and end boundary conditions.

View full abstract

In this paper, the plane problems of a single edge crack in two bonded elastic layers and in an elastic surface layer bonded to an elastic semi-infinite plane are analyzed by the body force method. The stress fields induced by a point force and a displacement discontinuity in two bonded elastic half-planes obtained by Hetenyi's solution, in other words, Green's functions in closed forms, are used as fundamental solutions to solve those problems. The boundary conditions for stress free-edges of the layers and the crack surface are satisfied by superposing the distributed fundamental solutions and adjusting their densities. The stress intensity factors are systematically calculated for the various geometrical conditions and the various stiffness ratios of the layers.

View full abstract

This study was carried out in relation to application of statistics of extremes to the evaluation of remaining life of high temperature components. When applying statistics of extremes, it is necessary to measure lengths of a large number of short cracks. This is time-consuming and laborious, and hence some automization of measurement is desirable. In the present paper, image processing was applied to measurement of short crack lengths and its usefulness was investigated. For this purpose, creep-fatigue tests were carried out on SUS 316 L. Smooth specimens were subjected to slow-fast cyclic strain of a total strain range of 1 % at 650°C in air with ε^^.=2×10^-5/s in tension and ε^^.=2×10^-3/s in compression. Taking replicas of the specimen surface, the images of the replicas were put into the image processing equipment through the CCD camera and the stereoscopic microscope. Using some new algorithms developed in this study as well as conventional algorithms, individual crack lengths were measured. The crack length measured by means of image processing was compared with that by the eye observation. It was found that the error of measurement of crack length by image processing was within ±10%. The distribution of the crack length was also studied. Furthermore, propagation and coallescence behaviors of cracks were investigated by means of image processing.

View full abstract

A method for theoretical analysis of loading effect on a. c. potential drop is proposed for measurement of cracks. By treating formulation to solve Maxwell's equations for an a.c. current flow in a specimen, it is found that the conductivity σ and the permeability μ of the specimen can be handled collectively in the form √(μ/σ). The quantity √(μ/σ) is denoted by γ. Loading causes a change in γ. Since the change is considered to be dominant along the crack front, the loading effect is analyzed by introducing a thin layer confining the change along the crack front. The layer is named line impedance. Its property is determined by a frequency of the current and the change in γ, which is expressed as a function of the stress intensity factor, K₁, through a 2-D experiment. Numerical results are shown for a semielliptical and a semicircular surface crack.

View full abstract

The impulsive responses of a finite circular cylindrical shell filled with fluid are analyzed in order to clarify the shell-fluid coupling effect. The shell clamped ends is subjected to impulsive axisym-metric loading. Flu^^(..)gge's dynamic bending shell theory and the potential theory of compressible perfect fluid are used in the analysis. The analytical solutions in the Laplace-transformed domain are obtained. The inversion of the solutions is performed numerically using the algorithm of FFT. As a result, it is thown that the amplitude of coupled responses decreases with the lapse of time. On the other hand those of uncoupled responses are constant. When the non dimensional parameter β, which denotes the ratio of fluid bulk modulus and the rigidity of the shell, is large or the shell is thin, the coupled influences increase remarkably.

View full abstract

In this paper, a finite difference method based on integration along bicharacteristics is extended for the numerical analysis of the propagation of two-dimensional torsional wave in an elastic circular cylinder composed of several cylinders bonded completely at each end. A numerical example is worked out for the case in which torsional impact is applied to the end face of the cylinder. By obtaining numerical results for various mechanical properties of the cylinders, it is shown that the propagation of the reflected and transmitted waves can be investigated in detail.

View full abstract

In a frontal collision of an automobile, side members are the most important factor for occupant protection. When the side members collapse axially, they absorb large amount of crash energy enough to avoid the deformation of compartments. To promote the axial collapse without bending, beads are generally provided on the side member. But it is very difficult to find the effective position of beading on the members because the relation between the buckling wave and the beads cannot be seen clearly in a bending crash mode. In this study, we show an analytical method of beading on the side members along the buckling mode calculated by a finite element analysis.

View full abstract

This paper provides an exact analysis for shear stress of a five-layered rectangular rod with damping viscoelastic layers, and presents its torsional rigidity and loss factor in the form of an explicit function. Numerical evaluations of both maximum shear stresses in the rod and in the viscoelastic layer are carried out, and these arising points are discussed. The optimum rod composition to maximize the loss factor ratio of the rod to viscoelastic material clearly exists in accordance with each layer thickness and its material constant. These maximum values are nearly equal to each other if the viscoelastic layer thickness is thinner than those of the outer and core elastic layers. These maximum values shift in the temperature range and become higher by means of the optimum adjustment of the viscoelastic layer thickness. Then, the design of the rod for the environmental temperature seems to be performed by such adjustment.

View full abstract

The purpose of this paper is to discuss solidification and mechanical characteristics in the strip casting process by the twin roll method. Interaction of temperature fields between the strip and the cooling roll is taken into account by the substructure method when solving the heat conduction equation under steady state. Simultaneously, in order to verify the results of a simulated solidification mode and temperature, experiments by a small-scale continuous casting system were carried out to determine the temperature in the strip of Sn-Bi alloy with low melting temperature. Moreover, a simulation method for the stress and deformation with a free boundary was developed by employing an elastic-viscoplastic constitutive relation relevant to describing both solid and liquid states, and the thickness of the strip along the free boundary was calculated and compared with the measured thickness.

View full abstract

A dislocation dynamical model for plastic deformation including increase and decrease in mobile dislocation density and increase in immobile dislocation density is proposed. The model takes account of dynamical internal stress due to mobile dislocations and statical internal stress (work hardening) due to immobile dislocations. Stress (applied stress and effective stress)-strain curves of silicon single crystals, through upper and lower yield points up to steady states of deformation, are calculated numerically by the model for various strain rates. The results of calculation agree well with the experiments by Yonenaga and Sumino [Phys. Stat. Solidi (a), 50 (1978), 685] except for upper yield points. Calculated dependence of lower yield stresses and of effective stress-strain curves after lower yield points on the initial density of dislocations also agrees with their experiments.

View full abstract

This paper deals with an inverse problem analysis of thermal deformation in a transient state for an infinitely long solid cylinder with nonhomogeneous material properties in a radial direction. Assuming that the solid cylinder is heated axisymmetrically by the unknown temperature function of the surrounding media, we have analyzed the inverse problem with the object of determination of the temperature function for the surrounding media so as to satisfy the prescribed transient thermal displacement component in the radial direction of the cylinder. The problems of transient heat conduction and the associated axisymmetric thermoelastic field are coupled for the theoretical developments, then the solutions of the temperature function for the surrounding media, the temperature change and the associated thermal stresses, thermal deformations for the cylinder are evaluated by making use of the method of Laplace transform. Thereafter, the numerical results for the temperature change and the stress distributions are shown in figures.

View full abstract

In order to investigate the basic mechanical properties of a bias-type two-way shape memory component, the thermal cyclic experiments on a TiNi shape memory alloy wire under a constant spring constant of a bias spring were carried out. The cyclic characteristics of recovery stress and recovery strain and the conditions for application of cyclic recovery stress and recovery strain were discussed. The main results are summarized as follows. The recovery stress and the recovery strain induced in the two-way component take the maximum values at the initial strain near the completing point of the stress-induced martensitic transformation. The recovery stress and the recovery strain decrease significantly in the early cycles, but take almost constant values after these cycles. In order to obtain large recovery stress, it is necessary to make the spring constant of the bias spring high and to make the initial strain small. In order to get large recovery strain, it is necessary to make the spring constant low and the initial strain large. The transformation temperatures of the stress-induced martensitic transformation and the reverse transformation rise with the number of thermal cycles. In progress of thermal cycles, the stress-induced martensitic transformation and the reverse transformation occur gradually with variation in temperature, and the starting and completing temperatures of each transformation become unclear.

View full abstract

The development of a reliable bonding technique for ceramics and metal is very important for applying ceramics to structure materials. However, there are some serious problems, one of which is the thermal stress concentration occurring near the edge of the interface bonding two materials with different mechanical and thermal properties. In this paper, residual thermal stresses near the edge of interface of bonded joints with several angles of wedge are computed by using the elastic and elastic-plastic finite element methods (FEM). The stress distribution of elastic analysis was in good agreement with Bogy's analysis. The elastic-plastic analysis was carried out for the case that temperature is altered from 800K to 300K. It was found that when stress in metal along the interface exceeds its yield strength, the order of stress singularity at the interface edge in ceramics side was nearly equal to the value obtained by Bogy's theory using very low Young's modulus. It was confirmed by using thermoelastic-plastic FEM that the stress singularity diminished at the wedge angle obtained by Bogy's theory

View full abstract

Thermal stress distribution in a butt adhesive joint composed of two finite thin plates, in which both end surfaces are kept at higher temperatures and the other surfaces at lower temperatures, is analyzed using the two-dimensional theory of elasticity. In case of the joint with similar adherends kept at the same higher temperature at both free end surfaces of the joint, it is seen that the maximum tensile stress occurs at the edge of the interface between the adherends and the adhesive and compressive stress around the center of the interface. The effects of the ratios of the coefficient of thermal expansion, Young's modulus and the thickness of the adherend to those of the adhesive on the thermal stress distributions are obtained by the numerical calculations. Moreover, in the case of the adherends kept at different temperatures at each free end surface, thermal stress distributions at each interface are obtained numerically. For verification, the thermal strain distribution near the interface was measured using strain gages and also the photoelastic experiment was carried out on the joint where an epoxy resin plate was used as an adhesive. The analytical results were satisfactorily consistent with the experimental ones.

View full abstract

The maximum principal stress in electrodeposited multilayer bellows (EDM bellows) is studied by the finite-element method. The EDM bellows are made of two kind of materials which have different Young's moduli and by an electrodeposition method. It is found that the maximum principal stress in EDM bellows is smaller than that in single-layer bellows. This is a similar behavior as is expected for formed multilayer bellows.

View full abstract

To estimate the influence of the strain history and the strain rate history on the subsequent loading simultaneously, the strain rate jump tests under the combined loading condition were carried out for copper and pure titanium. The obtained equi-equivalent plastic strain surface of both metals is carefully examined, then, the constitutive equations which include both the effect of the strain history and the strain rate history are proposed. Comparison of the numerical simulations with the corresponding experimental results is performed for the verfication of the proposed model.

View full abstract

A Clarification of the mechanical causes for low-back pain requires a knowledge of the states of stress and/or strain throughout the thoraco-lumbar spine. However, the deformation behavior of an intervertebral disc in vivo has not been investigated sufficiently. This paper presents a new method to evaluate the state of strain occurring on the sagittal plane of an intervertebral disc in vivo under flexion and extension. That is, the displacement on the plane of a disc was measured by FCR (Fuji computer radiography) at the medical hospital of Hokkaido University. The internal strain distribution of a disc was estimated by use of an isoparametric shape function introduced to express the deformation of a disc. Consequently, compressive strain at the anterior part of lumbar discs, and tensile strain at the posterior part were confirmed to occur in vivo under flexion. On the other hand, the strain distribution under extension was opposite to that under flexion.

View full abstract

The Fourier analysis of the broadened X-ray diffraction profiles from polycrystalline materials allows the determination of the microstrain in materials. To obtain an accurate microstrain, it is important to measure the Fourier coefficients of the diffraction profiles precisely. The equations for calculating the standard deviation of the Fourier coefficients are derived analytically. From these equations, the size of the variation in the Fourier coefficients due to counting statistics can be obtained. The variances in the measured Fourier coefficients result in an error in the microstrain calculated from them. An equation for calculating the standard deviation of the microstrain was derived by taking account of the variances of the Fourier coefficients. The Fourier coefficients and the microstrain calculated from the Fourier coefficients were determined for quenched and tempered steel, and their standard deviations were calculated by using the present equations. As a result, the relative error of the microstrain was remarkably larger than that of the Fourier coefficient.

View full abstract

The texture-anisotropy of rolled plate was measured using a scanning acoustic microscope (SAM) for acousto-elasticity research. A new technique to influence V(Z) curve by varying the power of ultrasonic signal was proposed for the accurate velocity measurement of a leaky Rayleigh wave. This technique improved the accuracy by one-fifth over the conventional method. The magnitude of the texture-anisotropy of cold-rolled aluminum plate distributes parabolically along the thickness of the plate. It remains almost constant in the center part of the plate, and becomes suddenly large at both surfaces of the plate. Comparing two kinds of plate of the same thickness which were rolled either by many passes of a small reduction roll, or by a few paths of a high reduction roll, the magnitude and the variation of texture-anisotropy of the former are larger than those of the latter.

View full abstract

This report deals with a method of estimating microstructures in multiphase composites by inverse analysis. First, the multiphase composites are modeled with a microstructure model which is composed of ellipsoidal reinforcements in a continuous matrix. Then, characteristic dimensions of the analytical model are determined so that the difference between experimental measurements on the macroscopic material properties, including anisotropy., and analytical values obtained from the above model is minimized by inverse analysis using mathematical programing. Finally, the present method is confirmed by application to metal/plastic composites.

View full abstract

The effect of the direction of the crystal axis (C-axis) on macrofractography of single-crystal ice specimens was investigated by means of the SEN bending test at -10°C. The loading rate K^^._I was about 200 kPa√(m)/s, Four types of single-crystal ice specimens, in which the C-axis direction of the crystal was placed at an angle θ of 0°, 45°(I), 45°(II) and 90°, respectively, to the notch plane, were used. The edge notch was made by pressing the edge of a razor blade on the bottom of the saw-cut slot. A macroreplica was made on one side of the fracture surface, using dental impression material. The C-axis direction was measured from the surface of the other side by the etch pit method. Macroreplicas clearly showed that the aspects of fracture surfaces were affected by the C-axis direction and that the fracture surface contained many traces of crack extension along the basal plane {0001} or the prismatic plane {101^^-0}.

View full abstract

A nondestructive electrochemical method is developed for detection and evaluation of temper embrittlement of Cr-Mo-V Steel and 2.25 Cr-1 Mo steel. In this method, temper embrittlement can be detected and evaluated by the second peak of current density "I_P2" which appears in the anodic polarization curves measured in a calcium nitrate solution. In contrast to nonembrittled materials (couplings and de-embrittling heat-treated materials) which exhibited the dull and low second peak, the sharp and high second peak current appeared in the anodic polarization curves measured on embrittled materials. This increase in I_P2 is caused by selective dissolution of embrittled (impurity-enriched) grain boundaries and that increase in I_P2 has a good correlation with the shifts in FATT. Temper embrittlement can be nondestructively evaluated by I_P2 measurements with high accuracy.

View full abstract

An ultrasonic bolt axial force inspection system is developed for the control of bolts used on turbine outercasings in a thermal power plant. The ultrasonic time of flight through the bolt under stress is analysed by using the acoustoelastic law. It turned out that the proportional constant, conventionally called the material constant, is determined by an acoustoelastic constant k and Young's modulus E as 1-kE. Measurements of k for some kinds of steel bolts revealed that k is approximately the same value (-1.1×10¹¹ m²/N) without concerning heat treatment. Based on these results, an ultrasonic bolt axial force inspection system is developed. Applying this system to turbine bolts, it is found that the bolts could be controlled more precisely than ever.

View full abstract

This paper deals with the design sensitivity calculation by the direct method for isoparametric curved shell elements. Sensitivity parameters include the intrinsic property of shell thickness, as well as geometric variables which influence the size and shape of a structure. The influence of design variables therefore may be separated into two distinct contributions. The influence of parametric mapping within an element, as well as the influence of geometric variables on the orientation of an element in space, is taken into consideration for the sensitivity calculation of geometric variables, and efficient formulations are derived. The methods presented here are applied to the sensitivity calculations of displacement, stress, buckling stress and natural frequency for typical basic examples such as a square plate and a cylindrical shell. The numerical results are compared with the theoretical solutions and finite difference values.

View full abstract

Gas flow in the airway and lung deformation due to tissue elasticity and surface tension are the main characteristics of respiratory dynamics. In this study, the simulation with the discretized model as a distributed parameter system is used extensively to identify some of the model parameters which could not be measured directly and to estimate the real situation of the lung respiration. By using 5 excised lungs from Japanese white rabbits, the pleural pressure vs. lung volume change (P_pt-ΔV)relations are observed in situ under oscillating pleural pressure. The elastic properties of the lung tissue are measured under biaxial tension. The unknown parameters, such as that for the surface tension in the alveoli, are identified so that the calculated P_pt-ΔV relations based on simulation coincide with the experimental results for each lung. The simulation studies with the identified parameter demonstrate the potential to estimate the real phenomena inside the excised lung.

View full abstract

A formulation is presented for the shape modification under equality and inequality constraint conditions which are expressed in linear form of design variables. Slack variables are introduced to modify the inequality conditions into the equality form. The formulation is based on the solution of the governing equation of the design variables and slack variables by means of Moore-Penrose generalized inverse. The complementary solution is added when the partidular solution fails to satisfy the requisite of the slack variables. Numerical examples based on the boundary element sensitivity analysis deal with the shape modification of a press yoke aiming at light-weight design under stress restriction and verify the validity of the proposed formulation.

View full abstract

An estimation method of the dynamic constitutive equations using the method of least squares is described. In this paper, the unknown parameters in the constitutive equations of which the functional form is known were estimated. The estimation is performed by using the method of nonlinear least squares utilizing the Gauss-Newton method because the constitutive equations are nonlinear. Numerical simulation were done such that the unknown parameters in the constitutive equation were estimated from the value of the strain wave in a long bar subjected to longitudinal impact. The stability of the estimation and influence of experimental error on the estimated parameters were examined. As a result, it was concluded that the present method is practically useful.

View full abstract

An optimization design of natural frequencies of a cantilever beam with rigid or elastic supporting springs is studied. The structure modification is optimized for a given set of the change of the eigenvalues of the frequencies. Design parameters are the position and rigidity of the supporting springs. An iterative calculation to obtain the objective expectations is carried out by means of a sensitivity analysis of the finite element solution and a Lagrange function to seek the objective structure as near to the baseline structure as possible. Useful results are obtained for a calculation method of the sensitivity, a treating method of the running-out support from the end of the cantilever, a restriction method of the change of the design parameters which brings a good convergency, and so on.

View full abstract