To clarify the behavior of corrosion fatigue crack propagation in squeeze-cast Al-Si aluminum alloy, AC8A-T6, corrosion fatigue crack propagation tests were performed in this investigation under the condition of the stress ratio of 0.1, triangular waveform and a testing frequency of 10Hz in 3.0% saline solution using compact tension specimens with side grooves. Judging from the experimental results, corrosion fatigue crack propagation behavior is affected by the initial stress intensity factor range, ΔK, at the start of the experiment. When the test was started at low ΔK, the crack propagation rate was enhanced immediately after starting the test and thereafter became retarded, compared with that in air. It was observed that the retardation of crack propagation is due to blunting by the secondary cracks. On the other hand, enhancement of crack propagation is caused by dissolution around the eutectic silicon. The path of crack propagation during corrosion fatigue depended on the microstructure of material which consisted of dendrite and eutectic spheroidal silicon. Corrosion fatigue cracks propagated mainly along the eutectic structure and deviations in the path were larger than those under the test in air.

View full abstract

High-cycle torsional fatigue tests were carried out on carbon steel plain specimens (S45C) and the local deformation in the crack initiation process was observed by an electron microscope. Fine grid lines were drawn on the specimen surface with a diamond needle. The lines were too fine to influence the structure of the slip bands. The local strains were measured from the distorted grids. The cracks generate in the area with large local deformation (several hundred % shearing strain), that is, in the grain boundary between pearlite and ferrite or in the ferrite crystal grain sandwiched between pearlites.

View full abstract

The lives required for crack initiation and crack growth should be separately dealt with in fatigue life prediction. The crack growth life estimated on the basis of the da/dN-ΔKε relationship exhibits a steeper slope on the log Δε/2-log N_P diagram than that in the experiment, where da/dN and ΔKε are crack growth rate and strain intensity factor range, and N_P is life required for crack growth. The da/dN-ΔKσ and da/dN-ΔKε relationships show the stress dependency, unless crack closure is considered, where the strain intensity factor range is denoted as ΔKε. This, however, disappears in the da/dN-EΔKε_eff relationship where ΔKε_eff is the effective strain intensity factor range. Furthermore, the crack growth life curve determined from the da/dN-ΔKε_eff relationship agrees well with the test data. The crack-initiation-to-final-failure-life ratio shows the stress dependency, ranging from 0.4 to 0.6, in the high-cycle-fatigue region under rotation bending. Consequently, we can predict the EΔε/2-N_f curve by combining the stress dependency of the crack initiation life and the da/dN-EΔKε_eff relationship under rotation bending.

View full abstract

The measurement of fatigue crack tip opening stress σ_op and closing stress σ_cl occurring in actual equipment is generally very difficult. This is the reason why the range of effective stress intensity factor is not sufficiently applied to the fatigue crack propagation behavior of actual equipment. In this study, the measurement of σ_op and σ_cl in large size specimen is carried out by means of the strain interference method and the crack opening and closing points of long crack are successfully measured. Futhermore, σ_op and σ_cl in a specimen subjected to a rotating bending stress is measured. From these experimental results, the strain interference method is expected to be applicable to the measurement of σ_op and σ_cl in actual equipment.

View full abstract

In order to examine the influence of fiber flow formed by forging on the fatigue strength of a microalloyed steel having ferrite-pearlite structures, the fatigue test was carried out on the connecting rod made of the steel and on test pieces (T.P.) of which the angles of fiber flow (A.F.F.) were 0°, 45°and 90°to the axis of the T.P. The fatigue property of the steel tested was comparable with that of tempered steels such as SCM435. However, it can be suggested as a result of testing the connecting rod that the fatigue strength of this steel depends on the A.F.F. Furthermore, for a 45°T.P. of which the A.F.F. corresponds to the direction of the largest shear-stress plane, the steel decreased slightly in fatigue strength, by a few percent. Based on fractography using SEM, the cause of the anisotropy in the fatigue strength was found to be that the fatigue crack initiated in the fiber flow of 45°T.P. However, besides 45°T.P., through "the accelerated testing method for emerging a retarded metallurgical crack" propossed by the author, for 0°and 90°T.P., the fatigue crack initiated in the fiber flow under the cyclic loadings below fatigue limit led to a fish eye fracture surface.

View full abstract

The evolution of surface roughness of aluminum foil caused by fatigue slip deformation is dependent on cyclic stress and the number of cycles. In order to use aluminum foil as a fatigue gauge by utilizing its surface roughness, isoroughness curves are obtained and an empirical formula for them is proposed. Effect of material factor, i.e. grain size and thickness of the foil, on the evolution of roughness is also investigated. It is found that evolution of roughness is well represented by R_a = (σ^a/2/k_o) N^1/2 + m, where a and k_o and m are material constants. Accordingly, if the number of cycles N and the center line average height R_a is measured, the applied stress is evaluated by this formula. It is also found that the effect of the thickness on the roughness is remarkable, while the effect of grain size is very small. This is beneficial to control the strain sensitivity without causing the gauge length.

View full abstract

The effects of inhomogeneity phase (IP) such as Si and Al-Fe-Mn in Al-Si eutectic alloys on fatigue strength are considered to be virtually equivalent to small defects or nonmetallic inclusions. The maximum size distribution of the cumulative frequency of IPs was investigated. The square root of the projection area of the maximum IPs (√(area)_max) was found to obey the statistics of extreme. The fatigue strength prediction equation was successfully applied to predict the scatter band of the fatigue strength of three kinds of Al-Si new eutectic alloys. These alloys are usually used in the piston component of the high-performance compressor used for car air conditioners. It is suggested also that the combination of the fatigue strength prediction equation with the statistics distribution data on √(area)_max can be used for the quality control of materials before the adoption for design.

View full abstract

The fracture toughnesses of Si₃N₄ (two lots), SiC, TiB₂, ZrO₂ and Al₂O₃ were obtained by bending precracked specimens in three different ways: by the bridge compression (BC) method, by fatigue precracking from a crack introduced by the BC method (FP method), and by the controlled surface flaw (CSF) method. For comparison, fracture toughness evaluations by the chevron notch (CN) method and the indentation fracture (IF) method were also carried out. Furthermore, the effect of loading rate on fracture toughness of FP specimens was examined together with a crack propagation test under static fatigue. The K_IC values evaluated by FP, CSF and CN methods showed a reasonable agreement with each other for all six of the ceramics tested, while the evaluation by BC and IF methods tended to give considerably lower values than those obtained by the three methods. The normalization of these K_IC and K_ISCC values was done by a theoretical K_IC value, which was derived according to the Griffith criterion for unstable fracture. The different fracture modes of these materials were successfully characterized in correlation to the normalized fracture toughness.

View full abstract

This paper describes an apparent difference in the fracture toughnesses obtaind by using JSME S 001 and ASTM E 813 test standards, and its influence on a fracture load prediction of carbon steel pipes. Fracture tougnhness tests were conducted in air at room temperature on 1CT specimens prepared from a carbon steel pipe STS 42 (20 B, sch. 100) for LWR plants. Using these fracture toughnesses, analyses were carried out to evaluate the fracture loads of a carbon steel pipe with a circumferential through-wall crack. It is noted that the predicted fracture loads for the two pipes with same geometry are almost the same in spite of large difference in the fracture toughnesses.

View full abstract

A two-dimensional finite-element simulation of an impact fracture toughness test under high loading rates is performed. A crack subject to a time-dependent uniform pressure on its surface is considered, and caustics in reflection are generated by making assumption of the plane stress condition. The crack is assumed to begin to propagate when the theoretically exact value of K_I (t) reaches a critical value K_crit. The time history of the "measured" dynamic stress intensity factor K_I (t) is determined from the peak value of the time history of the "measured" K_I (t). It is found that the impact fracture toughness thus determined increases with the decrease in time-to-fracture when the loading rate is high. This may explain some of the physical meaning of the fracture incubation time .

View full abstract

This paper is concerned with the elasticity solution of a bimaterial interface crack. First, a critical analysis was performed on various interface crack models which were proposed to resolve the well-known contradiction of overlapping of crack surfaces associated with the classical solution by England et al. It was shown that Comninou's model was the most reasonable. Second, the elastic stress field obtained by the classical solution was compared quantitatively with that by Comninou's model for the case of pure tension. It was shown that the former agrees with the latter in the region of r &gsirn; 2_S with accuracy of about 90 % where r is the distance from the crack tip and s is the contact zone size. This result, together with the fact that s is quite small, shows clearly that the classical solution is actually valid in the case of pure tension. Accuracy of the r^-1/2-type expression of the stress field of the interface crack was also examined. Finally, definition of the stress intensity factors based on the classical solution was discussed, and a reasonable system of definition was indicated.

View full abstract

This investigation presents a method determining the complex shear moduls of damping materials from torsional resonance test of a rectangular covered rod bonded damping layer materials on both sides of the rod. This method (Composite Beam Test Method) provides an available expression with such determination, which is based on an exact solution of a torsional rigidity of the three-layered rod. The experimental results carried out for a commercially available damping rubber show reliable determination up to its transition temperature range. An error analysis for the covered rod, and also a sandwich rod with a damping core layer, clarifies that the resultant errors of the determined values of both rods greatly depend on the accuracy of both measured values of the thickness and shear modulus of each elastic layer. Furthermore, the error analysis clarifies that the sandwich rod is superior in this method to the covered rod due to the exceeding reduction of its resultant error in comparison to the one of the covered rod over wide range of the damping material constants.

View full abstract

Non-steady-state three-dimensional deformation around the front and the rear ends in shape rolling with grooved rolls is simulated by the rigid-plastic finite element method using the isoparametric hexahedral elements with eight nodes. To simulate complex non-steady-state deformation in shape rolling, a non-steady-state scheme using a remeshing technique is proposed. The workpiece is remeshed at each step of deformation to locate nodal points at the boundary of the interface between the roll and the workpiece. The present method can calculate not only the non-steady state deformation around the front and the rear ends but also steady-state-deformation. As examples of non-steady state simulation, rolling processes of bars and beams with grooved rolls are chosen. The deforming shape and the distributions of stress and strain for the workpiece in non-steady and steady states are computed.

View full abstract

The constitutive relation between thermophysical properties, temperature and hydrostatic pressure of liquid materials is studied on the basis of the first author's inelastic deformation theory (combined deformation mechanism model). We consider, in this paper, that the structure of a material changes simultaneously when the deformation mechanism or the constitutive relation changes at some temperature and hydrostatic pressure. An examination of experimental data by this concept lead to the following conclusions : the constitutive relation for each thermal property at the liquid region is divided into subregions bounded by the broken points : the constitutive relations of water for viscosity, thermal conductivity and saturation pressure are well explained, quantitatively by the proposed equations which are of the same type as derived from Arrehnius equation : the temperature at which the isobaric specific heat capacity is maximum seems to be the boundary between liquid and gas.

View full abstract

When a transversely isotropic sphere is subjected suddenly to a uniform temperature rise over its cross section, a stress wave occurs at the surface the moment thermal impact is applied. The stress wave at the surface proceeds radially inward to the center of the sphere. The wave may accumulate at the center and give rise to very large stress magnitudes, even though the initial thermal stress is relatively small. This phenomenon is called the stress-focusing effect. This paper analyzes the effects of these waves precisely using the ray theory. The numerical results give clear indications of a stress-focusing effect in a transversely isotropic sphere.

View full abstract

In the present study, the elastic-plastic deformation of circular tubes compressed laterally between two rigid plates was investigated. Finite-element analysis was used to investigate the distribution of stress and strain in the tube specimens during their deformation. In the analysis, a new incremental technique which can control the increment of displacement or applied load was used to save calculation time. Since a recalculation involving a number of transient elements from the elastic region to the plastic is performed in an incremental step, the accuracy of this technique is quite reasonable. The load-deflection curves of tubes obtained from the FEM analysis showed good agreement with that obtained from experiments.

View full abstract

In the present study, the effects of tube thickness and the mechanical properties of tube materials on the elastic-plastic deformation of circular tubes compressed laterally between rigid plates, were investigated. A series of lateral compression tests were carried out for a commercially pure aluminium 1197-O, and two aluminium alloys, 6063-T5 and 2014-T1, at a quasi-static rate. In most of the specimens of 2041-T1 and 6063-T5, fracture occurs during lateral compression tests. The location of relatively thinner specimens was the outer surface close to the horizontal ends and that of thicker specimens was the inner surface close to the vertical ends. In order to investigate stress and strain in the deforming tube, finite-element analyses were also performed. The load/deflection curves of tubes obtained from the FEM analysis showed good agreement with those obtained from experiments. The FEM results could explain well the change of fracture location.

View full abstract

The commercially supplied pure iron which was rolled to various reductions was cathodically hydrogen-charged in sulfuric acid solution to investigate the blister growth behavior in hydrogen-induced cracking. In addition, a computer simulation was carried out to predict the blister growth characteristics. The experimental result showed that the diameter of blisters increased with increasing charging time and rolling reduction. By assuming an increase in dislocation density (hydrogen trapping site) and an increase in growth rate of blister with increasing rolling reduction, the computer simulation could well predict the blister growth process and the three-dimensional distribution of blisters inside the material.

View full abstract

In this paper, we deal with the formulation of the axisymmetric boundary integral equation for the transversely isotropic elastic medium. We assume that the axis of elastic symmetry is coincident with the axis of rotation and that the elastic body is subjected to an arbitrary axisymmetric loading without torsion. The fundamental solution we adopt is the freespace Green's functions for the transversely isotropic body subjected to a torsionless ring like loading. These Green's functions were derived in our previous work. The numerical calculations are carried out, and the results are compared with the results by the finite-element method.

View full abstract

The plastic packages of ICs have been changing from insertion types to surface mounted types. Reflow soldering is used for surface mounted ICs onto printed circuit boards. During this process, packages are heated to above 200°C. If the encapsulant absorbs moisture, package cracking may occur during the reflow soldering. It is known that the cracks are caused by vapor pressure generated inside the packages, causing excessive stress in the plastic. In the former report, moisture diffusion analysis in the plastic was performed and vapor pressure could be obtained. Since the edge tip of the chip pad where package cracking occurs is a singular point within the elastic stress field, strength evaluation for package cracking is not established. In this study, the quantitative strength evaluation of the plastic is investigated by using the stress singularity theory.

View full abstract

The stress fields near the tip of a V-notch have singularities. The order of the stress singularity is dependent on the deformation mode. For the I deformation, the stress singurarity is 1/r^1-λ₁, and for the mode II deformation, it is 1/r^1-λ₂. Thus, the stress fields around the singular point must be described in terms of two pairs of constants λ₁, K_{I, λ1} and λ₂, K_{II, λ2}. Therefore, such numerical stress analysis, in which neither of the stress singularities is taken into account or only one stress singularity is taken into account, are generally not effectiue. In this study, a new method is presented for analyzing the problem of the V-notch in a semi-infinite plate. In this method, the two singular stress fields are taken into account at the same time, and the parameters K_Iλ1 and K_{II, λ2} are determined by the solutions of the equations. This method is also applicable to the problem of a composite plate because the stress singularity at the corner of the composite plate is essentially the same one as in the present problem.

View full abstract

A simple and accurate finite cylindrical element method is formulated for obtaining stresses and deformations in thick-walled cross-ply laminated FRP cylindrical shells. The effects of thermal residual stresses, the boundary conditions at both ends and the stacking sequence of unidirectional prepregs on the design of thick-walled cross-ply laminated FRP cylindrical shells are discussed. Although the thermal residual strain reduces the maximum tensile radial strain, its maximum is still higher and causes an interlaminar failure of the thick-walled cross-ply laminated FRP cylindrical shell. The optimum stacking sequence of these FRP cylindrical shells among the five stacking sequence models is also proposed to give the smallest value of radial strain.

View full abstract

Stresses in a slab with three orthotropic laminated layers are analyzed with the three-dimensional theory of transversely isotropic elasticity. The slab is subjected to a transverse concentrated or distributed load at the center part on the top face. To obtain qualitative knowledge for delamination of CFRP laminates, numerical results of interlaminar stresses are given for [0°_n/90°_n]_sym graphite/epoxy square laminates. It is shown that shear stresses on the upper interlaminar boundary are larger than those on the lower one. On each interfacial boundary, shear stresses are maxima just below the concentrated load point or the edge of the distributed load area, though these stresses are zero at the center point.

View full abstract

This paper presents a LDU decomposition method of the stiffness submatrices for developing a microcomputer/personal computer-aided finite element method. The LDU method needs to use the computer disk memory equipment such as floppy and hard disks to store the data obtained. The finite element method (FEM) using the LDU method is applicable to an analysis of a three-dimensional elastic block under tension. This FEM is available for an analysis of structures of a large-size matrix required for a memory capacity larger than that of the machine itself. This LDU method also contributes to the decrease of memory capacity in the conventional FEM. Therefore, the LDU method is effective for a microcomputer/personal computer-aided FEM.

View full abstract

This paper presents a boundary element application to determine the optimum impressed current densities and optimum location of electrode in a cathodic protection system. The potential within the electrolyte is described by Poisson's equation with nonlinear boundary conditions which are enforced based on experimentally determined electrochemical polarization curves. The optimal impressed current densities are determined in order to minimize the power supply for protection under the protecting conditions that the electric potential of every part of the structure to be protected should be less than some critical value. The solution is obtained by using the conjugate gradient method in which the protecting conditions are taken into account by the penalty function method. The boundary element method is employed to discretize the governing equations. Several numerical examples are presented to demonstrate the practical applicability of the pro-posed method.

View full abstract

In the case of combined loads such as the simultaneous action of cyclic torsion and plane bending acting on the machine element, neither the first nor second principal stress amplitudes can be detected by the copper electroplating method of stress analysis because this method utilizes the phenomenon that grown grains appear when the maximum shearing stress amplitude in the deposited layer attains the proper value to the plating. In this paper, a unique method adopting copper foil with a microcircular hole that enables separation and measurement of the first and second principal stresses in the element subjected to combined loads at both ends is proposed.

View full abstract

The change in electrochemical properties caused by microstructural changes was utilized for the nondestructive detection of material degradation. S/H tubes of fossil boilers were studied where carbide, Laves and sigma phase formation were detected by the anodic current peak I_P in an anodic polarization measurement of 316H stainless steel in lM KOH electrolytes. For practical uses, a small and light-weight electrochemical probe and a portable anodic polarization measuring system were developed; their effectiveness was shown through on-site measurements in many actual plants. Using measured I_P, the equivalent operational temperature of tubes after long-term service operation was estimated with reference to the I_P-t exp (-Q/RT) curve of aged materials. It is suggested that their metallographic changes can be evaluated by a parameter of I_P/I_PC, where Ipc is estimated as to be the threshold value of I_P. It is also indicated that their mechanical damage markedly increases by nucleation and growth of Laves phases at I_P/I_PC value over 0.7.

View full abstract

Although photoelastic analysis has a great advantage in the visualization of the internal stress state which arises in a two-dimensional model, several difficulties still exist in accurate measurement of fringe distribution around a point of singularity such that concentrated load is applied, owing to the insufficient resolving power of ordinary equipment for image treatment. In order to overcome these kinds of difficulties, the authors propose a new concept and algorithm for complement or compensation of blurred photoelastic isochromatic data in an extremely condensed area. Taking the well-known theoretical solution near a singular point in two-dimensional elasticity into account, collapsed data can be reconstructed well by determination of unknown constants in the theoretical functional solution using an optimizing approximation method with accurately measured experimental data slightly apart from the point.

View full abstract

To clarify the mechanical efficacy of the reconstruction of the posterior element following the cervical multiple laminectomies, stresses around cervical vertebrae are calculated in this paper. The cervical vertebrae are coated by the compact bone, and the deformation of these can be negligible in comparision with those of ligaments, intervertebral discs and cartilages. Thereby, the stresses around the cervical vertebrae are successfully solved with use of the rigid body-spring model method. The numerical calculations are carried out and it is shown that the reconstruction of the spinoligamentous complex is indispensable to the cervical laminectomy which is performed for the decompression of the spinal cord.

View full abstract