In this study, the principle by which the ribs are formed in a leaf of tree is applied to resolve the problem of stiffeners in plate and shell structures. The generalized cellular automata are used to form a structure shape and the parameters are evolved by the genetic algorithm. The effectiveness of this method is confirmed through the numerical examples of the minimum compliance design problem of thin plates. Traditional constraint conditions such as the maximum plate thickness, plate volume and so on can be omitted by taking the ratio between the biggest deflection of plate with the stiffener and the corresponding deflection of uniform thickness plate under the equivalent volume as the objective function. The obtained optimum solutions are very similar to the leaf ribs and the shapes with more degrees of freedom.

View full abstract

For the optimization problem of large structure, especially of crash worthiness analysis, finite element method is often used to calculate the structure response. However, the crash worthiness optimization problem is characterized by the non-linearity and noisy behavior (numerical inaccuracies) of response, it leads to tremendous computation cost and will fail to find the correct optimal solution when the sensitivity based optimization strategies are used. In this paper, we propose a new function approximation method based on response surface methodology, to smooth and simulate the structure response, by using Holographic Neural Network (HNN), to solve the optimal design. Sequential training scheme is applied initially to increase the accuracy of function approximation. Then we discuss the initial designs and the added designs selection to effectively reduce the analysis times (function calls). This optimization approach is illustrated for two numerical examples comparing with the Sequential Quadratic Programming (SQP) and shows an effective procedure of reducing the function calls. Finally, an industrial crash worthiness optimization design problem is carried out.

View full abstract

Changes in surface shape and microstructure of a thin film induced by heating and cooling are simulated by means of molecular dynamics method. A simple three-dimensional model of iron atoms as an example of bcc metal with periodic boundary conditions is employed to identify the temperature for melting and crystallization and the rate by use of Finnis-Sinclair potential function. Simulations during heating and cooling processes of a thin film are carried out with a model which has free surface and fixed boundary in one direction and periodic boundaries in the other two directions being a model of optical disk of phase-change type used in DVD system. A rippled shape of the surface is observed when a concentrated part of the surface is heated to be molten state, while planer surface is achieved by flat heat source. Such behavior depends on the difference in the surface tension and surface energy. The surface tension decreases as increasing temperature, and the tension of solidified crystal reveals to be larger than that of liquid. The potential energy near the molten surface gives higher value than that of the bulk, and it may minimize the surface area. Amorphous structure is obtained by rapid cooling of the molten region while it becomes crystallized by moderate cooling, which causes the small amorphous spot on the surface corresponding to a memory digit, and the digit vanishes by the subsequent reheating.

View full abstract

In this study, the molecular dynamics (MD) is used to simulate an adhesive wear, where the simulation model is simplified as a single α-Fe (bcc material) under shear deformation. The adhesive plane is supposed to be (001) plane and the slip (shear) direction be [010] direction. First, MD analysis is carried out with/without periodic boundary conditions in [100] and [010] directions to clarify the influence of periodic boundary condition. Next, the shear deformation velocity is varied and its influence on the behavior of adhesive wear is examined. The results are summarized as follows : (a) The periodic boundary condition significantly affects the deformation of adhesive wear. That is, the existence of free boundary shows a very important roll in wear. (b) The deformation velocity does not affect the form of adhesive wear, only the time to adhesive wear becomes shorter as velocity becomes larger.

View full abstract

In order to understand precisely the structure and properties of interface in solids, it is important to elucidate the disordered atomic configuration on the basis of quantum mechanics. In this study, the structure and energy of Σ9 tilt grain boundary in aluminum is analyzed by ab initio molecular dynamics simulation based on the density functional theory. The results obtained are summarized as follows. (1) The grain boundary has a pair of atoms which are close to each other, and one of them moves along the grain boundary during relaxation. (2) The maximum valence charge density near the grain boundary is 0.03 a.u.^-3, which is nearly equal to that in single crystal. (3) The grain boundary energy is 0.431 J/m², which is smaller than that of Σ5 tilt grain boundary, 0.458 J/m². Moreover, the validity of Morse potential and EMT (Effective Medium Theory) potential is examined by comparing the atomic force and energy with those obtained in ab initio simulation. (1) The grain boundary structure and the force acting on each atom evaluated by EMT are similar to those in ab initio simulation. However, the grain boundary energy obtained by EMT potential is much smaller than that obtained by ab initio simulation. (2) Morse potential does not give adequate values of atomic force as well as grain boundary energy.

View full abstract

In the present investigation. a computational procedure is developed for behaviors of stress waves induced in elastic solid material. Finite volume method is applied to the elastodynamic equations and flux vectors for computations are evaluated according to Godunov method. In order to obtain second order accuracy, the two-stage Runge-Kutta method is used for time integration and the MUSCL approach is employed for flux evaluation. Simulations are conducted for stress waves in a rectangle solid and in a wedge at the appex angle of 165 degrees. The computational results are quite reasonable.

View full abstract

The boundary element method (BEM) does not require a domain integral in steady thermoelastic problems without heat generation. However, in problems with distributed heat generation, the domain integral is necessary. This paper shows that the axisymmetric problem of steady thermoelasticity with nonuniform heat generation over the region can be easily solved without a domain integral by means of the improved multiple-reciprocity boundary element method. The conventional multiple-reciprocity boundary element method is not suitable for the complicated heat generation. In the improved multiple-reciprocity boundary element method, the heat generation is interpolated by boundary integral equation.

View full abstract

In the previous papers we proposed a method for shape and topology optimization (i.e., layout optimization) of two-dimensional finite element models of mechanical structures using genetic algorithm. The method is contrived in such a way that optimal layouts are obtained by selecting the layouts of the models as individuals, and by adopting removal and addition parameters of the elements as chromosomes. A certain relationship between the both parameters are certified to exist in various experiments. We, therefore, proposed a finite-difference-type of sensitive optimal method using the relationship, which can be called a deterministic GA approach with a single individual. It is demonstrated that the final layouts obtained by the approach are similar to those gained by GA and cellular automaton.

View full abstract

This paper proposes a shape optimization method to reduce stress concentration of jointed bodies by paying attention to stress singularities at the ends of bonded interface. The shape geometry is expressed by the Bezier function and is optimized using the sequential linear programming. The effectiveness of the proposed shape optimization method is demonstrated through a numerical example of single-lap joint. Results can provide much information for optimum design of single-lap joint.

View full abstract

The J-integral has been used widely in treating elastic-plastic crack problems and in most cases it has been assumed that the value of J-integral is independent of integral path. In this paper, the physical meaning of J-integral and the dependence of it on integral path are examined in the cases of tension of strips with a central crack through the FEM elastic-plastic analysis based on the incremental deformation theory and the concept of non-linear crack mechanics developed recently by the authors.

View full abstract

Rotating bending fatigue tests were carried out on the plain and notched specimens of annealed and pre-strained 0.45% carbon steels. The obtained results are discussed based on the concept of linear notch mechanics. The main results are summarized as follows : (1) Fatigue limit of plain specimen σ_w0 increases monotonously with the increase in the value of Vickers hardness. (2) The value of notch radius ρ₀ at the branch point is almost independent of pre-strain. (3) The notch sensitivity for σ_w1 and σ_w2 are almost independent of pre-strain. Therefore, the values of σ_w1 and σ_w2 of pre-strained specimens can be predicted through the value of σ_w0 of pre-strained specimens and the relations between of K₁σ_w1-1/ρ and K₁σ_w2-1/ρ of annealed specimens.

View full abstract

The effects of low temperature annealing, bluing, on the S-N curve of eutectoid steel wires drawn up to the strain of 3.5 were studied. Vickers hardness tests and measurements of residual strain along the wire axis using Heyn's method were also conducted. The fatigue limit as well as the hardness and the residual tensile stress in the surface layer of the wire increased with increase in the drawing strain. Bluing increased the hardness at low temperature up to 300°C and then decreased with the temperature rise. The residual stress diminished monotonically to zero with the bluing temperature rise up to 500°C. The bluing increased the fatigue limit, being highest at 400°C, through combined effects of strain aging and recovery of the structure.

View full abstract

The effects of residual stress on fatigue strength at a weld toe in multi-pass fillet-welded joints were evaluated. The residual stresses in the specimen were varied by controlling the sequence of welding passes. The residual stress at the weld toe was 80 MPa in the specimen whose last welding pass was on the main plate, but it was 170 MPa in the specimen whose last pass was on the attachment. The fatigue strength of the specimens was nearly the same at high stress amplitude, but the fatigue strength of the specimen whose last welded pass on the main plate was larger than that of the other specimen at low stress amplitude. This difference is due to the magnitude of the initial residual stress and the relaxation of the residual stress under the cyclic loading. The effects of the residual stress were shown in the modified Goodman diagram, in which residual stress is treated as a mean stress.

View full abstract

To assess the effectiveness of various repair schemes for components of an overhead crane, weld repaired gusset specimens, several kinds of plate specimens with cracks repaired by welding, and cover plate specimens were fabricated, and the fatigue strength was determined for various specimens subjected to a simulated program loading of an overhead crane. The fatigue strength of weld repaired gusset plate specimens was found to be approximately the same as that of as-welded ones. The fatigue strength of weld repaired specimens with partial-penetration of welds decreased with an increase in the initial crack length. In case of the one-surface weld repaired specimens, the fatigue strength was not enough, while the specimens with cracks completely repaired by welding from both surfaces displayed sufficient fatigue strength to satisfy the requirement of a design code of cranes. In case of the one-surface cover plate specimens, the fatigue strength reduced significantly due to high stress concentration. Hence, this type of cover plate is an inadequate repair method.

View full abstract

The effect of stress ratio (R=σ_min/σ_msx) on subsurface fatigue crack initiation behavior in high-cycle fatigue of beta-type titanium alloy, Ti-15V-3Cr-3Sn-3Al, was investigated using smooth specimen subjected to axial reversed loading in air at room temperature. Two types of microstructure prepared using different aging time conditions included the partially precipitated alphaphase in beta-phase and the fully precipitated one. The first part of S-N curves at high stress amplitude region tested under R=-1 and 0.05 looks similar to curves reported in the literature, however after the curve flattens out there was a sharp drop in the fatigue strength at low stress amplitude region, accompanying a transition of crack initiation site from surface to interior. The number of cycles to the transition depended on R but did not depend on the microstructure. Fatigue fracture of all specimens tested under R=0.5 and 0.7 showed the subsurface crack initiation. No defect was observed at the subsurface crack nucleation sites, but the region of the nucleation site had a grain morphology of facet pattern. Limit of fatigue fracture occurred by surface crack initiation was controlled by the threshold of formation of slip in beta-phase due to cyclic hardening, and described by the Gerber-relationship characterizing mean stress effect.

View full abstract

In order to examine the precise fatigue mechanism of so-called "fish eye" type fatigue failure emerging in very high cycle fatigue, rotating bending fatigue tests were carried out on a high strength low alloy steel. Differently fatigue damaged specimens were prepared and their cross section was carefully observed so as to gather statistical information of initiation and growth of subcracks in subsurface region. The results showed that penny-shaped subcracks were nucleated at very small inclusions before stress cycles of 10⁴. All of the subcracks initiated at inclusions grew up at maximum to a length of the non-propagating crack, being formed in the matrix of subsurface and controlled the "true fatigue limit" of the material. To observe the internal crack growth on specimen surface, a surface-wrapped specimen was fatigued. The result revealed that the shielding of oxygen results in decreasing the fatigue strength.

View full abstract

In the present study, the static and cyclic fatigue behavior of alumina ceramics was investigated under different environments of air and ion exchanged water. Bend specimens, with an indentation induced flaw at the center, were tested under a static and cyclic load applied by four-point bending. The test results have shown that in the case of static fatigue, time-to-failure in water was somewhat lower than that in air, while, in the case of cyclic fatigue, there was no significant difference by environment. The K₁-V characteristics determined by double torsion method that utilizes relaxation technique were used to predict time-to-failure under static and cyclic loading of alumina. The predictions in lifetime were significantly longer than the experiments for both environments of air and water. The predictions in lifetime could explain the tendency of lifetimes in static fatigue qualitatively, but the lifetimes in cyclic fatigue were significantly lower than the predictions for both environments of air and water. These facts show that there may be any effect of stress cycling on time-to-failure in alumina.

View full abstract

The rotating-bending fatigue tests on a TiNi shape-memory alloy wire were carried out in air and in water, and low cycle fatigue was investigated. The fatigue life in water shortened with increasing temperature but did not depend on frequency. The temperature of the wire in air increased with increasing frequency and the fatigue life shortened. The fatigue equation was proposed to describe the fatigue life depending on strain amplitude, temperature and frequency. The fatigue life was estimated well by the proposed equation. The shape memory processing temperature did not affect the fatigue life.

View full abstract

In order to clarify the role of slip deformation on cavitation erosion, vibratory cavitation tests were carried out for a NiTi shape memory alloy in martensitic phase and austenitic phase. This alloy deforms plastically by the combination of slip and twinning deformation, and the erosion resistance of NiTi shape memory alloy is 500 to 1000 times higher than that of SUS316 stainless steel. The erosion resistance of austenitic phase is superior to martensitic phase, because the accumulation of slip in the austenitic phase is slower due to the superelasticity than that in the martensitic phase. The martensitic phase test with intermittent shape recovery shows a better resistance without the shape recovery, since the deformation twins produced by cavitation erosion test are eliminated by shape recovery, resulting in the delay of slip accumulation. However, the amount of shape recovery was decreased with test duration. The decreasing rate of shape recovery has a good linear relationship with the volume loss rate. It was concluded that the accumulation of slip deformation corresponds well with the volume loss rate.

View full abstract

A study of hydrogen embrittlement on high strength steels was carried out with thin walled hollow and thin plate specimens, of which stress state would be plane stress condition, to compare them with solid round bar specimens which had been subjected to plane strain condition. The specimens were cathodically charged with hydrogen under sustained load conditions. The delayed fracture life of thin walled hollow and thin plate specimens lasts longer compared with the solid bar specimens. SEM fractographic examination showed that the essential process of delayed fracture, in which a formation of quasi-cleavage (QC) crack triggers the intergranular (IG) crack after subcritical propagation into a critical size, is unchangeable irrespective of the stress condition. Consequently, total life of the hydrogen charged steel is mainly determined by the period of crack initiation and subcritical growth of the QC crack.

View full abstract

Size effect in microhardness measurement has been reported and discussed by many researchers. In spite of the efforts, decisive explanation for the size effect is not yet available. In this paper we discuss a new empirical hardness determination method for force-penetration depth curves. In order to avoid effects of grain boundaries, precipitations and deformed layer, electrolytically polished metallic single crystals of Ni, Fe, Mo and W were used. The analysis begins with relationship between force and Vickers Hardness for the same penetration depth. Then, four different types of functions are fitted to force-penetration curves to determine final forms with which one can calculate Vickers Hardness from force and penetration depth at any point of force curves.

View full abstract

In this paper, a method of evaluating the static strength of a V-shaped notch based on the singular stress field at the notch tip is studied. The singular stress field is defined by two parameters, K_{1, λ1} and K_{11, λ2}, which correspond to the intensities of the symmetric stress field and the skewsymmetric field, respectively. Four kinds of fracture criteria are considered ; two of them are based on the tensile strength δ_B and the other two are based on the fracture toughness K_1C. The usefullness of the criteria is investigated through the experiment results carried out on plane speciments of acrylic resin having a sharp notch for various notch configurations such as the opening angle, the inclined angle and the notch depth.

View full abstract

A molecular dynamics (MD) simulation was conducted on a thin wire of a single crystalline nickel subjected to tensile deformation in the [001] direction. When the applied strain, ε_zz, is smaller than 0.116, the crystal deforms uniformly and homogeneously, being perfectly elastic. A sudden drop in the applied load takes place at the moment of ε_zz=0.116 by a partial yielding. Dynamic process of the local strain concentration and the dislocation nucleation was analyzed through the measurement of a resolved shear strain, γ_rss, on every lattice plane in the direction of atom migration due to the partial dislocation : (1) When ε_zz reaches 0.116 at the first step of the present MD simulation, a high strain concentration region appears on the (010) surface, where the magnitude of γ_rss is larger than 0.13. (2) γ_rss increases during the relaxation time for the arrangement of atoms under a constant global strain of ε_zz=0.116, which is caused by the relaxation of the atomic potential energy. (3) A partial dislocation nucleates at the region when γ_rss reaches about 0.20. (4) The dislocation begins to glide on a (III) plane where γ_rss is most concentrated. It is noteworthy that the local strain increases during the relaxation of atomic arrangement before the dislocation nucleation, and that its concentration proceeds to nucleate the dislocation. In order to clarify the mechanism of the strain concentration on a smooth surface of the perfect crystal and the magnitude of the critical resolved shear strain : γ_rss=0.13, the crystal was subjected to an ideally homogeneous deformation, and its lattice instability criterion was evaluated based on the Born's concept. It was found that this homogeneously deformed lattice begins to be unstable in the direction of the resolved shear strain when γ_rss just reaches 0.13 being equal to the aforementioned crystal. Thus the localization of the resolved shear strain is attributed to the local lattice instability.

View full abstract

Thermal stress failure of square type surface mount device (SMD) such as ceramic capacitor has been evaluated in surface mounting process on metal base printed wiring board (PWB). For the evaluation, a series of transient thermal stress analyses with FEM model and thermal stress failure experiments for the surface mount assembly were carried out with complementary relation. As a result, thermal stress failure probability of the capacitor is estimated by utilizing the distribution function of three-point bending strength and the maximum value of major principal stress in the vicinity of failure initiation point. Additionally, influence of deformation properties of solder and insulation layer, which is formed on the surface of the PWB, on the failure probability is also evaluated quantitatively by introducing two parameters E_tg and σ_ys. E_tg is related to glass transition temperature and Young's modulus of the layer and σ_ys indicates yield stress of the solder at ambient temperature. Consequently, the present evaluation has led us to appropriate choice of the solder and the layer in order to prevent the capacitor from the thermal stress failure in the surface mounting process of square type SMD on metal base PWB.

View full abstract

Plastic shear-bending buckling under seismic loadings is one of the major problems in the structural design of FBR main vessels. Pseudo-dynamic and dynamic buckling tests of cylinders were performed in order to study the effects of nonlinear seismic response on buckling strength, ductility, and plastic response reduction. The buckling strength formulae and the rule for ductility factors both derived from static tests were confirmed to be valid for the tests under dynamic loads. The displacement-constant rule for response reduction effect was modified by acceleration amplification factor in order to maintain applicability for various spectral profiles of seismic excitations. The response reduction estimated by the proposed rule was reasonably conservative for all cases of the pseudo-dynamic and the dynamic tests. Finally, a seismic safety assessment rule was proposed for plastic shear-bending buckling of cylinders, which include the proposed response reduction rule.

View full abstract

Hardness test is one of the basic material testings. This investigates strain behavior in polycarbonate plate by loaded spherical indentator using the strain-freezing method to establish a method of evaluating the material properties of high polymer materials which have been widely used as machine parts and structural members because of a high elastic modulus and strength. As a result, the strain-freezing method was found to be effective for analyzing strains in polycarbonate by loaded spherical indentator. Furthermore, the relation between the photoelastic fringe order and principal strain difference is found to be proportional as well as the relation between the total strain.

View full abstract

A damper made of coiled steel is often used for base isolation systems of buildings. Since large relative displacements occur at the isolation system, large deformation elastic-plastic analysis is necessary in the designing of the damper. In this paper a method of the analysis using the degenerated beam element is presented. A simple uniaxial constitutive equation that can be used in cyclic loadings is proposed. The equation uses a bilinear function for a skeleton curve and the Ramberg-Osgood functions for hysteresis curves. A simple method for the modeling of slips at boundaries of the damper is also proposed. A method for evaluating parameters of the constitutive equation and the model of the slip is shown. With the use of the identified parameters, the results obtained by the analyses agreed well with those of experiments.

View full abstract

Recently, damage in the adhesive layer has been investigated in-situ microscopic observation under several load conditions. However, observations are limited on the surface of the adhesive layer. If the tensile test of the butt joint whose adherend thickness is approximately as same as the thickness of adhesive layer can be realized, damage in the interior of the adhesive layer may be observed using transmitted light. This will facilitate elucidation of the fracture mechanism of adheisvely bonded joints. In this study, adhesively bonded butt joint specimens were made using thin steel plate with thicknesses of 0.3 mm and 1 mm, where thiokol modified and rubber modified adhesives were used. Using these butt joints, two kinds of tensile tests were conducted ; One to obtain stress-strain curves of these butt joints, and the other to investigate the damage in the adhesive layer in-situ microscopic observation. Moreover, stress distributions in the adhesive layer were also analyzed using the finite element method. Main results obtained are as follows. (1) The stress-strain curves indicated that yielding behavior of the butt joints bonded by thiokol modified adhesive was mainly due to the plasticity of the adhesive layer, while that joint of joints using rubber modified adhesive was mainly due to damage of the adhesive layer. (2) In-situ observation indicated that for butt joints bonded with thiokol modified adhesive. initial debonding occurs at the adhesive/adherend interface near the end of the adhesive layer where maximum shear stress is concentrated. On the other hand, for butt joints bonded with rubber modified adhesive stress whitening was observed in the adhesive layer, where the whitening zone agreed well with the counter of the mean hydrostatic stress. Then, initial crack appeared in the middle of the adhesive layer where mean hydrostatic stress was higher than at the adhesive/adherend interface.

View full abstract

The stress distributions in joints combining adhesives with bolts are analyzed by using axisymmetric theory of elasticity as a three-body contact problem when torsions are applied to the joints. The joints consist of two hollow cylinders and they are fastened by a bolt and nut after being joined with an adhesive. When an external torsion is applied to a combination joint, an increment/decrement of torsion occurs in the bolt. A method for estimating the variation of torsion in the bolt is proposed. In addition, using the interface stress distribution, a method for predicting strength of the combination joint is proposed. The torsional load occurred in the bolt was measured using strain gauges. Strength of the combination joint was also measured. A fairly good agreement was seen between the analytical and the experimental results. It is shown that the increment of torsional load occurred in the bolt is small substantially. It is observed that the strength of combination joint is greater than that of adhesive joints. It is also found that the combination joint strength increases as the intial clamping force increases and that it is greater than that of bolted joints without adhesives.

View full abstract

This paper is concerned with the theoretical treatment of transient problem of three-dimensional piezothermoelasticity involving a rectangular plate of crystal class mm2 due to partially heat supply in a transient state. We obtain the exact solutions for the three-dimensional temperature change in a transient state and the three-dimensional transient piezothermoelasticity of a simple supported plate. As an example, numerical calculations are carried out for a cadmium selenide solid, and the piezoelectric effects and the effects of the electric surface charge on the displacements, stresses, electric potential, and electric displacements are examined precisely.

View full abstract

In this paper, a genetic algorithm is applied to an optimization problem of material composition for a hollow circular cylinder of step-formed functionally graded materials. In the first place, the one dimensional transient temperature distribution for a laminated composite hollow circular cylinder is analyzed theoretically. Furthermore, making use of Airy's stress function method, the thermal stress components are formulated under the mechanical condition of being traction free. As a numerical example, the hollow circular cylinder composed of zirconium oxide and titanium alloy is considered. And, as the optimization problem of minimizing the thermal stress distribution, the numerical calculations are carried out making use of genetic algorithm without supposing a distribution function of material composition, and the optimum material composition of each layer is determined taking into account the effect of temperature-dependency of material properties.

View full abstract

Nonlinear stress-strain (S-S) properties of a plain woven glass fabric composite were measured under three different loading paths which gave the same final stress state. Thin-walled tubular specimens were used. In all tests, acoustic emissions (AE) were measured to distinguish damage accumulation such as debonding, matrix cracking and fiber breakage. Internal damage of failed specimens was also examined using an optical microscope. The S-S relations up to failure were affected by the biaxial stress ratio α (the ratio between normal and shear stresses) and loading paths. Both damage accumulation and its sequence were also affected by loading paths. And they were characterized by microphotographs and AEs such as total AE event and primary frequency. However, the strength did not depend on the loading path. The material failure was governed by the final stress condition regardless of the difference of loading path.

View full abstract

In this research work, we investigated the torsional motion of the left ventricle about the long axis, which has been assumed to be one of the components of the normal systolic motion. The myocardial wall motions were analyzed for four normal volunteers and a patient with hypertrophic cardiomyopathy (HCM), by using the magnetic resonance tagging technique. It was recognized from the results for the normal volunteers that there was the counterclockwise rotation of the apical side plane with respect to the basal side plane during the ejection period when looked toward the base from the apex, and that the torsional motion in the septal wall was minimum. It was confirmed that this torsional motion did not appear in the diseased part of the patient with HCM. This study may suggest that the torsional motion of the left ventricle is an important component of the normal myocardial wall motion, and that the quantification of the torsional motion is useful for the evaluation of the cardiac contractility.

View full abstract