Toughening mechanisms of three types of rubber toughened poly (methyl methacrylate) (RT-PMMA) were investigated under mode I loading condition by optical and electron microscopies in conjunction with the quantitative evaluation of mode I fracture toughness. Polarized optical microscopy clearly exhibited damage zone development ahead of a crack-tip of the RT-PMMAs. The three RT-PMMAs revealed different shapes of the damage zone. Transmission electron microscopy exhibited microcrazes nucleated in the equator of rubber particles within the damage zones. Extensive deformation of rubber particles corresponding to localized shear yielding of the PMMA matrix was also found in a region close to a propagating crack-tip. In addition, cavitation of rubber particles was observed in the vicinity of the crack. It is therefore understood that the toughening of the RT-PMMAs is due to energy dissipation caused by the microdamage formations such as microcrazing, matrix shear deformation and rubber particle cavitation ahead of the crack-tip.

抄録全体を表示

In this work, we investigate the toughening mechanism of the rubber modified epoxy resin. The fracture toughness (K_IC) is measured using CT specimens for three kinds of rubber modified epoxy resin with different rubber content. The damage zone and rubber particles around a crack tip of a damaged specimen just before fracture are observed by a polarization microscope and an atomic force microscope (AFM). Both the fracture energy (G_IC) and the size of damage zone increase with the rubber content below 15 wt%. The size of the rubber particles can be qualitatively correlated with the G_IC and the size of damage zone. The cavitation of the rubber particles inside the damage zone is observed, which is expected to be main toughening mechanism by rubber particles. The stress which causes the cavitation of rubber particles is estimated by the Dugdale model.

抄録全体を表示

The relationship between miscibility and fracture toughness for poly (styrene-co-N-phenyl-male-imide) (SMI) and poly (styrene-co-acrylonitrile) (SAN) blends was studied as a function of acrlonitrile (AN) content in SAN. Miscibility and blend morphology were characterized using differential scanning calorimetry (DSC) and field emission scanning transmission electron microscopy (FE-STEM) respectively. Stress intensity factor at crack initiation, K_li, was obtained to evaluate fracture toughness. DSC results showed that SMI/SAN blends changed from a miscible to an immiscible state with an increase of AN content in SAN. FE-STEM provided excellent phase contrast for morphology of immiscible SMI/SAN blends. Generally speaking, Kn of miscible blends was higher than that of immiscible blends. It has been shown that toughness enhancement can be achieved with increases of AN content in SAN and molecular weight of SMI and SAN in a miscible blend.

抄録全体を表示

This paper describes the quasi-static lateral compression tests of high-strength, high-elastic-modulus fibers. The fibers examined were γ-alumina fiber (Altex) by Sumitomo Chemical Co. Ltd., two types of carbon fibers by Mitsubishi Rayon Co. Ltd., and four types of aramid fibers by du Pont and Teijin Ltd. Special attention was paid to the environmental influence of water absorption, electron or ultraviolet radiation on lateral compression behavior of aramid fibers. The mechanical testing machine employed was a specially designed testing machine (load range : 0.1mN∼5N, accuracy : 0.02 mN), which enables mechanical testing including fatigue of microelements. By using the testing machine, the lateral compression of fibers whose diameter ranged from 7 to 17 μm was achieved with sufficient precision. The lateral compression behavior of the carbon fibers and ceramic fiber exhibits brittle nature : most of load-displacement curve follows the elastic deformation, with a final catastrophic failure into small pieces. However, that of the aramid fibers shows more ductile nature : a very early stage of load-displacement curve is elastic, and the rest is plastic. Final catastrophic fracture that is observed in ceramic and carbon fibers do not occur, with large amount of plastic deformation. The influence of electron radiation on the transverse elastic modulus of aramid fibers is negligible. However, water absorption as well as ultraviolet radiation lowers the transverse elastic modulus. The fiber surface was closely examined by using an atomic force microscope, and the influence of environment on surface degradation and deformation behavior is discussed.

抄録全体を表示

In recent years, films have been widely used as the small-sized and light-weight materials for electronic parts. Therefore, it is important to know mechanical properties of the film. For reliability of these parts, it is particularly necessary to evaluate fatigue fracture properties of the film. In this study, the effect that rolling direction of the copper film influenced slip initiation, microcrack initiation, and fatigue life was studied. The two types of specimen that rolling direction was parallel and perpendicular to loading direction were prepared in film fatigue testing. moreover, stress intensity factors were calculated using BEM analysis. As a result, a slip and a crack initiate at less number of cycles and lower stress amplitude in the film loaded to the perpendicular direction than in that loaded to the parallel direction. Conversely, a crack in the film loaded to the perpendicular direction propagates slower than that in the film loaded to the parallel direction. There was a difference between the two types of specimen in crack observation on the specimen surface. The reason was probably that deformation restraint due to grain boundary was different in accordance with the aspect ratio of the grain stretched by rolling.

抄録全体を表示

A governing parameter of electomigration damage in metal lines of electronic packages for the prediction of void or hillock formation has been proposed considering two-dimensional distributions of current density and temperature, where not only the polycrystalline line but also the bamboo line structure was considered in a simplified manner. The usefulness of the parameter has also been verified through good agreement of predicted void volume with that measured in the experiment, where the singularity of current density at the corner of an angled polycrystalline line was varied. In the present paper, the effects due to reduction of current concentration at the corner of angled line and also the singularity of temperature gradient on the void formation are investigated experimentally. In addition to the experimental data obtained so far, the present experimental results are used for the verification of usefulness of the parameter in more detail. The predicted void volume is always found to be in good agreement with the experimental result and thereby the rationality of using the parameter is established.

抄録全体を表示

Nanoindentation measurement for TiN, Al₂O₃, SiO₂ and MnS inclusions in SCM440, SUP12 and SWOSC-V steels was carried out, using an AFM ultra-micro hardness tester and a special lever which has a three-sided pyramidal diamond indenter with the edge angle of 60°. Results obtained are as follows. (1) A force-penetration curve was obtained on the inclusion with the diameter smaller than 2 μm, because the AFM ultra-micro hardness tester has the imaging capability in a nanometer scale. (2) Vickers hardness values were estimated from force-penetration curves as 100, 400∼450, 900∼2 500 and 1 000∼4 000 for MnS, SiO₂, Al₂O₃ and TiN inclusions, respectively, using two experimental equations between nanoindentation measurement and micro-Vickers measurement. (3) The hardness measurement for the base metal showed that the Vickers hardness from the nanoindentation measurement was lower than that from the micro-Vickers measurement. This was because the effect of grain boundaries and precipitates on the hardness was stronger in the micro-Vickers measurement.

抄録全体を表示

Since the surface morphology of materials can be observed with atomic-scale resolution, scanning atomic force microscopy (AFM) is a powerful technique to study mechanisms of fatigue and fracture of metallic materials. In the present study, slip band formation and fatigue crack initiation processes in α-brass were observed by means of AFM. Surface of a fatigued specimen was observed under maximum stress. unloading state, and minimum stress at fatigue cycles. In the initial stage of the fatigue process, slip bands which formed only under tension stress or compression stress were observed. These kinds of slip bands, however, disappeared shortly. Under tension stress, cracks could be detected easily just after initiation. For specimens at the unloading state, cracks just after initiation were hardly detected from the AFM images, By measuring the size of slip bands at the unloading state, however, crack initiation from slip bands could be detected. Before crack initiation heights and width of extrusions and the depth and width of intrusions gradually increased with the number of cycles. After crack initiation, one of these values changed drastically, depending on the slip band angle relative to the stress axis and the shape of slip bands.

抄録全体を表示

Quite unique and interesting behaviors of shape memory alloy under complex loading conditions, including the complicated path dependency, have been observed in systematic experiments by applying some combiened loads of axial force and torque to the thin-walled tubular specimen of Cu-based polycrystalline shape memory alloy. A set of constitutive equations is proposed, which can describe the complicated behaviors observed in the experiments. The mesoscopic approach is employed in the formulation because the complicated behaviors are closely related with the microstructural changes of materials, and the obtained equations show a reasonable coincidence with the exprimental results. In this first paper, the process of modelling and the detail of formulated constitutive equations are described, and the comparison between experimental resutls and computed results by using the proposed constitutive equations are shown in the second report.

抄録全体を表示

In the numerical simulation of mechanical behaviors of composite materials and structures, a novel hierarchical modelling technique is proposed. Not only the global deformation but also the mesoscopic stresses are analyzed precisely with less human efforts than conventional methods. Four level hierarchy is defined for the textile composites, and the homogenization method and finite element mesh superposition technique are adopted. The latter technique makes it possible to overlay arbitrary local fine mesh on the global rough mesh. A panel with a rib made of knitted fabric composite material is analyzed by the proposed method. Three-dimensional modelling of the mesostructure of knitted fabric composite material is also shown. The validity of the proposed method was examined, and the mesoscopic stresses are calculated easily.

抄録全体を表示

A numerical method is presented to simulate the dynamics of particle dispersed systems in complex flows during an injection molding process of particle dispersed thermoplastics. The particle simulation method (PSM) is combined with a flow analysis of the finite volume method (FVM). In the PSM, a particle is modeled with an array of spheres. Each pair of adjacent spheres is connected and deformed by three types of springs ; stretch, bend, and twist. The motion of particles under flow field obtained from the flow analysis is followed by solving the translational and rotational equations of motion for each constituent sphere. The mobility matrix for each particle is calculated to obtain the hydrodynamic force and torque exerted on each sphere. For the hydrodynamic interaction among particles, the near-field lubrication force is considered when the separation between spheres belonging different particles is close, but the far-field part is neglected. The method was applied to motion of fillers (fiber and talc) during an injection molding process of thermoplastics. The calculated motions of fillers were in good agreement with the experimental results.

抄録全体を表示

In particle or short-fiber reinforced composites, cracking or debonding of the reinforcements is a significant damage mode because the damaged reinforcements lose load carrying capacity. This paper deals with a theory of the reinforcement damage in discontinuously-reinforced composites and its application. The composite with progressive cracking damage contains intact and cracked reinforcements in a matrix. To describe the load carrying capacity of the cracked reinforcement, the average stress of a broken ellipsoidal inhomogeneity in an infinite body which was proposed in the previous paper is introduced. An incremental constitutive relation of the composites with progressive cracking damage of the reinforcements has been developed based on the Eshelby's equivalent inclusion method and Mori and Tanaka's mean field concept. This damage theory can describe not only cracking damage but also debonding damage of the reinforcements by modifying the load carrying capacity of damaged reinforcements. Influence of the reinforcement damage on the stress-strain response and elastic stiffness of the composites is discussed. It is noted that the full-debonding damage gives the lower limit of the stress-strain relation of the composite with progressive reinforcement damage.

抄録全体を表示

An injection-molded short-fiber compostie often contains some fiber clusters due to a poor mixing process. An analytical model for a composite with such clusters is developed to estimate a composite stiffness and stresses inside and just outside a fiber of a composite. Firstly the stiffness of the cluster is predicted by applying the Eshelby's equivalent inclusion method to the cluster. Secondly the stiffness of the overall composite and the stress inside a fiber are assessed by applying the Eshelby's method to two kinds of inhomogeneities, i.e. cluster and fiber. Finally the stress just outside the fiber, namely the fiber-end stress, is evaluated using Hill-Walpole-Mura's jump condition. It is concluded from this parametric study that the effective stiffness of composite tends to decrease as the volume fraction of clusters increases, and cracks at fiber ends are more likely to occur at a lower applied stress level as the volume fraction of clusters increases since the magnitude of the fiber-end stress increases. It is found that a carbon-fiber, namely higher stiffness reinforcement, composite is more sensitive to the clusters than a glass-fiber one.

抄録全体を表示

In order to describe the relation of mesoscopic phenomena such as interfacial debonding, and breakage of fiber and matrix to macroscopic stress-strain curve of unidirectional composites, an approximate nondimensional solution method was presented using two-dimensional model. With application of this method to several examples in which the species and locations of broken components were varied to each other, the features that the fiber-breakage-induced debonding occurs at lower strain than the matrix-breakage-induced one, the overall debonding is accelerated due to the mechanical interactions under the existence of many broken fiber and matrix, the progress of overall debonding is dependent on species of the broken components and on the geometrical location of broken components and debonded interface, and stress-strain curve shows drops in stress due to progress of debonding were revealed. Also as an extended application of the present method, a nondimensional Monte Carlo Simulation method was presented to describe the behavior of the composite in which the number and location of broken components and debonded interface vary with increasing strain.

抄録全体を表示

Stress analysis was conducted for a transversely reinforced composite laminate with an edge crack in the matrix or along an interface by the boundary element method (BEM). The laminate is submitted to the tensile stress perpendicular to the laminate in plane strain condition. The results obtained are summarized as follows ; (1) When the laminate is composed of such two materials as don't show the free edge effect, the stress intensity factor (SIF) for a small edge crack coincides with that for a small inner crack. (2) In the case that the free edge effect is remarkable, when a crack is very small (or thort), the change in normalized SIFs, K₁ (K₁=K₁/K_lh, K_lh is SIF for a crack in a homogeneous orthotropic body) for a matrix crack and K₁, K₂ and K₁ for an interface crack depend on the magnitude of stress singularity by free edge effect. (3) As a crack becomes larger (or longer), K₁ for a matrix crack and K_i for an interface crack tends to converge on the constant values. The convergent values are calculated by the energy release rate for a crack in the homogeneous orthotropic body which has the same elastic constants as those of the composite macroscopically. (4) The magnitude of SIF for a small edge crack can be calculated by the stress distribution near the edge without a crack.

抄録全体を表示

Computer simulation of creep cavity behaviour in polycrystalline metal was attempted. In this model, Voronoi's polyhedrons were looked upon as a polycrystalline metal and thier facets were upon as grain boundaries. The values of grain-boundary sliding and normal stress were assigned to the each grain by their respective distribution functions which depended on the direction of the facet. The damaged grain boundary was defined as the facet whose sliding value and stress value were greater than critical values respectively. The nuclei of cavity were generated on the damaged grain boundaries and the number was decided by an exponential function of the normal stress of each facet. Cavity size was calculated by a differential equation of time under the influence of the normal stress. The result of the simulation with proper distribution function and material constants shows good agreement with the result of experiment.

抄録全体を表示

A method of evaluating the macroscopic shape of defects in microscopically inhomogeneous material such as inclusion-reinforced composites is proposed by considering the order of scales that describes the mechanical behaviors (or physical phenomena). By using the method of multi-scale asymptotic expansions for field variables, the averaging technique with representative volume elements for inhomogeneous media becomes a reasonable tool to deal with the medium as a homogeneous one in macroscale. Therefore the linearized inverse scattering method along with Kirchhoff approximations can be applied to identify the shape of defects in macroscopically homogeneous media since the homogenization procedure replaces the media by homogeneous ones. Numerical examples show the capability of the proposed method to determine the shape of cracks whose sizes are much larger than the microscopic heterogeneities, which are invisible for ultrasonic waves.

抄録全体を表示

Elastic analyses near a tip of interface crack between dissimilar solids have shown that non-self-similar singularity develops and an overlapping of the crack surfaces may occur near the crack tip. In the present paper microscopic deformation and fracture near the tip of an interface crack were simulated by applying the molecular dynamics. Size effect was found in the near-tip deformation of geometrically similar cracks subjected to remote tension. The effect can be explained in terms of local stress intensity factor defined using an atomic spacing as a representative length. The direction of fracture depended on the local stress intensity factor and the interfacial strength. Locally pure opening mode and pure shearing mode were defined based on the local stress intensity factor. Deformation and fracture were simulated for these modes. The critical energy release rate under the pure opening mode was comparable with that calculated using the atomic potential. In the so-called overlapping region slipping developed along crack surface and the fracture occurred almost perpendicular to the crack surface.

抄録全体を表示

In order to analyze the atom-order functionally gradient materials (FGM), the molecular dynamics (MD) simulations are proceeded. Two kinds of FGM models, which are constructed with nickel and copper atoms, are made by putting the atoms at lattice points with respect to the mixture ratio. Firstly, the uniform lattice constant, that is the average value of the lattice constant of Ni and Cu, is used. The elastic Constants are very different from those calculated by the mixture formulations for macro model, because the lattice constant of Ni and Cu is different each other. Secondly, the changing lattice constant, that is the lattice constant changes with respect to the mixture ratio, is used. By this model, the distribution of elastic constants along the gradient direction is in good agreement with that derived from the mixture formulations. Not only the elastic constants but also the stress distributions are obtained, and the difference between these two models are discussed.

抄録全体を表示

In the present study, the 2-dimensional discrete dislocation dynamics (DDD) simulation is applied to crack tip field under cyclic loading. The crack is located on bcc (body centered cubic) iron subjected to the mode I crack opening load. The crack front is taken in [11^^-0] direction and the plane strain assumption confines the number of active slip system to two ((1^^-1^^-2) [111], (112) [111^^-]) of the 48 proper bcc slip systems. The angles θ between the slip direction and the direction of the crack extension are set 54.7°and 35.3°. The plastic deformations, obtained by different computational methods (molecular dynamics, DDD, and FEM based on crystal plasticity) in different scales are compared mutually. The shapes of plastic region, obtained by the DDD and the FEM based on crystal plasticity qualitatively correspond with each other. Under cyclic loading, the number of nucleated dislocations is almost balanced to the number of annihilated/disappearing dislocation so that the number of current dislocations is almost constant through the simulation. The cleavage stress at crack tip is increasing during this steady state. This seems to give some useful informations of ductile-brittle transition in fatigue crack extension by the DDD approach.

抄録全体を表示

Fatigue crack growth increment per cycle in the near-threshold regime is nearly equal to the atomic spacing for metallic materials. Microscopic analyses using the molecular dynamics are, therefore, advantageous for investigating the mechanisms of fatigue crack propagation in the near-threshold regime over macroscopic analyses, such as the finite element method and the boundary element method. In the present paper molecular dynamic simulations were conducted for the near-threshold fatigue crack propagation in α-Fe at 600 K. The Johnson potential was used to calculate the interatomic forces. It was found that at certain level of the stress intensity factor the crack propagated gradually, while it ceased to propagate at lower stress intensity levels. This indicates that the near-threshold fatigue crack propagation can be simulated by using the molecular dynamic method. To take the effect of new fracture surface into account, modification of interatomic potential was introduced. The effect of crack advance was also considered in the simulations. The effectiveness of these modifications was demonstrated.

抄録全体を表示

Propagation of the leaky Rayleigh and creeping waves, which are excited by a line-focused transducer, have been analyzed numerically under plane strain by finite element method. The analyzed model is composed of a line-focused transducer, coupling water and an aluminum plate. The wave equation of pressure in water and those of displacements for a solid are discretized with linear triangular finite elements. A new data structure is constructed to improve efficiency of calculation by using reference index tables which minimize memory size. Nonreflecting boundaries are applied at the outer boundaries of the plate to remove reflection at the external boundaries of the solid. The calculated velocities of the leaky Rayleigh and creeping waves at the water-aluminum interface are close to the measured velocities. The attenuation constant of the leaky Rayleigh wave is in good agreement with the theoretical value at the central frequency of the transducer used.

抄録全体を表示

First, this paper presents the concept of mixed-phase simulation with fracture-path prediction mode which uses a postulated propagation-direction criterion together with experimentally obtained crack-propagation history. Furthermore, this paper presents the development of an automatic moving finite element method, that incorporates the Delaunay automatic triangulation. Using the automatic moving finite element method, the mixed-phase simulation with fracture-path prediction mode is carried out for mixed-mode impact fracture tests. Various dynamic fracture mechanics parameters are evaluated by the path-independent dynamic J integral. The moving finite element method in conjunction with the local symmetry criterion (the Kn=0 criterion) successfully predicts the numerical fracture path in excellent agreement with experimentally obtained actual fracture path.

抄録全体を表示

Dynamic creep tests were performed for the specimen with projections of sintered silicon nitride ceramics at test temperature, T=1300°C, where triangular stress wave was imposed at stress rates σ=0.1, 0.5 and 20 MPa/s. Then dynamic creep displacement was measured between the two projections by means of the laser beam type displacement measuring system. Cyclic strain range decreased monotonously with increase in the number of stress cycles in dynamic creep test and this attenuation was much more enhanced at the higher cyclic stress, suggesting that cyclic deformation resistance increases during dynamic creep at 1300°C. The equivalent stresses σ_ep.1 and σ_ep.2 were assessed to examine the relationships between the dynamic and static creep strain rates, and the dynamic and static creep life times to failure. Then minimum creep strain rate ε_cmln and creep life time to failure t_f were plotted against the equivalent stresses in the dynamic creep. It was shown that the relationships between ε_cmln and σ_ep.1, and t_f, and σ_ep.2 were in quite good agreement with those in the static creep.

抄録全体を表示

Genetic algorithms are applied to obtain the optimal stacking sequences for maximizations of buckling load of composite cylinders. In this problem, three rules for empirical stacking are implemented as constraints. The first is that the laminates have to be balanced. The second is that more than 4 contiguous plies of the same orientation is not allowed. The third is that difference of fiber orientations between adjacent plies must not exceed 45 degree. The difficulty of handling constraints in genetic optimizations are handled by adoption of a new repair system. The new repair system operates when the chromosome violating the constraints is decoded to a stacking sequence. This repair system does not alter genes of the chromosome but changes decoding rules to introduce the constraints. This is similar to recessive genes in biology. By using this repair system, a chromosome of {2 2 2 1 1 1 0 0 0} is decoded to [90/90/90/45/90/-45/-45/0/45] s. Design reliability of the genetic algorithm with the new repair system are examined in detail by comparing the reliability with a penalty method. As a result, the genetic algorithm with the repair system is shown to provide higher design reliability.

抄録全体を表示

The damage process of fiber reinforced glass composites were studied under tensile loading. The tensile damage initiation and growth in BN-coated HI-NICALON SiC fiber reinforced glass matrix cross-ply laminates were experimentally clarified by replica observations, An improved shear-lag model with two major cracks, which penetrate both 0 degree and 90 degree plies, were used to predict the stress-strain behavior and reasonably well predicted the stress-strain curve until matrix cracks in the 0 degree ply saturate. Ultimate tensile strength (UTS) could be predicted by using a probabilistic fracture model based on the strength distribution data of fibers extracted from fabricated composites. Difference between the experimental strength and the predicted one was attributed to a difference in the strength distribution data of fibers and the one of fibers in the real composites. Then, UTS was compared with the prediction using the modified strength distribution data of fibers.

抄録全体を表示

Using a plate made of A2017-T6 metal matrix composites reinforced with 10 vol. % and 20 vol. % Al₂O₃ particles, crack propagation rate da/dN of a mode I suface crack by the simultaneous action of plane bending and cyclic torsion is sought. And the effects of crack tip opening stress σ_top, crack opening displacement COD and biaxial stress ratio C on da/dN are examined. When ⊿K is lower than the peculiar level, da/dN decreases with the increase of volume fraction of Al₂O₃ in C=0 and C=-0.55. But da/dN of Al alloy becomes minimum in C=-1 and the effect of Al₂O₃ particles disappears. σ_top increases with Al₂O₃ particles and the decrease of C. On the other hand, COD not always increases with the decrease of C. These phenomena can be explained by the residual compressive stress formed at the surface layer of the specimen by the fatigue test and the surface roughness of crack section.

抄録全体を表示

The strength of fiber reinforced plastic (FRP) composite, E glass fiber/epoxy, was evaluated by the ring burst test. In order to control the fiber waviness in FW-FRP composite, the specimens were fabricated by a filament winding (FW) method, where the tension in fiber was controlled as the winding load was increased or decreased gradually. Experimental burst pressure was lower than the ideal value based on the law of mixture under homogeneous strain state. The cancellations of fiber waviness were observed by a high magnification video system. The cancellation behavior could be characterized using acoustic emission (AE) analysis and high magnification video system. In particular, the increasing winding tension specimens showed more remarkable cancellation behavior than the other specimens. Cancellations of fiber waviness seriously affected the reduction of burst pressure of FW-FRP multiply composite. In this study, the burst pressure of FW-FRP composite was enhanced by the control of winding tension as the load in fibers was decreased gradually. Then the experimental model of fracture behavior, in which fiber waviness could be considered, was proposed. The predicted values using the experimental model showed good agreement with experimental results.

抄録全体を表示

Elastic modulus of composite materials during fatigue tests was evaluated from short term creep tests and a modified visco-elastic model. In the elastic region of tensile tests, the fiber surface treatment did not affect stress-strain relations. The surface treatments, however, changed fatigue life, i. e., depending of surface treatment, difference with tens of times of fatigue life at the same stress amplitude was observed. In fatigue tests, the change of elastic modulus, evaluated from the modified visco-elastic model, had good correlation to residual strain. Except the last stage of the fatigue process, cracks were not observed in resin. Therefore, fatigue damage is considered to consist of initiation and propagation of interfacial debonding at the end of fibers, and increasing of residual strain which came from fiber pulling-out. From interfacial debonding length, which was evaluated from the reduction of elastic modulus, the effective fiber length was considered to be shortened by more than 50%.

抄録全体を表示

In order to investigate the effect of cyclic stress on fatigue behavior of mica glass ceramics and alumina, the specimens with an indentation induced flaw were tested under static tensile loading and cycling loading at stress ratio R=0 and -1. The frequencies in the cyclic loading tests were 5 and 20 Hz respectively. The main results were as follows ; (1) A significant decrease in fatigue lifetime by cyclic loading was observed on both materials, though there being large scatters on mica glass ceramics. (2) An appreciable decrease in fatigue lifetime due to the increase of frequency was observed on mica glass ceramics, while a little effect of frequency was observed on alumina. (3) The effect of stress ration under cyclic loading was small on both materials, though a somewhat significant decrease in fatigue lifetime being observed in the case of frequency of 20 Hz on alumina. (4) SEM fractography revealed a dominant interglanular fracture on both materials, but in the case of cyclic fatigue fracture of alumina more comminution was observed than in the case of static fatigue fracture. (5) It was considered that the decrease in fatigue lifetime due to cyclic loading in mica glass ceramics results on the degradation of grain briding effect by the pull-out of mica from glass matrix, while the effect of cyclic stress in alumina results on the degradation of grain bridging effect by comminution and fretting in the grain.

抄録全体を表示

Fatigue tests have been performed to investigate the effect of strain aging on the fatigue limit and fatigue crack behavior about two kinds of austenitic stainless steels, SUS 304 and nitrogen contained SUS 304 (SUS 304 N). From the testing results, the fatigue limits of SUS 304 and SUS 304 N increase by 5 and 30 MPa, respectively by the low temperature annealing (300°C×2h) during fatigue test by the specified cyclic ratio under constant higher stress amplitude than their original fatigue limits. It is considered that the effect of strain aging by nitrogen in SUS 304 N increases its fatigue limit. However, the specimens, in which the fatigue micro-cracks in the length of 1∼2 grains sizes exist under the higher stress amplitude than the fatigue limit of each material, propagate to be a final fracture before 1×10⁷ cycles under the same original stress amplitude of fatigue limit, even if the specimens were annealed in the vacuum at low temperature.

抄録全体を表示

The fatigue crack growth tests of the SUS 631 stainless steel foil (thickness : 0.05 mm) were carried out in a wide range of fatigue crack growth rate covering the threshold stress intensity factor range ⊿K_th. The effects of aging hardening treatment and the stress ratio on the fatigue crack growth characteristics were investigated on the basis of fracture mechanics. The crack growth rate was measured using the center notched specimens of solution heat treatment (1313K, 2s) foil (STF) and two kinds of aging treatment (703 K, 1 h and 753 K, 1 h) foils. Crack growth tests were conducted at 10 Hz and the stress ratio R=0.1 and R=0.5. The main results obtained are as follows ; (1) The Paris'law could also be applied between da/dN and ⊿K to the SUS 631 stainless steel foil in a special region of da/dN and it was shown to be da/dN=C (⊿K)^m. (2) In a low-⊿K region the crack growth rate da/dN of aging treatment foil was higher than that of the STF and was accompanied by a small threshold value ⊿K_th. On the other hand, in a high-⊿K region da/dN of ageing treatment foils was lower than that of STF. (3) Fatigue crack growth resistance of a higher temperature aging material was higher than that of a lower temperature aging material with a higher tensile strength. (4) Crack closure level (K_op/K_max) was rapidly increased as the maximum stress intensity factor K_max decrease for each material at R=0.1 and R=0.5.

抄録全体を表示

This paper deals with a general formula for singular boundary elements in the computation of two-and three-dimensional stress intensity factors using the boundary element method. The formula is based on "double-mapping", and the traction singular boundary elements can present any stress singularity and have any number of nodes. This singularity formula is applied to several crack problems. The results show its effectiveness in determining stress intensity factors.

抄録全体を表示

The friction welding for copper and aluminium is useful to minimize the diffusion thickness at the interface between copper and aluminium. The minimization of diffusion thickness is required for maintaining the joint strength, because brittle intermetallic compounds are formed at the friction weld interface between copper and aluminium. However, the improvement techniques on joint strength characteristics have not fully been clarified. In this study, stress singularities at the intersection of free surfaces and the interface between copper and aluminium are investigated using finite element analysis. It was confirmed that the no free-edge stress singularities existed in case of convex angle. In this experiment, copper and aluminium friction welded joints with convex apex angle were fabricated as to minimize the diffusion thickness, such ad > 1 μm at the interface between copper and aluminium. It was confirmed that the good tensile strength could be observed in any cases. The Charpy impact energy became comparatively low values in case of 90 degree of apex angle, and revealed a remarkable good impact energy can be obtained in case of the convex angle that showed no free-edge stress singularities.

抄録全体を表示

A method of structural synthesis is proposed to set disposed shape of a flexible frame within an allowable region. Members of the flexible frame, which are initially straight, are discretized into straight bar segments linked by virtual coil springs at nodes. Assuming total strain energy is independently separated into two contributions of bending deformation by the springs and rectilinear deformation by the segments, we find positions of the nodes in the disposed state by means of the Lagrange multiplier method that searches for a stationary point of the total strain energy under the constraints on geometrical boundary conditions. A new formulation of the sensitivity analysis is derived to approximate the change of the found nodal positions with respect to design variables in the first-order sense. A layout constraint to exclude all the nodes from convex prohibited regions in disposed state yields governing equation of design variables in the form of linear simultaneous equations with rectangular matrix of coefficients, which is dealt with by the Moore-Penrose generalized inverse. A numerical example concerning with Y-shaped frame whose disposed shape is constrained by three prohibited regions demonstrates the validity of the proposed formulation.

抄録全体を表示

With the evolution of magnetic tape to the Giga-byte level, smoother surface is required than before. The surface of the magnetic tape is smoothed in calendering process. It is necessary to estimate the surface roughness in the operation conditions of calender. This paper presents the analyzing method of the surface smoothing considering with the real contact area and the viscoelasticity of the magnetic layer at the temperature of calender roll. The calculated surface roughness is almost equal to experimental. However, in the case of soft calendering, calculated surface roughness is smaller than experimental. This is thought to be due to the visco-elastic deformation of the magnetic layer at the real temperature different from that of calender roll.

抄録全体を表示