A method for evaluating the distribution of electrical potential around multiple spherical defects was proposed. As the method is based on the known solution for a single defect, the electric field can be efficienly analyzed in comparison with the other methods, such as the finite element method. The electric field in a conductive material with multiple spherical defects at random locations is estimated by the method. The analysis shows that the increase in the potential difference normalized by the potential difference without defects, ΔV/V₀, is in proportion to the product of the volumetric density of defects and the mean of cubed defect radius, n_v [r³]_m. The universal relationship is independent of the location of defects and the distribution of defect radius. Thus, the damage due to the multiple defects can be evaluated by the increase in potential difference.

View full abstract

The radiated sound from the impacted body must have the information with respect to the impact force. We have proposed the method in order to identify the impact force by analyzing the radiated sound from the impacted body. This sound is measured by the microphone and the impact force can be given by the inverse analysis using FFT. In this paper, availability of the method is studied by the many experiments. It is confirmed that the impact force can be given when the microphone is set at far from the impacted body. Moreover, the efficiency of this method for the inelastic body is concerned.

View full abstract

In the present study, identification of boundary conditions for elastostatic thin plate bending problems is discussed. In the first step of the research, it is necessary to determine unknown quantities before the identification of the conditions. This paper is concerned with an inverse analysis for the unknown boundary values by the boundary element method (BEM). The unknown boundary values are estimated with additional data for the deflections or the strains given on some internal points of the plate. Some typical numerical calculations show the validity of the present approach, and the usefulness of rank decreasing technique based on singular value decomposition employed with Hansen's L curve method in order to stabilize the numerical solutions on inverse analysis.

View full abstract

This paper describes a method for identification of boundary conditions for elastostatic thin plate bending problems. In the present procedure, boundary conditions of the plate are represented by boundary integral relation including influence functions. The influence functions are defined as the deflection and rotation when shear force and bending moment are applied continuously on the supporting body. The influence functions approximated with polynomials can be determined numerically from these boundary quantities given by the method in the first report. Through some numerical demonstrations, it is shown that the boundary conditions of the plate can be identified well by the present inverse analysis based on the boundary element method.

View full abstract

The thermal elasto-plastic behavior of a pylon member in hot-dip galvanization is analyzed by the finite element method in order to predict the occurence of cracking due to zine embrittlement, in which the surrounding zone of a bolt hole is modeled by using axisymmetric solid elements with Bodner-Partom's viscoplastic constitutive equation. Numerical stuides are conducted for the influence of the residual stress due to welding on the cracking near the hole. The analysis is extended to the damage analysis using Bodner-Chan's damage evolution equation. The uncoupled, locally coupled and fully coupled analyses are conducted for the coupling of the damage constitutive equation and the structural analysis. Numerical results show that the locally-coupled analysis is the most efficient and accurate for the present problem.

View full abstract

A series of cyclic deformation tests was conducted on modified 9 Cr-1 Mo steel at various temperatures, including nonisothermal conditions to investigate inlastic behavior of the material. The material showed cyclic softening behavior under temperatures between 200°C and 600°C, and its degree got larger as the temperature was increased. Temperature history effect was scarcely found except the case where the test temperature was changed from high to low after large cyclic softening at the high temperature. Peak stress under thermomechanical conditions showed approximately the same value as those observed under corresponding isothermal condition. Furthermore, internal state variables used in the authors' unified constitutive model were experimentally measured. An aging stress, which corresponds to resistance by dynamic strain aging effect, and an effective stress were almost constant under cyclic deformation conditions, while a back stress showed eminent cyclic softening behavior.

View full abstract

Based on inelastic behavior of modified 9 Cr-1 Mo steel described in the first report, the unified constitutive model proposed by the authors was extended for cyclic deformations, including nonisothermal conditions. In the constitutive model, cyclic softening behavior was expressed by decreasing an asymptotic value of back stress with accumulation of inelastic strain. Most of the constants in the constitutive model were determined as a function of the temperature in order to calculate stress strain relationships under nonisothermal conditions. To evaluate the validity of the constitutive model, it was applied to analysis of inelastic behavior under mononotic tension, stress relaxation, creep, isothermal cyclic deformations including stress relaxation and nonisothermal cyclic deformations. It was found that the present constitutive model was successful in describing the inelastic behavior of the steel adequately.

View full abstract

Constitutive models for cyclic plasticity have been developed for about two decades. However, there are few constitutive models which can explain uniaxial ratchetting behavior accurately because of lack of knowledge of the correlation between uniaxial ratchetting and viscosity of materials. In this paper, to clarify the correlation, a series of experiments and numerical simulations are carried out. First, creep tests after the primary cyclic loading are conducted using Type 304 stainless steel at room temperature. Both uniaxial ratchetting and cyclic tension-compression loading controlled with strain amplitude are chosen for the primary cyclic loading. After the primary loading, subsequent creep tests are conducted at the maximum stress of the primary loading. Different creep curves are observed due to loading history of the primary loading, inspite of the same stress level of the subsequent creep tests. Moreover, the corresponding strain to creep during ratchetting deformation can be found out. Finally, the effect of the primary loading on the subsequent creep is also studied with the simulation using the dislocation based constitutive model.

View full abstract

Hill reformulates classical specification of plastic response in metal by introducing generalized variables, where the measures of stress are sets of the loading magnitudes, and the measures of strain rate are generated by work-conjugacy. The work hardening h_G in generalized variable is constructed in association with the original one. Eigenstates are defined as the state where h_G vanishes and they involve the state of simultaneous stationarity of generalized stress. The eigenstates are classified into latent eigenstate and active one. In the former the generalized stress rate vector is perpendicular to the normal of yield surface in generalized stress space, and the latter is named for the simultaneous stationarity of generalized stress. However, the eigenstates are given a priori without any explanation relevant to multiplicity. Therefore, doubt remains whether they satisfy solution multiplicity or not. Then, in this paper the doubt is examined from the view point of solution multiplicity. It is found that the latent eigenstate cannot satisfy the multiplicity criterion except for the case where the nominal stress is used as the generalized stress. Furthermore, in general, the active eigenstate cannot be realized. We propose here load instability instead of the active eigenstate.

View full abstract

Chemical structure and physical properties of Polyetherketone (b-PEK) synthesized from dichlorobenzophenone and sodium carbonate in the presence of silica/Cu salt catalysts have been investigated. After basic hydrolysis, the chemical structure was analyzed by high performance liquid chromatography (HPLC) and mass spectrometric analysis (MS). b-PEK had a few branched structures and was a different from Polyetherketone (1-PEK) which is commercially available or can be synthesized from 4-hydroxy-4'-fluorobenzophenone. b-PEK had good mechanical strength, modulus and fluidity under the high share rate compared with 1-PEK. It was thought to be due to the degree of crystalline or other micro-structure and it observed that the injection molded test specimen of b-PEK had higher crystalline and orientation than 1-PEK by measuring wide angle X-ray diffraction.

View full abstract

Mechanical alloying process (hereafter abbreviated as MA) is applied to polycaprolactone (PCL)/Cu powder mixture, in order to develop the copper base material with the characteristics of degradation in ground. The mean value as well as the dispersion of hardness decreases with increasing MA time. The value of hardness scarcely coincide with the theoretical prediction based on the law of mixture as MA time increase. The electric resistivity increases with increasing MA time up to the level depending on the copper content. The value at saturation decreases as the copper content increases. SEM observation shows that copper particles remain as the aggregates at shorter MA time and the particles wrapped by PCL disperse homogeneously at longer MA time. This indicates that the electric conduction and the force transmission originate from the aggregates at shorter MA time, and that no electric conduction and force transmission originate from the dispersed particles at longer MA time.

View full abstract

Recently, FRP composite structures have been applied to many structures of vehicles. Damage, such as delamination, by an accidental impact causes stiffness degradation of the structures. Since the damage is interior and invisible from the outside, it is important to inspect the damage by nondestructive inspection. In the present study, we propose an identification method of delamination in CFRP laminated beam using the natural frequency change and response surface methodology (RSM). Response surfaces are used for approximation of the natural frequency change in delaminated CFRP beam. In order to identify the delamination, we make a performance function using response surfaces. By minimizing the performance function, we can predict the delamination position and length of the delaminated CFRP beam easily and inexpensively. As a result, it is shown that there are good agreement between actual and predicted delamination positions and lengths.

View full abstract

Carbon/Carbon composites (C/C) have recently been studied and developed for structural applications in the field of aerospace and power generator technologies. However, C/C has low interlaminar strength. Therefore, it is important to evaluate interlaminar fracture properties and test methods for interlaminar strength. In the present paper, each of the ILSS (interlaminar shear strength), the ILTS (interlaminar tensile strength) and the interlaminar mixed mode strength with the coexistence of interlaminar shear and tensile or compressive stresses was studied by using several different test methods on C/C. The ILSS was obtained by using the SBS (short beam shear) test, the DNS (double notch shear) test, the Arcan test and the quasi-Arcan shear mode test. The ILTS was obtained by using the Arcan test and the quasi-Arcan tensile mode test. The interlaminar mixed-mode strength was obtained by the Arcan test. Comparisons of respective tests and several laminate configurations were conducted using microscopic failure mechanism and 3-dimensional FEM analysis. Both Von-Mises, Tsai-Hill and Tsai-Wu failure criterions were used to explain the mixed-mode fracture behavior. Effects of notch stress concentration were found negligible based on FEM analysis and experimental results.

View full abstract

Off-axis fatigue behavior of unidirectional continuous carbon fiber-reinforced composites has been studied for T800H/epoxy. AS4/PEEK, and T800H/Polyimide systems, with a particular emphasis on the effect of matrix and matrix/interface properties at elevated temperature. Tensiontension fatigue tests were performed at 100°C for various kinds of plain coupon specimens with different fiber orientation angles. It was observed for these composite systems that the fatigue strengths decreased with increasing off-axis angles. Regarding the epoxy- and PEEK-based laminae, almost linear off-axis S-N relationships were obtained over the whole range of fatigue loading up to 10⁶ cycles. The off-axis fatigue strength of the polyimide-based laminae significantly dropped near the fatigue cycle of 10⁴ which was ascribed to the initial defect of non-uniform fiber distribution. Normalizing the fatigue stress with the static strength at test temperature, the off-axis fatigue data eventually fell on a single S-N relationship for these composites. This characteristic was described well by a non-dimensional effective stress defined on the basis of the classical static failure theories. It was also demonstrated that the off-axis fatigue behavior of these unidirectional fiber composites was described by the classical composite fatigue failure models.

View full abstract

The effect of dilute acid solution on fatigue damage process of GFRP under tension-tension cyclic loading was investigated. Dilute sulfuric acid of pH=3.9 was used. This pH value is considered to be a level commonly found in acid rain. The conclusions are as follows. (1) When the specimen has been exposed to pH=3.9 acid for seventh months, the effect of pH=3.9 acid on fatigue strength is almost the same as that of city water. (2) The fatigue degradation of GFRP in pH 3.9 acid and city water is greater than that in air. (3) As the concentration of acid solution increases, the fatigue life becomes short due to fiber degradation. (4) As the immersion time increases, the fatigue life would decrease even in dilute acid solution whose concentration is as low as acid rain while city water gives no significant effect on the strength of glass fiber.

View full abstract

Fatigue properties of the nitrided SS 400 are investigated on the basis of the observation of dislocation structures, surface topography and fracture surface by means of optical and electron microscopy. Both slip bands and dislocation structures during stress cycling are not formed in the compound layer. The dislocation morphology of the diffusion layer is composed of the untangled and uniformly distributed dislocations and the clusters of the dislocation dipoles at lower stress amplitude, but well defined cell structures at higher stress amplitude. It is also found that the fatigue cracks of the nitrided specimen are initiated without the formation of slip bands and the separation between the compound layer and the diffusion layer occurs during cyclic loading. On the basis of these observations, a mechanism of fatigue fracture for the nitrided SS 400 is proposed.

View full abstract

The fatigue properties of austenitic stainless steels with different nitrogen content have been investigated in the high-cycles region using Ono-type rotating bending fatigue testing machine. The materials used in this test were three kinds of austenitic stainless steels, i.e. typical austenitic stainless steel, SUS 304 and nitrogen-contained SUS 304 (SUS 304 N and YUS 170). When considering the fatigue limit is the maximum stress amplitude by 1×10⁷ cycles, the fatigue limit of SUS 304 N becomes remarkably larger than that of SUS 304, while the fatigue limit of YUS 170 is smaller than that of SUS 304 irrespective of higher nitrogen content in YUS 170. The above difference would be caused by the difference of the effect of strain aging and deformation induced martensite transformation, because YUS 170 has more stable austenitic structure than SUS 304 according to higher nitrogen content. In addition, the non-propagating micro-cracks have not been observed in the not-broken specimens of SUS 304 and SUS 304 N subjected to the stress amplitude of fatigue limit by 1×10⁷ cycles. While a few micro-cracks were observed in the specimen of YUS 170 under same condition as the above and these specimens were broken under the following same stress amplitude by 1.2×10⁷ cycles.

View full abstract

Low cycle fatigue tests were carried out under push-pull loading conditions using Ferritic Ductile Cast Iron (FDI). In order to clarify the fatigue fracture mechanism of FDI in an Extremely Low Cycle Fatigue (ELCF) regime, an observation of the longitudinal sections of the partially fatigued specimens were performed. The fatigue process in the ELCF regime can be classified into three stages : (a) Generation of micro voids, caused by the debonding of the graphite-matrix interface. (b) Growth of the micro voids, perpendicular to the loading axis. (c) Formation of an internal macro crack, due to the coalescence of micro voids inside the material. Two types of microvoid coalescence processes have been observed. The first, due to the growth of internal micro-cracks and the second, due to internal necking of the matrix. The fatigue process of the FDI in the ELCF regime was accompained by the static fracture mechanism.

View full abstract

Fretting fatigue tests of carbon steel coated with titanium nitride (TiN) by PVD method were carried out to discuss the effect of ceramic coating on fretting fatigue behavior. Fretting fatigue strength of the present carbon steel significantly increased by TiN coating onto the specimen. The fretting fatigue fracture process in TiN coated steel was as follows. Fretting cracks initiate at a certain stage of fatigue life from the flaws or worn region in TiN film initiated by fretting action, and propagate until final unstable fracture occurs. The improvement of fretting fatigue strength by TiN coating resulted mainly from the retardation of fretting fatigue crack initiation due to existence of hard TiN film on the contact surface of the specimen.

View full abstract

This paper describes a mechanism of surface initiated crack growth in rolling contact fatigue. Experiments with epoxy resin disk specimens were carried out using a newly developed test rig for rolling contact fatigue. Rolling contact occurred between an epoxy resin disk of 100 mm dia. and steel disk of 290 mm dia. under normal force of 400 N/mm and tangential force of 40 N/mm. After rolling cycles of 2.4×10⁴, surface-initiated cracks were observed. The cracks initially grew along the radial direction up to depth of 2.8 mm, and propagated to the circumferential direction, which coincides with the direction of tangential force. By optical microscopic examinations, striation-like marks were observed on the radial crack surfaces and the striation spacing decreased with increasing crack length. Stress intensity factors K_I and K_II of a crack under rolling contact fatigue were analyzed by finite element method. When the crack length is short, the calculated value of stress intensity factor range ΔK₁ is larger than that of threshold stress intensity factor range ΔK_th for epoxy resin. It is shown that the cracks can propagate under mode-I mechanism. The crack growth transited from radial to circumferential direction was predicted by maximum circumferential stress around the crack tip.

View full abstract

Arrest function in fatigue crack growth at 25 and 550°C in air was investigated for a alloy with zirconium in 0.17 Wt%, Fe-27Cr-35Ni-0.17Zr, and a alloy without zirconium, Fe-21Cr-32Ni. The main conclusions obtained are as follows. (1) The fatigue threshold under closure-free condition was about 40% higher for Fe-27Cr-35Ni-0.17Zr alloy than for Fe-21Cr-32Ni alloy at 550°C, while the threshold was independent of the alloys at 25°C. (2) The fatigue limit at 550°C was about 40% higher for Fe-27Cr-35Ni-0.17Zr alloy than for Fe-21Cr-32Ni alloy. (3) The improvement in high-temperature fatigue threshold were explained, considering that the presence of zirconium strengthened the oxide film at the crack tip. Accordingly, we concluded that a small addition of Zr could gives rise to an arrest function of fatigue crack growth at elevated temperatures.

View full abstract

Crack growth behavior from pre-indentations and wear behavior in Si₃N₄ ceramics were investigated under ball-on-plate contact. Contact fatigue tests were carried out with the use of a newly developed apparatus under three contact conditions : sliding contact, sliding-including-rolling contact and rolling contact. A friction coefficient was monitored during the tests. It was found that the failure was strongly affected by the friction coefficient. The contact fatigue failure was dominated by the wear, for a higher friction coefficient above a threshold value of 0.13, while the failure was dominated by a crack growth than the wear for a lower friction coefficient than the threshold value. The crack growth behavior was observed under rolling contact and sliding-including-rolling contact. In the center of the contact area, the crack was found to grow only in a perpendicular direction to the motion of the ball. In the border of the contact area, on the other hand, the cracks were found to grow both in the parallel direction and in the perpendicular direction. The crack growth was accelerated by an increased friction force under rolling-including-sliding contact.

View full abstract

It has been accepted not only that effective proof testing of ceramic materials demands rapid loading and unloading to prevent strength degradation but also that especially rapid unloading brings truncated strength of σ_p (the stress just before very rapid unloading). Such common sense in proof testing as above has been rechecked with analytical considerations. The analysis reveals that as the loading and unloading rate increases, weaker samples tend to survive easily. As a consequence, the truncated strength is not characterized analytically, and Weibull curve will not have asymptotic feature. The present analytical results come to the conclusion that the common sense in proof testing will not be accepted from the physical point of view.

View full abstract

CFRP specimens have been exposed under accelerated exposure apparatus. The specimens are made of epoxy resin and the unidirectional carbon fiber prepreg having the fiber orientation angles of 0.45 and 90 degree, respectively. In order to clarify the dependence of the exposure period on the weatherability strength, the complicated accelerated exposure test is continued to 20 cycles. The variations of flexural strengths, flexural moduli, matrix volume fractions and thicknesses to the exposure period are reported in this paper. Furthermore, the experimental results of flexural strength and modulus for CFRP are compared with their estimated values calculated from the results of accelerated exposure of epoxy resin.

View full abstract

Stiffness is an important matter of concern, especially in automobile frame structure. When structure is made from hollow section beams, local deformation at jointed area reduces the stiffness remarkably. Local stiffness of T-joint of rectangular hollow section member of unequal width were studied under out of plane bending conditions. Three different analytical solutions were derived for chord tube flange distortion. The solutions were checked by FEM, in various branch size, shape and thickness of chord tube web, and also checked by experiments of welded T-joint made of aluminum extrusions.

View full abstract

Ti-Ni shape memory alloy was fabricated by spark-plasma sintering method, and isothermal tensile and recovery stress tests were carried out on the alloy. The effect of fabrication conditions, such as sintering and heat-treatment conditions, on the mechanical properties of the alloy were investigated. The appropriate sintering and heat-treatment conditions were found out by measuring the density, observing the microstructure and analyzing the X-ray diffraction of the sintered compacts. The relative density of the sintered compacts was around 98%. The obtained alloy showed apparent shape memory effect, but its tensile strength and elongation were much less than those of cast material of the same alloy. The recovery stresses for various stress-strain histories by a temperature rise from 293 K to 353 K were around 230 MPa.

View full abstract

Green's functions reported in the previous paper were the solutions for single root in the characteristic equation of a semi-infinite transversely isotropic thermoelastic solid with a steady heat source. The functions were expressed in a Cartesian coordinate system. In this paper, Green's functions for 2-ple and 3-ple roots in the characteristic equation of a semi-infinite transversely isotropic thermoelastic solid with a steady heat source are presented. A different term from the functions for single root in the characteristic equation appears in these solutions. For thermal boundary conditions, both cases of adiabatic condition and the temperature prescribed on the boundary plane are treated, and for the stress and the displacement boundary conditions the following three cases are treated ; (1) stresses are free, (2) all components of displacement are fixed, (3) only displacement in the normal direction is fixed.

View full abstract

The axial impact crushing behavior of square tubes with and without stiffeners are studied by Finite Element method, and comparisons of mean axial crushing force between numerical solutions and theoretical predications as well as qualitative comparisons of folding patterns between numerical simulations and experimental observations are made and discussed. Based on the numerical analyses, dynamic crushing energy absorption maximization problems of square tubes with and without stiffeners are respectively solved using the crashworthiness maximization technique for tubular structures which combined the techniques of design of experiment, response surface approximation as well as usual mathematical programming. In addition, the capacities of crushing energy absorbing of cylindrical tubes and square tubes with and without stiffeners are also compared.

View full abstract

New lap joint type shear specimen has been proposed to measure creep deformation properties of solder using for printed circuit board (PCB). The specimen has the following characters. The solder joint is fabricated easily by reflow soldering which is the same as surface mounting process of PCB and is kept in nearly pure shear stress state at any loading level onto the specimen. Additionally, the specimen is tested easily by utilizing usual Instron type testing machine. The experimental results of creep tests for Pb/Sn eutectic solder were compared with data shown in the literature. Consequently, the validity of the specimen was confirmed through the fact that the creep deformation properties were obtained with sufficient accuracy. Furthermore, the specimen was applied to the creep tests of some kinds of Sn/Pb solder with high thermal fatigue strength. The creep deformation properties of the solder were evaluated in comparison with the properties of Sn/Pb eutectic solder. As a result, the solder shows relatively small amount of creep deformation, namely, lower minimum creep strain rate in steady state creep.

View full abstract

This paper describes the tensile property of six kinds of Sn-Pb solders. Static tension tests were carried out at strain rates ranging from 0.001%/sec to 2.0% / sec using 5Sn-95Pb, 10Sn-90Pb, 40Sn-60Pb, 60Sn-40Pb, 63Sn-37Pb, and 62Sn-36Pb-2Ag solders at 313-398 K. Strain rates faster than 2.0%/sec were required to obtain time-independent Young's modulus and yield stress of the solders. In tests at the strain rates slower than 2.0%/sec, creep deformation was included in the stress-strain relationship, and Young's modulus and yield stress which do not depend on time could not be obtained. A new equation which can predict tensile strength of Sn-Pb solders as functions of temperature, strain rate and chemical composition were proposed. New equations were also proposed for predicting Young's modulus and yield stress as functions of temperature and chemical composition. These equations predict Young's modules and yield stress within a factor of 1.25 and 1.15, respectively. Low cycle fatigue lives were discussed based on the universal slope method.

View full abstract

Further fracture mechanics analyses to understand comprehensively the effects of material properties and package geometries on two configurations of delaminations along the interfaces at a) die-bonding layer and b) the bottom surface of the die pad in LSI plastic packages subjected to temperature cyclic loading were performed. Specifically, the effects on these delaminations of such package constituent material properties as 1) coefficients of thermal expansion and 2) Young's moduli of the encapsulant resin and die-bonding material were examined. Package geometry factors such as thickness and width, including chip size were also considered. The analyses were conducted on the basis of three models of delaminated packages, in which large stress intensity factors at the tips of growing delaminations were induced in our previous work. The impact of these parameters on the delaminations were determined and several design guides of Cu alloy and Fe-42% Ni alloy leadframe packages were pointed out.

View full abstract

The path-independent dynamic J integral originally derived by Nishioka and Atluri is valid for homogeneous materials. Recently, an extended path-independent dynamic J integral has been derived by the present authors, for non-homogeneous materials. They have also derived the dynamic J integrals separated into both material sides for an interfacial crack tip. Is this paper, dynamic finite element analyses are carried out for interfacial cracks in bimaterial plates subject to impact loading. Different cases of material mismatch are considered. The path independence of the dynamic J integral components and the separated dynamic J integrals are verified for impact interfacial crack problems by the numerical results. Several points were elucidated from the numerical simulations. For instance, (i) if the dynamic J integral response curves are normalized properly, the response curves for the same mismatch ratio with different rigidities coincide each other. (ii) The major responses of the dynamic J integral components always manifest themselves when the impact stress wave in the compliant material side impinges to the crack tip. (iii) The separated dynamic J integral or equivalently separated energy release rate from the compliant material dominates the fracture process of an interfacial crack with a large mismatch ratio.

View full abstract

When alternating current is applied to paramagnetic materials, such as austenitic stainless steels or aluminum alloys, the existence of cracks on the material surface or subsurface causes changes in current flow around them and, as a result, electromagnetic wave generated with this change in current flow can be detected on or near surface of a material. In this case, there is a possibility for detection of these defects by detecting electromagnetic wave by antenna. Hence, the measurements can be done without contact or without removing paint. In this paper, first results on detection of electromagnetic wave around defects with shielded loop antenna is described with a principle of this novel NDI method.

View full abstract

Dynamic buckling experiments on thin cylindrical shells placed inside a rigid liquid container were carried out using a shaking table. These shells and container represent the thermal baffles and the reactor vessel of a fast breeder reactor, respectively. The fluid pressure caused by the horizontal excitation distributes non-uniformly around the cylinders and causes external pressure buckling deformation on them. The buckling pressure was measured with various types of test cylinders under various seismic waves, and it was confirmed to be higher than that predicted by static buckling analyses. It was also found that sub-harmonic vibration occurs under a certain sinusoidal wave excitation and the response displacement increases suddenly at a lower pressure level than the buckling pressure obtained by seismic wave excitation tests. Based on the test results, it was pointed out that, in the seismic design to prevent the buckling of the thermal baffles, the static buckling analyses can be used as long as the sub-harmonic vibration does not occur.

View full abstract

In order to make erosion mechanism of spheroidal graphite cast iron clear, the authors have carried out erosion tests and observed the continuous structural change of vertical section near surface. The erosion mechanism of spheroidal graphite cast iron progresses like this : (1) the spheroidal graphites on the surface layer deform gradually (2) lips grow in the impact direction (3) they expand and drop out. Though the erosion mechanism progresses by this repetitive process of lip-growth and dropping out, the life span is different depending on material. The wear growth rate, which is a standard life span of lip-growth and dropout, becomes smaller according as pearlite ratio increases and corresponds well to the erosion rate. The authors have also confirmed that our continuous observation method is quite practical and useful to elucidate erosion mechanism.

View full abstract