The structures of giant fullerenes have been considered from the topological view point. As the fullerenes are formed of the pentagonal rings and hexagonal rings, Euler's theorem in the topology shows that 12 pentagonal rings are neccessary though the number of hexagonal ones is arbitrary for the formation of the closed polyhedron. This restraint predicts that symmetric giant fullerenes have the following shapes: the icosahedral-shaped, the tetrahedral-shaped, the pentagonal prism-shaped and the hexagonal prism-shaped ones. We have classified the icosahedral-shaped and the tetrahedral-shaped ones into three systems, respectively, and the pentagonal prism-shaped and the hexagonal prism-shaped ones into two systems, respectively, thus totaling ten systems. We have formulated the equations for calculating the number of atoms n in the giant fullerenes Cn. The results indicate that the forming of the round-shaped giant fullerenes requires the sets of heptagonal and pentagonal rings while keeping (the number of pentagonal rings)-(the number of heptagonal rings)=12.
Molecular dynamics (MD) method is applied to a one-dimensional harmonic oscillation system. The problem, when Verlet algorithm and Runge-Kutta method are adopted as difference calculuses, is solvable using linear algebra. The following are noted for the equilibrium (or balanced) structure calculation by means of MD method; (1) Time step h should be less than the threshold value hc for the calculi. The solution diverges with iteration for h>hc. The magnitude of hc is roughly estimated to be √M/κ, where M is the mass of a constituent particle (atom) and κ is the spring constant given by the second-order differential of potential d2φ/dr2. (2) Kinetic energy dissipation technique is conventionally used in Verlet algorithm in order that the total energy of the system is made minimal and the solution converges to an equilibrium point. The damping efficiency heavily depends on h. It becomes very little for h<<hc and so h should be in the same order of hc. (3) The total energy of system decreases monotonously with iteration in Runge-Kutta method. The damping effect is remarkable for h≈hc, while the energy is almost conservative for h<<hc. Numerical experiments showed that the above results (1)-(3) are analogically valid for one- and two-dimensional Lennard-Jones lattice, suggesting the applicability to a non-linear potential system as well as to a many-dimensional system.
Metal matrix composite with gradient volume fraction of particles was processed by centrifugal casting technique using an aluminum alloy as a matrix metal and SiC particles as reinforcements. The distribution of particles in the matrix under various conditions of centrifugal casting was observed. The simulation of the solidification process during the course of centrifugal casting was performed in order to predict the distribution of particles in the transverse direction of axis of mold. In the simulation, the heat conduction equation considering the air/metal free boundary was solved by the finite element technique based on the Galerkin method, and the distribution of temperature was elucidated. The viscosity of molten matrix metal and the solid fraction, both of which depend on temperature, were calculated, and the movement of particles by the centrifugal force and drag were simulated. The results of simulation was compared with the experimental results and the validity of simulation was confirmed.
An elastic anisotropy produced by uni-axial plastic compression of a steel plate was studied with ultrasonic transverse waves. By taking into account an initial anisotropy of the plate, specimens were cut along four different directions in its plane. Giving them plastic compressive strains within about 10%, the velocities and polarization directions of transverse waves were measured by the sing-around method. For propagation in the transverse direction of specimens, it was shown that the polarization directions were parallel and perpendicular to the compression direction irrespective of plastic strain, and that the velocities varied linearly with plastic strain. Slopes of these linear velocity changes were nearly constant for each of the two polarization directions. For propagation in the thickness direction, the quite similar results were obtained on the specimens parallel and perpendicular to the rolling direction, while the rotations of polarization directions, and nonlinear velocity changes were confirmed in the other specimens. These experimental results were considered based on theoretical expressions for velocities and polarization directions, which were derived by assuming a slight anisotropy of a material to be the sum of an initial orthotropy and a strain-induced transverse isotropy. As a result, it was concluded that plastic compression produces a transverse isotropy, and, in general, rotates the anisotropy axes of materials.
Superplastic deformation behaviors in an (α+γ) duplex stainless steel have been studied with particular emphasis on microstructural change and cavity formation during tensile deformation at 950°C. The main results obtained are as follows; (1) Homogeneous (α+γ) microduplex structure with very fine α and γ grains can be obtained during tensile testing at 950°C at a strain rate of 1.0×10-2/S in the specimens which were solution-treated in the α single phase region and subsequent heavily cold-rolled. The initial cold rolling after solution treatment can play an important role for the formation of homogeneous (α+γ) microduplex structure, that is, the heavy cold rolling promotes the recrystallization of α matrix during hot deformation, resulting in the homogeneous (α+γ) microduplex structure. (2) Superplasticity is considerably enhanced by the prior heavy cold rolling, that is, the maximum elongation at a strain rate of 1.0×10-2S-1 in 90% cold-rolled specimens was 930%. The improvement of superplasticity arises from grain refinement of α and γ during hot deformation. (3) Two types of cavities were observed during hot deformation, that is, the one initiated in the early stage of deformation at the inclusions which were elongated by the prior heavy cold rolling, and the other initiated at grain boundaries and interfaces after the intermediate stage of deformation, suggesting that these cavities especially the former type of large ones may have deleterious effect on the strength and ductility during room temperature service.
The influence of slip between inclusion and matrix on the plastic deformation of an inhomogeneous material with inclusions was investigated. A plane model of inhomogeneous material with a thin boundary layer between the elliptic inclusions and the matrix was adopted, where the material was assumed to be rigid-plastic. The boundary slip layer was modeled by assuming lower yield stress for the boundary layer than those of the inclusion and the matrix. The rigid-plastic finite element method was used for the numerical calculation under the plane strain condition. The effects of the aspect ratio of the inclusion, the yield stress of the boundary layer and the volume fraction of the inclusion were studied. The patterns of the shear bands produced along the boundary layer and the mean flow stress of the inhomogeneous material were also discussed.
The effect of reinforcement of elastic modulus in polymer composite materials filled with ultramicroscopic particles has been investigated by taking an example of polymethyl methacrylate-palladium cluster composites. The singularity that makes the elastic modulus increase twice has been indicated by filling up the microscopic fine particles of 10-20Å in spite of a little volume which is 0.005%. This singularity is caused by the fact that the microscopic particles and the matrix, are perfectly unificated in the materials. The reinforcement of elastic modulus can be explained by the perfect parallel model in consideration of the interface restriction regions. It is clearly shown that such singularity is brought in only by the relative size effect of microscopic particles in such perfect composite materials.
Low cycle fatigue tests were performed at RT and elevated (673K) temperatures for ceramic sprayed 1%Cr-0.5%Mo steels. The fatigue properties of ceramic-sprayed steels and the mechanism of fatigue-fracture were discussed. The main results obtained are as follows. (1) The fatigue life of the ceramic-sprayed steel at 673K was slightly longer than the life at RT, while it was shorter than those of the base metal at RT and elevated temperatures. (2) The crack initiation in the ceramic layer occurred at a significantly early stage of fatigue of the coated steel. As the strain range became smaller, the crack initiation life of the ceramic layer became longer. (3) The fatigue-fracture process of the coated steel is considered as follows. At an early stage of fatigue, a fatigue crack is initiated at the surface of the ceramic layer and propagates quickly to the metal layer. The propagating fatigue crack through the ceramic layer continues to grow into the substrate.
Fracture behavior of Al2O3, PSZ and Si3N4 ceramics was investigated under cyclic mode I/mode II mixed-mode loading in high purity water at room temperature. The time to fracture was shorter than that for cyclic mode I loading, and it decreased markedly with increasing the mode ratio ΔKII/ΔKI. The crack propagation rate was higher than that for cyclic mode I loading, and it increased markedly with increasing the mode ratio ΔKII/ΔKI. The time to fracture was much shorter than the predicted based on cyclic mode I loading. The difference between them became larger in the order, Al2O3, Si3N4, PSZ. In the case of PSZ, the phase transformation from tetragonal to monoclinic took place under the cyclic mixed-mode and mode I loading tests, and there was almost no difference in the quantity. It was observed that the morphology of fracture surface was different between them, and that the fracture surface in the region of SCG for the mixed-mode loading developed a flat surface which is considered to occur owing to the addition of ΔKII.
In the previous paper, it was reported that the crushing load of Si3N4 balls for rolling hearings followed a 2-parameter Weibull distribution. In the present study, the method is presented to estimate its shape parameter from the distribution of the fracture strength of the bending specimens which were prepared from the same raw materials as those used for the production of Si3N4 balls and to estimate the scale parameter from the crushing tests of few balls. Then based on these two parameters, the method is proposed to estimate the crushing load for Si3N4 balls. Furthermore, the method of sampling inspection for balls is examined, and the number of balls to be tested is suggested.
Since adhesive bonded joints generally possess a considerable scatter in their strength, elucidation of the strength distribuion properties of these joints is important for better utilization of these joints. In this study, to investigate the change in strength distribution of adhesive joints due to the size effect, tensile and tensile shear strengths of adhesive bonded butt and lap joints were statistically evaluated. The probability of failure was calculated for butt joints with a different diameter and for lap joints with a different lap length. Each data set was then plotted using normal, Weibull and doubly exponential functions. Furthermore, the effect of adhesive area on the mean and coefficient of variation of the joint strength was investigated. First, the strength distributions of butt and lap joints were simulated by the weakest link model based on the strength distribution of the butt joint having the smallest diameter and stress distribution of the adhesive layer obtained by FEM analysis. Then, the mean and coefficient of variation of these joints were compared with the simulation results. The main results obtained are summarized as follows; (1) The strength distribution of adhesive bonded butt joint was nearly approximated by a normal distribution. However, for lap joint the strength distribution was fitted for an extreme distribution rather than a normal ditribution. (2) The relationship between the mean strength and adhesive area plotted on a logarithmic graph paper showed experimental curves of butt and lap joints with a convex shape. However, the mean strength of the butt joint simulated by the weakest link model decreased linearly with increasing adhesive area, while the shape of the simulated curve of the lap joint was concave. (3) The coefficient of variation of the butt joint was independent of the adhesive area, as predicted from the model, whereas the coefficient of variation of the lap joint decreased with increasing adhesive area contrary to the prediction of the model. This discrepacy may have been caused by a variation in shear strength due to a measurement error in the lap length.
Ni-Ti is known as one of the most excellent shape memory alloys because it has large recovery strain and recovery stress. However, shape memory effect (SME) is much affected by its composition, heat treatment and strain at heat treatment. Five compositions of Ni-Ti alloys [50.0-52.0 (at%Ni)] were chosen and treated at various temperatures [573-873(K)] and for various times [15min-4hr]. Then the measurements of shape change, transformation temperature and tensile mechanical properties were carried out. Moreover, the all-round shape memory effect (ARSM) of Ni-rich specimen [51.0 at%Ni] was examined after thermal cycle, strain cycle and exposure to high temperature. The results obtained are summarized as follows: (1) Ti-50.0 at%Ni alloy could induce good SME. (2) Pseudo-elasticity (PE) was much affected by the composition of alloy and heat treatment. (3) Compositions in a range from 51.0 at%Ni to 51.5 at%Ni and heat treatment at a range from 673K to 723K were most effective in increasing reversible shape memory effect (RSM). While PE declined with increasing strain, strain below 1% was effective in enhancing ARSM. Moreover, heat treatment below 623K for a short time made RSM opposite. (4) Thermal cycle was effective in enhancing ARSM because it increased shape change and hysteresis. Exposure above 473K induced abnormal ARSM.
The viscosities of 0.1mol·kg-1 aqueous solutions of tetraalkylammonium bromide (R4NBr; R=Me, Et, n-Pr, and n-Bu) were measured at 5.0, 10.0, 15.0, 25.0, 40.0, and 50.0°C up to 375MPa, using a high pressure rolling-ball viscometer. The reproducibility of the viscosities under high pressure was within ±1%. The viscosity minimum observed at 5.0°C for pure water became shallow or disappeared by adding R4NBr salts and with increasing temperature. The additive effect became larger in order of the alkyl chain length of R4NBr salts. The activation energy (Ev) and activation volume (Vv) for the viscous flow of aqueous solutions of R4NBr were estimated at each pressure. The minimum of Ev at about 250MPa was observed and its minimum became shallow in order of the alkyl chain length of R4NBr. The Vv of R4NBr increased with increasing pressure.
Calcium silicate briquette using waste silica was lightened by the addition of colloidal silica to the raw material. The ratio of SiO2 in the colloidal silica against that in waste silica, r, was varied from 0 to 1.2. With an increase of r until reaching 0.4, the bulk density decreased. Above 0.4 the bulk density remained in 1.1 to 1.2Mg/m3. The bending strength dropped sharply with the small addition of colloidal silica and linearly decreased with an increase of r. The briquette of 13MPa strength and 1.17Mg/m3 bulk density was obtained at the optimum ratio, r=0.4.
In recent years, ultrarapid-hardening cement concrete has been widely used for the repairing works of concrete structures. The purpose of this study is to establish the mix design method for the ultrarapid-hardening cement concrete. The ultrarapid-hardening cement concrete was prepared with various unit cement contents, water-cement ratios, sand-aggregate ratios and retarder-cement ratios at 10, 20 and 30°C, and tested for slump, air content, handling time (which was defined as the elapsed time when the slump decreased to about 2cm) and compressive strength. The conclusions obtained from the test results are summarized as follows: (1) The slump of the ultrarapid-hardening cement concrete with various mix proportions may generally be expressed by the following equation at 10, 20 and 30°C: Sl=Ae-B(Va+Vc)/Vw where Sl is the slump of the ultrarapid-hardening cement concrete, Va, Vc and Vw, are the volumes of aggregate, cement and water per unit volume of the ultrarapid-hardening cement concrete respectively, and A and B are empirical constants. (2) The compressive strength and handling time of the ultrarapid-hardening cement concrete with various mix proportions may generally be expressed as a function of “water-cement ratio (α)” and “retarder-cement ratio (β)” by the following equations at 10, 20 and 30°C: F=A/(Bα·Cβ) Ht=A·Bα·Cβ where F and Ht are the handling time and compressive strength of the ultrarapid-hardening cement concrete respectively, α is the water-cement ratio (α=Vw/Vc), β is the retarder-cement ratio (β=Vr/Vc), and A, B and C are empirical constants. (3) A mix design system for the ultrarapid-hardening cement concrete is proposed by use of the respective equations for slump, handling time and compressive strength predictions.
In order to investigate a constitutive law for various materials, many combined tension-torsion tests have been performed at complex strain paths. During these tests, it is more difficult to measure the axial and the shear strain simultaneously than to obtain the load and the torque. Although some instruments to measure both strains have been deviced, these are considerably large in comparison with the specimen dimensions, and therefore, it may be difficult to attach them briefly to the cylindrical specimen. In this paper, a small-sized biaxial strain meter was newly developed to measure the axial and shear strains. The stress-strain curves were measured for solid polymers by use of the new meter and were compared with the numerical computation of an overstress theory.